diff --git "a/Qwen2-VL-7B-FULL-full/trainer_state.json" "b/Qwen2-VL-7B-FULL-full/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/Qwen2-VL-7B-FULL-full/trainer_state.json"
@@ -0,0 +1,66598 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9999605610843073,
+  "eval_steps": 500,
+  "global_step": 9508,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00010517044184731881,
+      "grad_norm": 59.92323260669162,
+      "learning_rate": 5.0000000000000004e-08,
+      "loss": 2.6475,
+      "step": 1
+    },
+    {
+      "epoch": 0.00021034088369463762,
+      "grad_norm": 63.10940340752038,
+      "learning_rate": 1.0000000000000001e-07,
+      "loss": 2.7602,
+      "step": 2
+    },
+    {
+      "epoch": 0.00031551132554195644,
+      "grad_norm": 57.05123636971969,
+      "learning_rate": 1.5000000000000002e-07,
+      "loss": 2.6286,
+      "step": 3
+    },
+    {
+      "epoch": 0.00042068176738927523,
+      "grad_norm": 58.99194594170873,
+      "learning_rate": 2.0000000000000002e-07,
+      "loss": 2.6448,
+      "step": 4
+    },
+    {
+      "epoch": 0.000525852209236594,
+      "grad_norm": 60.12716290697537,
+      "learning_rate": 2.5000000000000004e-07,
+      "loss": 2.7222,
+      "step": 5
+    },
+    {
+      "epoch": 0.0006310226510839129,
+      "grad_norm": 65.69643117171603,
+      "learning_rate": 3.0000000000000004e-07,
+      "loss": 2.7782,
+      "step": 6
+    },
+    {
+      "epoch": 0.0007361930929312317,
+      "grad_norm": 54.90002763855268,
+      "learning_rate": 3.5000000000000004e-07,
+      "loss": 2.5453,
+      "step": 7
+    },
+    {
+      "epoch": 0.0008413635347785505,
+      "grad_norm": 61.987907330986985,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 2.6941,
+      "step": 8
+    },
+    {
+      "epoch": 0.0009465339766258693,
+      "grad_norm": 56.061715666844634,
+      "learning_rate": 4.5000000000000003e-07,
+      "loss": 2.6707,
+      "step": 9
+    },
+    {
+      "epoch": 0.001051704418473188,
+      "grad_norm": 63.95430356271557,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 2.6813,
+      "step": 10
+    },
+    {
+      "epoch": 0.001156874860320507,
+      "grad_norm": 57.584908322695675,
+      "learning_rate": 5.5e-07,
+      "loss": 2.4722,
+      "step": 11
+    },
+    {
+      "epoch": 0.0012620453021678258,
+      "grad_norm": 54.9509876269393,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 2.4741,
+      "step": 12
+    },
+    {
+      "epoch": 0.0013672157440151446,
+      "grad_norm": 56.36235985124101,
+      "learning_rate": 6.5e-07,
+      "loss": 2.638,
+      "step": 13
+    },
+    {
+      "epoch": 0.0014723861858624634,
+      "grad_norm": 46.019883313287565,
+      "learning_rate": 7.000000000000001e-07,
+      "loss": 2.2947,
+      "step": 14
+    },
+    {
+      "epoch": 0.001577556627709782,
+      "grad_norm": 39.36043169435088,
+      "learning_rate": 7.5e-07,
+      "loss": 1.9651,
+      "step": 15
+    },
+    {
+      "epoch": 0.001682727069557101,
+      "grad_norm": 44.53064805034052,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 2.1088,
+      "step": 16
+    },
+    {
+      "epoch": 0.0017878975114044198,
+      "grad_norm": 40.77122432291506,
+      "learning_rate": 8.500000000000001e-07,
+      "loss": 1.9506,
+      "step": 17
+    },
+    {
+      "epoch": 0.0018930679532517386,
+      "grad_norm": 33.413887606907345,
+      "learning_rate": 9.000000000000001e-07,
+      "loss": 1.8277,
+      "step": 18
+    },
+    {
+      "epoch": 0.0019982383950990575,
+      "grad_norm": 35.333835421159975,
+      "learning_rate": 9.500000000000001e-07,
+      "loss": 1.7883,
+      "step": 19
+    },
+    {
+      "epoch": 0.002103408836946376,
+      "grad_norm": 19.922751692123384,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 1.4946,
+      "step": 20
+    },
+    {
+      "epoch": 0.002208579278793695,
+      "grad_norm": 18.4584352012479,
+      "learning_rate": 1.0500000000000001e-06,
+      "loss": 1.4215,
+      "step": 21
+    },
+    {
+      "epoch": 0.002313749720641014,
+      "grad_norm": 19.920485594493087,
+      "learning_rate": 1.1e-06,
+      "loss": 1.4416,
+      "step": 22
+    },
+    {
+      "epoch": 0.002418920162488333,
+      "grad_norm": 16.206117061616325,
+      "learning_rate": 1.1500000000000002e-06,
+      "loss": 1.408,
+      "step": 23
+    },
+    {
+      "epoch": 0.0025240906043356515,
+      "grad_norm": 14.345274194736033,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 1.3782,
+      "step": 24
+    },
+    {
+      "epoch": 0.00262926104618297,
+      "grad_norm": 14.507180671796032,
+      "learning_rate": 1.25e-06,
+      "loss": 1.3618,
+      "step": 25
+    },
+    {
+      "epoch": 0.002734431488030289,
+      "grad_norm": 9.60203081004738,
+      "learning_rate": 1.3e-06,
+      "loss": 1.2395,
+      "step": 26
+    },
+    {
+      "epoch": 0.002839601929877608,
+      "grad_norm": 8.95961863190965,
+      "learning_rate": 1.3500000000000002e-06,
+      "loss": 1.1838,
+      "step": 27
+    },
+    {
+      "epoch": 0.002944772371724927,
+      "grad_norm": 7.743322908623659,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 1.149,
+      "step": 28
+    },
+    {
+      "epoch": 0.0030499428135722455,
+      "grad_norm": 8.000577585626917,
+      "learning_rate": 1.45e-06,
+      "loss": 1.1665,
+      "step": 29
+    },
+    {
+      "epoch": 0.003155113255419564,
+      "grad_norm": 7.385939102175253,
+      "learning_rate": 1.5e-06,
+      "loss": 1.1347,
+      "step": 30
+    },
+    {
+      "epoch": 0.0032602836972668832,
+      "grad_norm": 7.716119057096808,
+      "learning_rate": 1.5500000000000002e-06,
+      "loss": 1.1274,
+      "step": 31
+    },
+    {
+      "epoch": 0.003365454139114202,
+      "grad_norm": 7.434862596192311,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 1.1286,
+      "step": 32
+    },
+    {
+      "epoch": 0.003470624580961521,
+      "grad_norm": 7.734788123462178,
+      "learning_rate": 1.6500000000000003e-06,
+      "loss": 1.1539,
+      "step": 33
+    },
+    {
+      "epoch": 0.0035757950228088396,
+      "grad_norm": 5.966202361098791,
+      "learning_rate": 1.7000000000000002e-06,
+      "loss": 1.1445,
+      "step": 34
+    },
+    {
+      "epoch": 0.003680965464656158,
+      "grad_norm": 5.487171731422804,
+      "learning_rate": 1.75e-06,
+      "loss": 1.1067,
+      "step": 35
+    },
+    {
+      "epoch": 0.0037861359065034773,
+      "grad_norm": 4.431591345917228,
+      "learning_rate": 1.8000000000000001e-06,
+      "loss": 1.0843,
+      "step": 36
+    },
+    {
+      "epoch": 0.003891306348350796,
+      "grad_norm": 6.362482878285757,
+      "learning_rate": 1.85e-06,
+      "loss": 1.1035,
+      "step": 37
+    },
+    {
+      "epoch": 0.003996476790198115,
+      "grad_norm": 5.97952819240149,
+      "learning_rate": 1.9000000000000002e-06,
+      "loss": 1.1151,
+      "step": 38
+    },
+    {
+      "epoch": 0.004101647232045434,
+      "grad_norm": 7.236076440399765,
+      "learning_rate": 1.9500000000000004e-06,
+      "loss": 1.111,
+      "step": 39
+    },
+    {
+      "epoch": 0.004206817673892752,
+      "grad_norm": 7.231000749346946,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.0934,
+      "step": 40
+    },
+    {
+      "epoch": 0.004311988115740071,
+      "grad_norm": 7.458173023837684,
+      "learning_rate": 2.05e-06,
+      "loss": 1.1124,
+      "step": 41
+    },
+    {
+      "epoch": 0.00441715855758739,
+      "grad_norm": 7.048786741399918,
+      "learning_rate": 2.1000000000000002e-06,
+      "loss": 1.0972,
+      "step": 42
+    },
+    {
+      "epoch": 0.0045223289994347085,
+      "grad_norm": 5.904771802109113,
+      "learning_rate": 2.15e-06,
+      "loss": 1.1162,
+      "step": 43
+    },
+    {
+      "epoch": 0.004627499441282028,
+      "grad_norm": 8.892328258421735,
+      "learning_rate": 2.2e-06,
+      "loss": 1.1236,
+      "step": 44
+    },
+    {
+      "epoch": 0.004732669883129347,
+      "grad_norm": 5.984008442771571,
+      "learning_rate": 2.25e-06,
+      "loss": 1.1132,
+      "step": 45
+    },
+    {
+      "epoch": 0.004837840324976666,
+      "grad_norm": 4.433024003668077,
+      "learning_rate": 2.3000000000000004e-06,
+      "loss": 1.0935,
+      "step": 46
+    },
+    {
+      "epoch": 0.004943010766823984,
+      "grad_norm": 6.204751514592563,
+      "learning_rate": 2.35e-06,
+      "loss": 1.0928,
+      "step": 47
+    },
+    {
+      "epoch": 0.005048181208671303,
+      "grad_norm": 5.759130507078808,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 1.129,
+      "step": 48
+    },
+    {
+      "epoch": 0.005153351650518622,
+      "grad_norm": 7.090689618351067,
+      "learning_rate": 2.4500000000000003e-06,
+      "loss": 1.1351,
+      "step": 49
+    },
+    {
+      "epoch": 0.00525852209236594,
+      "grad_norm": 5.9876797839663745,
+      "learning_rate": 2.5e-06,
+      "loss": 1.0683,
+      "step": 50
+    },
+    {
+      "epoch": 0.005363692534213259,
+      "grad_norm": 7.595469488307658,
+      "learning_rate": 2.55e-06,
+      "loss": 1.0764,
+      "step": 51
+    },
+    {
+      "epoch": 0.005468862976060578,
+      "grad_norm": 5.978805506171299,
+      "learning_rate": 2.6e-06,
+      "loss": 1.1101,
+      "step": 52
+    },
+    {
+      "epoch": 0.005574033417907897,
+      "grad_norm": 5.389419934729486,
+      "learning_rate": 2.6500000000000005e-06,
+      "loss": 1.1183,
+      "step": 53
+    },
+    {
+      "epoch": 0.005679203859755216,
+      "grad_norm": 4.867913228404486,
+      "learning_rate": 2.7000000000000004e-06,
+      "loss": 1.1145,
+      "step": 54
+    },
+    {
+      "epoch": 0.005784374301602535,
+      "grad_norm": 8.051057067872096,
+      "learning_rate": 2.7500000000000004e-06,
+      "loss": 1.0788,
+      "step": 55
+    },
+    {
+      "epoch": 0.005889544743449854,
+      "grad_norm": 3.529571123644078,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 1.0933,
+      "step": 56
+    },
+    {
+      "epoch": 0.005994715185297172,
+      "grad_norm": 5.92833470887133,
+      "learning_rate": 2.85e-06,
+      "loss": 1.0814,
+      "step": 57
+    },
+    {
+      "epoch": 0.006099885627144491,
+      "grad_norm": 4.04957198212851,
+      "learning_rate": 2.9e-06,
+      "loss": 1.0894,
+      "step": 58
+    },
+    {
+      "epoch": 0.00620505606899181,
+      "grad_norm": 5.75800280839568,
+      "learning_rate": 2.95e-06,
+      "loss": 1.1081,
+      "step": 59
+    },
+    {
+      "epoch": 0.006310226510839128,
+      "grad_norm": 7.099778116155824,
+      "learning_rate": 3e-06,
+      "loss": 1.0787,
+      "step": 60
+    },
+    {
+      "epoch": 0.006415396952686447,
+      "grad_norm": 6.963451888461215,
+      "learning_rate": 3.05e-06,
+      "loss": 1.0834,
+      "step": 61
+    },
+    {
+      "epoch": 0.0065205673945337665,
+      "grad_norm": 6.132860287326327,
+      "learning_rate": 3.1000000000000004e-06,
+      "loss": 1.1122,
+      "step": 62
+    },
+    {
+      "epoch": 0.0066257378363810855,
+      "grad_norm": 5.193732285933575,
+      "learning_rate": 3.1500000000000003e-06,
+      "loss": 1.0615,
+      "step": 63
+    },
+    {
+      "epoch": 0.006730908278228404,
+      "grad_norm": 8.634476668713994,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 1.1458,
+      "step": 64
+    },
+    {
+      "epoch": 0.006836078720075723,
+      "grad_norm": 6.06480211066226,
+      "learning_rate": 3.2500000000000002e-06,
+      "loss": 1.0445,
+      "step": 65
+    },
+    {
+      "epoch": 0.006941249161923042,
+      "grad_norm": 6.9138684741530305,
+      "learning_rate": 3.3000000000000006e-06,
+      "loss": 1.1032,
+      "step": 66
+    },
+    {
+      "epoch": 0.00704641960377036,
+      "grad_norm": 5.982376980555477,
+      "learning_rate": 3.3500000000000005e-06,
+      "loss": 1.0606,
+      "step": 67
+    },
+    {
+      "epoch": 0.007151590045617679,
+      "grad_norm": 5.730675646561972,
+      "learning_rate": 3.4000000000000005e-06,
+      "loss": 1.0767,
+      "step": 68
+    },
+    {
+      "epoch": 0.007256760487464998,
+      "grad_norm": 4.46760589551449,
+      "learning_rate": 3.45e-06,
+      "loss": 1.054,
+      "step": 69
+    },
+    {
+      "epoch": 0.007361930929312316,
+      "grad_norm": 6.883313991446145,
+      "learning_rate": 3.5e-06,
+      "loss": 1.06,
+      "step": 70
+    },
+    {
+      "epoch": 0.0074671013711596354,
+      "grad_norm": 6.893517348928994,
+      "learning_rate": 3.5500000000000003e-06,
+      "loss": 1.076,
+      "step": 71
+    },
+    {
+      "epoch": 0.0075722718130069545,
+      "grad_norm": 6.622296647960904,
+      "learning_rate": 3.6000000000000003e-06,
+      "loss": 1.0866,
+      "step": 72
+    },
+    {
+      "epoch": 0.007677442254854274,
+      "grad_norm": 7.375737365019887,
+      "learning_rate": 3.65e-06,
+      "loss": 1.1335,
+      "step": 73
+    },
+    {
+      "epoch": 0.007782612696701592,
+      "grad_norm": 6.842915774489913,
+      "learning_rate": 3.7e-06,
+      "loss": 1.0819,
+      "step": 74
+    },
+    {
+      "epoch": 0.00788778313854891,
+      "grad_norm": 4.643560225418101,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 1.0627,
+      "step": 75
+    },
+    {
+      "epoch": 0.00799295358039623,
+      "grad_norm": 6.239260554512991,
+      "learning_rate": 3.8000000000000005e-06,
+      "loss": 1.0939,
+      "step": 76
+    },
+    {
+      "epoch": 0.008098124022243549,
+      "grad_norm": 6.739902596879763,
+      "learning_rate": 3.85e-06,
+      "loss": 1.0527,
+      "step": 77
+    },
+    {
+      "epoch": 0.008203294464090868,
+      "grad_norm": 4.805461911549342,
+      "learning_rate": 3.900000000000001e-06,
+      "loss": 1.0995,
+      "step": 78
+    },
+    {
+      "epoch": 0.008308464905938185,
+      "grad_norm": 4.600713739789344,
+      "learning_rate": 3.95e-06,
+      "loss": 1.0967,
+      "step": 79
+    },
+    {
+      "epoch": 0.008413635347785504,
+      "grad_norm": 5.785549449267131,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.079,
+      "step": 80
+    },
+    {
+      "epoch": 0.008518805789632823,
+      "grad_norm": 5.974468879016202,
+      "learning_rate": 4.05e-06,
+      "loss": 1.0991,
+      "step": 81
+    },
+    {
+      "epoch": 0.008623976231480143,
+      "grad_norm": 6.971544736299156,
+      "learning_rate": 4.1e-06,
+      "loss": 1.0889,
+      "step": 82
+    },
+    {
+      "epoch": 0.008729146673327462,
+      "grad_norm": 5.443939888698424,
+      "learning_rate": 4.15e-06,
+      "loss": 1.0758,
+      "step": 83
+    },
+    {
+      "epoch": 0.00883431711517478,
+      "grad_norm": 4.987130690796422,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 1.1191,
+      "step": 84
+    },
+    {
+      "epoch": 0.0089394875570221,
+      "grad_norm": 4.91454726971417,
+      "learning_rate": 4.25e-06,
+      "loss": 1.0437,
+      "step": 85
+    },
+    {
+      "epoch": 0.009044657998869417,
+      "grad_norm": 4.1569197342316375,
+      "learning_rate": 4.3e-06,
+      "loss": 1.0681,
+      "step": 86
+    },
+    {
+      "epoch": 0.009149828440716736,
+      "grad_norm": 6.3276090379354795,
+      "learning_rate": 4.350000000000001e-06,
+      "loss": 1.0924,
+      "step": 87
+    },
+    {
+      "epoch": 0.009254998882564055,
+      "grad_norm": 7.136019535146,
+      "learning_rate": 4.4e-06,
+      "loss": 1.062,
+      "step": 88
+    },
+    {
+      "epoch": 0.009360169324411374,
+      "grad_norm": 7.000501847709018,
+      "learning_rate": 4.450000000000001e-06,
+      "loss": 1.0825,
+      "step": 89
+    },
+    {
+      "epoch": 0.009465339766258693,
+      "grad_norm": 8.348084064157154,
+      "learning_rate": 4.5e-06,
+      "loss": 1.098,
+      "step": 90
+    },
+    {
+      "epoch": 0.009570510208106012,
+      "grad_norm": 6.3365124926671585,
+      "learning_rate": 4.5500000000000005e-06,
+      "loss": 1.0722,
+      "step": 91
+    },
+    {
+      "epoch": 0.009675680649953331,
+      "grad_norm": 6.850521475132954,
+      "learning_rate": 4.600000000000001e-06,
+      "loss": 1.0531,
+      "step": 92
+    },
+    {
+      "epoch": 0.009780851091800649,
+      "grad_norm": 3.534183914997874,
+      "learning_rate": 4.65e-06,
+      "loss": 1.0619,
+      "step": 93
+    },
+    {
+      "epoch": 0.009886021533647968,
+      "grad_norm": 5.513355921851675,
+      "learning_rate": 4.7e-06,
+      "loss": 1.0853,
+      "step": 94
+    },
+    {
+      "epoch": 0.009991191975495287,
+      "grad_norm": 4.3299555896256745,
+      "learning_rate": 4.75e-06,
+      "loss": 1.0755,
+      "step": 95
+    },
+    {
+      "epoch": 0.010096362417342606,
+      "grad_norm": 6.325453913242973,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 1.0889,
+      "step": 96
+    },
+    {
+      "epoch": 0.010201532859189925,
+      "grad_norm": 4.789718329298915,
+      "learning_rate": 4.85e-06,
+      "loss": 1.0862,
+      "step": 97
+    },
+    {
+      "epoch": 0.010306703301037244,
+      "grad_norm": 3.8761128602402297,
+      "learning_rate": 4.9000000000000005e-06,
+      "loss": 1.0612,
+      "step": 98
+    },
+    {
+      "epoch": 0.010411873742884561,
+      "grad_norm": 5.009647930586521,
+      "learning_rate": 4.95e-06,
+      "loss": 1.0595,
+      "step": 99
+    },
+    {
+      "epoch": 0.01051704418473188,
+      "grad_norm": 4.418331270407165,
+      "learning_rate": 5e-06,
+      "loss": 1.0824,
+      "step": 100
+    },
+    {
+      "epoch": 0.0106222146265792,
+      "grad_norm": 3.9028162186690802,
+      "learning_rate": 4.99999986061529e-06,
+      "loss": 1.0651,
+      "step": 101
+    },
+    {
+      "epoch": 0.010727385068426519,
+      "grad_norm": 6.8929955441897075,
+      "learning_rate": 4.999999442461175e-06,
+      "loss": 1.0997,
+      "step": 102
+    },
+    {
+      "epoch": 0.010832555510273838,
+      "grad_norm": 4.7248893287683185,
+      "learning_rate": 4.999998745537701e-06,
+      "loss": 1.0686,
+      "step": 103
+    },
+    {
+      "epoch": 0.010937725952121157,
+      "grad_norm": 5.812256570251753,
+      "learning_rate": 4.999997769844946e-06,
+      "loss": 1.0618,
+      "step": 104
+    },
+    {
+      "epoch": 0.011042896393968476,
+      "grad_norm": 2.708772175167679,
+      "learning_rate": 4.999996515383019e-06,
+      "loss": 1.0804,
+      "step": 105
+    },
+    {
+      "epoch": 0.011148066835815793,
+      "grad_norm": 6.2527627142277815,
+      "learning_rate": 4.9999949821520596e-06,
+      "loss": 1.1207,
+      "step": 106
+    },
+    {
+      "epoch": 0.011253237277663112,
+      "grad_norm": 5.5457481955779455,
+      "learning_rate": 4.99999317015224e-06,
+      "loss": 1.091,
+      "step": 107
+    },
+    {
+      "epoch": 0.011358407719510431,
+      "grad_norm": 5.761317582160581,
+      "learning_rate": 4.999991079383761e-06,
+      "loss": 1.095,
+      "step": 108
+    },
+    {
+      "epoch": 0.01146357816135775,
+      "grad_norm": 4.220512736500614,
+      "learning_rate": 4.999988709846856e-06,
+      "loss": 1.053,
+      "step": 109
+    },
+    {
+      "epoch": 0.01156874860320507,
+      "grad_norm": 6.66166762678115,
+      "learning_rate": 4.999986061541789e-06,
+      "loss": 1.0754,
+      "step": 110
+    },
+    {
+      "epoch": 0.011673919045052389,
+      "grad_norm": 3.35655019228154,
+      "learning_rate": 4.999983134468855e-06,
+      "loss": 1.0594,
+      "step": 111
+    },
+    {
+      "epoch": 0.011779089486899708,
+      "grad_norm": 4.939800790239611,
+      "learning_rate": 4.999979928628382e-06,
+      "loss": 1.0732,
+      "step": 112
+    },
+    {
+      "epoch": 0.011884259928747025,
+      "grad_norm": 5.566242564632212,
+      "learning_rate": 4.999976444020727e-06,
+      "loss": 1.1101,
+      "step": 113
+    },
+    {
+      "epoch": 0.011989430370594344,
+      "grad_norm": 4.129321853432761,
+      "learning_rate": 4.999972680646276e-06,
+      "loss": 1.0468,
+      "step": 114
+    },
+    {
+      "epoch": 0.012094600812441663,
+      "grad_norm": 5.459895093767263,
+      "learning_rate": 4.999968638505452e-06,
+      "loss": 1.1098,
+      "step": 115
+    },
+    {
+      "epoch": 0.012199771254288982,
+      "grad_norm": 4.268158570515893,
+      "learning_rate": 4.9999643175987045e-06,
+      "loss": 1.0813,
+      "step": 116
+    },
+    {
+      "epoch": 0.012304941696136301,
+      "grad_norm": 5.11466661467585,
+      "learning_rate": 4.999959717926515e-06,
+      "loss": 1.0552,
+      "step": 117
+    },
+    {
+      "epoch": 0.01241011213798362,
+      "grad_norm": 3.4664159857194274,
+      "learning_rate": 4.999954839489396e-06,
+      "loss": 1.0541,
+      "step": 118
+    },
+    {
+      "epoch": 0.01251528257983094,
+      "grad_norm": 3.9511165586614445,
+      "learning_rate": 4.999949682287893e-06,
+      "loss": 1.058,
+      "step": 119
+    },
+    {
+      "epoch": 0.012620453021678257,
+      "grad_norm": 3.614330210857767,
+      "learning_rate": 4.999944246322579e-06,
+      "loss": 1.0401,
+      "step": 120
+    },
+    {
+      "epoch": 0.012725623463525576,
+      "grad_norm": 5.105632604786978,
+      "learning_rate": 4.999938531594062e-06,
+      "loss": 1.0793,
+      "step": 121
+    },
+    {
+      "epoch": 0.012830793905372895,
+      "grad_norm": 4.957380407595618,
+      "learning_rate": 4.999932538102978e-06,
+      "loss": 1.0787,
+      "step": 122
+    },
+    {
+      "epoch": 0.012935964347220214,
+      "grad_norm": 4.643940273061057,
+      "learning_rate": 4.999926265849997e-06,
+      "loss": 1.0568,
+      "step": 123
+    },
+    {
+      "epoch": 0.013041134789067533,
+      "grad_norm": 4.627911589515215,
+      "learning_rate": 4.999919714835816e-06,
+      "loss": 1.1011,
+      "step": 124
+    },
+    {
+      "epoch": 0.013146305230914852,
+      "grad_norm": 3.102924045859167,
+      "learning_rate": 4.999912885061167e-06,
+      "loss": 1.0854,
+      "step": 125
+    },
+    {
+      "epoch": 0.013251475672762171,
+      "grad_norm": 4.92620716306597,
+      "learning_rate": 4.999905776526811e-06,
+      "loss": 1.106,
+      "step": 126
+    },
+    {
+      "epoch": 0.013356646114609488,
+      "grad_norm": 3.527926170990561,
+      "learning_rate": 4.9998983892335416e-06,
+      "loss": 1.0882,
+      "step": 127
+    },
+    {
+      "epoch": 0.013461816556456807,
+      "grad_norm": 4.005511364388478,
+      "learning_rate": 4.999890723182183e-06,
+      "loss": 1.0608,
+      "step": 128
+    },
+    {
+      "epoch": 0.013566986998304126,
+      "grad_norm": 3.6947799424472887,
+      "learning_rate": 4.9998827783735875e-06,
+      "loss": 1.082,
+      "step": 129
+    },
+    {
+      "epoch": 0.013672157440151446,
+      "grad_norm": 4.8537336025866225,
+      "learning_rate": 4.999874554808643e-06,
+      "loss": 1.0154,
+      "step": 130
+    },
+    {
+      "epoch": 0.013777327881998765,
+      "grad_norm": 5.224073594225298,
+      "learning_rate": 4.999866052488265e-06,
+      "loss": 1.0672,
+      "step": 131
+    },
+    {
+      "epoch": 0.013882498323846084,
+      "grad_norm": 6.699186460471001,
+      "learning_rate": 4.999857271413403e-06,
+      "loss": 1.1189,
+      "step": 132
+    },
+    {
+      "epoch": 0.013987668765693401,
+      "grad_norm": 3.974289494553639,
+      "learning_rate": 4.999848211585035e-06,
+      "loss": 1.0751,
+      "step": 133
+    },
+    {
+      "epoch": 0.01409283920754072,
+      "grad_norm": 4.528291252084844,
+      "learning_rate": 4.999838873004173e-06,
+      "loss": 1.0478,
+      "step": 134
+    },
+    {
+      "epoch": 0.01419800964938804,
+      "grad_norm": 4.971558341946056,
+      "learning_rate": 4.999829255671857e-06,
+      "loss": 1.0358,
+      "step": 135
+    },
+    {
+      "epoch": 0.014303180091235358,
+      "grad_norm": 4.085942947769176,
+      "learning_rate": 4.99981935958916e-06,
+      "loss": 1.0944,
+      "step": 136
+    },
+    {
+      "epoch": 0.014408350533082677,
+      "grad_norm": 6.664417732538393,
+      "learning_rate": 4.999809184757185e-06,
+      "loss": 1.0917,
+      "step": 137
+    },
+    {
+      "epoch": 0.014513520974929996,
+      "grad_norm": 4.806426966250999,
+      "learning_rate": 4.999798731177066e-06,
+      "loss": 1.1051,
+      "step": 138
+    },
+    {
+      "epoch": 0.014618691416777315,
+      "grad_norm": 4.561326503607282,
+      "learning_rate": 4.9997879988499695e-06,
+      "loss": 1.0322,
+      "step": 139
+    },
+    {
+      "epoch": 0.014723861858624633,
+      "grad_norm": 4.264177348921955,
+      "learning_rate": 4.999776987777093e-06,
+      "loss": 1.0489,
+      "step": 140
+    },
+    {
+      "epoch": 0.014829032300471952,
+      "grad_norm": 4.72227071149687,
+      "learning_rate": 4.9997656979596624e-06,
+      "loss": 1.0356,
+      "step": 141
+    },
+    {
+      "epoch": 0.014934202742319271,
+      "grad_norm": 5.240830574426804,
+      "learning_rate": 4.999754129398938e-06,
+      "loss": 1.0517,
+      "step": 142
+    },
+    {
+      "epoch": 0.01503937318416659,
+      "grad_norm": 3.7824663189008674,
+      "learning_rate": 4.999742282096209e-06,
+      "loss": 1.0765,
+      "step": 143
+    },
+    {
+      "epoch": 0.015144543626013909,
+      "grad_norm": 4.031279478713754,
+      "learning_rate": 4.9997301560527976e-06,
+      "loss": 1.043,
+      "step": 144
+    },
+    {
+      "epoch": 0.015249714067861228,
+      "grad_norm": 3.4761815646490244,
+      "learning_rate": 4.999717751270055e-06,
+      "loss": 1.0808,
+      "step": 145
+    },
+    {
+      "epoch": 0.015354884509708547,
+      "grad_norm": 3.5505878417744525,
+      "learning_rate": 4.9997050677493646e-06,
+      "loss": 1.0576,
+      "step": 146
+    },
+    {
+      "epoch": 0.015460054951555864,
+      "grad_norm": 5.655354096949263,
+      "learning_rate": 4.999692105492141e-06,
+      "loss": 1.1257,
+      "step": 147
+    },
+    {
+      "epoch": 0.015565225393403184,
+      "grad_norm": 4.080899903125995,
+      "learning_rate": 4.999678864499828e-06,
+      "loss": 1.1199,
+      "step": 148
+    },
+    {
+      "epoch": 0.015670395835250504,
+      "grad_norm": 4.634309433383539,
+      "learning_rate": 4.9996653447739054e-06,
+      "loss": 1.105,
+      "step": 149
+    },
+    {
+      "epoch": 0.01577556627709782,
+      "grad_norm": 4.261945588917611,
+      "learning_rate": 4.999651546315877e-06,
+      "loss": 1.0543,
+      "step": 150
+    },
+    {
+      "epoch": 0.01588073671894514,
+      "grad_norm": 4.213617826403962,
+      "learning_rate": 4.999637469127284e-06,
+      "loss": 1.0604,
+      "step": 151
+    },
+    {
+      "epoch": 0.01598590716079246,
+      "grad_norm": 5.107781163495558,
+      "learning_rate": 4.9996231132096955e-06,
+      "loss": 1.075,
+      "step": 152
+    },
+    {
+      "epoch": 0.016091077602639777,
+      "grad_norm": 4.210150800964649,
+      "learning_rate": 4.999608478564713e-06,
+      "loss": 1.0672,
+      "step": 153
+    },
+    {
+      "epoch": 0.016196248044487098,
+      "grad_norm": 3.1431426378379084,
+      "learning_rate": 4.999593565193965e-06,
+      "loss": 1.0347,
+      "step": 154
+    },
+    {
+      "epoch": 0.016301418486334415,
+      "grad_norm": 4.579222492136831,
+      "learning_rate": 4.999578373099119e-06,
+      "loss": 1.0858,
+      "step": 155
+    },
+    {
+      "epoch": 0.016406588928181736,
+      "grad_norm": 5.066019223738902,
+      "learning_rate": 4.999562902281866e-06,
+      "loss": 1.0952,
+      "step": 156
+    },
+    {
+      "epoch": 0.016511759370029053,
+      "grad_norm": 2.6553215468038216,
+      "learning_rate": 4.999547152743933e-06,
+      "loss": 1.062,
+      "step": 157
+    },
+    {
+      "epoch": 0.01661692981187637,
+      "grad_norm": 2.465224547489308,
+      "learning_rate": 4.999531124487074e-06,
+      "loss": 1.0518,
+      "step": 158
+    },
+    {
+      "epoch": 0.01672210025372369,
+      "grad_norm": 3.446424772289735,
+      "learning_rate": 4.999514817513079e-06,
+      "loss": 1.0765,
+      "step": 159
+    },
+    {
+      "epoch": 0.01682727069557101,
+      "grad_norm": 5.154039712884113,
+      "learning_rate": 4.999498231823765e-06,
+      "loss": 1.0515,
+      "step": 160
+    },
+    {
+      "epoch": 0.01693244113741833,
+      "grad_norm": 3.327200405368395,
+      "learning_rate": 4.9994813674209805e-06,
+      "loss": 1.0734,
+      "step": 161
+    },
+    {
+      "epoch": 0.017037611579265647,
+      "grad_norm": 6.642307412221283,
+      "learning_rate": 4.999464224306607e-06,
+      "loss": 1.1043,
+      "step": 162
+    },
+    {
+      "epoch": 0.017142782021112968,
+      "grad_norm": 3.530091197441022,
+      "learning_rate": 4.999446802482556e-06,
+      "loss": 1.0833,
+      "step": 163
+    },
+    {
+      "epoch": 0.017247952462960285,
+      "grad_norm": 5.657095039085531,
+      "learning_rate": 4.99942910195077e-06,
+      "loss": 1.0632,
+      "step": 164
+    },
+    {
+      "epoch": 0.017353122904807602,
+      "grad_norm": 5.574109052251118,
+      "learning_rate": 4.999411122713223e-06,
+      "loss": 1.095,
+      "step": 165
+    },
+    {
+      "epoch": 0.017458293346654923,
+      "grad_norm": 5.919741121398457,
+      "learning_rate": 4.99939286477192e-06,
+      "loss": 1.0714,
+      "step": 166
+    },
+    {
+      "epoch": 0.01756346378850224,
+      "grad_norm": 5.72438831771799,
+      "learning_rate": 4.999374328128896e-06,
+      "loss": 1.1101,
+      "step": 167
+    },
+    {
+      "epoch": 0.01766863423034956,
+      "grad_norm": 3.6834749885484404,
+      "learning_rate": 4.99935551278622e-06,
+      "loss": 1.0671,
+      "step": 168
+    },
+    {
+      "epoch": 0.01777380467219688,
+      "grad_norm": 3.573832404245989,
+      "learning_rate": 4.999336418745987e-06,
+      "loss": 1.0852,
+      "step": 169
+    },
+    {
+      "epoch": 0.0178789751140442,
+      "grad_norm": 4.735678508134887,
+      "learning_rate": 4.9993170460103295e-06,
+      "loss": 1.0536,
+      "step": 170
+    },
+    {
+      "epoch": 0.017984145555891517,
+      "grad_norm": 4.056452634779431,
+      "learning_rate": 4.9992973945814045e-06,
+      "loss": 1.0438,
+      "step": 171
+    },
+    {
+      "epoch": 0.018089315997738834,
+      "grad_norm": 4.957195740828829,
+      "learning_rate": 4.999277464461405e-06,
+      "loss": 1.0839,
+      "step": 172
+    },
+    {
+      "epoch": 0.018194486439586155,
+      "grad_norm": 3.926000211879302,
+      "learning_rate": 4.9992572556525535e-06,
+      "loss": 1.0788,
+      "step": 173
+    },
+    {
+      "epoch": 0.018299656881433472,
+      "grad_norm": 4.037823469074097,
+      "learning_rate": 4.999236768157103e-06,
+      "loss": 1.0648,
+      "step": 174
+    },
+    {
+      "epoch": 0.018404827323280793,
+      "grad_norm": 3.5763594650594706,
+      "learning_rate": 4.999216001977338e-06,
+      "loss": 1.0776,
+      "step": 175
+    },
+    {
+      "epoch": 0.01850999776512811,
+      "grad_norm": 4.216315800530571,
+      "learning_rate": 4.999194957115574e-06,
+      "loss": 1.0727,
+      "step": 176
+    },
+    {
+      "epoch": 0.01861516820697543,
+      "grad_norm": 3.7534868688617897,
+      "learning_rate": 4.999173633574158e-06,
+      "loss": 1.0114,
+      "step": 177
+    },
+    {
+      "epoch": 0.01872033864882275,
+      "grad_norm": 3.2854824452012257,
+      "learning_rate": 4.999152031355468e-06,
+      "loss": 1.0757,
+      "step": 178
+    },
+    {
+      "epoch": 0.018825509090670066,
+      "grad_norm": 5.452335174224296,
+      "learning_rate": 4.999130150461912e-06,
+      "loss": 1.1222,
+      "step": 179
+    },
+    {
+      "epoch": 0.018930679532517387,
+      "grad_norm": 4.533901928234252,
+      "learning_rate": 4.999107990895931e-06,
+      "loss": 1.0774,
+      "step": 180
+    },
+    {
+      "epoch": 0.019035849974364704,
+      "grad_norm": 3.9049991918717053,
+      "learning_rate": 4.999085552659995e-06,
+      "loss": 1.0649,
+      "step": 181
+    },
+    {
+      "epoch": 0.019141020416212025,
+      "grad_norm": 3.813729543734002,
+      "learning_rate": 4.9990628357566055e-06,
+      "loss": 1.0448,
+      "step": 182
+    },
+    {
+      "epoch": 0.019246190858059342,
+      "grad_norm": 3.5500653660890524,
+      "learning_rate": 4.999039840188297e-06,
+      "loss": 1.0776,
+      "step": 183
+    },
+    {
+      "epoch": 0.019351361299906663,
+      "grad_norm": 4.714466443116458,
+      "learning_rate": 4.999016565957633e-06,
+      "loss": 1.1458,
+      "step": 184
+    },
+    {
+      "epoch": 0.01945653174175398,
+      "grad_norm": 5.858915587801188,
+      "learning_rate": 4.9989930130672085e-06,
+      "loss": 1.067,
+      "step": 185
+    },
+    {
+      "epoch": 0.019561702183601298,
+      "grad_norm": 6.034299083940081,
+      "learning_rate": 4.99896918151965e-06,
+      "loss": 1.0984,
+      "step": 186
+    },
+    {
+      "epoch": 0.01966687262544862,
+      "grad_norm": 4.074883949501879,
+      "learning_rate": 4.9989450713176156e-06,
+      "loss": 1.0182,
+      "step": 187
+    },
+    {
+      "epoch": 0.019772043067295936,
+      "grad_norm": 3.8546892311347984,
+      "learning_rate": 4.998920682463794e-06,
+      "loss": 1.1006,
+      "step": 188
+    },
+    {
+      "epoch": 0.019877213509143257,
+      "grad_norm": 2.8323107910480316,
+      "learning_rate": 4.998896014960904e-06,
+      "loss": 1.1025,
+      "step": 189
+    },
+    {
+      "epoch": 0.019982383950990574,
+      "grad_norm": 3.2647909530548818,
+      "learning_rate": 4.998871068811695e-06,
+      "loss": 1.0553,
+      "step": 190
+    },
+    {
+      "epoch": 0.02008755439283789,
+      "grad_norm": 4.8772055016374924,
+      "learning_rate": 4.99884584401895e-06,
+      "loss": 1.0435,
+      "step": 191
+    },
+    {
+      "epoch": 0.020192724834685212,
+      "grad_norm": 4.446998405870374,
+      "learning_rate": 4.998820340585482e-06,
+      "loss": 1.0841,
+      "step": 192
+    },
+    {
+      "epoch": 0.02029789527653253,
+      "grad_norm": 6.789642965482605,
+      "learning_rate": 4.998794558514135e-06,
+      "loss": 1.0501,
+      "step": 193
+    },
+    {
+      "epoch": 0.02040306571837985,
+      "grad_norm": 5.368972583293381,
+      "learning_rate": 4.998768497807783e-06,
+      "loss": 1.0759,
+      "step": 194
+    },
+    {
+      "epoch": 0.020508236160227167,
+      "grad_norm": 4.6463740756090175,
+      "learning_rate": 4.9987421584693316e-06,
+      "loss": 1.0359,
+      "step": 195
+    },
+    {
+      "epoch": 0.02061340660207449,
+      "grad_norm": 3.638853744840631,
+      "learning_rate": 4.99871554050172e-06,
+      "loss": 1.043,
+      "step": 196
+    },
+    {
+      "epoch": 0.020718577043921806,
+      "grad_norm": 5.963532277665872,
+      "learning_rate": 4.998688643907914e-06,
+      "loss": 1.0821,
+      "step": 197
+    },
+    {
+      "epoch": 0.020823747485769123,
+      "grad_norm": 4.211503968938633,
+      "learning_rate": 4.998661468690914e-06,
+      "loss": 1.0341,
+      "step": 198
+    },
+    {
+      "epoch": 0.020928917927616444,
+      "grad_norm": 5.9295459087612095,
+      "learning_rate": 4.9986340148537506e-06,
+      "loss": 1.0743,
+      "step": 199
+    },
+    {
+      "epoch": 0.02103408836946376,
+      "grad_norm": 3.4504029111986774,
+      "learning_rate": 4.998606282399484e-06,
+      "loss": 1.0975,
+      "step": 200
+    },
+    {
+      "epoch": 0.021139258811311082,
+      "grad_norm": 2.7556837965327654,
+      "learning_rate": 4.998578271331207e-06,
+      "loss": 1.1087,
+      "step": 201
+    },
+    {
+      "epoch": 0.0212444292531584,
+      "grad_norm": 4.467263377291867,
+      "learning_rate": 4.998549981652043e-06,
+      "loss": 1.0805,
+      "step": 202
+    },
+    {
+      "epoch": 0.02134959969500572,
+      "grad_norm": 3.12366809969675,
+      "learning_rate": 4.998521413365147e-06,
+      "loss": 1.1032,
+      "step": 203
+    },
+    {
+      "epoch": 0.021454770136853037,
+      "grad_norm": 3.678013910323886,
+      "learning_rate": 4.998492566473704e-06,
+      "loss": 1.0831,
+      "step": 204
+    },
+    {
+      "epoch": 0.021559940578700355,
+      "grad_norm": 4.534344934215397,
+      "learning_rate": 4.998463440980931e-06,
+      "loss": 1.0911,
+      "step": 205
+    },
+    {
+      "epoch": 0.021665111020547675,
+      "grad_norm": 5.977433768607247,
+      "learning_rate": 4.998434036890075e-06,
+      "loss": 1.0532,
+      "step": 206
+    },
+    {
+      "epoch": 0.021770281462394993,
+      "grad_norm": 3.097171042849419,
+      "learning_rate": 4.998404354204416e-06,
+      "loss": 1.0624,
+      "step": 207
+    },
+    {
+      "epoch": 0.021875451904242314,
+      "grad_norm": 4.741569481149735,
+      "learning_rate": 4.998374392927262e-06,
+      "loss": 1.0814,
+      "step": 208
+    },
+    {
+      "epoch": 0.02198062234608963,
+      "grad_norm": 5.387514870731206,
+      "learning_rate": 4.998344153061957e-06,
+      "loss": 1.0667,
+      "step": 209
+    },
+    {
+      "epoch": 0.02208579278793695,
+      "grad_norm": 3.107349100997203,
+      "learning_rate": 4.99831363461187e-06,
+      "loss": 1.0283,
+      "step": 210
+    },
+    {
+      "epoch": 0.02219096322978427,
+      "grad_norm": 4.446119580642245,
+      "learning_rate": 4.998282837580405e-06,
+      "loss": 1.0642,
+      "step": 211
+    },
+    {
+      "epoch": 0.022296133671631586,
+      "grad_norm": 3.6643593798924377,
+      "learning_rate": 4.998251761970997e-06,
+      "loss": 1.0556,
+      "step": 212
+    },
+    {
+      "epoch": 0.022401304113478907,
+      "grad_norm": 3.2516463208152917,
+      "learning_rate": 4.99822040778711e-06,
+      "loss": 1.0386,
+      "step": 213
+    },
+    {
+      "epoch": 0.022506474555326225,
+      "grad_norm": 5.733316290126696,
+      "learning_rate": 4.998188775032241e-06,
+      "loss": 1.0906,
+      "step": 214
+    },
+    {
+      "epoch": 0.022611644997173545,
+      "grad_norm": 4.91709171945248,
+      "learning_rate": 4.998156863709917e-06,
+      "loss": 1.0903,
+      "step": 215
+    },
+    {
+      "epoch": 0.022716815439020863,
+      "grad_norm": 4.657782710443831,
+      "learning_rate": 4.998124673823695e-06,
+      "loss": 1.1172,
+      "step": 216
+    },
+    {
+      "epoch": 0.022821985880868183,
+      "grad_norm": 6.169459224100382,
+      "learning_rate": 4.998092205377168e-06,
+      "loss": 1.0814,
+      "step": 217
+    },
+    {
+      "epoch": 0.0229271563227155,
+      "grad_norm": 4.948076959697061,
+      "learning_rate": 4.998059458373952e-06,
+      "loss": 1.1041,
+      "step": 218
+    },
+    {
+      "epoch": 0.023032326764562818,
+      "grad_norm": 5.515487044434082,
+      "learning_rate": 4.998026432817702e-06,
+      "loss": 1.0771,
+      "step": 219
+    },
+    {
+      "epoch": 0.02313749720641014,
+      "grad_norm": 3.6929446615249364,
+      "learning_rate": 4.9979931287121e-06,
+      "loss": 1.0612,
+      "step": 220
+    },
+    {
+      "epoch": 0.023242667648257456,
+      "grad_norm": 3.8867790075290976,
+      "learning_rate": 4.9979595460608575e-06,
+      "loss": 1.0381,
+      "step": 221
+    },
+    {
+      "epoch": 0.023347838090104777,
+      "grad_norm": 4.9882563091630825,
+      "learning_rate": 4.997925684867721e-06,
+      "loss": 1.096,
+      "step": 222
+    },
+    {
+      "epoch": 0.023453008531952094,
+      "grad_norm": 3.8165955011698727,
+      "learning_rate": 4.997891545136467e-06,
+      "loss": 1.04,
+      "step": 223
+    },
+    {
+      "epoch": 0.023558178973799415,
+      "grad_norm": 4.840235092033454,
+      "learning_rate": 4.997857126870902e-06,
+      "loss": 1.0896,
+      "step": 224
+    },
+    {
+      "epoch": 0.023663349415646732,
+      "grad_norm": 5.334008497621717,
+      "learning_rate": 4.9978224300748625e-06,
+      "loss": 1.1243,
+      "step": 225
+    },
+    {
+      "epoch": 0.02376851985749405,
+      "grad_norm": 3.6424315914693337,
+      "learning_rate": 4.997787454752217e-06,
+      "loss": 1.0253,
+      "step": 226
+    },
+    {
+      "epoch": 0.02387369029934137,
+      "grad_norm": 3.7986132911207617,
+      "learning_rate": 4.997752200906868e-06,
+      "loss": 1.0196,
+      "step": 227
+    },
+    {
+      "epoch": 0.023978860741188688,
+      "grad_norm": 3.5858200402131777,
+      "learning_rate": 4.997716668542746e-06,
+      "loss": 1.0516,
+      "step": 228
+    },
+    {
+      "epoch": 0.02408403118303601,
+      "grad_norm": 4.090414058227785,
+      "learning_rate": 4.997680857663812e-06,
+      "loss": 1.0889,
+      "step": 229
+    },
+    {
+      "epoch": 0.024189201624883326,
+      "grad_norm": 6.149411432586733,
+      "learning_rate": 4.997644768274059e-06,
+      "loss": 1.0915,
+      "step": 230
+    },
+    {
+      "epoch": 0.024294372066730647,
+      "grad_norm": 4.990447662747164,
+      "learning_rate": 4.997608400377513e-06,
+      "loss": 1.1035,
+      "step": 231
+    },
+    {
+      "epoch": 0.024399542508577964,
+      "grad_norm": 3.6379678181805626,
+      "learning_rate": 4.997571753978228e-06,
+      "loss": 1.0523,
+      "step": 232
+    },
+    {
+      "epoch": 0.02450471295042528,
+      "grad_norm": 4.063194976272717,
+      "learning_rate": 4.99753482908029e-06,
+      "loss": 1.0584,
+      "step": 233
+    },
+    {
+      "epoch": 0.024609883392272602,
+      "grad_norm": 4.805240814647609,
+      "learning_rate": 4.997497625687818e-06,
+      "loss": 1.1217,
+      "step": 234
+    },
+    {
+      "epoch": 0.02471505383411992,
+      "grad_norm": 4.213633680202407,
+      "learning_rate": 4.997460143804958e-06,
+      "loss": 1.0653,
+      "step": 235
+    },
+    {
+      "epoch": 0.02482022427596724,
+      "grad_norm": 4.605436966627413,
+      "learning_rate": 4.997422383435893e-06,
+      "loss": 1.0525,
+      "step": 236
+    },
+    {
+      "epoch": 0.024925394717814558,
+      "grad_norm": 3.403523309179731,
+      "learning_rate": 4.99738434458483e-06,
+      "loss": 1.0587,
+      "step": 237
+    },
+    {
+      "epoch": 0.02503056515966188,
+      "grad_norm": 3.2122262832368973,
+      "learning_rate": 4.997346027256013e-06,
+      "loss": 1.0795,
+      "step": 238
+    },
+    {
+      "epoch": 0.025135735601509196,
+      "grad_norm": 4.571309867532744,
+      "learning_rate": 4.997307431453713e-06,
+      "loss": 1.0673,
+      "step": 239
+    },
+    {
+      "epoch": 0.025240906043356513,
+      "grad_norm": 3.5447243507334396,
+      "learning_rate": 4.9972685571822355e-06,
+      "loss": 1.0261,
+      "step": 240
+    },
+    {
+      "epoch": 0.025346076485203834,
+      "grad_norm": 3.2377036811282336,
+      "learning_rate": 4.997229404445914e-06,
+      "loss": 1.0424,
+      "step": 241
+    },
+    {
+      "epoch": 0.02545124692705115,
+      "grad_norm": 3.6188960442806017,
+      "learning_rate": 4.997189973249115e-06,
+      "loss": 1.0553,
+      "step": 242
+    },
+    {
+      "epoch": 0.025556417368898472,
+      "grad_norm": 3.3956125214606336,
+      "learning_rate": 4.997150263596236e-06,
+      "loss": 1.0437,
+      "step": 243
+    },
+    {
+      "epoch": 0.02566158781074579,
+      "grad_norm": 4.3265152622652,
+      "learning_rate": 4.997110275491702e-06,
+      "loss": 1.0729,
+      "step": 244
+    },
+    {
+      "epoch": 0.02576675825259311,
+      "grad_norm": 4.350664171417144,
+      "learning_rate": 4.997070008939976e-06,
+      "loss": 1.0522,
+      "step": 245
+    },
+    {
+      "epoch": 0.025871928694440428,
+      "grad_norm": 3.8316671691883872,
+      "learning_rate": 4.997029463945545e-06,
+      "loss": 1.0685,
+      "step": 246
+    },
+    {
+      "epoch": 0.025977099136287745,
+      "grad_norm": 3.5671479642145076,
+      "learning_rate": 4.996988640512931e-06,
+      "loss": 1.066,
+      "step": 247
+    },
+    {
+      "epoch": 0.026082269578135066,
+      "grad_norm": 3.947898665647903,
+      "learning_rate": 4.996947538646687e-06,
+      "loss": 1.0524,
+      "step": 248
+    },
+    {
+      "epoch": 0.026187440019982383,
+      "grad_norm": 3.499962475992957,
+      "learning_rate": 4.996906158351396e-06,
+      "loss": 1.0395,
+      "step": 249
+    },
+    {
+      "epoch": 0.026292610461829704,
+      "grad_norm": 3.653224022870733,
+      "learning_rate": 4.99686449963167e-06,
+      "loss": 1.0245,
+      "step": 250
+    },
+    {
+      "epoch": 0.02639778090367702,
+      "grad_norm": 3.2905872706720634,
+      "learning_rate": 4.996822562492157e-06,
+      "loss": 1.1054,
+      "step": 251
+    },
+    {
+      "epoch": 0.026502951345524342,
+      "grad_norm": 4.813210566349707,
+      "learning_rate": 4.996780346937532e-06,
+      "loss": 1.1126,
+      "step": 252
+    },
+    {
+      "epoch": 0.02660812178737166,
+      "grad_norm": 3.8457085777908926,
+      "learning_rate": 4.996737852972503e-06,
+      "loss": 1.0783,
+      "step": 253
+    },
+    {
+      "epoch": 0.026713292229218977,
+      "grad_norm": 3.7993606994618316,
+      "learning_rate": 4.9966950806018075e-06,
+      "loss": 1.0461,
+      "step": 254
+    },
+    {
+      "epoch": 0.026818462671066298,
+      "grad_norm": 4.8200166677976375,
+      "learning_rate": 4.996652029830216e-06,
+      "loss": 1.0875,
+      "step": 255
+    },
+    {
+      "epoch": 0.026923633112913615,
+      "grad_norm": 4.328119154393376,
+      "learning_rate": 4.996608700662528e-06,
+      "loss": 1.0739,
+      "step": 256
+    },
+    {
+      "epoch": 0.027028803554760936,
+      "grad_norm": 3.1342866705412686,
+      "learning_rate": 4.996565093103576e-06,
+      "loss": 1.0948,
+      "step": 257
+    },
+    {
+      "epoch": 0.027133973996608253,
+      "grad_norm": 4.961920624574888,
+      "learning_rate": 4.996521207158222e-06,
+      "loss": 1.0635,
+      "step": 258
+    },
+    {
+      "epoch": 0.027239144438455574,
+      "grad_norm": 3.360160077037182,
+      "learning_rate": 4.99647704283136e-06,
+      "loss": 1.0652,
+      "step": 259
+    },
+    {
+      "epoch": 0.02734431488030289,
+      "grad_norm": 3.5370384271061077,
+      "learning_rate": 4.9964326001279145e-06,
+      "loss": 1.0602,
+      "step": 260
+    },
+    {
+      "epoch": 0.02744948532215021,
+      "grad_norm": 3.991312523764047,
+      "learning_rate": 4.996387879052841e-06,
+      "loss": 1.0563,
+      "step": 261
+    },
+    {
+      "epoch": 0.02755465576399753,
+      "grad_norm": 3.80460958358854,
+      "learning_rate": 4.996342879611126e-06,
+      "loss": 1.0713,
+      "step": 262
+    },
+    {
+      "epoch": 0.027659826205844847,
+      "grad_norm": 4.454127422134502,
+      "learning_rate": 4.996297601807788e-06,
+      "loss": 1.0725,
+      "step": 263
+    },
+    {
+      "epoch": 0.027764996647692167,
+      "grad_norm": 4.743749897023404,
+      "learning_rate": 4.996252045647875e-06,
+      "loss": 1.0079,
+      "step": 264
+    },
+    {
+      "epoch": 0.027870167089539485,
+      "grad_norm": 4.979818582007352,
+      "learning_rate": 4.9962062111364675e-06,
+      "loss": 1.0908,
+      "step": 265
+    },
+    {
+      "epoch": 0.027975337531386802,
+      "grad_norm": 3.8751951820571597,
+      "learning_rate": 4.996160098278676e-06,
+      "loss": 1.0515,
+      "step": 266
+    },
+    {
+      "epoch": 0.028080507973234123,
+      "grad_norm": 3.5344802127222783,
+      "learning_rate": 4.996113707079644e-06,
+      "loss": 1.0124,
+      "step": 267
+    },
+    {
+      "epoch": 0.02818567841508144,
+      "grad_norm": 4.089090640577849,
+      "learning_rate": 4.996067037544542e-06,
+      "loss": 1.0692,
+      "step": 268
+    },
+    {
+      "epoch": 0.02829084885692876,
+      "grad_norm": 3.2448736578587423,
+      "learning_rate": 4.996020089678575e-06,
+      "loss": 1.039,
+      "step": 269
+    },
+    {
+      "epoch": 0.02839601929877608,
+      "grad_norm": 4.037941021774882,
+      "learning_rate": 4.995972863486978e-06,
+      "loss": 1.08,
+      "step": 270
+    },
+    {
+      "epoch": 0.0285011897406234,
+      "grad_norm": 4.186559650612413,
+      "learning_rate": 4.9959253589750185e-06,
+      "loss": 1.0745,
+      "step": 271
+    },
+    {
+      "epoch": 0.028606360182470716,
+      "grad_norm": 2.7388820923632777,
+      "learning_rate": 4.995877576147993e-06,
+      "loss": 1.0651,
+      "step": 272
+    },
+    {
+      "epoch": 0.028711530624318034,
+      "grad_norm": 3.5856157401608666,
+      "learning_rate": 4.9958295150112275e-06,
+      "loss": 1.0874,
+      "step": 273
+    },
+    {
+      "epoch": 0.028816701066165355,
+      "grad_norm": 3.380461362281521,
+      "learning_rate": 4.995781175570083e-06,
+      "loss": 1.0709,
+      "step": 274
+    },
+    {
+      "epoch": 0.028921871508012672,
+      "grad_norm": 3.5335871406927724,
+      "learning_rate": 4.99573255782995e-06,
+      "loss": 1.0515,
+      "step": 275
+    },
+    {
+      "epoch": 0.029027041949859993,
+      "grad_norm": 4.463613237000542,
+      "learning_rate": 4.995683661796249e-06,
+      "loss": 1.0656,
+      "step": 276
+    },
+    {
+      "epoch": 0.02913221239170731,
+      "grad_norm": 3.4536408983322886,
+      "learning_rate": 4.995634487474433e-06,
+      "loss": 1.0725,
+      "step": 277
+    },
+    {
+      "epoch": 0.02923738283355463,
+      "grad_norm": 4.333343256417978,
+      "learning_rate": 4.995585034869984e-06,
+      "loss": 1.0408,
+      "step": 278
+    },
+    {
+      "epoch": 0.029342553275401948,
+      "grad_norm": 3.9790917572612847,
+      "learning_rate": 4.995535303988418e-06,
+      "loss": 1.0569,
+      "step": 279
+    },
+    {
+      "epoch": 0.029447723717249265,
+      "grad_norm": 3.5455400588602717,
+      "learning_rate": 4.99548529483528e-06,
+      "loss": 1.049,
+      "step": 280
+    },
+    {
+      "epoch": 0.029552894159096586,
+      "grad_norm": 2.336572942885481,
+      "learning_rate": 4.995435007416145e-06,
+      "loss": 1.0106,
+      "step": 281
+    },
+    {
+      "epoch": 0.029658064600943904,
+      "grad_norm": 4.019746471320661,
+      "learning_rate": 4.995384441736622e-06,
+      "loss": 1.0501,
+      "step": 282
+    },
+    {
+      "epoch": 0.029763235042791224,
+      "grad_norm": 5.0468126375485065,
+      "learning_rate": 4.995333597802349e-06,
+      "loss": 1.1031,
+      "step": 283
+    },
+    {
+      "epoch": 0.029868405484638542,
+      "grad_norm": 4.142494686730891,
+      "learning_rate": 4.995282475618995e-06,
+      "loss": 1.0887,
+      "step": 284
+    },
+    {
+      "epoch": 0.029973575926485863,
+      "grad_norm": 5.271308613066555,
+      "learning_rate": 4.9952310751922615e-06,
+      "loss": 1.0625,
+      "step": 285
+    },
+    {
+      "epoch": 0.03007874636833318,
+      "grad_norm": 4.001834907924138,
+      "learning_rate": 4.995179396527878e-06,
+      "loss": 1.0622,
+      "step": 286
+    },
+    {
+      "epoch": 0.030183916810180497,
+      "grad_norm": 3.9801706203568856,
+      "learning_rate": 4.99512743963161e-06,
+      "loss": 1.0791,
+      "step": 287
+    },
+    {
+      "epoch": 0.030289087252027818,
+      "grad_norm": 5.1051985033669425,
+      "learning_rate": 4.99507520450925e-06,
+      "loss": 1.0684,
+      "step": 288
+    },
+    {
+      "epoch": 0.030394257693875135,
+      "grad_norm": 3.936963708694301,
+      "learning_rate": 4.995022691166621e-06,
+      "loss": 1.047,
+      "step": 289
+    },
+    {
+      "epoch": 0.030499428135722456,
+      "grad_norm": 3.714404487265949,
+      "learning_rate": 4.994969899609581e-06,
+      "loss": 1.0303,
+      "step": 290
+    },
+    {
+      "epoch": 0.030604598577569773,
+      "grad_norm": 4.912241616953762,
+      "learning_rate": 4.994916829844015e-06,
+      "loss": 1.0836,
+      "step": 291
+    },
+    {
+      "epoch": 0.030709769019417094,
+      "grad_norm": 4.067161813822889,
+      "learning_rate": 4.994863481875842e-06,
+      "loss": 1.0335,
+      "step": 292
+    },
+    {
+      "epoch": 0.03081493946126441,
+      "grad_norm": 4.6662969482103644,
+      "learning_rate": 4.994809855711009e-06,
+      "loss": 1.0666,
+      "step": 293
+    },
+    {
+      "epoch": 0.03092010990311173,
+      "grad_norm": 4.517226678040155,
+      "learning_rate": 4.994755951355496e-06,
+      "loss": 1.0358,
+      "step": 294
+    },
+    {
+      "epoch": 0.03102528034495905,
+      "grad_norm": 3.7309332333912724,
+      "learning_rate": 4.994701768815317e-06,
+      "loss": 1.0611,
+      "step": 295
+    },
+    {
+      "epoch": 0.031130450786806367,
+      "grad_norm": 4.014104758735321,
+      "learning_rate": 4.994647308096509e-06,
+      "loss": 1.0727,
+      "step": 296
+    },
+    {
+      "epoch": 0.031235621228653688,
+      "grad_norm": 3.29886953781129,
+      "learning_rate": 4.994592569205148e-06,
+      "loss": 1.0919,
+      "step": 297
+    },
+    {
+      "epoch": 0.03134079167050101,
+      "grad_norm": 2.7078199003469257,
+      "learning_rate": 4.994537552147337e-06,
+      "loss": 1.0173,
+      "step": 298
+    },
+    {
+      "epoch": 0.031445962112348326,
+      "grad_norm": 2.8266073729377332,
+      "learning_rate": 4.994482256929211e-06,
+      "loss": 1.0709,
+      "step": 299
+    },
+    {
+      "epoch": 0.03155113255419564,
+      "grad_norm": 3.268397966576739,
+      "learning_rate": 4.994426683556935e-06,
+      "loss": 1.0217,
+      "step": 300
+    },
+    {
+      "epoch": 0.03165630299604296,
+      "grad_norm": 2.6632983909674994,
+      "learning_rate": 4.994370832036705e-06,
+      "loss": 1.0279,
+      "step": 301
+    },
+    {
+      "epoch": 0.03176147343789028,
+      "grad_norm": 4.0187186523919705,
+      "learning_rate": 4.994314702374753e-06,
+      "loss": 1.0646,
+      "step": 302
+    },
+    {
+      "epoch": 0.0318666438797376,
+      "grad_norm": 4.204278869511361,
+      "learning_rate": 4.994258294577333e-06,
+      "loss": 1.0762,
+      "step": 303
+    },
+    {
+      "epoch": 0.03197181432158492,
+      "grad_norm": 4.383109964538248,
+      "learning_rate": 4.9942016086507384e-06,
+      "loss": 1.0602,
+      "step": 304
+    },
+    {
+      "epoch": 0.03207698476343224,
+      "grad_norm": 5.459005697486776,
+      "learning_rate": 4.9941446446012874e-06,
+      "loss": 1.0854,
+      "step": 305
+    },
+    {
+      "epoch": 0.032182155205279554,
+      "grad_norm": 3.251384242833329,
+      "learning_rate": 4.994087402435335e-06,
+      "loss": 1.0001,
+      "step": 306
+    },
+    {
+      "epoch": 0.03228732564712687,
+      "grad_norm": 4.629466051075931,
+      "learning_rate": 4.994029882159261e-06,
+      "loss": 1.0552,
+      "step": 307
+    },
+    {
+      "epoch": 0.032392496088974196,
+      "grad_norm": 2.716529865958913,
+      "learning_rate": 4.993972083779482e-06,
+      "loss": 1.0108,
+      "step": 308
+    },
+    {
+      "epoch": 0.03249766653082151,
+      "grad_norm": 3.575967885671494,
+      "learning_rate": 4.993914007302441e-06,
+      "loss": 1.0709,
+      "step": 309
+    },
+    {
+      "epoch": 0.03260283697266883,
+      "grad_norm": 3.5669292643875186,
+      "learning_rate": 4.993855652734616e-06,
+      "loss": 1.0912,
+      "step": 310
+    },
+    {
+      "epoch": 0.03270800741451615,
+      "grad_norm": 4.010833762984559,
+      "learning_rate": 4.993797020082511e-06,
+      "loss": 1.0583,
+      "step": 311
+    },
+    {
+      "epoch": 0.03281317785636347,
+      "grad_norm": 4.104510377305269,
+      "learning_rate": 4.9937381093526675e-06,
+      "loss": 1.0087,
+      "step": 312
+    },
+    {
+      "epoch": 0.03291834829821079,
+      "grad_norm": 4.784979263043128,
+      "learning_rate": 4.993678920551651e-06,
+      "loss": 1.0615,
+      "step": 313
+    },
+    {
+      "epoch": 0.03302351874005811,
+      "grad_norm": 4.164191063473395,
+      "learning_rate": 4.993619453686065e-06,
+      "loss": 1.0409,
+      "step": 314
+    },
+    {
+      "epoch": 0.033128689181905424,
+      "grad_norm": 3.1649556775301733,
+      "learning_rate": 4.993559708762538e-06,
+      "loss": 1.0811,
+      "step": 315
+    },
+    {
+      "epoch": 0.03323385962375274,
+      "grad_norm": 3.6637842760946917,
+      "learning_rate": 4.993499685787732e-06,
+      "loss": 1.0539,
+      "step": 316
+    },
+    {
+      "epoch": 0.033339030065600066,
+      "grad_norm": 3.1821001996556157,
+      "learning_rate": 4.993439384768342e-06,
+      "loss": 1.0799,
+      "step": 317
+    },
+    {
+      "epoch": 0.03344420050744738,
+      "grad_norm": 3.3225742030256535,
+      "learning_rate": 4.993378805711091e-06,
+      "loss": 1.079,
+      "step": 318
+    },
+    {
+      "epoch": 0.0335493709492947,
+      "grad_norm": 3.7196848657078045,
+      "learning_rate": 4.993317948622733e-06,
+      "loss": 1.0537,
+      "step": 319
+    },
+    {
+      "epoch": 0.03365454139114202,
+      "grad_norm": 3.0347036015285385,
+      "learning_rate": 4.993256813510055e-06,
+      "loss": 1.0425,
+      "step": 320
+    },
+    {
+      "epoch": 0.033759711832989335,
+      "grad_norm": 3.4466877790068056,
+      "learning_rate": 4.993195400379875e-06,
+      "loss": 1.0275,
+      "step": 321
+    },
+    {
+      "epoch": 0.03386488227483666,
+      "grad_norm": 3.4937996524768393,
+      "learning_rate": 4.993133709239039e-06,
+      "loss": 1.0074,
+      "step": 322
+    },
+    {
+      "epoch": 0.03397005271668398,
+      "grad_norm": 3.961714485473967,
+      "learning_rate": 4.993071740094428e-06,
+      "loss": 1.0152,
+      "step": 323
+    },
+    {
+      "epoch": 0.034075223158531294,
+      "grad_norm": 3.882633562046403,
+      "learning_rate": 4.993009492952951e-06,
+      "loss": 1.0408,
+      "step": 324
+    },
+    {
+      "epoch": 0.03418039360037861,
+      "grad_norm": 4.777184945580783,
+      "learning_rate": 4.992946967821549e-06,
+      "loss": 1.095,
+      "step": 325
+    },
+    {
+      "epoch": 0.034285564042225936,
+      "grad_norm": 4.8588365098982775,
+      "learning_rate": 4.9928841647071936e-06,
+      "loss": 1.0568,
+      "step": 326
+    },
+    {
+      "epoch": 0.03439073448407325,
+      "grad_norm": 3.6839425202673297,
+      "learning_rate": 4.992821083616889e-06,
+      "loss": 1.0496,
+      "step": 327
+    },
+    {
+      "epoch": 0.03449590492592057,
+      "grad_norm": 3.66693362260413,
+      "learning_rate": 4.992757724557669e-06,
+      "loss": 1.0753,
+      "step": 328
+    },
+    {
+      "epoch": 0.03460107536776789,
+      "grad_norm": 4.978727167916742,
+      "learning_rate": 4.992694087536597e-06,
+      "loss": 1.0729,
+      "step": 329
+    },
+    {
+      "epoch": 0.034706245809615205,
+      "grad_norm": 5.114707679163974,
+      "learning_rate": 4.992630172560771e-06,
+      "loss": 1.0943,
+      "step": 330
+    },
+    {
+      "epoch": 0.03481141625146253,
+      "grad_norm": 3.163523736651089,
+      "learning_rate": 4.992565979637318e-06,
+      "loss": 1.0419,
+      "step": 331
+    },
+    {
+      "epoch": 0.034916586693309846,
+      "grad_norm": 3.636944273388002,
+      "learning_rate": 4.9925015087733945e-06,
+      "loss": 1.047,
+      "step": 332
+    },
+    {
+      "epoch": 0.035021757135157164,
+      "grad_norm": 4.259395764869889,
+      "learning_rate": 4.992436759976191e-06,
+      "loss": 1.0776,
+      "step": 333
+    },
+    {
+      "epoch": 0.03512692757700448,
+      "grad_norm": 3.920573922209101,
+      "learning_rate": 4.992371733252926e-06,
+      "loss": 1.0535,
+      "step": 334
+    },
+    {
+      "epoch": 0.0352320980188518,
+      "grad_norm": 4.293490908646825,
+      "learning_rate": 4.992306428610853e-06,
+      "loss": 1.0483,
+      "step": 335
+    },
+    {
+      "epoch": 0.03533726846069912,
+      "grad_norm": 3.6789152142394914,
+      "learning_rate": 4.9922408460572505e-06,
+      "loss": 1.0418,
+      "step": 336
+    },
+    {
+      "epoch": 0.03544243890254644,
+      "grad_norm": 3.3817970301847238,
+      "learning_rate": 4.992174985599434e-06,
+      "loss": 1.0514,
+      "step": 337
+    },
+    {
+      "epoch": 0.03554760934439376,
+      "grad_norm": 3.3702499323179596,
+      "learning_rate": 4.992108847244746e-06,
+      "loss": 1.0259,
+      "step": 338
+    },
+    {
+      "epoch": 0.035652779786241075,
+      "grad_norm": 4.5111317152835815,
+      "learning_rate": 4.9920424310005625e-06,
+      "loss": 1.0555,
+      "step": 339
+    },
+    {
+      "epoch": 0.0357579502280884,
+      "grad_norm": 4.693632339648465,
+      "learning_rate": 4.9919757368742895e-06,
+      "loss": 1.0464,
+      "step": 340
+    },
+    {
+      "epoch": 0.035863120669935716,
+      "grad_norm": 3.7052320429842736,
+      "learning_rate": 4.9919087648733634e-06,
+      "loss": 1.0429,
+      "step": 341
+    },
+    {
+      "epoch": 0.035968291111783034,
+      "grad_norm": 5.043659801808216,
+      "learning_rate": 4.991841515005253e-06,
+      "loss": 1.0648,
+      "step": 342
+    },
+    {
+      "epoch": 0.03607346155363035,
+      "grad_norm": 4.146978032877157,
+      "learning_rate": 4.991773987277455e-06,
+      "loss": 1.1074,
+      "step": 343
+    },
+    {
+      "epoch": 0.03617863199547767,
+      "grad_norm": 6.009958045778625,
+      "learning_rate": 4.991706181697501e-06,
+      "loss": 1.0836,
+      "step": 344
+    },
+    {
+      "epoch": 0.03628380243732499,
+      "grad_norm": 4.730758307272077,
+      "learning_rate": 4.991638098272951e-06,
+      "loss": 1.0436,
+      "step": 345
+    },
+    {
+      "epoch": 0.03638897287917231,
+      "grad_norm": 4.19923954684141,
+      "learning_rate": 4.991569737011398e-06,
+      "loss": 1.0276,
+      "step": 346
+    },
+    {
+      "epoch": 0.03649414332101963,
+      "grad_norm": 3.5491173146821247,
+      "learning_rate": 4.991501097920464e-06,
+      "loss": 1.0366,
+      "step": 347
+    },
+    {
+      "epoch": 0.036599313762866945,
+      "grad_norm": 3.4759309555067213,
+      "learning_rate": 4.991432181007803e-06,
+      "loss": 1.0173,
+      "step": 348
+    },
+    {
+      "epoch": 0.03670448420471426,
+      "grad_norm": 3.817356633667997,
+      "learning_rate": 4.9913629862811e-06,
+      "loss": 1.0628,
+      "step": 349
+    },
+    {
+      "epoch": 0.036809654646561586,
+      "grad_norm": 3.537846598304811,
+      "learning_rate": 4.99129351374807e-06,
+      "loss": 1.057,
+      "step": 350
+    },
+    {
+      "epoch": 0.036914825088408904,
+      "grad_norm": 5.421441758466292,
+      "learning_rate": 4.99122376341646e-06,
+      "loss": 1.0986,
+      "step": 351
+    },
+    {
+      "epoch": 0.03701999553025622,
+      "grad_norm": 4.12752414216299,
+      "learning_rate": 4.9911537352940485e-06,
+      "loss": 1.0398,
+      "step": 352
+    },
+    {
+      "epoch": 0.03712516597210354,
+      "grad_norm": 4.437838613524334,
+      "learning_rate": 4.991083429388643e-06,
+      "loss": 1.0359,
+      "step": 353
+    },
+    {
+      "epoch": 0.03723033641395086,
+      "grad_norm": 3.22408613692183,
+      "learning_rate": 4.991012845708084e-06,
+      "loss": 1.0653,
+      "step": 354
+    },
+    {
+      "epoch": 0.03733550685579818,
+      "grad_norm": 3.3399036692465884,
+      "learning_rate": 4.990941984260241e-06,
+      "loss": 1.0765,
+      "step": 355
+    },
+    {
+      "epoch": 0.0374406772976455,
+      "grad_norm": 3.936416890688034,
+      "learning_rate": 4.9908708450530174e-06,
+      "loss": 1.051,
+      "step": 356
+    },
+    {
+      "epoch": 0.037545847739492814,
+      "grad_norm": 3.990025349268498,
+      "learning_rate": 4.990799428094345e-06,
+      "loss": 1.0507,
+      "step": 357
+    },
+    {
+      "epoch": 0.03765101818134013,
+      "grad_norm": 4.239682909525811,
+      "learning_rate": 4.9907277333921865e-06,
+      "loss": 1.0568,
+      "step": 358
+    },
+    {
+      "epoch": 0.037756188623187456,
+      "grad_norm": 3.7354470248901133,
+      "learning_rate": 4.9906557609545375e-06,
+      "loss": 1.0847,
+      "step": 359
+    },
+    {
+      "epoch": 0.03786135906503477,
+      "grad_norm": 5.431410350419551,
+      "learning_rate": 4.990583510789423e-06,
+      "loss": 1.088,
+      "step": 360
+    },
+    {
+      "epoch": 0.03796652950688209,
+      "grad_norm": 4.049632017073229,
+      "learning_rate": 4.990510982904899e-06,
+      "loss": 1.0853,
+      "step": 361
+    },
+    {
+      "epoch": 0.03807169994872941,
+      "grad_norm": 3.8736875944016775,
+      "learning_rate": 4.990438177309054e-06,
+      "loss": 1.0628,
+      "step": 362
+    },
+    {
+      "epoch": 0.038176870390576725,
+      "grad_norm": 2.99774541371039,
+      "learning_rate": 4.9903650940100055e-06,
+      "loss": 1.0509,
+      "step": 363
+    },
+    {
+      "epoch": 0.03828204083242405,
+      "grad_norm": 5.319456457308596,
+      "learning_rate": 4.990291733015904e-06,
+      "loss": 1.037,
+      "step": 364
+    },
+    {
+      "epoch": 0.03838721127427137,
+      "grad_norm": 4.702856139460145,
+      "learning_rate": 4.9902180943349286e-06,
+      "loss": 1.0805,
+      "step": 365
+    },
+    {
+      "epoch": 0.038492381716118684,
+      "grad_norm": 3.8095265476696665,
+      "learning_rate": 4.99014417797529e-06,
+      "loss": 1.0669,
+      "step": 366
+    },
+    {
+      "epoch": 0.038597552157966,
+      "grad_norm": 4.961192575518297,
+      "learning_rate": 4.990069983945233e-06,
+      "loss": 1.0844,
+      "step": 367
+    },
+    {
+      "epoch": 0.038702722599813326,
+      "grad_norm": 4.891981596003543,
+      "learning_rate": 4.989995512253028e-06,
+      "loss": 1.0502,
+      "step": 368
+    },
+    {
+      "epoch": 0.03880789304166064,
+      "grad_norm": 3.6566189119364667,
+      "learning_rate": 4.989920762906981e-06,
+      "loss": 1.06,
+      "step": 369
+    },
+    {
+      "epoch": 0.03891306348350796,
+      "grad_norm": 5.900942628114732,
+      "learning_rate": 4.989845735915426e-06,
+      "loss": 1.068,
+      "step": 370
+    },
+    {
+      "epoch": 0.03901823392535528,
+      "grad_norm": 3.8108566487033477,
+      "learning_rate": 4.98977043128673e-06,
+      "loss": 1.0531,
+      "step": 371
+    },
+    {
+      "epoch": 0.039123404367202595,
+      "grad_norm": 3.9068966220872006,
+      "learning_rate": 4.9896948490292905e-06,
+      "loss": 1.064,
+      "step": 372
+    },
+    {
+      "epoch": 0.03922857480904992,
+      "grad_norm": 4.376179595496623,
+      "learning_rate": 4.989618989151533e-06,
+      "loss": 1.0025,
+      "step": 373
+    },
+    {
+      "epoch": 0.03933374525089724,
+      "grad_norm": 3.6380450371166155,
+      "learning_rate": 4.989542851661919e-06,
+      "loss": 1.0173,
+      "step": 374
+    },
+    {
+      "epoch": 0.039438915692744554,
+      "grad_norm": 4.554841578105061,
+      "learning_rate": 4.9894664365689385e-06,
+      "loss": 1.0903,
+      "step": 375
+    },
+    {
+      "epoch": 0.03954408613459187,
+      "grad_norm": 2.853296420735663,
+      "learning_rate": 4.9893897438811106e-06,
+      "loss": 1.0246,
+      "step": 376
+    },
+    {
+      "epoch": 0.03964925657643919,
+      "grad_norm": 4.257333189026485,
+      "learning_rate": 4.989312773606988e-06,
+      "loss": 1.097,
+      "step": 377
+    },
+    {
+      "epoch": 0.03975442701828651,
+      "grad_norm": 3.678981492068401,
+      "learning_rate": 4.989235525755154e-06,
+      "loss": 1.0611,
+      "step": 378
+    },
+    {
+      "epoch": 0.03985959746013383,
+      "grad_norm": 5.195313256569016,
+      "learning_rate": 4.989158000334221e-06,
+      "loss": 1.0454,
+      "step": 379
+    },
+    {
+      "epoch": 0.03996476790198115,
+      "grad_norm": 1.8104051625470303,
+      "learning_rate": 4.989080197352834e-06,
+      "loss": 1.0422,
+      "step": 380
+    },
+    {
+      "epoch": 0.040069938343828465,
+      "grad_norm": 3.538275104798571,
+      "learning_rate": 4.989002116819671e-06,
+      "loss": 1.0597,
+      "step": 381
+    },
+    {
+      "epoch": 0.04017510878567578,
+      "grad_norm": 2.8012444815542716,
+      "learning_rate": 4.9889237587434355e-06,
+      "loss": 1.0183,
+      "step": 382
+    },
+    {
+      "epoch": 0.04028027922752311,
+      "grad_norm": 3.367459724491466,
+      "learning_rate": 4.988845123132867e-06,
+      "loss": 1.0388,
+      "step": 383
+    },
+    {
+      "epoch": 0.040385449669370424,
+      "grad_norm": 4.145106659583553,
+      "learning_rate": 4.988766209996733e-06,
+      "loss": 1.0344,
+      "step": 384
+    },
+    {
+      "epoch": 0.04049062011121774,
+      "grad_norm": 3.40206472561461,
+      "learning_rate": 4.9886870193438325e-06,
+      "loss": 1.052,
+      "step": 385
+    },
+    {
+      "epoch": 0.04059579055306506,
+      "grad_norm": 3.725684215375892,
+      "learning_rate": 4.988607551182997e-06,
+      "loss": 1.0809,
+      "step": 386
+    },
+    {
+      "epoch": 0.04070096099491238,
+      "grad_norm": 3.731707512512842,
+      "learning_rate": 4.9885278055230875e-06,
+      "loss": 1.0266,
+      "step": 387
+    },
+    {
+      "epoch": 0.0408061314367597,
+      "grad_norm": 3.928396974689701,
+      "learning_rate": 4.988447782372996e-06,
+      "loss": 1.063,
+      "step": 388
+    },
+    {
+      "epoch": 0.04091130187860702,
+      "grad_norm": 4.31218060090711,
+      "learning_rate": 4.9883674817416464e-06,
+      "loss": 1.0864,
+      "step": 389
+    },
+    {
+      "epoch": 0.041016472320454335,
+      "grad_norm": 3.0718531395576583,
+      "learning_rate": 4.988286903637991e-06,
+      "loss": 1.0459,
+      "step": 390
+    },
+    {
+      "epoch": 0.04112164276230165,
+      "grad_norm": 2.9157630700754886,
+      "learning_rate": 4.988206048071017e-06,
+      "loss": 1.0581,
+      "step": 391
+    },
+    {
+      "epoch": 0.04122681320414898,
+      "grad_norm": 3.8247937216792,
+      "learning_rate": 4.988124915049739e-06,
+      "loss": 1.0286,
+      "step": 392
+    },
+    {
+      "epoch": 0.041331983645996294,
+      "grad_norm": 2.6874655896655097,
+      "learning_rate": 4.988043504583206e-06,
+      "loss": 1.0387,
+      "step": 393
+    },
+    {
+      "epoch": 0.04143715408784361,
+      "grad_norm": 2.9913671043398677,
+      "learning_rate": 4.987961816680493e-06,
+      "loss": 1.0536,
+      "step": 394
+    },
+    {
+      "epoch": 0.04154232452969093,
+      "grad_norm": 2.8088862124300404,
+      "learning_rate": 4.987879851350711e-06,
+      "loss": 1.078,
+      "step": 395
+    },
+    {
+      "epoch": 0.041647494971538246,
+      "grad_norm": 3.755585245451105,
+      "learning_rate": 4.987797608602997e-06,
+      "loss": 1.0792,
+      "step": 396
+    },
+    {
+      "epoch": 0.04175266541338557,
+      "grad_norm": 2.8770765754045846,
+      "learning_rate": 4.987715088446526e-06,
+      "loss": 1.0757,
+      "step": 397
+    },
+    {
+      "epoch": 0.04185783585523289,
+      "grad_norm": 3.273137502812656,
+      "learning_rate": 4.987632290890497e-06,
+      "loss": 1.0471,
+      "step": 398
+    },
+    {
+      "epoch": 0.041963006297080205,
+      "grad_norm": 3.636240730716032,
+      "learning_rate": 4.987549215944143e-06,
+      "loss": 1.0532,
+      "step": 399
+    },
+    {
+      "epoch": 0.04206817673892752,
+      "grad_norm": 4.208273895144222,
+      "learning_rate": 4.987465863616727e-06,
+      "loss": 1.0681,
+      "step": 400
+    },
+    {
+      "epoch": 0.042173347180774846,
+      "grad_norm": 3.5106533475647748,
+      "learning_rate": 4.987382233917545e-06,
+      "loss": 1.0538,
+      "step": 401
+    },
+    {
+      "epoch": 0.042278517622622164,
+      "grad_norm": 4.886913187088557,
+      "learning_rate": 4.98729832685592e-06,
+      "loss": 1.0417,
+      "step": 402
+    },
+    {
+      "epoch": 0.04238368806446948,
+      "grad_norm": 5.560665522659327,
+      "learning_rate": 4.98721414244121e-06,
+      "loss": 1.0699,
+      "step": 403
+    },
+    {
+      "epoch": 0.0424888585063168,
+      "grad_norm": 4.094323716122683,
+      "learning_rate": 4.987129680682802e-06,
+      "loss": 1.0764,
+      "step": 404
+    },
+    {
+      "epoch": 0.042594028948164116,
+      "grad_norm": 2.6629248986818075,
+      "learning_rate": 4.987044941590114e-06,
+      "loss": 1.0667,
+      "step": 405
+    },
+    {
+      "epoch": 0.04269919939001144,
+      "grad_norm": 3.4109848570390193,
+      "learning_rate": 4.986959925172594e-06,
+      "loss": 1.037,
+      "step": 406
+    },
+    {
+      "epoch": 0.04280436983185876,
+      "grad_norm": 3.0342413572937317,
+      "learning_rate": 4.986874631439724e-06,
+      "loss": 1.0709,
+      "step": 407
+    },
+    {
+      "epoch": 0.042909540273706075,
+      "grad_norm": 2.8390095815269465,
+      "learning_rate": 4.9867890604010125e-06,
+      "loss": 1.0196,
+      "step": 408
+    },
+    {
+      "epoch": 0.04301471071555339,
+      "grad_norm": 3.0707226746539464,
+      "learning_rate": 4.9867032120660045e-06,
+      "loss": 1.0393,
+      "step": 409
+    },
+    {
+      "epoch": 0.04311988115740071,
+      "grad_norm": 3.8038872687449765,
+      "learning_rate": 4.98661708644427e-06,
+      "loss": 1.032,
+      "step": 410
+    },
+    {
+      "epoch": 0.043225051599248034,
+      "grad_norm": 3.6068516146454193,
+      "learning_rate": 4.986530683545414e-06,
+      "loss": 1.0328,
+      "step": 411
+    },
+    {
+      "epoch": 0.04333022204109535,
+      "grad_norm": 3.8245339487596546,
+      "learning_rate": 4.98644400337907e-06,
+      "loss": 1.103,
+      "step": 412
+    },
+    {
+      "epoch": 0.04343539248294267,
+      "grad_norm": 3.4368283671133746,
+      "learning_rate": 4.986357045954905e-06,
+      "loss": 1.0507,
+      "step": 413
+    },
+    {
+      "epoch": 0.043540562924789986,
+      "grad_norm": 2.74188819425057,
+      "learning_rate": 4.986269811282615e-06,
+      "loss": 1.0316,
+      "step": 414
+    },
+    {
+      "epoch": 0.04364573336663731,
+      "grad_norm": 3.679825151599156,
+      "learning_rate": 4.9861822993719255e-06,
+      "loss": 1.0097,
+      "step": 415
+    },
+    {
+      "epoch": 0.04375090380848463,
+      "grad_norm": 3.297855659802533,
+      "learning_rate": 4.986094510232597e-06,
+      "loss": 1.0219,
+      "step": 416
+    },
+    {
+      "epoch": 0.043856074250331945,
+      "grad_norm": 3.4058245947376173,
+      "learning_rate": 4.986006443874419e-06,
+      "loss": 1.0377,
+      "step": 417
+    },
+    {
+      "epoch": 0.04396124469217926,
+      "grad_norm": 4.363555255178052,
+      "learning_rate": 4.9859181003072095e-06,
+      "loss": 1.0637,
+      "step": 418
+    },
+    {
+      "epoch": 0.04406641513402658,
+      "grad_norm": 3.9928133996636395,
+      "learning_rate": 4.985829479540821e-06,
+      "loss": 1.0591,
+      "step": 419
+    },
+    {
+      "epoch": 0.0441715855758739,
+      "grad_norm": 4.073405278783087,
+      "learning_rate": 4.985740581585135e-06,
+      "loss": 1.059,
+      "step": 420
+    },
+    {
+      "epoch": 0.04427675601772122,
+      "grad_norm": 4.084709791211826,
+      "learning_rate": 4.985651406450063e-06,
+      "loss": 1.0313,
+      "step": 421
+    },
+    {
+      "epoch": 0.04438192645956854,
+      "grad_norm": 3.715460139932838,
+      "learning_rate": 4.985561954145552e-06,
+      "loss": 1.0913,
+      "step": 422
+    },
+    {
+      "epoch": 0.044487096901415855,
+      "grad_norm": 3.7233678758471744,
+      "learning_rate": 4.985472224681573e-06,
+      "loss": 1.0577,
+      "step": 423
+    },
+    {
+      "epoch": 0.04459226734326317,
+      "grad_norm": 5.652763390578416,
+      "learning_rate": 4.985382218068133e-06,
+      "loss": 1.0382,
+      "step": 424
+    },
+    {
+      "epoch": 0.0446974377851105,
+      "grad_norm": 3.468271207036182,
+      "learning_rate": 4.985291934315269e-06,
+      "loss": 1.0498,
+      "step": 425
+    },
+    {
+      "epoch": 0.044802608226957814,
+      "grad_norm": 3.1935939850144233,
+      "learning_rate": 4.985201373433048e-06,
+      "loss": 1.0395,
+      "step": 426
+    },
+    {
+      "epoch": 0.04490777866880513,
+      "grad_norm": 3.714801883999893,
+      "learning_rate": 4.985110535431569e-06,
+      "loss": 1.063,
+      "step": 427
+    },
+    {
+      "epoch": 0.04501294911065245,
+      "grad_norm": 3.441136685461427,
+      "learning_rate": 4.985019420320959e-06,
+      "loss": 1.0503,
+      "step": 428
+    },
+    {
+      "epoch": 0.04511811955249977,
+      "grad_norm": 2.8419115986373042,
+      "learning_rate": 4.98492802811138e-06,
+      "loss": 1.0479,
+      "step": 429
+    },
+    {
+      "epoch": 0.04522328999434709,
+      "grad_norm": 3.039904416144775,
+      "learning_rate": 4.984836358813022e-06,
+      "loss": 1.0469,
+      "step": 430
+    },
+    {
+      "epoch": 0.04532846043619441,
+      "grad_norm": 4.691661020519824,
+      "learning_rate": 4.984744412436107e-06,
+      "loss": 1.0572,
+      "step": 431
+    },
+    {
+      "epoch": 0.045433630878041725,
+      "grad_norm": 4.540987997363801,
+      "learning_rate": 4.9846521889908885e-06,
+      "loss": 1.0781,
+      "step": 432
+    },
+    {
+      "epoch": 0.04553880131988904,
+      "grad_norm": 4.296517950327922,
+      "learning_rate": 4.984559688487649e-06,
+      "loss": 1.0635,
+      "step": 433
+    },
+    {
+      "epoch": 0.04564397176173637,
+      "grad_norm": 3.2572651182872585,
+      "learning_rate": 4.984466910936703e-06,
+      "loss": 1.0681,
+      "step": 434
+    },
+    {
+      "epoch": 0.045749142203583684,
+      "grad_norm": 2.9557462605367415,
+      "learning_rate": 4.9843738563483975e-06,
+      "loss": 1.0567,
+      "step": 435
+    },
+    {
+      "epoch": 0.045854312645431,
+      "grad_norm": 3.9469604897643022,
+      "learning_rate": 4.984280524733107e-06,
+      "loss": 1.0509,
+      "step": 436
+    },
+    {
+      "epoch": 0.04595948308727832,
+      "grad_norm": 2.7541457544733188,
+      "learning_rate": 4.984186916101239e-06,
+      "loss": 1.0414,
+      "step": 437
+    },
+    {
+      "epoch": 0.046064653529125636,
+      "grad_norm": 3.622690550239156,
+      "learning_rate": 4.984093030463233e-06,
+      "loss": 1.0864,
+      "step": 438
+    },
+    {
+      "epoch": 0.04616982397097296,
+      "grad_norm": 4.32000715987469,
+      "learning_rate": 4.983998867829557e-06,
+      "loss": 1.0398,
+      "step": 439
+    },
+    {
+      "epoch": 0.04627499441282028,
+      "grad_norm": 3.7777411421192792,
+      "learning_rate": 4.98390442821071e-06,
+      "loss": 1.0565,
+      "step": 440
+    },
+    {
+      "epoch": 0.046380164854667595,
+      "grad_norm": 3.47497239447561,
+      "learning_rate": 4.983809711617224e-06,
+      "loss": 1.0516,
+      "step": 441
+    },
+    {
+      "epoch": 0.04648533529651491,
+      "grad_norm": 3.8957477430994687,
+      "learning_rate": 4.98371471805966e-06,
+      "loss": 1.0513,
+      "step": 442
+    },
+    {
+      "epoch": 0.04659050573836224,
+      "grad_norm": 2.4570852022191603,
+      "learning_rate": 4.983619447548611e-06,
+      "loss": 1.0278,
+      "step": 443
+    },
+    {
+      "epoch": 0.046695676180209554,
+      "grad_norm": 4.079822189897958,
+      "learning_rate": 4.9835239000947005e-06,
+      "loss": 1.0827,
+      "step": 444
+    },
+    {
+      "epoch": 0.04680084662205687,
+      "grad_norm": 3.547742042978678,
+      "learning_rate": 4.983428075708582e-06,
+      "loss": 1.1007,
+      "step": 445
+    },
+    {
+      "epoch": 0.04690601706390419,
+      "grad_norm": 4.0885742408881125,
+      "learning_rate": 4.983331974400941e-06,
+      "loss": 1.0316,
+      "step": 446
+    },
+    {
+      "epoch": 0.047011187505751506,
+      "grad_norm": 3.103253885780263,
+      "learning_rate": 4.983235596182494e-06,
+      "loss": 1.0792,
+      "step": 447
+    },
+    {
+      "epoch": 0.04711635794759883,
+      "grad_norm": 2.6930299837503164,
+      "learning_rate": 4.983138941063988e-06,
+      "loss": 1.0769,
+      "step": 448
+    },
+    {
+      "epoch": 0.04722152838944615,
+      "grad_norm": 2.828513719063435,
+      "learning_rate": 4.9830420090562e-06,
+      "loss": 1.0662,
+      "step": 449
+    },
+    {
+      "epoch": 0.047326698831293465,
+      "grad_norm": 3.6605584013370245,
+      "learning_rate": 4.982944800169939e-06,
+      "loss": 1.0513,
+      "step": 450
+    },
+    {
+      "epoch": 0.04743186927314078,
+      "grad_norm": 2.902862443517808,
+      "learning_rate": 4.982847314416044e-06,
+      "loss": 1.0157,
+      "step": 451
+    },
+    {
+      "epoch": 0.0475370397149881,
+      "grad_norm": 2.8680844974252686,
+      "learning_rate": 4.982749551805387e-06,
+      "loss": 1.0493,
+      "step": 452
+    },
+    {
+      "epoch": 0.047642210156835424,
+      "grad_norm": 2.8491457922180197,
+      "learning_rate": 4.9826515123488675e-06,
+      "loss": 1.0603,
+      "step": 453
+    },
+    {
+      "epoch": 0.04774738059868274,
+      "grad_norm": 3.0486623847828334,
+      "learning_rate": 4.982553196057419e-06,
+      "loss": 1.0604,
+      "step": 454
+    },
+    {
+      "epoch": 0.04785255104053006,
+      "grad_norm": 2.2871258606215874,
+      "learning_rate": 4.982454602942004e-06,
+      "loss": 1.0367,
+      "step": 455
+    },
+    {
+      "epoch": 0.047957721482377376,
+      "grad_norm": 3.822855535156918,
+      "learning_rate": 4.982355733013616e-06,
+      "loss": 1.0289,
+      "step": 456
+    },
+    {
+      "epoch": 0.04806289192422469,
+      "grad_norm": 3.4057137793165486,
+      "learning_rate": 4.98225658628328e-06,
+      "loss": 1.0331,
+      "step": 457
+    },
+    {
+      "epoch": 0.04816806236607202,
+      "grad_norm": 4.014699719424169,
+      "learning_rate": 4.982157162762052e-06,
+      "loss": 1.0255,
+      "step": 458
+    },
+    {
+      "epoch": 0.048273232807919335,
+      "grad_norm": 4.028054004330074,
+      "learning_rate": 4.982057462461018e-06,
+      "loss": 1.0444,
+      "step": 459
+    },
+    {
+      "epoch": 0.04837840324976665,
+      "grad_norm": 5.391630404479916,
+      "learning_rate": 4.981957485391297e-06,
+      "loss": 1.0536,
+      "step": 460
+    },
+    {
+      "epoch": 0.04848357369161397,
+      "grad_norm": 2.3994992334050105,
+      "learning_rate": 4.981857231564034e-06,
+      "loss": 1.0103,
+      "step": 461
+    },
+    {
+      "epoch": 0.048588744133461294,
+      "grad_norm": 3.798097639454563,
+      "learning_rate": 4.9817567009904114e-06,
+      "loss": 1.0238,
+      "step": 462
+    },
+    {
+      "epoch": 0.04869391457530861,
+      "grad_norm": 4.714715898953462,
+      "learning_rate": 4.981655893681637e-06,
+      "loss": 1.0237,
+      "step": 463
+    },
+    {
+      "epoch": 0.04879908501715593,
+      "grad_norm": 4.388735872403405,
+      "learning_rate": 4.981554809648952e-06,
+      "loss": 1.076,
+      "step": 464
+    },
+    {
+      "epoch": 0.048904255459003246,
+      "grad_norm": 4.403273076780409,
+      "learning_rate": 4.981453448903629e-06,
+      "loss": 1.0737,
+      "step": 465
+    },
+    {
+      "epoch": 0.04900942590085056,
+      "grad_norm": 2.5521629736975404,
+      "learning_rate": 4.98135181145697e-06,
+      "loss": 1.0514,
+      "step": 466
+    },
+    {
+      "epoch": 0.04911459634269789,
+      "grad_norm": 3.8606552004883534,
+      "learning_rate": 4.981249897320307e-06,
+      "loss": 1.0413,
+      "step": 467
+    },
+    {
+      "epoch": 0.049219766784545205,
+      "grad_norm": 2.8692792806121536,
+      "learning_rate": 4.981147706505007e-06,
+      "loss": 1.0463,
+      "step": 468
+    },
+    {
+      "epoch": 0.04932493722639252,
+      "grad_norm": 2.7409626254110315,
+      "learning_rate": 4.981045239022462e-06,
+      "loss": 1.0609,
+      "step": 469
+    },
+    {
+      "epoch": 0.04943010766823984,
+      "grad_norm": 2.9586823141347707,
+      "learning_rate": 4.9809424948841e-06,
+      "loss": 1.0576,
+      "step": 470
+    },
+    {
+      "epoch": 0.04953527811008716,
+      "grad_norm": 2.6686230862505886,
+      "learning_rate": 4.980839474101378e-06,
+      "loss": 1.0157,
+      "step": 471
+    },
+    {
+      "epoch": 0.04964044855193448,
+      "grad_norm": 3.139097480801446,
+      "learning_rate": 4.980736176685782e-06,
+      "loss": 1.0454,
+      "step": 472
+    },
+    {
+      "epoch": 0.0497456189937818,
+      "grad_norm": 3.100054655774797,
+      "learning_rate": 4.980632602648831e-06,
+      "loss": 1.0296,
+      "step": 473
+    },
+    {
+      "epoch": 0.049850789435629116,
+      "grad_norm": 4.665650390709168,
+      "learning_rate": 4.980528752002075e-06,
+      "loss": 1.0421,
+      "step": 474
+    },
+    {
+      "epoch": 0.04995595987747643,
+      "grad_norm": 3.2001730382677067,
+      "learning_rate": 4.9804246247570935e-06,
+      "loss": 1.0599,
+      "step": 475
+    },
+    {
+      "epoch": 0.05006113031932376,
+      "grad_norm": 3.5026752591463493,
+      "learning_rate": 4.980320220925498e-06,
+      "loss": 1.0683,
+      "step": 476
+    },
+    {
+      "epoch": 0.050166300761171075,
+      "grad_norm": 4.538529420292572,
+      "learning_rate": 4.98021554051893e-06,
+      "loss": 1.0342,
+      "step": 477
+    },
+    {
+      "epoch": 0.05027147120301839,
+      "grad_norm": 2.9057622514006374,
+      "learning_rate": 4.980110583549062e-06,
+      "loss": 1.0419,
+      "step": 478
+    },
+    {
+      "epoch": 0.05037664164486571,
+      "grad_norm": 4.168498583654285,
+      "learning_rate": 4.980005350027598e-06,
+      "loss": 1.0235,
+      "step": 479
+    },
+    {
+      "epoch": 0.05048181208671303,
+      "grad_norm": 3.465721454429658,
+      "learning_rate": 4.979899839966273e-06,
+      "loss": 1.0825,
+      "step": 480
+    },
+    {
+      "epoch": 0.05058698252856035,
+      "grad_norm": 3.9874964500146994,
+      "learning_rate": 4.97979405337685e-06,
+      "loss": 1.0683,
+      "step": 481
+    },
+    {
+      "epoch": 0.05069215297040767,
+      "grad_norm": 3.256605985477986,
+      "learning_rate": 4.979687990271126e-06,
+      "loss": 1.0309,
+      "step": 482
+    },
+    {
+      "epoch": 0.050797323412254985,
+      "grad_norm": 2.822984342445266,
+      "learning_rate": 4.979581650660929e-06,
+      "loss": 1.0334,
+      "step": 483
+    },
+    {
+      "epoch": 0.0509024938541023,
+      "grad_norm": 4.047140521535769,
+      "learning_rate": 4.979475034558115e-06,
+      "loss": 1.0336,
+      "step": 484
+    },
+    {
+      "epoch": 0.05100766429594962,
+      "grad_norm": 3.121659309179012,
+      "learning_rate": 4.979368141974575e-06,
+      "loss": 1.0975,
+      "step": 485
+    },
+    {
+      "epoch": 0.051112834737796944,
+      "grad_norm": 4.436709845774703,
+      "learning_rate": 4.979260972922226e-06,
+      "loss": 1.0085,
+      "step": 486
+    },
+    {
+      "epoch": 0.05121800517964426,
+      "grad_norm": 5.131430337585738,
+      "learning_rate": 4.9791535274130185e-06,
+      "loss": 1.0919,
+      "step": 487
+    },
+    {
+      "epoch": 0.05132317562149158,
+      "grad_norm": 4.09555529702269,
+      "learning_rate": 4.9790458054589344e-06,
+      "loss": 1.0877,
+      "step": 488
+    },
+    {
+      "epoch": 0.051428346063338896,
+      "grad_norm": 3.287984548144607,
+      "learning_rate": 4.9789378070719854e-06,
+      "loss": 1.0707,
+      "step": 489
+    },
+    {
+      "epoch": 0.05153351650518622,
+      "grad_norm": 3.3946741219834977,
+      "learning_rate": 4.978829532264213e-06,
+      "loss": 1.055,
+      "step": 490
+    },
+    {
+      "epoch": 0.05163868694703354,
+      "grad_norm": 3.058713885707903,
+      "learning_rate": 4.978720981047692e-06,
+      "loss": 1.0804,
+      "step": 491
+    },
+    {
+      "epoch": 0.051743857388880855,
+      "grad_norm": 3.6709984313290933,
+      "learning_rate": 4.978612153434527e-06,
+      "loss": 1.0564,
+      "step": 492
+    },
+    {
+      "epoch": 0.05184902783072817,
+      "grad_norm": 5.006510707239013,
+      "learning_rate": 4.9785030494368515e-06,
+      "loss": 1.0852,
+      "step": 493
+    },
+    {
+      "epoch": 0.05195419827257549,
+      "grad_norm": 3.653015612885224,
+      "learning_rate": 4.978393669066832e-06,
+      "loss": 1.0751,
+      "step": 494
+    },
+    {
+      "epoch": 0.052059368714422814,
+      "grad_norm": 3.339203412325409,
+      "learning_rate": 4.978284012336666e-06,
+      "loss": 1.0509,
+      "step": 495
+    },
+    {
+      "epoch": 0.05216453915627013,
+      "grad_norm": 4.106088996601506,
+      "learning_rate": 4.978174079258581e-06,
+      "loss": 1.0331,
+      "step": 496
+    },
+    {
+      "epoch": 0.05226970959811745,
+      "grad_norm": 4.537839424849167,
+      "learning_rate": 4.978063869844834e-06,
+      "loss": 1.0399,
+      "step": 497
+    },
+    {
+      "epoch": 0.052374880039964766,
+      "grad_norm": 3.671123861891514,
+      "learning_rate": 4.977953384107716e-06,
+      "loss": 1.0404,
+      "step": 498
+    },
+    {
+      "epoch": 0.052480050481812084,
+      "grad_norm": 3.835556456175973,
+      "learning_rate": 4.977842622059546e-06,
+      "loss": 1.042,
+      "step": 499
+    },
+    {
+      "epoch": 0.05258522092365941,
+      "grad_norm": 3.810143572024863,
+      "learning_rate": 4.977731583712675e-06,
+      "loss": 1.0619,
+      "step": 500
+    },
+    {
+      "epoch": 0.052690391365506725,
+      "grad_norm": 2.977023375505081,
+      "learning_rate": 4.977620269079485e-06,
+      "loss": 1.0557,
+      "step": 501
+    },
+    {
+      "epoch": 0.05279556180735404,
+      "grad_norm": 2.785680113793684,
+      "learning_rate": 4.977508678172388e-06,
+      "loss": 1.0228,
+      "step": 502
+    },
+    {
+      "epoch": 0.05290073224920136,
+      "grad_norm": 3.1735915547497573,
+      "learning_rate": 4.977396811003828e-06,
+      "loss": 1.0574,
+      "step": 503
+    },
+    {
+      "epoch": 0.053005902691048684,
+      "grad_norm": 2.4497703751127227,
+      "learning_rate": 4.977284667586278e-06,
+      "loss": 1.0517,
+      "step": 504
+    },
+    {
+      "epoch": 0.053111073132896,
+      "grad_norm": 5.411811630465418,
+      "learning_rate": 4.977172247932243e-06,
+      "loss": 1.087,
+      "step": 505
+    },
+    {
+      "epoch": 0.05321624357474332,
+      "grad_norm": 3.763750928236047,
+      "learning_rate": 4.97705955205426e-06,
+      "loss": 1.0788,
+      "step": 506
+    },
+    {
+      "epoch": 0.053321414016590636,
+      "grad_norm": 3.322087579873048,
+      "learning_rate": 4.976946579964893e-06,
+      "loss": 1.0785,
+      "step": 507
+    },
+    {
+      "epoch": 0.05342658445843795,
+      "grad_norm": 4.615164294462559,
+      "learning_rate": 4.976833331676742e-06,
+      "loss": 1.0071,
+      "step": 508
+    },
+    {
+      "epoch": 0.05353175490028528,
+      "grad_norm": 4.4250976293850295,
+      "learning_rate": 4.976719807202434e-06,
+      "loss": 1.0403,
+      "step": 509
+    },
+    {
+      "epoch": 0.053636925342132595,
+      "grad_norm": 3.8283999846040877,
+      "learning_rate": 4.976606006554626e-06,
+      "loss": 1.0061,
+      "step": 510
+    },
+    {
+      "epoch": 0.05374209578397991,
+      "grad_norm": 3.441735254441529,
+      "learning_rate": 4.976491929746011e-06,
+      "loss": 1.0699,
+      "step": 511
+    },
+    {
+      "epoch": 0.05384726622582723,
+      "grad_norm": 3.3474483003361835,
+      "learning_rate": 4.976377576789307e-06,
+      "loss": 1.0513,
+      "step": 512
+    },
+    {
+      "epoch": 0.05395243666767455,
+      "grad_norm": 3.2087836619502603,
+      "learning_rate": 4.976262947697265e-06,
+      "loss": 1.015,
+      "step": 513
+    },
+    {
+      "epoch": 0.05405760710952187,
+      "grad_norm": 4.79038201957431,
+      "learning_rate": 4.97614804248267e-06,
+      "loss": 1.0619,
+      "step": 514
+    },
+    {
+      "epoch": 0.05416277755136919,
+      "grad_norm": 5.0334628294143595,
+      "learning_rate": 4.976032861158332e-06,
+      "loss": 1.0501,
+      "step": 515
+    },
+    {
+      "epoch": 0.054267947993216506,
+      "grad_norm": 5.716661298960797,
+      "learning_rate": 4.9759174037370955e-06,
+      "loss": 1.051,
+      "step": 516
+    },
+    {
+      "epoch": 0.05437311843506382,
+      "grad_norm": 3.219972550593159,
+      "learning_rate": 4.975801670231835e-06,
+      "loss": 1.0549,
+      "step": 517
+    },
+    {
+      "epoch": 0.05447828887691115,
+      "grad_norm": 4.042629628393561,
+      "learning_rate": 4.975685660655456e-06,
+      "loss": 1.0104,
+      "step": 518
+    },
+    {
+      "epoch": 0.054583459318758465,
+      "grad_norm": 4.588832225167543,
+      "learning_rate": 4.975569375020894e-06,
+      "loss": 1.0681,
+      "step": 519
+    },
+    {
+      "epoch": 0.05468862976060578,
+      "grad_norm": 3.6574107244998344,
+      "learning_rate": 4.9754528133411144e-06,
+      "loss": 1.0306,
+      "step": 520
+    },
+    {
+      "epoch": 0.0547938002024531,
+      "grad_norm": 3.064757154713279,
+      "learning_rate": 4.975335975629117e-06,
+      "loss": 0.9752,
+      "step": 521
+    },
+    {
+      "epoch": 0.05489897064430042,
+      "grad_norm": 3.7277635260897997,
+      "learning_rate": 4.975218861897929e-06,
+      "loss": 1.0457,
+      "step": 522
+    },
+    {
+      "epoch": 0.05500414108614774,
+      "grad_norm": 3.2935157357520124,
+      "learning_rate": 4.9751014721606104e-06,
+      "loss": 1.0834,
+      "step": 523
+    },
+    {
+      "epoch": 0.05510931152799506,
+      "grad_norm": 2.341096648916625,
+      "learning_rate": 4.97498380643025e-06,
+      "loss": 1.0159,
+      "step": 524
+    },
+    {
+      "epoch": 0.055214481969842376,
+      "grad_norm": 2.9750956985730785,
+      "learning_rate": 4.974865864719969e-06,
+      "loss": 1.0646,
+      "step": 525
+    },
+    {
+      "epoch": 0.05531965241168969,
+      "grad_norm": 2.712385516803827,
+      "learning_rate": 4.974747647042918e-06,
+      "loss": 1.0122,
+      "step": 526
+    },
+    {
+      "epoch": 0.05542482285353701,
+      "grad_norm": 3.225400308884659,
+      "learning_rate": 4.9746291534122805e-06,
+      "loss": 1.0523,
+      "step": 527
+    },
+    {
+      "epoch": 0.055529993295384335,
+      "grad_norm": 4.290571209392225,
+      "learning_rate": 4.974510383841269e-06,
+      "loss": 1.0738,
+      "step": 528
+    },
+    {
+      "epoch": 0.05563516373723165,
+      "grad_norm": 4.014053600914674,
+      "learning_rate": 4.974391338343126e-06,
+      "loss": 1.0592,
+      "step": 529
+    },
+    {
+      "epoch": 0.05574033417907897,
+      "grad_norm": 3.3268020956167166,
+      "learning_rate": 4.974272016931127e-06,
+      "loss": 1.043,
+      "step": 530
+    },
+    {
+      "epoch": 0.05584550462092629,
+      "grad_norm": 4.333148338760938,
+      "learning_rate": 4.974152419618579e-06,
+      "loss": 1.0473,
+      "step": 531
+    },
+    {
+      "epoch": 0.055950675062773604,
+      "grad_norm": 2.8778208791693616,
+      "learning_rate": 4.974032546418816e-06,
+      "loss": 1.0629,
+      "step": 532
+    },
+    {
+      "epoch": 0.05605584550462093,
+      "grad_norm": 3.2582876551113142,
+      "learning_rate": 4.973912397345204e-06,
+      "loss": 1.0486,
+      "step": 533
+    },
+    {
+      "epoch": 0.056161015946468246,
+      "grad_norm": 3.287154664003592,
+      "learning_rate": 4.973791972411142e-06,
+      "loss": 1.0428,
+      "step": 534
+    },
+    {
+      "epoch": 0.05626618638831556,
+      "grad_norm": 3.331118043139556,
+      "learning_rate": 4.973671271630057e-06,
+      "loss": 1.0282,
+      "step": 535
+    },
+    {
+      "epoch": 0.05637135683016288,
+      "grad_norm": 3.3363705578676464,
+      "learning_rate": 4.973550295015411e-06,
+      "loss": 1.0733,
+      "step": 536
+    },
+    {
+      "epoch": 0.056476527272010205,
+      "grad_norm": 3.191076118323749,
+      "learning_rate": 4.973429042580691e-06,
+      "loss": 1.0458,
+      "step": 537
+    },
+    {
+      "epoch": 0.05658169771385752,
+      "grad_norm": 3.2347419761171716,
+      "learning_rate": 4.973307514339418e-06,
+      "loss": 1.053,
+      "step": 538
+    },
+    {
+      "epoch": 0.05668686815570484,
+      "grad_norm": 3.4907812740771575,
+      "learning_rate": 4.973185710305145e-06,
+      "loss": 1.0548,
+      "step": 539
+    },
+    {
+      "epoch": 0.05679203859755216,
+      "grad_norm": 4.184646901965426,
+      "learning_rate": 4.973063630491451e-06,
+      "loss": 1.0761,
+      "step": 540
+    },
+    {
+      "epoch": 0.056897209039399474,
+      "grad_norm": 3.0281792628868645,
+      "learning_rate": 4.972941274911953e-06,
+      "loss": 1.0389,
+      "step": 541
+    },
+    {
+      "epoch": 0.0570023794812468,
+      "grad_norm": 3.6955546648249378,
+      "learning_rate": 4.972818643580291e-06,
+      "loss": 1.0145,
+      "step": 542
+    },
+    {
+      "epoch": 0.057107549923094116,
+      "grad_norm": 3.1190860521523813,
+      "learning_rate": 4.972695736510141e-06,
+      "loss": 0.9828,
+      "step": 543
+    },
+    {
+      "epoch": 0.05721272036494143,
+      "grad_norm": 3.100972513801203,
+      "learning_rate": 4.972572553715208e-06,
+      "loss": 1.035,
+      "step": 544
+    },
+    {
+      "epoch": 0.05731789080678875,
+      "grad_norm": 3.0257460963065297,
+      "learning_rate": 4.972449095209227e-06,
+      "loss": 1.0376,
+      "step": 545
+    },
+    {
+      "epoch": 0.05742306124863607,
+      "grad_norm": 3.932692512358207,
+      "learning_rate": 4.972325361005966e-06,
+      "loss": 1.0455,
+      "step": 546
+    },
+    {
+      "epoch": 0.05752823169048339,
+      "grad_norm": 3.196511181526071,
+      "learning_rate": 4.9722013511192215e-06,
+      "loss": 0.9909,
+      "step": 547
+    },
+    {
+      "epoch": 0.05763340213233071,
+      "grad_norm": 3.467930305235877,
+      "learning_rate": 4.9720770655628216e-06,
+      "loss": 1.044,
+      "step": 548
+    },
+    {
+      "epoch": 0.057738572574178026,
+      "grad_norm": 4.5794753754186495,
+      "learning_rate": 4.971952504350625e-06,
+      "loss": 1.0603,
+      "step": 549
+    },
+    {
+      "epoch": 0.057843743016025344,
+      "grad_norm": 3.2755087160087073,
+      "learning_rate": 4.9718276674965215e-06,
+      "loss": 1.066,
+      "step": 550
+    },
+    {
+      "epoch": 0.05794891345787267,
+      "grad_norm": 3.9349343797276197,
+      "learning_rate": 4.971702555014431e-06,
+      "loss": 1.0681,
+      "step": 551
+    },
+    {
+      "epoch": 0.058054083899719985,
+      "grad_norm": 4.361018090944349,
+      "learning_rate": 4.971577166918305e-06,
+      "loss": 1.0697,
+      "step": 552
+    },
+    {
+      "epoch": 0.0581592543415673,
+      "grad_norm": 4.570084487039526,
+      "learning_rate": 4.971451503222125e-06,
+      "loss": 1.0249,
+      "step": 553
+    },
+    {
+      "epoch": 0.05826442478341462,
+      "grad_norm": 4.362303063650197,
+      "learning_rate": 4.971325563939903e-06,
+      "loss": 1.0465,
+      "step": 554
+    },
+    {
+      "epoch": 0.05836959522526194,
+      "grad_norm": 2.2989031239751285,
+      "learning_rate": 4.971199349085683e-06,
+      "loss": 1.0107,
+      "step": 555
+    },
+    {
+      "epoch": 0.05847476566710926,
+      "grad_norm": 3.672401717225256,
+      "learning_rate": 4.971072858673539e-06,
+      "loss": 1.0765,
+      "step": 556
+    },
+    {
+      "epoch": 0.05857993610895658,
+      "grad_norm": 3.633753822132858,
+      "learning_rate": 4.970946092717574e-06,
+      "loss": 1.0491,
+      "step": 557
+    },
+    {
+      "epoch": 0.058685106550803896,
+      "grad_norm": 4.381583646223104,
+      "learning_rate": 4.970819051231927e-06,
+      "loss": 1.0819,
+      "step": 558
+    },
+    {
+      "epoch": 0.058790276992651214,
+      "grad_norm": 3.3145358767742987,
+      "learning_rate": 4.970691734230759e-06,
+      "loss": 1.0243,
+      "step": 559
+    },
+    {
+      "epoch": 0.05889544743449853,
+      "grad_norm": 2.16306091430307,
+      "learning_rate": 4.970564141728271e-06,
+      "loss": 1.0251,
+      "step": 560
+    },
+    {
+      "epoch": 0.059000617876345855,
+      "grad_norm": 4.923343581263921,
+      "learning_rate": 4.9704362737386894e-06,
+      "loss": 1.0501,
+      "step": 561
+    },
+    {
+      "epoch": 0.05910578831819317,
+      "grad_norm": 3.808471042980745,
+      "learning_rate": 4.970308130276273e-06,
+      "loss": 1.0398,
+      "step": 562
+    },
+    {
+      "epoch": 0.05921095876004049,
+      "grad_norm": 3.13104997246359,
+      "learning_rate": 4.970179711355308e-06,
+      "loss": 1.0325,
+      "step": 563
+    },
+    {
+      "epoch": 0.05931612920188781,
+      "grad_norm": 3.0500679981758436,
+      "learning_rate": 4.970051016990118e-06,
+      "loss": 1.0579,
+      "step": 564
+    },
+    {
+      "epoch": 0.05942129964373513,
+      "grad_norm": 2.9497172428170204,
+      "learning_rate": 4.96992204719505e-06,
+      "loss": 1.049,
+      "step": 565
+    },
+    {
+      "epoch": 0.05952647008558245,
+      "grad_norm": 1.9764188767604873,
+      "learning_rate": 4.9697928019844885e-06,
+      "loss": 1.0358,
+      "step": 566
+    },
+    {
+      "epoch": 0.059631640527429766,
+      "grad_norm": 3.9746447469142185,
+      "learning_rate": 4.969663281372842e-06,
+      "loss": 1.0577,
+      "step": 567
+    },
+    {
+      "epoch": 0.059736810969277084,
+      "grad_norm": 3.788137117647662,
+      "learning_rate": 4.969533485374556e-06,
+      "loss": 0.9947,
+      "step": 568
+    },
+    {
+      "epoch": 0.0598419814111244,
+      "grad_norm": 4.405021349521944,
+      "learning_rate": 4.9694034140041015e-06,
+      "loss": 1.0581,
+      "step": 569
+    },
+    {
+      "epoch": 0.059947151852971725,
+      "grad_norm": 2.0847517500095334,
+      "learning_rate": 4.9692730672759835e-06,
+      "loss": 1.0142,
+      "step": 570
+    },
+    {
+      "epoch": 0.06005232229481904,
+      "grad_norm": 5.811807304246141,
+      "learning_rate": 4.969142445204736e-06,
+      "loss": 1.0579,
+      "step": 571
+    },
+    {
+      "epoch": 0.06015749273666636,
+      "grad_norm": 3.222186946748591,
+      "learning_rate": 4.969011547804925e-06,
+      "loss": 1.0306,
+      "step": 572
+    },
+    {
+      "epoch": 0.06026266317851368,
+      "grad_norm": 3.7831417030237837,
+      "learning_rate": 4.968880375091147e-06,
+      "loss": 1.0391,
+      "step": 573
+    },
+    {
+      "epoch": 0.060367833620360994,
+      "grad_norm": 3.9362867297970143,
+      "learning_rate": 4.968748927078028e-06,
+      "loss": 1.0684,
+      "step": 574
+    },
+    {
+      "epoch": 0.06047300406220832,
+      "grad_norm": 3.329344719091068,
+      "learning_rate": 4.968617203780226e-06,
+      "loss": 1.0679,
+      "step": 575
+    },
+    {
+      "epoch": 0.060578174504055636,
+      "grad_norm": 2.693926328348047,
+      "learning_rate": 4.968485205212429e-06,
+      "loss": 1.0686,
+      "step": 576
+    },
+    {
+      "epoch": 0.06068334494590295,
+      "grad_norm": 4.752717425247337,
+      "learning_rate": 4.968352931389355e-06,
+      "loss": 1.0477,
+      "step": 577
+    },
+    {
+      "epoch": 0.06078851538775027,
+      "grad_norm": 5.382651519152381,
+      "learning_rate": 4.968220382325755e-06,
+      "loss": 1.0522,
+      "step": 578
+    },
+    {
+      "epoch": 0.060893685829597595,
+      "grad_norm": 5.095229123548204,
+      "learning_rate": 4.968087558036408e-06,
+      "loss": 1.0342,
+      "step": 579
+    },
+    {
+      "epoch": 0.06099885627144491,
+      "grad_norm": 2.9622063334614506,
+      "learning_rate": 4.967954458536126e-06,
+      "loss": 1.0545,
+      "step": 580
+    },
+    {
+      "epoch": 0.06110402671329223,
+      "grad_norm": 4.010704877984255,
+      "learning_rate": 4.96782108383975e-06,
+      "loss": 1.0396,
+      "step": 581
+    },
+    {
+      "epoch": 0.06120919715513955,
+      "grad_norm": 3.679353200958754,
+      "learning_rate": 4.967687433962152e-06,
+      "loss": 1.0354,
+      "step": 582
+    },
+    {
+      "epoch": 0.061314367596986864,
+      "grad_norm": 4.1840739624189345,
+      "learning_rate": 4.967553508918236e-06,
+      "loss": 1.0032,
+      "step": 583
+    },
+    {
+      "epoch": 0.06141953803883419,
+      "grad_norm": 3.8257222018053394,
+      "learning_rate": 4.967419308722935e-06,
+      "loss": 1.0402,
+      "step": 584
+    },
+    {
+      "epoch": 0.061524708480681506,
+      "grad_norm": 4.453654621625787,
+      "learning_rate": 4.967284833391213e-06,
+      "loss": 1.0446,
+      "step": 585
+    },
+    {
+      "epoch": 0.06162987892252882,
+      "grad_norm": 3.37619320579649,
+      "learning_rate": 4.967150082938066e-06,
+      "loss": 1.0381,
+      "step": 586
+    },
+    {
+      "epoch": 0.06173504936437614,
+      "grad_norm": 4.183795371545729,
+      "learning_rate": 4.9670150573785195e-06,
+      "loss": 1.0677,
+      "step": 587
+    },
+    {
+      "epoch": 0.06184021980622346,
+      "grad_norm": 2.901597121091973,
+      "learning_rate": 4.96687975672763e-06,
+      "loss": 1.0499,
+      "step": 588
+    },
+    {
+      "epoch": 0.06194539024807078,
+      "grad_norm": 3.838808993150415,
+      "learning_rate": 4.966744181000483e-06,
+      "loss": 1.0541,
+      "step": 589
+    },
+    {
+      "epoch": 0.0620505606899181,
+      "grad_norm": 3.750578964588987,
+      "learning_rate": 4.966608330212198e-06,
+      "loss": 1.0377,
+      "step": 590
+    },
+    {
+      "epoch": 0.06215573113176542,
+      "grad_norm": 3.6029951214777403,
+      "learning_rate": 4.9664722043779226e-06,
+      "loss": 1.0386,
+      "step": 591
+    },
+    {
+      "epoch": 0.062260901573612734,
+      "grad_norm": 3.430327053415612,
+      "learning_rate": 4.966335803512837e-06,
+      "loss": 1.0532,
+      "step": 592
+    },
+    {
+      "epoch": 0.06236607201546006,
+      "grad_norm": 2.7139325949372775,
+      "learning_rate": 4.96619912763215e-06,
+      "loss": 1.0649,
+      "step": 593
+    },
+    {
+      "epoch": 0.062471242457307376,
+      "grad_norm": 3.1492912949831884,
+      "learning_rate": 4.966062176751101e-06,
+      "loss": 1.0493,
+      "step": 594
+    },
+    {
+      "epoch": 0.0625764128991547,
+      "grad_norm": 4.125956595172075,
+      "learning_rate": 4.965924950884964e-06,
+      "loss": 1.0206,
+      "step": 595
+    },
+    {
+      "epoch": 0.06268158334100202,
+      "grad_norm": 3.7929128777641425,
+      "learning_rate": 4.965787450049038e-06,
+      "loss": 1.0299,
+      "step": 596
+    },
+    {
+      "epoch": 0.06278675378284933,
+      "grad_norm": 4.0810239450371215,
+      "learning_rate": 4.965649674258657e-06,
+      "loss": 1.1232,
+      "step": 597
+    },
+    {
+      "epoch": 0.06289192422469665,
+      "grad_norm": 4.6131544836052,
+      "learning_rate": 4.965511623529183e-06,
+      "loss": 1.0422,
+      "step": 598
+    },
+    {
+      "epoch": 0.06299709466654396,
+      "grad_norm": 4.798402879282435,
+      "learning_rate": 4.965373297876009e-06,
+      "loss": 1.0787,
+      "step": 599
+    },
+    {
+      "epoch": 0.06310226510839129,
+      "grad_norm": 4.335100341025516,
+      "learning_rate": 4.965234697314563e-06,
+      "loss": 0.9823,
+      "step": 600
+    },
+    {
+      "epoch": 0.06320743555023861,
+      "grad_norm": 3.842720236255419,
+      "learning_rate": 4.965095821860296e-06,
+      "loss": 1.0744,
+      "step": 601
+    },
+    {
+      "epoch": 0.06331260599208592,
+      "grad_norm": 5.2410971364749335,
+      "learning_rate": 4.964956671528696e-06,
+      "loss": 1.0372,
+      "step": 602
+    },
+    {
+      "epoch": 0.06341777643393325,
+      "grad_norm": 3.974109563837657,
+      "learning_rate": 4.964817246335278e-06,
+      "loss": 1.0889,
+      "step": 603
+    },
+    {
+      "epoch": 0.06352294687578056,
+      "grad_norm": 3.8457386538934326,
+      "learning_rate": 4.96467754629559e-06,
+      "loss": 1.0567,
+      "step": 604
+    },
+    {
+      "epoch": 0.06362811731762788,
+      "grad_norm": 4.314777488524298,
+      "learning_rate": 4.964537571425209e-06,
+      "loss": 1.0197,
+      "step": 605
+    },
+    {
+      "epoch": 0.0637332877594752,
+      "grad_norm": 3.1075473114740966,
+      "learning_rate": 4.9643973217397446e-06,
+      "loss": 1.042,
+      "step": 606
+    },
+    {
+      "epoch": 0.06383845820132251,
+      "grad_norm": 2.8685624789514637,
+      "learning_rate": 4.964256797254834e-06,
+      "loss": 1.0444,
+      "step": 607
+    },
+    {
+      "epoch": 0.06394362864316984,
+      "grad_norm": 3.0667602949659627,
+      "learning_rate": 4.964115997986148e-06,
+      "loss": 1.0434,
+      "step": 608
+    },
+    {
+      "epoch": 0.06404879908501715,
+      "grad_norm": 2.3511531120455955,
+      "learning_rate": 4.963974923949386e-06,
+      "loss": 1.0321,
+      "step": 609
+    },
+    {
+      "epoch": 0.06415396952686447,
+      "grad_norm": 2.7978514748891055,
+      "learning_rate": 4.963833575160278e-06,
+      "loss": 1.0402,
+      "step": 610
+    },
+    {
+      "epoch": 0.0642591399687118,
+      "grad_norm": 4.049018255338047,
+      "learning_rate": 4.963691951634588e-06,
+      "loss": 1.0826,
+      "step": 611
+    },
+    {
+      "epoch": 0.06436431041055911,
+      "grad_norm": 3.3678790577984072,
+      "learning_rate": 4.9635500533881065e-06,
+      "loss": 1.0306,
+      "step": 612
+    },
+    {
+      "epoch": 0.06446948085240643,
+      "grad_norm": 3.0634850122257786,
+      "learning_rate": 4.963407880436657e-06,
+      "loss": 1.0558,
+      "step": 613
+    },
+    {
+      "epoch": 0.06457465129425374,
+      "grad_norm": 3.0330037801287957,
+      "learning_rate": 4.963265432796092e-06,
+      "loss": 1.0743,
+      "step": 614
+    },
+    {
+      "epoch": 0.06467982173610107,
+      "grad_norm": 3.071000708505292,
+      "learning_rate": 4.963122710482295e-06,
+      "loss": 1.0543,
+      "step": 615
+    },
+    {
+      "epoch": 0.06478499217794839,
+      "grad_norm": 2.8610188182071385,
+      "learning_rate": 4.962979713511183e-06,
+      "loss": 1.0579,
+      "step": 616
+    },
+    {
+      "epoch": 0.0648901626197957,
+      "grad_norm": 4.206292540959885,
+      "learning_rate": 4.962836441898699e-06,
+      "loss": 1.0701,
+      "step": 617
+    },
+    {
+      "epoch": 0.06499533306164303,
+      "grad_norm": 4.356324104061252,
+      "learning_rate": 4.96269289566082e-06,
+      "loss": 1.0955,
+      "step": 618
+    },
+    {
+      "epoch": 0.06510050350349035,
+      "grad_norm": 3.319002510122104,
+      "learning_rate": 4.9625490748135525e-06,
+      "loss": 1.0207,
+      "step": 619
+    },
+    {
+      "epoch": 0.06520567394533766,
+      "grad_norm": 3.7127033614212026,
+      "learning_rate": 4.962404979372933e-06,
+      "loss": 1.0698,
+      "step": 620
+    },
+    {
+      "epoch": 0.06531084438718499,
+      "grad_norm": 4.18629805171437,
+      "learning_rate": 4.962260609355029e-06,
+      "loss": 1.0392,
+      "step": 621
+    },
+    {
+      "epoch": 0.0654160148290323,
+      "grad_norm": 4.993374261310267,
+      "learning_rate": 4.962115964775941e-06,
+      "loss": 1.008,
+      "step": 622
+    },
+    {
+      "epoch": 0.06552118527087962,
+      "grad_norm": 3.017269799303926,
+      "learning_rate": 4.961971045651796e-06,
+      "loss": 1.0126,
+      "step": 623
+    },
+    {
+      "epoch": 0.06562635571272694,
+      "grad_norm": 2.700741825605453,
+      "learning_rate": 4.961825851998754e-06,
+      "loss": 1.0446,
+      "step": 624
+    },
+    {
+      "epoch": 0.06573152615457425,
+      "grad_norm": 3.3134417945987567,
+      "learning_rate": 4.961680383833005e-06,
+      "loss": 1.0507,
+      "step": 625
+    },
+    {
+      "epoch": 0.06583669659642158,
+      "grad_norm": 4.294665174497018,
+      "learning_rate": 4.961534641170771e-06,
+      "loss": 1.0448,
+      "step": 626
+    },
+    {
+      "epoch": 0.06594186703826889,
+      "grad_norm": 3.659408792404163,
+      "learning_rate": 4.961388624028303e-06,
+      "loss": 1.0181,
+      "step": 627
+    },
+    {
+      "epoch": 0.06604703748011621,
+      "grad_norm": 3.640846682505268,
+      "learning_rate": 4.9612423324218816e-06,
+      "loss": 1.1181,
+      "step": 628
+    },
+    {
+      "epoch": 0.06615220792196354,
+      "grad_norm": 3.086905697715193,
+      "learning_rate": 4.961095766367821e-06,
+      "loss": 1.0573,
+      "step": 629
+    },
+    {
+      "epoch": 0.06625737836381085,
+      "grad_norm": 3.9597870998951468,
+      "learning_rate": 4.960948925882464e-06,
+      "loss": 1.0285,
+      "step": 630
+    },
+    {
+      "epoch": 0.06636254880565817,
+      "grad_norm": 4.292598646500938,
+      "learning_rate": 4.960801810982184e-06,
+      "loss": 1.0601,
+      "step": 631
+    },
+    {
+      "epoch": 0.06646771924750548,
+      "grad_norm": 3.600297435919102,
+      "learning_rate": 4.9606544216833865e-06,
+      "loss": 1.0509,
+      "step": 632
+    },
+    {
+      "epoch": 0.06657288968935281,
+      "grad_norm": 3.7580399038893755,
+      "learning_rate": 4.960506758002506e-06,
+      "loss": 1.0315,
+      "step": 633
+    },
+    {
+      "epoch": 0.06667806013120013,
+      "grad_norm": 3.2034136066067047,
+      "learning_rate": 4.960358819956007e-06,
+      "loss": 1.0486,
+      "step": 634
+    },
+    {
+      "epoch": 0.06678323057304744,
+      "grad_norm": 2.650453131312526,
+      "learning_rate": 4.960210607560387e-06,
+      "loss": 1.004,
+      "step": 635
+    },
+    {
+      "epoch": 0.06688840101489477,
+      "grad_norm": 3.3291779041483793,
+      "learning_rate": 4.960062120832174e-06,
+      "loss": 1.0592,
+      "step": 636
+    },
+    {
+      "epoch": 0.06699357145674208,
+      "grad_norm": 4.162217220561789,
+      "learning_rate": 4.959913359787923e-06,
+      "loss": 1.0178,
+      "step": 637
+    },
+    {
+      "epoch": 0.0670987418985894,
+      "grad_norm": 2.909767419364276,
+      "learning_rate": 4.959764324444224e-06,
+      "loss": 1.0572,
+      "step": 638
+    },
+    {
+      "epoch": 0.06720391234043673,
+      "grad_norm": 3.4590810268323127,
+      "learning_rate": 4.959615014817694e-06,
+      "loss": 0.9815,
+      "step": 639
+    },
+    {
+      "epoch": 0.06730908278228404,
+      "grad_norm": 5.476510291071485,
+      "learning_rate": 4.959465430924983e-06,
+      "loss": 1.0637,
+      "step": 640
+    },
+    {
+      "epoch": 0.06741425322413136,
+      "grad_norm": 3.499714211214465,
+      "learning_rate": 4.9593155727827705e-06,
+      "loss": 1.0513,
+      "step": 641
+    },
+    {
+      "epoch": 0.06751942366597867,
+      "grad_norm": 2.055004505673057,
+      "learning_rate": 4.9591654404077675e-06,
+      "loss": 1.043,
+      "step": 642
+    },
+    {
+      "epoch": 0.067624594107826,
+      "grad_norm": 3.4790021910325915,
+      "learning_rate": 4.959015033816714e-06,
+      "loss": 1.0246,
+      "step": 643
+    },
+    {
+      "epoch": 0.06772976454967332,
+      "grad_norm": 3.278908306612351,
+      "learning_rate": 4.958864353026382e-06,
+      "loss": 1.0671,
+      "step": 644
+    },
+    {
+      "epoch": 0.06783493499152063,
+      "grad_norm": 3.2634424157131585,
+      "learning_rate": 4.958713398053574e-06,
+      "loss": 1.052,
+      "step": 645
+    },
+    {
+      "epoch": 0.06794010543336795,
+      "grad_norm": 3.123492650449054,
+      "learning_rate": 4.958562168915121e-06,
+      "loss": 1.0305,
+      "step": 646
+    },
+    {
+      "epoch": 0.06804527587521528,
+      "grad_norm": 3.237487642730812,
+      "learning_rate": 4.9584106656278884e-06,
+      "loss": 1.0463,
+      "step": 647
+    },
+    {
+      "epoch": 0.06815044631706259,
+      "grad_norm": 4.187978143084541,
+      "learning_rate": 4.958258888208769e-06,
+      "loss": 1.0597,
+      "step": 648
+    },
+    {
+      "epoch": 0.06825561675890991,
+      "grad_norm": 2.9512732989254475,
+      "learning_rate": 4.958106836674686e-06,
+      "loss": 1.0864,
+      "step": 649
+    },
+    {
+      "epoch": 0.06836078720075722,
+      "grad_norm": 2.922076078908869,
+      "learning_rate": 4.957954511042598e-06,
+      "loss": 1.0638,
+      "step": 650
+    },
+    {
+      "epoch": 0.06846595764260455,
+      "grad_norm": 4.07040889167094,
+      "learning_rate": 4.957801911329485e-06,
+      "loss": 1.0417,
+      "step": 651
+    },
+    {
+      "epoch": 0.06857112808445187,
+      "grad_norm": 2.8281332452599495,
+      "learning_rate": 4.957649037552368e-06,
+      "loss": 1.0614,
+      "step": 652
+    },
+    {
+      "epoch": 0.06867629852629918,
+      "grad_norm": 2.6425711970959544,
+      "learning_rate": 4.957495889728291e-06,
+      "loss": 1.0922,
+      "step": 653
+    },
+    {
+      "epoch": 0.0687814689681465,
+      "grad_norm": 2.4951631686592157,
+      "learning_rate": 4.957342467874331e-06,
+      "loss": 1.017,
+      "step": 654
+    },
+    {
+      "epoch": 0.06888663940999382,
+      "grad_norm": 2.738893400348362,
+      "learning_rate": 4.957188772007597e-06,
+      "loss": 1.0489,
+      "step": 655
+    },
+    {
+      "epoch": 0.06899180985184114,
+      "grad_norm": 4.8906186641004785,
+      "learning_rate": 4.957034802145227e-06,
+      "loss": 1.0226,
+      "step": 656
+    },
+    {
+      "epoch": 0.06909698029368846,
+      "grad_norm": 3.9715716237558287,
+      "learning_rate": 4.9568805583043905e-06,
+      "loss": 1.0836,
+      "step": 657
+    },
+    {
+      "epoch": 0.06920215073553578,
+      "grad_norm": 2.7599656326533384,
+      "learning_rate": 4.956726040502285e-06,
+      "loss": 1.0429,
+      "step": 658
+    },
+    {
+      "epoch": 0.0693073211773831,
+      "grad_norm": 3.4427680533746687,
+      "learning_rate": 4.956571248756142e-06,
+      "loss": 1.03,
+      "step": 659
+    },
+    {
+      "epoch": 0.06941249161923041,
+      "grad_norm": 3.5854381263268666,
+      "learning_rate": 4.9564161830832214e-06,
+      "loss": 1.0513,
+      "step": 660
+    },
+    {
+      "epoch": 0.06951766206107773,
+      "grad_norm": 3.7704419908620355,
+      "learning_rate": 4.956260843500814e-06,
+      "loss": 1.0662,
+      "step": 661
+    },
+    {
+      "epoch": 0.06962283250292506,
+      "grad_norm": 4.3580242444386545,
+      "learning_rate": 4.9561052300262415e-06,
+      "loss": 1.0661,
+      "step": 662
+    },
+    {
+      "epoch": 0.06972800294477237,
+      "grad_norm": 3.720266896736387,
+      "learning_rate": 4.955949342676856e-06,
+      "loss": 1.0175,
+      "step": 663
+    },
+    {
+      "epoch": 0.06983317338661969,
+      "grad_norm": 4.7410376644337475,
+      "learning_rate": 4.955793181470041e-06,
+      "loss": 1.0149,
+      "step": 664
+    },
+    {
+      "epoch": 0.069938343828467,
+      "grad_norm": 3.32459207886422,
+      "learning_rate": 4.955636746423208e-06,
+      "loss": 1.0271,
+      "step": 665
+    },
+    {
+      "epoch": 0.07004351427031433,
+      "grad_norm": 4.773668981576176,
+      "learning_rate": 4.9554800375538026e-06,
+      "loss": 1.1019,
+      "step": 666
+    },
+    {
+      "epoch": 0.07014868471216165,
+      "grad_norm": 4.563403967070941,
+      "learning_rate": 4.955323054879298e-06,
+      "loss": 1.0346,
+      "step": 667
+    },
+    {
+      "epoch": 0.07025385515400896,
+      "grad_norm": 3.2801934792913885,
+      "learning_rate": 4.9551657984171995e-06,
+      "loss": 1.0036,
+      "step": 668
+    },
+    {
+      "epoch": 0.07035902559585629,
+      "grad_norm": 2.815254256813279,
+      "learning_rate": 4.955008268185041e-06,
+      "loss": 1.0693,
+      "step": 669
+    },
+    {
+      "epoch": 0.0704641960377036,
+      "grad_norm": 3.9847665137382737,
+      "learning_rate": 4.954850464200391e-06,
+      "loss": 1.0579,
+      "step": 670
+    },
+    {
+      "epoch": 0.07056936647955092,
+      "grad_norm": 3.8295707952048708,
+      "learning_rate": 4.954692386480843e-06,
+      "loss": 1.0618,
+      "step": 671
+    },
+    {
+      "epoch": 0.07067453692139825,
+      "grad_norm": 4.523650449147604,
+      "learning_rate": 4.954534035044025e-06,
+      "loss": 1.0736,
+      "step": 672
+    },
+    {
+      "epoch": 0.07077970736324556,
+      "grad_norm": 3.6692146168192235,
+      "learning_rate": 4.954375409907595e-06,
+      "loss": 1.0575,
+      "step": 673
+    },
+    {
+      "epoch": 0.07088487780509288,
+      "grad_norm": 3.325725179037857,
+      "learning_rate": 4.954216511089242e-06,
+      "loss": 1.0396,
+      "step": 674
+    },
+    {
+      "epoch": 0.07099004824694019,
+      "grad_norm": 3.3396872270434512,
+      "learning_rate": 4.954057338606681e-06,
+      "loss": 1.0595,
+      "step": 675
+    },
+    {
+      "epoch": 0.07109521868878751,
+      "grad_norm": 2.472910469422091,
+      "learning_rate": 4.953897892477664e-06,
+      "loss": 1.0438,
+      "step": 676
+    },
+    {
+      "epoch": 0.07120038913063484,
+      "grad_norm": 2.9528197412204227,
+      "learning_rate": 4.953738172719969e-06,
+      "loss": 1.0478,
+      "step": 677
+    },
+    {
+      "epoch": 0.07130555957248215,
+      "grad_norm": 3.1922126178334853,
+      "learning_rate": 4.953578179351407e-06,
+      "loss": 1.0706,
+      "step": 678
+    },
+    {
+      "epoch": 0.07141073001432947,
+      "grad_norm": 2.9060870572179778,
+      "learning_rate": 4.953417912389817e-06,
+      "loss": 1.0375,
+      "step": 679
+    },
+    {
+      "epoch": 0.0715159004561768,
+      "grad_norm": 2.2896504072768393,
+      "learning_rate": 4.9532573718530716e-06,
+      "loss": 1.0023,
+      "step": 680
+    },
+    {
+      "epoch": 0.07162107089802411,
+      "grad_norm": 3.4112113846941288,
+      "learning_rate": 4.953096557759072e-06,
+      "loss": 1.0565,
+      "step": 681
+    },
+    {
+      "epoch": 0.07172624133987143,
+      "grad_norm": 3.7035390337251837,
+      "learning_rate": 4.9529354701257486e-06,
+      "loss": 1.0406,
+      "step": 682
+    },
+    {
+      "epoch": 0.07183141178171874,
+      "grad_norm": 2.8453052460942705,
+      "learning_rate": 4.9527741089710664e-06,
+      "loss": 1.0652,
+      "step": 683
+    },
+    {
+      "epoch": 0.07193658222356607,
+      "grad_norm": 3.7620645383784233,
+      "learning_rate": 4.952612474313017e-06,
+      "loss": 1.047,
+      "step": 684
+    },
+    {
+      "epoch": 0.07204175266541339,
+      "grad_norm": 4.070346821177919,
+      "learning_rate": 4.9524505661696235e-06,
+      "loss": 1.0757,
+      "step": 685
+    },
+    {
+      "epoch": 0.0721469231072607,
+      "grad_norm": 3.5783791981742135,
+      "learning_rate": 4.952288384558941e-06,
+      "loss": 1.0512,
+      "step": 686
+    },
+    {
+      "epoch": 0.07225209354910803,
+      "grad_norm": 3.0673840522581046,
+      "learning_rate": 4.952125929499054e-06,
+      "loss": 1.0278,
+      "step": 687
+    },
+    {
+      "epoch": 0.07235726399095534,
+      "grad_norm": 3.2229108300056653,
+      "learning_rate": 4.9519632010080765e-06,
+      "loss": 1.0657,
+      "step": 688
+    },
+    {
+      "epoch": 0.07246243443280266,
+      "grad_norm": 2.652129171991843,
+      "learning_rate": 4.951800199104155e-06,
+      "loss": 1.0299,
+      "step": 689
+    },
+    {
+      "epoch": 0.07256760487464999,
+      "grad_norm": 3.364270267477337,
+      "learning_rate": 4.951636923805464e-06,
+      "loss": 1.0527,
+      "step": 690
+    },
+    {
+      "epoch": 0.0726727753164973,
+      "grad_norm": 3.4812038018158136,
+      "learning_rate": 4.951473375130212e-06,
+      "loss": 1.0331,
+      "step": 691
+    },
+    {
+      "epoch": 0.07277794575834462,
+      "grad_norm": 3.7937124089162704,
+      "learning_rate": 4.951309553096635e-06,
+      "loss": 1.0407,
+      "step": 692
+    },
+    {
+      "epoch": 0.07288311620019193,
+      "grad_norm": 3.1503406868653787,
+      "learning_rate": 4.951145457723e-06,
+      "loss": 1.0531,
+      "step": 693
+    },
+    {
+      "epoch": 0.07298828664203925,
+      "grad_norm": 2.8144705447628526,
+      "learning_rate": 4.9509810890276065e-06,
+      "loss": 1.1033,
+      "step": 694
+    },
+    {
+      "epoch": 0.07309345708388658,
+      "grad_norm": 3.3436807422524053,
+      "learning_rate": 4.95081644702878e-06,
+      "loss": 1.0837,
+      "step": 695
+    },
+    {
+      "epoch": 0.07319862752573389,
+      "grad_norm": 4.010147599775435,
+      "learning_rate": 4.950651531744882e-06,
+      "loss": 1.0769,
+      "step": 696
+    },
+    {
+      "epoch": 0.07330379796758121,
+      "grad_norm": 3.1326979849262195,
+      "learning_rate": 4.9504863431943004e-06,
+      "loss": 1.0019,
+      "step": 697
+    },
+    {
+      "epoch": 0.07340896840942852,
+      "grad_norm": 3.1431506128416995,
+      "learning_rate": 4.9503208813954565e-06,
+      "loss": 1.0326,
+      "step": 698
+    },
+    {
+      "epoch": 0.07351413885127585,
+      "grad_norm": 3.563862494912588,
+      "learning_rate": 4.9501551463667985e-06,
+      "loss": 1.0365,
+      "step": 699
+    },
+    {
+      "epoch": 0.07361930929312317,
+      "grad_norm": 2.9908810962971812,
+      "learning_rate": 4.949989138126809e-06,
+      "loss": 1.0655,
+      "step": 700
+    },
+    {
+      "epoch": 0.07372447973497048,
+      "grad_norm": 2.169128476617307,
+      "learning_rate": 4.949822856693997e-06,
+      "loss": 1.0292,
+      "step": 701
+    },
+    {
+      "epoch": 0.07382965017681781,
+      "grad_norm": 4.359267398693771,
+      "learning_rate": 4.949656302086907e-06,
+      "loss": 1.0228,
+      "step": 702
+    },
+    {
+      "epoch": 0.07393482061866512,
+      "grad_norm": 3.787799254738041,
+      "learning_rate": 4.949489474324108e-06,
+      "loss": 1.0041,
+      "step": 703
+    },
+    {
+      "epoch": 0.07403999106051244,
+      "grad_norm": 2.667883375421812,
+      "learning_rate": 4.949322373424206e-06,
+      "loss": 1.0423,
+      "step": 704
+    },
+    {
+      "epoch": 0.07414516150235977,
+      "grad_norm": 2.9302612190851787,
+      "learning_rate": 4.949154999405832e-06,
+      "loss": 1.0259,
+      "step": 705
+    },
+    {
+      "epoch": 0.07425033194420708,
+      "grad_norm": 4.283595912438445,
+      "learning_rate": 4.94898735228765e-06,
+      "loss": 1.0675,
+      "step": 706
+    },
+    {
+      "epoch": 0.0743555023860544,
+      "grad_norm": 3.2175291514292996,
+      "learning_rate": 4.948819432088353e-06,
+      "loss": 1.0215,
+      "step": 707
+    },
+    {
+      "epoch": 0.07446067282790172,
+      "grad_norm": 3.7422947687070645,
+      "learning_rate": 4.948651238826667e-06,
+      "loss": 1.0539,
+      "step": 708
+    },
+    {
+      "epoch": 0.07456584326974904,
+      "grad_norm": 3.6570635294285383,
+      "learning_rate": 4.948482772521346e-06,
+      "loss": 1.0375,
+      "step": 709
+    },
+    {
+      "epoch": 0.07467101371159636,
+      "grad_norm": 4.267142662139774,
+      "learning_rate": 4.948314033191175e-06,
+      "loss": 1.0495,
+      "step": 710
+    },
+    {
+      "epoch": 0.07477618415344367,
+      "grad_norm": 3.6286208637111277,
+      "learning_rate": 4.948145020854971e-06,
+      "loss": 1.0083,
+      "step": 711
+    },
+    {
+      "epoch": 0.074881354595291,
+      "grad_norm": 3.942094346556032,
+      "learning_rate": 4.947975735531578e-06,
+      "loss": 1.0355,
+      "step": 712
+    },
+    {
+      "epoch": 0.07498652503713832,
+      "grad_norm": 2.0042428048097256,
+      "learning_rate": 4.947806177239875e-06,
+      "loss": 1.066,
+      "step": 713
+    },
+    {
+      "epoch": 0.07509169547898563,
+      "grad_norm": 3.4172310190837143,
+      "learning_rate": 4.947636345998768e-06,
+      "loss": 1.0153,
+      "step": 714
+    },
+    {
+      "epoch": 0.07519686592083295,
+      "grad_norm": 3.2523213929196273,
+      "learning_rate": 4.947466241827194e-06,
+      "loss": 1.0243,
+      "step": 715
+    },
+    {
+      "epoch": 0.07530203636268026,
+      "grad_norm": 4.081933448185989,
+      "learning_rate": 4.947295864744121e-06,
+      "loss": 0.9925,
+      "step": 716
+    },
+    {
+      "epoch": 0.07540720680452759,
+      "grad_norm": 3.0031168886878583,
+      "learning_rate": 4.947125214768549e-06,
+      "loss": 1.0452,
+      "step": 717
+    },
+    {
+      "epoch": 0.07551237724637491,
+      "grad_norm": 2.8581225129342287,
+      "learning_rate": 4.946954291919505e-06,
+      "loss": 1.0324,
+      "step": 718
+    },
+    {
+      "epoch": 0.07561754768822222,
+      "grad_norm": 2.663605746652252,
+      "learning_rate": 4.946783096216049e-06,
+      "loss": 1.0029,
+      "step": 719
+    },
+    {
+      "epoch": 0.07572271813006955,
+      "grad_norm": 3.2647089266169984,
+      "learning_rate": 4.94661162767727e-06,
+      "loss": 1.0534,
+      "step": 720
+    },
+    {
+      "epoch": 0.07582788857191686,
+      "grad_norm": 3.6156070223917216,
+      "learning_rate": 4.94643988632229e-06,
+      "loss": 0.9935,
+      "step": 721
+    },
+    {
+      "epoch": 0.07593305901376418,
+      "grad_norm": 3.758495500937088,
+      "learning_rate": 4.946267872170256e-06,
+      "loss": 0.9976,
+      "step": 722
+    },
+    {
+      "epoch": 0.0760382294556115,
+      "grad_norm": 2.8459717369309736,
+      "learning_rate": 4.946095585240353e-06,
+      "loss": 1.0202,
+      "step": 723
+    },
+    {
+      "epoch": 0.07614339989745882,
+      "grad_norm": 4.228677921247087,
+      "learning_rate": 4.945923025551789e-06,
+      "loss": 1.0273,
+      "step": 724
+    },
+    {
+      "epoch": 0.07624857033930614,
+      "grad_norm": 2.354998022999138,
+      "learning_rate": 4.945750193123808e-06,
+      "loss": 1.0082,
+      "step": 725
+    },
+    {
+      "epoch": 0.07635374078115345,
+      "grad_norm": 3.1974238173146965,
+      "learning_rate": 4.94557708797568e-06,
+      "loss": 1.0845,
+      "step": 726
+    },
+    {
+      "epoch": 0.07645891122300078,
+      "grad_norm": 3.2991867123662186,
+      "learning_rate": 4.94540371012671e-06,
+      "loss": 1.0363,
+      "step": 727
+    },
+    {
+      "epoch": 0.0765640816648481,
+      "grad_norm": 4.790231421271602,
+      "learning_rate": 4.945230059596229e-06,
+      "loss": 1.0522,
+      "step": 728
+    },
+    {
+      "epoch": 0.07666925210669541,
+      "grad_norm": 2.985913413424653,
+      "learning_rate": 4.945056136403601e-06,
+      "loss": 1.0011,
+      "step": 729
+    },
+    {
+      "epoch": 0.07677442254854273,
+      "grad_norm": 3.4749105660247452,
+      "learning_rate": 4.944881940568219e-06,
+      "loss": 1.0506,
+      "step": 730
+    },
+    {
+      "epoch": 0.07687959299039004,
+      "grad_norm": 2.331821909393682,
+      "learning_rate": 4.94470747210951e-06,
+      "loss": 0.9939,
+      "step": 731
+    },
+    {
+      "epoch": 0.07698476343223737,
+      "grad_norm": 3.2284649198570965,
+      "learning_rate": 4.944532731046926e-06,
+      "loss": 1.0831,
+      "step": 732
+    },
+    {
+      "epoch": 0.07708993387408469,
+      "grad_norm": 3.902120982308078,
+      "learning_rate": 4.944357717399952e-06,
+      "loss": 1.1132,
+      "step": 733
+    },
+    {
+      "epoch": 0.077195104315932,
+      "grad_norm": 3.4172360346109785,
+      "learning_rate": 4.944182431188106e-06,
+      "loss": 1.073,
+      "step": 734
+    },
+    {
+      "epoch": 0.07730027475777933,
+      "grad_norm": 3.6093384619673254,
+      "learning_rate": 4.94400687243093e-06,
+      "loss": 1.0422,
+      "step": 735
+    },
+    {
+      "epoch": 0.07740544519962665,
+      "grad_norm": 2.938298485477721,
+      "learning_rate": 4.943831041148003e-06,
+      "loss": 0.991,
+      "step": 736
+    },
+    {
+      "epoch": 0.07751061564147396,
+      "grad_norm": 3.5768669215903386,
+      "learning_rate": 4.94365493735893e-06,
+      "loss": 1.0159,
+      "step": 737
+    },
+    {
+      "epoch": 0.07761578608332129,
+      "grad_norm": 3.236703601168475,
+      "learning_rate": 4.943478561083349e-06,
+      "loss": 1.0307,
+      "step": 738
+    },
+    {
+      "epoch": 0.0777209565251686,
+      "grad_norm": 4.136119049876218,
+      "learning_rate": 4.9433019123409264e-06,
+      "loss": 1.0184,
+      "step": 739
+    },
+    {
+      "epoch": 0.07782612696701592,
+      "grad_norm": 3.485691899686312,
+      "learning_rate": 4.94312499115136e-06,
+      "loss": 1.0624,
+      "step": 740
+    },
+    {
+      "epoch": 0.07793129740886325,
+      "grad_norm": 3.294064033644912,
+      "learning_rate": 4.942947797534379e-06,
+      "loss": 1.0091,
+      "step": 741
+    },
+    {
+      "epoch": 0.07803646785071056,
+      "grad_norm": 3.388312825573561,
+      "learning_rate": 4.942770331509741e-06,
+      "loss": 1.048,
+      "step": 742
+    },
+    {
+      "epoch": 0.07814163829255788,
+      "grad_norm": 3.8425212422184707,
+      "learning_rate": 4.942592593097235e-06,
+      "loss": 1.0264,
+      "step": 743
+    },
+    {
+      "epoch": 0.07824680873440519,
+      "grad_norm": 4.216912890195311,
+      "learning_rate": 4.942414582316679e-06,
+      "loss": 1.0031,
+      "step": 744
+    },
+    {
+      "epoch": 0.07835197917625251,
+      "grad_norm": 2.943110966674974,
+      "learning_rate": 4.942236299187926e-06,
+      "loss": 1.0654,
+      "step": 745
+    },
+    {
+      "epoch": 0.07845714961809984,
+      "grad_norm": 2.7561756500798076,
+      "learning_rate": 4.942057743730852e-06,
+      "loss": 1.0762,
+      "step": 746
+    },
+    {
+      "epoch": 0.07856232005994715,
+      "grad_norm": 2.5060436964593458,
+      "learning_rate": 4.941878915965369e-06,
+      "loss": 1.0292,
+      "step": 747
+    },
+    {
+      "epoch": 0.07866749050179447,
+      "grad_norm": 4.677558119647075,
+      "learning_rate": 4.941699815911418e-06,
+      "loss": 1.0099,
+      "step": 748
+    },
+    {
+      "epoch": 0.07877266094364178,
+      "grad_norm": 2.6303885402682554,
+      "learning_rate": 4.94152044358897e-06,
+      "loss": 1.0585,
+      "step": 749
+    },
+    {
+      "epoch": 0.07887783138548911,
+      "grad_norm": 2.9906552945797777,
+      "learning_rate": 4.941340799018026e-06,
+      "loss": 1.0716,
+      "step": 750
+    },
+    {
+      "epoch": 0.07898300182733643,
+      "grad_norm": 3.1557961688407,
+      "learning_rate": 4.941160882218617e-06,
+      "loss": 1.0114,
+      "step": 751
+    },
+    {
+      "epoch": 0.07908817226918374,
+      "grad_norm": 2.5082121223741614,
+      "learning_rate": 4.940980693210807e-06,
+      "loss": 1.0839,
+      "step": 752
+    },
+    {
+      "epoch": 0.07919334271103107,
+      "grad_norm": 4.115073295467908,
+      "learning_rate": 4.940800232014688e-06,
+      "loss": 1.0528,
+      "step": 753
+    },
+    {
+      "epoch": 0.07929851315287838,
+      "grad_norm": 4.221836693988338,
+      "learning_rate": 4.940619498650381e-06,
+      "loss": 1.0458,
+      "step": 754
+    },
+    {
+      "epoch": 0.0794036835947257,
+      "grad_norm": 3.43936918626118,
+      "learning_rate": 4.940438493138041e-06,
+      "loss": 1.1102,
+      "step": 755
+    },
+    {
+      "epoch": 0.07950885403657303,
+      "grad_norm": 2.534335147364592,
+      "learning_rate": 4.9402572154978515e-06,
+      "loss": 1.03,
+      "step": 756
+    },
+    {
+      "epoch": 0.07961402447842034,
+      "grad_norm": 3.6985951390943397,
+      "learning_rate": 4.940075665750026e-06,
+      "loss": 1.0355,
+      "step": 757
+    },
+    {
+      "epoch": 0.07971919492026766,
+      "grad_norm": 3.654741106636443,
+      "learning_rate": 4.939893843914808e-06,
+      "loss": 1.0647,
+      "step": 758
+    },
+    {
+      "epoch": 0.07982436536211497,
+      "grad_norm": 3.686292494310798,
+      "learning_rate": 4.9397117500124725e-06,
+      "loss": 1.0508,
+      "step": 759
+    },
+    {
+      "epoch": 0.0799295358039623,
+      "grad_norm": 4.219962087918095,
+      "learning_rate": 4.939529384063325e-06,
+      "loss": 1.0415,
+      "step": 760
+    },
+    {
+      "epoch": 0.08003470624580962,
+      "grad_norm": 3.5870732019483387,
+      "learning_rate": 4.9393467460876995e-06,
+      "loss": 1.0291,
+      "step": 761
+    },
+    {
+      "epoch": 0.08013987668765693,
+      "grad_norm": 3.5610335680344902,
+      "learning_rate": 4.939163836105964e-06,
+      "loss": 1.0283,
+      "step": 762
+    },
+    {
+      "epoch": 0.08024504712950425,
+      "grad_norm": 2.8440793536671802,
+      "learning_rate": 4.938980654138511e-06,
+      "loss": 1.0676,
+      "step": 763
+    },
+    {
+      "epoch": 0.08035021757135156,
+      "grad_norm": 3.5144508685059512,
+      "learning_rate": 4.93879720020577e-06,
+      "loss": 0.9927,
+      "step": 764
+    },
+    {
+      "epoch": 0.08045538801319889,
+      "grad_norm": 3.6070080778872353,
+      "learning_rate": 4.938613474328195e-06,
+      "loss": 1.0694,
+      "step": 765
+    },
+    {
+      "epoch": 0.08056055845504621,
+      "grad_norm": 4.242572556515164,
+      "learning_rate": 4.938429476526273e-06,
+      "loss": 1.027,
+      "step": 766
+    },
+    {
+      "epoch": 0.08066572889689352,
+      "grad_norm": 2.4803307039348605,
+      "learning_rate": 4.938245206820522e-06,
+      "loss": 1.0119,
+      "step": 767
+    },
+    {
+      "epoch": 0.08077089933874085,
+      "grad_norm": 2.9387959106412804,
+      "learning_rate": 4.938060665231491e-06,
+      "loss": 1.0404,
+      "step": 768
+    },
+    {
+      "epoch": 0.08087606978058817,
+      "grad_norm": 4.172139806499708,
+      "learning_rate": 4.937875851779755e-06,
+      "loss": 1.0385,
+      "step": 769
+    },
+    {
+      "epoch": 0.08098124022243548,
+      "grad_norm": 3.0840295037017786,
+      "learning_rate": 4.937690766485924e-06,
+      "loss": 1.0129,
+      "step": 770
+    },
+    {
+      "epoch": 0.08108641066428281,
+      "grad_norm": 3.070735572354231,
+      "learning_rate": 4.9375054093706356e-06,
+      "loss": 1.0741,
+      "step": 771
+    },
+    {
+      "epoch": 0.08119158110613012,
+      "grad_norm": 3.1878677650163247,
+      "learning_rate": 4.937319780454559e-06,
+      "loss": 1.0794,
+      "step": 772
+    },
+    {
+      "epoch": 0.08129675154797744,
+      "grad_norm": 2.863706515463834,
+      "learning_rate": 4.937133879758394e-06,
+      "loss": 1.0336,
+      "step": 773
+    },
+    {
+      "epoch": 0.08140192198982477,
+      "grad_norm": 3.0701280575553334,
+      "learning_rate": 4.936947707302868e-06,
+      "loss": 1.0586,
+      "step": 774
+    },
+    {
+      "epoch": 0.08150709243167208,
+      "grad_norm": 3.7990360279352218,
+      "learning_rate": 4.936761263108742e-06,
+      "loss": 1.013,
+      "step": 775
+    },
+    {
+      "epoch": 0.0816122628735194,
+      "grad_norm": 3.2463147605003493,
+      "learning_rate": 4.936574547196806e-06,
+      "loss": 1.0674,
+      "step": 776
+    },
+    {
+      "epoch": 0.08171743331536671,
+      "grad_norm": 3.346659843217337,
+      "learning_rate": 4.93638755958788e-06,
+      "loss": 1.0484,
+      "step": 777
+    },
+    {
+      "epoch": 0.08182260375721404,
+      "grad_norm": 3.421009993046732,
+      "learning_rate": 4.9362003003028135e-06,
+      "loss": 0.9867,
+      "step": 778
+    },
+    {
+      "epoch": 0.08192777419906136,
+      "grad_norm": 3.5732995197430086,
+      "learning_rate": 4.93601276936249e-06,
+      "loss": 1.0472,
+      "step": 779
+    },
+    {
+      "epoch": 0.08203294464090867,
+      "grad_norm": 2.5915581383584554,
+      "learning_rate": 4.935824966787818e-06,
+      "loss": 1.0096,
+      "step": 780
+    },
+    {
+      "epoch": 0.082138115082756,
+      "grad_norm": 2.6084843786308376,
+      "learning_rate": 4.935636892599741e-06,
+      "loss": 1.0011,
+      "step": 781
+    },
+    {
+      "epoch": 0.0822432855246033,
+      "grad_norm": 2.7979653896224375,
+      "learning_rate": 4.935448546819229e-06,
+      "loss": 1.0501,
+      "step": 782
+    },
+    {
+      "epoch": 0.08234845596645063,
+      "grad_norm": 3.1927150256696253,
+      "learning_rate": 4.935259929467285e-06,
+      "loss": 1.0239,
+      "step": 783
+    },
+    {
+      "epoch": 0.08245362640829795,
+      "grad_norm": 3.5723314450964945,
+      "learning_rate": 4.935071040564942e-06,
+      "loss": 1.06,
+      "step": 784
+    },
+    {
+      "epoch": 0.08255879685014526,
+      "grad_norm": 2.951913249700751,
+      "learning_rate": 4.93488188013326e-06,
+      "loss": 1.091,
+      "step": 785
+    },
+    {
+      "epoch": 0.08266396729199259,
+      "grad_norm": 2.897178403548757,
+      "learning_rate": 4.9346924481933345e-06,
+      "loss": 0.9997,
+      "step": 786
+    },
+    {
+      "epoch": 0.0827691377338399,
+      "grad_norm": 2.976619910088866,
+      "learning_rate": 4.9345027447662876e-06,
+      "loss": 1.0628,
+      "step": 787
+    },
+    {
+      "epoch": 0.08287430817568722,
+      "grad_norm": 5.562882028586305,
+      "learning_rate": 4.934312769873273e-06,
+      "loss": 1.0514,
+      "step": 788
+    },
+    {
+      "epoch": 0.08297947861753455,
+      "grad_norm": 3.3122007501483086,
+      "learning_rate": 4.934122523535474e-06,
+      "loss": 1.017,
+      "step": 789
+    },
+    {
+      "epoch": 0.08308464905938186,
+      "grad_norm": 4.484356852207891,
+      "learning_rate": 4.9339320057741045e-06,
+      "loss": 1.0628,
+      "step": 790
+    },
+    {
+      "epoch": 0.08318981950122918,
+      "grad_norm": 2.662851425317841,
+      "learning_rate": 4.933741216610409e-06,
+      "loss": 1.068,
+      "step": 791
+    },
+    {
+      "epoch": 0.08329498994307649,
+      "grad_norm": 3.0734598927453765,
+      "learning_rate": 4.933550156065662e-06,
+      "loss": 1.043,
+      "step": 792
+    },
+    {
+      "epoch": 0.08340016038492382,
+      "grad_norm": 2.692919866577266,
+      "learning_rate": 4.933358824161167e-06,
+      "loss": 1.0257,
+      "step": 793
+    },
+    {
+      "epoch": 0.08350533082677114,
+      "grad_norm": 3.656399597645738,
+      "learning_rate": 4.933167220918262e-06,
+      "loss": 1.0345,
+      "step": 794
+    },
+    {
+      "epoch": 0.08361050126861845,
+      "grad_norm": 3.587236612420999,
+      "learning_rate": 4.9329753463583095e-06,
+      "loss": 1.0455,
+      "step": 795
+    },
+    {
+      "epoch": 0.08371567171046577,
+      "grad_norm": 3.656718452993634,
+      "learning_rate": 4.932783200502705e-06,
+      "loss": 1.0298,
+      "step": 796
+    },
+    {
+      "epoch": 0.0838208421523131,
+      "grad_norm": 3.00036115403809,
+      "learning_rate": 4.932590783372877e-06,
+      "loss": 1.022,
+      "step": 797
+    },
+    {
+      "epoch": 0.08392601259416041,
+      "grad_norm": 2.5105535700401242,
+      "learning_rate": 4.9323980949902786e-06,
+      "loss": 1.0868,
+      "step": 798
+    },
+    {
+      "epoch": 0.08403118303600773,
+      "grad_norm": 3.7118503835471546,
+      "learning_rate": 4.9322051353763965e-06,
+      "loss": 1.03,
+      "step": 799
+    },
+    {
+      "epoch": 0.08413635347785504,
+      "grad_norm": 4.549244874514477,
+      "learning_rate": 4.932011904552749e-06,
+      "loss": 1.0312,
+      "step": 800
+    },
+    {
+      "epoch": 0.08424152391970237,
+      "grad_norm": 2.695319366140478,
+      "learning_rate": 4.931818402540881e-06,
+      "loss": 1.0456,
+      "step": 801
+    },
+    {
+      "epoch": 0.08434669436154969,
+      "grad_norm": 5.172251410102943,
+      "learning_rate": 4.93162462936237e-06,
+      "loss": 1.06,
+      "step": 802
+    },
+    {
+      "epoch": 0.084451864803397,
+      "grad_norm": 3.731288930108174,
+      "learning_rate": 4.931430585038823e-06,
+      "loss": 1.0429,
+      "step": 803
+    },
+    {
+      "epoch": 0.08455703524524433,
+      "grad_norm": 2.688017597159252,
+      "learning_rate": 4.931236269591878e-06,
+      "loss": 1.0244,
+      "step": 804
+    },
+    {
+      "epoch": 0.08466220568709164,
+      "grad_norm": 2.8764757418451796,
+      "learning_rate": 4.9310416830432025e-06,
+      "loss": 1.0227,
+      "step": 805
+    },
+    {
+      "epoch": 0.08476737612893896,
+      "grad_norm": 4.0235649503508455,
+      "learning_rate": 4.930846825414495e-06,
+      "loss": 1.0423,
+      "step": 806
+    },
+    {
+      "epoch": 0.08487254657078629,
+      "grad_norm": 3.536335146675604,
+      "learning_rate": 4.930651696727482e-06,
+      "loss": 1.0062,
+      "step": 807
+    },
+    {
+      "epoch": 0.0849777170126336,
+      "grad_norm": 2.957564708059679,
+      "learning_rate": 4.930456297003923e-06,
+      "loss": 1.0522,
+      "step": 808
+    },
+    {
+      "epoch": 0.08508288745448092,
+      "grad_norm": 3.219817117203254,
+      "learning_rate": 4.930260626265607e-06,
+      "loss": 1.0915,
+      "step": 809
+    },
+    {
+      "epoch": 0.08518805789632823,
+      "grad_norm": 2.2575110274834636,
+      "learning_rate": 4.930064684534352e-06,
+      "loss": 1.0395,
+      "step": 810
+    },
+    {
+      "epoch": 0.08529322833817556,
+      "grad_norm": 3.301667345209895,
+      "learning_rate": 4.929868471832007e-06,
+      "loss": 1.0394,
+      "step": 811
+    },
+    {
+      "epoch": 0.08539839878002288,
+      "grad_norm": 4.501375773442117,
+      "learning_rate": 4.929671988180452e-06,
+      "loss": 1.0537,
+      "step": 812
+    },
+    {
+      "epoch": 0.08550356922187019,
+      "grad_norm": 2.9447996895825437,
+      "learning_rate": 4.929475233601595e-06,
+      "loss": 1.0163,
+      "step": 813
+    },
+    {
+      "epoch": 0.08560873966371751,
+      "grad_norm": 3.890755121107722,
+      "learning_rate": 4.929278208117378e-06,
+      "loss": 1.0418,
+      "step": 814
+    },
+    {
+      "epoch": 0.08571391010556483,
+      "grad_norm": 2.55891195688046,
+      "learning_rate": 4.929080911749769e-06,
+      "loss": 1.0191,
+      "step": 815
+    },
+    {
+      "epoch": 0.08581908054741215,
+      "grad_norm": 4.672850050572631,
+      "learning_rate": 4.928883344520768e-06,
+      "loss": 1.0363,
+      "step": 816
+    },
+    {
+      "epoch": 0.08592425098925947,
+      "grad_norm": 3.1282491114827575,
+      "learning_rate": 4.928685506452407e-06,
+      "loss": 1.0421,
+      "step": 817
+    },
+    {
+      "epoch": 0.08602942143110678,
+      "grad_norm": 2.0646227484923143,
+      "learning_rate": 4.928487397566743e-06,
+      "loss": 1.0407,
+      "step": 818
+    },
+    {
+      "epoch": 0.08613459187295411,
+      "grad_norm": 3.2962481364973564,
+      "learning_rate": 4.928289017885871e-06,
+      "loss": 1.0724,
+      "step": 819
+    },
+    {
+      "epoch": 0.08623976231480142,
+      "grad_norm": 3.4645243615129897,
+      "learning_rate": 4.92809036743191e-06,
+      "loss": 1.0396,
+      "step": 820
+    },
+    {
+      "epoch": 0.08634493275664874,
+      "grad_norm": 2.3605676802931286,
+      "learning_rate": 4.92789144622701e-06,
+      "loss": 0.9944,
+      "step": 821
+    },
+    {
+      "epoch": 0.08645010319849607,
+      "grad_norm": 4.514015314071715,
+      "learning_rate": 4.927692254293354e-06,
+      "loss": 1.0792,
+      "step": 822
+    },
+    {
+      "epoch": 0.08655527364034338,
+      "grad_norm": 3.364110041018555,
+      "learning_rate": 4.927492791653153e-06,
+      "loss": 1.0558,
+      "step": 823
+    },
+    {
+      "epoch": 0.0866604440821907,
+      "grad_norm": 2.1273722043275995,
+      "learning_rate": 4.927293058328647e-06,
+      "loss": 1.0158,
+      "step": 824
+    },
+    {
+      "epoch": 0.08676561452403801,
+      "grad_norm": 3.5256816639423327,
+      "learning_rate": 4.92709305434211e-06,
+      "loss": 1.0353,
+      "step": 825
+    },
+    {
+      "epoch": 0.08687078496588534,
+      "grad_norm": 3.6339397393250095,
+      "learning_rate": 4.926892779715843e-06,
+      "loss": 1.0626,
+      "step": 826
+    },
+    {
+      "epoch": 0.08697595540773266,
+      "grad_norm": 2.9798523498933567,
+      "learning_rate": 4.926692234472178e-06,
+      "loss": 1.0116,
+      "step": 827
+    },
+    {
+      "epoch": 0.08708112584957997,
+      "grad_norm": 3.025215020843175,
+      "learning_rate": 4.9264914186334775e-06,
+      "loss": 1.0143,
+      "step": 828
+    },
+    {
+      "epoch": 0.0871862962914273,
+      "grad_norm": 2.931478722868882,
+      "learning_rate": 4.926290332222134e-06,
+      "loss": 1.0852,
+      "step": 829
+    },
+    {
+      "epoch": 0.08729146673327462,
+      "grad_norm": 3.179330588406845,
+      "learning_rate": 4.9260889752605715e-06,
+      "loss": 1.0382,
+      "step": 830
+    },
+    {
+      "epoch": 0.08739663717512193,
+      "grad_norm": 2.7220659147274264,
+      "learning_rate": 4.925887347771241e-06,
+      "loss": 1.0312,
+      "step": 831
+    },
+    {
+      "epoch": 0.08750180761696925,
+      "grad_norm": 4.610622268460886,
+      "learning_rate": 4.925685449776627e-06,
+      "loss": 1.046,
+      "step": 832
+    },
+    {
+      "epoch": 0.08760697805881656,
+      "grad_norm": 2.720186510106471,
+      "learning_rate": 4.925483281299242e-06,
+      "loss": 1.0246,
+      "step": 833
+    },
+    {
+      "epoch": 0.08771214850066389,
+      "grad_norm": 2.4258313763166854,
+      "learning_rate": 4.925280842361628e-06,
+      "loss": 1.0416,
+      "step": 834
+    },
+    {
+      "epoch": 0.08781731894251121,
+      "grad_norm": 2.5395043340810464,
+      "learning_rate": 4.925078132986361e-06,
+      "loss": 0.9997,
+      "step": 835
+    },
+    {
+      "epoch": 0.08792248938435852,
+      "grad_norm": 3.0748465405966168,
+      "learning_rate": 4.924875153196042e-06,
+      "loss": 1.0486,
+      "step": 836
+    },
+    {
+      "epoch": 0.08802765982620585,
+      "grad_norm": 3.676854554009859,
+      "learning_rate": 4.924671903013308e-06,
+      "loss": 1.0313,
+      "step": 837
+    },
+    {
+      "epoch": 0.08813283026805316,
+      "grad_norm": 5.11342505319437,
+      "learning_rate": 4.9244683824608205e-06,
+      "loss": 1.0866,
+      "step": 838
+    },
+    {
+      "epoch": 0.08823800070990048,
+      "grad_norm": 3.364235859687841,
+      "learning_rate": 4.924264591561275e-06,
+      "loss": 1.0374,
+      "step": 839
+    },
+    {
+      "epoch": 0.0883431711517478,
+      "grad_norm": 2.543264664118595,
+      "learning_rate": 4.924060530337394e-06,
+      "loss": 1.036,
+      "step": 840
+    },
+    {
+      "epoch": 0.08844834159359512,
+      "grad_norm": 2.2631599527781208,
+      "learning_rate": 4.9238561988119346e-06,
+      "loss": 1.0327,
+      "step": 841
+    },
+    {
+      "epoch": 0.08855351203544244,
+      "grad_norm": 3.3000605867237742,
+      "learning_rate": 4.923651597007679e-06,
+      "loss": 1.0459,
+      "step": 842
+    },
+    {
+      "epoch": 0.08865868247728975,
+      "grad_norm": 2.7652313110005657,
+      "learning_rate": 4.923446724947443e-06,
+      "loss": 1.0426,
+      "step": 843
+    },
+    {
+      "epoch": 0.08876385291913708,
+      "grad_norm": 3.1194029297763133,
+      "learning_rate": 4.923241582654071e-06,
+      "loss": 1.0286,
+      "step": 844
+    },
+    {
+      "epoch": 0.0888690233609844,
+      "grad_norm": 4.097579369777372,
+      "learning_rate": 4.923036170150438e-06,
+      "loss": 1.0021,
+      "step": 845
+    },
+    {
+      "epoch": 0.08897419380283171,
+      "grad_norm": 2.741301220555694,
+      "learning_rate": 4.922830487459449e-06,
+      "loss": 1.0333,
+      "step": 846
+    },
+    {
+      "epoch": 0.08907936424467904,
+      "grad_norm": 3.542962902887874,
+      "learning_rate": 4.92262453460404e-06,
+      "loss": 1.0349,
+      "step": 847
+    },
+    {
+      "epoch": 0.08918453468652635,
+      "grad_norm": 2.855697339700191,
+      "learning_rate": 4.922418311607176e-06,
+      "loss": 1.0311,
+      "step": 848
+    },
+    {
+      "epoch": 0.08928970512837367,
+      "grad_norm": 2.880135557426853,
+      "learning_rate": 4.922211818491852e-06,
+      "loss": 1.0054,
+      "step": 849
+    },
+    {
+      "epoch": 0.089394875570221,
+      "grad_norm": 4.213156822451781,
+      "learning_rate": 4.922005055281094e-06,
+      "loss": 1.0374,
+      "step": 850
+    },
+    {
+      "epoch": 0.0895000460120683,
+      "grad_norm": 3.6779910799103632,
+      "learning_rate": 4.921798021997957e-06,
+      "loss": 0.9997,
+      "step": 851
+    },
+    {
+      "epoch": 0.08960521645391563,
+      "grad_norm": 4.758221935382132,
+      "learning_rate": 4.921590718665527e-06,
+      "loss": 1.0371,
+      "step": 852
+    },
+    {
+      "epoch": 0.08971038689576294,
+      "grad_norm": 3.3975540812884444,
+      "learning_rate": 4.921383145306922e-06,
+      "loss": 1.0318,
+      "step": 853
+    },
+    {
+      "epoch": 0.08981555733761026,
+      "grad_norm": 2.530639515419742,
+      "learning_rate": 4.921175301945284e-06,
+      "loss": 1.0652,
+      "step": 854
+    },
+    {
+      "epoch": 0.08992072777945759,
+      "grad_norm": 2.8662710938283986,
+      "learning_rate": 4.920967188603794e-06,
+      "loss": 1.024,
+      "step": 855
+    },
+    {
+      "epoch": 0.0900258982213049,
+      "grad_norm": 3.4254687930282737,
+      "learning_rate": 4.920758805305654e-06,
+      "loss": 1.0682,
+      "step": 856
+    },
+    {
+      "epoch": 0.09013106866315222,
+      "grad_norm": 2.1300497289243707,
+      "learning_rate": 4.920550152074103e-06,
+      "loss": 1.0442,
+      "step": 857
+    },
+    {
+      "epoch": 0.09023623910499955,
+      "grad_norm": 2.7404728800806417,
+      "learning_rate": 4.920341228932406e-06,
+      "loss": 1.0344,
+      "step": 858
+    },
+    {
+      "epoch": 0.09034140954684686,
+      "grad_norm": 3.579891745395384,
+      "learning_rate": 4.9201320359038595e-06,
+      "loss": 1.016,
+      "step": 859
+    },
+    {
+      "epoch": 0.09044657998869418,
+      "grad_norm": 3.5079338774058355,
+      "learning_rate": 4.919922573011791e-06,
+      "loss": 1.017,
+      "step": 860
+    },
+    {
+      "epoch": 0.09055175043054149,
+      "grad_norm": 4.474680092802719,
+      "learning_rate": 4.919712840279559e-06,
+      "loss": 1.1165,
+      "step": 861
+    },
+    {
+      "epoch": 0.09065692087238882,
+      "grad_norm": 2.7950268131476217,
+      "learning_rate": 4.9195028377305465e-06,
+      "loss": 1.0025,
+      "step": 862
+    },
+    {
+      "epoch": 0.09076209131423614,
+      "grad_norm": 3.6168740028611377,
+      "learning_rate": 4.919292565388172e-06,
+      "loss": 1.0431,
+      "step": 863
+    },
+    {
+      "epoch": 0.09086726175608345,
+      "grad_norm": 3.569181810804004,
+      "learning_rate": 4.919082023275884e-06,
+      "loss": 0.9941,
+      "step": 864
+    },
+    {
+      "epoch": 0.09097243219793077,
+      "grad_norm": 2.687691480063549,
+      "learning_rate": 4.918871211417157e-06,
+      "loss": 1.0376,
+      "step": 865
+    },
+    {
+      "epoch": 0.09107760263977809,
+      "grad_norm": 3.481903769104546,
+      "learning_rate": 4.9186601298355e-06,
+      "loss": 1.0285,
+      "step": 866
+    },
+    {
+      "epoch": 0.09118277308162541,
+      "grad_norm": 3.2967751447777798,
+      "learning_rate": 4.91844877855445e-06,
+      "loss": 1.0255,
+      "step": 867
+    },
+    {
+      "epoch": 0.09128794352347273,
+      "grad_norm": 3.258456644130292,
+      "learning_rate": 4.918237157597574e-06,
+      "loss": 1.009,
+      "step": 868
+    },
+    {
+      "epoch": 0.09139311396532004,
+      "grad_norm": 4.155311114794585,
+      "learning_rate": 4.918025266988469e-06,
+      "loss": 1.0554,
+      "step": 869
+    },
+    {
+      "epoch": 0.09149828440716737,
+      "grad_norm": 2.719205408216921,
+      "learning_rate": 4.917813106750763e-06,
+      "loss": 1.0526,
+      "step": 870
+    },
+    {
+      "epoch": 0.09160345484901468,
+      "grad_norm": 3.7086484713956644,
+      "learning_rate": 4.917600676908114e-06,
+      "loss": 1.0841,
+      "step": 871
+    },
+    {
+      "epoch": 0.091708625290862,
+      "grad_norm": 3.6291790131528265,
+      "learning_rate": 4.9173879774842085e-06,
+      "loss": 1.0466,
+      "step": 872
+    },
+    {
+      "epoch": 0.09181379573270933,
+      "grad_norm": 4.901385161683924,
+      "learning_rate": 4.917175008502763e-06,
+      "loss": 1.0683,
+      "step": 873
+    },
+    {
+      "epoch": 0.09191896617455664,
+      "grad_norm": 2.6391942505342145,
+      "learning_rate": 4.91696176998753e-06,
+      "loss": 1.0368,
+      "step": 874
+    },
+    {
+      "epoch": 0.09202413661640396,
+      "grad_norm": 2.92397301023947,
+      "learning_rate": 4.916748261962282e-06,
+      "loss": 1.0502,
+      "step": 875
+    },
+    {
+      "epoch": 0.09212930705825127,
+      "grad_norm": 3.3487651603386643,
+      "learning_rate": 4.916534484450829e-06,
+      "loss": 1.0492,
+      "step": 876
+    },
+    {
+      "epoch": 0.0922344775000986,
+      "grad_norm": 3.4178169663489615,
+      "learning_rate": 4.9163204374770085e-06,
+      "loss": 1.0182,
+      "step": 877
+    },
+    {
+      "epoch": 0.09233964794194592,
+      "grad_norm": 2.660490436372227,
+      "learning_rate": 4.916106121064689e-06,
+      "loss": 1.0636,
+      "step": 878
+    },
+    {
+      "epoch": 0.09244481838379323,
+      "grad_norm": 3.647201920237282,
+      "learning_rate": 4.915891535237768e-06,
+      "loss": 1.0018,
+      "step": 879
+    },
+    {
+      "epoch": 0.09254998882564056,
+      "grad_norm": 2.7950649078172547,
+      "learning_rate": 4.915676680020173e-06,
+      "loss": 1.0415,
+      "step": 880
+    },
+    {
+      "epoch": 0.09265515926748787,
+      "grad_norm": 3.308392412587543,
+      "learning_rate": 4.915461555435863e-06,
+      "loss": 1.0095,
+      "step": 881
+    },
+    {
+      "epoch": 0.09276032970933519,
+      "grad_norm": 2.497275944616539,
+      "learning_rate": 4.915246161508825e-06,
+      "loss": 1.0251,
+      "step": 882
+    },
+    {
+      "epoch": 0.09286550015118251,
+      "grad_norm": 2.4296994999785975,
+      "learning_rate": 4.915030498263079e-06,
+      "loss": 1.0309,
+      "step": 883
+    },
+    {
+      "epoch": 0.09297067059302982,
+      "grad_norm": 4.151821039375308,
+      "learning_rate": 4.914814565722671e-06,
+      "loss": 1.0479,
+      "step": 884
+    },
+    {
+      "epoch": 0.09307584103487715,
+      "grad_norm": 2.935706432111893,
+      "learning_rate": 4.91459836391168e-06,
+      "loss": 1.0566,
+      "step": 885
+    },
+    {
+      "epoch": 0.09318101147672447,
+      "grad_norm": 3.4020244169255527,
+      "learning_rate": 4.914381892854214e-06,
+      "loss": 1.0392,
+      "step": 886
+    },
+    {
+      "epoch": 0.09328618191857178,
+      "grad_norm": 3.4742996375581985,
+      "learning_rate": 4.914165152574412e-06,
+      "loss": 1.0323,
+      "step": 887
+    },
+    {
+      "epoch": 0.09339135236041911,
+      "grad_norm": 4.274200371311763,
+      "learning_rate": 4.913948143096442e-06,
+      "loss": 1.0627,
+      "step": 888
+    },
+    {
+      "epoch": 0.09349652280226642,
+      "grad_norm": 5.300639145263014,
+      "learning_rate": 4.9137308644445e-06,
+      "loss": 1.015,
+      "step": 889
+    },
+    {
+      "epoch": 0.09360169324411374,
+      "grad_norm": 3.5689399063136884,
+      "learning_rate": 4.913513316642818e-06,
+      "loss": 1.0453,
+      "step": 890
+    },
+    {
+      "epoch": 0.09370686368596107,
+      "grad_norm": 4.104293284314525,
+      "learning_rate": 4.913295499715651e-06,
+      "loss": 1.039,
+      "step": 891
+    },
+    {
+      "epoch": 0.09381203412780838,
+      "grad_norm": 2.9248882626299046,
+      "learning_rate": 4.913077413687289e-06,
+      "loss": 1.0304,
+      "step": 892
+    },
+    {
+      "epoch": 0.0939172045696557,
+      "grad_norm": 2.565656733778235,
+      "learning_rate": 4.91285905858205e-06,
+      "loss": 1.0233,
+      "step": 893
+    },
+    {
+      "epoch": 0.09402237501150301,
+      "grad_norm": 4.317147708295007,
+      "learning_rate": 4.912640434424283e-06,
+      "loss": 1.0293,
+      "step": 894
+    },
+    {
+      "epoch": 0.09412754545335034,
+      "grad_norm": 4.1886808246014615,
+      "learning_rate": 4.912421541238365e-06,
+      "loss": 1.0664,
+      "step": 895
+    },
+    {
+      "epoch": 0.09423271589519766,
+      "grad_norm": 3.5710378646260996,
+      "learning_rate": 4.912202379048704e-06,
+      "loss": 1.0582,
+      "step": 896
+    },
+    {
+      "epoch": 0.09433788633704497,
+      "grad_norm": 3.8755773448584523,
+      "learning_rate": 4.91198294787974e-06,
+      "loss": 1.0265,
+      "step": 897
+    },
+    {
+      "epoch": 0.0944430567788923,
+      "grad_norm": 3.2106737884292484,
+      "learning_rate": 4.91176324775594e-06,
+      "loss": 1.0467,
+      "step": 898
+    },
+    {
+      "epoch": 0.0945482272207396,
+      "grad_norm": 2.239928932171705,
+      "learning_rate": 4.911543278701802e-06,
+      "loss": 1.0499,
+      "step": 899
+    },
+    {
+      "epoch": 0.09465339766258693,
+      "grad_norm": 4.27992709829204,
+      "learning_rate": 4.9113230407418565e-06,
+      "loss": 1.0314,
+      "step": 900
+    },
+    {
+      "epoch": 0.09475856810443425,
+      "grad_norm": 3.7991157017709134,
+      "learning_rate": 4.911102533900659e-06,
+      "loss": 1.019,
+      "step": 901
+    },
+    {
+      "epoch": 0.09486373854628156,
+      "grad_norm": 3.2843177713241345,
+      "learning_rate": 4.910881758202799e-06,
+      "loss": 1.0747,
+      "step": 902
+    },
+    {
+      "epoch": 0.09496890898812889,
+      "grad_norm": 3.3614094256905336,
+      "learning_rate": 4.910660713672895e-06,
+      "loss": 1.047,
+      "step": 903
+    },
+    {
+      "epoch": 0.0950740794299762,
+      "grad_norm": 2.546889529528352,
+      "learning_rate": 4.910439400335595e-06,
+      "loss": 1.0134,
+      "step": 904
+    },
+    {
+      "epoch": 0.09517924987182352,
+      "grad_norm": 3.5988909556040927,
+      "learning_rate": 4.910217818215576e-06,
+      "loss": 1.0236,
+      "step": 905
+    },
+    {
+      "epoch": 0.09528442031367085,
+      "grad_norm": 4.327738766483183,
+      "learning_rate": 4.909995967337548e-06,
+      "loss": 1.0353,
+      "step": 906
+    },
+    {
+      "epoch": 0.09538959075551816,
+      "grad_norm": 4.8588271006072326,
+      "learning_rate": 4.9097738477262466e-06,
+      "loss": 1.0939,
+      "step": 907
+    },
+    {
+      "epoch": 0.09549476119736548,
+      "grad_norm": 4.033803872334996,
+      "learning_rate": 4.9095514594064434e-06,
+      "loss": 1.0285,
+      "step": 908
+    },
+    {
+      "epoch": 0.09559993163921279,
+      "grad_norm": 3.75933079753237,
+      "learning_rate": 4.9093288024029325e-06,
+      "loss": 1.0483,
+      "step": 909
+    },
+    {
+      "epoch": 0.09570510208106012,
+      "grad_norm": 3.0548663786720534,
+      "learning_rate": 4.9091058767405455e-06,
+      "loss": 1.051,
+      "step": 910
+    },
+    {
+      "epoch": 0.09581027252290744,
+      "grad_norm": 2.784053866442591,
+      "learning_rate": 4.908882682444137e-06,
+      "loss": 1.059,
+      "step": 911
+    },
+    {
+      "epoch": 0.09591544296475475,
+      "grad_norm": 2.130590850889847,
+      "learning_rate": 4.908659219538598e-06,
+      "loss": 1.0208,
+      "step": 912
+    },
+    {
+      "epoch": 0.09602061340660208,
+      "grad_norm": 2.3000679174789456,
+      "learning_rate": 4.908435488048844e-06,
+      "loss": 1.0253,
+      "step": 913
+    },
+    {
+      "epoch": 0.09612578384844939,
+      "grad_norm": 3.0043936554319193,
+      "learning_rate": 4.908211487999825e-06,
+      "loss": 1.0579,
+      "step": 914
+    },
+    {
+      "epoch": 0.09623095429029671,
+      "grad_norm": 2.864142159496675,
+      "learning_rate": 4.9079872194165155e-06,
+      "loss": 1.0353,
+      "step": 915
+    },
+    {
+      "epoch": 0.09633612473214404,
+      "grad_norm": 3.8952660611202754,
+      "learning_rate": 4.907762682323926e-06,
+      "loss": 1.0701,
+      "step": 916
+    },
+    {
+      "epoch": 0.09644129517399135,
+      "grad_norm": 3.9160674317154913,
+      "learning_rate": 4.907537876747094e-06,
+      "loss": 1.0547,
+      "step": 917
+    },
+    {
+      "epoch": 0.09654646561583867,
+      "grad_norm": 2.7081904691236613,
+      "learning_rate": 4.907312802711086e-06,
+      "loss": 0.9969,
+      "step": 918
+    },
+    {
+      "epoch": 0.096651636057686,
+      "grad_norm": 4.209552166676453,
+      "learning_rate": 4.907087460240999e-06,
+      "loss": 1.0433,
+      "step": 919
+    },
+    {
+      "epoch": 0.0967568064995333,
+      "grad_norm": 2.3073683643555083,
+      "learning_rate": 4.906861849361962e-06,
+      "loss": 1.0019,
+      "step": 920
+    },
+    {
+      "epoch": 0.09686197694138063,
+      "grad_norm": 3.94540024204696,
+      "learning_rate": 4.906635970099131e-06,
+      "loss": 1.0257,
+      "step": 921
+    },
+    {
+      "epoch": 0.09696714738322794,
+      "grad_norm": 3.9569458882368895,
+      "learning_rate": 4.906409822477695e-06,
+      "loss": 1.0383,
+      "step": 922
+    },
+    {
+      "epoch": 0.09707231782507526,
+      "grad_norm": 3.4725916392841043,
+      "learning_rate": 4.906183406522869e-06,
+      "loss": 1.006,
+      "step": 923
+    },
+    {
+      "epoch": 0.09717748826692259,
+      "grad_norm": 3.5244971631589577,
+      "learning_rate": 4.9059567222599015e-06,
+      "loss": 0.992,
+      "step": 924
+    },
+    {
+      "epoch": 0.0972826587087699,
+      "grad_norm": 3.274079488886004,
+      "learning_rate": 4.90572976971407e-06,
+      "loss": 0.992,
+      "step": 925
+    },
+    {
+      "epoch": 0.09738782915061722,
+      "grad_norm": 2.5329141838670455,
+      "learning_rate": 4.905502548910681e-06,
+      "loss": 1.041,
+      "step": 926
+    },
+    {
+      "epoch": 0.09749299959246453,
+      "grad_norm": 2.8385757720404765,
+      "learning_rate": 4.90527505987507e-06,
+      "loss": 1.0303,
+      "step": 927
+    },
+    {
+      "epoch": 0.09759817003431186,
+      "grad_norm": 2.930548884609321,
+      "learning_rate": 4.905047302632606e-06,
+      "loss": 1.0359,
+      "step": 928
+    },
+    {
+      "epoch": 0.09770334047615918,
+      "grad_norm": 4.156000403439952,
+      "learning_rate": 4.904819277208685e-06,
+      "loss": 1.0979,
+      "step": 929
+    },
+    {
+      "epoch": 0.09780851091800649,
+      "grad_norm": 2.536688544557416,
+      "learning_rate": 4.904590983628732e-06,
+      "loss": 1.0609,
+      "step": 930
+    },
+    {
+      "epoch": 0.09791368135985382,
+      "grad_norm": 3.9076531761075897,
+      "learning_rate": 4.904362421918205e-06,
+      "loss": 1.0202,
+      "step": 931
+    },
+    {
+      "epoch": 0.09801885180170113,
+      "grad_norm": 3.5518902322982493,
+      "learning_rate": 4.904133592102591e-06,
+      "loss": 1.0553,
+      "step": 932
+    },
+    {
+      "epoch": 0.09812402224354845,
+      "grad_norm": 3.1936214651164696,
+      "learning_rate": 4.9039044942074055e-06,
+      "loss": 1.0646,
+      "step": 933
+    },
+    {
+      "epoch": 0.09822919268539577,
+      "grad_norm": 4.135090085556305,
+      "learning_rate": 4.903675128258194e-06,
+      "loss": 1.0229,
+      "step": 934
+    },
+    {
+      "epoch": 0.09833436312724309,
+      "grad_norm": 2.2494098003681673,
+      "learning_rate": 4.903445494280534e-06,
+      "loss": 1.0082,
+      "step": 935
+    },
+    {
+      "epoch": 0.09843953356909041,
+      "grad_norm": 3.148431264860253,
+      "learning_rate": 4.90321559230003e-06,
+      "loss": 1.03,
+      "step": 936
+    },
+    {
+      "epoch": 0.09854470401093772,
+      "grad_norm": 3.4526369350300383,
+      "learning_rate": 4.902985422342319e-06,
+      "loss": 1.0412,
+      "step": 937
+    },
+    {
+      "epoch": 0.09864987445278504,
+      "grad_norm": 1.9933016033441238,
+      "learning_rate": 4.902754984433067e-06,
+      "loss": 1.0032,
+      "step": 938
+    },
+    {
+      "epoch": 0.09875504489463237,
+      "grad_norm": 2.560711359722772,
+      "learning_rate": 4.902524278597969e-06,
+      "loss": 1.0586,
+      "step": 939
+    },
+    {
+      "epoch": 0.09886021533647968,
+      "grad_norm": 3.6578892637833404,
+      "learning_rate": 4.9022933048627496e-06,
+      "loss": 1.0093,
+      "step": 940
+    },
+    {
+      "epoch": 0.098965385778327,
+      "grad_norm": 3.2769623438585196,
+      "learning_rate": 4.902062063253165e-06,
+      "loss": 1.0349,
+      "step": 941
+    },
+    {
+      "epoch": 0.09907055622017431,
+      "grad_norm": 2.629492634710313,
+      "learning_rate": 4.901830553795001e-06,
+      "loss": 1.0607,
+      "step": 942
+    },
+    {
+      "epoch": 0.09917572666202164,
+      "grad_norm": 3.126871970690739,
+      "learning_rate": 4.9015987765140715e-06,
+      "loss": 1.0513,
+      "step": 943
+    },
+    {
+      "epoch": 0.09928089710386896,
+      "grad_norm": 5.290623571841186,
+      "learning_rate": 4.901366731436223e-06,
+      "loss": 1.0321,
+      "step": 944
+    },
+    {
+      "epoch": 0.09938606754571627,
+      "grad_norm": 2.549896186428418,
+      "learning_rate": 4.901134418587329e-06,
+      "loss": 1.0487,
+      "step": 945
+    },
+    {
+      "epoch": 0.0994912379875636,
+      "grad_norm": 3.3781283890627596,
+      "learning_rate": 4.900901837993295e-06,
+      "loss": 1.0193,
+      "step": 946
+    },
+    {
+      "epoch": 0.09959640842941092,
+      "grad_norm": 3.3355283132202738,
+      "learning_rate": 4.900668989680055e-06,
+      "loss": 0.9998,
+      "step": 947
+    },
+    {
+      "epoch": 0.09970157887125823,
+      "grad_norm": 2.6098798632254003,
+      "learning_rate": 4.900435873673574e-06,
+      "loss": 0.9725,
+      "step": 948
+    },
+    {
+      "epoch": 0.09980674931310556,
+      "grad_norm": 2.802666593583616,
+      "learning_rate": 4.900202489999845e-06,
+      "loss": 1.0438,
+      "step": 949
+    },
+    {
+      "epoch": 0.09991191975495287,
+      "grad_norm": 2.0938671069598858,
+      "learning_rate": 4.899968838684893e-06,
+      "loss": 1.0205,
+      "step": 950
+    },
+    {
+      "epoch": 0.10001709019680019,
+      "grad_norm": 3.6038776689006586,
+      "learning_rate": 4.8997349197547724e-06,
+      "loss": 1.0578,
+      "step": 951
+    },
+    {
+      "epoch": 0.10012226063864751,
+      "grad_norm": 4.970829658802114,
+      "learning_rate": 4.899500733235567e-06,
+      "loss": 1.025,
+      "step": 952
+    },
+    {
+      "epoch": 0.10022743108049482,
+      "grad_norm": 4.678353320775025,
+      "learning_rate": 4.899266279153388e-06,
+      "loss": 1.0622,
+      "step": 953
+    },
+    {
+      "epoch": 0.10033260152234215,
+      "grad_norm": 5.741743419399476,
+      "learning_rate": 4.899031557534383e-06,
+      "loss": 1.0874,
+      "step": 954
+    },
+    {
+      "epoch": 0.10043777196418946,
+      "grad_norm": 3.104804866637022,
+      "learning_rate": 4.8987965684047215e-06,
+      "loss": 1.0013,
+      "step": 955
+    },
+    {
+      "epoch": 0.10054294240603678,
+      "grad_norm": 3.03354521888864,
+      "learning_rate": 4.898561311790609e-06,
+      "loss": 1.004,
+      "step": 956
+    },
+    {
+      "epoch": 0.10064811284788411,
+      "grad_norm": 3.424553202851718,
+      "learning_rate": 4.898325787718277e-06,
+      "loss": 1.0363,
+      "step": 957
+    },
+    {
+      "epoch": 0.10075328328973142,
+      "grad_norm": 3.000149075657252,
+      "learning_rate": 4.898089996213988e-06,
+      "loss": 1.0371,
+      "step": 958
+    },
+    {
+      "epoch": 0.10085845373157874,
+      "grad_norm": 2.3914498005066194,
+      "learning_rate": 4.897853937304037e-06,
+      "loss": 1.0101,
+      "step": 959
+    },
+    {
+      "epoch": 0.10096362417342605,
+      "grad_norm": 3.7269730617951997,
+      "learning_rate": 4.897617611014744e-06,
+      "loss": 1.0741,
+      "step": 960
+    },
+    {
+      "epoch": 0.10106879461527338,
+      "grad_norm": 3.4685995382297,
+      "learning_rate": 4.897381017372462e-06,
+      "loss": 1.0635,
+      "step": 961
+    },
+    {
+      "epoch": 0.1011739650571207,
+      "grad_norm": 2.414965651072397,
+      "learning_rate": 4.897144156403573e-06,
+      "loss": 1.0601,
+      "step": 962
+    },
+    {
+      "epoch": 0.10127913549896801,
+      "grad_norm": 3.3913472210312117,
+      "learning_rate": 4.8969070281344895e-06,
+      "loss": 1.0682,
+      "step": 963
+    },
+    {
+      "epoch": 0.10138430594081534,
+      "grad_norm": 2.6184438328601387,
+      "learning_rate": 4.896669632591652e-06,
+      "loss": 1.0295,
+      "step": 964
+    },
+    {
+      "epoch": 0.10148947638266265,
+      "grad_norm": 3.3103117280466154,
+      "learning_rate": 4.8964319698015325e-06,
+      "loss": 1.0155,
+      "step": 965
+    },
+    {
+      "epoch": 0.10159464682450997,
+      "grad_norm": 2.7095856980343966,
+      "learning_rate": 4.896194039790632e-06,
+      "loss": 1.0172,
+      "step": 966
+    },
+    {
+      "epoch": 0.1016998172663573,
+      "grad_norm": 3.606502874104067,
+      "learning_rate": 4.895955842585483e-06,
+      "loss": 1.0368,
+      "step": 967
+    },
+    {
+      "epoch": 0.1018049877082046,
+      "grad_norm": 3.7818643892335024,
+      "learning_rate": 4.895717378212644e-06,
+      "loss": 1.0736,
+      "step": 968
+    },
+    {
+      "epoch": 0.10191015815005193,
+      "grad_norm": 3.9660970245154865,
+      "learning_rate": 4.895478646698707e-06,
+      "loss": 1.0143,
+      "step": 969
+    },
+    {
+      "epoch": 0.10201532859189924,
+      "grad_norm": 3.472975484177083,
+      "learning_rate": 4.895239648070292e-06,
+      "loss": 1.0414,
+      "step": 970
+    },
+    {
+      "epoch": 0.10212049903374656,
+      "grad_norm": 3.2923024045337277,
+      "learning_rate": 4.895000382354049e-06,
+      "loss": 1.0605,
+      "step": 971
+    },
+    {
+      "epoch": 0.10222566947559389,
+      "grad_norm": 4.414663168893738,
+      "learning_rate": 4.89476084957666e-06,
+      "loss": 1.0273,
+      "step": 972
+    },
+    {
+      "epoch": 0.1023308399174412,
+      "grad_norm": 5.295626697084187,
+      "learning_rate": 4.894521049764831e-06,
+      "loss": 1.0586,
+      "step": 973
+    },
+    {
+      "epoch": 0.10243601035928852,
+      "grad_norm": 2.9933688401634213,
+      "learning_rate": 4.8942809829453046e-06,
+      "loss": 1.0011,
+      "step": 974
+    },
+    {
+      "epoch": 0.10254118080113583,
+      "grad_norm": 3.1480184087773253,
+      "learning_rate": 4.894040649144849e-06,
+      "loss": 1.0485,
+      "step": 975
+    },
+    {
+      "epoch": 0.10264635124298316,
+      "grad_norm": 3.0397182716899254,
+      "learning_rate": 4.893800048390264e-06,
+      "loss": 1.0304,
+      "step": 976
+    },
+    {
+      "epoch": 0.10275152168483048,
+      "grad_norm": 3.4747456565622787,
+      "learning_rate": 4.893559180708378e-06,
+      "loss": 1.0068,
+      "step": 977
+    },
+    {
+      "epoch": 0.10285669212667779,
+      "grad_norm": 3.54567802483446,
+      "learning_rate": 4.8933180461260485e-06,
+      "loss": 1.0231,
+      "step": 978
+    },
+    {
+      "epoch": 0.10296186256852512,
+      "grad_norm": 3.4856402750957307,
+      "learning_rate": 4.893076644670166e-06,
+      "loss": 1.0518,
+      "step": 979
+    },
+    {
+      "epoch": 0.10306703301037244,
+      "grad_norm": 2.97684659916138,
+      "learning_rate": 4.892834976367647e-06,
+      "loss": 0.9815,
+      "step": 980
+    },
+    {
+      "epoch": 0.10317220345221975,
+      "grad_norm": 4.321536953737777,
+      "learning_rate": 4.8925930412454405e-06,
+      "loss": 1.0374,
+      "step": 981
+    },
+    {
+      "epoch": 0.10327737389406708,
+      "grad_norm": 3.6585034863508605,
+      "learning_rate": 4.8923508393305224e-06,
+      "loss": 0.9504,
+      "step": 982
+    },
+    {
+      "epoch": 0.10338254433591439,
+      "grad_norm": 3.3098381288492074,
+      "learning_rate": 4.892108370649902e-06,
+      "loss": 1.0276,
+      "step": 983
+    },
+    {
+      "epoch": 0.10348771477776171,
+      "grad_norm": 3.3462769999723383,
+      "learning_rate": 4.891865635230616e-06,
+      "loss": 1.0415,
+      "step": 984
+    },
+    {
+      "epoch": 0.10359288521960903,
+      "grad_norm": 3.1379133999705426,
+      "learning_rate": 4.891622633099731e-06,
+      "loss": 1.0503,
+      "step": 985
+    },
+    {
+      "epoch": 0.10369805566145635,
+      "grad_norm": 2.096943884410061,
+      "learning_rate": 4.8913793642843434e-06,
+      "loss": 1.0132,
+      "step": 986
+    },
+    {
+      "epoch": 0.10380322610330367,
+      "grad_norm": 3.49899699410188,
+      "learning_rate": 4.89113582881158e-06,
+      "loss": 1.036,
+      "step": 987
+    },
+    {
+      "epoch": 0.10390839654515098,
+      "grad_norm": 4.363397554483991,
+      "learning_rate": 4.890892026708596e-06,
+      "loss": 1.0641,
+      "step": 988
+    },
+    {
+      "epoch": 0.1040135669869983,
+      "grad_norm": 2.843253723260328,
+      "learning_rate": 4.8906479580025774e-06,
+      "loss": 1.0259,
+      "step": 989
+    },
+    {
+      "epoch": 0.10411873742884563,
+      "grad_norm": 3.720334378029494,
+      "learning_rate": 4.890403622720742e-06,
+      "loss": 1.0301,
+      "step": 990
+    },
+    {
+      "epoch": 0.10422390787069294,
+      "grad_norm": 3.4282471758253483,
+      "learning_rate": 4.890159020890333e-06,
+      "loss": 1.0388,
+      "step": 991
+    },
+    {
+      "epoch": 0.10432907831254026,
+      "grad_norm": 4.570380050609438,
+      "learning_rate": 4.889914152538625e-06,
+      "loss": 1.0445,
+      "step": 992
+    },
+    {
+      "epoch": 0.10443424875438757,
+      "grad_norm": 4.005408760979363,
+      "learning_rate": 4.889669017692924e-06,
+      "loss": 1.0779,
+      "step": 993
+    },
+    {
+      "epoch": 0.1045394191962349,
+      "grad_norm": 3.317322954004475,
+      "learning_rate": 4.889423616380564e-06,
+      "loss": 1.0502,
+      "step": 994
+    },
+    {
+      "epoch": 0.10464458963808222,
+      "grad_norm": 2.196763228351377,
+      "learning_rate": 4.889177948628908e-06,
+      "loss": 1.0265,
+      "step": 995
+    },
+    {
+      "epoch": 0.10474976007992953,
+      "grad_norm": 2.877333453688417,
+      "learning_rate": 4.8889320144653525e-06,
+      "loss": 1.0354,
+      "step": 996
+    },
+    {
+      "epoch": 0.10485493052177686,
+      "grad_norm": 1.912523311494541,
+      "learning_rate": 4.8886858139173185e-06,
+      "loss": 1.0046,
+      "step": 997
+    },
+    {
+      "epoch": 0.10496010096362417,
+      "grad_norm": 4.312918317607742,
+      "learning_rate": 4.88843934701226e-06,
+      "loss": 1.0117,
+      "step": 998
+    },
+    {
+      "epoch": 0.10506527140547149,
+      "grad_norm": 3.0476518763567144,
+      "learning_rate": 4.888192613777661e-06,
+      "loss": 1.0288,
+      "step": 999
+    },
+    {
+      "epoch": 0.10517044184731882,
+      "grad_norm": 2.9241631966551433,
+      "learning_rate": 4.887945614241034e-06,
+      "loss": 1.0364,
+      "step": 1000
+    },
+    {
+      "epoch": 0.10527561228916613,
+      "grad_norm": 2.0938595160121287,
+      "learning_rate": 4.88769834842992e-06,
+      "loss": 0.999,
+      "step": 1001
+    },
+    {
+      "epoch": 0.10538078273101345,
+      "grad_norm": 2.9042647406620397,
+      "learning_rate": 4.887450816371892e-06,
+      "loss": 1.0036,
+      "step": 1002
+    },
+    {
+      "epoch": 0.10548595317286076,
+      "grad_norm": 3.430761146907295,
+      "learning_rate": 4.887203018094552e-06,
+      "loss": 1.0187,
+      "step": 1003
+    },
+    {
+      "epoch": 0.10559112361470809,
+      "grad_norm": 3.564602317410496,
+      "learning_rate": 4.88695495362553e-06,
+      "loss": 1.0255,
+      "step": 1004
+    },
+    {
+      "epoch": 0.10569629405655541,
+      "grad_norm": 2.3078908692690443,
+      "learning_rate": 4.886706622992489e-06,
+      "loss": 1.0303,
+      "step": 1005
+    },
+    {
+      "epoch": 0.10580146449840272,
+      "grad_norm": 2.761094777536776,
+      "learning_rate": 4.886458026223118e-06,
+      "loss": 1.0259,
+      "step": 1006
+    },
+    {
+      "epoch": 0.10590663494025004,
+      "grad_norm": 3.4909322317604388,
+      "learning_rate": 4.88620916334514e-06,
+      "loss": 1.0701,
+      "step": 1007
+    },
+    {
+      "epoch": 0.10601180538209737,
+      "grad_norm": 3.042806287394382,
+      "learning_rate": 4.885960034386302e-06,
+      "loss": 1.0188,
+      "step": 1008
+    },
+    {
+      "epoch": 0.10611697582394468,
+      "grad_norm": 3.4715361900462773,
+      "learning_rate": 4.885710639374387e-06,
+      "loss": 1.0059,
+      "step": 1009
+    },
+    {
+      "epoch": 0.106222146265792,
+      "grad_norm": 2.7625550172439524,
+      "learning_rate": 4.885460978337201e-06,
+      "loss": 1.0592,
+      "step": 1010
+    },
+    {
+      "epoch": 0.10632731670763931,
+      "grad_norm": 2.7235208933430597,
+      "learning_rate": 4.885211051302586e-06,
+      "loss": 1.0214,
+      "step": 1011
+    },
+    {
+      "epoch": 0.10643248714948664,
+      "grad_norm": 2.7923910284329403,
+      "learning_rate": 4.88496085829841e-06,
+      "loss": 1.0122,
+      "step": 1012
+    },
+    {
+      "epoch": 0.10653765759133396,
+      "grad_norm": 4.09723932625378,
+      "learning_rate": 4.884710399352572e-06,
+      "loss": 1.0252,
+      "step": 1013
+    },
+    {
+      "epoch": 0.10664282803318127,
+      "grad_norm": 4.277407726951024,
+      "learning_rate": 4.884459674492997e-06,
+      "loss": 1.0445,
+      "step": 1014
+    },
+    {
+      "epoch": 0.1067479984750286,
+      "grad_norm": 2.741940824883468,
+      "learning_rate": 4.884208683747647e-06,
+      "loss": 1.0452,
+      "step": 1015
+    },
+    {
+      "epoch": 0.1068531689168759,
+      "grad_norm": 2.846019968822997,
+      "learning_rate": 4.883957427144507e-06,
+      "loss": 1.0386,
+      "step": 1016
+    },
+    {
+      "epoch": 0.10695833935872323,
+      "grad_norm": 3.0668518235856776,
+      "learning_rate": 4.8837059047115955e-06,
+      "loss": 1.073,
+      "step": 1017
+    },
+    {
+      "epoch": 0.10706350980057056,
+      "grad_norm": 2.7044021207813835,
+      "learning_rate": 4.883454116476957e-06,
+      "loss": 1.025,
+      "step": 1018
+    },
+    {
+      "epoch": 0.10716868024241787,
+      "grad_norm": 2.9884834738601533,
+      "learning_rate": 4.88320206246867e-06,
+      "loss": 1.0362,
+      "step": 1019
+    },
+    {
+      "epoch": 0.10727385068426519,
+      "grad_norm": 2.6459541792130823,
+      "learning_rate": 4.88294974271484e-06,
+      "loss": 1.035,
+      "step": 1020
+    },
+    {
+      "epoch": 0.1073790211261125,
+      "grad_norm": 3.296712064083779,
+      "learning_rate": 4.882697157243601e-06,
+      "loss": 1.0486,
+      "step": 1021
+    },
+    {
+      "epoch": 0.10748419156795982,
+      "grad_norm": 3.5876433786961885,
+      "learning_rate": 4.882444306083121e-06,
+      "loss": 1.0428,
+      "step": 1022
+    },
+    {
+      "epoch": 0.10758936200980715,
+      "grad_norm": 2.4991241646439426,
+      "learning_rate": 4.882191189261592e-06,
+      "loss": 1.0133,
+      "step": 1023
+    },
+    {
+      "epoch": 0.10769453245165446,
+      "grad_norm": 3.07812377252626,
+      "learning_rate": 4.881937806807241e-06,
+      "loss": 1.0337,
+      "step": 1024
+    },
+    {
+      "epoch": 0.10779970289350178,
+      "grad_norm": 2.229456998154068,
+      "learning_rate": 4.881684158748321e-06,
+      "loss": 1.0208,
+      "step": 1025
+    },
+    {
+      "epoch": 0.1079048733353491,
+      "grad_norm": 3.7922511732147144,
+      "learning_rate": 4.881430245113115e-06,
+      "loss": 1.0326,
+      "step": 1026
+    },
+    {
+      "epoch": 0.10801004377719642,
+      "grad_norm": 3.1647389772945695,
+      "learning_rate": 4.881176065929938e-06,
+      "loss": 1.0131,
+      "step": 1027
+    },
+    {
+      "epoch": 0.10811521421904374,
+      "grad_norm": 2.8627766739221787,
+      "learning_rate": 4.880921621227131e-06,
+      "loss": 1.024,
+      "step": 1028
+    },
+    {
+      "epoch": 0.10822038466089105,
+      "grad_norm": 2.5687462636056035,
+      "learning_rate": 4.880666911033068e-06,
+      "loss": 1.0409,
+      "step": 1029
+    },
+    {
+      "epoch": 0.10832555510273838,
+      "grad_norm": 2.916734582685727,
+      "learning_rate": 4.880411935376151e-06,
+      "loss": 1.0788,
+      "step": 1030
+    },
+    {
+      "epoch": 0.10843072554458569,
+      "grad_norm": 4.000486986234842,
+      "learning_rate": 4.880156694284811e-06,
+      "loss": 1.0599,
+      "step": 1031
+    },
+    {
+      "epoch": 0.10853589598643301,
+      "grad_norm": 3.5425566453452295,
+      "learning_rate": 4.87990118778751e-06,
+      "loss": 1.0306,
+      "step": 1032
+    },
+    {
+      "epoch": 0.10864106642828034,
+      "grad_norm": 3.259354111862751,
+      "learning_rate": 4.879645415912739e-06,
+      "loss": 1.0667,
+      "step": 1033
+    },
+    {
+      "epoch": 0.10874623687012765,
+      "grad_norm": 3.7603711436052585,
+      "learning_rate": 4.8793893786890186e-06,
+      "loss": 1.0272,
+      "step": 1034
+    },
+    {
+      "epoch": 0.10885140731197497,
+      "grad_norm": 2.650644535850584,
+      "learning_rate": 4.879133076144898e-06,
+      "loss": 1.0186,
+      "step": 1035
+    },
+    {
+      "epoch": 0.1089565777538223,
+      "grad_norm": 3.152266769786039,
+      "learning_rate": 4.8788765083089586e-06,
+      "loss": 1.0406,
+      "step": 1036
+    },
+    {
+      "epoch": 0.1090617481956696,
+      "grad_norm": 3.240897604655947,
+      "learning_rate": 4.878619675209809e-06,
+      "loss": 1.0398,
+      "step": 1037
+    },
+    {
+      "epoch": 0.10916691863751693,
+      "grad_norm": 2.145650055329262,
+      "learning_rate": 4.8783625768760865e-06,
+      "loss": 1.0149,
+      "step": 1038
+    },
+    {
+      "epoch": 0.10927208907936424,
+      "grad_norm": 2.495958954073448,
+      "learning_rate": 4.878105213336462e-06,
+      "loss": 1.0143,
+      "step": 1039
+    },
+    {
+      "epoch": 0.10937725952121156,
+      "grad_norm": 2.669064484615587,
+      "learning_rate": 4.877847584619632e-06,
+      "loss": 1.0499,
+      "step": 1040
+    },
+    {
+      "epoch": 0.10948242996305889,
+      "grad_norm": 2.877634072763941,
+      "learning_rate": 4.8775896907543245e-06,
+      "loss": 1.0111,
+      "step": 1041
+    },
+    {
+      "epoch": 0.1095876004049062,
+      "grad_norm": 2.928402468643089,
+      "learning_rate": 4.877331531769297e-06,
+      "loss": 1.0233,
+      "step": 1042
+    },
+    {
+      "epoch": 0.10969277084675352,
+      "grad_norm": 3.1475271796192756,
+      "learning_rate": 4.877073107693336e-06,
+      "loss": 1.0445,
+      "step": 1043
+    },
+    {
+      "epoch": 0.10979794128860083,
+      "grad_norm": 2.7416601116581,
+      "learning_rate": 4.876814418555257e-06,
+      "loss": 1.0513,
+      "step": 1044
+    },
+    {
+      "epoch": 0.10990311173044816,
+      "grad_norm": 2.468594567435823,
+      "learning_rate": 4.876555464383908e-06,
+      "loss": 1.0396,
+      "step": 1045
+    },
+    {
+      "epoch": 0.11000828217229548,
+      "grad_norm": 3.7894959288974306,
+      "learning_rate": 4.876296245208162e-06,
+      "loss": 1.0068,
+      "step": 1046
+    },
+    {
+      "epoch": 0.11011345261414279,
+      "grad_norm": 4.315993647092533,
+      "learning_rate": 4.876036761056925e-06,
+      "loss": 1.0244,
+      "step": 1047
+    },
+    {
+      "epoch": 0.11021862305599012,
+      "grad_norm": 2.7986247997518605,
+      "learning_rate": 4.875777011959131e-06,
+      "loss": 1.0252,
+      "step": 1048
+    },
+    {
+      "epoch": 0.11032379349783743,
+      "grad_norm": 2.872248737665653,
+      "learning_rate": 4.875516997943745e-06,
+      "loss": 1.0191,
+      "step": 1049
+    },
+    {
+      "epoch": 0.11042896393968475,
+      "grad_norm": 3.51973767363257,
+      "learning_rate": 4.8752567190397605e-06,
+      "loss": 1.0226,
+      "step": 1050
+    },
+    {
+      "epoch": 0.11053413438153208,
+      "grad_norm": 2.793611486570461,
+      "learning_rate": 4.874996175276199e-06,
+      "loss": 1.0259,
+      "step": 1051
+    },
+    {
+      "epoch": 0.11063930482337939,
+      "grad_norm": 3.5618775194605243,
+      "learning_rate": 4.8747353666821155e-06,
+      "loss": 1.0744,
+      "step": 1052
+    },
+    {
+      "epoch": 0.11074447526522671,
+      "grad_norm": 3.7992122350452693,
+      "learning_rate": 4.8744742932865905e-06,
+      "loss": 1.0727,
+      "step": 1053
+    },
+    {
+      "epoch": 0.11084964570707402,
+      "grad_norm": 2.1766727817716065,
+      "learning_rate": 4.874212955118736e-06,
+      "loss": 1.0345,
+      "step": 1054
+    },
+    {
+      "epoch": 0.11095481614892135,
+      "grad_norm": 2.988093572670979,
+      "learning_rate": 4.873951352207694e-06,
+      "loss": 0.9793,
+      "step": 1055
+    },
+    {
+      "epoch": 0.11105998659076867,
+      "grad_norm": 3.100468284349954,
+      "learning_rate": 4.873689484582634e-06,
+      "loss": 1.1039,
+      "step": 1056
+    },
+    {
+      "epoch": 0.11116515703261598,
+      "grad_norm": 2.234883911073863,
+      "learning_rate": 4.873427352272758e-06,
+      "loss": 1.0484,
+      "step": 1057
+    },
+    {
+      "epoch": 0.1112703274744633,
+      "grad_norm": 3.034027361644701,
+      "learning_rate": 4.8731649553072945e-06,
+      "loss": 1.0477,
+      "step": 1058
+    },
+    {
+      "epoch": 0.11137549791631061,
+      "grad_norm": 2.7701642330357976,
+      "learning_rate": 4.872902293715502e-06,
+      "loss": 0.9995,
+      "step": 1059
+    },
+    {
+      "epoch": 0.11148066835815794,
+      "grad_norm": 3.5596220939294545,
+      "learning_rate": 4.872639367526672e-06,
+      "loss": 1.0595,
+      "step": 1060
+    },
+    {
+      "epoch": 0.11158583880000526,
+      "grad_norm": 2.4396549311475755,
+      "learning_rate": 4.872376176770121e-06,
+      "loss": 1.0638,
+      "step": 1061
+    },
+    {
+      "epoch": 0.11169100924185257,
+      "grad_norm": 2.7581484936556913,
+      "learning_rate": 4.872112721475196e-06,
+      "loss": 1.0511,
+      "step": 1062
+    },
+    {
+      "epoch": 0.1117961796836999,
+      "grad_norm": 3.59528480669715,
+      "learning_rate": 4.871849001671276e-06,
+      "loss": 1.0209,
+      "step": 1063
+    },
+    {
+      "epoch": 0.11190135012554721,
+      "grad_norm": 3.0507361752299187,
+      "learning_rate": 4.871585017387767e-06,
+      "loss": 1.0476,
+      "step": 1064
+    },
+    {
+      "epoch": 0.11200652056739453,
+      "grad_norm": 3.6738695742416176,
+      "learning_rate": 4.871320768654105e-06,
+      "loss": 1.0659,
+      "step": 1065
+    },
+    {
+      "epoch": 0.11211169100924186,
+      "grad_norm": 3.8253088021331867,
+      "learning_rate": 4.871056255499758e-06,
+      "loss": 0.9856,
+      "step": 1066
+    },
+    {
+      "epoch": 0.11221686145108917,
+      "grad_norm": 3.1580111988558994,
+      "learning_rate": 4.870791477954218e-06,
+      "loss": 1.0082,
+      "step": 1067
+    },
+    {
+      "epoch": 0.11232203189293649,
+      "grad_norm": 3.0599335655636146,
+      "learning_rate": 4.87052643604701e-06,
+      "loss": 1.0662,
+      "step": 1068
+    },
+    {
+      "epoch": 0.11242720233478382,
+      "grad_norm": 2.2702607865189224,
+      "learning_rate": 4.870261129807692e-06,
+      "loss": 1.0522,
+      "step": 1069
+    },
+    {
+      "epoch": 0.11253237277663113,
+      "grad_norm": 2.6723836637796845,
+      "learning_rate": 4.869995559265844e-06,
+      "loss": 1.0286,
+      "step": 1070
+    },
+    {
+      "epoch": 0.11263754321847845,
+      "grad_norm": 4.21492156394967,
+      "learning_rate": 4.869729724451081e-06,
+      "loss": 1.0665,
+      "step": 1071
+    },
+    {
+      "epoch": 0.11274271366032576,
+      "grad_norm": 2.8588027598894703,
+      "learning_rate": 4.869463625393044e-06,
+      "loss": 1.0236,
+      "step": 1072
+    },
+    {
+      "epoch": 0.11284788410217308,
+      "grad_norm": 3.3016899697968882,
+      "learning_rate": 4.869197262121406e-06,
+      "loss": 1.0559,
+      "step": 1073
+    },
+    {
+      "epoch": 0.11295305454402041,
+      "grad_norm": 3.419706325986653,
+      "learning_rate": 4.8689306346658704e-06,
+      "loss": 1.011,
+      "step": 1074
+    },
+    {
+      "epoch": 0.11305822498586772,
+      "grad_norm": 2.0613879638549,
+      "learning_rate": 4.868663743056165e-06,
+      "loss": 0.9756,
+      "step": 1075
+    },
+    {
+      "epoch": 0.11316339542771504,
+      "grad_norm": 3.333749579337557,
+      "learning_rate": 4.868396587322053e-06,
+      "loss": 1.0821,
+      "step": 1076
+    },
+    {
+      "epoch": 0.11326856586956235,
+      "grad_norm": 3.0234091587963032,
+      "learning_rate": 4.868129167493322e-06,
+      "loss": 1.0416,
+      "step": 1077
+    },
+    {
+      "epoch": 0.11337373631140968,
+      "grad_norm": 3.1722297477760466,
+      "learning_rate": 4.867861483599793e-06,
+      "loss": 1.049,
+      "step": 1078
+    },
+    {
+      "epoch": 0.113478906753257,
+      "grad_norm": 3.123942723713283,
+      "learning_rate": 4.867593535671315e-06,
+      "loss": 1.0038,
+      "step": 1079
+    },
+    {
+      "epoch": 0.11358407719510431,
+      "grad_norm": 3.3849552815537054,
+      "learning_rate": 4.867325323737765e-06,
+      "loss": 0.9871,
+      "step": 1080
+    },
+    {
+      "epoch": 0.11368924763695164,
+      "grad_norm": 3.6000442180720498,
+      "learning_rate": 4.8670568478290515e-06,
+      "loss": 1.0303,
+      "step": 1081
+    },
+    {
+      "epoch": 0.11379441807879895,
+      "grad_norm": 3.0065860923788676,
+      "learning_rate": 4.866788107975111e-06,
+      "loss": 1.0284,
+      "step": 1082
+    },
+    {
+      "epoch": 0.11389958852064627,
+      "grad_norm": 3.307728595187707,
+      "learning_rate": 4.866519104205911e-06,
+      "loss": 1.0484,
+      "step": 1083
+    },
+    {
+      "epoch": 0.1140047589624936,
+      "grad_norm": 2.9561692789631078,
+      "learning_rate": 4.866249836551447e-06,
+      "loss": 1.012,
+      "step": 1084
+    },
+    {
+      "epoch": 0.1141099294043409,
+      "grad_norm": 2.5709168983222503,
+      "learning_rate": 4.865980305041746e-06,
+      "loss": 1.024,
+      "step": 1085
+    },
+    {
+      "epoch": 0.11421509984618823,
+      "grad_norm": 2.7856609101318104,
+      "learning_rate": 4.865710509706859e-06,
+      "loss": 1.0148,
+      "step": 1086
+    },
+    {
+      "epoch": 0.11432027028803554,
+      "grad_norm": 4.12784259860208,
+      "learning_rate": 4.8654404505768735e-06,
+      "loss": 1.0658,
+      "step": 1087
+    },
+    {
+      "epoch": 0.11442544072988287,
+      "grad_norm": 3.9354921027135186,
+      "learning_rate": 4.865170127681903e-06,
+      "loss": 1.0093,
+      "step": 1088
+    },
+    {
+      "epoch": 0.11453061117173019,
+      "grad_norm": 4.361516591990753,
+      "learning_rate": 4.8648995410520905e-06,
+      "loss": 1.0849,
+      "step": 1089
+    },
+    {
+      "epoch": 0.1146357816135775,
+      "grad_norm": 3.874224839358617,
+      "learning_rate": 4.864628690717607e-06,
+      "loss": 1.0004,
+      "step": 1090
+    },
+    {
+      "epoch": 0.11474095205542482,
+      "grad_norm": 2.713904318986021,
+      "learning_rate": 4.8643575767086555e-06,
+      "loss": 0.9988,
+      "step": 1091
+    },
+    {
+      "epoch": 0.11484612249727214,
+      "grad_norm": 4.127820079366728,
+      "learning_rate": 4.864086199055467e-06,
+      "loss": 1.0265,
+      "step": 1092
+    },
+    {
+      "epoch": 0.11495129293911946,
+      "grad_norm": 4.16373977878337,
+      "learning_rate": 4.863814557788303e-06,
+      "loss": 1.0445,
+      "step": 1093
+    },
+    {
+      "epoch": 0.11505646338096678,
+      "grad_norm": 4.340273179649079,
+      "learning_rate": 4.863542652937453e-06,
+      "loss": 1.0214,
+      "step": 1094
+    },
+    {
+      "epoch": 0.1151616338228141,
+      "grad_norm": 3.5979531619959593,
+      "learning_rate": 4.863270484533237e-06,
+      "loss": 0.9954,
+      "step": 1095
+    },
+    {
+      "epoch": 0.11526680426466142,
+      "grad_norm": 3.7563254164355624,
+      "learning_rate": 4.862998052606001e-06,
+      "loss": 1.0446,
+      "step": 1096
+    },
+    {
+      "epoch": 0.11537197470650874,
+      "grad_norm": 2.8311030399401904,
+      "learning_rate": 4.862725357186129e-06,
+      "loss": 0.9762,
+      "step": 1097
+    },
+    {
+      "epoch": 0.11547714514835605,
+      "grad_norm": 3.26148888613775,
+      "learning_rate": 4.862452398304024e-06,
+      "loss": 1.0667,
+      "step": 1098
+    },
+    {
+      "epoch": 0.11558231559020338,
+      "grad_norm": 2.2978414859803964,
+      "learning_rate": 4.862179175990124e-06,
+      "loss": 0.9832,
+      "step": 1099
+    },
+    {
+      "epoch": 0.11568748603205069,
+      "grad_norm": 2.8821374960941304,
+      "learning_rate": 4.861905690274896e-06,
+      "loss": 1.0225,
+      "step": 1100
+    },
+    {
+      "epoch": 0.11579265647389801,
+      "grad_norm": 2.657398753259678,
+      "learning_rate": 4.861631941188836e-06,
+      "loss": 1.036,
+      "step": 1101
+    },
+    {
+      "epoch": 0.11589782691574534,
+      "grad_norm": 1.7911673644478512,
+      "learning_rate": 4.861357928762468e-06,
+      "loss": 1.0355,
+      "step": 1102
+    },
+    {
+      "epoch": 0.11600299735759265,
+      "grad_norm": 3.1971189200731316,
+      "learning_rate": 4.8610836530263485e-06,
+      "loss": 1.0348,
+      "step": 1103
+    },
+    {
+      "epoch": 0.11610816779943997,
+      "grad_norm": 3.9670686015745438,
+      "learning_rate": 4.860809114011059e-06,
+      "loss": 1.0403,
+      "step": 1104
+    },
+    {
+      "epoch": 0.11621333824128728,
+      "grad_norm": 3.4009369343225426,
+      "learning_rate": 4.860534311747215e-06,
+      "loss": 1.0572,
+      "step": 1105
+    },
+    {
+      "epoch": 0.1163185086831346,
+      "grad_norm": 4.1414177645044745,
+      "learning_rate": 4.860259246265456e-06,
+      "loss": 1.0702,
+      "step": 1106
+    },
+    {
+      "epoch": 0.11642367912498193,
+      "grad_norm": 4.123107172342563,
+      "learning_rate": 4.859983917596458e-06,
+      "loss": 1.0605,
+      "step": 1107
+    },
+    {
+      "epoch": 0.11652884956682924,
+      "grad_norm": 3.969640922287132,
+      "learning_rate": 4.859708325770919e-06,
+      "loss": 1.008,
+      "step": 1108
+    },
+    {
+      "epoch": 0.11663402000867656,
+      "grad_norm": 2.90888044600641,
+      "learning_rate": 4.859432470819572e-06,
+      "loss": 1.0459,
+      "step": 1109
+    },
+    {
+      "epoch": 0.11673919045052387,
+      "grad_norm": 2.831787199338218,
+      "learning_rate": 4.859156352773174e-06,
+      "loss": 1.0261,
+      "step": 1110
+    },
+    {
+      "epoch": 0.1168443608923712,
+      "grad_norm": 3.987477080469216,
+      "learning_rate": 4.858879971662518e-06,
+      "loss": 1.0415,
+      "step": 1111
+    },
+    {
+      "epoch": 0.11694953133421852,
+      "grad_norm": 3.9586675261870274,
+      "learning_rate": 4.85860332751842e-06,
+      "loss": 1.0675,
+      "step": 1112
+    },
+    {
+      "epoch": 0.11705470177606583,
+      "grad_norm": 2.6528061023001386,
+      "learning_rate": 4.858326420371728e-06,
+      "loss": 1.0326,
+      "step": 1113
+    },
+    {
+      "epoch": 0.11715987221791316,
+      "grad_norm": 3.2831047642667297,
+      "learning_rate": 4.858049250253321e-06,
+      "loss": 1.0562,
+      "step": 1114
+    },
+    {
+      "epoch": 0.11726504265976047,
+      "grad_norm": 3.144259175603822,
+      "learning_rate": 4.8577718171941036e-06,
+      "loss": 1.0119,
+      "step": 1115
+    },
+    {
+      "epoch": 0.11737021310160779,
+      "grad_norm": 3.6872799055696244,
+      "learning_rate": 4.857494121225014e-06,
+      "loss": 1.0456,
+      "step": 1116
+    },
+    {
+      "epoch": 0.11747538354345512,
+      "grad_norm": 3.1976523712311273,
+      "learning_rate": 4.857216162377015e-06,
+      "loss": 1.0582,
+      "step": 1117
+    },
+    {
+      "epoch": 0.11758055398530243,
+      "grad_norm": 2.9008275495280342,
+      "learning_rate": 4.8569379406811034e-06,
+      "loss": 1.0335,
+      "step": 1118
+    },
+    {
+      "epoch": 0.11768572442714975,
+      "grad_norm": 3.5671660679987265,
+      "learning_rate": 4.856659456168301e-06,
+      "loss": 1.0251,
+      "step": 1119
+    },
+    {
+      "epoch": 0.11779089486899706,
+      "grad_norm": 3.9412822209537426,
+      "learning_rate": 4.856380708869663e-06,
+      "loss": 1.0671,
+      "step": 1120
+    },
+    {
+      "epoch": 0.11789606531084439,
+      "grad_norm": 3.191017359982901,
+      "learning_rate": 4.85610169881627e-06,
+      "loss": 1.0434,
+      "step": 1121
+    },
+    {
+      "epoch": 0.11800123575269171,
+      "grad_norm": 3.195874922814993,
+      "learning_rate": 4.855822426039236e-06,
+      "loss": 0.9924,
+      "step": 1122
+    },
+    {
+      "epoch": 0.11810640619453902,
+      "grad_norm": 2.751515168083024,
+      "learning_rate": 4.855542890569701e-06,
+      "loss": 1.0373,
+      "step": 1123
+    },
+    {
+      "epoch": 0.11821157663638635,
+      "grad_norm": 2.896579359878487,
+      "learning_rate": 4.855263092438834e-06,
+      "loss": 1.0463,
+      "step": 1124
+    },
+    {
+      "epoch": 0.11831674707823366,
+      "grad_norm": 3.367419965470842,
+      "learning_rate": 4.8549830316778365e-06,
+      "loss": 1.0156,
+      "step": 1125
+    },
+    {
+      "epoch": 0.11842191752008098,
+      "grad_norm": 2.2007559598381357,
+      "learning_rate": 4.854702708317937e-06,
+      "loss": 1.0342,
+      "step": 1126
+    },
+    {
+      "epoch": 0.1185270879619283,
+      "grad_norm": 2.729565964737428,
+      "learning_rate": 4.8544221223903925e-06,
+      "loss": 1.0192,
+      "step": 1127
+    },
+    {
+      "epoch": 0.11863225840377561,
+      "grad_norm": 4.234177007738469,
+      "learning_rate": 4.854141273926492e-06,
+      "loss": 0.9931,
+      "step": 1128
+    },
+    {
+      "epoch": 0.11873742884562294,
+      "grad_norm": 4.1159850072468025,
+      "learning_rate": 4.8538601629575525e-06,
+      "loss": 1.048,
+      "step": 1129
+    },
+    {
+      "epoch": 0.11884259928747026,
+      "grad_norm": 3.030352704362315,
+      "learning_rate": 4.8535787895149186e-06,
+      "loss": 1.0481,
+      "step": 1130
+    },
+    {
+      "epoch": 0.11894776972931757,
+      "grad_norm": 2.416745852348571,
+      "learning_rate": 4.853297153629967e-06,
+      "loss": 1.0426,
+      "step": 1131
+    },
+    {
+      "epoch": 0.1190529401711649,
+      "grad_norm": 2.953961483423199,
+      "learning_rate": 4.853015255334101e-06,
+      "loss": 1.0233,
+      "step": 1132
+    },
+    {
+      "epoch": 0.11915811061301221,
+      "grad_norm": 3.1479898039953103,
+      "learning_rate": 4.852733094658756e-06,
+      "loss": 1.0074,
+      "step": 1133
+    },
+    {
+      "epoch": 0.11926328105485953,
+      "grad_norm": 2.0774548323641397,
+      "learning_rate": 4.852450671635395e-06,
+      "loss": 1.01,
+      "step": 1134
+    },
+    {
+      "epoch": 0.11936845149670686,
+      "grad_norm": 2.51715458277697,
+      "learning_rate": 4.852167986295508e-06,
+      "loss": 1.0401,
+      "step": 1135
+    },
+    {
+      "epoch": 0.11947362193855417,
+      "grad_norm": 3.6879308333975587,
+      "learning_rate": 4.851885038670618e-06,
+      "loss": 1.0577,
+      "step": 1136
+    },
+    {
+      "epoch": 0.11957879238040149,
+      "grad_norm": 3.27065440253345,
+      "learning_rate": 4.851601828792278e-06,
+      "loss": 1.0163,
+      "step": 1137
+    },
+    {
+      "epoch": 0.1196839628222488,
+      "grad_norm": 2.2130774677775853,
+      "learning_rate": 4.8513183566920654e-06,
+      "loss": 1.0543,
+      "step": 1138
+    },
+    {
+      "epoch": 0.11978913326409613,
+      "grad_norm": 2.8314329893329093,
+      "learning_rate": 4.8510346224015896e-06,
+      "loss": 1.0796,
+      "step": 1139
+    },
+    {
+      "epoch": 0.11989430370594345,
+      "grad_norm": 3.7243358852096953,
+      "learning_rate": 4.85075062595249e-06,
+      "loss": 1.0379,
+      "step": 1140
+    },
+    {
+      "epoch": 0.11999947414779076,
+      "grad_norm": 3.7144900747943623,
+      "learning_rate": 4.850466367376435e-06,
+      "loss": 1.0259,
+      "step": 1141
+    },
+    {
+      "epoch": 0.12010464458963808,
+      "grad_norm": 3.231616766464792,
+      "learning_rate": 4.850181846705121e-06,
+      "loss": 1.0482,
+      "step": 1142
+    },
+    {
+      "epoch": 0.1202098150314854,
+      "grad_norm": 3.3082034220519416,
+      "learning_rate": 4.8498970639702745e-06,
+      "loss": 1.0284,
+      "step": 1143
+    },
+    {
+      "epoch": 0.12031498547333272,
+      "grad_norm": 2.9443493452344183,
+      "learning_rate": 4.84961201920365e-06,
+      "loss": 1.0511,
+      "step": 1144
+    },
+    {
+      "epoch": 0.12042015591518004,
+      "grad_norm": 3.196336464928972,
+      "learning_rate": 4.849326712437033e-06,
+      "loss": 1.0723,
+      "step": 1145
+    },
+    {
+      "epoch": 0.12052532635702735,
+      "grad_norm": 2.71928604807591,
+      "learning_rate": 4.849041143702238e-06,
+      "loss": 1.0112,
+      "step": 1146
+    },
+    {
+      "epoch": 0.12063049679887468,
+      "grad_norm": 3.533369997436239,
+      "learning_rate": 4.8487553130311065e-06,
+      "loss": 0.9867,
+      "step": 1147
+    },
+    {
+      "epoch": 0.12073566724072199,
+      "grad_norm": 2.455468182286396,
+      "learning_rate": 4.848469220455512e-06,
+      "loss": 0.9962,
+      "step": 1148
+    },
+    {
+      "epoch": 0.12084083768256931,
+      "grad_norm": 2.3850247504745465,
+      "learning_rate": 4.848182866007356e-06,
+      "loss": 1.0686,
+      "step": 1149
+    },
+    {
+      "epoch": 0.12094600812441664,
+      "grad_norm": 2.5387686707247843,
+      "learning_rate": 4.84789624971857e-06,
+      "loss": 1.0425,
+      "step": 1150
+    },
+    {
+      "epoch": 0.12105117856626395,
+      "grad_norm": 3.482360328905752,
+      "learning_rate": 4.8476093716211125e-06,
+      "loss": 1.0271,
+      "step": 1151
+    },
+    {
+      "epoch": 0.12115634900811127,
+      "grad_norm": 3.6287895041841702,
+      "learning_rate": 4.847322231746973e-06,
+      "loss": 1.0491,
+      "step": 1152
+    },
+    {
+      "epoch": 0.12126151944995858,
+      "grad_norm": 2.7406507151068404,
+      "learning_rate": 4.84703483012817e-06,
+      "loss": 1.0579,
+      "step": 1153
+    },
+    {
+      "epoch": 0.1213666898918059,
+      "grad_norm": 2.093402317749306,
+      "learning_rate": 4.846747166796751e-06,
+      "loss": 1.0114,
+      "step": 1154
+    },
+    {
+      "epoch": 0.12147186033365323,
+      "grad_norm": 2.7833406137397896,
+      "learning_rate": 4.846459241784793e-06,
+      "loss": 1.0164,
+      "step": 1155
+    },
+    {
+      "epoch": 0.12157703077550054,
+      "grad_norm": 3.5971545223060626,
+      "learning_rate": 4.846171055124401e-06,
+      "loss": 1.0388,
+      "step": 1156
+    },
+    {
+      "epoch": 0.12168220121734787,
+      "grad_norm": 4.388318975491837,
+      "learning_rate": 4.845882606847712e-06,
+      "loss": 1.0544,
+      "step": 1157
+    },
+    {
+      "epoch": 0.12178737165919519,
+      "grad_norm": 3.471550657627086,
+      "learning_rate": 4.845593896986888e-06,
+      "loss": 1.0936,
+      "step": 1158
+    },
+    {
+      "epoch": 0.1218925421010425,
+      "grad_norm": 3.0123013262369924,
+      "learning_rate": 4.845304925574122e-06,
+      "loss": 0.9964,
+      "step": 1159
+    },
+    {
+      "epoch": 0.12199771254288982,
+      "grad_norm": 2.2624703252236453,
+      "learning_rate": 4.84501569264164e-06,
+      "loss": 1.0276,
+      "step": 1160
+    },
+    {
+      "epoch": 0.12210288298473713,
+      "grad_norm": 2.675714197722898,
+      "learning_rate": 4.8447261982216905e-06,
+      "loss": 0.9954,
+      "step": 1161
+    },
+    {
+      "epoch": 0.12220805342658446,
+      "grad_norm": 4.13088015952959,
+      "learning_rate": 4.8444364423465555e-06,
+      "loss": 0.9864,
+      "step": 1162
+    },
+    {
+      "epoch": 0.12231322386843178,
+      "grad_norm": 4.229159717585945,
+      "learning_rate": 4.844146425048545e-06,
+      "loss": 1.0392,
+      "step": 1163
+    },
+    {
+      "epoch": 0.1224183943102791,
+      "grad_norm": 3.419828767103554,
+      "learning_rate": 4.843856146359999e-06,
+      "loss": 1.0395,
+      "step": 1164
+    },
+    {
+      "epoch": 0.12252356475212642,
+      "grad_norm": 4.317347247396286,
+      "learning_rate": 4.843565606313283e-06,
+      "loss": 1.0242,
+      "step": 1165
+    },
+    {
+      "epoch": 0.12262873519397373,
+      "grad_norm": 3.754270500912085,
+      "learning_rate": 4.843274804940798e-06,
+      "loss": 1.052,
+      "step": 1166
+    },
+    {
+      "epoch": 0.12273390563582105,
+      "grad_norm": 2.8312910701769294,
+      "learning_rate": 4.8429837422749695e-06,
+      "loss": 1.0404,
+      "step": 1167
+    },
+    {
+      "epoch": 0.12283907607766838,
+      "grad_norm": 3.823314861494254,
+      "learning_rate": 4.842692418348251e-06,
+      "loss": 1.0094,
+      "step": 1168
+    },
+    {
+      "epoch": 0.12294424651951569,
+      "grad_norm": 3.9925424351563095,
+      "learning_rate": 4.842400833193131e-06,
+      "loss": 1.0336,
+      "step": 1169
+    },
+    {
+      "epoch": 0.12304941696136301,
+      "grad_norm": 4.5156463343872755,
+      "learning_rate": 4.84210898684212e-06,
+      "loss": 1.0294,
+      "step": 1170
+    },
+    {
+      "epoch": 0.12315458740321032,
+      "grad_norm": 4.134735379560587,
+      "learning_rate": 4.841816879327764e-06,
+      "loss": 1.0681,
+      "step": 1171
+    },
+    {
+      "epoch": 0.12325975784505765,
+      "grad_norm": 3.5718684762008586,
+      "learning_rate": 4.8415245106826335e-06,
+      "loss": 1.0457,
+      "step": 1172
+    },
+    {
+      "epoch": 0.12336492828690497,
+      "grad_norm": 2.8524025783294835,
+      "learning_rate": 4.841231880939331e-06,
+      "loss": 1.051,
+      "step": 1173
+    },
+    {
+      "epoch": 0.12347009872875228,
+      "grad_norm": 2.1112242358528794,
+      "learning_rate": 4.840938990130486e-06,
+      "loss": 0.9749,
+      "step": 1174
+    },
+    {
+      "epoch": 0.1235752691705996,
+      "grad_norm": 3.489044644429071,
+      "learning_rate": 4.840645838288759e-06,
+      "loss": 1.0839,
+      "step": 1175
+    },
+    {
+      "epoch": 0.12368043961244692,
+      "grad_norm": 3.7851298708992007,
+      "learning_rate": 4.840352425446838e-06,
+      "loss": 1.0008,
+      "step": 1176
+    },
+    {
+      "epoch": 0.12378561005429424,
+      "grad_norm": 3.534928614423682,
+      "learning_rate": 4.840058751637441e-06,
+      "loss": 0.9969,
+      "step": 1177
+    },
+    {
+      "epoch": 0.12389078049614156,
+      "grad_norm": 3.866088509887153,
+      "learning_rate": 4.839764816893315e-06,
+      "loss": 1.0296,
+      "step": 1178
+    },
+    {
+      "epoch": 0.12399595093798887,
+      "grad_norm": 3.8493227086097836,
+      "learning_rate": 4.839470621247235e-06,
+      "loss": 0.9829,
+      "step": 1179
+    },
+    {
+      "epoch": 0.1241011213798362,
+      "grad_norm": 3.863296748292079,
+      "learning_rate": 4.839176164732009e-06,
+      "loss": 1.0043,
+      "step": 1180
+    },
+    {
+      "epoch": 0.12420629182168351,
+      "grad_norm": 3.044556390384978,
+      "learning_rate": 4.838881447380468e-06,
+      "loss": 1.0635,
+      "step": 1181
+    },
+    {
+      "epoch": 0.12431146226353083,
+      "grad_norm": 3.036859307808052,
+      "learning_rate": 4.838586469225477e-06,
+      "loss": 1.068,
+      "step": 1182
+    },
+    {
+      "epoch": 0.12441663270537816,
+      "grad_norm": 3.4852695563768936,
+      "learning_rate": 4.838291230299927e-06,
+      "loss": 1.0264,
+      "step": 1183
+    },
+    {
+      "epoch": 0.12452180314722547,
+      "grad_norm": 3.541328436650173,
+      "learning_rate": 4.837995730636742e-06,
+      "loss": 1.0692,
+      "step": 1184
+    },
+    {
+      "epoch": 0.12462697358907279,
+      "grad_norm": 2.403850262061819,
+      "learning_rate": 4.83769997026887e-06,
+      "loss": 1.0131,
+      "step": 1185
+    },
+    {
+      "epoch": 0.12473214403092012,
+      "grad_norm": 2.0179723281866204,
+      "learning_rate": 4.837403949229291e-06,
+      "loss": 1.033,
+      "step": 1186
+    },
+    {
+      "epoch": 0.12483731447276743,
+      "grad_norm": 3.288715529975734,
+      "learning_rate": 4.837107667551015e-06,
+      "loss": 1.0143,
+      "step": 1187
+    },
+    {
+      "epoch": 0.12494248491461475,
+      "grad_norm": 2.488652116770584,
+      "learning_rate": 4.8368111252670776e-06,
+      "loss": 1.0064,
+      "step": 1188
+    },
+    {
+      "epoch": 0.12504765535646206,
+      "grad_norm": 2.099768655554014,
+      "learning_rate": 4.836514322410548e-06,
+      "loss": 1.0311,
+      "step": 1189
+    },
+    {
+      "epoch": 0.1251528257983094,
+      "grad_norm": 2.61507115515271,
+      "learning_rate": 4.83621725901452e-06,
+      "loss": 1.0037,
+      "step": 1190
+    },
+    {
+      "epoch": 0.1252579962401567,
+      "grad_norm": 3.5627001839530887,
+      "learning_rate": 4.83591993511212e-06,
+      "loss": 1.0104,
+      "step": 1191
+    },
+    {
+      "epoch": 0.12536316668200403,
+      "grad_norm": 2.7611171047918854,
+      "learning_rate": 4.835622350736499e-06,
+      "loss": 1.0015,
+      "step": 1192
+    },
+    {
+      "epoch": 0.12546833712385133,
+      "grad_norm": 2.6167789911665333,
+      "learning_rate": 4.835324505920845e-06,
+      "loss": 1.0208,
+      "step": 1193
+    },
+    {
+      "epoch": 0.12557350756569866,
+      "grad_norm": 3.3090032658379624,
+      "learning_rate": 4.835026400698366e-06,
+      "loss": 1.0339,
+      "step": 1194
+    },
+    {
+      "epoch": 0.12567867800754598,
+      "grad_norm": 2.7592196627213172,
+      "learning_rate": 4.8347280351023044e-06,
+      "loss": 1.021,
+      "step": 1195
+    },
+    {
+      "epoch": 0.1257838484493933,
+      "grad_norm": 3.529997579629742,
+      "learning_rate": 4.83442940916593e-06,
+      "loss": 1.0269,
+      "step": 1196
+    },
+    {
+      "epoch": 0.12588901889124063,
+      "grad_norm": 3.0771754582486324,
+      "learning_rate": 4.834130522922541e-06,
+      "loss": 1.0385,
+      "step": 1197
+    },
+    {
+      "epoch": 0.12599418933308792,
+      "grad_norm": 3.019098612661189,
+      "learning_rate": 4.8338313764054676e-06,
+      "loss": 1.0067,
+      "step": 1198
+    },
+    {
+      "epoch": 0.12609935977493525,
+      "grad_norm": 3.1455190399414774,
+      "learning_rate": 4.8335319696480655e-06,
+      "loss": 0.9901,
+      "step": 1199
+    },
+    {
+      "epoch": 0.12620453021678257,
+      "grad_norm": 2.815592993177342,
+      "learning_rate": 4.833232302683721e-06,
+      "loss": 1.0099,
+      "step": 1200
+    },
+    {
+      "epoch": 0.1263097006586299,
+      "grad_norm": 2.1662241166960974,
+      "learning_rate": 4.83293237554585e-06,
+      "loss": 1.0209,
+      "step": 1201
+    },
+    {
+      "epoch": 0.12641487110047722,
+      "grad_norm": 2.5510337259609717,
+      "learning_rate": 4.832632188267896e-06,
+      "loss": 1.0417,
+      "step": 1202
+    },
+    {
+      "epoch": 0.12652004154232452,
+      "grad_norm": 2.7666819520333146,
+      "learning_rate": 4.832331740883333e-06,
+      "loss": 1.0352,
+      "step": 1203
+    },
+    {
+      "epoch": 0.12662521198417184,
+      "grad_norm": 3.0449829988980937,
+      "learning_rate": 4.832031033425663e-06,
+      "loss": 1.0291,
+      "step": 1204
+    },
+    {
+      "epoch": 0.12673038242601917,
+      "grad_norm": 2.6400119751853564,
+      "learning_rate": 4.831730065928416e-06,
+      "loss": 1.0378,
+      "step": 1205
+    },
+    {
+      "epoch": 0.1268355528678665,
+      "grad_norm": 2.94538447937034,
+      "learning_rate": 4.831428838425153e-06,
+      "loss": 1.0311,
+      "step": 1206
+    },
+    {
+      "epoch": 0.12694072330971382,
+      "grad_norm": 2.3594393856071614,
+      "learning_rate": 4.8311273509494635e-06,
+      "loss": 1.0124,
+      "step": 1207
+    },
+    {
+      "epoch": 0.1270458937515611,
+      "grad_norm": 2.153915292661592,
+      "learning_rate": 4.830825603534967e-06,
+      "loss": 1.0277,
+      "step": 1208
+    },
+    {
+      "epoch": 0.12715106419340844,
+      "grad_norm": 3.456220143305514,
+      "learning_rate": 4.8305235962153075e-06,
+      "loss": 1.0868,
+      "step": 1209
+    },
+    {
+      "epoch": 0.12725623463525576,
+      "grad_norm": 2.7043626503920337,
+      "learning_rate": 4.830221329024163e-06,
+      "loss": 1.0537,
+      "step": 1210
+    },
+    {
+      "epoch": 0.12736140507710308,
+      "grad_norm": 3.589165033013744,
+      "learning_rate": 4.829918801995239e-06,
+      "loss": 1.0796,
+      "step": 1211
+    },
+    {
+      "epoch": 0.1274665755189504,
+      "grad_norm": 3.373260416606364,
+      "learning_rate": 4.829616015162269e-06,
+      "loss": 1.0191,
+      "step": 1212
+    },
+    {
+      "epoch": 0.1275717459607977,
+      "grad_norm": 3.5871990636482796,
+      "learning_rate": 4.8293129685590164e-06,
+      "loss": 1.0407,
+      "step": 1213
+    },
+    {
+      "epoch": 0.12767691640264503,
+      "grad_norm": 4.2127544793207745,
+      "learning_rate": 4.8290096622192735e-06,
+      "loss": 1.0372,
+      "step": 1214
+    },
+    {
+      "epoch": 0.12778208684449235,
+      "grad_norm": 3.511106624322245,
+      "learning_rate": 4.82870609617686e-06,
+      "loss": 1.0408,
+      "step": 1215
+    },
+    {
+      "epoch": 0.12788725728633968,
+      "grad_norm": 2.8132894727294055,
+      "learning_rate": 4.828402270465628e-06,
+      "loss": 0.9943,
+      "step": 1216
+    },
+    {
+      "epoch": 0.127992427728187,
+      "grad_norm": 2.4459083353657807,
+      "learning_rate": 4.828098185119454e-06,
+      "loss": 1.0259,
+      "step": 1217
+    },
+    {
+      "epoch": 0.1280975981700343,
+      "grad_norm": 2.3329129365329337,
+      "learning_rate": 4.827793840172247e-06,
+      "loss": 1.0269,
+      "step": 1218
+    },
+    {
+      "epoch": 0.12820276861188162,
+      "grad_norm": 2.8128127508148015,
+      "learning_rate": 4.827489235657944e-06,
+      "loss": 1.0505,
+      "step": 1219
+    },
+    {
+      "epoch": 0.12830793905372895,
+      "grad_norm": 2.339660968358773,
+      "learning_rate": 4.827184371610511e-06,
+      "loss": 1.0047,
+      "step": 1220
+    },
+    {
+      "epoch": 0.12841310949557627,
+      "grad_norm": 3.3186117188660584,
+      "learning_rate": 4.826879248063943e-06,
+      "loss": 1.0416,
+      "step": 1221
+    },
+    {
+      "epoch": 0.1285182799374236,
+      "grad_norm": 3.0496345780537077,
+      "learning_rate": 4.826573865052261e-06,
+      "loss": 1.0285,
+      "step": 1222
+    },
+    {
+      "epoch": 0.1286234503792709,
+      "grad_norm": 3.0696030887083134,
+      "learning_rate": 4.8262682226095215e-06,
+      "loss": 1.0221,
+      "step": 1223
+    },
+    {
+      "epoch": 0.12872862082111822,
+      "grad_norm": 2.911026501529926,
+      "learning_rate": 4.825962320769804e-06,
+      "loss": 0.996,
+      "step": 1224
+    },
+    {
+      "epoch": 0.12883379126296554,
+      "grad_norm": 2.6415248186701414,
+      "learning_rate": 4.825656159567218e-06,
+      "loss": 1.0077,
+      "step": 1225
+    },
+    {
+      "epoch": 0.12893896170481287,
+      "grad_norm": 3.015861478144796,
+      "learning_rate": 4.8253497390359035e-06,
+      "loss": 1.0199,
+      "step": 1226
+    },
+    {
+      "epoch": 0.1290441321466602,
+      "grad_norm": 2.968448842373254,
+      "learning_rate": 4.82504305921003e-06,
+      "loss": 1.0586,
+      "step": 1227
+    },
+    {
+      "epoch": 0.1291493025885075,
+      "grad_norm": 3.6356255518516805,
+      "learning_rate": 4.824736120123794e-06,
+      "loss": 1.0877,
+      "step": 1228
+    },
+    {
+      "epoch": 0.1292544730303548,
+      "grad_norm": 2.2178443735382616,
+      "learning_rate": 4.824428921811421e-06,
+      "loss": 1.0224,
+      "step": 1229
+    },
+    {
+      "epoch": 0.12935964347220213,
+      "grad_norm": 2.5367079837422497,
+      "learning_rate": 4.824121464307168e-06,
+      "loss": 1.0118,
+      "step": 1230
+    },
+    {
+      "epoch": 0.12946481391404946,
+      "grad_norm": 3.130091176692925,
+      "learning_rate": 4.823813747645315e-06,
+      "loss": 1.0146,
+      "step": 1231
+    },
+    {
+      "epoch": 0.12956998435589678,
+      "grad_norm": 2.4669259065457902,
+      "learning_rate": 4.823505771860178e-06,
+      "loss": 1.0262,
+      "step": 1232
+    },
+    {
+      "epoch": 0.1296751547977441,
+      "grad_norm": 2.3946588572460015,
+      "learning_rate": 4.823197536986098e-06,
+      "loss": 1.0202,
+      "step": 1233
+    },
+    {
+      "epoch": 0.1297803252395914,
+      "grad_norm": 3.28718019202575,
+      "learning_rate": 4.822889043057446e-06,
+      "loss": 0.996,
+      "step": 1234
+    },
+    {
+      "epoch": 0.12988549568143873,
+      "grad_norm": 3.172840137269599,
+      "learning_rate": 4.82258029010862e-06,
+      "loss": 0.9892,
+      "step": 1235
+    },
+    {
+      "epoch": 0.12999066612328605,
+      "grad_norm": 3.469596633957251,
+      "learning_rate": 4.82227127817405e-06,
+      "loss": 1.0587,
+      "step": 1236
+    },
+    {
+      "epoch": 0.13009583656513338,
+      "grad_norm": 2.555342312316807,
+      "learning_rate": 4.821962007288191e-06,
+      "loss": 1.0267,
+      "step": 1237
+    },
+    {
+      "epoch": 0.1302010070069807,
+      "grad_norm": 2.5256812029218128,
+      "learning_rate": 4.821652477485531e-06,
+      "loss": 1.0289,
+      "step": 1238
+    },
+    {
+      "epoch": 0.130306177448828,
+      "grad_norm": 3.1983565580222875,
+      "learning_rate": 4.821342688800586e-06,
+      "loss": 1.0321,
+      "step": 1239
+    },
+    {
+      "epoch": 0.13041134789067532,
+      "grad_norm": 3.620474013802864,
+      "learning_rate": 4.821032641267897e-06,
+      "loss": 0.9785,
+      "step": 1240
+    },
+    {
+      "epoch": 0.13051651833252265,
+      "grad_norm": 2.661892927357515,
+      "learning_rate": 4.820722334922039e-06,
+      "loss": 1.039,
+      "step": 1241
+    },
+    {
+      "epoch": 0.13062168877436997,
+      "grad_norm": 3.34786582341296,
+      "learning_rate": 4.820411769797611e-06,
+      "loss": 1.0484,
+      "step": 1242
+    },
+    {
+      "epoch": 0.1307268592162173,
+      "grad_norm": 3.9689276628030985,
+      "learning_rate": 4.820100945929247e-06,
+      "loss": 1.0423,
+      "step": 1243
+    },
+    {
+      "epoch": 0.1308320296580646,
+      "grad_norm": 2.883967378632545,
+      "learning_rate": 4.819789863351603e-06,
+      "loss": 1.0503,
+      "step": 1244
+    },
+    {
+      "epoch": 0.13093720009991192,
+      "grad_norm": 3.105797028005016,
+      "learning_rate": 4.819478522099369e-06,
+      "loss": 1.0227,
+      "step": 1245
+    },
+    {
+      "epoch": 0.13104237054175924,
+      "grad_norm": 1.6808339529344005,
+      "learning_rate": 4.819166922207261e-06,
+      "loss": 1.0181,
+      "step": 1246
+    },
+    {
+      "epoch": 0.13114754098360656,
+      "grad_norm": 2.7847369838128326,
+      "learning_rate": 4.818855063710026e-06,
+      "loss": 1.0094,
+      "step": 1247
+    },
+    {
+      "epoch": 0.1312527114254539,
+      "grad_norm": 2.8619801152801365,
+      "learning_rate": 4.8185429466424375e-06,
+      "loss": 1.0377,
+      "step": 1248
+    },
+    {
+      "epoch": 0.13135788186730118,
+      "grad_norm": 3.00420826947291,
+      "learning_rate": 4.8182305710393e-06,
+      "loss": 1.0415,
+      "step": 1249
+    },
+    {
+      "epoch": 0.1314630523091485,
+      "grad_norm": 3.106603153157244,
+      "learning_rate": 4.8179179369354445e-06,
+      "loss": 1.0087,
+      "step": 1250
+    },
+    {
+      "epoch": 0.13156822275099583,
+      "grad_norm": 2.9567174545180936,
+      "learning_rate": 4.817605044365733e-06,
+      "loss": 1.0042,
+      "step": 1251
+    },
+    {
+      "epoch": 0.13167339319284316,
+      "grad_norm": 3.6684630626815578,
+      "learning_rate": 4.817291893365055e-06,
+      "loss": 1.0208,
+      "step": 1252
+    },
+    {
+      "epoch": 0.13177856363469048,
+      "grad_norm": 3.8333923409888406,
+      "learning_rate": 4.8169784839683295e-06,
+      "loss": 1.032,
+      "step": 1253
+    },
+    {
+      "epoch": 0.13188373407653778,
+      "grad_norm": 3.076774148237722,
+      "learning_rate": 4.816664816210505e-06,
+      "loss": 1.0345,
+      "step": 1254
+    },
+    {
+      "epoch": 0.1319889045183851,
+      "grad_norm": 2.528183859455226,
+      "learning_rate": 4.816350890126556e-06,
+      "loss": 1.0463,
+      "step": 1255
+    },
+    {
+      "epoch": 0.13209407496023243,
+      "grad_norm": 3.2080429059373765,
+      "learning_rate": 4.816036705751489e-06,
+      "loss": 1.0083,
+      "step": 1256
+    },
+    {
+      "epoch": 0.13219924540207975,
+      "grad_norm": 2.996843724364759,
+      "learning_rate": 4.8157222631203374e-06,
+      "loss": 0.9658,
+      "step": 1257
+    },
+    {
+      "epoch": 0.13230441584392708,
+      "grad_norm": 2.823099339288846,
+      "learning_rate": 4.815407562268165e-06,
+      "loss": 1.0478,
+      "step": 1258
+    },
+    {
+      "epoch": 0.13240958628577437,
+      "grad_norm": 2.5147629342097635,
+      "learning_rate": 4.815092603230063e-06,
+      "loss": 0.9942,
+      "step": 1259
+    },
+    {
+      "epoch": 0.1325147567276217,
+      "grad_norm": 3.1482493227003316,
+      "learning_rate": 4.81477738604115e-06,
+      "loss": 0.9881,
+      "step": 1260
+    },
+    {
+      "epoch": 0.13261992716946902,
+      "grad_norm": 3.020760053243842,
+      "learning_rate": 4.814461910736578e-06,
+      "loss": 1.0364,
+      "step": 1261
+    },
+    {
+      "epoch": 0.13272509761131634,
+      "grad_norm": 2.221968445422404,
+      "learning_rate": 4.814146177351523e-06,
+      "loss": 1.0265,
+      "step": 1262
+    },
+    {
+      "epoch": 0.13283026805316367,
+      "grad_norm": 3.0176270858205565,
+      "learning_rate": 4.8138301859211925e-06,
+      "loss": 1.0519,
+      "step": 1263
+    },
+    {
+      "epoch": 0.13293543849501097,
+      "grad_norm": 2.2986812981864317,
+      "learning_rate": 4.813513936480821e-06,
+      "loss": 1.03,
+      "step": 1264
+    },
+    {
+      "epoch": 0.1330406089368583,
+      "grad_norm": 2.8952298044973057,
+      "learning_rate": 4.8131974290656745e-06,
+      "loss": 1.0342,
+      "step": 1265
+    },
+    {
+      "epoch": 0.13314577937870561,
+      "grad_norm": 2.8246589734902225,
+      "learning_rate": 4.812880663711045e-06,
+      "loss": 1.0491,
+      "step": 1266
+    },
+    {
+      "epoch": 0.13325094982055294,
+      "grad_norm": 2.5406616352440134,
+      "learning_rate": 4.812563640452254e-06,
+      "loss": 1.0069,
+      "step": 1267
+    },
+    {
+      "epoch": 0.13335612026240026,
+      "grad_norm": 2.3100739684989167,
+      "learning_rate": 4.812246359324653e-06,
+      "loss": 1.002,
+      "step": 1268
+    },
+    {
+      "epoch": 0.13346129070424756,
+      "grad_norm": 2.9322165786317997,
+      "learning_rate": 4.811928820363621e-06,
+      "loss": 1.0207,
+      "step": 1269
+    },
+    {
+      "epoch": 0.13356646114609488,
+      "grad_norm": 2.3628042524685533,
+      "learning_rate": 4.811611023604564e-06,
+      "loss": 1.0314,
+      "step": 1270
+    },
+    {
+      "epoch": 0.1336716315879422,
+      "grad_norm": 3.3682424272817895,
+      "learning_rate": 4.811292969082922e-06,
+      "loss": 0.9997,
+      "step": 1271
+    },
+    {
+      "epoch": 0.13377680202978953,
+      "grad_norm": 2.796544714713144,
+      "learning_rate": 4.8109746568341595e-06,
+      "loss": 1.0163,
+      "step": 1272
+    },
+    {
+      "epoch": 0.13388197247163686,
+      "grad_norm": 3.2927205255203997,
+      "learning_rate": 4.81065608689377e-06,
+      "loss": 1.0652,
+      "step": 1273
+    },
+    {
+      "epoch": 0.13398714291348415,
+      "grad_norm": 4.2986772325983935,
+      "learning_rate": 4.810337259297277e-06,
+      "loss": 1.0425,
+      "step": 1274
+    },
+    {
+      "epoch": 0.13409231335533148,
+      "grad_norm": 2.5363498831942675,
+      "learning_rate": 4.810018174080233e-06,
+      "loss": 1.0201,
+      "step": 1275
+    },
+    {
+      "epoch": 0.1341974837971788,
+      "grad_norm": 3.5836252312590973,
+      "learning_rate": 4.809698831278217e-06,
+      "loss": 1.0456,
+      "step": 1276
+    },
+    {
+      "epoch": 0.13430265423902613,
+      "grad_norm": 3.425235368215492,
+      "learning_rate": 4.80937923092684e-06,
+      "loss": 1.0208,
+      "step": 1277
+    },
+    {
+      "epoch": 0.13440782468087345,
+      "grad_norm": 1.9745893346472898,
+      "learning_rate": 4.809059373061737e-06,
+      "loss": 0.9924,
+      "step": 1278
+    },
+    {
+      "epoch": 0.13451299512272075,
+      "grad_norm": 3.0646315393918835,
+      "learning_rate": 4.808739257718578e-06,
+      "loss": 0.9939,
+      "step": 1279
+    },
+    {
+      "epoch": 0.13461816556456807,
+      "grad_norm": 2.479949101456566,
+      "learning_rate": 4.808418884933056e-06,
+      "loss": 1.0327,
+      "step": 1280
+    },
+    {
+      "epoch": 0.1347233360064154,
+      "grad_norm": 2.922429890182779,
+      "learning_rate": 4.808098254740897e-06,
+      "loss": 1.0177,
+      "step": 1281
+    },
+    {
+      "epoch": 0.13482850644826272,
+      "grad_norm": 1.924186395677562,
+      "learning_rate": 4.807777367177851e-06,
+      "loss": 0.9983,
+      "step": 1282
+    },
+    {
+      "epoch": 0.13493367689011004,
+      "grad_norm": 3.935674804500333,
+      "learning_rate": 4.807456222279703e-06,
+      "loss": 1.0269,
+      "step": 1283
+    },
+    {
+      "epoch": 0.13503884733195734,
+      "grad_norm": 2.938713767635199,
+      "learning_rate": 4.80713482008226e-06,
+      "loss": 1.0239,
+      "step": 1284
+    },
+    {
+      "epoch": 0.13514401777380466,
+      "grad_norm": 3.1136862313700497,
+      "learning_rate": 4.806813160621362e-06,
+      "loss": 0.9802,
+      "step": 1285
+    },
+    {
+      "epoch": 0.135249188215652,
+      "grad_norm": 3.3108785416004722,
+      "learning_rate": 4.806491243932877e-06,
+      "loss": 1.0428,
+      "step": 1286
+    },
+    {
+      "epoch": 0.1353543586574993,
+      "grad_norm": 2.1210612289323048,
+      "learning_rate": 4.806169070052701e-06,
+      "loss": 0.9905,
+      "step": 1287
+    },
+    {
+      "epoch": 0.13545952909934664,
+      "grad_norm": 3.1193765406401797,
+      "learning_rate": 4.805846639016759e-06,
+      "loss": 1.0397,
+      "step": 1288
+    },
+    {
+      "epoch": 0.13556469954119393,
+      "grad_norm": 3.3591628388988757,
+      "learning_rate": 4.805523950861004e-06,
+      "loss": 1.0248,
+      "step": 1289
+    },
+    {
+      "epoch": 0.13566986998304126,
+      "grad_norm": 2.4188726077223746,
+      "learning_rate": 4.8052010056214184e-06,
+      "loss": 1.0516,
+      "step": 1290
+    },
+    {
+      "epoch": 0.13577504042488858,
+      "grad_norm": 3.4651111802218093,
+      "learning_rate": 4.804877803334013e-06,
+      "loss": 1.0804,
+      "step": 1291
+    },
+    {
+      "epoch": 0.1358802108667359,
+      "grad_norm": 2.919536255232778,
+      "learning_rate": 4.8045543440348276e-06,
+      "loss": 1.0286,
+      "step": 1292
+    },
+    {
+      "epoch": 0.13598538130858323,
+      "grad_norm": 2.2477800261870344,
+      "learning_rate": 4.804230627759931e-06,
+      "loss": 1.0185,
+      "step": 1293
+    },
+    {
+      "epoch": 0.13609055175043056,
+      "grad_norm": 2.571200906414569,
+      "learning_rate": 4.803906654545419e-06,
+      "loss": 1.018,
+      "step": 1294
+    },
+    {
+      "epoch": 0.13619572219227785,
+      "grad_norm": 2.760586584220439,
+      "learning_rate": 4.803582424427419e-06,
+      "loss": 1.0139,
+      "step": 1295
+    },
+    {
+      "epoch": 0.13630089263412518,
+      "grad_norm": 2.566832369387431,
+      "learning_rate": 4.803257937442082e-06,
+      "loss": 1.0429,
+      "step": 1296
+    },
+    {
+      "epoch": 0.1364060630759725,
+      "grad_norm": 4.222863669394409,
+      "learning_rate": 4.802933193625593e-06,
+      "loss": 1.0491,
+      "step": 1297
+    },
+    {
+      "epoch": 0.13651123351781982,
+      "grad_norm": 3.8303058287728664,
+      "learning_rate": 4.802608193014164e-06,
+      "loss": 1.0619,
+      "step": 1298
+    },
+    {
+      "epoch": 0.13661640395966715,
+      "grad_norm": 2.6093520415311167,
+      "learning_rate": 4.802282935644034e-06,
+      "loss": 1.0233,
+      "step": 1299
+    },
+    {
+      "epoch": 0.13672157440151445,
+      "grad_norm": 2.9601576956633706,
+      "learning_rate": 4.8019574215514705e-06,
+      "loss": 1.0153,
+      "step": 1300
+    },
+    {
+      "epoch": 0.13682674484336177,
+      "grad_norm": 2.1944445212698986,
+      "learning_rate": 4.801631650772774e-06,
+      "loss": 0.9904,
+      "step": 1301
+    },
+    {
+      "epoch": 0.1369319152852091,
+      "grad_norm": 2.6684642895304327,
+      "learning_rate": 4.801305623344268e-06,
+      "loss": 1.0708,
+      "step": 1302
+    },
+    {
+      "epoch": 0.13703708572705642,
+      "grad_norm": 2.686121392463823,
+      "learning_rate": 4.800979339302308e-06,
+      "loss": 1.0116,
+      "step": 1303
+    },
+    {
+      "epoch": 0.13714225616890374,
+      "grad_norm": 2.8185256267861836,
+      "learning_rate": 4.800652798683277e-06,
+      "loss": 1.055,
+      "step": 1304
+    },
+    {
+      "epoch": 0.13724742661075104,
+      "grad_norm": 2.793491565244081,
+      "learning_rate": 4.800326001523586e-06,
+      "loss": 0.9955,
+      "step": 1305
+    },
+    {
+      "epoch": 0.13735259705259836,
+      "grad_norm": 4.301260101751436,
+      "learning_rate": 4.799998947859678e-06,
+      "loss": 1.0288,
+      "step": 1306
+    },
+    {
+      "epoch": 0.1374577674944457,
+      "grad_norm": 2.1242732201639183,
+      "learning_rate": 4.799671637728019e-06,
+      "loss": 1.0248,
+      "step": 1307
+    },
+    {
+      "epoch": 0.137562937936293,
+      "grad_norm": 3.5698968807461364,
+      "learning_rate": 4.799344071165108e-06,
+      "loss": 1.0036,
+      "step": 1308
+    },
+    {
+      "epoch": 0.13766810837814034,
+      "grad_norm": 2.5835211133401237,
+      "learning_rate": 4.799016248207472e-06,
+      "loss": 1.0235,
+      "step": 1309
+    },
+    {
+      "epoch": 0.13777327881998763,
+      "grad_norm": 3.0261001648609454,
+      "learning_rate": 4.7986881688916646e-06,
+      "loss": 1.0384,
+      "step": 1310
+    },
+    {
+      "epoch": 0.13787844926183496,
+      "grad_norm": 3.474891168157717,
+      "learning_rate": 4.798359833254269e-06,
+      "loss": 1.0484,
+      "step": 1311
+    },
+    {
+      "epoch": 0.13798361970368228,
+      "grad_norm": 3.1716499888572036,
+      "learning_rate": 4.798031241331898e-06,
+      "loss": 1.0298,
+      "step": 1312
+    },
+    {
+      "epoch": 0.1380887901455296,
+      "grad_norm": 2.7269924098282514,
+      "learning_rate": 4.7977023931611916e-06,
+      "loss": 1.0619,
+      "step": 1313
+    },
+    {
+      "epoch": 0.13819396058737693,
+      "grad_norm": 3.0594179433515247,
+      "learning_rate": 4.797373288778819e-06,
+      "loss": 1.0194,
+      "step": 1314
+    },
+    {
+      "epoch": 0.13829913102922423,
+      "grad_norm": 2.6485215434411704,
+      "learning_rate": 4.797043928221479e-06,
+      "loss": 1.0012,
+      "step": 1315
+    },
+    {
+      "epoch": 0.13840430147107155,
+      "grad_norm": 2.387465962136887,
+      "learning_rate": 4.796714311525897e-06,
+      "loss": 0.9964,
+      "step": 1316
+    },
+    {
+      "epoch": 0.13850947191291887,
+      "grad_norm": 3.2239747245175883,
+      "learning_rate": 4.796384438728827e-06,
+      "loss": 1.0092,
+      "step": 1317
+    },
+    {
+      "epoch": 0.1386146423547662,
+      "grad_norm": 3.378894724491444,
+      "learning_rate": 4.796054309867053e-06,
+      "loss": 1.063,
+      "step": 1318
+    },
+    {
+      "epoch": 0.13871981279661352,
+      "grad_norm": 2.9523255723706203,
+      "learning_rate": 4.7957239249773876e-06,
+      "loss": 1.0547,
+      "step": 1319
+    },
+    {
+      "epoch": 0.13882498323846082,
+      "grad_norm": 2.7245192269142366,
+      "learning_rate": 4.795393284096671e-06,
+      "loss": 1.0321,
+      "step": 1320
+    },
+    {
+      "epoch": 0.13893015368030814,
+      "grad_norm": 3.402757468965259,
+      "learning_rate": 4.7950623872617705e-06,
+      "loss": 1.0365,
+      "step": 1321
+    },
+    {
+      "epoch": 0.13903532412215547,
+      "grad_norm": 2.0645918619669907,
+      "learning_rate": 4.7947312345095865e-06,
+      "loss": 1.0378,
+      "step": 1322
+    },
+    {
+      "epoch": 0.1391404945640028,
+      "grad_norm": 3.1649331722389027,
+      "learning_rate": 4.794399825877044e-06,
+      "loss": 1.0502,
+      "step": 1323
+    },
+    {
+      "epoch": 0.13924566500585012,
+      "grad_norm": 3.0438185017979316,
+      "learning_rate": 4.794068161401097e-06,
+      "loss": 1.0202,
+      "step": 1324
+    },
+    {
+      "epoch": 0.1393508354476974,
+      "grad_norm": 2.8567093139911126,
+      "learning_rate": 4.793736241118729e-06,
+      "loss": 0.9863,
+      "step": 1325
+    },
+    {
+      "epoch": 0.13945600588954474,
+      "grad_norm": 3.1861700342675783,
+      "learning_rate": 4.793404065066951e-06,
+      "loss": 1.052,
+      "step": 1326
+    },
+    {
+      "epoch": 0.13956117633139206,
+      "grad_norm": 2.268534254398867,
+      "learning_rate": 4.7930716332828055e-06,
+      "loss": 1.0266,
+      "step": 1327
+    },
+    {
+      "epoch": 0.13966634677323939,
+      "grad_norm": 2.7798949911303215,
+      "learning_rate": 4.792738945803358e-06,
+      "loss": 1.0042,
+      "step": 1328
+    },
+    {
+      "epoch": 0.1397715172150867,
+      "grad_norm": 2.9625802390516673,
+      "learning_rate": 4.792406002665708e-06,
+      "loss": 1.0692,
+      "step": 1329
+    },
+    {
+      "epoch": 0.139876687656934,
+      "grad_norm": 3.3097071364207076,
+      "learning_rate": 4.792072803906982e-06,
+      "loss": 1.0495,
+      "step": 1330
+    },
+    {
+      "epoch": 0.13998185809878133,
+      "grad_norm": 3.3911434469914252,
+      "learning_rate": 4.791739349564332e-06,
+      "loss": 1.0219,
+      "step": 1331
+    },
+    {
+      "epoch": 0.14008702854062866,
+      "grad_norm": 2.4415653316509323,
+      "learning_rate": 4.791405639674941e-06,
+      "loss": 1.0159,
+      "step": 1332
+    },
+    {
+      "epoch": 0.14019219898247598,
+      "grad_norm": 2.673194071720424,
+      "learning_rate": 4.791071674276021e-06,
+      "loss": 1.0428,
+      "step": 1333
+    },
+    {
+      "epoch": 0.1402973694243233,
+      "grad_norm": 3.4270530074992207,
+      "learning_rate": 4.790737453404812e-06,
+      "loss": 1.0427,
+      "step": 1334
+    },
+    {
+      "epoch": 0.1404025398661706,
+      "grad_norm": 2.370560923133824,
+      "learning_rate": 4.790402977098583e-06,
+      "loss": 1.0541,
+      "step": 1335
+    },
+    {
+      "epoch": 0.14050771030801792,
+      "grad_norm": 3.178115791238882,
+      "learning_rate": 4.7900682453946276e-06,
+      "loss": 1.0314,
+      "step": 1336
+    },
+    {
+      "epoch": 0.14061288074986525,
+      "grad_norm": 2.6802658373773176,
+      "learning_rate": 4.7897332583302745e-06,
+      "loss": 1.0589,
+      "step": 1337
+    },
+    {
+      "epoch": 0.14071805119171257,
+      "grad_norm": 2.848166188577203,
+      "learning_rate": 4.789398015942875e-06,
+      "loss": 1.0272,
+      "step": 1338
+    },
+    {
+      "epoch": 0.1408232216335599,
+      "grad_norm": 3.1907811105881922,
+      "learning_rate": 4.789062518269812e-06,
+      "loss": 1.034,
+      "step": 1339
+    },
+    {
+      "epoch": 0.1409283920754072,
+      "grad_norm": 2.9382816317350375,
+      "learning_rate": 4.7887267653484965e-06,
+      "loss": 1.015,
+      "step": 1340
+    },
+    {
+      "epoch": 0.14103356251725452,
+      "grad_norm": 4.496979260313411,
+      "learning_rate": 4.788390757216367e-06,
+      "loss": 1.0533,
+      "step": 1341
+    },
+    {
+      "epoch": 0.14113873295910184,
+      "grad_norm": 2.6947544952945037,
+      "learning_rate": 4.7880544939108915e-06,
+      "loss": 1.045,
+      "step": 1342
+    },
+    {
+      "epoch": 0.14124390340094917,
+      "grad_norm": 2.6878213842926644,
+      "learning_rate": 4.787717975469566e-06,
+      "loss": 1.0424,
+      "step": 1343
+    },
+    {
+      "epoch": 0.1413490738427965,
+      "grad_norm": 2.9207345197550323,
+      "learning_rate": 4.787381201929913e-06,
+      "loss": 1.0479,
+      "step": 1344
+    },
+    {
+      "epoch": 0.1414542442846438,
+      "grad_norm": 2.8600747613086024,
+      "learning_rate": 4.787044173329489e-06,
+      "loss": 1.0018,
+      "step": 1345
+    },
+    {
+      "epoch": 0.1415594147264911,
+      "grad_norm": 2.456864436992996,
+      "learning_rate": 4.7867068897058725e-06,
+      "loss": 1.0043,
+      "step": 1346
+    },
+    {
+      "epoch": 0.14166458516833844,
+      "grad_norm": 3.0962033894183087,
+      "learning_rate": 4.7863693510966735e-06,
+      "loss": 0.9958,
+      "step": 1347
+    },
+    {
+      "epoch": 0.14176975561018576,
+      "grad_norm": 2.6323675844789016,
+      "learning_rate": 4.786031557539532e-06,
+      "loss": 1.0502,
+      "step": 1348
+    },
+    {
+      "epoch": 0.14187492605203308,
+      "grad_norm": 2.694477249584249,
+      "learning_rate": 4.785693509072112e-06,
+      "loss": 1.0372,
+      "step": 1349
+    },
+    {
+      "epoch": 0.14198009649388038,
+      "grad_norm": 3.3629384103067452,
+      "learning_rate": 4.785355205732111e-06,
+      "loss": 0.9938,
+      "step": 1350
+    },
+    {
+      "epoch": 0.1420852669357277,
+      "grad_norm": 3.334830674838776,
+      "learning_rate": 4.78501664755725e-06,
+      "loss": 1.0405,
+      "step": 1351
+    },
+    {
+      "epoch": 0.14219043737757503,
+      "grad_norm": 2.2938533826137455,
+      "learning_rate": 4.784677834585283e-06,
+      "loss": 1.0207,
+      "step": 1352
+    },
+    {
+      "epoch": 0.14229560781942235,
+      "grad_norm": 2.5803341129447896,
+      "learning_rate": 4.784338766853989e-06,
+      "loss": 1.0151,
+      "step": 1353
+    },
+    {
+      "epoch": 0.14240077826126968,
+      "grad_norm": 2.9122499893707383,
+      "learning_rate": 4.783999444401178e-06,
+      "loss": 1.0644,
+      "step": 1354
+    },
+    {
+      "epoch": 0.142505948703117,
+      "grad_norm": 2.091992322725459,
+      "learning_rate": 4.783659867264685e-06,
+      "loss": 0.9981,
+      "step": 1355
+    },
+    {
+      "epoch": 0.1426111191449643,
+      "grad_norm": 3.065753050616509,
+      "learning_rate": 4.783320035482378e-06,
+      "loss": 0.9894,
+      "step": 1356
+    },
+    {
+      "epoch": 0.14271628958681162,
+      "grad_norm": 3.1525287256847903,
+      "learning_rate": 4.7829799490921485e-06,
+      "loss": 1.077,
+      "step": 1357
+    },
+    {
+      "epoch": 0.14282146002865895,
+      "grad_norm": 3.3013362649199065,
+      "learning_rate": 4.7826396081319206e-06,
+      "loss": 1.0439,
+      "step": 1358
+    },
+    {
+      "epoch": 0.14292663047050627,
+      "grad_norm": 2.6267213566754295,
+      "learning_rate": 4.782299012639644e-06,
+      "loss": 1.0254,
+      "step": 1359
+    },
+    {
+      "epoch": 0.1430318009123536,
+      "grad_norm": 3.664812917190621,
+      "learning_rate": 4.781958162653298e-06,
+      "loss": 1.0252,
+      "step": 1360
+    },
+    {
+      "epoch": 0.1431369713542009,
+      "grad_norm": 2.3021182933754103,
+      "learning_rate": 4.781617058210889e-06,
+      "loss": 1.0615,
+      "step": 1361
+    },
+    {
+      "epoch": 0.14324214179604822,
+      "grad_norm": 2.881195846130054,
+      "learning_rate": 4.781275699350455e-06,
+      "loss": 1.0005,
+      "step": 1362
+    },
+    {
+      "epoch": 0.14334731223789554,
+      "grad_norm": 3.7980632477387632,
+      "learning_rate": 4.780934086110059e-06,
+      "loss": 1.0062,
+      "step": 1363
+    },
+    {
+      "epoch": 0.14345248267974287,
+      "grad_norm": 2.4526767159057505,
+      "learning_rate": 4.780592218527793e-06,
+      "loss": 1.0588,
+      "step": 1364
+    },
+    {
+      "epoch": 0.1435576531215902,
+      "grad_norm": 2.7801413202576177,
+      "learning_rate": 4.780250096641778e-06,
+      "loss": 1.004,
+      "step": 1365
+    },
+    {
+      "epoch": 0.1436628235634375,
+      "grad_norm": 3.2606411773565442,
+      "learning_rate": 4.779907720490164e-06,
+      "loss": 1.0199,
+      "step": 1366
+    },
+    {
+      "epoch": 0.1437679940052848,
+      "grad_norm": 3.222765541930548,
+      "learning_rate": 4.779565090111129e-06,
+      "loss": 0.9907,
+      "step": 1367
+    },
+    {
+      "epoch": 0.14387316444713213,
+      "grad_norm": 3.1387995556419357,
+      "learning_rate": 4.779222205542877e-06,
+      "loss": 1.0304,
+      "step": 1368
+    },
+    {
+      "epoch": 0.14397833488897946,
+      "grad_norm": 2.389261805095194,
+      "learning_rate": 4.778879066823644e-06,
+      "loss": 1.0449,
+      "step": 1369
+    },
+    {
+      "epoch": 0.14408350533082678,
+      "grad_norm": 2.724316181670948,
+      "learning_rate": 4.778535673991692e-06,
+      "loss": 1.0199,
+      "step": 1370
+    },
+    {
+      "epoch": 0.14418867577267408,
+      "grad_norm": 3.0155113196278585,
+      "learning_rate": 4.7781920270853126e-06,
+      "loss": 1.0333,
+      "step": 1371
+    },
+    {
+      "epoch": 0.1442938462145214,
+      "grad_norm": 3.5146591280021164,
+      "learning_rate": 4.777848126142824e-06,
+      "loss": 1.0147,
+      "step": 1372
+    },
+    {
+      "epoch": 0.14439901665636873,
+      "grad_norm": 2.61346142352295,
+      "learning_rate": 4.777503971202574e-06,
+      "loss": 1.0076,
+      "step": 1373
+    },
+    {
+      "epoch": 0.14450418709821605,
+      "grad_norm": 2.7874977280860618,
+      "learning_rate": 4.77715956230294e-06,
+      "loss": 0.994,
+      "step": 1374
+    },
+    {
+      "epoch": 0.14460935754006338,
+      "grad_norm": 3.1736286121545536,
+      "learning_rate": 4.776814899482325e-06,
+      "loss": 1.0834,
+      "step": 1375
+    },
+    {
+      "epoch": 0.14471452798191067,
+      "grad_norm": 3.6191087311210732,
+      "learning_rate": 4.776469982779161e-06,
+      "loss": 1.0153,
+      "step": 1376
+    },
+    {
+      "epoch": 0.144819698423758,
+      "grad_norm": 3.745985677561027,
+      "learning_rate": 4.7761248122319105e-06,
+      "loss": 0.9798,
+      "step": 1377
+    },
+    {
+      "epoch": 0.14492486886560532,
+      "grad_norm": 2.3804794207514406,
+      "learning_rate": 4.775779387879061e-06,
+      "loss": 1.0106,
+      "step": 1378
+    },
+    {
+      "epoch": 0.14503003930745265,
+      "grad_norm": 2.0323842954289346,
+      "learning_rate": 4.7754337097591315e-06,
+      "loss": 1.031,
+      "step": 1379
+    },
+    {
+      "epoch": 0.14513520974929997,
+      "grad_norm": 2.633921230163831,
+      "learning_rate": 4.7750877779106666e-06,
+      "loss": 1.0546,
+      "step": 1380
+    },
+    {
+      "epoch": 0.14524038019114727,
+      "grad_norm": 3.477178349518341,
+      "learning_rate": 4.774741592372242e-06,
+      "loss": 1.0429,
+      "step": 1381
+    },
+    {
+      "epoch": 0.1453455506329946,
+      "grad_norm": 3.025833331172734,
+      "learning_rate": 4.774395153182459e-06,
+      "loss": 1.0074,
+      "step": 1382
+    },
+    {
+      "epoch": 0.14545072107484192,
+      "grad_norm": 2.8582583766175005,
+      "learning_rate": 4.774048460379947e-06,
+      "loss": 1.0239,
+      "step": 1383
+    },
+    {
+      "epoch": 0.14555589151668924,
+      "grad_norm": 3.5823549566490933,
+      "learning_rate": 4.773701514003367e-06,
+      "loss": 1.0257,
+      "step": 1384
+    },
+    {
+      "epoch": 0.14566106195853656,
+      "grad_norm": 2.6124292693903612,
+      "learning_rate": 4.773354314091405e-06,
+      "loss": 0.9493,
+      "step": 1385
+    },
+    {
+      "epoch": 0.14576623240038386,
+      "grad_norm": 2.8268960445709226,
+      "learning_rate": 4.773006860682777e-06,
+      "loss": 1.0065,
+      "step": 1386
+    },
+    {
+      "epoch": 0.14587140284223118,
+      "grad_norm": 3.9272556505800273,
+      "learning_rate": 4.772659153816228e-06,
+      "loss": 1.0455,
+      "step": 1387
+    },
+    {
+      "epoch": 0.1459765732840785,
+      "grad_norm": 3.071553519913472,
+      "learning_rate": 4.7723111935305275e-06,
+      "loss": 1.0151,
+      "step": 1388
+    },
+    {
+      "epoch": 0.14608174372592583,
+      "grad_norm": 2.713380086603046,
+      "learning_rate": 4.7719629798644775e-06,
+      "loss": 1.0394,
+      "step": 1389
+    },
+    {
+      "epoch": 0.14618691416777316,
+      "grad_norm": 2.5543229953092608,
+      "learning_rate": 4.7716145128569054e-06,
+      "loss": 1.0254,
+      "step": 1390
+    },
+    {
+      "epoch": 0.14629208460962045,
+      "grad_norm": 2.5719448084092216,
+      "learning_rate": 4.771265792546669e-06,
+      "loss": 0.9972,
+      "step": 1391
+    },
+    {
+      "epoch": 0.14639725505146778,
+      "grad_norm": 2.360610533233657,
+      "learning_rate": 4.770916818972653e-06,
+      "loss": 1.0448,
+      "step": 1392
+    },
+    {
+      "epoch": 0.1465024254933151,
+      "grad_norm": 2.8007366713065274,
+      "learning_rate": 4.77056759217377e-06,
+      "loss": 1.0058,
+      "step": 1393
+    },
+    {
+      "epoch": 0.14660759593516243,
+      "grad_norm": 2.524726744796562,
+      "learning_rate": 4.770218112188964e-06,
+      "loss": 1.0684,
+      "step": 1394
+    },
+    {
+      "epoch": 0.14671276637700975,
+      "grad_norm": 3.3903828921485943,
+      "learning_rate": 4.769868379057201e-06,
+      "loss": 1.1003,
+      "step": 1395
+    },
+    {
+      "epoch": 0.14681793681885705,
+      "grad_norm": 3.4978205266954037,
+      "learning_rate": 4.7695183928174804e-06,
+      "loss": 1.037,
+      "step": 1396
+    },
+    {
+      "epoch": 0.14692310726070437,
+      "grad_norm": 3.3446642465813152,
+      "learning_rate": 4.76916815350883e-06,
+      "loss": 1.0366,
+      "step": 1397
+    },
+    {
+      "epoch": 0.1470282777025517,
+      "grad_norm": 2.2392665764959836,
+      "learning_rate": 4.768817661170302e-06,
+      "loss": 1.0026,
+      "step": 1398
+    },
+    {
+      "epoch": 0.14713344814439902,
+      "grad_norm": 2.8612244152019963,
+      "learning_rate": 4.768466915840981e-06,
+      "loss": 1.0446,
+      "step": 1399
+    },
+    {
+      "epoch": 0.14723861858624634,
+      "grad_norm": 3.3281982579588085,
+      "learning_rate": 4.768115917559976e-06,
+      "loss": 1.0825,
+      "step": 1400
+    },
+    {
+      "epoch": 0.14734378902809364,
+      "grad_norm": 2.2985012611056597,
+      "learning_rate": 4.767764666366427e-06,
+      "loss": 1.0283,
+      "step": 1401
+    },
+    {
+      "epoch": 0.14744895946994097,
+      "grad_norm": 2.6342734587607755,
+      "learning_rate": 4.767413162299501e-06,
+      "loss": 0.9836,
+      "step": 1402
+    },
+    {
+      "epoch": 0.1475541299117883,
+      "grad_norm": 2.555811598919671,
+      "learning_rate": 4.7670614053983945e-06,
+      "loss": 0.9753,
+      "step": 1403
+    },
+    {
+      "epoch": 0.14765930035363561,
+      "grad_norm": 4.202733894529233,
+      "learning_rate": 4.766709395702329e-06,
+      "loss": 1.0445,
+      "step": 1404
+    },
+    {
+      "epoch": 0.14776447079548294,
+      "grad_norm": 2.8209004214002,
+      "learning_rate": 4.766357133250558e-06,
+      "loss": 1.0223,
+      "step": 1405
+    },
+    {
+      "epoch": 0.14786964123733023,
+      "grad_norm": 2.8446733034162657,
+      "learning_rate": 4.766004618082361e-06,
+      "loss": 1.0289,
+      "step": 1406
+    },
+    {
+      "epoch": 0.14797481167917756,
+      "grad_norm": 4.068733400424128,
+      "learning_rate": 4.765651850237046e-06,
+      "loss": 1.0016,
+      "step": 1407
+    },
+    {
+      "epoch": 0.14807998212102488,
+      "grad_norm": 3.844649508056965,
+      "learning_rate": 4.765298829753949e-06,
+      "loss": 0.9918,
+      "step": 1408
+    },
+    {
+      "epoch": 0.1481851525628722,
+      "grad_norm": 2.919584733622149,
+      "learning_rate": 4.764945556672435e-06,
+      "loss": 1.0238,
+      "step": 1409
+    },
+    {
+      "epoch": 0.14829032300471953,
+      "grad_norm": 3.7015807840351584,
+      "learning_rate": 4.764592031031898e-06,
+      "loss": 1.0315,
+      "step": 1410
+    },
+    {
+      "epoch": 0.14839549344656686,
+      "grad_norm": 4.273538161097895,
+      "learning_rate": 4.7642382528717565e-06,
+      "loss": 1.0426,
+      "step": 1411
+    },
+    {
+      "epoch": 0.14850066388841415,
+      "grad_norm": 3.509784403656781,
+      "learning_rate": 4.763884222231461e-06,
+      "loss": 0.9692,
+      "step": 1412
+    },
+    {
+      "epoch": 0.14860583433026148,
+      "grad_norm": 3.524891382474383,
+      "learning_rate": 4.763529939150489e-06,
+      "loss": 1.023,
+      "step": 1413
+    },
+    {
+      "epoch": 0.1487110047721088,
+      "grad_norm": 2.860810743236083,
+      "learning_rate": 4.763175403668344e-06,
+      "loss": 0.9906,
+      "step": 1414
+    },
+    {
+      "epoch": 0.14881617521395613,
+      "grad_norm": 3.051115350337341,
+      "learning_rate": 4.762820615824561e-06,
+      "loss": 1.0535,
+      "step": 1415
+    },
+    {
+      "epoch": 0.14892134565580345,
+      "grad_norm": 2.868633232025887,
+      "learning_rate": 4.7624655756587e-06,
+      "loss": 1.0226,
+      "step": 1416
+    },
+    {
+      "epoch": 0.14902651609765075,
+      "grad_norm": 3.2091074851672823,
+      "learning_rate": 4.762110283210353e-06,
+      "loss": 1.0125,
+      "step": 1417
+    },
+    {
+      "epoch": 0.14913168653949807,
+      "grad_norm": 3.4892610849853614,
+      "learning_rate": 4.761754738519136e-06,
+      "loss": 1.0046,
+      "step": 1418
+    },
+    {
+      "epoch": 0.1492368569813454,
+      "grad_norm": 3.5426727194444623,
+      "learning_rate": 4.761398941624696e-06,
+      "loss": 1.012,
+      "step": 1419
+    },
+    {
+      "epoch": 0.14934202742319272,
+      "grad_norm": 2.648969769760997,
+      "learning_rate": 4.761042892566707e-06,
+      "loss": 1.0218,
+      "step": 1420
+    },
+    {
+      "epoch": 0.14944719786504004,
+      "grad_norm": 3.2305776271347746,
+      "learning_rate": 4.7606865913848725e-06,
+      "loss": 1.0702,
+      "step": 1421
+    },
+    {
+      "epoch": 0.14955236830688734,
+      "grad_norm": 3.1832515293962027,
+      "learning_rate": 4.760330038118919e-06,
+      "loss": 1.0464,
+      "step": 1422
+    },
+    {
+      "epoch": 0.14965753874873466,
+      "grad_norm": 3.4113754972892743,
+      "learning_rate": 4.759973232808609e-06,
+      "loss": 1.0189,
+      "step": 1423
+    },
+    {
+      "epoch": 0.149762709190582,
+      "grad_norm": 2.720125937598134,
+      "learning_rate": 4.759616175493726e-06,
+      "loss": 0.9823,
+      "step": 1424
+    },
+    {
+      "epoch": 0.1498678796324293,
+      "grad_norm": 3.4131804917283537,
+      "learning_rate": 4.7592588662140876e-06,
+      "loss": 1.0335,
+      "step": 1425
+    },
+    {
+      "epoch": 0.14997305007427664,
+      "grad_norm": 2.5441205030353937,
+      "learning_rate": 4.7589013050095345e-06,
+      "loss": 1.0354,
+      "step": 1426
+    },
+    {
+      "epoch": 0.15007822051612393,
+      "grad_norm": 3.064947944270619,
+      "learning_rate": 4.758543491919938e-06,
+      "loss": 1.0068,
+      "step": 1427
+    },
+    {
+      "epoch": 0.15018339095797126,
+      "grad_norm": 3.2381546378652417,
+      "learning_rate": 4.7581854269851975e-06,
+      "loss": 1.0831,
+      "step": 1428
+    },
+    {
+      "epoch": 0.15028856139981858,
+      "grad_norm": 3.0616456434518002,
+      "learning_rate": 4.75782711024524e-06,
+      "loss": 1.0218,
+      "step": 1429
+    },
+    {
+      "epoch": 0.1503937318416659,
+      "grad_norm": 2.923899250348212,
+      "learning_rate": 4.757468541740019e-06,
+      "loss": 1.0323,
+      "step": 1430
+    },
+    {
+      "epoch": 0.15049890228351323,
+      "grad_norm": 2.8260428144309735,
+      "learning_rate": 4.75710972150952e-06,
+      "loss": 1.0136,
+      "step": 1431
+    },
+    {
+      "epoch": 0.15060407272536053,
+      "grad_norm": 2.9366911956981085,
+      "learning_rate": 4.756750649593753e-06,
+      "loss": 1.0514,
+      "step": 1432
+    },
+    {
+      "epoch": 0.15070924316720785,
+      "grad_norm": 2.4211017023085635,
+      "learning_rate": 4.756391326032757e-06,
+      "loss": 1.0258,
+      "step": 1433
+    },
+    {
+      "epoch": 0.15081441360905518,
+      "grad_norm": 2.819024865923823,
+      "learning_rate": 4.756031750866601e-06,
+      "loss": 1.0051,
+      "step": 1434
+    },
+    {
+      "epoch": 0.1509195840509025,
+      "grad_norm": 3.1999851809479063,
+      "learning_rate": 4.755671924135379e-06,
+      "loss": 1.0053,
+      "step": 1435
+    },
+    {
+      "epoch": 0.15102475449274982,
+      "grad_norm": 3.4482793978325272,
+      "learning_rate": 4.755311845879214e-06,
+      "loss": 1.0168,
+      "step": 1436
+    },
+    {
+      "epoch": 0.15112992493459712,
+      "grad_norm": 3.2963820376471755,
+      "learning_rate": 4.754951516138259e-06,
+      "loss": 1.0435,
+      "step": 1437
+    },
+    {
+      "epoch": 0.15123509537644445,
+      "grad_norm": 2.6226202887722674,
+      "learning_rate": 4.754590934952692e-06,
+      "loss": 1.0022,
+      "step": 1438
+    },
+    {
+      "epoch": 0.15134026581829177,
+      "grad_norm": 4.119115148771308,
+      "learning_rate": 4.754230102362723e-06,
+      "loss": 1.0604,
+      "step": 1439
+    },
+    {
+      "epoch": 0.1514454362601391,
+      "grad_norm": 3.3530136943675606,
+      "learning_rate": 4.7538690184085845e-06,
+      "loss": 1.0374,
+      "step": 1440
+    },
+    {
+      "epoch": 0.15155060670198642,
+      "grad_norm": 3.0643118980396307,
+      "learning_rate": 4.7535076831305425e-06,
+      "loss": 1.0299,
+      "step": 1441
+    },
+    {
+      "epoch": 0.15165577714383371,
+      "grad_norm": 2.6808250795383914,
+      "learning_rate": 4.753146096568888e-06,
+      "loss": 1.0616,
+      "step": 1442
+    },
+    {
+      "epoch": 0.15176094758568104,
+      "grad_norm": 2.762417572989777,
+      "learning_rate": 4.752784258763941e-06,
+      "loss": 1.043,
+      "step": 1443
+    },
+    {
+      "epoch": 0.15186611802752836,
+      "grad_norm": 2.069150811277505,
+      "learning_rate": 4.752422169756048e-06,
+      "loss": 0.9979,
+      "step": 1444
+    },
+    {
+      "epoch": 0.1519712884693757,
+      "grad_norm": 2.952492663200309,
+      "learning_rate": 4.7520598295855866e-06,
+      "loss": 1.0638,
+      "step": 1445
+    },
+    {
+      "epoch": 0.152076458911223,
+      "grad_norm": 2.4223572836685126,
+      "learning_rate": 4.751697238292959e-06,
+      "loss": 1.0165,
+      "step": 1446
+    },
+    {
+      "epoch": 0.1521816293530703,
+      "grad_norm": 3.035954168307138,
+      "learning_rate": 4.751334395918598e-06,
+      "loss": 1.0296,
+      "step": 1447
+    },
+    {
+      "epoch": 0.15228679979491763,
+      "grad_norm": 2.5050761120596396,
+      "learning_rate": 4.7509713025029624e-06,
+      "loss": 1.0469,
+      "step": 1448
+    },
+    {
+      "epoch": 0.15239197023676496,
+      "grad_norm": 3.6719733529575693,
+      "learning_rate": 4.750607958086541e-06,
+      "loss": 1.0405,
+      "step": 1449
+    },
+    {
+      "epoch": 0.15249714067861228,
+      "grad_norm": 2.122334436297666,
+      "learning_rate": 4.750244362709848e-06,
+      "loss": 1.0358,
+      "step": 1450
+    },
+    {
+      "epoch": 0.1526023111204596,
+      "grad_norm": 3.307644507612065,
+      "learning_rate": 4.749880516413428e-06,
+      "loss": 1.0013,
+      "step": 1451
+    },
+    {
+      "epoch": 0.1527074815623069,
+      "grad_norm": 2.26272280730536,
+      "learning_rate": 4.749516419237854e-06,
+      "loss": 1.0286,
+      "step": 1452
+    },
+    {
+      "epoch": 0.15281265200415423,
+      "grad_norm": 3.4033816929506444,
+      "learning_rate": 4.749152071223724e-06,
+      "loss": 1.0483,
+      "step": 1453
+    },
+    {
+      "epoch": 0.15291782244600155,
+      "grad_norm": 3.19081203769598,
+      "learning_rate": 4.748787472411665e-06,
+      "loss": 1.054,
+      "step": 1454
+    },
+    {
+      "epoch": 0.15302299288784887,
+      "grad_norm": 2.863548719789538,
+      "learning_rate": 4.748422622842335e-06,
+      "loss": 1.0378,
+      "step": 1455
+    },
+    {
+      "epoch": 0.1531281633296962,
+      "grad_norm": 2.918561339352643,
+      "learning_rate": 4.748057522556415e-06,
+      "loss": 1.064,
+      "step": 1456
+    },
+    {
+      "epoch": 0.1532333337715435,
+      "grad_norm": 3.235112382704938,
+      "learning_rate": 4.747692171594619e-06,
+      "loss": 1.0154,
+      "step": 1457
+    },
+    {
+      "epoch": 0.15333850421339082,
+      "grad_norm": 2.9715433832923392,
+      "learning_rate": 4.747326569997684e-06,
+      "loss": 0.9907,
+      "step": 1458
+    },
+    {
+      "epoch": 0.15344367465523814,
+      "grad_norm": 3.7107017840666696,
+      "learning_rate": 4.746960717806379e-06,
+      "loss": 0.998,
+      "step": 1459
+    },
+    {
+      "epoch": 0.15354884509708547,
+      "grad_norm": 3.1443639363904614,
+      "learning_rate": 4.7465946150615e-06,
+      "loss": 1.0366,
+      "step": 1460
+    },
+    {
+      "epoch": 0.1536540155389328,
+      "grad_norm": 3.281239643947147,
+      "learning_rate": 4.746228261803868e-06,
+      "loss": 1.0366,
+      "step": 1461
+    },
+    {
+      "epoch": 0.1537591859807801,
+      "grad_norm": 3.382772295417614,
+      "learning_rate": 4.745861658074336e-06,
+      "loss": 1.0344,
+      "step": 1462
+    },
+    {
+      "epoch": 0.1538643564226274,
+      "grad_norm": 2.466197221139934,
+      "learning_rate": 4.745494803913781e-06,
+      "loss": 1.0028,
+      "step": 1463
+    },
+    {
+      "epoch": 0.15396952686447474,
+      "grad_norm": 3.360579078456302,
+      "learning_rate": 4.745127699363115e-06,
+      "loss": 1.0104,
+      "step": 1464
+    },
+    {
+      "epoch": 0.15407469730632206,
+      "grad_norm": 4.356696217441136,
+      "learning_rate": 4.744760344463267e-06,
+      "loss": 1.0674,
+      "step": 1465
+    },
+    {
+      "epoch": 0.15417986774816939,
+      "grad_norm": 3.2306718414761044,
+      "learning_rate": 4.744392739255203e-06,
+      "loss": 1.0717,
+      "step": 1466
+    },
+    {
+      "epoch": 0.15428503819001668,
+      "grad_norm": 2.2164328768086197,
+      "learning_rate": 4.744024883779915e-06,
+      "loss": 1.0049,
+      "step": 1467
+    },
+    {
+      "epoch": 0.154390208631864,
+      "grad_norm": 2.932022946712518,
+      "learning_rate": 4.74365677807842e-06,
+      "loss": 1.0116,
+      "step": 1468
+    },
+    {
+      "epoch": 0.15449537907371133,
+      "grad_norm": 2.590086012736109,
+      "learning_rate": 4.743288422191764e-06,
+      "loss": 1.0302,
+      "step": 1469
+    },
+    {
+      "epoch": 0.15460054951555866,
+      "grad_norm": 2.1435708761875665,
+      "learning_rate": 4.7429198161610225e-06,
+      "loss": 1.0083,
+      "step": 1470
+    },
+    {
+      "epoch": 0.15470571995740598,
+      "grad_norm": 3.630046810094376,
+      "learning_rate": 4.7425509600272974e-06,
+      "loss": 0.9893,
+      "step": 1471
+    },
+    {
+      "epoch": 0.1548108903992533,
+      "grad_norm": 2.719889362421328,
+      "learning_rate": 4.742181853831721e-06,
+      "loss": 1.0572,
+      "step": 1472
+    },
+    {
+      "epoch": 0.1549160608411006,
+      "grad_norm": 2.327018336326447,
+      "learning_rate": 4.74181249761545e-06,
+      "loss": 0.9962,
+      "step": 1473
+    },
+    {
+      "epoch": 0.15502123128294792,
+      "grad_norm": 2.3785350704725112,
+      "learning_rate": 4.74144289141967e-06,
+      "loss": 0.9893,
+      "step": 1474
+    },
+    {
+      "epoch": 0.15512640172479525,
+      "grad_norm": 2.59518918661727,
+      "learning_rate": 4.741073035285595e-06,
+      "loss": 0.9894,
+      "step": 1475
+    },
+    {
+      "epoch": 0.15523157216664257,
+      "grad_norm": 3.4946347737501915,
+      "learning_rate": 4.7407029292544675e-06,
+      "loss": 1.0119,
+      "step": 1476
+    },
+    {
+      "epoch": 0.1553367426084899,
+      "grad_norm": 2.6605925838001627,
+      "learning_rate": 4.740332573367557e-06,
+      "loss": 1.0098,
+      "step": 1477
+    },
+    {
+      "epoch": 0.1554419130503372,
+      "grad_norm": 3.8403647907126643,
+      "learning_rate": 4.739961967666161e-06,
+      "loss": 1.0622,
+      "step": 1478
+    },
+    {
+      "epoch": 0.15554708349218452,
+      "grad_norm": 2.4061249592344356,
+      "learning_rate": 4.739591112191605e-06,
+      "loss": 0.9919,
+      "step": 1479
+    },
+    {
+      "epoch": 0.15565225393403184,
+      "grad_norm": 3.619438134390123,
+      "learning_rate": 4.739220006985243e-06,
+      "loss": 1.0495,
+      "step": 1480
+    },
+    {
+      "epoch": 0.15575742437587917,
+      "grad_norm": 2.7851891041711814,
+      "learning_rate": 4.738848652088454e-06,
+      "loss": 1.0299,
+      "step": 1481
+    },
+    {
+      "epoch": 0.1558625948177265,
+      "grad_norm": 2.975714954933627,
+      "learning_rate": 4.738477047542649e-06,
+      "loss": 1.0205,
+      "step": 1482
+    },
+    {
+      "epoch": 0.1559677652595738,
+      "grad_norm": 3.878474331685192,
+      "learning_rate": 4.738105193389264e-06,
+      "loss": 0.988,
+      "step": 1483
+    },
+    {
+      "epoch": 0.1560729357014211,
+      "grad_norm": 3.449910786924835,
+      "learning_rate": 4.737733089669764e-06,
+      "loss": 1.0652,
+      "step": 1484
+    },
+    {
+      "epoch": 0.15617810614326844,
+      "grad_norm": 2.3712070448593803,
+      "learning_rate": 4.737360736425641e-06,
+      "loss": 1.0411,
+      "step": 1485
+    },
+    {
+      "epoch": 0.15628327658511576,
+      "grad_norm": 2.6454413280340874,
+      "learning_rate": 4.736988133698416e-06,
+      "loss": 1.0204,
+      "step": 1486
+    },
+    {
+      "epoch": 0.15638844702696308,
+      "grad_norm": 3.004156559443189,
+      "learning_rate": 4.736615281529635e-06,
+      "loss": 1.0051,
+      "step": 1487
+    },
+    {
+      "epoch": 0.15649361746881038,
+      "grad_norm": 3.7915743827388186,
+      "learning_rate": 4.736242179960877e-06,
+      "loss": 1.0261,
+      "step": 1488
+    },
+    {
+      "epoch": 0.1565987879106577,
+      "grad_norm": 3.602538359343123,
+      "learning_rate": 4.735868829033744e-06,
+      "loss": 1.0642,
+      "step": 1489
+    },
+    {
+      "epoch": 0.15670395835250503,
+      "grad_norm": 1.966550689023779,
+      "learning_rate": 4.735495228789867e-06,
+      "loss": 1.0363,
+      "step": 1490
+    },
+    {
+      "epoch": 0.15680912879435235,
+      "grad_norm": 2.511000996166126,
+      "learning_rate": 4.735121379270907e-06,
+      "loss": 1.025,
+      "step": 1491
+    },
+    {
+      "epoch": 0.15691429923619968,
+      "grad_norm": 2.563979152422195,
+      "learning_rate": 4.734747280518549e-06,
+      "loss": 1.0361,
+      "step": 1492
+    },
+    {
+      "epoch": 0.15701946967804697,
+      "grad_norm": 2.2916478990627946,
+      "learning_rate": 4.73437293257451e-06,
+      "loss": 1.0462,
+      "step": 1493
+    },
+    {
+      "epoch": 0.1571246401198943,
+      "grad_norm": 2.2807919186264436,
+      "learning_rate": 4.733998335480532e-06,
+      "loss": 1.0184,
+      "step": 1494
+    },
+    {
+      "epoch": 0.15722981056174162,
+      "grad_norm": 2.7351488990346535,
+      "learning_rate": 4.733623489278385e-06,
+      "loss": 1.0148,
+      "step": 1495
+    },
+    {
+      "epoch": 0.15733498100358895,
+      "grad_norm": 2.893572157477807,
+      "learning_rate": 4.733248394009867e-06,
+      "loss": 1.0028,
+      "step": 1496
+    },
+    {
+      "epoch": 0.15744015144543627,
+      "grad_norm": 2.6060141894891595,
+      "learning_rate": 4.732873049716805e-06,
+      "loss": 0.998,
+      "step": 1497
+    },
+    {
+      "epoch": 0.15754532188728357,
+      "grad_norm": 2.7708906663010313,
+      "learning_rate": 4.732497456441052e-06,
+      "loss": 1.0671,
+      "step": 1498
+    },
+    {
+      "epoch": 0.1576504923291309,
+      "grad_norm": 3.472148650720611,
+      "learning_rate": 4.732121614224491e-06,
+      "loss": 1.0178,
+      "step": 1499
+    },
+    {
+      "epoch": 0.15775566277097822,
+      "grad_norm": 2.5207390547929003,
+      "learning_rate": 4.731745523109029e-06,
+      "loss": 1.0238,
+      "step": 1500
+    },
+    {
+      "epoch": 0.15786083321282554,
+      "grad_norm": 3.4771528545364316,
+      "learning_rate": 4.731369183136605e-06,
+      "loss": 1.0706,
+      "step": 1501
+    },
+    {
+      "epoch": 0.15796600365467287,
+      "grad_norm": 2.65521118845941,
+      "learning_rate": 4.730992594349183e-06,
+      "loss": 1.0402,
+      "step": 1502
+    },
+    {
+      "epoch": 0.15807117409652016,
+      "grad_norm": 3.10957683893925,
+      "learning_rate": 4.730615756788756e-06,
+      "loss": 0.9857,
+      "step": 1503
+    },
+    {
+      "epoch": 0.15817634453836749,
+      "grad_norm": 2.6542173527792463,
+      "learning_rate": 4.730238670497345e-06,
+      "loss": 1.0336,
+      "step": 1504
+    },
+    {
+      "epoch": 0.1582815149802148,
+      "grad_norm": 3.77211681230834,
+      "learning_rate": 4.729861335516995e-06,
+      "loss": 1.0299,
+      "step": 1505
+    },
+    {
+      "epoch": 0.15838668542206213,
+      "grad_norm": 2.6411474196043727,
+      "learning_rate": 4.7294837518897855e-06,
+      "loss": 1.0441,
+      "step": 1506
+    },
+    {
+      "epoch": 0.15849185586390946,
+      "grad_norm": 2.99873384055499,
+      "learning_rate": 4.729105919657818e-06,
+      "loss": 1.0002,
+      "step": 1507
+    },
+    {
+      "epoch": 0.15859702630575676,
+      "grad_norm": 2.8324948060160904,
+      "learning_rate": 4.728727838863224e-06,
+      "loss": 1.0513,
+      "step": 1508
+    },
+    {
+      "epoch": 0.15870219674760408,
+      "grad_norm": 2.455839993218039,
+      "learning_rate": 4.728349509548163e-06,
+      "loss": 1.0082,
+      "step": 1509
+    },
+    {
+      "epoch": 0.1588073671894514,
+      "grad_norm": 3.5575923268239027,
+      "learning_rate": 4.7279709317548215e-06,
+      "loss": 1.0516,
+      "step": 1510
+    },
+    {
+      "epoch": 0.15891253763129873,
+      "grad_norm": 2.940216951551406,
+      "learning_rate": 4.727592105525413e-06,
+      "loss": 1.0099,
+      "step": 1511
+    },
+    {
+      "epoch": 0.15901770807314605,
+      "grad_norm": 2.8061910489333397,
+      "learning_rate": 4.72721303090218e-06,
+      "loss": 1.0284,
+      "step": 1512
+    },
+    {
+      "epoch": 0.15912287851499335,
+      "grad_norm": 2.8488993729192567,
+      "learning_rate": 4.726833707927393e-06,
+      "loss": 1.0029,
+      "step": 1513
+    },
+    {
+      "epoch": 0.15922804895684067,
+      "grad_norm": 2.3760044861779015,
+      "learning_rate": 4.7264541366433495e-06,
+      "loss": 1.064,
+      "step": 1514
+    },
+    {
+      "epoch": 0.159333219398688,
+      "grad_norm": 3.03368894409861,
+      "learning_rate": 4.726074317092373e-06,
+      "loss": 1.057,
+      "step": 1515
+    },
+    {
+      "epoch": 0.15943838984053532,
+      "grad_norm": 2.6274139546172273,
+      "learning_rate": 4.7256942493168176e-06,
+      "loss": 1.0468,
+      "step": 1516
+    },
+    {
+      "epoch": 0.15954356028238265,
+      "grad_norm": 2.5208665813866533,
+      "learning_rate": 4.725313933359064e-06,
+      "loss": 1.01,
+      "step": 1517
+    },
+    {
+      "epoch": 0.15964873072422994,
+      "grad_norm": 3.0144732325896313,
+      "learning_rate": 4.724933369261519e-06,
+      "loss": 1.0093,
+      "step": 1518
+    },
+    {
+      "epoch": 0.15975390116607727,
+      "grad_norm": 3.3250090964686274,
+      "learning_rate": 4.72455255706662e-06,
+      "loss": 0.9936,
+      "step": 1519
+    },
+    {
+      "epoch": 0.1598590716079246,
+      "grad_norm": 3.306440325363034,
+      "learning_rate": 4.724171496816831e-06,
+      "loss": 1.002,
+      "step": 1520
+    },
+    {
+      "epoch": 0.15996424204977192,
+      "grad_norm": 2.978656372246788,
+      "learning_rate": 4.7237901885546405e-06,
+      "loss": 1.0175,
+      "step": 1521
+    },
+    {
+      "epoch": 0.16006941249161924,
+      "grad_norm": 3.4879449084728056,
+      "learning_rate": 4.72340863232257e-06,
+      "loss": 1.05,
+      "step": 1522
+    },
+    {
+      "epoch": 0.16017458293346654,
+      "grad_norm": 2.7810346030912405,
+      "learning_rate": 4.723026828163164e-06,
+      "loss": 1.0297,
+      "step": 1523
+    },
+    {
+      "epoch": 0.16027975337531386,
+      "grad_norm": 2.2616816284442205,
+      "learning_rate": 4.722644776118999e-06,
+      "loss": 1.0186,
+      "step": 1524
+    },
+    {
+      "epoch": 0.16038492381716118,
+      "grad_norm": 2.729508624247899,
+      "learning_rate": 4.722262476232674e-06,
+      "loss": 1.0272,
+      "step": 1525
+    },
+    {
+      "epoch": 0.1604900942590085,
+      "grad_norm": 3.219956094139774,
+      "learning_rate": 4.72187992854682e-06,
+      "loss": 1.0983,
+      "step": 1526
+    },
+    {
+      "epoch": 0.16059526470085583,
+      "grad_norm": 3.9250885931506914,
+      "learning_rate": 4.7214971331040945e-06,
+      "loss": 1.0105,
+      "step": 1527
+    },
+    {
+      "epoch": 0.16070043514270313,
+      "grad_norm": 4.60638365380386,
+      "learning_rate": 4.721114089947181e-06,
+      "loss": 1.0343,
+      "step": 1528
+    },
+    {
+      "epoch": 0.16080560558455045,
+      "grad_norm": 4.087070072305575,
+      "learning_rate": 4.720730799118792e-06,
+      "loss": 1.072,
+      "step": 1529
+    },
+    {
+      "epoch": 0.16091077602639778,
+      "grad_norm": 2.114182974470423,
+      "learning_rate": 4.7203472606616685e-06,
+      "loss": 1.0551,
+      "step": 1530
+    },
+    {
+      "epoch": 0.1610159464682451,
+      "grad_norm": 2.7564719333605305,
+      "learning_rate": 4.719963474618576e-06,
+      "loss": 1.0046,
+      "step": 1531
+    },
+    {
+      "epoch": 0.16112111691009243,
+      "grad_norm": 2.194526925536452,
+      "learning_rate": 4.719579441032312e-06,
+      "loss": 0.9904,
+      "step": 1532
+    },
+    {
+      "epoch": 0.16122628735193975,
+      "grad_norm": 3.069579977353765,
+      "learning_rate": 4.7191951599456974e-06,
+      "loss": 1.0293,
+      "step": 1533
+    },
+    {
+      "epoch": 0.16133145779378705,
+      "grad_norm": 2.6614677051330355,
+      "learning_rate": 4.718810631401584e-06,
+      "loss": 0.969,
+      "step": 1534
+    },
+    {
+      "epoch": 0.16143662823563437,
+      "grad_norm": 2.0239894866349384,
+      "learning_rate": 4.718425855442847e-06,
+      "loss": 1.0166,
+      "step": 1535
+    },
+    {
+      "epoch": 0.1615417986774817,
+      "grad_norm": 2.6814734955623747,
+      "learning_rate": 4.718040832112396e-06,
+      "loss": 0.9991,
+      "step": 1536
+    },
+    {
+      "epoch": 0.16164696911932902,
+      "grad_norm": 3.3899909411502156,
+      "learning_rate": 4.717655561453161e-06,
+      "loss": 1.0229,
+      "step": 1537
+    },
+    {
+      "epoch": 0.16175213956117634,
+      "grad_norm": 3.6416888404436514,
+      "learning_rate": 4.7172700435081024e-06,
+      "loss": 1.0336,
+      "step": 1538
+    },
+    {
+      "epoch": 0.16185731000302364,
+      "grad_norm": 3.297177936886574,
+      "learning_rate": 4.71688427832021e-06,
+      "loss": 1.0382,
+      "step": 1539
+    },
+    {
+      "epoch": 0.16196248044487097,
+      "grad_norm": 2.9251719307297983,
+      "learning_rate": 4.716498265932501e-06,
+      "loss": 1.0379,
+      "step": 1540
+    },
+    {
+      "epoch": 0.1620676508867183,
+      "grad_norm": 3.1867271581724332,
+      "learning_rate": 4.716112006388015e-06,
+      "loss": 1.0093,
+      "step": 1541
+    },
+    {
+      "epoch": 0.16217282132856561,
+      "grad_norm": 2.7382289471102927,
+      "learning_rate": 4.715725499729826e-06,
+      "loss": 1.0596,
+      "step": 1542
+    },
+    {
+      "epoch": 0.16227799177041294,
+      "grad_norm": 2.482681348080373,
+      "learning_rate": 4.715338746001031e-06,
+      "loss": 1.0241,
+      "step": 1543
+    },
+    {
+      "epoch": 0.16238316221226023,
+      "grad_norm": 3.321185423836697,
+      "learning_rate": 4.7149517452447565e-06,
+      "loss": 1.065,
+      "step": 1544
+    },
+    {
+      "epoch": 0.16248833265410756,
+      "grad_norm": 2.2553546129954034,
+      "learning_rate": 4.714564497504156e-06,
+      "loss": 1.0071,
+      "step": 1545
+    },
+    {
+      "epoch": 0.16259350309595488,
+      "grad_norm": 2.9289085282753473,
+      "learning_rate": 4.714177002822411e-06,
+      "loss": 1.0756,
+      "step": 1546
+    },
+    {
+      "epoch": 0.1626986735378022,
+      "grad_norm": 2.6229389236880514,
+      "learning_rate": 4.7137892612427296e-06,
+      "loss": 0.9867,
+      "step": 1547
+    },
+    {
+      "epoch": 0.16280384397964953,
+      "grad_norm": 2.5687055044071756,
+      "learning_rate": 4.713401272808348e-06,
+      "loss": 1.0312,
+      "step": 1548
+    },
+    {
+      "epoch": 0.16290901442149683,
+      "grad_norm": 3.075733116731394,
+      "learning_rate": 4.713013037562531e-06,
+      "loss": 0.999,
+      "step": 1549
+    },
+    {
+      "epoch": 0.16301418486334415,
+      "grad_norm": 2.4489321077973316,
+      "learning_rate": 4.712624555548568e-06,
+      "loss": 1.0424,
+      "step": 1550
+    },
+    {
+      "epoch": 0.16311935530519148,
+      "grad_norm": 2.786568331849303,
+      "learning_rate": 4.712235826809779e-06,
+      "loss": 1.0076,
+      "step": 1551
+    },
+    {
+      "epoch": 0.1632245257470388,
+      "grad_norm": 4.261177829125735,
+      "learning_rate": 4.711846851389511e-06,
+      "loss": 1.055,
+      "step": 1552
+    },
+    {
+      "epoch": 0.16332969618888613,
+      "grad_norm": 3.0252873549859878,
+      "learning_rate": 4.711457629331136e-06,
+      "loss": 1.0215,
+      "step": 1553
+    },
+    {
+      "epoch": 0.16343486663073342,
+      "grad_norm": 2.127523638438989,
+      "learning_rate": 4.711068160678056e-06,
+      "loss": 1.035,
+      "step": 1554
+    },
+    {
+      "epoch": 0.16354003707258075,
+      "grad_norm": 3.1748982619719097,
+      "learning_rate": 4.710678445473701e-06,
+      "loss": 1.0526,
+      "step": 1555
+    },
+    {
+      "epoch": 0.16364520751442807,
+      "grad_norm": 3.0204839143683087,
+      "learning_rate": 4.710288483761524e-06,
+      "loss": 1.0283,
+      "step": 1556
+    },
+    {
+      "epoch": 0.1637503779562754,
+      "grad_norm": 2.5309450768958883,
+      "learning_rate": 4.709898275585013e-06,
+      "loss": 1.0435,
+      "step": 1557
+    },
+    {
+      "epoch": 0.16385554839812272,
+      "grad_norm": 2.9972306044536636,
+      "learning_rate": 4.709507820987676e-06,
+      "loss": 0.9908,
+      "step": 1558
+    },
+    {
+      "epoch": 0.16396071883997002,
+      "grad_norm": 3.6133112485444188,
+      "learning_rate": 4.709117120013054e-06,
+      "loss": 1.0428,
+      "step": 1559
+    },
+    {
+      "epoch": 0.16406588928181734,
+      "grad_norm": 2.2177026935577637,
+      "learning_rate": 4.708726172704712e-06,
+      "loss": 1.0373,
+      "step": 1560
+    },
+    {
+      "epoch": 0.16417105972366466,
+      "grad_norm": 3.665100322013887,
+      "learning_rate": 4.708334979106243e-06,
+      "loss": 1.017,
+      "step": 1561
+    },
+    {
+      "epoch": 0.164276230165512,
+      "grad_norm": 2.120634042862793,
+      "learning_rate": 4.707943539261269e-06,
+      "loss": 1.0423,
+      "step": 1562
+    },
+    {
+      "epoch": 0.1643814006073593,
+      "grad_norm": 3.3096318778798346,
+      "learning_rate": 4.707551853213439e-06,
+      "loss": 1.0324,
+      "step": 1563
+    },
+    {
+      "epoch": 0.1644865710492066,
+      "grad_norm": 2.722439105462851,
+      "learning_rate": 4.7071599210064275e-06,
+      "loss": 1.0437,
+      "step": 1564
+    },
+    {
+      "epoch": 0.16459174149105393,
+      "grad_norm": 2.9680950262706105,
+      "learning_rate": 4.70676774268394e-06,
+      "loss": 1.0011,
+      "step": 1565
+    },
+    {
+      "epoch": 0.16469691193290126,
+      "grad_norm": 2.7946642826043,
+      "learning_rate": 4.706375318289706e-06,
+      "loss": 0.9863,
+      "step": 1566
+    },
+    {
+      "epoch": 0.16480208237474858,
+      "grad_norm": 3.008921981513297,
+      "learning_rate": 4.705982647867484e-06,
+      "loss": 1.0011,
+      "step": 1567
+    },
+    {
+      "epoch": 0.1649072528165959,
+      "grad_norm": 2.438967195615608,
+      "learning_rate": 4.705589731461061e-06,
+      "loss": 1.0553,
+      "step": 1568
+    },
+    {
+      "epoch": 0.1650124232584432,
+      "grad_norm": 2.0590569324727492,
+      "learning_rate": 4.705196569114248e-06,
+      "loss": 1.0026,
+      "step": 1569
+    },
+    {
+      "epoch": 0.16511759370029053,
+      "grad_norm": 2.751042822546601,
+      "learning_rate": 4.704803160870888e-06,
+      "loss": 1.0228,
+      "step": 1570
+    },
+    {
+      "epoch": 0.16522276414213785,
+      "grad_norm": 1.9114644938599432,
+      "learning_rate": 4.704409506774848e-06,
+      "loss": 1.0235,
+      "step": 1571
+    },
+    {
+      "epoch": 0.16532793458398518,
+      "grad_norm": 2.7725485760210433,
+      "learning_rate": 4.7040156068700225e-06,
+      "loss": 1.039,
+      "step": 1572
+    },
+    {
+      "epoch": 0.1654331050258325,
+      "grad_norm": 3.3207748456848956,
+      "learning_rate": 4.703621461200337e-06,
+      "loss": 1.0551,
+      "step": 1573
+    },
+    {
+      "epoch": 0.1655382754676798,
+      "grad_norm": 2.2142667900370805,
+      "learning_rate": 4.703227069809739e-06,
+      "loss": 1.0284,
+      "step": 1574
+    },
+    {
+      "epoch": 0.16564344590952712,
+      "grad_norm": 2.4762983701232666,
+      "learning_rate": 4.702832432742208e-06,
+      "loss": 1.0301,
+      "step": 1575
+    },
+    {
+      "epoch": 0.16574861635137444,
+      "grad_norm": 2.7474342790405464,
+      "learning_rate": 4.702437550041749e-06,
+      "loss": 1.0002,
+      "step": 1576
+    },
+    {
+      "epoch": 0.16585378679322177,
+      "grad_norm": 2.490632975494078,
+      "learning_rate": 4.702042421752393e-06,
+      "loss": 0.9741,
+      "step": 1577
+    },
+    {
+      "epoch": 0.1659589572350691,
+      "grad_norm": 2.7515934794089683,
+      "learning_rate": 4.701647047918202e-06,
+      "loss": 1.0435,
+      "step": 1578
+    },
+    {
+      "epoch": 0.1660641276769164,
+      "grad_norm": 2.3948584711688916,
+      "learning_rate": 4.701251428583261e-06,
+      "loss": 1.0609,
+      "step": 1579
+    },
+    {
+      "epoch": 0.16616929811876371,
+      "grad_norm": 2.792702623447517,
+      "learning_rate": 4.700855563791686e-06,
+      "loss": 1.0161,
+      "step": 1580
+    },
+    {
+      "epoch": 0.16627446856061104,
+      "grad_norm": 2.81166121024242,
+      "learning_rate": 4.700459453587619e-06,
+      "loss": 1.0005,
+      "step": 1581
+    },
+    {
+      "epoch": 0.16637963900245836,
+      "grad_norm": 3.2253366610207723,
+      "learning_rate": 4.70006309801523e-06,
+      "loss": 1.0329,
+      "step": 1582
+    },
+    {
+      "epoch": 0.1664848094443057,
+      "grad_norm": 3.306319989253321,
+      "learning_rate": 4.699666497118714e-06,
+      "loss": 1.0728,
+      "step": 1583
+    },
+    {
+      "epoch": 0.16658997988615298,
+      "grad_norm": 2.3503420314034082,
+      "learning_rate": 4.699269650942296e-06,
+      "loss": 1.006,
+      "step": 1584
+    },
+    {
+      "epoch": 0.1666951503280003,
+      "grad_norm": 2.8744464659676634,
+      "learning_rate": 4.6988725595302275e-06,
+      "loss": 1.06,
+      "step": 1585
+    },
+    {
+      "epoch": 0.16680032076984763,
+      "grad_norm": 3.3878922195539247,
+      "learning_rate": 4.698475222926788e-06,
+      "loss": 1.0133,
+      "step": 1586
+    },
+    {
+      "epoch": 0.16690549121169496,
+      "grad_norm": 2.9490960578463747,
+      "learning_rate": 4.698077641176282e-06,
+      "loss": 0.9812,
+      "step": 1587
+    },
+    {
+      "epoch": 0.16701066165354228,
+      "grad_norm": 3.898835572720005,
+      "learning_rate": 4.697679814323044e-06,
+      "loss": 1.0266,
+      "step": 1588
+    },
+    {
+      "epoch": 0.16711583209538958,
+      "grad_norm": 2.4198826439223278,
+      "learning_rate": 4.6972817424114335e-06,
+      "loss": 0.9853,
+      "step": 1589
+    },
+    {
+      "epoch": 0.1672210025372369,
+      "grad_norm": 3.2157863216591065,
+      "learning_rate": 4.6968834254858405e-06,
+      "loss": 0.9761,
+      "step": 1590
+    },
+    {
+      "epoch": 0.16732617297908423,
+      "grad_norm": 3.42945668773291,
+      "learning_rate": 4.69648486359068e-06,
+      "loss": 1.0218,
+      "step": 1591
+    },
+    {
+      "epoch": 0.16743134342093155,
+      "grad_norm": 3.8221661628298316,
+      "learning_rate": 4.6960860567703935e-06,
+      "loss": 1.0389,
+      "step": 1592
+    },
+    {
+      "epoch": 0.16753651386277887,
+      "grad_norm": 2.5084187696585563,
+      "learning_rate": 4.6956870050694524e-06,
+      "loss": 1.046,
+      "step": 1593
+    },
+    {
+      "epoch": 0.1676416843046262,
+      "grad_norm": 2.2140737319132415,
+      "learning_rate": 4.695287708532353e-06,
+      "loss": 1.0125,
+      "step": 1594
+    },
+    {
+      "epoch": 0.1677468547464735,
+      "grad_norm": 2.906928468843093,
+      "learning_rate": 4.6948881672036205e-06,
+      "loss": 1.0367,
+      "step": 1595
+    },
+    {
+      "epoch": 0.16785202518832082,
+      "grad_norm": 2.9231721544834257,
+      "learning_rate": 4.694488381127808e-06,
+      "loss": 1.0004,
+      "step": 1596
+    },
+    {
+      "epoch": 0.16795719563016814,
+      "grad_norm": 2.2518787810723015,
+      "learning_rate": 4.6940883503494925e-06,
+      "loss": 1.0029,
+      "step": 1597
+    },
+    {
+      "epoch": 0.16806236607201547,
+      "grad_norm": 2.75505444461088,
+      "learning_rate": 4.693688074913282e-06,
+      "loss": 1.0651,
+      "step": 1598
+    },
+    {
+      "epoch": 0.1681675365138628,
+      "grad_norm": 2.9638750280320094,
+      "learning_rate": 4.69328755486381e-06,
+      "loss": 0.981,
+      "step": 1599
+    },
+    {
+      "epoch": 0.1682727069557101,
+      "grad_norm": 2.2358315107683038,
+      "learning_rate": 4.692886790245738e-06,
+      "loss": 1.0402,
+      "step": 1600
+    },
+    {
+      "epoch": 0.1683778773975574,
+      "grad_norm": 2.554442640835414,
+      "learning_rate": 4.692485781103753e-06,
+      "loss": 0.9991,
+      "step": 1601
+    },
+    {
+      "epoch": 0.16848304783940474,
+      "grad_norm": 3.6587203735508886,
+      "learning_rate": 4.692084527482572e-06,
+      "loss": 1.0296,
+      "step": 1602
+    },
+    {
+      "epoch": 0.16858821828125206,
+      "grad_norm": 2.874856178936171,
+      "learning_rate": 4.691683029426938e-06,
+      "loss": 1.034,
+      "step": 1603
+    },
+    {
+      "epoch": 0.16869338872309939,
+      "grad_norm": 2.0504221476945466,
+      "learning_rate": 4.691281286981619e-06,
+      "loss": 1.0412,
+      "step": 1604
+    },
+    {
+      "epoch": 0.16879855916494668,
+      "grad_norm": 4.299453983662296,
+      "learning_rate": 4.690879300191416e-06,
+      "loss": 1.0454,
+      "step": 1605
+    },
+    {
+      "epoch": 0.168903729606794,
+      "grad_norm": 3.0645533844066875,
+      "learning_rate": 4.690477069101151e-06,
+      "loss": 1.0127,
+      "step": 1606
+    },
+    {
+      "epoch": 0.16900890004864133,
+      "grad_norm": 3.8020279455543893,
+      "learning_rate": 4.690074593755676e-06,
+      "loss": 1.0422,
+      "step": 1607
+    },
+    {
+      "epoch": 0.16911407049048865,
+      "grad_norm": 3.0926763271379034,
+      "learning_rate": 4.689671874199871e-06,
+      "loss": 0.9955,
+      "step": 1608
+    },
+    {
+      "epoch": 0.16921924093233598,
+      "grad_norm": 3.0460948789795594,
+      "learning_rate": 4.6892689104786425e-06,
+      "loss": 0.9942,
+      "step": 1609
+    },
+    {
+      "epoch": 0.16932441137418328,
+      "grad_norm": 2.562365194025863,
+      "learning_rate": 4.6888657026369235e-06,
+      "loss": 0.9697,
+      "step": 1610
+    },
+    {
+      "epoch": 0.1694295818160306,
+      "grad_norm": 2.8814563767503456,
+      "learning_rate": 4.688462250719675e-06,
+      "loss": 1.0186,
+      "step": 1611
+    },
+    {
+      "epoch": 0.16953475225787792,
+      "grad_norm": 3.0019075219247844,
+      "learning_rate": 4.6880585547718845e-06,
+      "loss": 1.0168,
+      "step": 1612
+    },
+    {
+      "epoch": 0.16963992269972525,
+      "grad_norm": 3.3566378728550226,
+      "learning_rate": 4.6876546148385685e-06,
+      "loss": 1.0138,
+      "step": 1613
+    },
+    {
+      "epoch": 0.16974509314157257,
+      "grad_norm": 3.815106274251855,
+      "learning_rate": 4.687250430964768e-06,
+      "loss": 1.0534,
+      "step": 1614
+    },
+    {
+      "epoch": 0.16985026358341987,
+      "grad_norm": 3.955757242671359,
+      "learning_rate": 4.6868460031955535e-06,
+      "loss": 1.0462,
+      "step": 1615
+    },
+    {
+      "epoch": 0.1699554340252672,
+      "grad_norm": 3.1546888936551762,
+      "learning_rate": 4.686441331576021e-06,
+      "loss": 0.9697,
+      "step": 1616
+    },
+    {
+      "epoch": 0.17006060446711452,
+      "grad_norm": 2.987499231967973,
+      "learning_rate": 4.686036416151296e-06,
+      "loss": 1.0134,
+      "step": 1617
+    },
+    {
+      "epoch": 0.17016577490896184,
+      "grad_norm": 3.2641730627495567,
+      "learning_rate": 4.6856312569665285e-06,
+      "loss": 1.0178,
+      "step": 1618
+    },
+    {
+      "epoch": 0.17027094535080917,
+      "grad_norm": 2.5005821127937096,
+      "learning_rate": 4.685225854066897e-06,
+      "loss": 1.0002,
+      "step": 1619
+    },
+    {
+      "epoch": 0.17037611579265646,
+      "grad_norm": 2.3448731740329434,
+      "learning_rate": 4.684820207497608e-06,
+      "loss": 1.0644,
+      "step": 1620
+    },
+    {
+      "epoch": 0.1704812862345038,
+      "grad_norm": 3.060151169504036,
+      "learning_rate": 4.684414317303894e-06,
+      "loss": 1.0934,
+      "step": 1621
+    },
+    {
+      "epoch": 0.1705864566763511,
+      "grad_norm": 2.00150049947049,
+      "learning_rate": 4.6840081835310135e-06,
+      "loss": 1.0504,
+      "step": 1622
+    },
+    {
+      "epoch": 0.17069162711819844,
+      "grad_norm": 2.708844063331739,
+      "learning_rate": 4.683601806224255e-06,
+      "loss": 1.0163,
+      "step": 1623
+    },
+    {
+      "epoch": 0.17079679756004576,
+      "grad_norm": 3.4472196380160978,
+      "learning_rate": 4.683195185428932e-06,
+      "loss": 0.9911,
+      "step": 1624
+    },
+    {
+      "epoch": 0.17090196800189306,
+      "grad_norm": 2.8820764062472026,
+      "learning_rate": 4.6827883211903865e-06,
+      "loss": 1.0349,
+      "step": 1625
+    },
+    {
+      "epoch": 0.17100713844374038,
+      "grad_norm": 2.2574433822521067,
+      "learning_rate": 4.682381213553986e-06,
+      "loss": 1.0159,
+      "step": 1626
+    },
+    {
+      "epoch": 0.1711123088855877,
+      "grad_norm": 3.089754095729008,
+      "learning_rate": 4.681973862565128e-06,
+      "loss": 1.0047,
+      "step": 1627
+    },
+    {
+      "epoch": 0.17121747932743503,
+      "grad_norm": 3.057788643389871,
+      "learning_rate": 4.681566268269233e-06,
+      "loss": 1.0697,
+      "step": 1628
+    },
+    {
+      "epoch": 0.17132264976928235,
+      "grad_norm": 2.5667766260752085,
+      "learning_rate": 4.681158430711753e-06,
+      "loss": 1.0402,
+      "step": 1629
+    },
+    {
+      "epoch": 0.17142782021112965,
+      "grad_norm": 2.9763268275620116,
+      "learning_rate": 4.680750349938164e-06,
+      "loss": 1.0168,
+      "step": 1630
+    },
+    {
+      "epoch": 0.17153299065297697,
+      "grad_norm": 2.9070930547134566,
+      "learning_rate": 4.68034202599397e-06,
+      "loss": 1.0487,
+      "step": 1631
+    },
+    {
+      "epoch": 0.1716381610948243,
+      "grad_norm": 2.6307238317161494,
+      "learning_rate": 4.679933458924702e-06,
+      "loss": 1.0303,
+      "step": 1632
+    },
+    {
+      "epoch": 0.17174333153667162,
+      "grad_norm": 3.6659807629189913,
+      "learning_rate": 4.67952464877592e-06,
+      "loss": 0.9981,
+      "step": 1633
+    },
+    {
+      "epoch": 0.17184850197851895,
+      "grad_norm": 2.736348178417095,
+      "learning_rate": 4.679115595593208e-06,
+      "loss": 0.9939,
+      "step": 1634
+    },
+    {
+      "epoch": 0.17195367242036624,
+      "grad_norm": 3.0409494428632993,
+      "learning_rate": 4.67870629942218e-06,
+      "loss": 1.0585,
+      "step": 1635
+    },
+    {
+      "epoch": 0.17205884286221357,
+      "grad_norm": 2.4454258516009593,
+      "learning_rate": 4.678296760308474e-06,
+      "loss": 1.0395,
+      "step": 1636
+    },
+    {
+      "epoch": 0.1721640133040609,
+      "grad_norm": 2.612925874097498,
+      "learning_rate": 4.677886978297758e-06,
+      "loss": 0.9964,
+      "step": 1637
+    },
+    {
+      "epoch": 0.17226918374590822,
+      "grad_norm": 3.0527819045438793,
+      "learning_rate": 4.677476953435725e-06,
+      "loss": 1.0306,
+      "step": 1638
+    },
+    {
+      "epoch": 0.17237435418775554,
+      "grad_norm": 3.7722857272845616,
+      "learning_rate": 4.677066685768097e-06,
+      "loss": 1.0256,
+      "step": 1639
+    },
+    {
+      "epoch": 0.17247952462960284,
+      "grad_norm": 4.671732853843867,
+      "learning_rate": 4.676656175340621e-06,
+      "loss": 1.0539,
+      "step": 1640
+    },
+    {
+      "epoch": 0.17258469507145016,
+      "grad_norm": 3.402158905492146,
+      "learning_rate": 4.676245422199073e-06,
+      "loss": 1.041,
+      "step": 1641
+    },
+    {
+      "epoch": 0.17268986551329749,
+      "grad_norm": 2.7043484924050913,
+      "learning_rate": 4.675834426389254e-06,
+      "loss": 0.9748,
+      "step": 1642
+    },
+    {
+      "epoch": 0.1727950359551448,
+      "grad_norm": 3.045874999700864,
+      "learning_rate": 4.675423187956995e-06,
+      "loss": 1.0199,
+      "step": 1643
+    },
+    {
+      "epoch": 0.17290020639699213,
+      "grad_norm": 2.0391498549902325,
+      "learning_rate": 4.675011706948151e-06,
+      "loss": 0.9807,
+      "step": 1644
+    },
+    {
+      "epoch": 0.17300537683883943,
+      "grad_norm": 2.8730116912157517,
+      "learning_rate": 4.674599983408605e-06,
+      "loss": 1.0442,
+      "step": 1645
+    },
+    {
+      "epoch": 0.17311054728068676,
+      "grad_norm": 3.6137100268246387,
+      "learning_rate": 4.674188017384269e-06,
+      "loss": 1.0149,
+      "step": 1646
+    },
+    {
+      "epoch": 0.17321571772253408,
+      "grad_norm": 3.388019574947697,
+      "learning_rate": 4.673775808921078e-06,
+      "loss": 1.0567,
+      "step": 1647
+    },
+    {
+      "epoch": 0.1733208881643814,
+      "grad_norm": 2.7619468248957775,
+      "learning_rate": 4.673363358065e-06,
+      "loss": 0.9992,
+      "step": 1648
+    },
+    {
+      "epoch": 0.17342605860622873,
+      "grad_norm": 3.196115448528693,
+      "learning_rate": 4.672950664862022e-06,
+      "loss": 1.0359,
+      "step": 1649
+    },
+    {
+      "epoch": 0.17353122904807602,
+      "grad_norm": 3.1251623630514054,
+      "learning_rate": 4.672537729358166e-06,
+      "loss": 1.0346,
+      "step": 1650
+    },
+    {
+      "epoch": 0.17363639948992335,
+      "grad_norm": 2.3482592044116024,
+      "learning_rate": 4.672124551599476e-06,
+      "loss": 0.99,
+      "step": 1651
+    },
+    {
+      "epoch": 0.17374156993177067,
+      "grad_norm": 3.2990493587317995,
+      "learning_rate": 4.671711131632025e-06,
+      "loss": 1.0203,
+      "step": 1652
+    },
+    {
+      "epoch": 0.173846740373618,
+      "grad_norm": 2.2221626940474923,
+      "learning_rate": 4.671297469501912e-06,
+      "loss": 1.0141,
+      "step": 1653
+    },
+    {
+      "epoch": 0.17395191081546532,
+      "grad_norm": 3.4954782095000776,
+      "learning_rate": 4.670883565255264e-06,
+      "loss": 1.003,
+      "step": 1654
+    },
+    {
+      "epoch": 0.17405708125731265,
+      "grad_norm": 4.077759847846728,
+      "learning_rate": 4.670469418938235e-06,
+      "loss": 1.04,
+      "step": 1655
+    },
+    {
+      "epoch": 0.17416225169915994,
+      "grad_norm": 3.7217727086743913,
+      "learning_rate": 4.670055030597004e-06,
+      "loss": 1.0416,
+      "step": 1656
+    },
+    {
+      "epoch": 0.17426742214100727,
+      "grad_norm": 2.924757634673209,
+      "learning_rate": 4.66964040027778e-06,
+      "loss": 1.0284,
+      "step": 1657
+    },
+    {
+      "epoch": 0.1743725925828546,
+      "grad_norm": 2.5219450852454037,
+      "learning_rate": 4.669225528026797e-06,
+      "loss": 1.0186,
+      "step": 1658
+    },
+    {
+      "epoch": 0.17447776302470192,
+      "grad_norm": 2.9803154223037223,
+      "learning_rate": 4.668810413890318e-06,
+      "loss": 1.0058,
+      "step": 1659
+    },
+    {
+      "epoch": 0.17458293346654924,
+      "grad_norm": 2.8822355518190004,
+      "learning_rate": 4.668395057914627e-06,
+      "loss": 1.0453,
+      "step": 1660
+    },
+    {
+      "epoch": 0.17468810390839654,
+      "grad_norm": 3.6402941948880283,
+      "learning_rate": 4.667979460146045e-06,
+      "loss": 1.0375,
+      "step": 1661
+    },
+    {
+      "epoch": 0.17479327435024386,
+      "grad_norm": 2.5715274816378746,
+      "learning_rate": 4.6675636206309105e-06,
+      "loss": 1.013,
+      "step": 1662
+    },
+    {
+      "epoch": 0.17489844479209118,
+      "grad_norm": 3.0432398948655734,
+      "learning_rate": 4.667147539415594e-06,
+      "loss": 1.0241,
+      "step": 1663
+    },
+    {
+      "epoch": 0.1750036152339385,
+      "grad_norm": 3.395961606680768,
+      "learning_rate": 4.666731216546492e-06,
+      "loss": 0.9814,
+      "step": 1664
+    },
+    {
+      "epoch": 0.17510878567578583,
+      "grad_norm": 2.863033871229343,
+      "learning_rate": 4.6663146520700275e-06,
+      "loss": 1.0259,
+      "step": 1665
+    },
+    {
+      "epoch": 0.17521395611763313,
+      "grad_norm": 2.905956162367102,
+      "learning_rate": 4.665897846032651e-06,
+      "loss": 1.0223,
+      "step": 1666
+    },
+    {
+      "epoch": 0.17531912655948045,
+      "grad_norm": 2.9044821170497563,
+      "learning_rate": 4.66548079848084e-06,
+      "loss": 1.0568,
+      "step": 1667
+    },
+    {
+      "epoch": 0.17542429700132778,
+      "grad_norm": 2.5251868732416978,
+      "learning_rate": 4.665063509461098e-06,
+      "loss": 1.0291,
+      "step": 1668
+    },
+    {
+      "epoch": 0.1755294674431751,
+      "grad_norm": 1.9745812484711869,
+      "learning_rate": 4.664645979019954e-06,
+      "loss": 1.0088,
+      "step": 1669
+    },
+    {
+      "epoch": 0.17563463788502243,
+      "grad_norm": 2.9571167671583933,
+      "learning_rate": 4.6642282072039694e-06,
+      "loss": 1.038,
+      "step": 1670
+    },
+    {
+      "epoch": 0.17573980832686972,
+      "grad_norm": 3.162873132329408,
+      "learning_rate": 4.663810194059727e-06,
+      "loss": 1.039,
+      "step": 1671
+    },
+    {
+      "epoch": 0.17584497876871705,
+      "grad_norm": 2.707745488716953,
+      "learning_rate": 4.663391939633839e-06,
+      "loss": 1.0412,
+      "step": 1672
+    },
+    {
+      "epoch": 0.17595014921056437,
+      "grad_norm": 2.390831077180832,
+      "learning_rate": 4.662973443972943e-06,
+      "loss": 1.0217,
+      "step": 1673
+    },
+    {
+      "epoch": 0.1760553196524117,
+      "grad_norm": 2.6697672577896103,
+      "learning_rate": 4.662554707123707e-06,
+      "loss": 0.9758,
+      "step": 1674
+    },
+    {
+      "epoch": 0.17616049009425902,
+      "grad_norm": 2.5536432355735417,
+      "learning_rate": 4.662135729132821e-06,
+      "loss": 1.0093,
+      "step": 1675
+    },
+    {
+      "epoch": 0.17626566053610632,
+      "grad_norm": 4.454966321168614,
+      "learning_rate": 4.661716510047005e-06,
+      "loss": 1.0656,
+      "step": 1676
+    },
+    {
+      "epoch": 0.17637083097795364,
+      "grad_norm": 2.9059225841826586,
+      "learning_rate": 4.661297049913005e-06,
+      "loss": 1.054,
+      "step": 1677
+    },
+    {
+      "epoch": 0.17647600141980097,
+      "grad_norm": 2.7078515707870063,
+      "learning_rate": 4.660877348777595e-06,
+      "loss": 0.9475,
+      "step": 1678
+    },
+    {
+      "epoch": 0.1765811718616483,
+      "grad_norm": 2.714819257387345,
+      "learning_rate": 4.660457406687574e-06,
+      "loss": 1.008,
+      "step": 1679
+    },
+    {
+      "epoch": 0.1766863423034956,
+      "grad_norm": 3.560183070646618,
+      "learning_rate": 4.66003722368977e-06,
+      "loss": 1.0123,
+      "step": 1680
+    },
+    {
+      "epoch": 0.1767915127453429,
+      "grad_norm": 4.011571407375856,
+      "learning_rate": 4.659616799831035e-06,
+      "loss": 1.0113,
+      "step": 1681
+    },
+    {
+      "epoch": 0.17689668318719023,
+      "grad_norm": 3.512773712732269,
+      "learning_rate": 4.659196135158251e-06,
+      "loss": 1.0847,
+      "step": 1682
+    },
+    {
+      "epoch": 0.17700185362903756,
+      "grad_norm": 3.5710928774903783,
+      "learning_rate": 4.658775229718323e-06,
+      "loss": 0.9836,
+      "step": 1683
+    },
+    {
+      "epoch": 0.17710702407088488,
+      "grad_norm": 2.576288670084694,
+      "learning_rate": 4.6583540835581885e-06,
+      "loss": 0.9977,
+      "step": 1684
+    },
+    {
+      "epoch": 0.1772121945127322,
+      "grad_norm": 2.67065462027682,
+      "learning_rate": 4.657932696724807e-06,
+      "loss": 1.0019,
+      "step": 1685
+    },
+    {
+      "epoch": 0.1773173649545795,
+      "grad_norm": 3.0304864854926286,
+      "learning_rate": 4.657511069265166e-06,
+      "loss": 1.0513,
+      "step": 1686
+    },
+    {
+      "epoch": 0.17742253539642683,
+      "grad_norm": 2.756757768961673,
+      "learning_rate": 4.6570892012262806e-06,
+      "loss": 1.055,
+      "step": 1687
+    },
+    {
+      "epoch": 0.17752770583827415,
+      "grad_norm": 3.3322689645027306,
+      "learning_rate": 4.656667092655192e-06,
+      "loss": 1.0246,
+      "step": 1688
+    },
+    {
+      "epoch": 0.17763287628012148,
+      "grad_norm": 2.51855374758776,
+      "learning_rate": 4.65624474359897e-06,
+      "loss": 0.9991,
+      "step": 1689
+    },
+    {
+      "epoch": 0.1777380467219688,
+      "grad_norm": 1.793782237595271,
+      "learning_rate": 4.655822154104708e-06,
+      "loss": 0.9936,
+      "step": 1690
+    },
+    {
+      "epoch": 0.1778432171638161,
+      "grad_norm": 2.06409532213173,
+      "learning_rate": 4.655399324219529e-06,
+      "loss": 1.0011,
+      "step": 1691
+    },
+    {
+      "epoch": 0.17794838760566342,
+      "grad_norm": 3.1724819528325843,
+      "learning_rate": 4.654976253990582e-06,
+      "loss": 1.0609,
+      "step": 1692
+    },
+    {
+      "epoch": 0.17805355804751075,
+      "grad_norm": 3.173700150317756,
+      "learning_rate": 4.654552943465042e-06,
+      "loss": 1.0357,
+      "step": 1693
+    },
+    {
+      "epoch": 0.17815872848935807,
+      "grad_norm": 2.630712209918402,
+      "learning_rate": 4.654129392690111e-06,
+      "loss": 1.0445,
+      "step": 1694
+    },
+    {
+      "epoch": 0.1782638989312054,
+      "grad_norm": 3.195364242863823,
+      "learning_rate": 4.653705601713019e-06,
+      "loss": 1.0546,
+      "step": 1695
+    },
+    {
+      "epoch": 0.1783690693730527,
+      "grad_norm": 2.9685233327726737,
+      "learning_rate": 4.653281570581023e-06,
+      "loss": 0.9853,
+      "step": 1696
+    },
+    {
+      "epoch": 0.17847423981490002,
+      "grad_norm": 2.683882737708656,
+      "learning_rate": 4.6528572993414036e-06,
+      "loss": 1.0162,
+      "step": 1697
+    },
+    {
+      "epoch": 0.17857941025674734,
+      "grad_norm": 3.4527829667595067,
+      "learning_rate": 4.652432788041471e-06,
+      "loss": 1.0187,
+      "step": 1698
+    },
+    {
+      "epoch": 0.17868458069859466,
+      "grad_norm": 3.5173119907210735,
+      "learning_rate": 4.652008036728563e-06,
+      "loss": 1.0463,
+      "step": 1699
+    },
+    {
+      "epoch": 0.178789751140442,
+      "grad_norm": 3.388433179291527,
+      "learning_rate": 4.651583045450041e-06,
+      "loss": 1.0248,
+      "step": 1700
+    },
+    {
+      "epoch": 0.17889492158228928,
+      "grad_norm": 2.1485938375897313,
+      "learning_rate": 4.651157814253295e-06,
+      "loss": 1.0051,
+      "step": 1701
+    },
+    {
+      "epoch": 0.1790000920241366,
+      "grad_norm": 3.1044973840896732,
+      "learning_rate": 4.650732343185743e-06,
+      "loss": 1.0274,
+      "step": 1702
+    },
+    {
+      "epoch": 0.17910526246598393,
+      "grad_norm": 2.6481193563942718,
+      "learning_rate": 4.6503066322948264e-06,
+      "loss": 1.0212,
+      "step": 1703
+    },
+    {
+      "epoch": 0.17921043290783126,
+      "grad_norm": 2.9567702455119496,
+      "learning_rate": 4.649880681628016e-06,
+      "loss": 1.0586,
+      "step": 1704
+    },
+    {
+      "epoch": 0.17931560334967858,
+      "grad_norm": 2.847681704947885,
+      "learning_rate": 4.649454491232809e-06,
+      "loss": 1.0353,
+      "step": 1705
+    },
+    {
+      "epoch": 0.17942077379152588,
+      "grad_norm": 2.544316932295647,
+      "learning_rate": 4.649028061156728e-06,
+      "loss": 1.0109,
+      "step": 1706
+    },
+    {
+      "epoch": 0.1795259442333732,
+      "grad_norm": 2.5651350900643863,
+      "learning_rate": 4.648601391447325e-06,
+      "loss": 1.0665,
+      "step": 1707
+    },
+    {
+      "epoch": 0.17963111467522053,
+      "grad_norm": 3.305216400055959,
+      "learning_rate": 4.648174482152176e-06,
+      "loss": 1.0119,
+      "step": 1708
+    },
+    {
+      "epoch": 0.17973628511706785,
+      "grad_norm": 2.907862648072626,
+      "learning_rate": 4.647747333318884e-06,
+      "loss": 1.0051,
+      "step": 1709
+    },
+    {
+      "epoch": 0.17984145555891518,
+      "grad_norm": 2.7406970823992785,
+      "learning_rate": 4.64731994499508e-06,
+      "loss": 0.9843,
+      "step": 1710
+    },
+    {
+      "epoch": 0.17994662600076247,
+      "grad_norm": 2.986361487648303,
+      "learning_rate": 4.646892317228422e-06,
+      "loss": 1.0133,
+      "step": 1711
+    },
+    {
+      "epoch": 0.1800517964426098,
+      "grad_norm": 2.897968238151226,
+      "learning_rate": 4.646464450066592e-06,
+      "loss": 1.0311,
+      "step": 1712
+    },
+    {
+      "epoch": 0.18015696688445712,
+      "grad_norm": 3.4099728159788008,
+      "learning_rate": 4.646036343557302e-06,
+      "loss": 0.9871,
+      "step": 1713
+    },
+    {
+      "epoch": 0.18026213732630444,
+      "grad_norm": 3.7576035434231745,
+      "learning_rate": 4.6456079977482885e-06,
+      "loss": 1.032,
+      "step": 1714
+    },
+    {
+      "epoch": 0.18036730776815177,
+      "grad_norm": 2.569078677988013,
+      "learning_rate": 4.645179412687316e-06,
+      "loss": 1.024,
+      "step": 1715
+    },
+    {
+      "epoch": 0.1804724782099991,
+      "grad_norm": 3.286151517668889,
+      "learning_rate": 4.644750588422174e-06,
+      "loss": 1.062,
+      "step": 1716
+    },
+    {
+      "epoch": 0.1805776486518464,
+      "grad_norm": 3.179805762126629,
+      "learning_rate": 4.644321525000681e-06,
+      "loss": 1.0809,
+      "step": 1717
+    },
+    {
+      "epoch": 0.18068281909369371,
+      "grad_norm": 2.154906179564819,
+      "learning_rate": 4.64389222247068e-06,
+      "loss": 1.0338,
+      "step": 1718
+    },
+    {
+      "epoch": 0.18078798953554104,
+      "grad_norm": 3.4239950119409195,
+      "learning_rate": 4.643462680880042e-06,
+      "loss": 1.0273,
+      "step": 1719
+    },
+    {
+      "epoch": 0.18089315997738836,
+      "grad_norm": 2.9055090835022632,
+      "learning_rate": 4.643032900276664e-06,
+      "loss": 1.0144,
+      "step": 1720
+    },
+    {
+      "epoch": 0.1809983304192357,
+      "grad_norm": 2.4602189314080776,
+      "learning_rate": 4.642602880708469e-06,
+      "loss": 0.9683,
+      "step": 1721
+    },
+    {
+      "epoch": 0.18110350086108298,
+      "grad_norm": 2.90218657552329,
+      "learning_rate": 4.642172622223409e-06,
+      "loss": 1.0738,
+      "step": 1722
+    },
+    {
+      "epoch": 0.1812086713029303,
+      "grad_norm": 2.3552760906435983,
+      "learning_rate": 4.641742124869461e-06,
+      "loss": 0.9947,
+      "step": 1723
+    },
+    {
+      "epoch": 0.18131384174477763,
+      "grad_norm": 2.8906978212652032,
+      "learning_rate": 4.641311388694629e-06,
+      "loss": 1.0405,
+      "step": 1724
+    },
+    {
+      "epoch": 0.18141901218662496,
+      "grad_norm": 2.2774591530752653,
+      "learning_rate": 4.640880413746942e-06,
+      "loss": 1.0631,
+      "step": 1725
+    },
+    {
+      "epoch": 0.18152418262847228,
+      "grad_norm": 2.909505385998881,
+      "learning_rate": 4.640449200074459e-06,
+      "loss": 1.0374,
+      "step": 1726
+    },
+    {
+      "epoch": 0.18162935307031958,
+      "grad_norm": 2.2639846101073484,
+      "learning_rate": 4.6400177477252615e-06,
+      "loss": 1.0135,
+      "step": 1727
+    },
+    {
+      "epoch": 0.1817345235121669,
+      "grad_norm": 3.409130796181564,
+      "learning_rate": 4.639586056747461e-06,
+      "loss": 1.0132,
+      "step": 1728
+    },
+    {
+      "epoch": 0.18183969395401423,
+      "grad_norm": 2.1982637333179915,
+      "learning_rate": 4.639154127189195e-06,
+      "loss": 0.989,
+      "step": 1729
+    },
+    {
+      "epoch": 0.18194486439586155,
+      "grad_norm": 3.8360953153189,
+      "learning_rate": 4.638721959098626e-06,
+      "loss": 1.0704,
+      "step": 1730
+    },
+    {
+      "epoch": 0.18205003483770887,
+      "grad_norm": 2.4398191847689934,
+      "learning_rate": 4.638289552523944e-06,
+      "loss": 1.0277,
+      "step": 1731
+    },
+    {
+      "epoch": 0.18215520527955617,
+      "grad_norm": 2.6670166772713646,
+      "learning_rate": 4.637856907513366e-06,
+      "loss": 1.0189,
+      "step": 1732
+    },
+    {
+      "epoch": 0.1822603757214035,
+      "grad_norm": 2.6746238004492495,
+      "learning_rate": 4.637424024115136e-06,
+      "loss": 1.0083,
+      "step": 1733
+    },
+    {
+      "epoch": 0.18236554616325082,
+      "grad_norm": 4.189496479103646,
+      "learning_rate": 4.636990902377523e-06,
+      "loss": 1.0884,
+      "step": 1734
+    },
+    {
+      "epoch": 0.18247071660509814,
+      "grad_norm": 3.617312839139946,
+      "learning_rate": 4.636557542348823e-06,
+      "loss": 1.02,
+      "step": 1735
+    },
+    {
+      "epoch": 0.18257588704694547,
+      "grad_norm": 2.7303718078659025,
+      "learning_rate": 4.6361239440773595e-06,
+      "loss": 1.033,
+      "step": 1736
+    },
+    {
+      "epoch": 0.18268105748879276,
+      "grad_norm": 2.7591255551753426,
+      "learning_rate": 4.635690107611483e-06,
+      "loss": 1.0249,
+      "step": 1737
+    },
+    {
+      "epoch": 0.1827862279306401,
+      "grad_norm": 2.698775360170187,
+      "learning_rate": 4.635256032999569e-06,
+      "loss": 1.0181,
+      "step": 1738
+    },
+    {
+      "epoch": 0.1828913983724874,
+      "grad_norm": 2.9039375552443776,
+      "learning_rate": 4.63482172029002e-06,
+      "loss": 1.0174,
+      "step": 1739
+    },
+    {
+      "epoch": 0.18299656881433474,
+      "grad_norm": 2.819682186528183,
+      "learning_rate": 4.6343871695312646e-06,
+      "loss": 0.9973,
+      "step": 1740
+    },
+    {
+      "epoch": 0.18310173925618206,
+      "grad_norm": 2.752650866563116,
+      "learning_rate": 4.63395238077176e-06,
+      "loss": 0.9959,
+      "step": 1741
+    },
+    {
+      "epoch": 0.18320690969802936,
+      "grad_norm": 2.037156561103997,
+      "learning_rate": 4.6335173540599875e-06,
+      "loss": 1.0232,
+      "step": 1742
+    },
+    {
+      "epoch": 0.18331208013987668,
+      "grad_norm": 2.954334928188035,
+      "learning_rate": 4.633082089444457e-06,
+      "loss": 0.9881,
+      "step": 1743
+    },
+    {
+      "epoch": 0.183417250581724,
+      "grad_norm": 2.7393783209030946,
+      "learning_rate": 4.632646586973702e-06,
+      "loss": 1.0439,
+      "step": 1744
+    },
+    {
+      "epoch": 0.18352242102357133,
+      "grad_norm": 3.1124274526696385,
+      "learning_rate": 4.6322108466962865e-06,
+      "loss": 0.9901,
+      "step": 1745
+    },
+    {
+      "epoch": 0.18362759146541865,
+      "grad_norm": 2.827497506273816,
+      "learning_rate": 4.631774868660798e-06,
+      "loss": 1.0557,
+      "step": 1746
+    },
+    {
+      "epoch": 0.18373276190726595,
+      "grad_norm": 2.7113839252801495,
+      "learning_rate": 4.631338652915851e-06,
+      "loss": 1.0363,
+      "step": 1747
+    },
+    {
+      "epoch": 0.18383793234911328,
+      "grad_norm": 3.0202399372617896,
+      "learning_rate": 4.6309021995100875e-06,
+      "loss": 0.9892,
+      "step": 1748
+    },
+    {
+      "epoch": 0.1839431027909606,
+      "grad_norm": 3.0853410173378117,
+      "learning_rate": 4.630465508492176e-06,
+      "loss": 0.9995,
+      "step": 1749
+    },
+    {
+      "epoch": 0.18404827323280792,
+      "grad_norm": 3.8771840606848347,
+      "learning_rate": 4.630028579910809e-06,
+      "loss": 1.0325,
+      "step": 1750
+    },
+    {
+      "epoch": 0.18415344367465525,
+      "grad_norm": 2.924444671474204,
+      "learning_rate": 4.629591413814709e-06,
+      "loss": 1.0269,
+      "step": 1751
+    },
+    {
+      "epoch": 0.18425861411650254,
+      "grad_norm": 3.855054949681042,
+      "learning_rate": 4.629154010252624e-06,
+      "loss": 1.0047,
+      "step": 1752
+    },
+    {
+      "epoch": 0.18436378455834987,
+      "grad_norm": 2.94820330025766,
+      "learning_rate": 4.628716369273326e-06,
+      "loss": 1.0513,
+      "step": 1753
+    },
+    {
+      "epoch": 0.1844689550001972,
+      "grad_norm": 3.1410382438209057,
+      "learning_rate": 4.628278490925617e-06,
+      "loss": 1.0178,
+      "step": 1754
+    },
+    {
+      "epoch": 0.18457412544204452,
+      "grad_norm": 3.430775164691166,
+      "learning_rate": 4.6278403752583235e-06,
+      "loss": 1.0179,
+      "step": 1755
+    },
+    {
+      "epoch": 0.18467929588389184,
+      "grad_norm": 3.935774284929484,
+      "learning_rate": 4.627402022320298e-06,
+      "loss": 1.0026,
+      "step": 1756
+    },
+    {
+      "epoch": 0.18478446632573914,
+      "grad_norm": 4.002646697081782,
+      "learning_rate": 4.626963432160421e-06,
+      "loss": 1.0347,
+      "step": 1757
+    },
+    {
+      "epoch": 0.18488963676758646,
+      "grad_norm": 2.5892402072879275,
+      "learning_rate": 4.626524604827598e-06,
+      "loss": 1.0222,
+      "step": 1758
+    },
+    {
+      "epoch": 0.1849948072094338,
+      "grad_norm": 3.069956064773937,
+      "learning_rate": 4.6260855403707625e-06,
+      "loss": 1.0151,
+      "step": 1759
+    },
+    {
+      "epoch": 0.1850999776512811,
+      "grad_norm": 2.7469752838222714,
+      "learning_rate": 4.625646238838873e-06,
+      "loss": 1.0214,
+      "step": 1760
+    },
+    {
+      "epoch": 0.18520514809312844,
+      "grad_norm": 2.5645297933313875,
+      "learning_rate": 4.6252067002809155e-06,
+      "loss": 1.007,
+      "step": 1761
+    },
+    {
+      "epoch": 0.18531031853497573,
+      "grad_norm": 3.4751336669098194,
+      "learning_rate": 4.6247669247459015e-06,
+      "loss": 1.0049,
+      "step": 1762
+    },
+    {
+      "epoch": 0.18541548897682306,
+      "grad_norm": 2.641984983795442,
+      "learning_rate": 4.62432691228287e-06,
+      "loss": 0.9999,
+      "step": 1763
+    },
+    {
+      "epoch": 0.18552065941867038,
+      "grad_norm": 2.4784360351295875,
+      "learning_rate": 4.623886662940885e-06,
+      "loss": 0.9663,
+      "step": 1764
+    },
+    {
+      "epoch": 0.1856258298605177,
+      "grad_norm": 3.7083059171189814,
+      "learning_rate": 4.6234461767690384e-06,
+      "loss": 0.9815,
+      "step": 1765
+    },
+    {
+      "epoch": 0.18573100030236503,
+      "grad_norm": 3.085101182950583,
+      "learning_rate": 4.623005453816447e-06,
+      "loss": 1.0183,
+      "step": 1766
+    },
+    {
+      "epoch": 0.18583617074421233,
+      "grad_norm": 3.8717416513609164,
+      "learning_rate": 4.622564494132256e-06,
+      "loss": 1.0394,
+      "step": 1767
+    },
+    {
+      "epoch": 0.18594134118605965,
+      "grad_norm": 2.6220312567865682,
+      "learning_rate": 4.622123297765636e-06,
+      "loss": 0.9954,
+      "step": 1768
+    },
+    {
+      "epoch": 0.18604651162790697,
+      "grad_norm": 3.2456038569135313,
+      "learning_rate": 4.621681864765783e-06,
+      "loss": 1.0068,
+      "step": 1769
+    },
+    {
+      "epoch": 0.1861516820697543,
+      "grad_norm": 2.99049524146547,
+      "learning_rate": 4.621240195181918e-06,
+      "loss": 1.0466,
+      "step": 1770
+    },
+    {
+      "epoch": 0.18625685251160162,
+      "grad_norm": 3.7023703331482074,
+      "learning_rate": 4.620798289063295e-06,
+      "loss": 1.028,
+      "step": 1771
+    },
+    {
+      "epoch": 0.18636202295344895,
+      "grad_norm": 2.284086379818769,
+      "learning_rate": 4.620356146459187e-06,
+      "loss": 1.0119,
+      "step": 1772
+    },
+    {
+      "epoch": 0.18646719339529624,
+      "grad_norm": 3.7163945954226514,
+      "learning_rate": 4.619913767418898e-06,
+      "loss": 1.0508,
+      "step": 1773
+    },
+    {
+      "epoch": 0.18657236383714357,
+      "grad_norm": 3.8367678119936324,
+      "learning_rate": 4.619471151991755e-06,
+      "loss": 1.0105,
+      "step": 1774
+    },
+    {
+      "epoch": 0.1866775342789909,
+      "grad_norm": 2.8688910090189923,
+      "learning_rate": 4.619028300227114e-06,
+      "loss": 1.0038,
+      "step": 1775
+    },
+    {
+      "epoch": 0.18678270472083822,
+      "grad_norm": 2.650967656775131,
+      "learning_rate": 4.618585212174357e-06,
+      "loss": 0.9438,
+      "step": 1776
+    },
+    {
+      "epoch": 0.18688787516268554,
+      "grad_norm": 4.345745813091422,
+      "learning_rate": 4.618141887882891e-06,
+      "loss": 1.025,
+      "step": 1777
+    },
+    {
+      "epoch": 0.18699304560453284,
+      "grad_norm": 2.868036629003295,
+      "learning_rate": 4.6176983274021495e-06,
+      "loss": 1.0249,
+      "step": 1778
+    },
+    {
+      "epoch": 0.18709821604638016,
+      "grad_norm": 2.372708709831453,
+      "learning_rate": 4.617254530781594e-06,
+      "loss": 1.0195,
+      "step": 1779
+    },
+    {
+      "epoch": 0.18720338648822749,
+      "grad_norm": 3.358894346037336,
+      "learning_rate": 4.6168104980707105e-06,
+      "loss": 1.0665,
+      "step": 1780
+    },
+    {
+      "epoch": 0.1873085569300748,
+      "grad_norm": 3.3819142018228643,
+      "learning_rate": 4.6163662293190135e-06,
+      "loss": 1.0316,
+      "step": 1781
+    },
+    {
+      "epoch": 0.18741372737192213,
+      "grad_norm": 2.7812505645074777,
+      "learning_rate": 4.61592172457604e-06,
+      "loss": 1.0267,
+      "step": 1782
+    },
+    {
+      "epoch": 0.18751889781376943,
+      "grad_norm": 2.666048271062157,
+      "learning_rate": 4.615476983891359e-06,
+      "loss": 1.0273,
+      "step": 1783
+    },
+    {
+      "epoch": 0.18762406825561675,
+      "grad_norm": 2.8496484883881994,
+      "learning_rate": 4.61503200731456e-06,
+      "loss": 0.9927,
+      "step": 1784
+    },
+    {
+      "epoch": 0.18772923869746408,
+      "grad_norm": 2.604297527248957,
+      "learning_rate": 4.6145867948952605e-06,
+      "loss": 1.0363,
+      "step": 1785
+    },
+    {
+      "epoch": 0.1878344091393114,
+      "grad_norm": 2.25491398075674,
+      "learning_rate": 4.614141346683109e-06,
+      "loss": 1.038,
+      "step": 1786
+    },
+    {
+      "epoch": 0.18793957958115873,
+      "grad_norm": 2.264636873435953,
+      "learning_rate": 4.6136956627277725e-06,
+      "loss": 0.9992,
+      "step": 1787
+    },
+    {
+      "epoch": 0.18804475002300602,
+      "grad_norm": 2.704549825776999,
+      "learning_rate": 4.61324974307895e-06,
+      "loss": 1.0082,
+      "step": 1788
+    },
+    {
+      "epoch": 0.18814992046485335,
+      "grad_norm": 2.353105958418805,
+      "learning_rate": 4.612803587786366e-06,
+      "loss": 1.0102,
+      "step": 1789
+    },
+    {
+      "epoch": 0.18825509090670067,
+      "grad_norm": 2.948441358243538,
+      "learning_rate": 4.612357196899768e-06,
+      "loss": 1.033,
+      "step": 1790
+    },
+    {
+      "epoch": 0.188360261348548,
+      "grad_norm": 1.9059658097973773,
+      "learning_rate": 4.611910570468933e-06,
+      "loss": 1.002,
+      "step": 1791
+    },
+    {
+      "epoch": 0.18846543179039532,
+      "grad_norm": 2.6032758990156712,
+      "learning_rate": 4.611463708543665e-06,
+      "loss": 1.0658,
+      "step": 1792
+    },
+    {
+      "epoch": 0.18857060223224262,
+      "grad_norm": 2.230556596976156,
+      "learning_rate": 4.61101661117379e-06,
+      "loss": 1.0214,
+      "step": 1793
+    },
+    {
+      "epoch": 0.18867577267408994,
+      "grad_norm": 3.0702976065770664,
+      "learning_rate": 4.610569278409164e-06,
+      "loss": 1.0007,
+      "step": 1794
+    },
+    {
+      "epoch": 0.18878094311593727,
+      "grad_norm": 3.007760617807363,
+      "learning_rate": 4.610121710299668e-06,
+      "loss": 1.0181,
+      "step": 1795
+    },
+    {
+      "epoch": 0.1888861135577846,
+      "grad_norm": 2.7344843254797127,
+      "learning_rate": 4.609673906895208e-06,
+      "loss": 1.0361,
+      "step": 1796
+    },
+    {
+      "epoch": 0.18899128399963191,
+      "grad_norm": 1.9281901603612084,
+      "learning_rate": 4.609225868245721e-06,
+      "loss": 1.0006,
+      "step": 1797
+    },
+    {
+      "epoch": 0.1890964544414792,
+      "grad_norm": 3.2114203189511152,
+      "learning_rate": 4.608777594401164e-06,
+      "loss": 1.0043,
+      "step": 1798
+    },
+    {
+      "epoch": 0.18920162488332654,
+      "grad_norm": 2.145414047326945,
+      "learning_rate": 4.608329085411523e-06,
+      "loss": 1.0424,
+      "step": 1799
+    },
+    {
+      "epoch": 0.18930679532517386,
+      "grad_norm": 2.51024154366095,
+      "learning_rate": 4.607880341326812e-06,
+      "loss": 1.0284,
+      "step": 1800
+    },
+    {
+      "epoch": 0.18941196576702118,
+      "grad_norm": 3.0454022502767804,
+      "learning_rate": 4.607431362197067e-06,
+      "loss": 0.9817,
+      "step": 1801
+    },
+    {
+      "epoch": 0.1895171362088685,
+      "grad_norm": 3.050295307561754,
+      "learning_rate": 4.6069821480723545e-06,
+      "loss": 1.0539,
+      "step": 1802
+    },
+    {
+      "epoch": 0.1896223066507158,
+      "grad_norm": 4.406095657800138,
+      "learning_rate": 4.6065326990027656e-06,
+      "loss": 1.0129,
+      "step": 1803
+    },
+    {
+      "epoch": 0.18972747709256313,
+      "grad_norm": 2.37443582116452,
+      "learning_rate": 4.606083015038416e-06,
+      "loss": 0.9949,
+      "step": 1804
+    },
+    {
+      "epoch": 0.18983264753441045,
+      "grad_norm": 3.2183447058232253,
+      "learning_rate": 4.6056330962294496e-06,
+      "loss": 1.0636,
+      "step": 1805
+    },
+    {
+      "epoch": 0.18993781797625778,
+      "grad_norm": 2.8177933591262407,
+      "learning_rate": 4.605182942626037e-06,
+      "loss": 1.0327,
+      "step": 1806
+    },
+    {
+      "epoch": 0.1900429884181051,
+      "grad_norm": 2.3566286140592654,
+      "learning_rate": 4.604732554278371e-06,
+      "loss": 1.0276,
+      "step": 1807
+    },
+    {
+      "epoch": 0.1901481588599524,
+      "grad_norm": 3.0654404715127024,
+      "learning_rate": 4.604281931236675e-06,
+      "loss": 1.0586,
+      "step": 1808
+    },
+    {
+      "epoch": 0.19025332930179972,
+      "grad_norm": 2.099265216978414,
+      "learning_rate": 4.6038310735511985e-06,
+      "loss": 1.0363,
+      "step": 1809
+    },
+    {
+      "epoch": 0.19035849974364705,
+      "grad_norm": 2.521236583898647,
+      "learning_rate": 4.603379981272213e-06,
+      "loss": 0.9839,
+      "step": 1810
+    },
+    {
+      "epoch": 0.19046367018549437,
+      "grad_norm": 2.528668148495521,
+      "learning_rate": 4.6029286544500205e-06,
+      "loss": 1.009,
+      "step": 1811
+    },
+    {
+      "epoch": 0.1905688406273417,
+      "grad_norm": 3.7980630261576125,
+      "learning_rate": 4.602477093134947e-06,
+      "loss": 0.9996,
+      "step": 1812
+    },
+    {
+      "epoch": 0.190674011069189,
+      "grad_norm": 2.965801661136027,
+      "learning_rate": 4.602025297377345e-06,
+      "loss": 1.0368,
+      "step": 1813
+    },
+    {
+      "epoch": 0.19077918151103632,
+      "grad_norm": 2.5523632367350104,
+      "learning_rate": 4.6015732672275925e-06,
+      "loss": 0.9936,
+      "step": 1814
+    },
+    {
+      "epoch": 0.19088435195288364,
+      "grad_norm": 3.02733722515206,
+      "learning_rate": 4.601121002736095e-06,
+      "loss": 1.0129,
+      "step": 1815
+    },
+    {
+      "epoch": 0.19098952239473097,
+      "grad_norm": 4.357521284066693,
+      "learning_rate": 4.600668503953285e-06,
+      "loss": 1.0212,
+      "step": 1816
+    },
+    {
+      "epoch": 0.1910946928365783,
+      "grad_norm": 3.1172216407619318,
+      "learning_rate": 4.600215770929617e-06,
+      "loss": 1.093,
+      "step": 1817
+    },
+    {
+      "epoch": 0.19119986327842559,
+      "grad_norm": 2.6332985511775493,
+      "learning_rate": 4.599762803715576e-06,
+      "loss": 1.0391,
+      "step": 1818
+    },
+    {
+      "epoch": 0.1913050337202729,
+      "grad_norm": 2.569865786330344,
+      "learning_rate": 4.599309602361671e-06,
+      "loss": 1.0291,
+      "step": 1819
+    },
+    {
+      "epoch": 0.19141020416212023,
+      "grad_norm": 3.2827113014482445,
+      "learning_rate": 4.5988561669184376e-06,
+      "loss": 1.0382,
+      "step": 1820
+    },
+    {
+      "epoch": 0.19151537460396756,
+      "grad_norm": 2.1719148645999584,
+      "learning_rate": 4.598402497436436e-06,
+      "loss": 1.0192,
+      "step": 1821
+    },
+    {
+      "epoch": 0.19162054504581488,
+      "grad_norm": 2.2765741322316546,
+      "learning_rate": 4.597948593966256e-06,
+      "loss": 1.0286,
+      "step": 1822
+    },
+    {
+      "epoch": 0.19172571548766218,
+      "grad_norm": 4.114857140083318,
+      "learning_rate": 4.59749445655851e-06,
+      "loss": 1.045,
+      "step": 1823
+    },
+    {
+      "epoch": 0.1918308859295095,
+      "grad_norm": 2.997696029602926,
+      "learning_rate": 4.597040085263838e-06,
+      "loss": 1.04,
+      "step": 1824
+    },
+    {
+      "epoch": 0.19193605637135683,
+      "grad_norm": 3.047085040895484,
+      "learning_rate": 4.596585480132906e-06,
+      "loss": 1.0447,
+      "step": 1825
+    },
+    {
+      "epoch": 0.19204122681320415,
+      "grad_norm": 3.9271378946979847,
+      "learning_rate": 4.596130641216406e-06,
+      "loss": 1.008,
+      "step": 1826
+    },
+    {
+      "epoch": 0.19214639725505148,
+      "grad_norm": 2.0196952752323414,
+      "learning_rate": 4.595675568565058e-06,
+      "loss": 0.9986,
+      "step": 1827
+    },
+    {
+      "epoch": 0.19225156769689877,
+      "grad_norm": 3.228450782508349,
+      "learning_rate": 4.5952202622296015e-06,
+      "loss": 1.0224,
+      "step": 1828
+    },
+    {
+      "epoch": 0.1923567381387461,
+      "grad_norm": 3.398130308359955,
+      "learning_rate": 4.594764722260812e-06,
+      "loss": 1.0222,
+      "step": 1829
+    },
+    {
+      "epoch": 0.19246190858059342,
+      "grad_norm": 2.3897945567374648,
+      "learning_rate": 4.594308948709482e-06,
+      "loss": 1.0258,
+      "step": 1830
+    },
+    {
+      "epoch": 0.19256707902244075,
+      "grad_norm": 2.867204854790798,
+      "learning_rate": 4.593852941626435e-06,
+      "loss": 0.9945,
+      "step": 1831
+    },
+    {
+      "epoch": 0.19267224946428807,
+      "grad_norm": 3.1790669432232317,
+      "learning_rate": 4.59339670106252e-06,
+      "loss": 1.0426,
+      "step": 1832
+    },
+    {
+      "epoch": 0.1927774199061354,
+      "grad_norm": 3.0542457539999694,
+      "learning_rate": 4.59294022706861e-06,
+      "loss": 1.0483,
+      "step": 1833
+    },
+    {
+      "epoch": 0.1928825903479827,
+      "grad_norm": 3.2664531106906685,
+      "learning_rate": 4.592483519695606e-06,
+      "loss": 0.989,
+      "step": 1834
+    },
+    {
+      "epoch": 0.19298776078983002,
+      "grad_norm": 2.4685050890530436,
+      "learning_rate": 4.592026578994435e-06,
+      "loss": 1.0167,
+      "step": 1835
+    },
+    {
+      "epoch": 0.19309293123167734,
+      "grad_norm": 2.9309814017337197,
+      "learning_rate": 4.59156940501605e-06,
+      "loss": 1.0402,
+      "step": 1836
+    },
+    {
+      "epoch": 0.19319810167352466,
+      "grad_norm": 3.121589794261374,
+      "learning_rate": 4.591111997811427e-06,
+      "loss": 1.0245,
+      "step": 1837
+    },
+    {
+      "epoch": 0.193303272115372,
+      "grad_norm": 3.620780997081411,
+      "learning_rate": 4.590654357431573e-06,
+      "loss": 1.0185,
+      "step": 1838
+    },
+    {
+      "epoch": 0.19340844255721928,
+      "grad_norm": 2.5408636903460775,
+      "learning_rate": 4.590196483927517e-06,
+      "loss": 1.0493,
+      "step": 1839
+    },
+    {
+      "epoch": 0.1935136129990666,
+      "grad_norm": 2.642817968832206,
+      "learning_rate": 4.589738377350316e-06,
+      "loss": 1.0139,
+      "step": 1840
+    },
+    {
+      "epoch": 0.19361878344091393,
+      "grad_norm": 2.461137436664243,
+      "learning_rate": 4.589280037751052e-06,
+      "loss": 1.048,
+      "step": 1841
+    },
+    {
+      "epoch": 0.19372395388276126,
+      "grad_norm": 2.2490071042179625,
+      "learning_rate": 4.5888214651808325e-06,
+      "loss": 1.014,
+      "step": 1842
+    },
+    {
+      "epoch": 0.19382912432460858,
+      "grad_norm": 2.4662682458707255,
+      "learning_rate": 4.5883626596907945e-06,
+      "loss": 1.0227,
+      "step": 1843
+    },
+    {
+      "epoch": 0.19393429476645588,
+      "grad_norm": 4.025300912872622,
+      "learning_rate": 4.587903621332097e-06,
+      "loss": 1.0394,
+      "step": 1844
+    },
+    {
+      "epoch": 0.1940394652083032,
+      "grad_norm": 3.0814943588959722,
+      "learning_rate": 4.5874443501559265e-06,
+      "loss": 1.0244,
+      "step": 1845
+    },
+    {
+      "epoch": 0.19414463565015053,
+      "grad_norm": 3.5317457168193074,
+      "learning_rate": 4.586984846213494e-06,
+      "loss": 1.0279,
+      "step": 1846
+    },
+    {
+      "epoch": 0.19424980609199785,
+      "grad_norm": 3.5877904213974934,
+      "learning_rate": 4.586525109556039e-06,
+      "loss": 1.0146,
+      "step": 1847
+    },
+    {
+      "epoch": 0.19435497653384518,
+      "grad_norm": 3.2081496657086075,
+      "learning_rate": 4.586065140234827e-06,
+      "loss": 1.022,
+      "step": 1848
+    },
+    {
+      "epoch": 0.19446014697569247,
+      "grad_norm": 2.0788025293634678,
+      "learning_rate": 4.585604938301146e-06,
+      "loss": 1.0051,
+      "step": 1849
+    },
+    {
+      "epoch": 0.1945653174175398,
+      "grad_norm": 2.894336944770527,
+      "learning_rate": 4.585144503806312e-06,
+      "loss": 1.0676,
+      "step": 1850
+    },
+    {
+      "epoch": 0.19467048785938712,
+      "grad_norm": 2.877696747200756,
+      "learning_rate": 4.584683836801669e-06,
+      "loss": 1.0292,
+      "step": 1851
+    },
+    {
+      "epoch": 0.19477565830123444,
+      "grad_norm": 4.55585031976607,
+      "learning_rate": 4.584222937338584e-06,
+      "loss": 1.0615,
+      "step": 1852
+    },
+    {
+      "epoch": 0.19488082874308177,
+      "grad_norm": 1.7959609995555694,
+      "learning_rate": 4.583761805468449e-06,
+      "loss": 0.967,
+      "step": 1853
+    },
+    {
+      "epoch": 0.19498599918492907,
+      "grad_norm": 2.742630215241344,
+      "learning_rate": 4.583300441242688e-06,
+      "loss": 0.9956,
+      "step": 1854
+    },
+    {
+      "epoch": 0.1950911696267764,
+      "grad_norm": 2.6981352318462974,
+      "learning_rate": 4.582838844712741e-06,
+      "loss": 1.0238,
+      "step": 1855
+    },
+    {
+      "epoch": 0.1951963400686237,
+      "grad_norm": 3.463326781342451,
+      "learning_rate": 4.582377015930085e-06,
+      "loss": 1.0759,
+      "step": 1856
+    },
+    {
+      "epoch": 0.19530151051047104,
+      "grad_norm": 2.4494104426262138,
+      "learning_rate": 4.581914954946215e-06,
+      "loss": 1.013,
+      "step": 1857
+    },
+    {
+      "epoch": 0.19540668095231836,
+      "grad_norm": 2.2066591737539865,
+      "learning_rate": 4.581452661812655e-06,
+      "loss": 1.0223,
+      "step": 1858
+    },
+    {
+      "epoch": 0.19551185139416566,
+      "grad_norm": 3.3773728745426803,
+      "learning_rate": 4.5809901365809524e-06,
+      "loss": 1.0246,
+      "step": 1859
+    },
+    {
+      "epoch": 0.19561702183601298,
+      "grad_norm": 2.5954532792428218,
+      "learning_rate": 4.580527379302685e-06,
+      "loss": 1.0264,
+      "step": 1860
+    },
+    {
+      "epoch": 0.1957221922778603,
+      "grad_norm": 2.651276443421835,
+      "learning_rate": 4.580064390029452e-06,
+      "loss": 1.0487,
+      "step": 1861
+    },
+    {
+      "epoch": 0.19582736271970763,
+      "grad_norm": 2.6830451292812834,
+      "learning_rate": 4.579601168812882e-06,
+      "loss": 1.037,
+      "step": 1862
+    },
+    {
+      "epoch": 0.19593253316155496,
+      "grad_norm": 3.2055413372462813,
+      "learning_rate": 4.579137715704626e-06,
+      "loss": 1.0315,
+      "step": 1863
+    },
+    {
+      "epoch": 0.19603770360340225,
+      "grad_norm": 2.144505459852746,
+      "learning_rate": 4.578674030756364e-06,
+      "loss": 0.9836,
+      "step": 1864
+    },
+    {
+      "epoch": 0.19614287404524958,
+      "grad_norm": 2.630253444185954,
+      "learning_rate": 4.578210114019799e-06,
+      "loss": 1.0294,
+      "step": 1865
+    },
+    {
+      "epoch": 0.1962480444870969,
+      "grad_norm": 3.8376888541418306,
+      "learning_rate": 4.577745965546662e-06,
+      "loss": 0.9988,
+      "step": 1866
+    },
+    {
+      "epoch": 0.19635321492894423,
+      "grad_norm": 2.068518479777229,
+      "learning_rate": 4.577281585388711e-06,
+      "loss": 1.0126,
+      "step": 1867
+    },
+    {
+      "epoch": 0.19645838537079155,
+      "grad_norm": 2.4680664224880497,
+      "learning_rate": 4.576816973597725e-06,
+      "loss": 1.013,
+      "step": 1868
+    },
+    {
+      "epoch": 0.19656355581263885,
+      "grad_norm": 3.2990236331625615,
+      "learning_rate": 4.576352130225513e-06,
+      "loss": 1.0228,
+      "step": 1869
+    },
+    {
+      "epoch": 0.19666872625448617,
+      "grad_norm": 2.772397789423621,
+      "learning_rate": 4.5758870553239095e-06,
+      "loss": 0.9804,
+      "step": 1870
+    },
+    {
+      "epoch": 0.1967738966963335,
+      "grad_norm": 3.1931620564914747,
+      "learning_rate": 4.575421748944773e-06,
+      "loss": 1.0534,
+      "step": 1871
+    },
+    {
+      "epoch": 0.19687906713818082,
+      "grad_norm": 2.4312586968940426,
+      "learning_rate": 4.574956211139989e-06,
+      "loss": 1.0221,
+      "step": 1872
+    },
+    {
+      "epoch": 0.19698423758002814,
+      "grad_norm": 1.9836984160196345,
+      "learning_rate": 4.574490441961469e-06,
+      "loss": 1.0625,
+      "step": 1873
+    },
+    {
+      "epoch": 0.19708940802187544,
+      "grad_norm": 2.687082682874049,
+      "learning_rate": 4.57402444146115e-06,
+      "loss": 1.0262,
+      "step": 1874
+    },
+    {
+      "epoch": 0.19719457846372276,
+      "grad_norm": 3.29954290699376,
+      "learning_rate": 4.573558209690993e-06,
+      "loss": 1.0292,
+      "step": 1875
+    },
+    {
+      "epoch": 0.1972997489055701,
+      "grad_norm": 2.5564424315027505,
+      "learning_rate": 4.573091746702988e-06,
+      "loss": 1.0151,
+      "step": 1876
+    },
+    {
+      "epoch": 0.1974049193474174,
+      "grad_norm": 3.4128408599390716,
+      "learning_rate": 4.572625052549149e-06,
+      "loss": 1.0509,
+      "step": 1877
+    },
+    {
+      "epoch": 0.19751008978926474,
+      "grad_norm": 2.754362006103999,
+      "learning_rate": 4.572158127281516e-06,
+      "loss": 0.9928,
+      "step": 1878
+    },
+    {
+      "epoch": 0.19761526023111203,
+      "grad_norm": 2.737808240634724,
+      "learning_rate": 4.571690970952155e-06,
+      "loss": 1.0393,
+      "step": 1879
+    },
+    {
+      "epoch": 0.19772043067295936,
+      "grad_norm": 2.5757192677948813,
+      "learning_rate": 4.571223583613157e-06,
+      "loss": 1.0617,
+      "step": 1880
+    },
+    {
+      "epoch": 0.19782560111480668,
+      "grad_norm": 2.622124067123704,
+      "learning_rate": 4.570755965316639e-06,
+      "loss": 0.9885,
+      "step": 1881
+    },
+    {
+      "epoch": 0.197930771556654,
+      "grad_norm": 2.5421671476785,
+      "learning_rate": 4.570288116114745e-06,
+      "loss": 0.9976,
+      "step": 1882
+    },
+    {
+      "epoch": 0.19803594199850133,
+      "grad_norm": 2.8283659577999583,
+      "learning_rate": 4.569820036059644e-06,
+      "loss": 1.054,
+      "step": 1883
+    },
+    {
+      "epoch": 0.19814111244034863,
+      "grad_norm": 2.4526889476942695,
+      "learning_rate": 4.56935172520353e-06,
+      "loss": 1.0015,
+      "step": 1884
+    },
+    {
+      "epoch": 0.19824628288219595,
+      "grad_norm": 1.8002824330773355,
+      "learning_rate": 4.568883183598623e-06,
+      "loss": 1.0031,
+      "step": 1885
+    },
+    {
+      "epoch": 0.19835145332404328,
+      "grad_norm": 1.7744689934585205,
+      "learning_rate": 4.56841441129717e-06,
+      "loss": 1.0815,
+      "step": 1886
+    },
+    {
+      "epoch": 0.1984566237658906,
+      "grad_norm": 3.315507706913469,
+      "learning_rate": 4.567945408351441e-06,
+      "loss": 1.007,
+      "step": 1887
+    },
+    {
+      "epoch": 0.19856179420773792,
+      "grad_norm": 2.5040098416445953,
+      "learning_rate": 4.567476174813735e-06,
+      "loss": 1.0166,
+      "step": 1888
+    },
+    {
+      "epoch": 0.19866696464958522,
+      "grad_norm": 3.0590946518561544,
+      "learning_rate": 4.567006710736375e-06,
+      "loss": 1.0333,
+      "step": 1889
+    },
+    {
+      "epoch": 0.19877213509143254,
+      "grad_norm": 1.9729874214632432,
+      "learning_rate": 4.5665370161717095e-06,
+      "loss": 1.0099,
+      "step": 1890
+    },
+    {
+      "epoch": 0.19887730553327987,
+      "grad_norm": 2.646128060765525,
+      "learning_rate": 4.566067091172114e-06,
+      "loss": 0.9994,
+      "step": 1891
+    },
+    {
+      "epoch": 0.1989824759751272,
+      "grad_norm": 2.39440777880957,
+      "learning_rate": 4.565596935789987e-06,
+      "loss": 1.0348,
+      "step": 1892
+    },
+    {
+      "epoch": 0.19908764641697452,
+      "grad_norm": 3.5153787075984586,
+      "learning_rate": 4.5651265500777564e-06,
+      "loss": 1.0805,
+      "step": 1893
+    },
+    {
+      "epoch": 0.19919281685882184,
+      "grad_norm": 2.5557058180370955,
+      "learning_rate": 4.564655934087873e-06,
+      "loss": 1.0109,
+      "step": 1894
+    },
+    {
+      "epoch": 0.19929798730066914,
+      "grad_norm": 2.541551289434198,
+      "learning_rate": 4.564185087872814e-06,
+      "loss": 1.0578,
+      "step": 1895
+    },
+    {
+      "epoch": 0.19940315774251646,
+      "grad_norm": 2.222619391054327,
+      "learning_rate": 4.563714011485082e-06,
+      "loss": 1.0541,
+      "step": 1896
+    },
+    {
+      "epoch": 0.1995083281843638,
+      "grad_norm": 3.635232836192664,
+      "learning_rate": 4.563242704977206e-06,
+      "loss": 1.0528,
+      "step": 1897
+    },
+    {
+      "epoch": 0.1996134986262111,
+      "grad_norm": 2.804337831993397,
+      "learning_rate": 4.562771168401742e-06,
+      "loss": 1.033,
+      "step": 1898
+    },
+    {
+      "epoch": 0.19971866906805844,
+      "grad_norm": 2.940386059188817,
+      "learning_rate": 4.562299401811268e-06,
+      "loss": 1.0787,
+      "step": 1899
+    },
+    {
+      "epoch": 0.19982383950990573,
+      "grad_norm": 2.516512929364484,
+      "learning_rate": 4.56182740525839e-06,
+      "loss": 1.0043,
+      "step": 1900
+    },
+    {
+      "epoch": 0.19992900995175306,
+      "grad_norm": 2.8419554559032596,
+      "learning_rate": 4.561355178795739e-06,
+      "loss": 1.0395,
+      "step": 1901
+    },
+    {
+      "epoch": 0.20003418039360038,
+      "grad_norm": 3.203241148860607,
+      "learning_rate": 4.560882722475973e-06,
+      "loss": 1.0354,
+      "step": 1902
+    },
+    {
+      "epoch": 0.2001393508354477,
+      "grad_norm": 2.9994247587115233,
+      "learning_rate": 4.560410036351774e-06,
+      "loss": 1.0102,
+      "step": 1903
+    },
+    {
+      "epoch": 0.20024452127729503,
+      "grad_norm": 2.5106634446501164,
+      "learning_rate": 4.559937120475849e-06,
+      "loss": 1.0177,
+      "step": 1904
+    },
+    {
+      "epoch": 0.20034969171914233,
+      "grad_norm": 3.201430008264517,
+      "learning_rate": 4.559463974900934e-06,
+      "loss": 1.0222,
+      "step": 1905
+    },
+    {
+      "epoch": 0.20045486216098965,
+      "grad_norm": 3.5626303781607938,
+      "learning_rate": 4.558990599679787e-06,
+      "loss": 1.021,
+      "step": 1906
+    },
+    {
+      "epoch": 0.20056003260283697,
+      "grad_norm": 2.3292692746627024,
+      "learning_rate": 4.558516994865194e-06,
+      "loss": 1.0364,
+      "step": 1907
+    },
+    {
+      "epoch": 0.2006652030446843,
+      "grad_norm": 2.747377159004723,
+      "learning_rate": 4.558043160509964e-06,
+      "loss": 1.0101,
+      "step": 1908
+    },
+    {
+      "epoch": 0.20077037348653162,
+      "grad_norm": 2.876129833597873,
+      "learning_rate": 4.557569096666934e-06,
+      "loss": 0.9977,
+      "step": 1909
+    },
+    {
+      "epoch": 0.20087554392837892,
+      "grad_norm": 3.8381922934160744,
+      "learning_rate": 4.5570948033889675e-06,
+      "loss": 1.0311,
+      "step": 1910
+    },
+    {
+      "epoch": 0.20098071437022624,
+      "grad_norm": 2.5502637033749447,
+      "learning_rate": 4.5566202807289485e-06,
+      "loss": 1.0031,
+      "step": 1911
+    },
+    {
+      "epoch": 0.20108588481207357,
+      "grad_norm": 2.07505839155443,
+      "learning_rate": 4.556145528739793e-06,
+      "loss": 1.0599,
+      "step": 1912
+    },
+    {
+      "epoch": 0.2011910552539209,
+      "grad_norm": 2.8177451858148355,
+      "learning_rate": 4.555670547474438e-06,
+      "loss": 1.0686,
+      "step": 1913
+    },
+    {
+      "epoch": 0.20129622569576822,
+      "grad_norm": 4.173041129558708,
+      "learning_rate": 4.555195336985848e-06,
+      "loss": 1.0301,
+      "step": 1914
+    },
+    {
+      "epoch": 0.2014013961376155,
+      "grad_norm": 2.552074293804037,
+      "learning_rate": 4.554719897327013e-06,
+      "loss": 1.0075,
+      "step": 1915
+    },
+    {
+      "epoch": 0.20150656657946284,
+      "grad_norm": 3.3799408110891975,
+      "learning_rate": 4.554244228550947e-06,
+      "loss": 1.0219,
+      "step": 1916
+    },
+    {
+      "epoch": 0.20161173702131016,
+      "grad_norm": 2.9673353024375415,
+      "learning_rate": 4.5537683307106924e-06,
+      "loss": 1.0484,
+      "step": 1917
+    },
+    {
+      "epoch": 0.20171690746315749,
+      "grad_norm": 2.8301509680066084,
+      "learning_rate": 4.553292203859314e-06,
+      "loss": 1.0586,
+      "step": 1918
+    },
+    {
+      "epoch": 0.2018220779050048,
+      "grad_norm": 3.757513825220459,
+      "learning_rate": 4.552815848049904e-06,
+      "loss": 0.9888,
+      "step": 1919
+    },
+    {
+      "epoch": 0.2019272483468521,
+      "grad_norm": 2.158466766123102,
+      "learning_rate": 4.552339263335581e-06,
+      "loss": 1.0265,
+      "step": 1920
+    },
+    {
+      "epoch": 0.20203241878869943,
+      "grad_norm": 4.206006355483029,
+      "learning_rate": 4.551862449769487e-06,
+      "loss": 1.0384,
+      "step": 1921
+    },
+    {
+      "epoch": 0.20213758923054675,
+      "grad_norm": 2.713465885037005,
+      "learning_rate": 4.5513854074047905e-06,
+      "loss": 1.0314,
+      "step": 1922
+    },
+    {
+      "epoch": 0.20224275967239408,
+      "grad_norm": 2.159072474282689,
+      "learning_rate": 4.550908136294685e-06,
+      "loss": 1.0454,
+      "step": 1923
+    },
+    {
+      "epoch": 0.2023479301142414,
+      "grad_norm": 2.5490149894200846,
+      "learning_rate": 4.55043063649239e-06,
+      "loss": 1.0143,
+      "step": 1924
+    },
+    {
+      "epoch": 0.2024531005560887,
+      "grad_norm": 2.9457442371208815,
+      "learning_rate": 4.549952908051151e-06,
+      "loss": 1.0303,
+      "step": 1925
+    },
+    {
+      "epoch": 0.20255827099793602,
+      "grad_norm": 3.717268820987839,
+      "learning_rate": 4.549474951024238e-06,
+      "loss": 1.0308,
+      "step": 1926
+    },
+    {
+      "epoch": 0.20266344143978335,
+      "grad_norm": 2.6781806036533222,
+      "learning_rate": 4.548996765464947e-06,
+      "loss": 1.0291,
+      "step": 1927
+    },
+    {
+      "epoch": 0.20276861188163067,
+      "grad_norm": 2.18546757845666,
+      "learning_rate": 4.5485183514266004e-06,
+      "loss": 1.0024,
+      "step": 1928
+    },
+    {
+      "epoch": 0.202873782323478,
+      "grad_norm": 2.510233734668508,
+      "learning_rate": 4.548039708962544e-06,
+      "loss": 1.0261,
+      "step": 1929
+    },
+    {
+      "epoch": 0.2029789527653253,
+      "grad_norm": 2.704997908914001,
+      "learning_rate": 4.547560838126149e-06,
+      "loss": 0.9926,
+      "step": 1930
+    },
+    {
+      "epoch": 0.20308412320717262,
+      "grad_norm": 2.5107441763705647,
+      "learning_rate": 4.5470817389708155e-06,
+      "loss": 1.0371,
+      "step": 1931
+    },
+    {
+      "epoch": 0.20318929364901994,
+      "grad_norm": 3.774440786379843,
+      "learning_rate": 4.546602411549966e-06,
+      "loss": 1.0473,
+      "step": 1932
+    },
+    {
+      "epoch": 0.20329446409086727,
+      "grad_norm": 2.3929427473890987,
+      "learning_rate": 4.546122855917049e-06,
+      "loss": 1.0211,
+      "step": 1933
+    },
+    {
+      "epoch": 0.2033996345327146,
+      "grad_norm": 2.150887513168131,
+      "learning_rate": 4.5456430721255384e-06,
+      "loss": 0.9851,
+      "step": 1934
+    },
+    {
+      "epoch": 0.2035048049745619,
+      "grad_norm": 2.4012826520584274,
+      "learning_rate": 4.545163060228934e-06,
+      "loss": 1.0105,
+      "step": 1935
+    },
+    {
+      "epoch": 0.2036099754164092,
+      "grad_norm": 2.1182604676351495,
+      "learning_rate": 4.544682820280762e-06,
+      "loss": 1.0485,
+      "step": 1936
+    },
+    {
+      "epoch": 0.20371514585825654,
+      "grad_norm": 3.9728966306261673,
+      "learning_rate": 4.54420235233457e-06,
+      "loss": 1.0238,
+      "step": 1937
+    },
+    {
+      "epoch": 0.20382031630010386,
+      "grad_norm": 3.2576390659692103,
+      "learning_rate": 4.543721656443938e-06,
+      "loss": 1.0105,
+      "step": 1938
+    },
+    {
+      "epoch": 0.20392548674195118,
+      "grad_norm": 2.7319424380077,
+      "learning_rate": 4.5432407326624635e-06,
+      "loss": 1.0294,
+      "step": 1939
+    },
+    {
+      "epoch": 0.20403065718379848,
+      "grad_norm": 3.6616075268684134,
+      "learning_rate": 4.542759581043775e-06,
+      "loss": 1.0375,
+      "step": 1940
+    },
+    {
+      "epoch": 0.2041358276256458,
+      "grad_norm": 2.286876999823875,
+      "learning_rate": 4.5422782016415255e-06,
+      "loss": 0.972,
+      "step": 1941
+    },
+    {
+      "epoch": 0.20424099806749313,
+      "grad_norm": 3.5108681231659533,
+      "learning_rate": 4.541796594509391e-06,
+      "loss": 1.0733,
+      "step": 1942
+    },
+    {
+      "epoch": 0.20434616850934045,
+      "grad_norm": 1.916935649789232,
+      "learning_rate": 4.541314759701075e-06,
+      "loss": 1.0334,
+      "step": 1943
+    },
+    {
+      "epoch": 0.20445133895118778,
+      "grad_norm": 3.1369552191984047,
+      "learning_rate": 4.540832697270305e-06,
+      "loss": 1.0138,
+      "step": 1944
+    },
+    {
+      "epoch": 0.20455650939303507,
+      "grad_norm": 2.8904365660760942,
+      "learning_rate": 4.540350407270836e-06,
+      "loss": 1.0579,
+      "step": 1945
+    },
+    {
+      "epoch": 0.2046616798348824,
+      "grad_norm": 2.678633223567234,
+      "learning_rate": 4.539867889756447e-06,
+      "loss": 1.0582,
+      "step": 1946
+    },
+    {
+      "epoch": 0.20476685027672972,
+      "grad_norm": 2.5400319091221664,
+      "learning_rate": 4.539385144780942e-06,
+      "loss": 1.0247,
+      "step": 1947
+    },
+    {
+      "epoch": 0.20487202071857705,
+      "grad_norm": 1.8249180058750547,
+      "learning_rate": 4.538902172398151e-06,
+      "loss": 1.0025,
+      "step": 1948
+    },
+    {
+      "epoch": 0.20497719116042437,
+      "grad_norm": 2.39557256136971,
+      "learning_rate": 4.5384189726619285e-06,
+      "loss": 1.0428,
+      "step": 1949
+    },
+    {
+      "epoch": 0.20508236160227167,
+      "grad_norm": 2.0902532581196818,
+      "learning_rate": 4.537935545626156e-06,
+      "loss": 1.0141,
+      "step": 1950
+    },
+    {
+      "epoch": 0.205187532044119,
+      "grad_norm": 2.6266101963885817,
+      "learning_rate": 4.5374518913447384e-06,
+      "loss": 1.0491,
+      "step": 1951
+    },
+    {
+      "epoch": 0.20529270248596632,
+      "grad_norm": 1.8895661108791229,
+      "learning_rate": 4.536968009871608e-06,
+      "loss": 1.0318,
+      "step": 1952
+    },
+    {
+      "epoch": 0.20539787292781364,
+      "grad_norm": 2.1586887149934095,
+      "learning_rate": 4.536483901260721e-06,
+      "loss": 1.0529,
+      "step": 1953
+    },
+    {
+      "epoch": 0.20550304336966096,
+      "grad_norm": 3.064938688866919,
+      "learning_rate": 4.535999565566058e-06,
+      "loss": 0.9881,
+      "step": 1954
+    },
+    {
+      "epoch": 0.2056082138115083,
+      "grad_norm": 3.4278129755911415,
+      "learning_rate": 4.535515002841628e-06,
+      "loss": 1.0366,
+      "step": 1955
+    },
+    {
+      "epoch": 0.20571338425335559,
+      "grad_norm": 2.442730036564765,
+      "learning_rate": 4.535030213141462e-06,
+      "loss": 1.0062,
+      "step": 1956
+    },
+    {
+      "epoch": 0.2058185546952029,
+      "grad_norm": 2.6315188808426813,
+      "learning_rate": 4.5345451965196196e-06,
+      "loss": 0.9853,
+      "step": 1957
+    },
+    {
+      "epoch": 0.20592372513705023,
+      "grad_norm": 3.773145979161978,
+      "learning_rate": 4.5340599530301826e-06,
+      "loss": 1.0389,
+      "step": 1958
+    },
+    {
+      "epoch": 0.20602889557889756,
+      "grad_norm": 2.7833840302672614,
+      "learning_rate": 4.53357448272726e-06,
+      "loss": 1.0276,
+      "step": 1959
+    },
+    {
+      "epoch": 0.20613406602074488,
+      "grad_norm": 3.1775766358378053,
+      "learning_rate": 4.5330887856649845e-06,
+      "loss": 0.9929,
+      "step": 1960
+    },
+    {
+      "epoch": 0.20623923646259218,
+      "grad_norm": 4.593991119183918,
+      "learning_rate": 4.532602861897516e-06,
+      "loss": 1.0597,
+      "step": 1961
+    },
+    {
+      "epoch": 0.2063444069044395,
+      "grad_norm": 2.310336282250504,
+      "learning_rate": 4.532116711479039e-06,
+      "loss": 1.0174,
+      "step": 1962
+    },
+    {
+      "epoch": 0.20644957734628683,
+      "grad_norm": 3.9022900668265668,
+      "learning_rate": 4.531630334463762e-06,
+      "loss": 1.079,
+      "step": 1963
+    },
+    {
+      "epoch": 0.20655474778813415,
+      "grad_norm": 3.760113016517964,
+      "learning_rate": 4.53114373090592e-06,
+      "loss": 0.992,
+      "step": 1964
+    },
+    {
+      "epoch": 0.20665991822998148,
+      "grad_norm": 2.583786767339631,
+      "learning_rate": 4.5306569008597745e-06,
+      "loss": 1.0444,
+      "step": 1965
+    },
+    {
+      "epoch": 0.20676508867182877,
+      "grad_norm": 1.7472575830950914,
+      "learning_rate": 4.53016984437961e-06,
+      "loss": 0.9853,
+      "step": 1966
+    },
+    {
+      "epoch": 0.2068702591136761,
+      "grad_norm": 2.6276414742188097,
+      "learning_rate": 4.529682561519736e-06,
+      "loss": 1.0232,
+      "step": 1967
+    },
+    {
+      "epoch": 0.20697542955552342,
+      "grad_norm": 2.4241124519098705,
+      "learning_rate": 4.52919505233449e-06,
+      "loss": 1.017,
+      "step": 1968
+    },
+    {
+      "epoch": 0.20708059999737075,
+      "grad_norm": 3.4436315276636815,
+      "learning_rate": 4.528707316878233e-06,
+      "loss": 1.0483,
+      "step": 1969
+    },
+    {
+      "epoch": 0.20718577043921807,
+      "grad_norm": 3.6255766199501904,
+      "learning_rate": 4.528219355205349e-06,
+      "loss": 1.007,
+      "step": 1970
+    },
+    {
+      "epoch": 0.20729094088106537,
+      "grad_norm": 2.5750935689757917,
+      "learning_rate": 4.527731167370252e-06,
+      "loss": 0.991,
+      "step": 1971
+    },
+    {
+      "epoch": 0.2073961113229127,
+      "grad_norm": 3.527871448351104,
+      "learning_rate": 4.527242753427378e-06,
+      "loss": 0.9817,
+      "step": 1972
+    },
+    {
+      "epoch": 0.20750128176476001,
+      "grad_norm": 3.0602580136672124,
+      "learning_rate": 4.526754113431188e-06,
+      "loss": 1.0392,
+      "step": 1973
+    },
+    {
+      "epoch": 0.20760645220660734,
+      "grad_norm": 2.7611033572598025,
+      "learning_rate": 4.526265247436171e-06,
+      "loss": 0.9994,
+      "step": 1974
+    },
+    {
+      "epoch": 0.20771162264845466,
+      "grad_norm": 2.6859585185179067,
+      "learning_rate": 4.525776155496838e-06,
+      "loss": 1.0119,
+      "step": 1975
+    },
+    {
+      "epoch": 0.20781679309030196,
+      "grad_norm": 2.782628475167393,
+      "learning_rate": 4.525286837667726e-06,
+      "loss": 1.0515,
+      "step": 1976
+    },
+    {
+      "epoch": 0.20792196353214928,
+      "grad_norm": 2.9155913927480275,
+      "learning_rate": 4.5247972940034e-06,
+      "loss": 0.9998,
+      "step": 1977
+    },
+    {
+      "epoch": 0.2080271339739966,
+      "grad_norm": 2.9307391999616335,
+      "learning_rate": 4.524307524558446e-06,
+      "loss": 1.0204,
+      "step": 1978
+    },
+    {
+      "epoch": 0.20813230441584393,
+      "grad_norm": 2.400026037357351,
+      "learning_rate": 4.523817529387478e-06,
+      "loss": 1.0126,
+      "step": 1979
+    },
+    {
+      "epoch": 0.20823747485769126,
+      "grad_norm": 2.116341446271874,
+      "learning_rate": 4.523327308545133e-06,
+      "loss": 1.0167,
+      "step": 1980
+    },
+    {
+      "epoch": 0.20834264529953855,
+      "grad_norm": 3.6019941589879414,
+      "learning_rate": 4.522836862086076e-06,
+      "loss": 1.0122,
+      "step": 1981
+    },
+    {
+      "epoch": 0.20844781574138588,
+      "grad_norm": 2.530136205218369,
+      "learning_rate": 4.5223461900649945e-06,
+      "loss": 1.0279,
+      "step": 1982
+    },
+    {
+      "epoch": 0.2085529861832332,
+      "grad_norm": 2.5456318862932923,
+      "learning_rate": 4.521855292536603e-06,
+      "loss": 1.0164,
+      "step": 1983
+    },
+    {
+      "epoch": 0.20865815662508053,
+      "grad_norm": 3.2036037703970095,
+      "learning_rate": 4.52136416955564e-06,
+      "loss": 1.0244,
+      "step": 1984
+    },
+    {
+      "epoch": 0.20876332706692785,
+      "grad_norm": 2.700422512134757,
+      "learning_rate": 4.5208728211768696e-06,
+      "loss": 1.0144,
+      "step": 1985
+    },
+    {
+      "epoch": 0.20886849750877515,
+      "grad_norm": 2.4414664331229265,
+      "learning_rate": 4.520381247455081e-06,
+      "loss": 1.0219,
+      "step": 1986
+    },
+    {
+      "epoch": 0.20897366795062247,
+      "grad_norm": 2.3814233051604368,
+      "learning_rate": 4.519889448445088e-06,
+      "loss": 1.0393,
+      "step": 1987
+    },
+    {
+      "epoch": 0.2090788383924698,
+      "grad_norm": 2.6970110365603985,
+      "learning_rate": 4.519397424201731e-06,
+      "loss": 1.0227,
+      "step": 1988
+    },
+    {
+      "epoch": 0.20918400883431712,
+      "grad_norm": 3.0579544571066384,
+      "learning_rate": 4.518905174779874e-06,
+      "loss": 1.0041,
+      "step": 1989
+    },
+    {
+      "epoch": 0.20928917927616444,
+      "grad_norm": 3.0701954087018533,
+      "learning_rate": 4.518412700234407e-06,
+      "loss": 1.0392,
+      "step": 1990
+    },
+    {
+      "epoch": 0.20939434971801174,
+      "grad_norm": 3.10208553663707,
+      "learning_rate": 4.5179200006202425e-06,
+      "loss": 0.9866,
+      "step": 1991
+    },
+    {
+      "epoch": 0.20949952015985907,
+      "grad_norm": 2.5379999852239687,
+      "learning_rate": 4.517427075992325e-06,
+      "loss": 0.9833,
+      "step": 1992
+    },
+    {
+      "epoch": 0.2096046906017064,
+      "grad_norm": 2.898863425583052,
+      "learning_rate": 4.516933926405614e-06,
+      "loss": 1.024,
+      "step": 1993
+    },
+    {
+      "epoch": 0.2097098610435537,
+      "grad_norm": 2.5016600905979134,
+      "learning_rate": 4.516440551915103e-06,
+      "loss": 1.0243,
+      "step": 1994
+    },
+    {
+      "epoch": 0.20981503148540104,
+      "grad_norm": 2.37701429193729,
+      "learning_rate": 4.5159469525758065e-06,
+      "loss": 1.0574,
+      "step": 1995
+    },
+    {
+      "epoch": 0.20992020192724833,
+      "grad_norm": 3.349993368242468,
+      "learning_rate": 4.515453128442764e-06,
+      "loss": 0.9991,
+      "step": 1996
+    },
+    {
+      "epoch": 0.21002537236909566,
+      "grad_norm": 3.041271688381229,
+      "learning_rate": 4.514959079571042e-06,
+      "loss": 1.0306,
+      "step": 1997
+    },
+    {
+      "epoch": 0.21013054281094298,
+      "grad_norm": 3.7606519650863452,
+      "learning_rate": 4.514464806015729e-06,
+      "loss": 1.0277,
+      "step": 1998
+    },
+    {
+      "epoch": 0.2102357132527903,
+      "grad_norm": 2.8528069179863307,
+      "learning_rate": 4.513970307831941e-06,
+      "loss": 0.9865,
+      "step": 1999
+    },
+    {
+      "epoch": 0.21034088369463763,
+      "grad_norm": 3.3798634505812974,
+      "learning_rate": 4.513475585074819e-06,
+      "loss": 1.0434,
+      "step": 2000
+    },
+    {
+      "epoch": 0.21044605413648493,
+      "grad_norm": 3.446977094091343,
+      "learning_rate": 4.512980637799529e-06,
+      "loss": 1.0129,
+      "step": 2001
+    },
+    {
+      "epoch": 0.21055122457833225,
+      "grad_norm": 2.5663183477604483,
+      "learning_rate": 4.512485466061258e-06,
+      "loss": 1.0074,
+      "step": 2002
+    },
+    {
+      "epoch": 0.21065639502017958,
+      "grad_norm": 2.713416190814693,
+      "learning_rate": 4.511990069915226e-06,
+      "loss": 1.0175,
+      "step": 2003
+    },
+    {
+      "epoch": 0.2107615654620269,
+      "grad_norm": 3.017467404478425,
+      "learning_rate": 4.511494449416671e-06,
+      "loss": 1.0022,
+      "step": 2004
+    },
+    {
+      "epoch": 0.21086673590387423,
+      "grad_norm": 2.869407764453783,
+      "learning_rate": 4.510998604620859e-06,
+      "loss": 1.0091,
+      "step": 2005
+    },
+    {
+      "epoch": 0.21097190634572152,
+      "grad_norm": 2.2914275171439624,
+      "learning_rate": 4.510502535583081e-06,
+      "loss": 1.027,
+      "step": 2006
+    },
+    {
+      "epoch": 0.21107707678756885,
+      "grad_norm": 3.1289559026795413,
+      "learning_rate": 4.510006242358652e-06,
+      "loss": 1.0192,
+      "step": 2007
+    },
+    {
+      "epoch": 0.21118224722941617,
+      "grad_norm": 3.365076590082289,
+      "learning_rate": 4.509509725002913e-06,
+      "loss": 1.0314,
+      "step": 2008
+    },
+    {
+      "epoch": 0.2112874176712635,
+      "grad_norm": 2.619162513458224,
+      "learning_rate": 4.5090129835712286e-06,
+      "loss": 0.9736,
+      "step": 2009
+    },
+    {
+      "epoch": 0.21139258811311082,
+      "grad_norm": 2.6581918245692973,
+      "learning_rate": 4.50851601811899e-06,
+      "loss": 0.9946,
+      "step": 2010
+    },
+    {
+      "epoch": 0.21149775855495812,
+      "grad_norm": 3.2402527360258713,
+      "learning_rate": 4.508018828701613e-06,
+      "loss": 1.0187,
+      "step": 2011
+    },
+    {
+      "epoch": 0.21160292899680544,
+      "grad_norm": 3.561750316675759,
+      "learning_rate": 4.5075214153745375e-06,
+      "loss": 1.0131,
+      "step": 2012
+    },
+    {
+      "epoch": 0.21170809943865276,
+      "grad_norm": 3.6081378607990815,
+      "learning_rate": 4.507023778193229e-06,
+      "loss": 1.0536,
+      "step": 2013
+    },
+    {
+      "epoch": 0.2118132698805001,
+      "grad_norm": 3.98061138379272,
+      "learning_rate": 4.506525917213178e-06,
+      "loss": 0.9886,
+      "step": 2014
+    },
+    {
+      "epoch": 0.2119184403223474,
+      "grad_norm": 3.6286847213673656,
+      "learning_rate": 4.506027832489901e-06,
+      "loss": 1.0802,
+      "step": 2015
+    },
+    {
+      "epoch": 0.21202361076419474,
+      "grad_norm": 2.970920799122295,
+      "learning_rate": 4.505529524078936e-06,
+      "loss": 0.9943,
+      "step": 2016
+    },
+    {
+      "epoch": 0.21212878120604203,
+      "grad_norm": 3.1998864392209034,
+      "learning_rate": 4.505030992035851e-06,
+      "loss": 1.0541,
+      "step": 2017
+    },
+    {
+      "epoch": 0.21223395164788936,
+      "grad_norm": 3.5531939185269117,
+      "learning_rate": 4.504532236416234e-06,
+      "loss": 1.0057,
+      "step": 2018
+    },
+    {
+      "epoch": 0.21233912208973668,
+      "grad_norm": 2.6807118334186173,
+      "learning_rate": 4.504033257275701e-06,
+      "loss": 0.989,
+      "step": 2019
+    },
+    {
+      "epoch": 0.212444292531584,
+      "grad_norm": 3.709260634562878,
+      "learning_rate": 4.5035340546698915e-06,
+      "loss": 1.0069,
+      "step": 2020
+    },
+    {
+      "epoch": 0.21254946297343133,
+      "grad_norm": 3.0800163674767655,
+      "learning_rate": 4.503034628654472e-06,
+      "loss": 1.0188,
+      "step": 2021
+    },
+    {
+      "epoch": 0.21265463341527863,
+      "grad_norm": 2.827473390632371,
+      "learning_rate": 4.50253497928513e-06,
+      "loss": 1.0189,
+      "step": 2022
+    },
+    {
+      "epoch": 0.21275980385712595,
+      "grad_norm": 3.293367596776558,
+      "learning_rate": 4.502035106617583e-06,
+      "loss": 0.9761,
+      "step": 2023
+    },
+    {
+      "epoch": 0.21286497429897328,
+      "grad_norm": 2.9764561882025093,
+      "learning_rate": 4.501535010707569e-06,
+      "loss": 1.0117,
+      "step": 2024
+    },
+    {
+      "epoch": 0.2129701447408206,
+      "grad_norm": 1.8974311339827867,
+      "learning_rate": 4.501034691610852e-06,
+      "loss": 1.0089,
+      "step": 2025
+    },
+    {
+      "epoch": 0.21307531518266792,
+      "grad_norm": 3.113046982013976,
+      "learning_rate": 4.500534149383222e-06,
+      "loss": 0.9953,
+      "step": 2026
+    },
+    {
+      "epoch": 0.21318048562451522,
+      "grad_norm": 2.5218275940026307,
+      "learning_rate": 4.5000333840804945e-06,
+      "loss": 1.0573,
+      "step": 2027
+    },
+    {
+      "epoch": 0.21328565606636254,
+      "grad_norm": 2.1504903869654974,
+      "learning_rate": 4.499532395758507e-06,
+      "loss": 1.0295,
+      "step": 2028
+    },
+    {
+      "epoch": 0.21339082650820987,
+      "grad_norm": 2.116485409339894,
+      "learning_rate": 4.499031184473125e-06,
+      "loss": 1.0145,
+      "step": 2029
+    },
+    {
+      "epoch": 0.2134959969500572,
+      "grad_norm": 2.1971557306143583,
+      "learning_rate": 4.498529750280237e-06,
+      "loss": 0.9975,
+      "step": 2030
+    },
+    {
+      "epoch": 0.21360116739190452,
+      "grad_norm": 2.441106240316848,
+      "learning_rate": 4.4980280932357565e-06,
+      "loss": 1.0324,
+      "step": 2031
+    },
+    {
+      "epoch": 0.2137063378337518,
+      "grad_norm": 1.8648832005177987,
+      "learning_rate": 4.4975262133956235e-06,
+      "loss": 0.976,
+      "step": 2032
+    },
+    {
+      "epoch": 0.21381150827559914,
+      "grad_norm": 2.393559217891105,
+      "learning_rate": 4.497024110815799e-06,
+      "loss": 1.0236,
+      "step": 2033
+    },
+    {
+      "epoch": 0.21391667871744646,
+      "grad_norm": 3.4843636423614175,
+      "learning_rate": 4.496521785552273e-06,
+      "loss": 1.0423,
+      "step": 2034
+    },
+    {
+      "epoch": 0.2140218491592938,
+      "grad_norm": 2.9511317676008297,
+      "learning_rate": 4.496019237661059e-06,
+      "loss": 1.0201,
+      "step": 2035
+    },
+    {
+      "epoch": 0.2141270196011411,
+      "grad_norm": 2.554507983535873,
+      "learning_rate": 4.495516467198193e-06,
+      "loss": 0.9985,
+      "step": 2036
+    },
+    {
+      "epoch": 0.2142321900429884,
+      "grad_norm": 2.5051574205030973,
+      "learning_rate": 4.49501347421974e-06,
+      "loss": 0.9961,
+      "step": 2037
+    },
+    {
+      "epoch": 0.21433736048483573,
+      "grad_norm": 2.148199842957758,
+      "learning_rate": 4.494510258781788e-06,
+      "loss": 0.9899,
+      "step": 2038
+    },
+    {
+      "epoch": 0.21444253092668306,
+      "grad_norm": 3.4634861798471466,
+      "learning_rate": 4.4940068209404465e-06,
+      "loss": 1.0501,
+      "step": 2039
+    },
+    {
+      "epoch": 0.21454770136853038,
+      "grad_norm": 3.022603989678606,
+      "learning_rate": 4.493503160751855e-06,
+      "loss": 0.999,
+      "step": 2040
+    },
+    {
+      "epoch": 0.2146528718103777,
+      "grad_norm": 2.595687398370017,
+      "learning_rate": 4.4929992782721755e-06,
+      "loss": 1.0535,
+      "step": 2041
+    },
+    {
+      "epoch": 0.214758042252225,
+      "grad_norm": 2.3610745008748033,
+      "learning_rate": 4.492495173557594e-06,
+      "loss": 0.9964,
+      "step": 2042
+    },
+    {
+      "epoch": 0.21486321269407233,
+      "grad_norm": 2.523798173897767,
+      "learning_rate": 4.491990846664322e-06,
+      "loss": 1.0458,
+      "step": 2043
+    },
+    {
+      "epoch": 0.21496838313591965,
+      "grad_norm": 2.768262718642694,
+      "learning_rate": 4.4914862976485975e-06,
+      "loss": 1.0412,
+      "step": 2044
+    },
+    {
+      "epoch": 0.21507355357776697,
+      "grad_norm": 3.5933411780857605,
+      "learning_rate": 4.4909815265666786e-06,
+      "loss": 1.0408,
+      "step": 2045
+    },
+    {
+      "epoch": 0.2151787240196143,
+      "grad_norm": 1.859821990629818,
+      "learning_rate": 4.490476533474854e-06,
+      "loss": 1.033,
+      "step": 2046
+    },
+    {
+      "epoch": 0.2152838944614616,
+      "grad_norm": 2.952652314844579,
+      "learning_rate": 4.4899713184294336e-06,
+      "loss": 1.013,
+      "step": 2047
+    },
+    {
+      "epoch": 0.21538906490330892,
+      "grad_norm": 3.0877949687525,
+      "learning_rate": 4.489465881486753e-06,
+      "loss": 1.065,
+      "step": 2048
+    },
+    {
+      "epoch": 0.21549423534515624,
+      "grad_norm": 2.1468116186309043,
+      "learning_rate": 4.48896022270317e-06,
+      "loss": 0.9845,
+      "step": 2049
+    },
+    {
+      "epoch": 0.21559940578700357,
+      "grad_norm": 3.0848353636050745,
+      "learning_rate": 4.488454342135074e-06,
+      "loss": 1.0412,
+      "step": 2050
+    },
+    {
+      "epoch": 0.2157045762288509,
+      "grad_norm": 2.827181553896088,
+      "learning_rate": 4.487948239838869e-06,
+      "loss": 0.957,
+      "step": 2051
+    },
+    {
+      "epoch": 0.2158097466706982,
+      "grad_norm": 2.722663933189815,
+      "learning_rate": 4.487441915870995e-06,
+      "loss": 0.9992,
+      "step": 2052
+    },
+    {
+      "epoch": 0.2159149171125455,
+      "grad_norm": 2.996172795915466,
+      "learning_rate": 4.486935370287907e-06,
+      "loss": 1.0284,
+      "step": 2053
+    },
+    {
+      "epoch": 0.21602008755439284,
+      "grad_norm": 3.6047336301727637,
+      "learning_rate": 4.48642860314609e-06,
+      "loss": 1.0249,
+      "step": 2054
+    },
+    {
+      "epoch": 0.21612525799624016,
+      "grad_norm": 2.867533696478624,
+      "learning_rate": 4.485921614502054e-06,
+      "loss": 1.0379,
+      "step": 2055
+    },
+    {
+      "epoch": 0.21623042843808749,
+      "grad_norm": 3.176123229551431,
+      "learning_rate": 4.485414404412329e-06,
+      "loss": 1.0427,
+      "step": 2056
+    },
+    {
+      "epoch": 0.21633559887993478,
+      "grad_norm": 2.9533715243960965,
+      "learning_rate": 4.484906972933476e-06,
+      "loss": 0.9631,
+      "step": 2057
+    },
+    {
+      "epoch": 0.2164407693217821,
+      "grad_norm": 2.8688315629005716,
+      "learning_rate": 4.484399320122075e-06,
+      "loss": 1.0105,
+      "step": 2058
+    },
+    {
+      "epoch": 0.21654593976362943,
+      "grad_norm": 3.5024671051619136,
+      "learning_rate": 4.483891446034736e-06,
+      "loss": 1.0587,
+      "step": 2059
+    },
+    {
+      "epoch": 0.21665111020547675,
+      "grad_norm": 1.901326742128866,
+      "learning_rate": 4.4833833507280884e-06,
+      "loss": 1.0266,
+      "step": 2060
+    },
+    {
+      "epoch": 0.21675628064732408,
+      "grad_norm": 2.5650512457429966,
+      "learning_rate": 4.4828750342587895e-06,
+      "loss": 1.0043,
+      "step": 2061
+    },
+    {
+      "epoch": 0.21686145108917138,
+      "grad_norm": 2.448189051315377,
+      "learning_rate": 4.482366496683521e-06,
+      "loss": 1.0479,
+      "step": 2062
+    },
+    {
+      "epoch": 0.2169666215310187,
+      "grad_norm": 2.824309733500601,
+      "learning_rate": 4.481857738058989e-06,
+      "loss": 0.988,
+      "step": 2063
+    },
+    {
+      "epoch": 0.21707179197286602,
+      "grad_norm": 3.056046040580293,
+      "learning_rate": 4.481348758441923e-06,
+      "loss": 1.017,
+      "step": 2064
+    },
+    {
+      "epoch": 0.21717696241471335,
+      "grad_norm": 2.0164976145859885,
+      "learning_rate": 4.480839557889079e-06,
+      "loss": 1.0477,
+      "step": 2065
+    },
+    {
+      "epoch": 0.21728213285656067,
+      "grad_norm": 2.9759205852579274,
+      "learning_rate": 4.480330136457237e-06,
+      "loss": 1.0313,
+      "step": 2066
+    },
+    {
+      "epoch": 0.21738730329840797,
+      "grad_norm": 4.396630703796807,
+      "learning_rate": 4.4798204942032005e-06,
+      "loss": 1.0554,
+      "step": 2067
+    },
+    {
+      "epoch": 0.2174924737402553,
+      "grad_norm": 3.2600670571924453,
+      "learning_rate": 4.4793106311838e-06,
+      "loss": 1.0142,
+      "step": 2068
+    },
+    {
+      "epoch": 0.21759764418210262,
+      "grad_norm": 2.554456195461736,
+      "learning_rate": 4.478800547455887e-06,
+      "loss": 1.033,
+      "step": 2069
+    },
+    {
+      "epoch": 0.21770281462394994,
+      "grad_norm": 1.8748589681364354,
+      "learning_rate": 4.478290243076342e-06,
+      "loss": 0.9785,
+      "step": 2070
+    },
+    {
+      "epoch": 0.21780798506579727,
+      "grad_norm": 3.558459347898174,
+      "learning_rate": 4.4777797181020675e-06,
+      "loss": 1.0148,
+      "step": 2071
+    },
+    {
+      "epoch": 0.2179131555076446,
+      "grad_norm": 2.4480561013052253,
+      "learning_rate": 4.477268972589989e-06,
+      "loss": 1.0506,
+      "step": 2072
+    },
+    {
+      "epoch": 0.2180183259494919,
+      "grad_norm": 3.167847874819189,
+      "learning_rate": 4.476758006597061e-06,
+      "loss": 1.0134,
+      "step": 2073
+    },
+    {
+      "epoch": 0.2181234963913392,
+      "grad_norm": 2.035096373195341,
+      "learning_rate": 4.476246820180259e-06,
+      "loss": 0.9828,
+      "step": 2074
+    },
+    {
+      "epoch": 0.21822866683318654,
+      "grad_norm": 3.0180664267026978,
+      "learning_rate": 4.475735413396585e-06,
+      "loss": 0.9717,
+      "step": 2075
+    },
+    {
+      "epoch": 0.21833383727503386,
+      "grad_norm": 2.937593336372816,
+      "learning_rate": 4.475223786303064e-06,
+      "loss": 1.0418,
+      "step": 2076
+    },
+    {
+      "epoch": 0.21843900771688118,
+      "grad_norm": 3.1853321519938125,
+      "learning_rate": 4.474711938956747e-06,
+      "loss": 1.0308,
+      "step": 2077
+    },
+    {
+      "epoch": 0.21854417815872848,
+      "grad_norm": 2.781640730371229,
+      "learning_rate": 4.4741998714147085e-06,
+      "loss": 0.9788,
+      "step": 2078
+    },
+    {
+      "epoch": 0.2186493486005758,
+      "grad_norm": 2.97926045372165,
+      "learning_rate": 4.473687583734048e-06,
+      "loss": 1.0543,
+      "step": 2079
+    },
+    {
+      "epoch": 0.21875451904242313,
+      "grad_norm": 3.427682301198862,
+      "learning_rate": 4.473175075971889e-06,
+      "loss": 1.0172,
+      "step": 2080
+    },
+    {
+      "epoch": 0.21885968948427045,
+      "grad_norm": 2.594403559813018,
+      "learning_rate": 4.472662348185382e-06,
+      "loss": 1.031,
+      "step": 2081
+    },
+    {
+      "epoch": 0.21896485992611778,
+      "grad_norm": 3.1273306327310193,
+      "learning_rate": 4.472149400431699e-06,
+      "loss": 1.0469,
+      "step": 2082
+    },
+    {
+      "epoch": 0.21907003036796507,
+      "grad_norm": 2.604066679542293,
+      "learning_rate": 4.471636232768038e-06,
+      "loss": 0.9904,
+      "step": 2083
+    },
+    {
+      "epoch": 0.2191752008098124,
+      "grad_norm": 2.4026133544836767,
+      "learning_rate": 4.47112284525162e-06,
+      "loss": 1.0221,
+      "step": 2084
+    },
+    {
+      "epoch": 0.21928037125165972,
+      "grad_norm": 3.599921055468287,
+      "learning_rate": 4.470609237939693e-06,
+      "loss": 0.9989,
+      "step": 2085
+    },
+    {
+      "epoch": 0.21938554169350705,
+      "grad_norm": 2.131221331900976,
+      "learning_rate": 4.470095410889528e-06,
+      "loss": 0.9947,
+      "step": 2086
+    },
+    {
+      "epoch": 0.21949071213535437,
+      "grad_norm": 2.1497035619992104,
+      "learning_rate": 4.46958136415842e-06,
+      "loss": 1.0332,
+      "step": 2087
+    },
+    {
+      "epoch": 0.21959588257720167,
+      "grad_norm": 2.3616472688438406,
+      "learning_rate": 4.469067097803689e-06,
+      "loss": 1.0208,
+      "step": 2088
+    },
+    {
+      "epoch": 0.219701053019049,
+      "grad_norm": 2.6359552878027674,
+      "learning_rate": 4.46855261188268e-06,
+      "loss": 1.0618,
+      "step": 2089
+    },
+    {
+      "epoch": 0.21980622346089632,
+      "grad_norm": 1.9018280297999302,
+      "learning_rate": 4.468037906452765e-06,
+      "loss": 1.006,
+      "step": 2090
+    },
+    {
+      "epoch": 0.21991139390274364,
+      "grad_norm": 2.5721362754088517,
+      "learning_rate": 4.4675229815713324e-06,
+      "loss": 1.0456,
+      "step": 2091
+    },
+    {
+      "epoch": 0.22001656434459096,
+      "grad_norm": 2.954980814994551,
+      "learning_rate": 4.467007837295805e-06,
+      "loss": 1.0167,
+      "step": 2092
+    },
+    {
+      "epoch": 0.22012173478643826,
+      "grad_norm": 3.3860189739598643,
+      "learning_rate": 4.466492473683624e-06,
+      "loss": 1.0261,
+      "step": 2093
+    },
+    {
+      "epoch": 0.22022690522828559,
+      "grad_norm": 2.9650931599791757,
+      "learning_rate": 4.465976890792254e-06,
+      "loss": 1.0466,
+      "step": 2094
+    },
+    {
+      "epoch": 0.2203320756701329,
+      "grad_norm": 3.104064301699326,
+      "learning_rate": 4.465461088679189e-06,
+      "loss": 1.0256,
+      "step": 2095
+    },
+    {
+      "epoch": 0.22043724611198023,
+      "grad_norm": 2.636686526867292,
+      "learning_rate": 4.464945067401945e-06,
+      "loss": 1.0324,
+      "step": 2096
+    },
+    {
+      "epoch": 0.22054241655382756,
+      "grad_norm": 2.9763466587493896,
+      "learning_rate": 4.464428827018062e-06,
+      "loss": 1.0048,
+      "step": 2097
+    },
+    {
+      "epoch": 0.22064758699567485,
+      "grad_norm": 2.658483317192696,
+      "learning_rate": 4.4639123675851035e-06,
+      "loss": 1.0172,
+      "step": 2098
+    },
+    {
+      "epoch": 0.22075275743752218,
+      "grad_norm": 2.70559891063467,
+      "learning_rate": 4.46339568916066e-06,
+      "loss": 1.0571,
+      "step": 2099
+    },
+    {
+      "epoch": 0.2208579278793695,
+      "grad_norm": 2.880535256255276,
+      "learning_rate": 4.462878791802345e-06,
+      "loss": 1.0431,
+      "step": 2100
+    },
+    {
+      "epoch": 0.22096309832121683,
+      "grad_norm": 3.233937870847447,
+      "learning_rate": 4.4623616755677965e-06,
+      "loss": 0.9888,
+      "step": 2101
+    },
+    {
+      "epoch": 0.22106826876306415,
+      "grad_norm": 2.804108237608663,
+      "learning_rate": 4.461844340514678e-06,
+      "loss": 1.0465,
+      "step": 2102
+    },
+    {
+      "epoch": 0.22117343920491145,
+      "grad_norm": 3.2318628882501295,
+      "learning_rate": 4.4613267867006745e-06,
+      "loss": 1.0353,
+      "step": 2103
+    },
+    {
+      "epoch": 0.22127860964675877,
+      "grad_norm": 3.0591146643434786,
+      "learning_rate": 4.460809014183498e-06,
+      "loss": 1.0467,
+      "step": 2104
+    },
+    {
+      "epoch": 0.2213837800886061,
+      "grad_norm": 3.1219953722891525,
+      "learning_rate": 4.460291023020885e-06,
+      "loss": 1.0335,
+      "step": 2105
+    },
+    {
+      "epoch": 0.22148895053045342,
+      "grad_norm": 2.837039568315429,
+      "learning_rate": 4.4597728132705944e-06,
+      "loss": 1.0017,
+      "step": 2106
+    },
+    {
+      "epoch": 0.22159412097230075,
+      "grad_norm": 2.401393140037482,
+      "learning_rate": 4.459254384990411e-06,
+      "loss": 1.0577,
+      "step": 2107
+    },
+    {
+      "epoch": 0.22169929141414804,
+      "grad_norm": 2.8215140434580905,
+      "learning_rate": 4.458735738238145e-06,
+      "loss": 0.9908,
+      "step": 2108
+    },
+    {
+      "epoch": 0.22180446185599537,
+      "grad_norm": 3.800737740073966,
+      "learning_rate": 4.4582168730716264e-06,
+      "loss": 1.0633,
+      "step": 2109
+    },
+    {
+      "epoch": 0.2219096322978427,
+      "grad_norm": 2.755738180922399,
+      "learning_rate": 4.457697789548715e-06,
+      "loss": 1.0299,
+      "step": 2110
+    },
+    {
+      "epoch": 0.22201480273969001,
+      "grad_norm": 3.5390554379093477,
+      "learning_rate": 4.457178487727292e-06,
+      "loss": 1.0707,
+      "step": 2111
+    },
+    {
+      "epoch": 0.22211997318153734,
+      "grad_norm": 4.020240288638399,
+      "learning_rate": 4.456658967665265e-06,
+      "loss": 1.0235,
+      "step": 2112
+    },
+    {
+      "epoch": 0.22222514362338464,
+      "grad_norm": 2.7427373031139166,
+      "learning_rate": 4.4561392294205615e-06,
+      "loss": 1.0166,
+      "step": 2113
+    },
+    {
+      "epoch": 0.22233031406523196,
+      "grad_norm": 3.0778943699917094,
+      "learning_rate": 4.455619273051139e-06,
+      "loss": 1.035,
+      "step": 2114
+    },
+    {
+      "epoch": 0.22243548450707928,
+      "grad_norm": 2.747240418850475,
+      "learning_rate": 4.455099098614975e-06,
+      "loss": 0.9829,
+      "step": 2115
+    },
+    {
+      "epoch": 0.2225406549489266,
+      "grad_norm": 3.7757497910125486,
+      "learning_rate": 4.454578706170075e-06,
+      "loss": 1.0399,
+      "step": 2116
+    },
+    {
+      "epoch": 0.22264582539077393,
+      "grad_norm": 3.683295020630938,
+      "learning_rate": 4.454058095774465e-06,
+      "loss": 1.0256,
+      "step": 2117
+    },
+    {
+      "epoch": 0.22275099583262123,
+      "grad_norm": 2.9562137863553235,
+      "learning_rate": 4.453537267486197e-06,
+      "loss": 1.027,
+      "step": 2118
+    },
+    {
+      "epoch": 0.22285616627446855,
+      "grad_norm": 3.658729786664914,
+      "learning_rate": 4.453016221363349e-06,
+      "loss": 1.0615,
+      "step": 2119
+    },
+    {
+      "epoch": 0.22296133671631588,
+      "grad_norm": 2.7902924257034796,
+      "learning_rate": 4.452494957464021e-06,
+      "loss": 1.0576,
+      "step": 2120
+    },
+    {
+      "epoch": 0.2230665071581632,
+      "grad_norm": 2.7787504278125286,
+      "learning_rate": 4.451973475846337e-06,
+      "loss": 0.9904,
+      "step": 2121
+    },
+    {
+      "epoch": 0.22317167760001053,
+      "grad_norm": 3.234868103009726,
+      "learning_rate": 4.451451776568446e-06,
+      "loss": 1.0415,
+      "step": 2122
+    },
+    {
+      "epoch": 0.22327684804185782,
+      "grad_norm": 3.1870487546909323,
+      "learning_rate": 4.450929859688524e-06,
+      "loss": 1.0222,
+      "step": 2123
+    },
+    {
+      "epoch": 0.22338201848370515,
+      "grad_norm": 2.4299350284656125,
+      "learning_rate": 4.450407725264768e-06,
+      "loss": 1.0192,
+      "step": 2124
+    },
+    {
+      "epoch": 0.22348718892555247,
+      "grad_norm": 2.0691045775857115,
+      "learning_rate": 4.449885373355398e-06,
+      "loss": 1.0423,
+      "step": 2125
+    },
+    {
+      "epoch": 0.2235923593673998,
+      "grad_norm": 2.2233933117212628,
+      "learning_rate": 4.449362804018661e-06,
+      "loss": 0.9819,
+      "step": 2126
+    },
+    {
+      "epoch": 0.22369752980924712,
+      "grad_norm": 3.019428274676109,
+      "learning_rate": 4.44884001731283e-06,
+      "loss": 1.0295,
+      "step": 2127
+    },
+    {
+      "epoch": 0.22380270025109442,
+      "grad_norm": 2.659057630935619,
+      "learning_rate": 4.4483170132961964e-06,
+      "loss": 0.9956,
+      "step": 2128
+    },
+    {
+      "epoch": 0.22390787069294174,
+      "grad_norm": 2.830751803296575,
+      "learning_rate": 4.447793792027082e-06,
+      "loss": 1.0471,
+      "step": 2129
+    },
+    {
+      "epoch": 0.22401304113478906,
+      "grad_norm": 3.464512110587927,
+      "learning_rate": 4.447270353563828e-06,
+      "loss": 1.0079,
+      "step": 2130
+    },
+    {
+      "epoch": 0.2241182115766364,
+      "grad_norm": 3.3402290513696724,
+      "learning_rate": 4.446746697964803e-06,
+      "loss": 1.0274,
+      "step": 2131
+    },
+    {
+      "epoch": 0.2242233820184837,
+      "grad_norm": 2.6721352938404586,
+      "learning_rate": 4.446222825288398e-06,
+      "loss": 0.9991,
+      "step": 2132
+    },
+    {
+      "epoch": 0.22432855246033104,
+      "grad_norm": 3.3974985750450815,
+      "learning_rate": 4.44569873559303e-06,
+      "loss": 1.0168,
+      "step": 2133
+    },
+    {
+      "epoch": 0.22443372290217833,
+      "grad_norm": 2.272453274425081,
+      "learning_rate": 4.445174428937137e-06,
+      "loss": 1.0018,
+      "step": 2134
+    },
+    {
+      "epoch": 0.22453889334402566,
+      "grad_norm": 2.733296898300132,
+      "learning_rate": 4.444649905379184e-06,
+      "loss": 1.0517,
+      "step": 2135
+    },
+    {
+      "epoch": 0.22464406378587298,
+      "grad_norm": 3.0055966087106563,
+      "learning_rate": 4.444125164977662e-06,
+      "loss": 1.0395,
+      "step": 2136
+    },
+    {
+      "epoch": 0.2247492342277203,
+      "grad_norm": 2.7417512364939682,
+      "learning_rate": 4.44360020779108e-06,
+      "loss": 1.0283,
+      "step": 2137
+    },
+    {
+      "epoch": 0.22485440466956763,
+      "grad_norm": 2.136270049883368,
+      "learning_rate": 4.443075033877978e-06,
+      "loss": 1.0313,
+      "step": 2138
+    },
+    {
+      "epoch": 0.22495957511141493,
+      "grad_norm": 2.623520046793813,
+      "learning_rate": 4.442549643296915e-06,
+      "loss": 1.0285,
+      "step": 2139
+    },
+    {
+      "epoch": 0.22506474555326225,
+      "grad_norm": 2.6185982076249306,
+      "learning_rate": 4.442024036106476e-06,
+      "loss": 1.0296,
+      "step": 2140
+    },
+    {
+      "epoch": 0.22516991599510958,
+      "grad_norm": 2.953502401243642,
+      "learning_rate": 4.441498212365271e-06,
+      "loss": 1.0403,
+      "step": 2141
+    },
+    {
+      "epoch": 0.2252750864369569,
+      "grad_norm": 5.131915620587942,
+      "learning_rate": 4.440972172131933e-06,
+      "loss": 1.0135,
+      "step": 2142
+    },
+    {
+      "epoch": 0.22538025687880422,
+      "grad_norm": 3.2510064788855537,
+      "learning_rate": 4.440445915465119e-06,
+      "loss": 1.0226,
+      "step": 2143
+    },
+    {
+      "epoch": 0.22548542732065152,
+      "grad_norm": 2.2690579029077007,
+      "learning_rate": 4.439919442423513e-06,
+      "loss": 1.0519,
+      "step": 2144
+    },
+    {
+      "epoch": 0.22559059776249885,
+      "grad_norm": 3.028677542365834,
+      "learning_rate": 4.439392753065818e-06,
+      "loss": 1.0209,
+      "step": 2145
+    },
+    {
+      "epoch": 0.22569576820434617,
+      "grad_norm": 2.6183652601612795,
+      "learning_rate": 4.438865847450766e-06,
+      "loss": 1.0234,
+      "step": 2146
+    },
+    {
+      "epoch": 0.2258009386461935,
+      "grad_norm": 2.655490949700378,
+      "learning_rate": 4.438338725637111e-06,
+      "loss": 0.9981,
+      "step": 2147
+    },
+    {
+      "epoch": 0.22590610908804082,
+      "grad_norm": 3.0919304406899104,
+      "learning_rate": 4.43781138768363e-06,
+      "loss": 1.0164,
+      "step": 2148
+    },
+    {
+      "epoch": 0.22601127952988811,
+      "grad_norm": 2.930774431302077,
+      "learning_rate": 4.437283833649125e-06,
+      "loss": 0.9999,
+      "step": 2149
+    },
+    {
+      "epoch": 0.22611644997173544,
+      "grad_norm": 1.6428635453927467,
+      "learning_rate": 4.436756063592424e-06,
+      "loss": 0.9752,
+      "step": 2150
+    },
+    {
+      "epoch": 0.22622162041358276,
+      "grad_norm": 2.3696606960139546,
+      "learning_rate": 4.4362280775723775e-06,
+      "loss": 1.0179,
+      "step": 2151
+    },
+    {
+      "epoch": 0.2263267908554301,
+      "grad_norm": 3.3313866190682817,
+      "learning_rate": 4.4356998756478585e-06,
+      "loss": 0.9765,
+      "step": 2152
+    },
+    {
+      "epoch": 0.2264319612972774,
+      "grad_norm": 2.577906154713765,
+      "learning_rate": 4.435171457877766e-06,
+      "loss": 0.97,
+      "step": 2153
+    },
+    {
+      "epoch": 0.2265371317391247,
+      "grad_norm": 2.503718726560151,
+      "learning_rate": 4.434642824321023e-06,
+      "loss": 0.9966,
+      "step": 2154
+    },
+    {
+      "epoch": 0.22664230218097203,
+      "grad_norm": 2.7189315507014298,
+      "learning_rate": 4.434113975036577e-06,
+      "loss": 1.0341,
+      "step": 2155
+    },
+    {
+      "epoch": 0.22674747262281936,
+      "grad_norm": 2.858149879919419,
+      "learning_rate": 4.433584910083397e-06,
+      "loss": 1.0271,
+      "step": 2156
+    },
+    {
+      "epoch": 0.22685264306466668,
+      "grad_norm": 2.2845656463626094,
+      "learning_rate": 4.4330556295204796e-06,
+      "loss": 0.9709,
+      "step": 2157
+    },
+    {
+      "epoch": 0.226957813506514,
+      "grad_norm": 2.2220751802170122,
+      "learning_rate": 4.432526133406843e-06,
+      "loss": 0.9941,
+      "step": 2158
+    },
+    {
+      "epoch": 0.2270629839483613,
+      "grad_norm": 2.710945616604557,
+      "learning_rate": 4.43199642180153e-06,
+      "loss": 1.0139,
+      "step": 2159
+    },
+    {
+      "epoch": 0.22716815439020863,
+      "grad_norm": 2.3634868040227235,
+      "learning_rate": 4.4314664947636075e-06,
+      "loss": 0.9786,
+      "step": 2160
+    },
+    {
+      "epoch": 0.22727332483205595,
+      "grad_norm": 2.4612164449548977,
+      "learning_rate": 4.430936352352167e-06,
+      "loss": 1.0174,
+      "step": 2161
+    },
+    {
+      "epoch": 0.22737849527390327,
+      "grad_norm": 2.0768737166088878,
+      "learning_rate": 4.430405994626323e-06,
+      "loss": 1.013,
+      "step": 2162
+    },
+    {
+      "epoch": 0.2274836657157506,
+      "grad_norm": 2.3532257856103374,
+      "learning_rate": 4.429875421645214e-06,
+      "loss": 1.0091,
+      "step": 2163
+    },
+    {
+      "epoch": 0.2275888361575979,
+      "grad_norm": 3.0325316947217362,
+      "learning_rate": 4.429344633468005e-06,
+      "loss": 1.028,
+      "step": 2164
+    },
+    {
+      "epoch": 0.22769400659944522,
+      "grad_norm": 3.017583155543902,
+      "learning_rate": 4.428813630153881e-06,
+      "loss": 1.0273,
+      "step": 2165
+    },
+    {
+      "epoch": 0.22779917704129254,
+      "grad_norm": 2.9673310296779034,
+      "learning_rate": 4.428282411762054e-06,
+      "loss": 1.0312,
+      "step": 2166
+    },
+    {
+      "epoch": 0.22790434748313987,
+      "grad_norm": 2.756113051254936,
+      "learning_rate": 4.427750978351758e-06,
+      "loss": 1.0329,
+      "step": 2167
+    },
+    {
+      "epoch": 0.2280095179249872,
+      "grad_norm": 3.197164204980888,
+      "learning_rate": 4.427219329982253e-06,
+      "loss": 0.999,
+      "step": 2168
+    },
+    {
+      "epoch": 0.2281146883668345,
+      "grad_norm": 2.8107489657605944,
+      "learning_rate": 4.426687466712822e-06,
+      "loss": 1.02,
+      "step": 2169
+    },
+    {
+      "epoch": 0.2282198588086818,
+      "grad_norm": 2.7507863352021547,
+      "learning_rate": 4.426155388602771e-06,
+      "loss": 0.9941,
+      "step": 2170
+    },
+    {
+      "epoch": 0.22832502925052914,
+      "grad_norm": 2.0778977549643103,
+      "learning_rate": 4.425623095711431e-06,
+      "loss": 1.0075,
+      "step": 2171
+    },
+    {
+      "epoch": 0.22843019969237646,
+      "grad_norm": 2.5218199437660487,
+      "learning_rate": 4.425090588098158e-06,
+      "loss": 1.0061,
+      "step": 2172
+    },
+    {
+      "epoch": 0.2285353701342238,
+      "grad_norm": 1.8947896556435786,
+      "learning_rate": 4.424557865822329e-06,
+      "loss": 1.0319,
+      "step": 2173
+    },
+    {
+      "epoch": 0.22864054057607108,
+      "grad_norm": 2.8580135935444715,
+      "learning_rate": 4.424024928943347e-06,
+      "loss": 1.0138,
+      "step": 2174
+    },
+    {
+      "epoch": 0.2287457110179184,
+      "grad_norm": 1.9185065863265085,
+      "learning_rate": 4.42349177752064e-06,
+      "loss": 1.0251,
+      "step": 2175
+    },
+    {
+      "epoch": 0.22885088145976573,
+      "grad_norm": 2.2887816350371484,
+      "learning_rate": 4.4229584116136574e-06,
+      "loss": 1.0304,
+      "step": 2176
+    },
+    {
+      "epoch": 0.22895605190161306,
+      "grad_norm": 2.114307828197282,
+      "learning_rate": 4.422424831281873e-06,
+      "loss": 1.0362,
+      "step": 2177
+    },
+    {
+      "epoch": 0.22906122234346038,
+      "grad_norm": 3.6302171769750364,
+      "learning_rate": 4.421891036584787e-06,
+      "loss": 1.0028,
+      "step": 2178
+    },
+    {
+      "epoch": 0.22916639278530768,
+      "grad_norm": 3.4568600537373784,
+      "learning_rate": 4.421357027581921e-06,
+      "loss": 0.9796,
+      "step": 2179
+    },
+    {
+      "epoch": 0.229271563227155,
+      "grad_norm": 3.3494243915757016,
+      "learning_rate": 4.42082280433282e-06,
+      "loss": 1.0287,
+      "step": 2180
+    },
+    {
+      "epoch": 0.22937673366900233,
+      "grad_norm": 3.815143472620564,
+      "learning_rate": 4.420288366897055e-06,
+      "loss": 1.0211,
+      "step": 2181
+    },
+    {
+      "epoch": 0.22948190411084965,
+      "grad_norm": 2.134153845893804,
+      "learning_rate": 4.419753715334219e-06,
+      "loss": 1.032,
+      "step": 2182
+    },
+    {
+      "epoch": 0.22958707455269697,
+      "grad_norm": 2.6909666767089333,
+      "learning_rate": 4.419218849703933e-06,
+      "loss": 0.9872,
+      "step": 2183
+    },
+    {
+      "epoch": 0.22969224499454427,
+      "grad_norm": 2.3377268022126994,
+      "learning_rate": 4.418683770065834e-06,
+      "loss": 1.0023,
+      "step": 2184
+    },
+    {
+      "epoch": 0.2297974154363916,
+      "grad_norm": 3.1485034054066783,
+      "learning_rate": 4.41814847647959e-06,
+      "loss": 1.0084,
+      "step": 2185
+    },
+    {
+      "epoch": 0.22990258587823892,
+      "grad_norm": 2.354211954549771,
+      "learning_rate": 4.4176129690048905e-06,
+      "loss": 1.006,
+      "step": 2186
+    },
+    {
+      "epoch": 0.23000775632008624,
+      "grad_norm": 2.463169295196819,
+      "learning_rate": 4.417077247701449e-06,
+      "loss": 1.0514,
+      "step": 2187
+    },
+    {
+      "epoch": 0.23011292676193357,
+      "grad_norm": 2.130669171298515,
+      "learning_rate": 4.4165413126290015e-06,
+      "loss": 1.0368,
+      "step": 2188
+    },
+    {
+      "epoch": 0.23021809720378086,
+      "grad_norm": 2.6129366976545105,
+      "learning_rate": 4.416005163847309e-06,
+      "loss": 0.9739,
+      "step": 2189
+    },
+    {
+      "epoch": 0.2303232676456282,
+      "grad_norm": 2.84939027329353,
+      "learning_rate": 4.415468801416158e-06,
+      "loss": 0.993,
+      "step": 2190
+    },
+    {
+      "epoch": 0.2304284380874755,
+      "grad_norm": 2.426269871426963,
+      "learning_rate": 4.4149322253953545e-06,
+      "loss": 1.0335,
+      "step": 2191
+    },
+    {
+      "epoch": 0.23053360852932284,
+      "grad_norm": 2.379802288203194,
+      "learning_rate": 4.414395435844732e-06,
+      "loss": 0.9636,
+      "step": 2192
+    },
+    {
+      "epoch": 0.23063877897117016,
+      "grad_norm": 2.400251646859452,
+      "learning_rate": 4.413858432824147e-06,
+      "loss": 1.0368,
+      "step": 2193
+    },
+    {
+      "epoch": 0.23074394941301749,
+      "grad_norm": 2.8763286217663677,
+      "learning_rate": 4.41332121639348e-06,
+      "loss": 1.0031,
+      "step": 2194
+    },
+    {
+      "epoch": 0.23084911985486478,
+      "grad_norm": 3.2398926109360526,
+      "learning_rate": 4.412783786612634e-06,
+      "loss": 1.0469,
+      "step": 2195
+    },
+    {
+      "epoch": 0.2309542902967121,
+      "grad_norm": 2.465573849120777,
+      "learning_rate": 4.412246143541536e-06,
+      "loss": 1.015,
+      "step": 2196
+    },
+    {
+      "epoch": 0.23105946073855943,
+      "grad_norm": 2.7811425882506873,
+      "learning_rate": 4.4117082872401395e-06,
+      "loss": 0.9901,
+      "step": 2197
+    },
+    {
+      "epoch": 0.23116463118040675,
+      "grad_norm": 2.990852585631329,
+      "learning_rate": 4.411170217768417e-06,
+      "loss": 1.0293,
+      "step": 2198
+    },
+    {
+      "epoch": 0.23126980162225408,
+      "grad_norm": 3.261551379361568,
+      "learning_rate": 4.410631935186369e-06,
+      "loss": 1.0079,
+      "step": 2199
+    },
+    {
+      "epoch": 0.23137497206410138,
+      "grad_norm": 2.7992235343655527,
+      "learning_rate": 4.410093439554019e-06,
+      "loss": 1.0397,
+      "step": 2200
+    },
+    {
+      "epoch": 0.2314801425059487,
+      "grad_norm": 2.4192682778964336,
+      "learning_rate": 4.409554730931412e-06,
+      "loss": 1.0289,
+      "step": 2201
+    },
+    {
+      "epoch": 0.23158531294779602,
+      "grad_norm": 2.5957977370487235,
+      "learning_rate": 4.4090158093786175e-06,
+      "loss": 1.0366,
+      "step": 2202
+    },
+    {
+      "epoch": 0.23169048338964335,
+      "grad_norm": 2.8959476464448923,
+      "learning_rate": 4.408476674955732e-06,
+      "loss": 1.0785,
+      "step": 2203
+    },
+    {
+      "epoch": 0.23179565383149067,
+      "grad_norm": 2.644482199192657,
+      "learning_rate": 4.407937327722871e-06,
+      "loss": 1.0254,
+      "step": 2204
+    },
+    {
+      "epoch": 0.23190082427333797,
+      "grad_norm": 3.442594882430092,
+      "learning_rate": 4.407397767740176e-06,
+      "loss": 1.0397,
+      "step": 2205
+    },
+    {
+      "epoch": 0.2320059947151853,
+      "grad_norm": 2.5341412795753566,
+      "learning_rate": 4.406857995067813e-06,
+      "loss": 1.006,
+      "step": 2206
+    },
+    {
+      "epoch": 0.23211116515703262,
+      "grad_norm": 2.2043374296304994,
+      "learning_rate": 4.406318009765971e-06,
+      "loss": 1.0111,
+      "step": 2207
+    },
+    {
+      "epoch": 0.23221633559887994,
+      "grad_norm": 3.0723931658417487,
+      "learning_rate": 4.405777811894861e-06,
+      "loss": 1.043,
+      "step": 2208
+    },
+    {
+      "epoch": 0.23232150604072727,
+      "grad_norm": 2.364705705885201,
+      "learning_rate": 4.405237401514722e-06,
+      "loss": 1.0548,
+      "step": 2209
+    },
+    {
+      "epoch": 0.23242667648257456,
+      "grad_norm": 3.1983834328748246,
+      "learning_rate": 4.404696778685811e-06,
+      "loss": 1.0139,
+      "step": 2210
+    },
+    {
+      "epoch": 0.2325318469244219,
+      "grad_norm": 2.746332139516773,
+      "learning_rate": 4.4041559434684135e-06,
+      "loss": 1.02,
+      "step": 2211
+    },
+    {
+      "epoch": 0.2326370173662692,
+      "grad_norm": 2.274325848928587,
+      "learning_rate": 4.4036148959228365e-06,
+      "loss": 1.0089,
+      "step": 2212
+    },
+    {
+      "epoch": 0.23274218780811654,
+      "grad_norm": 3.0278107573054336,
+      "learning_rate": 4.40307363610941e-06,
+      "loss": 1.0081,
+      "step": 2213
+    },
+    {
+      "epoch": 0.23284735824996386,
+      "grad_norm": 2.165972003847509,
+      "learning_rate": 4.40253216408849e-06,
+      "loss": 1.0138,
+      "step": 2214
+    },
+    {
+      "epoch": 0.23295252869181116,
+      "grad_norm": 2.8880774293438334,
+      "learning_rate": 4.401990479920455e-06,
+      "loss": 0.9974,
+      "step": 2215
+    },
+    {
+      "epoch": 0.23305769913365848,
+      "grad_norm": 2.105729606603633,
+      "learning_rate": 4.401448583665706e-06,
+      "loss": 1.0269,
+      "step": 2216
+    },
+    {
+      "epoch": 0.2331628695755058,
+      "grad_norm": 2.5957248301304108,
+      "learning_rate": 4.4009064753846685e-06,
+      "loss": 1.0,
+      "step": 2217
+    },
+    {
+      "epoch": 0.23326804001735313,
+      "grad_norm": 3.0915416585982336,
+      "learning_rate": 4.400364155137793e-06,
+      "loss": 1.0225,
+      "step": 2218
+    },
+    {
+      "epoch": 0.23337321045920045,
+      "grad_norm": 2.174158247873258,
+      "learning_rate": 4.3998216229855504e-06,
+      "loss": 0.9767,
+      "step": 2219
+    },
+    {
+      "epoch": 0.23347838090104775,
+      "grad_norm": 1.6915485879041205,
+      "learning_rate": 4.3992788789884385e-06,
+      "loss": 0.9691,
+      "step": 2220
+    },
+    {
+      "epoch": 0.23358355134289507,
+      "grad_norm": 3.1989567531239733,
+      "learning_rate": 4.398735923206978e-06,
+      "loss": 1.0219,
+      "step": 2221
+    },
+    {
+      "epoch": 0.2336887217847424,
+      "grad_norm": 2.4905020268160234,
+      "learning_rate": 4.398192755701713e-06,
+      "loss": 1.0105,
+      "step": 2222
+    },
+    {
+      "epoch": 0.23379389222658972,
+      "grad_norm": 3.061385255458901,
+      "learning_rate": 4.397649376533209e-06,
+      "loss": 1.0154,
+      "step": 2223
+    },
+    {
+      "epoch": 0.23389906266843705,
+      "grad_norm": 2.1173291128116647,
+      "learning_rate": 4.397105785762059e-06,
+      "loss": 1.0071,
+      "step": 2224
+    },
+    {
+      "epoch": 0.23400423311028434,
+      "grad_norm": 2.4753927815387264,
+      "learning_rate": 4.396561983448876e-06,
+      "loss": 1.0191,
+      "step": 2225
+    },
+    {
+      "epoch": 0.23410940355213167,
+      "grad_norm": 2.6295178770980465,
+      "learning_rate": 4.396017969654299e-06,
+      "loss": 1.0107,
+      "step": 2226
+    },
+    {
+      "epoch": 0.234214573993979,
+      "grad_norm": 2.555842843500082,
+      "learning_rate": 4.395473744438989e-06,
+      "loss": 1.0316,
+      "step": 2227
+    },
+    {
+      "epoch": 0.23431974443582632,
+      "grad_norm": 2.698211855081485,
+      "learning_rate": 4.394929307863633e-06,
+      "loss": 1.0002,
+      "step": 2228
+    },
+    {
+      "epoch": 0.23442491487767364,
+      "grad_norm": 3.0360295797113963,
+      "learning_rate": 4.394384659988938e-06,
+      "loss": 0.9551,
+      "step": 2229
+    },
+    {
+      "epoch": 0.23453008531952094,
+      "grad_norm": 3.1523775915462826,
+      "learning_rate": 4.393839800875638e-06,
+      "loss": 0.9874,
+      "step": 2230
+    },
+    {
+      "epoch": 0.23463525576136826,
+      "grad_norm": 2.900257288408909,
+      "learning_rate": 4.3932947305844875e-06,
+      "loss": 0.9747,
+      "step": 2231
+    },
+    {
+      "epoch": 0.23474042620321559,
+      "grad_norm": 3.0099034185771005,
+      "learning_rate": 4.392749449176268e-06,
+      "loss": 1.0314,
+      "step": 2232
+    },
+    {
+      "epoch": 0.2348455966450629,
+      "grad_norm": 3.82061371167963,
+      "learning_rate": 4.39220395671178e-06,
+      "loss": 1.0465,
+      "step": 2233
+    },
+    {
+      "epoch": 0.23495076708691023,
+      "grad_norm": 2.833323567535811,
+      "learning_rate": 4.391658253251853e-06,
+      "loss": 1.0376,
+      "step": 2234
+    },
+    {
+      "epoch": 0.23505593752875753,
+      "grad_norm": 4.222429513815148,
+      "learning_rate": 4.391112338857335e-06,
+      "loss": 0.9761,
+      "step": 2235
+    },
+    {
+      "epoch": 0.23516110797060485,
+      "grad_norm": 2.8239821014267354,
+      "learning_rate": 4.390566213589101e-06,
+      "loss": 1.0447,
+      "step": 2236
+    },
+    {
+      "epoch": 0.23526627841245218,
+      "grad_norm": 2.47061316638576,
+      "learning_rate": 4.390019877508048e-06,
+      "loss": 1.0365,
+      "step": 2237
+    },
+    {
+      "epoch": 0.2353714488542995,
+      "grad_norm": 2.504173310706496,
+      "learning_rate": 4.389473330675096e-06,
+      "loss": 1.0194,
+      "step": 2238
+    },
+    {
+      "epoch": 0.23547661929614683,
+      "grad_norm": 3.4045598206190935,
+      "learning_rate": 4.38892657315119e-06,
+      "loss": 1.0359,
+      "step": 2239
+    },
+    {
+      "epoch": 0.23558178973799412,
+      "grad_norm": 3.5442386693147396,
+      "learning_rate": 4.388379604997297e-06,
+      "loss": 0.9705,
+      "step": 2240
+    },
+    {
+      "epoch": 0.23568696017984145,
+      "grad_norm": 3.938459399524716,
+      "learning_rate": 4.387832426274409e-06,
+      "loss": 1.0083,
+      "step": 2241
+    },
+    {
+      "epoch": 0.23579213062168877,
+      "grad_norm": 2.7251348273983993,
+      "learning_rate": 4.38728503704354e-06,
+      "loss": 1.0247,
+      "step": 2242
+    },
+    {
+      "epoch": 0.2358973010635361,
+      "grad_norm": 3.119137566805799,
+      "learning_rate": 4.386737437365729e-06,
+      "loss": 0.9493,
+      "step": 2243
+    },
+    {
+      "epoch": 0.23600247150538342,
+      "grad_norm": 2.526082167109994,
+      "learning_rate": 4.386189627302037e-06,
+      "loss": 1.008,
+      "step": 2244
+    },
+    {
+      "epoch": 0.23610764194723072,
+      "grad_norm": 2.5608821293229105,
+      "learning_rate": 4.385641606913548e-06,
+      "loss": 0.9885,
+      "step": 2245
+    },
+    {
+      "epoch": 0.23621281238907804,
+      "grad_norm": 2.682222019417078,
+      "learning_rate": 4.385093376261373e-06,
+      "loss": 1.0404,
+      "step": 2246
+    },
+    {
+      "epoch": 0.23631798283092537,
+      "grad_norm": 2.258432603781061,
+      "learning_rate": 4.384544935406641e-06,
+      "loss": 1.0196,
+      "step": 2247
+    },
+    {
+      "epoch": 0.2364231532727727,
+      "grad_norm": 2.671642366757651,
+      "learning_rate": 4.383996284410511e-06,
+      "loss": 1.024,
+      "step": 2248
+    },
+    {
+      "epoch": 0.23652832371462001,
+      "grad_norm": 3.1327248445757463,
+      "learning_rate": 4.383447423334159e-06,
+      "loss": 1.0182,
+      "step": 2249
+    },
+    {
+      "epoch": 0.2366334941564673,
+      "grad_norm": 2.723883446970864,
+      "learning_rate": 4.382898352238788e-06,
+      "loss": 0.9918,
+      "step": 2250
+    },
+    {
+      "epoch": 0.23673866459831464,
+      "grad_norm": 3.0308816991498118,
+      "learning_rate": 4.382349071185624e-06,
+      "loss": 1.0059,
+      "step": 2251
+    },
+    {
+      "epoch": 0.23684383504016196,
+      "grad_norm": 2.223810923909976,
+      "learning_rate": 4.3817995802359155e-06,
+      "loss": 1.0591,
+      "step": 2252
+    },
+    {
+      "epoch": 0.23694900548200928,
+      "grad_norm": 2.198306040920323,
+      "learning_rate": 4.3812498794509365e-06,
+      "loss": 1.0037,
+      "step": 2253
+    },
+    {
+      "epoch": 0.2370541759238566,
+      "grad_norm": 2.823104481198863,
+      "learning_rate": 4.38069996889198e-06,
+      "loss": 1.0164,
+      "step": 2254
+    },
+    {
+      "epoch": 0.23715934636570393,
+      "grad_norm": 2.3927642257493975,
+      "learning_rate": 4.380149848620369e-06,
+      "loss": 1.0045,
+      "step": 2255
+    },
+    {
+      "epoch": 0.23726451680755123,
+      "grad_norm": 2.7176141928001334,
+      "learning_rate": 4.379599518697444e-06,
+      "loss": 0.9935,
+      "step": 2256
+    },
+    {
+      "epoch": 0.23736968724939855,
+      "grad_norm": 2.386268088932299,
+      "learning_rate": 4.379048979184572e-06,
+      "loss": 1.0106,
+      "step": 2257
+    },
+    {
+      "epoch": 0.23747485769124588,
+      "grad_norm": 2.327011792366536,
+      "learning_rate": 4.37849823014314e-06,
+      "loss": 0.9859,
+      "step": 2258
+    },
+    {
+      "epoch": 0.2375800281330932,
+      "grad_norm": 2.310094818927127,
+      "learning_rate": 4.377947271634564e-06,
+      "loss": 1.0425,
+      "step": 2259
+    },
+    {
+      "epoch": 0.23768519857494053,
+      "grad_norm": 3.7252277158290026,
+      "learning_rate": 4.3773961037202784e-06,
+      "loss": 1.0013,
+      "step": 2260
+    },
+    {
+      "epoch": 0.23779036901678782,
+      "grad_norm": 2.864574664034268,
+      "learning_rate": 4.376844726461743e-06,
+      "loss": 0.9962,
+      "step": 2261
+    },
+    {
+      "epoch": 0.23789553945863515,
+      "grad_norm": 3.3321853074983405,
+      "learning_rate": 4.376293139920441e-06,
+      "loss": 1.0182,
+      "step": 2262
+    },
+    {
+      "epoch": 0.23800070990048247,
+      "grad_norm": 2.2402296812420173,
+      "learning_rate": 4.375741344157879e-06,
+      "loss": 1.0232,
+      "step": 2263
+    },
+    {
+      "epoch": 0.2381058803423298,
+      "grad_norm": 3.332880747822815,
+      "learning_rate": 4.375189339235585e-06,
+      "loss": 0.9578,
+      "step": 2264
+    },
+    {
+      "epoch": 0.23821105078417712,
+      "grad_norm": 2.4681929360535872,
+      "learning_rate": 4.374637125215113e-06,
+      "loss": 0.9791,
+      "step": 2265
+    },
+    {
+      "epoch": 0.23831622122602442,
+      "grad_norm": 2.404293531281789,
+      "learning_rate": 4.374084702158039e-06,
+      "loss": 1.0218,
+      "step": 2266
+    },
+    {
+      "epoch": 0.23842139166787174,
+      "grad_norm": 2.4396191881750577,
+      "learning_rate": 4.373532070125962e-06,
+      "loss": 1.0236,
+      "step": 2267
+    },
+    {
+      "epoch": 0.23852656210971906,
+      "grad_norm": 2.9682403601811367,
+      "learning_rate": 4.3729792291805055e-06,
+      "loss": 1.0397,
+      "step": 2268
+    },
+    {
+      "epoch": 0.2386317325515664,
+      "grad_norm": 3.1442251603576037,
+      "learning_rate": 4.372426179383315e-06,
+      "loss": 0.9759,
+      "step": 2269
+    },
+    {
+      "epoch": 0.2387369029934137,
+      "grad_norm": 2.7480986026706664,
+      "learning_rate": 4.371872920796059e-06,
+      "loss": 1.0032,
+      "step": 2270
+    },
+    {
+      "epoch": 0.238842073435261,
+      "grad_norm": 3.520061750412933,
+      "learning_rate": 4.371319453480431e-06,
+      "loss": 1.0073,
+      "step": 2271
+    },
+    {
+      "epoch": 0.23894724387710833,
+      "grad_norm": 2.622605154078674,
+      "learning_rate": 4.370765777498149e-06,
+      "loss": 1.0505,
+      "step": 2272
+    },
+    {
+      "epoch": 0.23905241431895566,
+      "grad_norm": 2.995803194412082,
+      "learning_rate": 4.370211892910949e-06,
+      "loss": 1.0239,
+      "step": 2273
+    },
+    {
+      "epoch": 0.23915758476080298,
+      "grad_norm": 3.0130918185780664,
+      "learning_rate": 4.369657799780595e-06,
+      "loss": 1.0446,
+      "step": 2274
+    },
+    {
+      "epoch": 0.2392627552026503,
+      "grad_norm": 2.2351494176209736,
+      "learning_rate": 4.369103498168872e-06,
+      "loss": 1.0292,
+      "step": 2275
+    },
+    {
+      "epoch": 0.2393679256444976,
+      "grad_norm": 2.710072469645011,
+      "learning_rate": 4.368548988137589e-06,
+      "loss": 1.045,
+      "step": 2276
+    },
+    {
+      "epoch": 0.23947309608634493,
+      "grad_norm": 2.3805553686213075,
+      "learning_rate": 4.367994269748579e-06,
+      "loss": 0.9925,
+      "step": 2277
+    },
+    {
+      "epoch": 0.23957826652819225,
+      "grad_norm": 3.089542808347871,
+      "learning_rate": 4.367439343063696e-06,
+      "loss": 1.0466,
+      "step": 2278
+    },
+    {
+      "epoch": 0.23968343697003958,
+      "grad_norm": 1.6810878146896986,
+      "learning_rate": 4.3668842081448206e-06,
+      "loss": 1.0232,
+      "step": 2279
+    },
+    {
+      "epoch": 0.2397886074118869,
+      "grad_norm": 2.8351406317406522,
+      "learning_rate": 4.366328865053853e-06,
+      "loss": 1.0121,
+      "step": 2280
+    },
+    {
+      "epoch": 0.2398937778537342,
+      "grad_norm": 2.2094407063080648,
+      "learning_rate": 4.365773313852718e-06,
+      "loss": 1.0166,
+      "step": 2281
+    },
+    {
+      "epoch": 0.23999894829558152,
+      "grad_norm": 2.816907367682445,
+      "learning_rate": 4.365217554603365e-06,
+      "loss": 1.0322,
+      "step": 2282
+    },
+    {
+      "epoch": 0.24010411873742885,
+      "grad_norm": 3.0356646249148382,
+      "learning_rate": 4.364661587367765e-06,
+      "loss": 0.9952,
+      "step": 2283
+    },
+    {
+      "epoch": 0.24020928917927617,
+      "grad_norm": 3.4033015359175898,
+      "learning_rate": 4.364105412207914e-06,
+      "loss": 1.043,
+      "step": 2284
+    },
+    {
+      "epoch": 0.2403144596211235,
+      "grad_norm": 3.068575576133874,
+      "learning_rate": 4.3635490291858285e-06,
+      "loss": 1.0411,
+      "step": 2285
+    },
+    {
+      "epoch": 0.2404196300629708,
+      "grad_norm": 2.863399553567626,
+      "learning_rate": 4.362992438363549e-06,
+      "loss": 0.9437,
+      "step": 2286
+    },
+    {
+      "epoch": 0.24052480050481811,
+      "grad_norm": 3.0443021645648094,
+      "learning_rate": 4.362435639803141e-06,
+      "loss": 0.9712,
+      "step": 2287
+    },
+    {
+      "epoch": 0.24062997094666544,
+      "grad_norm": 2.100247783006265,
+      "learning_rate": 4.361878633566691e-06,
+      "loss": 0.9919,
+      "step": 2288
+    },
+    {
+      "epoch": 0.24073514138851276,
+      "grad_norm": 2.0331761263216053,
+      "learning_rate": 4.361321419716309e-06,
+      "loss": 1.0414,
+      "step": 2289
+    },
+    {
+      "epoch": 0.2408403118303601,
+      "grad_norm": 2.4507988887485017,
+      "learning_rate": 4.3607639983141305e-06,
+      "loss": 1.0253,
+      "step": 2290
+    },
+    {
+      "epoch": 0.24094548227220738,
+      "grad_norm": 2.3383275471123235,
+      "learning_rate": 4.360206369422311e-06,
+      "loss": 0.9672,
+      "step": 2291
+    },
+    {
+      "epoch": 0.2410506527140547,
+      "grad_norm": 2.985680116206487,
+      "learning_rate": 4.35964853310303e-06,
+      "loss": 1.0054,
+      "step": 2292
+    },
+    {
+      "epoch": 0.24115582315590203,
+      "grad_norm": 3.0674737355369768,
+      "learning_rate": 4.359090489418492e-06,
+      "loss": 1.0293,
+      "step": 2293
+    },
+    {
+      "epoch": 0.24126099359774936,
+      "grad_norm": 3.179508934892898,
+      "learning_rate": 4.358532238430922e-06,
+      "loss": 1.0232,
+      "step": 2294
+    },
+    {
+      "epoch": 0.24136616403959668,
+      "grad_norm": 2.68704103252696,
+      "learning_rate": 4.35797378020257e-06,
+      "loss": 1.0297,
+      "step": 2295
+    },
+    {
+      "epoch": 0.24147133448144398,
+      "grad_norm": 3.002150660941129,
+      "learning_rate": 4.357415114795709e-06,
+      "loss": 1.0245,
+      "step": 2296
+    },
+    {
+      "epoch": 0.2415765049232913,
+      "grad_norm": 2.258390927515721,
+      "learning_rate": 4.356856242272634e-06,
+      "loss": 1.0379,
+      "step": 2297
+    },
+    {
+      "epoch": 0.24168167536513863,
+      "grad_norm": 2.24898790122129,
+      "learning_rate": 4.356297162695662e-06,
+      "loss": 0.9842,
+      "step": 2298
+    },
+    {
+      "epoch": 0.24178684580698595,
+      "grad_norm": 2.0293860274575515,
+      "learning_rate": 4.3557378761271365e-06,
+      "loss": 0.9971,
+      "step": 2299
+    },
+    {
+      "epoch": 0.24189201624883327,
+      "grad_norm": 2.049737362518493,
+      "learning_rate": 4.3551783826294225e-06,
+      "loss": 0.9803,
+      "step": 2300
+    },
+    {
+      "epoch": 0.24199718669068057,
+      "grad_norm": 2.82131721819364,
+      "learning_rate": 4.3546186822649074e-06,
+      "loss": 1.0371,
+      "step": 2301
+    },
+    {
+      "epoch": 0.2421023571325279,
+      "grad_norm": 2.6674301569769794,
+      "learning_rate": 4.3540587750960015e-06,
+      "loss": 1.0335,
+      "step": 2302
+    },
+    {
+      "epoch": 0.24220752757437522,
+      "grad_norm": 2.5133457003010315,
+      "learning_rate": 4.353498661185139e-06,
+      "loss": 1.0216,
+      "step": 2303
+    },
+    {
+      "epoch": 0.24231269801622254,
+      "grad_norm": 2.4664866984104425,
+      "learning_rate": 4.352938340594778e-06,
+      "loss": 0.9618,
+      "step": 2304
+    },
+    {
+      "epoch": 0.24241786845806987,
+      "grad_norm": 2.5932114027385276,
+      "learning_rate": 4.352377813387398e-06,
+      "loss": 0.9812,
+      "step": 2305
+    },
+    {
+      "epoch": 0.24252303889991716,
+      "grad_norm": 2.9653618102280928,
+      "learning_rate": 4.351817079625502e-06,
+      "loss": 1.0269,
+      "step": 2306
+    },
+    {
+      "epoch": 0.2426282093417645,
+      "grad_norm": 2.724639989437581,
+      "learning_rate": 4.351256139371616e-06,
+      "loss": 1.0426,
+      "step": 2307
+    },
+    {
+      "epoch": 0.2427333797836118,
+      "grad_norm": 2.475922956963675,
+      "learning_rate": 4.350694992688289e-06,
+      "loss": 1.0309,
+      "step": 2308
+    },
+    {
+      "epoch": 0.24283855022545914,
+      "grad_norm": 2.3209110579795906,
+      "learning_rate": 4.350133639638094e-06,
+      "loss": 0.9894,
+      "step": 2309
+    },
+    {
+      "epoch": 0.24294372066730646,
+      "grad_norm": 3.9449479325880183,
+      "learning_rate": 4.349572080283626e-06,
+      "loss": 1.0032,
+      "step": 2310
+    },
+    {
+      "epoch": 0.24304889110915376,
+      "grad_norm": 3.591306291688279,
+      "learning_rate": 4.349010314687503e-06,
+      "loss": 1.0259,
+      "step": 2311
+    },
+    {
+      "epoch": 0.24315406155100108,
+      "grad_norm": 2.805682382653168,
+      "learning_rate": 4.348448342912365e-06,
+      "loss": 1.0302,
+      "step": 2312
+    },
+    {
+      "epoch": 0.2432592319928484,
+      "grad_norm": 2.540123317376257,
+      "learning_rate": 4.347886165020879e-06,
+      "loss": 1.0104,
+      "step": 2313
+    },
+    {
+      "epoch": 0.24336440243469573,
+      "grad_norm": 2.305099301071069,
+      "learning_rate": 4.347323781075731e-06,
+      "loss": 0.9888,
+      "step": 2314
+    },
+    {
+      "epoch": 0.24346957287654306,
+      "grad_norm": 3.599970133750129,
+      "learning_rate": 4.34676119113963e-06,
+      "loss": 1.0483,
+      "step": 2315
+    },
+    {
+      "epoch": 0.24357474331839038,
+      "grad_norm": 3.770809111412246,
+      "learning_rate": 4.3461983952753105e-06,
+      "loss": 1.018,
+      "step": 2316
+    },
+    {
+      "epoch": 0.24367991376023768,
+      "grad_norm": 3.0932950050010013,
+      "learning_rate": 4.3456353935455285e-06,
+      "loss": 0.9798,
+      "step": 2317
+    },
+    {
+      "epoch": 0.243785084202085,
+      "grad_norm": 2.2871843195096364,
+      "learning_rate": 4.345072186013063e-06,
+      "loss": 1.0237,
+      "step": 2318
+    },
+    {
+      "epoch": 0.24389025464393232,
+      "grad_norm": 2.957870000494886,
+      "learning_rate": 4.344508772740714e-06,
+      "loss": 1.01,
+      "step": 2319
+    },
+    {
+      "epoch": 0.24399542508577965,
+      "grad_norm": 1.8700377021220989,
+      "learning_rate": 4.343945153791309e-06,
+      "loss": 0.9877,
+      "step": 2320
+    },
+    {
+      "epoch": 0.24410059552762697,
+      "grad_norm": 1.7017015392411625,
+      "learning_rate": 4.343381329227695e-06,
+      "loss": 0.9863,
+      "step": 2321
+    },
+    {
+      "epoch": 0.24420576596947427,
+      "grad_norm": 2.0639304419621642,
+      "learning_rate": 4.342817299112744e-06,
+      "loss": 0.9882,
+      "step": 2322
+    },
+    {
+      "epoch": 0.2443109364113216,
+      "grad_norm": 3.6180446386514125,
+      "learning_rate": 4.342253063509348e-06,
+      "loss": 1.0368,
+      "step": 2323
+    },
+    {
+      "epoch": 0.24441610685316892,
+      "grad_norm": 2.843315371054092,
+      "learning_rate": 4.341688622480423e-06,
+      "loss": 1.0174,
+      "step": 2324
+    },
+    {
+      "epoch": 0.24452127729501624,
+      "grad_norm": 2.315783436822103,
+      "learning_rate": 4.341123976088912e-06,
+      "loss": 1.0636,
+      "step": 2325
+    },
+    {
+      "epoch": 0.24462644773686357,
+      "grad_norm": 2.584613916276836,
+      "learning_rate": 4.340559124397774e-06,
+      "loss": 1.0318,
+      "step": 2326
+    },
+    {
+      "epoch": 0.24473161817871086,
+      "grad_norm": 3.106997533126451,
+      "learning_rate": 4.339994067469996e-06,
+      "loss": 0.9838,
+      "step": 2327
+    },
+    {
+      "epoch": 0.2448367886205582,
+      "grad_norm": 2.931771272566417,
+      "learning_rate": 4.339428805368586e-06,
+      "loss": 1.026,
+      "step": 2328
+    },
+    {
+      "epoch": 0.2449419590624055,
+      "grad_norm": 3.097355703100136,
+      "learning_rate": 4.338863338156575e-06,
+      "loss": 1.0357,
+      "step": 2329
+    },
+    {
+      "epoch": 0.24504712950425284,
+      "grad_norm": 2.163293960868166,
+      "learning_rate": 4.338297665897016e-06,
+      "loss": 1.0129,
+      "step": 2330
+    },
+    {
+      "epoch": 0.24515229994610016,
+      "grad_norm": 2.6395195900185966,
+      "learning_rate": 4.3377317886529875e-06,
+      "loss": 1.0473,
+      "step": 2331
+    },
+    {
+      "epoch": 0.24525747038794746,
+      "grad_norm": 2.857130604203511,
+      "learning_rate": 4.337165706487589e-06,
+      "loss": 1.0317,
+      "step": 2332
+    },
+    {
+      "epoch": 0.24536264082979478,
+      "grad_norm": 2.49538696157863,
+      "learning_rate": 4.336599419463943e-06,
+      "loss": 0.9921,
+      "step": 2333
+    },
+    {
+      "epoch": 0.2454678112716421,
+      "grad_norm": 3.3688127588253645,
+      "learning_rate": 4.336032927645194e-06,
+      "loss": 1.0419,
+      "step": 2334
+    },
+    {
+      "epoch": 0.24557298171348943,
+      "grad_norm": 2.4873830553852536,
+      "learning_rate": 4.3354662310945105e-06,
+      "loss": 1.0384,
+      "step": 2335
+    },
+    {
+      "epoch": 0.24567815215533675,
+      "grad_norm": 4.367565068639842,
+      "learning_rate": 4.334899329875084e-06,
+      "loss": 1.0342,
+      "step": 2336
+    },
+    {
+      "epoch": 0.24578332259718405,
+      "grad_norm": 2.3864309309189227,
+      "learning_rate": 4.3343322240501276e-06,
+      "loss": 0.9888,
+      "step": 2337
+    },
+    {
+      "epoch": 0.24588849303903137,
+      "grad_norm": 2.2279995974847298,
+      "learning_rate": 4.33376491368288e-06,
+      "loss": 1.0035,
+      "step": 2338
+    },
+    {
+      "epoch": 0.2459936634808787,
+      "grad_norm": 2.8286281778671287,
+      "learning_rate": 4.333197398836598e-06,
+      "loss": 0.9848,
+      "step": 2339
+    },
+    {
+      "epoch": 0.24609883392272602,
+      "grad_norm": 1.9274410042281873,
+      "learning_rate": 4.332629679574566e-06,
+      "loss": 1.0054,
+      "step": 2340
+    },
+    {
+      "epoch": 0.24620400436457335,
+      "grad_norm": 2.5323024007120893,
+      "learning_rate": 4.332061755960089e-06,
+      "loss": 1.0341,
+      "step": 2341
+    },
+    {
+      "epoch": 0.24630917480642064,
+      "grad_norm": 2.8771741834278948,
+      "learning_rate": 4.331493628056494e-06,
+      "loss": 1.0287,
+      "step": 2342
+    },
+    {
+      "epoch": 0.24641434524826797,
+      "grad_norm": 3.559675628437571,
+      "learning_rate": 4.33092529592713e-06,
+      "loss": 1.0325,
+      "step": 2343
+    },
+    {
+      "epoch": 0.2465195156901153,
+      "grad_norm": 2.4388512232776023,
+      "learning_rate": 4.330356759635374e-06,
+      "loss": 1.0221,
+      "step": 2344
+    },
+    {
+      "epoch": 0.24662468613196262,
+      "grad_norm": 2.682329033109531,
+      "learning_rate": 4.329788019244621e-06,
+      "loss": 0.9862,
+      "step": 2345
+    },
+    {
+      "epoch": 0.24672985657380994,
+      "grad_norm": 2.5463446910483367,
+      "learning_rate": 4.329219074818288e-06,
+      "loss": 1.0284,
+      "step": 2346
+    },
+    {
+      "epoch": 0.24683502701565724,
+      "grad_norm": 2.0296846553885253,
+      "learning_rate": 4.32864992641982e-06,
+      "loss": 1.0119,
+      "step": 2347
+    },
+    {
+      "epoch": 0.24694019745750456,
+      "grad_norm": 2.048540561932806,
+      "learning_rate": 4.328080574112678e-06,
+      "loss": 1.01,
+      "step": 2348
+    },
+    {
+      "epoch": 0.2470453678993519,
+      "grad_norm": 3.4200047911393745,
+      "learning_rate": 4.327511017960353e-06,
+      "loss": 1.0236,
+      "step": 2349
+    },
+    {
+      "epoch": 0.2471505383411992,
+      "grad_norm": 2.777887461388932,
+      "learning_rate": 4.326941258026352e-06,
+      "loss": 1.0413,
+      "step": 2350
+    },
+    {
+      "epoch": 0.24725570878304653,
+      "grad_norm": 2.6334314381177175,
+      "learning_rate": 4.326371294374207e-06,
+      "loss": 1.0092,
+      "step": 2351
+    },
+    {
+      "epoch": 0.24736087922489383,
+      "grad_norm": 2.880908366865062,
+      "learning_rate": 4.325801127067477e-06,
+      "loss": 1.0373,
+      "step": 2352
+    },
+    {
+      "epoch": 0.24746604966674116,
+      "grad_norm": 3.588695174592861,
+      "learning_rate": 4.325230756169737e-06,
+      "loss": 1.0001,
+      "step": 2353
+    },
+    {
+      "epoch": 0.24757122010858848,
+      "grad_norm": 3.0543565117198535,
+      "learning_rate": 4.324660181744589e-06,
+      "loss": 0.9983,
+      "step": 2354
+    },
+    {
+      "epoch": 0.2476763905504358,
+      "grad_norm": 2.650395143277798,
+      "learning_rate": 4.324089403855657e-06,
+      "loss": 1.0016,
+      "step": 2355
+    },
+    {
+      "epoch": 0.24778156099228313,
+      "grad_norm": 2.339046826958962,
+      "learning_rate": 4.323518422566586e-06,
+      "loss": 1.021,
+      "step": 2356
+    },
+    {
+      "epoch": 0.24788673143413043,
+      "grad_norm": 2.6212259533171647,
+      "learning_rate": 4.322947237941045e-06,
+      "loss": 0.9759,
+      "step": 2357
+    },
+    {
+      "epoch": 0.24799190187597775,
+      "grad_norm": 3.383293390834341,
+      "learning_rate": 4.322375850042726e-06,
+      "loss": 0.9904,
+      "step": 2358
+    },
+    {
+      "epoch": 0.24809707231782507,
+      "grad_norm": 2.670237539236923,
+      "learning_rate": 4.321804258935343e-06,
+      "loss": 1.0199,
+      "step": 2359
+    },
+    {
+      "epoch": 0.2482022427596724,
+      "grad_norm": 3.250184677256926,
+      "learning_rate": 4.3212324646826334e-06,
+      "loss": 0.9948,
+      "step": 2360
+    },
+    {
+      "epoch": 0.24830741320151972,
+      "grad_norm": 2.5906224312948916,
+      "learning_rate": 4.320660467348355e-06,
+      "loss": 1.0054,
+      "step": 2361
+    },
+    {
+      "epoch": 0.24841258364336702,
+      "grad_norm": 2.3329999607620695,
+      "learning_rate": 4.320088266996292e-06,
+      "loss": 1.0111,
+      "step": 2362
+    },
+    {
+      "epoch": 0.24851775408521434,
+      "grad_norm": 3.94765814630664,
+      "learning_rate": 4.3195158636902475e-06,
+      "loss": 1.0006,
+      "step": 2363
+    },
+    {
+      "epoch": 0.24862292452706167,
+      "grad_norm": 2.247678998715992,
+      "learning_rate": 4.3189432574940506e-06,
+      "loss": 0.9504,
+      "step": 2364
+    },
+    {
+      "epoch": 0.248728094968909,
+      "grad_norm": 2.638346815741289,
+      "learning_rate": 4.3183704484715496e-06,
+      "loss": 1.016,
+      "step": 2365
+    },
+    {
+      "epoch": 0.24883326541075632,
+      "grad_norm": 2.9361318540962094,
+      "learning_rate": 4.3177974366866185e-06,
+      "loss": 1.0447,
+      "step": 2366
+    },
+    {
+      "epoch": 0.2489384358526036,
+      "grad_norm": 3.1042757839794515,
+      "learning_rate": 4.317224222203152e-06,
+      "loss": 1.0061,
+      "step": 2367
+    },
+    {
+      "epoch": 0.24904360629445094,
+      "grad_norm": 2.2154767610485426,
+      "learning_rate": 4.316650805085068e-06,
+      "loss": 1.005,
+      "step": 2368
+    },
+    {
+      "epoch": 0.24914877673629826,
+      "grad_norm": 3.623072584794595,
+      "learning_rate": 4.316077185396307e-06,
+      "loss": 1.0467,
+      "step": 2369
+    },
+    {
+      "epoch": 0.24925394717814559,
+      "grad_norm": 2.33674243869589,
+      "learning_rate": 4.3155033632008316e-06,
+      "loss": 1.0021,
+      "step": 2370
+    },
+    {
+      "epoch": 0.2493591176199929,
+      "grad_norm": 2.8007703287989907,
+      "learning_rate": 4.314929338562629e-06,
+      "loss": 1.0135,
+      "step": 2371
+    },
+    {
+      "epoch": 0.24946428806184023,
+      "grad_norm": 3.2196932437208923,
+      "learning_rate": 4.314355111545705e-06,
+      "loss": 0.9289,
+      "step": 2372
+    },
+    {
+      "epoch": 0.24956945850368753,
+      "grad_norm": 3.334365342086359,
+      "learning_rate": 4.313780682214093e-06,
+      "loss": 1.0507,
+      "step": 2373
+    },
+    {
+      "epoch": 0.24967462894553485,
+      "grad_norm": 2.4745573038357915,
+      "learning_rate": 4.313206050631844e-06,
+      "loss": 0.9509,
+      "step": 2374
+    },
+    {
+      "epoch": 0.24977979938738218,
+      "grad_norm": 2.0261724785390123,
+      "learning_rate": 4.312631216863035e-06,
+      "loss": 1.0073,
+      "step": 2375
+    },
+    {
+      "epoch": 0.2498849698292295,
+      "grad_norm": 3.786354346442156,
+      "learning_rate": 4.312056180971764e-06,
+      "loss": 1.0409,
+      "step": 2376
+    },
+    {
+      "epoch": 0.24999014027107683,
+      "grad_norm": 1.8293455049956973,
+      "learning_rate": 4.3114809430221534e-06,
+      "loss": 0.983,
+      "step": 2377
+    },
+    {
+      "epoch": 0.2500953107129241,
+      "grad_norm": 2.7551244030560023,
+      "learning_rate": 4.3109055030783445e-06,
+      "loss": 1.0099,
+      "step": 2378
+    },
+    {
+      "epoch": 0.25020048115477145,
+      "grad_norm": 3.713006572607176,
+      "learning_rate": 4.310329861204505e-06,
+      "loss": 0.9658,
+      "step": 2379
+    },
+    {
+      "epoch": 0.2503056515966188,
+      "grad_norm": 3.841601007385603,
+      "learning_rate": 4.309754017464823e-06,
+      "loss": 1.0511,
+      "step": 2380
+    },
+    {
+      "epoch": 0.2504108220384661,
+      "grad_norm": 1.7189850598284027,
+      "learning_rate": 4.309177971923509e-06,
+      "loss": 1.0225,
+      "step": 2381
+    },
+    {
+      "epoch": 0.2505159924803134,
+      "grad_norm": 2.3106913101555326,
+      "learning_rate": 4.308601724644797e-06,
+      "loss": 1.0436,
+      "step": 2382
+    },
+    {
+      "epoch": 0.25062116292216075,
+      "grad_norm": 1.685769997718884,
+      "learning_rate": 4.308025275692943e-06,
+      "loss": 0.9828,
+      "step": 2383
+    },
+    {
+      "epoch": 0.25072633336400807,
+      "grad_norm": 2.605151083100969,
+      "learning_rate": 4.307448625132225e-06,
+      "loss": 0.9986,
+      "step": 2384
+    },
+    {
+      "epoch": 0.25083150380585534,
+      "grad_norm": 2.070105328779204,
+      "learning_rate": 4.306871773026945e-06,
+      "loss": 0.9713,
+      "step": 2385
+    },
+    {
+      "epoch": 0.25093667424770266,
+      "grad_norm": 2.1511866343693007,
+      "learning_rate": 4.306294719441426e-06,
+      "loss": 1.0466,
+      "step": 2386
+    },
+    {
+      "epoch": 0.25104184468955,
+      "grad_norm": 2.7021103206947936,
+      "learning_rate": 4.305717464440015e-06,
+      "loss": 1.042,
+      "step": 2387
+    },
+    {
+      "epoch": 0.2511470151313973,
+      "grad_norm": 2.264135547093338,
+      "learning_rate": 4.305140008087078e-06,
+      "loss": 1.0355,
+      "step": 2388
+    },
+    {
+      "epoch": 0.25125218557324464,
+      "grad_norm": 2.6174496430229994,
+      "learning_rate": 4.304562350447008e-06,
+      "loss": 1.0403,
+      "step": 2389
+    },
+    {
+      "epoch": 0.25135735601509196,
+      "grad_norm": 2.5030793337590103,
+      "learning_rate": 4.303984491584217e-06,
+      "loss": 0.9831,
+      "step": 2390
+    },
+    {
+      "epoch": 0.2514625264569393,
+      "grad_norm": 3.246883328280729,
+      "learning_rate": 4.303406431563142e-06,
+      "loss": 1.0792,
+      "step": 2391
+    },
+    {
+      "epoch": 0.2515676968987866,
+      "grad_norm": 2.7687583296326004,
+      "learning_rate": 4.30282817044824e-06,
+      "loss": 1.0085,
+      "step": 2392
+    },
+    {
+      "epoch": 0.25167286734063393,
+      "grad_norm": 2.3272942288985554,
+      "learning_rate": 4.302249708303993e-06,
+      "loss": 1.0151,
+      "step": 2393
+    },
+    {
+      "epoch": 0.25177803778248126,
+      "grad_norm": 2.266231486995089,
+      "learning_rate": 4.301671045194902e-06,
+      "loss": 1.0047,
+      "step": 2394
+    },
+    {
+      "epoch": 0.2518832082243285,
+      "grad_norm": 2.754988205132692,
+      "learning_rate": 4.301092181185495e-06,
+      "loss": 1.0253,
+      "step": 2395
+    },
+    {
+      "epoch": 0.25198837866617585,
+      "grad_norm": 2.6929620972255255,
+      "learning_rate": 4.300513116340317e-06,
+      "loss": 1.0054,
+      "step": 2396
+    },
+    {
+      "epoch": 0.2520935491080232,
+      "grad_norm": 2.336085325218602,
+      "learning_rate": 4.2999338507239395e-06,
+      "loss": 0.9848,
+      "step": 2397
+    },
+    {
+      "epoch": 0.2521987195498705,
+      "grad_norm": 2.461142882427421,
+      "learning_rate": 4.299354384400957e-06,
+      "loss": 1.0059,
+      "step": 2398
+    },
+    {
+      "epoch": 0.2523038899917178,
+      "grad_norm": 2.532391792296108,
+      "learning_rate": 4.2987747174359805e-06,
+      "loss": 0.9754,
+      "step": 2399
+    },
+    {
+      "epoch": 0.25240906043356515,
+      "grad_norm": 2.621685341882318,
+      "learning_rate": 4.298194849893651e-06,
+      "loss": 1.0266,
+      "step": 2400
+    },
+    {
+      "epoch": 0.25251423087541247,
+      "grad_norm": 2.3181343970978356,
+      "learning_rate": 4.297614781838627e-06,
+      "loss": 1.0272,
+      "step": 2401
+    },
+    {
+      "epoch": 0.2526194013172598,
+      "grad_norm": 2.05334980419308,
+      "learning_rate": 4.297034513335591e-06,
+      "loss": 0.9885,
+      "step": 2402
+    },
+    {
+      "epoch": 0.2527245717591071,
+      "grad_norm": 2.364971791424021,
+      "learning_rate": 4.296454044449245e-06,
+      "loss": 1.009,
+      "step": 2403
+    },
+    {
+      "epoch": 0.25282974220095444,
+      "grad_norm": 2.0265896786829805,
+      "learning_rate": 4.295873375244319e-06,
+      "loss": 1.0018,
+      "step": 2404
+    },
+    {
+      "epoch": 0.25293491264280177,
+      "grad_norm": 4.311217157785901,
+      "learning_rate": 4.295292505785561e-06,
+      "loss": 1.0421,
+      "step": 2405
+    },
+    {
+      "epoch": 0.25304008308464904,
+      "grad_norm": 2.629094308586008,
+      "learning_rate": 4.294711436137742e-06,
+      "loss": 1.0348,
+      "step": 2406
+    },
+    {
+      "epoch": 0.25314525352649636,
+      "grad_norm": 2.680285943391119,
+      "learning_rate": 4.294130166365656e-06,
+      "loss": 1.057,
+      "step": 2407
+    },
+    {
+      "epoch": 0.2532504239683437,
+      "grad_norm": 2.5244192892893267,
+      "learning_rate": 4.293548696534119e-06,
+      "loss": 1.0008,
+      "step": 2408
+    },
+    {
+      "epoch": 0.253355594410191,
+      "grad_norm": 2.1567089358080076,
+      "learning_rate": 4.2929670267079695e-06,
+      "loss": 1.0427,
+      "step": 2409
+    },
+    {
+      "epoch": 0.25346076485203833,
+      "grad_norm": 2.4191444936394277,
+      "learning_rate": 4.292385156952069e-06,
+      "loss": 1.0596,
+      "step": 2410
+    },
+    {
+      "epoch": 0.25356593529388566,
+      "grad_norm": 2.260012880541108,
+      "learning_rate": 4.291803087331299e-06,
+      "loss": 1.0076,
+      "step": 2411
+    },
+    {
+      "epoch": 0.253671105735733,
+      "grad_norm": 1.9375892516696325,
+      "learning_rate": 4.291220817910566e-06,
+      "loss": 1.036,
+      "step": 2412
+    },
+    {
+      "epoch": 0.2537762761775803,
+      "grad_norm": 2.425680002791181,
+      "learning_rate": 4.290638348754797e-06,
+      "loss": 1.0195,
+      "step": 2413
+    },
+    {
+      "epoch": 0.25388144661942763,
+      "grad_norm": 2.1474551573197074,
+      "learning_rate": 4.2900556799289415e-06,
+      "loss": 0.9866,
+      "step": 2414
+    },
+    {
+      "epoch": 0.25398661706127496,
+      "grad_norm": 2.3436746364584877,
+      "learning_rate": 4.289472811497972e-06,
+      "loss": 1.0447,
+      "step": 2415
+    },
+    {
+      "epoch": 0.2540917875031222,
+      "grad_norm": 2.114334641282732,
+      "learning_rate": 4.288889743526884e-06,
+      "loss": 1.0044,
+      "step": 2416
+    },
+    {
+      "epoch": 0.25419695794496955,
+      "grad_norm": 2.5018267220432846,
+      "learning_rate": 4.288306476080691e-06,
+      "loss": 1.061,
+      "step": 2417
+    },
+    {
+      "epoch": 0.2543021283868169,
+      "grad_norm": 3.079700459263164,
+      "learning_rate": 4.287723009224436e-06,
+      "loss": 1.0394,
+      "step": 2418
+    },
+    {
+      "epoch": 0.2544072988286642,
+      "grad_norm": 2.6029768657221672,
+      "learning_rate": 4.2871393430231775e-06,
+      "loss": 1.0305,
+      "step": 2419
+    },
+    {
+      "epoch": 0.2545124692705115,
+      "grad_norm": 3.4032647927169966,
+      "learning_rate": 4.2865554775419985e-06,
+      "loss": 1.0707,
+      "step": 2420
+    },
+    {
+      "epoch": 0.25461763971235885,
+      "grad_norm": 2.9725437594396964,
+      "learning_rate": 4.2859714128460065e-06,
+      "loss": 0.9889,
+      "step": 2421
+    },
+    {
+      "epoch": 0.25472281015420617,
+      "grad_norm": 1.908112429787501,
+      "learning_rate": 4.2853871490003265e-06,
+      "loss": 0.9918,
+      "step": 2422
+    },
+    {
+      "epoch": 0.2548279805960535,
+      "grad_norm": 2.4844690233952913,
+      "learning_rate": 4.2848026860701104e-06,
+      "loss": 1.0429,
+      "step": 2423
+    },
+    {
+      "epoch": 0.2549331510379008,
+      "grad_norm": 2.7736489003252034,
+      "learning_rate": 4.284218024120531e-06,
+      "loss": 1.0414,
+      "step": 2424
+    },
+    {
+      "epoch": 0.25503832147974814,
+      "grad_norm": 2.532322414340705,
+      "learning_rate": 4.28363316321678e-06,
+      "loss": 1.001,
+      "step": 2425
+    },
+    {
+      "epoch": 0.2551434919215954,
+      "grad_norm": 2.970052877513903,
+      "learning_rate": 4.283048103424077e-06,
+      "loss": 1.0803,
+      "step": 2426
+    },
+    {
+      "epoch": 0.25524866236344274,
+      "grad_norm": 3.4153519083462185,
+      "learning_rate": 4.282462844807659e-06,
+      "loss": 1.01,
+      "step": 2427
+    },
+    {
+      "epoch": 0.25535383280529006,
+      "grad_norm": 3.1053802029271513,
+      "learning_rate": 4.281877387432787e-06,
+      "loss": 1.051,
+      "step": 2428
+    },
+    {
+      "epoch": 0.2554590032471374,
+      "grad_norm": 2.6052490168349975,
+      "learning_rate": 4.281291731364744e-06,
+      "loss": 0.9926,
+      "step": 2429
+    },
+    {
+      "epoch": 0.2555641736889847,
+      "grad_norm": 2.1996478485684903,
+      "learning_rate": 4.280705876668836e-06,
+      "loss": 1.0286,
+      "step": 2430
+    },
+    {
+      "epoch": 0.25566934413083203,
+      "grad_norm": 2.4448585973957906,
+      "learning_rate": 4.2801198234103895e-06,
+      "loss": 1.0489,
+      "step": 2431
+    },
+    {
+      "epoch": 0.25577451457267936,
+      "grad_norm": 2.605626730548949,
+      "learning_rate": 4.279533571654754e-06,
+      "loss": 1.0234,
+      "step": 2432
+    },
+    {
+      "epoch": 0.2558796850145267,
+      "grad_norm": 2.2571136636808773,
+      "learning_rate": 4.2789471214673e-06,
+      "loss": 1.049,
+      "step": 2433
+    },
+    {
+      "epoch": 0.255984855456374,
+      "grad_norm": 2.9495392792916615,
+      "learning_rate": 4.278360472913424e-06,
+      "loss": 0.9787,
+      "step": 2434
+    },
+    {
+      "epoch": 0.25609002589822133,
+      "grad_norm": 2.0896220524241884,
+      "learning_rate": 4.277773626058542e-06,
+      "loss": 1.0116,
+      "step": 2435
+    },
+    {
+      "epoch": 0.2561951963400686,
+      "grad_norm": 3.163234673182296,
+      "learning_rate": 4.277186580968088e-06,
+      "loss": 1.0366,
+      "step": 2436
+    },
+    {
+      "epoch": 0.2563003667819159,
+      "grad_norm": 2.462962363222071,
+      "learning_rate": 4.276599337707525e-06,
+      "loss": 1.0296,
+      "step": 2437
+    },
+    {
+      "epoch": 0.25640553722376325,
+      "grad_norm": 2.1865954009193524,
+      "learning_rate": 4.276011896342336e-06,
+      "loss": 1.0365,
+      "step": 2438
+    },
+    {
+      "epoch": 0.25651070766561057,
+      "grad_norm": 2.3965938792682144,
+      "learning_rate": 4.2754242569380226e-06,
+      "loss": 1.0514,
+      "step": 2439
+    },
+    {
+      "epoch": 0.2566158781074579,
+      "grad_norm": 2.428078705061389,
+      "learning_rate": 4.2748364195601135e-06,
+      "loss": 1.0428,
+      "step": 2440
+    },
+    {
+      "epoch": 0.2567210485493052,
+      "grad_norm": 2.7590755653461203,
+      "learning_rate": 4.274248384274156e-06,
+      "loss": 0.9769,
+      "step": 2441
+    },
+    {
+      "epoch": 0.25682621899115254,
+      "grad_norm": 1.9677366099461735,
+      "learning_rate": 4.27366015114572e-06,
+      "loss": 0.985,
+      "step": 2442
+    },
+    {
+      "epoch": 0.25693138943299987,
+      "grad_norm": 2.209987282212459,
+      "learning_rate": 4.2730717202404e-06,
+      "loss": 1.0309,
+      "step": 2443
+    },
+    {
+      "epoch": 0.2570365598748472,
+      "grad_norm": 1.9968714101047118,
+      "learning_rate": 4.272483091623809e-06,
+      "loss": 1.0394,
+      "step": 2444
+    },
+    {
+      "epoch": 0.2571417303166945,
+      "grad_norm": 2.579965658298572,
+      "learning_rate": 4.271894265361584e-06,
+      "loss": 0.9979,
+      "step": 2445
+    },
+    {
+      "epoch": 0.2572469007585418,
+      "grad_norm": 1.8660957651544707,
+      "learning_rate": 4.2713052415193844e-06,
+      "loss": 1.0358,
+      "step": 2446
+    },
+    {
+      "epoch": 0.2573520712003891,
+      "grad_norm": 2.4422132950894757,
+      "learning_rate": 4.27071602016289e-06,
+      "loss": 1.0568,
+      "step": 2447
+    },
+    {
+      "epoch": 0.25745724164223643,
+      "grad_norm": 3.129384441162122,
+      "learning_rate": 4.270126601357804e-06,
+      "loss": 1.085,
+      "step": 2448
+    },
+    {
+      "epoch": 0.25756241208408376,
+      "grad_norm": 2.813236158519783,
+      "learning_rate": 4.269536985169851e-06,
+      "loss": 1.0045,
+      "step": 2449
+    },
+    {
+      "epoch": 0.2576675825259311,
+      "grad_norm": 2.4100320645639863,
+      "learning_rate": 4.268947171664778e-06,
+      "loss": 1.0047,
+      "step": 2450
+    },
+    {
+      "epoch": 0.2577727529677784,
+      "grad_norm": 2.939524262190411,
+      "learning_rate": 4.268357160908354e-06,
+      "loss": 1.0213,
+      "step": 2451
+    },
+    {
+      "epoch": 0.25787792340962573,
+      "grad_norm": 2.450057161789921,
+      "learning_rate": 4.267766952966369e-06,
+      "loss": 1.0147,
+      "step": 2452
+    },
+    {
+      "epoch": 0.25798309385147306,
+      "grad_norm": 3.105852562406945,
+      "learning_rate": 4.267176547904636e-06,
+      "loss": 0.9983,
+      "step": 2453
+    },
+    {
+      "epoch": 0.2580882642933204,
+      "grad_norm": 3.0495170539629894,
+      "learning_rate": 4.266585945788991e-06,
+      "loss": 0.9887,
+      "step": 2454
+    },
+    {
+      "epoch": 0.2581934347351677,
+      "grad_norm": 1.9555953731591447,
+      "learning_rate": 4.26599514668529e-06,
+      "loss": 1.0166,
+      "step": 2455
+    },
+    {
+      "epoch": 0.258298605177015,
+      "grad_norm": 2.345172111269998,
+      "learning_rate": 4.26540415065941e-06,
+      "loss": 1.0157,
+      "step": 2456
+    },
+    {
+      "epoch": 0.2584037756188623,
+      "grad_norm": 2.648580133103009,
+      "learning_rate": 4.264812957777253e-06,
+      "loss": 1.059,
+      "step": 2457
+    },
+    {
+      "epoch": 0.2585089460607096,
+      "grad_norm": 2.8657246045783755,
+      "learning_rate": 4.2642215681047425e-06,
+      "loss": 1.0181,
+      "step": 2458
+    },
+    {
+      "epoch": 0.25861411650255695,
+      "grad_norm": 2.593705422659073,
+      "learning_rate": 4.263629981707823e-06,
+      "loss": 1.0275,
+      "step": 2459
+    },
+    {
+      "epoch": 0.25871928694440427,
+      "grad_norm": 2.733876182487878,
+      "learning_rate": 4.263038198652459e-06,
+      "loss": 0.9784,
+      "step": 2460
+    },
+    {
+      "epoch": 0.2588244573862516,
+      "grad_norm": 3.183024692281709,
+      "learning_rate": 4.26244621900464e-06,
+      "loss": 1.0185,
+      "step": 2461
+    },
+    {
+      "epoch": 0.2589296278280989,
+      "grad_norm": 2.7344593577023715,
+      "learning_rate": 4.261854042830377e-06,
+      "loss": 1.0289,
+      "step": 2462
+    },
+    {
+      "epoch": 0.25903479826994624,
+      "grad_norm": 1.9703850254076136,
+      "learning_rate": 4.261261670195701e-06,
+      "loss": 0.9684,
+      "step": 2463
+    },
+    {
+      "epoch": 0.25913996871179357,
+      "grad_norm": 2.910924468557452,
+      "learning_rate": 4.2606691011666675e-06,
+      "loss": 1.0207,
+      "step": 2464
+    },
+    {
+      "epoch": 0.2592451391536409,
+      "grad_norm": 3.5563615470552548,
+      "learning_rate": 4.260076335809351e-06,
+      "loss": 1.0237,
+      "step": 2465
+    },
+    {
+      "epoch": 0.2593503095954882,
+      "grad_norm": 3.665977688640787,
+      "learning_rate": 4.259483374189851e-06,
+      "loss": 0.9983,
+      "step": 2466
+    },
+    {
+      "epoch": 0.2594554800373355,
+      "grad_norm": 3.749669082591241,
+      "learning_rate": 4.258890216374286e-06,
+      "loss": 1.0253,
+      "step": 2467
+    },
+    {
+      "epoch": 0.2595606504791828,
+      "grad_norm": 2.2413255969070303,
+      "learning_rate": 4.258296862428799e-06,
+      "loss": 1.0029,
+      "step": 2468
+    },
+    {
+      "epoch": 0.25966582092103013,
+      "grad_norm": 2.267636209673735,
+      "learning_rate": 4.257703312419553e-06,
+      "loss": 1.0052,
+      "step": 2469
+    },
+    {
+      "epoch": 0.25977099136287746,
+      "grad_norm": 2.2484357476994177,
+      "learning_rate": 4.257109566412733e-06,
+      "loss": 1.045,
+      "step": 2470
+    },
+    {
+      "epoch": 0.2598761618047248,
+      "grad_norm": 2.933825598950081,
+      "learning_rate": 4.256515624474547e-06,
+      "loss": 1.0228,
+      "step": 2471
+    },
+    {
+      "epoch": 0.2599813322465721,
+      "grad_norm": 2.404569611194018,
+      "learning_rate": 4.255921486671223e-06,
+      "loss": 1.0117,
+      "step": 2472
+    },
+    {
+      "epoch": 0.26008650268841943,
+      "grad_norm": 2.6554997150272595,
+      "learning_rate": 4.255327153069014e-06,
+      "loss": 1.0379,
+      "step": 2473
+    },
+    {
+      "epoch": 0.26019167313026675,
+      "grad_norm": 2.912876480944375,
+      "learning_rate": 4.25473262373419e-06,
+      "loss": 1.0201,
+      "step": 2474
+    },
+    {
+      "epoch": 0.2602968435721141,
+      "grad_norm": 2.4808723598184925,
+      "learning_rate": 4.254137898733048e-06,
+      "loss": 1.0086,
+      "step": 2475
+    },
+    {
+      "epoch": 0.2604020140139614,
+      "grad_norm": 3.0077031830519627,
+      "learning_rate": 4.253542978131904e-06,
+      "loss": 1.0183,
+      "step": 2476
+    },
+    {
+      "epoch": 0.26050718445580867,
+      "grad_norm": 2.119327566527547,
+      "learning_rate": 4.252947861997096e-06,
+      "loss": 1.0165,
+      "step": 2477
+    },
+    {
+      "epoch": 0.260612354897656,
+      "grad_norm": 2.9506512336323585,
+      "learning_rate": 4.2523525503949835e-06,
+      "loss": 1.0186,
+      "step": 2478
+    },
+    {
+      "epoch": 0.2607175253395033,
+      "grad_norm": 3.4059461780090414,
+      "learning_rate": 4.251757043391949e-06,
+      "loss": 1.0065,
+      "step": 2479
+    },
+    {
+      "epoch": 0.26082269578135064,
+      "grad_norm": 3.159826449070252,
+      "learning_rate": 4.251161341054396e-06,
+      "loss": 1.0176,
+      "step": 2480
+    },
+    {
+      "epoch": 0.26092786622319797,
+      "grad_norm": 3.0165983438656765,
+      "learning_rate": 4.25056544344875e-06,
+      "loss": 1.0177,
+      "step": 2481
+    },
+    {
+      "epoch": 0.2610330366650453,
+      "grad_norm": 2.909172263879279,
+      "learning_rate": 4.249969350641459e-06,
+      "loss": 1.0157,
+      "step": 2482
+    },
+    {
+      "epoch": 0.2611382071068926,
+      "grad_norm": 2.629076914941897,
+      "learning_rate": 4.2493730626989905e-06,
+      "loss": 1.0504,
+      "step": 2483
+    },
+    {
+      "epoch": 0.26124337754873994,
+      "grad_norm": 2.790385279161908,
+      "learning_rate": 4.248776579687835e-06,
+      "loss": 1.0102,
+      "step": 2484
+    },
+    {
+      "epoch": 0.26134854799058727,
+      "grad_norm": 3.4197241373079224,
+      "learning_rate": 4.248179901674508e-06,
+      "loss": 1.0153,
+      "step": 2485
+    },
+    {
+      "epoch": 0.2614537184324346,
+      "grad_norm": 2.7065753512349597,
+      "learning_rate": 4.247583028725539e-06,
+      "loss": 0.9738,
+      "step": 2486
+    },
+    {
+      "epoch": 0.26155888887428186,
+      "grad_norm": 2.0539337784982474,
+      "learning_rate": 4.246985960907488e-06,
+      "loss": 1.0371,
+      "step": 2487
+    },
+    {
+      "epoch": 0.2616640593161292,
+      "grad_norm": 3.5039520400652115,
+      "learning_rate": 4.2463886982869316e-06,
+      "loss": 1.0311,
+      "step": 2488
+    },
+    {
+      "epoch": 0.2617692297579765,
+      "grad_norm": 2.260518449174576,
+      "learning_rate": 4.245791240930469e-06,
+      "loss": 1.0418,
+      "step": 2489
+    },
+    {
+      "epoch": 0.26187440019982383,
+      "grad_norm": 2.593178309237441,
+      "learning_rate": 4.245193588904721e-06,
+      "loss": 0.9884,
+      "step": 2490
+    },
+    {
+      "epoch": 0.26197957064167116,
+      "grad_norm": 2.5646604102114807,
+      "learning_rate": 4.244595742276329e-06,
+      "loss": 1.0198,
+      "step": 2491
+    },
+    {
+      "epoch": 0.2620847410835185,
+      "grad_norm": 2.1573764067407284,
+      "learning_rate": 4.243997701111961e-06,
+      "loss": 1.0283,
+      "step": 2492
+    },
+    {
+      "epoch": 0.2621899115253658,
+      "grad_norm": 2.706034557337528,
+      "learning_rate": 4.243399465478302e-06,
+      "loss": 0.9915,
+      "step": 2493
+    },
+    {
+      "epoch": 0.26229508196721313,
+      "grad_norm": 2.8335990248067704,
+      "learning_rate": 4.242801035442059e-06,
+      "loss": 0.9332,
+      "step": 2494
+    },
+    {
+      "epoch": 0.26240025240906045,
+      "grad_norm": 2.633396099123692,
+      "learning_rate": 4.242202411069961e-06,
+      "loss": 0.9786,
+      "step": 2495
+    },
+    {
+      "epoch": 0.2625054228509078,
+      "grad_norm": 3.360271499168492,
+      "learning_rate": 4.241603592428761e-06,
+      "loss": 1.0477,
+      "step": 2496
+    },
+    {
+      "epoch": 0.26261059329275505,
+      "grad_norm": 2.148918489099196,
+      "learning_rate": 4.241004579585231e-06,
+      "loss": 1.0063,
+      "step": 2497
+    },
+    {
+      "epoch": 0.26271576373460237,
+      "grad_norm": 3.2367845205717,
+      "learning_rate": 4.240405372606165e-06,
+      "loss": 1.0015,
+      "step": 2498
+    },
+    {
+      "epoch": 0.2628209341764497,
+      "grad_norm": 2.7895205043470988,
+      "learning_rate": 4.239805971558381e-06,
+      "loss": 1.0499,
+      "step": 2499
+    },
+    {
+      "epoch": 0.262926104618297,
+      "grad_norm": 1.9537882626779242,
+      "learning_rate": 4.239206376508716e-06,
+      "loss": 1.0381,
+      "step": 2500
+    },
+    {
+      "epoch": 0.26303127506014434,
+      "grad_norm": 2.784638128216924,
+      "learning_rate": 4.2386065875240295e-06,
+      "loss": 1.0069,
+      "step": 2501
+    },
+    {
+      "epoch": 0.26313644550199167,
+      "grad_norm": 2.9037401874074096,
+      "learning_rate": 4.238006604671202e-06,
+      "loss": 0.997,
+      "step": 2502
+    },
+    {
+      "epoch": 0.263241615943839,
+      "grad_norm": 3.7907319889378095,
+      "learning_rate": 4.2374064280171375e-06,
+      "loss": 1.0153,
+      "step": 2503
+    },
+    {
+      "epoch": 0.2633467863856863,
+      "grad_norm": 2.42146316099506,
+      "learning_rate": 4.2368060576287586e-06,
+      "loss": 0.9756,
+      "step": 2504
+    },
+    {
+      "epoch": 0.26345195682753364,
+      "grad_norm": 3.3487340200239943,
+      "learning_rate": 4.236205493573013e-06,
+      "loss": 1.0098,
+      "step": 2505
+    },
+    {
+      "epoch": 0.26355712726938096,
+      "grad_norm": 2.3850429416001337,
+      "learning_rate": 4.235604735916867e-06,
+      "loss": 1.0055,
+      "step": 2506
+    },
+    {
+      "epoch": 0.26366229771122823,
+      "grad_norm": 2.849867107035556,
+      "learning_rate": 4.235003784727311e-06,
+      "loss": 1.0175,
+      "step": 2507
+    },
+    {
+      "epoch": 0.26376746815307556,
+      "grad_norm": 3.019032768395256,
+      "learning_rate": 4.234402640071355e-06,
+      "loss": 1.0389,
+      "step": 2508
+    },
+    {
+      "epoch": 0.2638726385949229,
+      "grad_norm": 1.6410605133020546,
+      "learning_rate": 4.23380130201603e-06,
+      "loss": 1.0201,
+      "step": 2509
+    },
+    {
+      "epoch": 0.2639778090367702,
+      "grad_norm": 2.8300981371727656,
+      "learning_rate": 4.233199770628392e-06,
+      "loss": 1.0188,
+      "step": 2510
+    },
+    {
+      "epoch": 0.26408297947861753,
+      "grad_norm": 2.5403242188674,
+      "learning_rate": 4.2325980459755155e-06,
+      "loss": 1.0789,
+      "step": 2511
+    },
+    {
+      "epoch": 0.26418814992046485,
+      "grad_norm": 2.6388880305238147,
+      "learning_rate": 4.231996128124498e-06,
+      "loss": 0.9696,
+      "step": 2512
+    },
+    {
+      "epoch": 0.2642933203623122,
+      "grad_norm": 3.679088388538602,
+      "learning_rate": 4.231394017142458e-06,
+      "loss": 1.0035,
+      "step": 2513
+    },
+    {
+      "epoch": 0.2643984908041595,
+      "grad_norm": 2.153085146074569,
+      "learning_rate": 4.230791713096534e-06,
+      "loss": 0.9416,
+      "step": 2514
+    },
+    {
+      "epoch": 0.2645036612460068,
+      "grad_norm": 2.9035280609022696,
+      "learning_rate": 4.2301892160538895e-06,
+      "loss": 1.0398,
+      "step": 2515
+    },
+    {
+      "epoch": 0.26460883168785415,
+      "grad_norm": 3.382690191668748,
+      "learning_rate": 4.229586526081707e-06,
+      "loss": 1.0457,
+      "step": 2516
+    },
+    {
+      "epoch": 0.2647140021297014,
+      "grad_norm": 2.7928016735192167,
+      "learning_rate": 4.22898364324719e-06,
+      "loss": 0.9843,
+      "step": 2517
+    },
+    {
+      "epoch": 0.26481917257154874,
+      "grad_norm": 1.3854879435295921,
+      "learning_rate": 4.228380567617566e-06,
+      "loss": 1.06,
+      "step": 2518
+    },
+    {
+      "epoch": 0.26492434301339607,
+      "grad_norm": 2.3894512440051945,
+      "learning_rate": 4.227777299260083e-06,
+      "loss": 1.0271,
+      "step": 2519
+    },
+    {
+      "epoch": 0.2650295134552434,
+      "grad_norm": 2.409763930345137,
+      "learning_rate": 4.227173838242008e-06,
+      "loss": 1.0086,
+      "step": 2520
+    },
+    {
+      "epoch": 0.2651346838970907,
+      "grad_norm": 2.300365291967152,
+      "learning_rate": 4.2265701846306336e-06,
+      "loss": 1.0139,
+      "step": 2521
+    },
+    {
+      "epoch": 0.26523985433893804,
+      "grad_norm": 2.9413706238859936,
+      "learning_rate": 4.225966338493272e-06,
+      "loss": 1.0238,
+      "step": 2522
+    },
+    {
+      "epoch": 0.26534502478078537,
+      "grad_norm": 3.3804230105479385,
+      "learning_rate": 4.225362299897254e-06,
+      "loss": 1.0536,
+      "step": 2523
+    },
+    {
+      "epoch": 0.2654501952226327,
+      "grad_norm": 2.708985307345623,
+      "learning_rate": 4.2247580689099376e-06,
+      "loss": 0.9965,
+      "step": 2524
+    },
+    {
+      "epoch": 0.26555536566448,
+      "grad_norm": 2.8007246157283494,
+      "learning_rate": 4.224153645598698e-06,
+      "loss": 1.033,
+      "step": 2525
+    },
+    {
+      "epoch": 0.26566053610632734,
+      "grad_norm": 2.552503024762742,
+      "learning_rate": 4.2235490300309324e-06,
+      "loss": 1.0094,
+      "step": 2526
+    },
+    {
+      "epoch": 0.26576570654817466,
+      "grad_norm": 2.761858798258985,
+      "learning_rate": 4.222944222274062e-06,
+      "loss": 1.0003,
+      "step": 2527
+    },
+    {
+      "epoch": 0.26587087699002193,
+      "grad_norm": 2.5736511815926844,
+      "learning_rate": 4.222339222395526e-06,
+      "loss": 0.9731,
+      "step": 2528
+    },
+    {
+      "epoch": 0.26597604743186926,
+      "grad_norm": 2.6976780654834744,
+      "learning_rate": 4.221734030462787e-06,
+      "loss": 1.0515,
+      "step": 2529
+    },
+    {
+      "epoch": 0.2660812178737166,
+      "grad_norm": 3.4009547412639187,
+      "learning_rate": 4.22112864654333e-06,
+      "loss": 1.02,
+      "step": 2530
+    },
+    {
+      "epoch": 0.2661863883155639,
+      "grad_norm": 2.060001991686356,
+      "learning_rate": 4.220523070704657e-06,
+      "loss": 0.967,
+      "step": 2531
+    },
+    {
+      "epoch": 0.26629155875741123,
+      "grad_norm": 3.326975711226428,
+      "learning_rate": 4.219917303014297e-06,
+      "loss": 1.0475,
+      "step": 2532
+    },
+    {
+      "epoch": 0.26639672919925855,
+      "grad_norm": 3.6695180566563743,
+      "learning_rate": 4.219311343539797e-06,
+      "loss": 1.0338,
+      "step": 2533
+    },
+    {
+      "epoch": 0.2665018996411059,
+      "grad_norm": 3.402820484725127,
+      "learning_rate": 4.218705192348725e-06,
+      "loss": 1.0143,
+      "step": 2534
+    },
+    {
+      "epoch": 0.2666070700829532,
+      "grad_norm": 2.611669788607023,
+      "learning_rate": 4.2180988495086745e-06,
+      "loss": 1.0101,
+      "step": 2535
+    },
+    {
+      "epoch": 0.2667122405248005,
+      "grad_norm": 2.1839819006885364,
+      "learning_rate": 4.217492315087255e-06,
+      "loss": 1.0269,
+      "step": 2536
+    },
+    {
+      "epoch": 0.26681741096664785,
+      "grad_norm": 2.6726735252473715,
+      "learning_rate": 4.2168855891520995e-06,
+      "loss": 1.0632,
+      "step": 2537
+    },
+    {
+      "epoch": 0.2669225814084951,
+      "grad_norm": 2.8199626949758305,
+      "learning_rate": 4.216278671770865e-06,
+      "loss": 1.0688,
+      "step": 2538
+    },
+    {
+      "epoch": 0.26702775185034244,
+      "grad_norm": 3.170146232476931,
+      "learning_rate": 4.2156715630112245e-06,
+      "loss": 0.9838,
+      "step": 2539
+    },
+    {
+      "epoch": 0.26713292229218977,
+      "grad_norm": 3.0560237745595513,
+      "learning_rate": 4.215064262940878e-06,
+      "loss": 1.0116,
+      "step": 2540
+    },
+    {
+      "epoch": 0.2672380927340371,
+      "grad_norm": 2.510504357329309,
+      "learning_rate": 4.214456771627542e-06,
+      "loss": 0.9855,
+      "step": 2541
+    },
+    {
+      "epoch": 0.2673432631758844,
+      "grad_norm": 2.5311289381637163,
+      "learning_rate": 4.213849089138959e-06,
+      "loss": 1.0087,
+      "step": 2542
+    },
+    {
+      "epoch": 0.26744843361773174,
+      "grad_norm": 2.244207627746343,
+      "learning_rate": 4.213241215542888e-06,
+      "loss": 0.9875,
+      "step": 2543
+    },
+    {
+      "epoch": 0.26755360405957906,
+      "grad_norm": 1.9425305151864143,
+      "learning_rate": 4.2126331509071125e-06,
+      "loss": 0.9958,
+      "step": 2544
+    },
+    {
+      "epoch": 0.2676587745014264,
+      "grad_norm": 2.895384366283835,
+      "learning_rate": 4.212024895299437e-06,
+      "loss": 0.9963,
+      "step": 2545
+    },
+    {
+      "epoch": 0.2677639449432737,
+      "grad_norm": 2.6493035965793443,
+      "learning_rate": 4.211416448787686e-06,
+      "loss": 1.0128,
+      "step": 2546
+    },
+    {
+      "epoch": 0.26786911538512104,
+      "grad_norm": 2.303330006776538,
+      "learning_rate": 4.210807811439707e-06,
+      "loss": 1.0284,
+      "step": 2547
+    },
+    {
+      "epoch": 0.2679742858269683,
+      "grad_norm": 2.6701335223547833,
+      "learning_rate": 4.210198983323366e-06,
+      "loss": 1.0172,
+      "step": 2548
+    },
+    {
+      "epoch": 0.26807945626881563,
+      "grad_norm": 2.548255245355431,
+      "learning_rate": 4.209589964506553e-06,
+      "loss": 1.0094,
+      "step": 2549
+    },
+    {
+      "epoch": 0.26818462671066295,
+      "grad_norm": 2.5662458724791417,
+      "learning_rate": 4.2089807550571786e-06,
+      "loss": 1.0284,
+      "step": 2550
+    },
+    {
+      "epoch": 0.2682897971525103,
+      "grad_norm": 2.7555215764610255,
+      "learning_rate": 4.208371355043174e-06,
+      "loss": 0.9936,
+      "step": 2551
+    },
+    {
+      "epoch": 0.2683949675943576,
+      "grad_norm": 1.6391645215285977,
+      "learning_rate": 4.207761764532493e-06,
+      "loss": 1.0162,
+      "step": 2552
+    },
+    {
+      "epoch": 0.2685001380362049,
+      "grad_norm": 2.861451794419541,
+      "learning_rate": 4.207151983593109e-06,
+      "loss": 1.0493,
+      "step": 2553
+    },
+    {
+      "epoch": 0.26860530847805225,
+      "grad_norm": 3.000252612225211,
+      "learning_rate": 4.206542012293016e-06,
+      "loss": 0.9965,
+      "step": 2554
+    },
+    {
+      "epoch": 0.2687104789198996,
+      "grad_norm": 3.0326862394669813,
+      "learning_rate": 4.205931850700232e-06,
+      "loss": 1.0026,
+      "step": 2555
+    },
+    {
+      "epoch": 0.2688156493617469,
+      "grad_norm": 2.6432936556834026,
+      "learning_rate": 4.205321498882795e-06,
+      "loss": 1.0279,
+      "step": 2556
+    },
+    {
+      "epoch": 0.2689208198035942,
+      "grad_norm": 2.9469947026091146,
+      "learning_rate": 4.204710956908763e-06,
+      "loss": 0.9982,
+      "step": 2557
+    },
+    {
+      "epoch": 0.2690259902454415,
+      "grad_norm": 3.0253527353338447,
+      "learning_rate": 4.204100224846217e-06,
+      "loss": 1.0196,
+      "step": 2558
+    },
+    {
+      "epoch": 0.2691311606872888,
+      "grad_norm": 2.8517981562593753,
+      "learning_rate": 4.203489302763258e-06,
+      "loss": 1.0275,
+      "step": 2559
+    },
+    {
+      "epoch": 0.26923633112913614,
+      "grad_norm": 2.4348277733854418,
+      "learning_rate": 4.202878190728009e-06,
+      "loss": 0.9938,
+      "step": 2560
+    },
+    {
+      "epoch": 0.26934150157098347,
+      "grad_norm": 3.561918888528251,
+      "learning_rate": 4.202266888808613e-06,
+      "loss": 1.041,
+      "step": 2561
+    },
+    {
+      "epoch": 0.2694466720128308,
+      "grad_norm": 2.0281988514230394,
+      "learning_rate": 4.201655397073234e-06,
+      "loss": 1.0193,
+      "step": 2562
+    },
+    {
+      "epoch": 0.2695518424546781,
+      "grad_norm": 2.6293480142232317,
+      "learning_rate": 4.2010437155900605e-06,
+      "loss": 1.0125,
+      "step": 2563
+    },
+    {
+      "epoch": 0.26965701289652544,
+      "grad_norm": 3.250687419475308,
+      "learning_rate": 4.200431844427299e-06,
+      "loss": 0.993,
+      "step": 2564
+    },
+    {
+      "epoch": 0.26976218333837276,
+      "grad_norm": 2.42448998982233,
+      "learning_rate": 4.199819783653177e-06,
+      "loss": 1.0116,
+      "step": 2565
+    },
+    {
+      "epoch": 0.2698673537802201,
+      "grad_norm": 2.6337946712929936,
+      "learning_rate": 4.199207533335944e-06,
+      "loss": 0.9941,
+      "step": 2566
+    },
+    {
+      "epoch": 0.2699725242220674,
+      "grad_norm": 2.710054204566471,
+      "learning_rate": 4.198595093543871e-06,
+      "loss": 0.9949,
+      "step": 2567
+    },
+    {
+      "epoch": 0.2700776946639147,
+      "grad_norm": 2.697631288571504,
+      "learning_rate": 4.197982464345251e-06,
+      "loss": 0.9725,
+      "step": 2568
+    },
+    {
+      "epoch": 0.270182865105762,
+      "grad_norm": 2.5276904086562926,
+      "learning_rate": 4.197369645808394e-06,
+      "loss": 1.0171,
+      "step": 2569
+    },
+    {
+      "epoch": 0.27028803554760933,
+      "grad_norm": 2.46942882217501,
+      "learning_rate": 4.196756638001638e-06,
+      "loss": 1.0111,
+      "step": 2570
+    },
+    {
+      "epoch": 0.27039320598945665,
+      "grad_norm": 2.657116364982858,
+      "learning_rate": 4.196143440993335e-06,
+      "loss": 1.0135,
+      "step": 2571
+    },
+    {
+      "epoch": 0.270498376431304,
+      "grad_norm": 2.173209952959072,
+      "learning_rate": 4.195530054851863e-06,
+      "loss": 0.9677,
+      "step": 2572
+    },
+    {
+      "epoch": 0.2706035468731513,
+      "grad_norm": 2.22871510749567,
+      "learning_rate": 4.1949164796456174e-06,
+      "loss": 1.0164,
+      "step": 2573
+    },
+    {
+      "epoch": 0.2707087173149986,
+      "grad_norm": 1.903168210740841,
+      "learning_rate": 4.194302715443018e-06,
+      "loss": 0.9581,
+      "step": 2574
+    },
+    {
+      "epoch": 0.27081388775684595,
+      "grad_norm": 1.8872564936631262,
+      "learning_rate": 4.193688762312504e-06,
+      "loss": 0.9807,
+      "step": 2575
+    },
+    {
+      "epoch": 0.2709190581986933,
+      "grad_norm": 2.456914305309123,
+      "learning_rate": 4.193074620322536e-06,
+      "loss": 1.0195,
+      "step": 2576
+    },
+    {
+      "epoch": 0.2710242286405406,
+      "grad_norm": 3.185302938492518,
+      "learning_rate": 4.192460289541596e-06,
+      "loss": 0.9971,
+      "step": 2577
+    },
+    {
+      "epoch": 0.27112939908238787,
+      "grad_norm": 3.314584198817348,
+      "learning_rate": 4.191845770038186e-06,
+      "loss": 0.9779,
+      "step": 2578
+    },
+    {
+      "epoch": 0.2712345695242352,
+      "grad_norm": 2.876042479155621,
+      "learning_rate": 4.19123106188083e-06,
+      "loss": 0.9777,
+      "step": 2579
+    },
+    {
+      "epoch": 0.2713397399660825,
+      "grad_norm": 2.446043436959705,
+      "learning_rate": 4.1906161651380725e-06,
+      "loss": 1.0048,
+      "step": 2580
+    },
+    {
+      "epoch": 0.27144491040792984,
+      "grad_norm": 2.376061110547315,
+      "learning_rate": 4.19000107987848e-06,
+      "loss": 1.0115,
+      "step": 2581
+    },
+    {
+      "epoch": 0.27155008084977716,
+      "grad_norm": 2.681106684018155,
+      "learning_rate": 4.189385806170637e-06,
+      "loss": 1.027,
+      "step": 2582
+    },
+    {
+      "epoch": 0.2716552512916245,
+      "grad_norm": 3.098522232486938,
+      "learning_rate": 4.188770344083155e-06,
+      "loss": 0.9921,
+      "step": 2583
+    },
+    {
+      "epoch": 0.2717604217334718,
+      "grad_norm": 2.90935844056823,
+      "learning_rate": 4.188154693684659e-06,
+      "loss": 1.0271,
+      "step": 2584
+    },
+    {
+      "epoch": 0.27186559217531914,
+      "grad_norm": 3.1110671923281936,
+      "learning_rate": 4.187538855043802e-06,
+      "loss": 1.0198,
+      "step": 2585
+    },
+    {
+      "epoch": 0.27197076261716646,
+      "grad_norm": 2.682316070595872,
+      "learning_rate": 4.186922828229254e-06,
+      "loss": 1.0307,
+      "step": 2586
+    },
+    {
+      "epoch": 0.2720759330590138,
+      "grad_norm": 3.015097650865152,
+      "learning_rate": 4.186306613309704e-06,
+      "loss": 1.0326,
+      "step": 2587
+    },
+    {
+      "epoch": 0.2721811035008611,
+      "grad_norm": 2.366488727509157,
+      "learning_rate": 4.185690210353869e-06,
+      "loss": 1.0285,
+      "step": 2588
+    },
+    {
+      "epoch": 0.2722862739427084,
+      "grad_norm": 2.6756347836291523,
+      "learning_rate": 4.185073619430479e-06,
+      "loss": 1.0118,
+      "step": 2589
+    },
+    {
+      "epoch": 0.2723914443845557,
+      "grad_norm": 2.001448901496582,
+      "learning_rate": 4.184456840608291e-06,
+      "loss": 1.0067,
+      "step": 2590
+    },
+    {
+      "epoch": 0.272496614826403,
+      "grad_norm": 2.425491517107179,
+      "learning_rate": 4.183839873956081e-06,
+      "loss": 0.9963,
+      "step": 2591
+    },
+    {
+      "epoch": 0.27260178526825035,
+      "grad_norm": 3.4036870716254417,
+      "learning_rate": 4.183222719542643e-06,
+      "loss": 0.9999,
+      "step": 2592
+    },
+    {
+      "epoch": 0.2727069557100977,
+      "grad_norm": 2.720894946053787,
+      "learning_rate": 4.182605377436797e-06,
+      "loss": 1.0125,
+      "step": 2593
+    },
+    {
+      "epoch": 0.272812126151945,
+      "grad_norm": 2.6441136268156793,
+      "learning_rate": 4.18198784770738e-06,
+      "loss": 1.0377,
+      "step": 2594
+    },
+    {
+      "epoch": 0.2729172965937923,
+      "grad_norm": 3.0457102136253273,
+      "learning_rate": 4.1813701304232515e-06,
+      "loss": 1.0487,
+      "step": 2595
+    },
+    {
+      "epoch": 0.27302246703563965,
+      "grad_norm": 2.559547849904084,
+      "learning_rate": 4.1807522256532925e-06,
+      "loss": 0.9605,
+      "step": 2596
+    },
+    {
+      "epoch": 0.273127637477487,
+      "grad_norm": 1.8862838127732708,
+      "learning_rate": 4.1801341334664035e-06,
+      "loss": 0.9979,
+      "step": 2597
+    },
+    {
+      "epoch": 0.2732328079193343,
+      "grad_norm": 3.6119148619521337,
+      "learning_rate": 4.179515853931507e-06,
+      "loss": 1.0307,
+      "step": 2598
+    },
+    {
+      "epoch": 0.27333797836118157,
+      "grad_norm": 3.286509646757304,
+      "learning_rate": 4.178897387117547e-06,
+      "loss": 0.9789,
+      "step": 2599
+    },
+    {
+      "epoch": 0.2734431488030289,
+      "grad_norm": 2.7213519725866573,
+      "learning_rate": 4.178278733093485e-06,
+      "loss": 1.0111,
+      "step": 2600
+    },
+    {
+      "epoch": 0.2735483192448762,
+      "grad_norm": 2.352945473282292,
+      "learning_rate": 4.177659891928307e-06,
+      "loss": 1.0025,
+      "step": 2601
+    },
+    {
+      "epoch": 0.27365348968672354,
+      "grad_norm": 2.5229083023446015,
+      "learning_rate": 4.1770408636910185e-06,
+      "loss": 1.0267,
+      "step": 2602
+    },
+    {
+      "epoch": 0.27375866012857086,
+      "grad_norm": 2.818848626261983,
+      "learning_rate": 4.176421648450646e-06,
+      "loss": 0.9997,
+      "step": 2603
+    },
+    {
+      "epoch": 0.2738638305704182,
+      "grad_norm": 2.737597916134103,
+      "learning_rate": 4.175802246276237e-06,
+      "loss": 1.0182,
+      "step": 2604
+    },
+    {
+      "epoch": 0.2739690010122655,
+      "grad_norm": 2.3272831659875393,
+      "learning_rate": 4.1751826572368596e-06,
+      "loss": 1.0206,
+      "step": 2605
+    },
+    {
+      "epoch": 0.27407417145411284,
+      "grad_norm": 3.029868785399325,
+      "learning_rate": 4.174562881401602e-06,
+      "loss": 1.0163,
+      "step": 2606
+    },
+    {
+      "epoch": 0.27417934189596016,
+      "grad_norm": 3.336473793800413,
+      "learning_rate": 4.173942918839576e-06,
+      "loss": 0.976,
+      "step": 2607
+    },
+    {
+      "epoch": 0.2742845123378075,
+      "grad_norm": 2.6514956429533942,
+      "learning_rate": 4.17332276961991e-06,
+      "loss": 1.0177,
+      "step": 2608
+    },
+    {
+      "epoch": 0.27438968277965475,
+      "grad_norm": 2.679420445962811,
+      "learning_rate": 4.172702433811756e-06,
+      "loss": 1.0326,
+      "step": 2609
+    },
+    {
+      "epoch": 0.2744948532215021,
+      "grad_norm": 3.6351143708377616,
+      "learning_rate": 4.172081911484287e-06,
+      "loss": 1.005,
+      "step": 2610
+    },
+    {
+      "epoch": 0.2746000236633494,
+      "grad_norm": 2.9056209865867184,
+      "learning_rate": 4.171461202706696e-06,
+      "loss": 1.0271,
+      "step": 2611
+    },
+    {
+      "epoch": 0.2747051941051967,
+      "grad_norm": 2.417998566001529,
+      "learning_rate": 4.170840307548196e-06,
+      "loss": 1.02,
+      "step": 2612
+    },
+    {
+      "epoch": 0.27481036454704405,
+      "grad_norm": 2.9211736184925505,
+      "learning_rate": 4.170219226078023e-06,
+      "loss": 1.0503,
+      "step": 2613
+    },
+    {
+      "epoch": 0.2749155349888914,
+      "grad_norm": 2.418767483745108,
+      "learning_rate": 4.169597958365431e-06,
+      "loss": 1.0352,
+      "step": 2614
+    },
+    {
+      "epoch": 0.2750207054307387,
+      "grad_norm": 1.9237097504393978,
+      "learning_rate": 4.1689765044796965e-06,
+      "loss": 1.0098,
+      "step": 2615
+    },
+    {
+      "epoch": 0.275125875872586,
+      "grad_norm": 2.6838641339374902,
+      "learning_rate": 4.168354864490117e-06,
+      "loss": 1.0082,
+      "step": 2616
+    },
+    {
+      "epoch": 0.27523104631443335,
+      "grad_norm": 2.577137912600681,
+      "learning_rate": 4.16773303846601e-06,
+      "loss": 1.0114,
+      "step": 2617
+    },
+    {
+      "epoch": 0.27533621675628067,
+      "grad_norm": 3.0299539149461783,
+      "learning_rate": 4.167111026476714e-06,
+      "loss": 1.0265,
+      "step": 2618
+    },
+    {
+      "epoch": 0.27544138719812794,
+      "grad_norm": 2.3354635781135595,
+      "learning_rate": 4.166488828591587e-06,
+      "loss": 1.0034,
+      "step": 2619
+    },
+    {
+      "epoch": 0.27554655763997526,
+      "grad_norm": 3.2046716724463673,
+      "learning_rate": 4.1658664448800105e-06,
+      "loss": 1.0174,
+      "step": 2620
+    },
+    {
+      "epoch": 0.2756517280818226,
+      "grad_norm": 2.4271263749473597,
+      "learning_rate": 4.165243875411385e-06,
+      "loss": 1.038,
+      "step": 2621
+    },
+    {
+      "epoch": 0.2757568985236699,
+      "grad_norm": 2.946133587548101,
+      "learning_rate": 4.16462112025513e-06,
+      "loss": 1.0256,
+      "step": 2622
+    },
+    {
+      "epoch": 0.27586206896551724,
+      "grad_norm": 2.4337734453214823,
+      "learning_rate": 4.163998179480689e-06,
+      "loss": 1.0307,
+      "step": 2623
+    },
+    {
+      "epoch": 0.27596723940736456,
+      "grad_norm": 2.551909935384673,
+      "learning_rate": 4.163375053157526e-06,
+      "loss": 1.0084,
+      "step": 2624
+    },
+    {
+      "epoch": 0.2760724098492119,
+      "grad_norm": 3.4353272254389964,
+      "learning_rate": 4.162751741355122e-06,
+      "loss": 1.0516,
+      "step": 2625
+    },
+    {
+      "epoch": 0.2761775802910592,
+      "grad_norm": 2.070361498390321,
+      "learning_rate": 4.1621282441429824e-06,
+      "loss": 0.9945,
+      "step": 2626
+    },
+    {
+      "epoch": 0.27628275073290653,
+      "grad_norm": 2.6297532263169803,
+      "learning_rate": 4.161504561590632e-06,
+      "loss": 1.0422,
+      "step": 2627
+    },
+    {
+      "epoch": 0.27638792117475386,
+      "grad_norm": 2.225063604883171,
+      "learning_rate": 4.1608806937676156e-06,
+      "loss": 1.0377,
+      "step": 2628
+    },
+    {
+      "epoch": 0.2764930916166011,
+      "grad_norm": 2.1970819975722082,
+      "learning_rate": 4.160256640743499e-06,
+      "loss": 1.0056,
+      "step": 2629
+    },
+    {
+      "epoch": 0.27659826205844845,
+      "grad_norm": 1.8843326423778457,
+      "learning_rate": 4.159632402587871e-06,
+      "loss": 1.0027,
+      "step": 2630
+    },
+    {
+      "epoch": 0.2767034325002958,
+      "grad_norm": 2.384132121607812,
+      "learning_rate": 4.159007979370337e-06,
+      "loss": 0.9889,
+      "step": 2631
+    },
+    {
+      "epoch": 0.2768086029421431,
+      "grad_norm": 2.3542048671824496,
+      "learning_rate": 4.158383371160526e-06,
+      "loss": 0.9994,
+      "step": 2632
+    },
+    {
+      "epoch": 0.2769137733839904,
+      "grad_norm": 1.9109594616784749,
+      "learning_rate": 4.157758578028086e-06,
+      "loss": 0.9851,
+      "step": 2633
+    },
+    {
+      "epoch": 0.27701894382583775,
+      "grad_norm": 2.719043073466606,
+      "learning_rate": 4.1571336000426865e-06,
+      "loss": 0.9938,
+      "step": 2634
+    },
+    {
+      "epoch": 0.2771241142676851,
+      "grad_norm": 2.3398878575871818,
+      "learning_rate": 4.156508437274017e-06,
+      "loss": 0.9909,
+      "step": 2635
+    },
+    {
+      "epoch": 0.2772292847095324,
+      "grad_norm": 2.6076473869071077,
+      "learning_rate": 4.15588308979179e-06,
+      "loss": 1.0094,
+      "step": 2636
+    },
+    {
+      "epoch": 0.2773344551513797,
+      "grad_norm": 2.9186482297651373,
+      "learning_rate": 4.155257557665734e-06,
+      "loss": 1.0107,
+      "step": 2637
+    },
+    {
+      "epoch": 0.27743962559322705,
+      "grad_norm": 2.924548529106016,
+      "learning_rate": 4.154631840965603e-06,
+      "loss": 1.0273,
+      "step": 2638
+    },
+    {
+      "epoch": 0.2775447960350743,
+      "grad_norm": 2.79169988893619,
+      "learning_rate": 4.154005939761167e-06,
+      "loss": 1.045,
+      "step": 2639
+    },
+    {
+      "epoch": 0.27764996647692164,
+      "grad_norm": 2.5444212624663476,
+      "learning_rate": 4.15337985412222e-06,
+      "loss": 0.9968,
+      "step": 2640
+    },
+    {
+      "epoch": 0.27775513691876896,
+      "grad_norm": 2.770240170620273,
+      "learning_rate": 4.152753584118576e-06,
+      "loss": 1.0392,
+      "step": 2641
+    },
+    {
+      "epoch": 0.2778603073606163,
+      "grad_norm": 3.112319778752494,
+      "learning_rate": 4.152127129820067e-06,
+      "loss": 1.0249,
+      "step": 2642
+    },
+    {
+      "epoch": 0.2779654778024636,
+      "grad_norm": 2.310464927042958,
+      "learning_rate": 4.151500491296551e-06,
+      "loss": 0.9965,
+      "step": 2643
+    },
+    {
+      "epoch": 0.27807064824431094,
+      "grad_norm": 3.2829488475899558,
+      "learning_rate": 4.150873668617899e-06,
+      "loss": 0.9908,
+      "step": 2644
+    },
+    {
+      "epoch": 0.27817581868615826,
+      "grad_norm": 2.176507323056151,
+      "learning_rate": 4.150246661854009e-06,
+      "loss": 0.9991,
+      "step": 2645
+    },
+    {
+      "epoch": 0.2782809891280056,
+      "grad_norm": 1.5655734274913256,
+      "learning_rate": 4.149619471074797e-06,
+      "loss": 0.9418,
+      "step": 2646
+    },
+    {
+      "epoch": 0.2783861595698529,
+      "grad_norm": 2.108505326156569,
+      "learning_rate": 4.148992096350199e-06,
+      "loss": 1.0599,
+      "step": 2647
+    },
+    {
+      "epoch": 0.27849133001170023,
+      "grad_norm": 3.3136337973670202,
+      "learning_rate": 4.1483645377501726e-06,
+      "loss": 1.0149,
+      "step": 2648
+    },
+    {
+      "epoch": 0.27859650045354756,
+      "grad_norm": 2.54494923113061,
+      "learning_rate": 4.147736795344695e-06,
+      "loss": 0.9897,
+      "step": 2649
+    },
+    {
+      "epoch": 0.2787016708953948,
+      "grad_norm": 3.225755330401636,
+      "learning_rate": 4.147108869203765e-06,
+      "loss": 1.0042,
+      "step": 2650
+    },
+    {
+      "epoch": 0.27880684133724215,
+      "grad_norm": 2.7331184633216123,
+      "learning_rate": 4.146480759397401e-06,
+      "loss": 1.0194,
+      "step": 2651
+    },
+    {
+      "epoch": 0.2789120117790895,
+      "grad_norm": 3.0050939408575794,
+      "learning_rate": 4.145852465995642e-06,
+      "loss": 0.9754,
+      "step": 2652
+    },
+    {
+      "epoch": 0.2790171822209368,
+      "grad_norm": 1.8464056831154667,
+      "learning_rate": 4.145223989068547e-06,
+      "loss": 0.9801,
+      "step": 2653
+    },
+    {
+      "epoch": 0.2791223526627841,
+      "grad_norm": 3.288289011526489,
+      "learning_rate": 4.1445953286861976e-06,
+      "loss": 1.0099,
+      "step": 2654
+    },
+    {
+      "epoch": 0.27922752310463145,
+      "grad_norm": 2.9001394289482763,
+      "learning_rate": 4.143966484918692e-06,
+      "loss": 1.0246,
+      "step": 2655
+    },
+    {
+      "epoch": 0.27933269354647877,
+      "grad_norm": 2.7075837385137715,
+      "learning_rate": 4.143337457836154e-06,
+      "loss": 1.0293,
+      "step": 2656
+    },
+    {
+      "epoch": 0.2794378639883261,
+      "grad_norm": 3.4531421732250114,
+      "learning_rate": 4.142708247508723e-06,
+      "loss": 1.015,
+      "step": 2657
+    },
+    {
+      "epoch": 0.2795430344301734,
+      "grad_norm": 2.473925702976424,
+      "learning_rate": 4.142078854006561e-06,
+      "loss": 1.0234,
+      "step": 2658
+    },
+    {
+      "epoch": 0.27964820487202074,
+      "grad_norm": 2.533831100531155,
+      "learning_rate": 4.141449277399852e-06,
+      "loss": 1.04,
+      "step": 2659
+    },
+    {
+      "epoch": 0.279753375313868,
+      "grad_norm": 2.9141485074580458,
+      "learning_rate": 4.140819517758796e-06,
+      "loss": 1.0459,
+      "step": 2660
+    },
+    {
+      "epoch": 0.27985854575571534,
+      "grad_norm": 1.6275648359890877,
+      "learning_rate": 4.1401895751536185e-06,
+      "loss": 1.0112,
+      "step": 2661
+    },
+    {
+      "epoch": 0.27996371619756266,
+      "grad_norm": 2.3074858640345064,
+      "learning_rate": 4.139559449654561e-06,
+      "loss": 1.0349,
+      "step": 2662
+    },
+    {
+      "epoch": 0.28006888663941,
+      "grad_norm": 2.266029848888559,
+      "learning_rate": 4.138929141331888e-06,
+      "loss": 1.0054,
+      "step": 2663
+    },
+    {
+      "epoch": 0.2801740570812573,
+      "grad_norm": 2.30011851501213,
+      "learning_rate": 4.138298650255885e-06,
+      "loss": 1.0183,
+      "step": 2664
+    },
+    {
+      "epoch": 0.28027922752310463,
+      "grad_norm": 2.0840813611244378,
+      "learning_rate": 4.137667976496855e-06,
+      "loss": 1.0502,
+      "step": 2665
+    },
+    {
+      "epoch": 0.28038439796495196,
+      "grad_norm": 2.1068525223894343,
+      "learning_rate": 4.137037120125125e-06,
+      "loss": 0.9381,
+      "step": 2666
+    },
+    {
+      "epoch": 0.2804895684067993,
+      "grad_norm": 2.7828354238745967,
+      "learning_rate": 4.136406081211038e-06,
+      "loss": 1.019,
+      "step": 2667
+    },
+    {
+      "epoch": 0.2805947388486466,
+      "grad_norm": 2.6442032647808733,
+      "learning_rate": 4.135774859824962e-06,
+      "loss": 1.0206,
+      "step": 2668
+    },
+    {
+      "epoch": 0.28069990929049393,
+      "grad_norm": 2.5457929175843677,
+      "learning_rate": 4.13514345603728e-06,
+      "loss": 0.9947,
+      "step": 2669
+    },
+    {
+      "epoch": 0.2808050797323412,
+      "grad_norm": 3.348381030141236,
+      "learning_rate": 4.134511869918402e-06,
+      "loss": 1.0436,
+      "step": 2670
+    },
+    {
+      "epoch": 0.2809102501741885,
+      "grad_norm": 1.8730510261518738,
+      "learning_rate": 4.133880101538753e-06,
+      "loss": 0.9471,
+      "step": 2671
+    },
+    {
+      "epoch": 0.28101542061603585,
+      "grad_norm": 2.39799470254033,
+      "learning_rate": 4.13324815096878e-06,
+      "loss": 1.0186,
+      "step": 2672
+    },
+    {
+      "epoch": 0.2811205910578832,
+      "grad_norm": 2.626761498640316,
+      "learning_rate": 4.132616018278951e-06,
+      "loss": 0.9936,
+      "step": 2673
+    },
+    {
+      "epoch": 0.2812257614997305,
+      "grad_norm": 2.7240523919341504,
+      "learning_rate": 4.1319837035397525e-06,
+      "loss": 0.9984,
+      "step": 2674
+    },
+    {
+      "epoch": 0.2813309319415778,
+      "grad_norm": 2.856921462750168,
+      "learning_rate": 4.131351206821694e-06,
+      "loss": 1.0513,
+      "step": 2675
+    },
+    {
+      "epoch": 0.28143610238342515,
+      "grad_norm": 1.9943496303192936,
+      "learning_rate": 4.130718528195303e-06,
+      "loss": 1.0222,
+      "step": 2676
+    },
+    {
+      "epoch": 0.28154127282527247,
+      "grad_norm": 4.2289456038778,
+      "learning_rate": 4.130085667731129e-06,
+      "loss": 1.0211,
+      "step": 2677
+    },
+    {
+      "epoch": 0.2816464432671198,
+      "grad_norm": 2.9894225702147543,
+      "learning_rate": 4.129452625499739e-06,
+      "loss": 1.0277,
+      "step": 2678
+    },
+    {
+      "epoch": 0.2817516137089671,
+      "grad_norm": 1.992493935234122,
+      "learning_rate": 4.128819401571724e-06,
+      "loss": 1.0171,
+      "step": 2679
+    },
+    {
+      "epoch": 0.2818567841508144,
+      "grad_norm": 2.652762924314855,
+      "learning_rate": 4.128185996017692e-06,
+      "loss": 1.0551,
+      "step": 2680
+    },
+    {
+      "epoch": 0.2819619545926617,
+      "grad_norm": 2.604810015626862,
+      "learning_rate": 4.127552408908274e-06,
+      "loss": 1.0353,
+      "step": 2681
+    },
+    {
+      "epoch": 0.28206712503450904,
+      "grad_norm": 2.7095942486504505,
+      "learning_rate": 4.126918640314118e-06,
+      "loss": 1.0172,
+      "step": 2682
+    },
+    {
+      "epoch": 0.28217229547635636,
+      "grad_norm": 3.2119845572549863,
+      "learning_rate": 4.1262846903058966e-06,
+      "loss": 0.9824,
+      "step": 2683
+    },
+    {
+      "epoch": 0.2822774659182037,
+      "grad_norm": 2.4305337457552003,
+      "learning_rate": 4.1256505589542976e-06,
+      "loss": 1.0247,
+      "step": 2684
+    },
+    {
+      "epoch": 0.282382636360051,
+      "grad_norm": 2.0472831189235277,
+      "learning_rate": 4.125016246330034e-06,
+      "loss": 1.0256,
+      "step": 2685
+    },
+    {
+      "epoch": 0.28248780680189833,
+      "grad_norm": 2.1373038919099687,
+      "learning_rate": 4.124381752503834e-06,
+      "loss": 1.0105,
+      "step": 2686
+    },
+    {
+      "epoch": 0.28259297724374566,
+      "grad_norm": 1.5615996486183839,
+      "learning_rate": 4.123747077546451e-06,
+      "loss": 1.0378,
+      "step": 2687
+    },
+    {
+      "epoch": 0.282698147685593,
+      "grad_norm": 1.8388087927806853,
+      "learning_rate": 4.123112221528654e-06,
+      "loss": 1.003,
+      "step": 2688
+    },
+    {
+      "epoch": 0.2828033181274403,
+      "grad_norm": 3.487408850014367,
+      "learning_rate": 4.122477184521237e-06,
+      "loss": 1.0202,
+      "step": 2689
+    },
+    {
+      "epoch": 0.2829084885692876,
+      "grad_norm": 2.666547606564312,
+      "learning_rate": 4.121841966595009e-06,
+      "loss": 1.042,
+      "step": 2690
+    },
+    {
+      "epoch": 0.2830136590111349,
+      "grad_norm": 2.99015276586862,
+      "learning_rate": 4.121206567820803e-06,
+      "loss": 1.0559,
+      "step": 2691
+    },
+    {
+      "epoch": 0.2831188294529822,
+      "grad_norm": 2.866974473703025,
+      "learning_rate": 4.120570988269472e-06,
+      "loss": 0.9848,
+      "step": 2692
+    },
+    {
+      "epoch": 0.28322399989482955,
+      "grad_norm": 2.50728052143124,
+      "learning_rate": 4.119935228011885e-06,
+      "loss": 1.0286,
+      "step": 2693
+    },
+    {
+      "epoch": 0.28332917033667687,
+      "grad_norm": 2.3872886212034095,
+      "learning_rate": 4.119299287118937e-06,
+      "loss": 1.0258,
+      "step": 2694
+    },
+    {
+      "epoch": 0.2834343407785242,
+      "grad_norm": 1.7175832511041358,
+      "learning_rate": 4.118663165661538e-06,
+      "loss": 1.0324,
+      "step": 2695
+    },
+    {
+      "epoch": 0.2835395112203715,
+      "grad_norm": 2.657674607317461,
+      "learning_rate": 4.118026863710623e-06,
+      "loss": 1.0573,
+      "step": 2696
+    },
+    {
+      "epoch": 0.28364468166221884,
+      "grad_norm": 2.5589768139758755,
+      "learning_rate": 4.117390381337144e-06,
+      "loss": 0.9864,
+      "step": 2697
+    },
+    {
+      "epoch": 0.28374985210406617,
+      "grad_norm": 3.1191553451509093,
+      "learning_rate": 4.116753718612072e-06,
+      "loss": 0.9765,
+      "step": 2698
+    },
+    {
+      "epoch": 0.2838550225459135,
+      "grad_norm": 3.6343579637339842,
+      "learning_rate": 4.116116875606402e-06,
+      "loss": 1.0286,
+      "step": 2699
+    },
+    {
+      "epoch": 0.28396019298776076,
+      "grad_norm": 2.884280278934574,
+      "learning_rate": 4.1154798523911446e-06,
+      "loss": 1.0308,
+      "step": 2700
+    },
+    {
+      "epoch": 0.2840653634296081,
+      "grad_norm": 2.2973711289546603,
+      "learning_rate": 4.114842649037335e-06,
+      "loss": 0.965,
+      "step": 2701
+    },
+    {
+      "epoch": 0.2841705338714554,
+      "grad_norm": 2.870025826884082,
+      "learning_rate": 4.114205265616026e-06,
+      "loss": 0.9826,
+      "step": 2702
+    },
+    {
+      "epoch": 0.28427570431330274,
+      "grad_norm": 2.8685291174085332,
+      "learning_rate": 4.1135677021982885e-06,
+      "loss": 1.0087,
+      "step": 2703
+    },
+    {
+      "epoch": 0.28438087475515006,
+      "grad_norm": 3.000969023465314,
+      "learning_rate": 4.11292995885522e-06,
+      "loss": 1.0142,
+      "step": 2704
+    },
+    {
+      "epoch": 0.2844860451969974,
+      "grad_norm": 2.382052707066313,
+      "learning_rate": 4.11229203565793e-06,
+      "loss": 0.9957,
+      "step": 2705
+    },
+    {
+      "epoch": 0.2845912156388447,
+      "grad_norm": 2.5092385915195723,
+      "learning_rate": 4.111653932677553e-06,
+      "loss": 0.97,
+      "step": 2706
+    },
+    {
+      "epoch": 0.28469638608069203,
+      "grad_norm": 2.83339755995173,
+      "learning_rate": 4.111015649985243e-06,
+      "loss": 0.9873,
+      "step": 2707
+    },
+    {
+      "epoch": 0.28480155652253936,
+      "grad_norm": 2.7074863691662046,
+      "learning_rate": 4.110377187652174e-06,
+      "loss": 0.987,
+      "step": 2708
+    },
+    {
+      "epoch": 0.2849067269643867,
+      "grad_norm": 1.9740060538086266,
+      "learning_rate": 4.109738545749538e-06,
+      "loss": 0.9809,
+      "step": 2709
+    },
+    {
+      "epoch": 0.285011897406234,
+      "grad_norm": 3.3206180620517114,
+      "learning_rate": 4.1090997243485494e-06,
+      "loss": 1.0318,
+      "step": 2710
+    },
+    {
+      "epoch": 0.2851170678480813,
+      "grad_norm": 3.4193647547163435,
+      "learning_rate": 4.108460723520441e-06,
+      "loss": 1.0323,
+      "step": 2711
+    },
+    {
+      "epoch": 0.2852222382899286,
+      "grad_norm": 2.461999394482132,
+      "learning_rate": 4.107821543336468e-06,
+      "loss": 0.969,
+      "step": 2712
+    },
+    {
+      "epoch": 0.2853274087317759,
+      "grad_norm": 2.033813622679507,
+      "learning_rate": 4.107182183867903e-06,
+      "loss": 0.9876,
+      "step": 2713
+    },
+    {
+      "epoch": 0.28543257917362325,
+      "grad_norm": 2.7246823900429673,
+      "learning_rate": 4.106542645186039e-06,
+      "loss": 0.9927,
+      "step": 2714
+    },
+    {
+      "epoch": 0.28553774961547057,
+      "grad_norm": 2.869297378241626,
+      "learning_rate": 4.10590292736219e-06,
+      "loss": 1.0259,
+      "step": 2715
+    },
+    {
+      "epoch": 0.2856429200573179,
+      "grad_norm": 2.521121687829408,
+      "learning_rate": 4.105263030467689e-06,
+      "loss": 1.0261,
+      "step": 2716
+    },
+    {
+      "epoch": 0.2857480904991652,
+      "grad_norm": 2.599567696407617,
+      "learning_rate": 4.10462295457389e-06,
+      "loss": 1.0017,
+      "step": 2717
+    },
+    {
+      "epoch": 0.28585326094101254,
+      "grad_norm": 2.024220267749111,
+      "learning_rate": 4.103982699752167e-06,
+      "loss": 0.9992,
+      "step": 2718
+    },
+    {
+      "epoch": 0.28595843138285987,
+      "grad_norm": 2.6696411790056565,
+      "learning_rate": 4.103342266073913e-06,
+      "loss": 1.0149,
+      "step": 2719
+    },
+    {
+      "epoch": 0.2860636018247072,
+      "grad_norm": 1.8522315223803922,
+      "learning_rate": 4.102701653610541e-06,
+      "loss": 1.0337,
+      "step": 2720
+    },
+    {
+      "epoch": 0.28616877226655446,
+      "grad_norm": 2.7949963553772075,
+      "learning_rate": 4.102060862433484e-06,
+      "loss": 0.9954,
+      "step": 2721
+    },
+    {
+      "epoch": 0.2862739427084018,
+      "grad_norm": 2.229728603318455,
+      "learning_rate": 4.101419892614195e-06,
+      "loss": 1.0195,
+      "step": 2722
+    },
+    {
+      "epoch": 0.2863791131502491,
+      "grad_norm": 3.3540545324036914,
+      "learning_rate": 4.10077874422415e-06,
+      "loss": 0.9692,
+      "step": 2723
+    },
+    {
+      "epoch": 0.28648428359209643,
+      "grad_norm": 2.7750001031832183,
+      "learning_rate": 4.100137417334838e-06,
+      "loss": 1.0112,
+      "step": 2724
+    },
+    {
+      "epoch": 0.28658945403394376,
+      "grad_norm": 2.8776802158198462,
+      "learning_rate": 4.099495912017773e-06,
+      "loss": 1.0133,
+      "step": 2725
+    },
+    {
+      "epoch": 0.2866946244757911,
+      "grad_norm": 2.728837469705898,
+      "learning_rate": 4.09885422834449e-06,
+      "loss": 1.0062,
+      "step": 2726
+    },
+    {
+      "epoch": 0.2867997949176384,
+      "grad_norm": 3.151019351944655,
+      "learning_rate": 4.0982123663865394e-06,
+      "loss": 1.0328,
+      "step": 2727
+    },
+    {
+      "epoch": 0.28690496535948573,
+      "grad_norm": 2.278189959064822,
+      "learning_rate": 4.097570326215495e-06,
+      "loss": 0.9912,
+      "step": 2728
+    },
+    {
+      "epoch": 0.28701013580133306,
+      "grad_norm": 3.0741949886346394,
+      "learning_rate": 4.096928107902949e-06,
+      "loss": 1.0505,
+      "step": 2729
+    },
+    {
+      "epoch": 0.2871153062431804,
+      "grad_norm": 2.2526302976277495,
+      "learning_rate": 4.096285711520513e-06,
+      "loss": 1.0154,
+      "step": 2730
+    },
+    {
+      "epoch": 0.28722047668502765,
+      "grad_norm": 2.4076837142276974,
+      "learning_rate": 4.09564313713982e-06,
+      "loss": 1.0188,
+      "step": 2731
+    },
+    {
+      "epoch": 0.287325647126875,
+      "grad_norm": 1.8879117425120007,
+      "learning_rate": 4.095000384832522e-06,
+      "loss": 1.0139,
+      "step": 2732
+    },
+    {
+      "epoch": 0.2874308175687223,
+      "grad_norm": 3.1660647193201643,
+      "learning_rate": 4.09435745467029e-06,
+      "loss": 0.9862,
+      "step": 2733
+    },
+    {
+      "epoch": 0.2875359880105696,
+      "grad_norm": 2.4093841491214922,
+      "learning_rate": 4.0937143467248176e-06,
+      "loss": 1.038,
+      "step": 2734
+    },
+    {
+      "epoch": 0.28764115845241695,
+      "grad_norm": 2.3022115233534937,
+      "learning_rate": 4.093071061067815e-06,
+      "loss": 1.0139,
+      "step": 2735
+    },
+    {
+      "epoch": 0.28774632889426427,
+      "grad_norm": 2.285699802794061,
+      "learning_rate": 4.092427597771013e-06,
+      "loss": 0.9595,
+      "step": 2736
+    },
+    {
+      "epoch": 0.2878514993361116,
+      "grad_norm": 2.7549603896627004,
+      "learning_rate": 4.091783956906164e-06,
+      "loss": 1.0187,
+      "step": 2737
+    },
+    {
+      "epoch": 0.2879566697779589,
+      "grad_norm": 3.483788671598276,
+      "learning_rate": 4.091140138545037e-06,
+      "loss": 1.0093,
+      "step": 2738
+    },
+    {
+      "epoch": 0.28806184021980624,
+      "grad_norm": 3.2320477916369437,
+      "learning_rate": 4.090496142759425e-06,
+      "loss": 1.0223,
+      "step": 2739
+    },
+    {
+      "epoch": 0.28816701066165357,
+      "grad_norm": 2.4540897960845,
+      "learning_rate": 4.089851969621138e-06,
+      "loss": 1.0203,
+      "step": 2740
+    },
+    {
+      "epoch": 0.28827218110350084,
+      "grad_norm": 3.241726113530304,
+      "learning_rate": 4.089207619202006e-06,
+      "loss": 1.015,
+      "step": 2741
+    },
+    {
+      "epoch": 0.28837735154534816,
+      "grad_norm": 1.8502699365222972,
+      "learning_rate": 4.088563091573879e-06,
+      "loss": 1.0269,
+      "step": 2742
+    },
+    {
+      "epoch": 0.2884825219871955,
+      "grad_norm": 2.5427369122444423,
+      "learning_rate": 4.087918386808627e-06,
+      "loss": 1.0118,
+      "step": 2743
+    },
+    {
+      "epoch": 0.2885876924290428,
+      "grad_norm": 2.301419718364066,
+      "learning_rate": 4.087273504978139e-06,
+      "loss": 1.0222,
+      "step": 2744
+    },
+    {
+      "epoch": 0.28869286287089013,
+      "grad_norm": 2.917540225698338,
+      "learning_rate": 4.086628446154325e-06,
+      "loss": 1.0672,
+      "step": 2745
+    },
+    {
+      "epoch": 0.28879803331273746,
+      "grad_norm": 3.578547174167657,
+      "learning_rate": 4.085983210409114e-06,
+      "loss": 0.9953,
+      "step": 2746
+    },
+    {
+      "epoch": 0.2889032037545848,
+      "grad_norm": 3.3534719860895152,
+      "learning_rate": 4.085337797814455e-06,
+      "loss": 0.9982,
+      "step": 2747
+    },
+    {
+      "epoch": 0.2890083741964321,
+      "grad_norm": 2.676661209030138,
+      "learning_rate": 4.084692208442316e-06,
+      "loss": 1.0266,
+      "step": 2748
+    },
+    {
+      "epoch": 0.28911354463827943,
+      "grad_norm": 3.20816985126381,
+      "learning_rate": 4.084046442364686e-06,
+      "loss": 1.0083,
+      "step": 2749
+    },
+    {
+      "epoch": 0.28921871508012675,
+      "grad_norm": 4.219244894983786,
+      "learning_rate": 4.0834004996535706e-06,
+      "loss": 1.0561,
+      "step": 2750
+    },
+    {
+      "epoch": 0.289323885521974,
+      "grad_norm": 2.1890256985234298,
+      "learning_rate": 4.082754380381001e-06,
+      "loss": 0.9447,
+      "step": 2751
+    },
+    {
+      "epoch": 0.28942905596382135,
+      "grad_norm": 2.8643683405510396,
+      "learning_rate": 4.082108084619021e-06,
+      "loss": 0.9889,
+      "step": 2752
+    },
+    {
+      "epoch": 0.28953422640566867,
+      "grad_norm": 2.51789630360384,
+      "learning_rate": 4.0814616124397015e-06,
+      "loss": 0.991,
+      "step": 2753
+    },
+    {
+      "epoch": 0.289639396847516,
+      "grad_norm": 2.0882264359630436,
+      "learning_rate": 4.080814963915125e-06,
+      "loss": 0.9913,
+      "step": 2754
+    },
+    {
+      "epoch": 0.2897445672893633,
+      "grad_norm": 2.214555512262277,
+      "learning_rate": 4.0801681391174005e-06,
+      "loss": 0.9323,
+      "step": 2755
+    },
+    {
+      "epoch": 0.28984973773121064,
+      "grad_norm": 2.4270616412613424,
+      "learning_rate": 4.079521138118654e-06,
+      "loss": 1.0091,
+      "step": 2756
+    },
+    {
+      "epoch": 0.28995490817305797,
+      "grad_norm": 2.5107109147244344,
+      "learning_rate": 4.07887396099103e-06,
+      "loss": 1.0413,
+      "step": 2757
+    },
+    {
+      "epoch": 0.2900600786149053,
+      "grad_norm": 2.599915906438244,
+      "learning_rate": 4.078226607806694e-06,
+      "loss": 0.9893,
+      "step": 2758
+    },
+    {
+      "epoch": 0.2901652490567526,
+      "grad_norm": 2.2767518525790287,
+      "learning_rate": 4.077579078637831e-06,
+      "loss": 1.0241,
+      "step": 2759
+    },
+    {
+      "epoch": 0.29027041949859994,
+      "grad_norm": 2.794647276010669,
+      "learning_rate": 4.076931373556646e-06,
+      "loss": 1.037,
+      "step": 2760
+    },
+    {
+      "epoch": 0.2903755899404472,
+      "grad_norm": 3.253264621576212,
+      "learning_rate": 4.076283492635362e-06,
+      "loss": 1.0122,
+      "step": 2761
+    },
+    {
+      "epoch": 0.29048076038229453,
+      "grad_norm": 2.499980883018592,
+      "learning_rate": 4.075635435946225e-06,
+      "loss": 1.0027,
+      "step": 2762
+    },
+    {
+      "epoch": 0.29058593082414186,
+      "grad_norm": 2.972899947742262,
+      "learning_rate": 4.074987203561497e-06,
+      "loss": 0.9985,
+      "step": 2763
+    },
+    {
+      "epoch": 0.2906911012659892,
+      "grad_norm": 1.9737479143234022,
+      "learning_rate": 4.074338795553459e-06,
+      "loss": 1.0032,
+      "step": 2764
+    },
+    {
+      "epoch": 0.2907962717078365,
+      "grad_norm": 2.584186588539805,
+      "learning_rate": 4.073690211994417e-06,
+      "loss": 1.0373,
+      "step": 2765
+    },
+    {
+      "epoch": 0.29090144214968383,
+      "grad_norm": 3.1651907150502154,
+      "learning_rate": 4.0730414529566905e-06,
+      "loss": 0.9922,
+      "step": 2766
+    },
+    {
+      "epoch": 0.29100661259153116,
+      "grad_norm": 2.6878955915114253,
+      "learning_rate": 4.072392518512623e-06,
+      "loss": 1.0367,
+      "step": 2767
+    },
+    {
+      "epoch": 0.2911117830333785,
+      "grad_norm": 2.489228399908619,
+      "learning_rate": 4.071743408734574e-06,
+      "loss": 1.0302,
+      "step": 2768
+    },
+    {
+      "epoch": 0.2912169534752258,
+      "grad_norm": 3.060067782653973,
+      "learning_rate": 4.071094123694926e-06,
+      "loss": 0.9873,
+      "step": 2769
+    },
+    {
+      "epoch": 0.29132212391707313,
+      "grad_norm": 2.9513075554961388,
+      "learning_rate": 4.070444663466079e-06,
+      "loss": 1.0038,
+      "step": 2770
+    },
+    {
+      "epoch": 0.29142729435892045,
+      "grad_norm": 2.495540383919349,
+      "learning_rate": 4.069795028120452e-06,
+      "loss": 0.9954,
+      "step": 2771
+    },
+    {
+      "epoch": 0.2915324648007677,
+      "grad_norm": 4.0277271187757,
+      "learning_rate": 4.069145217730484e-06,
+      "loss": 0.9957,
+      "step": 2772
+    },
+    {
+      "epoch": 0.29163763524261505,
+      "grad_norm": 2.511687233295335,
+      "learning_rate": 4.068495232368635e-06,
+      "loss": 0.9685,
+      "step": 2773
+    },
+    {
+      "epoch": 0.29174280568446237,
+      "grad_norm": 2.1591231007772085,
+      "learning_rate": 4.067845072107384e-06,
+      "loss": 1.0597,
+      "step": 2774
+    },
+    {
+      "epoch": 0.2918479761263097,
+      "grad_norm": 2.4755744056169635,
+      "learning_rate": 4.0671947370192264e-06,
+      "loss": 1.0009,
+      "step": 2775
+    },
+    {
+      "epoch": 0.291953146568157,
+      "grad_norm": 1.7500400514158185,
+      "learning_rate": 4.066544227176683e-06,
+      "loss": 1.0042,
+      "step": 2776
+    },
+    {
+      "epoch": 0.29205831701000434,
+      "grad_norm": 2.7751869337926225,
+      "learning_rate": 4.065893542652288e-06,
+      "loss": 0.9955,
+      "step": 2777
+    },
+    {
+      "epoch": 0.29216348745185167,
+      "grad_norm": 2.0427141686414445,
+      "learning_rate": 4.065242683518599e-06,
+      "loss": 0.9956,
+      "step": 2778
+    },
+    {
+      "epoch": 0.292268657893699,
+      "grad_norm": 2.582374430746467,
+      "learning_rate": 4.0645916498481905e-06,
+      "loss": 0.9875,
+      "step": 2779
+    },
+    {
+      "epoch": 0.2923738283355463,
+      "grad_norm": 2.1183619829247573,
+      "learning_rate": 4.06394044171366e-06,
+      "loss": 0.9543,
+      "step": 2780
+    },
+    {
+      "epoch": 0.29247899877739364,
+      "grad_norm": 2.288872193618224,
+      "learning_rate": 4.063289059187621e-06,
+      "loss": 1.0294,
+      "step": 2781
+    },
+    {
+      "epoch": 0.2925841692192409,
+      "grad_norm": 2.9861250226649356,
+      "learning_rate": 4.062637502342708e-06,
+      "loss": 1.0364,
+      "step": 2782
+    },
+    {
+      "epoch": 0.29268933966108823,
+      "grad_norm": 2.263444569755448,
+      "learning_rate": 4.061985771251573e-06,
+      "loss": 1.0157,
+      "step": 2783
+    },
+    {
+      "epoch": 0.29279451010293556,
+      "grad_norm": 3.25482041267685,
+      "learning_rate": 4.061333865986892e-06,
+      "loss": 1.0169,
+      "step": 2784
+    },
+    {
+      "epoch": 0.2928996805447829,
+      "grad_norm": 2.993652291933884,
+      "learning_rate": 4.060681786621357e-06,
+      "loss": 1.0432,
+      "step": 2785
+    },
+    {
+      "epoch": 0.2930048509866302,
+      "grad_norm": 2.5325196207248517,
+      "learning_rate": 4.060029533227678e-06,
+      "loss": 0.9786,
+      "step": 2786
+    },
+    {
+      "epoch": 0.29311002142847753,
+      "grad_norm": 3.072755160863761,
+      "learning_rate": 4.059377105878586e-06,
+      "loss": 1.0081,
+      "step": 2787
+    },
+    {
+      "epoch": 0.29321519187032485,
+      "grad_norm": 2.9059852060216436,
+      "learning_rate": 4.058724504646834e-06,
+      "loss": 0.999,
+      "step": 2788
+    },
+    {
+      "epoch": 0.2933203623121722,
+      "grad_norm": 2.1664281660182154,
+      "learning_rate": 4.058071729605191e-06,
+      "loss": 1.0101,
+      "step": 2789
+    },
+    {
+      "epoch": 0.2934255327540195,
+      "grad_norm": 2.6755938701735014,
+      "learning_rate": 4.057418780826448e-06,
+      "loss": 0.9915,
+      "step": 2790
+    },
+    {
+      "epoch": 0.2935307031958668,
+      "grad_norm": 1.7269644244858031,
+      "learning_rate": 4.05676565838341e-06,
+      "loss": 1.0071,
+      "step": 2791
+    },
+    {
+      "epoch": 0.2936358736377141,
+      "grad_norm": 2.74362018348315,
+      "learning_rate": 4.0561123623489096e-06,
+      "loss": 1.0314,
+      "step": 2792
+    },
+    {
+      "epoch": 0.2937410440795614,
+      "grad_norm": 2.340557452024337,
+      "learning_rate": 4.0554588927957925e-06,
+      "loss": 0.9678,
+      "step": 2793
+    },
+    {
+      "epoch": 0.29384621452140874,
+      "grad_norm": 2.066040916919864,
+      "learning_rate": 4.054805249796925e-06,
+      "loss": 1.0263,
+      "step": 2794
+    },
+    {
+      "epoch": 0.29395138496325607,
+      "grad_norm": 2.8805200149811956,
+      "learning_rate": 4.054151433425194e-06,
+      "loss": 1.0603,
+      "step": 2795
+    },
+    {
+      "epoch": 0.2940565554051034,
+      "grad_norm": 2.434123603671788,
+      "learning_rate": 4.053497443753505e-06,
+      "loss": 1.0042,
+      "step": 2796
+    },
+    {
+      "epoch": 0.2941617258469507,
+      "grad_norm": 3.618193905329278,
+      "learning_rate": 4.052843280854783e-06,
+      "loss": 0.9702,
+      "step": 2797
+    },
+    {
+      "epoch": 0.29426689628879804,
+      "grad_norm": 3.1682364893441193,
+      "learning_rate": 4.052188944801972e-06,
+      "loss": 1.0143,
+      "step": 2798
+    },
+    {
+      "epoch": 0.29437206673064537,
+      "grad_norm": 2.3744837915482484,
+      "learning_rate": 4.051534435668038e-06,
+      "loss": 0.9982,
+      "step": 2799
+    },
+    {
+      "epoch": 0.2944772371724927,
+      "grad_norm": 2.1386699249557872,
+      "learning_rate": 4.050879753525959e-06,
+      "loss": 1.0242,
+      "step": 2800
+    },
+    {
+      "epoch": 0.29458240761434,
+      "grad_norm": 2.6446899519304314,
+      "learning_rate": 4.050224898448741e-06,
+      "loss": 0.9404,
+      "step": 2801
+    },
+    {
+      "epoch": 0.2946875780561873,
+      "grad_norm": 2.381644444210151,
+      "learning_rate": 4.049569870509404e-06,
+      "loss": 1.0128,
+      "step": 2802
+    },
+    {
+      "epoch": 0.2947927484980346,
+      "grad_norm": 2.0924851058105287,
+      "learning_rate": 4.048914669780989e-06,
+      "loss": 1.0263,
+      "step": 2803
+    },
+    {
+      "epoch": 0.29489791893988193,
+      "grad_norm": 2.8128059213176666,
+      "learning_rate": 4.048259296336556e-06,
+      "loss": 1.0109,
+      "step": 2804
+    },
+    {
+      "epoch": 0.29500308938172926,
+      "grad_norm": 2.230026087010622,
+      "learning_rate": 4.047603750249184e-06,
+      "loss": 0.9928,
+      "step": 2805
+    },
+    {
+      "epoch": 0.2951082598235766,
+      "grad_norm": 3.651123695031622,
+      "learning_rate": 4.0469480315919714e-06,
+      "loss": 1.0337,
+      "step": 2806
+    },
+    {
+      "epoch": 0.2952134302654239,
+      "grad_norm": 3.794948500887568,
+      "learning_rate": 4.0462921404380376e-06,
+      "loss": 1.0177,
+      "step": 2807
+    },
+    {
+      "epoch": 0.29531860070727123,
+      "grad_norm": 2.717172683131933,
+      "learning_rate": 4.045636076860517e-06,
+      "loss": 0.9645,
+      "step": 2808
+    },
+    {
+      "epoch": 0.29542377114911855,
+      "grad_norm": 2.8432922404185557,
+      "learning_rate": 4.044979840932567e-06,
+      "loss": 1.0147,
+      "step": 2809
+    },
+    {
+      "epoch": 0.2955289415909659,
+      "grad_norm": 2.466690276825208,
+      "learning_rate": 4.044323432727363e-06,
+      "loss": 1.04,
+      "step": 2810
+    },
+    {
+      "epoch": 0.2956341120328132,
+      "grad_norm": 1.898880562225239,
+      "learning_rate": 4.0436668523180985e-06,
+      "loss": 0.9796,
+      "step": 2811
+    },
+    {
+      "epoch": 0.29573928247466047,
+      "grad_norm": 2.2401687192743265,
+      "learning_rate": 4.04301009977799e-06,
+      "loss": 1.0075,
+      "step": 2812
+    },
+    {
+      "epoch": 0.2958444529165078,
+      "grad_norm": 2.779156433583653,
+      "learning_rate": 4.042353175180268e-06,
+      "loss": 1.0163,
+      "step": 2813
+    },
+    {
+      "epoch": 0.2959496233583551,
+      "grad_norm": 2.6233755857676617,
+      "learning_rate": 4.041696078598185e-06,
+      "loss": 1.0308,
+      "step": 2814
+    },
+    {
+      "epoch": 0.29605479380020244,
+      "grad_norm": 2.89215727215323,
+      "learning_rate": 4.041038810105014e-06,
+      "loss": 1.039,
+      "step": 2815
+    },
+    {
+      "epoch": 0.29615996424204977,
+      "grad_norm": 2.2937244141847275,
+      "learning_rate": 4.040381369774045e-06,
+      "loss": 0.9985,
+      "step": 2816
+    },
+    {
+      "epoch": 0.2962651346838971,
+      "grad_norm": 2.2753449571613507,
+      "learning_rate": 4.039723757678585e-06,
+      "loss": 0.9842,
+      "step": 2817
+    },
+    {
+      "epoch": 0.2963703051257444,
+      "grad_norm": 2.825575794098944,
+      "learning_rate": 4.0390659738919665e-06,
+      "loss": 1.0148,
+      "step": 2818
+    },
+    {
+      "epoch": 0.29647547556759174,
+      "grad_norm": 2.1237158283334088,
+      "learning_rate": 4.0384080184875355e-06,
+      "loss": 0.9896,
+      "step": 2819
+    },
+    {
+      "epoch": 0.29658064600943906,
+      "grad_norm": 2.8266979036329123,
+      "learning_rate": 4.037749891538661e-06,
+      "loss": 1.0097,
+      "step": 2820
+    },
+    {
+      "epoch": 0.2966858164512864,
+      "grad_norm": 2.2004271894165557,
+      "learning_rate": 4.037091593118726e-06,
+      "loss": 1.0418,
+      "step": 2821
+    },
+    {
+      "epoch": 0.2967909868931337,
+      "grad_norm": 2.0681432005760914,
+      "learning_rate": 4.036433123301139e-06,
+      "loss": 0.9897,
+      "step": 2822
+    },
+    {
+      "epoch": 0.296896157334981,
+      "grad_norm": 2.8211256820827404,
+      "learning_rate": 4.035774482159323e-06,
+      "loss": 1.0378,
+      "step": 2823
+    },
+    {
+      "epoch": 0.2970013277768283,
+      "grad_norm": 2.6606426976386675,
+      "learning_rate": 4.035115669766721e-06,
+      "loss": 1.0247,
+      "step": 2824
+    },
+    {
+      "epoch": 0.29710649821867563,
+      "grad_norm": 2.5354918405939366,
+      "learning_rate": 4.034456686196798e-06,
+      "loss": 0.9795,
+      "step": 2825
+    },
+    {
+      "epoch": 0.29721166866052295,
+      "grad_norm": 3.4840884239739487,
+      "learning_rate": 4.033797531523034e-06,
+      "loss": 0.9938,
+      "step": 2826
+    },
+    {
+      "epoch": 0.2973168391023703,
+      "grad_norm": 2.9035094968158135,
+      "learning_rate": 4.033138205818931e-06,
+      "loss": 0.9971,
+      "step": 2827
+    },
+    {
+      "epoch": 0.2974220095442176,
+      "grad_norm": 2.388147959432155,
+      "learning_rate": 4.032478709158007e-06,
+      "loss": 0.9786,
+      "step": 2828
+    },
+    {
+      "epoch": 0.2975271799860649,
+      "grad_norm": 2.5043218419305657,
+      "learning_rate": 4.0318190416138024e-06,
+      "loss": 1.013,
+      "step": 2829
+    },
+    {
+      "epoch": 0.29763235042791225,
+      "grad_norm": 2.630395705298903,
+      "learning_rate": 4.031159203259876e-06,
+      "loss": 0.9978,
+      "step": 2830
+    },
+    {
+      "epoch": 0.2977375208697596,
+      "grad_norm": 3.048566956011636,
+      "learning_rate": 4.030499194169803e-06,
+      "loss": 1.022,
+      "step": 2831
+    },
+    {
+      "epoch": 0.2978426913116069,
+      "grad_norm": 2.5379425325683656,
+      "learning_rate": 4.029839014417181e-06,
+      "loss": 1.0029,
+      "step": 2832
+    },
+    {
+      "epoch": 0.29794786175345417,
+      "grad_norm": 2.6960366196246857,
+      "learning_rate": 4.029178664075625e-06,
+      "loss": 1.0183,
+      "step": 2833
+    },
+    {
+      "epoch": 0.2980530321953015,
+      "grad_norm": 3.224224287628243,
+      "learning_rate": 4.028518143218768e-06,
+      "loss": 1.073,
+      "step": 2834
+    },
+    {
+      "epoch": 0.2981582026371488,
+      "grad_norm": 2.429803063051489,
+      "learning_rate": 4.027857451920264e-06,
+      "loss": 0.9988,
+      "step": 2835
+    },
+    {
+      "epoch": 0.29826337307899614,
+      "grad_norm": 3.126241311853215,
+      "learning_rate": 4.027196590253786e-06,
+      "loss": 1.0351,
+      "step": 2836
+    },
+    {
+      "epoch": 0.29836854352084347,
+      "grad_norm": 2.3886966033688557,
+      "learning_rate": 4.026535558293024e-06,
+      "loss": 1.0146,
+      "step": 2837
+    },
+    {
+      "epoch": 0.2984737139626908,
+      "grad_norm": 2.1691315598429544,
+      "learning_rate": 4.025874356111689e-06,
+      "loss": 0.9821,
+      "step": 2838
+    },
+    {
+      "epoch": 0.2985788844045381,
+      "grad_norm": 3.2536165239458774,
+      "learning_rate": 4.025212983783511e-06,
+      "loss": 1.0239,
+      "step": 2839
+    },
+    {
+      "epoch": 0.29868405484638544,
+      "grad_norm": 2.361345102232033,
+      "learning_rate": 4.024551441382235e-06,
+      "loss": 1.0125,
+      "step": 2840
+    },
+    {
+      "epoch": 0.29878922528823276,
+      "grad_norm": 2.797329995864098,
+      "learning_rate": 4.023889728981631e-06,
+      "loss": 1.0278,
+      "step": 2841
+    },
+    {
+      "epoch": 0.2988943957300801,
+      "grad_norm": 2.053294094600003,
+      "learning_rate": 4.023227846655484e-06,
+      "loss": 0.9636,
+      "step": 2842
+    },
+    {
+      "epoch": 0.29899956617192736,
+      "grad_norm": 2.0068541001608,
+      "learning_rate": 4.0225657944776e-06,
+      "loss": 0.9924,
+      "step": 2843
+    },
+    {
+      "epoch": 0.2991047366137747,
+      "grad_norm": 3.0489588147562947,
+      "learning_rate": 4.021903572521802e-06,
+      "loss": 1.059,
+      "step": 2844
+    },
+    {
+      "epoch": 0.299209907055622,
+      "grad_norm": 2.416659465285543,
+      "learning_rate": 4.021241180861933e-06,
+      "loss": 1.0491,
+      "step": 2845
+    },
+    {
+      "epoch": 0.29931507749746933,
+      "grad_norm": 1.8189123954689044,
+      "learning_rate": 4.0205786195718545e-06,
+      "loss": 0.997,
+      "step": 2846
+    },
+    {
+      "epoch": 0.29942024793931665,
+      "grad_norm": 2.2393968626632126,
+      "learning_rate": 4.0199158887254484e-06,
+      "loss": 1.0253,
+      "step": 2847
+    },
+    {
+      "epoch": 0.299525418381164,
+      "grad_norm": 2.7113732338095438,
+      "learning_rate": 4.019252988396613e-06,
+      "loss": 0.9651,
+      "step": 2848
+    },
+    {
+      "epoch": 0.2996305888230113,
+      "grad_norm": 3.219458618141807,
+      "learning_rate": 4.018589918659267e-06,
+      "loss": 1.047,
+      "step": 2849
+    },
+    {
+      "epoch": 0.2997357592648586,
+      "grad_norm": 2.485215246867124,
+      "learning_rate": 4.01792667958735e-06,
+      "loss": 0.995,
+      "step": 2850
+    },
+    {
+      "epoch": 0.29984092970670595,
+      "grad_norm": 2.5877817382570276,
+      "learning_rate": 4.0172632712548145e-06,
+      "loss": 0.9823,
+      "step": 2851
+    },
+    {
+      "epoch": 0.2999461001485533,
+      "grad_norm": 2.8920656586696043,
+      "learning_rate": 4.016599693735639e-06,
+      "loss": 1.0172,
+      "step": 2852
+    },
+    {
+      "epoch": 0.30005127059040054,
+      "grad_norm": 2.1377145970196025,
+      "learning_rate": 4.015935947103816e-06,
+      "loss": 0.9841,
+      "step": 2853
+    },
+    {
+      "epoch": 0.30015644103224787,
+      "grad_norm": 2.357987310739265,
+      "learning_rate": 4.015272031433358e-06,
+      "loss": 0.9969,
+      "step": 2854
+    },
+    {
+      "epoch": 0.3002616114740952,
+      "grad_norm": 2.6065064666162927,
+      "learning_rate": 4.014607946798298e-06,
+      "loss": 0.9742,
+      "step": 2855
+    },
+    {
+      "epoch": 0.3003667819159425,
+      "grad_norm": 2.5201558670387443,
+      "learning_rate": 4.013943693272686e-06,
+      "loss": 0.9801,
+      "step": 2856
+    },
+    {
+      "epoch": 0.30047195235778984,
+      "grad_norm": 3.0654655143598166,
+      "learning_rate": 4.013279270930592e-06,
+      "loss": 1.0204,
+      "step": 2857
+    },
+    {
+      "epoch": 0.30057712279963716,
+      "grad_norm": 4.142146640718669,
+      "learning_rate": 4.012614679846103e-06,
+      "loss": 1.0169,
+      "step": 2858
+    },
+    {
+      "epoch": 0.3006822932414845,
+      "grad_norm": 1.7013966226721964,
+      "learning_rate": 4.011949920093327e-06,
+      "loss": 1.0133,
+      "step": 2859
+    },
+    {
+      "epoch": 0.3007874636833318,
+      "grad_norm": 2.3493733773934036,
+      "learning_rate": 4.0112849917463905e-06,
+      "loss": 1.0396,
+      "step": 2860
+    },
+    {
+      "epoch": 0.30089263412517914,
+      "grad_norm": 2.9932760739027167,
+      "learning_rate": 4.010619894879436e-06,
+      "loss": 1.0482,
+      "step": 2861
+    },
+    {
+      "epoch": 0.30099780456702646,
+      "grad_norm": 3.4913376390190747,
+      "learning_rate": 4.009954629566629e-06,
+      "loss": 1.0478,
+      "step": 2862
+    },
+    {
+      "epoch": 0.30110297500887373,
+      "grad_norm": 2.4522472564185223,
+      "learning_rate": 4.0092891958821515e-06,
+      "loss": 1.0164,
+      "step": 2863
+    },
+    {
+      "epoch": 0.30120814545072105,
+      "grad_norm": 2.7224329510455996,
+      "learning_rate": 4.0086235939002024e-06,
+      "loss": 1.0279,
+      "step": 2864
+    },
+    {
+      "epoch": 0.3013133158925684,
+      "grad_norm": 2.7001843459526396,
+      "learning_rate": 4.007957823695005e-06,
+      "loss": 1.0319,
+      "step": 2865
+    },
+    {
+      "epoch": 0.3014184863344157,
+      "grad_norm": 2.9328100471570813,
+      "learning_rate": 4.007291885340796e-06,
+      "loss": 0.9669,
+      "step": 2866
+    },
+    {
+      "epoch": 0.301523656776263,
+      "grad_norm": 2.8953214503968705,
+      "learning_rate": 4.006625778911831e-06,
+      "loss": 0.9993,
+      "step": 2867
+    },
+    {
+      "epoch": 0.30162882721811035,
+      "grad_norm": 3.0833861909775333,
+      "learning_rate": 4.005959504482389e-06,
+      "loss": 0.9904,
+      "step": 2868
+    },
+    {
+      "epoch": 0.3017339976599577,
+      "grad_norm": 2.706249563006177,
+      "learning_rate": 4.005293062126764e-06,
+      "loss": 1.0137,
+      "step": 2869
+    },
+    {
+      "epoch": 0.301839168101805,
+      "grad_norm": 2.6523322513613996,
+      "learning_rate": 4.004626451919268e-06,
+      "loss": 1.022,
+      "step": 2870
+    },
+    {
+      "epoch": 0.3019443385436523,
+      "grad_norm": 2.363970677643061,
+      "learning_rate": 4.003959673934235e-06,
+      "loss": 0.9987,
+      "step": 2871
+    },
+    {
+      "epoch": 0.30204950898549965,
+      "grad_norm": 2.6706791504734824,
+      "learning_rate": 4.003292728246015e-06,
+      "loss": 1.0179,
+      "step": 2872
+    },
+    {
+      "epoch": 0.3021546794273469,
+      "grad_norm": 3.0479510186456977,
+      "learning_rate": 4.002625614928978e-06,
+      "loss": 0.956,
+      "step": 2873
+    },
+    {
+      "epoch": 0.30225984986919424,
+      "grad_norm": 1.8651319495727035,
+      "learning_rate": 4.001958334057512e-06,
+      "loss": 1.0283,
+      "step": 2874
+    },
+    {
+      "epoch": 0.30236502031104157,
+      "grad_norm": 2.065753591700377,
+      "learning_rate": 4.001290885706023e-06,
+      "loss": 0.9982,
+      "step": 2875
+    },
+    {
+      "epoch": 0.3024701907528889,
+      "grad_norm": 3.420963266297305,
+      "learning_rate": 4.0006232699489385e-06,
+      "loss": 0.9814,
+      "step": 2876
+    },
+    {
+      "epoch": 0.3025753611947362,
+      "grad_norm": 2.598806589411273,
+      "learning_rate": 3.9999554868607036e-06,
+      "loss": 1.0091,
+      "step": 2877
+    },
+    {
+      "epoch": 0.30268053163658354,
+      "grad_norm": 2.691137805944449,
+      "learning_rate": 3.999287536515778e-06,
+      "loss": 1.0782,
+      "step": 2878
+    },
+    {
+      "epoch": 0.30278570207843086,
+      "grad_norm": 2.5676564139134976,
+      "learning_rate": 3.998619418988646e-06,
+      "loss": 1.0563,
+      "step": 2879
+    },
+    {
+      "epoch": 0.3028908725202782,
+      "grad_norm": 2.5833344941215235,
+      "learning_rate": 3.997951134353808e-06,
+      "loss": 1.0403,
+      "step": 2880
+    },
+    {
+      "epoch": 0.3029960429621255,
+      "grad_norm": 2.1282219410524914,
+      "learning_rate": 3.99728268268578e-06,
+      "loss": 0.9861,
+      "step": 2881
+    },
+    {
+      "epoch": 0.30310121340397284,
+      "grad_norm": 3.4275643393055333,
+      "learning_rate": 3.996614064059104e-06,
+      "loss": 1.0173,
+      "step": 2882
+    },
+    {
+      "epoch": 0.30320638384582016,
+      "grad_norm": 2.2265143102649647,
+      "learning_rate": 3.995945278548331e-06,
+      "loss": 1.0226,
+      "step": 2883
+    },
+    {
+      "epoch": 0.30331155428766743,
+      "grad_norm": 2.8938023915675264,
+      "learning_rate": 3.99527632622804e-06,
+      "loss": 1.0185,
+      "step": 2884
+    },
+    {
+      "epoch": 0.30341672472951475,
+      "grad_norm": 2.854330050298019,
+      "learning_rate": 3.994607207172823e-06,
+      "loss": 1.0003,
+      "step": 2885
+    },
+    {
+      "epoch": 0.3035218951713621,
+      "grad_norm": 2.878769782756261,
+      "learning_rate": 3.993937921457292e-06,
+      "loss": 0.9944,
+      "step": 2886
+    },
+    {
+      "epoch": 0.3036270656132094,
+      "grad_norm": 2.8721086441370294,
+      "learning_rate": 3.993268469156077e-06,
+      "loss": 0.9825,
+      "step": 2887
+    },
+    {
+      "epoch": 0.3037322360550567,
+      "grad_norm": 2.2081747573437074,
+      "learning_rate": 3.992598850343827e-06,
+      "loss": 0.9873,
+      "step": 2888
+    },
+    {
+      "epoch": 0.30383740649690405,
+      "grad_norm": 2.9007140299444822,
+      "learning_rate": 3.991929065095211e-06,
+      "loss": 1.0232,
+      "step": 2889
+    },
+    {
+      "epoch": 0.3039425769387514,
+      "grad_norm": 1.3415032258707018,
+      "learning_rate": 3.991259113484916e-06,
+      "loss": 0.9871,
+      "step": 2890
+    },
+    {
+      "epoch": 0.3040477473805987,
+      "grad_norm": 2.7105364628823456,
+      "learning_rate": 3.990588995587643e-06,
+      "loss": 1.0259,
+      "step": 2891
+    },
+    {
+      "epoch": 0.304152917822446,
+      "grad_norm": 2.31954600362432,
+      "learning_rate": 3.989918711478118e-06,
+      "loss": 1.042,
+      "step": 2892
+    },
+    {
+      "epoch": 0.30425808826429335,
+      "grad_norm": 2.7954636771354346,
+      "learning_rate": 3.989248261231084e-06,
+      "loss": 1.0302,
+      "step": 2893
+    },
+    {
+      "epoch": 0.3043632587061406,
+      "grad_norm": 2.7537708698276813,
+      "learning_rate": 3.988577644921299e-06,
+      "loss": 1.0587,
+      "step": 2894
+    },
+    {
+      "epoch": 0.30446842914798794,
+      "grad_norm": 3.0962306715267935,
+      "learning_rate": 3.9879068626235425e-06,
+      "loss": 0.9759,
+      "step": 2895
+    },
+    {
+      "epoch": 0.30457359958983526,
+      "grad_norm": 2.3957225896167746,
+      "learning_rate": 3.987235914412614e-06,
+      "loss": 1.0023,
+      "step": 2896
+    },
+    {
+      "epoch": 0.3046787700316826,
+      "grad_norm": 2.771932833415502,
+      "learning_rate": 3.986564800363326e-06,
+      "loss": 1.0512,
+      "step": 2897
+    },
+    {
+      "epoch": 0.3047839404735299,
+      "grad_norm": 2.3606896051469537,
+      "learning_rate": 3.985893520550516e-06,
+      "loss": 1.0107,
+      "step": 2898
+    },
+    {
+      "epoch": 0.30488911091537724,
+      "grad_norm": 2.4174906103885943,
+      "learning_rate": 3.985222075049035e-06,
+      "loss": 0.9982,
+      "step": 2899
+    },
+    {
+      "epoch": 0.30499428135722456,
+      "grad_norm": 1.6550177073693109,
+      "learning_rate": 3.984550463933754e-06,
+      "loss": 0.9569,
+      "step": 2900
+    },
+    {
+      "epoch": 0.3050994517990719,
+      "grad_norm": 3.4294334384914213,
+      "learning_rate": 3.983878687279565e-06,
+      "loss": 1.0138,
+      "step": 2901
+    },
+    {
+      "epoch": 0.3052046222409192,
+      "grad_norm": 2.56368532622028,
+      "learning_rate": 3.9832067451613755e-06,
+      "loss": 1.0517,
+      "step": 2902
+    },
+    {
+      "epoch": 0.30530979268276653,
+      "grad_norm": 2.8083843993980953,
+      "learning_rate": 3.982534637654112e-06,
+      "loss": 0.9687,
+      "step": 2903
+    },
+    {
+      "epoch": 0.3054149631246138,
+      "grad_norm": 2.7297876966125156,
+      "learning_rate": 3.981862364832718e-06,
+      "loss": 1.014,
+      "step": 2904
+    },
+    {
+      "epoch": 0.3055201335664611,
+      "grad_norm": 2.103857181645903,
+      "learning_rate": 3.981189926772161e-06,
+      "loss": 0.9635,
+      "step": 2905
+    },
+    {
+      "epoch": 0.30562530400830845,
+      "grad_norm": 2.5474502602534175,
+      "learning_rate": 3.980517323547419e-06,
+      "loss": 1.0005,
+      "step": 2906
+    },
+    {
+      "epoch": 0.3057304744501558,
+      "grad_norm": 2.4631842991491637,
+      "learning_rate": 3.979844555233496e-06,
+      "loss": 1.0259,
+      "step": 2907
+    },
+    {
+      "epoch": 0.3058356448920031,
+      "grad_norm": 3.829976049615588,
+      "learning_rate": 3.979171621905409e-06,
+      "loss": 1.0053,
+      "step": 2908
+    },
+    {
+      "epoch": 0.3059408153338504,
+      "grad_norm": 3.0337382004641404,
+      "learning_rate": 3.978498523638194e-06,
+      "loss": 1.0447,
+      "step": 2909
+    },
+    {
+      "epoch": 0.30604598577569775,
+      "grad_norm": 2.4115719177112105,
+      "learning_rate": 3.97782526050691e-06,
+      "loss": 0.9999,
+      "step": 2910
+    },
+    {
+      "epoch": 0.3061511562175451,
+      "grad_norm": 2.5527608779229696,
+      "learning_rate": 3.977151832586628e-06,
+      "loss": 1.0281,
+      "step": 2911
+    },
+    {
+      "epoch": 0.3062563266593924,
+      "grad_norm": 2.0029391608583262,
+      "learning_rate": 3.976478239952444e-06,
+      "loss": 1.0366,
+      "step": 2912
+    },
+    {
+      "epoch": 0.3063614971012397,
+      "grad_norm": 2.447326314753849,
+      "learning_rate": 3.975804482679464e-06,
+      "loss": 1.0104,
+      "step": 2913
+    },
+    {
+      "epoch": 0.306466667543087,
+      "grad_norm": 2.2765266252277203,
+      "learning_rate": 3.975130560842821e-06,
+      "loss": 0.979,
+      "step": 2914
+    },
+    {
+      "epoch": 0.3065718379849343,
+      "grad_norm": 2.615082713312339,
+      "learning_rate": 3.974456474517661e-06,
+      "loss": 1.0355,
+      "step": 2915
+    },
+    {
+      "epoch": 0.30667700842678164,
+      "grad_norm": 2.380276345939876,
+      "learning_rate": 3.97378222377915e-06,
+      "loss": 1.0034,
+      "step": 2916
+    },
+    {
+      "epoch": 0.30678217886862896,
+      "grad_norm": 2.3073048459869168,
+      "learning_rate": 3.973107808702472e-06,
+      "loss": 1.0268,
+      "step": 2917
+    },
+    {
+      "epoch": 0.3068873493104763,
+      "grad_norm": 1.8941823202478179,
+      "learning_rate": 3.9724332293628295e-06,
+      "loss": 1.0306,
+      "step": 2918
+    },
+    {
+      "epoch": 0.3069925197523236,
+      "grad_norm": 2.896280116503032,
+      "learning_rate": 3.9717584858354454e-06,
+      "loss": 1.008,
+      "step": 2919
+    },
+    {
+      "epoch": 0.30709769019417094,
+      "grad_norm": 2.7180851301598103,
+      "learning_rate": 3.971083578195556e-06,
+      "loss": 1.0215,
+      "step": 2920
+    },
+    {
+      "epoch": 0.30720286063601826,
+      "grad_norm": 2.0401389712432487,
+      "learning_rate": 3.970408506518419e-06,
+      "loss": 0.9842,
+      "step": 2921
+    },
+    {
+      "epoch": 0.3073080310778656,
+      "grad_norm": 2.3667558391152426,
+      "learning_rate": 3.969733270879313e-06,
+      "loss": 0.9797,
+      "step": 2922
+    },
+    {
+      "epoch": 0.3074132015197129,
+      "grad_norm": 3.307599231062686,
+      "learning_rate": 3.969057871353529e-06,
+      "loss": 0.9895,
+      "step": 2923
+    },
+    {
+      "epoch": 0.3075183719615602,
+      "grad_norm": 2.8646693785344555,
+      "learning_rate": 3.96838230801638e-06,
+      "loss": 0.9618,
+      "step": 2924
+    },
+    {
+      "epoch": 0.3076235424034075,
+      "grad_norm": 2.2316449558106974,
+      "learning_rate": 3.967706580943197e-06,
+      "loss": 1.0306,
+      "step": 2925
+    },
+    {
+      "epoch": 0.3077287128452548,
+      "grad_norm": 2.771790047908696,
+      "learning_rate": 3.9670306902093286e-06,
+      "loss": 1.0021,
+      "step": 2926
+    },
+    {
+      "epoch": 0.30783388328710215,
+      "grad_norm": 2.7193389805440207,
+      "learning_rate": 3.9663546358901415e-06,
+      "loss": 0.9975,
+      "step": 2927
+    },
+    {
+      "epoch": 0.3079390537289495,
+      "grad_norm": 2.174531390393959,
+      "learning_rate": 3.965678418061023e-06,
+      "loss": 1.0174,
+      "step": 2928
+    },
+    {
+      "epoch": 0.3080442241707968,
+      "grad_norm": 2.473148623762536,
+      "learning_rate": 3.965002036797374e-06,
+      "loss": 0.9751,
+      "step": 2929
+    },
+    {
+      "epoch": 0.3081493946126441,
+      "grad_norm": 2.523481125448685,
+      "learning_rate": 3.9643254921746176e-06,
+      "loss": 1.0486,
+      "step": 2930
+    },
+    {
+      "epoch": 0.30825456505449145,
+      "grad_norm": 2.4231464073038103,
+      "learning_rate": 3.963648784268193e-06,
+      "loss": 1.0308,
+      "step": 2931
+    },
+    {
+      "epoch": 0.30835973549633877,
+      "grad_norm": 2.908693220754445,
+      "learning_rate": 3.9629719131535595e-06,
+      "loss": 1.019,
+      "step": 2932
+    },
+    {
+      "epoch": 0.3084649059381861,
+      "grad_norm": 2.831569138155086,
+      "learning_rate": 3.9622948789061935e-06,
+      "loss": 1.0282,
+      "step": 2933
+    },
+    {
+      "epoch": 0.30857007638003336,
+      "grad_norm": 2.1822044513817533,
+      "learning_rate": 3.961617681601588e-06,
+      "loss": 1.0471,
+      "step": 2934
+    },
+    {
+      "epoch": 0.3086752468218807,
+      "grad_norm": 1.968984775871048,
+      "learning_rate": 3.960940321315257e-06,
+      "loss": 0.9938,
+      "step": 2935
+    },
+    {
+      "epoch": 0.308780417263728,
+      "grad_norm": 3.0742676631782015,
+      "learning_rate": 3.960262798122731e-06,
+      "loss": 1.0384,
+      "step": 2936
+    },
+    {
+      "epoch": 0.30888558770557534,
+      "grad_norm": 1.7715236826173646,
+      "learning_rate": 3.95958511209956e-06,
+      "loss": 1.017,
+      "step": 2937
+    },
+    {
+      "epoch": 0.30899075814742266,
+      "grad_norm": 2.6875698356975297,
+      "learning_rate": 3.95890726332131e-06,
+      "loss": 1.0554,
+      "step": 2938
+    },
+    {
+      "epoch": 0.30909592858927,
+      "grad_norm": 1.9619216486997688,
+      "learning_rate": 3.958229251863567e-06,
+      "loss": 0.9878,
+      "step": 2939
+    },
+    {
+      "epoch": 0.3092010990311173,
+      "grad_norm": 2.7541620784435406,
+      "learning_rate": 3.957551077801935e-06,
+      "loss": 0.9557,
+      "step": 2940
+    },
+    {
+      "epoch": 0.30930626947296463,
+      "grad_norm": 1.8909424032199753,
+      "learning_rate": 3.956872741212035e-06,
+      "loss": 1.0064,
+      "step": 2941
+    },
+    {
+      "epoch": 0.30941143991481196,
+      "grad_norm": 2.623868633279091,
+      "learning_rate": 3.956194242169506e-06,
+      "loss": 1.0086,
+      "step": 2942
+    },
+    {
+      "epoch": 0.3095166103566593,
+      "grad_norm": 3.8473610355629195,
+      "learning_rate": 3.955515580750008e-06,
+      "loss": 1.0438,
+      "step": 2943
+    },
+    {
+      "epoch": 0.3096217807985066,
+      "grad_norm": 3.550700597509594,
+      "learning_rate": 3.954836757029214e-06,
+      "loss": 1.0497,
+      "step": 2944
+    },
+    {
+      "epoch": 0.3097269512403539,
+      "grad_norm": 2.9530235516637013,
+      "learning_rate": 3.9541577710828225e-06,
+      "loss": 1.0073,
+      "step": 2945
+    },
+    {
+      "epoch": 0.3098321216822012,
+      "grad_norm": 2.5132191578181384,
+      "learning_rate": 3.953478622986542e-06,
+      "loss": 1.0157,
+      "step": 2946
+    },
+    {
+      "epoch": 0.3099372921240485,
+      "grad_norm": 2.815147625525899,
+      "learning_rate": 3.952799312816105e-06,
+      "loss": 1.0341,
+      "step": 2947
+    },
+    {
+      "epoch": 0.31004246256589585,
+      "grad_norm": 2.3437305046069077,
+      "learning_rate": 3.9521198406472575e-06,
+      "loss": 1.0106,
+      "step": 2948
+    },
+    {
+      "epoch": 0.3101476330077432,
+      "grad_norm": 2.7478630820842724,
+      "learning_rate": 3.9514402065557675e-06,
+      "loss": 0.9561,
+      "step": 2949
+    },
+    {
+      "epoch": 0.3102528034495905,
+      "grad_norm": 3.1039712218296445,
+      "learning_rate": 3.950760410617421e-06,
+      "loss": 1.0222,
+      "step": 2950
+    },
+    {
+      "epoch": 0.3103579738914378,
+      "grad_norm": 2.758211476967507,
+      "learning_rate": 3.950080452908016e-06,
+      "loss": 0.9854,
+      "step": 2951
+    },
+    {
+      "epoch": 0.31046314433328515,
+      "grad_norm": 3.118482777087128,
+      "learning_rate": 3.949400333503378e-06,
+      "loss": 1.0438,
+      "step": 2952
+    },
+    {
+      "epoch": 0.31056831477513247,
+      "grad_norm": 2.3959007914603077,
+      "learning_rate": 3.948720052479343e-06,
+      "loss": 1.002,
+      "step": 2953
+    },
+    {
+      "epoch": 0.3106734852169798,
+      "grad_norm": 1.9575742067090993,
+      "learning_rate": 3.948039609911768e-06,
+      "loss": 0.9846,
+      "step": 2954
+    },
+    {
+      "epoch": 0.31077865565882706,
+      "grad_norm": 2.218966006005466,
+      "learning_rate": 3.947359005876527e-06,
+      "loss": 1.0369,
+      "step": 2955
+    },
+    {
+      "epoch": 0.3108838261006744,
+      "grad_norm": 2.957402283333697,
+      "learning_rate": 3.946678240449515e-06,
+      "loss": 0.9891,
+      "step": 2956
+    },
+    {
+      "epoch": 0.3109889965425217,
+      "grad_norm": 2.6314748348639805,
+      "learning_rate": 3.94599731370664e-06,
+      "loss": 1.0428,
+      "step": 2957
+    },
+    {
+      "epoch": 0.31109416698436904,
+      "grad_norm": 2.7377714026833866,
+      "learning_rate": 3.945316225723832e-06,
+      "loss": 1.035,
+      "step": 2958
+    },
+    {
+      "epoch": 0.31119933742621636,
+      "grad_norm": 2.053832855562837,
+      "learning_rate": 3.944634976577036e-06,
+      "loss": 0.9603,
+      "step": 2959
+    },
+    {
+      "epoch": 0.3113045078680637,
+      "grad_norm": 3.1158025154221125,
+      "learning_rate": 3.943953566342219e-06,
+      "loss": 0.9998,
+      "step": 2960
+    },
+    {
+      "epoch": 0.311409678309911,
+      "grad_norm": 2.396662606226732,
+      "learning_rate": 3.9432719950953625e-06,
+      "loss": 1.003,
+      "step": 2961
+    },
+    {
+      "epoch": 0.31151484875175833,
+      "grad_norm": 2.477520494380087,
+      "learning_rate": 3.942590262912466e-06,
+      "loss": 1.0024,
+      "step": 2962
+    },
+    {
+      "epoch": 0.31162001919360566,
+      "grad_norm": 2.966774835153069,
+      "learning_rate": 3.941908369869549e-06,
+      "loss": 1.024,
+      "step": 2963
+    },
+    {
+      "epoch": 0.311725189635453,
+      "grad_norm": 2.3612255349598286,
+      "learning_rate": 3.941226316042648e-06,
+      "loss": 0.9814,
+      "step": 2964
+    },
+    {
+      "epoch": 0.31183036007730025,
+      "grad_norm": 3.0645177496274334,
+      "learning_rate": 3.940544101507817e-06,
+      "loss": 0.9985,
+      "step": 2965
+    },
+    {
+      "epoch": 0.3119355305191476,
+      "grad_norm": 2.695073034424229,
+      "learning_rate": 3.939861726341128e-06,
+      "loss": 0.9781,
+      "step": 2966
+    },
+    {
+      "epoch": 0.3120407009609949,
+      "grad_norm": 2.731332674190935,
+      "learning_rate": 3.939179190618671e-06,
+      "loss": 0.9987,
+      "step": 2967
+    },
+    {
+      "epoch": 0.3121458714028422,
+      "grad_norm": 1.620215381519614,
+      "learning_rate": 3.938496494416554e-06,
+      "loss": 0.9743,
+      "step": 2968
+    },
+    {
+      "epoch": 0.31225104184468955,
+      "grad_norm": 2.3891831524051295,
+      "learning_rate": 3.937813637810904e-06,
+      "loss": 1.0118,
+      "step": 2969
+    },
+    {
+      "epoch": 0.31235621228653687,
+      "grad_norm": 2.2376707736140067,
+      "learning_rate": 3.937130620877863e-06,
+      "loss": 0.9824,
+      "step": 2970
+    },
+    {
+      "epoch": 0.3124613827283842,
+      "grad_norm": 2.4199521650953866,
+      "learning_rate": 3.936447443693595e-06,
+      "loss": 0.9938,
+      "step": 2971
+    },
+    {
+      "epoch": 0.3125665531702315,
+      "grad_norm": 2.3031295788347226,
+      "learning_rate": 3.935764106334278e-06,
+      "loss": 1.0102,
+      "step": 2972
+    },
+    {
+      "epoch": 0.31267172361207884,
+      "grad_norm": 3.5924980543328506,
+      "learning_rate": 3.935080608876109e-06,
+      "loss": 0.9912,
+      "step": 2973
+    },
+    {
+      "epoch": 0.31277689405392617,
+      "grad_norm": 2.8358224243128154,
+      "learning_rate": 3.934396951395305e-06,
+      "loss": 1.004,
+      "step": 2974
+    },
+    {
+      "epoch": 0.31288206449577344,
+      "grad_norm": 2.094776426203808,
+      "learning_rate": 3.933713133968097e-06,
+      "loss": 0.96,
+      "step": 2975
+    },
+    {
+      "epoch": 0.31298723493762076,
+      "grad_norm": 2.8795819218722722,
+      "learning_rate": 3.933029156670738e-06,
+      "loss": 0.99,
+      "step": 2976
+    },
+    {
+      "epoch": 0.3130924053794681,
+      "grad_norm": 3.1871471127409756,
+      "learning_rate": 3.9323450195794954e-06,
+      "loss": 0.9891,
+      "step": 2977
+    },
+    {
+      "epoch": 0.3131975758213154,
+      "grad_norm": 2.573983278943729,
+      "learning_rate": 3.9316607227706564e-06,
+      "loss": 0.995,
+      "step": 2978
+    },
+    {
+      "epoch": 0.31330274626316273,
+      "grad_norm": 2.4060298355716445,
+      "learning_rate": 3.930976266320525e-06,
+      "loss": 1.0275,
+      "step": 2979
+    },
+    {
+      "epoch": 0.31340791670501006,
+      "grad_norm": 2.4230905246024435,
+      "learning_rate": 3.930291650305424e-06,
+      "loss": 0.997,
+      "step": 2980
+    },
+    {
+      "epoch": 0.3135130871468574,
+      "grad_norm": 1.8561383044323279,
+      "learning_rate": 3.929606874801694e-06,
+      "loss": 0.9847,
+      "step": 2981
+    },
+    {
+      "epoch": 0.3136182575887047,
+      "grad_norm": 2.549475061635563,
+      "learning_rate": 3.9289219398856905e-06,
+      "loss": 0.9506,
+      "step": 2982
+    },
+    {
+      "epoch": 0.31372342803055203,
+      "grad_norm": 2.220462170609673,
+      "learning_rate": 3.928236845633791e-06,
+      "loss": 1.0019,
+      "step": 2983
+    },
+    {
+      "epoch": 0.31382859847239936,
+      "grad_norm": 2.856787147753265,
+      "learning_rate": 3.927551592122389e-06,
+      "loss": 1.0285,
+      "step": 2984
+    },
+    {
+      "epoch": 0.3139337689142466,
+      "grad_norm": 3.083922842525906,
+      "learning_rate": 3.926866179427894e-06,
+      "loss": 1.0602,
+      "step": 2985
+    },
+    {
+      "epoch": 0.31403893935609395,
+      "grad_norm": 2.597249942858582,
+      "learning_rate": 3.926180607626735e-06,
+      "loss": 1.0163,
+      "step": 2986
+    },
+    {
+      "epoch": 0.3141441097979413,
+      "grad_norm": 2.3456341225046278,
+      "learning_rate": 3.92549487679536e-06,
+      "loss": 1.0348,
+      "step": 2987
+    },
+    {
+      "epoch": 0.3142492802397886,
+      "grad_norm": 3.092563769825593,
+      "learning_rate": 3.924808987010234e-06,
+      "loss": 1.0002,
+      "step": 2988
+    },
+    {
+      "epoch": 0.3143544506816359,
+      "grad_norm": 2.1068768689816983,
+      "learning_rate": 3.924122938347835e-06,
+      "loss": 0.994,
+      "step": 2989
+    },
+    {
+      "epoch": 0.31445962112348325,
+      "grad_norm": 2.5596294401722846,
+      "learning_rate": 3.923436730884668e-06,
+      "loss": 1.0405,
+      "step": 2990
+    },
+    {
+      "epoch": 0.31456479156533057,
+      "grad_norm": 2.5979688517209687,
+      "learning_rate": 3.922750364697246e-06,
+      "loss": 1.0104,
+      "step": 2991
+    },
+    {
+      "epoch": 0.3146699620071779,
+      "grad_norm": 3.0395592306666175,
+      "learning_rate": 3.922063839862107e-06,
+      "loss": 1.0278,
+      "step": 2992
+    },
+    {
+      "epoch": 0.3147751324490252,
+      "grad_norm": 3.095491872011433,
+      "learning_rate": 3.921377156455802e-06,
+      "loss": 1.0113,
+      "step": 2993
+    },
+    {
+      "epoch": 0.31488030289087254,
+      "grad_norm": 3.2475174637972857,
+      "learning_rate": 3.920690314554903e-06,
+      "loss": 1.0456,
+      "step": 2994
+    },
+    {
+      "epoch": 0.3149854733327198,
+      "grad_norm": 2.4088622405130926,
+      "learning_rate": 3.9200033142359975e-06,
+      "loss": 1.0592,
+      "step": 2995
+    },
+    {
+      "epoch": 0.31509064377456714,
+      "grad_norm": 2.71257741566834,
+      "learning_rate": 3.919316155575692e-06,
+      "loss": 1.0116,
+      "step": 2996
+    },
+    {
+      "epoch": 0.31519581421641446,
+      "grad_norm": 2.450376163606724,
+      "learning_rate": 3.918628838650609e-06,
+      "loss": 1.0195,
+      "step": 2997
+    },
+    {
+      "epoch": 0.3153009846582618,
+      "grad_norm": 3.3118566816217427,
+      "learning_rate": 3.9179413635373895e-06,
+      "loss": 0.968,
+      "step": 2998
+    },
+    {
+      "epoch": 0.3154061551001091,
+      "grad_norm": 2.242881590172938,
+      "learning_rate": 3.917253730312694e-06,
+      "loss": 0.9813,
+      "step": 2999
+    },
+    {
+      "epoch": 0.31551132554195643,
+      "grad_norm": 2.427670249904579,
+      "learning_rate": 3.916565939053198e-06,
+      "loss": 1.018,
+      "step": 3000
+    },
+    {
+      "epoch": 0.31561649598380376,
+      "grad_norm": 3.0529942785276436,
+      "learning_rate": 3.915877989835595e-06,
+      "loss": 1.03,
+      "step": 3001
+    },
+    {
+      "epoch": 0.3157216664256511,
+      "grad_norm": 2.5578272107417317,
+      "learning_rate": 3.915189882736597e-06,
+      "loss": 1.0433,
+      "step": 3002
+    },
+    {
+      "epoch": 0.3158268368674984,
+      "grad_norm": 2.585770865143952,
+      "learning_rate": 3.914501617832935e-06,
+      "loss": 1.0069,
+      "step": 3003
+    },
+    {
+      "epoch": 0.31593200730934573,
+      "grad_norm": 3.2812648373996893,
+      "learning_rate": 3.9138131952013535e-06,
+      "loss": 1.0027,
+      "step": 3004
+    },
+    {
+      "epoch": 0.31603717775119305,
+      "grad_norm": 3.3571648099609654,
+      "learning_rate": 3.913124614918617e-06,
+      "loss": 1.0312,
+      "step": 3005
+    },
+    {
+      "epoch": 0.3161423481930403,
+      "grad_norm": 1.7227150762858836,
+      "learning_rate": 3.9124358770615094e-06,
+      "loss": 0.9955,
+      "step": 3006
+    },
+    {
+      "epoch": 0.31624751863488765,
+      "grad_norm": 3.5117489508527022,
+      "learning_rate": 3.911746981706829e-06,
+      "loss": 1.0127,
+      "step": 3007
+    },
+    {
+      "epoch": 0.31635268907673497,
+      "grad_norm": 2.651233249155203,
+      "learning_rate": 3.911057928931394e-06,
+      "loss": 0.9847,
+      "step": 3008
+    },
+    {
+      "epoch": 0.3164578595185823,
+      "grad_norm": 3.406708098715362,
+      "learning_rate": 3.910368718812037e-06,
+      "loss": 1.0204,
+      "step": 3009
+    },
+    {
+      "epoch": 0.3165630299604296,
+      "grad_norm": 2.6490697167241666,
+      "learning_rate": 3.909679351425612e-06,
+      "loss": 1.0232,
+      "step": 3010
+    },
+    {
+      "epoch": 0.31666820040227694,
+      "grad_norm": 3.0367731314433857,
+      "learning_rate": 3.90898982684899e-06,
+      "loss": 1.0102,
+      "step": 3011
+    },
+    {
+      "epoch": 0.31677337084412427,
+      "grad_norm": 1.99254564411588,
+      "learning_rate": 3.908300145159055e-06,
+      "loss": 0.9834,
+      "step": 3012
+    },
+    {
+      "epoch": 0.3168785412859716,
+      "grad_norm": 3.2342199206410482,
+      "learning_rate": 3.907610306432716e-06,
+      "loss": 0.9951,
+      "step": 3013
+    },
+    {
+      "epoch": 0.3169837117278189,
+      "grad_norm": 2.322687279586599,
+      "learning_rate": 3.906920310746891e-06,
+      "loss": 1.0255,
+      "step": 3014
+    },
+    {
+      "epoch": 0.31708888216966624,
+      "grad_norm": 2.2910797838024104,
+      "learning_rate": 3.906230158178523e-06,
+      "loss": 1.0091,
+      "step": 3015
+    },
+    {
+      "epoch": 0.3171940526115135,
+      "grad_norm": 2.5171087691012386,
+      "learning_rate": 3.9055398488045685e-06,
+      "loss": 0.935,
+      "step": 3016
+    },
+    {
+      "epoch": 0.31729922305336083,
+      "grad_norm": 1.9921759243129207,
+      "learning_rate": 3.9048493827020015e-06,
+      "loss": 1.0218,
+      "step": 3017
+    },
+    {
+      "epoch": 0.31740439349520816,
+      "grad_norm": 2.1147110832212928,
+      "learning_rate": 3.904158759947816e-06,
+      "loss": 1.0058,
+      "step": 3018
+    },
+    {
+      "epoch": 0.3175095639370555,
+      "grad_norm": 3.0475021542334555,
+      "learning_rate": 3.9034679806190204e-06,
+      "loss": 1.0187,
+      "step": 3019
+    },
+    {
+      "epoch": 0.3176147343789028,
+      "grad_norm": 2.745517766757522,
+      "learning_rate": 3.902777044792642e-06,
+      "loss": 1.0115,
+      "step": 3020
+    },
+    {
+      "epoch": 0.31771990482075013,
+      "grad_norm": 1.5206699032468878,
+      "learning_rate": 3.902085952545726e-06,
+      "loss": 0.9619,
+      "step": 3021
+    },
+    {
+      "epoch": 0.31782507526259746,
+      "grad_norm": 2.4452139648992395,
+      "learning_rate": 3.901394703955335e-06,
+      "loss": 0.9903,
+      "step": 3022
+    },
+    {
+      "epoch": 0.3179302457044448,
+      "grad_norm": 2.5184421966703594,
+      "learning_rate": 3.900703299098548e-06,
+      "loss": 0.9858,
+      "step": 3023
+    },
+    {
+      "epoch": 0.3180354161462921,
+      "grad_norm": 1.9652874285496023,
+      "learning_rate": 3.900011738052463e-06,
+      "loss": 1.0252,
+      "step": 3024
+    },
+    {
+      "epoch": 0.31814058658813943,
+      "grad_norm": 2.392255529966947,
+      "learning_rate": 3.899320020894192e-06,
+      "loss": 1.0227,
+      "step": 3025
+    },
+    {
+      "epoch": 0.3182457570299867,
+      "grad_norm": 2.1611953399013153,
+      "learning_rate": 3.898628147700869e-06,
+      "loss": 0.9738,
+      "step": 3026
+    },
+    {
+      "epoch": 0.318350927471834,
+      "grad_norm": 2.2430395870179063,
+      "learning_rate": 3.8979361185496426e-06,
+      "loss": 0.9749,
+      "step": 3027
+    },
+    {
+      "epoch": 0.31845609791368135,
+      "grad_norm": 3.541913920661668,
+      "learning_rate": 3.897243933517679e-06,
+      "loss": 1.0152,
+      "step": 3028
+    },
+    {
+      "epoch": 0.31856126835552867,
+      "grad_norm": 1.7600013581050258,
+      "learning_rate": 3.896551592682164e-06,
+      "loss": 0.9925,
+      "step": 3029
+    },
+    {
+      "epoch": 0.318666438797376,
+      "grad_norm": 3.0382688857854387,
+      "learning_rate": 3.895859096120296e-06,
+      "loss": 1.005,
+      "step": 3030
+    },
+    {
+      "epoch": 0.3187716092392233,
+      "grad_norm": 1.7559359887388504,
+      "learning_rate": 3.8951664439092966e-06,
+      "loss": 1.0027,
+      "step": 3031
+    },
+    {
+      "epoch": 0.31887677968107064,
+      "grad_norm": 2.41513756248943,
+      "learning_rate": 3.8944736361263996e-06,
+      "loss": 1.0058,
+      "step": 3032
+    },
+    {
+      "epoch": 0.31898195012291797,
+      "grad_norm": 3.824958571683499,
+      "learning_rate": 3.89378067284886e-06,
+      "loss": 0.9754,
+      "step": 3033
+    },
+    {
+      "epoch": 0.3190871205647653,
+      "grad_norm": 2.8869062987377174,
+      "learning_rate": 3.893087554153948e-06,
+      "loss": 0.974,
+      "step": 3034
+    },
+    {
+      "epoch": 0.3191922910066126,
+      "grad_norm": 3.3376996294364782,
+      "learning_rate": 3.892394280118952e-06,
+      "loss": 1.0322,
+      "step": 3035
+    },
+    {
+      "epoch": 0.3192974614484599,
+      "grad_norm": 2.1493242739855316,
+      "learning_rate": 3.891700850821178e-06,
+      "loss": 0.9793,
+      "step": 3036
+    },
+    {
+      "epoch": 0.3194026318903072,
+      "grad_norm": 2.0600928702389694,
+      "learning_rate": 3.891007266337947e-06,
+      "loss": 0.9667,
+      "step": 3037
+    },
+    {
+      "epoch": 0.31950780233215453,
+      "grad_norm": 2.422518969459341,
+      "learning_rate": 3.8903135267466e-06,
+      "loss": 1.0283,
+      "step": 3038
+    },
+    {
+      "epoch": 0.31961297277400186,
+      "grad_norm": 2.7095503103612764,
+      "learning_rate": 3.889619632124495e-06,
+      "loss": 1.0291,
+      "step": 3039
+    },
+    {
+      "epoch": 0.3197181432158492,
+      "grad_norm": 2.6136848210226833,
+      "learning_rate": 3.888925582549006e-06,
+      "loss": 1.0488,
+      "step": 3040
+    },
+    {
+      "epoch": 0.3198233136576965,
+      "grad_norm": 3.3309108315349962,
+      "learning_rate": 3.888231378097525e-06,
+      "loss": 1.0006,
+      "step": 3041
+    },
+    {
+      "epoch": 0.31992848409954383,
+      "grad_norm": 2.9848323149363716,
+      "learning_rate": 3.8875370188474606e-06,
+      "loss": 1.0346,
+      "step": 3042
+    },
+    {
+      "epoch": 0.32003365454139115,
+      "grad_norm": 1.8815097278144668,
+      "learning_rate": 3.88684250487624e-06,
+      "loss": 0.9791,
+      "step": 3043
+    },
+    {
+      "epoch": 0.3201388249832385,
+      "grad_norm": 2.975369543509321,
+      "learning_rate": 3.886147836261307e-06,
+      "loss": 1.0097,
+      "step": 3044
+    },
+    {
+      "epoch": 0.3202439954250858,
+      "grad_norm": 1.9729548982996215,
+      "learning_rate": 3.8854530130801226e-06,
+      "loss": 1.0049,
+      "step": 3045
+    },
+    {
+      "epoch": 0.32034916586693307,
+      "grad_norm": 2.964599401521195,
+      "learning_rate": 3.884758035410165e-06,
+      "loss": 0.9898,
+      "step": 3046
+    },
+    {
+      "epoch": 0.3204543363087804,
+      "grad_norm": 2.240600934404886,
+      "learning_rate": 3.884062903328929e-06,
+      "loss": 0.9923,
+      "step": 3047
+    },
+    {
+      "epoch": 0.3205595067506277,
+      "grad_norm": 2.598966467579801,
+      "learning_rate": 3.883367616913927e-06,
+      "loss": 1.0003,
+      "step": 3048
+    },
+    {
+      "epoch": 0.32066467719247504,
+      "grad_norm": 2.2641338646409883,
+      "learning_rate": 3.8826721762426885e-06,
+      "loss": 1.002,
+      "step": 3049
+    },
+    {
+      "epoch": 0.32076984763432237,
+      "grad_norm": 2.2997383920918897,
+      "learning_rate": 3.881976581392763e-06,
+      "loss": 1.0243,
+      "step": 3050
+    },
+    {
+      "epoch": 0.3208750180761697,
+      "grad_norm": 2.1550910823149136,
+      "learning_rate": 3.881280832441712e-06,
+      "loss": 0.991,
+      "step": 3051
+    },
+    {
+      "epoch": 0.320980188518017,
+      "grad_norm": 2.855557332734093,
+      "learning_rate": 3.880584929467119e-06,
+      "loss": 0.9982,
+      "step": 3052
+    },
+    {
+      "epoch": 0.32108535895986434,
+      "grad_norm": 3.5691659687435897,
+      "learning_rate": 3.879888872546581e-06,
+      "loss": 1.0354,
+      "step": 3053
+    },
+    {
+      "epoch": 0.32119052940171167,
+      "grad_norm": 2.2916019493138817,
+      "learning_rate": 3.879192661757715e-06,
+      "loss": 0.9934,
+      "step": 3054
+    },
+    {
+      "epoch": 0.321295699843559,
+      "grad_norm": 2.473847967975587,
+      "learning_rate": 3.878496297178151e-06,
+      "loss": 1.0425,
+      "step": 3055
+    },
+    {
+      "epoch": 0.32140087028540626,
+      "grad_norm": 3.677982616082223,
+      "learning_rate": 3.8777997788855435e-06,
+      "loss": 0.9895,
+      "step": 3056
+    },
+    {
+      "epoch": 0.3215060407272536,
+      "grad_norm": 2.7264358103840376,
+      "learning_rate": 3.877103106957559e-06,
+      "loss": 1.0241,
+      "step": 3057
+    },
+    {
+      "epoch": 0.3216112111691009,
+      "grad_norm": 2.2211588348254185,
+      "learning_rate": 3.876406281471877e-06,
+      "loss": 1.0182,
+      "step": 3058
+    },
+    {
+      "epoch": 0.32171638161094823,
+      "grad_norm": 3.091295524354873,
+      "learning_rate": 3.875709302506204e-06,
+      "loss": 1.0349,
+      "step": 3059
+    },
+    {
+      "epoch": 0.32182155205279556,
+      "grad_norm": 2.236128132871001,
+      "learning_rate": 3.875012170138256e-06,
+      "loss": 0.9954,
+      "step": 3060
+    },
+    {
+      "epoch": 0.3219267224946429,
+      "grad_norm": 2.66801073802577,
+      "learning_rate": 3.874314884445771e-06,
+      "loss": 0.9817,
+      "step": 3061
+    },
+    {
+      "epoch": 0.3220318929364902,
+      "grad_norm": 2.6169139547610816,
+      "learning_rate": 3.873617445506499e-06,
+      "loss": 1.0343,
+      "step": 3062
+    },
+    {
+      "epoch": 0.32213706337833753,
+      "grad_norm": 2.709426123429328,
+      "learning_rate": 3.872919853398212e-06,
+      "loss": 1.0071,
+      "step": 3063
+    },
+    {
+      "epoch": 0.32224223382018485,
+      "grad_norm": 2.299843384495024,
+      "learning_rate": 3.872222108198696e-06,
+      "loss": 1.0012,
+      "step": 3064
+    },
+    {
+      "epoch": 0.3223474042620322,
+      "grad_norm": 2.636252219736122,
+      "learning_rate": 3.871524209985755e-06,
+      "loss": 1.0427,
+      "step": 3065
+    },
+    {
+      "epoch": 0.3224525747038795,
+      "grad_norm": 3.117900115113946,
+      "learning_rate": 3.87082615883721e-06,
+      "loss": 1.0041,
+      "step": 3066
+    },
+    {
+      "epoch": 0.32255774514572677,
+      "grad_norm": 2.2357890840611465,
+      "learning_rate": 3.8701279548309e-06,
+      "loss": 0.9666,
+      "step": 3067
+    },
+    {
+      "epoch": 0.3226629155875741,
+      "grad_norm": 2.566115912683812,
+      "learning_rate": 3.869429598044679e-06,
+      "loss": 1.0166,
+      "step": 3068
+    },
+    {
+      "epoch": 0.3227680860294214,
+      "grad_norm": 2.7172969124996165,
+      "learning_rate": 3.868731088556419e-06,
+      "loss": 1.0099,
+      "step": 3069
+    },
+    {
+      "epoch": 0.32287325647126874,
+      "grad_norm": 1.5483818724179716,
+      "learning_rate": 3.868032426444012e-06,
+      "loss": 0.9894,
+      "step": 3070
+    },
+    {
+      "epoch": 0.32297842691311607,
+      "grad_norm": 2.318650017542439,
+      "learning_rate": 3.867333611785361e-06,
+      "loss": 1.05,
+      "step": 3071
+    },
+    {
+      "epoch": 0.3230835973549634,
+      "grad_norm": 1.917805124510381,
+      "learning_rate": 3.86663464465839e-06,
+      "loss": 0.9977,
+      "step": 3072
+    },
+    {
+      "epoch": 0.3231887677968107,
+      "grad_norm": 2.3161890326941275,
+      "learning_rate": 3.86593552514104e-06,
+      "loss": 1.003,
+      "step": 3073
+    },
+    {
+      "epoch": 0.32329393823865804,
+      "grad_norm": 2.1693043396128093,
+      "learning_rate": 3.865236253311268e-06,
+      "loss": 0.9695,
+      "step": 3074
+    },
+    {
+      "epoch": 0.32339910868050536,
+      "grad_norm": 2.8668301075021914,
+      "learning_rate": 3.864536829247049e-06,
+      "loss": 0.9888,
+      "step": 3075
+    },
+    {
+      "epoch": 0.3235042791223527,
+      "grad_norm": 2.4444547149619296,
+      "learning_rate": 3.863837253026372e-06,
+      "loss": 1.0067,
+      "step": 3076
+    },
+    {
+      "epoch": 0.32360944956419996,
+      "grad_norm": 3.3217478770609925,
+      "learning_rate": 3.863137524727248e-06,
+      "loss": 1.018,
+      "step": 3077
+    },
+    {
+      "epoch": 0.3237146200060473,
+      "grad_norm": 3.375894933254088,
+      "learning_rate": 3.862437644427699e-06,
+      "loss": 1.0314,
+      "step": 3078
+    },
+    {
+      "epoch": 0.3238197904478946,
+      "grad_norm": 3.1024834482299135,
+      "learning_rate": 3.86173761220577e-06,
+      "loss": 0.9857,
+      "step": 3079
+    },
+    {
+      "epoch": 0.32392496088974193,
+      "grad_norm": 1.8859603281941375,
+      "learning_rate": 3.8610374281395205e-06,
+      "loss": 1.0021,
+      "step": 3080
+    },
+    {
+      "epoch": 0.32403013133158926,
+      "grad_norm": 2.181930740279361,
+      "learning_rate": 3.860337092307023e-06,
+      "loss": 1.0024,
+      "step": 3081
+    },
+    {
+      "epoch": 0.3241353017734366,
+      "grad_norm": 2.2394462965094806,
+      "learning_rate": 3.859636604786372e-06,
+      "loss": 1.0499,
+      "step": 3082
+    },
+    {
+      "epoch": 0.3242404722152839,
+      "grad_norm": 2.429116428790114,
+      "learning_rate": 3.8589359656556775e-06,
+      "loss": 1.0016,
+      "step": 3083
+    },
+    {
+      "epoch": 0.32434564265713123,
+      "grad_norm": 3.2486086179267697,
+      "learning_rate": 3.858235174993067e-06,
+      "loss": 0.9778,
+      "step": 3084
+    },
+    {
+      "epoch": 0.32445081309897855,
+      "grad_norm": 2.168312151384131,
+      "learning_rate": 3.857534232876684e-06,
+      "loss": 0.9809,
+      "step": 3085
+    },
+    {
+      "epoch": 0.3245559835408259,
+      "grad_norm": 2.894850817884378,
+      "learning_rate": 3.856833139384687e-06,
+      "loss": 1.0092,
+      "step": 3086
+    },
+    {
+      "epoch": 0.32466115398267315,
+      "grad_norm": 3.1973922575609697,
+      "learning_rate": 3.856131894595255e-06,
+      "loss": 1.0425,
+      "step": 3087
+    },
+    {
+      "epoch": 0.32476632442452047,
+      "grad_norm": 2.6786962321285883,
+      "learning_rate": 3.855430498586582e-06,
+      "loss": 0.9848,
+      "step": 3088
+    },
+    {
+      "epoch": 0.3248714948663678,
+      "grad_norm": 1.7068699911559382,
+      "learning_rate": 3.8547289514368795e-06,
+      "loss": 0.9864,
+      "step": 3089
+    },
+    {
+      "epoch": 0.3249766653082151,
+      "grad_norm": 1.762520608645227,
+      "learning_rate": 3.854027253224375e-06,
+      "loss": 0.9843,
+      "step": 3090
+    },
+    {
+      "epoch": 0.32508183575006244,
+      "grad_norm": 1.756341086493816,
+      "learning_rate": 3.853325404027313e-06,
+      "loss": 0.9738,
+      "step": 3091
+    },
+    {
+      "epoch": 0.32518700619190977,
+      "grad_norm": 1.9319056465036724,
+      "learning_rate": 3.852623403923955e-06,
+      "loss": 1.0034,
+      "step": 3092
+    },
+    {
+      "epoch": 0.3252921766337571,
+      "grad_norm": 2.627736625312551,
+      "learning_rate": 3.851921252992581e-06,
+      "loss": 1.0046,
+      "step": 3093
+    },
+    {
+      "epoch": 0.3253973470756044,
+      "grad_norm": 2.387276425999192,
+      "learning_rate": 3.851218951311484e-06,
+      "loss": 1.0061,
+      "step": 3094
+    },
+    {
+      "epoch": 0.32550251751745174,
+      "grad_norm": 2.6657811160762868,
+      "learning_rate": 3.850516498958979e-06,
+      "loss": 0.989,
+      "step": 3095
+    },
+    {
+      "epoch": 0.32560768795929906,
+      "grad_norm": 3.7563246886039785,
+      "learning_rate": 3.849813896013392e-06,
+      "loss": 0.9911,
+      "step": 3096
+    },
+    {
+      "epoch": 0.32571285840114633,
+      "grad_norm": 3.186401005858273,
+      "learning_rate": 3.849111142553069e-06,
+      "loss": 0.9781,
+      "step": 3097
+    },
+    {
+      "epoch": 0.32581802884299366,
+      "grad_norm": 2.8120268250532128,
+      "learning_rate": 3.848408238656375e-06,
+      "loss": 1.0248,
+      "step": 3098
+    },
+    {
+      "epoch": 0.325923199284841,
+      "grad_norm": 2.5367126653996985,
+      "learning_rate": 3.8477051844016865e-06,
+      "loss": 1.0108,
+      "step": 3099
+    },
+    {
+      "epoch": 0.3260283697266883,
+      "grad_norm": 2.4405235194407804,
+      "learning_rate": 3.847001979867401e-06,
+      "loss": 0.9774,
+      "step": 3100
+    },
+    {
+      "epoch": 0.32613354016853563,
+      "grad_norm": 1.8950206598525756,
+      "learning_rate": 3.846298625131931e-06,
+      "loss": 1.0017,
+      "step": 3101
+    },
+    {
+      "epoch": 0.32623871061038295,
+      "grad_norm": 2.7265934900210964,
+      "learning_rate": 3.845595120273706e-06,
+      "loss": 1.0165,
+      "step": 3102
+    },
+    {
+      "epoch": 0.3263438810522303,
+      "grad_norm": 2.205211886185536,
+      "learning_rate": 3.8448914653711725e-06,
+      "loss": 1.037,
+      "step": 3103
+    },
+    {
+      "epoch": 0.3264490514940776,
+      "grad_norm": 1.9366743565239208,
+      "learning_rate": 3.844187660502792e-06,
+      "loss": 0.9738,
+      "step": 3104
+    },
+    {
+      "epoch": 0.3265542219359249,
+      "grad_norm": 2.593512907813824,
+      "learning_rate": 3.843483705747045e-06,
+      "loss": 1.0458,
+      "step": 3105
+    },
+    {
+      "epoch": 0.32665939237777225,
+      "grad_norm": 2.0167179831155115,
+      "learning_rate": 3.84277960118243e-06,
+      "loss": 1.0399,
+      "step": 3106
+    },
+    {
+      "epoch": 0.3267645628196195,
+      "grad_norm": 2.334539077805553,
+      "learning_rate": 3.842075346887457e-06,
+      "loss": 1.0077,
+      "step": 3107
+    },
+    {
+      "epoch": 0.32686973326146684,
+      "grad_norm": 2.604441795888962,
+      "learning_rate": 3.841370942940657e-06,
+      "loss": 1.0063,
+      "step": 3108
+    },
+    {
+      "epoch": 0.32697490370331417,
+      "grad_norm": 1.8763989090691047,
+      "learning_rate": 3.840666389420578e-06,
+      "loss": 1.0246,
+      "step": 3109
+    },
+    {
+      "epoch": 0.3270800741451615,
+      "grad_norm": 2.1829721118576337,
+      "learning_rate": 3.839961686405782e-06,
+      "loss": 1.0295,
+      "step": 3110
+    },
+    {
+      "epoch": 0.3271852445870088,
+      "grad_norm": 2.235743716349407,
+      "learning_rate": 3.839256833974848e-06,
+      "loss": 0.9813,
+      "step": 3111
+    },
+    {
+      "epoch": 0.32729041502885614,
+      "grad_norm": 2.682923955392616,
+      "learning_rate": 3.838551832206373e-06,
+      "loss": 0.9701,
+      "step": 3112
+    },
+    {
+      "epoch": 0.32739558547070347,
+      "grad_norm": 2.564345157324214,
+      "learning_rate": 3.8378466811789714e-06,
+      "loss": 1.0052,
+      "step": 3113
+    },
+    {
+      "epoch": 0.3275007559125508,
+      "grad_norm": 2.745614634832843,
+      "learning_rate": 3.837141380971271e-06,
+      "loss": 0.9663,
+      "step": 3114
+    },
+    {
+      "epoch": 0.3276059263543981,
+      "grad_norm": 4.102665613851121,
+      "learning_rate": 3.83643593166192e-06,
+      "loss": 1.0145,
+      "step": 3115
+    },
+    {
+      "epoch": 0.32771109679624544,
+      "grad_norm": 1.9187924233062632,
+      "learning_rate": 3.835730333329581e-06,
+      "loss": 1.0537,
+      "step": 3116
+    },
+    {
+      "epoch": 0.3278162672380927,
+      "grad_norm": 2.0366789642980856,
+      "learning_rate": 3.835024586052933e-06,
+      "loss": 1.0184,
+      "step": 3117
+    },
+    {
+      "epoch": 0.32792143767994003,
+      "grad_norm": 2.584247749331405,
+      "learning_rate": 3.834318689910673e-06,
+      "loss": 1.0199,
+      "step": 3118
+    },
+    {
+      "epoch": 0.32802660812178736,
+      "grad_norm": 2.3869440994924385,
+      "learning_rate": 3.833612644981514e-06,
+      "loss": 1.0261,
+      "step": 3119
+    },
+    {
+      "epoch": 0.3281317785636347,
+      "grad_norm": 1.744981904406401,
+      "learning_rate": 3.832906451344185e-06,
+      "loss": 0.9887,
+      "step": 3120
+    },
+    {
+      "epoch": 0.328236949005482,
+      "grad_norm": 2.205884774883593,
+      "learning_rate": 3.832200109077433e-06,
+      "loss": 1.0138,
+      "step": 3121
+    },
+    {
+      "epoch": 0.32834211944732933,
+      "grad_norm": 2.8960434627989566,
+      "learning_rate": 3.831493618260019e-06,
+      "loss": 1.0183,
+      "step": 3122
+    },
+    {
+      "epoch": 0.32844728988917665,
+      "grad_norm": 2.6153093943252803,
+      "learning_rate": 3.830786978970723e-06,
+      "loss": 1.0074,
+      "step": 3123
+    },
+    {
+      "epoch": 0.328552460331024,
+      "grad_norm": 3.2605234910491787,
+      "learning_rate": 3.830080191288342e-06,
+      "loss": 0.9695,
+      "step": 3124
+    },
+    {
+      "epoch": 0.3286576307728713,
+      "grad_norm": 2.3212838057371696,
+      "learning_rate": 3.829373255291686e-06,
+      "loss": 1.0006,
+      "step": 3125
+    },
+    {
+      "epoch": 0.3287628012147186,
+      "grad_norm": 3.1048512567149156,
+      "learning_rate": 3.828666171059586e-06,
+      "loss": 1.0273,
+      "step": 3126
+    },
+    {
+      "epoch": 0.32886797165656595,
+      "grad_norm": 3.432285163654643,
+      "learning_rate": 3.827958938670886e-06,
+      "loss": 1.0458,
+      "step": 3127
+    },
+    {
+      "epoch": 0.3289731420984132,
+      "grad_norm": 2.819849325554979,
+      "learning_rate": 3.827251558204449e-06,
+      "loss": 0.9533,
+      "step": 3128
+    },
+    {
+      "epoch": 0.32907831254026054,
+      "grad_norm": 1.6507578523510593,
+      "learning_rate": 3.826544029739152e-06,
+      "loss": 0.9536,
+      "step": 3129
+    },
+    {
+      "epoch": 0.32918348298210787,
+      "grad_norm": 2.4830132435715537,
+      "learning_rate": 3.825836353353891e-06,
+      "loss": 1.0009,
+      "step": 3130
+    },
+    {
+      "epoch": 0.3292886534239552,
+      "grad_norm": 2.4797089042570533,
+      "learning_rate": 3.825128529127577e-06,
+      "loss": 0.9801,
+      "step": 3131
+    },
+    {
+      "epoch": 0.3293938238658025,
+      "grad_norm": 2.4239798237039594,
+      "learning_rate": 3.824420557139139e-06,
+      "loss": 1.0182,
+      "step": 3132
+    },
+    {
+      "epoch": 0.32949899430764984,
+      "grad_norm": 1.945405865543695,
+      "learning_rate": 3.82371243746752e-06,
+      "loss": 1.0321,
+      "step": 3133
+    },
+    {
+      "epoch": 0.32960416474949716,
+      "grad_norm": 2.084748306095817,
+      "learning_rate": 3.823004170191681e-06,
+      "loss": 1.0155,
+      "step": 3134
+    },
+    {
+      "epoch": 0.3297093351913445,
+      "grad_norm": 2.251126194835179,
+      "learning_rate": 3.8222957553906e-06,
+      "loss": 1.0392,
+      "step": 3135
+    },
+    {
+      "epoch": 0.3298145056331918,
+      "grad_norm": 2.2034664670180626,
+      "learning_rate": 3.821587193143269e-06,
+      "loss": 0.9897,
+      "step": 3136
+    },
+    {
+      "epoch": 0.32991967607503914,
+      "grad_norm": 2.9409945673975626,
+      "learning_rate": 3.8208784835287e-06,
+      "loss": 0.9982,
+      "step": 3137
+    },
+    {
+      "epoch": 0.3300248465168864,
+      "grad_norm": 3.058005825612,
+      "learning_rate": 3.82016962662592e-06,
+      "loss": 1.0317,
+      "step": 3138
+    },
+    {
+      "epoch": 0.33013001695873373,
+      "grad_norm": 3.490847243165153,
+      "learning_rate": 3.819460622513971e-06,
+      "loss": 1.0036,
+      "step": 3139
+    },
+    {
+      "epoch": 0.33023518740058105,
+      "grad_norm": 2.071993271550719,
+      "learning_rate": 3.818751471271912e-06,
+      "loss": 0.9886,
+      "step": 3140
+    },
+    {
+      "epoch": 0.3303403578424284,
+      "grad_norm": 2.799536363219469,
+      "learning_rate": 3.818042172978821e-06,
+      "loss": 0.9794,
+      "step": 3141
+    },
+    {
+      "epoch": 0.3304455282842757,
+      "grad_norm": 3.262735659956398,
+      "learning_rate": 3.8173327277137876e-06,
+      "loss": 1.0189,
+      "step": 3142
+    },
+    {
+      "epoch": 0.330550698726123,
+      "grad_norm": 2.1874156683363664,
+      "learning_rate": 3.816623135555921e-06,
+      "loss": 0.963,
+      "step": 3143
+    },
+    {
+      "epoch": 0.33065586916797035,
+      "grad_norm": 2.487129588775537,
+      "learning_rate": 3.815913396584348e-06,
+      "loss": 1.0097,
+      "step": 3144
+    },
+    {
+      "epoch": 0.3307610396098177,
+      "grad_norm": 2.8581708249255096,
+      "learning_rate": 3.815203510878209e-06,
+      "loss": 1.0382,
+      "step": 3145
+    },
+    {
+      "epoch": 0.330866210051665,
+      "grad_norm": 2.353074717228548,
+      "learning_rate": 3.814493478516661e-06,
+      "loss": 1.0092,
+      "step": 3146
+    },
+    {
+      "epoch": 0.3309713804935123,
+      "grad_norm": 2.7540007639438375,
+      "learning_rate": 3.813783299578879e-06,
+      "loss": 1.0189,
+      "step": 3147
+    },
+    {
+      "epoch": 0.3310765509353596,
+      "grad_norm": 2.4522274567147657,
+      "learning_rate": 3.8130729741440536e-06,
+      "loss": 1.0141,
+      "step": 3148
+    },
+    {
+      "epoch": 0.3311817213772069,
+      "grad_norm": 2.849670426659251,
+      "learning_rate": 3.8123625022913915e-06,
+      "loss": 1.0118,
+      "step": 3149
+    },
+    {
+      "epoch": 0.33128689181905424,
+      "grad_norm": 2.3301835319177373,
+      "learning_rate": 3.811651884100115e-06,
+      "loss": 0.9961,
+      "step": 3150
+    },
+    {
+      "epoch": 0.33139206226090157,
+      "grad_norm": 2.3601708357586006,
+      "learning_rate": 3.8109411196494638e-06,
+      "loss": 0.9735,
+      "step": 3151
+    },
+    {
+      "epoch": 0.3314972327027489,
+      "grad_norm": 2.0573991989526474,
+      "learning_rate": 3.810230209018694e-06,
+      "loss": 1.021,
+      "step": 3152
+    },
+    {
+      "epoch": 0.3316024031445962,
+      "grad_norm": 2.9354196677378037,
+      "learning_rate": 3.809519152287079e-06,
+      "loss": 0.992,
+      "step": 3153
+    },
+    {
+      "epoch": 0.33170757358644354,
+      "grad_norm": 2.544344310226173,
+      "learning_rate": 3.8088079495339046e-06,
+      "loss": 0.957,
+      "step": 3154
+    },
+    {
+      "epoch": 0.33181274402829086,
+      "grad_norm": 2.2971377313074393,
+      "learning_rate": 3.8080966008384775e-06,
+      "loss": 1.0143,
+      "step": 3155
+    },
+    {
+      "epoch": 0.3319179144701382,
+      "grad_norm": 2.9198098245236026,
+      "learning_rate": 3.807385106280117e-06,
+      "loss": 0.966,
+      "step": 3156
+    },
+    {
+      "epoch": 0.3320230849119855,
+      "grad_norm": 2.787741411848439,
+      "learning_rate": 3.806673465938161e-06,
+      "loss": 1.0119,
+      "step": 3157
+    },
+    {
+      "epoch": 0.3321282553538328,
+      "grad_norm": 2.139566525593868,
+      "learning_rate": 3.805961679891964e-06,
+      "loss": 1.0094,
+      "step": 3158
+    },
+    {
+      "epoch": 0.3322334257956801,
+      "grad_norm": 2.55889761213254,
+      "learning_rate": 3.805249748220893e-06,
+      "loss": 1.0046,
+      "step": 3159
+    },
+    {
+      "epoch": 0.33233859623752743,
+      "grad_norm": 2.811994900697266,
+      "learning_rate": 3.804537671004337e-06,
+      "loss": 1.0026,
+      "step": 3160
+    },
+    {
+      "epoch": 0.33244376667937475,
+      "grad_norm": 2.6857582832549745,
+      "learning_rate": 3.8038254483216962e-06,
+      "loss": 0.9833,
+      "step": 3161
+    },
+    {
+      "epoch": 0.3325489371212221,
+      "grad_norm": 2.6228532392360866,
+      "learning_rate": 3.8031130802523896e-06,
+      "loss": 1.0335,
+      "step": 3162
+    },
+    {
+      "epoch": 0.3326541075630694,
+      "grad_norm": 2.6508818240324836,
+      "learning_rate": 3.802400566875851e-06,
+      "loss": 1.0416,
+      "step": 3163
+    },
+    {
+      "epoch": 0.3327592780049167,
+      "grad_norm": 2.708891165068549,
+      "learning_rate": 3.8016879082715326e-06,
+      "loss": 1.0311,
+      "step": 3164
+    },
+    {
+      "epoch": 0.33286444844676405,
+      "grad_norm": 1.9791714208493958,
+      "learning_rate": 3.8009751045189004e-06,
+      "loss": 1.0022,
+      "step": 3165
+    },
+    {
+      "epoch": 0.3329696188886114,
+      "grad_norm": 3.4480804252433592,
+      "learning_rate": 3.8002621556974367e-06,
+      "loss": 1.0225,
+      "step": 3166
+    },
+    {
+      "epoch": 0.3330747893304587,
+      "grad_norm": 1.9335962357260972,
+      "learning_rate": 3.7995490618866425e-06,
+      "loss": 0.9969,
+      "step": 3167
+    },
+    {
+      "epoch": 0.33317995977230597,
+      "grad_norm": 2.388818127758868,
+      "learning_rate": 3.7988358231660333e-06,
+      "loss": 1.023,
+      "step": 3168
+    },
+    {
+      "epoch": 0.3332851302141533,
+      "grad_norm": 3.4063184395962445,
+      "learning_rate": 3.7981224396151393e-06,
+      "loss": 1.0006,
+      "step": 3169
+    },
+    {
+      "epoch": 0.3333903006560006,
+      "grad_norm": 3.1228965688498764,
+      "learning_rate": 3.7974089113135094e-06,
+      "loss": 1.0256,
+      "step": 3170
+    },
+    {
+      "epoch": 0.33349547109784794,
+      "grad_norm": 2.969409956688769,
+      "learning_rate": 3.7966952383407075e-06,
+      "loss": 0.9796,
+      "step": 3171
+    },
+    {
+      "epoch": 0.33360064153969526,
+      "grad_norm": 1.772482770164931,
+      "learning_rate": 3.7959814207763134e-06,
+      "loss": 0.9964,
+      "step": 3172
+    },
+    {
+      "epoch": 0.3337058119815426,
+      "grad_norm": 2.5201047196046034,
+      "learning_rate": 3.7952674586999226e-06,
+      "loss": 1.0227,
+      "step": 3173
+    },
+    {
+      "epoch": 0.3338109824233899,
+      "grad_norm": 2.848018681472792,
+      "learning_rate": 3.794553352191149e-06,
+      "loss": 1.0123,
+      "step": 3174
+    },
+    {
+      "epoch": 0.33391615286523724,
+      "grad_norm": 3.1213302281106547,
+      "learning_rate": 3.7938391013296193e-06,
+      "loss": 1.0099,
+      "step": 3175
+    },
+    {
+      "epoch": 0.33402132330708456,
+      "grad_norm": 2.132897712038548,
+      "learning_rate": 3.79312470619498e-06,
+      "loss": 1.0306,
+      "step": 3176
+    },
+    {
+      "epoch": 0.3341264937489319,
+      "grad_norm": 1.6980061002378293,
+      "learning_rate": 3.79241016686689e-06,
+      "loss": 0.9981,
+      "step": 3177
+    },
+    {
+      "epoch": 0.33423166419077915,
+      "grad_norm": 2.133136371300042,
+      "learning_rate": 3.791695483425026e-06,
+      "loss": 0.9986,
+      "step": 3178
+    },
+    {
+      "epoch": 0.3343368346326265,
+      "grad_norm": 2.1135831702313888,
+      "learning_rate": 3.7909806559490827e-06,
+      "loss": 0.9966,
+      "step": 3179
+    },
+    {
+      "epoch": 0.3344420050744738,
+      "grad_norm": 2.517571111822009,
+      "learning_rate": 3.790265684518767e-06,
+      "loss": 1.0228,
+      "step": 3180
+    },
+    {
+      "epoch": 0.3345471755163211,
+      "grad_norm": 2.98686980884583,
+      "learning_rate": 3.7895505692138045e-06,
+      "loss": 1.0223,
+      "step": 3181
+    },
+    {
+      "epoch": 0.33465234595816845,
+      "grad_norm": 2.261356671644225,
+      "learning_rate": 3.7888353101139353e-06,
+      "loss": 1.0245,
+      "step": 3182
+    },
+    {
+      "epoch": 0.3347575164000158,
+      "grad_norm": 2.409763840683323,
+      "learning_rate": 3.7881199072989176e-06,
+      "loss": 0.9886,
+      "step": 3183
+    },
+    {
+      "epoch": 0.3348626868418631,
+      "grad_norm": 1.9087618932889743,
+      "learning_rate": 3.7874043608485234e-06,
+      "loss": 1.0211,
+      "step": 3184
+    },
+    {
+      "epoch": 0.3349678572837104,
+      "grad_norm": 1.998174510767556,
+      "learning_rate": 3.7866886708425427e-06,
+      "loss": 1.0038,
+      "step": 3185
+    },
+    {
+      "epoch": 0.33507302772555775,
+      "grad_norm": 1.9103622576526025,
+      "learning_rate": 3.785972837360779e-06,
+      "loss": 1.0043,
+      "step": 3186
+    },
+    {
+      "epoch": 0.3351781981674051,
+      "grad_norm": 1.9379810577362075,
+      "learning_rate": 3.785256860483054e-06,
+      "loss": 1.0059,
+      "step": 3187
+    },
+    {
+      "epoch": 0.3352833686092524,
+      "grad_norm": 2.9610384556218,
+      "learning_rate": 3.7845407402892066e-06,
+      "loss": 0.9998,
+      "step": 3188
+    },
+    {
+      "epoch": 0.33538853905109967,
+      "grad_norm": 2.2025225245305076,
+      "learning_rate": 3.7838244768590866e-06,
+      "loss": 1.0089,
+      "step": 3189
+    },
+    {
+      "epoch": 0.335493709492947,
+      "grad_norm": 2.8328379269098076,
+      "learning_rate": 3.7831080702725643e-06,
+      "loss": 1.0317,
+      "step": 3190
+    },
+    {
+      "epoch": 0.3355988799347943,
+      "grad_norm": 1.9624846245151617,
+      "learning_rate": 3.7823915206095246e-06,
+      "loss": 0.9584,
+      "step": 3191
+    },
+    {
+      "epoch": 0.33570405037664164,
+      "grad_norm": 2.3934098571822546,
+      "learning_rate": 3.7816748279498687e-06,
+      "loss": 1.0102,
+      "step": 3192
+    },
+    {
+      "epoch": 0.33580922081848896,
+      "grad_norm": 2.1927978381480053,
+      "learning_rate": 3.780957992373513e-06,
+      "loss": 1.0098,
+      "step": 3193
+    },
+    {
+      "epoch": 0.3359143912603363,
+      "grad_norm": 2.7690159806961896,
+      "learning_rate": 3.7802410139603908e-06,
+      "loss": 0.9457,
+      "step": 3194
+    },
+    {
+      "epoch": 0.3360195617021836,
+      "grad_norm": 2.5415137912355137,
+      "learning_rate": 3.7795238927904497e-06,
+      "loss": 0.9811,
+      "step": 3195
+    },
+    {
+      "epoch": 0.33612473214403094,
+      "grad_norm": 2.895165075185927,
+      "learning_rate": 3.778806628943655e-06,
+      "loss": 1.0291,
+      "step": 3196
+    },
+    {
+      "epoch": 0.33622990258587826,
+      "grad_norm": 3.0795576587129676,
+      "learning_rate": 3.778089222499987e-06,
+      "loss": 1.0233,
+      "step": 3197
+    },
+    {
+      "epoch": 0.3363350730277256,
+      "grad_norm": 2.4751867790109943,
+      "learning_rate": 3.7773716735394415e-06,
+      "loss": 1.0516,
+      "step": 3198
+    },
+    {
+      "epoch": 0.33644024346957285,
+      "grad_norm": 2.4430066060566134,
+      "learning_rate": 3.776653982142033e-06,
+      "loss": 0.9791,
+      "step": 3199
+    },
+    {
+      "epoch": 0.3365454139114202,
+      "grad_norm": 2.5324655109273606,
+      "learning_rate": 3.7759361483877865e-06,
+      "loss": 0.9815,
+      "step": 3200
+    },
+    {
+      "epoch": 0.3366505843532675,
+      "grad_norm": 2.7523454023910365,
+      "learning_rate": 3.7752181723567484e-06,
+      "loss": 1.0079,
+      "step": 3201
+    },
+    {
+      "epoch": 0.3367557547951148,
+      "grad_norm": 2.3630412261657607,
+      "learning_rate": 3.7745000541289777e-06,
+      "loss": 1.0065,
+      "step": 3202
+    },
+    {
+      "epoch": 0.33686092523696215,
+      "grad_norm": 2.775762382934883,
+      "learning_rate": 3.7737817937845504e-06,
+      "loss": 0.9977,
+      "step": 3203
+    },
+    {
+      "epoch": 0.3369660956788095,
+      "grad_norm": 2.4895504075619606,
+      "learning_rate": 3.7730633914035585e-06,
+      "loss": 1.0495,
+      "step": 3204
+    },
+    {
+      "epoch": 0.3370712661206568,
+      "grad_norm": 2.721766046225593,
+      "learning_rate": 3.7723448470661084e-06,
+      "loss": 0.949,
+      "step": 3205
+    },
+    {
+      "epoch": 0.3371764365625041,
+      "grad_norm": 2.9668981245557053,
+      "learning_rate": 3.771626160852324e-06,
+      "loss": 0.9979,
+      "step": 3206
+    },
+    {
+      "epoch": 0.33728160700435145,
+      "grad_norm": 3.793048897027738,
+      "learning_rate": 3.770907332842344e-06,
+      "loss": 1.0005,
+      "step": 3207
+    },
+    {
+      "epoch": 0.33738677744619877,
+      "grad_norm": 2.2379198509213,
+      "learning_rate": 3.770188363116324e-06,
+      "loss": 0.9751,
+      "step": 3208
+    },
+    {
+      "epoch": 0.33749194788804604,
+      "grad_norm": 3.13559481810738,
+      "learning_rate": 3.769469251754434e-06,
+      "loss": 0.9865,
+      "step": 3209
+    },
+    {
+      "epoch": 0.33759711832989336,
+      "grad_norm": 2.717848142691034,
+      "learning_rate": 3.7687499988368613e-06,
+      "loss": 1.0444,
+      "step": 3210
+    },
+    {
+      "epoch": 0.3377022887717407,
+      "grad_norm": 3.9961109065173073,
+      "learning_rate": 3.7680306044438074e-06,
+      "loss": 1.0281,
+      "step": 3211
+    },
+    {
+      "epoch": 0.337807459213588,
+      "grad_norm": 2.3111397830579867,
+      "learning_rate": 3.7673110686554915e-06,
+      "loss": 0.9546,
+      "step": 3212
+    },
+    {
+      "epoch": 0.33791262965543534,
+      "grad_norm": 2.556022750068411,
+      "learning_rate": 3.766591391552146e-06,
+      "loss": 1.0199,
+      "step": 3213
+    },
+    {
+      "epoch": 0.33801780009728266,
+      "grad_norm": 2.8546999903249555,
+      "learning_rate": 3.7658715732140206e-06,
+      "loss": 0.985,
+      "step": 3214
+    },
+    {
+      "epoch": 0.33812297053913,
+      "grad_norm": 3.3523241411809104,
+      "learning_rate": 3.7651516137213816e-06,
+      "loss": 1.0066,
+      "step": 3215
+    },
+    {
+      "epoch": 0.3382281409809773,
+      "grad_norm": 1.9532386436425269,
+      "learning_rate": 3.7644315131545096e-06,
+      "loss": 0.972,
+      "step": 3216
+    },
+    {
+      "epoch": 0.33833331142282463,
+      "grad_norm": 2.721629663326664,
+      "learning_rate": 3.763711271593702e-06,
+      "loss": 0.9979,
+      "step": 3217
+    },
+    {
+      "epoch": 0.33843848186467196,
+      "grad_norm": 3.4541692663759904,
+      "learning_rate": 3.76299088911927e-06,
+      "loss": 1.0571,
+      "step": 3218
+    },
+    {
+      "epoch": 0.3385436523065192,
+      "grad_norm": 2.976934045713143,
+      "learning_rate": 3.7622703658115435e-06,
+      "loss": 1.0059,
+      "step": 3219
+    },
+    {
+      "epoch": 0.33864882274836655,
+      "grad_norm": 3.0140365934725843,
+      "learning_rate": 3.761549701750865e-06,
+      "loss": 1.0244,
+      "step": 3220
+    },
+    {
+      "epoch": 0.3387539931902139,
+      "grad_norm": 1.9982975993871346,
+      "learning_rate": 3.7608288970175943e-06,
+      "loss": 1.0215,
+      "step": 3221
+    },
+    {
+      "epoch": 0.3388591636320612,
+      "grad_norm": 2.8744873417988788,
+      "learning_rate": 3.7601079516921076e-06,
+      "loss": 1.0015,
+      "step": 3222
+    },
+    {
+      "epoch": 0.3389643340739085,
+      "grad_norm": 2.150725684186714,
+      "learning_rate": 3.759386865854795e-06,
+      "loss": 1.0129,
+      "step": 3223
+    },
+    {
+      "epoch": 0.33906950451575585,
+      "grad_norm": 2.6319253916407,
+      "learning_rate": 3.758665639586064e-06,
+      "loss": 0.9982,
+      "step": 3224
+    },
+    {
+      "epoch": 0.3391746749576032,
+      "grad_norm": 2.1192499513393357,
+      "learning_rate": 3.7579442729663364e-06,
+      "loss": 0.9852,
+      "step": 3225
+    },
+    {
+      "epoch": 0.3392798453994505,
+      "grad_norm": 1.842379992675114,
+      "learning_rate": 3.75722276607605e-06,
+      "loss": 0.9908,
+      "step": 3226
+    },
+    {
+      "epoch": 0.3393850158412978,
+      "grad_norm": 2.055975701068765,
+      "learning_rate": 3.7565011189956597e-06,
+      "loss": 0.9898,
+      "step": 3227
+    },
+    {
+      "epoch": 0.33949018628314515,
+      "grad_norm": 2.4934374887026474,
+      "learning_rate": 3.7557793318056323e-06,
+      "loss": 0.9996,
+      "step": 3228
+    },
+    {
+      "epoch": 0.3395953567249924,
+      "grad_norm": 2.845049089637477,
+      "learning_rate": 3.755057404586455e-06,
+      "loss": 1.0016,
+      "step": 3229
+    },
+    {
+      "epoch": 0.33970052716683974,
+      "grad_norm": 3.0414746615738957,
+      "learning_rate": 3.7543353374186274e-06,
+      "loss": 1.0384,
+      "step": 3230
+    },
+    {
+      "epoch": 0.33980569760868706,
+      "grad_norm": 2.3969907781571793,
+      "learning_rate": 3.753613130382666e-06,
+      "loss": 1.0169,
+      "step": 3231
+    },
+    {
+      "epoch": 0.3399108680505344,
+      "grad_norm": 3.100902023873532,
+      "learning_rate": 3.752890783559102e-06,
+      "loss": 0.9903,
+      "step": 3232
+    },
+    {
+      "epoch": 0.3400160384923817,
+      "grad_norm": 2.515669648178997,
+      "learning_rate": 3.7521682970284827e-06,
+      "loss": 1.014,
+      "step": 3233
+    },
+    {
+      "epoch": 0.34012120893422904,
+      "grad_norm": 2.0868339076531215,
+      "learning_rate": 3.7514456708713717e-06,
+      "loss": 0.9637,
+      "step": 3234
+    },
+    {
+      "epoch": 0.34022637937607636,
+      "grad_norm": 2.837174645842647,
+      "learning_rate": 3.750722905168346e-06,
+      "loss": 0.9769,
+      "step": 3235
+    },
+    {
+      "epoch": 0.3403315498179237,
+      "grad_norm": 2.3872202244360547,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 1.003,
+      "step": 3236
+    },
+    {
+      "epoch": 0.340436720259771,
+      "grad_norm": 2.512997542175295,
+      "learning_rate": 3.7492769554469443e-06,
+      "loss": 0.9801,
+      "step": 3237
+    },
+    {
+      "epoch": 0.34054189070161833,
+      "grad_norm": 2.2402243033296507,
+      "learning_rate": 3.7485537715898036e-06,
+      "loss": 1.0344,
+      "step": 3238
+    },
+    {
+      "epoch": 0.3406470611434656,
+      "grad_norm": 1.788467788663251,
+      "learning_rate": 3.7478304485092174e-06,
+      "loss": 0.9945,
+      "step": 3239
+    },
+    {
+      "epoch": 0.3407522315853129,
+      "grad_norm": 2.438950326014985,
+      "learning_rate": 3.7471069862858446e-06,
+      "loss": 1.0212,
+      "step": 3240
+    },
+    {
+      "epoch": 0.34085740202716025,
+      "grad_norm": 2.625325390678325,
+      "learning_rate": 3.7463833850003532e-06,
+      "loss": 1.0471,
+      "step": 3241
+    },
+    {
+      "epoch": 0.3409625724690076,
+      "grad_norm": 2.298721915181398,
+      "learning_rate": 3.7456596447334324e-06,
+      "loss": 1.0091,
+      "step": 3242
+    },
+    {
+      "epoch": 0.3410677429108549,
+      "grad_norm": 2.393036922682323,
+      "learning_rate": 3.7449357655657854e-06,
+      "loss": 1.0367,
+      "step": 3243
+    },
+    {
+      "epoch": 0.3411729133527022,
+      "grad_norm": 2.439476214622686,
+      "learning_rate": 3.744211747578129e-06,
+      "loss": 1.0245,
+      "step": 3244
+    },
+    {
+      "epoch": 0.34127808379454955,
+      "grad_norm": 2.3404240819465536,
+      "learning_rate": 3.743487590851197e-06,
+      "loss": 0.9768,
+      "step": 3245
+    },
+    {
+      "epoch": 0.34138325423639687,
+      "grad_norm": 2.4310810439248467,
+      "learning_rate": 3.742763295465739e-06,
+      "loss": 1.0063,
+      "step": 3246
+    },
+    {
+      "epoch": 0.3414884246782442,
+      "grad_norm": 2.7737923645649105,
+      "learning_rate": 3.7420388615025194e-06,
+      "loss": 1.004,
+      "step": 3247
+    },
+    {
+      "epoch": 0.3415935951200915,
+      "grad_norm": 2.525162719248003,
+      "learning_rate": 3.7413142890423187e-06,
+      "loss": 1.0317,
+      "step": 3248
+    },
+    {
+      "epoch": 0.34169876556193884,
+      "grad_norm": 2.453175198423566,
+      "learning_rate": 3.740589578165932e-06,
+      "loss": 1.0053,
+      "step": 3249
+    },
+    {
+      "epoch": 0.3418039360037861,
+      "grad_norm": 2.412643165410365,
+      "learning_rate": 3.7398647289541703e-06,
+      "loss": 1.0517,
+      "step": 3250
+    },
+    {
+      "epoch": 0.34190910644563344,
+      "grad_norm": 1.8217035736085005,
+      "learning_rate": 3.7391397414878583e-06,
+      "loss": 0.9601,
+      "step": 3251
+    },
+    {
+      "epoch": 0.34201427688748076,
+      "grad_norm": 2.542322284034248,
+      "learning_rate": 3.73841461584784e-06,
+      "loss": 0.9978,
+      "step": 3252
+    },
+    {
+      "epoch": 0.3421194473293281,
+      "grad_norm": 1.9724168119288237,
+      "learning_rate": 3.7376893521149713e-06,
+      "loss": 0.9739,
+      "step": 3253
+    },
+    {
+      "epoch": 0.3422246177711754,
+      "grad_norm": 2.3405127012102325,
+      "learning_rate": 3.736963950370126e-06,
+      "loss": 1.0283,
+      "step": 3254
+    },
+    {
+      "epoch": 0.34232978821302273,
+      "grad_norm": 2.373486349876508,
+      "learning_rate": 3.7362384106941906e-06,
+      "loss": 0.9996,
+      "step": 3255
+    },
+    {
+      "epoch": 0.34243495865487006,
+      "grad_norm": 2.312609174057517,
+      "learning_rate": 3.7355127331680695e-06,
+      "loss": 1.0166,
+      "step": 3256
+    },
+    {
+      "epoch": 0.3425401290967174,
+      "grad_norm": 1.9211058932583842,
+      "learning_rate": 3.7347869178726804e-06,
+      "loss": 0.9955,
+      "step": 3257
+    },
+    {
+      "epoch": 0.3426452995385647,
+      "grad_norm": 6.881539412158539,
+      "learning_rate": 3.7340609648889575e-06,
+      "loss": 1.0193,
+      "step": 3258
+    },
+    {
+      "epoch": 0.34275046998041203,
+      "grad_norm": 2.0321978311810445,
+      "learning_rate": 3.7333348742978515e-06,
+      "loss": 0.9756,
+      "step": 3259
+    },
+    {
+      "epoch": 0.3428556404222593,
+      "grad_norm": 2.6613297784968886,
+      "learning_rate": 3.7326086461803247e-06,
+      "loss": 0.9983,
+      "step": 3260
+    },
+    {
+      "epoch": 0.3429608108641066,
+      "grad_norm": 2.0304428187193784,
+      "learning_rate": 3.7318822806173594e-06,
+      "loss": 1.0042,
+      "step": 3261
+    },
+    {
+      "epoch": 0.34306598130595395,
+      "grad_norm": 2.5112230342073842,
+      "learning_rate": 3.73115577768995e-06,
+      "loss": 0.9995,
+      "step": 3262
+    },
+    {
+      "epoch": 0.3431711517478013,
+      "grad_norm": 2.0313985757227706,
+      "learning_rate": 3.7304291374791085e-06,
+      "loss": 0.9986,
+      "step": 3263
+    },
+    {
+      "epoch": 0.3432763221896486,
+      "grad_norm": 2.5690827796961737,
+      "learning_rate": 3.7297023600658586e-06,
+      "loss": 0.9691,
+      "step": 3264
+    },
+    {
+      "epoch": 0.3433814926314959,
+      "grad_norm": 2.193452123288259,
+      "learning_rate": 3.7289754455312434e-06,
+      "loss": 0.9885,
+      "step": 3265
+    },
+    {
+      "epoch": 0.34348666307334325,
+      "grad_norm": 2.6601267632676575,
+      "learning_rate": 3.728248393956319e-06,
+      "loss": 1.0152,
+      "step": 3266
+    },
+    {
+      "epoch": 0.34359183351519057,
+      "grad_norm": 2.42078911309029,
+      "learning_rate": 3.7275212054221566e-06,
+      "loss": 1.0108,
+      "step": 3267
+    },
+    {
+      "epoch": 0.3436970039570379,
+      "grad_norm": 2.269243127837986,
+      "learning_rate": 3.7267938800098454e-06,
+      "loss": 0.9881,
+      "step": 3268
+    },
+    {
+      "epoch": 0.3438021743988852,
+      "grad_norm": 2.694411424356494,
+      "learning_rate": 3.7260664178004858e-06,
+      "loss": 1.0177,
+      "step": 3269
+    },
+    {
+      "epoch": 0.3439073448407325,
+      "grad_norm": 3.1436582660691896,
+      "learning_rate": 3.725338818875196e-06,
+      "loss": 1.0144,
+      "step": 3270
+    },
+    {
+      "epoch": 0.3440125152825798,
+      "grad_norm": 3.099436287839511,
+      "learning_rate": 3.724611083315109e-06,
+      "loss": 1.0101,
+      "step": 3271
+    },
+    {
+      "epoch": 0.34411768572442714,
+      "grad_norm": 2.7296113243028284,
+      "learning_rate": 3.7238832112013734e-06,
+      "loss": 1.023,
+      "step": 3272
+    },
+    {
+      "epoch": 0.34422285616627446,
+      "grad_norm": 2.2215641054092155,
+      "learning_rate": 3.723155202615153e-06,
+      "loss": 1.0077,
+      "step": 3273
+    },
+    {
+      "epoch": 0.3443280266081218,
+      "grad_norm": 2.627762150099357,
+      "learning_rate": 3.7224270576376243e-06,
+      "loss": 1.0031,
+      "step": 3274
+    },
+    {
+      "epoch": 0.3444331970499691,
+      "grad_norm": 2.368570833568216,
+      "learning_rate": 3.7216987763499835e-06,
+      "loss": 1.0204,
+      "step": 3275
+    },
+    {
+      "epoch": 0.34453836749181643,
+      "grad_norm": 2.1971410365900406,
+      "learning_rate": 3.7209703588334384e-06,
+      "loss": 1.0001,
+      "step": 3276
+    },
+    {
+      "epoch": 0.34464353793366376,
+      "grad_norm": 3.176673998132595,
+      "learning_rate": 3.7202418051692134e-06,
+      "loss": 1.0113,
+      "step": 3277
+    },
+    {
+      "epoch": 0.3447487083755111,
+      "grad_norm": 2.304560441480544,
+      "learning_rate": 3.719513115438548e-06,
+      "loss": 1.0286,
+      "step": 3278
+    },
+    {
+      "epoch": 0.3448538788173584,
+      "grad_norm": 2.3874420483075878,
+      "learning_rate": 3.7187842897226965e-06,
+      "loss": 1.0086,
+      "step": 3279
+    },
+    {
+      "epoch": 0.3449590492592057,
+      "grad_norm": 2.6394476203678696,
+      "learning_rate": 3.7180553281029295e-06,
+      "loss": 0.9717,
+      "step": 3280
+    },
+    {
+      "epoch": 0.345064219701053,
+      "grad_norm": 2.4270461247707034,
+      "learning_rate": 3.7173262306605307e-06,
+      "loss": 1.0398,
+      "step": 3281
+    },
+    {
+      "epoch": 0.3451693901429003,
+      "grad_norm": 3.626716130228977,
+      "learning_rate": 3.7165969974768005e-06,
+      "loss": 1.0027,
+      "step": 3282
+    },
+    {
+      "epoch": 0.34527456058474765,
+      "grad_norm": 3.345320093587124,
+      "learning_rate": 3.715867628633055e-06,
+      "loss": 0.9806,
+      "step": 3283
+    },
+    {
+      "epoch": 0.34537973102659497,
+      "grad_norm": 2.4384035873967025,
+      "learning_rate": 3.7151381242106232e-06,
+      "loss": 0.9817,
+      "step": 3284
+    },
+    {
+      "epoch": 0.3454849014684423,
+      "grad_norm": 2.6159320854024397,
+      "learning_rate": 3.7144084842908506e-06,
+      "loss": 0.9761,
+      "step": 3285
+    },
+    {
+      "epoch": 0.3455900719102896,
+      "grad_norm": 2.8847114804108696,
+      "learning_rate": 3.7136787089550986e-06,
+      "loss": 0.9864,
+      "step": 3286
+    },
+    {
+      "epoch": 0.34569524235213694,
+      "grad_norm": 3.029486595806022,
+      "learning_rate": 3.7129487982847422e-06,
+      "loss": 1.0003,
+      "step": 3287
+    },
+    {
+      "epoch": 0.34580041279398427,
+      "grad_norm": 3.109537564490848,
+      "learning_rate": 3.7122187523611724e-06,
+      "loss": 1.009,
+      "step": 3288
+    },
+    {
+      "epoch": 0.3459055832358316,
+      "grad_norm": 2.9156800483209326,
+      "learning_rate": 3.711488571265795e-06,
+      "loss": 0.9898,
+      "step": 3289
+    },
+    {
+      "epoch": 0.34601075367767886,
+      "grad_norm": 2.711869560964191,
+      "learning_rate": 3.71075825508003e-06,
+      "loss": 0.9939,
+      "step": 3290
+    },
+    {
+      "epoch": 0.3461159241195262,
+      "grad_norm": 1.4908550438638197,
+      "learning_rate": 3.7100278038853157e-06,
+      "loss": 1.0175,
+      "step": 3291
+    },
+    {
+      "epoch": 0.3462210945613735,
+      "grad_norm": 2.7237754282190534,
+      "learning_rate": 3.7092972177630998e-06,
+      "loss": 0.999,
+      "step": 3292
+    },
+    {
+      "epoch": 0.34632626500322083,
+      "grad_norm": 2.1709243952269577,
+      "learning_rate": 3.708566496794851e-06,
+      "loss": 0.9716,
+      "step": 3293
+    },
+    {
+      "epoch": 0.34643143544506816,
+      "grad_norm": 2.316753104081801,
+      "learning_rate": 3.7078356410620484e-06,
+      "loss": 1.0317,
+      "step": 3294
+    },
+    {
+      "epoch": 0.3465366058869155,
+      "grad_norm": 2.3542621040463567,
+      "learning_rate": 3.7071046506461893e-06,
+      "loss": 1.0029,
+      "step": 3295
+    },
+    {
+      "epoch": 0.3466417763287628,
+      "grad_norm": 1.9708389723051476,
+      "learning_rate": 3.7063735256287854e-06,
+      "loss": 0.9978,
+      "step": 3296
+    },
+    {
+      "epoch": 0.34674694677061013,
+      "grad_norm": 2.8033237545972414,
+      "learning_rate": 3.705642266091361e-06,
+      "loss": 0.9875,
+      "step": 3297
+    },
+    {
+      "epoch": 0.34685211721245746,
+      "grad_norm": 2.6769139841372196,
+      "learning_rate": 3.7049108721154586e-06,
+      "loss": 1.0103,
+      "step": 3298
+    },
+    {
+      "epoch": 0.3469572876543048,
+      "grad_norm": 2.2403444086333053,
+      "learning_rate": 3.7041793437826336e-06,
+      "loss": 1.0481,
+      "step": 3299
+    },
+    {
+      "epoch": 0.34706245809615205,
+      "grad_norm": 2.260213376943109,
+      "learning_rate": 3.703447681174458e-06,
+      "loss": 1.0359,
+      "step": 3300
+    },
+    {
+      "epoch": 0.3471676285379994,
+      "grad_norm": 2.367948574095755,
+      "learning_rate": 3.7027158843725164e-06,
+      "loss": 1.0121,
+      "step": 3301
+    },
+    {
+      "epoch": 0.3472727989798467,
+      "grad_norm": 2.250568883640813,
+      "learning_rate": 3.7019839534584113e-06,
+      "loss": 0.9927,
+      "step": 3302
+    },
+    {
+      "epoch": 0.347377969421694,
+      "grad_norm": 3.1053336133122396,
+      "learning_rate": 3.7012518885137572e-06,
+      "loss": 1.0122,
+      "step": 3303
+    },
+    {
+      "epoch": 0.34748313986354135,
+      "grad_norm": 2.1013260166293732,
+      "learning_rate": 3.7005196896201867e-06,
+      "loss": 1.0169,
+      "step": 3304
+    },
+    {
+      "epoch": 0.34758831030538867,
+      "grad_norm": 1.959095033611609,
+      "learning_rate": 3.6997873568593446e-06,
+      "loss": 1.0015,
+      "step": 3305
+    },
+    {
+      "epoch": 0.347693480747236,
+      "grad_norm": 2.511723319973884,
+      "learning_rate": 3.699054890312892e-06,
+      "loss": 0.9941,
+      "step": 3306
+    },
+    {
+      "epoch": 0.3477986511890833,
+      "grad_norm": 2.5395710231112902,
+      "learning_rate": 3.6983222900625047e-06,
+      "loss": 0.9469,
+      "step": 3307
+    },
+    {
+      "epoch": 0.34790382163093064,
+      "grad_norm": 2.7171955825527783,
+      "learning_rate": 3.697589556189873e-06,
+      "loss": 1.0311,
+      "step": 3308
+    },
+    {
+      "epoch": 0.34800899207277797,
+      "grad_norm": 2.5690913311688997,
+      "learning_rate": 3.6968566887767027e-06,
+      "loss": 1.0443,
+      "step": 3309
+    },
+    {
+      "epoch": 0.3481141625146253,
+      "grad_norm": 3.0893677787025147,
+      "learning_rate": 3.6961236879047135e-06,
+      "loss": 1.0376,
+      "step": 3310
+    },
+    {
+      "epoch": 0.34821933295647256,
+      "grad_norm": 2.3846973615171243,
+      "learning_rate": 3.695390553655642e-06,
+      "loss": 1.0051,
+      "step": 3311
+    },
+    {
+      "epoch": 0.3483245033983199,
+      "grad_norm": 2.6749794579639006,
+      "learning_rate": 3.6946572861112373e-06,
+      "loss": 0.9813,
+      "step": 3312
+    },
+    {
+      "epoch": 0.3484296738401672,
+      "grad_norm": 2.07377234986436,
+      "learning_rate": 3.6939238853532657e-06,
+      "loss": 0.9667,
+      "step": 3313
+    },
+    {
+      "epoch": 0.34853484428201453,
+      "grad_norm": 3.060207599086186,
+      "learning_rate": 3.693190351463505e-06,
+      "loss": 1.0142,
+      "step": 3314
+    },
+    {
+      "epoch": 0.34864001472386186,
+      "grad_norm": 2.936425505722273,
+      "learning_rate": 3.6924566845237508e-06,
+      "loss": 0.9837,
+      "step": 3315
+    },
+    {
+      "epoch": 0.3487451851657092,
+      "grad_norm": 3.587810211037129,
+      "learning_rate": 3.691722884615814e-06,
+      "loss": 1.0268,
+      "step": 3316
+    },
+    {
+      "epoch": 0.3488503556075565,
+      "grad_norm": 2.1046353489321077,
+      "learning_rate": 3.690988951821517e-06,
+      "loss": 1.0083,
+      "step": 3317
+    },
+    {
+      "epoch": 0.34895552604940383,
+      "grad_norm": 2.5495061254512805,
+      "learning_rate": 3.6902548862227007e-06,
+      "loss": 1.0373,
+      "step": 3318
+    },
+    {
+      "epoch": 0.34906069649125115,
+      "grad_norm": 3.116414932564101,
+      "learning_rate": 3.689520687901218e-06,
+      "loss": 1.039,
+      "step": 3319
+    },
+    {
+      "epoch": 0.3491658669330985,
+      "grad_norm": 2.0740724861509197,
+      "learning_rate": 3.6887863569389388e-06,
+      "loss": 1.0328,
+      "step": 3320
+    },
+    {
+      "epoch": 0.34927103737494575,
+      "grad_norm": 3.0783194643210865,
+      "learning_rate": 3.688051893417745e-06,
+      "loss": 1.0286,
+      "step": 3321
+    },
+    {
+      "epoch": 0.34937620781679307,
+      "grad_norm": 2.66641944778877,
+      "learning_rate": 3.687317297419536e-06,
+      "loss": 1.037,
+      "step": 3322
+    },
+    {
+      "epoch": 0.3494813782586404,
+      "grad_norm": 2.7985552282879804,
+      "learning_rate": 3.6865825690262256e-06,
+      "loss": 1.0158,
+      "step": 3323
+    },
+    {
+      "epoch": 0.3495865487004877,
+      "grad_norm": 2.3794909564668623,
+      "learning_rate": 3.6858477083197403e-06,
+      "loss": 1.0267,
+      "step": 3324
+    },
+    {
+      "epoch": 0.34969171914233504,
+      "grad_norm": 2.6094670978251115,
+      "learning_rate": 3.6851127153820243e-06,
+      "loss": 1.0293,
+      "step": 3325
+    },
+    {
+      "epoch": 0.34979688958418237,
+      "grad_norm": 3.154815924297272,
+      "learning_rate": 3.684377590295034e-06,
+      "loss": 0.9958,
+      "step": 3326
+    },
+    {
+      "epoch": 0.3499020600260297,
+      "grad_norm": 2.383540391351354,
+      "learning_rate": 3.6836423331407416e-06,
+      "loss": 1.0109,
+      "step": 3327
+    },
+    {
+      "epoch": 0.350007230467877,
+      "grad_norm": 1.9040090397223854,
+      "learning_rate": 3.6829069440011343e-06,
+      "loss": 0.9934,
+      "step": 3328
+    },
+    {
+      "epoch": 0.35011240090972434,
+      "grad_norm": 2.779759189993991,
+      "learning_rate": 3.682171422958214e-06,
+      "loss": 1.007,
+      "step": 3329
+    },
+    {
+      "epoch": 0.35021757135157167,
+      "grad_norm": 3.2401364529589824,
+      "learning_rate": 3.681435770093996e-06,
+      "loss": 1.0305,
+      "step": 3330
+    },
+    {
+      "epoch": 0.35032274179341893,
+      "grad_norm": 3.727910090979757,
+      "learning_rate": 3.6806999854905117e-06,
+      "loss": 1.0082,
+      "step": 3331
+    },
+    {
+      "epoch": 0.35042791223526626,
+      "grad_norm": 2.5069421827992153,
+      "learning_rate": 3.6799640692298076e-06,
+      "loss": 0.9588,
+      "step": 3332
+    },
+    {
+      "epoch": 0.3505330826771136,
+      "grad_norm": 3.2086550817958877,
+      "learning_rate": 3.6792280213939433e-06,
+      "loss": 1.0142,
+      "step": 3333
+    },
+    {
+      "epoch": 0.3506382531189609,
+      "grad_norm": 1.736538961113674,
+      "learning_rate": 3.6784918420649952e-06,
+      "loss": 0.9666,
+      "step": 3334
+    },
+    {
+      "epoch": 0.35074342356080823,
+      "grad_norm": 1.8790958449036552,
+      "learning_rate": 3.6777555313250505e-06,
+      "loss": 1.0285,
+      "step": 3335
+    },
+    {
+      "epoch": 0.35084859400265556,
+      "grad_norm": 2.3400912078637677,
+      "learning_rate": 3.6770190892562154e-06,
+      "loss": 1.0094,
+      "step": 3336
+    },
+    {
+      "epoch": 0.3509537644445029,
+      "grad_norm": 1.9690870053528573,
+      "learning_rate": 3.676282515940608e-06,
+      "loss": 0.9816,
+      "step": 3337
+    },
+    {
+      "epoch": 0.3510589348863502,
+      "grad_norm": 2.517883459106263,
+      "learning_rate": 3.675545811460362e-06,
+      "loss": 1.0008,
+      "step": 3338
+    },
+    {
+      "epoch": 0.35116410532819753,
+      "grad_norm": 4.054324173962527,
+      "learning_rate": 3.6748089758976267e-06,
+      "loss": 0.9958,
+      "step": 3339
+    },
+    {
+      "epoch": 0.35126927577004485,
+      "grad_norm": 3.008203369375549,
+      "learning_rate": 3.6740720093345646e-06,
+      "loss": 0.9959,
+      "step": 3340
+    },
+    {
+      "epoch": 0.3513744462118921,
+      "grad_norm": 2.2054094909667543,
+      "learning_rate": 3.6733349118533524e-06,
+      "loss": 0.9362,
+      "step": 3341
+    },
+    {
+      "epoch": 0.35147961665373945,
+      "grad_norm": 2.147723777459706,
+      "learning_rate": 3.6725976835361832e-06,
+      "loss": 1.0227,
+      "step": 3342
+    },
+    {
+      "epoch": 0.35158478709558677,
+      "grad_norm": 3.4189072148500466,
+      "learning_rate": 3.6718603244652634e-06,
+      "loss": 1.0129,
+      "step": 3343
+    },
+    {
+      "epoch": 0.3516899575374341,
+      "grad_norm": 2.39516829951863,
+      "learning_rate": 3.671122834722814e-06,
+      "loss": 1.0158,
+      "step": 3344
+    },
+    {
+      "epoch": 0.3517951279792814,
+      "grad_norm": 2.9471635187278786,
+      "learning_rate": 3.67038521439107e-06,
+      "loss": 0.9988,
+      "step": 3345
+    },
+    {
+      "epoch": 0.35190029842112874,
+      "grad_norm": 2.7956147749271136,
+      "learning_rate": 3.669647463552284e-06,
+      "loss": 1.0219,
+      "step": 3346
+    },
+    {
+      "epoch": 0.35200546886297607,
+      "grad_norm": 2.549496541093747,
+      "learning_rate": 3.6689095822887188e-06,
+      "loss": 1.0342,
+      "step": 3347
+    },
+    {
+      "epoch": 0.3521106393048234,
+      "grad_norm": 3.3421799590829755,
+      "learning_rate": 3.6681715706826555e-06,
+      "loss": 0.9693,
+      "step": 3348
+    },
+    {
+      "epoch": 0.3522158097466707,
+      "grad_norm": 2.7877765286056575,
+      "learning_rate": 3.6674334288163872e-06,
+      "loss": 0.9698,
+      "step": 3349
+    },
+    {
+      "epoch": 0.35232098018851804,
+      "grad_norm": 2.850956387184573,
+      "learning_rate": 3.666695156772222e-06,
+      "loss": 1.0254,
+      "step": 3350
+    },
+    {
+      "epoch": 0.3524261506303653,
+      "grad_norm": 2.5846160246925156,
+      "learning_rate": 3.6659567546324843e-06,
+      "loss": 0.9714,
+      "step": 3351
+    },
+    {
+      "epoch": 0.35253132107221263,
+      "grad_norm": 3.0980328608661667,
+      "learning_rate": 3.6652182224795108e-06,
+      "loss": 1.0372,
+      "step": 3352
+    },
+    {
+      "epoch": 0.35263649151405996,
+      "grad_norm": 2.2612646017522002,
+      "learning_rate": 3.6644795603956535e-06,
+      "loss": 0.978,
+      "step": 3353
+    },
+    {
+      "epoch": 0.3527416619559073,
+      "grad_norm": 2.549186265914262,
+      "learning_rate": 3.663740768463279e-06,
+      "loss": 1.0212,
+      "step": 3354
+    },
+    {
+      "epoch": 0.3528468323977546,
+      "grad_norm": 2.747876314122542,
+      "learning_rate": 3.663001846764769e-06,
+      "loss": 1.0214,
+      "step": 3355
+    },
+    {
+      "epoch": 0.35295200283960193,
+      "grad_norm": 2.235824760868568,
+      "learning_rate": 3.6622627953825187e-06,
+      "loss": 0.9917,
+      "step": 3356
+    },
+    {
+      "epoch": 0.35305717328144925,
+      "grad_norm": 2.4010747810798603,
+      "learning_rate": 3.661523614398938e-06,
+      "loss": 0.9649,
+      "step": 3357
+    },
+    {
+      "epoch": 0.3531623437232966,
+      "grad_norm": 2.828908496537335,
+      "learning_rate": 3.6607843038964515e-06,
+      "loss": 0.9981,
+      "step": 3358
+    },
+    {
+      "epoch": 0.3532675141651439,
+      "grad_norm": 2.555400473058865,
+      "learning_rate": 3.660044863957497e-06,
+      "loss": 1.0143,
+      "step": 3359
+    },
+    {
+      "epoch": 0.3533726846069912,
+      "grad_norm": 2.617313449912603,
+      "learning_rate": 3.6593052946645293e-06,
+      "loss": 1.0176,
+      "step": 3360
+    },
+    {
+      "epoch": 0.3534778550488385,
+      "grad_norm": 1.6229208590309265,
+      "learning_rate": 3.6585655961000144e-06,
+      "loss": 0.9889,
+      "step": 3361
+    },
+    {
+      "epoch": 0.3535830254906858,
+      "grad_norm": 2.6242669270631156,
+      "learning_rate": 3.6578257683464363e-06,
+      "loss": 0.9977,
+      "step": 3362
+    },
+    {
+      "epoch": 0.35368819593253314,
+      "grad_norm": 2.730302244599849,
+      "learning_rate": 3.6570858114862905e-06,
+      "loss": 0.9535,
+      "step": 3363
+    },
+    {
+      "epoch": 0.35379336637438047,
+      "grad_norm": 2.5049239813451427,
+      "learning_rate": 3.656345725602089e-06,
+      "loss": 1.0363,
+      "step": 3364
+    },
+    {
+      "epoch": 0.3538985368162278,
+      "grad_norm": 2.3239114473104454,
+      "learning_rate": 3.6556055107763553e-06,
+      "loss": 1.003,
+      "step": 3365
+    },
+    {
+      "epoch": 0.3540037072580751,
+      "grad_norm": 3.122736859914737,
+      "learning_rate": 3.6548651670916302e-06,
+      "loss": 1.013,
+      "step": 3366
+    },
+    {
+      "epoch": 0.35410887769992244,
+      "grad_norm": 3.046846196486878,
+      "learning_rate": 3.654124694630468e-06,
+      "loss": 1.0142,
+      "step": 3367
+    },
+    {
+      "epoch": 0.35421404814176977,
+      "grad_norm": 2.636989697235664,
+      "learning_rate": 3.653384093475436e-06,
+      "loss": 1.0422,
+      "step": 3368
+    },
+    {
+      "epoch": 0.3543192185836171,
+      "grad_norm": 2.61843008906023,
+      "learning_rate": 3.6526433637091193e-06,
+      "loss": 1.037,
+      "step": 3369
+    },
+    {
+      "epoch": 0.3544243890254644,
+      "grad_norm": 1.915867272619924,
+      "learning_rate": 3.651902505414112e-06,
+      "loss": 0.9925,
+      "step": 3370
+    },
+    {
+      "epoch": 0.35452955946731174,
+      "grad_norm": 2.9271823410792797,
+      "learning_rate": 3.651161518673028e-06,
+      "loss": 1.0368,
+      "step": 3371
+    },
+    {
+      "epoch": 0.354634729909159,
+      "grad_norm": 2.409589901528896,
+      "learning_rate": 3.6504204035684915e-06,
+      "loss": 1.0037,
+      "step": 3372
+    },
+    {
+      "epoch": 0.35473990035100633,
+      "grad_norm": 2.1387568533332395,
+      "learning_rate": 3.649679160183144e-06,
+      "loss": 1.0186,
+      "step": 3373
+    },
+    {
+      "epoch": 0.35484507079285366,
+      "grad_norm": 2.4112770744675314,
+      "learning_rate": 3.6489377885996385e-06,
+      "loss": 1.0667,
+      "step": 3374
+    },
+    {
+      "epoch": 0.354950241234701,
+      "grad_norm": 2.4610952866669096,
+      "learning_rate": 3.648196288900644e-06,
+      "loss": 1.0212,
+      "step": 3375
+    },
+    {
+      "epoch": 0.3550554116765483,
+      "grad_norm": 3.23300202577758,
+      "learning_rate": 3.6474546611688446e-06,
+      "loss": 1.0027,
+      "step": 3376
+    },
+    {
+      "epoch": 0.35516058211839563,
+      "grad_norm": 2.121250895820615,
+      "learning_rate": 3.646712905486936e-06,
+      "loss": 1.0263,
+      "step": 3377
+    },
+    {
+      "epoch": 0.35526575256024295,
+      "grad_norm": 3.2579471158022195,
+      "learning_rate": 3.6459710219376317e-06,
+      "loss": 1.019,
+      "step": 3378
+    },
+    {
+      "epoch": 0.3553709230020903,
+      "grad_norm": 2.283154927593308,
+      "learning_rate": 3.645229010603655e-06,
+      "loss": 1.0052,
+      "step": 3379
+    },
+    {
+      "epoch": 0.3554760934439376,
+      "grad_norm": 2.844999630621272,
+      "learning_rate": 3.6444868715677475e-06,
+      "loss": 0.9872,
+      "step": 3380
+    },
+    {
+      "epoch": 0.3555812638857849,
+      "grad_norm": 2.110778090946952,
+      "learning_rate": 3.6437446049126636e-06,
+      "loss": 0.9785,
+      "step": 3381
+    },
+    {
+      "epoch": 0.3556864343276322,
+      "grad_norm": 3.104483713885235,
+      "learning_rate": 3.6430022107211705e-06,
+      "loss": 1.058,
+      "step": 3382
+    },
+    {
+      "epoch": 0.3557916047694795,
+      "grad_norm": 3.698736872503185,
+      "learning_rate": 3.642259689076052e-06,
+      "loss": 0.9923,
+      "step": 3383
+    },
+    {
+      "epoch": 0.35589677521132684,
+      "grad_norm": 2.6468434275616657,
+      "learning_rate": 3.641517040060105e-06,
+      "loss": 1.0115,
+      "step": 3384
+    },
+    {
+      "epoch": 0.35600194565317417,
+      "grad_norm": 3.429273915761346,
+      "learning_rate": 3.6407742637561407e-06,
+      "loss": 0.9963,
+      "step": 3385
+    },
+    {
+      "epoch": 0.3561071160950215,
+      "grad_norm": 2.2796812206143393,
+      "learning_rate": 3.6400313602469835e-06,
+      "loss": 0.9957,
+      "step": 3386
+    },
+    {
+      "epoch": 0.3562122865368688,
+      "grad_norm": 2.2684651957658435,
+      "learning_rate": 3.639288329615474e-06,
+      "loss": 0.9733,
+      "step": 3387
+    },
+    {
+      "epoch": 0.35631745697871614,
+      "grad_norm": 2.7538582807610092,
+      "learning_rate": 3.6385451719444655e-06,
+      "loss": 0.9873,
+      "step": 3388
+    },
+    {
+      "epoch": 0.35642262742056346,
+      "grad_norm": 2.1667916415130244,
+      "learning_rate": 3.6378018873168254e-06,
+      "loss": 0.9706,
+      "step": 3389
+    },
+    {
+      "epoch": 0.3565277978624108,
+      "grad_norm": 2.6101805735697092,
+      "learning_rate": 3.6370584758154366e-06,
+      "loss": 1.0247,
+      "step": 3390
+    },
+    {
+      "epoch": 0.3566329683042581,
+      "grad_norm": 2.6339328347622066,
+      "learning_rate": 3.6363149375231937e-06,
+      "loss": 1.0401,
+      "step": 3391
+    },
+    {
+      "epoch": 0.3567381387461054,
+      "grad_norm": 2.3753106980243843,
+      "learning_rate": 3.6355712725230093e-06,
+      "loss": 0.9831,
+      "step": 3392
+    },
+    {
+      "epoch": 0.3568433091879527,
+      "grad_norm": 4.0920704573771305,
+      "learning_rate": 3.6348274808978063e-06,
+      "loss": 1.0566,
+      "step": 3393
+    },
+    {
+      "epoch": 0.35694847962980003,
+      "grad_norm": 2.1861120277976256,
+      "learning_rate": 3.634083562730523e-06,
+      "loss": 1.0138,
+      "step": 3394
+    },
+    {
+      "epoch": 0.35705365007164735,
+      "grad_norm": 2.4463244703652367,
+      "learning_rate": 3.633339518104113e-06,
+      "loss": 0.9825,
+      "step": 3395
+    },
+    {
+      "epoch": 0.3571588205134947,
+      "grad_norm": 2.2656990738852016,
+      "learning_rate": 3.632595347101543e-06,
+      "loss": 1.0078,
+      "step": 3396
+    },
+    {
+      "epoch": 0.357263990955342,
+      "grad_norm": 2.8104237008735944,
+      "learning_rate": 3.631851049805793e-06,
+      "loss": 1.0039,
+      "step": 3397
+    },
+    {
+      "epoch": 0.3573691613971893,
+      "grad_norm": 1.6073481837060561,
+      "learning_rate": 3.6311066262998585e-06,
+      "loss": 0.9729,
+      "step": 3398
+    },
+    {
+      "epoch": 0.35747433183903665,
+      "grad_norm": 2.1246171429605836,
+      "learning_rate": 3.630362076666748e-06,
+      "loss": 0.9902,
+      "step": 3399
+    },
+    {
+      "epoch": 0.357579502280884,
+      "grad_norm": 2.0248681289481048,
+      "learning_rate": 3.6296174009894856e-06,
+      "loss": 1.0255,
+      "step": 3400
+    },
+    {
+      "epoch": 0.3576846727227313,
+      "grad_norm": 2.938773915640707,
+      "learning_rate": 3.628872599351108e-06,
+      "loss": 1.0085,
+      "step": 3401
+    },
+    {
+      "epoch": 0.35778984316457857,
+      "grad_norm": 2.8664540166582944,
+      "learning_rate": 3.628127671834665e-06,
+      "loss": 1.0408,
+      "step": 3402
+    },
+    {
+      "epoch": 0.3578950136064259,
+      "grad_norm": 1.626968288052293,
+      "learning_rate": 3.627382618523224e-06,
+      "loss": 0.9455,
+      "step": 3403
+    },
+    {
+      "epoch": 0.3580001840482732,
+      "grad_norm": 2.748931967001064,
+      "learning_rate": 3.626637439499864e-06,
+      "loss": 1.0007,
+      "step": 3404
+    },
+    {
+      "epoch": 0.35810535449012054,
+      "grad_norm": 2.474980683402302,
+      "learning_rate": 3.6258921348476754e-06,
+      "loss": 0.9864,
+      "step": 3405
+    },
+    {
+      "epoch": 0.35821052493196787,
+      "grad_norm": 2.7019560736927595,
+      "learning_rate": 3.625146704649769e-06,
+      "loss": 0.9493,
+      "step": 3406
+    },
+    {
+      "epoch": 0.3583156953738152,
+      "grad_norm": 2.901227686551035,
+      "learning_rate": 3.6244011489892645e-06,
+      "loss": 1.0309,
+      "step": 3407
+    },
+    {
+      "epoch": 0.3584208658156625,
+      "grad_norm": 2.4436082587640886,
+      "learning_rate": 3.623655467949297e-06,
+      "loss": 1.0221,
+      "step": 3408
+    },
+    {
+      "epoch": 0.35852603625750984,
+      "grad_norm": 3.134715394017839,
+      "learning_rate": 3.6229096616130154e-06,
+      "loss": 0.9888,
+      "step": 3409
+    },
+    {
+      "epoch": 0.35863120669935716,
+      "grad_norm": 2.396056272037577,
+      "learning_rate": 3.6221637300635844e-06,
+      "loss": 1.056,
+      "step": 3410
+    },
+    {
+      "epoch": 0.3587363771412045,
+      "grad_norm": 2.0197687516513647,
+      "learning_rate": 3.6214176733841792e-06,
+      "loss": 0.9827,
+      "step": 3411
+    },
+    {
+      "epoch": 0.35884154758305176,
+      "grad_norm": 2.934753315186612,
+      "learning_rate": 3.6206714916579925e-06,
+      "loss": 0.9909,
+      "step": 3412
+    },
+    {
+      "epoch": 0.3589467180248991,
+      "grad_norm": 2.136669283224677,
+      "learning_rate": 3.619925184968229e-06,
+      "loss": 1.0157,
+      "step": 3413
+    },
+    {
+      "epoch": 0.3590518884667464,
+      "grad_norm": 2.5314749644598344,
+      "learning_rate": 3.6191787533981075e-06,
+      "loss": 1.0389,
+      "step": 3414
+    },
+    {
+      "epoch": 0.35915705890859373,
+      "grad_norm": 2.4095375951453004,
+      "learning_rate": 3.618432197030861e-06,
+      "loss": 1.0118,
+      "step": 3415
+    },
+    {
+      "epoch": 0.35926222935044105,
+      "grad_norm": 2.325532442349783,
+      "learning_rate": 3.617685515949736e-06,
+      "loss": 1.0057,
+      "step": 3416
+    },
+    {
+      "epoch": 0.3593673997922884,
+      "grad_norm": 2.4488118992083354,
+      "learning_rate": 3.6169387102379935e-06,
+      "loss": 1.0158,
+      "step": 3417
+    },
+    {
+      "epoch": 0.3594725702341357,
+      "grad_norm": 2.624600631200174,
+      "learning_rate": 3.6161917799789076e-06,
+      "loss": 1.0063,
+      "step": 3418
+    },
+    {
+      "epoch": 0.359577740675983,
+      "grad_norm": 2.643328190762798,
+      "learning_rate": 3.615444725255768e-06,
+      "loss": 0.9978,
+      "step": 3419
+    },
+    {
+      "epoch": 0.35968291111783035,
+      "grad_norm": 2.407743326176636,
+      "learning_rate": 3.6146975461518765e-06,
+      "loss": 0.981,
+      "step": 3420
+    },
+    {
+      "epoch": 0.3597880815596777,
+      "grad_norm": 2.311733214584104,
+      "learning_rate": 3.613950242750549e-06,
+      "loss": 0.9686,
+      "step": 3421
+    },
+    {
+      "epoch": 0.35989325200152494,
+      "grad_norm": 2.1921054999722123,
+      "learning_rate": 3.6132028151351163e-06,
+      "loss": 0.9877,
+      "step": 3422
+    },
+    {
+      "epoch": 0.35999842244337227,
+      "grad_norm": 1.9973124149153578,
+      "learning_rate": 3.6124552633889217e-06,
+      "loss": 0.9562,
+      "step": 3423
+    },
+    {
+      "epoch": 0.3601035928852196,
+      "grad_norm": 2.2987560515883567,
+      "learning_rate": 3.6117075875953233e-06,
+      "loss": 1.0459,
+      "step": 3424
+    },
+    {
+      "epoch": 0.3602087633270669,
+      "grad_norm": 3.20659343785868,
+      "learning_rate": 3.610959787837693e-06,
+      "loss": 1.0328,
+      "step": 3425
+    },
+    {
+      "epoch": 0.36031393376891424,
+      "grad_norm": 2.7744645057080355,
+      "learning_rate": 3.6102118641994166e-06,
+      "loss": 0.9981,
+      "step": 3426
+    },
+    {
+      "epoch": 0.36041910421076157,
+      "grad_norm": 1.4642559942958495,
+      "learning_rate": 3.6094638167638924e-06,
+      "loss": 0.9947,
+      "step": 3427
+    },
+    {
+      "epoch": 0.3605242746526089,
+      "grad_norm": 2.386402726019697,
+      "learning_rate": 3.608715645614534e-06,
+      "loss": 1.0195,
+      "step": 3428
+    },
+    {
+      "epoch": 0.3606294450944562,
+      "grad_norm": 2.6634628273386465,
+      "learning_rate": 3.607967350834769e-06,
+      "loss": 1.0473,
+      "step": 3429
+    },
+    {
+      "epoch": 0.36073461553630354,
+      "grad_norm": 3.495355937233298,
+      "learning_rate": 3.6072189325080364e-06,
+      "loss": 1.0102,
+      "step": 3430
+    },
+    {
+      "epoch": 0.36083978597815086,
+      "grad_norm": 1.9319582065248462,
+      "learning_rate": 3.6064703907177923e-06,
+      "loss": 0.9659,
+      "step": 3431
+    },
+    {
+      "epoch": 0.3609449564199982,
+      "grad_norm": 2.3818832688536973,
+      "learning_rate": 3.6057217255475034e-06,
+      "loss": 0.9997,
+      "step": 3432
+    },
+    {
+      "epoch": 0.36105012686184546,
+      "grad_norm": 1.7493814880171161,
+      "learning_rate": 3.6049729370806534e-06,
+      "loss": 1.0365,
+      "step": 3433
+    },
+    {
+      "epoch": 0.3611552973036928,
+      "grad_norm": 2.8422167028984644,
+      "learning_rate": 3.6042240254007367e-06,
+      "loss": 0.9969,
+      "step": 3434
+    },
+    {
+      "epoch": 0.3612604677455401,
+      "grad_norm": 2.27792880127726,
+      "learning_rate": 3.6034749905912637e-06,
+      "loss": 1.0197,
+      "step": 3435
+    },
+    {
+      "epoch": 0.36136563818738743,
+      "grad_norm": 3.38034146624978,
+      "learning_rate": 3.6027258327357573e-06,
+      "loss": 1.0358,
+      "step": 3436
+    },
+    {
+      "epoch": 0.36147080862923475,
+      "grad_norm": 1.915172769306197,
+      "learning_rate": 3.6019765519177536e-06,
+      "loss": 1.0184,
+      "step": 3437
+    },
+    {
+      "epoch": 0.3615759790710821,
+      "grad_norm": 3.7012827312157732,
+      "learning_rate": 3.6012271482208043e-06,
+      "loss": 1.0315,
+      "step": 3438
+    },
+    {
+      "epoch": 0.3616811495129294,
+      "grad_norm": 2.8348305338817665,
+      "learning_rate": 3.600477621728473e-06,
+      "loss": 1.0226,
+      "step": 3439
+    },
+    {
+      "epoch": 0.3617863199547767,
+      "grad_norm": 2.226936028972617,
+      "learning_rate": 3.5997279725243382e-06,
+      "loss": 0.9864,
+      "step": 3440
+    },
+    {
+      "epoch": 0.36189149039662405,
+      "grad_norm": 2.8560998684935988,
+      "learning_rate": 3.5989782006919915e-06,
+      "loss": 1.0336,
+      "step": 3441
+    },
+    {
+      "epoch": 0.3619966608384714,
+      "grad_norm": 2.229618061905727,
+      "learning_rate": 3.5982283063150388e-06,
+      "loss": 0.9575,
+      "step": 3442
+    },
+    {
+      "epoch": 0.36210183128031864,
+      "grad_norm": 1.5501631952100456,
+      "learning_rate": 3.5974782894770983e-06,
+      "loss": 1.0172,
+      "step": 3443
+    },
+    {
+      "epoch": 0.36220700172216597,
+      "grad_norm": 2.803536569294912,
+      "learning_rate": 3.5967281502618035e-06,
+      "loss": 1.0584,
+      "step": 3444
+    },
+    {
+      "epoch": 0.3623121721640133,
+      "grad_norm": 2.1481647342373957,
+      "learning_rate": 3.5959778887527995e-06,
+      "loss": 1.0145,
+      "step": 3445
+    },
+    {
+      "epoch": 0.3624173426058606,
+      "grad_norm": 3.0517742861625896,
+      "learning_rate": 3.595227505033747e-06,
+      "loss": 1.0183,
+      "step": 3446
+    },
+    {
+      "epoch": 0.36252251304770794,
+      "grad_norm": 2.6479450779631217,
+      "learning_rate": 3.5944769991883197e-06,
+      "loss": 0.9737,
+      "step": 3447
+    },
+    {
+      "epoch": 0.36262768348955526,
+      "grad_norm": 2.5993294962929605,
+      "learning_rate": 3.5937263713002046e-06,
+      "loss": 1.0112,
+      "step": 3448
+    },
+    {
+      "epoch": 0.3627328539314026,
+      "grad_norm": 2.2975185450286353,
+      "learning_rate": 3.5929756214531035e-06,
+      "loss": 1.0724,
+      "step": 3449
+    },
+    {
+      "epoch": 0.3628380243732499,
+      "grad_norm": 3.128724604849356,
+      "learning_rate": 3.59222474973073e-06,
+      "loss": 1.0058,
+      "step": 3450
+    },
+    {
+      "epoch": 0.36294319481509724,
+      "grad_norm": 2.551432018630201,
+      "learning_rate": 3.5914737562168116e-06,
+      "loss": 1.0239,
+      "step": 3451
+    },
+    {
+      "epoch": 0.36304836525694456,
+      "grad_norm": 2.505173837642558,
+      "learning_rate": 3.590722640995091e-06,
+      "loss": 0.9775,
+      "step": 3452
+    },
+    {
+      "epoch": 0.36315353569879183,
+      "grad_norm": 2.5187594676149434,
+      "learning_rate": 3.589971404149323e-06,
+      "loss": 1.0362,
+      "step": 3453
+    },
+    {
+      "epoch": 0.36325870614063915,
+      "grad_norm": 2.239747984102128,
+      "learning_rate": 3.589220045763276e-06,
+      "loss": 1.0336,
+      "step": 3454
+    },
+    {
+      "epoch": 0.3633638765824865,
+      "grad_norm": 2.9744557369992184,
+      "learning_rate": 3.588468565920732e-06,
+      "loss": 0.9874,
+      "step": 3455
+    },
+    {
+      "epoch": 0.3634690470243338,
+      "grad_norm": 2.0034051449703902,
+      "learning_rate": 3.5877169647054875e-06,
+      "loss": 0.9759,
+      "step": 3456
+    },
+    {
+      "epoch": 0.3635742174661811,
+      "grad_norm": 2.4849493412176065,
+      "learning_rate": 3.586965242201353e-06,
+      "loss": 1.0216,
+      "step": 3457
+    },
+    {
+      "epoch": 0.36367938790802845,
+      "grad_norm": 2.4957735320651775,
+      "learning_rate": 3.586213398492149e-06,
+      "loss": 1.0352,
+      "step": 3458
+    },
+    {
+      "epoch": 0.3637845583498758,
+      "grad_norm": 2.966574153623298,
+      "learning_rate": 3.585461433661714e-06,
+      "loss": 1.0324,
+      "step": 3459
+    },
+    {
+      "epoch": 0.3638897287917231,
+      "grad_norm": 2.0761534835735658,
+      "learning_rate": 3.5847093477938955e-06,
+      "loss": 0.9981,
+      "step": 3460
+    },
+    {
+      "epoch": 0.3639948992335704,
+      "grad_norm": 2.400801379499366,
+      "learning_rate": 3.5839571409725593e-06,
+      "loss": 1.0048,
+      "step": 3461
+    },
+    {
+      "epoch": 0.36410006967541775,
+      "grad_norm": 2.871894855158759,
+      "learning_rate": 3.583204813281581e-06,
+      "loss": 1.0233,
+      "step": 3462
+    },
+    {
+      "epoch": 0.364205240117265,
+      "grad_norm": 2.7896234095062895,
+      "learning_rate": 3.582452364804852e-06,
+      "loss": 1.0171,
+      "step": 3463
+    },
+    {
+      "epoch": 0.36431041055911234,
+      "grad_norm": 2.1993954405877942,
+      "learning_rate": 3.581699795626275e-06,
+      "loss": 0.9991,
+      "step": 3464
+    },
+    {
+      "epoch": 0.36441558100095967,
+      "grad_norm": 2.054066351626218,
+      "learning_rate": 3.580947105829769e-06,
+      "loss": 1.0175,
+      "step": 3465
+    },
+    {
+      "epoch": 0.364520751442807,
+      "grad_norm": 3.0539518525223213,
+      "learning_rate": 3.580194295499263e-06,
+      "loss": 1.0232,
+      "step": 3466
+    },
+    {
+      "epoch": 0.3646259218846543,
+      "grad_norm": 3.168788256078409,
+      "learning_rate": 3.5794413647187008e-06,
+      "loss": 0.9885,
+      "step": 3467
+    },
+    {
+      "epoch": 0.36473109232650164,
+      "grad_norm": 1.775472584199268,
+      "learning_rate": 3.578688313572042e-06,
+      "loss": 0.9917,
+      "step": 3468
+    },
+    {
+      "epoch": 0.36483626276834896,
+      "grad_norm": 2.6600939684935625,
+      "learning_rate": 3.577935142143256e-06,
+      "loss": 0.9925,
+      "step": 3469
+    },
+    {
+      "epoch": 0.3649414332101963,
+      "grad_norm": 3.0526490958871295,
+      "learning_rate": 3.577181850516329e-06,
+      "loss": 1.029,
+      "step": 3470
+    },
+    {
+      "epoch": 0.3650466036520436,
+      "grad_norm": 2.5112352942954383,
+      "learning_rate": 3.576428438775257e-06,
+      "loss": 1.0038,
+      "step": 3471
+    },
+    {
+      "epoch": 0.36515177409389094,
+      "grad_norm": 2.4210733566582645,
+      "learning_rate": 3.575674907004052e-06,
+      "loss": 1.0058,
+      "step": 3472
+    },
+    {
+      "epoch": 0.3652569445357382,
+      "grad_norm": 2.513245854367,
+      "learning_rate": 3.5749212552867397e-06,
+      "loss": 1.0608,
+      "step": 3473
+    },
+    {
+      "epoch": 0.36536211497758553,
+      "grad_norm": 2.7039176655465473,
+      "learning_rate": 3.5741674837073563e-06,
+      "loss": 0.9913,
+      "step": 3474
+    },
+    {
+      "epoch": 0.36546728541943285,
+      "grad_norm": 3.0641705314770693,
+      "learning_rate": 3.5734135923499548e-06,
+      "loss": 1.0289,
+      "step": 3475
+    },
+    {
+      "epoch": 0.3655724558612802,
+      "grad_norm": 2.7553733431065064,
+      "learning_rate": 3.572659581298598e-06,
+      "loss": 0.9869,
+      "step": 3476
+    },
+    {
+      "epoch": 0.3656776263031275,
+      "grad_norm": 2.876816353464465,
+      "learning_rate": 3.571905450637366e-06,
+      "loss": 1.0163,
+      "step": 3477
+    },
+    {
+      "epoch": 0.3657827967449748,
+      "grad_norm": 2.7493296725215712,
+      "learning_rate": 3.571151200450349e-06,
+      "loss": 1.0349,
+      "step": 3478
+    },
+    {
+      "epoch": 0.36588796718682215,
+      "grad_norm": 2.7429718758626693,
+      "learning_rate": 3.5703968308216523e-06,
+      "loss": 0.9933,
+      "step": 3479
+    },
+    {
+      "epoch": 0.3659931376286695,
+      "grad_norm": 2.7233182700143024,
+      "learning_rate": 3.569642341835394e-06,
+      "loss": 0.9801,
+      "step": 3480
+    },
+    {
+      "epoch": 0.3660983080705168,
+      "grad_norm": 2.898665070653661,
+      "learning_rate": 3.5688877335757055e-06,
+      "loss": 1.0327,
+      "step": 3481
+    },
+    {
+      "epoch": 0.3662034785123641,
+      "grad_norm": 3.3379320870265663,
+      "learning_rate": 3.5681330061267317e-06,
+      "loss": 1.0193,
+      "step": 3482
+    },
+    {
+      "epoch": 0.36630864895421145,
+      "grad_norm": 2.6749109290262223,
+      "learning_rate": 3.5673781595726286e-06,
+      "loss": 0.9996,
+      "step": 3483
+    },
+    {
+      "epoch": 0.3664138193960587,
+      "grad_norm": 2.316180353529294,
+      "learning_rate": 3.5666231939975702e-06,
+      "loss": 1.044,
+      "step": 3484
+    },
+    {
+      "epoch": 0.36651898983790604,
+      "grad_norm": 2.343425285023339,
+      "learning_rate": 3.5658681094857394e-06,
+      "loss": 0.9644,
+      "step": 3485
+    },
+    {
+      "epoch": 0.36662416027975336,
+      "grad_norm": 2.8549983640033703,
+      "learning_rate": 3.5651129061213345e-06,
+      "loss": 1.0054,
+      "step": 3486
+    },
+    {
+      "epoch": 0.3667293307216007,
+      "grad_norm": 2.8622892093311907,
+      "learning_rate": 3.5643575839885667e-06,
+      "loss": 1.0235,
+      "step": 3487
+    },
+    {
+      "epoch": 0.366834501163448,
+      "grad_norm": 2.9948754783989253,
+      "learning_rate": 3.5636021431716604e-06,
+      "loss": 1.0314,
+      "step": 3488
+    },
+    {
+      "epoch": 0.36693967160529534,
+      "grad_norm": 2.1124215741074,
+      "learning_rate": 3.5628465837548526e-06,
+      "loss": 1.0036,
+      "step": 3489
+    },
+    {
+      "epoch": 0.36704484204714266,
+      "grad_norm": 2.985214300844437,
+      "learning_rate": 3.5620909058223933e-06,
+      "loss": 1.0203,
+      "step": 3490
+    },
+    {
+      "epoch": 0.36715001248899,
+      "grad_norm": 3.111175604817724,
+      "learning_rate": 3.5613351094585484e-06,
+      "loss": 1.0222,
+      "step": 3491
+    },
+    {
+      "epoch": 0.3672551829308373,
+      "grad_norm": 2.3456559320452195,
+      "learning_rate": 3.5605791947475934e-06,
+      "loss": 0.9925,
+      "step": 3492
+    },
+    {
+      "epoch": 0.36736035337268463,
+      "grad_norm": 2.1181973524224675,
+      "learning_rate": 3.55982316177382e-06,
+      "loss": 0.992,
+      "step": 3493
+    },
+    {
+      "epoch": 0.3674655238145319,
+      "grad_norm": 3.0413234842482844,
+      "learning_rate": 3.5590670106215307e-06,
+      "loss": 1.0221,
+      "step": 3494
+    },
+    {
+      "epoch": 0.3675706942563792,
+      "grad_norm": 2.564889170052048,
+      "learning_rate": 3.5583107413750427e-06,
+      "loss": 1.0259,
+      "step": 3495
+    },
+    {
+      "epoch": 0.36767586469822655,
+      "grad_norm": 2.626728709218182,
+      "learning_rate": 3.5575543541186853e-06,
+      "loss": 1.0114,
+      "step": 3496
+    },
+    {
+      "epoch": 0.3677810351400739,
+      "grad_norm": 2.1294990688555417,
+      "learning_rate": 3.5567978489368026e-06,
+      "loss": 1.01,
+      "step": 3497
+    },
+    {
+      "epoch": 0.3678862055819212,
+      "grad_norm": 2.502867155421716,
+      "learning_rate": 3.55604122591375e-06,
+      "loss": 1.0494,
+      "step": 3498
+    },
+    {
+      "epoch": 0.3679913760237685,
+      "grad_norm": 3.537928483150429,
+      "learning_rate": 3.5552844851338973e-06,
+      "loss": 1.0098,
+      "step": 3499
+    },
+    {
+      "epoch": 0.36809654646561585,
+      "grad_norm": 2.1233082823635665,
+      "learning_rate": 3.5545276266816265e-06,
+      "loss": 1.0037,
+      "step": 3500
+    },
+    {
+      "epoch": 0.3682017169074632,
+      "grad_norm": 2.065621137457995,
+      "learning_rate": 3.5537706506413338e-06,
+      "loss": 1.0426,
+      "step": 3501
+    },
+    {
+      "epoch": 0.3683068873493105,
+      "grad_norm": 2.382029858431779,
+      "learning_rate": 3.553013557097428e-06,
+      "loss": 1.0496,
+      "step": 3502
+    },
+    {
+      "epoch": 0.3684120577911578,
+      "grad_norm": 1.6297673370173709,
+      "learning_rate": 3.552256346134329e-06,
+      "loss": 1.012,
+      "step": 3503
+    },
+    {
+      "epoch": 0.3685172282330051,
+      "grad_norm": 2.575845849912579,
+      "learning_rate": 3.5514990178364746e-06,
+      "loss": 0.9622,
+      "step": 3504
+    },
+    {
+      "epoch": 0.3686223986748524,
+      "grad_norm": 1.8413351349315177,
+      "learning_rate": 3.550741572288311e-06,
+      "loss": 1.0271,
+      "step": 3505
+    },
+    {
+      "epoch": 0.36872756911669974,
+      "grad_norm": 2.2981599308221745,
+      "learning_rate": 3.5499840095742987e-06,
+      "loss": 1.0653,
+      "step": 3506
+    },
+    {
+      "epoch": 0.36883273955854706,
+      "grad_norm": 2.770027195890931,
+      "learning_rate": 3.549226329778914e-06,
+      "loss": 1.0416,
+      "step": 3507
+    },
+    {
+      "epoch": 0.3689379100003944,
+      "grad_norm": 2.172783977718339,
+      "learning_rate": 3.5484685329866424e-06,
+      "loss": 1.0382,
+      "step": 3508
+    },
+    {
+      "epoch": 0.3690430804422417,
+      "grad_norm": 2.7336060159802953,
+      "learning_rate": 3.547710619281985e-06,
+      "loss": 1.0135,
+      "step": 3509
+    },
+    {
+      "epoch": 0.36914825088408904,
+      "grad_norm": 3.5054838300829463,
+      "learning_rate": 3.546952588749454e-06,
+      "loss": 0.9847,
+      "step": 3510
+    },
+    {
+      "epoch": 0.36925342132593636,
+      "grad_norm": 3.127625095566459,
+      "learning_rate": 3.5461944414735766e-06,
+      "loss": 0.9888,
+      "step": 3511
+    },
+    {
+      "epoch": 0.3693585917677837,
+      "grad_norm": 2.4488246566949448,
+      "learning_rate": 3.545436177538892e-06,
+      "loss": 0.9815,
+      "step": 3512
+    },
+    {
+      "epoch": 0.369463762209631,
+      "grad_norm": 3.6498915209813,
+      "learning_rate": 3.544677797029952e-06,
+      "loss": 0.998,
+      "step": 3513
+    },
+    {
+      "epoch": 0.3695689326514783,
+      "grad_norm": 2.8541317738893306,
+      "learning_rate": 3.5439193000313226e-06,
+      "loss": 1.0012,
+      "step": 3514
+    },
+    {
+      "epoch": 0.3696741030933256,
+      "grad_norm": 2.912304761031327,
+      "learning_rate": 3.5431606866275812e-06,
+      "loss": 0.984,
+      "step": 3515
+    },
+    {
+      "epoch": 0.3697792735351729,
+      "grad_norm": 2.2203764999621467,
+      "learning_rate": 3.542401956903321e-06,
+      "loss": 1.0176,
+      "step": 3516
+    },
+    {
+      "epoch": 0.36988444397702025,
+      "grad_norm": 2.439171891410255,
+      "learning_rate": 3.5416431109431437e-06,
+      "loss": 1.0556,
+      "step": 3517
+    },
+    {
+      "epoch": 0.3699896144188676,
+      "grad_norm": 2.797287931857089,
+      "learning_rate": 3.540884148831668e-06,
+      "loss": 1.0811,
+      "step": 3518
+    },
+    {
+      "epoch": 0.3700947848607149,
+      "grad_norm": 2.6586762875266614,
+      "learning_rate": 3.540125070653524e-06,
+      "loss": 1.0006,
+      "step": 3519
+    },
+    {
+      "epoch": 0.3701999553025622,
+      "grad_norm": 2.9224263695935817,
+      "learning_rate": 3.5393658764933546e-06,
+      "loss": 1.0327,
+      "step": 3520
+    },
+    {
+      "epoch": 0.37030512574440955,
+      "grad_norm": 2.7522305202993405,
+      "learning_rate": 3.5386065664358164e-06,
+      "loss": 1.0175,
+      "step": 3521
+    },
+    {
+      "epoch": 0.37041029618625687,
+      "grad_norm": 2.195682252666581,
+      "learning_rate": 3.5378471405655768e-06,
+      "loss": 0.958,
+      "step": 3522
+    },
+    {
+      "epoch": 0.3705154666281042,
+      "grad_norm": 2.4183425267496985,
+      "learning_rate": 3.5370875989673198e-06,
+      "loss": 1.0164,
+      "step": 3523
+    },
+    {
+      "epoch": 0.37062063706995146,
+      "grad_norm": 2.315338386367098,
+      "learning_rate": 3.536327941725739e-06,
+      "loss": 1.0022,
+      "step": 3524
+    },
+    {
+      "epoch": 0.3707258075117988,
+      "grad_norm": 2.3589678245581815,
+      "learning_rate": 3.5355681689255417e-06,
+      "loss": 0.9943,
+      "step": 3525
+    },
+    {
+      "epoch": 0.3708309779536461,
+      "grad_norm": 2.575779344588469,
+      "learning_rate": 3.5348082806514496e-06,
+      "loss": 1.0326,
+      "step": 3526
+    },
+    {
+      "epoch": 0.37093614839549344,
+      "grad_norm": 2.172676292007147,
+      "learning_rate": 3.534048276988195e-06,
+      "loss": 0.9872,
+      "step": 3527
+    },
+    {
+      "epoch": 0.37104131883734076,
+      "grad_norm": 2.676488772279958,
+      "learning_rate": 3.5332881580205246e-06,
+      "loss": 0.9993,
+      "step": 3528
+    },
+    {
+      "epoch": 0.3711464892791881,
+      "grad_norm": 2.137926314108314,
+      "learning_rate": 3.5325279238331977e-06,
+      "loss": 1.0049,
+      "step": 3529
+    },
+    {
+      "epoch": 0.3712516597210354,
+      "grad_norm": 2.7103603439452075,
+      "learning_rate": 3.531767574510987e-06,
+      "loss": 0.957,
+      "step": 3530
+    },
+    {
+      "epoch": 0.37135683016288273,
+      "grad_norm": 2.558627297098295,
+      "learning_rate": 3.5310071101386766e-06,
+      "loss": 1.0162,
+      "step": 3531
+    },
+    {
+      "epoch": 0.37146200060473006,
+      "grad_norm": 2.610302442944242,
+      "learning_rate": 3.530246530801064e-06,
+      "loss": 0.9467,
+      "step": 3532
+    },
+    {
+      "epoch": 0.3715671710465774,
+      "grad_norm": 2.5992181909547543,
+      "learning_rate": 3.52948583658296e-06,
+      "loss": 0.997,
+      "step": 3533
+    },
+    {
+      "epoch": 0.37167234148842465,
+      "grad_norm": 2.943011406158968,
+      "learning_rate": 3.528725027569188e-06,
+      "loss": 1.0061,
+      "step": 3534
+    },
+    {
+      "epoch": 0.371777511930272,
+      "grad_norm": 2.0335971995059103,
+      "learning_rate": 3.527964103844583e-06,
+      "loss": 1.0035,
+      "step": 3535
+    },
+    {
+      "epoch": 0.3718826823721193,
+      "grad_norm": 2.290860416398897,
+      "learning_rate": 3.5272030654939967e-06,
+      "loss": 1.0089,
+      "step": 3536
+    },
+    {
+      "epoch": 0.3719878528139666,
+      "grad_norm": 3.705983857104791,
+      "learning_rate": 3.526441912602288e-06,
+      "loss": 1.0453,
+      "step": 3537
+    },
+    {
+      "epoch": 0.37209302325581395,
+      "grad_norm": 2.910250579235628,
+      "learning_rate": 3.525680645254333e-06,
+      "loss": 1.0259,
+      "step": 3538
+    },
+    {
+      "epoch": 0.3721981936976613,
+      "grad_norm": 2.7958380455087366,
+      "learning_rate": 3.5249192635350178e-06,
+      "loss": 1.0077,
+      "step": 3539
+    },
+    {
+      "epoch": 0.3723033641395086,
+      "grad_norm": 2.415766053951578,
+      "learning_rate": 3.5241577675292426e-06,
+      "loss": 0.9892,
+      "step": 3540
+    },
+    {
+      "epoch": 0.3724085345813559,
+      "grad_norm": 3.1343561189937725,
+      "learning_rate": 3.5233961573219203e-06,
+      "loss": 1.0475,
+      "step": 3541
+    },
+    {
+      "epoch": 0.37251370502320325,
+      "grad_norm": 2.476762825422412,
+      "learning_rate": 3.5226344329979767e-06,
+      "loss": 1.01,
+      "step": 3542
+    },
+    {
+      "epoch": 0.37261887546505057,
+      "grad_norm": 2.4464917162110815,
+      "learning_rate": 3.52187259464235e-06,
+      "loss": 1.0814,
+      "step": 3543
+    },
+    {
+      "epoch": 0.3727240459068979,
+      "grad_norm": 2.54651260509731,
+      "learning_rate": 3.521110642339991e-06,
+      "loss": 1.0308,
+      "step": 3544
+    },
+    {
+      "epoch": 0.37282921634874516,
+      "grad_norm": 2.757074308896569,
+      "learning_rate": 3.5203485761758627e-06,
+      "loss": 0.9986,
+      "step": 3545
+    },
+    {
+      "epoch": 0.3729343867905925,
+      "grad_norm": 2.2476503257651643,
+      "learning_rate": 3.519586396234942e-06,
+      "loss": 1.045,
+      "step": 3546
+    },
+    {
+      "epoch": 0.3730395572324398,
+      "grad_norm": 3.4160260399141134,
+      "learning_rate": 3.518824102602217e-06,
+      "loss": 0.9736,
+      "step": 3547
+    },
+    {
+      "epoch": 0.37314472767428714,
+      "grad_norm": 2.3235809456017242,
+      "learning_rate": 3.5180616953626905e-06,
+      "loss": 1.026,
+      "step": 3548
+    },
+    {
+      "epoch": 0.37324989811613446,
+      "grad_norm": 3.1866772084593977,
+      "learning_rate": 3.5172991746013764e-06,
+      "loss": 1.0147,
+      "step": 3549
+    },
+    {
+      "epoch": 0.3733550685579818,
+      "grad_norm": 1.7405784529022383,
+      "learning_rate": 3.516536540403302e-06,
+      "loss": 0.9932,
+      "step": 3550
+    },
+    {
+      "epoch": 0.3734602389998291,
+      "grad_norm": 2.4307045279664665,
+      "learning_rate": 3.5157737928535065e-06,
+      "loss": 1.0486,
+      "step": 3551
+    },
+    {
+      "epoch": 0.37356540944167643,
+      "grad_norm": 2.4827334628389925,
+      "learning_rate": 3.5150109320370425e-06,
+      "loss": 1.0143,
+      "step": 3552
+    },
+    {
+      "epoch": 0.37367057988352376,
+      "grad_norm": 2.17954336422211,
+      "learning_rate": 3.5142479580389744e-06,
+      "loss": 0.9822,
+      "step": 3553
+    },
+    {
+      "epoch": 0.3737757503253711,
+      "grad_norm": 2.009056281408665,
+      "learning_rate": 3.5134848709443793e-06,
+      "loss": 1.0011,
+      "step": 3554
+    },
+    {
+      "epoch": 0.37388092076721835,
+      "grad_norm": 2.70716892667249,
+      "learning_rate": 3.5127216708383495e-06,
+      "loss": 1.0292,
+      "step": 3555
+    },
+    {
+      "epoch": 0.3739860912090657,
+      "grad_norm": 1.9705536861580673,
+      "learning_rate": 3.5119583578059845e-06,
+      "loss": 1.005,
+      "step": 3556
+    },
+    {
+      "epoch": 0.374091261650913,
+      "grad_norm": 2.6162266436594415,
+      "learning_rate": 3.5111949319324027e-06,
+      "loss": 1.0156,
+      "step": 3557
+    },
+    {
+      "epoch": 0.3741964320927603,
+      "grad_norm": 2.360521949456891,
+      "learning_rate": 3.51043139330273e-06,
+      "loss": 0.9847,
+      "step": 3558
+    },
+    {
+      "epoch": 0.37430160253460765,
+      "grad_norm": 3.006787570352432,
+      "learning_rate": 3.5096677420021092e-06,
+      "loss": 1.0322,
+      "step": 3559
+    },
+    {
+      "epoch": 0.37440677297645497,
+      "grad_norm": 2.2412182060746964,
+      "learning_rate": 3.508903978115691e-06,
+      "loss": 0.986,
+      "step": 3560
+    },
+    {
+      "epoch": 0.3745119434183023,
+      "grad_norm": 2.3336931601024395,
+      "learning_rate": 3.508140101728641e-06,
+      "loss": 1.0246,
+      "step": 3561
+    },
+    {
+      "epoch": 0.3746171138601496,
+      "grad_norm": 2.562229225549696,
+      "learning_rate": 3.5073761129261393e-06,
+      "loss": 0.9975,
+      "step": 3562
+    },
+    {
+      "epoch": 0.37472228430199694,
+      "grad_norm": 3.1758702644776156,
+      "learning_rate": 3.5066120117933743e-06,
+      "loss": 1.02,
+      "step": 3563
+    },
+    {
+      "epoch": 0.37482745474384427,
+      "grad_norm": 2.228109854729782,
+      "learning_rate": 3.505847798415551e-06,
+      "loss": 1.0244,
+      "step": 3564
+    },
+    {
+      "epoch": 0.37493262518569154,
+      "grad_norm": 2.355915255145524,
+      "learning_rate": 3.505083472877884e-06,
+      "loss": 1.0427,
+      "step": 3565
+    },
+    {
+      "epoch": 0.37503779562753886,
+      "grad_norm": 2.71234930662769,
+      "learning_rate": 3.5043190352656026e-06,
+      "loss": 0.9927,
+      "step": 3566
+    },
+    {
+      "epoch": 0.3751429660693862,
+      "grad_norm": 3.125287938693249,
+      "learning_rate": 3.503554485663947e-06,
+      "loss": 0.9746,
+      "step": 3567
+    },
+    {
+      "epoch": 0.3752481365112335,
+      "grad_norm": 2.832305455594945,
+      "learning_rate": 3.50278982415817e-06,
+      "loss": 0.9868,
+      "step": 3568
+    },
+    {
+      "epoch": 0.37535330695308083,
+      "grad_norm": 2.16624778915707,
+      "learning_rate": 3.502025050833538e-06,
+      "loss": 0.9823,
+      "step": 3569
+    },
+    {
+      "epoch": 0.37545847739492816,
+      "grad_norm": 2.6723466436176517,
+      "learning_rate": 3.5012601657753283e-06,
+      "loss": 0.9912,
+      "step": 3570
+    },
+    {
+      "epoch": 0.3755636478367755,
+      "grad_norm": 2.894125932929896,
+      "learning_rate": 3.500495169068832e-06,
+      "loss": 0.9625,
+      "step": 3571
+    },
+    {
+      "epoch": 0.3756688182786228,
+      "grad_norm": 2.349145898282107,
+      "learning_rate": 3.499730060799352e-06,
+      "loss": 0.9749,
+      "step": 3572
+    },
+    {
+      "epoch": 0.37577398872047013,
+      "grad_norm": 2.2715747784546636,
+      "learning_rate": 3.4989648410522053e-06,
+      "loss": 1.0067,
+      "step": 3573
+    },
+    {
+      "epoch": 0.37587915916231746,
+      "grad_norm": 2.7664541979733435,
+      "learning_rate": 3.498199509912718e-06,
+      "loss": 1.0479,
+      "step": 3574
+    },
+    {
+      "epoch": 0.3759843296041647,
+      "grad_norm": 2.120229567329133,
+      "learning_rate": 3.497434067466231e-06,
+      "loss": 1.0106,
+      "step": 3575
+    },
+    {
+      "epoch": 0.37608950004601205,
+      "grad_norm": 2.1692031969688608,
+      "learning_rate": 3.4966685137980972e-06,
+      "loss": 0.9755,
+      "step": 3576
+    },
+    {
+      "epoch": 0.3761946704878594,
+      "grad_norm": 2.304539050607103,
+      "learning_rate": 3.4959028489936815e-06,
+      "loss": 1.0251,
+      "step": 3577
+    },
+    {
+      "epoch": 0.3762998409297067,
+      "grad_norm": 3.3999675880797784,
+      "learning_rate": 3.4951370731383615e-06,
+      "loss": 0.9857,
+      "step": 3578
+    },
+    {
+      "epoch": 0.376405011371554,
+      "grad_norm": 2.254087700165053,
+      "learning_rate": 3.494371186317528e-06,
+      "loss": 0.9828,
+      "step": 3579
+    },
+    {
+      "epoch": 0.37651018181340135,
+      "grad_norm": 2.8458404135362048,
+      "learning_rate": 3.4936051886165825e-06,
+      "loss": 1.0431,
+      "step": 3580
+    },
+    {
+      "epoch": 0.37661535225524867,
+      "grad_norm": 2.4716575338002214,
+      "learning_rate": 3.4928390801209395e-06,
+      "loss": 0.9775,
+      "step": 3581
+    },
+    {
+      "epoch": 0.376720522697096,
+      "grad_norm": 2.6453228157698714,
+      "learning_rate": 3.492072860916027e-06,
+      "loss": 1.0048,
+      "step": 3582
+    },
+    {
+      "epoch": 0.3768256931389433,
+      "grad_norm": 2.7675242977400454,
+      "learning_rate": 3.4913065310872834e-06,
+      "loss": 1.0323,
+      "step": 3583
+    },
+    {
+      "epoch": 0.37693086358079064,
+      "grad_norm": 2.204262494467231,
+      "learning_rate": 3.4905400907201604e-06,
+      "loss": 1.0081,
+      "step": 3584
+    },
+    {
+      "epoch": 0.3770360340226379,
+      "grad_norm": 2.9513125690747883,
+      "learning_rate": 3.489773539900123e-06,
+      "loss": 0.9778,
+      "step": 3585
+    },
+    {
+      "epoch": 0.37714120446448524,
+      "grad_norm": 2.363516897078722,
+      "learning_rate": 3.4890068787126475e-06,
+      "loss": 1.0189,
+      "step": 3586
+    },
+    {
+      "epoch": 0.37724637490633256,
+      "grad_norm": 3.122013375469763,
+      "learning_rate": 3.4882401072432215e-06,
+      "loss": 1.0105,
+      "step": 3587
+    },
+    {
+      "epoch": 0.3773515453481799,
+      "grad_norm": 2.36989045728494,
+      "learning_rate": 3.487473225577347e-06,
+      "loss": 0.9898,
+      "step": 3588
+    },
+    {
+      "epoch": 0.3774567157900272,
+      "grad_norm": 2.9849863757199784,
+      "learning_rate": 3.486706233800537e-06,
+      "loss": 1.0172,
+      "step": 3589
+    },
+    {
+      "epoch": 0.37756188623187453,
+      "grad_norm": 2.6450922737910116,
+      "learning_rate": 3.4859391319983165e-06,
+      "loss": 1.0169,
+      "step": 3590
+    },
+    {
+      "epoch": 0.37766705667372186,
+      "grad_norm": 2.521241818346483,
+      "learning_rate": 3.485171920256224e-06,
+      "loss": 1.0099,
+      "step": 3591
+    },
+    {
+      "epoch": 0.3777722271155692,
+      "grad_norm": 3.926184947651807,
+      "learning_rate": 3.484404598659809e-06,
+      "loss": 1.026,
+      "step": 3592
+    },
+    {
+      "epoch": 0.3778773975574165,
+      "grad_norm": 2.5873477697553944,
+      "learning_rate": 3.483637167294634e-06,
+      "loss": 0.9857,
+      "step": 3593
+    },
+    {
+      "epoch": 0.37798256799926383,
+      "grad_norm": 2.890617758160231,
+      "learning_rate": 3.4828696262462743e-06,
+      "loss": 1.0289,
+      "step": 3594
+    },
+    {
+      "epoch": 0.3780877384411111,
+      "grad_norm": 2.7950755400661755,
+      "learning_rate": 3.482101975600316e-06,
+      "loss": 0.9954,
+      "step": 3595
+    },
+    {
+      "epoch": 0.3781929088829584,
+      "grad_norm": 1.9102337369098177,
+      "learning_rate": 3.481334215442358e-06,
+      "loss": 0.9304,
+      "step": 3596
+    },
+    {
+      "epoch": 0.37829807932480575,
+      "grad_norm": 3.195058034843074,
+      "learning_rate": 3.4805663458580113e-06,
+      "loss": 1.0091,
+      "step": 3597
+    },
+    {
+      "epoch": 0.37840324976665307,
+      "grad_norm": 2.1115956503169757,
+      "learning_rate": 3.4797983669329004e-06,
+      "loss": 1.0348,
+      "step": 3598
+    },
+    {
+      "epoch": 0.3785084202085004,
+      "grad_norm": 2.3416744434406813,
+      "learning_rate": 3.47903027875266e-06,
+      "loss": 0.9704,
+      "step": 3599
+    },
+    {
+      "epoch": 0.3786135906503477,
+      "grad_norm": 2.363772195609984,
+      "learning_rate": 3.4782620814029376e-06,
+      "loss": 0.9948,
+      "step": 3600
+    },
+    {
+      "epoch": 0.37871876109219504,
+      "grad_norm": 1.9628721511439282,
+      "learning_rate": 3.4774937749693947e-06,
+      "loss": 1.0062,
+      "step": 3601
+    },
+    {
+      "epoch": 0.37882393153404237,
+      "grad_norm": 2.426600466574066,
+      "learning_rate": 3.4767253595377015e-06,
+      "loss": 0.9645,
+      "step": 3602
+    },
+    {
+      "epoch": 0.3789291019758897,
+      "grad_norm": 2.6977870839090645,
+      "learning_rate": 3.4759568351935446e-06,
+      "loss": 1.0291,
+      "step": 3603
+    },
+    {
+      "epoch": 0.379034272417737,
+      "grad_norm": 2.676605696492157,
+      "learning_rate": 3.4751882020226174e-06,
+      "loss": 1.0354,
+      "step": 3604
+    },
+    {
+      "epoch": 0.37913944285958434,
+      "grad_norm": 3.363753270722036,
+      "learning_rate": 3.4744194601106314e-06,
+      "loss": 0.9662,
+      "step": 3605
+    },
+    {
+      "epoch": 0.3792446133014316,
+      "grad_norm": 2.063579478740677,
+      "learning_rate": 3.4736506095433053e-06,
+      "loss": 1.0097,
+      "step": 3606
+    },
+    {
+      "epoch": 0.37934978374327893,
+      "grad_norm": 1.9293491925383035,
+      "learning_rate": 3.472881650406373e-06,
+      "loss": 0.9955,
+      "step": 3607
+    },
+    {
+      "epoch": 0.37945495418512626,
+      "grad_norm": 2.0019157816680124,
+      "learning_rate": 3.472112582785579e-06,
+      "loss": 0.958,
+      "step": 3608
+    },
+    {
+      "epoch": 0.3795601246269736,
+      "grad_norm": 1.9340739804799547,
+      "learning_rate": 3.4713434067666803e-06,
+      "loss": 0.9936,
+      "step": 3609
+    },
+    {
+      "epoch": 0.3796652950688209,
+      "grad_norm": 2.621910541531949,
+      "learning_rate": 3.4705741224354463e-06,
+      "loss": 1.0269,
+      "step": 3610
+    },
+    {
+      "epoch": 0.37977046551066823,
+      "grad_norm": 2.319507440485628,
+      "learning_rate": 3.4698047298776578e-06,
+      "loss": 0.9785,
+      "step": 3611
+    },
+    {
+      "epoch": 0.37987563595251556,
+      "grad_norm": 3.5152668337700885,
+      "learning_rate": 3.4690352291791084e-06,
+      "loss": 1.0375,
+      "step": 3612
+    },
+    {
+      "epoch": 0.3799808063943629,
+      "grad_norm": 2.56130563777068,
+      "learning_rate": 3.4682656204256033e-06,
+      "loss": 0.9976,
+      "step": 3613
+    },
+    {
+      "epoch": 0.3800859768362102,
+      "grad_norm": 3.0184814971313294,
+      "learning_rate": 3.4674959037029593e-06,
+      "loss": 1.028,
+      "step": 3614
+    },
+    {
+      "epoch": 0.38019114727805753,
+      "grad_norm": 2.846303670335661,
+      "learning_rate": 3.4667260790970065e-06,
+      "loss": 0.9842,
+      "step": 3615
+    },
+    {
+      "epoch": 0.3802963177199048,
+      "grad_norm": 2.4611491100975558,
+      "learning_rate": 3.465956146693586e-06,
+      "loss": 0.9622,
+      "step": 3616
+    },
+    {
+      "epoch": 0.3804014881617521,
+      "grad_norm": 2.615213958589618,
+      "learning_rate": 3.4651861065785515e-06,
+      "loss": 1.0227,
+      "step": 3617
+    },
+    {
+      "epoch": 0.38050665860359945,
+      "grad_norm": 2.541915515993435,
+      "learning_rate": 3.464415958837768e-06,
+      "loss": 1.0021,
+      "step": 3618
+    },
+    {
+      "epoch": 0.38061182904544677,
+      "grad_norm": 2.3551759643491184,
+      "learning_rate": 3.4636457035571135e-06,
+      "loss": 0.9849,
+      "step": 3619
+    },
+    {
+      "epoch": 0.3807169994872941,
+      "grad_norm": 3.3489736874070846,
+      "learning_rate": 3.462875340822477e-06,
+      "loss": 0.9757,
+      "step": 3620
+    },
+    {
+      "epoch": 0.3808221699291414,
+      "grad_norm": 3.2681188585536973,
+      "learning_rate": 3.4621048707197607e-06,
+      "loss": 1.0341,
+      "step": 3621
+    },
+    {
+      "epoch": 0.38092734037098874,
+      "grad_norm": 3.4147560699832744,
+      "learning_rate": 3.461334293334877e-06,
+      "loss": 1.0442,
+      "step": 3622
+    },
+    {
+      "epoch": 0.38103251081283607,
+      "grad_norm": 2.2145079608120346,
+      "learning_rate": 3.4605636087537514e-06,
+      "loss": 0.998,
+      "step": 3623
+    },
+    {
+      "epoch": 0.3811376812546834,
+      "grad_norm": 1.8033358102527874,
+      "learning_rate": 3.4597928170623217e-06,
+      "loss": 1.0069,
+      "step": 3624
+    },
+    {
+      "epoch": 0.3812428516965307,
+      "grad_norm": 2.7695898589555465,
+      "learning_rate": 3.459021918346537e-06,
+      "loss": 1.018,
+      "step": 3625
+    },
+    {
+      "epoch": 0.381348022138378,
+      "grad_norm": 3.306734754623489,
+      "learning_rate": 3.458250912692359e-06,
+      "loss": 1.006,
+      "step": 3626
+    },
+    {
+      "epoch": 0.3814531925802253,
+      "grad_norm": 2.986143628061839,
+      "learning_rate": 3.4574798001857598e-06,
+      "loss": 1.0543,
+      "step": 3627
+    },
+    {
+      "epoch": 0.38155836302207263,
+      "grad_norm": 2.3585732168057922,
+      "learning_rate": 3.4567085809127247e-06,
+      "loss": 1.0696,
+      "step": 3628
+    },
+    {
+      "epoch": 0.38166353346391996,
+      "grad_norm": 2.876800068207621,
+      "learning_rate": 3.4559372549592513e-06,
+      "loss": 1.0084,
+      "step": 3629
+    },
+    {
+      "epoch": 0.3817687039057673,
+      "grad_norm": 2.7444522756355534,
+      "learning_rate": 3.455165822411347e-06,
+      "loss": 1.0238,
+      "step": 3630
+    },
+    {
+      "epoch": 0.3818738743476146,
+      "grad_norm": 2.73044610583874,
+      "learning_rate": 3.4543942833550347e-06,
+      "loss": 0.982,
+      "step": 3631
+    },
+    {
+      "epoch": 0.38197904478946193,
+      "grad_norm": 3.4323322120847712,
+      "learning_rate": 3.453622637876346e-06,
+      "loss": 1.0404,
+      "step": 3632
+    },
+    {
+      "epoch": 0.38208421523130925,
+      "grad_norm": 3.64343859536563,
+      "learning_rate": 3.4528508860613243e-06,
+      "loss": 1.0047,
+      "step": 3633
+    },
+    {
+      "epoch": 0.3821893856731566,
+      "grad_norm": 2.3123411149550885,
+      "learning_rate": 3.452079027996027e-06,
+      "loss": 0.9852,
+      "step": 3634
+    },
+    {
+      "epoch": 0.3822945561150039,
+      "grad_norm": 2.2206426400316,
+      "learning_rate": 3.451307063766522e-06,
+      "loss": 1.0439,
+      "step": 3635
+    },
+    {
+      "epoch": 0.38239972655685117,
+      "grad_norm": 2.0863520447051855,
+      "learning_rate": 3.45053499345889e-06,
+      "loss": 0.9948,
+      "step": 3636
+    },
+    {
+      "epoch": 0.3825048969986985,
+      "grad_norm": 1.7414750885086039,
+      "learning_rate": 3.449762817159221e-06,
+      "loss": 1.0387,
+      "step": 3637
+    },
+    {
+      "epoch": 0.3826100674405458,
+      "grad_norm": 2.8068429720399295,
+      "learning_rate": 3.4489905349536207e-06,
+      "loss": 1.0214,
+      "step": 3638
+    },
+    {
+      "epoch": 0.38271523788239314,
+      "grad_norm": 2.3872404074955216,
+      "learning_rate": 3.448218146928204e-06,
+      "loss": 1.0203,
+      "step": 3639
+    },
+    {
+      "epoch": 0.38282040832424047,
+      "grad_norm": 3.3031263883141846,
+      "learning_rate": 3.4474456531690976e-06,
+      "loss": 1.0072,
+      "step": 3640
+    },
+    {
+      "epoch": 0.3829255787660878,
+      "grad_norm": 2.6702786020242044,
+      "learning_rate": 3.4466730537624406e-06,
+      "loss": 1.0556,
+      "step": 3641
+    },
+    {
+      "epoch": 0.3830307492079351,
+      "grad_norm": 2.2459257036074343,
+      "learning_rate": 3.4459003487943842e-06,
+      "loss": 0.9824,
+      "step": 3642
+    },
+    {
+      "epoch": 0.38313591964978244,
+      "grad_norm": 2.809514281952401,
+      "learning_rate": 3.4451275383510905e-06,
+      "loss": 1.0374,
+      "step": 3643
+    },
+    {
+      "epoch": 0.38324109009162977,
+      "grad_norm": 2.6516762564393455,
+      "learning_rate": 3.444354622518735e-06,
+      "loss": 0.9732,
+      "step": 3644
+    },
+    {
+      "epoch": 0.3833462605334771,
+      "grad_norm": 2.7733996194687043,
+      "learning_rate": 3.443581601383503e-06,
+      "loss": 1.007,
+      "step": 3645
+    },
+    {
+      "epoch": 0.38345143097532436,
+      "grad_norm": 2.7799328910992718,
+      "learning_rate": 3.4428084750315925e-06,
+      "loss": 1.0295,
+      "step": 3646
+    },
+    {
+      "epoch": 0.3835566014171717,
+      "grad_norm": 2.195837771998775,
+      "learning_rate": 3.442035243549213e-06,
+      "loss": 1.0024,
+      "step": 3647
+    },
+    {
+      "epoch": 0.383661771859019,
+      "grad_norm": 1.9838499795852829,
+      "learning_rate": 3.441261907022585e-06,
+      "loss": 0.9941,
+      "step": 3648
+    },
+    {
+      "epoch": 0.38376694230086633,
+      "grad_norm": 2.8357895335106758,
+      "learning_rate": 3.4404884655379433e-06,
+      "loss": 1.0466,
+      "step": 3649
+    },
+    {
+      "epoch": 0.38387211274271366,
+      "grad_norm": 3.490737509449956,
+      "learning_rate": 3.4397149191815317e-06,
+      "loss": 0.9822,
+      "step": 3650
+    },
+    {
+      "epoch": 0.383977283184561,
+      "grad_norm": 3.0020658384708194,
+      "learning_rate": 3.438941268039606e-06,
+      "loss": 1.0228,
+      "step": 3651
+    },
+    {
+      "epoch": 0.3840824536264083,
+      "grad_norm": 2.580567442677589,
+      "learning_rate": 3.438167512198436e-06,
+      "loss": 0.9814,
+      "step": 3652
+    },
+    {
+      "epoch": 0.38418762406825563,
+      "grad_norm": 2.6240587564325732,
+      "learning_rate": 3.4373936517442996e-06,
+      "loss": 0.9582,
+      "step": 3653
+    },
+    {
+      "epoch": 0.38429279451010295,
+      "grad_norm": 2.80704824521018,
+      "learning_rate": 3.43661968676349e-06,
+      "loss": 1.0803,
+      "step": 3654
+    },
+    {
+      "epoch": 0.3843979649519503,
+      "grad_norm": 2.019265289979809,
+      "learning_rate": 3.4358456173423084e-06,
+      "loss": 0.9665,
+      "step": 3655
+    },
+    {
+      "epoch": 0.38450313539379755,
+      "grad_norm": 2.130869954578095,
+      "learning_rate": 3.4350714435670706e-06,
+      "loss": 0.9747,
+      "step": 3656
+    },
+    {
+      "epoch": 0.38460830583564487,
+      "grad_norm": 2.6188509617557267,
+      "learning_rate": 3.434297165524103e-06,
+      "loss": 0.9966,
+      "step": 3657
+    },
+    {
+      "epoch": 0.3847134762774922,
+      "grad_norm": 3.1899105918775748,
+      "learning_rate": 3.433522783299744e-06,
+      "loss": 1.0068,
+      "step": 3658
+    },
+    {
+      "epoch": 0.3848186467193395,
+      "grad_norm": 2.6244822778152224,
+      "learning_rate": 3.432748296980343e-06,
+      "loss": 0.998,
+      "step": 3659
+    },
+    {
+      "epoch": 0.38492381716118684,
+      "grad_norm": 2.1576133938956885,
+      "learning_rate": 3.4319737066522603e-06,
+      "loss": 1.0098,
+      "step": 3660
+    },
+    {
+      "epoch": 0.38502898760303417,
+      "grad_norm": 2.218275955657075,
+      "learning_rate": 3.43119901240187e-06,
+      "loss": 0.9752,
+      "step": 3661
+    },
+    {
+      "epoch": 0.3851341580448815,
+      "grad_norm": 3.06330157636448,
+      "learning_rate": 3.430424214315556e-06,
+      "loss": 0.9907,
+      "step": 3662
+    },
+    {
+      "epoch": 0.3852393284867288,
+      "grad_norm": 2.559143144501293,
+      "learning_rate": 3.429649312479714e-06,
+      "loss": 1.0291,
+      "step": 3663
+    },
+    {
+      "epoch": 0.38534449892857614,
+      "grad_norm": 2.7644506663996316,
+      "learning_rate": 3.4288743069807516e-06,
+      "loss": 0.9873,
+      "step": 3664
+    },
+    {
+      "epoch": 0.38544966937042346,
+      "grad_norm": 2.511298257225812,
+      "learning_rate": 3.4280991979050892e-06,
+      "loss": 1.0193,
+      "step": 3665
+    },
+    {
+      "epoch": 0.3855548398122708,
+      "grad_norm": 3.51758390169175,
+      "learning_rate": 3.427323985339156e-06,
+      "loss": 1.0199,
+      "step": 3666
+    },
+    {
+      "epoch": 0.38566001025411806,
+      "grad_norm": 2.8645279133183448,
+      "learning_rate": 3.4265486693693944e-06,
+      "loss": 0.9971,
+      "step": 3667
+    },
+    {
+      "epoch": 0.3857651806959654,
+      "grad_norm": 2.9018074235287155,
+      "learning_rate": 3.4257732500822592e-06,
+      "loss": 0.9952,
+      "step": 3668
+    },
+    {
+      "epoch": 0.3858703511378127,
+      "grad_norm": 2.0588163183524357,
+      "learning_rate": 3.4249977275642147e-06,
+      "loss": 0.9869,
+      "step": 3669
+    },
+    {
+      "epoch": 0.38597552157966003,
+      "grad_norm": 2.686052380256393,
+      "learning_rate": 3.424222101901738e-06,
+      "loss": 1.0185,
+      "step": 3670
+    },
+    {
+      "epoch": 0.38608069202150735,
+      "grad_norm": 2.3591003789868448,
+      "learning_rate": 3.423446373181317e-06,
+      "loss": 1.018,
+      "step": 3671
+    },
+    {
+      "epoch": 0.3861858624633547,
+      "grad_norm": 4.273636944885451,
+      "learning_rate": 3.4226705414894517e-06,
+      "loss": 1.0388,
+      "step": 3672
+    },
+    {
+      "epoch": 0.386291032905202,
+      "grad_norm": 2.9970837835983386,
+      "learning_rate": 3.4218946069126534e-06,
+      "loss": 1.0069,
+      "step": 3673
+    },
+    {
+      "epoch": 0.3863962033470493,
+      "grad_norm": 1.7215807328411599,
+      "learning_rate": 3.4211185695374454e-06,
+      "loss": 0.9614,
+      "step": 3674
+    },
+    {
+      "epoch": 0.38650137378889665,
+      "grad_norm": 3.4161522944359737,
+      "learning_rate": 3.4203424294503617e-06,
+      "loss": 1.0311,
+      "step": 3675
+    },
+    {
+      "epoch": 0.386606544230744,
+      "grad_norm": 2.734529352859378,
+      "learning_rate": 3.419566186737947e-06,
+      "loss": 0.9555,
+      "step": 3676
+    },
+    {
+      "epoch": 0.38671171467259124,
+      "grad_norm": 2.091680566056881,
+      "learning_rate": 3.4187898414867594e-06,
+      "loss": 0.998,
+      "step": 3677
+    },
+    {
+      "epoch": 0.38681688511443857,
+      "grad_norm": 2.5504380693881066,
+      "learning_rate": 3.418013393783367e-06,
+      "loss": 1.0147,
+      "step": 3678
+    },
+    {
+      "epoch": 0.3869220555562859,
+      "grad_norm": 2.4427536161465766,
+      "learning_rate": 3.4172368437143495e-06,
+      "loss": 1.0142,
+      "step": 3679
+    },
+    {
+      "epoch": 0.3870272259981332,
+      "grad_norm": 1.9842607417290865,
+      "learning_rate": 3.4164601913662985e-06,
+      "loss": 1.0323,
+      "step": 3680
+    },
+    {
+      "epoch": 0.38713239643998054,
+      "grad_norm": 2.65586599370129,
+      "learning_rate": 3.4156834368258175e-06,
+      "loss": 1.0053,
+      "step": 3681
+    },
+    {
+      "epoch": 0.38723756688182787,
+      "grad_norm": 2.0732316500738306,
+      "learning_rate": 3.4149065801795196e-06,
+      "loss": 0.9952,
+      "step": 3682
+    },
+    {
+      "epoch": 0.3873427373236752,
+      "grad_norm": 2.7781109716188843,
+      "learning_rate": 3.4141296215140307e-06,
+      "loss": 1.0386,
+      "step": 3683
+    },
+    {
+      "epoch": 0.3874479077655225,
+      "grad_norm": 2.0689240430312554,
+      "learning_rate": 3.4133525609159883e-06,
+      "loss": 0.9956,
+      "step": 3684
+    },
+    {
+      "epoch": 0.38755307820736984,
+      "grad_norm": 1.9357441976440448,
+      "learning_rate": 3.4125753984720392e-06,
+      "loss": 1.0061,
+      "step": 3685
+    },
+    {
+      "epoch": 0.38765824864921716,
+      "grad_norm": 3.5270821137244472,
+      "learning_rate": 3.411798134268845e-06,
+      "loss": 0.9688,
+      "step": 3686
+    },
+    {
+      "epoch": 0.38776341909106443,
+      "grad_norm": 2.7246597051655637,
+      "learning_rate": 3.4110207683930754e-06,
+      "loss": 1.018,
+      "step": 3687
+    },
+    {
+      "epoch": 0.38786858953291176,
+      "grad_norm": 2.84204082065096,
+      "learning_rate": 3.410243300931413e-06,
+      "loss": 1.0036,
+      "step": 3688
+    },
+    {
+      "epoch": 0.3879737599747591,
+      "grad_norm": 2.8073515518648535,
+      "learning_rate": 3.409465731970551e-06,
+      "loss": 0.987,
+      "step": 3689
+    },
+    {
+      "epoch": 0.3880789304166064,
+      "grad_norm": 2.203385890249724,
+      "learning_rate": 3.408688061597196e-06,
+      "loss": 1.0379,
+      "step": 3690
+    },
+    {
+      "epoch": 0.38818410085845373,
+      "grad_norm": 1.9853510788665791,
+      "learning_rate": 3.4079102898980633e-06,
+      "loss": 1.008,
+      "step": 3691
+    },
+    {
+      "epoch": 0.38828927130030105,
+      "grad_norm": 2.0269843653629502,
+      "learning_rate": 3.407132416959879e-06,
+      "loss": 0.9997,
+      "step": 3692
+    },
+    {
+      "epoch": 0.3883944417421484,
+      "grad_norm": 2.513094917455823,
+      "learning_rate": 3.406354442869384e-06,
+      "loss": 1.033,
+      "step": 3693
+    },
+    {
+      "epoch": 0.3884996121839957,
+      "grad_norm": 2.1492929808384917,
+      "learning_rate": 3.405576367713328e-06,
+      "loss": 0.9705,
+      "step": 3694
+    },
+    {
+      "epoch": 0.388604782625843,
+      "grad_norm": 2.4692888812089064,
+      "learning_rate": 3.404798191578472e-06,
+      "loss": 1.0106,
+      "step": 3695
+    },
+    {
+      "epoch": 0.38870995306769035,
+      "grad_norm": 3.2356374351548935,
+      "learning_rate": 3.4040199145515882e-06,
+      "loss": 0.9841,
+      "step": 3696
+    },
+    {
+      "epoch": 0.3888151235095376,
+      "grad_norm": 2.6289580492215268,
+      "learning_rate": 3.4032415367194628e-06,
+      "loss": 1.0185,
+      "step": 3697
+    },
+    {
+      "epoch": 0.38892029395138494,
+      "grad_norm": 2.66254807289422,
+      "learning_rate": 3.4024630581688895e-06,
+      "loss": 1.022,
+      "step": 3698
+    },
+    {
+      "epoch": 0.38902546439323227,
+      "grad_norm": 2.312866252842055,
+      "learning_rate": 3.4016844789866733e-06,
+      "loss": 1.0001,
+      "step": 3699
+    },
+    {
+      "epoch": 0.3891306348350796,
+      "grad_norm": 2.748756895399873,
+      "learning_rate": 3.400905799259634e-06,
+      "loss": 1.0213,
+      "step": 3700
+    },
+    {
+      "epoch": 0.3892358052769269,
+      "grad_norm": 2.422215122838384,
+      "learning_rate": 3.4001270190745988e-06,
+      "loss": 0.9726,
+      "step": 3701
+    },
+    {
+      "epoch": 0.38934097571877424,
+      "grad_norm": 2.3324336660284413,
+      "learning_rate": 3.3993481385184097e-06,
+      "loss": 1.0022,
+      "step": 3702
+    },
+    {
+      "epoch": 0.38944614616062156,
+      "grad_norm": 3.3738941893509407,
+      "learning_rate": 3.398569157677916e-06,
+      "loss": 1.0698,
+      "step": 3703
+    },
+    {
+      "epoch": 0.3895513166024689,
+      "grad_norm": 1.9996879578710953,
+      "learning_rate": 3.3977900766399807e-06,
+      "loss": 1.0164,
+      "step": 3704
+    },
+    {
+      "epoch": 0.3896564870443162,
+      "grad_norm": 3.7223647380800693,
+      "learning_rate": 3.3970108954914782e-06,
+      "loss": 1.0151,
+      "step": 3705
+    },
+    {
+      "epoch": 0.38976165748616354,
+      "grad_norm": 2.9087827912362934,
+      "learning_rate": 3.3962316143192923e-06,
+      "loss": 1.0174,
+      "step": 3706
+    },
+    {
+      "epoch": 0.3898668279280108,
+      "grad_norm": 2.222505824431483,
+      "learning_rate": 3.3954522332103197e-06,
+      "loss": 1.0393,
+      "step": 3707
+    },
+    {
+      "epoch": 0.38997199836985813,
+      "grad_norm": 2.110921600030713,
+      "learning_rate": 3.394672752251466e-06,
+      "loss": 0.9682,
+      "step": 3708
+    },
+    {
+      "epoch": 0.39007716881170545,
+      "grad_norm": 2.6718880567786507,
+      "learning_rate": 3.3938931715296515e-06,
+      "loss": 0.9888,
+      "step": 3709
+    },
+    {
+      "epoch": 0.3901823392535528,
+      "grad_norm": 3.1737106251642637,
+      "learning_rate": 3.393113491131803e-06,
+      "loss": 0.9869,
+      "step": 3710
+    },
+    {
+      "epoch": 0.3902875096954001,
+      "grad_norm": 3.316157096089004,
+      "learning_rate": 3.3923337111448635e-06,
+      "loss": 1.0071,
+      "step": 3711
+    },
+    {
+      "epoch": 0.3903926801372474,
+      "grad_norm": 2.393156968914557,
+      "learning_rate": 3.391553831655783e-06,
+      "loss": 1.0382,
+      "step": 3712
+    },
+    {
+      "epoch": 0.39049785057909475,
+      "grad_norm": 2.8352293224211986,
+      "learning_rate": 3.3907738527515243e-06,
+      "loss": 0.9929,
+      "step": 3713
+    },
+    {
+      "epoch": 0.3906030210209421,
+      "grad_norm": 2.950692846867964,
+      "learning_rate": 3.3899937745190616e-06,
+      "loss": 1.0297,
+      "step": 3714
+    },
+    {
+      "epoch": 0.3907081914627894,
+      "grad_norm": 1.9201546485115777,
+      "learning_rate": 3.3892135970453787e-06,
+      "loss": 0.9889,
+      "step": 3715
+    },
+    {
+      "epoch": 0.3908133619046367,
+      "grad_norm": 3.0185805923153746,
+      "learning_rate": 3.3884333204174722e-06,
+      "loss": 1.0277,
+      "step": 3716
+    },
+    {
+      "epoch": 0.390918532346484,
+      "grad_norm": 2.574076886352291,
+      "learning_rate": 3.387652944722349e-06,
+      "loss": 1.0126,
+      "step": 3717
+    },
+    {
+      "epoch": 0.3910237027883313,
+      "grad_norm": 2.9898536035964702,
+      "learning_rate": 3.3868724700470273e-06,
+      "loss": 1.0308,
+      "step": 3718
+    },
+    {
+      "epoch": 0.39112887323017864,
+      "grad_norm": 3.170586149546781,
+      "learning_rate": 3.3860918964785354e-06,
+      "loss": 1.012,
+      "step": 3719
+    },
+    {
+      "epoch": 0.39123404367202597,
+      "grad_norm": 2.162024803561226,
+      "learning_rate": 3.3853112241039136e-06,
+      "loss": 1.0032,
+      "step": 3720
+    },
+    {
+      "epoch": 0.3913392141138733,
+      "grad_norm": 2.800456326423131,
+      "learning_rate": 3.3845304530102137e-06,
+      "loss": 0.9699,
+      "step": 3721
+    },
+    {
+      "epoch": 0.3914443845557206,
+      "grad_norm": 2.230103160597217,
+      "learning_rate": 3.3837495832844958e-06,
+      "loss": 0.9907,
+      "step": 3722
+    },
+    {
+      "epoch": 0.39154955499756794,
+      "grad_norm": 1.9475989647841914,
+      "learning_rate": 3.382968615013835e-06,
+      "loss": 0.9762,
+      "step": 3723
+    },
+    {
+      "epoch": 0.39165472543941526,
+      "grad_norm": 1.690195126900537,
+      "learning_rate": 3.382187548285314e-06,
+      "loss": 0.9697,
+      "step": 3724
+    },
+    {
+      "epoch": 0.3917598958812626,
+      "grad_norm": 2.043759641357441,
+      "learning_rate": 3.3814063831860282e-06,
+      "loss": 0.9833,
+      "step": 3725
+    },
+    {
+      "epoch": 0.3918650663231099,
+      "grad_norm": 2.798655205453248,
+      "learning_rate": 3.3806251198030843e-06,
+      "loss": 0.9836,
+      "step": 3726
+    },
+    {
+      "epoch": 0.39197023676495724,
+      "grad_norm": 2.8030545959030553,
+      "learning_rate": 3.3798437582235984e-06,
+      "loss": 0.9983,
+      "step": 3727
+    },
+    {
+      "epoch": 0.3920754072068045,
+      "grad_norm": 1.88892717549588,
+      "learning_rate": 3.3790622985346982e-06,
+      "loss": 0.9806,
+      "step": 3728
+    },
+    {
+      "epoch": 0.39218057764865183,
+      "grad_norm": 2.9555789052928256,
+      "learning_rate": 3.378280740823523e-06,
+      "loss": 1.0005,
+      "step": 3729
+    },
+    {
+      "epoch": 0.39228574809049915,
+      "grad_norm": 2.999318912879433,
+      "learning_rate": 3.377499085177223e-06,
+      "loss": 1.037,
+      "step": 3730
+    },
+    {
+      "epoch": 0.3923909185323465,
+      "grad_norm": 3.187290454242505,
+      "learning_rate": 3.3767173316829577e-06,
+      "loss": 1.0223,
+      "step": 3731
+    },
+    {
+      "epoch": 0.3924960889741938,
+      "grad_norm": 2.739342019449803,
+      "learning_rate": 3.3759354804279e-06,
+      "loss": 1.0258,
+      "step": 3732
+    },
+    {
+      "epoch": 0.3926012594160411,
+      "grad_norm": 2.358011734083977,
+      "learning_rate": 3.375153531499231e-06,
+      "loss": 1.0051,
+      "step": 3733
+    },
+    {
+      "epoch": 0.39270642985788845,
+      "grad_norm": 2.4014349663864345,
+      "learning_rate": 3.3743714849841457e-06,
+      "loss": 1.0244,
+      "step": 3734
+    },
+    {
+      "epoch": 0.3928116002997358,
+      "grad_norm": 2.621545965673851,
+      "learning_rate": 3.373589340969847e-06,
+      "loss": 0.9887,
+      "step": 3735
+    },
+    {
+      "epoch": 0.3929167707415831,
+      "grad_norm": 1.4862634418648064,
+      "learning_rate": 3.37280709954355e-06,
+      "loss": 1.0241,
+      "step": 3736
+    },
+    {
+      "epoch": 0.3930219411834304,
+      "grad_norm": 2.3418718121471893,
+      "learning_rate": 3.3720247607924816e-06,
+      "loss": 1.0129,
+      "step": 3737
+    },
+    {
+      "epoch": 0.3931271116252777,
+      "grad_norm": 2.029673077254618,
+      "learning_rate": 3.3712423248038785e-06,
+      "loss": 1.0308,
+      "step": 3738
+    },
+    {
+      "epoch": 0.393232282067125,
+      "grad_norm": 2.15890181900642,
+      "learning_rate": 3.370459791664988e-06,
+      "loss": 0.9973,
+      "step": 3739
+    },
+    {
+      "epoch": 0.39333745250897234,
+      "grad_norm": 2.790195503492097,
+      "learning_rate": 3.369677161463068e-06,
+      "loss": 1.0607,
+      "step": 3740
+    },
+    {
+      "epoch": 0.39344262295081966,
+      "grad_norm": 2.3702375741582253,
+      "learning_rate": 3.3688944342853904e-06,
+      "loss": 0.9909,
+      "step": 3741
+    },
+    {
+      "epoch": 0.393547793392667,
+      "grad_norm": 2.6321034738325975,
+      "learning_rate": 3.368111610219232e-06,
+      "loss": 0.9887,
+      "step": 3742
+    },
+    {
+      "epoch": 0.3936529638345143,
+      "grad_norm": 3.016851750333276,
+      "learning_rate": 3.367328689351886e-06,
+      "loss": 1.0107,
+      "step": 3743
+    },
+    {
+      "epoch": 0.39375813427636164,
+      "grad_norm": 2.2979782597568774,
+      "learning_rate": 3.3665456717706522e-06,
+      "loss": 1.0155,
+      "step": 3744
+    },
+    {
+      "epoch": 0.39386330471820896,
+      "grad_norm": 2.695466999857748,
+      "learning_rate": 3.3657625575628462e-06,
+      "loss": 1.0072,
+      "step": 3745
+    },
+    {
+      "epoch": 0.3939684751600563,
+      "grad_norm": 1.8645001661672012,
+      "learning_rate": 3.3649793468157887e-06,
+      "loss": 0.99,
+      "step": 3746
+    },
+    {
+      "epoch": 0.3940736456019036,
+      "grad_norm": 2.324980227855501,
+      "learning_rate": 3.364196039616815e-06,
+      "loss": 0.9843,
+      "step": 3747
+    },
+    {
+      "epoch": 0.3941788160437509,
+      "grad_norm": 3.100195028994209,
+      "learning_rate": 3.3634126360532694e-06,
+      "loss": 0.9901,
+      "step": 3748
+    },
+    {
+      "epoch": 0.3942839864855982,
+      "grad_norm": 2.3643331155710534,
+      "learning_rate": 3.3626291362125075e-06,
+      "loss": 1.0829,
+      "step": 3749
+    },
+    {
+      "epoch": 0.39438915692744553,
+      "grad_norm": 2.2726875865978444,
+      "learning_rate": 3.3618455401818963e-06,
+      "loss": 0.9908,
+      "step": 3750
+    },
+    {
+      "epoch": 0.39449432736929285,
+      "grad_norm": 2.259480175273652,
+      "learning_rate": 3.3610618480488115e-06,
+      "loss": 0.9894,
+      "step": 3751
+    },
+    {
+      "epoch": 0.3945994978111402,
+      "grad_norm": 2.6256746329463154,
+      "learning_rate": 3.3602780599006425e-06,
+      "loss": 1.0088,
+      "step": 3752
+    },
+    {
+      "epoch": 0.3947046682529875,
+      "grad_norm": 2.8544786484384534,
+      "learning_rate": 3.359494175824787e-06,
+      "loss": 1.0236,
+      "step": 3753
+    },
+    {
+      "epoch": 0.3948098386948348,
+      "grad_norm": 2.055256613594115,
+      "learning_rate": 3.358710195908653e-06,
+      "loss": 1.0129,
+      "step": 3754
+    },
+    {
+      "epoch": 0.39491500913668215,
+      "grad_norm": 2.6762903150895165,
+      "learning_rate": 3.3579261202396624e-06,
+      "loss": 1.0084,
+      "step": 3755
+    },
+    {
+      "epoch": 0.3950201795785295,
+      "grad_norm": 3.105067449951284,
+      "learning_rate": 3.357141948905244e-06,
+      "loss": 1.0396,
+      "step": 3756
+    },
+    {
+      "epoch": 0.3951253500203768,
+      "grad_norm": 2.7357680717640736,
+      "learning_rate": 3.356357681992841e-06,
+      "loss": 1.0104,
+      "step": 3757
+    },
+    {
+      "epoch": 0.39523052046222407,
+      "grad_norm": 2.5272957675398615,
+      "learning_rate": 3.3555733195899034e-06,
+      "loss": 0.9918,
+      "step": 3758
+    },
+    {
+      "epoch": 0.3953356909040714,
+      "grad_norm": 2.2839849999981823,
+      "learning_rate": 3.354788861783894e-06,
+      "loss": 0.9403,
+      "step": 3759
+    },
+    {
+      "epoch": 0.3954408613459187,
+      "grad_norm": 2.68537039553363,
+      "learning_rate": 3.3540043086622865e-06,
+      "loss": 1.0048,
+      "step": 3760
+    },
+    {
+      "epoch": 0.39554603178776604,
+      "grad_norm": 1.95762032240909,
+      "learning_rate": 3.3532196603125646e-06,
+      "loss": 1.0022,
+      "step": 3761
+    },
+    {
+      "epoch": 0.39565120222961336,
+      "grad_norm": 2.4783053417448278,
+      "learning_rate": 3.3524349168222224e-06,
+      "loss": 1.0038,
+      "step": 3762
+    },
+    {
+      "epoch": 0.3957563726714607,
+      "grad_norm": 2.2138346751080373,
+      "learning_rate": 3.3516500782787653e-06,
+      "loss": 1.0094,
+      "step": 3763
+    },
+    {
+      "epoch": 0.395861543113308,
+      "grad_norm": 2.936093102972425,
+      "learning_rate": 3.3508651447697083e-06,
+      "loss": 1.0291,
+      "step": 3764
+    },
+    {
+      "epoch": 0.39596671355515534,
+      "grad_norm": 2.971080492228416,
+      "learning_rate": 3.350080116382578e-06,
+      "loss": 1.0148,
+      "step": 3765
+    },
+    {
+      "epoch": 0.39607188399700266,
+      "grad_norm": 2.4766119317435167,
+      "learning_rate": 3.3492949932049115e-06,
+      "loss": 0.9808,
+      "step": 3766
+    },
+    {
+      "epoch": 0.39617705443885,
+      "grad_norm": 2.0419716907450707,
+      "learning_rate": 3.3485097753242546e-06,
+      "loss": 0.9947,
+      "step": 3767
+    },
+    {
+      "epoch": 0.39628222488069725,
+      "grad_norm": 2.6590218339263787,
+      "learning_rate": 3.3477244628281667e-06,
+      "loss": 0.9918,
+      "step": 3768
+    },
+    {
+      "epoch": 0.3963873953225446,
+      "grad_norm": 2.711175294896815,
+      "learning_rate": 3.3469390558042163e-06,
+      "loss": 0.9712,
+      "step": 3769
+    },
+    {
+      "epoch": 0.3964925657643919,
+      "grad_norm": 3.5550676117965385,
+      "learning_rate": 3.346153554339982e-06,
+      "loss": 1.0413,
+      "step": 3770
+    },
+    {
+      "epoch": 0.3965977362062392,
+      "grad_norm": 2.9614092689255473,
+      "learning_rate": 3.3453679585230532e-06,
+      "loss": 1.0076,
+      "step": 3771
+    },
+    {
+      "epoch": 0.39670290664808655,
+      "grad_norm": 2.3025973292096507,
+      "learning_rate": 3.3445822684410295e-06,
+      "loss": 0.9737,
+      "step": 3772
+    },
+    {
+      "epoch": 0.3968080770899339,
+      "grad_norm": 2.2173191627458158,
+      "learning_rate": 3.3437964841815226e-06,
+      "loss": 0.9836,
+      "step": 3773
+    },
+    {
+      "epoch": 0.3969132475317812,
+      "grad_norm": 2.3309333133092243,
+      "learning_rate": 3.3430106058321517e-06,
+      "loss": 1.0039,
+      "step": 3774
+    },
+    {
+      "epoch": 0.3970184179736285,
+      "grad_norm": 1.9871063924389158,
+      "learning_rate": 3.3422246334805504e-06,
+      "loss": 0.9684,
+      "step": 3775
+    },
+    {
+      "epoch": 0.39712358841547585,
+      "grad_norm": 2.4572887046393204,
+      "learning_rate": 3.34143856721436e-06,
+      "loss": 0.9899,
+      "step": 3776
+    },
+    {
+      "epoch": 0.39722875885732317,
+      "grad_norm": 2.365568699436195,
+      "learning_rate": 3.3406524071212327e-06,
+      "loss": 0.9614,
+      "step": 3777
+    },
+    {
+      "epoch": 0.39733392929917044,
+      "grad_norm": 1.7818280646816138,
+      "learning_rate": 3.3398661532888314e-06,
+      "loss": 0.9849,
+      "step": 3778
+    },
+    {
+      "epoch": 0.39743909974101777,
+      "grad_norm": 1.9046459086271046,
+      "learning_rate": 3.339079805804829e-06,
+      "loss": 1.0143,
+      "step": 3779
+    },
+    {
+      "epoch": 0.3975442701828651,
+      "grad_norm": 2.3309655746278333,
+      "learning_rate": 3.3382933647569115e-06,
+      "loss": 0.9559,
+      "step": 3780
+    },
+    {
+      "epoch": 0.3976494406247124,
+      "grad_norm": 1.9564505802602077,
+      "learning_rate": 3.337506830232771e-06,
+      "loss": 1.0231,
+      "step": 3781
+    },
+    {
+      "epoch": 0.39775461106655974,
+      "grad_norm": 2.474905869601421,
+      "learning_rate": 3.3367202023201128e-06,
+      "loss": 1.0038,
+      "step": 3782
+    },
+    {
+      "epoch": 0.39785978150840706,
+      "grad_norm": 2.598347095424969,
+      "learning_rate": 3.3359334811066524e-06,
+      "loss": 1.0089,
+      "step": 3783
+    },
+    {
+      "epoch": 0.3979649519502544,
+      "grad_norm": 2.164211187048747,
+      "learning_rate": 3.3351466666801147e-06,
+      "loss": 1.0008,
+      "step": 3784
+    },
+    {
+      "epoch": 0.3980701223921017,
+      "grad_norm": 2.7808448041427845,
+      "learning_rate": 3.3343597591282366e-06,
+      "loss": 1.0307,
+      "step": 3785
+    },
+    {
+      "epoch": 0.39817529283394903,
+      "grad_norm": 3.473039231593379,
+      "learning_rate": 3.3335727585387636e-06,
+      "loss": 1.0376,
+      "step": 3786
+    },
+    {
+      "epoch": 0.39828046327579636,
+      "grad_norm": 2.8034840356520485,
+      "learning_rate": 3.332785664999453e-06,
+      "loss": 1.0261,
+      "step": 3787
+    },
+    {
+      "epoch": 0.3983856337176437,
+      "grad_norm": 2.7072024059878474,
+      "learning_rate": 3.3319984785980703e-06,
+      "loss": 1.0098,
+      "step": 3788
+    },
+    {
+      "epoch": 0.39849080415949095,
+      "grad_norm": 2.4114719388071872,
+      "learning_rate": 3.331211199422395e-06,
+      "loss": 1.0064,
+      "step": 3789
+    },
+    {
+      "epoch": 0.3985959746013383,
+      "grad_norm": 1.7540733706136926,
+      "learning_rate": 3.330423827560213e-06,
+      "loss": 0.9798,
+      "step": 3790
+    },
+    {
+      "epoch": 0.3987011450431856,
+      "grad_norm": 2.370248082622637,
+      "learning_rate": 3.3296363630993245e-06,
+      "loss": 0.999,
+      "step": 3791
+    },
+    {
+      "epoch": 0.3988063154850329,
+      "grad_norm": 2.51665004694854,
+      "learning_rate": 3.3288488061275365e-06,
+      "loss": 0.9833,
+      "step": 3792
+    },
+    {
+      "epoch": 0.39891148592688025,
+      "grad_norm": 2.5790260976139705,
+      "learning_rate": 3.328061156732667e-06,
+      "loss": 1.0166,
+      "step": 3793
+    },
+    {
+      "epoch": 0.3990166563687276,
+      "grad_norm": 2.090788901444674,
+      "learning_rate": 3.3272734150025467e-06,
+      "loss": 0.9734,
+      "step": 3794
+    },
+    {
+      "epoch": 0.3991218268105749,
+      "grad_norm": 2.0061905903896005,
+      "learning_rate": 3.3264855810250137e-06,
+      "loss": 1.0343,
+      "step": 3795
+    },
+    {
+      "epoch": 0.3992269972524222,
+      "grad_norm": 2.829412278323214,
+      "learning_rate": 3.3256976548879183e-06,
+      "loss": 0.9787,
+      "step": 3796
+    },
+    {
+      "epoch": 0.39933216769426955,
+      "grad_norm": 2.2031781208991723,
+      "learning_rate": 3.3249096366791196e-06,
+      "loss": 0.984,
+      "step": 3797
+    },
+    {
+      "epoch": 0.39943733813611687,
+      "grad_norm": 2.6789806053409944,
+      "learning_rate": 3.324121526486489e-06,
+      "loss": 1.0294,
+      "step": 3798
+    },
+    {
+      "epoch": 0.39954250857796414,
+      "grad_norm": 2.2888330035394273,
+      "learning_rate": 3.323333324397906e-06,
+      "loss": 1.0565,
+      "step": 3799
+    },
+    {
+      "epoch": 0.39964767901981146,
+      "grad_norm": 2.247074905242807,
+      "learning_rate": 3.3225450305012614e-06,
+      "loss": 1.012,
+      "step": 3800
+    },
+    {
+      "epoch": 0.3997528494616588,
+      "grad_norm": 2.694527096396108,
+      "learning_rate": 3.321756644884456e-06,
+      "loss": 1.0569,
+      "step": 3801
+    },
+    {
+      "epoch": 0.3998580199035061,
+      "grad_norm": 1.7511580004306093,
+      "learning_rate": 3.3209681676354012e-06,
+      "loss": 1.018,
+      "step": 3802
+    },
+    {
+      "epoch": 0.39996319034535344,
+      "grad_norm": 2.8140466000278463,
+      "learning_rate": 3.3201795988420184e-06,
+      "loss": 1.0134,
+      "step": 3803
+    },
+    {
+      "epoch": 0.40006836078720076,
+      "grad_norm": 2.892670193426442,
+      "learning_rate": 3.3193909385922385e-06,
+      "loss": 0.9829,
+      "step": 3804
+    },
+    {
+      "epoch": 0.4001735312290481,
+      "grad_norm": 2.0932756519682356,
+      "learning_rate": 3.3186021869740036e-06,
+      "loss": 1.0221,
+      "step": 3805
+    },
+    {
+      "epoch": 0.4002787016708954,
+      "grad_norm": 2.690319929330406,
+      "learning_rate": 3.3178133440752664e-06,
+      "loss": 1.0595,
+      "step": 3806
+    },
+    {
+      "epoch": 0.40038387211274273,
+      "grad_norm": 3.1784389270535365,
+      "learning_rate": 3.317024409983988e-06,
+      "loss": 1.0084,
+      "step": 3807
+    },
+    {
+      "epoch": 0.40048904255459006,
+      "grad_norm": 1.8851550338498042,
+      "learning_rate": 3.3162353847881414e-06,
+      "loss": 0.9487,
+      "step": 3808
+    },
+    {
+      "epoch": 0.4005942129964373,
+      "grad_norm": 2.5639704456014574,
+      "learning_rate": 3.3154462685757077e-06,
+      "loss": 0.9514,
+      "step": 3809
+    },
+    {
+      "epoch": 0.40069938343828465,
+      "grad_norm": 2.3978339213837856,
+      "learning_rate": 3.3146570614346814e-06,
+      "loss": 1.0111,
+      "step": 3810
+    },
+    {
+      "epoch": 0.400804553880132,
+      "grad_norm": 2.6853925911053285,
+      "learning_rate": 3.3138677634530637e-06,
+      "loss": 1.0216,
+      "step": 3811
+    },
+    {
+      "epoch": 0.4009097243219793,
+      "grad_norm": 2.138313165527399,
+      "learning_rate": 3.313078374718868e-06,
+      "loss": 0.9739,
+      "step": 3812
+    },
+    {
+      "epoch": 0.4010148947638266,
+      "grad_norm": 2.1570278204299753,
+      "learning_rate": 3.3122888953201176e-06,
+      "loss": 0.9708,
+      "step": 3813
+    },
+    {
+      "epoch": 0.40112006520567395,
+      "grad_norm": 2.128469182048962,
+      "learning_rate": 3.3114993253448457e-06,
+      "loss": 1.0079,
+      "step": 3814
+    },
+    {
+      "epoch": 0.40122523564752127,
+      "grad_norm": 3.0955012345320623,
+      "learning_rate": 3.3107096648810945e-06,
+      "loss": 1.0154,
+      "step": 3815
+    },
+    {
+      "epoch": 0.4013304060893686,
+      "grad_norm": 3.5600261230012413,
+      "learning_rate": 3.309919914016918e-06,
+      "loss": 1.0091,
+      "step": 3816
+    },
+    {
+      "epoch": 0.4014355765312159,
+      "grad_norm": 3.6965619154449274,
+      "learning_rate": 3.309130072840379e-06,
+      "loss": 0.9899,
+      "step": 3817
+    },
+    {
+      "epoch": 0.40154074697306325,
+      "grad_norm": 3.1497857281534882,
+      "learning_rate": 3.3083401414395516e-06,
+      "loss": 1.0138,
+      "step": 3818
+    },
+    {
+      "epoch": 0.4016459174149105,
+      "grad_norm": 2.752987084802028,
+      "learning_rate": 3.3075501199025194e-06,
+      "loss": 1.0456,
+      "step": 3819
+    },
+    {
+      "epoch": 0.40175108785675784,
+      "grad_norm": 2.3240624321572456,
+      "learning_rate": 3.3067600083173752e-06,
+      "loss": 0.963,
+      "step": 3820
+    },
+    {
+      "epoch": 0.40185625829860516,
+      "grad_norm": 3.583739472871665,
+      "learning_rate": 3.3059698067722236e-06,
+      "loss": 0.9924,
+      "step": 3821
+    },
+    {
+      "epoch": 0.4019614287404525,
+      "grad_norm": 2.1565141093715754,
+      "learning_rate": 3.305179515355177e-06,
+      "loss": 1.0152,
+      "step": 3822
+    },
+    {
+      "epoch": 0.4020665991822998,
+      "grad_norm": 2.697943981821852,
+      "learning_rate": 3.304389134154359e-06,
+      "loss": 1.0333,
+      "step": 3823
+    },
+    {
+      "epoch": 0.40217176962414714,
+      "grad_norm": 2.7143533256140215,
+      "learning_rate": 3.303598663257904e-06,
+      "loss": 0.9832,
+      "step": 3824
+    },
+    {
+      "epoch": 0.40227694006599446,
+      "grad_norm": 2.750860124589298,
+      "learning_rate": 3.3028081027539562e-06,
+      "loss": 1.0203,
+      "step": 3825
+    },
+    {
+      "epoch": 0.4023821105078418,
+      "grad_norm": 3.058694203710453,
+      "learning_rate": 3.302017452730668e-06,
+      "loss": 0.9606,
+      "step": 3826
+    },
+    {
+      "epoch": 0.4024872809496891,
+      "grad_norm": 2.6222751926965073,
+      "learning_rate": 3.301226713276203e-06,
+      "loss": 1.0546,
+      "step": 3827
+    },
+    {
+      "epoch": 0.40259245139153643,
+      "grad_norm": 3.5103197855371935,
+      "learning_rate": 3.300435884478737e-06,
+      "loss": 1.0299,
+      "step": 3828
+    },
+    {
+      "epoch": 0.4026976218333837,
+      "grad_norm": 2.8240018305917975,
+      "learning_rate": 3.2996449664264506e-06,
+      "loss": 1.0326,
+      "step": 3829
+    },
+    {
+      "epoch": 0.402802792275231,
+      "grad_norm": 2.5409628619006646,
+      "learning_rate": 3.298853959207539e-06,
+      "loss": 1.0357,
+      "step": 3830
+    },
+    {
+      "epoch": 0.40290796271707835,
+      "grad_norm": 2.6396149602766235,
+      "learning_rate": 3.298062862910205e-06,
+      "loss": 1.0238,
+      "step": 3831
+    },
+    {
+      "epoch": 0.4030131331589257,
+      "grad_norm": 2.2641809884701716,
+      "learning_rate": 3.297271677622662e-06,
+      "loss": 1.0312,
+      "step": 3832
+    },
+    {
+      "epoch": 0.403118303600773,
+      "grad_norm": 2.1404633046302535,
+      "learning_rate": 3.2964804034331344e-06,
+      "loss": 1.0002,
+      "step": 3833
+    },
+    {
+      "epoch": 0.4032234740426203,
+      "grad_norm": 3.02584736223,
+      "learning_rate": 3.2956890404298537e-06,
+      "loss": 1.0608,
+      "step": 3834
+    },
+    {
+      "epoch": 0.40332864448446765,
+      "grad_norm": 1.916867136991986,
+      "learning_rate": 3.2948975887010643e-06,
+      "loss": 0.9875,
+      "step": 3835
+    },
+    {
+      "epoch": 0.40343381492631497,
+      "grad_norm": 3.057690418288368,
+      "learning_rate": 3.2941060483350183e-06,
+      "loss": 0.9777,
+      "step": 3836
+    },
+    {
+      "epoch": 0.4035389853681623,
+      "grad_norm": 2.0359787957493336,
+      "learning_rate": 3.29331441941998e-06,
+      "loss": 0.9759,
+      "step": 3837
+    },
+    {
+      "epoch": 0.4036441558100096,
+      "grad_norm": 2.565513639029297,
+      "learning_rate": 3.292522702044221e-06,
+      "loss": 1.0005,
+      "step": 3838
+    },
+    {
+      "epoch": 0.4037493262518569,
+      "grad_norm": 2.6810471066425743,
+      "learning_rate": 3.2917308962960232e-06,
+      "loss": 1.0066,
+      "step": 3839
+    },
+    {
+      "epoch": 0.4038544966937042,
+      "grad_norm": 3.16105856831009,
+      "learning_rate": 3.2909390022636813e-06,
+      "loss": 1.0143,
+      "step": 3840
+    },
+    {
+      "epoch": 0.40395966713555154,
+      "grad_norm": 2.9822084656388923,
+      "learning_rate": 3.2901470200354953e-06,
+      "loss": 0.998,
+      "step": 3841
+    },
+    {
+      "epoch": 0.40406483757739886,
+      "grad_norm": 2.19021483535333,
+      "learning_rate": 3.289354949699779e-06,
+      "loss": 0.9732,
+      "step": 3842
+    },
+    {
+      "epoch": 0.4041700080192462,
+      "grad_norm": 3.090611395109951,
+      "learning_rate": 3.288562791344854e-06,
+      "loss": 1.0176,
+      "step": 3843
+    },
+    {
+      "epoch": 0.4042751784610935,
+      "grad_norm": 2.468177042950385,
+      "learning_rate": 3.2877705450590525e-06,
+      "loss": 1.04,
+      "step": 3844
+    },
+    {
+      "epoch": 0.40438034890294083,
+      "grad_norm": 3.2507866817267086,
+      "learning_rate": 3.2869782109307147e-06,
+      "loss": 1.024,
+      "step": 3845
+    },
+    {
+      "epoch": 0.40448551934478816,
+      "grad_norm": 2.749363551983819,
+      "learning_rate": 3.2861857890481928e-06,
+      "loss": 1.0086,
+      "step": 3846
+    },
+    {
+      "epoch": 0.4045906897866355,
+      "grad_norm": 3.3223260711708282,
+      "learning_rate": 3.2853932794998487e-06,
+      "loss": 1.026,
+      "step": 3847
+    },
+    {
+      "epoch": 0.4046958602284828,
+      "grad_norm": 2.0959836207172917,
+      "learning_rate": 3.2846006823740527e-06,
+      "loss": 0.963,
+      "step": 3848
+    },
+    {
+      "epoch": 0.40480103067033013,
+      "grad_norm": 2.601661822818262,
+      "learning_rate": 3.283807997759186e-06,
+      "loss": 1.0286,
+      "step": 3849
+    },
+    {
+      "epoch": 0.4049062011121774,
+      "grad_norm": 2.223265802293675,
+      "learning_rate": 3.283015225743638e-06,
+      "loss": 0.9474,
+      "step": 3850
+    },
+    {
+      "epoch": 0.4050113715540247,
+      "grad_norm": 3.994077202727601,
+      "learning_rate": 3.2822223664158103e-06,
+      "loss": 0.9945,
+      "step": 3851
+    },
+    {
+      "epoch": 0.40511654199587205,
+      "grad_norm": 2.035477588921751,
+      "learning_rate": 3.281429419864112e-06,
+      "loss": 0.9734,
+      "step": 3852
+    },
+    {
+      "epoch": 0.4052217124377194,
+      "grad_norm": 3.15506782103542,
+      "learning_rate": 3.2806363861769626e-06,
+      "loss": 1.0234,
+      "step": 3853
+    },
+    {
+      "epoch": 0.4053268828795667,
+      "grad_norm": 3.0102187470478428,
+      "learning_rate": 3.2798432654427925e-06,
+      "loss": 0.9851,
+      "step": 3854
+    },
+    {
+      "epoch": 0.405432053321414,
+      "grad_norm": 1.8770708121120634,
+      "learning_rate": 3.2790500577500393e-06,
+      "loss": 0.9907,
+      "step": 3855
+    },
+    {
+      "epoch": 0.40553722376326135,
+      "grad_norm": 2.6017298144724914,
+      "learning_rate": 3.2782567631871536e-06,
+      "loss": 1.0101,
+      "step": 3856
+    },
+    {
+      "epoch": 0.40564239420510867,
+      "grad_norm": 2.3770058244751833,
+      "learning_rate": 3.2774633818425932e-06,
+      "loss": 1.0182,
+      "step": 3857
+    },
+    {
+      "epoch": 0.405747564646956,
+      "grad_norm": 2.2714004871527003,
+      "learning_rate": 3.2766699138048265e-06,
+      "loss": 0.9867,
+      "step": 3858
+    },
+    {
+      "epoch": 0.4058527350888033,
+      "grad_norm": 2.0842568889627793,
+      "learning_rate": 3.27587635916233e-06,
+      "loss": 0.95,
+      "step": 3859
+    },
+    {
+      "epoch": 0.4059579055306506,
+      "grad_norm": 2.0840650081817884,
+      "learning_rate": 3.2750827180035927e-06,
+      "loss": 1.0121,
+      "step": 3860
+    },
+    {
+      "epoch": 0.4060630759724979,
+      "grad_norm": 1.9672860296600982,
+      "learning_rate": 3.2742889904171114e-06,
+      "loss": 1.0028,
+      "step": 3861
+    },
+    {
+      "epoch": 0.40616824641434524,
+      "grad_norm": 2.885167297028863,
+      "learning_rate": 3.2734951764913926e-06,
+      "loss": 0.9963,
+      "step": 3862
+    },
+    {
+      "epoch": 0.40627341685619256,
+      "grad_norm": 2.704006997564469,
+      "learning_rate": 3.2727012763149535e-06,
+      "loss": 1.0243,
+      "step": 3863
+    },
+    {
+      "epoch": 0.4063785872980399,
+      "grad_norm": 2.3809414848304398,
+      "learning_rate": 3.2719072899763186e-06,
+      "loss": 0.9912,
+      "step": 3864
+    },
+    {
+      "epoch": 0.4064837577398872,
+      "grad_norm": 2.6369184204411287,
+      "learning_rate": 3.271113217564025e-06,
+      "loss": 0.9984,
+      "step": 3865
+    },
+    {
+      "epoch": 0.40658892818173453,
+      "grad_norm": 2.8295484338886734,
+      "learning_rate": 3.2703190591666174e-06,
+      "loss": 1.0434,
+      "step": 3866
+    },
+    {
+      "epoch": 0.40669409862358186,
+      "grad_norm": 2.925767740990065,
+      "learning_rate": 3.2695248148726513e-06,
+      "loss": 1.032,
+      "step": 3867
+    },
+    {
+      "epoch": 0.4067992690654292,
+      "grad_norm": 2.6452717628361406,
+      "learning_rate": 3.268730484770689e-06,
+      "loss": 1.0057,
+      "step": 3868
+    },
+    {
+      "epoch": 0.4069044395072765,
+      "grad_norm": 2.0361868641093652,
+      "learning_rate": 3.2679360689493067e-06,
+      "loss": 0.9866,
+      "step": 3869
+    },
+    {
+      "epoch": 0.4070096099491238,
+      "grad_norm": 2.8896506691583648,
+      "learning_rate": 3.2671415674970874e-06,
+      "loss": 0.9859,
+      "step": 3870
+    },
+    {
+      "epoch": 0.4071147803909711,
+      "grad_norm": 2.3143761737562305,
+      "learning_rate": 3.266346980502624e-06,
+      "loss": 0.9912,
+      "step": 3871
+    },
+    {
+      "epoch": 0.4072199508328184,
+      "grad_norm": 2.821772166631868,
+      "learning_rate": 3.2655523080545188e-06,
+      "loss": 1.0074,
+      "step": 3872
+    },
+    {
+      "epoch": 0.40732512127466575,
+      "grad_norm": 2.539562382624043,
+      "learning_rate": 3.264757550241384e-06,
+      "loss": 1.0218,
+      "step": 3873
+    },
+    {
+      "epoch": 0.40743029171651307,
+      "grad_norm": 2.401844779407129,
+      "learning_rate": 3.263962707151842e-06,
+      "loss": 0.9949,
+      "step": 3874
+    },
+    {
+      "epoch": 0.4075354621583604,
+      "grad_norm": 1.6438140292650527,
+      "learning_rate": 3.263167778874523e-06,
+      "loss": 0.9944,
+      "step": 3875
+    },
+    {
+      "epoch": 0.4076406326002077,
+      "grad_norm": 2.1162759607114006,
+      "learning_rate": 3.2623727654980686e-06,
+      "loss": 1.0216,
+      "step": 3876
+    },
+    {
+      "epoch": 0.40774580304205504,
+      "grad_norm": 2.4748412890168354,
+      "learning_rate": 3.2615776671111284e-06,
+      "loss": 0.9946,
+      "step": 3877
+    },
+    {
+      "epoch": 0.40785097348390237,
+      "grad_norm": 2.4536684362236123,
+      "learning_rate": 3.2607824838023616e-06,
+      "loss": 1.0266,
+      "step": 3878
+    },
+    {
+      "epoch": 0.4079561439257497,
+      "grad_norm": 2.62038083169509,
+      "learning_rate": 3.259987215660439e-06,
+      "loss": 1.0633,
+      "step": 3879
+    },
+    {
+      "epoch": 0.40806131436759696,
+      "grad_norm": 3.030423450158621,
+      "learning_rate": 3.259191862774037e-06,
+      "loss": 1.0133,
+      "step": 3880
+    },
+    {
+      "epoch": 0.4081664848094443,
+      "grad_norm": 2.629007143942859,
+      "learning_rate": 3.2583964252318457e-06,
+      "loss": 0.986,
+      "step": 3881
+    },
+    {
+      "epoch": 0.4082716552512916,
+      "grad_norm": 2.7588881309590634,
+      "learning_rate": 3.25760090312256e-06,
+      "loss": 0.9893,
+      "step": 3882
+    },
+    {
+      "epoch": 0.40837682569313893,
+      "grad_norm": 1.8943250545350006,
+      "learning_rate": 3.25680529653489e-06,
+      "loss": 1.0228,
+      "step": 3883
+    },
+    {
+      "epoch": 0.40848199613498626,
+      "grad_norm": 2.1333974399767763,
+      "learning_rate": 3.2560096055575495e-06,
+      "loss": 1.0113,
+      "step": 3884
+    },
+    {
+      "epoch": 0.4085871665768336,
+      "grad_norm": 2.963505372102945,
+      "learning_rate": 3.2552138302792652e-06,
+      "loss": 0.9818,
+      "step": 3885
+    },
+    {
+      "epoch": 0.4086923370186809,
+      "grad_norm": 1.9028453643067,
+      "learning_rate": 3.254417970788772e-06,
+      "loss": 1.0293,
+      "step": 3886
+    },
+    {
+      "epoch": 0.40879750746052823,
+      "grad_norm": 2.733815135749297,
+      "learning_rate": 3.2536220271748154e-06,
+      "loss": 1.0211,
+      "step": 3887
+    },
+    {
+      "epoch": 0.40890267790237556,
+      "grad_norm": 2.7610883994558195,
+      "learning_rate": 3.252825999526148e-06,
+      "loss": 1.0258,
+      "step": 3888
+    },
+    {
+      "epoch": 0.4090078483442229,
+      "grad_norm": 1.5656271357264868,
+      "learning_rate": 3.252029887931533e-06,
+      "loss": 0.9811,
+      "step": 3889
+    },
+    {
+      "epoch": 0.40911301878607015,
+      "grad_norm": 1.9443105818052695,
+      "learning_rate": 3.251233692479744e-06,
+      "loss": 0.9607,
+      "step": 3890
+    },
+    {
+      "epoch": 0.4092181892279175,
+      "grad_norm": 2.4504615742047497,
+      "learning_rate": 3.2504374132595617e-06,
+      "loss": 0.9724,
+      "step": 3891
+    },
+    {
+      "epoch": 0.4093233596697648,
+      "grad_norm": 2.209862220160882,
+      "learning_rate": 3.249641050359779e-06,
+      "loss": 1.0212,
+      "step": 3892
+    },
+    {
+      "epoch": 0.4094285301116121,
+      "grad_norm": 3.0205649466760898,
+      "learning_rate": 3.2488446038691962e-06,
+      "loss": 1.0367,
+      "step": 3893
+    },
+    {
+      "epoch": 0.40953370055345945,
+      "grad_norm": 2.6987146362653918,
+      "learning_rate": 3.2480480738766222e-06,
+      "loss": 1.0034,
+      "step": 3894
+    },
+    {
+      "epoch": 0.40963887099530677,
+      "grad_norm": 2.147797158916421,
+      "learning_rate": 3.247251460470877e-06,
+      "loss": 0.9863,
+      "step": 3895
+    },
+    {
+      "epoch": 0.4097440414371541,
+      "grad_norm": 2.7248827223178025,
+      "learning_rate": 3.246454763740789e-06,
+      "loss": 1.0234,
+      "step": 3896
+    },
+    {
+      "epoch": 0.4098492118790014,
+      "grad_norm": 2.221702170406418,
+      "learning_rate": 3.2456579837751964e-06,
+      "loss": 1.0635,
+      "step": 3897
+    },
+    {
+      "epoch": 0.40995438232084874,
+      "grad_norm": 2.3438438321838726,
+      "learning_rate": 3.2448611206629456e-06,
+      "loss": 0.9872,
+      "step": 3898
+    },
+    {
+      "epoch": 0.41005955276269607,
+      "grad_norm": 1.8808604121161838,
+      "learning_rate": 3.2440641744928942e-06,
+      "loss": 0.9978,
+      "step": 3899
+    },
+    {
+      "epoch": 0.41016472320454334,
+      "grad_norm": 2.5185303976659172,
+      "learning_rate": 3.2432671453539074e-06,
+      "loss": 1.0102,
+      "step": 3900
+    },
+    {
+      "epoch": 0.41026989364639066,
+      "grad_norm": 3.0639940895334576,
+      "learning_rate": 3.242470033334859e-06,
+      "loss": 1.0326,
+      "step": 3901
+    },
+    {
+      "epoch": 0.410375064088238,
+      "grad_norm": 2.5153081413305505,
+      "learning_rate": 3.2416728385246354e-06,
+      "loss": 0.9968,
+      "step": 3902
+    },
+    {
+      "epoch": 0.4104802345300853,
+      "grad_norm": 2.3345795279528536,
+      "learning_rate": 3.2408755610121277e-06,
+      "loss": 1.0283,
+      "step": 3903
+    },
+    {
+      "epoch": 0.41058540497193263,
+      "grad_norm": 2.797650533754245,
+      "learning_rate": 3.2400782008862402e-06,
+      "loss": 1.0079,
+      "step": 3904
+    },
+    {
+      "epoch": 0.41069057541377996,
+      "grad_norm": 1.9715704104207485,
+      "learning_rate": 3.2392807582358838e-06,
+      "loss": 1.0157,
+      "step": 3905
+    },
+    {
+      "epoch": 0.4107957458556273,
+      "grad_norm": 2.3447329388547105,
+      "learning_rate": 3.2384832331499804e-06,
+      "loss": 1.039,
+      "step": 3906
+    },
+    {
+      "epoch": 0.4109009162974746,
+      "grad_norm": 2.7086394522324007,
+      "learning_rate": 3.2376856257174596e-06,
+      "loss": 1.0044,
+      "step": 3907
+    },
+    {
+      "epoch": 0.41100608673932193,
+      "grad_norm": 2.2145844261459966,
+      "learning_rate": 3.236887936027261e-06,
+      "loss": 1.0186,
+      "step": 3908
+    },
+    {
+      "epoch": 0.41111125718116925,
+      "grad_norm": 2.144833660377574,
+      "learning_rate": 3.2360901641683335e-06,
+      "loss": 0.9982,
+      "step": 3909
+    },
+    {
+      "epoch": 0.4112164276230166,
+      "grad_norm": 3.0103082487238346,
+      "learning_rate": 3.2352923102296336e-06,
+      "loss": 0.981,
+      "step": 3910
+    },
+    {
+      "epoch": 0.41132159806486385,
+      "grad_norm": 3.4747483035986786,
+      "learning_rate": 3.23449437430013e-06,
+      "loss": 1.0245,
+      "step": 3911
+    },
+    {
+      "epoch": 0.41142676850671117,
+      "grad_norm": 3.1896133858883626,
+      "learning_rate": 3.233696356468797e-06,
+      "loss": 0.9647,
+      "step": 3912
+    },
+    {
+      "epoch": 0.4115319389485585,
+      "grad_norm": 2.630457285599166,
+      "learning_rate": 3.2328982568246213e-06,
+      "loss": 1.0004,
+      "step": 3913
+    },
+    {
+      "epoch": 0.4116371093904058,
+      "grad_norm": 2.523456894015238,
+      "learning_rate": 3.2321000754565972e-06,
+      "loss": 1.0134,
+      "step": 3914
+    },
+    {
+      "epoch": 0.41174227983225314,
+      "grad_norm": 2.1853258787221623,
+      "learning_rate": 3.2313018124537273e-06,
+      "loss": 0.9849,
+      "step": 3915
+    },
+    {
+      "epoch": 0.41184745027410047,
+      "grad_norm": 2.207439047305829,
+      "learning_rate": 3.230503467905024e-06,
+      "loss": 0.9982,
+      "step": 3916
+    },
+    {
+      "epoch": 0.4119526207159478,
+      "grad_norm": 2.0123219656783427,
+      "learning_rate": 3.229705041899509e-06,
+      "loss": 0.9751,
+      "step": 3917
+    },
+    {
+      "epoch": 0.4120577911577951,
+      "grad_norm": 2.4532617794552687,
+      "learning_rate": 3.228906534526214e-06,
+      "loss": 1.0182,
+      "step": 3918
+    },
+    {
+      "epoch": 0.41216296159964244,
+      "grad_norm": 3.1167065793191897,
+      "learning_rate": 3.228107945874177e-06,
+      "loss": 0.9758,
+      "step": 3919
+    },
+    {
+      "epoch": 0.41226813204148977,
+      "grad_norm": 2.01605320694452,
+      "learning_rate": 3.2273092760324488e-06,
+      "loss": 1.0439,
+      "step": 3920
+    },
+    {
+      "epoch": 0.41237330248333703,
+      "grad_norm": 2.0503930931336134,
+      "learning_rate": 3.2265105250900857e-06,
+      "loss": 1.0149,
+      "step": 3921
+    },
+    {
+      "epoch": 0.41247847292518436,
+      "grad_norm": 2.535946058767984,
+      "learning_rate": 3.225711693136156e-06,
+      "loss": 1.0071,
+      "step": 3922
+    },
+    {
+      "epoch": 0.4125836433670317,
+      "grad_norm": 3.0494247439611684,
+      "learning_rate": 3.224912780259736e-06,
+      "loss": 1.0153,
+      "step": 3923
+    },
+    {
+      "epoch": 0.412688813808879,
+      "grad_norm": 1.9610941081321251,
+      "learning_rate": 3.2241137865499073e-06,
+      "loss": 1.0368,
+      "step": 3924
+    },
+    {
+      "epoch": 0.41279398425072633,
+      "grad_norm": 2.187054374833785,
+      "learning_rate": 3.2233147120957674e-06,
+      "loss": 0.9997,
+      "step": 3925
+    },
+    {
+      "epoch": 0.41289915469257366,
+      "grad_norm": 2.032475210558614,
+      "learning_rate": 3.2225155569864187e-06,
+      "loss": 0.9925,
+      "step": 3926
+    },
+    {
+      "epoch": 0.413004325134421,
+      "grad_norm": 2.778151381349566,
+      "learning_rate": 3.221716321310972e-06,
+      "loss": 0.9881,
+      "step": 3927
+    },
+    {
+      "epoch": 0.4131094955762683,
+      "grad_norm": 2.84516595462591,
+      "learning_rate": 3.2209170051585486e-06,
+      "loss": 1.0279,
+      "step": 3928
+    },
+    {
+      "epoch": 0.41321466601811563,
+      "grad_norm": 2.706803603589453,
+      "learning_rate": 3.2201176086182796e-06,
+      "loss": 1.0087,
+      "step": 3929
+    },
+    {
+      "epoch": 0.41331983645996295,
+      "grad_norm": 2.3853452577702137,
+      "learning_rate": 3.2193181317793034e-06,
+      "loss": 0.9627,
+      "step": 3930
+    },
+    {
+      "epoch": 0.4134250069018102,
+      "grad_norm": 2.8455662869523275,
+      "learning_rate": 3.2185185747307675e-06,
+      "loss": 1.0292,
+      "step": 3931
+    },
+    {
+      "epoch": 0.41353017734365755,
+      "grad_norm": 2.861868010646851,
+      "learning_rate": 3.217718937561829e-06,
+      "loss": 1.0368,
+      "step": 3932
+    },
+    {
+      "epoch": 0.41363534778550487,
+      "grad_norm": 2.665200169651637,
+      "learning_rate": 3.2169192203616525e-06,
+      "loss": 1.0477,
+      "step": 3933
+    },
+    {
+      "epoch": 0.4137405182273522,
+      "grad_norm": 2.7937901564994334,
+      "learning_rate": 3.2161194232194144e-06,
+      "loss": 0.9551,
+      "step": 3934
+    },
+    {
+      "epoch": 0.4138456886691995,
+      "grad_norm": 2.488699321040164,
+      "learning_rate": 3.2153195462242968e-06,
+      "loss": 1.0131,
+      "step": 3935
+    },
+    {
+      "epoch": 0.41395085911104684,
+      "grad_norm": 1.861341029966568,
+      "learning_rate": 3.2145195894654947e-06,
+      "loss": 1.0209,
+      "step": 3936
+    },
+    {
+      "epoch": 0.41405602955289417,
+      "grad_norm": 2.814340851769455,
+      "learning_rate": 3.2137195530322067e-06,
+      "loss": 1.0021,
+      "step": 3937
+    },
+    {
+      "epoch": 0.4141611999947415,
+      "grad_norm": 3.4877888465072187,
+      "learning_rate": 3.2129194370136447e-06,
+      "loss": 0.9719,
+      "step": 3938
+    },
+    {
+      "epoch": 0.4142663704365888,
+      "grad_norm": 1.8405494233280453,
+      "learning_rate": 3.2121192414990273e-06,
+      "loss": 0.9498,
+      "step": 3939
+    },
+    {
+      "epoch": 0.41437154087843614,
+      "grad_norm": 2.2875747831185893,
+      "learning_rate": 3.2113189665775812e-06,
+      "loss": 1.0207,
+      "step": 3940
+    },
+    {
+      "epoch": 0.4144767113202834,
+      "grad_norm": 1.7886487845468368,
+      "learning_rate": 3.2105186123385457e-06,
+      "loss": 0.9669,
+      "step": 3941
+    },
+    {
+      "epoch": 0.41458188176213073,
+      "grad_norm": 3.5285401524389775,
+      "learning_rate": 3.209718178871165e-06,
+      "loss": 1.0066,
+      "step": 3942
+    },
+    {
+      "epoch": 0.41468705220397806,
+      "grad_norm": 2.8080439416391902,
+      "learning_rate": 3.2089176662646946e-06,
+      "loss": 1.0376,
+      "step": 3943
+    },
+    {
+      "epoch": 0.4147922226458254,
+      "grad_norm": 2.8063762324719694,
+      "learning_rate": 3.2081170746083972e-06,
+      "loss": 0.9974,
+      "step": 3944
+    },
+    {
+      "epoch": 0.4148973930876727,
+      "grad_norm": 2.7926830931363478,
+      "learning_rate": 3.2073164039915457e-06,
+      "loss": 1.0089,
+      "step": 3945
+    },
+    {
+      "epoch": 0.41500256352952003,
+      "grad_norm": 3.0491045661913265,
+      "learning_rate": 3.20651565450342e-06,
+      "loss": 0.97,
+      "step": 3946
+    },
+    {
+      "epoch": 0.41510773397136735,
+      "grad_norm": 3.0426107615622855,
+      "learning_rate": 3.20571482623331e-06,
+      "loss": 1.0066,
+      "step": 3947
+    },
+    {
+      "epoch": 0.4152129044132147,
+      "grad_norm": 2.396107694194826,
+      "learning_rate": 3.2049139192705146e-06,
+      "loss": 1.0248,
+      "step": 3948
+    },
+    {
+      "epoch": 0.415318074855062,
+      "grad_norm": 2.149480243498675,
+      "learning_rate": 3.2041129337043415e-06,
+      "loss": 1.0269,
+      "step": 3949
+    },
+    {
+      "epoch": 0.4154232452969093,
+      "grad_norm": 2.619422758873159,
+      "learning_rate": 3.203311869624107e-06,
+      "loss": 1.0147,
+      "step": 3950
+    },
+    {
+      "epoch": 0.4155284157387566,
+      "grad_norm": 2.575801899112643,
+      "learning_rate": 3.202510727119135e-06,
+      "loss": 1.0062,
+      "step": 3951
+    },
+    {
+      "epoch": 0.4156335861806039,
+      "grad_norm": 2.2109556260255743,
+      "learning_rate": 3.20170950627876e-06,
+      "loss": 0.9984,
+      "step": 3952
+    },
+    {
+      "epoch": 0.41573875662245124,
+      "grad_norm": 2.664759138194936,
+      "learning_rate": 3.2009082071923243e-06,
+      "loss": 0.9412,
+      "step": 3953
+    },
+    {
+      "epoch": 0.41584392706429857,
+      "grad_norm": 3.5647874059605105,
+      "learning_rate": 3.2001068299491777e-06,
+      "loss": 1.0679,
+      "step": 3954
+    },
+    {
+      "epoch": 0.4159490975061459,
+      "grad_norm": 1.9762308009149232,
+      "learning_rate": 3.1993053746386817e-06,
+      "loss": 0.9786,
+      "step": 3955
+    },
+    {
+      "epoch": 0.4160542679479932,
+      "grad_norm": 2.685694396514915,
+      "learning_rate": 3.1985038413502035e-06,
+      "loss": 0.9744,
+      "step": 3956
+    },
+    {
+      "epoch": 0.41615943838984054,
+      "grad_norm": 1.788013679923769,
+      "learning_rate": 3.197702230173121e-06,
+      "loss": 0.9708,
+      "step": 3957
+    },
+    {
+      "epoch": 0.41626460883168787,
+      "grad_norm": 2.7969444556997947,
+      "learning_rate": 3.19690054119682e-06,
+      "loss": 1.0122,
+      "step": 3958
+    },
+    {
+      "epoch": 0.4163697792735352,
+      "grad_norm": 3.247388808518861,
+      "learning_rate": 3.1960987745106954e-06,
+      "loss": 1.023,
+      "step": 3959
+    },
+    {
+      "epoch": 0.4164749497153825,
+      "grad_norm": 2.891195304951478,
+      "learning_rate": 3.195296930204149e-06,
+      "loss": 1.0265,
+      "step": 3960
+    },
+    {
+      "epoch": 0.4165801201572298,
+      "grad_norm": 2.3527631916312184,
+      "learning_rate": 3.1944950083665943e-06,
+      "loss": 0.9563,
+      "step": 3961
+    },
+    {
+      "epoch": 0.4166852905990771,
+      "grad_norm": 2.1596860621844423,
+      "learning_rate": 3.1936930090874506e-06,
+      "loss": 0.9826,
+      "step": 3962
+    },
+    {
+      "epoch": 0.41679046104092443,
+      "grad_norm": 2.6617240165685665,
+      "learning_rate": 3.192890932456148e-06,
+      "loss": 1.0048,
+      "step": 3963
+    },
+    {
+      "epoch": 0.41689563148277176,
+      "grad_norm": 2.092563955479962,
+      "learning_rate": 3.1920887785621233e-06,
+      "loss": 1.0019,
+      "step": 3964
+    },
+    {
+      "epoch": 0.4170008019246191,
+      "grad_norm": 2.2736516404112326,
+      "learning_rate": 3.1912865474948242e-06,
+      "loss": 1.0307,
+      "step": 3965
+    },
+    {
+      "epoch": 0.4171059723664664,
+      "grad_norm": 3.3937785745994526,
+      "learning_rate": 3.190484239343705e-06,
+      "loss": 1.0387,
+      "step": 3966
+    },
+    {
+      "epoch": 0.41721114280831373,
+      "grad_norm": 3.086439637922143,
+      "learning_rate": 3.1896818541982288e-06,
+      "loss": 0.9966,
+      "step": 3967
+    },
+    {
+      "epoch": 0.41731631325016105,
+      "grad_norm": 2.161092398681452,
+      "learning_rate": 3.188879392147869e-06,
+      "loss": 0.9979,
+      "step": 3968
+    },
+    {
+      "epoch": 0.4174214836920084,
+      "grad_norm": 2.732897747501538,
+      "learning_rate": 3.188076853282105e-06,
+      "loss": 0.9888,
+      "step": 3969
+    },
+    {
+      "epoch": 0.4175266541338557,
+      "grad_norm": 2.3755042681989584,
+      "learning_rate": 3.187274237690426e-06,
+      "loss": 0.9807,
+      "step": 3970
+    },
+    {
+      "epoch": 0.417631824575703,
+      "grad_norm": 3.236466063215594,
+      "learning_rate": 3.1864715454623312e-06,
+      "loss": 0.9789,
+      "step": 3971
+    },
+    {
+      "epoch": 0.4177369950175503,
+      "grad_norm": 1.9463819496777226,
+      "learning_rate": 3.1856687766873262e-06,
+      "loss": 0.9797,
+      "step": 3972
+    },
+    {
+      "epoch": 0.4178421654593976,
+      "grad_norm": 2.2662145233490234,
+      "learning_rate": 3.184865931454926e-06,
+      "loss": 1.0288,
+      "step": 3973
+    },
+    {
+      "epoch": 0.41794733590124494,
+      "grad_norm": 1.866662882697477,
+      "learning_rate": 3.184063009854655e-06,
+      "loss": 1.0038,
+      "step": 3974
+    },
+    {
+      "epoch": 0.41805250634309227,
+      "grad_norm": 2.650269702092644,
+      "learning_rate": 3.183260011976044e-06,
+      "loss": 1.0019,
+      "step": 3975
+    },
+    {
+      "epoch": 0.4181576767849396,
+      "grad_norm": 2.8276642060293913,
+      "learning_rate": 3.1824569379086334e-06,
+      "loss": 1.0599,
+      "step": 3976
+    },
+    {
+      "epoch": 0.4182628472267869,
+      "grad_norm": 3.5709056704281057,
+      "learning_rate": 3.1816537877419735e-06,
+      "loss": 1.0059,
+      "step": 3977
+    },
+    {
+      "epoch": 0.41836801766863424,
+      "grad_norm": 2.071198284904167,
+      "learning_rate": 3.180850561565621e-06,
+      "loss": 1.0026,
+      "step": 3978
+    },
+    {
+      "epoch": 0.41847318811048156,
+      "grad_norm": 2.9106218215179527,
+      "learning_rate": 3.180047259469141e-06,
+      "loss": 1.0052,
+      "step": 3979
+    },
+    {
+      "epoch": 0.4185783585523289,
+      "grad_norm": 3.6515550019028513,
+      "learning_rate": 3.1792438815421094e-06,
+      "loss": 1.0022,
+      "step": 3980
+    },
+    {
+      "epoch": 0.4186835289941762,
+      "grad_norm": 2.730798220509796,
+      "learning_rate": 3.178440427874109e-06,
+      "loss": 1.0587,
+      "step": 3981
+    },
+    {
+      "epoch": 0.4187886994360235,
+      "grad_norm": 1.887706779651145,
+      "learning_rate": 3.1776368985547296e-06,
+      "loss": 0.9977,
+      "step": 3982
+    },
+    {
+      "epoch": 0.4188938698778708,
+      "grad_norm": 3.609708212124221,
+      "learning_rate": 3.176833293673572e-06,
+      "loss": 1.001,
+      "step": 3983
+    },
+    {
+      "epoch": 0.41899904031971813,
+      "grad_norm": 2.8850263626073764,
+      "learning_rate": 3.1760296133202444e-06,
+      "loss": 1.0177,
+      "step": 3984
+    },
+    {
+      "epoch": 0.41910421076156545,
+      "grad_norm": 2.131234221005343,
+      "learning_rate": 3.175225857584364e-06,
+      "loss": 1.0055,
+      "step": 3985
+    },
+    {
+      "epoch": 0.4192093812034128,
+      "grad_norm": 2.991084252493002,
+      "learning_rate": 3.174422026555554e-06,
+      "loss": 1.0366,
+      "step": 3986
+    },
+    {
+      "epoch": 0.4193145516452601,
+      "grad_norm": 2.6654269600235274,
+      "learning_rate": 3.17361812032345e-06,
+      "loss": 0.9736,
+      "step": 3987
+    },
+    {
+      "epoch": 0.4194197220871074,
+      "grad_norm": 2.6426415681720643,
+      "learning_rate": 3.172814138977692e-06,
+      "loss": 0.9885,
+      "step": 3988
+    },
+    {
+      "epoch": 0.41952489252895475,
+      "grad_norm": 2.684767657076028,
+      "learning_rate": 3.172010082607932e-06,
+      "loss": 0.9982,
+      "step": 3989
+    },
+    {
+      "epoch": 0.4196300629708021,
+      "grad_norm": 2.891495817722396,
+      "learning_rate": 3.1712059513038264e-06,
+      "loss": 1.016,
+      "step": 3990
+    },
+    {
+      "epoch": 0.4197352334126494,
+      "grad_norm": 1.9233695211634947,
+      "learning_rate": 3.1704017451550435e-06,
+      "loss": 0.9776,
+      "step": 3991
+    },
+    {
+      "epoch": 0.41984040385449667,
+      "grad_norm": 2.8276668356522165,
+      "learning_rate": 3.169597464251258e-06,
+      "loss": 0.958,
+      "step": 3992
+    },
+    {
+      "epoch": 0.419945574296344,
+      "grad_norm": 2.2819287728424387,
+      "learning_rate": 3.1687931086821534e-06,
+      "loss": 0.9661,
+      "step": 3993
+    },
+    {
+      "epoch": 0.4200507447381913,
+      "grad_norm": 1.9335515420518428,
+      "learning_rate": 3.1679886785374227e-06,
+      "loss": 1.0086,
+      "step": 3994
+    },
+    {
+      "epoch": 0.42015591518003864,
+      "grad_norm": 2.703224110233739,
+      "learning_rate": 3.1671841739067645e-06,
+      "loss": 0.987,
+      "step": 3995
+    },
+    {
+      "epoch": 0.42026108562188597,
+      "grad_norm": 2.366911505986237,
+      "learning_rate": 3.1663795948798883e-06,
+      "loss": 1.0131,
+      "step": 3996
+    },
+    {
+      "epoch": 0.4203662560637333,
+      "grad_norm": 2.1981340850481317,
+      "learning_rate": 3.1655749415465098e-06,
+      "loss": 1.0267,
+      "step": 3997
+    },
+    {
+      "epoch": 0.4204714265055806,
+      "grad_norm": 2.2878927859138662,
+      "learning_rate": 3.164770213996356e-06,
+      "loss": 0.9874,
+      "step": 3998
+    },
+    {
+      "epoch": 0.42057659694742794,
+      "grad_norm": 2.797600970455762,
+      "learning_rate": 3.1639654123191586e-06,
+      "loss": 1.0073,
+      "step": 3999
+    },
+    {
+      "epoch": 0.42068176738927526,
+      "grad_norm": 2.587687977989116,
+      "learning_rate": 3.1631605366046604e-06,
+      "loss": 0.9942,
+      "step": 4000
+    },
+    {
+      "epoch": 0.4207869378311226,
+      "grad_norm": 2.8519043232197503,
+      "learning_rate": 3.1623555869426105e-06,
+      "loss": 1.0209,
+      "step": 4001
+    },
+    {
+      "epoch": 0.42089210827296986,
+      "grad_norm": 2.7590859333924436,
+      "learning_rate": 3.161550563422768e-06,
+      "loss": 1.0017,
+      "step": 4002
+    },
+    {
+      "epoch": 0.4209972787148172,
+      "grad_norm": 2.422336890671471,
+      "learning_rate": 3.160745466134898e-06,
+      "loss": 1.0184,
+      "step": 4003
+    },
+    {
+      "epoch": 0.4211024491566645,
+      "grad_norm": 2.632748801794489,
+      "learning_rate": 3.1599402951687745e-06,
+      "loss": 1.0328,
+      "step": 4004
+    },
+    {
+      "epoch": 0.42120761959851183,
+      "grad_norm": 2.6624952216264592,
+      "learning_rate": 3.1591350506141834e-06,
+      "loss": 0.9996,
+      "step": 4005
+    },
+    {
+      "epoch": 0.42131279004035915,
+      "grad_norm": 2.181289439115408,
+      "learning_rate": 3.1583297325609117e-06,
+      "loss": 0.9675,
+      "step": 4006
+    },
+    {
+      "epoch": 0.4214179604822065,
+      "grad_norm": 1.9476495215981124,
+      "learning_rate": 3.1575243410987627e-06,
+      "loss": 0.9822,
+      "step": 4007
+    },
+    {
+      "epoch": 0.4215231309240538,
+      "grad_norm": 2.363361592022842,
+      "learning_rate": 3.1567188763175417e-06,
+      "loss": 0.9819,
+      "step": 4008
+    },
+    {
+      "epoch": 0.4216283013659011,
+      "grad_norm": 2.8584886252924653,
+      "learning_rate": 3.155913338307064e-06,
+      "loss": 1.0167,
+      "step": 4009
+    },
+    {
+      "epoch": 0.42173347180774845,
+      "grad_norm": 2.1134205793143463,
+      "learning_rate": 3.1551077271571533e-06,
+      "loss": 1.0171,
+      "step": 4010
+    },
+    {
+      "epoch": 0.4218386422495958,
+      "grad_norm": 3.1621683092136545,
+      "learning_rate": 3.154302042957642e-06,
+      "loss": 1.0072,
+      "step": 4011
+    },
+    {
+      "epoch": 0.42194381269144304,
+      "grad_norm": 3.3120782048978805,
+      "learning_rate": 3.153496285798371e-06,
+      "loss": 0.9971,
+      "step": 4012
+    },
+    {
+      "epoch": 0.42204898313329037,
+      "grad_norm": 2.0326302294261644,
+      "learning_rate": 3.152690455769186e-06,
+      "loss": 1.0114,
+      "step": 4013
+    },
+    {
+      "epoch": 0.4221541535751377,
+      "grad_norm": 2.0910034446035146,
+      "learning_rate": 3.1518845529599464e-06,
+      "loss": 0.995,
+      "step": 4014
+    },
+    {
+      "epoch": 0.422259324016985,
+      "grad_norm": 2.7438097015848326,
+      "learning_rate": 3.1510785774605148e-06,
+      "loss": 1.0412,
+      "step": 4015
+    },
+    {
+      "epoch": 0.42236449445883234,
+      "grad_norm": 1.9719647600817969,
+      "learning_rate": 3.1502725293607635e-06,
+      "loss": 0.9841,
+      "step": 4016
+    },
+    {
+      "epoch": 0.42246966490067966,
+      "grad_norm": 2.5438613247357154,
+      "learning_rate": 3.1494664087505743e-06,
+      "loss": 1.002,
+      "step": 4017
+    },
+    {
+      "epoch": 0.422574835342527,
+      "grad_norm": 2.3918717555442424,
+      "learning_rate": 3.1486602157198347e-06,
+      "loss": 0.986,
+      "step": 4018
+    },
+    {
+      "epoch": 0.4226800057843743,
+      "grad_norm": 2.6686455761422594,
+      "learning_rate": 3.1478539503584427e-06,
+      "loss": 0.9643,
+      "step": 4019
+    },
+    {
+      "epoch": 0.42278517622622164,
+      "grad_norm": 2.966460155715894,
+      "learning_rate": 3.147047612756302e-06,
+      "loss": 1.0212,
+      "step": 4020
+    },
+    {
+      "epoch": 0.42289034666806896,
+      "grad_norm": 1.8811689133408926,
+      "learning_rate": 3.1462412030033264e-06,
+      "loss": 1.0067,
+      "step": 4021
+    },
+    {
+      "epoch": 0.42299551710991623,
+      "grad_norm": 2.425415315035811,
+      "learning_rate": 3.1454347211894364e-06,
+      "loss": 1.031,
+      "step": 4022
+    },
+    {
+      "epoch": 0.42310068755176355,
+      "grad_norm": 2.0840405211523074,
+      "learning_rate": 3.144628167404561e-06,
+      "loss": 0.98,
+      "step": 4023
+    },
+    {
+      "epoch": 0.4232058579936109,
+      "grad_norm": 2.1726194511797052,
+      "learning_rate": 3.1438215417386377e-06,
+      "loss": 0.9629,
+      "step": 4024
+    },
+    {
+      "epoch": 0.4233110284354582,
+      "grad_norm": 2.3856593069057714,
+      "learning_rate": 3.1430148442816112e-06,
+      "loss": 0.9939,
+      "step": 4025
+    },
+    {
+      "epoch": 0.4234161988773055,
+      "grad_norm": 3.3977729251669184,
+      "learning_rate": 3.1422080751234342e-06,
+      "loss": 1.0636,
+      "step": 4026
+    },
+    {
+      "epoch": 0.42352136931915285,
+      "grad_norm": 1.9288928690265361,
+      "learning_rate": 3.1414012343540667e-06,
+      "loss": 0.9848,
+      "step": 4027
+    },
+    {
+      "epoch": 0.4236265397610002,
+      "grad_norm": 1.9840004899690646,
+      "learning_rate": 3.1405943220634805e-06,
+      "loss": 0.9988,
+      "step": 4028
+    },
+    {
+      "epoch": 0.4237317102028475,
+      "grad_norm": 2.442923625799874,
+      "learning_rate": 3.1397873383416503e-06,
+      "loss": 1.0168,
+      "step": 4029
+    },
+    {
+      "epoch": 0.4238368806446948,
+      "grad_norm": 2.739607689137626,
+      "learning_rate": 3.138980283278562e-06,
+      "loss": 1.0001,
+      "step": 4030
+    },
+    {
+      "epoch": 0.42394205108654215,
+      "grad_norm": 1.8732837933056634,
+      "learning_rate": 3.1381731569642093e-06,
+      "loss": 1.0149,
+      "step": 4031
+    },
+    {
+      "epoch": 0.4240472215283895,
+      "grad_norm": 2.1861276788817694,
+      "learning_rate": 3.137365959488591e-06,
+      "loss": 1.0425,
+      "step": 4032
+    },
+    {
+      "epoch": 0.42415239197023674,
+      "grad_norm": 2.227889990859692,
+      "learning_rate": 3.1365586909417177e-06,
+      "loss": 1.0049,
+      "step": 4033
+    },
+    {
+      "epoch": 0.42425756241208407,
+      "grad_norm": 2.401704867671759,
+      "learning_rate": 3.1357513514136044e-06,
+      "loss": 1.0011,
+      "step": 4034
+    },
+    {
+      "epoch": 0.4243627328539314,
+      "grad_norm": 2.572582774921797,
+      "learning_rate": 3.1349439409942778e-06,
+      "loss": 1.0292,
+      "step": 4035
+    },
+    {
+      "epoch": 0.4244679032957787,
+      "grad_norm": 2.464307334674958,
+      "learning_rate": 3.1341364597737684e-06,
+      "loss": 1.0012,
+      "step": 4036
+    },
+    {
+      "epoch": 0.42457307373762604,
+      "grad_norm": 2.1963762256824078,
+      "learning_rate": 3.133328907842118e-06,
+      "loss": 0.9715,
+      "step": 4037
+    },
+    {
+      "epoch": 0.42467824417947336,
+      "grad_norm": 3.0073856142249267,
+      "learning_rate": 3.132521285289375e-06,
+      "loss": 0.9854,
+      "step": 4038
+    },
+    {
+      "epoch": 0.4247834146213207,
+      "grad_norm": 2.7319358749557625,
+      "learning_rate": 3.1317135922055954e-06,
+      "loss": 1.0202,
+      "step": 4039
+    },
+    {
+      "epoch": 0.424888585063168,
+      "grad_norm": 1.7350134734136733,
+      "learning_rate": 3.130905828680842e-06,
+      "loss": 0.9956,
+      "step": 4040
+    },
+    {
+      "epoch": 0.42499375550501534,
+      "grad_norm": 2.3777384845962257,
+      "learning_rate": 3.1300979948051874e-06,
+      "loss": 1.0045,
+      "step": 4041
+    },
+    {
+      "epoch": 0.42509892594686266,
+      "grad_norm": 2.601237373112988,
+      "learning_rate": 3.1292900906687123e-06,
+      "loss": 0.9874,
+      "step": 4042
+    },
+    {
+      "epoch": 0.42520409638870993,
+      "grad_norm": 2.2851187750948054,
+      "learning_rate": 3.128482116361503e-06,
+      "loss": 0.9632,
+      "step": 4043
+    },
+    {
+      "epoch": 0.42530926683055725,
+      "grad_norm": 1.7435113684957382,
+      "learning_rate": 3.1276740719736565e-06,
+      "loss": 1.0032,
+      "step": 4044
+    },
+    {
+      "epoch": 0.4254144372724046,
+      "grad_norm": 2.623135138337169,
+      "learning_rate": 3.126865957595274e-06,
+      "loss": 1.0326,
+      "step": 4045
+    },
+    {
+      "epoch": 0.4255196077142519,
+      "grad_norm": 3.0108763917408683,
+      "learning_rate": 3.126057773316469e-06,
+      "loss": 0.9891,
+      "step": 4046
+    },
+    {
+      "epoch": 0.4256247781560992,
+      "grad_norm": 2.580885149154926,
+      "learning_rate": 3.1252495192273575e-06,
+      "loss": 1.0105,
+      "step": 4047
+    },
+    {
+      "epoch": 0.42572994859794655,
+      "grad_norm": 2.1650235060129694,
+      "learning_rate": 3.1244411954180677e-06,
+      "loss": 1.033,
+      "step": 4048
+    },
+    {
+      "epoch": 0.4258351190397939,
+      "grad_norm": 2.756079053013681,
+      "learning_rate": 3.1236328019787338e-06,
+      "loss": 0.9757,
+      "step": 4049
+    },
+    {
+      "epoch": 0.4259402894816412,
+      "grad_norm": 2.9476755695727452,
+      "learning_rate": 3.1228243389994976e-06,
+      "loss": 0.9734,
+      "step": 4050
+    },
+    {
+      "epoch": 0.4260454599234885,
+      "grad_norm": 1.816970545742394,
+      "learning_rate": 3.1220158065705104e-06,
+      "loss": 0.9716,
+      "step": 4051
+    },
+    {
+      "epoch": 0.42615063036533585,
+      "grad_norm": 2.1446993232730724,
+      "learning_rate": 3.1212072047819276e-06,
+      "loss": 0.9774,
+      "step": 4052
+    },
+    {
+      "epoch": 0.4262558008071831,
+      "grad_norm": 3.315132448487326,
+      "learning_rate": 3.1203985337239174e-06,
+      "loss": 1.0018,
+      "step": 4053
+    },
+    {
+      "epoch": 0.42636097124903044,
+      "grad_norm": 2.3941591823008643,
+      "learning_rate": 3.1195897934866504e-06,
+      "loss": 0.9595,
+      "step": 4054
+    },
+    {
+      "epoch": 0.42646614169087776,
+      "grad_norm": 2.0757761997376627,
+      "learning_rate": 3.118780984160307e-06,
+      "loss": 0.9928,
+      "step": 4055
+    },
+    {
+      "epoch": 0.4265713121327251,
+      "grad_norm": 1.8275624288227292,
+      "learning_rate": 3.1179721058350787e-06,
+      "loss": 1.0441,
+      "step": 4056
+    },
+    {
+      "epoch": 0.4266764825745724,
+      "grad_norm": 2.5156642086205534,
+      "learning_rate": 3.117163158601159e-06,
+      "loss": 1.0065,
+      "step": 4057
+    },
+    {
+      "epoch": 0.42678165301641974,
+      "grad_norm": 2.851734935278207,
+      "learning_rate": 3.1163541425487535e-06,
+      "loss": 0.9887,
+      "step": 4058
+    },
+    {
+      "epoch": 0.42688682345826706,
+      "grad_norm": 2.4798006522437617,
+      "learning_rate": 3.1155450577680725e-06,
+      "loss": 1.0047,
+      "step": 4059
+    },
+    {
+      "epoch": 0.4269919939001144,
+      "grad_norm": 1.5625163069318968,
+      "learning_rate": 3.1147359043493365e-06,
+      "loss": 0.96,
+      "step": 4060
+    },
+    {
+      "epoch": 0.4270971643419617,
+      "grad_norm": 2.125796790580701,
+      "learning_rate": 3.113926682382771e-06,
+      "loss": 0.9713,
+      "step": 4061
+    },
+    {
+      "epoch": 0.42720233478380903,
+      "grad_norm": 2.7077037498810754,
+      "learning_rate": 3.113117391958612e-06,
+      "loss": 0.995,
+      "step": 4062
+    },
+    {
+      "epoch": 0.4273075052256563,
+      "grad_norm": 2.611706521027208,
+      "learning_rate": 3.1123080331671015e-06,
+      "loss": 1.0046,
+      "step": 4063
+    },
+    {
+      "epoch": 0.4274126756675036,
+      "grad_norm": 1.909431132438331,
+      "learning_rate": 3.111498606098487e-06,
+      "loss": 1.0073,
+      "step": 4064
+    },
+    {
+      "epoch": 0.42751784610935095,
+      "grad_norm": 2.7980952079855665,
+      "learning_rate": 3.110689110843029e-06,
+      "loss": 1.0122,
+      "step": 4065
+    },
+    {
+      "epoch": 0.4276230165511983,
+      "grad_norm": 2.1381354002981037,
+      "learning_rate": 3.10987954749099e-06,
+      "loss": 0.9777,
+      "step": 4066
+    },
+    {
+      "epoch": 0.4277281869930456,
+      "grad_norm": 3.062933496654773,
+      "learning_rate": 3.1090699161326442e-06,
+      "loss": 1.0339,
+      "step": 4067
+    },
+    {
+      "epoch": 0.4278333574348929,
+      "grad_norm": 2.555658076041788,
+      "learning_rate": 3.108260216858272e-06,
+      "loss": 1.0222,
+      "step": 4068
+    },
+    {
+      "epoch": 0.42793852787674025,
+      "grad_norm": 2.924322622102227,
+      "learning_rate": 3.10745044975816e-06,
+      "loss": 0.9848,
+      "step": 4069
+    },
+    {
+      "epoch": 0.4280436983185876,
+      "grad_norm": 2.7919031185411525,
+      "learning_rate": 3.1066406149226046e-06,
+      "loss": 1.026,
+      "step": 4070
+    },
+    {
+      "epoch": 0.4281488687604349,
+      "grad_norm": 2.3674479918799087,
+      "learning_rate": 3.105830712441907e-06,
+      "loss": 1.0281,
+      "step": 4071
+    },
+    {
+      "epoch": 0.4282540392022822,
+      "grad_norm": 2.258912241147774,
+      "learning_rate": 3.1050207424063793e-06,
+      "loss": 0.9928,
+      "step": 4072
+    },
+    {
+      "epoch": 0.4283592096441295,
+      "grad_norm": 2.831893269330568,
+      "learning_rate": 3.1042107049063385e-06,
+      "loss": 1.011,
+      "step": 4073
+    },
+    {
+      "epoch": 0.4284643800859768,
+      "grad_norm": 2.9136787478525115,
+      "learning_rate": 3.103400600032111e-06,
+      "loss": 0.9885,
+      "step": 4074
+    },
+    {
+      "epoch": 0.42856955052782414,
+      "grad_norm": 2.5104840959586308,
+      "learning_rate": 3.1025904278740286e-06,
+      "loss": 1.0101,
+      "step": 4075
+    },
+    {
+      "epoch": 0.42867472096967146,
+      "grad_norm": 2.831638306335408,
+      "learning_rate": 3.1017801885224332e-06,
+      "loss": 0.9507,
+      "step": 4076
+    },
+    {
+      "epoch": 0.4287798914115188,
+      "grad_norm": 2.2190444522869814,
+      "learning_rate": 3.1009698820676714e-06,
+      "loss": 0.9653,
+      "step": 4077
+    },
+    {
+      "epoch": 0.4288850618533661,
+      "grad_norm": 3.176813249530208,
+      "learning_rate": 3.100159508600099e-06,
+      "loss": 0.9871,
+      "step": 4078
+    },
+    {
+      "epoch": 0.42899023229521344,
+      "grad_norm": 3.3715464732818154,
+      "learning_rate": 3.0993490682100797e-06,
+      "loss": 1.0664,
+      "step": 4079
+    },
+    {
+      "epoch": 0.42909540273706076,
+      "grad_norm": 2.187534155940015,
+      "learning_rate": 3.0985385609879832e-06,
+      "loss": 1.0389,
+      "step": 4080
+    },
+    {
+      "epoch": 0.4292005731789081,
+      "grad_norm": 1.6249601238371671,
+      "learning_rate": 3.097727987024187e-06,
+      "loss": 1.0206,
+      "step": 4081
+    },
+    {
+      "epoch": 0.4293057436207554,
+      "grad_norm": 2.028832037009087,
+      "learning_rate": 3.096917346409078e-06,
+      "loss": 1.0082,
+      "step": 4082
+    },
+    {
+      "epoch": 0.42941091406260273,
+      "grad_norm": 2.5626702289369585,
+      "learning_rate": 3.0961066392330475e-06,
+      "loss": 1.0013,
+      "step": 4083
+    },
+    {
+      "epoch": 0.42951608450445,
+      "grad_norm": 2.6483344911125952,
+      "learning_rate": 3.0952958655864957e-06,
+      "loss": 0.9299,
+      "step": 4084
+    },
+    {
+      "epoch": 0.4296212549462973,
+      "grad_norm": 2.240406701709471,
+      "learning_rate": 3.0944850255598307e-06,
+      "loss": 1.0073,
+      "step": 4085
+    },
+    {
+      "epoch": 0.42972642538814465,
+      "grad_norm": 3.2212875181907155,
+      "learning_rate": 3.0936741192434673e-06,
+      "loss": 1.0506,
+      "step": 4086
+    },
+    {
+      "epoch": 0.429831595829992,
+      "grad_norm": 2.8050004911994466,
+      "learning_rate": 3.0928631467278275e-06,
+      "loss": 1.045,
+      "step": 4087
+    },
+    {
+      "epoch": 0.4299367662718393,
+      "grad_norm": 2.3637943800552037,
+      "learning_rate": 3.0920521081033418e-06,
+      "loss": 1.0016,
+      "step": 4088
+    },
+    {
+      "epoch": 0.4300419367136866,
+      "grad_norm": 2.4022684297375942,
+      "learning_rate": 3.091241003460446e-06,
+      "loss": 1.0081,
+      "step": 4089
+    },
+    {
+      "epoch": 0.43014710715553395,
+      "grad_norm": 2.6213335210562643,
+      "learning_rate": 3.0904298328895865e-06,
+      "loss": 1.0066,
+      "step": 4090
+    },
+    {
+      "epoch": 0.43025227759738127,
+      "grad_norm": 2.405980637873229,
+      "learning_rate": 3.089618596481213e-06,
+      "loss": 0.9793,
+      "step": 4091
+    },
+    {
+      "epoch": 0.4303574480392286,
+      "grad_norm": 2.452115188274156,
+      "learning_rate": 3.088807294325786e-06,
+      "loss": 0.93,
+      "step": 4092
+    },
+    {
+      "epoch": 0.4304626184810759,
+      "grad_norm": 2.0884019140145065,
+      "learning_rate": 3.0879959265137722e-06,
+      "loss": 0.9783,
+      "step": 4093
+    },
+    {
+      "epoch": 0.4305677889229232,
+      "grad_norm": 2.9202655391972714,
+      "learning_rate": 3.0871844931356437e-06,
+      "loss": 0.9959,
+      "step": 4094
+    },
+    {
+      "epoch": 0.4306729593647705,
+      "grad_norm": 2.0360566333873136,
+      "learning_rate": 3.0863729942818835e-06,
+      "loss": 0.985,
+      "step": 4095
+    },
+    {
+      "epoch": 0.43077812980661784,
+      "grad_norm": 2.203833572789431,
+      "learning_rate": 3.08556143004298e-06,
+      "loss": 1.0127,
+      "step": 4096
+    },
+    {
+      "epoch": 0.43088330024846516,
+      "grad_norm": 3.060914354625758,
+      "learning_rate": 3.0847498005094277e-06,
+      "loss": 1.0552,
+      "step": 4097
+    },
+    {
+      "epoch": 0.4309884706903125,
+      "grad_norm": 2.127677108709688,
+      "learning_rate": 3.0839381057717295e-06,
+      "loss": 1.027,
+      "step": 4098
+    },
+    {
+      "epoch": 0.4310936411321598,
+      "grad_norm": 2.185558039100322,
+      "learning_rate": 3.083126345920397e-06,
+      "loss": 0.9747,
+      "step": 4099
+    },
+    {
+      "epoch": 0.43119881157400713,
+      "grad_norm": 2.502125904858561,
+      "learning_rate": 3.082314521045947e-06,
+      "loss": 0.9718,
+      "step": 4100
+    },
+    {
+      "epoch": 0.43130398201585446,
+      "grad_norm": 2.249386844041367,
+      "learning_rate": 3.081502631238904e-06,
+      "loss": 1.0284,
+      "step": 4101
+    },
+    {
+      "epoch": 0.4314091524577018,
+      "grad_norm": 2.510215362561653,
+      "learning_rate": 3.080690676589801e-06,
+      "loss": 1.0115,
+      "step": 4102
+    },
+    {
+      "epoch": 0.4315143228995491,
+      "grad_norm": 2.540632786506748,
+      "learning_rate": 3.0798786571891754e-06,
+      "loss": 1.0009,
+      "step": 4103
+    },
+    {
+      "epoch": 0.4316194933413964,
+      "grad_norm": 3.268212533873219,
+      "learning_rate": 3.0790665731275764e-06,
+      "loss": 1.0034,
+      "step": 4104
+    },
+    {
+      "epoch": 0.4317246637832437,
+      "grad_norm": 2.2631828866412564,
+      "learning_rate": 3.0782544244955546e-06,
+      "loss": 0.986,
+      "step": 4105
+    },
+    {
+      "epoch": 0.431829834225091,
+      "grad_norm": 2.8101017497715945,
+      "learning_rate": 3.077442211383674e-06,
+      "loss": 1.0227,
+      "step": 4106
+    },
+    {
+      "epoch": 0.43193500466693835,
+      "grad_norm": 2.590129436859189,
+      "learning_rate": 3.0766299338825003e-06,
+      "loss": 0.9566,
+      "step": 4107
+    },
+    {
+      "epoch": 0.4320401751087857,
+      "grad_norm": 2.6443948940324655,
+      "learning_rate": 3.0758175920826096e-06,
+      "loss": 0.9694,
+      "step": 4108
+    },
+    {
+      "epoch": 0.432145345550633,
+      "grad_norm": 3.141552378052775,
+      "learning_rate": 3.075005186074584e-06,
+      "loss": 1.0031,
+      "step": 4109
+    },
+    {
+      "epoch": 0.4322505159924803,
+      "grad_norm": 3.8498143175202366,
+      "learning_rate": 3.0741927159490133e-06,
+      "loss": 1.042,
+      "step": 4110
+    },
+    {
+      "epoch": 0.43235568643432765,
+      "grad_norm": 1.586336922468649,
+      "learning_rate": 3.073380181796495e-06,
+      "loss": 0.9673,
+      "step": 4111
+    },
+    {
+      "epoch": 0.43246085687617497,
+      "grad_norm": 3.3046763352664676,
+      "learning_rate": 3.072567583707632e-06,
+      "loss": 0.9643,
+      "step": 4112
+    },
+    {
+      "epoch": 0.4325660273180223,
+      "grad_norm": 3.225434065889587,
+      "learning_rate": 3.071754921773035e-06,
+      "loss": 1.0441,
+      "step": 4113
+    },
+    {
+      "epoch": 0.43267119775986956,
+      "grad_norm": 2.4386638186184277,
+      "learning_rate": 3.070942196083323e-06,
+      "loss": 0.9961,
+      "step": 4114
+    },
+    {
+      "epoch": 0.4327763682017169,
+      "grad_norm": 2.2998360002188307,
+      "learning_rate": 3.070129406729121e-06,
+      "loss": 1.0012,
+      "step": 4115
+    },
+    {
+      "epoch": 0.4328815386435642,
+      "grad_norm": 2.2206282124156593,
+      "learning_rate": 3.0693165538010617e-06,
+      "loss": 1.001,
+      "step": 4116
+    },
+    {
+      "epoch": 0.43298670908541154,
+      "grad_norm": 2.616939405371355,
+      "learning_rate": 3.0685036373897833e-06,
+      "loss": 0.9975,
+      "step": 4117
+    },
+    {
+      "epoch": 0.43309187952725886,
+      "grad_norm": 2.1392510552663477,
+      "learning_rate": 3.0676906575859335e-06,
+      "loss": 1.0204,
+      "step": 4118
+    },
+    {
+      "epoch": 0.4331970499691062,
+      "grad_norm": 2.4770512512838523,
+      "learning_rate": 3.0668776144801653e-06,
+      "loss": 1.0116,
+      "step": 4119
+    },
+    {
+      "epoch": 0.4333022204109535,
+      "grad_norm": 2.7619726953432227,
+      "learning_rate": 3.0660645081631396e-06,
+      "loss": 1.0338,
+      "step": 4120
+    },
+    {
+      "epoch": 0.43340739085280083,
+      "grad_norm": 3.6075982772582718,
+      "learning_rate": 3.0652513387255227e-06,
+      "loss": 0.9864,
+      "step": 4121
+    },
+    {
+      "epoch": 0.43351256129464816,
+      "grad_norm": 2.0025698189454344,
+      "learning_rate": 3.064438106257992e-06,
+      "loss": 1.0167,
+      "step": 4122
+    },
+    {
+      "epoch": 0.4336177317364955,
+      "grad_norm": 2.4436854317627286,
+      "learning_rate": 3.063624810851227e-06,
+      "loss": 1.0045,
+      "step": 4123
+    },
+    {
+      "epoch": 0.43372290217834275,
+      "grad_norm": 2.485396787287376,
+      "learning_rate": 3.0628114525959175e-06,
+      "loss": 1.001,
+      "step": 4124
+    },
+    {
+      "epoch": 0.4338280726201901,
+      "grad_norm": 3.1025874629215706,
+      "learning_rate": 3.0619980315827585e-06,
+      "loss": 1.0026,
+      "step": 4125
+    },
+    {
+      "epoch": 0.4339332430620374,
+      "grad_norm": 3.037225307324419,
+      "learning_rate": 3.061184547902454e-06,
+      "loss": 0.9949,
+      "step": 4126
+    },
+    {
+      "epoch": 0.4340384135038847,
+      "grad_norm": 2.8257071700615133,
+      "learning_rate": 3.060371001645713e-06,
+      "loss": 1.0168,
+      "step": 4127
+    },
+    {
+      "epoch": 0.43414358394573205,
+      "grad_norm": 1.4769811886831412,
+      "learning_rate": 3.0595573929032513e-06,
+      "loss": 0.9953,
+      "step": 4128
+    },
+    {
+      "epoch": 0.43424875438757937,
+      "grad_norm": 2.3056756386496047,
+      "learning_rate": 3.0587437217657937e-06,
+      "loss": 0.9979,
+      "step": 4129
+    },
+    {
+      "epoch": 0.4343539248294267,
+      "grad_norm": 2.530987884033079,
+      "learning_rate": 3.057929988324071e-06,
+      "loss": 1.0052,
+      "step": 4130
+    },
+    {
+      "epoch": 0.434459095271274,
+      "grad_norm": 2.669533462808313,
+      "learning_rate": 3.0571161926688204e-06,
+      "loss": 1.0023,
+      "step": 4131
+    },
+    {
+      "epoch": 0.43456426571312134,
+      "grad_norm": 2.0814334987484417,
+      "learning_rate": 3.056302334890786e-06,
+      "loss": 0.9791,
+      "step": 4132
+    },
+    {
+      "epoch": 0.43466943615496867,
+      "grad_norm": 2.8696601264581654,
+      "learning_rate": 3.0554884150807208e-06,
+      "loss": 1.0174,
+      "step": 4133
+    },
+    {
+      "epoch": 0.43477460659681594,
+      "grad_norm": 2.2470344072413533,
+      "learning_rate": 3.0546744333293815e-06,
+      "loss": 0.974,
+      "step": 4134
+    },
+    {
+      "epoch": 0.43487977703866326,
+      "grad_norm": 2.6293475640576642,
+      "learning_rate": 3.053860389727534e-06,
+      "loss": 1.0178,
+      "step": 4135
+    },
+    {
+      "epoch": 0.4349849474805106,
+      "grad_norm": 2.023015252677748,
+      "learning_rate": 3.0530462843659504e-06,
+      "loss": 0.991,
+      "step": 4136
+    },
+    {
+      "epoch": 0.4350901179223579,
+      "grad_norm": 3.028146066432831,
+      "learning_rate": 3.0522321173354095e-06,
+      "loss": 1.0114,
+      "step": 4137
+    },
+    {
+      "epoch": 0.43519528836420523,
+      "grad_norm": 2.198855179319069,
+      "learning_rate": 3.0514178887266986e-06,
+      "loss": 1.0585,
+      "step": 4138
+    },
+    {
+      "epoch": 0.43530045880605256,
+      "grad_norm": 1.9696831755080155,
+      "learning_rate": 3.0506035986306095e-06,
+      "loss": 1.0022,
+      "step": 4139
+    },
+    {
+      "epoch": 0.4354056292478999,
+      "grad_norm": 3.374841364809197,
+      "learning_rate": 3.0497892471379415e-06,
+      "loss": 1.0142,
+      "step": 4140
+    },
+    {
+      "epoch": 0.4355107996897472,
+      "grad_norm": 1.8557258077383334,
+      "learning_rate": 3.0489748343395013e-06,
+      "loss": 0.9776,
+      "step": 4141
+    },
+    {
+      "epoch": 0.43561597013159453,
+      "grad_norm": 2.47059315827509,
+      "learning_rate": 3.0481603603261023e-06,
+      "loss": 1.0122,
+      "step": 4142
+    },
+    {
+      "epoch": 0.43572114057344186,
+      "grad_norm": 2.7641667162449974,
+      "learning_rate": 3.0473458251885658e-06,
+      "loss": 0.9913,
+      "step": 4143
+    },
+    {
+      "epoch": 0.4358263110152892,
+      "grad_norm": 1.910330796679931,
+      "learning_rate": 3.0465312290177166e-06,
+      "loss": 1.0036,
+      "step": 4144
+    },
+    {
+      "epoch": 0.43593148145713645,
+      "grad_norm": 2.118224274596519,
+      "learning_rate": 3.0457165719043906e-06,
+      "loss": 1.0305,
+      "step": 4145
+    },
+    {
+      "epoch": 0.4360366518989838,
+      "grad_norm": 3.610473332887411,
+      "learning_rate": 3.0449018539394274e-06,
+      "loss": 1.0083,
+      "step": 4146
+    },
+    {
+      "epoch": 0.4361418223408311,
+      "grad_norm": 2.3957683382072847,
+      "learning_rate": 3.044087075213675e-06,
+      "loss": 0.9799,
+      "step": 4147
+    },
+    {
+      "epoch": 0.4362469927826784,
+      "grad_norm": 2.6043783680389865,
+      "learning_rate": 3.0432722358179873e-06,
+      "loss": 1.0203,
+      "step": 4148
+    },
+    {
+      "epoch": 0.43635216322452575,
+      "grad_norm": 2.0797439186751654,
+      "learning_rate": 3.0424573358432243e-06,
+      "loss": 0.9899,
+      "step": 4149
+    },
+    {
+      "epoch": 0.43645733366637307,
+      "grad_norm": 2.5250166716374918,
+      "learning_rate": 3.041642375380254e-06,
+      "loss": 1.0332,
+      "step": 4150
+    },
+    {
+      "epoch": 0.4365625041082204,
+      "grad_norm": 2.4929052295492213,
+      "learning_rate": 3.0408273545199517e-06,
+      "loss": 1.0466,
+      "step": 4151
+    },
+    {
+      "epoch": 0.4366676745500677,
+      "grad_norm": 1.9109696456025282,
+      "learning_rate": 3.0400122733531984e-06,
+      "loss": 0.9933,
+      "step": 4152
+    },
+    {
+      "epoch": 0.43677284499191504,
+      "grad_norm": 1.9084755489133232,
+      "learning_rate": 3.039197131970881e-06,
+      "loss": 0.986,
+      "step": 4153
+    },
+    {
+      "epoch": 0.43687801543376237,
+      "grad_norm": 2.669040697159479,
+      "learning_rate": 3.0383819304638953e-06,
+      "loss": 1.0244,
+      "step": 4154
+    },
+    {
+      "epoch": 0.43698318587560964,
+      "grad_norm": 2.3914620652281022,
+      "learning_rate": 3.0375666689231424e-06,
+      "loss": 1.0093,
+      "step": 4155
+    },
+    {
+      "epoch": 0.43708835631745696,
+      "grad_norm": 3.002326215365401,
+      "learning_rate": 3.0367513474395293e-06,
+      "loss": 1.0092,
+      "step": 4156
+    },
+    {
+      "epoch": 0.4371935267593043,
+      "grad_norm": 2.611694901531462,
+      "learning_rate": 3.035935966103972e-06,
+      "loss": 0.9976,
+      "step": 4157
+    },
+    {
+      "epoch": 0.4372986972011516,
+      "grad_norm": 3.2906065240695264,
+      "learning_rate": 3.0351205250073897e-06,
+      "loss": 1.0235,
+      "step": 4158
+    },
+    {
+      "epoch": 0.43740386764299893,
+      "grad_norm": 2.7111569861223654,
+      "learning_rate": 3.034305024240713e-06,
+      "loss": 0.9998,
+      "step": 4159
+    },
+    {
+      "epoch": 0.43750903808484626,
+      "grad_norm": 2.8104041489630553,
+      "learning_rate": 3.0334894638948753e-06,
+      "loss": 0.9441,
+      "step": 4160
+    },
+    {
+      "epoch": 0.4376142085266936,
+      "grad_norm": 2.088309040465824,
+      "learning_rate": 3.0326738440608176e-06,
+      "loss": 1.0236,
+      "step": 4161
+    },
+    {
+      "epoch": 0.4377193789685409,
+      "grad_norm": 2.4797799924774955,
+      "learning_rate": 3.031858164829489e-06,
+      "loss": 0.9917,
+      "step": 4162
+    },
+    {
+      "epoch": 0.43782454941038823,
+      "grad_norm": 2.2176787604079258,
+      "learning_rate": 3.0310424262918437e-06,
+      "loss": 0.9509,
+      "step": 4163
+    },
+    {
+      "epoch": 0.43792971985223555,
+      "grad_norm": 2.2461741633185994,
+      "learning_rate": 3.030226628538843e-06,
+      "loss": 0.9834,
+      "step": 4164
+    },
+    {
+      "epoch": 0.4380348902940828,
+      "grad_norm": 3.2683363657888775,
+      "learning_rate": 3.0294107716614522e-06,
+      "loss": 0.9626,
+      "step": 4165
+    },
+    {
+      "epoch": 0.43814006073593015,
+      "grad_norm": 2.101521859677799,
+      "learning_rate": 3.0285948557506497e-06,
+      "loss": 1.0173,
+      "step": 4166
+    },
+    {
+      "epoch": 0.4382452311777775,
+      "grad_norm": 2.5072346713043774,
+      "learning_rate": 3.027778880897413e-06,
+      "loss": 0.9717,
+      "step": 4167
+    },
+    {
+      "epoch": 0.4383504016196248,
+      "grad_norm": 2.886458825011365,
+      "learning_rate": 3.026962847192732e-06,
+      "loss": 0.9941,
+      "step": 4168
+    },
+    {
+      "epoch": 0.4384555720614721,
+      "grad_norm": 2.250039487973679,
+      "learning_rate": 3.0261467547276e-06,
+      "loss": 0.999,
+      "step": 4169
+    },
+    {
+      "epoch": 0.43856074250331945,
+      "grad_norm": 3.166565115166139,
+      "learning_rate": 3.0253306035930173e-06,
+      "loss": 0.9735,
+      "step": 4170
+    },
+    {
+      "epoch": 0.43866591294516677,
+      "grad_norm": 2.798487567270618,
+      "learning_rate": 3.024514393879992e-06,
+      "loss": 1.0038,
+      "step": 4171
+    },
+    {
+      "epoch": 0.4387710833870141,
+      "grad_norm": 2.7668838586778435,
+      "learning_rate": 3.0236981256795367e-06,
+      "loss": 1.0429,
+      "step": 4172
+    },
+    {
+      "epoch": 0.4388762538288614,
+      "grad_norm": 2.3144035959674554,
+      "learning_rate": 3.022881799082672e-06,
+      "loss": 0.9873,
+      "step": 4173
+    },
+    {
+      "epoch": 0.43898142427070874,
+      "grad_norm": 2.492637324980567,
+      "learning_rate": 3.022065414180425e-06,
+      "loss": 0.994,
+      "step": 4174
+    },
+    {
+      "epoch": 0.439086594712556,
+      "grad_norm": 2.551661508974668,
+      "learning_rate": 3.021248971063829e-06,
+      "loss": 1.0159,
+      "step": 4175
+    },
+    {
+      "epoch": 0.43919176515440334,
+      "grad_norm": 2.0455759141104584,
+      "learning_rate": 3.0204324698239236e-06,
+      "loss": 1.0531,
+      "step": 4176
+    },
+    {
+      "epoch": 0.43929693559625066,
+      "grad_norm": 2.798083383983236,
+      "learning_rate": 3.019615910551755e-06,
+      "loss": 0.9885,
+      "step": 4177
+    },
+    {
+      "epoch": 0.439402106038098,
+      "grad_norm": 1.9753015864782129,
+      "learning_rate": 3.0187992933383754e-06,
+      "loss": 0.981,
+      "step": 4178
+    },
+    {
+      "epoch": 0.4395072764799453,
+      "grad_norm": 2.0935177190552077,
+      "learning_rate": 3.017982618274844e-06,
+      "loss": 0.9449,
+      "step": 4179
+    },
+    {
+      "epoch": 0.43961244692179263,
+      "grad_norm": 2.3446418568178315,
+      "learning_rate": 3.0171658854522274e-06,
+      "loss": 1.0141,
+      "step": 4180
+    },
+    {
+      "epoch": 0.43971761736363996,
+      "grad_norm": 1.3397303520976527,
+      "learning_rate": 3.016349094961597e-06,
+      "loss": 1.0155,
+      "step": 4181
+    },
+    {
+      "epoch": 0.4398227878054873,
+      "grad_norm": 2.58181226066027,
+      "learning_rate": 3.015532246894031e-06,
+      "loss": 1.0046,
+      "step": 4182
+    },
+    {
+      "epoch": 0.4399279582473346,
+      "grad_norm": 2.4829085003143923,
+      "learning_rate": 3.0147153413406154e-06,
+      "loss": 1.0414,
+      "step": 4183
+    },
+    {
+      "epoch": 0.44003312868918193,
+      "grad_norm": 3.42576868411267,
+      "learning_rate": 3.013898378392441e-06,
+      "loss": 1.0368,
+      "step": 4184
+    },
+    {
+      "epoch": 0.4401382991310292,
+      "grad_norm": 2.252207757095223,
+      "learning_rate": 3.0130813581406044e-06,
+      "loss": 1.0187,
+      "step": 4185
+    },
+    {
+      "epoch": 0.4402434695728765,
+      "grad_norm": 2.5342374508667707,
+      "learning_rate": 3.0122642806762114e-06,
+      "loss": 0.9934,
+      "step": 4186
+    },
+    {
+      "epoch": 0.44034864001472385,
+      "grad_norm": 2.8671681025292246,
+      "learning_rate": 3.0114471460903714e-06,
+      "loss": 0.9852,
+      "step": 4187
+    },
+    {
+      "epoch": 0.44045381045657117,
+      "grad_norm": 2.5693510866729388,
+      "learning_rate": 3.0106299544742013e-06,
+      "loss": 1.0354,
+      "step": 4188
+    },
+    {
+      "epoch": 0.4405589808984185,
+      "grad_norm": 3.364495917648472,
+      "learning_rate": 3.0098127059188247e-06,
+      "loss": 1.0339,
+      "step": 4189
+    },
+    {
+      "epoch": 0.4406641513402658,
+      "grad_norm": 2.7990251477607146,
+      "learning_rate": 3.0089954005153706e-06,
+      "loss": 1.0116,
+      "step": 4190
+    },
+    {
+      "epoch": 0.44076932178211314,
+      "grad_norm": 1.7727662574506695,
+      "learning_rate": 3.0081780383549763e-06,
+      "loss": 1.012,
+      "step": 4191
+    },
+    {
+      "epoch": 0.44087449222396047,
+      "grad_norm": 3.2222272812715245,
+      "learning_rate": 3.0073606195287825e-06,
+      "loss": 0.9633,
+      "step": 4192
+    },
+    {
+      "epoch": 0.4409796626658078,
+      "grad_norm": 3.6290145756150722,
+      "learning_rate": 3.0065431441279386e-06,
+      "loss": 1.0499,
+      "step": 4193
+    },
+    {
+      "epoch": 0.4410848331076551,
+      "grad_norm": 3.1982957510410515,
+      "learning_rate": 3.005725612243599e-06,
+      "loss": 0.9487,
+      "step": 4194
+    },
+    {
+      "epoch": 0.4411900035495024,
+      "grad_norm": 2.8434896830214123,
+      "learning_rate": 3.0049080239669243e-06,
+      "loss": 1.0009,
+      "step": 4195
+    },
+    {
+      "epoch": 0.4412951739913497,
+      "grad_norm": 2.523909980989135,
+      "learning_rate": 3.0040903793890834e-06,
+      "loss": 1.0032,
+      "step": 4196
+    },
+    {
+      "epoch": 0.44140034443319703,
+      "grad_norm": 2.523114131380918,
+      "learning_rate": 3.003272678601249e-06,
+      "loss": 1.0288,
+      "step": 4197
+    },
+    {
+      "epoch": 0.44150551487504436,
+      "grad_norm": 2.8908811657988247,
+      "learning_rate": 3.002454921694602e-06,
+      "loss": 1.0062,
+      "step": 4198
+    },
+    {
+      "epoch": 0.4416106853168917,
+      "grad_norm": 2.7991620994148376,
+      "learning_rate": 3.0016371087603274e-06,
+      "loss": 1.0093,
+      "step": 4199
+    },
+    {
+      "epoch": 0.441715855758739,
+      "grad_norm": 2.6659000123176773,
+      "learning_rate": 3.0008192398896185e-06,
+      "loss": 1.0206,
+      "step": 4200
+    },
+    {
+      "epoch": 0.44182102620058633,
+      "grad_norm": 2.2057443650750863,
+      "learning_rate": 3.0000013151736747e-06,
+      "loss": 1.0263,
+      "step": 4201
+    },
+    {
+      "epoch": 0.44192619664243366,
+      "grad_norm": 2.7169424176961896,
+      "learning_rate": 2.999183334703699e-06,
+      "loss": 1.0041,
+      "step": 4202
+    },
+    {
+      "epoch": 0.442031367084281,
+      "grad_norm": 2.950358292631194,
+      "learning_rate": 2.9983652985709037e-06,
+      "loss": 0.9955,
+      "step": 4203
+    },
+    {
+      "epoch": 0.4421365375261283,
+      "grad_norm": 2.806264087299609,
+      "learning_rate": 2.9975472068665063e-06,
+      "loss": 1.0283,
+      "step": 4204
+    },
+    {
+      "epoch": 0.44224170796797563,
+      "grad_norm": 2.711842000947828,
+      "learning_rate": 2.9967290596817308e-06,
+      "loss": 1.0311,
+      "step": 4205
+    },
+    {
+      "epoch": 0.4423468784098229,
+      "grad_norm": 2.7709733812366664,
+      "learning_rate": 2.9959108571078056e-06,
+      "loss": 0.9947,
+      "step": 4206
+    },
+    {
+      "epoch": 0.4424520488516702,
+      "grad_norm": 3.2885009923549515,
+      "learning_rate": 2.995092599235968e-06,
+      "loss": 1.0179,
+      "step": 4207
+    },
+    {
+      "epoch": 0.44255721929351755,
+      "grad_norm": 2.955161655367558,
+      "learning_rate": 2.994274286157459e-06,
+      "loss": 0.9709,
+      "step": 4208
+    },
+    {
+      "epoch": 0.44266238973536487,
+      "grad_norm": 2.953888729382429,
+      "learning_rate": 2.9934559179635282e-06,
+      "loss": 1.0252,
+      "step": 4209
+    },
+    {
+      "epoch": 0.4427675601772122,
+      "grad_norm": 3.003042556969862,
+      "learning_rate": 2.992637494745429e-06,
+      "loss": 0.9944,
+      "step": 4210
+    },
+    {
+      "epoch": 0.4428727306190595,
+      "grad_norm": 2.7657547018738637,
+      "learning_rate": 2.9918190165944217e-06,
+      "loss": 0.9603,
+      "step": 4211
+    },
+    {
+      "epoch": 0.44297790106090684,
+      "grad_norm": 2.0161297281440613,
+      "learning_rate": 2.991000483601774e-06,
+      "loss": 1.0033,
+      "step": 4212
+    },
+    {
+      "epoch": 0.44308307150275417,
+      "grad_norm": 2.023154902922774,
+      "learning_rate": 2.9901818958587587e-06,
+      "loss": 0.9801,
+      "step": 4213
+    },
+    {
+      "epoch": 0.4431882419446015,
+      "grad_norm": 1.8645927902550326,
+      "learning_rate": 2.9893632534566534e-06,
+      "loss": 1.0054,
+      "step": 4214
+    },
+    {
+      "epoch": 0.4432934123864488,
+      "grad_norm": 3.5301225857507417,
+      "learning_rate": 2.9885445564867438e-06,
+      "loss": 0.9769,
+      "step": 4215
+    },
+    {
+      "epoch": 0.4433985828282961,
+      "grad_norm": 2.8515371337141744,
+      "learning_rate": 2.9877258050403214e-06,
+      "loss": 0.9968,
+      "step": 4216
+    },
+    {
+      "epoch": 0.4435037532701434,
+      "grad_norm": 2.404740251256232,
+      "learning_rate": 2.9869069992086825e-06,
+      "loss": 0.9942,
+      "step": 4217
+    },
+    {
+      "epoch": 0.44360892371199073,
+      "grad_norm": 2.046523753796662,
+      "learning_rate": 2.9860881390831303e-06,
+      "loss": 1.0065,
+      "step": 4218
+    },
+    {
+      "epoch": 0.44371409415383806,
+      "grad_norm": 3.103804612920373,
+      "learning_rate": 2.985269224754975e-06,
+      "loss": 1.061,
+      "step": 4219
+    },
+    {
+      "epoch": 0.4438192645956854,
+      "grad_norm": 2.221775249113264,
+      "learning_rate": 2.9844502563155324e-06,
+      "loss": 0.9729,
+      "step": 4220
+    },
+    {
+      "epoch": 0.4439244350375327,
+      "grad_norm": 2.36487818709531,
+      "learning_rate": 2.9836312338561223e-06,
+      "loss": 1.0165,
+      "step": 4221
+    },
+    {
+      "epoch": 0.44402960547938003,
+      "grad_norm": 2.5320619109662053,
+      "learning_rate": 2.9828121574680717e-06,
+      "loss": 1.0447,
+      "step": 4222
+    },
+    {
+      "epoch": 0.44413477592122735,
+      "grad_norm": 2.2988633433443515,
+      "learning_rate": 2.9819930272427162e-06,
+      "loss": 0.9668,
+      "step": 4223
+    },
+    {
+      "epoch": 0.4442399463630747,
+      "grad_norm": 2.9595824678582545,
+      "learning_rate": 2.981173843271393e-06,
+      "loss": 1.015,
+      "step": 4224
+    },
+    {
+      "epoch": 0.444345116804922,
+      "grad_norm": 2.950867108730787,
+      "learning_rate": 2.9803546056454487e-06,
+      "loss": 1.0077,
+      "step": 4225
+    },
+    {
+      "epoch": 0.44445028724676927,
+      "grad_norm": 1.6132738036228578,
+      "learning_rate": 2.9795353144562344e-06,
+      "loss": 0.9943,
+      "step": 4226
+    },
+    {
+      "epoch": 0.4445554576886166,
+      "grad_norm": 2.7813805140120333,
+      "learning_rate": 2.978715969795108e-06,
+      "loss": 0.9776,
+      "step": 4227
+    },
+    {
+      "epoch": 0.4446606281304639,
+      "grad_norm": 2.562516330449887,
+      "learning_rate": 2.9778965717534314e-06,
+      "loss": 0.9551,
+      "step": 4228
+    },
+    {
+      "epoch": 0.44476579857231124,
+      "grad_norm": 3.1737333083955583,
+      "learning_rate": 2.9770771204225744e-06,
+      "loss": 0.9925,
+      "step": 4229
+    },
+    {
+      "epoch": 0.44487096901415857,
+      "grad_norm": 1.8507674535018093,
+      "learning_rate": 2.9762576158939127e-06,
+      "loss": 1.0234,
+      "step": 4230
+    },
+    {
+      "epoch": 0.4449761394560059,
+      "grad_norm": 2.6261455818000408,
+      "learning_rate": 2.975438058258827e-06,
+      "loss": 1.0195,
+      "step": 4231
+    },
+    {
+      "epoch": 0.4450813098978532,
+      "grad_norm": 1.8245806335822785,
+      "learning_rate": 2.974618447608705e-06,
+      "loss": 0.9718,
+      "step": 4232
+    },
+    {
+      "epoch": 0.44518648033970054,
+      "grad_norm": 2.2479753999337815,
+      "learning_rate": 2.9737987840349393e-06,
+      "loss": 0.9852,
+      "step": 4233
+    },
+    {
+      "epoch": 0.44529165078154787,
+      "grad_norm": 2.35768351945659,
+      "learning_rate": 2.9729790676289276e-06,
+      "loss": 1.0047,
+      "step": 4234
+    },
+    {
+      "epoch": 0.4453968212233952,
+      "grad_norm": 1.6696931548669964,
+      "learning_rate": 2.972159298482076e-06,
+      "loss": 0.9698,
+      "step": 4235
+    },
+    {
+      "epoch": 0.44550199166524246,
+      "grad_norm": 2.0882289069459548,
+      "learning_rate": 2.971339476685795e-06,
+      "loss": 0.9712,
+      "step": 4236
+    },
+    {
+      "epoch": 0.4456071621070898,
+      "grad_norm": 2.093444735376746,
+      "learning_rate": 2.9705196023315007e-06,
+      "loss": 1.0109,
+      "step": 4237
+    },
+    {
+      "epoch": 0.4457123325489371,
+      "grad_norm": 2.5021177606762377,
+      "learning_rate": 2.9696996755106155e-06,
+      "loss": 1.0178,
+      "step": 4238
+    },
+    {
+      "epoch": 0.44581750299078443,
+      "grad_norm": 2.334943268674044,
+      "learning_rate": 2.968879696314568e-06,
+      "loss": 0.9909,
+      "step": 4239
+    },
+    {
+      "epoch": 0.44592267343263176,
+      "grad_norm": 2.3017506240577448,
+      "learning_rate": 2.968059664834792e-06,
+      "loss": 1.023,
+      "step": 4240
+    },
+    {
+      "epoch": 0.4460278438744791,
+      "grad_norm": 2.769841880116001,
+      "learning_rate": 2.967239581162727e-06,
+      "loss": 0.9995,
+      "step": 4241
+    },
+    {
+      "epoch": 0.4461330143163264,
+      "grad_norm": 3.7992104532494775,
+      "learning_rate": 2.96641944538982e-06,
+      "loss": 1.0083,
+      "step": 4242
+    },
+    {
+      "epoch": 0.44623818475817373,
+      "grad_norm": 2.1347949459797393,
+      "learning_rate": 2.9655992576075198e-06,
+      "loss": 0.9633,
+      "step": 4243
+    },
+    {
+      "epoch": 0.44634335520002105,
+      "grad_norm": 2.951928502125178,
+      "learning_rate": 2.964779017907287e-06,
+      "loss": 0.9934,
+      "step": 4244
+    },
+    {
+      "epoch": 0.4464485256418684,
+      "grad_norm": 2.263878251484875,
+      "learning_rate": 2.9639587263805824e-06,
+      "loss": 1.0103,
+      "step": 4245
+    },
+    {
+      "epoch": 0.44655369608371565,
+      "grad_norm": 3.2110964246131064,
+      "learning_rate": 2.963138383118876e-06,
+      "loss": 0.9281,
+      "step": 4246
+    },
+    {
+      "epoch": 0.44665886652556297,
+      "grad_norm": 2.7206342576353966,
+      "learning_rate": 2.962317988213642e-06,
+      "loss": 1.0208,
+      "step": 4247
+    },
+    {
+      "epoch": 0.4467640369674103,
+      "grad_norm": 2.569325435677048,
+      "learning_rate": 2.961497541756361e-06,
+      "loss": 0.9949,
+      "step": 4248
+    },
+    {
+      "epoch": 0.4468692074092576,
+      "grad_norm": 2.0106169690636726,
+      "learning_rate": 2.960677043838519e-06,
+      "loss": 1.0168,
+      "step": 4249
+    },
+    {
+      "epoch": 0.44697437785110494,
+      "grad_norm": 3.047440201325554,
+      "learning_rate": 2.959856494551608e-06,
+      "loss": 1.0501,
+      "step": 4250
+    },
+    {
+      "epoch": 0.44707954829295227,
+      "grad_norm": 2.9897469916545525,
+      "learning_rate": 2.9590358939871255e-06,
+      "loss": 1.0254,
+      "step": 4251
+    },
+    {
+      "epoch": 0.4471847187347996,
+      "grad_norm": 2.0389005489963674,
+      "learning_rate": 2.9582152422365745e-06,
+      "loss": 0.9759,
+      "step": 4252
+    },
+    {
+      "epoch": 0.4472898891766469,
+      "grad_norm": 2.791018390371004,
+      "learning_rate": 2.957394539391465e-06,
+      "loss": 1.0193,
+      "step": 4253
+    },
+    {
+      "epoch": 0.44739505961849424,
+      "grad_norm": 2.3959514096484473,
+      "learning_rate": 2.956573785543311e-06,
+      "loss": 1.0402,
+      "step": 4254
+    },
+    {
+      "epoch": 0.44750023006034156,
+      "grad_norm": 2.952041521825401,
+      "learning_rate": 2.955752980783633e-06,
+      "loss": 1.0327,
+      "step": 4255
+    },
+    {
+      "epoch": 0.44760540050218883,
+      "grad_norm": 2.834122618770755,
+      "learning_rate": 2.9549321252039577e-06,
+      "loss": 1.0195,
+      "step": 4256
+    },
+    {
+      "epoch": 0.44771057094403616,
+      "grad_norm": 2.6232922045756832,
+      "learning_rate": 2.954111218895816e-06,
+      "loss": 0.9946,
+      "step": 4257
+    },
+    {
+      "epoch": 0.4478157413858835,
+      "grad_norm": 2.890337682768083,
+      "learning_rate": 2.9532902619507465e-06,
+      "loss": 1.0015,
+      "step": 4258
+    },
+    {
+      "epoch": 0.4479209118277308,
+      "grad_norm": 2.0462008188888023,
+      "learning_rate": 2.95246925446029e-06,
+      "loss": 1.0015,
+      "step": 4259
+    },
+    {
+      "epoch": 0.44802608226957813,
+      "grad_norm": 2.3153362347848634,
+      "learning_rate": 2.951648196515998e-06,
+      "loss": 1.0159,
+      "step": 4260
+    },
+    {
+      "epoch": 0.44813125271142545,
+      "grad_norm": 3.294187755137225,
+      "learning_rate": 2.9508270882094227e-06,
+      "loss": 1.0014,
+      "step": 4261
+    },
+    {
+      "epoch": 0.4482364231532728,
+      "grad_norm": 2.8942766762795453,
+      "learning_rate": 2.9500059296321254e-06,
+      "loss": 1.015,
+      "step": 4262
+    },
+    {
+      "epoch": 0.4483415935951201,
+      "grad_norm": 2.8362356385644976,
+      "learning_rate": 2.9491847208756713e-06,
+      "loss": 1.0182,
+      "step": 4263
+    },
+    {
+      "epoch": 0.4484467640369674,
+      "grad_norm": 2.016987936853375,
+      "learning_rate": 2.9483634620316314e-06,
+      "loss": 0.9975,
+      "step": 4264
+    },
+    {
+      "epoch": 0.44855193447881475,
+      "grad_norm": 2.3257680119464226,
+      "learning_rate": 2.947542153191583e-06,
+      "loss": 0.9592,
+      "step": 4265
+    },
+    {
+      "epoch": 0.4486571049206621,
+      "grad_norm": 2.7528064005499218,
+      "learning_rate": 2.946720794447106e-06,
+      "loss": 1.0271,
+      "step": 4266
+    },
+    {
+      "epoch": 0.44876227536250934,
+      "grad_norm": 2.4897714526055252,
+      "learning_rate": 2.945899385889792e-06,
+      "loss": 0.9719,
+      "step": 4267
+    },
+    {
+      "epoch": 0.44886744580435667,
+      "grad_norm": 2.8675188349876675,
+      "learning_rate": 2.9450779276112313e-06,
+      "loss": 1.0455,
+      "step": 4268
+    },
+    {
+      "epoch": 0.448972616246204,
+      "grad_norm": 2.63772057013864,
+      "learning_rate": 2.944256419703025e-06,
+      "loss": 1.0008,
+      "step": 4269
+    },
+    {
+      "epoch": 0.4490777866880513,
+      "grad_norm": 2.767101795516205,
+      "learning_rate": 2.9434348622567773e-06,
+      "loss": 1.0132,
+      "step": 4270
+    },
+    {
+      "epoch": 0.44918295712989864,
+      "grad_norm": 2.8503438719512353,
+      "learning_rate": 2.942613255364097e-06,
+      "loss": 1.0182,
+      "step": 4271
+    },
+    {
+      "epoch": 0.44928812757174597,
+      "grad_norm": 3.264311463794955,
+      "learning_rate": 2.941791599116601e-06,
+      "loss": 0.9823,
+      "step": 4272
+    },
+    {
+      "epoch": 0.4493932980135933,
+      "grad_norm": 1.9470160801562246,
+      "learning_rate": 2.9409698936059083e-06,
+      "loss": 1.0072,
+      "step": 4273
+    },
+    {
+      "epoch": 0.4494984684554406,
+      "grad_norm": 2.578131298488729,
+      "learning_rate": 2.940148138923648e-06,
+      "loss": 0.9903,
+      "step": 4274
+    },
+    {
+      "epoch": 0.44960363889728794,
+      "grad_norm": 2.5479584159562036,
+      "learning_rate": 2.9393263351614503e-06,
+      "loss": 1.0249,
+      "step": 4275
+    },
+    {
+      "epoch": 0.44970880933913526,
+      "grad_norm": 2.3256148001185672,
+      "learning_rate": 2.9385044824109544e-06,
+      "loss": 1.0325,
+      "step": 4276
+    },
+    {
+      "epoch": 0.44981397978098253,
+      "grad_norm": 3.1496173601363933,
+      "learning_rate": 2.9376825807638016e-06,
+      "loss": 1.0063,
+      "step": 4277
+    },
+    {
+      "epoch": 0.44991915022282986,
+      "grad_norm": 2.851586489375221,
+      "learning_rate": 2.936860630311642e-06,
+      "loss": 1.029,
+      "step": 4278
+    },
+    {
+      "epoch": 0.4500243206646772,
+      "grad_norm": 3.088679797466413,
+      "learning_rate": 2.9360386311461276e-06,
+      "loss": 1.0058,
+      "step": 4279
+    },
+    {
+      "epoch": 0.4501294911065245,
+      "grad_norm": 2.054451946303782,
+      "learning_rate": 2.9352165833589188e-06,
+      "loss": 0.9734,
+      "step": 4280
+    },
+    {
+      "epoch": 0.45023466154837183,
+      "grad_norm": 2.718591132670405,
+      "learning_rate": 2.9343944870416798e-06,
+      "loss": 1.0024,
+      "step": 4281
+    },
+    {
+      "epoch": 0.45033983199021915,
+      "grad_norm": 2.746230280528559,
+      "learning_rate": 2.9335723422860807e-06,
+      "loss": 0.9915,
+      "step": 4282
+    },
+    {
+      "epoch": 0.4504450024320665,
+      "grad_norm": 2.537502797059733,
+      "learning_rate": 2.9327501491837977e-06,
+      "loss": 0.9917,
+      "step": 4283
+    },
+    {
+      "epoch": 0.4505501728739138,
+      "grad_norm": 3.1284325108868782,
+      "learning_rate": 2.931927907826511e-06,
+      "loss": 1.0431,
+      "step": 4284
+    },
+    {
+      "epoch": 0.4506553433157611,
+      "grad_norm": 1.812931921725533,
+      "learning_rate": 2.9311056183059085e-06,
+      "loss": 0.9778,
+      "step": 4285
+    },
+    {
+      "epoch": 0.45076051375760845,
+      "grad_norm": 2.851522179977043,
+      "learning_rate": 2.93028328071368e-06,
+      "loss": 0.9874,
+      "step": 4286
+    },
+    {
+      "epoch": 0.4508656841994557,
+      "grad_norm": 2.766819002757065,
+      "learning_rate": 2.9294608951415225e-06,
+      "loss": 0.9944,
+      "step": 4287
+    },
+    {
+      "epoch": 0.45097085464130304,
+      "grad_norm": 3.163681281030067,
+      "learning_rate": 2.9286384616811397e-06,
+      "loss": 1.031,
+      "step": 4288
+    },
+    {
+      "epoch": 0.45107602508315037,
+      "grad_norm": 2.485916557061216,
+      "learning_rate": 2.927815980424238e-06,
+      "loss": 1.0086,
+      "step": 4289
+    },
+    {
+      "epoch": 0.4511811955249977,
+      "grad_norm": 2.109206916955106,
+      "learning_rate": 2.926993451462532e-06,
+      "loss": 0.9913,
+      "step": 4290
+    },
+    {
+      "epoch": 0.451286365966845,
+      "grad_norm": 2.457120848888426,
+      "learning_rate": 2.926170874887738e-06,
+      "loss": 0.9919,
+      "step": 4291
+    },
+    {
+      "epoch": 0.45139153640869234,
+      "grad_norm": 2.303663013614732,
+      "learning_rate": 2.925348250791582e-06,
+      "loss": 0.9718,
+      "step": 4292
+    },
+    {
+      "epoch": 0.45149670685053966,
+      "grad_norm": 3.1484358472259335,
+      "learning_rate": 2.924525579265791e-06,
+      "loss": 0.9857,
+      "step": 4293
+    },
+    {
+      "epoch": 0.451601877292387,
+      "grad_norm": 3.189808105502437,
+      "learning_rate": 2.9237028604021008e-06,
+      "loss": 1.0218,
+      "step": 4294
+    },
+    {
+      "epoch": 0.4517070477342343,
+      "grad_norm": 2.7759170958312116,
+      "learning_rate": 2.9228800942922497e-06,
+      "loss": 1.0121,
+      "step": 4295
+    },
+    {
+      "epoch": 0.45181221817608164,
+      "grad_norm": 2.1847673720857,
+      "learning_rate": 2.922057281027983e-06,
+      "loss": 1.0119,
+      "step": 4296
+    },
+    {
+      "epoch": 0.4519173886179289,
+      "grad_norm": 2.2941536309406096,
+      "learning_rate": 2.921234420701051e-06,
+      "loss": 0.9933,
+      "step": 4297
+    },
+    {
+      "epoch": 0.45202255905977623,
+      "grad_norm": 1.7883355539808496,
+      "learning_rate": 2.9204115134032086e-06,
+      "loss": 0.9594,
+      "step": 4298
+    },
+    {
+      "epoch": 0.45212772950162355,
+      "grad_norm": 2.9394681120602506,
+      "learning_rate": 2.9195885592262167e-06,
+      "loss": 1.0014,
+      "step": 4299
+    },
+    {
+      "epoch": 0.4522328999434709,
+      "grad_norm": 3.0533456324247195,
+      "learning_rate": 2.9187655582618413e-06,
+      "loss": 0.9617,
+      "step": 4300
+    },
+    {
+      "epoch": 0.4523380703853182,
+      "grad_norm": 1.6982849296696754,
+      "learning_rate": 2.9179425106018532e-06,
+      "loss": 0.9429,
+      "step": 4301
+    },
+    {
+      "epoch": 0.4524432408271655,
+      "grad_norm": 3.1396740609931446,
+      "learning_rate": 2.9171194163380277e-06,
+      "loss": 1.0107,
+      "step": 4302
+    },
+    {
+      "epoch": 0.45254841126901285,
+      "grad_norm": 2.811681718875989,
+      "learning_rate": 2.916296275562147e-06,
+      "loss": 1.0572,
+      "step": 4303
+    },
+    {
+      "epoch": 0.4526535817108602,
+      "grad_norm": 3.2640718105886735,
+      "learning_rate": 2.9154730883659988e-06,
+      "loss": 1.025,
+      "step": 4304
+    },
+    {
+      "epoch": 0.4527587521527075,
+      "grad_norm": 2.69588504505022,
+      "learning_rate": 2.9146498548413725e-06,
+      "loss": 0.9978,
+      "step": 4305
+    },
+    {
+      "epoch": 0.4528639225945548,
+      "grad_norm": 2.2342874790888843,
+      "learning_rate": 2.913826575080067e-06,
+      "loss": 0.9304,
+      "step": 4306
+    },
+    {
+      "epoch": 0.4529690930364021,
+      "grad_norm": 3.0565709108799037,
+      "learning_rate": 2.9130032491738837e-06,
+      "loss": 1.0373,
+      "step": 4307
+    },
+    {
+      "epoch": 0.4530742634782494,
+      "grad_norm": 2.853116287639515,
+      "learning_rate": 2.9121798772146293e-06,
+      "loss": 1.0201,
+      "step": 4308
+    },
+    {
+      "epoch": 0.45317943392009674,
+      "grad_norm": 2.3744681471107607,
+      "learning_rate": 2.911356459294117e-06,
+      "loss": 0.9437,
+      "step": 4309
+    },
+    {
+      "epoch": 0.45328460436194407,
+      "grad_norm": 2.164104788250051,
+      "learning_rate": 2.910532995504163e-06,
+      "loss": 0.9791,
+      "step": 4310
+    },
+    {
+      "epoch": 0.4533897748037914,
+      "grad_norm": 1.6203464986260847,
+      "learning_rate": 2.9097094859365926e-06,
+      "loss": 0.9675,
+      "step": 4311
+    },
+    {
+      "epoch": 0.4534949452456387,
+      "grad_norm": 2.9303086296052934,
+      "learning_rate": 2.90888593068323e-06,
+      "loss": 1.0381,
+      "step": 4312
+    },
+    {
+      "epoch": 0.45360011568748604,
+      "grad_norm": 2.0626838173774376,
+      "learning_rate": 2.908062329835911e-06,
+      "loss": 0.9685,
+      "step": 4313
+    },
+    {
+      "epoch": 0.45370528612933336,
+      "grad_norm": 2.2697295104611115,
+      "learning_rate": 2.9072386834864723e-06,
+      "loss": 0.9784,
+      "step": 4314
+    },
+    {
+      "epoch": 0.4538104565711807,
+      "grad_norm": 3.0376685513758077,
+      "learning_rate": 2.9064149917267565e-06,
+      "loss": 1.0353,
+      "step": 4315
+    },
+    {
+      "epoch": 0.453915627013028,
+      "grad_norm": 2.657090557042862,
+      "learning_rate": 2.905591254648612e-06,
+      "loss": 0.9427,
+      "step": 4316
+    },
+    {
+      "epoch": 0.4540207974548753,
+      "grad_norm": 2.3929879751726872,
+      "learning_rate": 2.904767472343892e-06,
+      "loss": 1.0099,
+      "step": 4317
+    },
+    {
+      "epoch": 0.4541259678967226,
+      "grad_norm": 2.695393831318368,
+      "learning_rate": 2.9039436449044543e-06,
+      "loss": 1.0132,
+      "step": 4318
+    },
+    {
+      "epoch": 0.45423113833856993,
+      "grad_norm": 2.1469397224850417,
+      "learning_rate": 2.903119772422162e-06,
+      "loss": 0.9917,
+      "step": 4319
+    },
+    {
+      "epoch": 0.45433630878041725,
+      "grad_norm": 1.9890341493391972,
+      "learning_rate": 2.902295854988884e-06,
+      "loss": 1.0114,
+      "step": 4320
+    },
+    {
+      "epoch": 0.4544414792222646,
+      "grad_norm": 2.6430862936585617,
+      "learning_rate": 2.901471892696493e-06,
+      "loss": 1.0101,
+      "step": 4321
+    },
+    {
+      "epoch": 0.4545466496641119,
+      "grad_norm": 2.0147983041213253,
+      "learning_rate": 2.900647885636867e-06,
+      "loss": 1.0124,
+      "step": 4322
+    },
+    {
+      "epoch": 0.4546518201059592,
+      "grad_norm": 2.1685192867151057,
+      "learning_rate": 2.899823833901889e-06,
+      "loss": 1.0,
+      "step": 4323
+    },
+    {
+      "epoch": 0.45475699054780655,
+      "grad_norm": 2.332868281748152,
+      "learning_rate": 2.8989997375834485e-06,
+      "loss": 1.0148,
+      "step": 4324
+    },
+    {
+      "epoch": 0.4548621609896539,
+      "grad_norm": 1.3686107234017948,
+      "learning_rate": 2.8981755967734377e-06,
+      "loss": 0.9342,
+      "step": 4325
+    },
+    {
+      "epoch": 0.4549673314315012,
+      "grad_norm": 2.087757270716191,
+      "learning_rate": 2.8973514115637534e-06,
+      "loss": 0.9817,
+      "step": 4326
+    },
+    {
+      "epoch": 0.4550725018733485,
+      "grad_norm": 2.271350804832439,
+      "learning_rate": 2.8965271820463016e-06,
+      "loss": 1.0087,
+      "step": 4327
+    },
+    {
+      "epoch": 0.4551776723151958,
+      "grad_norm": 1.643573599294807,
+      "learning_rate": 2.895702908312987e-06,
+      "loss": 1.0125,
+      "step": 4328
+    },
+    {
+      "epoch": 0.4552828427570431,
+      "grad_norm": 2.554616326554526,
+      "learning_rate": 2.8948785904557262e-06,
+      "loss": 0.9899,
+      "step": 4329
+    },
+    {
+      "epoch": 0.45538801319889044,
+      "grad_norm": 2.91621589953988,
+      "learning_rate": 2.8940542285664337e-06,
+      "loss": 1.0189,
+      "step": 4330
+    },
+    {
+      "epoch": 0.45549318364073776,
+      "grad_norm": 1.9347991717626811,
+      "learning_rate": 2.8932298227370335e-06,
+      "loss": 0.966,
+      "step": 4331
+    },
+    {
+      "epoch": 0.4555983540825851,
+      "grad_norm": 3.5268488608131996,
+      "learning_rate": 2.8924053730594536e-06,
+      "loss": 0.9827,
+      "step": 4332
+    },
+    {
+      "epoch": 0.4557035245244324,
+      "grad_norm": 3.3032634100912515,
+      "learning_rate": 2.8915808796256268e-06,
+      "loss": 1.0037,
+      "step": 4333
+    },
+    {
+      "epoch": 0.45580869496627974,
+      "grad_norm": 2.594478528315312,
+      "learning_rate": 2.89075634252749e-06,
+      "loss": 0.9854,
+      "step": 4334
+    },
+    {
+      "epoch": 0.45591386540812706,
+      "grad_norm": 2.142538714749134,
+      "learning_rate": 2.8899317618569843e-06,
+      "loss": 1.0015,
+      "step": 4335
+    },
+    {
+      "epoch": 0.4560190358499744,
+      "grad_norm": 2.336631372632609,
+      "learning_rate": 2.889107137706059e-06,
+      "loss": 1.0257,
+      "step": 4336
+    },
+    {
+      "epoch": 0.4561242062918217,
+      "grad_norm": 3.5495497144192676,
+      "learning_rate": 2.8882824701666657e-06,
+      "loss": 1.0268,
+      "step": 4337
+    },
+    {
+      "epoch": 0.456229376733669,
+      "grad_norm": 1.8932999054002906,
+      "learning_rate": 2.88745775933076e-06,
+      "loss": 0.9928,
+      "step": 4338
+    },
+    {
+      "epoch": 0.4563345471755163,
+      "grad_norm": 3.0148596426527643,
+      "learning_rate": 2.8866330052903042e-06,
+      "loss": 1.024,
+      "step": 4339
+    },
+    {
+      "epoch": 0.4564397176173636,
+      "grad_norm": 2.0856993515410323,
+      "learning_rate": 2.885808208137265e-06,
+      "loss": 0.9969,
+      "step": 4340
+    },
+    {
+      "epoch": 0.45654488805921095,
+      "grad_norm": 2.095948181530454,
+      "learning_rate": 2.8849833679636137e-06,
+      "loss": 0.9824,
+      "step": 4341
+    },
+    {
+      "epoch": 0.4566500585010583,
+      "grad_norm": 2.5170281229756384,
+      "learning_rate": 2.8841584848613254e-06,
+      "loss": 1.002,
+      "step": 4342
+    },
+    {
+      "epoch": 0.4567552289429056,
+      "grad_norm": 2.1378769455831304,
+      "learning_rate": 2.883333558922383e-06,
+      "loss": 1.0271,
+      "step": 4343
+    },
+    {
+      "epoch": 0.4568603993847529,
+      "grad_norm": 2.3821344983844495,
+      "learning_rate": 2.88250859023877e-06,
+      "loss": 0.9948,
+      "step": 4344
+    },
+    {
+      "epoch": 0.45696556982660025,
+      "grad_norm": 2.5487726722527357,
+      "learning_rate": 2.8816835789024783e-06,
+      "loss": 1.0075,
+      "step": 4345
+    },
+    {
+      "epoch": 0.4570707402684476,
+      "grad_norm": 3.0550451214956023,
+      "learning_rate": 2.880858525005502e-06,
+      "loss": 1.0063,
+      "step": 4346
+    },
+    {
+      "epoch": 0.4571759107102949,
+      "grad_norm": 2.798668732759541,
+      "learning_rate": 2.880033428639842e-06,
+      "loss": 0.9896,
+      "step": 4347
+    },
+    {
+      "epoch": 0.45728108115214217,
+      "grad_norm": 2.980102984068875,
+      "learning_rate": 2.8792082898975028e-06,
+      "loss": 1.03,
+      "step": 4348
+    },
+    {
+      "epoch": 0.4573862515939895,
+      "grad_norm": 1.9173203998180592,
+      "learning_rate": 2.8783831088704923e-06,
+      "loss": 1.001,
+      "step": 4349
+    },
+    {
+      "epoch": 0.4574914220358368,
+      "grad_norm": 2.2818375318772843,
+      "learning_rate": 2.877557885650827e-06,
+      "loss": 1.0082,
+      "step": 4350
+    },
+    {
+      "epoch": 0.45759659247768414,
+      "grad_norm": 2.3754880401750844,
+      "learning_rate": 2.876732620330524e-06,
+      "loss": 1.0056,
+      "step": 4351
+    },
+    {
+      "epoch": 0.45770176291953146,
+      "grad_norm": 2.6543669930498495,
+      "learning_rate": 2.8759073130016073e-06,
+      "loss": 0.9431,
+      "step": 4352
+    },
+    {
+      "epoch": 0.4578069333613788,
+      "grad_norm": 2.6939766890582453,
+      "learning_rate": 2.8750819637561045e-06,
+      "loss": 0.9748,
+      "step": 4353
+    },
+    {
+      "epoch": 0.4579121038032261,
+      "grad_norm": 1.9175919955114296,
+      "learning_rate": 2.87425657268605e-06,
+      "loss": 1.0176,
+      "step": 4354
+    },
+    {
+      "epoch": 0.45801727424507344,
+      "grad_norm": 2.439681288815302,
+      "learning_rate": 2.873431139883479e-06,
+      "loss": 0.9385,
+      "step": 4355
+    },
+    {
+      "epoch": 0.45812244468692076,
+      "grad_norm": 2.8212284142415505,
+      "learning_rate": 2.872605665440436e-06,
+      "loss": 0.9857,
+      "step": 4356
+    },
+    {
+      "epoch": 0.4582276151287681,
+      "grad_norm": 2.1470640317341196,
+      "learning_rate": 2.8717801494489673e-06,
+      "loss": 0.9626,
+      "step": 4357
+    },
+    {
+      "epoch": 0.45833278557061535,
+      "grad_norm": 2.2492255805625514,
+      "learning_rate": 2.8709545920011233e-06,
+      "loss": 1.0123,
+      "step": 4358
+    },
+    {
+      "epoch": 0.4584379560124627,
+      "grad_norm": 2.0002199257891906,
+      "learning_rate": 2.8701289931889602e-06,
+      "loss": 1.0467,
+      "step": 4359
+    },
+    {
+      "epoch": 0.45854312645431,
+      "grad_norm": 2.981959655065841,
+      "learning_rate": 2.8693033531045395e-06,
+      "loss": 0.953,
+      "step": 4360
+    },
+    {
+      "epoch": 0.4586482968961573,
+      "grad_norm": 3.851394505625027,
+      "learning_rate": 2.868477671839926e-06,
+      "loss": 1.0175,
+      "step": 4361
+    },
+    {
+      "epoch": 0.45875346733800465,
+      "grad_norm": 2.31415083896268,
+      "learning_rate": 2.867651949487189e-06,
+      "loss": 1.0186,
+      "step": 4362
+    },
+    {
+      "epoch": 0.458858637779852,
+      "grad_norm": 3.8867757736496027,
+      "learning_rate": 2.8668261861384045e-06,
+      "loss": 1.0274,
+      "step": 4363
+    },
+    {
+      "epoch": 0.4589638082216993,
+      "grad_norm": 2.1989997123351555,
+      "learning_rate": 2.8660003818856506e-06,
+      "loss": 0.9769,
+      "step": 4364
+    },
+    {
+      "epoch": 0.4590689786635466,
+      "grad_norm": 2.0392697344733546,
+      "learning_rate": 2.865174536821011e-06,
+      "loss": 0.9493,
+      "step": 4365
+    },
+    {
+      "epoch": 0.45917414910539395,
+      "grad_norm": 2.4063204089586097,
+      "learning_rate": 2.864348651036574e-06,
+      "loss": 0.9577,
+      "step": 4366
+    },
+    {
+      "epoch": 0.45927931954724127,
+      "grad_norm": 2.285261233874211,
+      "learning_rate": 2.8635227246244306e-06,
+      "loss": 0.9948,
+      "step": 4367
+    },
+    {
+      "epoch": 0.45938448998908854,
+      "grad_norm": 2.2549083660343383,
+      "learning_rate": 2.8626967576766808e-06,
+      "loss": 1.024,
+      "step": 4368
+    },
+    {
+      "epoch": 0.45948966043093586,
+      "grad_norm": 2.408292400780605,
+      "learning_rate": 2.8618707502854243e-06,
+      "loss": 0.9919,
+      "step": 4369
+    },
+    {
+      "epoch": 0.4595948308727832,
+      "grad_norm": 2.147791593115681,
+      "learning_rate": 2.8610447025427685e-06,
+      "loss": 0.9478,
+      "step": 4370
+    },
+    {
+      "epoch": 0.4597000013146305,
+      "grad_norm": 2.0763303161056292,
+      "learning_rate": 2.8602186145408235e-06,
+      "loss": 1.0489,
+      "step": 4371
+    },
+    {
+      "epoch": 0.45980517175647784,
+      "grad_norm": 2.849591364090468,
+      "learning_rate": 2.859392486371705e-06,
+      "loss": 0.989,
+      "step": 4372
+    },
+    {
+      "epoch": 0.45991034219832516,
+      "grad_norm": 2.003930986515399,
+      "learning_rate": 2.858566318127532e-06,
+      "loss": 0.9761,
+      "step": 4373
+    },
+    {
+      "epoch": 0.4600155126401725,
+      "grad_norm": 1.9945701603167976,
+      "learning_rate": 2.8577401099004285e-06,
+      "loss": 1.0204,
+      "step": 4374
+    },
+    {
+      "epoch": 0.4601206830820198,
+      "grad_norm": 3.075579516233079,
+      "learning_rate": 2.856913861782525e-06,
+      "loss": 1.0004,
+      "step": 4375
+    },
+    {
+      "epoch": 0.46022585352386713,
+      "grad_norm": 2.351283254441647,
+      "learning_rate": 2.856087573865952e-06,
+      "loss": 1.0144,
+      "step": 4376
+    },
+    {
+      "epoch": 0.46033102396571446,
+      "grad_norm": 3.339158339951501,
+      "learning_rate": 2.8552612462428497e-06,
+      "loss": 1.006,
+      "step": 4377
+    },
+    {
+      "epoch": 0.4604361944075617,
+      "grad_norm": 2.309435332828068,
+      "learning_rate": 2.854434879005357e-06,
+      "loss": 1.0221,
+      "step": 4378
+    },
+    {
+      "epoch": 0.46054136484940905,
+      "grad_norm": 3.2211925972738333,
+      "learning_rate": 2.853608472245624e-06,
+      "loss": 1.0109,
+      "step": 4379
+    },
+    {
+      "epoch": 0.4606465352912564,
+      "grad_norm": 2.1892482332044327,
+      "learning_rate": 2.8527820260557986e-06,
+      "loss": 1.0088,
+      "step": 4380
+    },
+    {
+      "epoch": 0.4607517057331037,
+      "grad_norm": 2.410757032252132,
+      "learning_rate": 2.851955540528036e-06,
+      "loss": 0.9959,
+      "step": 4381
+    },
+    {
+      "epoch": 0.460856876174951,
+      "grad_norm": 2.359864321372521,
+      "learning_rate": 2.8511290157544976e-06,
+      "loss": 1.0216,
+      "step": 4382
+    },
+    {
+      "epoch": 0.46096204661679835,
+      "grad_norm": 2.1948898074829355,
+      "learning_rate": 2.8503024518273455e-06,
+      "loss": 0.989,
+      "step": 4383
+    },
+    {
+      "epoch": 0.4610672170586457,
+      "grad_norm": 2.4426656861817997,
+      "learning_rate": 2.849475848838749e-06,
+      "loss": 0.9701,
+      "step": 4384
+    },
+    {
+      "epoch": 0.461172387500493,
+      "grad_norm": 2.5319847009537946,
+      "learning_rate": 2.84864920688088e-06,
+      "loss": 1.0233,
+      "step": 4385
+    },
+    {
+      "epoch": 0.4612775579423403,
+      "grad_norm": 2.133328823397027,
+      "learning_rate": 2.847822526045917e-06,
+      "loss": 0.9656,
+      "step": 4386
+    },
+    {
+      "epoch": 0.46138272838418765,
+      "grad_norm": 2.639337179377181,
+      "learning_rate": 2.8469958064260405e-06,
+      "loss": 1.0025,
+      "step": 4387
+    },
+    {
+      "epoch": 0.46148789882603497,
+      "grad_norm": 2.0901303479901516,
+      "learning_rate": 2.846169048113435e-06,
+      "loss": 0.9886,
+      "step": 4388
+    },
+    {
+      "epoch": 0.46159306926788224,
+      "grad_norm": 2.7143229057209957,
+      "learning_rate": 2.8453422512002925e-06,
+      "loss": 1.0069,
+      "step": 4389
+    },
+    {
+      "epoch": 0.46169823970972956,
+      "grad_norm": 2.2128312082996158,
+      "learning_rate": 2.844515415778806e-06,
+      "loss": 1.0177,
+      "step": 4390
+    },
+    {
+      "epoch": 0.4618034101515769,
+      "grad_norm": 2.1207444779689135,
+      "learning_rate": 2.843688541941174e-06,
+      "loss": 1.0304,
+      "step": 4391
+    },
+    {
+      "epoch": 0.4619085805934242,
+      "grad_norm": 2.6321023824250025,
+      "learning_rate": 2.8428616297795998e-06,
+      "loss": 0.9876,
+      "step": 4392
+    },
+    {
+      "epoch": 0.46201375103527154,
+      "grad_norm": 2.5462159065092242,
+      "learning_rate": 2.84203467938629e-06,
+      "loss": 1.0227,
+      "step": 4393
+    },
+    {
+      "epoch": 0.46211892147711886,
+      "grad_norm": 3.258389125088761,
+      "learning_rate": 2.8412076908534574e-06,
+      "loss": 0.983,
+      "step": 4394
+    },
+    {
+      "epoch": 0.4622240919189662,
+      "grad_norm": 2.3212982515302603,
+      "learning_rate": 2.840380664273316e-06,
+      "loss": 0.9358,
+      "step": 4395
+    },
+    {
+      "epoch": 0.4623292623608135,
+      "grad_norm": 2.4763075667324657,
+      "learning_rate": 2.839553599738087e-06,
+      "loss": 1.0202,
+      "step": 4396
+    },
+    {
+      "epoch": 0.46243443280266083,
+      "grad_norm": 2.0484739974347534,
+      "learning_rate": 2.838726497339993e-06,
+      "loss": 1.0134,
+      "step": 4397
+    },
+    {
+      "epoch": 0.46253960324450816,
+      "grad_norm": 1.9402963476386015,
+      "learning_rate": 2.8378993571712638e-06,
+      "loss": 0.9729,
+      "step": 4398
+    },
+    {
+      "epoch": 0.4626447736863554,
+      "grad_norm": 1.8355127361037,
+      "learning_rate": 2.8370721793241314e-06,
+      "loss": 1.0101,
+      "step": 4399
+    },
+    {
+      "epoch": 0.46274994412820275,
+      "grad_norm": 2.994356739022344,
+      "learning_rate": 2.8362449638908324e-06,
+      "loss": 1.0008,
+      "step": 4400
+    },
+    {
+      "epoch": 0.4628551145700501,
+      "grad_norm": 2.910116749399226,
+      "learning_rate": 2.8354177109636076e-06,
+      "loss": 0.994,
+      "step": 4401
+    },
+    {
+      "epoch": 0.4629602850118974,
+      "grad_norm": 2.223042107881437,
+      "learning_rate": 2.834590420634703e-06,
+      "loss": 1.016,
+      "step": 4402
+    },
+    {
+      "epoch": 0.4630654554537447,
+      "grad_norm": 2.6324446425874073,
+      "learning_rate": 2.8337630929963672e-06,
+      "loss": 1.0376,
+      "step": 4403
+    },
+    {
+      "epoch": 0.46317062589559205,
+      "grad_norm": 2.8252122929662153,
+      "learning_rate": 2.832935728140853e-06,
+      "loss": 0.9432,
+      "step": 4404
+    },
+    {
+      "epoch": 0.46327579633743937,
+      "grad_norm": 2.0038921955516105,
+      "learning_rate": 2.83210832616042e-06,
+      "loss": 1.0019,
+      "step": 4405
+    },
+    {
+      "epoch": 0.4633809667792867,
+      "grad_norm": 2.549003738454214,
+      "learning_rate": 2.8312808871473275e-06,
+      "loss": 1.0029,
+      "step": 4406
+    },
+    {
+      "epoch": 0.463486137221134,
+      "grad_norm": 2.2351716118016474,
+      "learning_rate": 2.830453411193843e-06,
+      "loss": 0.9699,
+      "step": 4407
+    },
+    {
+      "epoch": 0.46359130766298134,
+      "grad_norm": 2.8680061904201173,
+      "learning_rate": 2.829625898392237e-06,
+      "loss": 1.0212,
+      "step": 4408
+    },
+    {
+      "epoch": 0.4636964781048286,
+      "grad_norm": 2.0089360810313908,
+      "learning_rate": 2.828798348834782e-06,
+      "loss": 1.0066,
+      "step": 4409
+    },
+    {
+      "epoch": 0.46380164854667594,
+      "grad_norm": 2.127532937133814,
+      "learning_rate": 2.827970762613757e-06,
+      "loss": 0.9945,
+      "step": 4410
+    },
+    {
+      "epoch": 0.46390681898852326,
+      "grad_norm": 2.089558020765553,
+      "learning_rate": 2.827143139821444e-06,
+      "loss": 1.0153,
+      "step": 4411
+    },
+    {
+      "epoch": 0.4640119894303706,
+      "grad_norm": 3.252402836184958,
+      "learning_rate": 2.82631548055013e-06,
+      "loss": 0.9513,
+      "step": 4412
+    },
+    {
+      "epoch": 0.4641171598722179,
+      "grad_norm": 2.135962227586247,
+      "learning_rate": 2.8254877848921036e-06,
+      "loss": 0.9607,
+      "step": 4413
+    },
+    {
+      "epoch": 0.46422233031406523,
+      "grad_norm": 2.41667417366253,
+      "learning_rate": 2.824660052939662e-06,
+      "loss": 0.9908,
+      "step": 4414
+    },
+    {
+      "epoch": 0.46432750075591256,
+      "grad_norm": 2.380362760812115,
+      "learning_rate": 2.823832284785102e-06,
+      "loss": 0.9779,
+      "step": 4415
+    },
+    {
+      "epoch": 0.4644326711977599,
+      "grad_norm": 2.0345880558081944,
+      "learning_rate": 2.8230044805207275e-06,
+      "loss": 0.9734,
+      "step": 4416
+    },
+    {
+      "epoch": 0.4645378416396072,
+      "grad_norm": 2.3167139827947776,
+      "learning_rate": 2.8221766402388436e-06,
+      "loss": 0.96,
+      "step": 4417
+    },
+    {
+      "epoch": 0.46464301208145453,
+      "grad_norm": 2.3878293198539224,
+      "learning_rate": 2.8213487640317615e-06,
+      "loss": 0.97,
+      "step": 4418
+    },
+    {
+      "epoch": 0.4647481825233018,
+      "grad_norm": 2.1973785313368426,
+      "learning_rate": 2.820520851991796e-06,
+      "loss": 0.9816,
+      "step": 4419
+    },
+    {
+      "epoch": 0.4648533529651491,
+      "grad_norm": 3.0125702231483262,
+      "learning_rate": 2.8196929042112652e-06,
+      "loss": 1.0055,
+      "step": 4420
+    },
+    {
+      "epoch": 0.46495852340699645,
+      "grad_norm": 2.3918412139996,
+      "learning_rate": 2.8188649207824925e-06,
+      "loss": 0.9632,
+      "step": 4421
+    },
+    {
+      "epoch": 0.4650636938488438,
+      "grad_norm": 1.9278006868749584,
+      "learning_rate": 2.8180369017978037e-06,
+      "loss": 0.9645,
+      "step": 4422
+    },
+    {
+      "epoch": 0.4651688642906911,
+      "grad_norm": 2.85997981754498,
+      "learning_rate": 2.8172088473495306e-06,
+      "loss": 0.9861,
+      "step": 4423
+    },
+    {
+      "epoch": 0.4652740347325384,
+      "grad_norm": 2.4986220629370344,
+      "learning_rate": 2.816380757530006e-06,
+      "loss": 0.957,
+      "step": 4424
+    },
+    {
+      "epoch": 0.46537920517438575,
+      "grad_norm": 2.3157951259747316,
+      "learning_rate": 2.8155526324315704e-06,
+      "loss": 0.9893,
+      "step": 4425
+    },
+    {
+      "epoch": 0.46548437561623307,
+      "grad_norm": 2.9956280856636384,
+      "learning_rate": 2.814724472146564e-06,
+      "loss": 1.0325,
+      "step": 4426
+    },
+    {
+      "epoch": 0.4655895460580804,
+      "grad_norm": 2.7269340487252345,
+      "learning_rate": 2.813896276767334e-06,
+      "loss": 0.9351,
+      "step": 4427
+    },
+    {
+      "epoch": 0.4656947164999277,
+      "grad_norm": 2.4606920746541014,
+      "learning_rate": 2.8130680463862315e-06,
+      "loss": 1.0123,
+      "step": 4428
+    },
+    {
+      "epoch": 0.465799886941775,
+      "grad_norm": 2.848249960105425,
+      "learning_rate": 2.8122397810956086e-06,
+      "loss": 0.9924,
+      "step": 4429
+    },
+    {
+      "epoch": 0.4659050573836223,
+      "grad_norm": 2.8446842951304783,
+      "learning_rate": 2.811411480987825e-06,
+      "loss": 1.0287,
+      "step": 4430
+    },
+    {
+      "epoch": 0.46601022782546964,
+      "grad_norm": 2.393650778819736,
+      "learning_rate": 2.810583146155243e-06,
+      "loss": 1.0055,
+      "step": 4431
+    },
+    {
+      "epoch": 0.46611539826731696,
+      "grad_norm": 3.0125679338832065,
+      "learning_rate": 2.8097547766902273e-06,
+      "loss": 0.9402,
+      "step": 4432
+    },
+    {
+      "epoch": 0.4662205687091643,
+      "grad_norm": 2.4891599650912886,
+      "learning_rate": 2.8089263726851474e-06,
+      "loss": 0.9905,
+      "step": 4433
+    },
+    {
+      "epoch": 0.4663257391510116,
+      "grad_norm": 2.6495533894705616,
+      "learning_rate": 2.8080979342323765e-06,
+      "loss": 0.9873,
+      "step": 4434
+    },
+    {
+      "epoch": 0.46643090959285893,
+      "grad_norm": 2.5985892014306375,
+      "learning_rate": 2.8072694614242935e-06,
+      "loss": 1.0312,
+      "step": 4435
+    },
+    {
+      "epoch": 0.46653608003470626,
+      "grad_norm": 2.636695159781973,
+      "learning_rate": 2.8064409543532776e-06,
+      "loss": 1.0396,
+      "step": 4436
+    },
+    {
+      "epoch": 0.4666412504765536,
+      "grad_norm": 3.244510424950399,
+      "learning_rate": 2.805612413111716e-06,
+      "loss": 0.9766,
+      "step": 4437
+    },
+    {
+      "epoch": 0.4667464209184009,
+      "grad_norm": 2.2447570199099154,
+      "learning_rate": 2.8047838377919952e-06,
+      "loss": 0.9901,
+      "step": 4438
+    },
+    {
+      "epoch": 0.4668515913602482,
+      "grad_norm": 2.5976977350088415,
+      "learning_rate": 2.8039552284865094e-06,
+      "loss": 0.9973,
+      "step": 4439
+    },
+    {
+      "epoch": 0.4669567618020955,
+      "grad_norm": 3.2454932031526944,
+      "learning_rate": 2.8031265852876537e-06,
+      "loss": 1.0331,
+      "step": 4440
+    },
+    {
+      "epoch": 0.4670619322439428,
+      "grad_norm": 2.523586811336105,
+      "learning_rate": 2.8022979082878297e-06,
+      "loss": 0.9869,
+      "step": 4441
+    },
+    {
+      "epoch": 0.46716710268579015,
+      "grad_norm": 2.302211346608485,
+      "learning_rate": 2.801469197579441e-06,
+      "loss": 1.0341,
+      "step": 4442
+    },
+    {
+      "epoch": 0.46727227312763747,
+      "grad_norm": 2.348259230268178,
+      "learning_rate": 2.800640453254894e-06,
+      "loss": 0.9683,
+      "step": 4443
+    },
+    {
+      "epoch": 0.4673774435694848,
+      "grad_norm": 2.366754132665982,
+      "learning_rate": 2.799811675406601e-06,
+      "loss": 0.9587,
+      "step": 4444
+    },
+    {
+      "epoch": 0.4674826140113321,
+      "grad_norm": 2.075592911455706,
+      "learning_rate": 2.7989828641269778e-06,
+      "loss": 1.0085,
+      "step": 4445
+    },
+    {
+      "epoch": 0.46758778445317944,
+      "grad_norm": 2.6326964601948877,
+      "learning_rate": 2.798154019508443e-06,
+      "loss": 0.9919,
+      "step": 4446
+    },
+    {
+      "epoch": 0.46769295489502677,
+      "grad_norm": 2.8302081829382835,
+      "learning_rate": 2.7973251416434176e-06,
+      "loss": 1.0534,
+      "step": 4447
+    },
+    {
+      "epoch": 0.4677981253368741,
+      "grad_norm": 2.639359868066465,
+      "learning_rate": 2.7964962306243305e-06,
+      "loss": 1.0037,
+      "step": 4448
+    },
+    {
+      "epoch": 0.4679032957787214,
+      "grad_norm": 2.292040628888014,
+      "learning_rate": 2.7956672865436095e-06,
+      "loss": 1.0244,
+      "step": 4449
+    },
+    {
+      "epoch": 0.4680084662205687,
+      "grad_norm": 3.4389539874184334,
+      "learning_rate": 2.794838309493689e-06,
+      "loss": 1.0291,
+      "step": 4450
+    },
+    {
+      "epoch": 0.468113636662416,
+      "grad_norm": 4.035449967348666,
+      "learning_rate": 2.7940092995670075e-06,
+      "loss": 1.0501,
+      "step": 4451
+    },
+    {
+      "epoch": 0.46821880710426333,
+      "grad_norm": 2.7831049279874653,
+      "learning_rate": 2.7931802568560053e-06,
+      "loss": 1.0194,
+      "step": 4452
+    },
+    {
+      "epoch": 0.46832397754611066,
+      "grad_norm": 1.8206417464439855,
+      "learning_rate": 2.792351181453126e-06,
+      "loss": 0.9495,
+      "step": 4453
+    },
+    {
+      "epoch": 0.468429147987958,
+      "grad_norm": 1.9541921059032041,
+      "learning_rate": 2.791522073450819e-06,
+      "loss": 1.0107,
+      "step": 4454
+    },
+    {
+      "epoch": 0.4685343184298053,
+      "grad_norm": 2.8892598367431086,
+      "learning_rate": 2.790692932941537e-06,
+      "loss": 1.0612,
+      "step": 4455
+    },
+    {
+      "epoch": 0.46863948887165263,
+      "grad_norm": 3.171614211998949,
+      "learning_rate": 2.7898637600177343e-06,
+      "loss": 1.0538,
+      "step": 4456
+    },
+    {
+      "epoch": 0.46874465931349996,
+      "grad_norm": 2.424092647751607,
+      "learning_rate": 2.7890345547718707e-06,
+      "loss": 1.0326,
+      "step": 4457
+    },
+    {
+      "epoch": 0.4688498297553473,
+      "grad_norm": 1.8602198107845023,
+      "learning_rate": 2.7882053172964085e-06,
+      "loss": 0.9663,
+      "step": 4458
+    },
+    {
+      "epoch": 0.4689550001971946,
+      "grad_norm": 2.3246688695244226,
+      "learning_rate": 2.787376047683815e-06,
+      "loss": 0.9459,
+      "step": 4459
+    },
+    {
+      "epoch": 0.4690601706390419,
+      "grad_norm": 2.762737851787094,
+      "learning_rate": 2.78654674602656e-06,
+      "loss": 1.0057,
+      "step": 4460
+    },
+    {
+      "epoch": 0.4691653410808892,
+      "grad_norm": 2.3224208280852427,
+      "learning_rate": 2.7857174124171166e-06,
+      "loss": 0.988,
+      "step": 4461
+    },
+    {
+      "epoch": 0.4692705115227365,
+      "grad_norm": 1.6891392684017292,
+      "learning_rate": 2.7848880469479623e-06,
+      "loss": 1.0095,
+      "step": 4462
+    },
+    {
+      "epoch": 0.46937568196458385,
+      "grad_norm": 2.3788041616512814,
+      "learning_rate": 2.7840586497115767e-06,
+      "loss": 1.0111,
+      "step": 4463
+    },
+    {
+      "epoch": 0.46948085240643117,
+      "grad_norm": 2.5464866689627996,
+      "learning_rate": 2.783229220800446e-06,
+      "loss": 0.9748,
+      "step": 4464
+    },
+    {
+      "epoch": 0.4695860228482785,
+      "grad_norm": 2.6905461256581114,
+      "learning_rate": 2.7823997603070573e-06,
+      "loss": 1.0209,
+      "step": 4465
+    },
+    {
+      "epoch": 0.4696911932901258,
+      "grad_norm": 2.767917863313156,
+      "learning_rate": 2.7815702683239002e-06,
+      "loss": 1.0259,
+      "step": 4466
+    },
+    {
+      "epoch": 0.46979636373197314,
+      "grad_norm": 2.347477742569025,
+      "learning_rate": 2.7807407449434726e-06,
+      "loss": 0.9851,
+      "step": 4467
+    },
+    {
+      "epoch": 0.46990153417382047,
+      "grad_norm": 1.90764637954435,
+      "learning_rate": 2.7799111902582697e-06,
+      "loss": 1.0211,
+      "step": 4468
+    },
+    {
+      "epoch": 0.4700067046156678,
+      "grad_norm": 2.0075620650917667,
+      "learning_rate": 2.7790816043607953e-06,
+      "loss": 0.987,
+      "step": 4469
+    },
+    {
+      "epoch": 0.47011187505751506,
+      "grad_norm": 2.7682956389063973,
+      "learning_rate": 2.7782519873435538e-06,
+      "loss": 0.9998,
+      "step": 4470
+    },
+    {
+      "epoch": 0.4702170454993624,
+      "grad_norm": 2.292023839054441,
+      "learning_rate": 2.7774223392990544e-06,
+      "loss": 1.0166,
+      "step": 4471
+    },
+    {
+      "epoch": 0.4703222159412097,
+      "grad_norm": 3.2215154014236353,
+      "learning_rate": 2.776592660319809e-06,
+      "loss": 1.0095,
+      "step": 4472
+    },
+    {
+      "epoch": 0.47042738638305703,
+      "grad_norm": 2.719209058008712,
+      "learning_rate": 2.775762950498333e-06,
+      "loss": 0.9651,
+      "step": 4473
+    },
+    {
+      "epoch": 0.47053255682490436,
+      "grad_norm": 3.210513336735668,
+      "learning_rate": 2.7749332099271466e-06,
+      "loss": 1.0112,
+      "step": 4474
+    },
+    {
+      "epoch": 0.4706377272667517,
+      "grad_norm": 2.6300858424014715,
+      "learning_rate": 2.7741034386987707e-06,
+      "loss": 1.0085,
+      "step": 4475
+    },
+    {
+      "epoch": 0.470742897708599,
+      "grad_norm": 3.14495333627856,
+      "learning_rate": 2.7732736369057326e-06,
+      "loss": 0.9815,
+      "step": 4476
+    },
+    {
+      "epoch": 0.47084806815044633,
+      "grad_norm": 2.5526520052973973,
+      "learning_rate": 2.772443804640561e-06,
+      "loss": 0.9963,
+      "step": 4477
+    },
+    {
+      "epoch": 0.47095323859229365,
+      "grad_norm": 1.9617055562355878,
+      "learning_rate": 2.771613941995789e-06,
+      "loss": 1.0017,
+      "step": 4478
+    },
+    {
+      "epoch": 0.471058409034141,
+      "grad_norm": 2.688578495945201,
+      "learning_rate": 2.7707840490639517e-06,
+      "loss": 0.9938,
+      "step": 4479
+    },
+    {
+      "epoch": 0.47116357947598825,
+      "grad_norm": 2.4345239857502046,
+      "learning_rate": 2.76995412593759e-06,
+      "loss": 0.996,
+      "step": 4480
+    },
+    {
+      "epoch": 0.47126874991783557,
+      "grad_norm": 3.6250029270038393,
+      "learning_rate": 2.769124172709246e-06,
+      "loss": 0.9917,
+      "step": 4481
+    },
+    {
+      "epoch": 0.4713739203596829,
+      "grad_norm": 2.2005515373005156,
+      "learning_rate": 2.7682941894714664e-06,
+      "loss": 0.9691,
+      "step": 4482
+    },
+    {
+      "epoch": 0.4714790908015302,
+      "grad_norm": 2.3295932635087224,
+      "learning_rate": 2.7674641763168003e-06,
+      "loss": 1.017,
+      "step": 4483
+    },
+    {
+      "epoch": 0.47158426124337754,
+      "grad_norm": 3.0234424271551417,
+      "learning_rate": 2.7666341333378005e-06,
+      "loss": 0.9685,
+      "step": 4484
+    },
+    {
+      "epoch": 0.47168943168522487,
+      "grad_norm": 2.486438862191606,
+      "learning_rate": 2.7658040606270244e-06,
+      "loss": 0.998,
+      "step": 4485
+    },
+    {
+      "epoch": 0.4717946021270722,
+      "grad_norm": 2.675116684244302,
+      "learning_rate": 2.7649739582770297e-06,
+      "loss": 1.0076,
+      "step": 4486
+    },
+    {
+      "epoch": 0.4718997725689195,
+      "grad_norm": 2.7072141444203908,
+      "learning_rate": 2.7641438263803804e-06,
+      "loss": 1.0033,
+      "step": 4487
+    },
+    {
+      "epoch": 0.47200494301076684,
+      "grad_norm": 2.5861797482973516,
+      "learning_rate": 2.7633136650296434e-06,
+      "loss": 0.9744,
+      "step": 4488
+    },
+    {
+      "epoch": 0.47211011345261417,
+      "grad_norm": 2.224720554082226,
+      "learning_rate": 2.762483474317387e-06,
+      "loss": 1.0217,
+      "step": 4489
+    },
+    {
+      "epoch": 0.47221528389446144,
+      "grad_norm": 2.245262509992629,
+      "learning_rate": 2.7616532543361834e-06,
+      "loss": 0.9912,
+      "step": 4490
+    },
+    {
+      "epoch": 0.47232045433630876,
+      "grad_norm": 2.7604293363117023,
+      "learning_rate": 2.7608230051786094e-06,
+      "loss": 0.9939,
+      "step": 4491
+    },
+    {
+      "epoch": 0.4724256247781561,
+      "grad_norm": 2.4111390221536224,
+      "learning_rate": 2.7599927269372453e-06,
+      "loss": 0.9487,
+      "step": 4492
+    },
+    {
+      "epoch": 0.4725307952200034,
+      "grad_norm": 1.891935956225179,
+      "learning_rate": 2.759162419704671e-06,
+      "loss": 0.9603,
+      "step": 4493
+    },
+    {
+      "epoch": 0.47263596566185073,
+      "grad_norm": 1.8123928584801083,
+      "learning_rate": 2.7583320835734743e-06,
+      "loss": 0.9951,
+      "step": 4494
+    },
+    {
+      "epoch": 0.47274113610369806,
+      "grad_norm": 2.3993983584972343,
+      "learning_rate": 2.757501718636244e-06,
+      "loss": 0.9939,
+      "step": 4495
+    },
+    {
+      "epoch": 0.4728463065455454,
+      "grad_norm": 2.0288832292698746,
+      "learning_rate": 2.7566713249855715e-06,
+      "loss": 1.0161,
+      "step": 4496
+    },
+    {
+      "epoch": 0.4729514769873927,
+      "grad_norm": 2.323445513583333,
+      "learning_rate": 2.7558409027140525e-06,
+      "loss": 0.9886,
+      "step": 4497
+    },
+    {
+      "epoch": 0.47305664742924003,
+      "grad_norm": 2.68445355500332,
+      "learning_rate": 2.7550104519142846e-06,
+      "loss": 0.9494,
+      "step": 4498
+    },
+    {
+      "epoch": 0.47316181787108735,
+      "grad_norm": 2.640136937856992,
+      "learning_rate": 2.754179972678871e-06,
+      "loss": 1.0562,
+      "step": 4499
+    },
+    {
+      "epoch": 0.4732669883129346,
+      "grad_norm": 3.324792657639246,
+      "learning_rate": 2.753349465100415e-06,
+      "loss": 0.9979,
+      "step": 4500
+    },
+    {
+      "epoch": 0.47337215875478195,
+      "grad_norm": 2.810264675470383,
+      "learning_rate": 2.7525189292715264e-06,
+      "loss": 1.0266,
+      "step": 4501
+    },
+    {
+      "epoch": 0.47347732919662927,
+      "grad_norm": 2.003647833018789,
+      "learning_rate": 2.751688365284816e-06,
+      "loss": 1.0359,
+      "step": 4502
+    },
+    {
+      "epoch": 0.4735824996384766,
+      "grad_norm": 2.7806320532702697,
+      "learning_rate": 2.7508577732328975e-06,
+      "loss": 1.0402,
+      "step": 4503
+    },
+    {
+      "epoch": 0.4736876700803239,
+      "grad_norm": 1.713725247330321,
+      "learning_rate": 2.750027153208388e-06,
+      "loss": 1.0154,
+      "step": 4504
+    },
+    {
+      "epoch": 0.47379284052217124,
+      "grad_norm": 2.8013510184260864,
+      "learning_rate": 2.7491965053039084e-06,
+      "loss": 1.026,
+      "step": 4505
+    },
+    {
+      "epoch": 0.47389801096401857,
+      "grad_norm": 2.7346431525668917,
+      "learning_rate": 2.7483658296120828e-06,
+      "loss": 1.0076,
+      "step": 4506
+    },
+    {
+      "epoch": 0.4740031814058659,
+      "grad_norm": 2.805919154601316,
+      "learning_rate": 2.747535126225538e-06,
+      "loss": 0.9873,
+      "step": 4507
+    },
+    {
+      "epoch": 0.4741083518477132,
+      "grad_norm": 2.950564613008157,
+      "learning_rate": 2.746704395236904e-06,
+      "loss": 1.0062,
+      "step": 4508
+    },
+    {
+      "epoch": 0.47421352228956054,
+      "grad_norm": 2.751482814652587,
+      "learning_rate": 2.745873636738813e-06,
+      "loss": 0.9653,
+      "step": 4509
+    },
+    {
+      "epoch": 0.47431869273140786,
+      "grad_norm": 2.5665028667710517,
+      "learning_rate": 2.7450428508239024e-06,
+      "loss": 0.9883,
+      "step": 4510
+    },
+    {
+      "epoch": 0.47442386317325513,
+      "grad_norm": 3.3938183374193205,
+      "learning_rate": 2.7442120375848096e-06,
+      "loss": 0.9968,
+      "step": 4511
+    },
+    {
+      "epoch": 0.47452903361510246,
+      "grad_norm": 2.1396244469160774,
+      "learning_rate": 2.7433811971141772e-06,
+      "loss": 0.964,
+      "step": 4512
+    },
+    {
+      "epoch": 0.4746342040569498,
+      "grad_norm": 2.34919376628943,
+      "learning_rate": 2.742550329504651e-06,
+      "loss": 0.9981,
+      "step": 4513
+    },
+    {
+      "epoch": 0.4747393744987971,
+      "grad_norm": 2.9534207402307864,
+      "learning_rate": 2.7417194348488786e-06,
+      "loss": 0.9802,
+      "step": 4514
+    },
+    {
+      "epoch": 0.47484454494064443,
+      "grad_norm": 3.164085329529094,
+      "learning_rate": 2.7408885132395117e-06,
+      "loss": 1.0135,
+      "step": 4515
+    },
+    {
+      "epoch": 0.47494971538249176,
+      "grad_norm": 2.2533801639022406,
+      "learning_rate": 2.7400575647692046e-06,
+      "loss": 0.9697,
+      "step": 4516
+    },
+    {
+      "epoch": 0.4750548858243391,
+      "grad_norm": 2.362830237698888,
+      "learning_rate": 2.7392265895306142e-06,
+      "loss": 1.0129,
+      "step": 4517
+    },
+    {
+      "epoch": 0.4751600562661864,
+      "grad_norm": 3.088808161086836,
+      "learning_rate": 2.7383955876164004e-06,
+      "loss": 0.9999,
+      "step": 4518
+    },
+    {
+      "epoch": 0.47526522670803373,
+      "grad_norm": 3.1253457323179767,
+      "learning_rate": 2.737564559119227e-06,
+      "loss": 1.0031,
+      "step": 4519
+    },
+    {
+      "epoch": 0.47537039714988105,
+      "grad_norm": 2.366491108309749,
+      "learning_rate": 2.7367335041317593e-06,
+      "loss": 1.0039,
+      "step": 4520
+    },
+    {
+      "epoch": 0.4754755675917283,
+      "grad_norm": 2.451828923555723,
+      "learning_rate": 2.7359024227466668e-06,
+      "loss": 0.9815,
+      "step": 4521
+    },
+    {
+      "epoch": 0.47558073803357565,
+      "grad_norm": 2.0943618962936945,
+      "learning_rate": 2.735071315056622e-06,
+      "loss": 0.985,
+      "step": 4522
+    },
+    {
+      "epoch": 0.47568590847542297,
+      "grad_norm": 2.978757177852028,
+      "learning_rate": 2.734240181154299e-06,
+      "loss": 1.0017,
+      "step": 4523
+    },
+    {
+      "epoch": 0.4757910789172703,
+      "grad_norm": 2.880549752555551,
+      "learning_rate": 2.733409021132377e-06,
+      "loss": 0.9694,
+      "step": 4524
+    },
+    {
+      "epoch": 0.4758962493591176,
+      "grad_norm": 1.9750573431598037,
+      "learning_rate": 2.7325778350835353e-06,
+      "loss": 0.9964,
+      "step": 4525
+    },
+    {
+      "epoch": 0.47600141980096494,
+      "grad_norm": 2.672025240947568,
+      "learning_rate": 2.7317466231004584e-06,
+      "loss": 0.9814,
+      "step": 4526
+    },
+    {
+      "epoch": 0.47610659024281227,
+      "grad_norm": 2.397919630397351,
+      "learning_rate": 2.7309153852758335e-06,
+      "loss": 1.0414,
+      "step": 4527
+    },
+    {
+      "epoch": 0.4762117606846596,
+      "grad_norm": 3.589638718925268,
+      "learning_rate": 2.730084121702348e-06,
+      "loss": 0.9689,
+      "step": 4528
+    },
+    {
+      "epoch": 0.4763169311265069,
+      "grad_norm": 2.383301359474508,
+      "learning_rate": 2.7292528324726963e-06,
+      "loss": 0.9526,
+      "step": 4529
+    },
+    {
+      "epoch": 0.47642210156835424,
+      "grad_norm": 1.7141669642552233,
+      "learning_rate": 2.7284215176795724e-06,
+      "loss": 0.9945,
+      "step": 4530
+    },
+    {
+      "epoch": 0.4765272720102015,
+      "grad_norm": 2.2151545678563083,
+      "learning_rate": 2.7275901774156753e-06,
+      "loss": 1.0156,
+      "step": 4531
+    },
+    {
+      "epoch": 0.47663244245204883,
+      "grad_norm": 2.362263151552294,
+      "learning_rate": 2.726758811773706e-06,
+      "loss": 0.9864,
+      "step": 4532
+    },
+    {
+      "epoch": 0.47673761289389616,
+      "grad_norm": 1.719772591675931,
+      "learning_rate": 2.725927420846367e-06,
+      "loss": 0.9876,
+      "step": 4533
+    },
+    {
+      "epoch": 0.4768427833357435,
+      "grad_norm": 2.803143954819945,
+      "learning_rate": 2.7250960047263658e-06,
+      "loss": 1.0204,
+      "step": 4534
+    },
+    {
+      "epoch": 0.4769479537775908,
+      "grad_norm": 1.9487171453914527,
+      "learning_rate": 2.7242645635064107e-06,
+      "loss": 0.9932,
+      "step": 4535
+    },
+    {
+      "epoch": 0.47705312421943813,
+      "grad_norm": 2.208339642159797,
+      "learning_rate": 2.7234330972792157e-06,
+      "loss": 0.919,
+      "step": 4536
+    },
+    {
+      "epoch": 0.47715829466128545,
+      "grad_norm": 2.2756389370959207,
+      "learning_rate": 2.7226016061374934e-06,
+      "loss": 1.0314,
+      "step": 4537
+    },
+    {
+      "epoch": 0.4772634651031328,
+      "grad_norm": 3.099273899280179,
+      "learning_rate": 2.7217700901739637e-06,
+      "loss": 1.0085,
+      "step": 4538
+    },
+    {
+      "epoch": 0.4773686355449801,
+      "grad_norm": 2.6724241504223354,
+      "learning_rate": 2.720938549481346e-06,
+      "loss": 0.9733,
+      "step": 4539
+    },
+    {
+      "epoch": 0.4774738059868274,
+      "grad_norm": 2.502055605870524,
+      "learning_rate": 2.720106984152364e-06,
+      "loss": 0.9623,
+      "step": 4540
+    },
+    {
+      "epoch": 0.4775789764286747,
+      "grad_norm": 2.0664420087785245,
+      "learning_rate": 2.719275394279743e-06,
+      "loss": 1.0268,
+      "step": 4541
+    },
+    {
+      "epoch": 0.477684146870522,
+      "grad_norm": 1.8769297433106866,
+      "learning_rate": 2.718443779956212e-06,
+      "loss": 1.0,
+      "step": 4542
+    },
+    {
+      "epoch": 0.47778931731236934,
+      "grad_norm": 2.458314650760072,
+      "learning_rate": 2.7176121412745026e-06,
+      "loss": 1.0313,
+      "step": 4543
+    },
+    {
+      "epoch": 0.47789448775421667,
+      "grad_norm": 2.630804381758795,
+      "learning_rate": 2.7167804783273495e-06,
+      "loss": 0.9763,
+      "step": 4544
+    },
+    {
+      "epoch": 0.477999658196064,
+      "grad_norm": 2.9835995375477418,
+      "learning_rate": 2.715948791207489e-06,
+      "loss": 0.9961,
+      "step": 4545
+    },
+    {
+      "epoch": 0.4781048286379113,
+      "grad_norm": 2.5063558471282446,
+      "learning_rate": 2.7151170800076603e-06,
+      "loss": 1.0074,
+      "step": 4546
+    },
+    {
+      "epoch": 0.47820999907975864,
+      "grad_norm": 2.827999520063029,
+      "learning_rate": 2.714285344820606e-06,
+      "loss": 0.9995,
+      "step": 4547
+    },
+    {
+      "epoch": 0.47831516952160597,
+      "grad_norm": 2.8368404828358442,
+      "learning_rate": 2.7134535857390714e-06,
+      "loss": 0.9858,
+      "step": 4548
+    },
+    {
+      "epoch": 0.4784203399634533,
+      "grad_norm": 2.807519023318031,
+      "learning_rate": 2.7126218028558037e-06,
+      "loss": 0.9945,
+      "step": 4549
+    },
+    {
+      "epoch": 0.4785255104053006,
+      "grad_norm": 2.3253069567766764,
+      "learning_rate": 2.711789996263554e-06,
+      "loss": 0.9841,
+      "step": 4550
+    },
+    {
+      "epoch": 0.4786306808471479,
+      "grad_norm": 2.400222728107831,
+      "learning_rate": 2.7109581660550733e-06,
+      "loss": 0.992,
+      "step": 4551
+    },
+    {
+      "epoch": 0.4787358512889952,
+      "grad_norm": 1.8434853526832968,
+      "learning_rate": 2.710126312323119e-06,
+      "loss": 0.9798,
+      "step": 4552
+    },
+    {
+      "epoch": 0.47884102173084253,
+      "grad_norm": 2.339371298345462,
+      "learning_rate": 2.7092944351604482e-06,
+      "loss": 0.9666,
+      "step": 4553
+    },
+    {
+      "epoch": 0.47894619217268986,
+      "grad_norm": 2.0644879283853754,
+      "learning_rate": 2.7084625346598232e-06,
+      "loss": 0.9738,
+      "step": 4554
+    },
+    {
+      "epoch": 0.4790513626145372,
+      "grad_norm": 3.0372309396395987,
+      "learning_rate": 2.707630610914005e-06,
+      "loss": 0.9611,
+      "step": 4555
+    },
+    {
+      "epoch": 0.4791565330563845,
+      "grad_norm": 2.7578247233652973,
+      "learning_rate": 2.706798664015761e-06,
+      "loss": 1.0389,
+      "step": 4556
+    },
+    {
+      "epoch": 0.47926170349823183,
+      "grad_norm": 3.0623439608759,
+      "learning_rate": 2.70596669405786e-06,
+      "loss": 1.0041,
+      "step": 4557
+    },
+    {
+      "epoch": 0.47936687394007915,
+      "grad_norm": 2.4941925421930433,
+      "learning_rate": 2.7051347011330716e-06,
+      "loss": 0.9829,
+      "step": 4558
+    },
+    {
+      "epoch": 0.4794720443819265,
+      "grad_norm": 3.0490408838472103,
+      "learning_rate": 2.704302685334171e-06,
+      "loss": 1.0366,
+      "step": 4559
+    },
+    {
+      "epoch": 0.4795772148237738,
+      "grad_norm": 2.3813045311909553,
+      "learning_rate": 2.7034706467539335e-06,
+      "loss": 0.9829,
+      "step": 4560
+    },
+    {
+      "epoch": 0.47968238526562107,
+      "grad_norm": 2.845538555861447,
+      "learning_rate": 2.702638585485139e-06,
+      "loss": 0.9978,
+      "step": 4561
+    },
+    {
+      "epoch": 0.4797875557074684,
+      "grad_norm": 3.2438971299974955,
+      "learning_rate": 2.701806501620568e-06,
+      "loss": 1.0011,
+      "step": 4562
+    },
+    {
+      "epoch": 0.4798927261493157,
+      "grad_norm": 2.3425851673049585,
+      "learning_rate": 2.700974395253004e-06,
+      "loss": 1.0052,
+      "step": 4563
+    },
+    {
+      "epoch": 0.47999789659116304,
+      "grad_norm": 2.8884679088128324,
+      "learning_rate": 2.7001422664752338e-06,
+      "loss": 0.969,
+      "step": 4564
+    },
+    {
+      "epoch": 0.48010306703301037,
+      "grad_norm": 2.7955695336255615,
+      "learning_rate": 2.699310115380046e-06,
+      "loss": 1.0197,
+      "step": 4565
+    },
+    {
+      "epoch": 0.4802082374748577,
+      "grad_norm": 3.190960632752079,
+      "learning_rate": 2.6984779420602324e-06,
+      "loss": 1.0072,
+      "step": 4566
+    },
+    {
+      "epoch": 0.480313407916705,
+      "grad_norm": 2.5976595875210506,
+      "learning_rate": 2.697645746608586e-06,
+      "loss": 0.9822,
+      "step": 4567
+    },
+    {
+      "epoch": 0.48041857835855234,
+      "grad_norm": 2.0169697804572664,
+      "learning_rate": 2.6968135291179036e-06,
+      "loss": 0.9938,
+      "step": 4568
+    },
+    {
+      "epoch": 0.48052374880039966,
+      "grad_norm": 2.9736885221988794,
+      "learning_rate": 2.6959812896809843e-06,
+      "loss": 1.0062,
+      "step": 4569
+    },
+    {
+      "epoch": 0.480628919242247,
+      "grad_norm": 1.4651386680966676,
+      "learning_rate": 2.6951490283906285e-06,
+      "loss": 0.9801,
+      "step": 4570
+    },
+    {
+      "epoch": 0.4807340896840943,
+      "grad_norm": 2.0786679973928197,
+      "learning_rate": 2.6943167453396397e-06,
+      "loss": 1.0099,
+      "step": 4571
+    },
+    {
+      "epoch": 0.4808392601259416,
+      "grad_norm": 2.3231766743941784,
+      "learning_rate": 2.6934844406208243e-06,
+      "loss": 0.9628,
+      "step": 4572
+    },
+    {
+      "epoch": 0.4809444305677889,
+      "grad_norm": 2.358885296144915,
+      "learning_rate": 2.6926521143269914e-06,
+      "loss": 1.016,
+      "step": 4573
+    },
+    {
+      "epoch": 0.48104960100963623,
+      "grad_norm": 2.4699479211845197,
+      "learning_rate": 2.6918197665509506e-06,
+      "loss": 0.9918,
+      "step": 4574
+    },
+    {
+      "epoch": 0.48115477145148355,
+      "grad_norm": 2.3154126074282173,
+      "learning_rate": 2.690987397385516e-06,
+      "loss": 1.0321,
+      "step": 4575
+    },
+    {
+      "epoch": 0.4812599418933309,
+      "grad_norm": 2.6809121807728493,
+      "learning_rate": 2.690155006923503e-06,
+      "loss": 1.0223,
+      "step": 4576
+    },
+    {
+      "epoch": 0.4813651123351782,
+      "grad_norm": 2.8409825087081644,
+      "learning_rate": 2.6893225952577294e-06,
+      "loss": 1.0256,
+      "step": 4577
+    },
+    {
+      "epoch": 0.4814702827770255,
+      "grad_norm": 2.235796045311807,
+      "learning_rate": 2.688490162481015e-06,
+      "loss": 1.0431,
+      "step": 4578
+    },
+    {
+      "epoch": 0.48157545321887285,
+      "grad_norm": 2.3738201066374747,
+      "learning_rate": 2.6876577086861844e-06,
+      "loss": 1.0046,
+      "step": 4579
+    },
+    {
+      "epoch": 0.4816806236607202,
+      "grad_norm": 2.2521790012523413,
+      "learning_rate": 2.686825233966061e-06,
+      "loss": 1.0065,
+      "step": 4580
+    },
+    {
+      "epoch": 0.4817857941025675,
+      "grad_norm": 2.680999712329763,
+      "learning_rate": 2.6859927384134727e-06,
+      "loss": 0.9777,
+      "step": 4581
+    },
+    {
+      "epoch": 0.48189096454441477,
+      "grad_norm": 2.4999906862595678,
+      "learning_rate": 2.685160222121249e-06,
+      "loss": 1.0089,
+      "step": 4582
+    },
+    {
+      "epoch": 0.4819961349862621,
+      "grad_norm": 1.5800457189959451,
+      "learning_rate": 2.6843276851822233e-06,
+      "loss": 0.9998,
+      "step": 4583
+    },
+    {
+      "epoch": 0.4821013054281094,
+      "grad_norm": 2.6596489372561725,
+      "learning_rate": 2.6834951276892273e-06,
+      "loss": 0.9915,
+      "step": 4584
+    },
+    {
+      "epoch": 0.48220647586995674,
+      "grad_norm": 2.701721201218792,
+      "learning_rate": 2.6826625497351e-06,
+      "loss": 0.9542,
+      "step": 4585
+    },
+    {
+      "epoch": 0.48231164631180407,
+      "grad_norm": 2.755401995048588,
+      "learning_rate": 2.681829951412679e-06,
+      "loss": 1.0366,
+      "step": 4586
+    },
+    {
+      "epoch": 0.4824168167536514,
+      "grad_norm": 2.6480098597209576,
+      "learning_rate": 2.6809973328148058e-06,
+      "loss": 0.9671,
+      "step": 4587
+    },
+    {
+      "epoch": 0.4825219871954987,
+      "grad_norm": 3.376948552074513,
+      "learning_rate": 2.6801646940343245e-06,
+      "loss": 1.0068,
+      "step": 4588
+    },
+    {
+      "epoch": 0.48262715763734604,
+      "grad_norm": 2.6687667258804013,
+      "learning_rate": 2.67933203516408e-06,
+      "loss": 0.9519,
+      "step": 4589
+    },
+    {
+      "epoch": 0.48273232807919336,
+      "grad_norm": 1.648325858154635,
+      "learning_rate": 2.6784993562969207e-06,
+      "loss": 1.0098,
+      "step": 4590
+    },
+    {
+      "epoch": 0.4828374985210407,
+      "grad_norm": 2.22475778779079,
+      "learning_rate": 2.6776666575256965e-06,
+      "loss": 1.0006,
+      "step": 4591
+    },
+    {
+      "epoch": 0.48294266896288796,
+      "grad_norm": 2.3818188632021093,
+      "learning_rate": 2.676833938943259e-06,
+      "loss": 1.0092,
+      "step": 4592
+    },
+    {
+      "epoch": 0.4830478394047353,
+      "grad_norm": 2.408034896184432,
+      "learning_rate": 2.6760012006424647e-06,
+      "loss": 0.998,
+      "step": 4593
+    },
+    {
+      "epoch": 0.4831530098465826,
+      "grad_norm": 1.8677903443913986,
+      "learning_rate": 2.6751684427161684e-06,
+      "loss": 0.9938,
+      "step": 4594
+    },
+    {
+      "epoch": 0.48325818028842993,
+      "grad_norm": 3.2959095169625856,
+      "learning_rate": 2.6743356652572304e-06,
+      "loss": 0.9984,
+      "step": 4595
+    },
+    {
+      "epoch": 0.48336335073027725,
+      "grad_norm": 3.1404363237116604,
+      "learning_rate": 2.673502868358512e-06,
+      "loss": 1.0293,
+      "step": 4596
+    },
+    {
+      "epoch": 0.4834685211721246,
+      "grad_norm": 2.5720636875148197,
+      "learning_rate": 2.6726700521128757e-06,
+      "loss": 1.0043,
+      "step": 4597
+    },
+    {
+      "epoch": 0.4835736916139719,
+      "grad_norm": 2.363912157861966,
+      "learning_rate": 2.671837216613187e-06,
+      "loss": 0.974,
+      "step": 4598
+    },
+    {
+      "epoch": 0.4836788620558192,
+      "grad_norm": 2.618997975520033,
+      "learning_rate": 2.6710043619523128e-06,
+      "loss": 1.0115,
+      "step": 4599
+    },
+    {
+      "epoch": 0.48378403249766655,
+      "grad_norm": 2.2993164157541632,
+      "learning_rate": 2.6701714882231256e-06,
+      "loss": 0.9936,
+      "step": 4600
+    },
+    {
+      "epoch": 0.4838892029395139,
+      "grad_norm": 2.7590745627058864,
+      "learning_rate": 2.669338595518494e-06,
+      "loss": 0.9784,
+      "step": 4601
+    },
+    {
+      "epoch": 0.48399437338136114,
+      "grad_norm": 2.7995862197909447,
+      "learning_rate": 2.6685056839312944e-06,
+      "loss": 0.976,
+      "step": 4602
+    },
+    {
+      "epoch": 0.48409954382320847,
+      "grad_norm": 2.571103128838185,
+      "learning_rate": 2.667672753554402e-06,
+      "loss": 0.9439,
+      "step": 4603
+    },
+    {
+      "epoch": 0.4842047142650558,
+      "grad_norm": 1.8408434670986793,
+      "learning_rate": 2.6668398044806946e-06,
+      "loss": 0.9928,
+      "step": 4604
+    },
+    {
+      "epoch": 0.4843098847069031,
+      "grad_norm": 2.135680773980519,
+      "learning_rate": 2.666006836803054e-06,
+      "loss": 1.0166,
+      "step": 4605
+    },
+    {
+      "epoch": 0.48441505514875044,
+      "grad_norm": 2.2802759432661186,
+      "learning_rate": 2.6651738506143603e-06,
+      "loss": 1.0261,
+      "step": 4606
+    },
+    {
+      "epoch": 0.48452022559059776,
+      "grad_norm": 2.220392952522567,
+      "learning_rate": 2.6643408460075e-06,
+      "loss": 1.0148,
+      "step": 4607
+    },
+    {
+      "epoch": 0.4846253960324451,
+      "grad_norm": 2.5358133229575097,
+      "learning_rate": 2.663507823075358e-06,
+      "loss": 0.9698,
+      "step": 4608
+    },
+    {
+      "epoch": 0.4847305664742924,
+      "grad_norm": 2.375319857666318,
+      "learning_rate": 2.662674781910824e-06,
+      "loss": 0.965,
+      "step": 4609
+    },
+    {
+      "epoch": 0.48483573691613974,
+      "grad_norm": 2.213027783818853,
+      "learning_rate": 2.6618417226067877e-06,
+      "loss": 1.0248,
+      "step": 4610
+    },
+    {
+      "epoch": 0.48494090735798706,
+      "grad_norm": 2.148489365280874,
+      "learning_rate": 2.6610086452561423e-06,
+      "loss": 0.9983,
+      "step": 4611
+    },
+    {
+      "epoch": 0.48504607779983433,
+      "grad_norm": 2.5409516705940627,
+      "learning_rate": 2.6601755499517826e-06,
+      "loss": 0.9994,
+      "step": 4612
+    },
+    {
+      "epoch": 0.48515124824168165,
+      "grad_norm": 2.363278403171672,
+      "learning_rate": 2.6593424367866042e-06,
+      "loss": 1.0018,
+      "step": 4613
+    },
+    {
+      "epoch": 0.485256418683529,
+      "grad_norm": 2.1133831998111945,
+      "learning_rate": 2.658509305853507e-06,
+      "loss": 1.0488,
+      "step": 4614
+    },
+    {
+      "epoch": 0.4853615891253763,
+      "grad_norm": 2.42236880112027,
+      "learning_rate": 2.6576761572453903e-06,
+      "loss": 0.9917,
+      "step": 4615
+    },
+    {
+      "epoch": 0.4854667595672236,
+      "grad_norm": 2.772025975679734,
+      "learning_rate": 2.6568429910551574e-06,
+      "loss": 1.026,
+      "step": 4616
+    },
+    {
+      "epoch": 0.48557193000907095,
+      "grad_norm": 2.5143222232504487,
+      "learning_rate": 2.6560098073757122e-06,
+      "loss": 0.9878,
+      "step": 4617
+    },
+    {
+      "epoch": 0.4856771004509183,
+      "grad_norm": 2.073400372306121,
+      "learning_rate": 2.6551766062999624e-06,
+      "loss": 0.9997,
+      "step": 4618
+    },
+    {
+      "epoch": 0.4857822708927656,
+      "grad_norm": 1.9864972586645486,
+      "learning_rate": 2.6543433879208147e-06,
+      "loss": 1.0091,
+      "step": 4619
+    },
+    {
+      "epoch": 0.4858874413346129,
+      "grad_norm": 2.751681846656289,
+      "learning_rate": 2.6535101523311806e-06,
+      "loss": 0.9822,
+      "step": 4620
+    },
+    {
+      "epoch": 0.48599261177646025,
+      "grad_norm": 2.313188102376116,
+      "learning_rate": 2.6526768996239725e-06,
+      "loss": 0.9828,
+      "step": 4621
+    },
+    {
+      "epoch": 0.4860977822183075,
+      "grad_norm": 1.8303821979796089,
+      "learning_rate": 2.651843629892103e-06,
+      "loss": 1.0083,
+      "step": 4622
+    },
+    {
+      "epoch": 0.48620295266015484,
+      "grad_norm": 1.6838185182278091,
+      "learning_rate": 2.65101034322849e-06,
+      "loss": 1.0225,
+      "step": 4623
+    },
+    {
+      "epoch": 0.48630812310200217,
+      "grad_norm": 2.3240841743762193,
+      "learning_rate": 2.6501770397260503e-06,
+      "loss": 0.9677,
+      "step": 4624
+    },
+    {
+      "epoch": 0.4864132935438495,
+      "grad_norm": 2.0762876706178477,
+      "learning_rate": 2.6493437194777038e-06,
+      "loss": 0.9767,
+      "step": 4625
+    },
+    {
+      "epoch": 0.4865184639856968,
+      "grad_norm": 2.7261027597048373,
+      "learning_rate": 2.648510382576373e-06,
+      "loss": 1.0191,
+      "step": 4626
+    },
+    {
+      "epoch": 0.48662363442754414,
+      "grad_norm": 2.5618337156491235,
+      "learning_rate": 2.647677029114981e-06,
+      "loss": 0.9701,
+      "step": 4627
+    },
+    {
+      "epoch": 0.48672880486939146,
+      "grad_norm": 2.703008948964744,
+      "learning_rate": 2.6468436591864534e-06,
+      "loss": 1.0215,
+      "step": 4628
+    },
+    {
+      "epoch": 0.4868339753112388,
+      "grad_norm": 2.0399542166728675,
+      "learning_rate": 2.6460102728837156e-06,
+      "loss": 1.0245,
+      "step": 4629
+    },
+    {
+      "epoch": 0.4869391457530861,
+      "grad_norm": 1.561628488644531,
+      "learning_rate": 2.6451768702996987e-06,
+      "loss": 0.9581,
+      "step": 4630
+    },
+    {
+      "epoch": 0.48704431619493344,
+      "grad_norm": 1.8892703105994535,
+      "learning_rate": 2.644343451527333e-06,
+      "loss": 0.9982,
+      "step": 4631
+    },
+    {
+      "epoch": 0.48714948663678076,
+      "grad_norm": 2.6663357784839676,
+      "learning_rate": 2.6435100166595516e-06,
+      "loss": 0.9515,
+      "step": 4632
+    },
+    {
+      "epoch": 0.48725465707862803,
+      "grad_norm": 2.71454659482501,
+      "learning_rate": 2.6426765657892883e-06,
+      "loss": 1.0505,
+      "step": 4633
+    },
+    {
+      "epoch": 0.48735982752047535,
+      "grad_norm": 2.4346295714541246,
+      "learning_rate": 2.64184309900948e-06,
+      "loss": 0.9909,
+      "step": 4634
+    },
+    {
+      "epoch": 0.4874649979623227,
+      "grad_norm": 2.884081836500342,
+      "learning_rate": 2.641009616413064e-06,
+      "loss": 1.0068,
+      "step": 4635
+    },
+    {
+      "epoch": 0.48757016840417,
+      "grad_norm": 3.1422402377931924,
+      "learning_rate": 2.6401761180929798e-06,
+      "loss": 1.0288,
+      "step": 4636
+    },
+    {
+      "epoch": 0.4876753388460173,
+      "grad_norm": 2.524963864157816,
+      "learning_rate": 2.63934260414217e-06,
+      "loss": 1.0381,
+      "step": 4637
+    },
+    {
+      "epoch": 0.48778050928786465,
+      "grad_norm": 2.009876109861969,
+      "learning_rate": 2.638509074653577e-06,
+      "loss": 0.9859,
+      "step": 4638
+    },
+    {
+      "epoch": 0.487885679729712,
+      "grad_norm": 2.0009976732169847,
+      "learning_rate": 2.637675529720147e-06,
+      "loss": 0.9932,
+      "step": 4639
+    },
+    {
+      "epoch": 0.4879908501715593,
+      "grad_norm": 2.0651436252017388,
+      "learning_rate": 2.6368419694348248e-06,
+      "loss": 1.0164,
+      "step": 4640
+    },
+    {
+      "epoch": 0.4880960206134066,
+      "grad_norm": 2.498095061638486,
+      "learning_rate": 2.6360083938905612e-06,
+      "loss": 1.0162,
+      "step": 4641
+    },
+    {
+      "epoch": 0.48820119105525395,
+      "grad_norm": 2.305734098685978,
+      "learning_rate": 2.635174803180305e-06,
+      "loss": 1.0235,
+      "step": 4642
+    },
+    {
+      "epoch": 0.4883063614971012,
+      "grad_norm": 1.9716369355731187,
+      "learning_rate": 2.6343411973970075e-06,
+      "loss": 0.9755,
+      "step": 4643
+    },
+    {
+      "epoch": 0.48841153193894854,
+      "grad_norm": 2.453378278184514,
+      "learning_rate": 2.633507576633623e-06,
+      "loss": 1.0,
+      "step": 4644
+    },
+    {
+      "epoch": 0.48851670238079586,
+      "grad_norm": 2.868869437829872,
+      "learning_rate": 2.632673940983106e-06,
+      "loss": 1.0469,
+      "step": 4645
+    },
+    {
+      "epoch": 0.4886218728226432,
+      "grad_norm": 3.2665412006936227,
+      "learning_rate": 2.631840290538415e-06,
+      "loss": 0.998,
+      "step": 4646
+    },
+    {
+      "epoch": 0.4887270432644905,
+      "grad_norm": 2.280933184203751,
+      "learning_rate": 2.6310066253925067e-06,
+      "loss": 0.9999,
+      "step": 4647
+    },
+    {
+      "epoch": 0.48883221370633784,
+      "grad_norm": 2.389682732235324,
+      "learning_rate": 2.6301729456383425e-06,
+      "loss": 0.9938,
+      "step": 4648
+    },
+    {
+      "epoch": 0.48893738414818516,
+      "grad_norm": 2.257361752513528,
+      "learning_rate": 2.629339251368884e-06,
+      "loss": 1.0198,
+      "step": 4649
+    },
+    {
+      "epoch": 0.4890425545900325,
+      "grad_norm": 3.1433460780930336,
+      "learning_rate": 2.6285055426770935e-06,
+      "loss": 1.022,
+      "step": 4650
+    },
+    {
+      "epoch": 0.4891477250318798,
+      "grad_norm": 2.1429693200460904,
+      "learning_rate": 2.627671819655937e-06,
+      "loss": 0.999,
+      "step": 4651
+    },
+    {
+      "epoch": 0.48925289547372713,
+      "grad_norm": 2.3878543730972206,
+      "learning_rate": 2.62683808239838e-06,
+      "loss": 1.0023,
+      "step": 4652
+    },
+    {
+      "epoch": 0.4893580659155744,
+      "grad_norm": 2.7691239385645683,
+      "learning_rate": 2.6260043309973925e-06,
+      "loss": 0.9745,
+      "step": 4653
+    },
+    {
+      "epoch": 0.4894632363574217,
+      "grad_norm": 2.499584360819442,
+      "learning_rate": 2.625170565545943e-06,
+      "loss": 1.0039,
+      "step": 4654
+    },
+    {
+      "epoch": 0.48956840679926905,
+      "grad_norm": 2.3633614479755547,
+      "learning_rate": 2.624336786137003e-06,
+      "loss": 0.9862,
+      "step": 4655
+    },
+    {
+      "epoch": 0.4896735772411164,
+      "grad_norm": 1.7313247618731031,
+      "learning_rate": 2.6235029928635457e-06,
+      "loss": 1.0106,
+      "step": 4656
+    },
+    {
+      "epoch": 0.4897787476829637,
+      "grad_norm": 2.176190189031968,
+      "learning_rate": 2.6226691858185454e-06,
+      "loss": 0.9691,
+      "step": 4657
+    },
+    {
+      "epoch": 0.489883918124811,
+      "grad_norm": 3.0580118331682637,
+      "learning_rate": 2.621835365094978e-06,
+      "loss": 1.0148,
+      "step": 4658
+    },
+    {
+      "epoch": 0.48998908856665835,
+      "grad_norm": 2.3238272613013864,
+      "learning_rate": 2.6210015307858207e-06,
+      "loss": 0.9906,
+      "step": 4659
+    },
+    {
+      "epoch": 0.4900942590085057,
+      "grad_norm": 2.6038315918171047,
+      "learning_rate": 2.620167682984052e-06,
+      "loss": 0.982,
+      "step": 4660
+    },
+    {
+      "epoch": 0.490199429450353,
+      "grad_norm": 3.173624125525723,
+      "learning_rate": 2.6193338217826536e-06,
+      "loss": 1.0282,
+      "step": 4661
+    },
+    {
+      "epoch": 0.4903045998922003,
+      "grad_norm": 2.4485387128876868,
+      "learning_rate": 2.6184999472746076e-06,
+      "loss": 1.019,
+      "step": 4662
+    },
+    {
+      "epoch": 0.4904097703340476,
+      "grad_norm": 2.2273055407932096,
+      "learning_rate": 2.6176660595528967e-06,
+      "loss": 1.0285,
+      "step": 4663
+    },
+    {
+      "epoch": 0.4905149407758949,
+      "grad_norm": 2.3699531308839004,
+      "learning_rate": 2.616832158710506e-06,
+      "loss": 0.988,
+      "step": 4664
+    },
+    {
+      "epoch": 0.49062011121774224,
+      "grad_norm": 2.5297910043499283,
+      "learning_rate": 2.6159982448404227e-06,
+      "loss": 1.0217,
+      "step": 4665
+    },
+    {
+      "epoch": 0.49072528165958956,
+      "grad_norm": 2.29822026488786,
+      "learning_rate": 2.615164318035633e-06,
+      "loss": 0.9885,
+      "step": 4666
+    },
+    {
+      "epoch": 0.4908304521014369,
+      "grad_norm": 2.3420596691918796,
+      "learning_rate": 2.6143303783891278e-06,
+      "loss": 1.0152,
+      "step": 4667
+    },
+    {
+      "epoch": 0.4909356225432842,
+      "grad_norm": 2.6066387189345885,
+      "learning_rate": 2.6134964259938967e-06,
+      "loss": 1.0215,
+      "step": 4668
+    },
+    {
+      "epoch": 0.49104079298513154,
+      "grad_norm": 2.579302269115743,
+      "learning_rate": 2.6126624609429325e-06,
+      "loss": 0.9935,
+      "step": 4669
+    },
+    {
+      "epoch": 0.49114596342697886,
+      "grad_norm": 3.1504338937696814,
+      "learning_rate": 2.611828483329229e-06,
+      "loss": 0.9786,
+      "step": 4670
+    },
+    {
+      "epoch": 0.4912511338688262,
+      "grad_norm": 2.3543913083014365,
+      "learning_rate": 2.6109944932457813e-06,
+      "loss": 0.9647,
+      "step": 4671
+    },
+    {
+      "epoch": 0.4913563043106735,
+      "grad_norm": 2.586644471709751,
+      "learning_rate": 2.610160490785584e-06,
+      "loss": 1.0352,
+      "step": 4672
+    },
+    {
+      "epoch": 0.4914614747525208,
+      "grad_norm": 2.135602362463421,
+      "learning_rate": 2.6093264760416377e-06,
+      "loss": 1.0012,
+      "step": 4673
+    },
+    {
+      "epoch": 0.4915666451943681,
+      "grad_norm": 2.8951974224499795,
+      "learning_rate": 2.6084924491069397e-06,
+      "loss": 0.989,
+      "step": 4674
+    },
+    {
+      "epoch": 0.4916718156362154,
+      "grad_norm": 2.670407042084373,
+      "learning_rate": 2.60765841007449e-06,
+      "loss": 0.9966,
+      "step": 4675
+    },
+    {
+      "epoch": 0.49177698607806275,
+      "grad_norm": 2.5070143385941073,
+      "learning_rate": 2.606824359037292e-06,
+      "loss": 0.9903,
+      "step": 4676
+    },
+    {
+      "epoch": 0.4918821565199101,
+      "grad_norm": 3.1780468314666246,
+      "learning_rate": 2.605990296088348e-06,
+      "loss": 0.9961,
+      "step": 4677
+    },
+    {
+      "epoch": 0.4919873269617574,
+      "grad_norm": 3.0212424807578233,
+      "learning_rate": 2.6051562213206633e-06,
+      "loss": 1.0063,
+      "step": 4678
+    },
+    {
+      "epoch": 0.4920924974036047,
+      "grad_norm": 2.1752773833274732,
+      "learning_rate": 2.604322134827242e-06,
+      "loss": 0.9692,
+      "step": 4679
+    },
+    {
+      "epoch": 0.49219766784545205,
+      "grad_norm": 2.101331063752568,
+      "learning_rate": 2.603488036701093e-06,
+      "loss": 0.9853,
+      "step": 4680
+    },
+    {
+      "epoch": 0.49230283828729937,
+      "grad_norm": 2.7504154040985385,
+      "learning_rate": 2.6026539270352234e-06,
+      "loss": 0.977,
+      "step": 4681
+    },
+    {
+      "epoch": 0.4924080087291467,
+      "grad_norm": 2.283095151367684,
+      "learning_rate": 2.6018198059226433e-06,
+      "loss": 0.9963,
+      "step": 4682
+    },
+    {
+      "epoch": 0.492513179170994,
+      "grad_norm": 3.6836502509688196,
+      "learning_rate": 2.600985673456364e-06,
+      "loss": 1.0127,
+      "step": 4683
+    },
+    {
+      "epoch": 0.4926183496128413,
+      "grad_norm": 1.7196183166216403,
+      "learning_rate": 2.6001515297293982e-06,
+      "loss": 0.96,
+      "step": 4684
+    },
+    {
+      "epoch": 0.4927235200546886,
+      "grad_norm": 2.0308139618206105,
+      "learning_rate": 2.599317374834759e-06,
+      "loss": 0.9981,
+      "step": 4685
+    },
+    {
+      "epoch": 0.49282869049653594,
+      "grad_norm": 2.7245202798321766,
+      "learning_rate": 2.59848320886546e-06,
+      "loss": 1.0274,
+      "step": 4686
+    },
+    {
+      "epoch": 0.49293386093838326,
+      "grad_norm": 2.2582233947427786,
+      "learning_rate": 2.597649031914519e-06,
+      "loss": 1.031,
+      "step": 4687
+    },
+    {
+      "epoch": 0.4930390313802306,
+      "grad_norm": 2.875499760161925,
+      "learning_rate": 2.5968148440749526e-06,
+      "loss": 1.0076,
+      "step": 4688
+    },
+    {
+      "epoch": 0.4931442018220779,
+      "grad_norm": 2.3873631179045076,
+      "learning_rate": 2.595980645439778e-06,
+      "loss": 1.0003,
+      "step": 4689
+    },
+    {
+      "epoch": 0.49324937226392523,
+      "grad_norm": 1.832107108457019,
+      "learning_rate": 2.595146436102016e-06,
+      "loss": 0.9479,
+      "step": 4690
+    },
+    {
+      "epoch": 0.49335454270577256,
+      "grad_norm": 2.1390781829339893,
+      "learning_rate": 2.5943122161546874e-06,
+      "loss": 0.9757,
+      "step": 4691
+    },
+    {
+      "epoch": 0.4934597131476199,
+      "grad_norm": 2.678011051923976,
+      "learning_rate": 2.593477985690815e-06,
+      "loss": 1.0185,
+      "step": 4692
+    },
+    {
+      "epoch": 0.4935648835894672,
+      "grad_norm": 2.655131659501105,
+      "learning_rate": 2.59264374480342e-06,
+      "loss": 0.9632,
+      "step": 4693
+    },
+    {
+      "epoch": 0.4936700540313145,
+      "grad_norm": 2.387020654145177,
+      "learning_rate": 2.5918094935855275e-06,
+      "loss": 0.9935,
+      "step": 4694
+    },
+    {
+      "epoch": 0.4937752244731618,
+      "grad_norm": 2.484922013212365,
+      "learning_rate": 2.5909752321301633e-06,
+      "loss": 0.9588,
+      "step": 4695
+    },
+    {
+      "epoch": 0.4938803949150091,
+      "grad_norm": 3.105536483019693,
+      "learning_rate": 2.590140960530355e-06,
+      "loss": 0.9992,
+      "step": 4696
+    },
+    {
+      "epoch": 0.49398556535685645,
+      "grad_norm": 2.4101649578596445,
+      "learning_rate": 2.589306678879129e-06,
+      "loss": 1.039,
+      "step": 4697
+    },
+    {
+      "epoch": 0.4940907357987038,
+      "grad_norm": 2.383606369665489,
+      "learning_rate": 2.5884723872695138e-06,
+      "loss": 1.0335,
+      "step": 4698
+    },
+    {
+      "epoch": 0.4941959062405511,
+      "grad_norm": 2.5503747168717874,
+      "learning_rate": 2.587638085794541e-06,
+      "loss": 1.0034,
+      "step": 4699
+    },
+    {
+      "epoch": 0.4943010766823984,
+      "grad_norm": 2.0251324980808287,
+      "learning_rate": 2.5868037745472408e-06,
+      "loss": 0.9547,
+      "step": 4700
+    },
+    {
+      "epoch": 0.49440624712424575,
+      "grad_norm": 2.2254913031928596,
+      "learning_rate": 2.5859694536206455e-06,
+      "loss": 1.018,
+      "step": 4701
+    },
+    {
+      "epoch": 0.49451141756609307,
+      "grad_norm": 1.7896906696794816,
+      "learning_rate": 2.5851351231077876e-06,
+      "loss": 0.9859,
+      "step": 4702
+    },
+    {
+      "epoch": 0.4946165880079404,
+      "grad_norm": 2.6959960617755168,
+      "learning_rate": 2.5843007831017024e-06,
+      "loss": 1.0311,
+      "step": 4703
+    },
+    {
+      "epoch": 0.49472175844978766,
+      "grad_norm": 3.497101772230584,
+      "learning_rate": 2.583466433695425e-06,
+      "loss": 0.98,
+      "step": 4704
+    },
+    {
+      "epoch": 0.494826928891635,
+      "grad_norm": 2.220969128258192,
+      "learning_rate": 2.5826320749819917e-06,
+      "loss": 1.0059,
+      "step": 4705
+    },
+    {
+      "epoch": 0.4949320993334823,
+      "grad_norm": 2.477231282570078,
+      "learning_rate": 2.5817977070544408e-06,
+      "loss": 1.0255,
+      "step": 4706
+    },
+    {
+      "epoch": 0.49503726977532964,
+      "grad_norm": 2.43917889374586,
+      "learning_rate": 2.5809633300058095e-06,
+      "loss": 1.0001,
+      "step": 4707
+    },
+    {
+      "epoch": 0.49514244021717696,
+      "grad_norm": 3.203459546979018,
+      "learning_rate": 2.580128943929139e-06,
+      "loss": 0.9777,
+      "step": 4708
+    },
+    {
+      "epoch": 0.4952476106590243,
+      "grad_norm": 2.072545796219728,
+      "learning_rate": 2.579294548917467e-06,
+      "loss": 0.9962,
+      "step": 4709
+    },
+    {
+      "epoch": 0.4953527811008716,
+      "grad_norm": 2.5943743382810465,
+      "learning_rate": 2.578460145063838e-06,
+      "loss": 0.9941,
+      "step": 4710
+    },
+    {
+      "epoch": 0.49545795154271893,
+      "grad_norm": 1.9553371503993335,
+      "learning_rate": 2.577625732461293e-06,
+      "loss": 0.9708,
+      "step": 4711
+    },
+    {
+      "epoch": 0.49556312198456626,
+      "grad_norm": 2.5527175195599345,
+      "learning_rate": 2.576791311202876e-06,
+      "loss": 1.0314,
+      "step": 4712
+    },
+    {
+      "epoch": 0.4956682924264136,
+      "grad_norm": 2.7301034938935858,
+      "learning_rate": 2.575956881381631e-06,
+      "loss": 1.0063,
+      "step": 4713
+    },
+    {
+      "epoch": 0.49577346286826085,
+      "grad_norm": 2.3128243301857165,
+      "learning_rate": 2.575122443090604e-06,
+      "loss": 1.0161,
+      "step": 4714
+    },
+    {
+      "epoch": 0.4958786333101082,
+      "grad_norm": 2.1154594017025357,
+      "learning_rate": 2.574287996422841e-06,
+      "loss": 1.0245,
+      "step": 4715
+    },
+    {
+      "epoch": 0.4959838037519555,
+      "grad_norm": 2.2840660853868533,
+      "learning_rate": 2.573453541471389e-06,
+      "loss": 1.0142,
+      "step": 4716
+    },
+    {
+      "epoch": 0.4960889741938028,
+      "grad_norm": 2.5091203779297104,
+      "learning_rate": 2.572619078329297e-06,
+      "loss": 0.9737,
+      "step": 4717
+    },
+    {
+      "epoch": 0.49619414463565015,
+      "grad_norm": 2.004908508620825,
+      "learning_rate": 2.571784607089613e-06,
+      "loss": 0.9905,
+      "step": 4718
+    },
+    {
+      "epoch": 0.49629931507749747,
+      "grad_norm": 2.525170667494257,
+      "learning_rate": 2.570950127845388e-06,
+      "loss": 0.9913,
+      "step": 4719
+    },
+    {
+      "epoch": 0.4964044855193448,
+      "grad_norm": 2.3933251418961414,
+      "learning_rate": 2.5701156406896726e-06,
+      "loss": 1.0001,
+      "step": 4720
+    },
+    {
+      "epoch": 0.4965096559611921,
+      "grad_norm": 2.7182770035408166,
+      "learning_rate": 2.5692811457155186e-06,
+      "loss": 0.9676,
+      "step": 4721
+    },
+    {
+      "epoch": 0.49661482640303944,
+      "grad_norm": 2.8066431382335386,
+      "learning_rate": 2.5684466430159794e-06,
+      "loss": 0.9872,
+      "step": 4722
+    },
+    {
+      "epoch": 0.49671999684488677,
+      "grad_norm": 2.255340942233427,
+      "learning_rate": 2.5676121326841063e-06,
+      "loss": 0.9676,
+      "step": 4723
+    },
+    {
+      "epoch": 0.49682516728673404,
+      "grad_norm": 2.0906594545737267,
+      "learning_rate": 2.566777614812956e-06,
+      "loss": 0.9905,
+      "step": 4724
+    },
+    {
+      "epoch": 0.49693033772858136,
+      "grad_norm": 3.0950706282178184,
+      "learning_rate": 2.5659430894955827e-06,
+      "loss": 1.0,
+      "step": 4725
+    },
+    {
+      "epoch": 0.4970355081704287,
+      "grad_norm": 2.7642796610888065,
+      "learning_rate": 2.5651085568250426e-06,
+      "loss": 0.9555,
+      "step": 4726
+    },
+    {
+      "epoch": 0.497140678612276,
+      "grad_norm": 2.584998095593244,
+      "learning_rate": 2.564274016894393e-06,
+      "loss": 0.987,
+      "step": 4727
+    },
+    {
+      "epoch": 0.49724584905412333,
+      "grad_norm": 2.688374289525621,
+      "learning_rate": 2.5634394697966915e-06,
+      "loss": 0.9615,
+      "step": 4728
+    },
+    {
+      "epoch": 0.49735101949597066,
+      "grad_norm": 2.9732686786827056,
+      "learning_rate": 2.562604915624996e-06,
+      "loss": 0.9954,
+      "step": 4729
+    },
+    {
+      "epoch": 0.497456189937818,
+      "grad_norm": 2.6909679945625635,
+      "learning_rate": 2.5617703544723653e-06,
+      "loss": 0.9819,
+      "step": 4730
+    },
+    {
+      "epoch": 0.4975613603796653,
+      "grad_norm": 2.397135959093128,
+      "learning_rate": 2.5609357864318614e-06,
+      "loss": 1.0235,
+      "step": 4731
+    },
+    {
+      "epoch": 0.49766653082151263,
+      "grad_norm": 2.7889907201615185,
+      "learning_rate": 2.5601012115965425e-06,
+      "loss": 1.0192,
+      "step": 4732
+    },
+    {
+      "epoch": 0.49777170126335996,
+      "grad_norm": 2.4282752934234417,
+      "learning_rate": 2.559266630059473e-06,
+      "loss": 1.0139,
+      "step": 4733
+    },
+    {
+      "epoch": 0.4978768717052072,
+      "grad_norm": 2.8902978095243874,
+      "learning_rate": 2.5584320419137127e-06,
+      "loss": 1.0143,
+      "step": 4734
+    },
+    {
+      "epoch": 0.49798204214705455,
+      "grad_norm": 3.286409074951358,
+      "learning_rate": 2.5575974472523273e-06,
+      "loss": 1.0451,
+      "step": 4735
+    },
+    {
+      "epoch": 0.4980872125889019,
+      "grad_norm": 2.009314598243189,
+      "learning_rate": 2.556762846168378e-06,
+      "loss": 0.9562,
+      "step": 4736
+    },
+    {
+      "epoch": 0.4981923830307492,
+      "grad_norm": 1.832895791185105,
+      "learning_rate": 2.5559282387549305e-06,
+      "loss": 1.0244,
+      "step": 4737
+    },
+    {
+      "epoch": 0.4982975534725965,
+      "grad_norm": 2.1324340102401798,
+      "learning_rate": 2.5550936251050503e-06,
+      "loss": 1.0173,
+      "step": 4738
+    },
+    {
+      "epoch": 0.49840272391444385,
+      "grad_norm": 2.608030569296082,
+      "learning_rate": 2.5542590053118022e-06,
+      "loss": 0.986,
+      "step": 4739
+    },
+    {
+      "epoch": 0.49850789435629117,
+      "grad_norm": 2.7802760127585473,
+      "learning_rate": 2.553424379468254e-06,
+      "loss": 1.0234,
+      "step": 4740
+    },
+    {
+      "epoch": 0.4986130647981385,
+      "grad_norm": 2.7173683170964082,
+      "learning_rate": 2.5525897476674722e-06,
+      "loss": 1.0201,
+      "step": 4741
+    },
+    {
+      "epoch": 0.4987182352399858,
+      "grad_norm": 2.67038694261878,
+      "learning_rate": 2.5517551100025257e-06,
+      "loss": 1.004,
+      "step": 4742
+    },
+    {
+      "epoch": 0.49882340568183314,
+      "grad_norm": 2.9249794319362192,
+      "learning_rate": 2.550920466566483e-06,
+      "loss": 1.011,
+      "step": 4743
+    },
+    {
+      "epoch": 0.49892857612368047,
+      "grad_norm": 2.6543517747008405,
+      "learning_rate": 2.5500858174524105e-06,
+      "loss": 0.9933,
+      "step": 4744
+    },
+    {
+      "epoch": 0.49903374656552774,
+      "grad_norm": 2.3829909949524537,
+      "learning_rate": 2.5492511627533816e-06,
+      "loss": 1.0197,
+      "step": 4745
+    },
+    {
+      "epoch": 0.49913891700737506,
+      "grad_norm": 2.850982046315611,
+      "learning_rate": 2.548416502562465e-06,
+      "loss": 0.9597,
+      "step": 4746
+    },
+    {
+      "epoch": 0.4992440874492224,
+      "grad_norm": 2.1710344138345365,
+      "learning_rate": 2.5475818369727328e-06,
+      "loss": 0.9962,
+      "step": 4747
+    },
+    {
+      "epoch": 0.4993492578910697,
+      "grad_norm": 2.3265309831025265,
+      "learning_rate": 2.5467471660772557e-06,
+      "loss": 1.0114,
+      "step": 4748
+    },
+    {
+      "epoch": 0.49945442833291703,
+      "grad_norm": 2.701708578526532,
+      "learning_rate": 2.545912489969107e-06,
+      "loss": 1.0285,
+      "step": 4749
+    },
+    {
+      "epoch": 0.49955959877476436,
+      "grad_norm": 2.650784964191637,
+      "learning_rate": 2.5450778087413588e-06,
+      "loss": 0.9867,
+      "step": 4750
+    },
+    {
+      "epoch": 0.4996647692166117,
+      "grad_norm": 2.5087138546199355,
+      "learning_rate": 2.5442431224870847e-06,
+      "loss": 1.0158,
+      "step": 4751
+    },
+    {
+      "epoch": 0.499769939658459,
+      "grad_norm": 2.995804939311976,
+      "learning_rate": 2.5434084312993582e-06,
+      "loss": 1.0099,
+      "step": 4752
+    },
+    {
+      "epoch": 0.49987511010030633,
+      "grad_norm": 2.6588509171871553,
+      "learning_rate": 2.542573735271255e-06,
+      "loss": 1.0047,
+      "step": 4753
+    },
+    {
+      "epoch": 0.49998028054215365,
+      "grad_norm": 1.8191787178519514,
+      "learning_rate": 2.5417390344958494e-06,
+      "loss": 1.0107,
+      "step": 4754
+    },
+    {
+      "epoch": 0.5000854509840009,
+      "grad_norm": 1.8754132743779173,
+      "learning_rate": 2.5409043290662173e-06,
+      "loss": 0.988,
+      "step": 4755
+    },
+    {
+      "epoch": 0.5001906214258482,
+      "grad_norm": 2.631056408985051,
+      "learning_rate": 2.5400696190754347e-06,
+      "loss": 1.0357,
+      "step": 4756
+    },
+    {
+      "epoch": 0.5002957918676956,
+      "grad_norm": 2.032133232558368,
+      "learning_rate": 2.5392349046165783e-06,
+      "loss": 0.9912,
+      "step": 4757
+    },
+    {
+      "epoch": 0.5004009623095429,
+      "grad_norm": 2.0586532168086187,
+      "learning_rate": 2.5384001857827256e-06,
+      "loss": 0.9694,
+      "step": 4758
+    },
+    {
+      "epoch": 0.5005061327513902,
+      "grad_norm": 2.644307665365883,
+      "learning_rate": 2.5375654626669537e-06,
+      "loss": 0.9973,
+      "step": 4759
+    },
+    {
+      "epoch": 0.5006113031932375,
+      "grad_norm": 1.773857858460592,
+      "learning_rate": 2.5367307353623404e-06,
+      "loss": 0.9653,
+      "step": 4760
+    },
+    {
+      "epoch": 0.5007164736350849,
+      "grad_norm": 2.3144669620091585,
+      "learning_rate": 2.5358960039619653e-06,
+      "loss": 0.9818,
+      "step": 4761
+    },
+    {
+      "epoch": 0.5008216440769322,
+      "grad_norm": 3.4633187371850513,
+      "learning_rate": 2.535061268558906e-06,
+      "loss": 1.0327,
+      "step": 4762
+    },
+    {
+      "epoch": 0.5009268145187795,
+      "grad_norm": 1.7464824159525518,
+      "learning_rate": 2.5342265292462437e-06,
+      "loss": 0.9835,
+      "step": 4763
+    },
+    {
+      "epoch": 0.5010319849606268,
+      "grad_norm": 2.5356175046255336,
+      "learning_rate": 2.5333917861170576e-06,
+      "loss": 0.9872,
+      "step": 4764
+    },
+    {
+      "epoch": 0.5011371554024742,
+      "grad_norm": 2.4355449363964503,
+      "learning_rate": 2.5325570392644282e-06,
+      "loss": 1.0068,
+      "step": 4765
+    },
+    {
+      "epoch": 0.5012423258443215,
+      "grad_norm": 2.415671397242718,
+      "learning_rate": 2.531722288781436e-06,
+      "loss": 0.9525,
+      "step": 4766
+    },
+    {
+      "epoch": 0.5013474962861688,
+      "grad_norm": 3.179290978551529,
+      "learning_rate": 2.5308875347611613e-06,
+      "loss": 1.0196,
+      "step": 4767
+    },
+    {
+      "epoch": 0.5014526667280161,
+      "grad_norm": 2.0301483199514765,
+      "learning_rate": 2.530052777296687e-06,
+      "loss": 1.0312,
+      "step": 4768
+    },
+    {
+      "epoch": 0.5015578371698635,
+      "grad_norm": 1.7988911344950873,
+      "learning_rate": 2.5292180164810944e-06,
+      "loss": 0.9735,
+      "step": 4769
+    },
+    {
+      "epoch": 0.5016630076117107,
+      "grad_norm": 2.91048308505893,
+      "learning_rate": 2.5283832524074657e-06,
+      "loss": 1.0213,
+      "step": 4770
+    },
+    {
+      "epoch": 0.501768178053558,
+      "grad_norm": 2.4229460616322136,
+      "learning_rate": 2.5275484851688848e-06,
+      "loss": 0.9999,
+      "step": 4771
+    },
+    {
+      "epoch": 0.5018733484954053,
+      "grad_norm": 2.4418059020235003,
+      "learning_rate": 2.5267137148584335e-06,
+      "loss": 0.9837,
+      "step": 4772
+    },
+    {
+      "epoch": 0.5019785189372526,
+      "grad_norm": 2.3825700805364,
+      "learning_rate": 2.5258789415691947e-06,
+      "loss": 1.039,
+      "step": 4773
+    },
+    {
+      "epoch": 0.5020836893791,
+      "grad_norm": 2.5170794669290286,
+      "learning_rate": 2.5250441653942536e-06,
+      "loss": 0.9986,
+      "step": 4774
+    },
+    {
+      "epoch": 0.5021888598209473,
+      "grad_norm": 2.894554710829124,
+      "learning_rate": 2.5242093864266936e-06,
+      "loss": 0.9848,
+      "step": 4775
+    },
+    {
+      "epoch": 0.5022940302627946,
+      "grad_norm": 2.3297677080374064,
+      "learning_rate": 2.5233746047595984e-06,
+      "loss": 0.9898,
+      "step": 4776
+    },
+    {
+      "epoch": 0.502399200704642,
+      "grad_norm": 2.7131245724405604,
+      "learning_rate": 2.5225398204860534e-06,
+      "loss": 1.0009,
+      "step": 4777
+    },
+    {
+      "epoch": 0.5025043711464893,
+      "grad_norm": 2.264411534263541,
+      "learning_rate": 2.5217050336991434e-06,
+      "loss": 0.9784,
+      "step": 4778
+    },
+    {
+      "epoch": 0.5026095415883366,
+      "grad_norm": 2.2425268597601833,
+      "learning_rate": 2.520870244491954e-06,
+      "loss": 0.9914,
+      "step": 4779
+    },
+    {
+      "epoch": 0.5027147120301839,
+      "grad_norm": 1.5512671176052935,
+      "learning_rate": 2.5200354529575693e-06,
+      "loss": 0.9783,
+      "step": 4780
+    },
+    {
+      "epoch": 0.5028198824720312,
+      "grad_norm": 2.4180849734923515,
+      "learning_rate": 2.5192006591890767e-06,
+      "loss": 1.0115,
+      "step": 4781
+    },
+    {
+      "epoch": 0.5029250529138786,
+      "grad_norm": 2.0161358166931813,
+      "learning_rate": 2.5183658632795614e-06,
+      "loss": 1.011,
+      "step": 4782
+    },
+    {
+      "epoch": 0.5030302233557259,
+      "grad_norm": 2.9226670044367222,
+      "learning_rate": 2.5175310653221092e-06,
+      "loss": 0.9739,
+      "step": 4783
+    },
+    {
+      "epoch": 0.5031353937975732,
+      "grad_norm": 2.421297281935595,
+      "learning_rate": 2.5166962654098075e-06,
+      "loss": 1.0089,
+      "step": 4784
+    },
+    {
+      "epoch": 0.5032405642394205,
+      "grad_norm": 1.820279687289599,
+      "learning_rate": 2.515861463635742e-06,
+      "loss": 1.0076,
+      "step": 4785
+    },
+    {
+      "epoch": 0.5033457346812679,
+      "grad_norm": 2.13363221480922,
+      "learning_rate": 2.515026660093e-06,
+      "loss": 1.0352,
+      "step": 4786
+    },
+    {
+      "epoch": 0.5034509051231152,
+      "grad_norm": 2.2192281908305596,
+      "learning_rate": 2.51419185487467e-06,
+      "loss": 0.9947,
+      "step": 4787
+    },
+    {
+      "epoch": 0.5035560755649625,
+      "grad_norm": 2.6940645632774833,
+      "learning_rate": 2.5133570480738367e-06,
+      "loss": 0.9647,
+      "step": 4788
+    },
+    {
+      "epoch": 0.5036612460068098,
+      "grad_norm": 2.92639520094335,
+      "learning_rate": 2.5125222397835893e-06,
+      "loss": 0.9956,
+      "step": 4789
+    },
+    {
+      "epoch": 0.503766416448657,
+      "grad_norm": 1.796576628883353,
+      "learning_rate": 2.511687430097014e-06,
+      "loss": 0.9822,
+      "step": 4790
+    },
+    {
+      "epoch": 0.5038715868905044,
+      "grad_norm": 2.66563998878901,
+      "learning_rate": 2.5108526191072e-06,
+      "loss": 0.9973,
+      "step": 4791
+    },
+    {
+      "epoch": 0.5039767573323517,
+      "grad_norm": 1.8895115699475404,
+      "learning_rate": 2.5100178069072347e-06,
+      "loss": 0.9552,
+      "step": 4792
+    },
+    {
+      "epoch": 0.504081927774199,
+      "grad_norm": 3.798229950568707,
+      "learning_rate": 2.5091829935902063e-06,
+      "loss": 0.9752,
+      "step": 4793
+    },
+    {
+      "epoch": 0.5041870982160463,
+      "grad_norm": 2.8809636374230667,
+      "learning_rate": 2.508348179249203e-06,
+      "loss": 0.983,
+      "step": 4794
+    },
+    {
+      "epoch": 0.5042922686578937,
+      "grad_norm": 2.5746810183718436,
+      "learning_rate": 2.5075133639773126e-06,
+      "loss": 0.9668,
+      "step": 4795
+    },
+    {
+      "epoch": 0.504397439099741,
+      "grad_norm": 1.8118205407198673,
+      "learning_rate": 2.506678547867623e-06,
+      "loss": 0.999,
+      "step": 4796
+    },
+    {
+      "epoch": 0.5045026095415883,
+      "grad_norm": 1.5885999010984382,
+      "learning_rate": 2.5058437310132244e-06,
+      "loss": 0.9747,
+      "step": 4797
+    },
+    {
+      "epoch": 0.5046077799834356,
+      "grad_norm": 2.1825556187323114,
+      "learning_rate": 2.5050089135072044e-06,
+      "loss": 0.9754,
+      "step": 4798
+    },
+    {
+      "epoch": 0.504712950425283,
+      "grad_norm": 2.1535118451086257,
+      "learning_rate": 2.5041740954426508e-06,
+      "loss": 1.0055,
+      "step": 4799
+    },
+    {
+      "epoch": 0.5048181208671303,
+      "grad_norm": 2.1050204876713177,
+      "learning_rate": 2.5033392769126543e-06,
+      "loss": 1.0276,
+      "step": 4800
+    },
+    {
+      "epoch": 0.5049232913089776,
+      "grad_norm": 2.139583763580259,
+      "learning_rate": 2.5025044580103012e-06,
+      "loss": 1.0135,
+      "step": 4801
+    },
+    {
+      "epoch": 0.5050284617508249,
+      "grad_norm": 2.1492851865676337,
+      "learning_rate": 2.5016696388286827e-06,
+      "loss": 1.014,
+      "step": 4802
+    },
+    {
+      "epoch": 0.5051336321926723,
+      "grad_norm": 2.1397490291747014,
+      "learning_rate": 2.5008348194608855e-06,
+      "loss": 0.991,
+      "step": 4803
+    },
+    {
+      "epoch": 0.5052388026345196,
+      "grad_norm": 1.9983819330325248,
+      "learning_rate": 2.5e-06,
+      "loss": 1.009,
+      "step": 4804
+    },
+    {
+      "epoch": 0.5053439730763669,
+      "grad_norm": 2.5956563261204013,
+      "learning_rate": 2.499165180539115e-06,
+      "loss": 0.9449,
+      "step": 4805
+    },
+    {
+      "epoch": 0.5054491435182142,
+      "grad_norm": 2.1548335567068073,
+      "learning_rate": 2.498330361171318e-06,
+      "loss": 0.9516,
+      "step": 4806
+    },
+    {
+      "epoch": 0.5055543139600616,
+      "grad_norm": 2.73508525509217,
+      "learning_rate": 2.497495541989699e-06,
+      "loss": 0.9602,
+      "step": 4807
+    },
+    {
+      "epoch": 0.5056594844019089,
+      "grad_norm": 2.3580994201959253,
+      "learning_rate": 2.496660723087347e-06,
+      "loss": 0.9723,
+      "step": 4808
+    },
+    {
+      "epoch": 0.5057646548437562,
+      "grad_norm": 3.0042626219250823,
+      "learning_rate": 2.49582590455735e-06,
+      "loss": 0.9715,
+      "step": 4809
+    },
+    {
+      "epoch": 0.5058698252856035,
+      "grad_norm": 2.3224530834303185,
+      "learning_rate": 2.494991086492797e-06,
+      "loss": 1.0099,
+      "step": 4810
+    },
+    {
+      "epoch": 0.5059749957274507,
+      "grad_norm": 2.5954016653855962,
+      "learning_rate": 2.4941562689867755e-06,
+      "loss": 1.0442,
+      "step": 4811
+    },
+    {
+      "epoch": 0.5060801661692981,
+      "grad_norm": 2.561011878419228,
+      "learning_rate": 2.493321452132377e-06,
+      "loss": 0.9306,
+      "step": 4812
+    },
+    {
+      "epoch": 0.5061853366111454,
+      "grad_norm": 2.6220115240229416,
+      "learning_rate": 2.4924866360226883e-06,
+      "loss": 0.9854,
+      "step": 4813
+    },
+    {
+      "epoch": 0.5062905070529927,
+      "grad_norm": 2.529146248465547,
+      "learning_rate": 2.4916518207507977e-06,
+      "loss": 1.0399,
+      "step": 4814
+    },
+    {
+      "epoch": 0.50639567749484,
+      "grad_norm": 2.308532895773023,
+      "learning_rate": 2.490817006409794e-06,
+      "loss": 0.974,
+      "step": 4815
+    },
+    {
+      "epoch": 0.5065008479366874,
+      "grad_norm": 2.8751837068490085,
+      "learning_rate": 2.4899821930927653e-06,
+      "loss": 0.9367,
+      "step": 4816
+    },
+    {
+      "epoch": 0.5066060183785347,
+      "grad_norm": 1.8788646946917793,
+      "learning_rate": 2.4891473808928006e-06,
+      "loss": 0.9854,
+      "step": 4817
+    },
+    {
+      "epoch": 0.506711188820382,
+      "grad_norm": 2.1679354845155387,
+      "learning_rate": 2.488312569902987e-06,
+      "loss": 0.9615,
+      "step": 4818
+    },
+    {
+      "epoch": 0.5068163592622293,
+      "grad_norm": 2.3508798049934403,
+      "learning_rate": 2.487477760216412e-06,
+      "loss": 1.0156,
+      "step": 4819
+    },
+    {
+      "epoch": 0.5069215297040767,
+      "grad_norm": 1.7607835727712602,
+      "learning_rate": 2.4866429519261646e-06,
+      "loss": 0.994,
+      "step": 4820
+    },
+    {
+      "epoch": 0.507026700145924,
+      "grad_norm": 3.0523263238936593,
+      "learning_rate": 2.4858081451253306e-06,
+      "loss": 0.975,
+      "step": 4821
+    },
+    {
+      "epoch": 0.5071318705877713,
+      "grad_norm": 2.3526180310210454,
+      "learning_rate": 2.4849733399070002e-06,
+      "loss": 0.9872,
+      "step": 4822
+    },
+    {
+      "epoch": 0.5072370410296186,
+      "grad_norm": 2.263463303823174,
+      "learning_rate": 2.484138536364259e-06,
+      "loss": 0.9537,
+      "step": 4823
+    },
+    {
+      "epoch": 0.507342211471466,
+      "grad_norm": 2.1671169893440982,
+      "learning_rate": 2.4833037345901937e-06,
+      "loss": 0.9872,
+      "step": 4824
+    },
+    {
+      "epoch": 0.5074473819133133,
+      "grad_norm": 2.73902112314044,
+      "learning_rate": 2.482468934677891e-06,
+      "loss": 1.0494,
+      "step": 4825
+    },
+    {
+      "epoch": 0.5075525523551606,
+      "grad_norm": 2.6521521906637107,
+      "learning_rate": 2.481634136720439e-06,
+      "loss": 0.9943,
+      "step": 4826
+    },
+    {
+      "epoch": 0.5076577227970079,
+      "grad_norm": 3.004342981120889,
+      "learning_rate": 2.4807993408109237e-06,
+      "loss": 0.9704,
+      "step": 4827
+    },
+    {
+      "epoch": 0.5077628932388553,
+      "grad_norm": 3.205741832679773,
+      "learning_rate": 2.479964547042431e-06,
+      "loss": 0.9874,
+      "step": 4828
+    },
+    {
+      "epoch": 0.5078680636807026,
+      "grad_norm": 3.0608467513431106,
+      "learning_rate": 2.479129755508047e-06,
+      "loss": 1.0236,
+      "step": 4829
+    },
+    {
+      "epoch": 0.5079732341225499,
+      "grad_norm": 2.2772013134452402,
+      "learning_rate": 2.478294966300857e-06,
+      "loss": 0.993,
+      "step": 4830
+    },
+    {
+      "epoch": 0.5080784045643971,
+      "grad_norm": 2.143379541365547,
+      "learning_rate": 2.477460179513947e-06,
+      "loss": 0.9831,
+      "step": 4831
+    },
+    {
+      "epoch": 0.5081835750062444,
+      "grad_norm": 2.086617872315913,
+      "learning_rate": 2.4766253952404024e-06,
+      "loss": 0.9932,
+      "step": 4832
+    },
+    {
+      "epoch": 0.5082887454480918,
+      "grad_norm": 2.6508833715078386,
+      "learning_rate": 2.4757906135733077e-06,
+      "loss": 1.0029,
+      "step": 4833
+    },
+    {
+      "epoch": 0.5083939158899391,
+      "grad_norm": 2.8727794559719513,
+      "learning_rate": 2.4749558346057464e-06,
+      "loss": 1.0128,
+      "step": 4834
+    },
+    {
+      "epoch": 0.5084990863317864,
+      "grad_norm": 2.1203097915270797,
+      "learning_rate": 2.4741210584308053e-06,
+      "loss": 0.9912,
+      "step": 4835
+    },
+    {
+      "epoch": 0.5086042567736337,
+      "grad_norm": 2.6936248414531594,
+      "learning_rate": 2.4732862851415674e-06,
+      "loss": 0.9887,
+      "step": 4836
+    },
+    {
+      "epoch": 0.5087094272154811,
+      "grad_norm": 2.388064075424561,
+      "learning_rate": 2.472451514831116e-06,
+      "loss": 0.9818,
+      "step": 4837
+    },
+    {
+      "epoch": 0.5088145976573284,
+      "grad_norm": 2.4879710422290318,
+      "learning_rate": 2.4716167475925356e-06,
+      "loss": 1.0028,
+      "step": 4838
+    },
+    {
+      "epoch": 0.5089197680991757,
+      "grad_norm": 2.35770735993766,
+      "learning_rate": 2.470781983518906e-06,
+      "loss": 1.0071,
+      "step": 4839
+    },
+    {
+      "epoch": 0.509024938541023,
+      "grad_norm": 2.0072570921567605,
+      "learning_rate": 2.4699472227033137e-06,
+      "loss": 1.0177,
+      "step": 4840
+    },
+    {
+      "epoch": 0.5091301089828704,
+      "grad_norm": 2.926954873115335,
+      "learning_rate": 2.46911246523884e-06,
+      "loss": 0.9914,
+      "step": 4841
+    },
+    {
+      "epoch": 0.5092352794247177,
+      "grad_norm": 1.8985162910533802,
+      "learning_rate": 2.4682777112185657e-06,
+      "loss": 0.96,
+      "step": 4842
+    },
+    {
+      "epoch": 0.509340449866565,
+      "grad_norm": 2.961749775540693,
+      "learning_rate": 2.467442960735573e-06,
+      "loss": 1.011,
+      "step": 4843
+    },
+    {
+      "epoch": 0.5094456203084123,
+      "grad_norm": 2.828236340352394,
+      "learning_rate": 2.466608213882943e-06,
+      "loss": 1.0186,
+      "step": 4844
+    },
+    {
+      "epoch": 0.5095507907502597,
+      "grad_norm": 2.933167988889152,
+      "learning_rate": 2.4657734707537567e-06,
+      "loss": 0.997,
+      "step": 4845
+    },
+    {
+      "epoch": 0.509655961192107,
+      "grad_norm": 2.4675557778050945,
+      "learning_rate": 2.464938731441095e-06,
+      "loss": 0.947,
+      "step": 4846
+    },
+    {
+      "epoch": 0.5097611316339543,
+      "grad_norm": 2.084400197522081,
+      "learning_rate": 2.464103996038036e-06,
+      "loss": 0.9936,
+      "step": 4847
+    },
+    {
+      "epoch": 0.5098663020758016,
+      "grad_norm": 2.5000064897079946,
+      "learning_rate": 2.46326926463766e-06,
+      "loss": 0.9836,
+      "step": 4848
+    },
+    {
+      "epoch": 0.509971472517649,
+      "grad_norm": 2.4556021731452953,
+      "learning_rate": 2.462434537333047e-06,
+      "loss": 0.9781,
+      "step": 4849
+    },
+    {
+      "epoch": 0.5100766429594963,
+      "grad_norm": 3.10178764217554,
+      "learning_rate": 2.461599814217275e-06,
+      "loss": 0.9933,
+      "step": 4850
+    },
+    {
+      "epoch": 0.5101818134013435,
+      "grad_norm": 2.50622015636115,
+      "learning_rate": 2.4607650953834225e-06,
+      "loss": 1.0053,
+      "step": 4851
+    },
+    {
+      "epoch": 0.5102869838431908,
+      "grad_norm": 2.9146926707614864,
+      "learning_rate": 2.459930380924566e-06,
+      "loss": 1.0261,
+      "step": 4852
+    },
+    {
+      "epoch": 0.5103921542850381,
+      "grad_norm": 2.2712260753133715,
+      "learning_rate": 2.459095670933783e-06,
+      "loss": 1.0111,
+      "step": 4853
+    },
+    {
+      "epoch": 0.5104973247268855,
+      "grad_norm": 2.0427151120190445,
+      "learning_rate": 2.458260965504151e-06,
+      "loss": 1.0178,
+      "step": 4854
+    },
+    {
+      "epoch": 0.5106024951687328,
+      "grad_norm": 2.615519242206928,
+      "learning_rate": 2.457426264728746e-06,
+      "loss": 1.0209,
+      "step": 4855
+    },
+    {
+      "epoch": 0.5107076656105801,
+      "grad_norm": 2.8016701953883216,
+      "learning_rate": 2.4565915687006426e-06,
+      "loss": 1.0498,
+      "step": 4856
+    },
+    {
+      "epoch": 0.5108128360524274,
+      "grad_norm": 1.909826636778149,
+      "learning_rate": 2.4557568775129157e-06,
+      "loss": 0.9668,
+      "step": 4857
+    },
+    {
+      "epoch": 0.5109180064942748,
+      "grad_norm": 2.0558742173386637,
+      "learning_rate": 2.454922191258642e-06,
+      "loss": 0.9879,
+      "step": 4858
+    },
+    {
+      "epoch": 0.5110231769361221,
+      "grad_norm": 2.0257159398608646,
+      "learning_rate": 2.454087510030894e-06,
+      "loss": 0.9859,
+      "step": 4859
+    },
+    {
+      "epoch": 0.5111283473779694,
+      "grad_norm": 1.7754121266644842,
+      "learning_rate": 2.453252833922745e-06,
+      "loss": 0.9699,
+      "step": 4860
+    },
+    {
+      "epoch": 0.5112335178198167,
+      "grad_norm": 2.4413398534310375,
+      "learning_rate": 2.4524181630272685e-06,
+      "loss": 1.0299,
+      "step": 4861
+    },
+    {
+      "epoch": 0.5113386882616641,
+      "grad_norm": 2.715928878533507,
+      "learning_rate": 2.451583497437535e-06,
+      "loss": 0.9741,
+      "step": 4862
+    },
+    {
+      "epoch": 0.5114438587035114,
+      "grad_norm": 2.3218910000415365,
+      "learning_rate": 2.450748837246619e-06,
+      "loss": 1.0065,
+      "step": 4863
+    },
+    {
+      "epoch": 0.5115490291453587,
+      "grad_norm": 2.2782925199881214,
+      "learning_rate": 2.44991418254759e-06,
+      "loss": 0.9806,
+      "step": 4864
+    },
+    {
+      "epoch": 0.511654199587206,
+      "grad_norm": 2.7149913688207135,
+      "learning_rate": 2.449079533433519e-06,
+      "loss": 1.0209,
+      "step": 4865
+    },
+    {
+      "epoch": 0.5117593700290534,
+      "grad_norm": 2.3390153312282185,
+      "learning_rate": 2.448244889997475e-06,
+      "loss": 0.999,
+      "step": 4866
+    },
+    {
+      "epoch": 0.5118645404709007,
+      "grad_norm": 3.0557721293095157,
+      "learning_rate": 2.447410252332528e-06,
+      "loss": 1.0012,
+      "step": 4867
+    },
+    {
+      "epoch": 0.511969710912748,
+      "grad_norm": 2.866535350728344,
+      "learning_rate": 2.4465756205317466e-06,
+      "loss": 0.992,
+      "step": 4868
+    },
+    {
+      "epoch": 0.5120748813545953,
+      "grad_norm": 2.9450377971992023,
+      "learning_rate": 2.4457409946881986e-06,
+      "loss": 1.0557,
+      "step": 4869
+    },
+    {
+      "epoch": 0.5121800517964427,
+      "grad_norm": 2.786445259165392,
+      "learning_rate": 2.444906374894951e-06,
+      "loss": 1.0568,
+      "step": 4870
+    },
+    {
+      "epoch": 0.51228522223829,
+      "grad_norm": 1.9248634237188782,
+      "learning_rate": 2.4440717612450695e-06,
+      "loss": 0.9872,
+      "step": 4871
+    },
+    {
+      "epoch": 0.5123903926801372,
+      "grad_norm": 3.2932080298333166,
+      "learning_rate": 2.4432371538316226e-06,
+      "loss": 0.9767,
+      "step": 4872
+    },
+    {
+      "epoch": 0.5124955631219845,
+      "grad_norm": 2.5355754300367486,
+      "learning_rate": 2.4424025527476735e-06,
+      "loss": 1.0017,
+      "step": 4873
+    },
+    {
+      "epoch": 0.5126007335638318,
+      "grad_norm": 2.598457779940784,
+      "learning_rate": 2.441567958086288e-06,
+      "loss": 0.9825,
+      "step": 4874
+    },
+    {
+      "epoch": 0.5127059040056792,
+      "grad_norm": 2.252477222253449,
+      "learning_rate": 2.4407333699405285e-06,
+      "loss": 1.0299,
+      "step": 4875
+    },
+    {
+      "epoch": 0.5128110744475265,
+      "grad_norm": 1.61574628720382,
+      "learning_rate": 2.4398987884034574e-06,
+      "loss": 0.9719,
+      "step": 4876
+    },
+    {
+      "epoch": 0.5129162448893738,
+      "grad_norm": 2.9461045046867795,
+      "learning_rate": 2.43906421356814e-06,
+      "loss": 1.0177,
+      "step": 4877
+    },
+    {
+      "epoch": 0.5130214153312211,
+      "grad_norm": 2.6086241827887715,
+      "learning_rate": 2.4382296455276355e-06,
+      "loss": 0.9679,
+      "step": 4878
+    },
+    {
+      "epoch": 0.5131265857730685,
+      "grad_norm": 2.469431110622542,
+      "learning_rate": 2.4373950843750053e-06,
+      "loss": 0.9654,
+      "step": 4879
+    },
+    {
+      "epoch": 0.5132317562149158,
+      "grad_norm": 2.4965960753098253,
+      "learning_rate": 2.4365605302033094e-06,
+      "loss": 0.9802,
+      "step": 4880
+    },
+    {
+      "epoch": 0.5133369266567631,
+      "grad_norm": 2.0319790709927505,
+      "learning_rate": 2.435725983105608e-06,
+      "loss": 0.9885,
+      "step": 4881
+    },
+    {
+      "epoch": 0.5134420970986104,
+      "grad_norm": 2.4000275308260544,
+      "learning_rate": 2.4348914431749578e-06,
+      "loss": 0.9208,
+      "step": 4882
+    },
+    {
+      "epoch": 0.5135472675404578,
+      "grad_norm": 2.317952586045972,
+      "learning_rate": 2.434056910504418e-06,
+      "loss": 0.9975,
+      "step": 4883
+    },
+    {
+      "epoch": 0.5136524379823051,
+      "grad_norm": 2.2760790757419778,
+      "learning_rate": 2.4332223851870453e-06,
+      "loss": 1.0071,
+      "step": 4884
+    },
+    {
+      "epoch": 0.5137576084241524,
+      "grad_norm": 2.4027564608452723,
+      "learning_rate": 2.4323878673158937e-06,
+      "loss": 1.0331,
+      "step": 4885
+    },
+    {
+      "epoch": 0.5138627788659997,
+      "grad_norm": 1.8639452403440218,
+      "learning_rate": 2.431553356984022e-06,
+      "loss": 0.9873,
+      "step": 4886
+    },
+    {
+      "epoch": 0.5139679493078471,
+      "grad_norm": 2.5864359775236245,
+      "learning_rate": 2.430718854284482e-06,
+      "loss": 0.9993,
+      "step": 4887
+    },
+    {
+      "epoch": 0.5140731197496944,
+      "grad_norm": 1.8024207909165022,
+      "learning_rate": 2.429884359310328e-06,
+      "loss": 1.0033,
+      "step": 4888
+    },
+    {
+      "epoch": 0.5141782901915417,
+      "grad_norm": 2.4114408144418538,
+      "learning_rate": 2.4290498721546123e-06,
+      "loss": 1.0138,
+      "step": 4889
+    },
+    {
+      "epoch": 0.514283460633389,
+      "grad_norm": 2.8587403555199553,
+      "learning_rate": 2.4282153929103874e-06,
+      "loss": 0.9866,
+      "step": 4890
+    },
+    {
+      "epoch": 0.5143886310752364,
+      "grad_norm": 2.2763471355900706,
+      "learning_rate": 2.427380921670704e-06,
+      "loss": 0.9761,
+      "step": 4891
+    },
+    {
+      "epoch": 0.5144938015170836,
+      "grad_norm": 2.7630996690467073,
+      "learning_rate": 2.426546458528612e-06,
+      "loss": 1.0034,
+      "step": 4892
+    },
+    {
+      "epoch": 0.5145989719589309,
+      "grad_norm": 2.5503825990043034,
+      "learning_rate": 2.42571200357716e-06,
+      "loss": 0.9842,
+      "step": 4893
+    },
+    {
+      "epoch": 0.5147041424007782,
+      "grad_norm": 1.707781885008851,
+      "learning_rate": 2.4248775569093968e-06,
+      "loss": 0.9673,
+      "step": 4894
+    },
+    {
+      "epoch": 0.5148093128426255,
+      "grad_norm": 2.372412499450598,
+      "learning_rate": 2.4240431186183695e-06,
+      "loss": 1.0197,
+      "step": 4895
+    },
+    {
+      "epoch": 0.5149144832844729,
+      "grad_norm": 2.6726883190728916,
+      "learning_rate": 2.423208688797125e-06,
+      "loss": 0.9981,
+      "step": 4896
+    },
+    {
+      "epoch": 0.5150196537263202,
+      "grad_norm": 2.672102772646679,
+      "learning_rate": 2.422374267538708e-06,
+      "loss": 0.9785,
+      "step": 4897
+    },
+    {
+      "epoch": 0.5151248241681675,
+      "grad_norm": 3.3567918073562586,
+      "learning_rate": 2.4215398549361632e-06,
+      "loss": 0.9869,
+      "step": 4898
+    },
+    {
+      "epoch": 0.5152299946100148,
+      "grad_norm": 3.2146007931473184,
+      "learning_rate": 2.420705451082533e-06,
+      "loss": 0.9755,
+      "step": 4899
+    },
+    {
+      "epoch": 0.5153351650518622,
+      "grad_norm": 2.8601027840979563,
+      "learning_rate": 2.4198710560708623e-06,
+      "loss": 1.0102,
+      "step": 4900
+    },
+    {
+      "epoch": 0.5154403354937095,
+      "grad_norm": 2.49782936985743,
+      "learning_rate": 2.419036669994191e-06,
+      "loss": 1.0102,
+      "step": 4901
+    },
+    {
+      "epoch": 0.5155455059355568,
+      "grad_norm": 2.666513374876871,
+      "learning_rate": 2.41820229294556e-06,
+      "loss": 0.9846,
+      "step": 4902
+    },
+    {
+      "epoch": 0.5156506763774041,
+      "grad_norm": 2.1553652633705274,
+      "learning_rate": 2.4173679250180083e-06,
+      "loss": 0.9994,
+      "step": 4903
+    },
+    {
+      "epoch": 0.5157558468192515,
+      "grad_norm": 1.9176350923655092,
+      "learning_rate": 2.4165335663045753e-06,
+      "loss": 0.9929,
+      "step": 4904
+    },
+    {
+      "epoch": 0.5158610172610988,
+      "grad_norm": 2.8881593227917377,
+      "learning_rate": 2.4156992168982985e-06,
+      "loss": 1.0123,
+      "step": 4905
+    },
+    {
+      "epoch": 0.5159661877029461,
+      "grad_norm": 1.7026322357589312,
+      "learning_rate": 2.4148648768922133e-06,
+      "loss": 0.9522,
+      "step": 4906
+    },
+    {
+      "epoch": 0.5160713581447934,
+      "grad_norm": 1.9182156402806023,
+      "learning_rate": 2.4140305463793557e-06,
+      "loss": 0.9721,
+      "step": 4907
+    },
+    {
+      "epoch": 0.5161765285866408,
+      "grad_norm": 2.630435818921885,
+      "learning_rate": 2.4131962254527592e-06,
+      "loss": 0.9998,
+      "step": 4908
+    },
+    {
+      "epoch": 0.5162816990284881,
+      "grad_norm": 2.639541492793874,
+      "learning_rate": 2.41236191420546e-06,
+      "loss": 1.0166,
+      "step": 4909
+    },
+    {
+      "epoch": 0.5163868694703354,
+      "grad_norm": 2.193440397829881,
+      "learning_rate": 2.4115276127304866e-06,
+      "loss": 1.0148,
+      "step": 4910
+    },
+    {
+      "epoch": 0.5164920399121827,
+      "grad_norm": 1.9956942722452817,
+      "learning_rate": 2.4106933211208723e-06,
+      "loss": 0.9871,
+      "step": 4911
+    },
+    {
+      "epoch": 0.51659721035403,
+      "grad_norm": 2.841990398589798,
+      "learning_rate": 2.4098590394696452e-06,
+      "loss": 1.01,
+      "step": 4912
+    },
+    {
+      "epoch": 0.5167023807958773,
+      "grad_norm": 2.4733117430723324,
+      "learning_rate": 2.4090247678698362e-06,
+      "loss": 1.0348,
+      "step": 4913
+    },
+    {
+      "epoch": 0.5168075512377246,
+      "grad_norm": 1.852569961733493,
+      "learning_rate": 2.408190506414473e-06,
+      "loss": 0.9695,
+      "step": 4914
+    },
+    {
+      "epoch": 0.5169127216795719,
+      "grad_norm": 2.4532174848303683,
+      "learning_rate": 2.4073562551965814e-06,
+      "loss": 0.9901,
+      "step": 4915
+    },
+    {
+      "epoch": 0.5170178921214192,
+      "grad_norm": 2.5251790955216697,
+      "learning_rate": 2.4065220143091863e-06,
+      "loss": 1.0092,
+      "step": 4916
+    },
+    {
+      "epoch": 0.5171230625632666,
+      "grad_norm": 2.76475785288874,
+      "learning_rate": 2.405687783845313e-06,
+      "loss": 0.9948,
+      "step": 4917
+    },
+    {
+      "epoch": 0.5172282330051139,
+      "grad_norm": 2.7685770717204776,
+      "learning_rate": 2.4048535638979844e-06,
+      "loss": 0.9939,
+      "step": 4918
+    },
+    {
+      "epoch": 0.5173334034469612,
+      "grad_norm": 2.2946067444852036,
+      "learning_rate": 2.4040193545602232e-06,
+      "loss": 0.9687,
+      "step": 4919
+    },
+    {
+      "epoch": 0.5174385738888085,
+      "grad_norm": 2.645304674423841,
+      "learning_rate": 2.403185155925049e-06,
+      "loss": 1.0173,
+      "step": 4920
+    },
+    {
+      "epoch": 0.5175437443306559,
+      "grad_norm": 2.629339889738927,
+      "learning_rate": 2.4023509680854822e-06,
+      "loss": 0.9829,
+      "step": 4921
+    },
+    {
+      "epoch": 0.5176489147725032,
+      "grad_norm": 2.1954075878052035,
+      "learning_rate": 2.40151679113454e-06,
+      "loss": 0.9908,
+      "step": 4922
+    },
+    {
+      "epoch": 0.5177540852143505,
+      "grad_norm": 2.095867879804772,
+      "learning_rate": 2.400682625165242e-06,
+      "loss": 0.9969,
+      "step": 4923
+    },
+    {
+      "epoch": 0.5178592556561978,
+      "grad_norm": 1.9455070815665647,
+      "learning_rate": 2.399848470270602e-06,
+      "loss": 0.9971,
+      "step": 4924
+    },
+    {
+      "epoch": 0.5179644260980452,
+      "grad_norm": 2.507117651664635,
+      "learning_rate": 2.399014326543637e-06,
+      "loss": 1.0045,
+      "step": 4925
+    },
+    {
+      "epoch": 0.5180695965398925,
+      "grad_norm": 2.4607100953155503,
+      "learning_rate": 2.398180194077357e-06,
+      "loss": 1.0235,
+      "step": 4926
+    },
+    {
+      "epoch": 0.5181747669817398,
+      "grad_norm": 2.777674225709361,
+      "learning_rate": 2.3973460729647775e-06,
+      "loss": 1.0138,
+      "step": 4927
+    },
+    {
+      "epoch": 0.5182799374235871,
+      "grad_norm": 2.313723654298759,
+      "learning_rate": 2.396511963298908e-06,
+      "loss": 0.9688,
+      "step": 4928
+    },
+    {
+      "epoch": 0.5183851078654345,
+      "grad_norm": 2.3190874101733803,
+      "learning_rate": 2.395677865172759e-06,
+      "loss": 1.0122,
+      "step": 4929
+    },
+    {
+      "epoch": 0.5184902783072818,
+      "grad_norm": 2.7640107446846773,
+      "learning_rate": 2.394843778679338e-06,
+      "loss": 0.9832,
+      "step": 4930
+    },
+    {
+      "epoch": 0.5185954487491291,
+      "grad_norm": 2.4081953739998534,
+      "learning_rate": 2.3940097039116523e-06,
+      "loss": 1.014,
+      "step": 4931
+    },
+    {
+      "epoch": 0.5187006191909764,
+      "grad_norm": 1.923457778269713,
+      "learning_rate": 2.3931756409627084e-06,
+      "loss": 0.9874,
+      "step": 4932
+    },
+    {
+      "epoch": 0.5188057896328236,
+      "grad_norm": 1.8553693404082456,
+      "learning_rate": 2.3923415899255105e-06,
+      "loss": 0.9887,
+      "step": 4933
+    },
+    {
+      "epoch": 0.518910960074671,
+      "grad_norm": 2.7194260650322684,
+      "learning_rate": 2.3915075508930615e-06,
+      "loss": 0.9387,
+      "step": 4934
+    },
+    {
+      "epoch": 0.5190161305165183,
+      "grad_norm": 2.652242421666457,
+      "learning_rate": 2.3906735239583623e-06,
+      "loss": 1.0244,
+      "step": 4935
+    },
+    {
+      "epoch": 0.5191213009583656,
+      "grad_norm": 2.549188854335063,
+      "learning_rate": 2.389839509214416e-06,
+      "loss": 1.0246,
+      "step": 4936
+    },
+    {
+      "epoch": 0.5192264714002129,
+      "grad_norm": 2.2665938281306155,
+      "learning_rate": 2.3890055067542195e-06,
+      "loss": 1.0106,
+      "step": 4937
+    },
+    {
+      "epoch": 0.5193316418420603,
+      "grad_norm": 2.8241998887282795,
+      "learning_rate": 2.3881715166707716e-06,
+      "loss": 1.0155,
+      "step": 4938
+    },
+    {
+      "epoch": 0.5194368122839076,
+      "grad_norm": 2.152806090447936,
+      "learning_rate": 2.387337539057068e-06,
+      "loss": 1.0189,
+      "step": 4939
+    },
+    {
+      "epoch": 0.5195419827257549,
+      "grad_norm": 2.172323988446627,
+      "learning_rate": 2.3865035740061037e-06,
+      "loss": 1.0695,
+      "step": 4940
+    },
+    {
+      "epoch": 0.5196471531676022,
+      "grad_norm": 2.728428245649283,
+      "learning_rate": 2.385669621610873e-06,
+      "loss": 1.0112,
+      "step": 4941
+    },
+    {
+      "epoch": 0.5197523236094496,
+      "grad_norm": 2.124385885202202,
+      "learning_rate": 2.384835681964368e-06,
+      "loss": 1.0248,
+      "step": 4942
+    },
+    {
+      "epoch": 0.5198574940512969,
+      "grad_norm": 2.313862649372938,
+      "learning_rate": 2.3840017551595785e-06,
+      "loss": 0.9969,
+      "step": 4943
+    },
+    {
+      "epoch": 0.5199626644931442,
+      "grad_norm": 2.4524723823538954,
+      "learning_rate": 2.3831678412894947e-06,
+      "loss": 0.9844,
+      "step": 4944
+    },
+    {
+      "epoch": 0.5200678349349915,
+      "grad_norm": 2.5132178605505486,
+      "learning_rate": 2.3823339404471037e-06,
+      "loss": 1.04,
+      "step": 4945
+    },
+    {
+      "epoch": 0.5201730053768389,
+      "grad_norm": 2.4402520311081197,
+      "learning_rate": 2.3815000527253933e-06,
+      "loss": 0.9824,
+      "step": 4946
+    },
+    {
+      "epoch": 0.5202781758186862,
+      "grad_norm": 2.1382688241450247,
+      "learning_rate": 2.380666178217347e-06,
+      "loss": 1.0087,
+      "step": 4947
+    },
+    {
+      "epoch": 0.5203833462605335,
+      "grad_norm": 2.538021983793176,
+      "learning_rate": 2.3798323170159487e-06,
+      "loss": 0.9869,
+      "step": 4948
+    },
+    {
+      "epoch": 0.5204885167023808,
+      "grad_norm": 1.851342803348307,
+      "learning_rate": 2.3789984692141798e-06,
+      "loss": 0.9851,
+      "step": 4949
+    },
+    {
+      "epoch": 0.5205936871442282,
+      "grad_norm": 2.176853442535673,
+      "learning_rate": 2.3781646349050227e-06,
+      "loss": 0.9821,
+      "step": 4950
+    },
+    {
+      "epoch": 0.5206988575860755,
+      "grad_norm": 2.3979243648827984,
+      "learning_rate": 2.377330814181455e-06,
+      "loss": 0.9397,
+      "step": 4951
+    },
+    {
+      "epoch": 0.5208040280279228,
+      "grad_norm": 2.0633810413062212,
+      "learning_rate": 2.3764970071364548e-06,
+      "loss": 1.0225,
+      "step": 4952
+    },
+    {
+      "epoch": 0.52090919846977,
+      "grad_norm": 1.7807728183386664,
+      "learning_rate": 2.3756632138629977e-06,
+      "loss": 0.9851,
+      "step": 4953
+    },
+    {
+      "epoch": 0.5210143689116173,
+      "grad_norm": 2.2949369278058382,
+      "learning_rate": 2.3748294344540575e-06,
+      "loss": 0.9848,
+      "step": 4954
+    },
+    {
+      "epoch": 0.5211195393534647,
+      "grad_norm": 2.8758134936809965,
+      "learning_rate": 2.373995669002608e-06,
+      "loss": 1.0296,
+      "step": 4955
+    },
+    {
+      "epoch": 0.521224709795312,
+      "grad_norm": 2.7047106011767563,
+      "learning_rate": 2.37316191760162e-06,
+      "loss": 1.0089,
+      "step": 4956
+    },
+    {
+      "epoch": 0.5213298802371593,
+      "grad_norm": 2.4083545951327463,
+      "learning_rate": 2.3723281803440642e-06,
+      "loss": 0.9828,
+      "step": 4957
+    },
+    {
+      "epoch": 0.5214350506790066,
+      "grad_norm": 2.3483276709030703,
+      "learning_rate": 2.371494457322907e-06,
+      "loss": 0.9683,
+      "step": 4958
+    },
+    {
+      "epoch": 0.521540221120854,
+      "grad_norm": 2.6165413672922004,
+      "learning_rate": 2.3706607486311166e-06,
+      "loss": 0.9807,
+      "step": 4959
+    },
+    {
+      "epoch": 0.5216453915627013,
+      "grad_norm": 2.7199298338176434,
+      "learning_rate": 2.3698270543616583e-06,
+      "loss": 0.9758,
+      "step": 4960
+    },
+    {
+      "epoch": 0.5217505620045486,
+      "grad_norm": 1.6020078122627697,
+      "learning_rate": 2.368993374607494e-06,
+      "loss": 0.993,
+      "step": 4961
+    },
+    {
+      "epoch": 0.5218557324463959,
+      "grad_norm": 3.0119525509546476,
+      "learning_rate": 2.3681597094615863e-06,
+      "loss": 1.04,
+      "step": 4962
+    },
+    {
+      "epoch": 0.5219609028882433,
+      "grad_norm": 2.8336389633739376,
+      "learning_rate": 2.367326059016894e-06,
+      "loss": 0.9782,
+      "step": 4963
+    },
+    {
+      "epoch": 0.5220660733300906,
+      "grad_norm": 3.302155303408267,
+      "learning_rate": 2.3664924233663776e-06,
+      "loss": 0.9895,
+      "step": 4964
+    },
+    {
+      "epoch": 0.5221712437719379,
+      "grad_norm": 2.0928625598039674,
+      "learning_rate": 2.3656588026029938e-06,
+      "loss": 1.042,
+      "step": 4965
+    },
+    {
+      "epoch": 0.5222764142137852,
+      "grad_norm": 2.5158469098657816,
+      "learning_rate": 2.3648251968196964e-06,
+      "loss": 0.9834,
+      "step": 4966
+    },
+    {
+      "epoch": 0.5223815846556326,
+      "grad_norm": 2.8574951074085377,
+      "learning_rate": 2.36399160610944e-06,
+      "loss": 0.9998,
+      "step": 4967
+    },
+    {
+      "epoch": 0.5224867550974799,
+      "grad_norm": 1.7337887982119702,
+      "learning_rate": 2.3631580305651756e-06,
+      "loss": 1.0026,
+      "step": 4968
+    },
+    {
+      "epoch": 0.5225919255393272,
+      "grad_norm": 3.308910933603135,
+      "learning_rate": 2.362324470279854e-06,
+      "loss": 0.9672,
+      "step": 4969
+    },
+    {
+      "epoch": 0.5226970959811745,
+      "grad_norm": 2.6333972565437556,
+      "learning_rate": 2.3614909253464235e-06,
+      "loss": 0.9774,
+      "step": 4970
+    },
+    {
+      "epoch": 0.5228022664230219,
+      "grad_norm": 2.09659946546361,
+      "learning_rate": 2.360657395857831e-06,
+      "loss": 0.9807,
+      "step": 4971
+    },
+    {
+      "epoch": 0.5229074368648692,
+      "grad_norm": 2.6391317053576806,
+      "learning_rate": 2.3598238819070206e-06,
+      "loss": 1.0443,
+      "step": 4972
+    },
+    {
+      "epoch": 0.5230126073067164,
+      "grad_norm": 2.242550715788799,
+      "learning_rate": 2.358990383586937e-06,
+      "loss": 0.966,
+      "step": 4973
+    },
+    {
+      "epoch": 0.5231177777485637,
+      "grad_norm": 3.108913453931425,
+      "learning_rate": 2.3581569009905204e-06,
+      "loss": 0.9724,
+      "step": 4974
+    },
+    {
+      "epoch": 0.523222948190411,
+      "grad_norm": 1.9901033912663888,
+      "learning_rate": 2.357323434210712e-06,
+      "loss": 0.9695,
+      "step": 4975
+    },
+    {
+      "epoch": 0.5233281186322584,
+      "grad_norm": 2.168744011583421,
+      "learning_rate": 2.3564899833404496e-06,
+      "loss": 0.9924,
+      "step": 4976
+    },
+    {
+      "epoch": 0.5234332890741057,
+      "grad_norm": 2.549473243082255,
+      "learning_rate": 2.355656548472667e-06,
+      "loss": 1.0382,
+      "step": 4977
+    },
+    {
+      "epoch": 0.523538459515953,
+      "grad_norm": 2.5199612478036957,
+      "learning_rate": 2.3548231297003017e-06,
+      "loss": 1.0082,
+      "step": 4978
+    },
+    {
+      "epoch": 0.5236436299578003,
+      "grad_norm": 2.3189840650193148,
+      "learning_rate": 2.3539897271162853e-06,
+      "loss": 0.9743,
+      "step": 4979
+    },
+    {
+      "epoch": 0.5237488003996477,
+      "grad_norm": 1.8363174051887219,
+      "learning_rate": 2.3531563408135482e-06,
+      "loss": 1.014,
+      "step": 4980
+    },
+    {
+      "epoch": 0.523853970841495,
+      "grad_norm": 2.136662771433713,
+      "learning_rate": 2.3523229708850194e-06,
+      "loss": 1.0037,
+      "step": 4981
+    },
+    {
+      "epoch": 0.5239591412833423,
+      "grad_norm": 2.5855613260565393,
+      "learning_rate": 2.351489617423627e-06,
+      "loss": 0.9987,
+      "step": 4982
+    },
+    {
+      "epoch": 0.5240643117251896,
+      "grad_norm": 2.3432603718739737,
+      "learning_rate": 2.3506562805222966e-06,
+      "loss": 0.9765,
+      "step": 4983
+    },
+    {
+      "epoch": 0.524169482167037,
+      "grad_norm": 2.6466370290433874,
+      "learning_rate": 2.3498229602739506e-06,
+      "loss": 0.99,
+      "step": 4984
+    },
+    {
+      "epoch": 0.5242746526088843,
+      "grad_norm": 3.0276801019190107,
+      "learning_rate": 2.348989656771511e-06,
+      "loss": 0.9581,
+      "step": 4985
+    },
+    {
+      "epoch": 0.5243798230507316,
+      "grad_norm": 3.1030135072125486,
+      "learning_rate": 2.348156370107897e-06,
+      "loss": 0.9684,
+      "step": 4986
+    },
+    {
+      "epoch": 0.5244849934925789,
+      "grad_norm": 1.6705531610760058,
+      "learning_rate": 2.3473231003760283e-06,
+      "loss": 1.0041,
+      "step": 4987
+    },
+    {
+      "epoch": 0.5245901639344263,
+      "grad_norm": 3.195727549016848,
+      "learning_rate": 2.3464898476688198e-06,
+      "loss": 1.0116,
+      "step": 4988
+    },
+    {
+      "epoch": 0.5246953343762736,
+      "grad_norm": 2.4173618786228532,
+      "learning_rate": 2.345656612079186e-06,
+      "loss": 0.9629,
+      "step": 4989
+    },
+    {
+      "epoch": 0.5248005048181209,
+      "grad_norm": 2.6428711784981602,
+      "learning_rate": 2.3448233937000385e-06,
+      "loss": 1.0233,
+      "step": 4990
+    },
+    {
+      "epoch": 0.5249056752599682,
+      "grad_norm": 1.912207664947132,
+      "learning_rate": 2.343990192624288e-06,
+      "loss": 0.9782,
+      "step": 4991
+    },
+    {
+      "epoch": 0.5250108457018156,
+      "grad_norm": 2.2091019995514523,
+      "learning_rate": 2.3431570089448434e-06,
+      "loss": 0.9875,
+      "step": 4992
+    },
+    {
+      "epoch": 0.5251160161436629,
+      "grad_norm": 2.7977308048194285,
+      "learning_rate": 2.3423238427546105e-06,
+      "loss": 1.0392,
+      "step": 4993
+    },
+    {
+      "epoch": 0.5252211865855101,
+      "grad_norm": 2.639342545737909,
+      "learning_rate": 2.341490694146494e-06,
+      "loss": 1.0452,
+      "step": 4994
+    },
+    {
+      "epoch": 0.5253263570273574,
+      "grad_norm": 2.2851218048922695,
+      "learning_rate": 2.3406575632133953e-06,
+      "loss": 0.9841,
+      "step": 4995
+    },
+    {
+      "epoch": 0.5254315274692047,
+      "grad_norm": 1.9739459077968367,
+      "learning_rate": 2.339824450048218e-06,
+      "loss": 0.9819,
+      "step": 4996
+    },
+    {
+      "epoch": 0.5255366979110521,
+      "grad_norm": 2.0317174932305333,
+      "learning_rate": 2.3389913547438586e-06,
+      "loss": 1.0313,
+      "step": 4997
+    },
+    {
+      "epoch": 0.5256418683528994,
+      "grad_norm": 2.7337990453530154,
+      "learning_rate": 2.338158277393213e-06,
+      "loss": 1.0307,
+      "step": 4998
+    },
+    {
+      "epoch": 0.5257470387947467,
+      "grad_norm": 2.665374217415374,
+      "learning_rate": 2.337325218089177e-06,
+      "loss": 0.9793,
+      "step": 4999
+    },
+    {
+      "epoch": 0.525852209236594,
+      "grad_norm": 2.167783407256349,
+      "learning_rate": 2.3364921769246423e-06,
+      "loss": 0.9815,
+      "step": 5000
+    },
+    {
+      "epoch": 0.5259573796784414,
+      "grad_norm": 1.9578480148487534,
+      "learning_rate": 2.335659153992501e-06,
+      "loss": 0.9851,
+      "step": 5001
+    },
+    {
+      "epoch": 0.5260625501202887,
+      "grad_norm": 1.8750731661000946,
+      "learning_rate": 2.33482614938564e-06,
+      "loss": 1.0107,
+      "step": 5002
+    },
+    {
+      "epoch": 0.526167720562136,
+      "grad_norm": 2.7268491605591705,
+      "learning_rate": 2.3339931631969473e-06,
+      "loss": 1.0255,
+      "step": 5003
+    },
+    {
+      "epoch": 0.5262728910039833,
+      "grad_norm": 2.3624254794486292,
+      "learning_rate": 2.333160195519306e-06,
+      "loss": 1.0034,
+      "step": 5004
+    },
+    {
+      "epoch": 0.5263780614458307,
+      "grad_norm": 2.0278498795912956,
+      "learning_rate": 2.3323272464455987e-06,
+      "loss": 0.9947,
+      "step": 5005
+    },
+    {
+      "epoch": 0.526483231887678,
+      "grad_norm": 2.2289155760280592,
+      "learning_rate": 2.331494316068706e-06,
+      "loss": 0.9749,
+      "step": 5006
+    },
+    {
+      "epoch": 0.5265884023295253,
+      "grad_norm": 2.4790681587132335,
+      "learning_rate": 2.3306614044815066e-06,
+      "loss": 0.9922,
+      "step": 5007
+    },
+    {
+      "epoch": 0.5266935727713726,
+      "grad_norm": 2.253254009909184,
+      "learning_rate": 2.329828511776876e-06,
+      "loss": 0.9891,
+      "step": 5008
+    },
+    {
+      "epoch": 0.52679874321322,
+      "grad_norm": 2.4459741279261102,
+      "learning_rate": 2.328995638047687e-06,
+      "loss": 1.0001,
+      "step": 5009
+    },
+    {
+      "epoch": 0.5269039136550673,
+      "grad_norm": 2.9185918204182313,
+      "learning_rate": 2.328162783386814e-06,
+      "loss": 0.9973,
+      "step": 5010
+    },
+    {
+      "epoch": 0.5270090840969146,
+      "grad_norm": 2.0661644888279582,
+      "learning_rate": 2.327329947887125e-06,
+      "loss": 1.0438,
+      "step": 5011
+    },
+    {
+      "epoch": 0.5271142545387619,
+      "grad_norm": 1.9485234437575496,
+      "learning_rate": 2.3264971316414893e-06,
+      "loss": 1.0256,
+      "step": 5012
+    },
+    {
+      "epoch": 0.5272194249806093,
+      "grad_norm": 2.178434629119489,
+      "learning_rate": 2.325664334742771e-06,
+      "loss": 1.0014,
+      "step": 5013
+    },
+    {
+      "epoch": 0.5273245954224565,
+      "grad_norm": 2.0382608544824095,
+      "learning_rate": 2.3248315572838316e-06,
+      "loss": 0.9811,
+      "step": 5014
+    },
+    {
+      "epoch": 0.5274297658643038,
+      "grad_norm": 2.2529667292304634,
+      "learning_rate": 2.323998799357536e-06,
+      "loss": 0.9599,
+      "step": 5015
+    },
+    {
+      "epoch": 0.5275349363061511,
+      "grad_norm": 2.1875904622022673,
+      "learning_rate": 2.3231660610567415e-06,
+      "loss": 0.9548,
+      "step": 5016
+    },
+    {
+      "epoch": 0.5276401067479984,
+      "grad_norm": 2.294988151424705,
+      "learning_rate": 2.322333342474305e-06,
+      "loss": 1.0398,
+      "step": 5017
+    },
+    {
+      "epoch": 0.5277452771898458,
+      "grad_norm": 2.593613259374422,
+      "learning_rate": 2.32150064370308e-06,
+      "loss": 0.9981,
+      "step": 5018
+    },
+    {
+      "epoch": 0.5278504476316931,
+      "grad_norm": 2.4843528327017097,
+      "learning_rate": 2.3206679648359206e-06,
+      "loss": 0.9869,
+      "step": 5019
+    },
+    {
+      "epoch": 0.5279556180735404,
+      "grad_norm": 2.368580138964457,
+      "learning_rate": 2.3198353059656763e-06,
+      "loss": 0.942,
+      "step": 5020
+    },
+    {
+      "epoch": 0.5280607885153877,
+      "grad_norm": 3.1046397593748103,
+      "learning_rate": 2.319002667185195e-06,
+      "loss": 0.9921,
+      "step": 5021
+    },
+    {
+      "epoch": 0.5281659589572351,
+      "grad_norm": 2.386039450298401,
+      "learning_rate": 2.318170048587322e-06,
+      "loss": 0.9925,
+      "step": 5022
+    },
+    {
+      "epoch": 0.5282711293990824,
+      "grad_norm": 2.1962062400668736,
+      "learning_rate": 2.3173374502649006e-06,
+      "loss": 1.0175,
+      "step": 5023
+    },
+    {
+      "epoch": 0.5283762998409297,
+      "grad_norm": 1.897357930598202,
+      "learning_rate": 2.316504872310773e-06,
+      "loss": 1.015,
+      "step": 5024
+    },
+    {
+      "epoch": 0.528481470282777,
+      "grad_norm": 2.699356451977988,
+      "learning_rate": 2.315672314817778e-06,
+      "loss": 1.0146,
+      "step": 5025
+    },
+    {
+      "epoch": 0.5285866407246244,
+      "grad_norm": 2.291400597221481,
+      "learning_rate": 2.3148397778787514e-06,
+      "loss": 1.003,
+      "step": 5026
+    },
+    {
+      "epoch": 0.5286918111664717,
+      "grad_norm": 1.81066716415222,
+      "learning_rate": 2.3140072615865273e-06,
+      "loss": 0.9671,
+      "step": 5027
+    },
+    {
+      "epoch": 0.528796981608319,
+      "grad_norm": 2.428779670922376,
+      "learning_rate": 2.3131747660339396e-06,
+      "loss": 1.0034,
+      "step": 5028
+    },
+    {
+      "epoch": 0.5289021520501663,
+      "grad_norm": 2.2304748543670074,
+      "learning_rate": 2.3123422913138165e-06,
+      "loss": 0.9739,
+      "step": 5029
+    },
+    {
+      "epoch": 0.5290073224920137,
+      "grad_norm": 2.107958029699162,
+      "learning_rate": 2.3115098375189854e-06,
+      "loss": 1.0084,
+      "step": 5030
+    },
+    {
+      "epoch": 0.529112492933861,
+      "grad_norm": 3.179847886514185,
+      "learning_rate": 2.310677404742272e-06,
+      "loss": 1.0061,
+      "step": 5031
+    },
+    {
+      "epoch": 0.5292176633757083,
+      "grad_norm": 1.720669464426383,
+      "learning_rate": 2.309844993076498e-06,
+      "loss": 0.9939,
+      "step": 5032
+    },
+    {
+      "epoch": 0.5293228338175556,
+      "grad_norm": 2.00459599458911,
+      "learning_rate": 2.3090126026144847e-06,
+      "loss": 0.9521,
+      "step": 5033
+    },
+    {
+      "epoch": 0.5294280042594028,
+      "grad_norm": 2.3013300137074255,
+      "learning_rate": 2.3081802334490502e-06,
+      "loss": 0.9883,
+      "step": 5034
+    },
+    {
+      "epoch": 0.5295331747012502,
+      "grad_norm": 2.3045482143367697,
+      "learning_rate": 2.30734788567301e-06,
+      "loss": 0.9919,
+      "step": 5035
+    },
+    {
+      "epoch": 0.5296383451430975,
+      "grad_norm": 1.67169847842081,
+      "learning_rate": 2.3065155593791756e-06,
+      "loss": 0.9767,
+      "step": 5036
+    },
+    {
+      "epoch": 0.5297435155849448,
+      "grad_norm": 2.894233184600396,
+      "learning_rate": 2.3056832546603607e-06,
+      "loss": 1.0191,
+      "step": 5037
+    },
+    {
+      "epoch": 0.5298486860267921,
+      "grad_norm": 2.011314434528289,
+      "learning_rate": 2.3048509716093723e-06,
+      "loss": 0.9874,
+      "step": 5038
+    },
+    {
+      "epoch": 0.5299538564686395,
+      "grad_norm": 2.2683865903528293,
+      "learning_rate": 2.3040187103190165e-06,
+      "loss": 0.9704,
+      "step": 5039
+    },
+    {
+      "epoch": 0.5300590269104868,
+      "grad_norm": 2.092218000336117,
+      "learning_rate": 2.303186470882097e-06,
+      "loss": 0.9837,
+      "step": 5040
+    },
+    {
+      "epoch": 0.5301641973523341,
+      "grad_norm": 2.5135197465867765,
+      "learning_rate": 2.302354253391414e-06,
+      "loss": 0.9526,
+      "step": 5041
+    },
+    {
+      "epoch": 0.5302693677941814,
+      "grad_norm": 2.572619453041509,
+      "learning_rate": 2.301522057939768e-06,
+      "loss": 0.9711,
+      "step": 5042
+    },
+    {
+      "epoch": 0.5303745382360288,
+      "grad_norm": 2.9257982055535323,
+      "learning_rate": 2.3006898846199544e-06,
+      "loss": 1.0119,
+      "step": 5043
+    },
+    {
+      "epoch": 0.5304797086778761,
+      "grad_norm": 2.7098826216208995,
+      "learning_rate": 2.299857733524767e-06,
+      "loss": 1.0066,
+      "step": 5044
+    },
+    {
+      "epoch": 0.5305848791197234,
+      "grad_norm": 2.727684751730271,
+      "learning_rate": 2.299025604746997e-06,
+      "loss": 0.9946,
+      "step": 5045
+    },
+    {
+      "epoch": 0.5306900495615707,
+      "grad_norm": 2.3317306284613957,
+      "learning_rate": 2.2981934983794324e-06,
+      "loss": 0.9265,
+      "step": 5046
+    },
+    {
+      "epoch": 0.5307952200034181,
+      "grad_norm": 2.5826246634345753,
+      "learning_rate": 2.2973614145148616e-06,
+      "loss": 1.0235,
+      "step": 5047
+    },
+    {
+      "epoch": 0.5309003904452654,
+      "grad_norm": 2.433635645176636,
+      "learning_rate": 2.2965293532460673e-06,
+      "loss": 0.9693,
+      "step": 5048
+    },
+    {
+      "epoch": 0.5310055608871127,
+      "grad_norm": 2.259538998753554,
+      "learning_rate": 2.2956973146658303e-06,
+      "loss": 0.9975,
+      "step": 5049
+    },
+    {
+      "epoch": 0.53111073132896,
+      "grad_norm": 2.3054114447327847,
+      "learning_rate": 2.294865298866929e-06,
+      "loss": 0.991,
+      "step": 5050
+    },
+    {
+      "epoch": 0.5312159017708074,
+      "grad_norm": 3.420364728806349,
+      "learning_rate": 2.294033305942141e-06,
+      "loss": 1.0031,
+      "step": 5051
+    },
+    {
+      "epoch": 0.5313210722126547,
+      "grad_norm": 2.8657151853369234,
+      "learning_rate": 2.2932013359842398e-06,
+      "loss": 0.9782,
+      "step": 5052
+    },
+    {
+      "epoch": 0.531426242654502,
+      "grad_norm": 3.1777989573404217,
+      "learning_rate": 2.292369389085996e-06,
+      "loss": 0.9521,
+      "step": 5053
+    },
+    {
+      "epoch": 0.5315314130963493,
+      "grad_norm": 2.635587308253445,
+      "learning_rate": 2.291537465340178e-06,
+      "loss": 0.9735,
+      "step": 5054
+    },
+    {
+      "epoch": 0.5316365835381965,
+      "grad_norm": 2.7400667145766846,
+      "learning_rate": 2.2907055648395517e-06,
+      "loss": 1.0065,
+      "step": 5055
+    },
+    {
+      "epoch": 0.5317417539800439,
+      "grad_norm": 2.1418010167978383,
+      "learning_rate": 2.2898736876768816e-06,
+      "loss": 1.0043,
+      "step": 5056
+    },
+    {
+      "epoch": 0.5318469244218912,
+      "grad_norm": 2.4204236378757296,
+      "learning_rate": 2.289041833944927e-06,
+      "loss": 0.9886,
+      "step": 5057
+    },
+    {
+      "epoch": 0.5319520948637385,
+      "grad_norm": 2.2258054408956847,
+      "learning_rate": 2.2882100037364472e-06,
+      "loss": 0.9754,
+      "step": 5058
+    },
+    {
+      "epoch": 0.5320572653055858,
+      "grad_norm": 1.698663415933468,
+      "learning_rate": 2.2873781971441963e-06,
+      "loss": 1.0052,
+      "step": 5059
+    },
+    {
+      "epoch": 0.5321624357474332,
+      "grad_norm": 2.39347765845993,
+      "learning_rate": 2.2865464142609286e-06,
+      "loss": 1.0169,
+      "step": 5060
+    },
+    {
+      "epoch": 0.5322676061892805,
+      "grad_norm": 2.1009995798487497,
+      "learning_rate": 2.2857146551793943e-06,
+      "loss": 1.0106,
+      "step": 5061
+    },
+    {
+      "epoch": 0.5323727766311278,
+      "grad_norm": 2.695049651098317,
+      "learning_rate": 2.2848829199923405e-06,
+      "loss": 0.9646,
+      "step": 5062
+    },
+    {
+      "epoch": 0.5324779470729751,
+      "grad_norm": 2.092568062548765,
+      "learning_rate": 2.2840512087925127e-06,
+      "loss": 0.988,
+      "step": 5063
+    },
+    {
+      "epoch": 0.5325831175148225,
+      "grad_norm": 2.183783767757615,
+      "learning_rate": 2.283219521672651e-06,
+      "loss": 0.9897,
+      "step": 5064
+    },
+    {
+      "epoch": 0.5326882879566698,
+      "grad_norm": 1.8926321090318932,
+      "learning_rate": 2.282387858725498e-06,
+      "loss": 0.999,
+      "step": 5065
+    },
+    {
+      "epoch": 0.5327934583985171,
+      "grad_norm": 2.3234643524419205,
+      "learning_rate": 2.281556220043789e-06,
+      "loss": 0.9972,
+      "step": 5066
+    },
+    {
+      "epoch": 0.5328986288403644,
+      "grad_norm": 2.311574355596325,
+      "learning_rate": 2.280724605720258e-06,
+      "loss": 1.0388,
+      "step": 5067
+    },
+    {
+      "epoch": 0.5330037992822118,
+      "grad_norm": 2.6794835776991985,
+      "learning_rate": 2.2798930158476375e-06,
+      "loss": 1.0157,
+      "step": 5068
+    },
+    {
+      "epoch": 0.5331089697240591,
+      "grad_norm": 2.314690603027226,
+      "learning_rate": 2.279061450518655e-06,
+      "loss": 1.0114,
+      "step": 5069
+    },
+    {
+      "epoch": 0.5332141401659064,
+      "grad_norm": 2.199288704139533,
+      "learning_rate": 2.278229909826037e-06,
+      "loss": 0.9784,
+      "step": 5070
+    },
+    {
+      "epoch": 0.5333193106077537,
+      "grad_norm": 1.817939098985619,
+      "learning_rate": 2.2773983938625074e-06,
+      "loss": 0.9549,
+      "step": 5071
+    },
+    {
+      "epoch": 0.533424481049601,
+      "grad_norm": 2.0687055899360303,
+      "learning_rate": 2.276566902720786e-06,
+      "loss": 0.9939,
+      "step": 5072
+    },
+    {
+      "epoch": 0.5335296514914484,
+      "grad_norm": 2.9464457058239257,
+      "learning_rate": 2.2757354364935893e-06,
+      "loss": 0.9716,
+      "step": 5073
+    },
+    {
+      "epoch": 0.5336348219332957,
+      "grad_norm": 1.5594113007408739,
+      "learning_rate": 2.274903995273635e-06,
+      "loss": 0.9705,
+      "step": 5074
+    },
+    {
+      "epoch": 0.5337399923751429,
+      "grad_norm": 2.6090779806826836,
+      "learning_rate": 2.2740725791536337e-06,
+      "loss": 0.9881,
+      "step": 5075
+    },
+    {
+      "epoch": 0.5338451628169902,
+      "grad_norm": 1.8377197278317314,
+      "learning_rate": 2.2732411882262946e-06,
+      "loss": 0.9562,
+      "step": 5076
+    },
+    {
+      "epoch": 0.5339503332588376,
+      "grad_norm": 2.7435287336866505,
+      "learning_rate": 2.272409822584325e-06,
+      "loss": 1.0032,
+      "step": 5077
+    },
+    {
+      "epoch": 0.5340555037006849,
+      "grad_norm": 2.212795732909546,
+      "learning_rate": 2.2715784823204275e-06,
+      "loss": 0.9326,
+      "step": 5078
+    },
+    {
+      "epoch": 0.5341606741425322,
+      "grad_norm": 2.3267493730330373,
+      "learning_rate": 2.270747167527304e-06,
+      "loss": 1.0236,
+      "step": 5079
+    },
+    {
+      "epoch": 0.5342658445843795,
+      "grad_norm": 2.622033615860494,
+      "learning_rate": 2.2699158782976527e-06,
+      "loss": 0.9834,
+      "step": 5080
+    },
+    {
+      "epoch": 0.5343710150262269,
+      "grad_norm": 2.210083057529942,
+      "learning_rate": 2.269084614724168e-06,
+      "loss": 1.0015,
+      "step": 5081
+    },
+    {
+      "epoch": 0.5344761854680742,
+      "grad_norm": 2.8063796227677877,
+      "learning_rate": 2.268253376899542e-06,
+      "loss": 0.9861,
+      "step": 5082
+    },
+    {
+      "epoch": 0.5345813559099215,
+      "grad_norm": 2.54873266219812,
+      "learning_rate": 2.267422164916465e-06,
+      "loss": 0.9637,
+      "step": 5083
+    },
+    {
+      "epoch": 0.5346865263517688,
+      "grad_norm": 2.9683253595659123,
+      "learning_rate": 2.2665909788676236e-06,
+      "loss": 1.0272,
+      "step": 5084
+    },
+    {
+      "epoch": 0.5347916967936162,
+      "grad_norm": 2.795406678292719,
+      "learning_rate": 2.2657598188457015e-06,
+      "loss": 0.9601,
+      "step": 5085
+    },
+    {
+      "epoch": 0.5348968672354635,
+      "grad_norm": 2.423988503624797,
+      "learning_rate": 2.264928684943379e-06,
+      "loss": 1.016,
+      "step": 5086
+    },
+    {
+      "epoch": 0.5350020376773108,
+      "grad_norm": 3.0389490068520244,
+      "learning_rate": 2.264097577253333e-06,
+      "loss": 0.98,
+      "step": 5087
+    },
+    {
+      "epoch": 0.5351072081191581,
+      "grad_norm": 2.591628315768945,
+      "learning_rate": 2.263266495868241e-06,
+      "loss": 0.9451,
+      "step": 5088
+    },
+    {
+      "epoch": 0.5352123785610055,
+      "grad_norm": 2.85809426107325,
+      "learning_rate": 2.262435440880774e-06,
+      "loss": 1.0456,
+      "step": 5089
+    },
+    {
+      "epoch": 0.5353175490028528,
+      "grad_norm": 2.4229876711875367,
+      "learning_rate": 2.2616044123836005e-06,
+      "loss": 0.962,
+      "step": 5090
+    },
+    {
+      "epoch": 0.5354227194447001,
+      "grad_norm": 2.272935313170715,
+      "learning_rate": 2.2607734104693866e-06,
+      "loss": 1.0093,
+      "step": 5091
+    },
+    {
+      "epoch": 0.5355278898865474,
+      "grad_norm": 2.9194015986993644,
+      "learning_rate": 2.2599424352307958e-06,
+      "loss": 0.9677,
+      "step": 5092
+    },
+    {
+      "epoch": 0.5356330603283948,
+      "grad_norm": 2.5574503210340755,
+      "learning_rate": 2.2591114867604887e-06,
+      "loss": 0.9673,
+      "step": 5093
+    },
+    {
+      "epoch": 0.5357382307702421,
+      "grad_norm": 1.81531777330057,
+      "learning_rate": 2.258280565151122e-06,
+      "loss": 1.005,
+      "step": 5094
+    },
+    {
+      "epoch": 0.5358434012120893,
+      "grad_norm": 2.8458384104973278,
+      "learning_rate": 2.25744967049535e-06,
+      "loss": 0.9489,
+      "step": 5095
+    },
+    {
+      "epoch": 0.5359485716539366,
+      "grad_norm": 2.1465556016022735,
+      "learning_rate": 2.2566188028858228e-06,
+      "loss": 0.995,
+      "step": 5096
+    },
+    {
+      "epoch": 0.5360537420957839,
+      "grad_norm": 2.3813284857921637,
+      "learning_rate": 2.2557879624151912e-06,
+      "loss": 1.0476,
+      "step": 5097
+    },
+    {
+      "epoch": 0.5361589125376313,
+      "grad_norm": 2.2720674366303735,
+      "learning_rate": 2.2549571491760985e-06,
+      "loss": 0.9967,
+      "step": 5098
+    },
+    {
+      "epoch": 0.5362640829794786,
+      "grad_norm": 2.1256470660057305,
+      "learning_rate": 2.254126363261188e-06,
+      "loss": 1.0248,
+      "step": 5099
+    },
+    {
+      "epoch": 0.5363692534213259,
+      "grad_norm": 1.430140563761567,
+      "learning_rate": 2.2532956047630973e-06,
+      "loss": 0.9834,
+      "step": 5100
+    },
+    {
+      "epoch": 0.5364744238631732,
+      "grad_norm": 2.2091020729494324,
+      "learning_rate": 2.252464873774462e-06,
+      "loss": 0.9964,
+      "step": 5101
+    },
+    {
+      "epoch": 0.5365795943050206,
+      "grad_norm": 3.05315371393753,
+      "learning_rate": 2.2516341703879176e-06,
+      "loss": 1.0133,
+      "step": 5102
+    },
+    {
+      "epoch": 0.5366847647468679,
+      "grad_norm": 2.67327232115545,
+      "learning_rate": 2.2508034946960924e-06,
+      "loss": 1.0006,
+      "step": 5103
+    },
+    {
+      "epoch": 0.5367899351887152,
+      "grad_norm": 2.990185755256598,
+      "learning_rate": 2.2499728467916133e-06,
+      "loss": 0.9785,
+      "step": 5104
+    },
+    {
+      "epoch": 0.5368951056305625,
+      "grad_norm": 2.391793174882368,
+      "learning_rate": 2.249142226767104e-06,
+      "loss": 0.9347,
+      "step": 5105
+    },
+    {
+      "epoch": 0.5370002760724099,
+      "grad_norm": 2.5608418817555014,
+      "learning_rate": 2.248311634715185e-06,
+      "loss": 1.0156,
+      "step": 5106
+    },
+    {
+      "epoch": 0.5371054465142572,
+      "grad_norm": 2.8618225744834067,
+      "learning_rate": 2.247481070728474e-06,
+      "loss": 1.029,
+      "step": 5107
+    },
+    {
+      "epoch": 0.5372106169561045,
+      "grad_norm": 1.5332374205582922,
+      "learning_rate": 2.2466505348995854e-06,
+      "loss": 0.9641,
+      "step": 5108
+    },
+    {
+      "epoch": 0.5373157873979518,
+      "grad_norm": 2.445318078146189,
+      "learning_rate": 2.24582002732113e-06,
+      "loss": 1.0037,
+      "step": 5109
+    },
+    {
+      "epoch": 0.5374209578397992,
+      "grad_norm": 1.948245870338096,
+      "learning_rate": 2.244989548085716e-06,
+      "loss": 0.9739,
+      "step": 5110
+    },
+    {
+      "epoch": 0.5375261282816465,
+      "grad_norm": 1.9108232074625904,
+      "learning_rate": 2.2441590972859484e-06,
+      "loss": 1.0196,
+      "step": 5111
+    },
+    {
+      "epoch": 0.5376312987234938,
+      "grad_norm": 2.239416100740601,
+      "learning_rate": 2.2433286750144293e-06,
+      "loss": 0.9678,
+      "step": 5112
+    },
+    {
+      "epoch": 0.5377364691653411,
+      "grad_norm": 2.205078947975664,
+      "learning_rate": 2.2424982813637567e-06,
+      "loss": 0.9857,
+      "step": 5113
+    },
+    {
+      "epoch": 0.5378416396071884,
+      "grad_norm": 2.253513657680878,
+      "learning_rate": 2.241667916426526e-06,
+      "loss": 0.9517,
+      "step": 5114
+    },
+    {
+      "epoch": 0.5379468100490358,
+      "grad_norm": 3.06969161503822,
+      "learning_rate": 2.240837580295329e-06,
+      "loss": 1.0048,
+      "step": 5115
+    },
+    {
+      "epoch": 0.538051980490883,
+      "grad_norm": 2.435849474370094,
+      "learning_rate": 2.2400072730627556e-06,
+      "loss": 0.9674,
+      "step": 5116
+    },
+    {
+      "epoch": 0.5381571509327303,
+      "grad_norm": 2.5677719895294433,
+      "learning_rate": 2.239176994821391e-06,
+      "loss": 0.9729,
+      "step": 5117
+    },
+    {
+      "epoch": 0.5382623213745776,
+      "grad_norm": 2.258970213664192,
+      "learning_rate": 2.2383467456638175e-06,
+      "loss": 0.9593,
+      "step": 5118
+    },
+    {
+      "epoch": 0.538367491816425,
+      "grad_norm": 2.4559982555422057,
+      "learning_rate": 2.237516525682614e-06,
+      "loss": 1.0028,
+      "step": 5119
+    },
+    {
+      "epoch": 0.5384726622582723,
+      "grad_norm": 3.0661627782794625,
+      "learning_rate": 2.2366863349703574e-06,
+      "loss": 0.9849,
+      "step": 5120
+    },
+    {
+      "epoch": 0.5385778327001196,
+      "grad_norm": 2.208903743452161,
+      "learning_rate": 2.23585617361962e-06,
+      "loss": 0.9887,
+      "step": 5121
+    },
+    {
+      "epoch": 0.5386830031419669,
+      "grad_norm": 2.5131994229325,
+      "learning_rate": 2.2350260417229715e-06,
+      "loss": 1.0212,
+      "step": 5122
+    },
+    {
+      "epoch": 0.5387881735838143,
+      "grad_norm": 2.513688036331137,
+      "learning_rate": 2.234195939372977e-06,
+      "loss": 1.011,
+      "step": 5123
+    },
+    {
+      "epoch": 0.5388933440256616,
+      "grad_norm": 1.9065741666317384,
+      "learning_rate": 2.2333658666621995e-06,
+      "loss": 0.9949,
+      "step": 5124
+    },
+    {
+      "epoch": 0.5389985144675089,
+      "grad_norm": 2.2999327606370494,
+      "learning_rate": 2.2325358236832e-06,
+      "loss": 0.9497,
+      "step": 5125
+    },
+    {
+      "epoch": 0.5391036849093562,
+      "grad_norm": 3.0371065795572543,
+      "learning_rate": 2.231705810528534e-06,
+      "loss": 1.0135,
+      "step": 5126
+    },
+    {
+      "epoch": 0.5392088553512036,
+      "grad_norm": 3.086589128157086,
+      "learning_rate": 2.230875827290755e-06,
+      "loss": 1.0293,
+      "step": 5127
+    },
+    {
+      "epoch": 0.5393140257930509,
+      "grad_norm": 2.6399411648602977,
+      "learning_rate": 2.2300458740624102e-06,
+      "loss": 0.9888,
+      "step": 5128
+    },
+    {
+      "epoch": 0.5394191962348982,
+      "grad_norm": 1.58725667452964,
+      "learning_rate": 2.2292159509360487e-06,
+      "loss": 0.9702,
+      "step": 5129
+    },
+    {
+      "epoch": 0.5395243666767455,
+      "grad_norm": 2.7241128959636027,
+      "learning_rate": 2.228386058004212e-06,
+      "loss": 1.0053,
+      "step": 5130
+    },
+    {
+      "epoch": 0.5396295371185929,
+      "grad_norm": 1.7958176439133673,
+      "learning_rate": 2.22755619535944e-06,
+      "loss": 0.9931,
+      "step": 5131
+    },
+    {
+      "epoch": 0.5397347075604402,
+      "grad_norm": 2.215221909482859,
+      "learning_rate": 2.2267263630942682e-06,
+      "loss": 1.0076,
+      "step": 5132
+    },
+    {
+      "epoch": 0.5398398780022875,
+      "grad_norm": 2.3342570809884267,
+      "learning_rate": 2.2258965613012293e-06,
+      "loss": 1.0169,
+      "step": 5133
+    },
+    {
+      "epoch": 0.5399450484441348,
+      "grad_norm": 2.2114565815704097,
+      "learning_rate": 2.2250667900728543e-06,
+      "loss": 1.0046,
+      "step": 5134
+    },
+    {
+      "epoch": 0.5400502188859821,
+      "grad_norm": 2.32994956342613,
+      "learning_rate": 2.224237049501668e-06,
+      "loss": 1.0162,
+      "step": 5135
+    },
+    {
+      "epoch": 0.5401553893278294,
+      "grad_norm": 2.2005296173151296,
+      "learning_rate": 2.223407339680192e-06,
+      "loss": 1.0203,
+      "step": 5136
+    },
+    {
+      "epoch": 0.5402605597696767,
+      "grad_norm": 2.71416231138807,
+      "learning_rate": 2.222577660700947e-06,
+      "loss": 0.9696,
+      "step": 5137
+    },
+    {
+      "epoch": 0.540365730211524,
+      "grad_norm": 2.3021954271008283,
+      "learning_rate": 2.2217480126564462e-06,
+      "loss": 1.0134,
+      "step": 5138
+    },
+    {
+      "epoch": 0.5404709006533713,
+      "grad_norm": 2.069021577288529,
+      "learning_rate": 2.220918395639205e-06,
+      "loss": 0.9861,
+      "step": 5139
+    },
+    {
+      "epoch": 0.5405760710952187,
+      "grad_norm": 2.113347825343487,
+      "learning_rate": 2.2200888097417308e-06,
+      "loss": 0.9842,
+      "step": 5140
+    },
+    {
+      "epoch": 0.540681241537066,
+      "grad_norm": 2.862415076458137,
+      "learning_rate": 2.219259255056528e-06,
+      "loss": 0.9512,
+      "step": 5141
+    },
+    {
+      "epoch": 0.5407864119789133,
+      "grad_norm": 2.494202998333855,
+      "learning_rate": 2.2184297316760998e-06,
+      "loss": 0.9957,
+      "step": 5142
+    },
+    {
+      "epoch": 0.5408915824207606,
+      "grad_norm": 2.5037856881206646,
+      "learning_rate": 2.2176002396929435e-06,
+      "loss": 1.0187,
+      "step": 5143
+    },
+    {
+      "epoch": 0.540996752862608,
+      "grad_norm": 2.300224957499141,
+      "learning_rate": 2.2167707791995547e-06,
+      "loss": 0.9993,
+      "step": 5144
+    },
+    {
+      "epoch": 0.5411019233044553,
+      "grad_norm": 1.7557662844334694,
+      "learning_rate": 2.2159413502884237e-06,
+      "loss": 0.9592,
+      "step": 5145
+    },
+    {
+      "epoch": 0.5412070937463026,
+      "grad_norm": 2.5946831278066744,
+      "learning_rate": 2.2151119530520394e-06,
+      "loss": 1.0363,
+      "step": 5146
+    },
+    {
+      "epoch": 0.5413122641881499,
+      "grad_norm": 2.6143394641387623,
+      "learning_rate": 2.2142825875828838e-06,
+      "loss": 0.9531,
+      "step": 5147
+    },
+    {
+      "epoch": 0.5414174346299973,
+      "grad_norm": 2.381679279813613,
+      "learning_rate": 2.2134532539734406e-06,
+      "loss": 1.0057,
+      "step": 5148
+    },
+    {
+      "epoch": 0.5415226050718446,
+      "grad_norm": 1.8202176335628701,
+      "learning_rate": 2.2126239523161854e-06,
+      "loss": 1.0132,
+      "step": 5149
+    },
+    {
+      "epoch": 0.5416277755136919,
+      "grad_norm": 2.30712857841413,
+      "learning_rate": 2.2117946827035927e-06,
+      "loss": 0.9888,
+      "step": 5150
+    },
+    {
+      "epoch": 0.5417329459555392,
+      "grad_norm": 2.5917572577099084,
+      "learning_rate": 2.2109654452281297e-06,
+      "loss": 0.9757,
+      "step": 5151
+    },
+    {
+      "epoch": 0.5418381163973865,
+      "grad_norm": 1.9879341436149915,
+      "learning_rate": 2.210136239982266e-06,
+      "loss": 1.0149,
+      "step": 5152
+    },
+    {
+      "epoch": 0.5419432868392339,
+      "grad_norm": 2.784187511162275,
+      "learning_rate": 2.2093070670584636e-06,
+      "loss": 0.9713,
+      "step": 5153
+    },
+    {
+      "epoch": 0.5420484572810812,
+      "grad_norm": 2.8215281669994905,
+      "learning_rate": 2.2084779265491817e-06,
+      "loss": 0.996,
+      "step": 5154
+    },
+    {
+      "epoch": 0.5421536277229285,
+      "grad_norm": 2.659872718452111,
+      "learning_rate": 2.207648818546875e-06,
+      "loss": 0.9979,
+      "step": 5155
+    },
+    {
+      "epoch": 0.5422587981647757,
+      "grad_norm": 2.4973238896568617,
+      "learning_rate": 2.206819743143996e-06,
+      "loss": 1.0018,
+      "step": 5156
+    },
+    {
+      "epoch": 0.5423639686066231,
+      "grad_norm": 3.3899471789949573,
+      "learning_rate": 2.2059907004329934e-06,
+      "loss": 1.0076,
+      "step": 5157
+    },
+    {
+      "epoch": 0.5424691390484704,
+      "grad_norm": 3.008953501860152,
+      "learning_rate": 2.2051616905063112e-06,
+      "loss": 0.9814,
+      "step": 5158
+    },
+    {
+      "epoch": 0.5425743094903177,
+      "grad_norm": 2.4382683163743804,
+      "learning_rate": 2.2043327134563917e-06,
+      "loss": 0.9694,
+      "step": 5159
+    },
+    {
+      "epoch": 0.542679479932165,
+      "grad_norm": 2.4112838713428606,
+      "learning_rate": 2.203503769375671e-06,
+      "loss": 0.9944,
+      "step": 5160
+    },
+    {
+      "epoch": 0.5427846503740124,
+      "grad_norm": 2.3975472141330703,
+      "learning_rate": 2.2026748583565824e-06,
+      "loss": 0.9987,
+      "step": 5161
+    },
+    {
+      "epoch": 0.5428898208158597,
+      "grad_norm": 2.4732650547216,
+      "learning_rate": 2.201845980491558e-06,
+      "loss": 1.0106,
+      "step": 5162
+    },
+    {
+      "epoch": 0.542994991257707,
+      "grad_norm": 1.9440375264246743,
+      "learning_rate": 2.2010171358730227e-06,
+      "loss": 0.9711,
+      "step": 5163
+    },
+    {
+      "epoch": 0.5431001616995543,
+      "grad_norm": 2.950910241043643,
+      "learning_rate": 2.2001883245933992e-06,
+      "loss": 1.0359,
+      "step": 5164
+    },
+    {
+      "epoch": 0.5432053321414017,
+      "grad_norm": 1.7374181043854764,
+      "learning_rate": 2.199359546745106e-06,
+      "loss": 0.9894,
+      "step": 5165
+    },
+    {
+      "epoch": 0.543310502583249,
+      "grad_norm": 1.919204253825503,
+      "learning_rate": 2.19853080242056e-06,
+      "loss": 0.9888,
+      "step": 5166
+    },
+    {
+      "epoch": 0.5434156730250963,
+      "grad_norm": 2.194559285983916,
+      "learning_rate": 2.1977020917121707e-06,
+      "loss": 0.9831,
+      "step": 5167
+    },
+    {
+      "epoch": 0.5435208434669436,
+      "grad_norm": 2.616426607003015,
+      "learning_rate": 2.1968734147123467e-06,
+      "loss": 1.0058,
+      "step": 5168
+    },
+    {
+      "epoch": 0.543626013908791,
+      "grad_norm": 2.9485745152973513,
+      "learning_rate": 2.196044771513492e-06,
+      "loss": 1.0286,
+      "step": 5169
+    },
+    {
+      "epoch": 0.5437311843506383,
+      "grad_norm": 2.3426077436696096,
+      "learning_rate": 2.195216162208005e-06,
+      "loss": 0.9494,
+      "step": 5170
+    },
+    {
+      "epoch": 0.5438363547924856,
+      "grad_norm": 2.350276659584285,
+      "learning_rate": 2.1943875868882853e-06,
+      "loss": 1.0144,
+      "step": 5171
+    },
+    {
+      "epoch": 0.5439415252343329,
+      "grad_norm": 1.9725474379833872,
+      "learning_rate": 2.1935590456467232e-06,
+      "loss": 0.9755,
+      "step": 5172
+    },
+    {
+      "epoch": 0.5440466956761802,
+      "grad_norm": 2.4143702510240055,
+      "learning_rate": 2.192730538575708e-06,
+      "loss": 0.9701,
+      "step": 5173
+    },
+    {
+      "epoch": 0.5441518661180276,
+      "grad_norm": 2.1110930182865064,
+      "learning_rate": 2.191902065767624e-06,
+      "loss": 1.0355,
+      "step": 5174
+    },
+    {
+      "epoch": 0.5442570365598749,
+      "grad_norm": 4.361748721313425,
+      "learning_rate": 2.191073627314854e-06,
+      "loss": 0.9864,
+      "step": 5175
+    },
+    {
+      "epoch": 0.5443622070017222,
+      "grad_norm": 2.501366700470722,
+      "learning_rate": 2.1902452233097736e-06,
+      "loss": 0.9758,
+      "step": 5176
+    },
+    {
+      "epoch": 0.5444673774435694,
+      "grad_norm": 2.235469201389308,
+      "learning_rate": 2.1894168538447576e-06,
+      "loss": 1.0104,
+      "step": 5177
+    },
+    {
+      "epoch": 0.5445725478854168,
+      "grad_norm": 2.6134379475508367,
+      "learning_rate": 2.1885885190121753e-06,
+      "loss": 0.9644,
+      "step": 5178
+    },
+    {
+      "epoch": 0.5446777183272641,
+      "grad_norm": 2.6083073446693876,
+      "learning_rate": 2.187760218904392e-06,
+      "loss": 0.9883,
+      "step": 5179
+    },
+    {
+      "epoch": 0.5447828887691114,
+      "grad_norm": 2.888337664510672,
+      "learning_rate": 2.1869319536137693e-06,
+      "loss": 0.9995,
+      "step": 5180
+    },
+    {
+      "epoch": 0.5448880592109587,
+      "grad_norm": 2.1326399057874914,
+      "learning_rate": 2.1861037232326666e-06,
+      "loss": 0.9857,
+      "step": 5181
+    },
+    {
+      "epoch": 0.544993229652806,
+      "grad_norm": 2.545298533806266,
+      "learning_rate": 2.1852755278534373e-06,
+      "loss": 0.9809,
+      "step": 5182
+    },
+    {
+      "epoch": 0.5450984000946534,
+      "grad_norm": 2.0108239660027656,
+      "learning_rate": 2.18444736756843e-06,
+      "loss": 1.0229,
+      "step": 5183
+    },
+    {
+      "epoch": 0.5452035705365007,
+      "grad_norm": 2.6337533910173057,
+      "learning_rate": 2.183619242469994e-06,
+      "loss": 0.9919,
+      "step": 5184
+    },
+    {
+      "epoch": 0.545308740978348,
+      "grad_norm": 2.5146353222564235,
+      "learning_rate": 2.1827911526504702e-06,
+      "loss": 0.9946,
+      "step": 5185
+    },
+    {
+      "epoch": 0.5454139114201954,
+      "grad_norm": 2.613024635280223,
+      "learning_rate": 2.1819630982021967e-06,
+      "loss": 0.9718,
+      "step": 5186
+    },
+    {
+      "epoch": 0.5455190818620427,
+      "grad_norm": 2.26349598516355,
+      "learning_rate": 2.1811350792175084e-06,
+      "loss": 1.0053,
+      "step": 5187
+    },
+    {
+      "epoch": 0.54562425230389,
+      "grad_norm": 2.3470184717219422,
+      "learning_rate": 2.1803070957887348e-06,
+      "loss": 0.9996,
+      "step": 5188
+    },
+    {
+      "epoch": 0.5457294227457373,
+      "grad_norm": 2.476693756773956,
+      "learning_rate": 2.1794791480082046e-06,
+      "loss": 1.0368,
+      "step": 5189
+    },
+    {
+      "epoch": 0.5458345931875846,
+      "grad_norm": 2.836739750714608,
+      "learning_rate": 2.1786512359682394e-06,
+      "loss": 0.9856,
+      "step": 5190
+    },
+    {
+      "epoch": 0.545939763629432,
+      "grad_norm": 2.120438515765004,
+      "learning_rate": 2.1778233597611576e-06,
+      "loss": 0.9753,
+      "step": 5191
+    },
+    {
+      "epoch": 0.5460449340712793,
+      "grad_norm": 2.085336690843402,
+      "learning_rate": 2.1769955194792737e-06,
+      "loss": 1.0106,
+      "step": 5192
+    },
+    {
+      "epoch": 0.5461501045131266,
+      "grad_norm": 2.388620073312146,
+      "learning_rate": 2.176167715214898e-06,
+      "loss": 0.9767,
+      "step": 5193
+    },
+    {
+      "epoch": 0.546255274954974,
+      "grad_norm": 2.71199718772841,
+      "learning_rate": 2.1753399470603387e-06,
+      "loss": 1.0116,
+      "step": 5194
+    },
+    {
+      "epoch": 0.5463604453968213,
+      "grad_norm": 2.6031851920934397,
+      "learning_rate": 2.174512215107897e-06,
+      "loss": 0.997,
+      "step": 5195
+    },
+    {
+      "epoch": 0.5464656158386686,
+      "grad_norm": 2.5270157415798815,
+      "learning_rate": 2.173684519449872e-06,
+      "loss": 1.0268,
+      "step": 5196
+    },
+    {
+      "epoch": 0.5465707862805158,
+      "grad_norm": 3.0251465048345607,
+      "learning_rate": 2.1728568601785564e-06,
+      "loss": 1.0076,
+      "step": 5197
+    },
+    {
+      "epoch": 0.5466759567223631,
+      "grad_norm": 3.213521672648237,
+      "learning_rate": 2.172029237386244e-06,
+      "loss": 1.0297,
+      "step": 5198
+    },
+    {
+      "epoch": 0.5467811271642105,
+      "grad_norm": 2.4599343033374774,
+      "learning_rate": 2.1712016511652187e-06,
+      "loss": 1.0106,
+      "step": 5199
+    },
+    {
+      "epoch": 0.5468862976060578,
+      "grad_norm": 2.744190371006768,
+      "learning_rate": 2.170374101607764e-06,
+      "loss": 1.0298,
+      "step": 5200
+    },
+    {
+      "epoch": 0.5469914680479051,
+      "grad_norm": 2.49369766763488,
+      "learning_rate": 2.169546588806158e-06,
+      "loss": 1.0502,
+      "step": 5201
+    },
+    {
+      "epoch": 0.5470966384897524,
+      "grad_norm": 2.381587412730638,
+      "learning_rate": 2.168719112852673e-06,
+      "loss": 0.9862,
+      "step": 5202
+    },
+    {
+      "epoch": 0.5472018089315998,
+      "grad_norm": 3.9803983911874736,
+      "learning_rate": 2.167891673839581e-06,
+      "loss": 1.0081,
+      "step": 5203
+    },
+    {
+      "epoch": 0.5473069793734471,
+      "grad_norm": 2.8513338869013287,
+      "learning_rate": 2.1670642718591477e-06,
+      "loss": 0.9999,
+      "step": 5204
+    },
+    {
+      "epoch": 0.5474121498152944,
+      "grad_norm": 2.087249392253148,
+      "learning_rate": 2.166236907003634e-06,
+      "loss": 0.9908,
+      "step": 5205
+    },
+    {
+      "epoch": 0.5475173202571417,
+      "grad_norm": 2.1537694107930956,
+      "learning_rate": 2.1654095793652975e-06,
+      "loss": 0.9787,
+      "step": 5206
+    },
+    {
+      "epoch": 0.547622490698989,
+      "grad_norm": 2.398002826051045,
+      "learning_rate": 2.1645822890363928e-06,
+      "loss": 1.0128,
+      "step": 5207
+    },
+    {
+      "epoch": 0.5477276611408364,
+      "grad_norm": 1.8014888313043216,
+      "learning_rate": 2.1637550361091685e-06,
+      "loss": 0.9712,
+      "step": 5208
+    },
+    {
+      "epoch": 0.5478328315826837,
+      "grad_norm": 2.6606854949977783,
+      "learning_rate": 2.16292782067587e-06,
+      "loss": 1.0116,
+      "step": 5209
+    },
+    {
+      "epoch": 0.547938002024531,
+      "grad_norm": 2.688100293619295,
+      "learning_rate": 2.162100642828737e-06,
+      "loss": 1.0044,
+      "step": 5210
+    },
+    {
+      "epoch": 0.5480431724663783,
+      "grad_norm": 2.4430646701220544,
+      "learning_rate": 2.161273502660007e-06,
+      "loss": 1.0112,
+      "step": 5211
+    },
+    {
+      "epoch": 0.5481483429082257,
+      "grad_norm": 2.0241794197127927,
+      "learning_rate": 2.1604464002619135e-06,
+      "loss": 1.0267,
+      "step": 5212
+    },
+    {
+      "epoch": 0.548253513350073,
+      "grad_norm": 3.178008506690804,
+      "learning_rate": 2.1596193357266844e-06,
+      "loss": 0.9995,
+      "step": 5213
+    },
+    {
+      "epoch": 0.5483586837919203,
+      "grad_norm": 2.161378205518831,
+      "learning_rate": 2.1587923091465434e-06,
+      "loss": 0.976,
+      "step": 5214
+    },
+    {
+      "epoch": 0.5484638542337676,
+      "grad_norm": 2.2458856326219885,
+      "learning_rate": 2.1579653206137104e-06,
+      "loss": 1.0095,
+      "step": 5215
+    },
+    {
+      "epoch": 0.548569024675615,
+      "grad_norm": 2.276838712981099,
+      "learning_rate": 2.1571383702204006e-06,
+      "loss": 0.9981,
+      "step": 5216
+    },
+    {
+      "epoch": 0.5486741951174622,
+      "grad_norm": 2.5745480726564693,
+      "learning_rate": 2.1563114580588267e-06,
+      "loss": 1.0043,
+      "step": 5217
+    },
+    {
+      "epoch": 0.5487793655593095,
+      "grad_norm": 1.8966939253835509,
+      "learning_rate": 2.1554845842211954e-06,
+      "loss": 1.0158,
+      "step": 5218
+    },
+    {
+      "epoch": 0.5488845360011568,
+      "grad_norm": 2.0953028195766423,
+      "learning_rate": 2.1546577487997087e-06,
+      "loss": 0.9859,
+      "step": 5219
+    },
+    {
+      "epoch": 0.5489897064430042,
+      "grad_norm": 2.110354152880763,
+      "learning_rate": 2.1538309518865646e-06,
+      "loss": 0.9917,
+      "step": 5220
+    },
+    {
+      "epoch": 0.5490948768848515,
+      "grad_norm": 2.3921433816101945,
+      "learning_rate": 2.1530041935739604e-06,
+      "loss": 0.9732,
+      "step": 5221
+    },
+    {
+      "epoch": 0.5492000473266988,
+      "grad_norm": 1.5432785190992437,
+      "learning_rate": 2.1521774739540833e-06,
+      "loss": 0.9604,
+      "step": 5222
+    },
+    {
+      "epoch": 0.5493052177685461,
+      "grad_norm": 2.6627689190811914,
+      "learning_rate": 2.1513507931191203e-06,
+      "loss": 1.0003,
+      "step": 5223
+    },
+    {
+      "epoch": 0.5494103882103935,
+      "grad_norm": 2.4129431035360427,
+      "learning_rate": 2.1505241511612522e-06,
+      "loss": 0.9286,
+      "step": 5224
+    },
+    {
+      "epoch": 0.5495155586522408,
+      "grad_norm": 2.9528788317691554,
+      "learning_rate": 2.149697548172655e-06,
+      "loss": 0.9514,
+      "step": 5225
+    },
+    {
+      "epoch": 0.5496207290940881,
+      "grad_norm": 2.2730624529958146,
+      "learning_rate": 2.1488709842455033e-06,
+      "loss": 0.9983,
+      "step": 5226
+    },
+    {
+      "epoch": 0.5497258995359354,
+      "grad_norm": 3.166867500015208,
+      "learning_rate": 2.1480444594719647e-06,
+      "loss": 1.0307,
+      "step": 5227
+    },
+    {
+      "epoch": 0.5498310699777827,
+      "grad_norm": 2.4113920354615916,
+      "learning_rate": 2.1472179739442027e-06,
+      "loss": 0.9667,
+      "step": 5228
+    },
+    {
+      "epoch": 0.5499362404196301,
+      "grad_norm": 2.518675037782677,
+      "learning_rate": 2.1463915277543766e-06,
+      "loss": 1.0238,
+      "step": 5229
+    },
+    {
+      "epoch": 0.5500414108614774,
+      "grad_norm": 2.6250890442322454,
+      "learning_rate": 2.1455651209946428e-06,
+      "loss": 0.9742,
+      "step": 5230
+    },
+    {
+      "epoch": 0.5501465813033247,
+      "grad_norm": 2.02618334080473,
+      "learning_rate": 2.144738753757151e-06,
+      "loss": 0.9678,
+      "step": 5231
+    },
+    {
+      "epoch": 0.550251751745172,
+      "grad_norm": 2.667267963473209,
+      "learning_rate": 2.1439124261340484e-06,
+      "loss": 0.9861,
+      "step": 5232
+    },
+    {
+      "epoch": 0.5503569221870194,
+      "grad_norm": 2.1122311481039597,
+      "learning_rate": 2.1430861382174763e-06,
+      "loss": 0.9488,
+      "step": 5233
+    },
+    {
+      "epoch": 0.5504620926288667,
+      "grad_norm": 1.7454212223945111,
+      "learning_rate": 2.1422598900995715e-06,
+      "loss": 0.9995,
+      "step": 5234
+    },
+    {
+      "epoch": 0.550567263070714,
+      "grad_norm": 2.5491840982703975,
+      "learning_rate": 2.1414336818724685e-06,
+      "loss": 0.9631,
+      "step": 5235
+    },
+    {
+      "epoch": 0.5506724335125613,
+      "grad_norm": 2.0258659355875044,
+      "learning_rate": 2.140607513628296e-06,
+      "loss": 0.9681,
+      "step": 5236
+    },
+    {
+      "epoch": 0.5507776039544087,
+      "grad_norm": 3.187827818577956,
+      "learning_rate": 2.1397813854591778e-06,
+      "loss": 0.962,
+      "step": 5237
+    },
+    {
+      "epoch": 0.5508827743962559,
+      "grad_norm": 2.1053841231191734,
+      "learning_rate": 2.138955297457233e-06,
+      "loss": 0.9881,
+      "step": 5238
+    },
+    {
+      "epoch": 0.5509879448381032,
+      "grad_norm": 2.0350361564304102,
+      "learning_rate": 2.138129249714576e-06,
+      "loss": 0.978,
+      "step": 5239
+    },
+    {
+      "epoch": 0.5510931152799505,
+      "grad_norm": 2.624945121824399,
+      "learning_rate": 2.13730324232332e-06,
+      "loss": 0.9699,
+      "step": 5240
+    },
+    {
+      "epoch": 0.5511982857217979,
+      "grad_norm": 3.242345363363925,
+      "learning_rate": 2.13647727537557e-06,
+      "loss": 0.9856,
+      "step": 5241
+    },
+    {
+      "epoch": 0.5513034561636452,
+      "grad_norm": 2.8837592921549278,
+      "learning_rate": 2.135651348963428e-06,
+      "loss": 1.0201,
+      "step": 5242
+    },
+    {
+      "epoch": 0.5514086266054925,
+      "grad_norm": 2.3013474493107795,
+      "learning_rate": 2.1348254631789895e-06,
+      "loss": 0.9774,
+      "step": 5243
+    },
+    {
+      "epoch": 0.5515137970473398,
+      "grad_norm": 2.113009140432173,
+      "learning_rate": 2.13399961811435e-06,
+      "loss": 0.976,
+      "step": 5244
+    },
+    {
+      "epoch": 0.5516189674891872,
+      "grad_norm": 2.6604980643288236,
+      "learning_rate": 2.133173813861596e-06,
+      "loss": 1.0091,
+      "step": 5245
+    },
+    {
+      "epoch": 0.5517241379310345,
+      "grad_norm": 3.1134737984113303,
+      "learning_rate": 2.1323480505128115e-06,
+      "loss": 0.9967,
+      "step": 5246
+    },
+    {
+      "epoch": 0.5518293083728818,
+      "grad_norm": 2.167403323911424,
+      "learning_rate": 2.131522328160075e-06,
+      "loss": 0.9711,
+      "step": 5247
+    },
+    {
+      "epoch": 0.5519344788147291,
+      "grad_norm": 2.3379019799741236,
+      "learning_rate": 2.130696646895461e-06,
+      "loss": 0.9972,
+      "step": 5248
+    },
+    {
+      "epoch": 0.5520396492565764,
+      "grad_norm": 2.059963339031156,
+      "learning_rate": 2.12987100681104e-06,
+      "loss": 0.9488,
+      "step": 5249
+    },
+    {
+      "epoch": 0.5521448196984238,
+      "grad_norm": 2.4196742713351194,
+      "learning_rate": 2.1290454079988775e-06,
+      "loss": 1.0105,
+      "step": 5250
+    },
+    {
+      "epoch": 0.5522499901402711,
+      "grad_norm": 2.462036105246982,
+      "learning_rate": 2.128219850551034e-06,
+      "loss": 0.9945,
+      "step": 5251
+    },
+    {
+      "epoch": 0.5523551605821184,
+      "grad_norm": 1.8416505320697634,
+      "learning_rate": 2.1273943345595637e-06,
+      "loss": 1.0043,
+      "step": 5252
+    },
+    {
+      "epoch": 0.5524603310239657,
+      "grad_norm": 3.251515812829299,
+      "learning_rate": 2.1265688601165206e-06,
+      "loss": 1.0242,
+      "step": 5253
+    },
+    {
+      "epoch": 0.5525655014658131,
+      "grad_norm": 2.6527329475849313,
+      "learning_rate": 2.125743427313951e-06,
+      "loss": 1.0045,
+      "step": 5254
+    },
+    {
+      "epoch": 0.5526706719076604,
+      "grad_norm": 1.9950343766089205,
+      "learning_rate": 2.124918036243896e-06,
+      "loss": 0.9664,
+      "step": 5255
+    },
+    {
+      "epoch": 0.5527758423495077,
+      "grad_norm": 2.2943708516221455,
+      "learning_rate": 2.124092686998394e-06,
+      "loss": 1.0111,
+      "step": 5256
+    },
+    {
+      "epoch": 0.552881012791355,
+      "grad_norm": 1.6382639178089367,
+      "learning_rate": 2.123267379669477e-06,
+      "loss": 1.0042,
+      "step": 5257
+    },
+    {
+      "epoch": 0.5529861832332023,
+      "grad_norm": 2.9901639984780255,
+      "learning_rate": 2.122442114349174e-06,
+      "loss": 1.0122,
+      "step": 5258
+    },
+    {
+      "epoch": 0.5530913536750496,
+      "grad_norm": 2.1998167103396287,
+      "learning_rate": 2.1216168911295085e-06,
+      "loss": 1.0353,
+      "step": 5259
+    },
+    {
+      "epoch": 0.5531965241168969,
+      "grad_norm": 2.9897721880211736,
+      "learning_rate": 2.120791710102499e-06,
+      "loss": 1.0083,
+      "step": 5260
+    },
+    {
+      "epoch": 0.5533016945587442,
+      "grad_norm": 2.054680773983714,
+      "learning_rate": 2.1199665713601593e-06,
+      "loss": 1.0036,
+      "step": 5261
+    },
+    {
+      "epoch": 0.5534068650005916,
+      "grad_norm": 2.617746462194278,
+      "learning_rate": 2.1191414749944985e-06,
+      "loss": 0.9665,
+      "step": 5262
+    },
+    {
+      "epoch": 0.5535120354424389,
+      "grad_norm": 1.9936196114564972,
+      "learning_rate": 2.1183164210975226e-06,
+      "loss": 1.0104,
+      "step": 5263
+    },
+    {
+      "epoch": 0.5536172058842862,
+      "grad_norm": 2.469231832845493,
+      "learning_rate": 2.1174914097612308e-06,
+      "loss": 0.994,
+      "step": 5264
+    },
+    {
+      "epoch": 0.5537223763261335,
+      "grad_norm": 2.93259002955845,
+      "learning_rate": 2.1166664410776184e-06,
+      "loss": 1.0127,
+      "step": 5265
+    },
+    {
+      "epoch": 0.5538275467679808,
+      "grad_norm": 2.0537797985124326,
+      "learning_rate": 2.1158415151386746e-06,
+      "loss": 0.9471,
+      "step": 5266
+    },
+    {
+      "epoch": 0.5539327172098282,
+      "grad_norm": 2.134532328734205,
+      "learning_rate": 2.115016632036387e-06,
+      "loss": 0.9859,
+      "step": 5267
+    },
+    {
+      "epoch": 0.5540378876516755,
+      "grad_norm": 2.2964327708883134,
+      "learning_rate": 2.1141917918627357e-06,
+      "loss": 1.0047,
+      "step": 5268
+    },
+    {
+      "epoch": 0.5541430580935228,
+      "grad_norm": 2.8516902511053535,
+      "learning_rate": 2.113366994709697e-06,
+      "loss": 1.0079,
+      "step": 5269
+    },
+    {
+      "epoch": 0.5542482285353701,
+      "grad_norm": 2.3546571378436743,
+      "learning_rate": 2.1125422406692416e-06,
+      "loss": 0.9691,
+      "step": 5270
+    },
+    {
+      "epoch": 0.5543533989772175,
+      "grad_norm": 2.630944240951188,
+      "learning_rate": 2.1117175298333347e-06,
+      "loss": 0.983,
+      "step": 5271
+    },
+    {
+      "epoch": 0.5544585694190648,
+      "grad_norm": 1.8823762990892248,
+      "learning_rate": 2.1108928622939413e-06,
+      "loss": 0.9626,
+      "step": 5272
+    },
+    {
+      "epoch": 0.5545637398609121,
+      "grad_norm": 2.355887099553958,
+      "learning_rate": 2.110068238143016e-06,
+      "loss": 0.9441,
+      "step": 5273
+    },
+    {
+      "epoch": 0.5546689103027594,
+      "grad_norm": 2.4481500085464205,
+      "learning_rate": 2.1092436574725113e-06,
+      "loss": 1.0182,
+      "step": 5274
+    },
+    {
+      "epoch": 0.5547740807446068,
+      "grad_norm": 2.088588569858711,
+      "learning_rate": 2.1084191203743732e-06,
+      "loss": 0.9869,
+      "step": 5275
+    },
+    {
+      "epoch": 0.5548792511864541,
+      "grad_norm": 2.935034768068429,
+      "learning_rate": 2.1075946269405464e-06,
+      "loss": 1.0415,
+      "step": 5276
+    },
+    {
+      "epoch": 0.5549844216283014,
+      "grad_norm": 2.511909936791334,
+      "learning_rate": 2.106770177262967e-06,
+      "loss": 1.0,
+      "step": 5277
+    },
+    {
+      "epoch": 0.5550895920701486,
+      "grad_norm": 2.489430938185479,
+      "learning_rate": 2.105945771433567e-06,
+      "loss": 1.013,
+      "step": 5278
+    },
+    {
+      "epoch": 0.555194762511996,
+      "grad_norm": 2.8817643395461388,
+      "learning_rate": 2.105121409544275e-06,
+      "loss": 1.002,
+      "step": 5279
+    },
+    {
+      "epoch": 0.5552999329538433,
+      "grad_norm": 2.9876800647989525,
+      "learning_rate": 2.104297091687013e-06,
+      "loss": 0.9837,
+      "step": 5280
+    },
+    {
+      "epoch": 0.5554051033956906,
+      "grad_norm": 2.521676022126096,
+      "learning_rate": 2.1034728179536996e-06,
+      "loss": 0.9923,
+      "step": 5281
+    },
+    {
+      "epoch": 0.5555102738375379,
+      "grad_norm": 2.3179342164652317,
+      "learning_rate": 2.102648588436247e-06,
+      "loss": 1.0151,
+      "step": 5282
+    },
+    {
+      "epoch": 0.5556154442793853,
+      "grad_norm": 2.328687099694681,
+      "learning_rate": 2.101824403226564e-06,
+      "loss": 0.9904,
+      "step": 5283
+    },
+    {
+      "epoch": 0.5557206147212326,
+      "grad_norm": 2.4299365708893346,
+      "learning_rate": 2.1010002624165528e-06,
+      "loss": 1.0372,
+      "step": 5284
+    },
+    {
+      "epoch": 0.5558257851630799,
+      "grad_norm": 1.9606388049139083,
+      "learning_rate": 2.100176166098111e-06,
+      "loss": 0.9814,
+      "step": 5285
+    },
+    {
+      "epoch": 0.5559309556049272,
+      "grad_norm": 1.6767864003425093,
+      "learning_rate": 2.0993521143631335e-06,
+      "loss": 0.9458,
+      "step": 5286
+    },
+    {
+      "epoch": 0.5560361260467745,
+      "grad_norm": 2.1286537540888175,
+      "learning_rate": 2.098528107303508e-06,
+      "loss": 0.962,
+      "step": 5287
+    },
+    {
+      "epoch": 0.5561412964886219,
+      "grad_norm": 2.5220581659039674,
+      "learning_rate": 2.0977041450111173e-06,
+      "loss": 1.0144,
+      "step": 5288
+    },
+    {
+      "epoch": 0.5562464669304692,
+      "grad_norm": 2.168639158568607,
+      "learning_rate": 2.0968802275778384e-06,
+      "loss": 1.0465,
+      "step": 5289
+    },
+    {
+      "epoch": 0.5563516373723165,
+      "grad_norm": 2.760268469404148,
+      "learning_rate": 2.0960563550955465e-06,
+      "loss": 0.9649,
+      "step": 5290
+    },
+    {
+      "epoch": 0.5564568078141638,
+      "grad_norm": 2.1553647272545158,
+      "learning_rate": 2.095232527656109e-06,
+      "loss": 0.999,
+      "step": 5291
+    },
+    {
+      "epoch": 0.5565619782560112,
+      "grad_norm": 1.619929018860887,
+      "learning_rate": 2.0944087453513887e-06,
+      "loss": 0.977,
+      "step": 5292
+    },
+    {
+      "epoch": 0.5566671486978585,
+      "grad_norm": 3.121330671866382,
+      "learning_rate": 2.0935850082732444e-06,
+      "loss": 1.0139,
+      "step": 5293
+    },
+    {
+      "epoch": 0.5567723191397058,
+      "grad_norm": 2.3387671861808363,
+      "learning_rate": 2.0927613165135285e-06,
+      "loss": 1.0007,
+      "step": 5294
+    },
+    {
+      "epoch": 0.5568774895815531,
+      "grad_norm": 2.200825022390463,
+      "learning_rate": 2.0919376701640897e-06,
+      "loss": 0.9796,
+      "step": 5295
+    },
+    {
+      "epoch": 0.5569826600234005,
+      "grad_norm": 2.3884861298168962,
+      "learning_rate": 2.0911140693167703e-06,
+      "loss": 0.9875,
+      "step": 5296
+    },
+    {
+      "epoch": 0.5570878304652478,
+      "grad_norm": 2.5987592283460597,
+      "learning_rate": 2.0902905140634087e-06,
+      "loss": 1.0331,
+      "step": 5297
+    },
+    {
+      "epoch": 0.5571930009070951,
+      "grad_norm": 3.301654294448108,
+      "learning_rate": 2.0894670044958364e-06,
+      "loss": 1.0113,
+      "step": 5298
+    },
+    {
+      "epoch": 0.5572981713489423,
+      "grad_norm": 2.711878855343128,
+      "learning_rate": 2.0886435407058836e-06,
+      "loss": 1.0027,
+      "step": 5299
+    },
+    {
+      "epoch": 0.5574033417907897,
+      "grad_norm": 2.8480009398658406,
+      "learning_rate": 2.087820122785371e-06,
+      "loss": 0.974,
+      "step": 5300
+    },
+    {
+      "epoch": 0.557508512232637,
+      "grad_norm": 2.3312290115794108,
+      "learning_rate": 2.0869967508261175e-06,
+      "loss": 0.9928,
+      "step": 5301
+    },
+    {
+      "epoch": 0.5576136826744843,
+      "grad_norm": 1.936637576330423,
+      "learning_rate": 2.086173424919934e-06,
+      "loss": 1.0031,
+      "step": 5302
+    },
+    {
+      "epoch": 0.5577188531163316,
+      "grad_norm": 1.902158601451205,
+      "learning_rate": 2.085350145158628e-06,
+      "loss": 1.0094,
+      "step": 5303
+    },
+    {
+      "epoch": 0.557824023558179,
+      "grad_norm": 2.1117431500701827,
+      "learning_rate": 2.084526911634002e-06,
+      "loss": 0.9817,
+      "step": 5304
+    },
+    {
+      "epoch": 0.5579291940000263,
+      "grad_norm": 2.541061650720203,
+      "learning_rate": 2.0837037244378534e-06,
+      "loss": 0.9666,
+      "step": 5305
+    },
+    {
+      "epoch": 0.5580343644418736,
+      "grad_norm": 2.1751713964862716,
+      "learning_rate": 2.082880583661973e-06,
+      "loss": 0.9924,
+      "step": 5306
+    },
+    {
+      "epoch": 0.5581395348837209,
+      "grad_norm": 2.0010450338415353,
+      "learning_rate": 2.082057489398148e-06,
+      "loss": 1.012,
+      "step": 5307
+    },
+    {
+      "epoch": 0.5582447053255682,
+      "grad_norm": 2.476716525076265,
+      "learning_rate": 2.0812344417381595e-06,
+      "loss": 1.0308,
+      "step": 5308
+    },
+    {
+      "epoch": 0.5583498757674156,
+      "grad_norm": 2.563557102937592,
+      "learning_rate": 2.0804114407737837e-06,
+      "loss": 0.9702,
+      "step": 5309
+    },
+    {
+      "epoch": 0.5584550462092629,
+      "grad_norm": 2.7438116844536573,
+      "learning_rate": 2.0795884865967922e-06,
+      "loss": 1.0164,
+      "step": 5310
+    },
+    {
+      "epoch": 0.5585602166511102,
+      "grad_norm": 2.8471554577442273,
+      "learning_rate": 2.07876557929895e-06,
+      "loss": 1.007,
+      "step": 5311
+    },
+    {
+      "epoch": 0.5586653870929575,
+      "grad_norm": 2.018132015140999,
+      "learning_rate": 2.077942718972017e-06,
+      "loss": 1.0139,
+      "step": 5312
+    },
+    {
+      "epoch": 0.5587705575348049,
+      "grad_norm": 2.125285679282089,
+      "learning_rate": 2.0771199057077507e-06,
+      "loss": 1.028,
+      "step": 5313
+    },
+    {
+      "epoch": 0.5588757279766522,
+      "grad_norm": 2.7625535043659775,
+      "learning_rate": 2.0762971395978996e-06,
+      "loss": 1.0062,
+      "step": 5314
+    },
+    {
+      "epoch": 0.5589808984184995,
+      "grad_norm": 2.1017174794213997,
+      "learning_rate": 2.0754744207342097e-06,
+      "loss": 1.0082,
+      "step": 5315
+    },
+    {
+      "epoch": 0.5590860688603468,
+      "grad_norm": 1.782677163365059,
+      "learning_rate": 2.074651749208419e-06,
+      "loss": 0.9765,
+      "step": 5316
+    },
+    {
+      "epoch": 0.5591912393021942,
+      "grad_norm": 3.167958590956396,
+      "learning_rate": 2.0738291251122624e-06,
+      "loss": 0.9923,
+      "step": 5317
+    },
+    {
+      "epoch": 0.5592964097440415,
+      "grad_norm": 1.907619393947842,
+      "learning_rate": 2.073006548537469e-06,
+      "loss": 1.0287,
+      "step": 5318
+    },
+    {
+      "epoch": 0.5594015801858887,
+      "grad_norm": 2.7345986219575256,
+      "learning_rate": 2.0721840195757626e-06,
+      "loss": 0.9527,
+      "step": 5319
+    },
+    {
+      "epoch": 0.559506750627736,
+      "grad_norm": 2.4886965308181384,
+      "learning_rate": 2.0713615383188615e-06,
+      "loss": 1.0136,
+      "step": 5320
+    },
+    {
+      "epoch": 0.5596119210695834,
+      "grad_norm": 2.6055536744435512,
+      "learning_rate": 2.0705391048584775e-06,
+      "loss": 0.9965,
+      "step": 5321
+    },
+    {
+      "epoch": 0.5597170915114307,
+      "grad_norm": 2.087005530742932,
+      "learning_rate": 2.0697167192863205e-06,
+      "loss": 1.0241,
+      "step": 5322
+    },
+    {
+      "epoch": 0.559822261953278,
+      "grad_norm": 2.5369112749685554,
+      "learning_rate": 2.0688943816940927e-06,
+      "loss": 1.0075,
+      "step": 5323
+    },
+    {
+      "epoch": 0.5599274323951253,
+      "grad_norm": 2.7221541004536953,
+      "learning_rate": 2.0680720921734894e-06,
+      "loss": 0.9315,
+      "step": 5324
+    },
+    {
+      "epoch": 0.5600326028369726,
+      "grad_norm": 2.022091119318397,
+      "learning_rate": 2.067249850816203e-06,
+      "loss": 0.9864,
+      "step": 5325
+    },
+    {
+      "epoch": 0.56013777327882,
+      "grad_norm": 2.3688556004256567,
+      "learning_rate": 2.0664276577139193e-06,
+      "loss": 1.0195,
+      "step": 5326
+    },
+    {
+      "epoch": 0.5602429437206673,
+      "grad_norm": 2.4773968134541096,
+      "learning_rate": 2.065605512958321e-06,
+      "loss": 1.0232,
+      "step": 5327
+    },
+    {
+      "epoch": 0.5603481141625146,
+      "grad_norm": 2.462817643536204,
+      "learning_rate": 2.0647834166410825e-06,
+      "loss": 0.9874,
+      "step": 5328
+    },
+    {
+      "epoch": 0.560453284604362,
+      "grad_norm": 2.305518728419192,
+      "learning_rate": 2.0639613688538733e-06,
+      "loss": 0.9921,
+      "step": 5329
+    },
+    {
+      "epoch": 0.5605584550462093,
+      "grad_norm": 2.6367842405171005,
+      "learning_rate": 2.063139369688359e-06,
+      "loss": 0.9613,
+      "step": 5330
+    },
+    {
+      "epoch": 0.5606636254880566,
+      "grad_norm": 1.8818850010645942,
+      "learning_rate": 2.062317419236199e-06,
+      "loss": 0.9777,
+      "step": 5331
+    },
+    {
+      "epoch": 0.5607687959299039,
+      "grad_norm": 2.1790160668617444,
+      "learning_rate": 2.0614955175890464e-06,
+      "loss": 1.0092,
+      "step": 5332
+    },
+    {
+      "epoch": 0.5608739663717512,
+      "grad_norm": 2.0639573437892857,
+      "learning_rate": 2.06067366483855e-06,
+      "loss": 1.0138,
+      "step": 5333
+    },
+    {
+      "epoch": 0.5609791368135986,
+      "grad_norm": 1.9579365333347298,
+      "learning_rate": 2.0598518610763534e-06,
+      "loss": 0.9952,
+      "step": 5334
+    },
+    {
+      "epoch": 0.5610843072554459,
+      "grad_norm": 2.2186014741559656,
+      "learning_rate": 2.0590301063940917e-06,
+      "loss": 0.9515,
+      "step": 5335
+    },
+    {
+      "epoch": 0.5611894776972932,
+      "grad_norm": 1.9011562436125449,
+      "learning_rate": 2.0582084008834003e-06,
+      "loss": 0.9342,
+      "step": 5336
+    },
+    {
+      "epoch": 0.5612946481391405,
+      "grad_norm": 1.8095902323216666,
+      "learning_rate": 2.057386744635904e-06,
+      "loss": 1.0039,
+      "step": 5337
+    },
+    {
+      "epoch": 0.5613998185809879,
+      "grad_norm": 1.8780136670652687,
+      "learning_rate": 2.056565137743224e-06,
+      "loss": 0.9827,
+      "step": 5338
+    },
+    {
+      "epoch": 0.5615049890228351,
+      "grad_norm": 3.2992844080330874,
+      "learning_rate": 2.055743580296976e-06,
+      "loss": 0.9805,
+      "step": 5339
+    },
+    {
+      "epoch": 0.5616101594646824,
+      "grad_norm": 3.092310843970352,
+      "learning_rate": 2.0549220723887687e-06,
+      "loss": 1.0094,
+      "step": 5340
+    },
+    {
+      "epoch": 0.5617153299065297,
+      "grad_norm": 2.4046006250279435,
+      "learning_rate": 2.0541006141102086e-06,
+      "loss": 0.967,
+      "step": 5341
+    },
+    {
+      "epoch": 0.561820500348377,
+      "grad_norm": 2.1278856112151336,
+      "learning_rate": 2.0532792055528946e-06,
+      "loss": 1.012,
+      "step": 5342
+    },
+    {
+      "epoch": 0.5619256707902244,
+      "grad_norm": 3.1700362603693475,
+      "learning_rate": 2.052457846808419e-06,
+      "loss": 0.9602,
+      "step": 5343
+    },
+    {
+      "epoch": 0.5620308412320717,
+      "grad_norm": 1.9326932220836064,
+      "learning_rate": 2.0516365379683694e-06,
+      "loss": 0.9952,
+      "step": 5344
+    },
+    {
+      "epoch": 0.562136011673919,
+      "grad_norm": 2.7114145309970086,
+      "learning_rate": 2.0508152791243296e-06,
+      "loss": 0.948,
+      "step": 5345
+    },
+    {
+      "epoch": 0.5622411821157663,
+      "grad_norm": 2.3203144056712706,
+      "learning_rate": 2.0499940703678755e-06,
+      "loss": 0.9971,
+      "step": 5346
+    },
+    {
+      "epoch": 0.5623463525576137,
+      "grad_norm": 2.3494148393308154,
+      "learning_rate": 2.049172911790578e-06,
+      "loss": 0.9742,
+      "step": 5347
+    },
+    {
+      "epoch": 0.562451522999461,
+      "grad_norm": 2.0875975574978782,
+      "learning_rate": 2.0483518034840034e-06,
+      "loss": 0.9948,
+      "step": 5348
+    },
+    {
+      "epoch": 0.5625566934413083,
+      "grad_norm": 1.970777139152171,
+      "learning_rate": 2.0475307455397103e-06,
+      "loss": 0.946,
+      "step": 5349
+    },
+    {
+      "epoch": 0.5626618638831556,
+      "grad_norm": 2.6489850276180555,
+      "learning_rate": 2.0467097380492547e-06,
+      "loss": 1.0242,
+      "step": 5350
+    },
+    {
+      "epoch": 0.562767034325003,
+      "grad_norm": 2.199452020540128,
+      "learning_rate": 2.0458887811041842e-06,
+      "loss": 1.0094,
+      "step": 5351
+    },
+    {
+      "epoch": 0.5628722047668503,
+      "grad_norm": 2.3720136627286084,
+      "learning_rate": 2.045067874796043e-06,
+      "loss": 0.9981,
+      "step": 5352
+    },
+    {
+      "epoch": 0.5629773752086976,
+      "grad_norm": 2.5209154043777073,
+      "learning_rate": 2.044247019216367e-06,
+      "loss": 1.0028,
+      "step": 5353
+    },
+    {
+      "epoch": 0.5630825456505449,
+      "grad_norm": 2.5610925238059186,
+      "learning_rate": 2.0434262144566895e-06,
+      "loss": 1.0665,
+      "step": 5354
+    },
+    {
+      "epoch": 0.5631877160923923,
+      "grad_norm": 2.897831403216789,
+      "learning_rate": 2.0426054606085356e-06,
+      "loss": 1.0275,
+      "step": 5355
+    },
+    {
+      "epoch": 0.5632928865342396,
+      "grad_norm": 2.376676302583607,
+      "learning_rate": 2.0417847577634263e-06,
+      "loss": 0.9664,
+      "step": 5356
+    },
+    {
+      "epoch": 0.5633980569760869,
+      "grad_norm": 1.902345338626602,
+      "learning_rate": 2.040964106012876e-06,
+      "loss": 0.964,
+      "step": 5357
+    },
+    {
+      "epoch": 0.5635032274179342,
+      "grad_norm": 2.8141159096441513,
+      "learning_rate": 2.0401435054483925e-06,
+      "loss": 0.959,
+      "step": 5358
+    },
+    {
+      "epoch": 0.5636083978597816,
+      "grad_norm": 2.213475597622369,
+      "learning_rate": 2.0393229561614817e-06,
+      "loss": 0.9913,
+      "step": 5359
+    },
+    {
+      "epoch": 0.5637135683016288,
+      "grad_norm": 2.8077710198008616,
+      "learning_rate": 2.03850245824364e-06,
+      "loss": 0.9795,
+      "step": 5360
+    },
+    {
+      "epoch": 0.5638187387434761,
+      "grad_norm": 1.7009560943465891,
+      "learning_rate": 2.037682011786359e-06,
+      "loss": 0.9519,
+      "step": 5361
+    },
+    {
+      "epoch": 0.5639239091853234,
+      "grad_norm": 1.9158342012022915,
+      "learning_rate": 2.036861616881125e-06,
+      "loss": 1.0132,
+      "step": 5362
+    },
+    {
+      "epoch": 0.5640290796271707,
+      "grad_norm": 2.4899098257579384,
+      "learning_rate": 2.036041273619418e-06,
+      "loss": 0.9862,
+      "step": 5363
+    },
+    {
+      "epoch": 0.5641342500690181,
+      "grad_norm": 2.8756220189009487,
+      "learning_rate": 2.035220982092714e-06,
+      "loss": 0.9549,
+      "step": 5364
+    },
+    {
+      "epoch": 0.5642394205108654,
+      "grad_norm": 2.466341733072218,
+      "learning_rate": 2.0344007423924807e-06,
+      "loss": 0.9962,
+      "step": 5365
+    },
+    {
+      "epoch": 0.5643445909527127,
+      "grad_norm": 2.1782294813736254,
+      "learning_rate": 2.0335805546101817e-06,
+      "loss": 0.9483,
+      "step": 5366
+    },
+    {
+      "epoch": 0.56444976139456,
+      "grad_norm": 2.117012458223317,
+      "learning_rate": 2.0327604188372735e-06,
+      "loss": 1.0088,
+      "step": 5367
+    },
+    {
+      "epoch": 0.5645549318364074,
+      "grad_norm": 1.9302647942163518,
+      "learning_rate": 2.0319403351652086e-06,
+      "loss": 0.9985,
+      "step": 5368
+    },
+    {
+      "epoch": 0.5646601022782547,
+      "grad_norm": 2.2737930556012915,
+      "learning_rate": 2.0311203036854326e-06,
+      "loss": 0.9287,
+      "step": 5369
+    },
+    {
+      "epoch": 0.564765272720102,
+      "grad_norm": 1.6722842954733659,
+      "learning_rate": 2.0303003244893853e-06,
+      "loss": 1.0125,
+      "step": 5370
+    },
+    {
+      "epoch": 0.5648704431619493,
+      "grad_norm": 2.9762602394882434,
+      "learning_rate": 2.0294803976685006e-06,
+      "loss": 0.9461,
+      "step": 5371
+    },
+    {
+      "epoch": 0.5649756136037967,
+      "grad_norm": 2.979884850333689,
+      "learning_rate": 2.028660523314205e-06,
+      "loss": 1.0165,
+      "step": 5372
+    },
+    {
+      "epoch": 0.565080784045644,
+      "grad_norm": 2.4718005740112305,
+      "learning_rate": 2.0278407015179243e-06,
+      "loss": 0.9953,
+      "step": 5373
+    },
+    {
+      "epoch": 0.5651859544874913,
+      "grad_norm": 2.038230117240658,
+      "learning_rate": 2.027020932371073e-06,
+      "loss": 0.9773,
+      "step": 5374
+    },
+    {
+      "epoch": 0.5652911249293386,
+      "grad_norm": 2.3536455408799988,
+      "learning_rate": 2.0262012159650624e-06,
+      "loss": 0.9947,
+      "step": 5375
+    },
+    {
+      "epoch": 0.565396295371186,
+      "grad_norm": 2.4699699581202252,
+      "learning_rate": 2.0253815523912955e-06,
+      "loss": 1.0202,
+      "step": 5376
+    },
+    {
+      "epoch": 0.5655014658130333,
+      "grad_norm": 2.7568128217452608,
+      "learning_rate": 2.024561941741173e-06,
+      "loss": 0.9816,
+      "step": 5377
+    },
+    {
+      "epoch": 0.5656066362548806,
+      "grad_norm": 2.5880748676931025,
+      "learning_rate": 2.0237423841060877e-06,
+      "loss": 0.9817,
+      "step": 5378
+    },
+    {
+      "epoch": 0.5657118066967279,
+      "grad_norm": 2.3807827153101138,
+      "learning_rate": 2.0229228795774264e-06,
+      "loss": 0.9829,
+      "step": 5379
+    },
+    {
+      "epoch": 0.5658169771385752,
+      "grad_norm": 2.419880425354416,
+      "learning_rate": 2.02210342824657e-06,
+      "loss": 0.9412,
+      "step": 5380
+    },
+    {
+      "epoch": 0.5659221475804225,
+      "grad_norm": 3.206944366362426,
+      "learning_rate": 2.021284030204893e-06,
+      "loss": 1.0397,
+      "step": 5381
+    },
+    {
+      "epoch": 0.5660273180222698,
+      "grad_norm": 2.48547978811849,
+      "learning_rate": 2.020464685543766e-06,
+      "loss": 0.971,
+      "step": 5382
+    },
+    {
+      "epoch": 0.5661324884641171,
+      "grad_norm": 2.6366008661640525,
+      "learning_rate": 2.0196453943545517e-06,
+      "loss": 1.0017,
+      "step": 5383
+    },
+    {
+      "epoch": 0.5662376589059644,
+      "grad_norm": 2.6842165173464867,
+      "learning_rate": 2.0188261567286076e-06,
+      "loss": 1.0313,
+      "step": 5384
+    },
+    {
+      "epoch": 0.5663428293478118,
+      "grad_norm": 2.6982167121358684,
+      "learning_rate": 2.018006972757285e-06,
+      "loss": 0.9802,
+      "step": 5385
+    },
+    {
+      "epoch": 0.5664479997896591,
+      "grad_norm": 2.4433736862921362,
+      "learning_rate": 2.0171878425319283e-06,
+      "loss": 0.962,
+      "step": 5386
+    },
+    {
+      "epoch": 0.5665531702315064,
+      "grad_norm": 2.475980829378924,
+      "learning_rate": 2.0163687661438786e-06,
+      "loss": 0.9586,
+      "step": 5387
+    },
+    {
+      "epoch": 0.5666583406733537,
+      "grad_norm": 2.6704965382072063,
+      "learning_rate": 2.0155497436844684e-06,
+      "loss": 1.0241,
+      "step": 5388
+    },
+    {
+      "epoch": 0.5667635111152011,
+      "grad_norm": 2.6472388985915325,
+      "learning_rate": 2.0147307752450253e-06,
+      "loss": 0.9816,
+      "step": 5389
+    },
+    {
+      "epoch": 0.5668686815570484,
+      "grad_norm": 2.2155377465887067,
+      "learning_rate": 2.0139118609168697e-06,
+      "loss": 0.996,
+      "step": 5390
+    },
+    {
+      "epoch": 0.5669738519988957,
+      "grad_norm": 2.6569851803737827,
+      "learning_rate": 2.0130930007913184e-06,
+      "loss": 1.003,
+      "step": 5391
+    },
+    {
+      "epoch": 0.567079022440743,
+      "grad_norm": 2.2619946690776236,
+      "learning_rate": 2.01227419495968e-06,
+      "loss": 1.0121,
+      "step": 5392
+    },
+    {
+      "epoch": 0.5671841928825904,
+      "grad_norm": 3.0146876977488497,
+      "learning_rate": 2.011455443513257e-06,
+      "loss": 1.0017,
+      "step": 5393
+    },
+    {
+      "epoch": 0.5672893633244377,
+      "grad_norm": 2.7358969902861303,
+      "learning_rate": 2.010636746543348e-06,
+      "loss": 1.0223,
+      "step": 5394
+    },
+    {
+      "epoch": 0.567394533766285,
+      "grad_norm": 2.5446999109646975,
+      "learning_rate": 2.009818104141242e-06,
+      "loss": 0.9872,
+      "step": 5395
+    },
+    {
+      "epoch": 0.5674997042081323,
+      "grad_norm": 2.5508462818540094,
+      "learning_rate": 2.0089995163982263e-06,
+      "loss": 1.0033,
+      "step": 5396
+    },
+    {
+      "epoch": 0.5676048746499797,
+      "grad_norm": 1.9993969226847457,
+      "learning_rate": 2.0081809834055787e-06,
+      "loss": 0.9702,
+      "step": 5397
+    },
+    {
+      "epoch": 0.567710045091827,
+      "grad_norm": 2.836401177949353,
+      "learning_rate": 2.007362505254572e-06,
+      "loss": 0.9877,
+      "step": 5398
+    },
+    {
+      "epoch": 0.5678152155336743,
+      "grad_norm": 2.653491277012509,
+      "learning_rate": 2.0065440820364718e-06,
+      "loss": 0.9803,
+      "step": 5399
+    },
+    {
+      "epoch": 0.5679203859755215,
+      "grad_norm": 3.4470366242642854,
+      "learning_rate": 2.005725713842541e-06,
+      "loss": 1.0328,
+      "step": 5400
+    },
+    {
+      "epoch": 0.5680255564173688,
+      "grad_norm": 2.9653592066566667,
+      "learning_rate": 2.0049074007640324e-06,
+      "loss": 1.0015,
+      "step": 5401
+    },
+    {
+      "epoch": 0.5681307268592162,
+      "grad_norm": 1.4100311048508116,
+      "learning_rate": 2.004089142892195e-06,
+      "loss": 0.9846,
+      "step": 5402
+    },
+    {
+      "epoch": 0.5682358973010635,
+      "grad_norm": 2.9910996927005007,
+      "learning_rate": 2.0032709403182705e-06,
+      "loss": 1.0175,
+      "step": 5403
+    },
+    {
+      "epoch": 0.5683410677429108,
+      "grad_norm": 1.6159196986325306,
+      "learning_rate": 2.002452793133494e-06,
+      "loss": 0.958,
+      "step": 5404
+    },
+    {
+      "epoch": 0.5684462381847581,
+      "grad_norm": 2.343270925363438,
+      "learning_rate": 2.001634701429097e-06,
+      "loss": 1.0338,
+      "step": 5405
+    },
+    {
+      "epoch": 0.5685514086266055,
+      "grad_norm": 2.6339615242917187,
+      "learning_rate": 2.000816665296302e-06,
+      "loss": 0.9807,
+      "step": 5406
+    },
+    {
+      "epoch": 0.5686565790684528,
+      "grad_norm": 2.928072201278944,
+      "learning_rate": 1.999998684826327e-06,
+      "loss": 1.0176,
+      "step": 5407
+    },
+    {
+      "epoch": 0.5687617495103001,
+      "grad_norm": 2.4139743488107976,
+      "learning_rate": 1.9991807601103823e-06,
+      "loss": 0.984,
+      "step": 5408
+    },
+    {
+      "epoch": 0.5688669199521474,
+      "grad_norm": 2.171791427192439,
+      "learning_rate": 1.9983628912396726e-06,
+      "loss": 1.0245,
+      "step": 5409
+    },
+    {
+      "epoch": 0.5689720903939948,
+      "grad_norm": 2.804934183732406,
+      "learning_rate": 1.997545078305399e-06,
+      "loss": 1.0397,
+      "step": 5410
+    },
+    {
+      "epoch": 0.5690772608358421,
+      "grad_norm": 1.974490805904351,
+      "learning_rate": 1.9967273213987515e-06,
+      "loss": 0.9931,
+      "step": 5411
+    },
+    {
+      "epoch": 0.5691824312776894,
+      "grad_norm": 2.2394117401253917,
+      "learning_rate": 1.995909620610918e-06,
+      "loss": 1.0064,
+      "step": 5412
+    },
+    {
+      "epoch": 0.5692876017195367,
+      "grad_norm": 2.7031568108174393,
+      "learning_rate": 1.9950919760330757e-06,
+      "loss": 0.9608,
+      "step": 5413
+    },
+    {
+      "epoch": 0.5693927721613841,
+      "grad_norm": 1.7767471050410926,
+      "learning_rate": 1.9942743877564018e-06,
+      "loss": 0.9177,
+      "step": 5414
+    },
+    {
+      "epoch": 0.5694979426032314,
+      "grad_norm": 1.9741959110358989,
+      "learning_rate": 1.9934568558720623e-06,
+      "loss": 1.0155,
+      "step": 5415
+    },
+    {
+      "epoch": 0.5696031130450787,
+      "grad_norm": 2.120718009384387,
+      "learning_rate": 1.9926393804712183e-06,
+      "loss": 0.9933,
+      "step": 5416
+    },
+    {
+      "epoch": 0.569708283486926,
+      "grad_norm": 2.3893886698304554,
+      "learning_rate": 1.9918219616450246e-06,
+      "loss": 0.9487,
+      "step": 5417
+    },
+    {
+      "epoch": 0.5698134539287734,
+      "grad_norm": 2.1203012449126217,
+      "learning_rate": 1.9910045994846294e-06,
+      "loss": 1.0161,
+      "step": 5418
+    },
+    {
+      "epoch": 0.5699186243706207,
+      "grad_norm": 2.402604527740921,
+      "learning_rate": 1.990187294081176e-06,
+      "loss": 0.9735,
+      "step": 5419
+    },
+    {
+      "epoch": 0.570023794812468,
+      "grad_norm": 2.3672965950739213,
+      "learning_rate": 1.9893700455257996e-06,
+      "loss": 0.9659,
+      "step": 5420
+    },
+    {
+      "epoch": 0.5701289652543152,
+      "grad_norm": 2.210201260927024,
+      "learning_rate": 1.98855285390963e-06,
+      "loss": 0.9688,
+      "step": 5421
+    },
+    {
+      "epoch": 0.5702341356961625,
+      "grad_norm": 2.653764999915913,
+      "learning_rate": 1.987735719323789e-06,
+      "loss": 0.9776,
+      "step": 5422
+    },
+    {
+      "epoch": 0.5703393061380099,
+      "grad_norm": 2.5642420040701692,
+      "learning_rate": 1.986918641859396e-06,
+      "loss": 0.995,
+      "step": 5423
+    },
+    {
+      "epoch": 0.5704444765798572,
+      "grad_norm": 2.2961949153030443,
+      "learning_rate": 1.9861016216075596e-06,
+      "loss": 0.9892,
+      "step": 5424
+    },
+    {
+      "epoch": 0.5705496470217045,
+      "grad_norm": 2.767041920692721,
+      "learning_rate": 1.985284658659385e-06,
+      "loss": 1.0365,
+      "step": 5425
+    },
+    {
+      "epoch": 0.5706548174635518,
+      "grad_norm": 1.648569046551329,
+      "learning_rate": 1.9844677531059698e-06,
+      "loss": 0.9728,
+      "step": 5426
+    },
+    {
+      "epoch": 0.5707599879053992,
+      "grad_norm": 2.6969141927182005,
+      "learning_rate": 1.9836509050384035e-06,
+      "loss": 0.9942,
+      "step": 5427
+    },
+    {
+      "epoch": 0.5708651583472465,
+      "grad_norm": 2.2808653469120084,
+      "learning_rate": 1.982834114547773e-06,
+      "loss": 0.9663,
+      "step": 5428
+    },
+    {
+      "epoch": 0.5709703287890938,
+      "grad_norm": 2.167902550863519,
+      "learning_rate": 1.982017381725157e-06,
+      "loss": 0.992,
+      "step": 5429
+    },
+    {
+      "epoch": 0.5710754992309411,
+      "grad_norm": 2.3499387994244323,
+      "learning_rate": 1.981200706661626e-06,
+      "loss": 0.9837,
+      "step": 5430
+    },
+    {
+      "epoch": 0.5711806696727885,
+      "grad_norm": 2.182644078032054,
+      "learning_rate": 1.9803840894482468e-06,
+      "loss": 0.9578,
+      "step": 5431
+    },
+    {
+      "epoch": 0.5712858401146358,
+      "grad_norm": 3.3676977139942434,
+      "learning_rate": 1.9795675301760776e-06,
+      "loss": 0.9728,
+      "step": 5432
+    },
+    {
+      "epoch": 0.5713910105564831,
+      "grad_norm": 2.353096165552454,
+      "learning_rate": 1.978751028936172e-06,
+      "loss": 0.9739,
+      "step": 5433
+    },
+    {
+      "epoch": 0.5714961809983304,
+      "grad_norm": 2.5627941445914373,
+      "learning_rate": 1.9779345858195757e-06,
+      "loss": 0.9801,
+      "step": 5434
+    },
+    {
+      "epoch": 0.5716013514401778,
+      "grad_norm": 2.8224949763649367,
+      "learning_rate": 1.977118200917329e-06,
+      "loss": 1.0196,
+      "step": 5435
+    },
+    {
+      "epoch": 0.5717065218820251,
+      "grad_norm": 2.41926342232906,
+      "learning_rate": 1.9763018743204637e-06,
+      "loss": 0.9623,
+      "step": 5436
+    },
+    {
+      "epoch": 0.5718116923238724,
+      "grad_norm": 2.5803768185373457,
+      "learning_rate": 1.9754856061200085e-06,
+      "loss": 1.0174,
+      "step": 5437
+    },
+    {
+      "epoch": 0.5719168627657197,
+      "grad_norm": 2.6612045523496857,
+      "learning_rate": 1.974669396406983e-06,
+      "loss": 0.9732,
+      "step": 5438
+    },
+    {
+      "epoch": 0.5720220332075671,
+      "grad_norm": 1.9777273507957978,
+      "learning_rate": 1.9738532452724007e-06,
+      "loss": 0.9457,
+      "step": 5439
+    },
+    {
+      "epoch": 0.5721272036494144,
+      "grad_norm": 3.0778735684461056,
+      "learning_rate": 1.9730371528072687e-06,
+      "loss": 1.0039,
+      "step": 5440
+    },
+    {
+      "epoch": 0.5722323740912616,
+      "grad_norm": 2.7734640366863537,
+      "learning_rate": 1.972221119102587e-06,
+      "loss": 1.0047,
+      "step": 5441
+    },
+    {
+      "epoch": 0.5723375445331089,
+      "grad_norm": 2.360250328220414,
+      "learning_rate": 1.9714051442493516e-06,
+      "loss": 0.9788,
+      "step": 5442
+    },
+    {
+      "epoch": 0.5724427149749562,
+      "grad_norm": 2.027404275154293,
+      "learning_rate": 1.970589228338548e-06,
+      "loss": 0.9784,
+      "step": 5443
+    },
+    {
+      "epoch": 0.5725478854168036,
+      "grad_norm": 2.3164867712708825,
+      "learning_rate": 1.969773371461159e-06,
+      "loss": 0.9658,
+      "step": 5444
+    },
+    {
+      "epoch": 0.5726530558586509,
+      "grad_norm": 1.9144239908740546,
+      "learning_rate": 1.9689575737081567e-06,
+      "loss": 0.9609,
+      "step": 5445
+    },
+    {
+      "epoch": 0.5727582263004982,
+      "grad_norm": 2.938174043030208,
+      "learning_rate": 1.9681418351705116e-06,
+      "loss": 1.0395,
+      "step": 5446
+    },
+    {
+      "epoch": 0.5728633967423455,
+      "grad_norm": 2.1982822359165013,
+      "learning_rate": 1.967326155939183e-06,
+      "loss": 0.9507,
+      "step": 5447
+    },
+    {
+      "epoch": 0.5729685671841929,
+      "grad_norm": 2.795386799438768,
+      "learning_rate": 1.9665105361051255e-06,
+      "loss": 0.9534,
+      "step": 5448
+    },
+    {
+      "epoch": 0.5730737376260402,
+      "grad_norm": 2.748854629184378,
+      "learning_rate": 1.965694975759288e-06,
+      "loss": 0.99,
+      "step": 5449
+    },
+    {
+      "epoch": 0.5731789080678875,
+      "grad_norm": 2.829975197200465,
+      "learning_rate": 1.9648794749926103e-06,
+      "loss": 1.0186,
+      "step": 5450
+    },
+    {
+      "epoch": 0.5732840785097348,
+      "grad_norm": 2.7529320683500447,
+      "learning_rate": 1.9640640338960294e-06,
+      "loss": 0.9807,
+      "step": 5451
+    },
+    {
+      "epoch": 0.5733892489515822,
+      "grad_norm": 2.558464051548766,
+      "learning_rate": 1.9632486525604715e-06,
+      "loss": 0.9763,
+      "step": 5452
+    },
+    {
+      "epoch": 0.5734944193934295,
+      "grad_norm": 2.2378183775994893,
+      "learning_rate": 1.962433331076859e-06,
+      "loss": 1.0049,
+      "step": 5453
+    },
+    {
+      "epoch": 0.5735995898352768,
+      "grad_norm": 1.9794409306196663,
+      "learning_rate": 1.961618069536105e-06,
+      "loss": 0.9516,
+      "step": 5454
+    },
+    {
+      "epoch": 0.5737047602771241,
+      "grad_norm": 2.5990704751367244,
+      "learning_rate": 1.960802868029119e-06,
+      "loss": 0.9981,
+      "step": 5455
+    },
+    {
+      "epoch": 0.5738099307189715,
+      "grad_norm": 2.111642340238224,
+      "learning_rate": 1.9599877266468024e-06,
+      "loss": 0.9729,
+      "step": 5456
+    },
+    {
+      "epoch": 0.5739151011608188,
+      "grad_norm": 2.337460712610555,
+      "learning_rate": 1.959172645480049e-06,
+      "loss": 0.9886,
+      "step": 5457
+    },
+    {
+      "epoch": 0.5740202716026661,
+      "grad_norm": 3.42592059314716,
+      "learning_rate": 1.958357624619747e-06,
+      "loss": 0.9792,
+      "step": 5458
+    },
+    {
+      "epoch": 0.5741254420445134,
+      "grad_norm": 3.04481036795497,
+      "learning_rate": 1.957542664156776e-06,
+      "loss": 1.0186,
+      "step": 5459
+    },
+    {
+      "epoch": 0.5742306124863608,
+      "grad_norm": 3.566327107470445,
+      "learning_rate": 1.9567277641820136e-06,
+      "loss": 1.0102,
+      "step": 5460
+    },
+    {
+      "epoch": 0.574335782928208,
+      "grad_norm": 2.262412810337984,
+      "learning_rate": 1.9559129247863253e-06,
+      "loss": 0.9987,
+      "step": 5461
+    },
+    {
+      "epoch": 0.5744409533700553,
+      "grad_norm": 2.230615471687081,
+      "learning_rate": 1.9550981460605734e-06,
+      "loss": 0.9789,
+      "step": 5462
+    },
+    {
+      "epoch": 0.5745461238119026,
+      "grad_norm": 2.3226218833907217,
+      "learning_rate": 1.9542834280956102e-06,
+      "loss": 0.9984,
+      "step": 5463
+    },
+    {
+      "epoch": 0.57465129425375,
+      "grad_norm": 3.038633583612286,
+      "learning_rate": 1.9534687709822834e-06,
+      "loss": 1.0057,
+      "step": 5464
+    },
+    {
+      "epoch": 0.5747564646955973,
+      "grad_norm": 2.271266550812082,
+      "learning_rate": 1.952654174811435e-06,
+      "loss": 0.9788,
+      "step": 5465
+    },
+    {
+      "epoch": 0.5748616351374446,
+      "grad_norm": 1.7249233590279198,
+      "learning_rate": 1.951839639673898e-06,
+      "loss": 1.0057,
+      "step": 5466
+    },
+    {
+      "epoch": 0.5749668055792919,
+      "grad_norm": 2.7803541788635417,
+      "learning_rate": 1.9510251656605e-06,
+      "loss": 1.0254,
+      "step": 5467
+    },
+    {
+      "epoch": 0.5750719760211392,
+      "grad_norm": 1.8011230248841785,
+      "learning_rate": 1.9502107528620593e-06,
+      "loss": 0.9893,
+      "step": 5468
+    },
+    {
+      "epoch": 0.5751771464629866,
+      "grad_norm": 2.82475143858172,
+      "learning_rate": 1.949396401369392e-06,
+      "loss": 0.9615,
+      "step": 5469
+    },
+    {
+      "epoch": 0.5752823169048339,
+      "grad_norm": 2.6365865937523014,
+      "learning_rate": 1.9485821112733023e-06,
+      "loss": 0.9794,
+      "step": 5470
+    },
+    {
+      "epoch": 0.5753874873466812,
+      "grad_norm": 2.8901960806978737,
+      "learning_rate": 1.947767882664591e-06,
+      "loss": 1.0382,
+      "step": 5471
+    },
+    {
+      "epoch": 0.5754926577885285,
+      "grad_norm": 1.6028594756674903,
+      "learning_rate": 1.946953715634051e-06,
+      "loss": 1.0084,
+      "step": 5472
+    },
+    {
+      "epoch": 0.5755978282303759,
+      "grad_norm": 2.680056010624629,
+      "learning_rate": 1.9461396102724666e-06,
+      "loss": 0.9539,
+      "step": 5473
+    },
+    {
+      "epoch": 0.5757029986722232,
+      "grad_norm": 1.9522822778958524,
+      "learning_rate": 1.9453255666706193e-06,
+      "loss": 0.9717,
+      "step": 5474
+    },
+    {
+      "epoch": 0.5758081691140705,
+      "grad_norm": 1.8400125921360837,
+      "learning_rate": 1.94451158491928e-06,
+      "loss": 1.0238,
+      "step": 5475
+    },
+    {
+      "epoch": 0.5759133395559178,
+      "grad_norm": 2.9729085425856856,
+      "learning_rate": 1.9436976651092143e-06,
+      "loss": 0.9872,
+      "step": 5476
+    },
+    {
+      "epoch": 0.5760185099977652,
+      "grad_norm": 3.0463506648492644,
+      "learning_rate": 1.94288380733118e-06,
+      "loss": 1.0292,
+      "step": 5477
+    },
+    {
+      "epoch": 0.5761236804396125,
+      "grad_norm": 3.312807191743345,
+      "learning_rate": 1.9420700116759295e-06,
+      "loss": 0.9883,
+      "step": 5478
+    },
+    {
+      "epoch": 0.5762288508814598,
+      "grad_norm": 2.193873965057158,
+      "learning_rate": 1.9412562782342067e-06,
+      "loss": 0.9968,
+      "step": 5479
+    },
+    {
+      "epoch": 0.5763340213233071,
+      "grad_norm": 2.4838716286811677,
+      "learning_rate": 1.9404426070967495e-06,
+      "loss": 0.9702,
+      "step": 5480
+    },
+    {
+      "epoch": 0.5764391917651545,
+      "grad_norm": 3.7749376463759794,
+      "learning_rate": 1.9396289983542884e-06,
+      "loss": 1.0172,
+      "step": 5481
+    },
+    {
+      "epoch": 0.5765443622070017,
+      "grad_norm": 2.5023157843427355,
+      "learning_rate": 1.9388154520975465e-06,
+      "loss": 1.0102,
+      "step": 5482
+    },
+    {
+      "epoch": 0.576649532648849,
+      "grad_norm": 2.7098943623571663,
+      "learning_rate": 1.938001968417242e-06,
+      "loss": 1.0124,
+      "step": 5483
+    },
+    {
+      "epoch": 0.5767547030906963,
+      "grad_norm": 2.548253174199779,
+      "learning_rate": 1.9371885474040838e-06,
+      "loss": 1.0292,
+      "step": 5484
+    },
+    {
+      "epoch": 0.5768598735325436,
+      "grad_norm": 2.5033277887085417,
+      "learning_rate": 1.936375189148774e-06,
+      "loss": 0.9901,
+      "step": 5485
+    },
+    {
+      "epoch": 0.576965043974391,
+      "grad_norm": 2.114247029054567,
+      "learning_rate": 1.9355618937420092e-06,
+      "loss": 0.9394,
+      "step": 5486
+    },
+    {
+      "epoch": 0.5770702144162383,
+      "grad_norm": 1.8131362680388996,
+      "learning_rate": 1.934748661274477e-06,
+      "loss": 0.957,
+      "step": 5487
+    },
+    {
+      "epoch": 0.5771753848580856,
+      "grad_norm": 2.203134930907001,
+      "learning_rate": 1.9339354918368613e-06,
+      "loss": 0.9839,
+      "step": 5488
+    },
+    {
+      "epoch": 0.5772805552999329,
+      "grad_norm": 1.8390231714993754,
+      "learning_rate": 1.9331223855198355e-06,
+      "loss": 0.9737,
+      "step": 5489
+    },
+    {
+      "epoch": 0.5773857257417803,
+      "grad_norm": 2.67902585867112,
+      "learning_rate": 1.9323093424140673e-06,
+      "loss": 0.9992,
+      "step": 5490
+    },
+    {
+      "epoch": 0.5774908961836276,
+      "grad_norm": 1.7899460341786475,
+      "learning_rate": 1.931496362610217e-06,
+      "loss": 1.0174,
+      "step": 5491
+    },
+    {
+      "epoch": 0.5775960666254749,
+      "grad_norm": 2.056265293390317,
+      "learning_rate": 1.930683446198939e-06,
+      "loss": 1.0018,
+      "step": 5492
+    },
+    {
+      "epoch": 0.5777012370673222,
+      "grad_norm": 2.1358816380940886,
+      "learning_rate": 1.9298705932708793e-06,
+      "loss": 0.9516,
+      "step": 5493
+    },
+    {
+      "epoch": 0.5778064075091696,
+      "grad_norm": 1.7026983219962548,
+      "learning_rate": 1.9290578039166775e-06,
+      "loss": 1.0184,
+      "step": 5494
+    },
+    {
+      "epoch": 0.5779115779510169,
+      "grad_norm": 2.710604777509262,
+      "learning_rate": 1.9282450782269657e-06,
+      "loss": 0.9627,
+      "step": 5495
+    },
+    {
+      "epoch": 0.5780167483928642,
+      "grad_norm": 2.4275052932179864,
+      "learning_rate": 1.9274324162923685e-06,
+      "loss": 0.981,
+      "step": 5496
+    },
+    {
+      "epoch": 0.5781219188347115,
+      "grad_norm": 2.639347852054867,
+      "learning_rate": 1.926619818203506e-06,
+      "loss": 1.0225,
+      "step": 5497
+    },
+    {
+      "epoch": 0.5782270892765589,
+      "grad_norm": 3.013422310390402,
+      "learning_rate": 1.925807284050987e-06,
+      "loss": 1.0114,
+      "step": 5498
+    },
+    {
+      "epoch": 0.5783322597184062,
+      "grad_norm": 2.5584374470603453,
+      "learning_rate": 1.9249948139254173e-06,
+      "loss": 0.9257,
+      "step": 5499
+    },
+    {
+      "epoch": 0.5784374301602535,
+      "grad_norm": 2.6663867612637335,
+      "learning_rate": 1.924182407917391e-06,
+      "loss": 0.9599,
+      "step": 5500
+    },
+    {
+      "epoch": 0.5785426006021008,
+      "grad_norm": 3.6833651707957387,
+      "learning_rate": 1.9233700661175e-06,
+      "loss": 0.9995,
+      "step": 5501
+    },
+    {
+      "epoch": 0.578647771043948,
+      "grad_norm": 1.9915527610671544,
+      "learning_rate": 1.922557788616327e-06,
+      "loss": 1.0072,
+      "step": 5502
+    },
+    {
+      "epoch": 0.5787529414857954,
+      "grad_norm": 2.3509129428046993,
+      "learning_rate": 1.921745575504446e-06,
+      "loss": 0.9915,
+      "step": 5503
+    },
+    {
+      "epoch": 0.5788581119276427,
+      "grad_norm": 2.4528194678305097,
+      "learning_rate": 1.920933426872425e-06,
+      "loss": 0.9599,
+      "step": 5504
+    },
+    {
+      "epoch": 0.57896328236949,
+      "grad_norm": 2.6177871712631915,
+      "learning_rate": 1.9201213428108246e-06,
+      "loss": 0.9898,
+      "step": 5505
+    },
+    {
+      "epoch": 0.5790684528113373,
+      "grad_norm": 2.5053984489921155,
+      "learning_rate": 1.9193093234102e-06,
+      "loss": 1.0196,
+      "step": 5506
+    },
+    {
+      "epoch": 0.5791736232531847,
+      "grad_norm": 2.498558815345497,
+      "learning_rate": 1.9184973687610965e-06,
+      "loss": 0.9645,
+      "step": 5507
+    },
+    {
+      "epoch": 0.579278793695032,
+      "grad_norm": 3.05754856518924,
+      "learning_rate": 1.917685478954054e-06,
+      "loss": 1.0264,
+      "step": 5508
+    },
+    {
+      "epoch": 0.5793839641368793,
+      "grad_norm": 2.2653910526244214,
+      "learning_rate": 1.9168736540796043e-06,
+      "loss": 0.9704,
+      "step": 5509
+    },
+    {
+      "epoch": 0.5794891345787266,
+      "grad_norm": 2.48443214283071,
+      "learning_rate": 1.9160618942282705e-06,
+      "loss": 1.0255,
+      "step": 5510
+    },
+    {
+      "epoch": 0.579594305020574,
+      "grad_norm": 3.004247233784531,
+      "learning_rate": 1.915250199490573e-06,
+      "loss": 1.0294,
+      "step": 5511
+    },
+    {
+      "epoch": 0.5796994754624213,
+      "grad_norm": 2.4418258201583103,
+      "learning_rate": 1.914438569957021e-06,
+      "loss": 1.0082,
+      "step": 5512
+    },
+    {
+      "epoch": 0.5798046459042686,
+      "grad_norm": 2.4279113250200823,
+      "learning_rate": 1.9136270057181173e-06,
+      "loss": 0.951,
+      "step": 5513
+    },
+    {
+      "epoch": 0.5799098163461159,
+      "grad_norm": 2.2619878885329263,
+      "learning_rate": 1.9128155068643563e-06,
+      "loss": 1.0286,
+      "step": 5514
+    },
+    {
+      "epoch": 0.5800149867879633,
+      "grad_norm": 3.0238638965690576,
+      "learning_rate": 1.912004073486229e-06,
+      "loss": 0.998,
+      "step": 5515
+    },
+    {
+      "epoch": 0.5801201572298106,
+      "grad_norm": 2.392433746833724,
+      "learning_rate": 1.9111927056742146e-06,
+      "loss": 0.971,
+      "step": 5516
+    },
+    {
+      "epoch": 0.5802253276716579,
+      "grad_norm": 2.2813223909499403,
+      "learning_rate": 1.9103814035187877e-06,
+      "loss": 1.0173,
+      "step": 5517
+    },
+    {
+      "epoch": 0.5803304981135052,
+      "grad_norm": 2.7032063827189092,
+      "learning_rate": 1.909570167110415e-06,
+      "loss": 1.0182,
+      "step": 5518
+    },
+    {
+      "epoch": 0.5804356685553526,
+      "grad_norm": 2.30490410235801,
+      "learning_rate": 1.9087589965395547e-06,
+      "loss": 0.9975,
+      "step": 5519
+    },
+    {
+      "epoch": 0.5805408389971999,
+      "grad_norm": 2.769410328624913,
+      "learning_rate": 1.9079478918966595e-06,
+      "loss": 0.9609,
+      "step": 5520
+    },
+    {
+      "epoch": 0.5806460094390472,
+      "grad_norm": 2.3183308581320694,
+      "learning_rate": 1.9071368532721734e-06,
+      "loss": 1.0247,
+      "step": 5521
+    },
+    {
+      "epoch": 0.5807511798808944,
+      "grad_norm": 3.0996340248511767,
+      "learning_rate": 1.9063258807565338e-06,
+      "loss": 0.9765,
+      "step": 5522
+    },
+    {
+      "epoch": 0.5808563503227417,
+      "grad_norm": 2.6197973686870553,
+      "learning_rate": 1.9055149744401693e-06,
+      "loss": 0.9358,
+      "step": 5523
+    },
+    {
+      "epoch": 0.5809615207645891,
+      "grad_norm": 2.241372830499171,
+      "learning_rate": 1.9047041344135045e-06,
+      "loss": 0.9821,
+      "step": 5524
+    },
+    {
+      "epoch": 0.5810666912064364,
+      "grad_norm": 2.9668398883596905,
+      "learning_rate": 1.903893360766953e-06,
+      "loss": 1.0332,
+      "step": 5525
+    },
+    {
+      "epoch": 0.5811718616482837,
+      "grad_norm": 2.2000463719499064,
+      "learning_rate": 1.9030826535909225e-06,
+      "loss": 0.9604,
+      "step": 5526
+    },
+    {
+      "epoch": 0.581277032090131,
+      "grad_norm": 2.212791966762905,
+      "learning_rate": 1.9022720129758132e-06,
+      "loss": 0.9615,
+      "step": 5527
+    },
+    {
+      "epoch": 0.5813822025319784,
+      "grad_norm": 2.2976157836326805,
+      "learning_rate": 1.9014614390120174e-06,
+      "loss": 0.9992,
+      "step": 5528
+    },
+    {
+      "epoch": 0.5814873729738257,
+      "grad_norm": 3.038579357250074,
+      "learning_rate": 1.9006509317899207e-06,
+      "loss": 1.0264,
+      "step": 5529
+    },
+    {
+      "epoch": 0.581592543415673,
+      "grad_norm": 2.8863407093044735,
+      "learning_rate": 1.8998404913999016e-06,
+      "loss": 0.9813,
+      "step": 5530
+    },
+    {
+      "epoch": 0.5816977138575203,
+      "grad_norm": 2.8846661345060953,
+      "learning_rate": 1.8990301179323298e-06,
+      "loss": 0.9626,
+      "step": 5531
+    },
+    {
+      "epoch": 0.5818028842993677,
+      "grad_norm": 2.4582156857408557,
+      "learning_rate": 1.8982198114775683e-06,
+      "loss": 0.9984,
+      "step": 5532
+    },
+    {
+      "epoch": 0.581908054741215,
+      "grad_norm": 2.178455937450243,
+      "learning_rate": 1.8974095721259718e-06,
+      "loss": 0.9677,
+      "step": 5533
+    },
+    {
+      "epoch": 0.5820132251830623,
+      "grad_norm": 2.4539058224075823,
+      "learning_rate": 1.89659939996789e-06,
+      "loss": 0.947,
+      "step": 5534
+    },
+    {
+      "epoch": 0.5821183956249096,
+      "grad_norm": 2.9776623960749364,
+      "learning_rate": 1.8957892950936623e-06,
+      "loss": 1.0148,
+      "step": 5535
+    },
+    {
+      "epoch": 0.582223566066757,
+      "grad_norm": 3.114821126421381,
+      "learning_rate": 1.8949792575936222e-06,
+      "loss": 0.9682,
+      "step": 5536
+    },
+    {
+      "epoch": 0.5823287365086043,
+      "grad_norm": 1.7570932709447848,
+      "learning_rate": 1.8941692875580934e-06,
+      "loss": 0.9917,
+      "step": 5537
+    },
+    {
+      "epoch": 0.5824339069504516,
+      "grad_norm": 2.612838069227518,
+      "learning_rate": 1.8933593850773963e-06,
+      "loss": 1.0041,
+      "step": 5538
+    },
+    {
+      "epoch": 0.5825390773922989,
+      "grad_norm": 2.0944013087449957,
+      "learning_rate": 1.8925495502418407e-06,
+      "loss": 0.9842,
+      "step": 5539
+    },
+    {
+      "epoch": 0.5826442478341463,
+      "grad_norm": 1.8402090587110502,
+      "learning_rate": 1.8917397831417285e-06,
+      "loss": 0.9741,
+      "step": 5540
+    },
+    {
+      "epoch": 0.5827494182759936,
+      "grad_norm": 2.3971865882038634,
+      "learning_rate": 1.8909300838673562e-06,
+      "loss": 0.9897,
+      "step": 5541
+    },
+    {
+      "epoch": 0.5828545887178409,
+      "grad_norm": 2.7386768450510988,
+      "learning_rate": 1.89012045250901e-06,
+      "loss": 1.0536,
+      "step": 5542
+    },
+    {
+      "epoch": 0.5829597591596881,
+      "grad_norm": 2.4719585825244157,
+      "learning_rate": 1.889310889156972e-06,
+      "loss": 0.9932,
+      "step": 5543
+    },
+    {
+      "epoch": 0.5830649296015354,
+      "grad_norm": 1.9873200200685415,
+      "learning_rate": 1.8885013939015133e-06,
+      "loss": 1.0105,
+      "step": 5544
+    },
+    {
+      "epoch": 0.5831701000433828,
+      "grad_norm": 2.0027592555824563,
+      "learning_rate": 1.8876919668329002e-06,
+      "loss": 0.9932,
+      "step": 5545
+    },
+    {
+      "epoch": 0.5832752704852301,
+      "grad_norm": 2.9904414990454433,
+      "learning_rate": 1.8868826080413876e-06,
+      "loss": 1.0046,
+      "step": 5546
+    },
+    {
+      "epoch": 0.5833804409270774,
+      "grad_norm": 2.7824243964175452,
+      "learning_rate": 1.8860733176172286e-06,
+      "loss": 1.0286,
+      "step": 5547
+    },
+    {
+      "epoch": 0.5834856113689247,
+      "grad_norm": 2.1679565097639917,
+      "learning_rate": 1.8852640956506643e-06,
+      "loss": 0.9693,
+      "step": 5548
+    },
+    {
+      "epoch": 0.5835907818107721,
+      "grad_norm": 2.0987314485767423,
+      "learning_rate": 1.884454942231928e-06,
+      "loss": 0.9306,
+      "step": 5549
+    },
+    {
+      "epoch": 0.5836959522526194,
+      "grad_norm": 2.74265547993407,
+      "learning_rate": 1.8836458574512478e-06,
+      "loss": 0.9962,
+      "step": 5550
+    },
+    {
+      "epoch": 0.5838011226944667,
+      "grad_norm": 1.8785431998355895,
+      "learning_rate": 1.882836841398841e-06,
+      "loss": 0.9127,
+      "step": 5551
+    },
+    {
+      "epoch": 0.583906293136314,
+      "grad_norm": 2.4599246039003364,
+      "learning_rate": 1.882027894164922e-06,
+      "loss": 0.9913,
+      "step": 5552
+    },
+    {
+      "epoch": 0.5840114635781614,
+      "grad_norm": 2.296968019204936,
+      "learning_rate": 1.8812190158396931e-06,
+      "loss": 0.9596,
+      "step": 5553
+    },
+    {
+      "epoch": 0.5841166340200087,
+      "grad_norm": 2.65439428097566,
+      "learning_rate": 1.880410206513351e-06,
+      "loss": 1.025,
+      "step": 5554
+    },
+    {
+      "epoch": 0.584221804461856,
+      "grad_norm": 2.1109947099788995,
+      "learning_rate": 1.8796014662760842e-06,
+      "loss": 1.0029,
+      "step": 5555
+    },
+    {
+      "epoch": 0.5843269749037033,
+      "grad_norm": 2.7664051470715405,
+      "learning_rate": 1.8787927952180724e-06,
+      "loss": 0.971,
+      "step": 5556
+    },
+    {
+      "epoch": 0.5844321453455507,
+      "grad_norm": 2.202625098363754,
+      "learning_rate": 1.8779841934294904e-06,
+      "loss": 0.9731,
+      "step": 5557
+    },
+    {
+      "epoch": 0.584537315787398,
+      "grad_norm": 2.2262523402556313,
+      "learning_rate": 1.8771756610005028e-06,
+      "loss": 0.9645,
+      "step": 5558
+    },
+    {
+      "epoch": 0.5846424862292453,
+      "grad_norm": 2.203570080279773,
+      "learning_rate": 1.8763671980212673e-06,
+      "loss": 0.9762,
+      "step": 5559
+    },
+    {
+      "epoch": 0.5847476566710926,
+      "grad_norm": 3.406979001707062,
+      "learning_rate": 1.8755588045819325e-06,
+      "loss": 0.988,
+      "step": 5560
+    },
+    {
+      "epoch": 0.58485282711294,
+      "grad_norm": 2.3316323807570685,
+      "learning_rate": 1.8747504807726433e-06,
+      "loss": 0.9723,
+      "step": 5561
+    },
+    {
+      "epoch": 0.5849579975547873,
+      "grad_norm": 2.5275255811504995,
+      "learning_rate": 1.8739422266835321e-06,
+      "loss": 1.0129,
+      "step": 5562
+    },
+    {
+      "epoch": 0.5850631679966345,
+      "grad_norm": 2.700141617045467,
+      "learning_rate": 1.8731340424047265e-06,
+      "loss": 0.9889,
+      "step": 5563
+    },
+    {
+      "epoch": 0.5851683384384818,
+      "grad_norm": 2.367236088955419,
+      "learning_rate": 1.8723259280263451e-06,
+      "loss": 0.9935,
+      "step": 5564
+    },
+    {
+      "epoch": 0.5852735088803291,
+      "grad_norm": 2.0726305566915997,
+      "learning_rate": 1.871517883638497e-06,
+      "loss": 0.9489,
+      "step": 5565
+    },
+    {
+      "epoch": 0.5853786793221765,
+      "grad_norm": 2.140334180895899,
+      "learning_rate": 1.8707099093312883e-06,
+      "loss": 0.9789,
+      "step": 5566
+    },
+    {
+      "epoch": 0.5854838497640238,
+      "grad_norm": 2.3801510907541035,
+      "learning_rate": 1.869902005194813e-06,
+      "loss": 0.9661,
+      "step": 5567
+    },
+    {
+      "epoch": 0.5855890202058711,
+      "grad_norm": 1.94348395585535,
+      "learning_rate": 1.869094171319159e-06,
+      "loss": 0.9675,
+      "step": 5568
+    },
+    {
+      "epoch": 0.5856941906477184,
+      "grad_norm": 1.8197908570293184,
+      "learning_rate": 1.8682864077944058e-06,
+      "loss": 0.9888,
+      "step": 5569
+    },
+    {
+      "epoch": 0.5857993610895658,
+      "grad_norm": 2.3503761909453478,
+      "learning_rate": 1.8674787147106255e-06,
+      "loss": 0.9775,
+      "step": 5570
+    },
+    {
+      "epoch": 0.5859045315314131,
+      "grad_norm": 2.424973046847033,
+      "learning_rate": 1.8666710921578823e-06,
+      "loss": 1.0109,
+      "step": 5571
+    },
+    {
+      "epoch": 0.5860097019732604,
+      "grad_norm": 3.1180462687524346,
+      "learning_rate": 1.865863540226232e-06,
+      "loss": 1.001,
+      "step": 5572
+    },
+    {
+      "epoch": 0.5861148724151077,
+      "grad_norm": 2.5532427885272946,
+      "learning_rate": 1.8650560590057235e-06,
+      "loss": 1.0227,
+      "step": 5573
+    },
+    {
+      "epoch": 0.5862200428569551,
+      "grad_norm": 2.2339931967688704,
+      "learning_rate": 1.8642486485863954e-06,
+      "loss": 1.0216,
+      "step": 5574
+    },
+    {
+      "epoch": 0.5863252132988024,
+      "grad_norm": 1.9277968092888864,
+      "learning_rate": 1.8634413090582831e-06,
+      "loss": 0.98,
+      "step": 5575
+    },
+    {
+      "epoch": 0.5864303837406497,
+      "grad_norm": 2.7536242032196583,
+      "learning_rate": 1.8626340405114097e-06,
+      "loss": 0.9137,
+      "step": 5576
+    },
+    {
+      "epoch": 0.586535554182497,
+      "grad_norm": 2.4161018794939157,
+      "learning_rate": 1.8618268430357916e-06,
+      "loss": 1.0122,
+      "step": 5577
+    },
+    {
+      "epoch": 0.5866407246243444,
+      "grad_norm": 2.243061062715193,
+      "learning_rate": 1.8610197167214383e-06,
+      "loss": 0.9863,
+      "step": 5578
+    },
+    {
+      "epoch": 0.5867458950661917,
+      "grad_norm": 2.371022287772955,
+      "learning_rate": 1.8602126616583499e-06,
+      "loss": 0.9608,
+      "step": 5579
+    },
+    {
+      "epoch": 0.586851065508039,
+      "grad_norm": 1.9392011011181172,
+      "learning_rate": 1.8594056779365202e-06,
+      "loss": 0.9605,
+      "step": 5580
+    },
+    {
+      "epoch": 0.5869562359498863,
+      "grad_norm": 2.688618985702904,
+      "learning_rate": 1.8585987656459337e-06,
+      "loss": 0.9798,
+      "step": 5581
+    },
+    {
+      "epoch": 0.5870614063917337,
+      "grad_norm": 2.0172735352124884,
+      "learning_rate": 1.8577919248765675e-06,
+      "loss": 0.9679,
+      "step": 5582
+    },
+    {
+      "epoch": 0.5871665768335809,
+      "grad_norm": 2.4283093844913197,
+      "learning_rate": 1.8569851557183894e-06,
+      "loss": 1.0153,
+      "step": 5583
+    },
+    {
+      "epoch": 0.5872717472754282,
+      "grad_norm": 2.4357436603887823,
+      "learning_rate": 1.8561784582613631e-06,
+      "loss": 0.9547,
+      "step": 5584
+    },
+    {
+      "epoch": 0.5873769177172755,
+      "grad_norm": 1.6496752603527884,
+      "learning_rate": 1.8553718325954395e-06,
+      "loss": 1.0061,
+      "step": 5585
+    },
+    {
+      "epoch": 0.5874820881591228,
+      "grad_norm": 2.3550989815363454,
+      "learning_rate": 1.8545652788105644e-06,
+      "loss": 1.0181,
+      "step": 5586
+    },
+    {
+      "epoch": 0.5875872586009702,
+      "grad_norm": 2.4247760719245592,
+      "learning_rate": 1.8537587969966746e-06,
+      "loss": 1.0417,
+      "step": 5587
+    },
+    {
+      "epoch": 0.5876924290428175,
+      "grad_norm": 2.7483529202080748,
+      "learning_rate": 1.852952387243698e-06,
+      "loss": 0.9984,
+      "step": 5588
+    },
+    {
+      "epoch": 0.5877975994846648,
+      "grad_norm": 2.3988717114809655,
+      "learning_rate": 1.8521460496415577e-06,
+      "loss": 1.021,
+      "step": 5589
+    },
+    {
+      "epoch": 0.5879027699265121,
+      "grad_norm": 3.4921295132942594,
+      "learning_rate": 1.8513397842801655e-06,
+      "loss": 0.9935,
+      "step": 5590
+    },
+    {
+      "epoch": 0.5880079403683595,
+      "grad_norm": 2.9336016166705066,
+      "learning_rate": 1.8505335912494265e-06,
+      "loss": 0.9968,
+      "step": 5591
+    },
+    {
+      "epoch": 0.5881131108102068,
+      "grad_norm": 2.8340234509298563,
+      "learning_rate": 1.8497274706392369e-06,
+      "loss": 1.0128,
+      "step": 5592
+    },
+    {
+      "epoch": 0.5882182812520541,
+      "grad_norm": 3.1961181107327,
+      "learning_rate": 1.848921422539486e-06,
+      "loss": 1.022,
+      "step": 5593
+    },
+    {
+      "epoch": 0.5883234516939014,
+      "grad_norm": 1.9909100410145897,
+      "learning_rate": 1.8481154470400545e-06,
+      "loss": 0.9937,
+      "step": 5594
+    },
+    {
+      "epoch": 0.5884286221357488,
+      "grad_norm": 3.066382990655515,
+      "learning_rate": 1.8473095442308145e-06,
+      "loss": 0.9469,
+      "step": 5595
+    },
+    {
+      "epoch": 0.5885337925775961,
+      "grad_norm": 2.6179963095077716,
+      "learning_rate": 1.8465037142016306e-06,
+      "loss": 0.9608,
+      "step": 5596
+    },
+    {
+      "epoch": 0.5886389630194434,
+      "grad_norm": 2.3037057873169524,
+      "learning_rate": 1.8456979570423583e-06,
+      "loss": 1.0054,
+      "step": 5597
+    },
+    {
+      "epoch": 0.5887441334612907,
+      "grad_norm": 2.2460590369251268,
+      "learning_rate": 1.8448922728428474e-06,
+      "loss": 0.9827,
+      "step": 5598
+    },
+    {
+      "epoch": 0.588849303903138,
+      "grad_norm": 2.7120869322911623,
+      "learning_rate": 1.844086661692937e-06,
+      "loss": 0.9871,
+      "step": 5599
+    },
+    {
+      "epoch": 0.5889544743449854,
+      "grad_norm": 3.41715465139414,
+      "learning_rate": 1.84328112368246e-06,
+      "loss": 0.9804,
+      "step": 5600
+    },
+    {
+      "epoch": 0.5890596447868327,
+      "grad_norm": 3.3908096687842604,
+      "learning_rate": 1.8424756589012385e-06,
+      "loss": 1.0025,
+      "step": 5601
+    },
+    {
+      "epoch": 0.58916481522868,
+      "grad_norm": 2.0383431427479506,
+      "learning_rate": 1.841670267439088e-06,
+      "loss": 0.9688,
+      "step": 5602
+    },
+    {
+      "epoch": 0.5892699856705274,
+      "grad_norm": 2.6775576431658092,
+      "learning_rate": 1.8408649493858176e-06,
+      "loss": 0.966,
+      "step": 5603
+    },
+    {
+      "epoch": 0.5893751561123746,
+      "grad_norm": 2.571976853064117,
+      "learning_rate": 1.8400597048312257e-06,
+      "loss": 1.0121,
+      "step": 5604
+    },
+    {
+      "epoch": 0.5894803265542219,
+      "grad_norm": 2.4059753404598236,
+      "learning_rate": 1.8392545338651036e-06,
+      "loss": 0.9888,
+      "step": 5605
+    },
+    {
+      "epoch": 0.5895854969960692,
+      "grad_norm": 2.069543872655015,
+      "learning_rate": 1.8384494365772333e-06,
+      "loss": 0.9747,
+      "step": 5606
+    },
+    {
+      "epoch": 0.5896906674379165,
+      "grad_norm": 2.30354396158409,
+      "learning_rate": 1.8376444130573901e-06,
+      "loss": 0.9904,
+      "step": 5607
+    },
+    {
+      "epoch": 0.5897958378797639,
+      "grad_norm": 2.2282342710722456,
+      "learning_rate": 1.8368394633953402e-06,
+      "loss": 0.9951,
+      "step": 5608
+    },
+    {
+      "epoch": 0.5899010083216112,
+      "grad_norm": 2.523786425607499,
+      "learning_rate": 1.8360345876808422e-06,
+      "loss": 1.0192,
+      "step": 5609
+    },
+    {
+      "epoch": 0.5900061787634585,
+      "grad_norm": 2.435738401461322,
+      "learning_rate": 1.835229786003645e-06,
+      "loss": 0.9807,
+      "step": 5610
+    },
+    {
+      "epoch": 0.5901113492053058,
+      "grad_norm": 2.4101222248980094,
+      "learning_rate": 1.83442505845349e-06,
+      "loss": 0.9915,
+      "step": 5611
+    },
+    {
+      "epoch": 0.5902165196471532,
+      "grad_norm": 1.8954669177419814,
+      "learning_rate": 1.8336204051201124e-06,
+      "loss": 0.9815,
+      "step": 5612
+    },
+    {
+      "epoch": 0.5903216900890005,
+      "grad_norm": 2.367357060810497,
+      "learning_rate": 1.832815826093236e-06,
+      "loss": 1.0251,
+      "step": 5613
+    },
+    {
+      "epoch": 0.5904268605308478,
+      "grad_norm": 2.4881726831979085,
+      "learning_rate": 1.8320113214625783e-06,
+      "loss": 0.9472,
+      "step": 5614
+    },
+    {
+      "epoch": 0.5905320309726951,
+      "grad_norm": 2.7845683502036715,
+      "learning_rate": 1.8312068913178466e-06,
+      "loss": 0.9873,
+      "step": 5615
+    },
+    {
+      "epoch": 0.5906372014145425,
+      "grad_norm": 1.968736042696628,
+      "learning_rate": 1.8304025357487427e-06,
+      "loss": 0.9252,
+      "step": 5616
+    },
+    {
+      "epoch": 0.5907423718563898,
+      "grad_norm": 2.8097759990413094,
+      "learning_rate": 1.829598254844957e-06,
+      "loss": 0.9349,
+      "step": 5617
+    },
+    {
+      "epoch": 0.5908475422982371,
+      "grad_norm": 2.375269564916978,
+      "learning_rate": 1.8287940486961744e-06,
+      "loss": 1.0258,
+      "step": 5618
+    },
+    {
+      "epoch": 0.5909527127400844,
+      "grad_norm": 2.3758359793157484,
+      "learning_rate": 1.8279899173920692e-06,
+      "loss": 0.9658,
+      "step": 5619
+    },
+    {
+      "epoch": 0.5910578831819318,
+      "grad_norm": 2.9877292184232345,
+      "learning_rate": 1.827185861022308e-06,
+      "loss": 1.0124,
+      "step": 5620
+    },
+    {
+      "epoch": 0.5911630536237791,
+      "grad_norm": 1.9064000817314557,
+      "learning_rate": 1.8263818796765506e-06,
+      "loss": 0.9492,
+      "step": 5621
+    },
+    {
+      "epoch": 0.5912682240656264,
+      "grad_norm": 2.4552260061481768,
+      "learning_rate": 1.8255779734444462e-06,
+      "loss": 0.9983,
+      "step": 5622
+    },
+    {
+      "epoch": 0.5913733945074737,
+      "grad_norm": 2.652810955909284,
+      "learning_rate": 1.8247741424156373e-06,
+      "loss": 1.0046,
+      "step": 5623
+    },
+    {
+      "epoch": 0.5914785649493209,
+      "grad_norm": 2.4957564210840797,
+      "learning_rate": 1.8239703866797553e-06,
+      "loss": 1.0137,
+      "step": 5624
+    },
+    {
+      "epoch": 0.5915837353911683,
+      "grad_norm": 2.776375551754447,
+      "learning_rate": 1.8231667063264282e-06,
+      "loss": 1.016,
+      "step": 5625
+    },
+    {
+      "epoch": 0.5916889058330156,
+      "grad_norm": 2.2349936262589067,
+      "learning_rate": 1.822363101445271e-06,
+      "loss": 0.99,
+      "step": 5626
+    },
+    {
+      "epoch": 0.5917940762748629,
+      "grad_norm": 2.136732395843005,
+      "learning_rate": 1.8215595721258921e-06,
+      "loss": 0.9894,
+      "step": 5627
+    },
+    {
+      "epoch": 0.5918992467167102,
+      "grad_norm": 2.695017472417158,
+      "learning_rate": 1.8207561184578915e-06,
+      "loss": 1.0002,
+      "step": 5628
+    },
+    {
+      "epoch": 0.5920044171585576,
+      "grad_norm": 2.69308063651352,
+      "learning_rate": 1.8199527405308593e-06,
+      "loss": 1.0041,
+      "step": 5629
+    },
+    {
+      "epoch": 0.5921095876004049,
+      "grad_norm": 2.534730264118862,
+      "learning_rate": 1.81914943843438e-06,
+      "loss": 1.0193,
+      "step": 5630
+    },
+    {
+      "epoch": 0.5922147580422522,
+      "grad_norm": 2.4582659603999564,
+      "learning_rate": 1.818346212258027e-06,
+      "loss": 1.0088,
+      "step": 5631
+    },
+    {
+      "epoch": 0.5923199284840995,
+      "grad_norm": 2.495714807866576,
+      "learning_rate": 1.817543062091367e-06,
+      "loss": 0.9929,
+      "step": 5632
+    },
+    {
+      "epoch": 0.5924250989259469,
+      "grad_norm": 2.58654760819409,
+      "learning_rate": 1.8167399880239572e-06,
+      "loss": 0.995,
+      "step": 5633
+    },
+    {
+      "epoch": 0.5925302693677942,
+      "grad_norm": 2.876672103469995,
+      "learning_rate": 1.815936990145345e-06,
+      "loss": 0.954,
+      "step": 5634
+    },
+    {
+      "epoch": 0.5926354398096415,
+      "grad_norm": 2.384145352263863,
+      "learning_rate": 1.8151340685450745e-06,
+      "loss": 0.9896,
+      "step": 5635
+    },
+    {
+      "epoch": 0.5927406102514888,
+      "grad_norm": 3.345693108309314,
+      "learning_rate": 1.8143312233126748e-06,
+      "loss": 1.0084,
+      "step": 5636
+    },
+    {
+      "epoch": 0.5928457806933362,
+      "grad_norm": 3.5913950173149782,
+      "learning_rate": 1.81352845453767e-06,
+      "loss": 0.983,
+      "step": 5637
+    },
+    {
+      "epoch": 0.5929509511351835,
+      "grad_norm": 2.3914332678489827,
+      "learning_rate": 1.8127257623095743e-06,
+      "loss": 1.002,
+      "step": 5638
+    },
+    {
+      "epoch": 0.5930561215770308,
+      "grad_norm": 3.185510171534475,
+      "learning_rate": 1.811923146717896e-06,
+      "loss": 0.9976,
+      "step": 5639
+    },
+    {
+      "epoch": 0.5931612920188781,
+      "grad_norm": 2.0993299580165767,
+      "learning_rate": 1.811120607852132e-06,
+      "loss": 0.9655,
+      "step": 5640
+    },
+    {
+      "epoch": 0.5932664624607255,
+      "grad_norm": 2.8049029671520116,
+      "learning_rate": 1.8103181458017719e-06,
+      "loss": 0.9852,
+      "step": 5641
+    },
+    {
+      "epoch": 0.5933716329025728,
+      "grad_norm": 2.0996772552523675,
+      "learning_rate": 1.8095157606562957e-06,
+      "loss": 0.9822,
+      "step": 5642
+    },
+    {
+      "epoch": 0.5934768033444201,
+      "grad_norm": 2.3389036391566878,
+      "learning_rate": 1.8087134525051762e-06,
+      "loss": 1.0309,
+      "step": 5643
+    },
+    {
+      "epoch": 0.5935819737862674,
+      "grad_norm": 2.3139277809580077,
+      "learning_rate": 1.8079112214378769e-06,
+      "loss": 0.9498,
+      "step": 5644
+    },
+    {
+      "epoch": 0.5936871442281146,
+      "grad_norm": 2.6546675909969113,
+      "learning_rate": 1.807109067543853e-06,
+      "loss": 1.0182,
+      "step": 5645
+    },
+    {
+      "epoch": 0.593792314669962,
+      "grad_norm": 2.591901764829732,
+      "learning_rate": 1.8063069909125502e-06,
+      "loss": 1.0092,
+      "step": 5646
+    },
+    {
+      "epoch": 0.5938974851118093,
+      "grad_norm": 1.5887306890493018,
+      "learning_rate": 1.805504991633406e-06,
+      "loss": 0.9922,
+      "step": 5647
+    },
+    {
+      "epoch": 0.5940026555536566,
+      "grad_norm": 2.160624868604629,
+      "learning_rate": 1.8047030697958513e-06,
+      "loss": 0.9504,
+      "step": 5648
+    },
+    {
+      "epoch": 0.5941078259955039,
+      "grad_norm": 2.0657716812683518,
+      "learning_rate": 1.8039012254893054e-06,
+      "loss": 0.993,
+      "step": 5649
+    },
+    {
+      "epoch": 0.5942129964373513,
+      "grad_norm": 2.827332461738818,
+      "learning_rate": 1.8030994588031804e-06,
+      "loss": 1.009,
+      "step": 5650
+    },
+    {
+      "epoch": 0.5943181668791986,
+      "grad_norm": 2.5335501756770125,
+      "learning_rate": 1.80229776982688e-06,
+      "loss": 0.9963,
+      "step": 5651
+    },
+    {
+      "epoch": 0.5944233373210459,
+      "grad_norm": 2.394767186447968,
+      "learning_rate": 1.801496158649797e-06,
+      "loss": 1.0434,
+      "step": 5652
+    },
+    {
+      "epoch": 0.5945285077628932,
+      "grad_norm": 3.032987653965299,
+      "learning_rate": 1.800694625361319e-06,
+      "loss": 1.0317,
+      "step": 5653
+    },
+    {
+      "epoch": 0.5946336782047406,
+      "grad_norm": 2.3486440028502384,
+      "learning_rate": 1.799893170050823e-06,
+      "loss": 0.9893,
+      "step": 5654
+    },
+    {
+      "epoch": 0.5947388486465879,
+      "grad_norm": 2.014009694496243,
+      "learning_rate": 1.7990917928076768e-06,
+      "loss": 0.9973,
+      "step": 5655
+    },
+    {
+      "epoch": 0.5948440190884352,
+      "grad_norm": 1.9133594346364118,
+      "learning_rate": 1.7982904937212409e-06,
+      "loss": 0.9813,
+      "step": 5656
+    },
+    {
+      "epoch": 0.5949491895302825,
+      "grad_norm": 3.1233038474070423,
+      "learning_rate": 1.7974892728808653e-06,
+      "loss": 0.9512,
+      "step": 5657
+    },
+    {
+      "epoch": 0.5950543599721299,
+      "grad_norm": 2.19478691298938,
+      "learning_rate": 1.7966881303758938e-06,
+      "loss": 0.9901,
+      "step": 5658
+    },
+    {
+      "epoch": 0.5951595304139772,
+      "grad_norm": 2.1551749316228093,
+      "learning_rate": 1.795887066295659e-06,
+      "loss": 1.0007,
+      "step": 5659
+    },
+    {
+      "epoch": 0.5952647008558245,
+      "grad_norm": 2.5085227649724957,
+      "learning_rate": 1.7950860807294863e-06,
+      "loss": 1.0337,
+      "step": 5660
+    },
+    {
+      "epoch": 0.5953698712976718,
+      "grad_norm": 2.385692972652722,
+      "learning_rate": 1.7942851737666906e-06,
+      "loss": 0.9654,
+      "step": 5661
+    },
+    {
+      "epoch": 0.5954750417395192,
+      "grad_norm": 2.221445423122614,
+      "learning_rate": 1.7934843454965808e-06,
+      "loss": 0.9642,
+      "step": 5662
+    },
+    {
+      "epoch": 0.5955802121813665,
+      "grad_norm": 2.745661167238812,
+      "learning_rate": 1.7926835960084555e-06,
+      "loss": 1.0256,
+      "step": 5663
+    },
+    {
+      "epoch": 0.5956853826232138,
+      "grad_norm": 2.227270973619197,
+      "learning_rate": 1.7918829253916032e-06,
+      "loss": 1.0145,
+      "step": 5664
+    },
+    {
+      "epoch": 0.595790553065061,
+      "grad_norm": 1.9107730031675045,
+      "learning_rate": 1.7910823337353062e-06,
+      "loss": 0.9939,
+      "step": 5665
+    },
+    {
+      "epoch": 0.5958957235069083,
+      "grad_norm": 2.273694926717976,
+      "learning_rate": 1.790281821128835e-06,
+      "loss": 1.0495,
+      "step": 5666
+    },
+    {
+      "epoch": 0.5960008939487557,
+      "grad_norm": 2.3672345502650383,
+      "learning_rate": 1.7894813876614547e-06,
+      "loss": 0.9918,
+      "step": 5667
+    },
+    {
+      "epoch": 0.596106064390603,
+      "grad_norm": 2.338013846800265,
+      "learning_rate": 1.7886810334224192e-06,
+      "loss": 1.0049,
+      "step": 5668
+    },
+    {
+      "epoch": 0.5962112348324503,
+      "grad_norm": 2.2878641074827413,
+      "learning_rate": 1.7878807585009744e-06,
+      "loss": 0.9282,
+      "step": 5669
+    },
+    {
+      "epoch": 0.5963164052742976,
+      "grad_norm": 2.447388144552986,
+      "learning_rate": 1.7870805629863563e-06,
+      "loss": 0.9712,
+      "step": 5670
+    },
+    {
+      "epoch": 0.596421575716145,
+      "grad_norm": 2.4810306355833025,
+      "learning_rate": 1.7862804469677942e-06,
+      "loss": 0.9651,
+      "step": 5671
+    },
+    {
+      "epoch": 0.5965267461579923,
+      "grad_norm": 2.049575677232311,
+      "learning_rate": 1.7854804105345064e-06,
+      "loss": 0.971,
+      "step": 5672
+    },
+    {
+      "epoch": 0.5966319165998396,
+      "grad_norm": 2.715949255809658,
+      "learning_rate": 1.7846804537757034e-06,
+      "loss": 0.9999,
+      "step": 5673
+    },
+    {
+      "epoch": 0.5967370870416869,
+      "grad_norm": 2.384190947460149,
+      "learning_rate": 1.7838805767805866e-06,
+      "loss": 0.9835,
+      "step": 5674
+    },
+    {
+      "epoch": 0.5968422574835343,
+      "grad_norm": 1.9117408819820847,
+      "learning_rate": 1.7830807796383475e-06,
+      "loss": 0.969,
+      "step": 5675
+    },
+    {
+      "epoch": 0.5969474279253816,
+      "grad_norm": 2.26677544024507,
+      "learning_rate": 1.782281062438172e-06,
+      "loss": 0.9658,
+      "step": 5676
+    },
+    {
+      "epoch": 0.5970525983672289,
+      "grad_norm": 2.0051996758671806,
+      "learning_rate": 1.7814814252692333e-06,
+      "loss": 0.9929,
+      "step": 5677
+    },
+    {
+      "epoch": 0.5971577688090762,
+      "grad_norm": 2.6826187119200426,
+      "learning_rate": 1.7806818682206972e-06,
+      "loss": 1.0851,
+      "step": 5678
+    },
+    {
+      "epoch": 0.5972629392509236,
+      "grad_norm": 2.619403217077916,
+      "learning_rate": 1.779882391381721e-06,
+      "loss": 0.9448,
+      "step": 5679
+    },
+    {
+      "epoch": 0.5973681096927709,
+      "grad_norm": 2.5065928671207316,
+      "learning_rate": 1.7790829948414512e-06,
+      "loss": 1.0461,
+      "step": 5680
+    },
+    {
+      "epoch": 0.5974732801346182,
+      "grad_norm": 2.450270322938859,
+      "learning_rate": 1.778283678689029e-06,
+      "loss": 0.9973,
+      "step": 5681
+    },
+    {
+      "epoch": 0.5975784505764655,
+      "grad_norm": 2.519020526501517,
+      "learning_rate": 1.7774844430135823e-06,
+      "loss": 0.9834,
+      "step": 5682
+    },
+    {
+      "epoch": 0.5976836210183128,
+      "grad_norm": 1.6420026627989386,
+      "learning_rate": 1.7766852879042335e-06,
+      "loss": 0.9535,
+      "step": 5683
+    },
+    {
+      "epoch": 0.5977887914601602,
+      "grad_norm": 2.0627308895943623,
+      "learning_rate": 1.7758862134500926e-06,
+      "loss": 0.9999,
+      "step": 5684
+    },
+    {
+      "epoch": 0.5978939619020074,
+      "grad_norm": 2.6591210105862904,
+      "learning_rate": 1.7750872197402652e-06,
+      "loss": 0.9714,
+      "step": 5685
+    },
+    {
+      "epoch": 0.5979991323438547,
+      "grad_norm": 2.4500227653148294,
+      "learning_rate": 1.7742883068638447e-06,
+      "loss": 0.9585,
+      "step": 5686
+    },
+    {
+      "epoch": 0.598104302785702,
+      "grad_norm": 2.740294615631952,
+      "learning_rate": 1.773489474909915e-06,
+      "loss": 1.014,
+      "step": 5687
+    },
+    {
+      "epoch": 0.5982094732275494,
+      "grad_norm": 3.1109859813510115,
+      "learning_rate": 1.7726907239675523e-06,
+      "loss": 0.9821,
+      "step": 5688
+    },
+    {
+      "epoch": 0.5983146436693967,
+      "grad_norm": 2.1772995997746176,
+      "learning_rate": 1.771892054125823e-06,
+      "loss": 1.0209,
+      "step": 5689
+    },
+    {
+      "epoch": 0.598419814111244,
+      "grad_norm": 2.8946158732426333,
+      "learning_rate": 1.7710934654737868e-06,
+      "loss": 1.0099,
+      "step": 5690
+    },
+    {
+      "epoch": 0.5985249845530913,
+      "grad_norm": 2.1027519425308787,
+      "learning_rate": 1.7702949581004917e-06,
+      "loss": 1.0164,
+      "step": 5691
+    },
+    {
+      "epoch": 0.5986301549949387,
+      "grad_norm": 2.931727859528439,
+      "learning_rate": 1.769496532094977e-06,
+      "loss": 0.9897,
+      "step": 5692
+    },
+    {
+      "epoch": 0.598735325436786,
+      "grad_norm": 2.3348996193471274,
+      "learning_rate": 1.7686981875462733e-06,
+      "loss": 1.0096,
+      "step": 5693
+    },
+    {
+      "epoch": 0.5988404958786333,
+      "grad_norm": 2.4906181074969207,
+      "learning_rate": 1.7678999245434036e-06,
+      "loss": 0.985,
+      "step": 5694
+    },
+    {
+      "epoch": 0.5989456663204806,
+      "grad_norm": 1.712480678425093,
+      "learning_rate": 1.7671017431753789e-06,
+      "loss": 0.9953,
+      "step": 5695
+    },
+    {
+      "epoch": 0.599050836762328,
+      "grad_norm": 1.9969216227932245,
+      "learning_rate": 1.7663036435312037e-06,
+      "loss": 1.0244,
+      "step": 5696
+    },
+    {
+      "epoch": 0.5991560072041753,
+      "grad_norm": 2.663976765479451,
+      "learning_rate": 1.7655056256998712e-06,
+      "loss": 1.0014,
+      "step": 5697
+    },
+    {
+      "epoch": 0.5992611776460226,
+      "grad_norm": 2.1061164190156885,
+      "learning_rate": 1.7647076897703664e-06,
+      "loss": 0.9526,
+      "step": 5698
+    },
+    {
+      "epoch": 0.5993663480878699,
+      "grad_norm": 2.164328415014977,
+      "learning_rate": 1.7639098358316673e-06,
+      "loss": 0.9751,
+      "step": 5699
+    },
+    {
+      "epoch": 0.5994715185297173,
+      "grad_norm": 2.926709142359428,
+      "learning_rate": 1.7631120639727396e-06,
+      "loss": 0.993,
+      "step": 5700
+    },
+    {
+      "epoch": 0.5995766889715646,
+      "grad_norm": 2.5586116221192676,
+      "learning_rate": 1.762314374282541e-06,
+      "loss": 0.9988,
+      "step": 5701
+    },
+    {
+      "epoch": 0.5996818594134119,
+      "grad_norm": 1.9550872865008981,
+      "learning_rate": 1.7615167668500205e-06,
+      "loss": 0.9676,
+      "step": 5702
+    },
+    {
+      "epoch": 0.5997870298552592,
+      "grad_norm": 2.7161983180737646,
+      "learning_rate": 1.7607192417641164e-06,
+      "loss": 0.9666,
+      "step": 5703
+    },
+    {
+      "epoch": 0.5998922002971065,
+      "grad_norm": 2.6847054400836576,
+      "learning_rate": 1.7599217991137604e-06,
+      "loss": 1.0108,
+      "step": 5704
+    },
+    {
+      "epoch": 0.5999973707389539,
+      "grad_norm": 2.3979400888890416,
+      "learning_rate": 1.759124438987873e-06,
+      "loss": 0.9661,
+      "step": 5705
+    },
+    {
+      "epoch": 0.6001025411808011,
+      "grad_norm": 2.2431642149168467,
+      "learning_rate": 1.758327161475366e-06,
+      "loss": 0.982,
+      "step": 5706
+    },
+    {
+      "epoch": 0.6002077116226484,
+      "grad_norm": 2.177608584163875,
+      "learning_rate": 1.7575299666651413e-06,
+      "loss": 0.9424,
+      "step": 5707
+    },
+    {
+      "epoch": 0.6003128820644957,
+      "grad_norm": 1.7290319576886468,
+      "learning_rate": 1.7567328546460939e-06,
+      "loss": 0.944,
+      "step": 5708
+    },
+    {
+      "epoch": 0.6004180525063431,
+      "grad_norm": 3.080894919454507,
+      "learning_rate": 1.7559358255071068e-06,
+      "loss": 1.0085,
+      "step": 5709
+    },
+    {
+      "epoch": 0.6005232229481904,
+      "grad_norm": 2.060775484827086,
+      "learning_rate": 1.755138879337055e-06,
+      "loss": 0.9477,
+      "step": 5710
+    },
+    {
+      "epoch": 0.6006283933900377,
+      "grad_norm": 2.6684216105836183,
+      "learning_rate": 1.754342016224805e-06,
+      "loss": 1.0047,
+      "step": 5711
+    },
+    {
+      "epoch": 0.600733563831885,
+      "grad_norm": 2.1366313634883154,
+      "learning_rate": 1.7535452362592116e-06,
+      "loss": 0.9851,
+      "step": 5712
+    },
+    {
+      "epoch": 0.6008387342737324,
+      "grad_norm": 2.3649889418446284,
+      "learning_rate": 1.7527485395291234e-06,
+      "loss": 1.0063,
+      "step": 5713
+    },
+    {
+      "epoch": 0.6009439047155797,
+      "grad_norm": 1.9751341643467382,
+      "learning_rate": 1.7519519261233786e-06,
+      "loss": 0.9699,
+      "step": 5714
+    },
+    {
+      "epoch": 0.601049075157427,
+      "grad_norm": 3.1232686070504982,
+      "learning_rate": 1.7511553961308048e-06,
+      "loss": 0.9794,
+      "step": 5715
+    },
+    {
+      "epoch": 0.6011542455992743,
+      "grad_norm": 1.7374660242199027,
+      "learning_rate": 1.750358949640221e-06,
+      "loss": 0.9681,
+      "step": 5716
+    },
+    {
+      "epoch": 0.6012594160411217,
+      "grad_norm": 2.5925063706238216,
+      "learning_rate": 1.749562586740438e-06,
+      "loss": 1.0276,
+      "step": 5717
+    },
+    {
+      "epoch": 0.601364586482969,
+      "grad_norm": 2.2871091706939097,
+      "learning_rate": 1.7487663075202565e-06,
+      "loss": 0.9771,
+      "step": 5718
+    },
+    {
+      "epoch": 0.6014697569248163,
+      "grad_norm": 2.64298782172665,
+      "learning_rate": 1.7479701120684678e-06,
+      "loss": 1.0117,
+      "step": 5719
+    },
+    {
+      "epoch": 0.6015749273666636,
+      "grad_norm": 2.2690844026727097,
+      "learning_rate": 1.747174000473853e-06,
+      "loss": 0.9819,
+      "step": 5720
+    },
+    {
+      "epoch": 0.601680097808511,
+      "grad_norm": 3.5195991147155135,
+      "learning_rate": 1.7463779728251844e-06,
+      "loss": 1.0233,
+      "step": 5721
+    },
+    {
+      "epoch": 0.6017852682503583,
+      "grad_norm": 2.001707103161917,
+      "learning_rate": 1.745582029211228e-06,
+      "loss": 1.0275,
+      "step": 5722
+    },
+    {
+      "epoch": 0.6018904386922056,
+      "grad_norm": 2.008284524458236,
+      "learning_rate": 1.7447861697207352e-06,
+      "loss": 1.0409,
+      "step": 5723
+    },
+    {
+      "epoch": 0.6019956091340529,
+      "grad_norm": 2.814557277899713,
+      "learning_rate": 1.7439903944424513e-06,
+      "loss": 0.9048,
+      "step": 5724
+    },
+    {
+      "epoch": 0.6021007795759002,
+      "grad_norm": 2.115344313736399,
+      "learning_rate": 1.7431947034651111e-06,
+      "loss": 1.0137,
+      "step": 5725
+    },
+    {
+      "epoch": 0.6022059500177475,
+      "grad_norm": 1.708140537223153,
+      "learning_rate": 1.7423990968774397e-06,
+      "loss": 0.93,
+      "step": 5726
+    },
+    {
+      "epoch": 0.6023111204595948,
+      "grad_norm": 3.0743193234151365,
+      "learning_rate": 1.7416035747681554e-06,
+      "loss": 1.0175,
+      "step": 5727
+    },
+    {
+      "epoch": 0.6024162909014421,
+      "grad_norm": 2.109346929311083,
+      "learning_rate": 1.7408081372259633e-06,
+      "loss": 0.957,
+      "step": 5728
+    },
+    {
+      "epoch": 0.6025214613432894,
+      "grad_norm": 2.63470957119076,
+      "learning_rate": 1.740012784339562e-06,
+      "loss": 0.9973,
+      "step": 5729
+    },
+    {
+      "epoch": 0.6026266317851368,
+      "grad_norm": 2.789184435937574,
+      "learning_rate": 1.7392175161976384e-06,
+      "loss": 0.9654,
+      "step": 5730
+    },
+    {
+      "epoch": 0.6027318022269841,
+      "grad_norm": 2.891274776951523,
+      "learning_rate": 1.7384223328888724e-06,
+      "loss": 0.9714,
+      "step": 5731
+    },
+    {
+      "epoch": 0.6028369726688314,
+      "grad_norm": 2.3509105765329568,
+      "learning_rate": 1.7376272345019325e-06,
+      "loss": 1.0159,
+      "step": 5732
+    },
+    {
+      "epoch": 0.6029421431106787,
+      "grad_norm": 2.8871291599462374,
+      "learning_rate": 1.736832221125478e-06,
+      "loss": 0.9938,
+      "step": 5733
+    },
+    {
+      "epoch": 0.603047313552526,
+      "grad_norm": 2.5476087471404476,
+      "learning_rate": 1.7360372928481594e-06,
+      "loss": 1.0238,
+      "step": 5734
+    },
+    {
+      "epoch": 0.6031524839943734,
+      "grad_norm": 2.607771150913734,
+      "learning_rate": 1.7352424497586163e-06,
+      "loss": 0.9731,
+      "step": 5735
+    },
+    {
+      "epoch": 0.6032576544362207,
+      "grad_norm": 1.834225254634065,
+      "learning_rate": 1.7344476919454816e-06,
+      "loss": 0.9735,
+      "step": 5736
+    },
+    {
+      "epoch": 0.603362824878068,
+      "grad_norm": 2.1538425332658617,
+      "learning_rate": 1.7336530194973766e-06,
+      "loss": 0.9937,
+      "step": 5737
+    },
+    {
+      "epoch": 0.6034679953199154,
+      "grad_norm": 2.0723158686851604,
+      "learning_rate": 1.732858432502914e-06,
+      "loss": 0.9705,
+      "step": 5738
+    },
+    {
+      "epoch": 0.6035731657617627,
+      "grad_norm": 2.7600166097053385,
+      "learning_rate": 1.7320639310506935e-06,
+      "loss": 0.9406,
+      "step": 5739
+    },
+    {
+      "epoch": 0.60367833620361,
+      "grad_norm": 2.4094461693884144,
+      "learning_rate": 1.7312695152293112e-06,
+      "loss": 1.017,
+      "step": 5740
+    },
+    {
+      "epoch": 0.6037835066454573,
+      "grad_norm": 3.320433240431043,
+      "learning_rate": 1.73047518512735e-06,
+      "loss": 1.0132,
+      "step": 5741
+    },
+    {
+      "epoch": 0.6038886770873046,
+      "grad_norm": 2.3770891741620512,
+      "learning_rate": 1.729680940833383e-06,
+      "loss": 1.0349,
+      "step": 5742
+    },
+    {
+      "epoch": 0.603993847529152,
+      "grad_norm": 2.2580054578011373,
+      "learning_rate": 1.7288867824359757e-06,
+      "loss": 0.9242,
+      "step": 5743
+    },
+    {
+      "epoch": 0.6040990179709993,
+      "grad_norm": 2.280728334297776,
+      "learning_rate": 1.7280927100236816e-06,
+      "loss": 0.9499,
+      "step": 5744
+    },
+    {
+      "epoch": 0.6042041884128466,
+      "grad_norm": 3.1522624706614413,
+      "learning_rate": 1.7272987236850475e-06,
+      "loss": 1.0292,
+      "step": 5745
+    },
+    {
+      "epoch": 0.6043093588546938,
+      "grad_norm": 2.9642417340517553,
+      "learning_rate": 1.7265048235086078e-06,
+      "loss": 1.0452,
+      "step": 5746
+    },
+    {
+      "epoch": 0.6044145292965412,
+      "grad_norm": 2.6256509593397146,
+      "learning_rate": 1.7257110095828894e-06,
+      "loss": 0.9713,
+      "step": 5747
+    },
+    {
+      "epoch": 0.6045196997383885,
+      "grad_norm": 2.363877425800695,
+      "learning_rate": 1.7249172819964084e-06,
+      "loss": 0.9749,
+      "step": 5748
+    },
+    {
+      "epoch": 0.6046248701802358,
+      "grad_norm": 2.8124098873748755,
+      "learning_rate": 1.7241236408376702e-06,
+      "loss": 1.031,
+      "step": 5749
+    },
+    {
+      "epoch": 0.6047300406220831,
+      "grad_norm": 1.9881986505849607,
+      "learning_rate": 1.7233300861951743e-06,
+      "loss": 1.024,
+      "step": 5750
+    },
+    {
+      "epoch": 0.6048352110639305,
+      "grad_norm": 2.6635556816470602,
+      "learning_rate": 1.7225366181574072e-06,
+      "loss": 1.0287,
+      "step": 5751
+    },
+    {
+      "epoch": 0.6049403815057778,
+      "grad_norm": 2.277134066909746,
+      "learning_rate": 1.7217432368128468e-06,
+      "loss": 0.9973,
+      "step": 5752
+    },
+    {
+      "epoch": 0.6050455519476251,
+      "grad_norm": 2.3189335868813776,
+      "learning_rate": 1.7209499422499607e-06,
+      "loss": 1.0081,
+      "step": 5753
+    },
+    {
+      "epoch": 0.6051507223894724,
+      "grad_norm": 2.2043466967505334,
+      "learning_rate": 1.7201567345572084e-06,
+      "loss": 1.0241,
+      "step": 5754
+    },
+    {
+      "epoch": 0.6052558928313198,
+      "grad_norm": 2.1236697874477737,
+      "learning_rate": 1.7193636138230382e-06,
+      "loss": 0.9812,
+      "step": 5755
+    },
+    {
+      "epoch": 0.6053610632731671,
+      "grad_norm": 2.774793111808096,
+      "learning_rate": 1.7185705801358892e-06,
+      "loss": 0.9876,
+      "step": 5756
+    },
+    {
+      "epoch": 0.6054662337150144,
+      "grad_norm": 2.7424681309962864,
+      "learning_rate": 1.7177776335841912e-06,
+      "loss": 1.0057,
+      "step": 5757
+    },
+    {
+      "epoch": 0.6055714041568617,
+      "grad_norm": 1.6016238303844932,
+      "learning_rate": 1.7169847742563624e-06,
+      "loss": 0.9981,
+      "step": 5758
+    },
+    {
+      "epoch": 0.605676574598709,
+      "grad_norm": 2.5812078320001364,
+      "learning_rate": 1.7161920022408153e-06,
+      "loss": 0.9899,
+      "step": 5759
+    },
+    {
+      "epoch": 0.6057817450405564,
+      "grad_norm": 2.099760762558072,
+      "learning_rate": 1.7153993176259481e-06,
+      "loss": 0.9424,
+      "step": 5760
+    },
+    {
+      "epoch": 0.6058869154824037,
+      "grad_norm": 2.7673202803512353,
+      "learning_rate": 1.714606720500152e-06,
+      "loss": 0.9918,
+      "step": 5761
+    },
+    {
+      "epoch": 0.605992085924251,
+      "grad_norm": 2.0971157563339053,
+      "learning_rate": 1.7138142109518068e-06,
+      "loss": 1.0119,
+      "step": 5762
+    },
+    {
+      "epoch": 0.6060972563660983,
+      "grad_norm": 2.4244381472334124,
+      "learning_rate": 1.7130217890692857e-06,
+      "loss": 0.9751,
+      "step": 5763
+    },
+    {
+      "epoch": 0.6062024268079457,
+      "grad_norm": 1.875983236644177,
+      "learning_rate": 1.7122294549409486e-06,
+      "loss": 0.9974,
+      "step": 5764
+    },
+    {
+      "epoch": 0.606307597249793,
+      "grad_norm": 1.9342289332061562,
+      "learning_rate": 1.7114372086551466e-06,
+      "loss": 0.9757,
+      "step": 5765
+    },
+    {
+      "epoch": 0.6064127676916403,
+      "grad_norm": 2.6297959475374175,
+      "learning_rate": 1.7106450503002214e-06,
+      "loss": 0.9784,
+      "step": 5766
+    },
+    {
+      "epoch": 0.6065179381334875,
+      "grad_norm": 3.026075897619725,
+      "learning_rate": 1.709852979964505e-06,
+      "loss": 0.9993,
+      "step": 5767
+    },
+    {
+      "epoch": 0.6066231085753349,
+      "grad_norm": 2.099309327822422,
+      "learning_rate": 1.7090609977363198e-06,
+      "loss": 0.9941,
+      "step": 5768
+    },
+    {
+      "epoch": 0.6067282790171822,
+      "grad_norm": 2.839764446363711,
+      "learning_rate": 1.7082691037039772e-06,
+      "loss": 0.9789,
+      "step": 5769
+    },
+    {
+      "epoch": 0.6068334494590295,
+      "grad_norm": 2.6592502093600325,
+      "learning_rate": 1.7074772979557802e-06,
+      "loss": 1.0025,
+      "step": 5770
+    },
+    {
+      "epoch": 0.6069386199008768,
+      "grad_norm": 2.3396615859703527,
+      "learning_rate": 1.7066855805800203e-06,
+      "loss": 0.945,
+      "step": 5771
+    },
+    {
+      "epoch": 0.6070437903427242,
+      "grad_norm": 2.967177827861926,
+      "learning_rate": 1.7058939516649814e-06,
+      "loss": 0.9851,
+      "step": 5772
+    },
+    {
+      "epoch": 0.6071489607845715,
+      "grad_norm": 2.8406257980156466,
+      "learning_rate": 1.7051024112989365e-06,
+      "loss": 0.9895,
+      "step": 5773
+    },
+    {
+      "epoch": 0.6072541312264188,
+      "grad_norm": 2.539983854743361,
+      "learning_rate": 1.7043109595701472e-06,
+      "loss": 1.0268,
+      "step": 5774
+    },
+    {
+      "epoch": 0.6073593016682661,
+      "grad_norm": 2.991656526457313,
+      "learning_rate": 1.7035195965668669e-06,
+      "loss": 0.9855,
+      "step": 5775
+    },
+    {
+      "epoch": 0.6074644721101135,
+      "grad_norm": 1.9105665351289631,
+      "learning_rate": 1.7027283223773378e-06,
+      "loss": 0.9728,
+      "step": 5776
+    },
+    {
+      "epoch": 0.6075696425519608,
+      "grad_norm": 2.2526519815756303,
+      "learning_rate": 1.7019371370897953e-06,
+      "loss": 0.9993,
+      "step": 5777
+    },
+    {
+      "epoch": 0.6076748129938081,
+      "grad_norm": 1.9285557068670938,
+      "learning_rate": 1.7011460407924616e-06,
+      "loss": 0.9362,
+      "step": 5778
+    },
+    {
+      "epoch": 0.6077799834356554,
+      "grad_norm": 3.059051701475386,
+      "learning_rate": 1.70035503357355e-06,
+      "loss": 1.0401,
+      "step": 5779
+    },
+    {
+      "epoch": 0.6078851538775027,
+      "grad_norm": 2.7521768591254525,
+      "learning_rate": 1.6995641155212638e-06,
+      "loss": 1.0341,
+      "step": 5780
+    },
+    {
+      "epoch": 0.6079903243193501,
+      "grad_norm": 2.680907512225668,
+      "learning_rate": 1.6987732867237967e-06,
+      "loss": 0.9964,
+      "step": 5781
+    },
+    {
+      "epoch": 0.6080954947611974,
+      "grad_norm": 2.8137952010868705,
+      "learning_rate": 1.6979825472693325e-06,
+      "loss": 0.9617,
+      "step": 5782
+    },
+    {
+      "epoch": 0.6082006652030447,
+      "grad_norm": 3.038100520437388,
+      "learning_rate": 1.6971918972460446e-06,
+      "loss": 1.0006,
+      "step": 5783
+    },
+    {
+      "epoch": 0.608305835644892,
+      "grad_norm": 2.754354820698447,
+      "learning_rate": 1.6964013367420967e-06,
+      "loss": 0.9527,
+      "step": 5784
+    },
+    {
+      "epoch": 0.6084110060867394,
+      "grad_norm": 2.8270349910908354,
+      "learning_rate": 1.695610865845641e-06,
+      "loss": 0.9745,
+      "step": 5785
+    },
+    {
+      "epoch": 0.6085161765285867,
+      "grad_norm": 2.7388567829030115,
+      "learning_rate": 1.694820484644824e-06,
+      "loss": 0.9969,
+      "step": 5786
+    },
+    {
+      "epoch": 0.6086213469704339,
+      "grad_norm": 2.0265884465537396,
+      "learning_rate": 1.6940301932277775e-06,
+      "loss": 0.982,
+      "step": 5787
+    },
+    {
+      "epoch": 0.6087265174122812,
+      "grad_norm": 2.679646606508221,
+      "learning_rate": 1.6932399916826254e-06,
+      "loss": 1.0164,
+      "step": 5788
+    },
+    {
+      "epoch": 0.6088316878541286,
+      "grad_norm": 2.4228792065312215,
+      "learning_rate": 1.692449880097482e-06,
+      "loss": 0.9963,
+      "step": 5789
+    },
+    {
+      "epoch": 0.6089368582959759,
+      "grad_norm": 2.5768372333502376,
+      "learning_rate": 1.6916598585604488e-06,
+      "loss": 0.9351,
+      "step": 5790
+    },
+    {
+      "epoch": 0.6090420287378232,
+      "grad_norm": 2.1737942757800415,
+      "learning_rate": 1.6908699271596213e-06,
+      "loss": 0.9949,
+      "step": 5791
+    },
+    {
+      "epoch": 0.6091471991796705,
+      "grad_norm": 3.049935748235343,
+      "learning_rate": 1.690080085983083e-06,
+      "loss": 0.9867,
+      "step": 5792
+    },
+    {
+      "epoch": 0.6092523696215179,
+      "grad_norm": 2.1746481010421723,
+      "learning_rate": 1.6892903351189065e-06,
+      "loss": 0.982,
+      "step": 5793
+    },
+    {
+      "epoch": 0.6093575400633652,
+      "grad_norm": 2.9706327728285493,
+      "learning_rate": 1.6885006746551551e-06,
+      "loss": 0.9631,
+      "step": 5794
+    },
+    {
+      "epoch": 0.6094627105052125,
+      "grad_norm": 1.935639999637283,
+      "learning_rate": 1.6877111046798829e-06,
+      "loss": 0.9725,
+      "step": 5795
+    },
+    {
+      "epoch": 0.6095678809470598,
+      "grad_norm": 1.936700358641335,
+      "learning_rate": 1.6869216252811327e-06,
+      "loss": 0.939,
+      "step": 5796
+    },
+    {
+      "epoch": 0.6096730513889072,
+      "grad_norm": 2.0817904454728176,
+      "learning_rate": 1.6861322365469373e-06,
+      "loss": 0.9761,
+      "step": 5797
+    },
+    {
+      "epoch": 0.6097782218307545,
+      "grad_norm": 1.5332951079387047,
+      "learning_rate": 1.6853429385653196e-06,
+      "loss": 1.0034,
+      "step": 5798
+    },
+    {
+      "epoch": 0.6098833922726018,
+      "grad_norm": 1.995502353302454,
+      "learning_rate": 1.6845537314242925e-06,
+      "loss": 0.9639,
+      "step": 5799
+    },
+    {
+      "epoch": 0.6099885627144491,
+      "grad_norm": 2.0050706202560553,
+      "learning_rate": 1.6837646152118592e-06,
+      "loss": 0.9861,
+      "step": 5800
+    },
+    {
+      "epoch": 0.6100937331562964,
+      "grad_norm": 2.331889223890956,
+      "learning_rate": 1.6829755900160127e-06,
+      "loss": 0.9623,
+      "step": 5801
+    },
+    {
+      "epoch": 0.6101989035981438,
+      "grad_norm": 2.77022089749751,
+      "learning_rate": 1.6821866559247344e-06,
+      "loss": 1.0013,
+      "step": 5802
+    },
+    {
+      "epoch": 0.6103040740399911,
+      "grad_norm": 2.550781452325086,
+      "learning_rate": 1.681397813025997e-06,
+      "loss": 1.0148,
+      "step": 5803
+    },
+    {
+      "epoch": 0.6104092444818384,
+      "grad_norm": 2.978580897273411,
+      "learning_rate": 1.6806090614077619e-06,
+      "loss": 0.9679,
+      "step": 5804
+    },
+    {
+      "epoch": 0.6105144149236857,
+      "grad_norm": 2.394625694101748,
+      "learning_rate": 1.6798204011579824e-06,
+      "loss": 0.9528,
+      "step": 5805
+    },
+    {
+      "epoch": 0.6106195853655331,
+      "grad_norm": 2.0977205974223407,
+      "learning_rate": 1.6790318323645996e-06,
+      "loss": 1.0424,
+      "step": 5806
+    },
+    {
+      "epoch": 0.6107247558073803,
+      "grad_norm": 2.7073708250144715,
+      "learning_rate": 1.678243355115545e-06,
+      "loss": 0.9837,
+      "step": 5807
+    },
+    {
+      "epoch": 0.6108299262492276,
+      "grad_norm": 2.815173180361358,
+      "learning_rate": 1.6774549694987388e-06,
+      "loss": 0.9832,
+      "step": 5808
+    },
+    {
+      "epoch": 0.6109350966910749,
+      "grad_norm": 2.3860074360187573,
+      "learning_rate": 1.676666675602095e-06,
+      "loss": 1.0108,
+      "step": 5809
+    },
+    {
+      "epoch": 0.6110402671329223,
+      "grad_norm": 2.5298738306057746,
+      "learning_rate": 1.6758784735135119e-06,
+      "loss": 1.0427,
+      "step": 5810
+    },
+    {
+      "epoch": 0.6111454375747696,
+      "grad_norm": 2.5000170631689826,
+      "learning_rate": 1.6750903633208812e-06,
+      "loss": 0.9571,
+      "step": 5811
+    },
+    {
+      "epoch": 0.6112506080166169,
+      "grad_norm": 2.50169544611617,
+      "learning_rate": 1.6743023451120831e-06,
+      "loss": 0.9463,
+      "step": 5812
+    },
+    {
+      "epoch": 0.6113557784584642,
+      "grad_norm": 1.7081141405864757,
+      "learning_rate": 1.673514418974987e-06,
+      "loss": 0.9924,
+      "step": 5813
+    },
+    {
+      "epoch": 0.6114609489003116,
+      "grad_norm": 3.2506880206165825,
+      "learning_rate": 1.672726584997454e-06,
+      "loss": 0.9991,
+      "step": 5814
+    },
+    {
+      "epoch": 0.6115661193421589,
+      "grad_norm": 2.953416905081632,
+      "learning_rate": 1.6719388432673336e-06,
+      "loss": 1.0007,
+      "step": 5815
+    },
+    {
+      "epoch": 0.6116712897840062,
+      "grad_norm": 2.818662791521142,
+      "learning_rate": 1.6711511938724648e-06,
+      "loss": 0.9789,
+      "step": 5816
+    },
+    {
+      "epoch": 0.6117764602258535,
+      "grad_norm": 2.3521060058811187,
+      "learning_rate": 1.670363636900676e-06,
+      "loss": 0.9882,
+      "step": 5817
+    },
+    {
+      "epoch": 0.6118816306677008,
+      "grad_norm": 2.293305321670319,
+      "learning_rate": 1.669576172439787e-06,
+      "loss": 1.0534,
+      "step": 5818
+    },
+    {
+      "epoch": 0.6119868011095482,
+      "grad_norm": 1.6726020449012442,
+      "learning_rate": 1.668788800577606e-06,
+      "loss": 0.9545,
+      "step": 5819
+    },
+    {
+      "epoch": 0.6120919715513955,
+      "grad_norm": 2.76117276419402,
+      "learning_rate": 1.6680015214019305e-06,
+      "loss": 0.977,
+      "step": 5820
+    },
+    {
+      "epoch": 0.6121971419932428,
+      "grad_norm": 2.1802762486981258,
+      "learning_rate": 1.6672143350005488e-06,
+      "loss": 1.0135,
+      "step": 5821
+    },
+    {
+      "epoch": 0.6123023124350901,
+      "grad_norm": 3.9598469389280613,
+      "learning_rate": 1.6664272414612368e-06,
+      "loss": 1.0158,
+      "step": 5822
+    },
+    {
+      "epoch": 0.6124074828769375,
+      "grad_norm": 2.2549595245033145,
+      "learning_rate": 1.6656402408717636e-06,
+      "loss": 0.9378,
+      "step": 5823
+    },
+    {
+      "epoch": 0.6125126533187848,
+      "grad_norm": 2.196920678193086,
+      "learning_rate": 1.6648533333198858e-06,
+      "loss": 1.0162,
+      "step": 5824
+    },
+    {
+      "epoch": 0.6126178237606321,
+      "grad_norm": 2.6642892656099866,
+      "learning_rate": 1.664066518893349e-06,
+      "loss": 1.0044,
+      "step": 5825
+    },
+    {
+      "epoch": 0.6127229942024794,
+      "grad_norm": 2.7734023454217027,
+      "learning_rate": 1.6632797976798887e-06,
+      "loss": 0.9841,
+      "step": 5826
+    },
+    {
+      "epoch": 0.6128281646443268,
+      "grad_norm": 2.3104109605714718,
+      "learning_rate": 1.6624931697672298e-06,
+      "loss": 0.973,
+      "step": 5827
+    },
+    {
+      "epoch": 0.612933335086174,
+      "grad_norm": 2.9928279355445984,
+      "learning_rate": 1.6617066352430893e-06,
+      "loss": 0.984,
+      "step": 5828
+    },
+    {
+      "epoch": 0.6130385055280213,
+      "grad_norm": 2.5371426562781116,
+      "learning_rate": 1.6609201941951715e-06,
+      "loss": 0.9845,
+      "step": 5829
+    },
+    {
+      "epoch": 0.6131436759698686,
+      "grad_norm": 1.8754458670700893,
+      "learning_rate": 1.6601338467111699e-06,
+      "loss": 0.9817,
+      "step": 5830
+    },
+    {
+      "epoch": 0.613248846411716,
+      "grad_norm": 2.322347493475784,
+      "learning_rate": 1.6593475928787683e-06,
+      "loss": 0.9833,
+      "step": 5831
+    },
+    {
+      "epoch": 0.6133540168535633,
+      "grad_norm": 2.604820493235049,
+      "learning_rate": 1.6585614327856408e-06,
+      "loss": 0.9869,
+      "step": 5832
+    },
+    {
+      "epoch": 0.6134591872954106,
+      "grad_norm": 2.2184504978618222,
+      "learning_rate": 1.6577753665194502e-06,
+      "loss": 0.9635,
+      "step": 5833
+    },
+    {
+      "epoch": 0.6135643577372579,
+      "grad_norm": 2.8750863642027644,
+      "learning_rate": 1.656989394167849e-06,
+      "loss": 1.0122,
+      "step": 5834
+    },
+    {
+      "epoch": 0.6136695281791053,
+      "grad_norm": 2.429229413863499,
+      "learning_rate": 1.6562035158184791e-06,
+      "loss": 0.9418,
+      "step": 5835
+    },
+    {
+      "epoch": 0.6137746986209526,
+      "grad_norm": 2.404031095251293,
+      "learning_rate": 1.655417731558971e-06,
+      "loss": 0.9868,
+      "step": 5836
+    },
+    {
+      "epoch": 0.6138798690627999,
+      "grad_norm": 3.657059627061674,
+      "learning_rate": 1.6546320414769474e-06,
+      "loss": 1.0213,
+      "step": 5837
+    },
+    {
+      "epoch": 0.6139850395046472,
+      "grad_norm": 2.3207503976843564,
+      "learning_rate": 1.6538464456600186e-06,
+      "loss": 0.9973,
+      "step": 5838
+    },
+    {
+      "epoch": 0.6140902099464945,
+      "grad_norm": 2.1274755433279586,
+      "learning_rate": 1.6530609441957844e-06,
+      "loss": 0.9843,
+      "step": 5839
+    },
+    {
+      "epoch": 0.6141953803883419,
+      "grad_norm": 1.7062502169444627,
+      "learning_rate": 1.6522755371718333e-06,
+      "loss": 0.9573,
+      "step": 5840
+    },
+    {
+      "epoch": 0.6143005508301892,
+      "grad_norm": 2.7514773804668002,
+      "learning_rate": 1.6514902246757458e-06,
+      "loss": 1.0072,
+      "step": 5841
+    },
+    {
+      "epoch": 0.6144057212720365,
+      "grad_norm": 2.251985168187469,
+      "learning_rate": 1.6507050067950898e-06,
+      "loss": 0.9661,
+      "step": 5842
+    },
+    {
+      "epoch": 0.6145108917138838,
+      "grad_norm": 1.8065977765727852,
+      "learning_rate": 1.649919883617423e-06,
+      "loss": 0.9683,
+      "step": 5843
+    },
+    {
+      "epoch": 0.6146160621557312,
+      "grad_norm": 2.9561496412074106,
+      "learning_rate": 1.6491348552302927e-06,
+      "loss": 1.0008,
+      "step": 5844
+    },
+    {
+      "epoch": 0.6147212325975785,
+      "grad_norm": 1.9869101556732869,
+      "learning_rate": 1.6483499217212357e-06,
+      "loss": 1.0071,
+      "step": 5845
+    },
+    {
+      "epoch": 0.6148264030394258,
+      "grad_norm": 2.1835075291524184,
+      "learning_rate": 1.6475650831777784e-06,
+      "loss": 0.9863,
+      "step": 5846
+    },
+    {
+      "epoch": 0.6149315734812731,
+      "grad_norm": 2.430561118173312,
+      "learning_rate": 1.6467803396874365e-06,
+      "loss": 0.9679,
+      "step": 5847
+    },
+    {
+      "epoch": 0.6150367439231204,
+      "grad_norm": 2.28403426164745,
+      "learning_rate": 1.6459956913377145e-06,
+      "loss": 1.02,
+      "step": 5848
+    },
+    {
+      "epoch": 0.6151419143649677,
+      "grad_norm": 2.6620061740930017,
+      "learning_rate": 1.6452111382161074e-06,
+      "loss": 0.9631,
+      "step": 5849
+    },
+    {
+      "epoch": 0.615247084806815,
+      "grad_norm": 1.6904530329363399,
+      "learning_rate": 1.6444266804100972e-06,
+      "loss": 0.9713,
+      "step": 5850
+    },
+    {
+      "epoch": 0.6153522552486623,
+      "grad_norm": 2.3212203494074455,
+      "learning_rate": 1.6436423180071598e-06,
+      "loss": 0.968,
+      "step": 5851
+    },
+    {
+      "epoch": 0.6154574256905097,
+      "grad_norm": 2.5364871090662837,
+      "learning_rate": 1.6428580510947563e-06,
+      "loss": 0.9652,
+      "step": 5852
+    },
+    {
+      "epoch": 0.615562596132357,
+      "grad_norm": 3.030037498205161,
+      "learning_rate": 1.6420738797603386e-06,
+      "loss": 0.9879,
+      "step": 5853
+    },
+    {
+      "epoch": 0.6156677665742043,
+      "grad_norm": 2.822749830603122,
+      "learning_rate": 1.6412898040913471e-06,
+      "loss": 1.0141,
+      "step": 5854
+    },
+    {
+      "epoch": 0.6157729370160516,
+      "grad_norm": 2.5948534835106645,
+      "learning_rate": 1.6405058241752142e-06,
+      "loss": 0.9961,
+      "step": 5855
+    },
+    {
+      "epoch": 0.615878107457899,
+      "grad_norm": 1.7892323129704704,
+      "learning_rate": 1.6397219400993584e-06,
+      "loss": 0.9357,
+      "step": 5856
+    },
+    {
+      "epoch": 0.6159832778997463,
+      "grad_norm": 2.4386090952796935,
+      "learning_rate": 1.6389381519511893e-06,
+      "loss": 0.977,
+      "step": 5857
+    },
+    {
+      "epoch": 0.6160884483415936,
+      "grad_norm": 2.621745541745822,
+      "learning_rate": 1.6381544598181048e-06,
+      "loss": 1.0185,
+      "step": 5858
+    },
+    {
+      "epoch": 0.6161936187834409,
+      "grad_norm": 2.232535333948487,
+      "learning_rate": 1.6373708637874925e-06,
+      "loss": 1.0113,
+      "step": 5859
+    },
+    {
+      "epoch": 0.6162987892252882,
+      "grad_norm": 3.2749764704846553,
+      "learning_rate": 1.6365873639467314e-06,
+      "loss": 0.9737,
+      "step": 5860
+    },
+    {
+      "epoch": 0.6164039596671356,
+      "grad_norm": 1.869914950095561,
+      "learning_rate": 1.635803960383186e-06,
+      "loss": 1.0177,
+      "step": 5861
+    },
+    {
+      "epoch": 0.6165091301089829,
+      "grad_norm": 2.285719824488398,
+      "learning_rate": 1.6350206531842122e-06,
+      "loss": 1.0154,
+      "step": 5862
+    },
+    {
+      "epoch": 0.6166143005508302,
+      "grad_norm": 2.4241270283535825,
+      "learning_rate": 1.634237442437154e-06,
+      "loss": 1.0028,
+      "step": 5863
+    },
+    {
+      "epoch": 0.6167194709926775,
+      "grad_norm": 2.2622451802360932,
+      "learning_rate": 1.6334543282293475e-06,
+      "loss": 0.9626,
+      "step": 5864
+    },
+    {
+      "epoch": 0.6168246414345249,
+      "grad_norm": 3.2500786695737562,
+      "learning_rate": 1.6326713106481148e-06,
+      "loss": 0.9625,
+      "step": 5865
+    },
+    {
+      "epoch": 0.6169298118763722,
+      "grad_norm": 3.008447944721862,
+      "learning_rate": 1.6318883897807688e-06,
+      "loss": 0.9794,
+      "step": 5866
+    },
+    {
+      "epoch": 0.6170349823182195,
+      "grad_norm": 2.7479663807080907,
+      "learning_rate": 1.631105565714611e-06,
+      "loss": 0.988,
+      "step": 5867
+    },
+    {
+      "epoch": 0.6171401527600667,
+      "grad_norm": 2.784120946238147,
+      "learning_rate": 1.6303228385369318e-06,
+      "loss": 1.0219,
+      "step": 5868
+    },
+    {
+      "epoch": 0.617245323201914,
+      "grad_norm": 2.3151431857814155,
+      "learning_rate": 1.629540208335013e-06,
+      "loss": 1.0192,
+      "step": 5869
+    },
+    {
+      "epoch": 0.6173504936437614,
+      "grad_norm": 2.5308826576755656,
+      "learning_rate": 1.6287576751961225e-06,
+      "loss": 0.9902,
+      "step": 5870
+    },
+    {
+      "epoch": 0.6174556640856087,
+      "grad_norm": 2.3429901480764976,
+      "learning_rate": 1.6279752392075192e-06,
+      "loss": 0.9569,
+      "step": 5871
+    },
+    {
+      "epoch": 0.617560834527456,
+      "grad_norm": 2.181050276365485,
+      "learning_rate": 1.6271929004564514e-06,
+      "loss": 1.0071,
+      "step": 5872
+    },
+    {
+      "epoch": 0.6176660049693034,
+      "grad_norm": 2.601702484291357,
+      "learning_rate": 1.6264106590301538e-06,
+      "loss": 0.9857,
+      "step": 5873
+    },
+    {
+      "epoch": 0.6177711754111507,
+      "grad_norm": 2.459923668745263,
+      "learning_rate": 1.6256285150158551e-06,
+      "loss": 0.9918,
+      "step": 5874
+    },
+    {
+      "epoch": 0.617876345852998,
+      "grad_norm": 2.197411706775619,
+      "learning_rate": 1.6248464685007692e-06,
+      "loss": 0.9421,
+      "step": 5875
+    },
+    {
+      "epoch": 0.6179815162948453,
+      "grad_norm": 2.095985627899745,
+      "learning_rate": 1.6240645195721017e-06,
+      "loss": 1.0025,
+      "step": 5876
+    },
+    {
+      "epoch": 0.6180866867366926,
+      "grad_norm": 2.749362024760833,
+      "learning_rate": 1.6232826683170425e-06,
+      "loss": 1.034,
+      "step": 5877
+    },
+    {
+      "epoch": 0.61819185717854,
+      "grad_norm": 2.274015303414764,
+      "learning_rate": 1.6225009148227778e-06,
+      "loss": 1.0224,
+      "step": 5878
+    },
+    {
+      "epoch": 0.6182970276203873,
+      "grad_norm": 2.895694165275508,
+      "learning_rate": 1.6217192591764774e-06,
+      "loss": 1.0139,
+      "step": 5879
+    },
+    {
+      "epoch": 0.6184021980622346,
+      "grad_norm": 2.7394375689175225,
+      "learning_rate": 1.6209377014653028e-06,
+      "loss": 1.001,
+      "step": 5880
+    },
+    {
+      "epoch": 0.618507368504082,
+      "grad_norm": 2.9670712927952763,
+      "learning_rate": 1.6201562417764028e-06,
+      "loss": 1.0286,
+      "step": 5881
+    },
+    {
+      "epoch": 0.6186125389459293,
+      "grad_norm": 2.313393062040808,
+      "learning_rate": 1.6193748801969164e-06,
+      "loss": 1.0052,
+      "step": 5882
+    },
+    {
+      "epoch": 0.6187177093877766,
+      "grad_norm": 2.578512587304975,
+      "learning_rate": 1.6185936168139724e-06,
+      "loss": 1.0067,
+      "step": 5883
+    },
+    {
+      "epoch": 0.6188228798296239,
+      "grad_norm": 2.489037978706978,
+      "learning_rate": 1.617812451714687e-06,
+      "loss": 0.9524,
+      "step": 5884
+    },
+    {
+      "epoch": 0.6189280502714712,
+      "grad_norm": 2.056449619227656,
+      "learning_rate": 1.6170313849861662e-06,
+      "loss": 0.9762,
+      "step": 5885
+    },
+    {
+      "epoch": 0.6190332207133186,
+      "grad_norm": 2.7761192879641396,
+      "learning_rate": 1.6162504167155044e-06,
+      "loss": 0.9772,
+      "step": 5886
+    },
+    {
+      "epoch": 0.6191383911551659,
+      "grad_norm": 2.2429168320600237,
+      "learning_rate": 1.6154695469897873e-06,
+      "loss": 0.9714,
+      "step": 5887
+    },
+    {
+      "epoch": 0.6192435615970132,
+      "grad_norm": 2.0592296766661806,
+      "learning_rate": 1.6146887758960865e-06,
+      "loss": 1.0256,
+      "step": 5888
+    },
+    {
+      "epoch": 0.6193487320388604,
+      "grad_norm": 2.393849999328136,
+      "learning_rate": 1.6139081035214654e-06,
+      "loss": 0.9837,
+      "step": 5889
+    },
+    {
+      "epoch": 0.6194539024807078,
+      "grad_norm": 2.618576013105238,
+      "learning_rate": 1.6131275299529736e-06,
+      "loss": 1.0011,
+      "step": 5890
+    },
+    {
+      "epoch": 0.6195590729225551,
+      "grad_norm": 2.2774723438717595,
+      "learning_rate": 1.612347055277651e-06,
+      "loss": 0.9881,
+      "step": 5891
+    },
+    {
+      "epoch": 0.6196642433644024,
+      "grad_norm": 1.7906851084234103,
+      "learning_rate": 1.6115666795825284e-06,
+      "loss": 0.9792,
+      "step": 5892
+    },
+    {
+      "epoch": 0.6197694138062497,
+      "grad_norm": 2.148647148559524,
+      "learning_rate": 1.6107864029546221e-06,
+      "loss": 1.0106,
+      "step": 5893
+    },
+    {
+      "epoch": 0.619874584248097,
+      "grad_norm": 2.6630752042411907,
+      "learning_rate": 1.6100062254809395e-06,
+      "loss": 1.0022,
+      "step": 5894
+    },
+    {
+      "epoch": 0.6199797546899444,
+      "grad_norm": 2.1012206126791355,
+      "learning_rate": 1.6092261472484765e-06,
+      "loss": 0.9314,
+      "step": 5895
+    },
+    {
+      "epoch": 0.6200849251317917,
+      "grad_norm": 2.581191874335221,
+      "learning_rate": 1.6084461683442176e-06,
+      "loss": 0.9863,
+      "step": 5896
+    },
+    {
+      "epoch": 0.620190095573639,
+      "grad_norm": 2.4711381450267513,
+      "learning_rate": 1.6076662888551373e-06,
+      "loss": 0.9523,
+      "step": 5897
+    },
+    {
+      "epoch": 0.6202952660154863,
+      "grad_norm": 1.9627152201554907,
+      "learning_rate": 1.6068865088681975e-06,
+      "loss": 0.9788,
+      "step": 5898
+    },
+    {
+      "epoch": 0.6204004364573337,
+      "grad_norm": 3.1900135916304433,
+      "learning_rate": 1.6061068284703502e-06,
+      "loss": 0.9554,
+      "step": 5899
+    },
+    {
+      "epoch": 0.620505606899181,
+      "grad_norm": 2.4902737582100762,
+      "learning_rate": 1.605327247748534e-06,
+      "loss": 0.9749,
+      "step": 5900
+    },
+    {
+      "epoch": 0.6206107773410283,
+      "grad_norm": 1.979528958259765,
+      "learning_rate": 1.604547766789681e-06,
+      "loss": 0.9627,
+      "step": 5901
+    },
+    {
+      "epoch": 0.6207159477828756,
+      "grad_norm": 2.5553977998948945,
+      "learning_rate": 1.6037683856807083e-06,
+      "loss": 0.9983,
+      "step": 5902
+    },
+    {
+      "epoch": 0.620821118224723,
+      "grad_norm": 2.5672010413971638,
+      "learning_rate": 1.6029891045085224e-06,
+      "loss": 1.0212,
+      "step": 5903
+    },
+    {
+      "epoch": 0.6209262886665703,
+      "grad_norm": 2.4249488163123267,
+      "learning_rate": 1.60220992336002e-06,
+      "loss": 1.0261,
+      "step": 5904
+    },
+    {
+      "epoch": 0.6210314591084176,
+      "grad_norm": 2.6586835053107656,
+      "learning_rate": 1.6014308423220848e-06,
+      "loss": 0.9869,
+      "step": 5905
+    },
+    {
+      "epoch": 0.6211366295502649,
+      "grad_norm": 1.9453115647301724,
+      "learning_rate": 1.6006518614815913e-06,
+      "loss": 1.002,
+      "step": 5906
+    },
+    {
+      "epoch": 0.6212417999921123,
+      "grad_norm": 2.0323748172785425,
+      "learning_rate": 1.5998729809254017e-06,
+      "loss": 0.9806,
+      "step": 5907
+    },
+    {
+      "epoch": 0.6213469704339596,
+      "grad_norm": 2.1843910280641445,
+      "learning_rate": 1.599094200740367e-06,
+      "loss": 0.9464,
+      "step": 5908
+    },
+    {
+      "epoch": 0.6214521408758068,
+      "grad_norm": 2.7758359003572113,
+      "learning_rate": 1.5983155210133267e-06,
+      "loss": 1.0226,
+      "step": 5909
+    },
+    {
+      "epoch": 0.6215573113176541,
+      "grad_norm": 2.6474203799852445,
+      "learning_rate": 1.5975369418311113e-06,
+      "loss": 1.0131,
+      "step": 5910
+    },
+    {
+      "epoch": 0.6216624817595015,
+      "grad_norm": 2.2015849233799907,
+      "learning_rate": 1.5967584632805378e-06,
+      "loss": 0.9715,
+      "step": 5911
+    },
+    {
+      "epoch": 0.6217676522013488,
+      "grad_norm": 2.5493391965343144,
+      "learning_rate": 1.595980085448412e-06,
+      "loss": 0.9627,
+      "step": 5912
+    },
+    {
+      "epoch": 0.6218728226431961,
+      "grad_norm": 2.116283213919637,
+      "learning_rate": 1.5952018084215293e-06,
+      "loss": 0.9863,
+      "step": 5913
+    },
+    {
+      "epoch": 0.6219779930850434,
+      "grad_norm": 2.4987388936618853,
+      "learning_rate": 1.5944236322866725e-06,
+      "loss": 0.9764,
+      "step": 5914
+    },
+    {
+      "epoch": 0.6220831635268907,
+      "grad_norm": 2.327898890577115,
+      "learning_rate": 1.5936455571306164e-06,
+      "loss": 1.0053,
+      "step": 5915
+    },
+    {
+      "epoch": 0.6221883339687381,
+      "grad_norm": 2.0393178342923215,
+      "learning_rate": 1.592867583040122e-06,
+      "loss": 0.9501,
+      "step": 5916
+    },
+    {
+      "epoch": 0.6222935044105854,
+      "grad_norm": 2.5407681947430527,
+      "learning_rate": 1.5920897101019384e-06,
+      "loss": 1.0122,
+      "step": 5917
+    },
+    {
+      "epoch": 0.6223986748524327,
+      "grad_norm": 2.4916758411681226,
+      "learning_rate": 1.5913119384028048e-06,
+      "loss": 0.9687,
+      "step": 5918
+    },
+    {
+      "epoch": 0.62250384529428,
+      "grad_norm": 1.9567570424692333,
+      "learning_rate": 1.5905342680294495e-06,
+      "loss": 0.9773,
+      "step": 5919
+    },
+    {
+      "epoch": 0.6226090157361274,
+      "grad_norm": 3.263714903866836,
+      "learning_rate": 1.589756699068588e-06,
+      "loss": 0.9846,
+      "step": 5920
+    },
+    {
+      "epoch": 0.6227141861779747,
+      "grad_norm": 2.9575517862827936,
+      "learning_rate": 1.5889792316069259e-06,
+      "loss": 1.0221,
+      "step": 5921
+    },
+    {
+      "epoch": 0.622819356619822,
+      "grad_norm": 2.606415686984147,
+      "learning_rate": 1.5882018657311563e-06,
+      "loss": 0.9988,
+      "step": 5922
+    },
+    {
+      "epoch": 0.6229245270616693,
+      "grad_norm": 2.446109177024214,
+      "learning_rate": 1.587424601527961e-06,
+      "loss": 0.9587,
+      "step": 5923
+    },
+    {
+      "epoch": 0.6230296975035167,
+      "grad_norm": 2.553224170320699,
+      "learning_rate": 1.5866474390840126e-06,
+      "loss": 0.9538,
+      "step": 5924
+    },
+    {
+      "epoch": 0.623134867945364,
+      "grad_norm": 2.5094380956851507,
+      "learning_rate": 1.5858703784859697e-06,
+      "loss": 1.03,
+      "step": 5925
+    },
+    {
+      "epoch": 0.6232400383872113,
+      "grad_norm": 2.0855074263373163,
+      "learning_rate": 1.5850934198204812e-06,
+      "loss": 0.98,
+      "step": 5926
+    },
+    {
+      "epoch": 0.6233452088290586,
+      "grad_norm": 2.6961341409896455,
+      "learning_rate": 1.5843165631741841e-06,
+      "loss": 1.0272,
+      "step": 5927
+    },
+    {
+      "epoch": 0.623450379270906,
+      "grad_norm": 2.4500422545848313,
+      "learning_rate": 1.5835398086337017e-06,
+      "loss": 0.9762,
+      "step": 5928
+    },
+    {
+      "epoch": 0.6235555497127532,
+      "grad_norm": 1.9103182982629987,
+      "learning_rate": 1.582763156285651e-06,
+      "loss": 0.9472,
+      "step": 5929
+    },
+    {
+      "epoch": 0.6236607201546005,
+      "grad_norm": 2.412571799050157,
+      "learning_rate": 1.581986606216634e-06,
+      "loss": 0.9921,
+      "step": 5930
+    },
+    {
+      "epoch": 0.6237658905964478,
+      "grad_norm": 2.5521277009233536,
+      "learning_rate": 1.5812101585132416e-06,
+      "loss": 0.9579,
+      "step": 5931
+    },
+    {
+      "epoch": 0.6238710610382951,
+      "grad_norm": 2.6571937101267458,
+      "learning_rate": 1.5804338132620535e-06,
+      "loss": 1.017,
+      "step": 5932
+    },
+    {
+      "epoch": 0.6239762314801425,
+      "grad_norm": 2.0622319382949676,
+      "learning_rate": 1.5796575705496392e-06,
+      "loss": 0.9466,
+      "step": 5933
+    },
+    {
+      "epoch": 0.6240814019219898,
+      "grad_norm": 3.2155213422084974,
+      "learning_rate": 1.578881430462555e-06,
+      "loss": 1.0423,
+      "step": 5934
+    },
+    {
+      "epoch": 0.6241865723638371,
+      "grad_norm": 2.8066759471700085,
+      "learning_rate": 1.5781053930873468e-06,
+      "loss": 1.0102,
+      "step": 5935
+    },
+    {
+      "epoch": 0.6242917428056844,
+      "grad_norm": 2.3021253355837206,
+      "learning_rate": 1.5773294585105492e-06,
+      "loss": 0.9828,
+      "step": 5936
+    },
+    {
+      "epoch": 0.6243969132475318,
+      "grad_norm": 2.476492213812445,
+      "learning_rate": 1.5765536268186832e-06,
+      "loss": 1.0049,
+      "step": 5937
+    },
+    {
+      "epoch": 0.6245020836893791,
+      "grad_norm": 2.5139739664998375,
+      "learning_rate": 1.5757778980982627e-06,
+      "loss": 1.0335,
+      "step": 5938
+    },
+    {
+      "epoch": 0.6246072541312264,
+      "grad_norm": 2.2258196498122844,
+      "learning_rate": 1.5750022724357861e-06,
+      "loss": 0.9761,
+      "step": 5939
+    },
+    {
+      "epoch": 0.6247124245730737,
+      "grad_norm": 2.726588286992223,
+      "learning_rate": 1.5742267499177416e-06,
+      "loss": 1.0145,
+      "step": 5940
+    },
+    {
+      "epoch": 0.6248175950149211,
+      "grad_norm": 2.6669317872534544,
+      "learning_rate": 1.5734513306306053e-06,
+      "loss": 0.9992,
+      "step": 5941
+    },
+    {
+      "epoch": 0.6249227654567684,
+      "grad_norm": 2.084506573805443,
+      "learning_rate": 1.5726760146608443e-06,
+      "loss": 1.0112,
+      "step": 5942
+    },
+    {
+      "epoch": 0.6250279358986157,
+      "grad_norm": 2.6171495295222855,
+      "learning_rate": 1.5719008020949114e-06,
+      "loss": 0.9778,
+      "step": 5943
+    },
+    {
+      "epoch": 0.625133106340463,
+      "grad_norm": 2.140554094520001,
+      "learning_rate": 1.5711256930192486e-06,
+      "loss": 1.0035,
+      "step": 5944
+    },
+    {
+      "epoch": 0.6252382767823104,
+      "grad_norm": 1.9264447426695412,
+      "learning_rate": 1.5703506875202868e-06,
+      "loss": 0.9721,
+      "step": 5945
+    },
+    {
+      "epoch": 0.6253434472241577,
+      "grad_norm": 2.5055312262881144,
+      "learning_rate": 1.569575785684444e-06,
+      "loss": 0.9955,
+      "step": 5946
+    },
+    {
+      "epoch": 0.625448617666005,
+      "grad_norm": 2.807271348200614,
+      "learning_rate": 1.5688009875981308e-06,
+      "loss": 1.0316,
+      "step": 5947
+    },
+    {
+      "epoch": 0.6255537881078523,
+      "grad_norm": 2.0493312335861074,
+      "learning_rate": 1.5680262933477405e-06,
+      "loss": 0.9913,
+      "step": 5948
+    },
+    {
+      "epoch": 0.6256589585496997,
+      "grad_norm": 2.7559029051850708,
+      "learning_rate": 1.5672517030196582e-06,
+      "loss": 1.0043,
+      "step": 5949
+    },
+    {
+      "epoch": 0.6257641289915469,
+      "grad_norm": 2.1694704480396445,
+      "learning_rate": 1.5664772167002573e-06,
+      "loss": 0.9878,
+      "step": 5950
+    },
+    {
+      "epoch": 0.6258692994333942,
+      "grad_norm": 2.1988058558900323,
+      "learning_rate": 1.565702834475897e-06,
+      "loss": 0.9623,
+      "step": 5951
+    },
+    {
+      "epoch": 0.6259744698752415,
+      "grad_norm": 2.6284645651941245,
+      "learning_rate": 1.5649285564329296e-06,
+      "loss": 0.9351,
+      "step": 5952
+    },
+    {
+      "epoch": 0.6260796403170888,
+      "grad_norm": 2.2987819649896393,
+      "learning_rate": 1.5641543826576926e-06,
+      "loss": 0.9896,
+      "step": 5953
+    },
+    {
+      "epoch": 0.6261848107589362,
+      "grad_norm": 3.0571010198179485,
+      "learning_rate": 1.5633803132365117e-06,
+      "loss": 1.0123,
+      "step": 5954
+    },
+    {
+      "epoch": 0.6262899812007835,
+      "grad_norm": 3.3037893530878386,
+      "learning_rate": 1.5626063482557009e-06,
+      "loss": 0.9803,
+      "step": 5955
+    },
+    {
+      "epoch": 0.6263951516426308,
+      "grad_norm": 2.9959633481925416,
+      "learning_rate": 1.561832487801565e-06,
+      "loss": 0.9568,
+      "step": 5956
+    },
+    {
+      "epoch": 0.6265003220844781,
+      "grad_norm": 2.3258563684955638,
+      "learning_rate": 1.5610587319603942e-06,
+      "loss": 0.9807,
+      "step": 5957
+    },
+    {
+      "epoch": 0.6266054925263255,
+      "grad_norm": 2.434441203619329,
+      "learning_rate": 1.5602850808184695e-06,
+      "loss": 0.9661,
+      "step": 5958
+    },
+    {
+      "epoch": 0.6267106629681728,
+      "grad_norm": 2.33530559179327,
+      "learning_rate": 1.5595115344620575e-06,
+      "loss": 0.9838,
+      "step": 5959
+    },
+    {
+      "epoch": 0.6268158334100201,
+      "grad_norm": 2.2518352944630355,
+      "learning_rate": 1.558738092977415e-06,
+      "loss": 1.0295,
+      "step": 5960
+    },
+    {
+      "epoch": 0.6269210038518674,
+      "grad_norm": 2.4291496261842473,
+      "learning_rate": 1.5579647564507878e-06,
+      "loss": 1.0062,
+      "step": 5961
+    },
+    {
+      "epoch": 0.6270261742937148,
+      "grad_norm": 2.436150506595444,
+      "learning_rate": 1.5571915249684084e-06,
+      "loss": 0.9884,
+      "step": 5962
+    },
+    {
+      "epoch": 0.6271313447355621,
+      "grad_norm": 2.3348752680999665,
+      "learning_rate": 1.5564183986164982e-06,
+      "loss": 0.9936,
+      "step": 5963
+    },
+    {
+      "epoch": 0.6272365151774094,
+      "grad_norm": 2.694260375344963,
+      "learning_rate": 1.5556453774812652e-06,
+      "loss": 0.9632,
+      "step": 5964
+    },
+    {
+      "epoch": 0.6273416856192567,
+      "grad_norm": 3.2038875399300526,
+      "learning_rate": 1.5548724616489095e-06,
+      "loss": 0.9407,
+      "step": 5965
+    },
+    {
+      "epoch": 0.6274468560611041,
+      "grad_norm": 2.7091848613454936,
+      "learning_rate": 1.5540996512056164e-06,
+      "loss": 0.9873,
+      "step": 5966
+    },
+    {
+      "epoch": 0.6275520265029514,
+      "grad_norm": 2.8068933011488033,
+      "learning_rate": 1.5533269462375603e-06,
+      "loss": 0.9585,
+      "step": 5967
+    },
+    {
+      "epoch": 0.6276571969447987,
+      "grad_norm": 2.7111104148324543,
+      "learning_rate": 1.5525543468309035e-06,
+      "loss": 0.987,
+      "step": 5968
+    },
+    {
+      "epoch": 0.627762367386646,
+      "grad_norm": 2.744267573867375,
+      "learning_rate": 1.5517818530717965e-06,
+      "loss": 0.9686,
+      "step": 5969
+    },
+    {
+      "epoch": 0.6278675378284932,
+      "grad_norm": 2.830012535610845,
+      "learning_rate": 1.5510094650463797e-06,
+      "loss": 0.9701,
+      "step": 5970
+    },
+    {
+      "epoch": 0.6279727082703406,
+      "grad_norm": 3.230771720949463,
+      "learning_rate": 1.5502371828407797e-06,
+      "loss": 1.073,
+      "step": 5971
+    },
+    {
+      "epoch": 0.6280778787121879,
+      "grad_norm": 2.3409480978908896,
+      "learning_rate": 1.5494650065411116e-06,
+      "loss": 0.9638,
+      "step": 5972
+    },
+    {
+      "epoch": 0.6281830491540352,
+      "grad_norm": 2.444047229295055,
+      "learning_rate": 1.548692936233479e-06,
+      "loss": 0.9757,
+      "step": 5973
+    },
+    {
+      "epoch": 0.6282882195958825,
+      "grad_norm": 2.00504969984711,
+      "learning_rate": 1.5479209720039733e-06,
+      "loss": 0.9657,
+      "step": 5974
+    },
+    {
+      "epoch": 0.6283933900377299,
+      "grad_norm": 3.516884926223067,
+      "learning_rate": 1.5471491139386763e-06,
+      "loss": 0.9693,
+      "step": 5975
+    },
+    {
+      "epoch": 0.6284985604795772,
+      "grad_norm": 2.794487173825674,
+      "learning_rate": 1.546377362123655e-06,
+      "loss": 0.9452,
+      "step": 5976
+    },
+    {
+      "epoch": 0.6286037309214245,
+      "grad_norm": 2.1935018379858984,
+      "learning_rate": 1.5456057166449657e-06,
+      "loss": 0.9745,
+      "step": 5977
+    },
+    {
+      "epoch": 0.6287089013632718,
+      "grad_norm": 2.297663462287113,
+      "learning_rate": 1.5448341775886528e-06,
+      "loss": 0.9751,
+      "step": 5978
+    },
+    {
+      "epoch": 0.6288140718051192,
+      "grad_norm": 2.4320433680260085,
+      "learning_rate": 1.544062745040749e-06,
+      "loss": 0.9293,
+      "step": 5979
+    },
+    {
+      "epoch": 0.6289192422469665,
+      "grad_norm": 2.872577702758348,
+      "learning_rate": 1.5432914190872757e-06,
+      "loss": 0.9477,
+      "step": 5980
+    },
+    {
+      "epoch": 0.6290244126888138,
+      "grad_norm": 2.6721948871558783,
+      "learning_rate": 1.542520199814241e-06,
+      "loss": 0.9874,
+      "step": 5981
+    },
+    {
+      "epoch": 0.6291295831306611,
+      "grad_norm": 2.5749173147193996,
+      "learning_rate": 1.541749087307642e-06,
+      "loss": 0.9875,
+      "step": 5982
+    },
+    {
+      "epoch": 0.6292347535725085,
+      "grad_norm": 3.239662190499185,
+      "learning_rate": 1.540978081653463e-06,
+      "loss": 0.9966,
+      "step": 5983
+    },
+    {
+      "epoch": 0.6293399240143558,
+      "grad_norm": 2.4691533579634926,
+      "learning_rate": 1.5402071829376785e-06,
+      "loss": 0.9739,
+      "step": 5984
+    },
+    {
+      "epoch": 0.6294450944562031,
+      "grad_norm": 2.59932511245003,
+      "learning_rate": 1.5394363912462492e-06,
+      "loss": 0.9414,
+      "step": 5985
+    },
+    {
+      "epoch": 0.6295502648980504,
+      "grad_norm": 3.223852238411686,
+      "learning_rate": 1.5386657066651242e-06,
+      "loss": 1.0248,
+      "step": 5986
+    },
+    {
+      "epoch": 0.6296554353398978,
+      "grad_norm": 2.6080630359778993,
+      "learning_rate": 1.5378951292802397e-06,
+      "loss": 0.9842,
+      "step": 5987
+    },
+    {
+      "epoch": 0.6297606057817451,
+      "grad_norm": 2.133772821565045,
+      "learning_rate": 1.537124659177523e-06,
+      "loss": 0.9879,
+      "step": 5988
+    },
+    {
+      "epoch": 0.6298657762235924,
+      "grad_norm": 2.3424669533242573,
+      "learning_rate": 1.5363542964428869e-06,
+      "loss": 0.981,
+      "step": 5989
+    },
+    {
+      "epoch": 0.6299709466654396,
+      "grad_norm": 2.805150094282492,
+      "learning_rate": 1.5355840411622324e-06,
+      "loss": 0.9469,
+      "step": 5990
+    },
+    {
+      "epoch": 0.630076117107287,
+      "grad_norm": 2.1906730624214235,
+      "learning_rate": 1.5348138934214493e-06,
+      "loss": 0.9742,
+      "step": 5991
+    },
+    {
+      "epoch": 0.6301812875491343,
+      "grad_norm": 2.1142812888162434,
+      "learning_rate": 1.5340438533064144e-06,
+      "loss": 0.9788,
+      "step": 5992
+    },
+    {
+      "epoch": 0.6302864579909816,
+      "grad_norm": 3.1567558738222607,
+      "learning_rate": 1.533273920902994e-06,
+      "loss": 0.9616,
+      "step": 5993
+    },
+    {
+      "epoch": 0.6303916284328289,
+      "grad_norm": 2.9030050091513866,
+      "learning_rate": 1.5325040962970417e-06,
+      "loss": 0.9866,
+      "step": 5994
+    },
+    {
+      "epoch": 0.6304967988746762,
+      "grad_norm": 2.7460052058389817,
+      "learning_rate": 1.5317343795743978e-06,
+      "loss": 0.9868,
+      "step": 5995
+    },
+    {
+      "epoch": 0.6306019693165236,
+      "grad_norm": 2.777830833328807,
+      "learning_rate": 1.5309647708208928e-06,
+      "loss": 1.0023,
+      "step": 5996
+    },
+    {
+      "epoch": 0.6307071397583709,
+      "grad_norm": 2.5120232310444104,
+      "learning_rate": 1.5301952701223422e-06,
+      "loss": 0.972,
+      "step": 5997
+    },
+    {
+      "epoch": 0.6308123102002182,
+      "grad_norm": 2.5363484810380212,
+      "learning_rate": 1.5294258775645545e-06,
+      "loss": 0.9653,
+      "step": 5998
+    },
+    {
+      "epoch": 0.6309174806420655,
+      "grad_norm": 2.3239370528340144,
+      "learning_rate": 1.5286565932333206e-06,
+      "loss": 0.9899,
+      "step": 5999
+    },
+    {
+      "epoch": 0.6310226510839129,
+      "grad_norm": 1.8488945840537439,
+      "learning_rate": 1.527887417214422e-06,
+      "loss": 0.9603,
+      "step": 6000
+    },
+    {
+      "epoch": 0.6311278215257602,
+      "grad_norm": 3.48906837044783,
+      "learning_rate": 1.5271183495936273e-06,
+      "loss": 1.0177,
+      "step": 6001
+    },
+    {
+      "epoch": 0.6312329919676075,
+      "grad_norm": 2.163852345437768,
+      "learning_rate": 1.526349390456695e-06,
+      "loss": 0.9964,
+      "step": 6002
+    },
+    {
+      "epoch": 0.6313381624094548,
+      "grad_norm": 2.5997209044493865,
+      "learning_rate": 1.5255805398893694e-06,
+      "loss": 1.0004,
+      "step": 6003
+    },
+    {
+      "epoch": 0.6314433328513022,
+      "grad_norm": 2.375034238569322,
+      "learning_rate": 1.524811797977383e-06,
+      "loss": 0.9733,
+      "step": 6004
+    },
+    {
+      "epoch": 0.6315485032931495,
+      "grad_norm": 2.496467895172032,
+      "learning_rate": 1.524043164806457e-06,
+      "loss": 1.0152,
+      "step": 6005
+    },
+    {
+      "epoch": 0.6316536737349968,
+      "grad_norm": 1.9885390550524185,
+      "learning_rate": 1.523274640462299e-06,
+      "loss": 0.9841,
+      "step": 6006
+    },
+    {
+      "epoch": 0.6317588441768441,
+      "grad_norm": 3.025548010215934,
+      "learning_rate": 1.5225062250306061e-06,
+      "loss": 1.0333,
+      "step": 6007
+    },
+    {
+      "epoch": 0.6318640146186915,
+      "grad_norm": 2.3886268665683668,
+      "learning_rate": 1.521737918597063e-06,
+      "loss": 0.9801,
+      "step": 6008
+    },
+    {
+      "epoch": 0.6319691850605388,
+      "grad_norm": 2.787511865407236,
+      "learning_rate": 1.5209697212473411e-06,
+      "loss": 0.9432,
+      "step": 6009
+    },
+    {
+      "epoch": 0.6320743555023861,
+      "grad_norm": 1.9376239960228194,
+      "learning_rate": 1.5202016330670998e-06,
+      "loss": 0.9643,
+      "step": 6010
+    },
+    {
+      "epoch": 0.6321795259442333,
+      "grad_norm": 2.5039360525755274,
+      "learning_rate": 1.5194336541419889e-06,
+      "loss": 1.0006,
+      "step": 6011
+    },
+    {
+      "epoch": 0.6322846963860806,
+      "grad_norm": 2.2219077385221486,
+      "learning_rate": 1.5186657845576428e-06,
+      "loss": 0.9957,
+      "step": 6012
+    },
+    {
+      "epoch": 0.632389866827928,
+      "grad_norm": 2.1820409541460544,
+      "learning_rate": 1.517898024399685e-06,
+      "loss": 1.001,
+      "step": 6013
+    },
+    {
+      "epoch": 0.6324950372697753,
+      "grad_norm": 3.0044306866211405,
+      "learning_rate": 1.5171303737537268e-06,
+      "loss": 1.0513,
+      "step": 6014
+    },
+    {
+      "epoch": 0.6326002077116226,
+      "grad_norm": 2.353451046725791,
+      "learning_rate": 1.5163628327053661e-06,
+      "loss": 0.9949,
+      "step": 6015
+    },
+    {
+      "epoch": 0.6327053781534699,
+      "grad_norm": 1.7590698297866734,
+      "learning_rate": 1.5155954013401916e-06,
+      "loss": 1.0115,
+      "step": 6016
+    },
+    {
+      "epoch": 0.6328105485953173,
+      "grad_norm": 2.903364189025722,
+      "learning_rate": 1.514828079743777e-06,
+      "loss": 1.0058,
+      "step": 6017
+    },
+    {
+      "epoch": 0.6329157190371646,
+      "grad_norm": 2.331986775898218,
+      "learning_rate": 1.5140608680016843e-06,
+      "loss": 0.9509,
+      "step": 6018
+    },
+    {
+      "epoch": 0.6330208894790119,
+      "grad_norm": 2.2153310954039225,
+      "learning_rate": 1.5132937661994642e-06,
+      "loss": 0.9977,
+      "step": 6019
+    },
+    {
+      "epoch": 0.6331260599208592,
+      "grad_norm": 2.5817333311290285,
+      "learning_rate": 1.5125267744226538e-06,
+      "loss": 1.0117,
+      "step": 6020
+    },
+    {
+      "epoch": 0.6332312303627066,
+      "grad_norm": 2.1972799748746903,
+      "learning_rate": 1.5117598927567791e-06,
+      "loss": 0.9816,
+      "step": 6021
+    },
+    {
+      "epoch": 0.6333364008045539,
+      "grad_norm": 2.501611715704273,
+      "learning_rate": 1.5109931212873535e-06,
+      "loss": 0.9624,
+      "step": 6022
+    },
+    {
+      "epoch": 0.6334415712464012,
+      "grad_norm": 3.048683807069792,
+      "learning_rate": 1.510226460099878e-06,
+      "loss": 1.0291,
+      "step": 6023
+    },
+    {
+      "epoch": 0.6335467416882485,
+      "grad_norm": 1.8577271918034934,
+      "learning_rate": 1.5094599092798396e-06,
+      "loss": 0.9834,
+      "step": 6024
+    },
+    {
+      "epoch": 0.6336519121300959,
+      "grad_norm": 2.360244200259059,
+      "learning_rate": 1.5086934689127173e-06,
+      "loss": 1.0152,
+      "step": 6025
+    },
+    {
+      "epoch": 0.6337570825719432,
+      "grad_norm": 3.3673054887669793,
+      "learning_rate": 1.5079271390839739e-06,
+      "loss": 1.0165,
+      "step": 6026
+    },
+    {
+      "epoch": 0.6338622530137905,
+      "grad_norm": 2.0209824979538773,
+      "learning_rate": 1.507160919879061e-06,
+      "loss": 1.0005,
+      "step": 6027
+    },
+    {
+      "epoch": 0.6339674234556378,
+      "grad_norm": 2.4114885968260156,
+      "learning_rate": 1.5063948113834184e-06,
+      "loss": 1.0082,
+      "step": 6028
+    },
+    {
+      "epoch": 0.6340725938974852,
+      "grad_norm": 2.185300297703002,
+      "learning_rate": 1.5056288136824725e-06,
+      "loss": 0.9766,
+      "step": 6029
+    },
+    {
+      "epoch": 0.6341777643393325,
+      "grad_norm": 2.0177323129193434,
+      "learning_rate": 1.5048629268616387e-06,
+      "loss": 0.9728,
+      "step": 6030
+    },
+    {
+      "epoch": 0.6342829347811797,
+      "grad_norm": 2.653274446710612,
+      "learning_rate": 1.5040971510063194e-06,
+      "loss": 0.9947,
+      "step": 6031
+    },
+    {
+      "epoch": 0.634388105223027,
+      "grad_norm": 2.579001884562927,
+      "learning_rate": 1.5033314862019038e-06,
+      "loss": 1.0079,
+      "step": 6032
+    },
+    {
+      "epoch": 0.6344932756648743,
+      "grad_norm": 2.5583814952866994,
+      "learning_rate": 1.5025659325337691e-06,
+      "loss": 1.0113,
+      "step": 6033
+    },
+    {
+      "epoch": 0.6345984461067217,
+      "grad_norm": 2.626887953110027,
+      "learning_rate": 1.5018004900872826e-06,
+      "loss": 1.0277,
+      "step": 6034
+    },
+    {
+      "epoch": 0.634703616548569,
+      "grad_norm": 2.1091067755253032,
+      "learning_rate": 1.5010351589477955e-06,
+      "loss": 0.9559,
+      "step": 6035
+    },
+    {
+      "epoch": 0.6348087869904163,
+      "grad_norm": 1.8200915464301803,
+      "learning_rate": 1.500269939200648e-06,
+      "loss": 0.9921,
+      "step": 6036
+    },
+    {
+      "epoch": 0.6349139574322636,
+      "grad_norm": 3.206419511627608,
+      "learning_rate": 1.4995048309311689e-06,
+      "loss": 1.0505,
+      "step": 6037
+    },
+    {
+      "epoch": 0.635019127874111,
+      "grad_norm": 3.2074930349618525,
+      "learning_rate": 1.4987398342246723e-06,
+      "loss": 0.9945,
+      "step": 6038
+    },
+    {
+      "epoch": 0.6351242983159583,
+      "grad_norm": 2.533006871994517,
+      "learning_rate": 1.4979749491664631e-06,
+      "loss": 0.9701,
+      "step": 6039
+    },
+    {
+      "epoch": 0.6352294687578056,
+      "grad_norm": 2.7185236711474596,
+      "learning_rate": 1.4972101758418307e-06,
+      "loss": 1.0024,
+      "step": 6040
+    },
+    {
+      "epoch": 0.6353346391996529,
+      "grad_norm": 2.946674199556197,
+      "learning_rate": 1.496445514336054e-06,
+      "loss": 0.9976,
+      "step": 6041
+    },
+    {
+      "epoch": 0.6354398096415003,
+      "grad_norm": 1.7428228667592487,
+      "learning_rate": 1.4956809647343984e-06,
+      "loss": 1.0117,
+      "step": 6042
+    },
+    {
+      "epoch": 0.6355449800833476,
+      "grad_norm": 2.3626314845775838,
+      "learning_rate": 1.494916527122116e-06,
+      "loss": 1.0219,
+      "step": 6043
+    },
+    {
+      "epoch": 0.6356501505251949,
+      "grad_norm": 2.6996995243978272,
+      "learning_rate": 1.4941522015844497e-06,
+      "loss": 0.9868,
+      "step": 6044
+    },
+    {
+      "epoch": 0.6357553209670422,
+      "grad_norm": 2.428347148699208,
+      "learning_rate": 1.4933879882066265e-06,
+      "loss": 0.9514,
+      "step": 6045
+    },
+    {
+      "epoch": 0.6358604914088896,
+      "grad_norm": 1.4275506316675006,
+      "learning_rate": 1.4926238870738624e-06,
+      "loss": 0.965,
+      "step": 6046
+    },
+    {
+      "epoch": 0.6359656618507369,
+      "grad_norm": 2.520049309698355,
+      "learning_rate": 1.4918598982713591e-06,
+      "loss": 1.0004,
+      "step": 6047
+    },
+    {
+      "epoch": 0.6360708322925842,
+      "grad_norm": 2.0394100491968654,
+      "learning_rate": 1.4910960218843099e-06,
+      "loss": 0.981,
+      "step": 6048
+    },
+    {
+      "epoch": 0.6361760027344315,
+      "grad_norm": 2.761031291865611,
+      "learning_rate": 1.4903322579978916e-06,
+      "loss": 0.9439,
+      "step": 6049
+    },
+    {
+      "epoch": 0.6362811731762789,
+      "grad_norm": 1.7561053210029363,
+      "learning_rate": 1.4895686066972703e-06,
+      "loss": 0.9956,
+      "step": 6050
+    },
+    {
+      "epoch": 0.6363863436181261,
+      "grad_norm": 1.662203724317949,
+      "learning_rate": 1.4888050680675983e-06,
+      "loss": 0.9915,
+      "step": 6051
+    },
+    {
+      "epoch": 0.6364915140599734,
+      "grad_norm": 2.6919552015061026,
+      "learning_rate": 1.4880416421940155e-06,
+      "loss": 1.0345,
+      "step": 6052
+    },
+    {
+      "epoch": 0.6365966845018207,
+      "grad_norm": 2.0994310916738197,
+      "learning_rate": 1.4872783291616516e-06,
+      "loss": 0.9567,
+      "step": 6053
+    },
+    {
+      "epoch": 0.636701854943668,
+      "grad_norm": 2.8279048616917497,
+      "learning_rate": 1.4865151290556213e-06,
+      "loss": 1.0237,
+      "step": 6054
+    },
+    {
+      "epoch": 0.6368070253855154,
+      "grad_norm": 1.997694763719727,
+      "learning_rate": 1.4857520419610269e-06,
+      "loss": 1.0362,
+      "step": 6055
+    },
+    {
+      "epoch": 0.6369121958273627,
+      "grad_norm": 2.8035596983277227,
+      "learning_rate": 1.4849890679629585e-06,
+      "loss": 1.0114,
+      "step": 6056
+    },
+    {
+      "epoch": 0.63701736626921,
+      "grad_norm": 3.067338117412502,
+      "learning_rate": 1.4842262071464941e-06,
+      "loss": 0.9986,
+      "step": 6057
+    },
+    {
+      "epoch": 0.6371225367110573,
+      "grad_norm": 2.258041094300344,
+      "learning_rate": 1.4834634595966987e-06,
+      "loss": 0.9978,
+      "step": 6058
+    },
+    {
+      "epoch": 0.6372277071529047,
+      "grad_norm": 1.7649409916767491,
+      "learning_rate": 1.4827008253986242e-06,
+      "loss": 0.9676,
+      "step": 6059
+    },
+    {
+      "epoch": 0.637332877594752,
+      "grad_norm": 2.156642745687781,
+      "learning_rate": 1.4819383046373103e-06,
+      "loss": 1.0234,
+      "step": 6060
+    },
+    {
+      "epoch": 0.6374380480365993,
+      "grad_norm": 1.9425720258392243,
+      "learning_rate": 1.481175897397783e-06,
+      "loss": 0.9994,
+      "step": 6061
+    },
+    {
+      "epoch": 0.6375432184784466,
+      "grad_norm": 1.747020487159967,
+      "learning_rate": 1.4804136037650587e-06,
+      "loss": 0.9864,
+      "step": 6062
+    },
+    {
+      "epoch": 0.637648388920294,
+      "grad_norm": 2.71636510704543,
+      "learning_rate": 1.4796514238241384e-06,
+      "loss": 1.0384,
+      "step": 6063
+    },
+    {
+      "epoch": 0.6377535593621413,
+      "grad_norm": 2.528649446654802,
+      "learning_rate": 1.47888935766001e-06,
+      "loss": 0.9637,
+      "step": 6064
+    },
+    {
+      "epoch": 0.6378587298039886,
+      "grad_norm": 2.2556064050120743,
+      "learning_rate": 1.4781274053576502e-06,
+      "loss": 1.008,
+      "step": 6065
+    },
+    {
+      "epoch": 0.6379639002458359,
+      "grad_norm": 1.750081160005925,
+      "learning_rate": 1.4773655670020235e-06,
+      "loss": 0.9957,
+      "step": 6066
+    },
+    {
+      "epoch": 0.6380690706876833,
+      "grad_norm": 2.250883027340165,
+      "learning_rate": 1.47660384267808e-06,
+      "loss": 0.9868,
+      "step": 6067
+    },
+    {
+      "epoch": 0.6381742411295306,
+      "grad_norm": 2.060415813689357,
+      "learning_rate": 1.4758422324707583e-06,
+      "loss": 0.9704,
+      "step": 6068
+    },
+    {
+      "epoch": 0.6382794115713779,
+      "grad_norm": 1.9461628426655384,
+      "learning_rate": 1.4750807364649833e-06,
+      "loss": 0.9545,
+      "step": 6069
+    },
+    {
+      "epoch": 0.6383845820132252,
+      "grad_norm": 2.1407118008400485,
+      "learning_rate": 1.474319354745668e-06,
+      "loss": 0.9842,
+      "step": 6070
+    },
+    {
+      "epoch": 0.6384897524550726,
+      "grad_norm": 2.469491976655806,
+      "learning_rate": 1.4735580873977125e-06,
+      "loss": 0.9548,
+      "step": 6071
+    },
+    {
+      "epoch": 0.6385949228969198,
+      "grad_norm": 2.115135764982275,
+      "learning_rate": 1.4727969345060041e-06,
+      "loss": 1.0031,
+      "step": 6072
+    },
+    {
+      "epoch": 0.6387000933387671,
+      "grad_norm": 3.4636565981184195,
+      "learning_rate": 1.472035896155417e-06,
+      "loss": 1.0143,
+      "step": 6073
+    },
+    {
+      "epoch": 0.6388052637806144,
+      "grad_norm": 2.4623855590153383,
+      "learning_rate": 1.4712749724308135e-06,
+      "loss": 1.0005,
+      "step": 6074
+    },
+    {
+      "epoch": 0.6389104342224617,
+      "grad_norm": 3.4935329001539603,
+      "learning_rate": 1.4705141634170402e-06,
+      "loss": 0.9797,
+      "step": 6075
+    },
+    {
+      "epoch": 0.6390156046643091,
+      "grad_norm": 2.981348278086113,
+      "learning_rate": 1.4697534691989362e-06,
+      "loss": 0.9747,
+      "step": 6076
+    },
+    {
+      "epoch": 0.6391207751061564,
+      "grad_norm": 2.3924446707785187,
+      "learning_rate": 1.468992889861324e-06,
+      "loss": 0.9882,
+      "step": 6077
+    },
+    {
+      "epoch": 0.6392259455480037,
+      "grad_norm": 2.23110800034559,
+      "learning_rate": 1.4682324254890135e-06,
+      "loss": 1.009,
+      "step": 6078
+    },
+    {
+      "epoch": 0.639331115989851,
+      "grad_norm": 2.468541532847342,
+      "learning_rate": 1.467472076166802e-06,
+      "loss": 0.9841,
+      "step": 6079
+    },
+    {
+      "epoch": 0.6394362864316984,
+      "grad_norm": 2.1517955193808342,
+      "learning_rate": 1.4667118419794756e-06,
+      "loss": 0.9967,
+      "step": 6080
+    },
+    {
+      "epoch": 0.6395414568735457,
+      "grad_norm": 2.5661813956881527,
+      "learning_rate": 1.4659517230118059e-06,
+      "loss": 0.9628,
+      "step": 6081
+    },
+    {
+      "epoch": 0.639646627315393,
+      "grad_norm": 2.181840326831563,
+      "learning_rate": 1.4651917193485516e-06,
+      "loss": 0.9237,
+      "step": 6082
+    },
+    {
+      "epoch": 0.6397517977572403,
+      "grad_norm": 2.1097264744955604,
+      "learning_rate": 1.4644318310744593e-06,
+      "loss": 1.0078,
+      "step": 6083
+    },
+    {
+      "epoch": 0.6398569681990877,
+      "grad_norm": 3.493695271673113,
+      "learning_rate": 1.4636720582742614e-06,
+      "loss": 0.9912,
+      "step": 6084
+    },
+    {
+      "epoch": 0.639962138640935,
+      "grad_norm": 1.9541985454696953,
+      "learning_rate": 1.4629124010326808e-06,
+      "loss": 0.9942,
+      "step": 6085
+    },
+    {
+      "epoch": 0.6400673090827823,
+      "grad_norm": 1.814035910685774,
+      "learning_rate": 1.4621528594344237e-06,
+      "loss": 1.0028,
+      "step": 6086
+    },
+    {
+      "epoch": 0.6401724795246296,
+      "grad_norm": 2.696801122775463,
+      "learning_rate": 1.461393433564185e-06,
+      "loss": 0.9912,
+      "step": 6087
+    },
+    {
+      "epoch": 0.640277649966477,
+      "grad_norm": 1.8963092165199675,
+      "learning_rate": 1.4606341235066452e-06,
+      "loss": 1.0165,
+      "step": 6088
+    },
+    {
+      "epoch": 0.6403828204083243,
+      "grad_norm": 3.669943765578645,
+      "learning_rate": 1.4598749293464763e-06,
+      "loss": 1.0253,
+      "step": 6089
+    },
+    {
+      "epoch": 0.6404879908501716,
+      "grad_norm": 2.8563817058680616,
+      "learning_rate": 1.459115851168333e-06,
+      "loss": 0.9864,
+      "step": 6090
+    },
+    {
+      "epoch": 0.6405931612920189,
+      "grad_norm": 3.028542769912519,
+      "learning_rate": 1.458356889056857e-06,
+      "loss": 0.9806,
+      "step": 6091
+    },
+    {
+      "epoch": 0.6406983317338661,
+      "grad_norm": 2.5809797491743653,
+      "learning_rate": 1.4575980430966808e-06,
+      "loss": 1.0162,
+      "step": 6092
+    },
+    {
+      "epoch": 0.6408035021757135,
+      "grad_norm": 2.040383893422859,
+      "learning_rate": 1.4568393133724185e-06,
+      "loss": 0.9846,
+      "step": 6093
+    },
+    {
+      "epoch": 0.6409086726175608,
+      "grad_norm": 2.4533017761010236,
+      "learning_rate": 1.4560806999686782e-06,
+      "loss": 1.0269,
+      "step": 6094
+    },
+    {
+      "epoch": 0.6410138430594081,
+      "grad_norm": 3.402665196483549,
+      "learning_rate": 1.4553222029700483e-06,
+      "loss": 1.0354,
+      "step": 6095
+    },
+    {
+      "epoch": 0.6411190135012554,
+      "grad_norm": 1.8564212218833198,
+      "learning_rate": 1.4545638224611091e-06,
+      "loss": 0.969,
+      "step": 6096
+    },
+    {
+      "epoch": 0.6412241839431028,
+      "grad_norm": 1.8685811138285715,
+      "learning_rate": 1.453805558526424e-06,
+      "loss": 0.9675,
+      "step": 6097
+    },
+    {
+      "epoch": 0.6413293543849501,
+      "grad_norm": 2.0516034951437496,
+      "learning_rate": 1.4530474112505462e-06,
+      "loss": 0.9577,
+      "step": 6098
+    },
+    {
+      "epoch": 0.6414345248267974,
+      "grad_norm": 2.632085890015541,
+      "learning_rate": 1.4522893807180163e-06,
+      "loss": 1.0097,
+      "step": 6099
+    },
+    {
+      "epoch": 0.6415396952686447,
+      "grad_norm": 2.3357227858145326,
+      "learning_rate": 1.4515314670133582e-06,
+      "loss": 0.99,
+      "step": 6100
+    },
+    {
+      "epoch": 0.6416448657104921,
+      "grad_norm": 2.031648424124929,
+      "learning_rate": 1.4507736702210872e-06,
+      "loss": 0.9714,
+      "step": 6101
+    },
+    {
+      "epoch": 0.6417500361523394,
+      "grad_norm": 2.9746248980463488,
+      "learning_rate": 1.4500159904257008e-06,
+      "loss": 0.995,
+      "step": 6102
+    },
+    {
+      "epoch": 0.6418552065941867,
+      "grad_norm": 2.4550674213700825,
+      "learning_rate": 1.4492584277116901e-06,
+      "loss": 1.0096,
+      "step": 6103
+    },
+    {
+      "epoch": 0.641960377036034,
+      "grad_norm": 2.5455704526475222,
+      "learning_rate": 1.4485009821635269e-06,
+      "loss": 0.9562,
+      "step": 6104
+    },
+    {
+      "epoch": 0.6420655474778814,
+      "grad_norm": 2.233105772105395,
+      "learning_rate": 1.4477436538656715e-06,
+      "loss": 0.9256,
+      "step": 6105
+    },
+    {
+      "epoch": 0.6421707179197287,
+      "grad_norm": 2.738769620298446,
+      "learning_rate": 1.446986442902574e-06,
+      "loss": 1.0144,
+      "step": 6106
+    },
+    {
+      "epoch": 0.642275888361576,
+      "grad_norm": 1.7815706347021845,
+      "learning_rate": 1.4462293493586662e-06,
+      "loss": 0.9717,
+      "step": 6107
+    },
+    {
+      "epoch": 0.6423810588034233,
+      "grad_norm": 2.0497666168497335,
+      "learning_rate": 1.445472373318374e-06,
+      "loss": 0.9561,
+      "step": 6108
+    },
+    {
+      "epoch": 0.6424862292452707,
+      "grad_norm": 2.6274997137270244,
+      "learning_rate": 1.444715514866103e-06,
+      "loss": 1.0196,
+      "step": 6109
+    },
+    {
+      "epoch": 0.642591399687118,
+      "grad_norm": 3.440273225742282,
+      "learning_rate": 1.443958774086251e-06,
+      "loss": 1.0102,
+      "step": 6110
+    },
+    {
+      "epoch": 0.6426965701289653,
+      "grad_norm": 2.483691809703666,
+      "learning_rate": 1.443202151063198e-06,
+      "loss": 1.0265,
+      "step": 6111
+    },
+    {
+      "epoch": 0.6428017405708125,
+      "grad_norm": 2.447339258191458,
+      "learning_rate": 1.4424456458813147e-06,
+      "loss": 1.0198,
+      "step": 6112
+    },
+    {
+      "epoch": 0.6429069110126598,
+      "grad_norm": 2.47035563695523,
+      "learning_rate": 1.4416892586249586e-06,
+      "loss": 1.0044,
+      "step": 6113
+    },
+    {
+      "epoch": 0.6430120814545072,
+      "grad_norm": 2.1915716837499333,
+      "learning_rate": 1.4409329893784702e-06,
+      "loss": 0.9784,
+      "step": 6114
+    },
+    {
+      "epoch": 0.6431172518963545,
+      "grad_norm": 2.2232422956545825,
+      "learning_rate": 1.4401768382261813e-06,
+      "loss": 1.0041,
+      "step": 6115
+    },
+    {
+      "epoch": 0.6432224223382018,
+      "grad_norm": 2.022035643282418,
+      "learning_rate": 1.4394208052524062e-06,
+      "loss": 1.0323,
+      "step": 6116
+    },
+    {
+      "epoch": 0.6433275927800491,
+      "grad_norm": 2.1999594819259496,
+      "learning_rate": 1.4386648905414525e-06,
+      "loss": 0.9813,
+      "step": 6117
+    },
+    {
+      "epoch": 0.6434327632218965,
+      "grad_norm": 2.4472550369492936,
+      "learning_rate": 1.4379090941776067e-06,
+      "loss": 0.9525,
+      "step": 6118
+    },
+    {
+      "epoch": 0.6435379336637438,
+      "grad_norm": 2.4214125950190817,
+      "learning_rate": 1.4371534162451487e-06,
+      "loss": 0.9848,
+      "step": 6119
+    },
+    {
+      "epoch": 0.6436431041055911,
+      "grad_norm": 2.6024268782469013,
+      "learning_rate": 1.4363978568283412e-06,
+      "loss": 1.0264,
+      "step": 6120
+    },
+    {
+      "epoch": 0.6437482745474384,
+      "grad_norm": 2.842563298958492,
+      "learning_rate": 1.4356424160114332e-06,
+      "loss": 0.9535,
+      "step": 6121
+    },
+    {
+      "epoch": 0.6438534449892858,
+      "grad_norm": 2.086520974335246,
+      "learning_rate": 1.4348870938786657e-06,
+      "loss": 0.9933,
+      "step": 6122
+    },
+    {
+      "epoch": 0.6439586154311331,
+      "grad_norm": 2.6018674216608306,
+      "learning_rate": 1.434131890514261e-06,
+      "loss": 0.9847,
+      "step": 6123
+    },
+    {
+      "epoch": 0.6440637858729804,
+      "grad_norm": 1.7399725974174627,
+      "learning_rate": 1.4333768060024308e-06,
+      "loss": 0.9887,
+      "step": 6124
+    },
+    {
+      "epoch": 0.6441689563148277,
+      "grad_norm": 2.33460686855743,
+      "learning_rate": 1.4326218404273718e-06,
+      "loss": 1.0107,
+      "step": 6125
+    },
+    {
+      "epoch": 0.6442741267566751,
+      "grad_norm": 2.304286713750231,
+      "learning_rate": 1.4318669938732694e-06,
+      "loss": 0.9888,
+      "step": 6126
+    },
+    {
+      "epoch": 0.6443792971985224,
+      "grad_norm": 2.617604600587451,
+      "learning_rate": 1.4311122664242955e-06,
+      "loss": 0.9804,
+      "step": 6127
+    },
+    {
+      "epoch": 0.6444844676403697,
+      "grad_norm": 2.884591813362857,
+      "learning_rate": 1.430357658164606e-06,
+      "loss": 0.9886,
+      "step": 6128
+    },
+    {
+      "epoch": 0.644589638082217,
+      "grad_norm": 2.437029381096377,
+      "learning_rate": 1.4296031691783485e-06,
+      "loss": 1.0295,
+      "step": 6129
+    },
+    {
+      "epoch": 0.6446948085240644,
+      "grad_norm": 2.9913792325943303,
+      "learning_rate": 1.4288487995496508e-06,
+      "loss": 0.9888,
+      "step": 6130
+    },
+    {
+      "epoch": 0.6447999789659117,
+      "grad_norm": 2.4797754234444715,
+      "learning_rate": 1.4280945493626347e-06,
+      "loss": 0.9409,
+      "step": 6131
+    },
+    {
+      "epoch": 0.644905149407759,
+      "grad_norm": 3.0033762892288545,
+      "learning_rate": 1.427340418701402e-06,
+      "loss": 1.0166,
+      "step": 6132
+    },
+    {
+      "epoch": 0.6450103198496062,
+      "grad_norm": 2.37731076377192,
+      "learning_rate": 1.4265864076500465e-06,
+      "loss": 0.963,
+      "step": 6133
+    },
+    {
+      "epoch": 0.6451154902914535,
+      "grad_norm": 2.6651860260554954,
+      "learning_rate": 1.4258325162926441e-06,
+      "loss": 1.0274,
+      "step": 6134
+    },
+    {
+      "epoch": 0.6452206607333009,
+      "grad_norm": 2.9869829981562384,
+      "learning_rate": 1.4250787447132607e-06,
+      "loss": 1.0177,
+      "step": 6135
+    },
+    {
+      "epoch": 0.6453258311751482,
+      "grad_norm": 2.170891852808566,
+      "learning_rate": 1.4243250929959484e-06,
+      "loss": 0.982,
+      "step": 6136
+    },
+    {
+      "epoch": 0.6454310016169955,
+      "grad_norm": 2.170159054445325,
+      "learning_rate": 1.4235715612247435e-06,
+      "loss": 1.0116,
+      "step": 6137
+    },
+    {
+      "epoch": 0.6455361720588428,
+      "grad_norm": 2.4494534962313868,
+      "learning_rate": 1.4228181494836724e-06,
+      "loss": 0.959,
+      "step": 6138
+    },
+    {
+      "epoch": 0.6456413425006902,
+      "grad_norm": 1.7490013283349262,
+      "learning_rate": 1.4220648578567444e-06,
+      "loss": 1.0183,
+      "step": 6139
+    },
+    {
+      "epoch": 0.6457465129425375,
+      "grad_norm": 2.8377357573800515,
+      "learning_rate": 1.4213116864279586e-06,
+      "loss": 0.977,
+      "step": 6140
+    },
+    {
+      "epoch": 0.6458516833843848,
+      "grad_norm": 1.933344788178783,
+      "learning_rate": 1.4205586352813e-06,
+      "loss": 0.9859,
+      "step": 6141
+    },
+    {
+      "epoch": 0.6459568538262321,
+      "grad_norm": 2.280844764369398,
+      "learning_rate": 1.4198057045007384e-06,
+      "loss": 1.0191,
+      "step": 6142
+    },
+    {
+      "epoch": 0.6460620242680795,
+      "grad_norm": 2.2048132233498654,
+      "learning_rate": 1.4190528941702328e-06,
+      "loss": 1.0413,
+      "step": 6143
+    },
+    {
+      "epoch": 0.6461671947099268,
+      "grad_norm": 2.4477864525484363,
+      "learning_rate": 1.4183002043737246e-06,
+      "loss": 1.0105,
+      "step": 6144
+    },
+    {
+      "epoch": 0.6462723651517741,
+      "grad_norm": 2.3566748243820625,
+      "learning_rate": 1.4175476351951484e-06,
+      "loss": 0.9748,
+      "step": 6145
+    },
+    {
+      "epoch": 0.6463775355936214,
+      "grad_norm": 2.636120161039787,
+      "learning_rate": 1.4167951867184187e-06,
+      "loss": 0.984,
+      "step": 6146
+    },
+    {
+      "epoch": 0.6464827060354688,
+      "grad_norm": 2.367508226857458,
+      "learning_rate": 1.4160428590274416e-06,
+      "loss": 0.9811,
+      "step": 6147
+    },
+    {
+      "epoch": 0.6465878764773161,
+      "grad_norm": 2.2275791479433167,
+      "learning_rate": 1.415290652206105e-06,
+      "loss": 0.9852,
+      "step": 6148
+    },
+    {
+      "epoch": 0.6466930469191634,
+      "grad_norm": 2.9242032394552138,
+      "learning_rate": 1.414538566338287e-06,
+      "loss": 1.0,
+      "step": 6149
+    },
+    {
+      "epoch": 0.6467982173610107,
+      "grad_norm": 2.200563913281183,
+      "learning_rate": 1.4137866015078523e-06,
+      "loss": 0.9766,
+      "step": 6150
+    },
+    {
+      "epoch": 0.646903387802858,
+      "grad_norm": 2.136169934182686,
+      "learning_rate": 1.4130347577986481e-06,
+      "loss": 0.9891,
+      "step": 6151
+    },
+    {
+      "epoch": 0.6470085582447054,
+      "grad_norm": 2.4522731115485916,
+      "learning_rate": 1.4122830352945133e-06,
+      "loss": 0.9605,
+      "step": 6152
+    },
+    {
+      "epoch": 0.6471137286865526,
+      "grad_norm": 1.8397790333899535,
+      "learning_rate": 1.411531434079268e-06,
+      "loss": 1.0013,
+      "step": 6153
+    },
+    {
+      "epoch": 0.6472188991283999,
+      "grad_norm": 2.2803600413168237,
+      "learning_rate": 1.410779954236725e-06,
+      "loss": 0.9917,
+      "step": 6154
+    },
+    {
+      "epoch": 0.6473240695702472,
+      "grad_norm": 2.8291558323999912,
+      "learning_rate": 1.4100285958506785e-06,
+      "loss": 1.0154,
+      "step": 6155
+    },
+    {
+      "epoch": 0.6474292400120946,
+      "grad_norm": 2.737018049431013,
+      "learning_rate": 1.4092773590049098e-06,
+      "loss": 0.9962,
+      "step": 6156
+    },
+    {
+      "epoch": 0.6475344104539419,
+      "grad_norm": 2.5922522490787583,
+      "learning_rate": 1.4085262437831886e-06,
+      "loss": 0.9314,
+      "step": 6157
+    },
+    {
+      "epoch": 0.6476395808957892,
+      "grad_norm": 2.1557844949857463,
+      "learning_rate": 1.4077752502692704e-06,
+      "loss": 0.9797,
+      "step": 6158
+    },
+    {
+      "epoch": 0.6477447513376365,
+      "grad_norm": 2.208795077852058,
+      "learning_rate": 1.4070243785468974e-06,
+      "loss": 0.9637,
+      "step": 6159
+    },
+    {
+      "epoch": 0.6478499217794839,
+      "grad_norm": 2.7684955688133965,
+      "learning_rate": 1.4062736286997952e-06,
+      "loss": 0.9932,
+      "step": 6160
+    },
+    {
+      "epoch": 0.6479550922213312,
+      "grad_norm": 2.7322932255125925,
+      "learning_rate": 1.4055230008116813e-06,
+      "loss": 0.9904,
+      "step": 6161
+    },
+    {
+      "epoch": 0.6480602626631785,
+      "grad_norm": 2.64744883621316,
+      "learning_rate": 1.404772494966254e-06,
+      "loss": 1.0088,
+      "step": 6162
+    },
+    {
+      "epoch": 0.6481654331050258,
+      "grad_norm": 2.5697944667740513,
+      "learning_rate": 1.4040221112472014e-06,
+      "loss": 0.9693,
+      "step": 6163
+    },
+    {
+      "epoch": 0.6482706035468732,
+      "grad_norm": 2.108197664933067,
+      "learning_rate": 1.4032718497381981e-06,
+      "loss": 1.032,
+      "step": 6164
+    },
+    {
+      "epoch": 0.6483757739887205,
+      "grad_norm": 3.875251629143974,
+      "learning_rate": 1.4025217105229021e-06,
+      "loss": 1.0005,
+      "step": 6165
+    },
+    {
+      "epoch": 0.6484809444305678,
+      "grad_norm": 2.3973712746700664,
+      "learning_rate": 1.4017716936849623e-06,
+      "loss": 0.9738,
+      "step": 6166
+    },
+    {
+      "epoch": 0.6485861148724151,
+      "grad_norm": 2.2650362281817866,
+      "learning_rate": 1.4010217993080076e-06,
+      "loss": 0.9662,
+      "step": 6167
+    },
+    {
+      "epoch": 0.6486912853142625,
+      "grad_norm": 2.1269159039754277,
+      "learning_rate": 1.400272027475662e-06,
+      "loss": 1.0022,
+      "step": 6168
+    },
+    {
+      "epoch": 0.6487964557561098,
+      "grad_norm": 2.177929012659701,
+      "learning_rate": 1.399522378271527e-06,
+      "loss": 0.9988,
+      "step": 6169
+    },
+    {
+      "epoch": 0.6489016261979571,
+      "grad_norm": 2.684842639627906,
+      "learning_rate": 1.3987728517791966e-06,
+      "loss": 0.9721,
+      "step": 6170
+    },
+    {
+      "epoch": 0.6490067966398044,
+      "grad_norm": 2.838066575594318,
+      "learning_rate": 1.3980234480822468e-06,
+      "loss": 0.9978,
+      "step": 6171
+    },
+    {
+      "epoch": 0.6491119670816518,
+      "grad_norm": 2.6540051552459465,
+      "learning_rate": 1.397274167264243e-06,
+      "loss": 0.9959,
+      "step": 6172
+    },
+    {
+      "epoch": 0.649217137523499,
+      "grad_norm": 2.4639663876901814,
+      "learning_rate": 1.3965250094087373e-06,
+      "loss": 0.9836,
+      "step": 6173
+    },
+    {
+      "epoch": 0.6493223079653463,
+      "grad_norm": 2.447427698722012,
+      "learning_rate": 1.3957759745992637e-06,
+      "loss": 0.9602,
+      "step": 6174
+    },
+    {
+      "epoch": 0.6494274784071936,
+      "grad_norm": 2.2679043767007983,
+      "learning_rate": 1.395027062919348e-06,
+      "loss": 1.0334,
+      "step": 6175
+    },
+    {
+      "epoch": 0.6495326488490409,
+      "grad_norm": 3.5789206773428317,
+      "learning_rate": 1.3942782744524974e-06,
+      "loss": 1.0175,
+      "step": 6176
+    },
+    {
+      "epoch": 0.6496378192908883,
+      "grad_norm": 2.729641701499508,
+      "learning_rate": 1.3935296092822087e-06,
+      "loss": 1.0013,
+      "step": 6177
+    },
+    {
+      "epoch": 0.6497429897327356,
+      "grad_norm": 2.3392616870926886,
+      "learning_rate": 1.392781067491965e-06,
+      "loss": 0.9527,
+      "step": 6178
+    },
+    {
+      "epoch": 0.6498481601745829,
+      "grad_norm": 2.969259025385436,
+      "learning_rate": 1.3920326491652325e-06,
+      "loss": 0.9911,
+      "step": 6179
+    },
+    {
+      "epoch": 0.6499533306164302,
+      "grad_norm": 2.055601110395823,
+      "learning_rate": 1.3912843543854664e-06,
+      "loss": 0.9653,
+      "step": 6180
+    },
+    {
+      "epoch": 0.6500585010582776,
+      "grad_norm": 2.808238408960534,
+      "learning_rate": 1.3905361832361078e-06,
+      "loss": 0.9675,
+      "step": 6181
+    },
+    {
+      "epoch": 0.6501636715001249,
+      "grad_norm": 2.2954725127353854,
+      "learning_rate": 1.3897881358005843e-06,
+      "loss": 0.9861,
+      "step": 6182
+    },
+    {
+      "epoch": 0.6502688419419722,
+      "grad_norm": 2.4627639209709113,
+      "learning_rate": 1.389040212162307e-06,
+      "loss": 0.9841,
+      "step": 6183
+    },
+    {
+      "epoch": 0.6503740123838195,
+      "grad_norm": 3.075750096193752,
+      "learning_rate": 1.3882924124046775e-06,
+      "loss": 1.0039,
+      "step": 6184
+    },
+    {
+      "epoch": 0.6504791828256669,
+      "grad_norm": 2.715707058368657,
+      "learning_rate": 1.387544736611079e-06,
+      "loss": 0.9638,
+      "step": 6185
+    },
+    {
+      "epoch": 0.6505843532675142,
+      "grad_norm": 1.977183746259915,
+      "learning_rate": 1.3867971848648843e-06,
+      "loss": 0.9912,
+      "step": 6186
+    },
+    {
+      "epoch": 0.6506895237093615,
+      "grad_norm": 2.7109584352412353,
+      "learning_rate": 1.386049757249452e-06,
+      "loss": 1.016,
+      "step": 6187
+    },
+    {
+      "epoch": 0.6507946941512088,
+      "grad_norm": 2.56279869947222,
+      "learning_rate": 1.3853024538481241e-06,
+      "loss": 0.9732,
+      "step": 6188
+    },
+    {
+      "epoch": 0.6508998645930562,
+      "grad_norm": 2.1123247416434574,
+      "learning_rate": 1.384555274744233e-06,
+      "loss": 1.0225,
+      "step": 6189
+    },
+    {
+      "epoch": 0.6510050350349035,
+      "grad_norm": 3.0471903006110517,
+      "learning_rate": 1.3838082200210932e-06,
+      "loss": 0.9687,
+      "step": 6190
+    },
+    {
+      "epoch": 0.6511102054767508,
+      "grad_norm": 2.272404748311318,
+      "learning_rate": 1.3830612897620072e-06,
+      "loss": 1.004,
+      "step": 6191
+    },
+    {
+      "epoch": 0.6512153759185981,
+      "grad_norm": 2.2705689260324875,
+      "learning_rate": 1.3823144840502656e-06,
+      "loss": 0.9655,
+      "step": 6192
+    },
+    {
+      "epoch": 0.6513205463604455,
+      "grad_norm": 2.549896761841074,
+      "learning_rate": 1.3815678029691399e-06,
+      "loss": 0.951,
+      "step": 6193
+    },
+    {
+      "epoch": 0.6514257168022927,
+      "grad_norm": 2.7098082789675817,
+      "learning_rate": 1.3808212466018927e-06,
+      "loss": 0.9638,
+      "step": 6194
+    },
+    {
+      "epoch": 0.65153088724414,
+      "grad_norm": 2.473211401630005,
+      "learning_rate": 1.3800748150317709e-06,
+      "loss": 0.9972,
+      "step": 6195
+    },
+    {
+      "epoch": 0.6516360576859873,
+      "grad_norm": 2.3649286529172047,
+      "learning_rate": 1.3793285083420077e-06,
+      "loss": 0.991,
+      "step": 6196
+    },
+    {
+      "epoch": 0.6517412281278346,
+      "grad_norm": 2.074820659875855,
+      "learning_rate": 1.378582326615821e-06,
+      "loss": 0.9987,
+      "step": 6197
+    },
+    {
+      "epoch": 0.651846398569682,
+      "grad_norm": 2.2908355538023453,
+      "learning_rate": 1.3778362699364167e-06,
+      "loss": 0.8914,
+      "step": 6198
+    },
+    {
+      "epoch": 0.6519515690115293,
+      "grad_norm": 2.495491364615591,
+      "learning_rate": 1.377090338386985e-06,
+      "loss": 0.9856,
+      "step": 6199
+    },
+    {
+      "epoch": 0.6520567394533766,
+      "grad_norm": 1.941710625750233,
+      "learning_rate": 1.3763445320507034e-06,
+      "loss": 1.0263,
+      "step": 6200
+    },
+    {
+      "epoch": 0.6521619098952239,
+      "grad_norm": 2.2537782874692978,
+      "learning_rate": 1.3755988510107365e-06,
+      "loss": 0.9837,
+      "step": 6201
+    },
+    {
+      "epoch": 0.6522670803370713,
+      "grad_norm": 2.14058569239182,
+      "learning_rate": 1.3748532953502317e-06,
+      "loss": 0.9417,
+      "step": 6202
+    },
+    {
+      "epoch": 0.6523722507789186,
+      "grad_norm": 2.2055386916726043,
+      "learning_rate": 1.3741078651523242e-06,
+      "loss": 0.9859,
+      "step": 6203
+    },
+    {
+      "epoch": 0.6524774212207659,
+      "grad_norm": 2.59614689675084,
+      "learning_rate": 1.3733625605001365e-06,
+      "loss": 0.9907,
+      "step": 6204
+    },
+    {
+      "epoch": 0.6525825916626132,
+      "grad_norm": 2.4124820848680453,
+      "learning_rate": 1.3726173814767763e-06,
+      "loss": 1.0189,
+      "step": 6205
+    },
+    {
+      "epoch": 0.6526877621044606,
+      "grad_norm": 2.350472903283974,
+      "learning_rate": 1.3718723281653357e-06,
+      "loss": 0.9483,
+      "step": 6206
+    },
+    {
+      "epoch": 0.6527929325463079,
+      "grad_norm": 2.728569075199262,
+      "learning_rate": 1.3711274006488935e-06,
+      "loss": 1.0009,
+      "step": 6207
+    },
+    {
+      "epoch": 0.6528981029881552,
+      "grad_norm": 2.1797869690723095,
+      "learning_rate": 1.370382599010515e-06,
+      "loss": 0.9916,
+      "step": 6208
+    },
+    {
+      "epoch": 0.6530032734300025,
+      "grad_norm": 2.34014133371477,
+      "learning_rate": 1.3696379233332518e-06,
+      "loss": 0.9545,
+      "step": 6209
+    },
+    {
+      "epoch": 0.6531084438718499,
+      "grad_norm": 2.760835763478809,
+      "learning_rate": 1.3688933737001425e-06,
+      "loss": 0.9907,
+      "step": 6210
+    },
+    {
+      "epoch": 0.6532136143136972,
+      "grad_norm": 2.0489131366244666,
+      "learning_rate": 1.3681489501942077e-06,
+      "loss": 0.9922,
+      "step": 6211
+    },
+    {
+      "epoch": 0.6533187847555445,
+      "grad_norm": 2.74268259260636,
+      "learning_rate": 1.3674046528984576e-06,
+      "loss": 0.9946,
+      "step": 6212
+    },
+    {
+      "epoch": 0.6534239551973918,
+      "grad_norm": 2.3837332766442016,
+      "learning_rate": 1.3666604818958878e-06,
+      "loss": 0.9978,
+      "step": 6213
+    },
+    {
+      "epoch": 0.653529125639239,
+      "grad_norm": 2.286480330832379,
+      "learning_rate": 1.3659164372694771e-06,
+      "loss": 1.0212,
+      "step": 6214
+    },
+    {
+      "epoch": 0.6536342960810864,
+      "grad_norm": 3.110138767109496,
+      "learning_rate": 1.365172519102195e-06,
+      "loss": 1.0182,
+      "step": 6215
+    },
+    {
+      "epoch": 0.6537394665229337,
+      "grad_norm": 2.3763826353787456,
+      "learning_rate": 1.3644287274769915e-06,
+      "loss": 0.9838,
+      "step": 6216
+    },
+    {
+      "epoch": 0.653844636964781,
+      "grad_norm": 2.147543965542234,
+      "learning_rate": 1.3636850624768065e-06,
+      "loss": 1.065,
+      "step": 6217
+    },
+    {
+      "epoch": 0.6539498074066283,
+      "grad_norm": 1.9128638684889196,
+      "learning_rate": 1.362941524184564e-06,
+      "loss": 0.99,
+      "step": 6218
+    },
+    {
+      "epoch": 0.6540549778484757,
+      "grad_norm": 2.701396037832273,
+      "learning_rate": 1.3621981126831755e-06,
+      "loss": 1.0051,
+      "step": 6219
+    },
+    {
+      "epoch": 0.654160148290323,
+      "grad_norm": 2.8040498677343324,
+      "learning_rate": 1.3614548280555351e-06,
+      "loss": 1.0138,
+      "step": 6220
+    },
+    {
+      "epoch": 0.6542653187321703,
+      "grad_norm": 2.0771737084674466,
+      "learning_rate": 1.3607116703845273e-06,
+      "loss": 1.0207,
+      "step": 6221
+    },
+    {
+      "epoch": 0.6543704891740176,
+      "grad_norm": 1.906154408918418,
+      "learning_rate": 1.3599686397530171e-06,
+      "loss": 1.0037,
+      "step": 6222
+    },
+    {
+      "epoch": 0.654475659615865,
+      "grad_norm": 2.6242007085780337,
+      "learning_rate": 1.35922573624386e-06,
+      "loss": 0.9953,
+      "step": 6223
+    },
+    {
+      "epoch": 0.6545808300577123,
+      "grad_norm": 2.536900445914889,
+      "learning_rate": 1.3584829599398958e-06,
+      "loss": 0.9774,
+      "step": 6224
+    },
+    {
+      "epoch": 0.6546860004995596,
+      "grad_norm": 2.3679575872686933,
+      "learning_rate": 1.3577403109239485e-06,
+      "loss": 0.9633,
+      "step": 6225
+    },
+    {
+      "epoch": 0.6547911709414069,
+      "grad_norm": 2.7750626080040086,
+      "learning_rate": 1.35699778927883e-06,
+      "loss": 0.9824,
+      "step": 6226
+    },
+    {
+      "epoch": 0.6548963413832543,
+      "grad_norm": 2.2512171030808386,
+      "learning_rate": 1.3562553950873377e-06,
+      "loss": 0.9846,
+      "step": 6227
+    },
+    {
+      "epoch": 0.6550015118251016,
+      "grad_norm": 3.2875448840376817,
+      "learning_rate": 1.3555131284322532e-06,
+      "loss": 0.9589,
+      "step": 6228
+    },
+    {
+      "epoch": 0.6551066822669489,
+      "grad_norm": 2.026779321691147,
+      "learning_rate": 1.3547709893963462e-06,
+      "loss": 0.9613,
+      "step": 6229
+    },
+    {
+      "epoch": 0.6552118527087962,
+      "grad_norm": 2.4724040803859197,
+      "learning_rate": 1.3540289780623697e-06,
+      "loss": 0.985,
+      "step": 6230
+    },
+    {
+      "epoch": 0.6553170231506436,
+      "grad_norm": 2.210229193121631,
+      "learning_rate": 1.3532870945130642e-06,
+      "loss": 1.0203,
+      "step": 6231
+    },
+    {
+      "epoch": 0.6554221935924909,
+      "grad_norm": 2.7088718437080894,
+      "learning_rate": 1.3525453388311554e-06,
+      "loss": 0.9994,
+      "step": 6232
+    },
+    {
+      "epoch": 0.6555273640343382,
+      "grad_norm": 2.6674805083220607,
+      "learning_rate": 1.3518037110993565e-06,
+      "loss": 1.0023,
+      "step": 6233
+    },
+    {
+      "epoch": 0.6556325344761854,
+      "grad_norm": 2.4368790272797347,
+      "learning_rate": 1.3510622114003619e-06,
+      "loss": 1.024,
+      "step": 6234
+    },
+    {
+      "epoch": 0.6557377049180327,
+      "grad_norm": 2.355980409234641,
+      "learning_rate": 1.3503208398168567e-06,
+      "loss": 1.0132,
+      "step": 6235
+    },
+    {
+      "epoch": 0.6558428753598801,
+      "grad_norm": 2.617783917085451,
+      "learning_rate": 1.349579596431509e-06,
+      "loss": 1.0071,
+      "step": 6236
+    },
+    {
+      "epoch": 0.6559480458017274,
+      "grad_norm": 1.908042391800999,
+      "learning_rate": 1.3488384813269726e-06,
+      "loss": 0.9537,
+      "step": 6237
+    },
+    {
+      "epoch": 0.6560532162435747,
+      "grad_norm": 2.4517991712108955,
+      "learning_rate": 1.3480974945858887e-06,
+      "loss": 0.9943,
+      "step": 6238
+    },
+    {
+      "epoch": 0.656158386685422,
+      "grad_norm": 2.125250466725107,
+      "learning_rate": 1.347356636290882e-06,
+      "loss": 0.9884,
+      "step": 6239
+    },
+    {
+      "epoch": 0.6562635571272694,
+      "grad_norm": 2.9442428748019256,
+      "learning_rate": 1.3466159065245637e-06,
+      "loss": 0.9988,
+      "step": 6240
+    },
+    {
+      "epoch": 0.6563687275691167,
+      "grad_norm": 2.1219839011084813,
+      "learning_rate": 1.3458753053695332e-06,
+      "loss": 0.9689,
+      "step": 6241
+    },
+    {
+      "epoch": 0.656473898010964,
+      "grad_norm": 2.175678143174604,
+      "learning_rate": 1.3451348329083702e-06,
+      "loss": 0.9788,
+      "step": 6242
+    },
+    {
+      "epoch": 0.6565790684528113,
+      "grad_norm": 1.8922658899883003,
+      "learning_rate": 1.344394489223646e-06,
+      "loss": 0.9898,
+      "step": 6243
+    },
+    {
+      "epoch": 0.6566842388946587,
+      "grad_norm": 2.6214577423300742,
+      "learning_rate": 1.3436542743979125e-06,
+      "loss": 0.9652,
+      "step": 6244
+    },
+    {
+      "epoch": 0.656789409336506,
+      "grad_norm": 2.105636623831596,
+      "learning_rate": 1.3429141885137097e-06,
+      "loss": 0.9862,
+      "step": 6245
+    },
+    {
+      "epoch": 0.6568945797783533,
+      "grad_norm": 1.9675801964204271,
+      "learning_rate": 1.342174231653564e-06,
+      "loss": 0.936,
+      "step": 6246
+    },
+    {
+      "epoch": 0.6569997502202006,
+      "grad_norm": 2.1697661785388473,
+      "learning_rate": 1.3414344038999862e-06,
+      "loss": 0.9785,
+      "step": 6247
+    },
+    {
+      "epoch": 0.657104920662048,
+      "grad_norm": 2.187917791791845,
+      "learning_rate": 1.340694705335472e-06,
+      "loss": 1.0148,
+      "step": 6248
+    },
+    {
+      "epoch": 0.6572100911038953,
+      "grad_norm": 3.026719311049528,
+      "learning_rate": 1.3399551360425035e-06,
+      "loss": 0.9711,
+      "step": 6249
+    },
+    {
+      "epoch": 0.6573152615457426,
+      "grad_norm": 1.8530024264442755,
+      "learning_rate": 1.33921569610355e-06,
+      "loss": 0.9588,
+      "step": 6250
+    },
+    {
+      "epoch": 0.6574204319875899,
+      "grad_norm": 2.1627719960544014,
+      "learning_rate": 1.3384763856010626e-06,
+      "loss": 0.9773,
+      "step": 6251
+    },
+    {
+      "epoch": 0.6575256024294373,
+      "grad_norm": 2.049389943315943,
+      "learning_rate": 1.3377372046174826e-06,
+      "loss": 0.9599,
+      "step": 6252
+    },
+    {
+      "epoch": 0.6576307728712846,
+      "grad_norm": 2.0628952108415333,
+      "learning_rate": 1.3369981532352317e-06,
+      "loss": 1.0007,
+      "step": 6253
+    },
+    {
+      "epoch": 0.6577359433131319,
+      "grad_norm": 2.491700162689158,
+      "learning_rate": 1.3362592315367212e-06,
+      "loss": 0.9925,
+      "step": 6254
+    },
+    {
+      "epoch": 0.6578411137549791,
+      "grad_norm": 2.2480692531605677,
+      "learning_rate": 1.3355204396043467e-06,
+      "loss": 0.9865,
+      "step": 6255
+    },
+    {
+      "epoch": 0.6579462841968264,
+      "grad_norm": 2.1560294381554073,
+      "learning_rate": 1.3347817775204903e-06,
+      "loss": 0.9463,
+      "step": 6256
+    },
+    {
+      "epoch": 0.6580514546386738,
+      "grad_norm": 2.1594218843164397,
+      "learning_rate": 1.3340432453675173e-06,
+      "loss": 0.9732,
+      "step": 6257
+    },
+    {
+      "epoch": 0.6581566250805211,
+      "grad_norm": 2.3800077094356666,
+      "learning_rate": 1.3333048432277779e-06,
+      "loss": 1.0126,
+      "step": 6258
+    },
+    {
+      "epoch": 0.6582617955223684,
+      "grad_norm": 2.109956422004809,
+      "learning_rate": 1.3325665711836138e-06,
+      "loss": 0.9533,
+      "step": 6259
+    },
+    {
+      "epoch": 0.6583669659642157,
+      "grad_norm": 3.0683493261098493,
+      "learning_rate": 1.331828429317345e-06,
+      "loss": 0.9883,
+      "step": 6260
+    },
+    {
+      "epoch": 0.6584721364060631,
+      "grad_norm": 2.686877306764361,
+      "learning_rate": 1.3310904177112819e-06,
+      "loss": 0.9882,
+      "step": 6261
+    },
+    {
+      "epoch": 0.6585773068479104,
+      "grad_norm": 3.1631141233592492,
+      "learning_rate": 1.3303525364477166e-06,
+      "loss": 1.0159,
+      "step": 6262
+    },
+    {
+      "epoch": 0.6586824772897577,
+      "grad_norm": 3.0131960777287388,
+      "learning_rate": 1.3296147856089298e-06,
+      "loss": 0.9658,
+      "step": 6263
+    },
+    {
+      "epoch": 0.658787647731605,
+      "grad_norm": 2.449844623908453,
+      "learning_rate": 1.328877165277187e-06,
+      "loss": 0.9998,
+      "step": 6264
+    },
+    {
+      "epoch": 0.6588928181734524,
+      "grad_norm": 2.452655678014848,
+      "learning_rate": 1.328139675534737e-06,
+      "loss": 0.9983,
+      "step": 6265
+    },
+    {
+      "epoch": 0.6589979886152997,
+      "grad_norm": 2.2755161044714,
+      "learning_rate": 1.3274023164638178e-06,
+      "loss": 1.0014,
+      "step": 6266
+    },
+    {
+      "epoch": 0.659103159057147,
+      "grad_norm": 1.9204300885679346,
+      "learning_rate": 1.326665088146648e-06,
+      "loss": 0.9859,
+      "step": 6267
+    },
+    {
+      "epoch": 0.6592083294989943,
+      "grad_norm": 3.1297354355899274,
+      "learning_rate": 1.325927990665436e-06,
+      "loss": 0.9901,
+      "step": 6268
+    },
+    {
+      "epoch": 0.6593134999408417,
+      "grad_norm": 2.6490272337574448,
+      "learning_rate": 1.325191024102373e-06,
+      "loss": 1.006,
+      "step": 6269
+    },
+    {
+      "epoch": 0.659418670382689,
+      "grad_norm": 3.235289643820615,
+      "learning_rate": 1.3244541885396384e-06,
+      "loss": 0.9851,
+      "step": 6270
+    },
+    {
+      "epoch": 0.6595238408245363,
+      "grad_norm": 2.232938536691584,
+      "learning_rate": 1.3237174840593927e-06,
+      "loss": 0.9571,
+      "step": 6271
+    },
+    {
+      "epoch": 0.6596290112663836,
+      "grad_norm": 2.7467215459350838,
+      "learning_rate": 1.3229809107437852e-06,
+      "loss": 0.9742,
+      "step": 6272
+    },
+    {
+      "epoch": 0.659734181708231,
+      "grad_norm": 3.265655439864346,
+      "learning_rate": 1.3222444686749508e-06,
+      "loss": 0.9426,
+      "step": 6273
+    },
+    {
+      "epoch": 0.6598393521500783,
+      "grad_norm": 2.7142378977870782,
+      "learning_rate": 1.3215081579350058e-06,
+      "loss": 0.9626,
+      "step": 6274
+    },
+    {
+      "epoch": 0.6599445225919255,
+      "grad_norm": 2.424253609885278,
+      "learning_rate": 1.3207719786060575e-06,
+      "loss": 0.9681,
+      "step": 6275
+    },
+    {
+      "epoch": 0.6600496930337728,
+      "grad_norm": 2.650710054297971,
+      "learning_rate": 1.3200359307701926e-06,
+      "loss": 1.0027,
+      "step": 6276
+    },
+    {
+      "epoch": 0.6601548634756201,
+      "grad_norm": 2.106049814182672,
+      "learning_rate": 1.319300014509488e-06,
+      "loss": 0.9976,
+      "step": 6277
+    },
+    {
+      "epoch": 0.6602600339174675,
+      "grad_norm": 3.0478589082735597,
+      "learning_rate": 1.318564229906005e-06,
+      "loss": 0.9812,
+      "step": 6278
+    },
+    {
+      "epoch": 0.6603652043593148,
+      "grad_norm": 2.644175851321169,
+      "learning_rate": 1.317828577041787e-06,
+      "loss": 0.9984,
+      "step": 6279
+    },
+    {
+      "epoch": 0.6604703748011621,
+      "grad_norm": 4.149275333858121,
+      "learning_rate": 1.317093055998867e-06,
+      "loss": 0.9905,
+      "step": 6280
+    },
+    {
+      "epoch": 0.6605755452430094,
+      "grad_norm": 2.1753727776576413,
+      "learning_rate": 1.3163576668592581e-06,
+      "loss": 0.9645,
+      "step": 6281
+    },
+    {
+      "epoch": 0.6606807156848568,
+      "grad_norm": 2.1321309814458433,
+      "learning_rate": 1.315622409704967e-06,
+      "loss": 0.9764,
+      "step": 6282
+    },
+    {
+      "epoch": 0.6607858861267041,
+      "grad_norm": 2.3284143235104104,
+      "learning_rate": 1.3148872846179761e-06,
+      "loss": 0.9685,
+      "step": 6283
+    },
+    {
+      "epoch": 0.6608910565685514,
+      "grad_norm": 2.20797024529712,
+      "learning_rate": 1.31415229168026e-06,
+      "loss": 0.9727,
+      "step": 6284
+    },
+    {
+      "epoch": 0.6609962270103987,
+      "grad_norm": 2.27464739045528,
+      "learning_rate": 1.3134174309737752e-06,
+      "loss": 0.9636,
+      "step": 6285
+    },
+    {
+      "epoch": 0.661101397452246,
+      "grad_norm": 2.564977023089282,
+      "learning_rate": 1.312682702580464e-06,
+      "loss": 0.9836,
+      "step": 6286
+    },
+    {
+      "epoch": 0.6612065678940934,
+      "grad_norm": 2.088470460153833,
+      "learning_rate": 1.3119481065822559e-06,
+      "loss": 0.8773,
+      "step": 6287
+    },
+    {
+      "epoch": 0.6613117383359407,
+      "grad_norm": 2.191541507414501,
+      "learning_rate": 1.3112136430610623e-06,
+      "loss": 1.0141,
+      "step": 6288
+    },
+    {
+      "epoch": 0.661416908777788,
+      "grad_norm": 2.459055525272844,
+      "learning_rate": 1.310479312098783e-06,
+      "loss": 0.9724,
+      "step": 6289
+    },
+    {
+      "epoch": 0.6615220792196354,
+      "grad_norm": 2.5763375289572004,
+      "learning_rate": 1.3097451137772999e-06,
+      "loss": 1.0105,
+      "step": 6290
+    },
+    {
+      "epoch": 0.6616272496614827,
+      "grad_norm": 3.1403073156867753,
+      "learning_rate": 1.3090110481784831e-06,
+      "loss": 0.9903,
+      "step": 6291
+    },
+    {
+      "epoch": 0.66173242010333,
+      "grad_norm": 2.718007886394719,
+      "learning_rate": 1.3082771153841872e-06,
+      "loss": 1.0049,
+      "step": 6292
+    },
+    {
+      "epoch": 0.6618375905451773,
+      "grad_norm": 2.551223778632464,
+      "learning_rate": 1.3075433154762496e-06,
+      "loss": 0.9796,
+      "step": 6293
+    },
+    {
+      "epoch": 0.6619427609870246,
+      "grad_norm": 2.1469053961592754,
+      "learning_rate": 1.3068096485364967e-06,
+      "loss": 1.0102,
+      "step": 6294
+    },
+    {
+      "epoch": 0.6620479314288719,
+      "grad_norm": 3.4328998554698007,
+      "learning_rate": 1.306076114646735e-06,
+      "loss": 0.9514,
+      "step": 6295
+    },
+    {
+      "epoch": 0.6621531018707192,
+      "grad_norm": 2.9569815422885153,
+      "learning_rate": 1.3053427138887631e-06,
+      "loss": 0.9742,
+      "step": 6296
+    },
+    {
+      "epoch": 0.6622582723125665,
+      "grad_norm": 2.8570811225791637,
+      "learning_rate": 1.3046094463443582e-06,
+      "loss": 1.0205,
+      "step": 6297
+    },
+    {
+      "epoch": 0.6623634427544138,
+      "grad_norm": 2.44062115758859,
+      "learning_rate": 1.3038763120952871e-06,
+      "loss": 1.004,
+      "step": 6298
+    },
+    {
+      "epoch": 0.6624686131962612,
+      "grad_norm": 2.1860706563332344,
+      "learning_rate": 1.303143311223298e-06,
+      "loss": 0.9852,
+      "step": 6299
+    },
+    {
+      "epoch": 0.6625737836381085,
+      "grad_norm": 2.305746587101714,
+      "learning_rate": 1.302410443810127e-06,
+      "loss": 0.9965,
+      "step": 6300
+    },
+    {
+      "epoch": 0.6626789540799558,
+      "grad_norm": 1.9272014050310644,
+      "learning_rate": 1.3016777099374962e-06,
+      "loss": 1.025,
+      "step": 6301
+    },
+    {
+      "epoch": 0.6627841245218031,
+      "grad_norm": 2.647801095541516,
+      "learning_rate": 1.3009451096871084e-06,
+      "loss": 0.9724,
+      "step": 6302
+    },
+    {
+      "epoch": 0.6628892949636505,
+      "grad_norm": 2.479510774074513,
+      "learning_rate": 1.3002126431406565e-06,
+      "loss": 0.9775,
+      "step": 6303
+    },
+    {
+      "epoch": 0.6629944654054978,
+      "grad_norm": 3.0571345818594478,
+      "learning_rate": 1.2994803103798131e-06,
+      "loss": 0.986,
+      "step": 6304
+    },
+    {
+      "epoch": 0.6630996358473451,
+      "grad_norm": 2.7508548007010054,
+      "learning_rate": 1.2987481114862427e-06,
+      "loss": 0.9855,
+      "step": 6305
+    },
+    {
+      "epoch": 0.6632048062891924,
+      "grad_norm": 1.9774859354202081,
+      "learning_rate": 1.2980160465415891e-06,
+      "loss": 0.9723,
+      "step": 6306
+    },
+    {
+      "epoch": 0.6633099767310398,
+      "grad_norm": 2.2732675979550114,
+      "learning_rate": 1.2972841156274843e-06,
+      "loss": 0.9814,
+      "step": 6307
+    },
+    {
+      "epoch": 0.6634151471728871,
+      "grad_norm": 2.362143185441358,
+      "learning_rate": 1.2965523188255438e-06,
+      "loss": 1.0208,
+      "step": 6308
+    },
+    {
+      "epoch": 0.6635203176147344,
+      "grad_norm": 2.2701352939481096,
+      "learning_rate": 1.2958206562173664e-06,
+      "loss": 0.9388,
+      "step": 6309
+    },
+    {
+      "epoch": 0.6636254880565817,
+      "grad_norm": 2.3921549016140102,
+      "learning_rate": 1.2950891278845423e-06,
+      "loss": 0.9988,
+      "step": 6310
+    },
+    {
+      "epoch": 0.663730658498429,
+      "grad_norm": 2.573032844055122,
+      "learning_rate": 1.2943577339086395e-06,
+      "loss": 0.9928,
+      "step": 6311
+    },
+    {
+      "epoch": 0.6638358289402764,
+      "grad_norm": 1.896933627943948,
+      "learning_rate": 1.2936264743712159e-06,
+      "loss": 0.9814,
+      "step": 6312
+    },
+    {
+      "epoch": 0.6639409993821237,
+      "grad_norm": 2.511447955098937,
+      "learning_rate": 1.292895349353811e-06,
+      "loss": 0.9846,
+      "step": 6313
+    },
+    {
+      "epoch": 0.664046169823971,
+      "grad_norm": 2.391646179219941,
+      "learning_rate": 1.2921643589379517e-06,
+      "loss": 0.9703,
+      "step": 6314
+    },
+    {
+      "epoch": 0.6641513402658183,
+      "grad_norm": 2.067946485876376,
+      "learning_rate": 1.2914335032051502e-06,
+      "loss": 0.971,
+      "step": 6315
+    },
+    {
+      "epoch": 0.6642565107076656,
+      "grad_norm": 3.1972490953679165,
+      "learning_rate": 1.2907027822369006e-06,
+      "loss": 0.9421,
+      "step": 6316
+    },
+    {
+      "epoch": 0.6643616811495129,
+      "grad_norm": 1.7933848682194085,
+      "learning_rate": 1.289972196114686e-06,
+      "loss": 1.0276,
+      "step": 6317
+    },
+    {
+      "epoch": 0.6644668515913602,
+      "grad_norm": 2.1050494682029797,
+      "learning_rate": 1.2892417449199696e-06,
+      "loss": 0.9682,
+      "step": 6318
+    },
+    {
+      "epoch": 0.6645720220332075,
+      "grad_norm": 2.5224497949789235,
+      "learning_rate": 1.2885114287342058e-06,
+      "loss": 1.0111,
+      "step": 6319
+    },
+    {
+      "epoch": 0.6646771924750549,
+      "grad_norm": 1.9853269766464046,
+      "learning_rate": 1.287781247638828e-06,
+      "loss": 0.9859,
+      "step": 6320
+    },
+    {
+      "epoch": 0.6647823629169022,
+      "grad_norm": 3.571360120955266,
+      "learning_rate": 1.2870512017152586e-06,
+      "loss": 1.0006,
+      "step": 6321
+    },
+    {
+      "epoch": 0.6648875333587495,
+      "grad_norm": 2.8210890079031303,
+      "learning_rate": 1.286321291044902e-06,
+      "loss": 0.9834,
+      "step": 6322
+    },
+    {
+      "epoch": 0.6649927038005968,
+      "grad_norm": 2.1350217532955544,
+      "learning_rate": 1.2855915157091498e-06,
+      "loss": 1.0235,
+      "step": 6323
+    },
+    {
+      "epoch": 0.6650978742424442,
+      "grad_norm": 2.939945404933564,
+      "learning_rate": 1.2848618757893782e-06,
+      "loss": 0.986,
+      "step": 6324
+    },
+    {
+      "epoch": 0.6652030446842915,
+      "grad_norm": 2.8261629770042602,
+      "learning_rate": 1.284132371366946e-06,
+      "loss": 1.0038,
+      "step": 6325
+    },
+    {
+      "epoch": 0.6653082151261388,
+      "grad_norm": 2.52942309852291,
+      "learning_rate": 1.2834030025232006e-06,
+      "loss": 0.9817,
+      "step": 6326
+    },
+    {
+      "epoch": 0.6654133855679861,
+      "grad_norm": 2.432243507022762,
+      "learning_rate": 1.2826737693394693e-06,
+      "loss": 1.0329,
+      "step": 6327
+    },
+    {
+      "epoch": 0.6655185560098335,
+      "grad_norm": 2.2735077443497937,
+      "learning_rate": 1.2819446718970713e-06,
+      "loss": 1.0334,
+      "step": 6328
+    },
+    {
+      "epoch": 0.6656237264516808,
+      "grad_norm": 2.6901417303834467,
+      "learning_rate": 1.2812157102773043e-06,
+      "loss": 1.0319,
+      "step": 6329
+    },
+    {
+      "epoch": 0.6657288968935281,
+      "grad_norm": 1.9151652735311289,
+      "learning_rate": 1.2804868845614527e-06,
+      "loss": 1.0053,
+      "step": 6330
+    },
+    {
+      "epoch": 0.6658340673353754,
+      "grad_norm": 2.923158306662884,
+      "learning_rate": 1.279758194830788e-06,
+      "loss": 0.963,
+      "step": 6331
+    },
+    {
+      "epoch": 0.6659392377772227,
+      "grad_norm": 3.174548375700343,
+      "learning_rate": 1.2790296411665618e-06,
+      "loss": 0.9674,
+      "step": 6332
+    },
+    {
+      "epoch": 0.6660444082190701,
+      "grad_norm": 3.0051084482405717,
+      "learning_rate": 1.2783012236500173e-06,
+      "loss": 0.9453,
+      "step": 6333
+    },
+    {
+      "epoch": 0.6661495786609174,
+      "grad_norm": 2.594123710120902,
+      "learning_rate": 1.2775729423623759e-06,
+      "loss": 0.9764,
+      "step": 6334
+    },
+    {
+      "epoch": 0.6662547491027647,
+      "grad_norm": 2.5531587705993126,
+      "learning_rate": 1.2768447973848485e-06,
+      "loss": 0.9683,
+      "step": 6335
+    },
+    {
+      "epoch": 0.6663599195446119,
+      "grad_norm": 3.1748751741089167,
+      "learning_rate": 1.276116788798627e-06,
+      "loss": 1.02,
+      "step": 6336
+    },
+    {
+      "epoch": 0.6664650899864593,
+      "grad_norm": 2.7632150853823148,
+      "learning_rate": 1.2753889166848909e-06,
+      "loss": 0.9777,
+      "step": 6337
+    },
+    {
+      "epoch": 0.6665702604283066,
+      "grad_norm": 2.0037733131489106,
+      "learning_rate": 1.274661181124805e-06,
+      "loss": 0.9881,
+      "step": 6338
+    },
+    {
+      "epoch": 0.6666754308701539,
+      "grad_norm": 2.4827855341616885,
+      "learning_rate": 1.2739335821995153e-06,
+      "loss": 0.94,
+      "step": 6339
+    },
+    {
+      "epoch": 0.6667806013120012,
+      "grad_norm": 2.8119810771826517,
+      "learning_rate": 1.2732061199901563e-06,
+      "loss": 0.994,
+      "step": 6340
+    },
+    {
+      "epoch": 0.6668857717538486,
+      "grad_norm": 2.752059269835914,
+      "learning_rate": 1.2724787945778427e-06,
+      "loss": 1.0225,
+      "step": 6341
+    },
+    {
+      "epoch": 0.6669909421956959,
+      "grad_norm": 2.2241542929333216,
+      "learning_rate": 1.271751606043682e-06,
+      "loss": 0.9708,
+      "step": 6342
+    },
+    {
+      "epoch": 0.6670961126375432,
+      "grad_norm": 2.6235709049395997,
+      "learning_rate": 1.2710245544687568e-06,
+      "loss": 1.0189,
+      "step": 6343
+    },
+    {
+      "epoch": 0.6672012830793905,
+      "grad_norm": 3.156266897222367,
+      "learning_rate": 1.2702976399341422e-06,
+      "loss": 0.9985,
+      "step": 6344
+    },
+    {
+      "epoch": 0.6673064535212379,
+      "grad_norm": 2.230980642118251,
+      "learning_rate": 1.2695708625208933e-06,
+      "loss": 0.9875,
+      "step": 6345
+    },
+    {
+      "epoch": 0.6674116239630852,
+      "grad_norm": 2.5767378584379155,
+      "learning_rate": 1.2688442223100494e-06,
+      "loss": 0.9804,
+      "step": 6346
+    },
+    {
+      "epoch": 0.6675167944049325,
+      "grad_norm": 2.029400978615484,
+      "learning_rate": 1.268117719382641e-06,
+      "loss": 0.9311,
+      "step": 6347
+    },
+    {
+      "epoch": 0.6676219648467798,
+      "grad_norm": 2.6045821189170235,
+      "learning_rate": 1.2673913538196753e-06,
+      "loss": 1.0034,
+      "step": 6348
+    },
+    {
+      "epoch": 0.6677271352886271,
+      "grad_norm": 2.0673514221546965,
+      "learning_rate": 1.26666512570215e-06,
+      "loss": 0.9702,
+      "step": 6349
+    },
+    {
+      "epoch": 0.6678323057304745,
+      "grad_norm": 2.227452564501476,
+      "learning_rate": 1.265939035111043e-06,
+      "loss": 0.9654,
+      "step": 6350
+    },
+    {
+      "epoch": 0.6679374761723218,
+      "grad_norm": 2.389880074299462,
+      "learning_rate": 1.26521308212732e-06,
+      "loss": 1.024,
+      "step": 6351
+    },
+    {
+      "epoch": 0.6680426466141691,
+      "grad_norm": 1.9298756073215733,
+      "learning_rate": 1.2644872668319317e-06,
+      "loss": 1.0022,
+      "step": 6352
+    },
+    {
+      "epoch": 0.6681478170560164,
+      "grad_norm": 2.8622091569435577,
+      "learning_rate": 1.2637615893058098e-06,
+      "loss": 0.9935,
+      "step": 6353
+    },
+    {
+      "epoch": 0.6682529874978638,
+      "grad_norm": 3.031543303955931,
+      "learning_rate": 1.263036049629875e-06,
+      "loss": 0.9945,
+      "step": 6354
+    },
+    {
+      "epoch": 0.6683581579397111,
+      "grad_norm": 3.1179126127334835,
+      "learning_rate": 1.262310647885028e-06,
+      "loss": 1.0373,
+      "step": 6355
+    },
+    {
+      "epoch": 0.6684633283815583,
+      "grad_norm": 2.301408947362086,
+      "learning_rate": 1.2615853841521602e-06,
+      "loss": 1.0232,
+      "step": 6356
+    },
+    {
+      "epoch": 0.6685684988234056,
+      "grad_norm": 2.9564009207787585,
+      "learning_rate": 1.2608602585121419e-06,
+      "loss": 1.0009,
+      "step": 6357
+    },
+    {
+      "epoch": 0.668673669265253,
+      "grad_norm": 2.338028385808441,
+      "learning_rate": 1.2601352710458314e-06,
+      "loss": 0.9979,
+      "step": 6358
+    },
+    {
+      "epoch": 0.6687788397071003,
+      "grad_norm": 1.9749844430021521,
+      "learning_rate": 1.2594104218340686e-06,
+      "loss": 0.9774,
+      "step": 6359
+    },
+    {
+      "epoch": 0.6688840101489476,
+      "grad_norm": 2.0803838048430623,
+      "learning_rate": 1.2586857109576814e-06,
+      "loss": 1.0137,
+      "step": 6360
+    },
+    {
+      "epoch": 0.6689891805907949,
+      "grad_norm": 2.561718404866887,
+      "learning_rate": 1.257961138497481e-06,
+      "loss": 1.0203,
+      "step": 6361
+    },
+    {
+      "epoch": 0.6690943510326423,
+      "grad_norm": 2.110231753776184,
+      "learning_rate": 1.2572367045342615e-06,
+      "loss": 0.9843,
+      "step": 6362
+    },
+    {
+      "epoch": 0.6691995214744896,
+      "grad_norm": 3.728842150551063,
+      "learning_rate": 1.256512409148804e-06,
+      "loss": 1.0121,
+      "step": 6363
+    },
+    {
+      "epoch": 0.6693046919163369,
+      "grad_norm": 2.6450699087235745,
+      "learning_rate": 1.2557882524218722e-06,
+      "loss": 1.0254,
+      "step": 6364
+    },
+    {
+      "epoch": 0.6694098623581842,
+      "grad_norm": 3.3862188687250474,
+      "learning_rate": 1.2550642344342155e-06,
+      "loss": 0.9904,
+      "step": 6365
+    },
+    {
+      "epoch": 0.6695150328000316,
+      "grad_norm": 2.6560070020655164,
+      "learning_rate": 1.2543403552665684e-06,
+      "loss": 1.001,
+      "step": 6366
+    },
+    {
+      "epoch": 0.6696202032418789,
+      "grad_norm": 2.066321011865616,
+      "learning_rate": 1.2536166149996476e-06,
+      "loss": 1.0294,
+      "step": 6367
+    },
+    {
+      "epoch": 0.6697253736837262,
+      "grad_norm": 1.841460947605558,
+      "learning_rate": 1.252893013714157e-06,
+      "loss": 0.955,
+      "step": 6368
+    },
+    {
+      "epoch": 0.6698305441255735,
+      "grad_norm": 2.6195237655803427,
+      "learning_rate": 1.2521695514907817e-06,
+      "loss": 0.981,
+      "step": 6369
+    },
+    {
+      "epoch": 0.6699357145674208,
+      "grad_norm": 2.5282441482759155,
+      "learning_rate": 1.2514462284101969e-06,
+      "loss": 1.0073,
+      "step": 6370
+    },
+    {
+      "epoch": 0.6700408850092682,
+      "grad_norm": 2.420722358809797,
+      "learning_rate": 1.2507230445530554e-06,
+      "loss": 1.0199,
+      "step": 6371
+    },
+    {
+      "epoch": 0.6701460554511155,
+      "grad_norm": 3.210358669515261,
+      "learning_rate": 1.2500000000000007e-06,
+      "loss": 0.935,
+      "step": 6372
+    },
+    {
+      "epoch": 0.6702512258929628,
+      "grad_norm": 2.6066278908101244,
+      "learning_rate": 1.2492770948316548e-06,
+      "loss": 0.9807,
+      "step": 6373
+    },
+    {
+      "epoch": 0.6703563963348101,
+      "grad_norm": 2.166982928665858,
+      "learning_rate": 1.2485543291286292e-06,
+      "loss": 1.0028,
+      "step": 6374
+    },
+    {
+      "epoch": 0.6704615667766575,
+      "grad_norm": 2.3061126697659953,
+      "learning_rate": 1.2478317029715182e-06,
+      "loss": 0.972,
+      "step": 6375
+    },
+    {
+      "epoch": 0.6705667372185048,
+      "grad_norm": 2.574290187989157,
+      "learning_rate": 1.2471092164408985e-06,
+      "loss": 0.9746,
+      "step": 6376
+    },
+    {
+      "epoch": 0.670671907660352,
+      "grad_norm": 2.5423268036429754,
+      "learning_rate": 1.2463868696173351e-06,
+      "loss": 1.0209,
+      "step": 6377
+    },
+    {
+      "epoch": 0.6707770781021993,
+      "grad_norm": 2.2164659344968767,
+      "learning_rate": 1.245664662581372e-06,
+      "loss": 0.9704,
+      "step": 6378
+    },
+    {
+      "epoch": 0.6708822485440467,
+      "grad_norm": 2.582570427543782,
+      "learning_rate": 1.2449425954135452e-06,
+      "loss": 0.9712,
+      "step": 6379
+    },
+    {
+      "epoch": 0.670987418985894,
+      "grad_norm": 2.124683883692389,
+      "learning_rate": 1.2442206681943685e-06,
+      "loss": 0.9725,
+      "step": 6380
+    },
+    {
+      "epoch": 0.6710925894277413,
+      "grad_norm": 2.3639142162437095,
+      "learning_rate": 1.2434988810043416e-06,
+      "loss": 0.9681,
+      "step": 6381
+    },
+    {
+      "epoch": 0.6711977598695886,
+      "grad_norm": 2.1442126123743024,
+      "learning_rate": 1.2427772339239502e-06,
+      "loss": 1.0037,
+      "step": 6382
+    },
+    {
+      "epoch": 0.671302930311436,
+      "grad_norm": 2.625018523809141,
+      "learning_rate": 1.2420557270336638e-06,
+      "loss": 1.0121,
+      "step": 6383
+    },
+    {
+      "epoch": 0.6714081007532833,
+      "grad_norm": 2.4345832011147217,
+      "learning_rate": 1.241334360413937e-06,
+      "loss": 0.972,
+      "step": 6384
+    },
+    {
+      "epoch": 0.6715132711951306,
+      "grad_norm": 2.7727329895954993,
+      "learning_rate": 1.2406131341452054e-06,
+      "loss": 1.0481,
+      "step": 6385
+    },
+    {
+      "epoch": 0.6716184416369779,
+      "grad_norm": 2.7456790530416173,
+      "learning_rate": 1.2398920483078939e-06,
+      "loss": 0.9972,
+      "step": 6386
+    },
+    {
+      "epoch": 0.6717236120788252,
+      "grad_norm": 2.2868278468077765,
+      "learning_rate": 1.2391711029824065e-06,
+      "loss": 0.981,
+      "step": 6387
+    },
+    {
+      "epoch": 0.6718287825206726,
+      "grad_norm": 2.193127747109827,
+      "learning_rate": 1.2384502982491359e-06,
+      "loss": 0.9802,
+      "step": 6388
+    },
+    {
+      "epoch": 0.6719339529625199,
+      "grad_norm": 2.525039802068027,
+      "learning_rate": 1.2377296341884578e-06,
+      "loss": 1.0594,
+      "step": 6389
+    },
+    {
+      "epoch": 0.6720391234043672,
+      "grad_norm": 2.2688658169002793,
+      "learning_rate": 1.2370091108807307e-06,
+      "loss": 0.9756,
+      "step": 6390
+    },
+    {
+      "epoch": 0.6721442938462145,
+      "grad_norm": 2.8518070010439383,
+      "learning_rate": 1.2362887284062994e-06,
+      "loss": 1.0145,
+      "step": 6391
+    },
+    {
+      "epoch": 0.6722494642880619,
+      "grad_norm": 3.120055638002566,
+      "learning_rate": 1.23556848684549e-06,
+      "loss": 1.0178,
+      "step": 6392
+    },
+    {
+      "epoch": 0.6723546347299092,
+      "grad_norm": 2.677188223326482,
+      "learning_rate": 1.2348483862786188e-06,
+      "loss": 0.9663,
+      "step": 6393
+    },
+    {
+      "epoch": 0.6724598051717565,
+      "grad_norm": 2.764351849759641,
+      "learning_rate": 1.2341284267859796e-06,
+      "loss": 0.9915,
+      "step": 6394
+    },
+    {
+      "epoch": 0.6725649756136038,
+      "grad_norm": 1.9868623247964174,
+      "learning_rate": 1.2334086084478553e-06,
+      "loss": 0.9902,
+      "step": 6395
+    },
+    {
+      "epoch": 0.6726701460554512,
+      "grad_norm": 1.88432409309354,
+      "learning_rate": 1.2326889313445095e-06,
+      "loss": 1.0214,
+      "step": 6396
+    },
+    {
+      "epoch": 0.6727753164972984,
+      "grad_norm": 2.505160726077182,
+      "learning_rate": 1.2319693955561926e-06,
+      "loss": 0.996,
+      "step": 6397
+    },
+    {
+      "epoch": 0.6728804869391457,
+      "grad_norm": 2.4305679692895326,
+      "learning_rate": 1.2312500011631396e-06,
+      "loss": 0.9957,
+      "step": 6398
+    },
+    {
+      "epoch": 0.672985657380993,
+      "grad_norm": 2.26838492918931,
+      "learning_rate": 1.2305307482455661e-06,
+      "loss": 0.9879,
+      "step": 6399
+    },
+    {
+      "epoch": 0.6730908278228404,
+      "grad_norm": 2.7005160503685772,
+      "learning_rate": 1.2298116368836772e-06,
+      "loss": 1.0049,
+      "step": 6400
+    },
+    {
+      "epoch": 0.6731959982646877,
+      "grad_norm": 2.5263479523166814,
+      "learning_rate": 1.2290926671576567e-06,
+      "loss": 1.0359,
+      "step": 6401
+    },
+    {
+      "epoch": 0.673301168706535,
+      "grad_norm": 2.1790008720432605,
+      "learning_rate": 1.2283738391476766e-06,
+      "loss": 1.0109,
+      "step": 6402
+    },
+    {
+      "epoch": 0.6734063391483823,
+      "grad_norm": 2.6192166675598467,
+      "learning_rate": 1.2276551529338929e-06,
+      "loss": 0.9518,
+      "step": 6403
+    },
+    {
+      "epoch": 0.6735115095902297,
+      "grad_norm": 2.905972519031942,
+      "learning_rate": 1.2269366085964424e-06,
+      "loss": 1.0174,
+      "step": 6404
+    },
+    {
+      "epoch": 0.673616680032077,
+      "grad_norm": 2.9312474143114664,
+      "learning_rate": 1.2262182062154498e-06,
+      "loss": 1.0203,
+      "step": 6405
+    },
+    {
+      "epoch": 0.6737218504739243,
+      "grad_norm": 2.394870761083195,
+      "learning_rate": 1.225499945871022e-06,
+      "loss": 0.9592,
+      "step": 6406
+    },
+    {
+      "epoch": 0.6738270209157716,
+      "grad_norm": 2.025708578404046,
+      "learning_rate": 1.2247818276432522e-06,
+      "loss": 0.9798,
+      "step": 6407
+    },
+    {
+      "epoch": 0.673932191357619,
+      "grad_norm": 2.183993520411737,
+      "learning_rate": 1.2240638516122135e-06,
+      "loss": 0.9284,
+      "step": 6408
+    },
+    {
+      "epoch": 0.6740373617994663,
+      "grad_norm": 1.9540801713327498,
+      "learning_rate": 1.2233460178579683e-06,
+      "loss": 0.9748,
+      "step": 6409
+    },
+    {
+      "epoch": 0.6741425322413136,
+      "grad_norm": 2.397767392403271,
+      "learning_rate": 1.2226283264605587e-06,
+      "loss": 0.9958,
+      "step": 6410
+    },
+    {
+      "epoch": 0.6742477026831609,
+      "grad_norm": 2.63791023453177,
+      "learning_rate": 1.2219107775000136e-06,
+      "loss": 0.9764,
+      "step": 6411
+    },
+    {
+      "epoch": 0.6743528731250082,
+      "grad_norm": 2.6612509498444563,
+      "learning_rate": 1.2211933710563462e-06,
+      "loss": 0.9665,
+      "step": 6412
+    },
+    {
+      "epoch": 0.6744580435668556,
+      "grad_norm": 2.1922298975484,
+      "learning_rate": 1.2204761072095511e-06,
+      "loss": 1.0004,
+      "step": 6413
+    },
+    {
+      "epoch": 0.6745632140087029,
+      "grad_norm": 2.3286284733059666,
+      "learning_rate": 1.219758986039611e-06,
+      "loss": 0.9977,
+      "step": 6414
+    },
+    {
+      "epoch": 0.6746683844505502,
+      "grad_norm": 2.8936932199687746,
+      "learning_rate": 1.2190420076264877e-06,
+      "loss": 1.0052,
+      "step": 6415
+    },
+    {
+      "epoch": 0.6747735548923975,
+      "grad_norm": 2.9466430859373043,
+      "learning_rate": 1.2183251720501317e-06,
+      "loss": 1.0172,
+      "step": 6416
+    },
+    {
+      "epoch": 0.6748787253342448,
+      "grad_norm": 2.6112081025461276,
+      "learning_rate": 1.2176084793904764e-06,
+      "loss": 1.026,
+      "step": 6417
+    },
+    {
+      "epoch": 0.6749838957760921,
+      "grad_norm": 2.623855271718345,
+      "learning_rate": 1.2168919297274368e-06,
+      "loss": 0.998,
+      "step": 6418
+    },
+    {
+      "epoch": 0.6750890662179394,
+      "grad_norm": 2.4575121259688033,
+      "learning_rate": 1.2161755231409142e-06,
+      "loss": 0.9709,
+      "step": 6419
+    },
+    {
+      "epoch": 0.6751942366597867,
+      "grad_norm": 2.5971196149724416,
+      "learning_rate": 1.2154592597107942e-06,
+      "loss": 0.9818,
+      "step": 6420
+    },
+    {
+      "epoch": 0.675299407101634,
+      "grad_norm": 1.8871909749453315,
+      "learning_rate": 1.214743139516946e-06,
+      "loss": 1.0128,
+      "step": 6421
+    },
+    {
+      "epoch": 0.6754045775434814,
+      "grad_norm": 2.2047792064214824,
+      "learning_rate": 1.2140271626392215e-06,
+      "loss": 0.9645,
+      "step": 6422
+    },
+    {
+      "epoch": 0.6755097479853287,
+      "grad_norm": 2.1961869799992435,
+      "learning_rate": 1.2133113291574586e-06,
+      "loss": 0.9703,
+      "step": 6423
+    },
+    {
+      "epoch": 0.675614918427176,
+      "grad_norm": 2.958708144601371,
+      "learning_rate": 1.212595639151477e-06,
+      "loss": 0.9813,
+      "step": 6424
+    },
+    {
+      "epoch": 0.6757200888690233,
+      "grad_norm": 2.1479518310219854,
+      "learning_rate": 1.211880092701083e-06,
+      "loss": 1.0119,
+      "step": 6425
+    },
+    {
+      "epoch": 0.6758252593108707,
+      "grad_norm": 2.390810440217156,
+      "learning_rate": 1.2111646898860654e-06,
+      "loss": 0.9381,
+      "step": 6426
+    },
+    {
+      "epoch": 0.675930429752718,
+      "grad_norm": 2.2064670963516835,
+      "learning_rate": 1.2104494307861963e-06,
+      "loss": 0.9644,
+      "step": 6427
+    },
+    {
+      "epoch": 0.6760356001945653,
+      "grad_norm": 2.429723636327678,
+      "learning_rate": 1.2097343154812332e-06,
+      "loss": 0.9703,
+      "step": 6428
+    },
+    {
+      "epoch": 0.6761407706364126,
+      "grad_norm": 3.083948528744446,
+      "learning_rate": 1.2090193440509173e-06,
+      "loss": 1.0232,
+      "step": 6429
+    },
+    {
+      "epoch": 0.67624594107826,
+      "grad_norm": 2.9234421217471187,
+      "learning_rate": 1.208304516574974e-06,
+      "loss": 0.9884,
+      "step": 6430
+    },
+    {
+      "epoch": 0.6763511115201073,
+      "grad_norm": 1.6972930816606224,
+      "learning_rate": 1.2075898331331112e-06,
+      "loss": 1.0111,
+      "step": 6431
+    },
+    {
+      "epoch": 0.6764562819619546,
+      "grad_norm": 3.532197380851382,
+      "learning_rate": 1.206875293805021e-06,
+      "loss": 1.0103,
+      "step": 6432
+    },
+    {
+      "epoch": 0.6765614524038019,
+      "grad_norm": 2.534725775044316,
+      "learning_rate": 1.206160898670381e-06,
+      "loss": 0.9684,
+      "step": 6433
+    },
+    {
+      "epoch": 0.6766666228456493,
+      "grad_norm": 2.6453642866329408,
+      "learning_rate": 1.2054466478088515e-06,
+      "loss": 1.0008,
+      "step": 6434
+    },
+    {
+      "epoch": 0.6767717932874966,
+      "grad_norm": 2.312434495761649,
+      "learning_rate": 1.2047325413000782e-06,
+      "loss": 0.9743,
+      "step": 6435
+    },
+    {
+      "epoch": 0.6768769637293439,
+      "grad_norm": 1.9592460033323447,
+      "learning_rate": 1.2040185792236874e-06,
+      "loss": 1.004,
+      "step": 6436
+    },
+    {
+      "epoch": 0.6769821341711912,
+      "grad_norm": 1.8537125451220031,
+      "learning_rate": 1.2033047616592938e-06,
+      "loss": 0.9589,
+      "step": 6437
+    },
+    {
+      "epoch": 0.6770873046130385,
+      "grad_norm": 2.3821524850998137,
+      "learning_rate": 1.2025910886864914e-06,
+      "loss": 1.0142,
+      "step": 6438
+    },
+    {
+      "epoch": 0.6771924750548858,
+      "grad_norm": 2.6671405915081805,
+      "learning_rate": 1.2018775603848613e-06,
+      "loss": 0.9977,
+      "step": 6439
+    },
+    {
+      "epoch": 0.6772976454967331,
+      "grad_norm": 2.9782145090252388,
+      "learning_rate": 1.201164176833968e-06,
+      "loss": 0.9522,
+      "step": 6440
+    },
+    {
+      "epoch": 0.6774028159385804,
+      "grad_norm": 2.3275648903630803,
+      "learning_rate": 1.2004509381133577e-06,
+      "loss": 0.9778,
+      "step": 6441
+    },
+    {
+      "epoch": 0.6775079863804278,
+      "grad_norm": 2.2558601171743953,
+      "learning_rate": 1.1997378443025633e-06,
+      "loss": 1.0001,
+      "step": 6442
+    },
+    {
+      "epoch": 0.6776131568222751,
+      "grad_norm": 2.7943480960191405,
+      "learning_rate": 1.1990248954811002e-06,
+      "loss": 1.0329,
+      "step": 6443
+    },
+    {
+      "epoch": 0.6777183272641224,
+      "grad_norm": 1.9980446949092163,
+      "learning_rate": 1.1983120917284682e-06,
+      "loss": 1.0,
+      "step": 6444
+    },
+    {
+      "epoch": 0.6778234977059697,
+      "grad_norm": 2.5703128757840967,
+      "learning_rate": 1.1975994331241491e-06,
+      "loss": 1.0233,
+      "step": 6445
+    },
+    {
+      "epoch": 0.677928668147817,
+      "grad_norm": 2.454703747591987,
+      "learning_rate": 1.1968869197476116e-06,
+      "loss": 0.9941,
+      "step": 6446
+    },
+    {
+      "epoch": 0.6780338385896644,
+      "grad_norm": 2.1653067794036995,
+      "learning_rate": 1.1961745516783044e-06,
+      "loss": 0.9694,
+      "step": 6447
+    },
+    {
+      "epoch": 0.6781390090315117,
+      "grad_norm": 2.2038359344425396,
+      "learning_rate": 1.1954623289956633e-06,
+      "loss": 0.977,
+      "step": 6448
+    },
+    {
+      "epoch": 0.678244179473359,
+      "grad_norm": 2.693813467095616,
+      "learning_rate": 1.1947502517791073e-06,
+      "loss": 0.984,
+      "step": 6449
+    },
+    {
+      "epoch": 0.6783493499152063,
+      "grad_norm": 2.1406308204159874,
+      "learning_rate": 1.194038320108037e-06,
+      "loss": 0.9603,
+      "step": 6450
+    },
+    {
+      "epoch": 0.6784545203570537,
+      "grad_norm": 2.1294455644301076,
+      "learning_rate": 1.1933265340618389e-06,
+      "loss": 0.9735,
+      "step": 6451
+    },
+    {
+      "epoch": 0.678559690798901,
+      "grad_norm": 2.103773447891833,
+      "learning_rate": 1.192614893719884e-06,
+      "loss": 0.9834,
+      "step": 6452
+    },
+    {
+      "epoch": 0.6786648612407483,
+      "grad_norm": 2.0922753017568336,
+      "learning_rate": 1.1919033991615234e-06,
+      "loss": 1.0113,
+      "step": 6453
+    },
+    {
+      "epoch": 0.6787700316825956,
+      "grad_norm": 1.9815603762507759,
+      "learning_rate": 1.1911920504660963e-06,
+      "loss": 0.9569,
+      "step": 6454
+    },
+    {
+      "epoch": 0.678875202124443,
+      "grad_norm": 1.797769460062252,
+      "learning_rate": 1.190480847712922e-06,
+      "loss": 0.9873,
+      "step": 6455
+    },
+    {
+      "epoch": 0.6789803725662903,
+      "grad_norm": 2.003693077137846,
+      "learning_rate": 1.1897697909813058e-06,
+      "loss": 1.0133,
+      "step": 6456
+    },
+    {
+      "epoch": 0.6790855430081376,
+      "grad_norm": 2.1003840733140686,
+      "learning_rate": 1.1890588803505362e-06,
+      "loss": 0.9632,
+      "step": 6457
+    },
+    {
+      "epoch": 0.6791907134499848,
+      "grad_norm": 2.080990514432014,
+      "learning_rate": 1.1883481158998862e-06,
+      "loss": 0.9783,
+      "step": 6458
+    },
+    {
+      "epoch": 0.6792958838918322,
+      "grad_norm": 3.374047031904028,
+      "learning_rate": 1.1876374977086094e-06,
+      "loss": 1.0202,
+      "step": 6459
+    },
+    {
+      "epoch": 0.6794010543336795,
+      "grad_norm": 2.4938920235535047,
+      "learning_rate": 1.1869270258559477e-06,
+      "loss": 1.008,
+      "step": 6460
+    },
+    {
+      "epoch": 0.6795062247755268,
+      "grad_norm": 2.186419415119328,
+      "learning_rate": 1.1862167004211217e-06,
+      "loss": 0.9835,
+      "step": 6461
+    },
+    {
+      "epoch": 0.6796113952173741,
+      "grad_norm": 2.9582048165354378,
+      "learning_rate": 1.1855065214833394e-06,
+      "loss": 0.9733,
+      "step": 6462
+    },
+    {
+      "epoch": 0.6797165656592214,
+      "grad_norm": 3.8575153613107016,
+      "learning_rate": 1.1847964891217923e-06,
+      "loss": 1.0115,
+      "step": 6463
+    },
+    {
+      "epoch": 0.6798217361010688,
+      "grad_norm": 2.242861038730955,
+      "learning_rate": 1.1840866034156526e-06,
+      "loss": 0.9677,
+      "step": 6464
+    },
+    {
+      "epoch": 0.6799269065429161,
+      "grad_norm": 2.0912260201336648,
+      "learning_rate": 1.1833768644440787e-06,
+      "loss": 0.9556,
+      "step": 6465
+    },
+    {
+      "epoch": 0.6800320769847634,
+      "grad_norm": 3.3182406948209993,
+      "learning_rate": 1.1826672722862137e-06,
+      "loss": 0.9788,
+      "step": 6466
+    },
+    {
+      "epoch": 0.6801372474266107,
+      "grad_norm": 2.887753294344934,
+      "learning_rate": 1.1819578270211801e-06,
+      "loss": 0.9623,
+      "step": 6467
+    },
+    {
+      "epoch": 0.6802424178684581,
+      "grad_norm": 3.1634544519317966,
+      "learning_rate": 1.1812485287280886e-06,
+      "loss": 1.0159,
+      "step": 6468
+    },
+    {
+      "epoch": 0.6803475883103054,
+      "grad_norm": 2.210225867067167,
+      "learning_rate": 1.1805393774860296e-06,
+      "loss": 0.9579,
+      "step": 6469
+    },
+    {
+      "epoch": 0.6804527587521527,
+      "grad_norm": 2.3215390072281874,
+      "learning_rate": 1.1798303733740801e-06,
+      "loss": 0.9572,
+      "step": 6470
+    },
+    {
+      "epoch": 0.680557929194,
+      "grad_norm": 2.531063961551066,
+      "learning_rate": 1.1791215164712993e-06,
+      "loss": 1.0024,
+      "step": 6471
+    },
+    {
+      "epoch": 0.6806630996358474,
+      "grad_norm": 2.6959358550758883,
+      "learning_rate": 1.1784128068567316e-06,
+      "loss": 0.996,
+      "step": 6472
+    },
+    {
+      "epoch": 0.6807682700776947,
+      "grad_norm": 2.0196092769679828,
+      "learning_rate": 1.1777042446094011e-06,
+      "loss": 0.9526,
+      "step": 6473
+    },
+    {
+      "epoch": 0.680873440519542,
+      "grad_norm": 2.020236502773026,
+      "learning_rate": 1.1769958298083192e-06,
+      "loss": 0.9758,
+      "step": 6474
+    },
+    {
+      "epoch": 0.6809786109613893,
+      "grad_norm": 2.1941180121097825,
+      "learning_rate": 1.176287562532481e-06,
+      "loss": 0.9882,
+      "step": 6475
+    },
+    {
+      "epoch": 0.6810837814032367,
+      "grad_norm": 2.44848887745427,
+      "learning_rate": 1.1755794428608614e-06,
+      "loss": 1.0088,
+      "step": 6476
+    },
+    {
+      "epoch": 0.681188951845084,
+      "grad_norm": 1.6532880562870076,
+      "learning_rate": 1.1748714708724232e-06,
+      "loss": 0.9774,
+      "step": 6477
+    },
+    {
+      "epoch": 0.6812941222869312,
+      "grad_norm": 2.6504903012905494,
+      "learning_rate": 1.1741636466461093e-06,
+      "loss": 0.9639,
+      "step": 6478
+    },
+    {
+      "epoch": 0.6813992927287785,
+      "grad_norm": 2.1833707527018653,
+      "learning_rate": 1.173455970260848e-06,
+      "loss": 0.9325,
+      "step": 6479
+    },
+    {
+      "epoch": 0.6815044631706259,
+      "grad_norm": 1.8991991840090796,
+      "learning_rate": 1.1727484417955512e-06,
+      "loss": 0.9709,
+      "step": 6480
+    },
+    {
+      "epoch": 0.6816096336124732,
+      "grad_norm": 2.7755476284341594,
+      "learning_rate": 1.1720410613291144e-06,
+      "loss": 1.0033,
+      "step": 6481
+    },
+    {
+      "epoch": 0.6817148040543205,
+      "grad_norm": 2.4340897257786587,
+      "learning_rate": 1.1713338289404152e-06,
+      "loss": 0.9761,
+      "step": 6482
+    },
+    {
+      "epoch": 0.6818199744961678,
+      "grad_norm": 2.6322074888977873,
+      "learning_rate": 1.1706267447083145e-06,
+      "loss": 0.9855,
+      "step": 6483
+    },
+    {
+      "epoch": 0.6819251449380151,
+      "grad_norm": 2.5435211848927963,
+      "learning_rate": 1.169919808711659e-06,
+      "loss": 0.9968,
+      "step": 6484
+    },
+    {
+      "epoch": 0.6820303153798625,
+      "grad_norm": 2.3401833639741376,
+      "learning_rate": 1.1692130210292767e-06,
+      "loss": 0.9764,
+      "step": 6485
+    },
+    {
+      "epoch": 0.6821354858217098,
+      "grad_norm": 2.757772162036146,
+      "learning_rate": 1.1685063817399818e-06,
+      "loss": 0.9863,
+      "step": 6486
+    },
+    {
+      "epoch": 0.6822406562635571,
+      "grad_norm": 2.2309395979337774,
+      "learning_rate": 1.167799890922568e-06,
+      "loss": 0.9525,
+      "step": 6487
+    },
+    {
+      "epoch": 0.6823458267054044,
+      "grad_norm": 2.2761181614261123,
+      "learning_rate": 1.167093548655815e-06,
+      "loss": 0.9839,
+      "step": 6488
+    },
+    {
+      "epoch": 0.6824509971472518,
+      "grad_norm": 2.4073807837683474,
+      "learning_rate": 1.1663873550184864e-06,
+      "loss": 0.9943,
+      "step": 6489
+    },
+    {
+      "epoch": 0.6825561675890991,
+      "grad_norm": 2.344425968364666,
+      "learning_rate": 1.1656813100893271e-06,
+      "loss": 1.0342,
+      "step": 6490
+    },
+    {
+      "epoch": 0.6826613380309464,
+      "grad_norm": 3.3083451221904077,
+      "learning_rate": 1.1649754139470679e-06,
+      "loss": 0.9833,
+      "step": 6491
+    },
+    {
+      "epoch": 0.6827665084727937,
+      "grad_norm": 2.101235305234725,
+      "learning_rate": 1.16426966667042e-06,
+      "loss": 1.0127,
+      "step": 6492
+    },
+    {
+      "epoch": 0.6828716789146411,
+      "grad_norm": 2.4853559361171267,
+      "learning_rate": 1.1635640683380803e-06,
+      "loss": 0.9932,
+      "step": 6493
+    },
+    {
+      "epoch": 0.6829768493564884,
+      "grad_norm": 2.823621227962317,
+      "learning_rate": 1.1628586190287289e-06,
+      "loss": 0.9721,
+      "step": 6494
+    },
+    {
+      "epoch": 0.6830820197983357,
+      "grad_norm": 2.1446739627758924,
+      "learning_rate": 1.1621533188210296e-06,
+      "loss": 0.9461,
+      "step": 6495
+    },
+    {
+      "epoch": 0.683187190240183,
+      "grad_norm": 2.5010687220751877,
+      "learning_rate": 1.1614481677936274e-06,
+      "loss": 0.9768,
+      "step": 6496
+    },
+    {
+      "epoch": 0.6832923606820304,
+      "grad_norm": 2.5827579517567134,
+      "learning_rate": 1.1607431660251523e-06,
+      "loss": 0.9719,
+      "step": 6497
+    },
+    {
+      "epoch": 0.6833975311238777,
+      "grad_norm": 2.163533594993246,
+      "learning_rate": 1.160038313594219e-06,
+      "loss": 0.9844,
+      "step": 6498
+    },
+    {
+      "epoch": 0.6835027015657249,
+      "grad_norm": 1.6505556607260339,
+      "learning_rate": 1.1593336105794222e-06,
+      "loss": 0.9816,
+      "step": 6499
+    },
+    {
+      "epoch": 0.6836078720075722,
+      "grad_norm": 2.5798234985766753,
+      "learning_rate": 1.158629057059343e-06,
+      "loss": 0.9543,
+      "step": 6500
+    },
+    {
+      "epoch": 0.6837130424494196,
+      "grad_norm": 2.847502464596311,
+      "learning_rate": 1.1579246531125435e-06,
+      "loss": 0.9713,
+      "step": 6501
+    },
+    {
+      "epoch": 0.6838182128912669,
+      "grad_norm": 2.664122177395312,
+      "learning_rate": 1.1572203988175706e-06,
+      "loss": 1.0001,
+      "step": 6502
+    },
+    {
+      "epoch": 0.6839233833331142,
+      "grad_norm": 2.3631903419020603,
+      "learning_rate": 1.1565162942529553e-06,
+      "loss": 0.9806,
+      "step": 6503
+    },
+    {
+      "epoch": 0.6840285537749615,
+      "grad_norm": 2.140387924285255,
+      "learning_rate": 1.155812339497209e-06,
+      "loss": 1.0032,
+      "step": 6504
+    },
+    {
+      "epoch": 0.6841337242168088,
+      "grad_norm": 2.481719404717142,
+      "learning_rate": 1.1551085346288296e-06,
+      "loss": 1.0173,
+      "step": 6505
+    },
+    {
+      "epoch": 0.6842388946586562,
+      "grad_norm": 2.508106903316508,
+      "learning_rate": 1.154404879726294e-06,
+      "loss": 0.9962,
+      "step": 6506
+    },
+    {
+      "epoch": 0.6843440651005035,
+      "grad_norm": 3.216269422744731,
+      "learning_rate": 1.1537013748680694e-06,
+      "loss": 1.0013,
+      "step": 6507
+    },
+    {
+      "epoch": 0.6844492355423508,
+      "grad_norm": 2.0004248330968206,
+      "learning_rate": 1.152998020132599e-06,
+      "loss": 1.0044,
+      "step": 6508
+    },
+    {
+      "epoch": 0.6845544059841981,
+      "grad_norm": 1.933485559067847,
+      "learning_rate": 1.1522948155983143e-06,
+      "loss": 0.9502,
+      "step": 6509
+    },
+    {
+      "epoch": 0.6846595764260455,
+      "grad_norm": 2.6662844264965972,
+      "learning_rate": 1.1515917613436258e-06,
+      "loss": 0.9901,
+      "step": 6510
+    },
+    {
+      "epoch": 0.6847647468678928,
+      "grad_norm": 1.9043626170295296,
+      "learning_rate": 1.1508888574469308e-06,
+      "loss": 0.9483,
+      "step": 6511
+    },
+    {
+      "epoch": 0.6848699173097401,
+      "grad_norm": 2.889546124540106,
+      "learning_rate": 1.1501861039866095e-06,
+      "loss": 0.9652,
+      "step": 6512
+    },
+    {
+      "epoch": 0.6849750877515874,
+      "grad_norm": 2.297426343613282,
+      "learning_rate": 1.1494835010410222e-06,
+      "loss": 0.9644,
+      "step": 6513
+    },
+    {
+      "epoch": 0.6850802581934348,
+      "grad_norm": 2.5762672587290227,
+      "learning_rate": 1.1487810486885164e-06,
+      "loss": 0.919,
+      "step": 6514
+    },
+    {
+      "epoch": 0.6851854286352821,
+      "grad_norm": 3.1574284790806653,
+      "learning_rate": 1.1480787470074197e-06,
+      "loss": 0.9916,
+      "step": 6515
+    },
+    {
+      "epoch": 0.6852905990771294,
+      "grad_norm": 2.7014959221760697,
+      "learning_rate": 1.147376596076045e-06,
+      "loss": 0.9377,
+      "step": 6516
+    },
+    {
+      "epoch": 0.6853957695189767,
+      "grad_norm": 2.5585301645643783,
+      "learning_rate": 1.146674595972688e-06,
+      "loss": 0.9876,
+      "step": 6517
+    },
+    {
+      "epoch": 0.6855009399608241,
+      "grad_norm": 2.2873581952413944,
+      "learning_rate": 1.1459727467756257e-06,
+      "loss": 1.0167,
+      "step": 6518
+    },
+    {
+      "epoch": 0.6856061104026713,
+      "grad_norm": 2.0427362279367567,
+      "learning_rate": 1.1452710485631216e-06,
+      "loss": 1.0053,
+      "step": 6519
+    },
+    {
+      "epoch": 0.6857112808445186,
+      "grad_norm": 1.6816352696983445,
+      "learning_rate": 1.1445695014134175e-06,
+      "loss": 0.9847,
+      "step": 6520
+    },
+    {
+      "epoch": 0.6858164512863659,
+      "grad_norm": 2.8015068825343077,
+      "learning_rate": 1.1438681054047454e-06,
+      "loss": 0.9762,
+      "step": 6521
+    },
+    {
+      "epoch": 0.6859216217282132,
+      "grad_norm": 2.8607684877029493,
+      "learning_rate": 1.143166860615313e-06,
+      "loss": 0.9794,
+      "step": 6522
+    },
+    {
+      "epoch": 0.6860267921700606,
+      "grad_norm": 1.8264702204384304,
+      "learning_rate": 1.1424657671233175e-06,
+      "loss": 1.0061,
+      "step": 6523
+    },
+    {
+      "epoch": 0.6861319626119079,
+      "grad_norm": 2.5820509572402357,
+      "learning_rate": 1.1417648250069332e-06,
+      "loss": 1.0065,
+      "step": 6524
+    },
+    {
+      "epoch": 0.6862371330537552,
+      "grad_norm": 2.131759160671369,
+      "learning_rate": 1.1410640343443222e-06,
+      "loss": 0.9638,
+      "step": 6525
+    },
+    {
+      "epoch": 0.6863423034956025,
+      "grad_norm": 2.427313320161486,
+      "learning_rate": 1.140363395213629e-06,
+      "loss": 0.9769,
+      "step": 6526
+    },
+    {
+      "epoch": 0.6864474739374499,
+      "grad_norm": 2.762517347392917,
+      "learning_rate": 1.139662907692978e-06,
+      "loss": 1.0102,
+      "step": 6527
+    },
+    {
+      "epoch": 0.6865526443792972,
+      "grad_norm": 1.928573110997746,
+      "learning_rate": 1.1389625718604816e-06,
+      "loss": 0.9984,
+      "step": 6528
+    },
+    {
+      "epoch": 0.6866578148211445,
+      "grad_norm": 2.890165560340471,
+      "learning_rate": 1.1382623877942291e-06,
+      "loss": 1.0161,
+      "step": 6529
+    },
+    {
+      "epoch": 0.6867629852629918,
+      "grad_norm": 2.287013440890568,
+      "learning_rate": 1.1375623555723008e-06,
+      "loss": 1.0233,
+      "step": 6530
+    },
+    {
+      "epoch": 0.6868681557048392,
+      "grad_norm": 1.865259627410808,
+      "learning_rate": 1.1368624752727529e-06,
+      "loss": 0.9824,
+      "step": 6531
+    },
+    {
+      "epoch": 0.6869733261466865,
+      "grad_norm": 2.865076892904702,
+      "learning_rate": 1.1361627469736286e-06,
+      "loss": 0.9781,
+      "step": 6532
+    },
+    {
+      "epoch": 0.6870784965885338,
+      "grad_norm": 1.8570863276832512,
+      "learning_rate": 1.1354631707529532e-06,
+      "loss": 0.9515,
+      "step": 6533
+    },
+    {
+      "epoch": 0.6871836670303811,
+      "grad_norm": 2.1759244504535022,
+      "learning_rate": 1.1347637466887324e-06,
+      "loss": 0.9344,
+      "step": 6534
+    },
+    {
+      "epoch": 0.6872888374722285,
+      "grad_norm": 2.696477773718701,
+      "learning_rate": 1.134064474858961e-06,
+      "loss": 0.9849,
+      "step": 6535
+    },
+    {
+      "epoch": 0.6873940079140758,
+      "grad_norm": 2.0827168686133066,
+      "learning_rate": 1.1333653553416107e-06,
+      "loss": 0.948,
+      "step": 6536
+    },
+    {
+      "epoch": 0.6874991783559231,
+      "grad_norm": 2.5449063654411237,
+      "learning_rate": 1.1326663882146407e-06,
+      "loss": 0.9875,
+      "step": 6537
+    },
+    {
+      "epoch": 0.6876043487977704,
+      "grad_norm": 2.1105732172457667,
+      "learning_rate": 1.1319675735559894e-06,
+      "loss": 1.0455,
+      "step": 6538
+    },
+    {
+      "epoch": 0.6877095192396177,
+      "grad_norm": 2.6112809368518044,
+      "learning_rate": 1.1312689114435806e-06,
+      "loss": 0.9998,
+      "step": 6539
+    },
+    {
+      "epoch": 0.687814689681465,
+      "grad_norm": 2.6962129988071917,
+      "learning_rate": 1.130570401955322e-06,
+      "loss": 0.974,
+      "step": 6540
+    },
+    {
+      "epoch": 0.6879198601233123,
+      "grad_norm": 2.3841373338690954,
+      "learning_rate": 1.129872045169101e-06,
+      "loss": 0.9668,
+      "step": 6541
+    },
+    {
+      "epoch": 0.6880250305651596,
+      "grad_norm": 2.00322624680125,
+      "learning_rate": 1.1291738411627913e-06,
+      "loss": 0.9997,
+      "step": 6542
+    },
+    {
+      "epoch": 0.688130201007007,
+      "grad_norm": 2.391244626370067,
+      "learning_rate": 1.1284757900142451e-06,
+      "loss": 0.9691,
+      "step": 6543
+    },
+    {
+      "epoch": 0.6882353714488543,
+      "grad_norm": 1.7816408797129248,
+      "learning_rate": 1.1277778918013046e-06,
+      "loss": 1.0039,
+      "step": 6544
+    },
+    {
+      "epoch": 0.6883405418907016,
+      "grad_norm": 1.9496538742999459,
+      "learning_rate": 1.127080146601788e-06,
+      "loss": 0.9769,
+      "step": 6545
+    },
+    {
+      "epoch": 0.6884457123325489,
+      "grad_norm": 2.7789172568168166,
+      "learning_rate": 1.1263825544935015e-06,
+      "loss": 0.9584,
+      "step": 6546
+    },
+    {
+      "epoch": 0.6885508827743962,
+      "grad_norm": 2.116800668579364,
+      "learning_rate": 1.1256851155542297e-06,
+      "loss": 0.9908,
+      "step": 6547
+    },
+    {
+      "epoch": 0.6886560532162436,
+      "grad_norm": 2.942405980467766,
+      "learning_rate": 1.1249878298617436e-06,
+      "loss": 0.9636,
+      "step": 6548
+    },
+    {
+      "epoch": 0.6887612236580909,
+      "grad_norm": 2.213773047892503,
+      "learning_rate": 1.124290697493797e-06,
+      "loss": 0.9824,
+      "step": 6549
+    },
+    {
+      "epoch": 0.6888663940999382,
+      "grad_norm": 2.1291578558657394,
+      "learning_rate": 1.1235937185281234e-06,
+      "loss": 0.9922,
+      "step": 6550
+    },
+    {
+      "epoch": 0.6889715645417855,
+      "grad_norm": 2.724578811906178,
+      "learning_rate": 1.1228968930424433e-06,
+      "loss": 0.9929,
+      "step": 6551
+    },
+    {
+      "epoch": 0.6890767349836329,
+      "grad_norm": 2.0628702408800677,
+      "learning_rate": 1.1222002211144567e-06,
+      "loss": 0.9745,
+      "step": 6552
+    },
+    {
+      "epoch": 0.6891819054254802,
+      "grad_norm": 2.1610050278428083,
+      "learning_rate": 1.1215037028218484e-06,
+      "loss": 1.005,
+      "step": 6553
+    },
+    {
+      "epoch": 0.6892870758673275,
+      "grad_norm": 2.7787423829228874,
+      "learning_rate": 1.1208073382422866e-06,
+      "loss": 1.0103,
+      "step": 6554
+    },
+    {
+      "epoch": 0.6893922463091748,
+      "grad_norm": 2.3348483217669633,
+      "learning_rate": 1.1201111274534198e-06,
+      "loss": 0.9533,
+      "step": 6555
+    },
+    {
+      "epoch": 0.6894974167510222,
+      "grad_norm": 2.11428494125051,
+      "learning_rate": 1.1194150705328825e-06,
+      "loss": 1.0074,
+      "step": 6556
+    },
+    {
+      "epoch": 0.6896025871928695,
+      "grad_norm": 2.424483283901889,
+      "learning_rate": 1.1187191675582878e-06,
+      "loss": 1.0236,
+      "step": 6557
+    },
+    {
+      "epoch": 0.6897077576347168,
+      "grad_norm": 2.041193732298862,
+      "learning_rate": 1.1180234186072379e-06,
+      "loss": 0.9219,
+      "step": 6558
+    },
+    {
+      "epoch": 0.6898129280765641,
+      "grad_norm": 2.7736259010967075,
+      "learning_rate": 1.1173278237573113e-06,
+      "loss": 0.9917,
+      "step": 6559
+    },
+    {
+      "epoch": 0.6899180985184113,
+      "grad_norm": 2.479292231955384,
+      "learning_rate": 1.1166323830860745e-06,
+      "loss": 0.9836,
+      "step": 6560
+    },
+    {
+      "epoch": 0.6900232689602587,
+      "grad_norm": 2.286822176950764,
+      "learning_rate": 1.1159370966710723e-06,
+      "loss": 0.9802,
+      "step": 6561
+    },
+    {
+      "epoch": 0.690128439402106,
+      "grad_norm": 1.7894272908985807,
+      "learning_rate": 1.1152419645898355e-06,
+      "loss": 0.9678,
+      "step": 6562
+    },
+    {
+      "epoch": 0.6902336098439533,
+      "grad_norm": 3.264730137920692,
+      "learning_rate": 1.114546986919878e-06,
+      "loss": 1.0223,
+      "step": 6563
+    },
+    {
+      "epoch": 0.6903387802858006,
+      "grad_norm": 2.1228771331780125,
+      "learning_rate": 1.1138521637386928e-06,
+      "loss": 0.9932,
+      "step": 6564
+    },
+    {
+      "epoch": 0.690443950727648,
+      "grad_norm": 2.83441954898417,
+      "learning_rate": 1.1131574951237607e-06,
+      "loss": 0.9904,
+      "step": 6565
+    },
+    {
+      "epoch": 0.6905491211694953,
+      "grad_norm": 2.3866086648057,
+      "learning_rate": 1.112462981152539e-06,
+      "loss": 0.9736,
+      "step": 6566
+    },
+    {
+      "epoch": 0.6906542916113426,
+      "grad_norm": 2.0765111597644133,
+      "learning_rate": 1.1117686219024756e-06,
+      "loss": 0.9855,
+      "step": 6567
+    },
+    {
+      "epoch": 0.6907594620531899,
+      "grad_norm": 1.9430515714587882,
+      "learning_rate": 1.1110744174509952e-06,
+      "loss": 0.9586,
+      "step": 6568
+    },
+    {
+      "epoch": 0.6908646324950373,
+      "grad_norm": 2.722560114399417,
+      "learning_rate": 1.1103803678755058e-06,
+      "loss": 0.944,
+      "step": 6569
+    },
+    {
+      "epoch": 0.6909698029368846,
+      "grad_norm": 2.1402961489701267,
+      "learning_rate": 1.109686473253401e-06,
+      "loss": 0.995,
+      "step": 6570
+    },
+    {
+      "epoch": 0.6910749733787319,
+      "grad_norm": 2.7043321565248677,
+      "learning_rate": 1.1089927336620531e-06,
+      "loss": 1.0437,
+      "step": 6571
+    },
+    {
+      "epoch": 0.6911801438205792,
+      "grad_norm": 2.434297743887707,
+      "learning_rate": 1.108299149178823e-06,
+      "loss": 0.9765,
+      "step": 6572
+    },
+    {
+      "epoch": 0.6912853142624266,
+      "grad_norm": 1.8390287310817142,
+      "learning_rate": 1.107605719881048e-06,
+      "loss": 0.9612,
+      "step": 6573
+    },
+    {
+      "epoch": 0.6913904847042739,
+      "grad_norm": 2.270753478584313,
+      "learning_rate": 1.1069124458460528e-06,
+      "loss": 0.8822,
+      "step": 6574
+    },
+    {
+      "epoch": 0.6914956551461212,
+      "grad_norm": 2.1749999487851523,
+      "learning_rate": 1.1062193271511408e-06,
+      "loss": 0.9452,
+      "step": 6575
+    },
+    {
+      "epoch": 0.6916008255879685,
+      "grad_norm": 2.202741500698131,
+      "learning_rate": 1.1055263638736008e-06,
+      "loss": 0.9954,
+      "step": 6576
+    },
+    {
+      "epoch": 0.6917059960298159,
+      "grad_norm": 3.070731100268695,
+      "learning_rate": 1.1048335560907047e-06,
+      "loss": 1.0214,
+      "step": 6577
+    },
+    {
+      "epoch": 0.6918111664716632,
+      "grad_norm": 2.7005615503787426,
+      "learning_rate": 1.1041409038797047e-06,
+      "loss": 1.0112,
+      "step": 6578
+    },
+    {
+      "epoch": 0.6919163369135105,
+      "grad_norm": 2.7644388909239135,
+      "learning_rate": 1.1034484073178377e-06,
+      "loss": 0.9633,
+      "step": 6579
+    },
+    {
+      "epoch": 0.6920215073553577,
+      "grad_norm": 2.01408411568597,
+      "learning_rate": 1.1027560664823208e-06,
+      "loss": 1.0022,
+      "step": 6580
+    },
+    {
+      "epoch": 0.692126677797205,
+      "grad_norm": 2.4415146472180593,
+      "learning_rate": 1.102063881450358e-06,
+      "loss": 0.9889,
+      "step": 6581
+    },
+    {
+      "epoch": 0.6922318482390524,
+      "grad_norm": 2.1342536088258277,
+      "learning_rate": 1.1013718522991315e-06,
+      "loss": 1.0117,
+      "step": 6582
+    },
+    {
+      "epoch": 0.6923370186808997,
+      "grad_norm": 2.4132414775288296,
+      "learning_rate": 1.100679979105809e-06,
+      "loss": 0.9468,
+      "step": 6583
+    },
+    {
+      "epoch": 0.692442189122747,
+      "grad_norm": 2.2328463531395295,
+      "learning_rate": 1.0999882619475382e-06,
+      "loss": 0.9449,
+      "step": 6584
+    },
+    {
+      "epoch": 0.6925473595645943,
+      "grad_norm": 2.3588285134864524,
+      "learning_rate": 1.0992967009014522e-06,
+      "loss": 0.9846,
+      "step": 6585
+    },
+    {
+      "epoch": 0.6926525300064417,
+      "grad_norm": 2.3173283765307326,
+      "learning_rate": 1.0986052960446658e-06,
+      "loss": 1.011,
+      "step": 6586
+    },
+    {
+      "epoch": 0.692757700448289,
+      "grad_norm": 2.300519093867645,
+      "learning_rate": 1.0979140474542743e-06,
+      "loss": 1.0277,
+      "step": 6587
+    },
+    {
+      "epoch": 0.6928628708901363,
+      "grad_norm": 2.392157404408049,
+      "learning_rate": 1.0972229552073594e-06,
+      "loss": 0.9473,
+      "step": 6588
+    },
+    {
+      "epoch": 0.6929680413319836,
+      "grad_norm": 2.5842800419125784,
+      "learning_rate": 1.0965320193809808e-06,
+      "loss": 0.9278,
+      "step": 6589
+    },
+    {
+      "epoch": 0.693073211773831,
+      "grad_norm": 1.891437035514997,
+      "learning_rate": 1.0958412400521851e-06,
+      "loss": 0.9947,
+      "step": 6590
+    },
+    {
+      "epoch": 0.6931783822156783,
+      "grad_norm": 2.6484100957216365,
+      "learning_rate": 1.0951506172979998e-06,
+      "loss": 1.0276,
+      "step": 6591
+    },
+    {
+      "epoch": 0.6932835526575256,
+      "grad_norm": 2.0840632815442985,
+      "learning_rate": 1.0944601511954328e-06,
+      "loss": 0.993,
+      "step": 6592
+    },
+    {
+      "epoch": 0.6933887230993729,
+      "grad_norm": 2.956513595108928,
+      "learning_rate": 1.0937698418214784e-06,
+      "loss": 1.0338,
+      "step": 6593
+    },
+    {
+      "epoch": 0.6934938935412203,
+      "grad_norm": 1.9243522516943474,
+      "learning_rate": 1.0930796892531092e-06,
+      "loss": 0.9915,
+      "step": 6594
+    },
+    {
+      "epoch": 0.6935990639830676,
+      "grad_norm": 2.1698046683173136,
+      "learning_rate": 1.0923896935672856e-06,
+      "loss": 0.9747,
+      "step": 6595
+    },
+    {
+      "epoch": 0.6937042344249149,
+      "grad_norm": 2.5425082569792092,
+      "learning_rate": 1.0916998548409449e-06,
+      "loss": 0.9799,
+      "step": 6596
+    },
+    {
+      "epoch": 0.6938094048667622,
+      "grad_norm": 2.577847600360515,
+      "learning_rate": 1.0910101731510113e-06,
+      "loss": 0.9921,
+      "step": 6597
+    },
+    {
+      "epoch": 0.6939145753086096,
+      "grad_norm": 1.8248209965382798,
+      "learning_rate": 1.090320648574388e-06,
+      "loss": 0.969,
+      "step": 6598
+    },
+    {
+      "epoch": 0.6940197457504569,
+      "grad_norm": 1.7602736317885663,
+      "learning_rate": 1.0896312811879634e-06,
+      "loss": 0.9835,
+      "step": 6599
+    },
+    {
+      "epoch": 0.6941249161923041,
+      "grad_norm": 2.958715201212913,
+      "learning_rate": 1.0889420710686077e-06,
+      "loss": 0.9786,
+      "step": 6600
+    },
+    {
+      "epoch": 0.6942300866341514,
+      "grad_norm": 2.029850889738827,
+      "learning_rate": 1.0882530182931717e-06,
+      "loss": 0.9684,
+      "step": 6601
+    },
+    {
+      "epoch": 0.6943352570759987,
+      "grad_norm": 2.032629169019649,
+      "learning_rate": 1.0875641229384918e-06,
+      "loss": 0.9604,
+      "step": 6602
+    },
+    {
+      "epoch": 0.6944404275178461,
+      "grad_norm": 2.6284812089984864,
+      "learning_rate": 1.0868753850813826e-06,
+      "loss": 0.9786,
+      "step": 6603
+    },
+    {
+      "epoch": 0.6945455979596934,
+      "grad_norm": 2.0592326838128967,
+      "learning_rate": 1.0861868047986473e-06,
+      "loss": 0.942,
+      "step": 6604
+    },
+    {
+      "epoch": 0.6946507684015407,
+      "grad_norm": 1.850794804171779,
+      "learning_rate": 1.0854983821670665e-06,
+      "loss": 0.9761,
+      "step": 6605
+    },
+    {
+      "epoch": 0.694755938843388,
+      "grad_norm": 2.5669897777189434,
+      "learning_rate": 1.0848101172634028e-06,
+      "loss": 1.021,
+      "step": 6606
+    },
+    {
+      "epoch": 0.6948611092852354,
+      "grad_norm": 2.6120005510022666,
+      "learning_rate": 1.0841220101644063e-06,
+      "loss": 1.0407,
+      "step": 6607
+    },
+    {
+      "epoch": 0.6949662797270827,
+      "grad_norm": 2.119486727780465,
+      "learning_rate": 1.0834340609468022e-06,
+      "loss": 1.0054,
+      "step": 6608
+    },
+    {
+      "epoch": 0.69507145016893,
+      "grad_norm": 2.2087002638922493,
+      "learning_rate": 1.0827462696873065e-06,
+      "loss": 0.9827,
+      "step": 6609
+    },
+    {
+      "epoch": 0.6951766206107773,
+      "grad_norm": 2.0129214447538137,
+      "learning_rate": 1.0820586364626103e-06,
+      "loss": 0.9977,
+      "step": 6610
+    },
+    {
+      "epoch": 0.6952817910526247,
+      "grad_norm": 2.199166327772394,
+      "learning_rate": 1.0813711613493922e-06,
+      "loss": 0.9758,
+      "step": 6611
+    },
+    {
+      "epoch": 0.695386961494472,
+      "grad_norm": 3.246740679995433,
+      "learning_rate": 1.080683844424309e-06,
+      "loss": 0.9628,
+      "step": 6612
+    },
+    {
+      "epoch": 0.6954921319363193,
+      "grad_norm": 2.2023486928412592,
+      "learning_rate": 1.0799966857640027e-06,
+      "loss": 0.9798,
+      "step": 6613
+    },
+    {
+      "epoch": 0.6955973023781666,
+      "grad_norm": 2.548062183974951,
+      "learning_rate": 1.0793096854450979e-06,
+      "loss": 0.9919,
+      "step": 6614
+    },
+    {
+      "epoch": 0.695702472820014,
+      "grad_norm": 3.1434898532802324,
+      "learning_rate": 1.0786228435441984e-06,
+      "loss": 1.0162,
+      "step": 6615
+    },
+    {
+      "epoch": 0.6958076432618613,
+      "grad_norm": 2.310428364039209,
+      "learning_rate": 1.0779361601378946e-06,
+      "loss": 0.9854,
+      "step": 6616
+    },
+    {
+      "epoch": 0.6959128137037086,
+      "grad_norm": 2.688905242521263,
+      "learning_rate": 1.077249635302754e-06,
+      "loss": 1.0092,
+      "step": 6617
+    },
+    {
+      "epoch": 0.6960179841455559,
+      "grad_norm": 2.192140291940976,
+      "learning_rate": 1.0765632691153333e-06,
+      "loss": 0.9757,
+      "step": 6618
+    },
+    {
+      "epoch": 0.6961231545874033,
+      "grad_norm": 3.107859618660363,
+      "learning_rate": 1.0758770616521646e-06,
+      "loss": 0.9874,
+      "step": 6619
+    },
+    {
+      "epoch": 0.6962283250292506,
+      "grad_norm": 1.8725342958729618,
+      "learning_rate": 1.0751910129897678e-06,
+      "loss": 0.9366,
+      "step": 6620
+    },
+    {
+      "epoch": 0.6963334954710978,
+      "grad_norm": 2.3738236012939358,
+      "learning_rate": 1.0745051232046399e-06,
+      "loss": 0.9436,
+      "step": 6621
+    },
+    {
+      "epoch": 0.6964386659129451,
+      "grad_norm": 2.438216862745785,
+      "learning_rate": 1.0738193923732648e-06,
+      "loss": 0.9679,
+      "step": 6622
+    },
+    {
+      "epoch": 0.6965438363547924,
+      "grad_norm": 2.1364655426341947,
+      "learning_rate": 1.0731338205721072e-06,
+      "loss": 0.9434,
+      "step": 6623
+    },
+    {
+      "epoch": 0.6966490067966398,
+      "grad_norm": 3.5579171402660625,
+      "learning_rate": 1.0724484078776121e-06,
+      "loss": 1.0199,
+      "step": 6624
+    },
+    {
+      "epoch": 0.6967541772384871,
+      "grad_norm": 3.215088447682091,
+      "learning_rate": 1.0717631543662098e-06,
+      "loss": 0.9623,
+      "step": 6625
+    },
+    {
+      "epoch": 0.6968593476803344,
+      "grad_norm": 2.7336030353955234,
+      "learning_rate": 1.07107806011431e-06,
+      "loss": 1.0457,
+      "step": 6626
+    },
+    {
+      "epoch": 0.6969645181221817,
+      "grad_norm": 1.8278294216486561,
+      "learning_rate": 1.0703931251983068e-06,
+      "loss": 0.9591,
+      "step": 6627
+    },
+    {
+      "epoch": 0.6970696885640291,
+      "grad_norm": 2.316066385624242,
+      "learning_rate": 1.0697083496945766e-06,
+      "loss": 0.9709,
+      "step": 6628
+    },
+    {
+      "epoch": 0.6971748590058764,
+      "grad_norm": 2.7923043921417308,
+      "learning_rate": 1.0690237336794753e-06,
+      "loss": 1.0243,
+      "step": 6629
+    },
+    {
+      "epoch": 0.6972800294477237,
+      "grad_norm": 2.0978384840603974,
+      "learning_rate": 1.0683392772293446e-06,
+      "loss": 0.9568,
+      "step": 6630
+    },
+    {
+      "epoch": 0.697385199889571,
+      "grad_norm": 2.465884366366025,
+      "learning_rate": 1.0676549804205048e-06,
+      "loss": 0.9826,
+      "step": 6631
+    },
+    {
+      "epoch": 0.6974903703314184,
+      "grad_norm": 2.218104914348849,
+      "learning_rate": 1.0669708433292628e-06,
+      "loss": 0.9982,
+      "step": 6632
+    },
+    {
+      "epoch": 0.6975955407732657,
+      "grad_norm": 2.3301810291362233,
+      "learning_rate": 1.0662868660319031e-06,
+      "loss": 0.9371,
+      "step": 6633
+    },
+    {
+      "epoch": 0.697700711215113,
+      "grad_norm": 2.1441525236150634,
+      "learning_rate": 1.0656030486046965e-06,
+      "loss": 0.9976,
+      "step": 6634
+    },
+    {
+      "epoch": 0.6978058816569603,
+      "grad_norm": 2.557259001781231,
+      "learning_rate": 1.0649193911238918e-06,
+      "loss": 1.0343,
+      "step": 6635
+    },
+    {
+      "epoch": 0.6979110520988077,
+      "grad_norm": 2.5638852296182266,
+      "learning_rate": 1.064235893665723e-06,
+      "loss": 1.0068,
+      "step": 6636
+    },
+    {
+      "epoch": 0.698016222540655,
+      "grad_norm": 2.584073642881432,
+      "learning_rate": 1.063552556306406e-06,
+      "loss": 0.9623,
+      "step": 6637
+    },
+    {
+      "epoch": 0.6981213929825023,
+      "grad_norm": 2.3545330761686962,
+      "learning_rate": 1.0628693791221373e-06,
+      "loss": 1.0107,
+      "step": 6638
+    },
+    {
+      "epoch": 0.6982265634243496,
+      "grad_norm": 1.945064530584297,
+      "learning_rate": 1.0621863621890976e-06,
+      "loss": 0.9861,
+      "step": 6639
+    },
+    {
+      "epoch": 0.698331733866197,
+      "grad_norm": 2.0623306033113145,
+      "learning_rate": 1.061503505583447e-06,
+      "loss": 0.9795,
+      "step": 6640
+    },
+    {
+      "epoch": 0.6984369043080442,
+      "grad_norm": 2.077722057621392,
+      "learning_rate": 1.0608208093813299e-06,
+      "loss": 0.9851,
+      "step": 6641
+    },
+    {
+      "epoch": 0.6985420747498915,
+      "grad_norm": 2.304749737440805,
+      "learning_rate": 1.0601382736588735e-06,
+      "loss": 0.9362,
+      "step": 6642
+    },
+    {
+      "epoch": 0.6986472451917388,
+      "grad_norm": 2.6096925989976385,
+      "learning_rate": 1.059455898492184e-06,
+      "loss": 1.0156,
+      "step": 6643
+    },
+    {
+      "epoch": 0.6987524156335861,
+      "grad_norm": 2.201610817630821,
+      "learning_rate": 1.0587736839573525e-06,
+      "loss": 0.9828,
+      "step": 6644
+    },
+    {
+      "epoch": 0.6988575860754335,
+      "grad_norm": 1.8056953171438117,
+      "learning_rate": 1.058091630130451e-06,
+      "loss": 0.9644,
+      "step": 6645
+    },
+    {
+      "epoch": 0.6989627565172808,
+      "grad_norm": 1.966122087361216,
+      "learning_rate": 1.0574097370875346e-06,
+      "loss": 0.997,
+      "step": 6646
+    },
+    {
+      "epoch": 0.6990679269591281,
+      "grad_norm": 2.879109107208205,
+      "learning_rate": 1.0567280049046383e-06,
+      "loss": 0.9991,
+      "step": 6647
+    },
+    {
+      "epoch": 0.6991730974009754,
+      "grad_norm": 2.8573357829655404,
+      "learning_rate": 1.056046433657782e-06,
+      "loss": 1.0177,
+      "step": 6648
+    },
+    {
+      "epoch": 0.6992782678428228,
+      "grad_norm": 2.8507110133723863,
+      "learning_rate": 1.0553650234229642e-06,
+      "loss": 0.9884,
+      "step": 6649
+    },
+    {
+      "epoch": 0.6993834382846701,
+      "grad_norm": 1.8137801088961556,
+      "learning_rate": 1.054683774276169e-06,
+      "loss": 0.9959,
+      "step": 6650
+    },
+    {
+      "epoch": 0.6994886087265174,
+      "grad_norm": 2.6038698060831513,
+      "learning_rate": 1.0540026862933612e-06,
+      "loss": 1.0342,
+      "step": 6651
+    },
+    {
+      "epoch": 0.6995937791683647,
+      "grad_norm": 3.033587811054729,
+      "learning_rate": 1.0533217595504859e-06,
+      "loss": 0.9652,
+      "step": 6652
+    },
+    {
+      "epoch": 0.6996989496102121,
+      "grad_norm": 2.1979861729861923,
+      "learning_rate": 1.0526409941234728e-06,
+      "loss": 0.9613,
+      "step": 6653
+    },
+    {
+      "epoch": 0.6998041200520594,
+      "grad_norm": 1.958123885683921,
+      "learning_rate": 1.0519603900882322e-06,
+      "loss": 0.9348,
+      "step": 6654
+    },
+    {
+      "epoch": 0.6999092904939067,
+      "grad_norm": 2.6695995950616047,
+      "learning_rate": 1.0512799475206576e-06,
+      "loss": 1.0061,
+      "step": 6655
+    },
+    {
+      "epoch": 0.700014460935754,
+      "grad_norm": 2.3933085010130206,
+      "learning_rate": 1.050599666496623e-06,
+      "loss": 0.9999,
+      "step": 6656
+    },
+    {
+      "epoch": 0.7001196313776014,
+      "grad_norm": 2.4484506390441987,
+      "learning_rate": 1.0499195470919844e-06,
+      "loss": 0.9801,
+      "step": 6657
+    },
+    {
+      "epoch": 0.7002248018194487,
+      "grad_norm": 2.10318195808962,
+      "learning_rate": 1.0492395893825804e-06,
+      "loss": 0.9614,
+      "step": 6658
+    },
+    {
+      "epoch": 0.700329972261296,
+      "grad_norm": 2.354963154046998,
+      "learning_rate": 1.0485597934442323e-06,
+      "loss": 0.9967,
+      "step": 6659
+    },
+    {
+      "epoch": 0.7004351427031433,
+      "grad_norm": 2.4833792329948308,
+      "learning_rate": 1.0478801593527436e-06,
+      "loss": 0.9805,
+      "step": 6660
+    },
+    {
+      "epoch": 0.7005403131449905,
+      "grad_norm": 2.319460276654493,
+      "learning_rate": 1.0472006871838963e-06,
+      "loss": 0.9926,
+      "step": 6661
+    },
+    {
+      "epoch": 0.7006454835868379,
+      "grad_norm": 2.6492931586569033,
+      "learning_rate": 1.0465213770134591e-06,
+      "loss": 1.0135,
+      "step": 6662
+    },
+    {
+      "epoch": 0.7007506540286852,
+      "grad_norm": 3.0029420857299183,
+      "learning_rate": 1.0458422289171786e-06,
+      "loss": 0.9789,
+      "step": 6663
+    },
+    {
+      "epoch": 0.7008558244705325,
+      "grad_norm": 1.924682101684733,
+      "learning_rate": 1.0451632429707856e-06,
+      "loss": 0.9996,
+      "step": 6664
+    },
+    {
+      "epoch": 0.7009609949123798,
+      "grad_norm": 2.5935534112052725,
+      "learning_rate": 1.0444844192499938e-06,
+      "loss": 0.9762,
+      "step": 6665
+    },
+    {
+      "epoch": 0.7010661653542272,
+      "grad_norm": 2.4891701580984837,
+      "learning_rate": 1.043805757830495e-06,
+      "loss": 0.9801,
+      "step": 6666
+    },
+    {
+      "epoch": 0.7011713357960745,
+      "grad_norm": 3.171560808484776,
+      "learning_rate": 1.0431272587879662e-06,
+      "loss": 0.9521,
+      "step": 6667
+    },
+    {
+      "epoch": 0.7012765062379218,
+      "grad_norm": 2.360067551193987,
+      "learning_rate": 1.0424489221980657e-06,
+      "loss": 0.9547,
+      "step": 6668
+    },
+    {
+      "epoch": 0.7013816766797691,
+      "grad_norm": 2.136337297008342,
+      "learning_rate": 1.0417707481364338e-06,
+      "loss": 0.9991,
+      "step": 6669
+    },
+    {
+      "epoch": 0.7014868471216165,
+      "grad_norm": 2.070557413137076,
+      "learning_rate": 1.0410927366786904e-06,
+      "loss": 0.9918,
+      "step": 6670
+    },
+    {
+      "epoch": 0.7015920175634638,
+      "grad_norm": 2.0832696047671546,
+      "learning_rate": 1.0404148879004415e-06,
+      "loss": 1.0046,
+      "step": 6671
+    },
+    {
+      "epoch": 0.7016971880053111,
+      "grad_norm": 3.3334838695797115,
+      "learning_rate": 1.0397372018772694e-06,
+      "loss": 0.996,
+      "step": 6672
+    },
+    {
+      "epoch": 0.7018023584471584,
+      "grad_norm": 3.2162155727197406,
+      "learning_rate": 1.0390596786847435e-06,
+      "loss": 0.9941,
+      "step": 6673
+    },
+    {
+      "epoch": 0.7019075288890058,
+      "grad_norm": 2.9010911694149053,
+      "learning_rate": 1.0383823183984133e-06,
+      "loss": 1.0102,
+      "step": 6674
+    },
+    {
+      "epoch": 0.7020126993308531,
+      "grad_norm": 1.9879071940536923,
+      "learning_rate": 1.0377051210938077e-06,
+      "loss": 1.0026,
+      "step": 6675
+    },
+    {
+      "epoch": 0.7021178697727004,
+      "grad_norm": 2.911710480192307,
+      "learning_rate": 1.0370280868464405e-06,
+      "loss": 1.016,
+      "step": 6676
+    },
+    {
+      "epoch": 0.7022230402145477,
+      "grad_norm": 2.47624061900751,
+      "learning_rate": 1.0363512157318076e-06,
+      "loss": 1.0157,
+      "step": 6677
+    },
+    {
+      "epoch": 0.7023282106563951,
+      "grad_norm": 2.7429443996244687,
+      "learning_rate": 1.0356745078253833e-06,
+      "loss": 0.9474,
+      "step": 6678
+    },
+    {
+      "epoch": 0.7024333810982424,
+      "grad_norm": 1.6759259613705626,
+      "learning_rate": 1.0349979632026272e-06,
+      "loss": 0.9684,
+      "step": 6679
+    },
+    {
+      "epoch": 0.7025385515400897,
+      "grad_norm": 2.1862048185264458,
+      "learning_rate": 1.0343215819389782e-06,
+      "loss": 1.0272,
+      "step": 6680
+    },
+    {
+      "epoch": 0.702643721981937,
+      "grad_norm": 2.8521745744648492,
+      "learning_rate": 1.0336453641098584e-06,
+      "loss": 0.9889,
+      "step": 6681
+    },
+    {
+      "epoch": 0.7027488924237842,
+      "grad_norm": 2.6870393262822474,
+      "learning_rate": 1.0329693097906714e-06,
+      "loss": 1.0578,
+      "step": 6682
+    },
+    {
+      "epoch": 0.7028540628656316,
+      "grad_norm": 2.6616207546943484,
+      "learning_rate": 1.0322934190568037e-06,
+      "loss": 0.9533,
+      "step": 6683
+    },
+    {
+      "epoch": 0.7029592333074789,
+      "grad_norm": 1.7603354792631345,
+      "learning_rate": 1.0316176919836207e-06,
+      "loss": 0.9708,
+      "step": 6684
+    },
+    {
+      "epoch": 0.7030644037493262,
+      "grad_norm": 2.3418486553788216,
+      "learning_rate": 1.0309421286464724e-06,
+      "loss": 0.9603,
+      "step": 6685
+    },
+    {
+      "epoch": 0.7031695741911735,
+      "grad_norm": 2.6938896407952133,
+      "learning_rate": 1.030266729120688e-06,
+      "loss": 1.0238,
+      "step": 6686
+    },
+    {
+      "epoch": 0.7032747446330209,
+      "grad_norm": 2.035299190868889,
+      "learning_rate": 1.0295914934815806e-06,
+      "loss": 0.9898,
+      "step": 6687
+    },
+    {
+      "epoch": 0.7033799150748682,
+      "grad_norm": 2.0038429608901978,
+      "learning_rate": 1.0289164218044452e-06,
+      "loss": 0.9689,
+      "step": 6688
+    },
+    {
+      "epoch": 0.7034850855167155,
+      "grad_norm": 2.4755202430037913,
+      "learning_rate": 1.0282415141645554e-06,
+      "loss": 0.9555,
+      "step": 6689
+    },
+    {
+      "epoch": 0.7035902559585628,
+      "grad_norm": 3.468729544937565,
+      "learning_rate": 1.02756677063717e-06,
+      "loss": 0.998,
+      "step": 6690
+    },
+    {
+      "epoch": 0.7036954264004102,
+      "grad_norm": 2.6675685372943,
+      "learning_rate": 1.0268921912975288e-06,
+      "loss": 1.0056,
+      "step": 6691
+    },
+    {
+      "epoch": 0.7038005968422575,
+      "grad_norm": 2.500797177465148,
+      "learning_rate": 1.0262177762208508e-06,
+      "loss": 0.9976,
+      "step": 6692
+    },
+    {
+      "epoch": 0.7039057672841048,
+      "grad_norm": 1.843407519469925,
+      "learning_rate": 1.0255435254823404e-06,
+      "loss": 0.9219,
+      "step": 6693
+    },
+    {
+      "epoch": 0.7040109377259521,
+      "grad_norm": 2.646055721221685,
+      "learning_rate": 1.02486943915718e-06,
+      "loss": 0.9785,
+      "step": 6694
+    },
+    {
+      "epoch": 0.7041161081677995,
+      "grad_norm": 2.4428289065934554,
+      "learning_rate": 1.0241955173205366e-06,
+      "loss": 1.0164,
+      "step": 6695
+    },
+    {
+      "epoch": 0.7042212786096468,
+      "grad_norm": 2.4160293322018522,
+      "learning_rate": 1.0235217600475569e-06,
+      "loss": 1.0088,
+      "step": 6696
+    },
+    {
+      "epoch": 0.7043264490514941,
+      "grad_norm": 2.6869630614056756,
+      "learning_rate": 1.0228481674133719e-06,
+      "loss": 0.9751,
+      "step": 6697
+    },
+    {
+      "epoch": 0.7044316194933414,
+      "grad_norm": 2.230312539280299,
+      "learning_rate": 1.0221747394930904e-06,
+      "loss": 1.0008,
+      "step": 6698
+    },
+    {
+      "epoch": 0.7045367899351888,
+      "grad_norm": 2.218987057193651,
+      "learning_rate": 1.0215014763618054e-06,
+      "loss": 0.9617,
+      "step": 6699
+    },
+    {
+      "epoch": 0.7046419603770361,
+      "grad_norm": 2.5556328251578284,
+      "learning_rate": 1.020828378094592e-06,
+      "loss": 1.0079,
+      "step": 6700
+    },
+    {
+      "epoch": 0.7047471308188834,
+      "grad_norm": 2.1111202262894917,
+      "learning_rate": 1.0201554447665044e-06,
+      "loss": 0.9656,
+      "step": 6701
+    },
+    {
+      "epoch": 0.7048523012607306,
+      "grad_norm": 2.3057500589924422,
+      "learning_rate": 1.0194826764525811e-06,
+      "loss": 0.9667,
+      "step": 6702
+    },
+    {
+      "epoch": 0.7049574717025779,
+      "grad_norm": 2.438881329131322,
+      "learning_rate": 1.01881007322784e-06,
+      "loss": 1.0032,
+      "step": 6703
+    },
+    {
+      "epoch": 0.7050626421444253,
+      "grad_norm": 1.9015647486903267,
+      "learning_rate": 1.0181376351672817e-06,
+      "loss": 0.9893,
+      "step": 6704
+    },
+    {
+      "epoch": 0.7051678125862726,
+      "grad_norm": 2.3923551756484422,
+      "learning_rate": 1.0174653623458886e-06,
+      "loss": 0.987,
+      "step": 6705
+    },
+    {
+      "epoch": 0.7052729830281199,
+      "grad_norm": 2.4314151560926165,
+      "learning_rate": 1.0167932548386253e-06,
+      "loss": 1.0085,
+      "step": 6706
+    },
+    {
+      "epoch": 0.7053781534699672,
+      "grad_norm": 1.781610905130927,
+      "learning_rate": 1.016121312720436e-06,
+      "loss": 0.9624,
+      "step": 6707
+    },
+    {
+      "epoch": 0.7054833239118146,
+      "grad_norm": 2.3479526084995106,
+      "learning_rate": 1.0154495360662464e-06,
+      "loss": 1.0281,
+      "step": 6708
+    },
+    {
+      "epoch": 0.7055884943536619,
+      "grad_norm": 1.787929279606114,
+      "learning_rate": 1.0147779249509662e-06,
+      "loss": 0.9802,
+      "step": 6709
+    },
+    {
+      "epoch": 0.7056936647955092,
+      "grad_norm": 2.5085426770017407,
+      "learning_rate": 1.014106479449485e-06,
+      "loss": 0.9289,
+      "step": 6710
+    },
+    {
+      "epoch": 0.7057988352373565,
+      "grad_norm": 2.2725044679555713,
+      "learning_rate": 1.0134351996366749e-06,
+      "loss": 0.9641,
+      "step": 6711
+    },
+    {
+      "epoch": 0.7059040056792039,
+      "grad_norm": 2.364113081705728,
+      "learning_rate": 1.0127640855873874e-06,
+      "loss": 0.9668,
+      "step": 6712
+    },
+    {
+      "epoch": 0.7060091761210512,
+      "grad_norm": 3.297691941833725,
+      "learning_rate": 1.0120931373764572e-06,
+      "loss": 1.0055,
+      "step": 6713
+    },
+    {
+      "epoch": 0.7061143465628985,
+      "grad_norm": 2.1743176526856316,
+      "learning_rate": 1.011422355078702e-06,
+      "loss": 0.993,
+      "step": 6714
+    },
+    {
+      "epoch": 0.7062195170047458,
+      "grad_norm": 2.8680555666625676,
+      "learning_rate": 1.0107517387689168e-06,
+      "loss": 1.0052,
+      "step": 6715
+    },
+    {
+      "epoch": 0.7063246874465932,
+      "grad_norm": 2.3554559333451874,
+      "learning_rate": 1.0100812885218824e-06,
+      "loss": 0.9962,
+      "step": 6716
+    },
+    {
+      "epoch": 0.7064298578884405,
+      "grad_norm": 2.094277187480027,
+      "learning_rate": 1.0094110044123578e-06,
+      "loss": 0.9985,
+      "step": 6717
+    },
+    {
+      "epoch": 0.7065350283302878,
+      "grad_norm": 2.515371556231332,
+      "learning_rate": 1.0087408865150852e-06,
+      "loss": 1.0295,
+      "step": 6718
+    },
+    {
+      "epoch": 0.7066401987721351,
+      "grad_norm": 3.0627093185609064,
+      "learning_rate": 1.0080709349047885e-06,
+      "loss": 0.9567,
+      "step": 6719
+    },
+    {
+      "epoch": 0.7067453692139825,
+      "grad_norm": 2.8156486671988894,
+      "learning_rate": 1.007401149656173e-06,
+      "loss": 0.986,
+      "step": 6720
+    },
+    {
+      "epoch": 0.7068505396558298,
+      "grad_norm": 2.2979514616422376,
+      "learning_rate": 1.0067315308439235e-06,
+      "loss": 1.0037,
+      "step": 6721
+    },
+    {
+      "epoch": 0.706955710097677,
+      "grad_norm": 3.19623467418787,
+      "learning_rate": 1.0060620785427083e-06,
+      "loss": 1.021,
+      "step": 6722
+    },
+    {
+      "epoch": 0.7070608805395243,
+      "grad_norm": 2.3549166089077667,
+      "learning_rate": 1.0053927928271775e-06,
+      "loss": 0.9937,
+      "step": 6723
+    },
+    {
+      "epoch": 0.7071660509813716,
+      "grad_norm": 2.822764978609374,
+      "learning_rate": 1.00472367377196e-06,
+      "loss": 1.0179,
+      "step": 6724
+    },
+    {
+      "epoch": 0.707271221423219,
+      "grad_norm": 2.659932951697124,
+      "learning_rate": 1.0040547214516698e-06,
+      "loss": 0.9869,
+      "step": 6725
+    },
+    {
+      "epoch": 0.7073763918650663,
+      "grad_norm": 3.185470330945337,
+      "learning_rate": 1.0033859359408977e-06,
+      "loss": 0.9823,
+      "step": 6726
+    },
+    {
+      "epoch": 0.7074815623069136,
+      "grad_norm": 2.609628396699824,
+      "learning_rate": 1.00271731731422e-06,
+      "loss": 0.9795,
+      "step": 6727
+    },
+    {
+      "epoch": 0.7075867327487609,
+      "grad_norm": 1.9750399309722628,
+      "learning_rate": 1.0020488656461936e-06,
+      "loss": 0.9989,
+      "step": 6728
+    },
+    {
+      "epoch": 0.7076919031906083,
+      "grad_norm": 2.7806599676410335,
+      "learning_rate": 1.001380581011354e-06,
+      "loss": 1.015,
+      "step": 6729
+    },
+    {
+      "epoch": 0.7077970736324556,
+      "grad_norm": 2.4361344876429527,
+      "learning_rate": 1.0007124634842227e-06,
+      "loss": 0.9652,
+      "step": 6730
+    },
+    {
+      "epoch": 0.7079022440743029,
+      "grad_norm": 2.232929886411633,
+      "learning_rate": 1.0000445131392975e-06,
+      "loss": 0.9799,
+      "step": 6731
+    },
+    {
+      "epoch": 0.7080074145161502,
+      "grad_norm": 2.1763015794617924,
+      "learning_rate": 9.993767300510613e-07,
+      "loss": 0.9916,
+      "step": 6732
+    },
+    {
+      "epoch": 0.7081125849579976,
+      "grad_norm": 2.054775097950991,
+      "learning_rate": 9.987091142939766e-07,
+      "loss": 0.9821,
+      "step": 6733
+    },
+    {
+      "epoch": 0.7082177553998449,
+      "grad_norm": 3.014980389348752,
+      "learning_rate": 9.980416659424894e-07,
+      "loss": 0.9988,
+      "step": 6734
+    },
+    {
+      "epoch": 0.7083229258416922,
+      "grad_norm": 2.2677152224867507,
+      "learning_rate": 9.97374385071023e-07,
+      "loss": 0.9856,
+      "step": 6735
+    },
+    {
+      "epoch": 0.7084280962835395,
+      "grad_norm": 2.235863059602925,
+      "learning_rate": 9.967072717539852e-07,
+      "loss": 0.9791,
+      "step": 6736
+    },
+    {
+      "epoch": 0.7085332667253869,
+      "grad_norm": 2.2178818893528867,
+      "learning_rate": 9.960403260657658e-07,
+      "loss": 0.9635,
+      "step": 6737
+    },
+    {
+      "epoch": 0.7086384371672342,
+      "grad_norm": 2.2511311829581184,
+      "learning_rate": 9.953735480807322e-07,
+      "loss": 0.9609,
+      "step": 6738
+    },
+    {
+      "epoch": 0.7087436076090815,
+      "grad_norm": 2.385032459651381,
+      "learning_rate": 9.947069378732372e-07,
+      "loss": 0.9845,
+      "step": 6739
+    },
+    {
+      "epoch": 0.7088487780509288,
+      "grad_norm": 1.9674072764447825,
+      "learning_rate": 9.940404955176114e-07,
+      "loss": 0.9351,
+      "step": 6740
+    },
+    {
+      "epoch": 0.7089539484927762,
+      "grad_norm": 2.4743599480951404,
+      "learning_rate": 9.933742210881688e-07,
+      "loss": 0.9413,
+      "step": 6741
+    },
+    {
+      "epoch": 0.7090591189346235,
+      "grad_norm": 2.0926323669690037,
+      "learning_rate": 9.927081146592058e-07,
+      "loss": 1.0089,
+      "step": 6742
+    },
+    {
+      "epoch": 0.7091642893764707,
+      "grad_norm": 2.5941026743219453,
+      "learning_rate": 9.920421763049957e-07,
+      "loss": 1.0047,
+      "step": 6743
+    },
+    {
+      "epoch": 0.709269459818318,
+      "grad_norm": 2.8240047908465216,
+      "learning_rate": 9.913764060997982e-07,
+      "loss": 0.972,
+      "step": 6744
+    },
+    {
+      "epoch": 0.7093746302601653,
+      "grad_norm": 2.284340896157482,
+      "learning_rate": 9.90710804117849e-07,
+      "loss": 0.9517,
+      "step": 6745
+    },
+    {
+      "epoch": 0.7094798007020127,
+      "grad_norm": 2.508946771542956,
+      "learning_rate": 9.900453704333718e-07,
+      "loss": 0.9857,
+      "step": 6746
+    },
+    {
+      "epoch": 0.70958497114386,
+      "grad_norm": 2.836809183347872,
+      "learning_rate": 9.893801051205643e-07,
+      "loss": 0.9961,
+      "step": 6747
+    },
+    {
+      "epoch": 0.7096901415857073,
+      "grad_norm": 2.2219358660428927,
+      "learning_rate": 9.88715008253611e-07,
+      "loss": 0.9792,
+      "step": 6748
+    },
+    {
+      "epoch": 0.7097953120275546,
+      "grad_norm": 2.7231291268770046,
+      "learning_rate": 9.880500799066734e-07,
+      "loss": 1.0135,
+      "step": 6749
+    },
+    {
+      "epoch": 0.709900482469402,
+      "grad_norm": 2.2397683772339847,
+      "learning_rate": 9.873853201538972e-07,
+      "loss": 0.9937,
+      "step": 6750
+    },
+    {
+      "epoch": 0.7100056529112493,
+      "grad_norm": 2.4531999818310926,
+      "learning_rate": 9.86720729069409e-07,
+      "loss": 0.9821,
+      "step": 6751
+    },
+    {
+      "epoch": 0.7101108233530966,
+      "grad_norm": 2.895236414074528,
+      "learning_rate": 9.860563067273142e-07,
+      "loss": 0.9877,
+      "step": 6752
+    },
+    {
+      "epoch": 0.7102159937949439,
+      "grad_norm": 2.142507657879859,
+      "learning_rate": 9.853920532017027e-07,
+      "loss": 1.0331,
+      "step": 6753
+    },
+    {
+      "epoch": 0.7103211642367913,
+      "grad_norm": 2.1738833228712044,
+      "learning_rate": 9.847279685666425e-07,
+      "loss": 0.9896,
+      "step": 6754
+    },
+    {
+      "epoch": 0.7104263346786386,
+      "grad_norm": 4.407965262891168,
+      "learning_rate": 9.840640528961849e-07,
+      "loss": 1.0162,
+      "step": 6755
+    },
+    {
+      "epoch": 0.7105315051204859,
+      "grad_norm": 3.380279233475454,
+      "learning_rate": 9.834003062643616e-07,
+      "loss": 0.9877,
+      "step": 6756
+    },
+    {
+      "epoch": 0.7106366755623332,
+      "grad_norm": 2.2302312037350207,
+      "learning_rate": 9.82736728745186e-07,
+      "loss": 0.9992,
+      "step": 6757
+    },
+    {
+      "epoch": 0.7107418460041806,
+      "grad_norm": 2.0718248730773996,
+      "learning_rate": 9.82073320412652e-07,
+      "loss": 0.9631,
+      "step": 6758
+    },
+    {
+      "epoch": 0.7108470164460279,
+      "grad_norm": 2.766749356649403,
+      "learning_rate": 9.814100813407326e-07,
+      "loss": 1.0268,
+      "step": 6759
+    },
+    {
+      "epoch": 0.7109521868878752,
+      "grad_norm": 2.7394051021844597,
+      "learning_rate": 9.807470116033879e-07,
+      "loss": 0.9684,
+      "step": 6760
+    },
+    {
+      "epoch": 0.7110573573297225,
+      "grad_norm": 2.576083662654087,
+      "learning_rate": 9.800841112745524e-07,
+      "loss": 0.9926,
+      "step": 6761
+    },
+    {
+      "epoch": 0.7111625277715699,
+      "grad_norm": 1.998383130036782,
+      "learning_rate": 9.794213804281463e-07,
+      "loss": 0.9218,
+      "step": 6762
+    },
+    {
+      "epoch": 0.7112676982134171,
+      "grad_norm": 2.613732532405646,
+      "learning_rate": 9.78758819138068e-07,
+      "loss": 1.0229,
+      "step": 6763
+    },
+    {
+      "epoch": 0.7113728686552644,
+      "grad_norm": 2.421476761648599,
+      "learning_rate": 9.780964274781984e-07,
+      "loss": 1.0062,
+      "step": 6764
+    },
+    {
+      "epoch": 0.7114780390971117,
+      "grad_norm": 2.13715013448252,
+      "learning_rate": 9.774342055224006e-07,
+      "loss": 1.0194,
+      "step": 6765
+    },
+    {
+      "epoch": 0.711583209538959,
+      "grad_norm": 2.3786521850695044,
+      "learning_rate": 9.76772153344516e-07,
+      "loss": 0.9514,
+      "step": 6766
+    },
+    {
+      "epoch": 0.7116883799808064,
+      "grad_norm": 2.2713978142889677,
+      "learning_rate": 9.761102710183698e-07,
+      "loss": 0.9796,
+      "step": 6767
+    },
+    {
+      "epoch": 0.7117935504226537,
+      "grad_norm": 2.165799232271969,
+      "learning_rate": 9.754485586177648e-07,
+      "loss": 0.9959,
+      "step": 6768
+    },
+    {
+      "epoch": 0.711898720864501,
+      "grad_norm": 1.8842675327288372,
+      "learning_rate": 9.747870162164903e-07,
+      "loss": 1.0099,
+      "step": 6769
+    },
+    {
+      "epoch": 0.7120038913063483,
+      "grad_norm": 1.9989490315740839,
+      "learning_rate": 9.741256438883108e-07,
+      "loss": 0.9783,
+      "step": 6770
+    },
+    {
+      "epoch": 0.7121090617481957,
+      "grad_norm": 3.0192380084749053,
+      "learning_rate": 9.734644417069764e-07,
+      "loss": 0.9645,
+      "step": 6771
+    },
+    {
+      "epoch": 0.712214232190043,
+      "grad_norm": 2.5407462322629355,
+      "learning_rate": 9.728034097462144e-07,
+      "loss": 0.9896,
+      "step": 6772
+    },
+    {
+      "epoch": 0.7123194026318903,
+      "grad_norm": 2.4113869913201023,
+      "learning_rate": 9.721425480797358e-07,
+      "loss": 0.9704,
+      "step": 6773
+    },
+    {
+      "epoch": 0.7124245730737376,
+      "grad_norm": 2.1235306653145734,
+      "learning_rate": 9.714818567812329e-07,
+      "loss": 1.0002,
+      "step": 6774
+    },
+    {
+      "epoch": 0.712529743515585,
+      "grad_norm": 2.141043980222307,
+      "learning_rate": 9.708213359243762e-07,
+      "loss": 0.9717,
+      "step": 6775
+    },
+    {
+      "epoch": 0.7126349139574323,
+      "grad_norm": 2.246816431557711,
+      "learning_rate": 9.701609855828202e-07,
+      "loss": 0.9864,
+      "step": 6776
+    },
+    {
+      "epoch": 0.7127400843992796,
+      "grad_norm": 2.2375134836918664,
+      "learning_rate": 9.695008058301978e-07,
+      "loss": 0.9901,
+      "step": 6777
+    },
+    {
+      "epoch": 0.7128452548411269,
+      "grad_norm": 2.9228849737443614,
+      "learning_rate": 9.688407967401248e-07,
+      "loss": 0.9737,
+      "step": 6778
+    },
+    {
+      "epoch": 0.7129504252829743,
+      "grad_norm": 2.8328425019223253,
+      "learning_rate": 9.681809583861982e-07,
+      "loss": 0.992,
+      "step": 6779
+    },
+    {
+      "epoch": 0.7130555957248216,
+      "grad_norm": 2.7116111720143743,
+      "learning_rate": 9.675212908419937e-07,
+      "loss": 0.9816,
+      "step": 6780
+    },
+    {
+      "epoch": 0.7131607661666689,
+      "grad_norm": 2.5984134719702983,
+      "learning_rate": 9.668617941810708e-07,
+      "loss": 0.9631,
+      "step": 6781
+    },
+    {
+      "epoch": 0.7132659366085162,
+      "grad_norm": 1.9628235706550243,
+      "learning_rate": 9.662024684769658e-07,
+      "loss": 0.9091,
+      "step": 6782
+    },
+    {
+      "epoch": 0.7133711070503634,
+      "grad_norm": 2.953890424536143,
+      "learning_rate": 9.655433138032022e-07,
+      "loss": 0.9763,
+      "step": 6783
+    },
+    {
+      "epoch": 0.7134762774922108,
+      "grad_norm": 1.873848783671468,
+      "learning_rate": 9.648843302332786e-07,
+      "loss": 0.9946,
+      "step": 6784
+    },
+    {
+      "epoch": 0.7135814479340581,
+      "grad_norm": 2.812780437855715,
+      "learning_rate": 9.642255178406782e-07,
+      "loss": 0.9467,
+      "step": 6785
+    },
+    {
+      "epoch": 0.7136866183759054,
+      "grad_norm": 2.0608044605763083,
+      "learning_rate": 9.635668766988618e-07,
+      "loss": 0.9883,
+      "step": 6786
+    },
+    {
+      "epoch": 0.7137917888177527,
+      "grad_norm": 2.38366077999555,
+      "learning_rate": 9.629084068812742e-07,
+      "loss": 0.9696,
+      "step": 6787
+    },
+    {
+      "epoch": 0.7138969592596001,
+      "grad_norm": 2.7383376317409844,
+      "learning_rate": 9.622501084613407e-07,
+      "loss": 0.9579,
+      "step": 6788
+    },
+    {
+      "epoch": 0.7140021297014474,
+      "grad_norm": 3.313699091586876,
+      "learning_rate": 9.615919815124647e-07,
+      "loss": 0.9894,
+      "step": 6789
+    },
+    {
+      "epoch": 0.7141073001432947,
+      "grad_norm": 2.0105423002645364,
+      "learning_rate": 9.609340261080343e-07,
+      "loss": 0.9843,
+      "step": 6790
+    },
+    {
+      "epoch": 0.714212470585142,
+      "grad_norm": 3.588279258448056,
+      "learning_rate": 9.602762423214146e-07,
+      "loss": 1.0235,
+      "step": 6791
+    },
+    {
+      "epoch": 0.7143176410269894,
+      "grad_norm": 2.430588136927392,
+      "learning_rate": 9.596186302259563e-07,
+      "loss": 1.0016,
+      "step": 6792
+    },
+    {
+      "epoch": 0.7144228114688367,
+      "grad_norm": 2.1694545531760423,
+      "learning_rate": 9.589611898949868e-07,
+      "loss": 0.9588,
+      "step": 6793
+    },
+    {
+      "epoch": 0.714527981910684,
+      "grad_norm": 2.4149358677574773,
+      "learning_rate": 9.583039214018152e-07,
+      "loss": 0.9814,
+      "step": 6794
+    },
+    {
+      "epoch": 0.7146331523525313,
+      "grad_norm": 2.272690222416524,
+      "learning_rate": 9.576468248197335e-07,
+      "loss": 0.9828,
+      "step": 6795
+    },
+    {
+      "epoch": 0.7147383227943787,
+      "grad_norm": 2.4404468267875363,
+      "learning_rate": 9.569899002220104e-07,
+      "loss": 0.9592,
+      "step": 6796
+    },
+    {
+      "epoch": 0.714843493236226,
+      "grad_norm": 2.481227510357708,
+      "learning_rate": 9.563331476819019e-07,
+      "loss": 0.9432,
+      "step": 6797
+    },
+    {
+      "epoch": 0.7149486636780733,
+      "grad_norm": 2.1951734767832924,
+      "learning_rate": 9.55676567272638e-07,
+      "loss": 1.0097,
+      "step": 6798
+    },
+    {
+      "epoch": 0.7150538341199206,
+      "grad_norm": 2.4639332073414137,
+      "learning_rate": 9.550201590674343e-07,
+      "loss": 0.9852,
+      "step": 6799
+    },
+    {
+      "epoch": 0.715159004561768,
+      "grad_norm": 2.1444939700922254,
+      "learning_rate": 9.54363923139484e-07,
+      "loss": 0.9932,
+      "step": 6800
+    },
+    {
+      "epoch": 0.7152641750036153,
+      "grad_norm": 2.9435311101102988,
+      "learning_rate": 9.53707859561963e-07,
+      "loss": 0.9227,
+      "step": 6801
+    },
+    {
+      "epoch": 0.7153693454454626,
+      "grad_norm": 2.7613347609604024,
+      "learning_rate": 9.530519684080289e-07,
+      "loss": 1.0551,
+      "step": 6802
+    },
+    {
+      "epoch": 0.7154745158873099,
+      "grad_norm": 3.138642130105925,
+      "learning_rate": 9.523962497508163e-07,
+      "loss": 0.9815,
+      "step": 6803
+    },
+    {
+      "epoch": 0.7155796863291571,
+      "grad_norm": 2.6016824562031577,
+      "learning_rate": 9.517407036634449e-07,
+      "loss": 0.9368,
+      "step": 6804
+    },
+    {
+      "epoch": 0.7156848567710045,
+      "grad_norm": 2.9399166189880686,
+      "learning_rate": 9.510853302190107e-07,
+      "loss": 0.9755,
+      "step": 6805
+    },
+    {
+      "epoch": 0.7157900272128518,
+      "grad_norm": 2.3854715791155123,
+      "learning_rate": 9.504301294905966e-07,
+      "loss": 1.0069,
+      "step": 6806
+    },
+    {
+      "epoch": 0.7158951976546991,
+      "grad_norm": 2.3367169758597504,
+      "learning_rate": 9.497751015512593e-07,
+      "loss": 0.9986,
+      "step": 6807
+    },
+    {
+      "epoch": 0.7160003680965464,
+      "grad_norm": 2.8377663825658592,
+      "learning_rate": 9.491202464740415e-07,
+      "loss": 1.0053,
+      "step": 6808
+    },
+    {
+      "epoch": 0.7161055385383938,
+      "grad_norm": 2.390482144562161,
+      "learning_rate": 9.484655643319643e-07,
+      "loss": 0.9992,
+      "step": 6809
+    },
+    {
+      "epoch": 0.7162107089802411,
+      "grad_norm": 2.942094456519339,
+      "learning_rate": 9.478110551980274e-07,
+      "loss": 1.0179,
+      "step": 6810
+    },
+    {
+      "epoch": 0.7163158794220884,
+      "grad_norm": 2.360695205905589,
+      "learning_rate": 9.471567191452175e-07,
+      "loss": 1.0237,
+      "step": 6811
+    },
+    {
+      "epoch": 0.7164210498639357,
+      "grad_norm": 2.403024984734448,
+      "learning_rate": 9.465025562464952e-07,
+      "loss": 0.99,
+      "step": 6812
+    },
+    {
+      "epoch": 0.7165262203057831,
+      "grad_norm": 2.3364437770617554,
+      "learning_rate": 9.458485665748071e-07,
+      "loss": 0.9672,
+      "step": 6813
+    },
+    {
+      "epoch": 0.7166313907476304,
+      "grad_norm": 2.158609622375604,
+      "learning_rate": 9.451947502030759e-07,
+      "loss": 0.9517,
+      "step": 6814
+    },
+    {
+      "epoch": 0.7167365611894777,
+      "grad_norm": 2.1055972278729715,
+      "learning_rate": 9.445411072042083e-07,
+      "loss": 0.964,
+      "step": 6815
+    },
+    {
+      "epoch": 0.716841731631325,
+      "grad_norm": 2.6035591950083155,
+      "learning_rate": 9.438876376510911e-07,
+      "loss": 0.9857,
+      "step": 6816
+    },
+    {
+      "epoch": 0.7169469020731724,
+      "grad_norm": 2.834112744073984,
+      "learning_rate": 9.432343416165899e-07,
+      "loss": 1.0336,
+      "step": 6817
+    },
+    {
+      "epoch": 0.7170520725150197,
+      "grad_norm": 2.479880902942861,
+      "learning_rate": 9.425812191735539e-07,
+      "loss": 1.0126,
+      "step": 6818
+    },
+    {
+      "epoch": 0.717157242956867,
+      "grad_norm": 1.8559851857923326,
+      "learning_rate": 9.419282703948085e-07,
+      "loss": 0.9631,
+      "step": 6819
+    },
+    {
+      "epoch": 0.7172624133987143,
+      "grad_norm": 1.8082543081529459,
+      "learning_rate": 9.412754953531664e-07,
+      "loss": 0.99,
+      "step": 6820
+    },
+    {
+      "epoch": 0.7173675838405617,
+      "grad_norm": 2.2523841295243123,
+      "learning_rate": 9.406228941214143e-07,
+      "loss": 0.9886,
+      "step": 6821
+    },
+    {
+      "epoch": 0.717472754282409,
+      "grad_norm": 2.457867045065379,
+      "learning_rate": 9.399704667723239e-07,
+      "loss": 1.0095,
+      "step": 6822
+    },
+    {
+      "epoch": 0.7175779247242563,
+      "grad_norm": 3.355058501321726,
+      "learning_rate": 9.393182133786443e-07,
+      "loss": 1.0292,
+      "step": 6823
+    },
+    {
+      "epoch": 0.7176830951661035,
+      "grad_norm": 2.3934295140430466,
+      "learning_rate": 9.386661340131078e-07,
+      "loss": 0.9797,
+      "step": 6824
+    },
+    {
+      "epoch": 0.7177882656079508,
+      "grad_norm": 4.421047798769204,
+      "learning_rate": 9.380142287484273e-07,
+      "loss": 0.9442,
+      "step": 6825
+    },
+    {
+      "epoch": 0.7178934360497982,
+      "grad_norm": 2.1419410112275603,
+      "learning_rate": 9.373624976572931e-07,
+      "loss": 1.0034,
+      "step": 6826
+    },
+    {
+      "epoch": 0.7179986064916455,
+      "grad_norm": 2.446158128193031,
+      "learning_rate": 9.367109408123803e-07,
+      "loss": 0.9845,
+      "step": 6827
+    },
+    {
+      "epoch": 0.7181037769334928,
+      "grad_norm": 2.8329699515453965,
+      "learning_rate": 9.3605955828634e-07,
+      "loss": 1.0246,
+      "step": 6828
+    },
+    {
+      "epoch": 0.7182089473753401,
+      "grad_norm": 2.780651175983165,
+      "learning_rate": 9.354083501518097e-07,
+      "loss": 1.0073,
+      "step": 6829
+    },
+    {
+      "epoch": 0.7183141178171875,
+      "grad_norm": 2.4642646857763753,
+      "learning_rate": 9.347573164814025e-07,
+      "loss": 0.9969,
+      "step": 6830
+    },
+    {
+      "epoch": 0.7184192882590348,
+      "grad_norm": 2.8457696249018243,
+      "learning_rate": 9.34106457347713e-07,
+      "loss": 1.0158,
+      "step": 6831
+    },
+    {
+      "epoch": 0.7185244587008821,
+      "grad_norm": 2.5543405389378275,
+      "learning_rate": 9.334557728233185e-07,
+      "loss": 0.9736,
+      "step": 6832
+    },
+    {
+      "epoch": 0.7186296291427294,
+      "grad_norm": 1.8641262528029885,
+      "learning_rate": 9.328052629807729e-07,
+      "loss": 0.98,
+      "step": 6833
+    },
+    {
+      "epoch": 0.7187347995845768,
+      "grad_norm": 2.5847950268121416,
+      "learning_rate": 9.32154927892617e-07,
+      "loss": 0.9441,
+      "step": 6834
+    },
+    {
+      "epoch": 0.7188399700264241,
+      "grad_norm": 2.129043724063401,
+      "learning_rate": 9.315047676313648e-07,
+      "loss": 0.9554,
+      "step": 6835
+    },
+    {
+      "epoch": 0.7189451404682714,
+      "grad_norm": 2.324839688594734,
+      "learning_rate": 9.308547822695166e-07,
+      "loss": 1.0237,
+      "step": 6836
+    },
+    {
+      "epoch": 0.7190503109101187,
+      "grad_norm": 1.9345404654516318,
+      "learning_rate": 9.302049718795489e-07,
+      "loss": 0.9591,
+      "step": 6837
+    },
+    {
+      "epoch": 0.719155481351966,
+      "grad_norm": 1.7931929866422445,
+      "learning_rate": 9.295553365339213e-07,
+      "loss": 0.9902,
+      "step": 6838
+    },
+    {
+      "epoch": 0.7192606517938134,
+      "grad_norm": 2.244608243473127,
+      "learning_rate": 9.289058763050743e-07,
+      "loss": 0.9788,
+      "step": 6839
+    },
+    {
+      "epoch": 0.7193658222356607,
+      "grad_norm": 2.381440258340103,
+      "learning_rate": 9.282565912654257e-07,
+      "loss": 0.9811,
+      "step": 6840
+    },
+    {
+      "epoch": 0.719470992677508,
+      "grad_norm": 2.645643314379612,
+      "learning_rate": 9.276074814873778e-07,
+      "loss": 0.9656,
+      "step": 6841
+    },
+    {
+      "epoch": 0.7195761631193553,
+      "grad_norm": 2.2621136426873374,
+      "learning_rate": 9.26958547043309e-07,
+      "loss": 0.9748,
+      "step": 6842
+    },
+    {
+      "epoch": 0.7196813335612027,
+      "grad_norm": 2.3967219964591155,
+      "learning_rate": 9.263097880055835e-07,
+      "loss": 0.9744,
+      "step": 6843
+    },
+    {
+      "epoch": 0.7197865040030499,
+      "grad_norm": 3.4195444597169553,
+      "learning_rate": 9.256612044465407e-07,
+      "loss": 1.0398,
+      "step": 6844
+    },
+    {
+      "epoch": 0.7198916744448972,
+      "grad_norm": 1.8349056835941067,
+      "learning_rate": 9.250127964385045e-07,
+      "loss": 0.9782,
+      "step": 6845
+    },
+    {
+      "epoch": 0.7199968448867445,
+      "grad_norm": 3.080343200685503,
+      "learning_rate": 9.243645640537755e-07,
+      "loss": 0.9775,
+      "step": 6846
+    },
+    {
+      "epoch": 0.7201020153285919,
+      "grad_norm": 2.207569417842817,
+      "learning_rate": 9.237165073646376e-07,
+      "loss": 0.9864,
+      "step": 6847
+    },
+    {
+      "epoch": 0.7202071857704392,
+      "grad_norm": 1.9935368383262349,
+      "learning_rate": 9.230686264433547e-07,
+      "loss": 0.962,
+      "step": 6848
+    },
+    {
+      "epoch": 0.7203123562122865,
+      "grad_norm": 3.0768788487950114,
+      "learning_rate": 9.224209213621693e-07,
+      "loss": 0.9913,
+      "step": 6849
+    },
+    {
+      "epoch": 0.7204175266541338,
+      "grad_norm": 3.593231825237885,
+      "learning_rate": 9.217733921933073e-07,
+      "loss": 1.0568,
+      "step": 6850
+    },
+    {
+      "epoch": 0.7205226970959812,
+      "grad_norm": 2.35712856290603,
+      "learning_rate": 9.21126039008971e-07,
+      "loss": 1.0247,
+      "step": 6851
+    },
+    {
+      "epoch": 0.7206278675378285,
+      "grad_norm": 2.060779548724852,
+      "learning_rate": 9.204788618813468e-07,
+      "loss": 0.986,
+      "step": 6852
+    },
+    {
+      "epoch": 0.7207330379796758,
+      "grad_norm": 1.8881478246910306,
+      "learning_rate": 9.198318608826001e-07,
+      "loss": 0.9918,
+      "step": 6853
+    },
+    {
+      "epoch": 0.7208382084215231,
+      "grad_norm": 1.9601162572993023,
+      "learning_rate": 9.191850360848756e-07,
+      "loss": 0.9542,
+      "step": 6854
+    },
+    {
+      "epoch": 0.7209433788633705,
+      "grad_norm": 1.982841411610949,
+      "learning_rate": 9.185383875603004e-07,
+      "loss": 1.0137,
+      "step": 6855
+    },
+    {
+      "epoch": 0.7210485493052178,
+      "grad_norm": 2.5279779328379943,
+      "learning_rate": 9.178919153809787e-07,
+      "loss": 1.0106,
+      "step": 6856
+    },
+    {
+      "epoch": 0.7211537197470651,
+      "grad_norm": 3.6580469329228826,
+      "learning_rate": 9.172456196190002e-07,
+      "loss": 0.988,
+      "step": 6857
+    },
+    {
+      "epoch": 0.7212588901889124,
+      "grad_norm": 2.1551774200448075,
+      "learning_rate": 9.165995003464295e-07,
+      "loss": 0.998,
+      "step": 6858
+    },
+    {
+      "epoch": 0.7213640606307598,
+      "grad_norm": 2.7261206696045357,
+      "learning_rate": 9.15953557635316e-07,
+      "loss": 0.9893,
+      "step": 6859
+    },
+    {
+      "epoch": 0.7214692310726071,
+      "grad_norm": 2.651636066544782,
+      "learning_rate": 9.153077915576849e-07,
+      "loss": 0.9691,
+      "step": 6860
+    },
+    {
+      "epoch": 0.7215744015144544,
+      "grad_norm": 3.4772896538999896,
+      "learning_rate": 9.146622021855455e-07,
+      "loss": 1.0076,
+      "step": 6861
+    },
+    {
+      "epoch": 0.7216795719563017,
+      "grad_norm": 2.378757397637525,
+      "learning_rate": 9.140167895908867e-07,
+      "loss": 1.0138,
+      "step": 6862
+    },
+    {
+      "epoch": 0.721784742398149,
+      "grad_norm": 1.9309802042614799,
+      "learning_rate": 9.133715538456753e-07,
+      "loss": 0.9607,
+      "step": 6863
+    },
+    {
+      "epoch": 0.7218899128399964,
+      "grad_norm": 2.4260010858714534,
+      "learning_rate": 9.12726495021862e-07,
+      "loss": 0.995,
+      "step": 6864
+    },
+    {
+      "epoch": 0.7219950832818436,
+      "grad_norm": 2.755072867504919,
+      "learning_rate": 9.12081613191374e-07,
+      "loss": 0.9962,
+      "step": 6865
+    },
+    {
+      "epoch": 0.7221002537236909,
+      "grad_norm": 2.318026979345631,
+      "learning_rate": 9.114369084261215e-07,
+      "loss": 0.9444,
+      "step": 6866
+    },
+    {
+      "epoch": 0.7222054241655382,
+      "grad_norm": 2.6082295680717023,
+      "learning_rate": 9.107923807979948e-07,
+      "loss": 1.0035,
+      "step": 6867
+    },
+    {
+      "epoch": 0.7223105946073856,
+      "grad_norm": 2.3969924183402185,
+      "learning_rate": 9.101480303788623e-07,
+      "loss": 0.9672,
+      "step": 6868
+    },
+    {
+      "epoch": 0.7224157650492329,
+      "grad_norm": 2.649977228442466,
+      "learning_rate": 9.095038572405751e-07,
+      "loss": 1.029,
+      "step": 6869
+    },
+    {
+      "epoch": 0.7225209354910802,
+      "grad_norm": 2.0300044348072186,
+      "learning_rate": 9.088598614549629e-07,
+      "loss": 0.9735,
+      "step": 6870
+    },
+    {
+      "epoch": 0.7226261059329275,
+      "grad_norm": 1.9663312006008373,
+      "learning_rate": 9.082160430938375e-07,
+      "loss": 0.9986,
+      "step": 6871
+    },
+    {
+      "epoch": 0.7227312763747749,
+      "grad_norm": 2.09299815663197,
+      "learning_rate": 9.075724022289878e-07,
+      "loss": 1.0071,
+      "step": 6872
+    },
+    {
+      "epoch": 0.7228364468166222,
+      "grad_norm": 2.8696877774023886,
+      "learning_rate": 9.069289389321864e-07,
+      "loss": 0.9965,
+      "step": 6873
+    },
+    {
+      "epoch": 0.7229416172584695,
+      "grad_norm": 1.8588664189025497,
+      "learning_rate": 9.062856532751832e-07,
+      "loss": 0.9622,
+      "step": 6874
+    },
+    {
+      "epoch": 0.7230467877003168,
+      "grad_norm": 2.477197723105134,
+      "learning_rate": 9.056425453297099e-07,
+      "loss": 0.9691,
+      "step": 6875
+    },
+    {
+      "epoch": 0.7231519581421642,
+      "grad_norm": 3.2970809410578554,
+      "learning_rate": 9.04999615167479e-07,
+      "loss": 0.9987,
+      "step": 6876
+    },
+    {
+      "epoch": 0.7232571285840115,
+      "grad_norm": 3.0574404452300183,
+      "learning_rate": 9.043568628601807e-07,
+      "loss": 1.0004,
+      "step": 6877
+    },
+    {
+      "epoch": 0.7233622990258588,
+      "grad_norm": 2.3164982947871096,
+      "learning_rate": 9.037142884794881e-07,
+      "loss": 0.9689,
+      "step": 6878
+    },
+    {
+      "epoch": 0.7234674694677061,
+      "grad_norm": 1.705348459141377,
+      "learning_rate": 9.030718920970513e-07,
+      "loss": 0.99,
+      "step": 6879
+    },
+    {
+      "epoch": 0.7235726399095535,
+      "grad_norm": 1.8734205753292643,
+      "learning_rate": 9.024296737845056e-07,
+      "loss": 1.0241,
+      "step": 6880
+    },
+    {
+      "epoch": 0.7236778103514008,
+      "grad_norm": 2.226219420147319,
+      "learning_rate": 9.017876336134615e-07,
+      "loss": 0.9465,
+      "step": 6881
+    },
+    {
+      "epoch": 0.7237829807932481,
+      "grad_norm": 2.3638808966827827,
+      "learning_rate": 9.011457716555108e-07,
+      "loss": 0.9356,
+      "step": 6882
+    },
+    {
+      "epoch": 0.7238881512350954,
+      "grad_norm": 2.274163453037241,
+      "learning_rate": 9.005040879822269e-07,
+      "loss": 0.9991,
+      "step": 6883
+    },
+    {
+      "epoch": 0.7239933216769427,
+      "grad_norm": 2.3270361322264423,
+      "learning_rate": 8.998625826651624e-07,
+      "loss": 0.9753,
+      "step": 6884
+    },
+    {
+      "epoch": 0.72409849211879,
+      "grad_norm": 2.380779478990076,
+      "learning_rate": 8.992212557758515e-07,
+      "loss": 0.9749,
+      "step": 6885
+    },
+    {
+      "epoch": 0.7242036625606373,
+      "grad_norm": 2.5279858961997146,
+      "learning_rate": 8.985801073858047e-07,
+      "loss": 1.0178,
+      "step": 6886
+    },
+    {
+      "epoch": 0.7243088330024846,
+      "grad_norm": 2.264259809683486,
+      "learning_rate": 8.979391375665169e-07,
+      "loss": 0.9952,
+      "step": 6887
+    },
+    {
+      "epoch": 0.7244140034443319,
+      "grad_norm": 2.3465532374579468,
+      "learning_rate": 8.972983463894599e-07,
+      "loss": 0.9717,
+      "step": 6888
+    },
+    {
+      "epoch": 0.7245191738861793,
+      "grad_norm": 2.682561586616032,
+      "learning_rate": 8.966577339260874e-07,
+      "loss": 0.9911,
+      "step": 6889
+    },
+    {
+      "epoch": 0.7246243443280266,
+      "grad_norm": 3.0023908265682917,
+      "learning_rate": 8.960173002478336e-07,
+      "loss": 0.9918,
+      "step": 6890
+    },
+    {
+      "epoch": 0.7247295147698739,
+      "grad_norm": 2.2766548656250625,
+      "learning_rate": 8.953770454261102e-07,
+      "loss": 0.9655,
+      "step": 6891
+    },
+    {
+      "epoch": 0.7248346852117212,
+      "grad_norm": 2.83034363859963,
+      "learning_rate": 8.947369695323113e-07,
+      "loss": 0.9809,
+      "step": 6892
+    },
+    {
+      "epoch": 0.7249398556535686,
+      "grad_norm": 2.4788072341230527,
+      "learning_rate": 8.940970726378106e-07,
+      "loss": 1.0237,
+      "step": 6893
+    },
+    {
+      "epoch": 0.7250450260954159,
+      "grad_norm": 2.711458177748327,
+      "learning_rate": 8.934573548139621e-07,
+      "loss": 0.9478,
+      "step": 6894
+    },
+    {
+      "epoch": 0.7251501965372632,
+      "grad_norm": 2.9684298621784375,
+      "learning_rate": 8.928178161320977e-07,
+      "loss": 1.0247,
+      "step": 6895
+    },
+    {
+      "epoch": 0.7252553669791105,
+      "grad_norm": 1.6958237240093583,
+      "learning_rate": 8.921784566635328e-07,
+      "loss": 1.0107,
+      "step": 6896
+    },
+    {
+      "epoch": 0.7253605374209579,
+      "grad_norm": 2.608563466979351,
+      "learning_rate": 8.915392764795592e-07,
+      "loss": 0.9734,
+      "step": 6897
+    },
+    {
+      "epoch": 0.7254657078628052,
+      "grad_norm": 2.465172273583475,
+      "learning_rate": 8.909002756514509e-07,
+      "loss": 0.9729,
+      "step": 6898
+    },
+    {
+      "epoch": 0.7255708783046525,
+      "grad_norm": 2.179996715866093,
+      "learning_rate": 8.902614542504631e-07,
+      "loss": 0.9561,
+      "step": 6899
+    },
+    {
+      "epoch": 0.7256760487464998,
+      "grad_norm": 1.9710483091820064,
+      "learning_rate": 8.896228123478268e-07,
+      "loss": 0.9673,
+      "step": 6900
+    },
+    {
+      "epoch": 0.7257812191883471,
+      "grad_norm": 2.3895996851177923,
+      "learning_rate": 8.889843500147577e-07,
+      "loss": 0.9992,
+      "step": 6901
+    },
+    {
+      "epoch": 0.7258863896301945,
+      "grad_norm": 2.4070929276655026,
+      "learning_rate": 8.883460673224478e-07,
+      "loss": 1.0059,
+      "step": 6902
+    },
+    {
+      "epoch": 0.7259915600720418,
+      "grad_norm": 2.52826203988613,
+      "learning_rate": 8.877079643420708e-07,
+      "loss": 0.9768,
+      "step": 6903
+    },
+    {
+      "epoch": 0.7260967305138891,
+      "grad_norm": 2.6173671239261616,
+      "learning_rate": 8.870700411447817e-07,
+      "loss": 0.9739,
+      "step": 6904
+    },
+    {
+      "epoch": 0.7262019009557364,
+      "grad_norm": 2.022268173256108,
+      "learning_rate": 8.864322978017114e-07,
+      "loss": 0.9741,
+      "step": 6905
+    },
+    {
+      "epoch": 0.7263070713975837,
+      "grad_norm": 3.336022838388577,
+      "learning_rate": 8.857947343839749e-07,
+      "loss": 0.9924,
+      "step": 6906
+    },
+    {
+      "epoch": 0.726412241839431,
+      "grad_norm": 3.1049563608044135,
+      "learning_rate": 8.851573509626649e-07,
+      "loss": 0.9864,
+      "step": 6907
+    },
+    {
+      "epoch": 0.7265174122812783,
+      "grad_norm": 2.2360381413354036,
+      "learning_rate": 8.845201476088558e-07,
+      "loss": 0.9817,
+      "step": 6908
+    },
+    {
+      "epoch": 0.7266225827231256,
+      "grad_norm": 2.431440384321841,
+      "learning_rate": 8.838831243935988e-07,
+      "loss": 1.006,
+      "step": 6909
+    },
+    {
+      "epoch": 0.726727753164973,
+      "grad_norm": 2.672270703566811,
+      "learning_rate": 8.832462813879289e-07,
+      "loss": 0.993,
+      "step": 6910
+    },
+    {
+      "epoch": 0.7268329236068203,
+      "grad_norm": 2.3586907446017147,
+      "learning_rate": 8.826096186628568e-07,
+      "loss": 0.9827,
+      "step": 6911
+    },
+    {
+      "epoch": 0.7269380940486676,
+      "grad_norm": 2.3158398975904197,
+      "learning_rate": 8.819731362893769e-07,
+      "loss": 0.9767,
+      "step": 6912
+    },
+    {
+      "epoch": 0.7270432644905149,
+      "grad_norm": 2.7399363636718146,
+      "learning_rate": 8.813368343384621e-07,
+      "loss": 1.014,
+      "step": 6913
+    },
+    {
+      "epoch": 0.7271484349323623,
+      "grad_norm": 2.308433695829236,
+      "learning_rate": 8.807007128810638e-07,
+      "loss": 1.0087,
+      "step": 6914
+    },
+    {
+      "epoch": 0.7272536053742096,
+      "grad_norm": 3.0347699529423333,
+      "learning_rate": 8.800647719881153e-07,
+      "loss": 0.9805,
+      "step": 6915
+    },
+    {
+      "epoch": 0.7273587758160569,
+      "grad_norm": 2.387065454939189,
+      "learning_rate": 8.794290117305296e-07,
+      "loss": 0.9813,
+      "step": 6916
+    },
+    {
+      "epoch": 0.7274639462579042,
+      "grad_norm": 2.8287122692548468,
+      "learning_rate": 8.787934321791972e-07,
+      "loss": 0.9892,
+      "step": 6917
+    },
+    {
+      "epoch": 0.7275691166997516,
+      "grad_norm": 2.377152435206834,
+      "learning_rate": 8.781580334049919e-07,
+      "loss": 0.9678,
+      "step": 6918
+    },
+    {
+      "epoch": 0.7276742871415989,
+      "grad_norm": 3.7839274906805667,
+      "learning_rate": 8.77522815478764e-07,
+      "loss": 1.0138,
+      "step": 6919
+    },
+    {
+      "epoch": 0.7277794575834462,
+      "grad_norm": 2.3333914671728175,
+      "learning_rate": 8.768877784713458e-07,
+      "loss": 1.0377,
+      "step": 6920
+    },
+    {
+      "epoch": 0.7278846280252935,
+      "grad_norm": 2.4232211693938637,
+      "learning_rate": 8.762529224535496e-07,
+      "loss": 0.9752,
+      "step": 6921
+    },
+    {
+      "epoch": 0.7279897984671408,
+      "grad_norm": 2.3166357731448946,
+      "learning_rate": 8.756182474961666e-07,
+      "loss": 0.9823,
+      "step": 6922
+    },
+    {
+      "epoch": 0.7280949689089882,
+      "grad_norm": 2.599714079191184,
+      "learning_rate": 8.749837536699671e-07,
+      "loss": 1.0219,
+      "step": 6923
+    },
+    {
+      "epoch": 0.7282001393508355,
+      "grad_norm": 2.758481393241142,
+      "learning_rate": 8.743494410457032e-07,
+      "loss": 1.0016,
+      "step": 6924
+    },
+    {
+      "epoch": 0.7283053097926828,
+      "grad_norm": 2.801903529605539,
+      "learning_rate": 8.737153096941045e-07,
+      "loss": 0.9807,
+      "step": 6925
+    },
+    {
+      "epoch": 0.72841048023453,
+      "grad_norm": 2.132005701744301,
+      "learning_rate": 8.730813596858823e-07,
+      "loss": 1.0284,
+      "step": 6926
+    },
+    {
+      "epoch": 0.7285156506763774,
+      "grad_norm": 2.645513420655628,
+      "learning_rate": 8.724475910917274e-07,
+      "loss": 1.0083,
+      "step": 6927
+    },
+    {
+      "epoch": 0.7286208211182247,
+      "grad_norm": 2.7201477772204816,
+      "learning_rate": 8.718140039823086e-07,
+      "loss": 0.9533,
+      "step": 6928
+    },
+    {
+      "epoch": 0.728725991560072,
+      "grad_norm": 1.9346726122158562,
+      "learning_rate": 8.711805984282767e-07,
+      "loss": 0.972,
+      "step": 6929
+    },
+    {
+      "epoch": 0.7288311620019193,
+      "grad_norm": 2.5900806982874673,
+      "learning_rate": 8.70547374500261e-07,
+      "loss": 1.0068,
+      "step": 6930
+    },
+    {
+      "epoch": 0.7289363324437667,
+      "grad_norm": 2.269593721723228,
+      "learning_rate": 8.699143322688719e-07,
+      "loss": 0.982,
+      "step": 6931
+    },
+    {
+      "epoch": 0.729041502885614,
+      "grad_norm": 3.3136833165123636,
+      "learning_rate": 8.692814718046979e-07,
+      "loss": 0.9883,
+      "step": 6932
+    },
+    {
+      "epoch": 0.7291466733274613,
+      "grad_norm": 2.360602310285996,
+      "learning_rate": 8.686487931783067e-07,
+      "loss": 0.9928,
+      "step": 6933
+    },
+    {
+      "epoch": 0.7292518437693086,
+      "grad_norm": 2.231311892385931,
+      "learning_rate": 8.680162964602479e-07,
+      "loss": 1.0011,
+      "step": 6934
+    },
+    {
+      "epoch": 0.729357014211156,
+      "grad_norm": 2.2504241332774653,
+      "learning_rate": 8.673839817210497e-07,
+      "loss": 0.9558,
+      "step": 6935
+    },
+    {
+      "epoch": 0.7294621846530033,
+      "grad_norm": 1.758728210506665,
+      "learning_rate": 8.66751849031221e-07,
+      "loss": 0.9341,
+      "step": 6936
+    },
+    {
+      "epoch": 0.7295673550948506,
+      "grad_norm": 2.618947911589737,
+      "learning_rate": 8.661198984612476e-07,
+      "loss": 0.9822,
+      "step": 6937
+    },
+    {
+      "epoch": 0.7296725255366979,
+      "grad_norm": 2.6255176262261877,
+      "learning_rate": 8.654881300815981e-07,
+      "loss": 0.9986,
+      "step": 6938
+    },
+    {
+      "epoch": 0.7297776959785452,
+      "grad_norm": 2.858506049180034,
+      "learning_rate": 8.648565439627205e-07,
+      "loss": 0.9505,
+      "step": 6939
+    },
+    {
+      "epoch": 0.7298828664203926,
+      "grad_norm": 2.0177047337756906,
+      "learning_rate": 8.642251401750395e-07,
+      "loss": 0.9817,
+      "step": 6940
+    },
+    {
+      "epoch": 0.7299880368622399,
+      "grad_norm": 2.414021437249756,
+      "learning_rate": 8.635939187889633e-07,
+      "loss": 0.9816,
+      "step": 6941
+    },
+    {
+      "epoch": 0.7300932073040872,
+      "grad_norm": 2.398138605350569,
+      "learning_rate": 8.629628798748763e-07,
+      "loss": 0.9552,
+      "step": 6942
+    },
+    {
+      "epoch": 0.7301983777459345,
+      "grad_norm": 3.202783456390097,
+      "learning_rate": 8.623320235031452e-07,
+      "loss": 0.98,
+      "step": 6943
+    },
+    {
+      "epoch": 0.7303035481877819,
+      "grad_norm": 1.9008924047025504,
+      "learning_rate": 8.617013497441154e-07,
+      "loss": 0.9407,
+      "step": 6944
+    },
+    {
+      "epoch": 0.7304087186296292,
+      "grad_norm": 2.0495458812583873,
+      "learning_rate": 8.610708586681127e-07,
+      "loss": 1.0026,
+      "step": 6945
+    },
+    {
+      "epoch": 0.7305138890714764,
+      "grad_norm": 2.281006292586853,
+      "learning_rate": 8.604405503454399e-07,
+      "loss": 0.968,
+      "step": 6946
+    },
+    {
+      "epoch": 0.7306190595133237,
+      "grad_norm": 2.8232403754432376,
+      "learning_rate": 8.598104248463823e-07,
+      "loss": 0.9745,
+      "step": 6947
+    },
+    {
+      "epoch": 0.7307242299551711,
+      "grad_norm": 3.0990957513005486,
+      "learning_rate": 8.591804822412048e-07,
+      "loss": 0.9712,
+      "step": 6948
+    },
+    {
+      "epoch": 0.7308294003970184,
+      "grad_norm": 2.948097662188505,
+      "learning_rate": 8.585507226001488e-07,
+      "loss": 0.972,
+      "step": 6949
+    },
+    {
+      "epoch": 0.7309345708388657,
+      "grad_norm": 2.4648039136536304,
+      "learning_rate": 8.579211459934394e-07,
+      "loss": 0.9927,
+      "step": 6950
+    },
+    {
+      "epoch": 0.731039741280713,
+      "grad_norm": 3.8091080476673755,
+      "learning_rate": 8.572917524912777e-07,
+      "loss": 0.9932,
+      "step": 6951
+    },
+    {
+      "epoch": 0.7311449117225604,
+      "grad_norm": 2.2912070211331597,
+      "learning_rate": 8.566625421638464e-07,
+      "loss": 1.0069,
+      "step": 6952
+    },
+    {
+      "epoch": 0.7312500821644077,
+      "grad_norm": 2.769316491698749,
+      "learning_rate": 8.560335150813081e-07,
+      "loss": 0.9821,
+      "step": 6953
+    },
+    {
+      "epoch": 0.731355252606255,
+      "grad_norm": 4.834866721749002,
+      "learning_rate": 8.554046713138034e-07,
+      "loss": 0.96,
+      "step": 6954
+    },
+    {
+      "epoch": 0.7314604230481023,
+      "grad_norm": 2.262988850257133,
+      "learning_rate": 8.54776010931454e-07,
+      "loss": 0.97,
+      "step": 6955
+    },
+    {
+      "epoch": 0.7315655934899497,
+      "grad_norm": 2.4207174619006913,
+      "learning_rate": 8.54147534004359e-07,
+      "loss": 0.9787,
+      "step": 6956
+    },
+    {
+      "epoch": 0.731670763931797,
+      "grad_norm": 2.6566870598983052,
+      "learning_rate": 8.535192406025997e-07,
+      "loss": 0.9971,
+      "step": 6957
+    },
+    {
+      "epoch": 0.7317759343736443,
+      "grad_norm": 2.3775122168529137,
+      "learning_rate": 8.52891130796235e-07,
+      "loss": 0.9409,
+      "step": 6958
+    },
+    {
+      "epoch": 0.7318811048154916,
+      "grad_norm": 2.5975495470680183,
+      "learning_rate": 8.522632046553056e-07,
+      "loss": 1.0159,
+      "step": 6959
+    },
+    {
+      "epoch": 0.731986275257339,
+      "grad_norm": 2.551497070086668,
+      "learning_rate": 8.516354622498279e-07,
+      "loss": 0.9776,
+      "step": 6960
+    },
+    {
+      "epoch": 0.7320914456991863,
+      "grad_norm": 2.030647013013727,
+      "learning_rate": 8.510079036498012e-07,
+      "loss": 0.9495,
+      "step": 6961
+    },
+    {
+      "epoch": 0.7321966161410336,
+      "grad_norm": 2.530716610861503,
+      "learning_rate": 8.503805289252037e-07,
+      "loss": 1.0091,
+      "step": 6962
+    },
+    {
+      "epoch": 0.7323017865828809,
+      "grad_norm": 2.8009461610833113,
+      "learning_rate": 8.497533381459914e-07,
+      "loss": 1.0142,
+      "step": 6963
+    },
+    {
+      "epoch": 0.7324069570247282,
+      "grad_norm": 2.7103271804572553,
+      "learning_rate": 8.491263313821021e-07,
+      "loss": 0.9479,
+      "step": 6964
+    },
+    {
+      "epoch": 0.7325121274665756,
+      "grad_norm": 3.0620007557697995,
+      "learning_rate": 8.484995087034506e-07,
+      "loss": 0.9684,
+      "step": 6965
+    },
+    {
+      "epoch": 0.7326172979084229,
+      "grad_norm": 2.310056025863538,
+      "learning_rate": 8.47872870179933e-07,
+      "loss": 1.0122,
+      "step": 6966
+    },
+    {
+      "epoch": 0.7327224683502701,
+      "grad_norm": 2.2752199543586933,
+      "learning_rate": 8.472464158814256e-07,
+      "loss": 0.9815,
+      "step": 6967
+    },
+    {
+      "epoch": 0.7328276387921174,
+      "grad_norm": 2.499533329716282,
+      "learning_rate": 8.466201458777809e-07,
+      "loss": 0.9966,
+      "step": 6968
+    },
+    {
+      "epoch": 0.7329328092339648,
+      "grad_norm": 3.0605576051148433,
+      "learning_rate": 8.459940602388345e-07,
+      "loss": 0.9442,
+      "step": 6969
+    },
+    {
+      "epoch": 0.7330379796758121,
+      "grad_norm": 3.0303954415055334,
+      "learning_rate": 8.453681590343979e-07,
+      "loss": 1.012,
+      "step": 6970
+    },
+    {
+      "epoch": 0.7331431501176594,
+      "grad_norm": 2.3529258666071855,
+      "learning_rate": 8.447424423342665e-07,
+      "loss": 1.0134,
+      "step": 6971
+    },
+    {
+      "epoch": 0.7332483205595067,
+      "grad_norm": 1.653851753950748,
+      "learning_rate": 8.441169102082106e-07,
+      "loss": 0.9661,
+      "step": 6972
+    },
+    {
+      "epoch": 0.733353491001354,
+      "grad_norm": 2.3418082947379775,
+      "learning_rate": 8.434915627259832e-07,
+      "loss": 1.0021,
+      "step": 6973
+    },
+    {
+      "epoch": 0.7334586614432014,
+      "grad_norm": 3.8928938699639724,
+      "learning_rate": 8.428663999573142e-07,
+      "loss": 0.9314,
+      "step": 6974
+    },
+    {
+      "epoch": 0.7335638318850487,
+      "grad_norm": 3.099834663643491,
+      "learning_rate": 8.422414219719147e-07,
+      "loss": 0.9996,
+      "step": 6975
+    },
+    {
+      "epoch": 0.733669002326896,
+      "grad_norm": 1.8811632859332335,
+      "learning_rate": 8.416166288394751e-07,
+      "loss": 0.9822,
+      "step": 6976
+    },
+    {
+      "epoch": 0.7337741727687433,
+      "grad_norm": 2.5292346499214986,
+      "learning_rate": 8.409920206296635e-07,
+      "loss": 1.0268,
+      "step": 6977
+    },
+    {
+      "epoch": 0.7338793432105907,
+      "grad_norm": 2.3412938045005416,
+      "learning_rate": 8.4036759741213e-07,
+      "loss": 0.9831,
+      "step": 6978
+    },
+    {
+      "epoch": 0.733984513652438,
+      "grad_norm": 2.3394006760695123,
+      "learning_rate": 8.39743359256501e-07,
+      "loss": 0.9884,
+      "step": 6979
+    },
+    {
+      "epoch": 0.7340896840942853,
+      "grad_norm": 1.9244006117602965,
+      "learning_rate": 8.39119306232385e-07,
+      "loss": 0.9443,
+      "step": 6980
+    },
+    {
+      "epoch": 0.7341948545361326,
+      "grad_norm": 2.478475134799539,
+      "learning_rate": 8.384954384093682e-07,
+      "loss": 0.9719,
+      "step": 6981
+    },
+    {
+      "epoch": 0.73430002497798,
+      "grad_norm": 3.0512883023668933,
+      "learning_rate": 8.378717558570182e-07,
+      "loss": 0.969,
+      "step": 6982
+    },
+    {
+      "epoch": 0.7344051954198273,
+      "grad_norm": 3.1758406178171295,
+      "learning_rate": 8.37248258644879e-07,
+      "loss": 0.9856,
+      "step": 6983
+    },
+    {
+      "epoch": 0.7345103658616746,
+      "grad_norm": 2.738349461307976,
+      "learning_rate": 8.366249468424742e-07,
+      "loss": 0.9978,
+      "step": 6984
+    },
+    {
+      "epoch": 0.7346155363035219,
+      "grad_norm": 2.321797129857226,
+      "learning_rate": 8.36001820519311e-07,
+      "loss": 1.0093,
+      "step": 6985
+    },
+    {
+      "epoch": 0.7347207067453693,
+      "grad_norm": 2.1761741288187255,
+      "learning_rate": 8.353788797448703e-07,
+      "loss": 0.9997,
+      "step": 6986
+    },
+    {
+      "epoch": 0.7348258771872165,
+      "grad_norm": 2.706187150898931,
+      "learning_rate": 8.347561245886169e-07,
+      "loss": 0.9516,
+      "step": 6987
+    },
+    {
+      "epoch": 0.7349310476290638,
+      "grad_norm": 2.061774866814489,
+      "learning_rate": 8.341335551199903e-07,
+      "loss": 0.9517,
+      "step": 6988
+    },
+    {
+      "epoch": 0.7350362180709111,
+      "grad_norm": 1.9863690617075271,
+      "learning_rate": 8.335111714084135e-07,
+      "loss": 0.9947,
+      "step": 6989
+    },
+    {
+      "epoch": 0.7351413885127585,
+      "grad_norm": 2.197421006470657,
+      "learning_rate": 8.328889735232876e-07,
+      "loss": 0.9866,
+      "step": 6990
+    },
+    {
+      "epoch": 0.7352465589546058,
+      "grad_norm": 3.8124829291968165,
+      "learning_rate": 8.322669615339909e-07,
+      "loss": 0.984,
+      "step": 6991
+    },
+    {
+      "epoch": 0.7353517293964531,
+      "grad_norm": 3.2431775985547375,
+      "learning_rate": 8.316451355098842e-07,
+      "loss": 1.0148,
+      "step": 6992
+    },
+    {
+      "epoch": 0.7354568998383004,
+      "grad_norm": 2.3326759532763464,
+      "learning_rate": 8.310234955203036e-07,
+      "loss": 0.9669,
+      "step": 6993
+    },
+    {
+      "epoch": 0.7355620702801478,
+      "grad_norm": 2.13604126329195,
+      "learning_rate": 8.304020416345698e-07,
+      "loss": 1.0038,
+      "step": 6994
+    },
+    {
+      "epoch": 0.7356672407219951,
+      "grad_norm": 2.5186724644933234,
+      "learning_rate": 8.297807739219777e-07,
+      "loss": 1.0161,
+      "step": 6995
+    },
+    {
+      "epoch": 0.7357724111638424,
+      "grad_norm": 2.8521649286716224,
+      "learning_rate": 8.291596924518048e-07,
+      "loss": 1.0343,
+      "step": 6996
+    },
+    {
+      "epoch": 0.7358775816056897,
+      "grad_norm": 2.7048039596817137,
+      "learning_rate": 8.285387972933045e-07,
+      "loss": 0.9705,
+      "step": 6997
+    },
+    {
+      "epoch": 0.735982752047537,
+      "grad_norm": 2.332648947659513,
+      "learning_rate": 8.279180885157129e-07,
+      "loss": 0.9918,
+      "step": 6998
+    },
+    {
+      "epoch": 0.7360879224893844,
+      "grad_norm": 2.206028825985984,
+      "learning_rate": 8.272975661882446e-07,
+      "loss": 0.9916,
+      "step": 6999
+    },
+    {
+      "epoch": 0.7361930929312317,
+      "grad_norm": 2.8257632521399847,
+      "learning_rate": 8.266772303800907e-07,
+      "loss": 0.9935,
+      "step": 7000
+    },
+    {
+      "epoch": 0.736298263373079,
+      "grad_norm": 1.9735295694119066,
+      "learning_rate": 8.260570811604252e-07,
+      "loss": 0.9716,
+      "step": 7001
+    },
+    {
+      "epoch": 0.7364034338149263,
+      "grad_norm": 2.766348296710771,
+      "learning_rate": 8.254371185983981e-07,
+      "loss": 0.9867,
+      "step": 7002
+    },
+    {
+      "epoch": 0.7365086042567737,
+      "grad_norm": 2.5669860597872582,
+      "learning_rate": 8.248173427631406e-07,
+      "loss": 0.9607,
+      "step": 7003
+    },
+    {
+      "epoch": 0.736613774698621,
+      "grad_norm": 2.011650558328637,
+      "learning_rate": 8.241977537237639e-07,
+      "loss": 0.9772,
+      "step": 7004
+    },
+    {
+      "epoch": 0.7367189451404683,
+      "grad_norm": 4.378445426045064,
+      "learning_rate": 8.235783515493545e-07,
+      "loss": 1.0191,
+      "step": 7005
+    },
+    {
+      "epoch": 0.7368241155823156,
+      "grad_norm": 3.0930192829825094,
+      "learning_rate": 8.229591363089826e-07,
+      "loss": 0.9812,
+      "step": 7006
+    },
+    {
+      "epoch": 0.7369292860241629,
+      "grad_norm": 2.8858200910410408,
+      "learning_rate": 8.223401080716934e-07,
+      "loss": 1.012,
+      "step": 7007
+    },
+    {
+      "epoch": 0.7370344564660102,
+      "grad_norm": 2.3565977478267763,
+      "learning_rate": 8.217212669065161e-07,
+      "loss": 0.9891,
+      "step": 7008
+    },
+    {
+      "epoch": 0.7371396269078575,
+      "grad_norm": 2.90432653736149,
+      "learning_rate": 8.21102612882454e-07,
+      "loss": 0.9853,
+      "step": 7009
+    },
+    {
+      "epoch": 0.7372447973497048,
+      "grad_norm": 2.636696481329649,
+      "learning_rate": 8.204841460684934e-07,
+      "loss": 0.9837,
+      "step": 7010
+    },
+    {
+      "epoch": 0.7373499677915522,
+      "grad_norm": 2.401750673418409,
+      "learning_rate": 8.198658665335968e-07,
+      "loss": 0.9897,
+      "step": 7011
+    },
+    {
+      "epoch": 0.7374551382333995,
+      "grad_norm": 2.187244354384058,
+      "learning_rate": 8.192477743467078e-07,
+      "loss": 0.9739,
+      "step": 7012
+    },
+    {
+      "epoch": 0.7375603086752468,
+      "grad_norm": 2.504902025352448,
+      "learning_rate": 8.186298695767494e-07,
+      "loss": 1.0082,
+      "step": 7013
+    },
+    {
+      "epoch": 0.7376654791170941,
+      "grad_norm": 2.656973818003429,
+      "learning_rate": 8.18012152292621e-07,
+      "loss": 1.0404,
+      "step": 7014
+    },
+    {
+      "epoch": 0.7377706495589414,
+      "grad_norm": 1.9657665226192536,
+      "learning_rate": 8.173946225632046e-07,
+      "loss": 0.9986,
+      "step": 7015
+    },
+    {
+      "epoch": 0.7378758200007888,
+      "grad_norm": 2.458503304437768,
+      "learning_rate": 8.16777280457357e-07,
+      "loss": 0.9499,
+      "step": 7016
+    },
+    {
+      "epoch": 0.7379809904426361,
+      "grad_norm": 2.8152286475357635,
+      "learning_rate": 8.1616012604392e-07,
+      "loss": 1.0263,
+      "step": 7017
+    },
+    {
+      "epoch": 0.7380861608844834,
+      "grad_norm": 2.7808079810706197,
+      "learning_rate": 8.155431593917096e-07,
+      "loss": 0.9589,
+      "step": 7018
+    },
+    {
+      "epoch": 0.7381913313263307,
+      "grad_norm": 2.2105204331938384,
+      "learning_rate": 8.149263805695215e-07,
+      "loss": 1.0189,
+      "step": 7019
+    },
+    {
+      "epoch": 0.7382965017681781,
+      "grad_norm": 2.152001818336773,
+      "learning_rate": 8.143097896461327e-07,
+      "loss": 0.9507,
+      "step": 7020
+    },
+    {
+      "epoch": 0.7384016722100254,
+      "grad_norm": 2.5837569059762746,
+      "learning_rate": 8.136933866902957e-07,
+      "loss": 0.9991,
+      "step": 7021
+    },
+    {
+      "epoch": 0.7385068426518727,
+      "grad_norm": 2.0815321522556953,
+      "learning_rate": 8.130771717707476e-07,
+      "loss": 0.9999,
+      "step": 7022
+    },
+    {
+      "epoch": 0.73861201309372,
+      "grad_norm": 2.4905114183126327,
+      "learning_rate": 8.124611449561981e-07,
+      "loss": 0.9588,
+      "step": 7023
+    },
+    {
+      "epoch": 0.7387171835355674,
+      "grad_norm": 2.0780844688809545,
+      "learning_rate": 8.118453063153412e-07,
+      "loss": 0.9142,
+      "step": 7024
+    },
+    {
+      "epoch": 0.7388223539774147,
+      "grad_norm": 2.1224882708812336,
+      "learning_rate": 8.112296559168459e-07,
+      "loss": 0.9753,
+      "step": 7025
+    },
+    {
+      "epoch": 0.738927524419262,
+      "grad_norm": 2.8656944207274395,
+      "learning_rate": 8.106141938293627e-07,
+      "loss": 0.979,
+      "step": 7026
+    },
+    {
+      "epoch": 0.7390326948611093,
+      "grad_norm": 2.5552855508838763,
+      "learning_rate": 8.099989201215214e-07,
+      "loss": 0.9649,
+      "step": 7027
+    },
+    {
+      "epoch": 0.7391378653029566,
+      "grad_norm": 1.591509673958794,
+      "learning_rate": 8.093838348619282e-07,
+      "loss": 0.9729,
+      "step": 7028
+    },
+    {
+      "epoch": 0.7392430357448039,
+      "grad_norm": 2.3016915010058825,
+      "learning_rate": 8.087689381191713e-07,
+      "loss": 0.9691,
+      "step": 7029
+    },
+    {
+      "epoch": 0.7393482061866512,
+      "grad_norm": 2.697962007116117,
+      "learning_rate": 8.081542299618139e-07,
+      "loss": 1.0254,
+      "step": 7030
+    },
+    {
+      "epoch": 0.7394533766284985,
+      "grad_norm": 2.585421932407672,
+      "learning_rate": 8.075397104584045e-07,
+      "loss": 0.9704,
+      "step": 7031
+    },
+    {
+      "epoch": 0.7395585470703459,
+      "grad_norm": 2.370762176558943,
+      "learning_rate": 8.069253796774639e-07,
+      "loss": 0.9573,
+      "step": 7032
+    },
+    {
+      "epoch": 0.7396637175121932,
+      "grad_norm": 2.7384589402328117,
+      "learning_rate": 8.063112376874965e-07,
+      "loss": 0.9659,
+      "step": 7033
+    },
+    {
+      "epoch": 0.7397688879540405,
+      "grad_norm": 2.3648476588909197,
+      "learning_rate": 8.056972845569833e-07,
+      "loss": 0.9979,
+      "step": 7034
+    },
+    {
+      "epoch": 0.7398740583958878,
+      "grad_norm": 2.363044938151124,
+      "learning_rate": 8.05083520354383e-07,
+      "loss": 0.9984,
+      "step": 7035
+    },
+    {
+      "epoch": 0.7399792288377351,
+      "grad_norm": 2.7144839085362498,
+      "learning_rate": 8.044699451481383e-07,
+      "loss": 0.98,
+      "step": 7036
+    },
+    {
+      "epoch": 0.7400843992795825,
+      "grad_norm": 1.991910306378637,
+      "learning_rate": 8.038565590066652e-07,
+      "loss": 1.0083,
+      "step": 7037
+    },
+    {
+      "epoch": 0.7401895697214298,
+      "grad_norm": 2.268190827360275,
+      "learning_rate": 8.032433619983628e-07,
+      "loss": 0.9803,
+      "step": 7038
+    },
+    {
+      "epoch": 0.7402947401632771,
+      "grad_norm": 2.9919518122332582,
+      "learning_rate": 8.026303541916056e-07,
+      "loss": 1.0184,
+      "step": 7039
+    },
+    {
+      "epoch": 0.7403999106051244,
+      "grad_norm": 1.9029884195579985,
+      "learning_rate": 8.020175356547497e-07,
+      "loss": 1.0364,
+      "step": 7040
+    },
+    {
+      "epoch": 0.7405050810469718,
+      "grad_norm": 2.8980438411255025,
+      "learning_rate": 8.014049064561294e-07,
+      "loss": 0.9867,
+      "step": 7041
+    },
+    {
+      "epoch": 0.7406102514888191,
+      "grad_norm": 2.9396255932824475,
+      "learning_rate": 8.007924666640565e-07,
+      "loss": 1.0225,
+      "step": 7042
+    },
+    {
+      "epoch": 0.7407154219306664,
+      "grad_norm": 3.0222446896596415,
+      "learning_rate": 8.001802163468244e-07,
+      "loss": 0.9761,
+      "step": 7043
+    },
+    {
+      "epoch": 0.7408205923725137,
+      "grad_norm": 3.017678825360528,
+      "learning_rate": 7.995681555727011e-07,
+      "loss": 1.0636,
+      "step": 7044
+    },
+    {
+      "epoch": 0.7409257628143611,
+      "grad_norm": 2.240379927077817,
+      "learning_rate": 7.989562844099396e-07,
+      "loss": 0.9825,
+      "step": 7045
+    },
+    {
+      "epoch": 0.7410309332562084,
+      "grad_norm": 3.079831156511648,
+      "learning_rate": 7.983446029267658e-07,
+      "loss": 1.0057,
+      "step": 7046
+    },
+    {
+      "epoch": 0.7411361036980557,
+      "grad_norm": 3.2771958422605723,
+      "learning_rate": 7.977331111913883e-07,
+      "loss": 0.9918,
+      "step": 7047
+    },
+    {
+      "epoch": 0.7412412741399029,
+      "grad_norm": 2.545219586879537,
+      "learning_rate": 7.971218092719921e-07,
+      "loss": 0.9816,
+      "step": 7048
+    },
+    {
+      "epoch": 0.7413464445817503,
+      "grad_norm": 1.8833415674316218,
+      "learning_rate": 7.965106972367423e-07,
+      "loss": 0.9534,
+      "step": 7049
+    },
+    {
+      "epoch": 0.7414516150235976,
+      "grad_norm": 2.2734406517544934,
+      "learning_rate": 7.958997751537836e-07,
+      "loss": 0.9753,
+      "step": 7050
+    },
+    {
+      "epoch": 0.7415567854654449,
+      "grad_norm": 2.119993555931499,
+      "learning_rate": 7.952890430912374e-07,
+      "loss": 0.9472,
+      "step": 7051
+    },
+    {
+      "epoch": 0.7416619559072922,
+      "grad_norm": 3.0421738792041166,
+      "learning_rate": 7.946785011172062e-07,
+      "loss": 1.0013,
+      "step": 7052
+    },
+    {
+      "epoch": 0.7417671263491395,
+      "grad_norm": 2.473943886643507,
+      "learning_rate": 7.940681492997678e-07,
+      "loss": 1.0045,
+      "step": 7053
+    },
+    {
+      "epoch": 0.7418722967909869,
+      "grad_norm": 3.1688667102775616,
+      "learning_rate": 7.934579877069848e-07,
+      "loss": 0.9387,
+      "step": 7054
+    },
+    {
+      "epoch": 0.7419774672328342,
+      "grad_norm": 2.5868684633157324,
+      "learning_rate": 7.92848016406893e-07,
+      "loss": 1.0002,
+      "step": 7055
+    },
+    {
+      "epoch": 0.7420826376746815,
+      "grad_norm": 2.2449015856520873,
+      "learning_rate": 7.922382354675079e-07,
+      "loss": 0.9367,
+      "step": 7056
+    },
+    {
+      "epoch": 0.7421878081165288,
+      "grad_norm": 3.009362045627099,
+      "learning_rate": 7.91628644956827e-07,
+      "loss": 0.9526,
+      "step": 7057
+    },
+    {
+      "epoch": 0.7422929785583762,
+      "grad_norm": 2.4357656876283156,
+      "learning_rate": 7.910192449428216e-07,
+      "loss": 1.0009,
+      "step": 7058
+    },
+    {
+      "epoch": 0.7423981490002235,
+      "grad_norm": 1.7009514930395582,
+      "learning_rate": 7.904100354934477e-07,
+      "loss": 0.9459,
+      "step": 7059
+    },
+    {
+      "epoch": 0.7425033194420708,
+      "grad_norm": 1.686132473222285,
+      "learning_rate": 7.898010166766348e-07,
+      "loss": 0.9756,
+      "step": 7060
+    },
+    {
+      "epoch": 0.7426084898839181,
+      "grad_norm": 2.580697246706924,
+      "learning_rate": 7.891921885602946e-07,
+      "loss": 1.014,
+      "step": 7061
+    },
+    {
+      "epoch": 0.7427136603257655,
+      "grad_norm": 1.8915629581945324,
+      "learning_rate": 7.885835512123144e-07,
+      "loss": 0.9774,
+      "step": 7062
+    },
+    {
+      "epoch": 0.7428188307676128,
+      "grad_norm": 2.819291640040572,
+      "learning_rate": 7.879751047005632e-07,
+      "loss": 0.9752,
+      "step": 7063
+    },
+    {
+      "epoch": 0.7429240012094601,
+      "grad_norm": 2.019716408325097,
+      "learning_rate": 7.87366849092888e-07,
+      "loss": 0.9477,
+      "step": 7064
+    },
+    {
+      "epoch": 0.7430291716513074,
+      "grad_norm": 2.179030867839558,
+      "learning_rate": 7.867587844571126e-07,
+      "loss": 0.9877,
+      "step": 7065
+    },
+    {
+      "epoch": 0.7431343420931548,
+      "grad_norm": 2.348644789987132,
+      "learning_rate": 7.861509108610423e-07,
+      "loss": 0.9943,
+      "step": 7066
+    },
+    {
+      "epoch": 0.7432395125350021,
+      "grad_norm": 1.9825526177265085,
+      "learning_rate": 7.855432283724576e-07,
+      "loss": 0.9825,
+      "step": 7067
+    },
+    {
+      "epoch": 0.7433446829768493,
+      "grad_norm": 2.171603481982258,
+      "learning_rate": 7.849357370591229e-07,
+      "loss": 1.0117,
+      "step": 7068
+    },
+    {
+      "epoch": 0.7434498534186966,
+      "grad_norm": 2.648552295188202,
+      "learning_rate": 7.843284369887757e-07,
+      "loss": 0.9668,
+      "step": 7069
+    },
+    {
+      "epoch": 0.743555023860544,
+      "grad_norm": 2.849512019966635,
+      "learning_rate": 7.837213282291365e-07,
+      "loss": 0.9952,
+      "step": 7070
+    },
+    {
+      "epoch": 0.7436601943023913,
+      "grad_norm": 2.368367081297725,
+      "learning_rate": 7.831144108479016e-07,
+      "loss": 0.9449,
+      "step": 7071
+    },
+    {
+      "epoch": 0.7437653647442386,
+      "grad_norm": 2.465131065995525,
+      "learning_rate": 7.825076849127458e-07,
+      "loss": 1.018,
+      "step": 7072
+    },
+    {
+      "epoch": 0.7438705351860859,
+      "grad_norm": 2.883105606748885,
+      "learning_rate": 7.819011504913266e-07,
+      "loss": 0.9762,
+      "step": 7073
+    },
+    {
+      "epoch": 0.7439757056279332,
+      "grad_norm": 3.0353049995873707,
+      "learning_rate": 7.812948076512747e-07,
+      "loss": 1.005,
+      "step": 7074
+    },
+    {
+      "epoch": 0.7440808760697806,
+      "grad_norm": 2.2610828766405118,
+      "learning_rate": 7.806886564602043e-07,
+      "loss": 0.9639,
+      "step": 7075
+    },
+    {
+      "epoch": 0.7441860465116279,
+      "grad_norm": 2.1841830172268466,
+      "learning_rate": 7.800826969857036e-07,
+      "loss": 0.9841,
+      "step": 7076
+    },
+    {
+      "epoch": 0.7442912169534752,
+      "grad_norm": 2.129396215048809,
+      "learning_rate": 7.79476929295343e-07,
+      "loss": 0.9898,
+      "step": 7077
+    },
+    {
+      "epoch": 0.7443963873953225,
+      "grad_norm": 2.3303104402152384,
+      "learning_rate": 7.788713534566714e-07,
+      "loss": 1.0202,
+      "step": 7078
+    },
+    {
+      "epoch": 0.7445015578371699,
+      "grad_norm": 2.278648106170102,
+      "learning_rate": 7.78265969537213e-07,
+      "loss": 0.9724,
+      "step": 7079
+    },
+    {
+      "epoch": 0.7446067282790172,
+      "grad_norm": 2.3135184335473524,
+      "learning_rate": 7.776607776044748e-07,
+      "loss": 0.9536,
+      "step": 7080
+    },
+    {
+      "epoch": 0.7447118987208645,
+      "grad_norm": 2.4775548340535365,
+      "learning_rate": 7.77055777725938e-07,
+      "loss": 0.9794,
+      "step": 7081
+    },
+    {
+      "epoch": 0.7448170691627118,
+      "grad_norm": 2.5496006086788356,
+      "learning_rate": 7.764509699690679e-07,
+      "loss": 1.017,
+      "step": 7082
+    },
+    {
+      "epoch": 0.7449222396045592,
+      "grad_norm": 2.1357477443739787,
+      "learning_rate": 7.758463544013026e-07,
+      "loss": 0.9812,
+      "step": 7083
+    },
+    {
+      "epoch": 0.7450274100464065,
+      "grad_norm": 2.5264510633127824,
+      "learning_rate": 7.752419310900636e-07,
+      "loss": 0.9786,
+      "step": 7084
+    },
+    {
+      "epoch": 0.7451325804882538,
+      "grad_norm": 2.7553954412125945,
+      "learning_rate": 7.746377001027466e-07,
+      "loss": 0.9899,
+      "step": 7085
+    },
+    {
+      "epoch": 0.7452377509301011,
+      "grad_norm": 2.1345306806782225,
+      "learning_rate": 7.740336615067293e-07,
+      "loss": 0.9952,
+      "step": 7086
+    },
+    {
+      "epoch": 0.7453429213719485,
+      "grad_norm": 2.0732183389576258,
+      "learning_rate": 7.734298153693671e-07,
+      "loss": 0.9757,
+      "step": 7087
+    },
+    {
+      "epoch": 0.7454480918137958,
+      "grad_norm": 2.513807387074567,
+      "learning_rate": 7.728261617579922e-07,
+      "loss": 0.9688,
+      "step": 7088
+    },
+    {
+      "epoch": 0.745553262255643,
+      "grad_norm": 1.6046019938022835,
+      "learning_rate": 7.722227007399183e-07,
+      "loss": 0.9902,
+      "step": 7089
+    },
+    {
+      "epoch": 0.7456584326974903,
+      "grad_norm": 2.5953536113957365,
+      "learning_rate": 7.716194323824345e-07,
+      "loss": 0.9866,
+      "step": 7090
+    },
+    {
+      "epoch": 0.7457636031393376,
+      "grad_norm": 2.326926439775202,
+      "learning_rate": 7.710163567528101e-07,
+      "loss": 0.9848,
+      "step": 7091
+    },
+    {
+      "epoch": 0.745868773581185,
+      "grad_norm": 2.8094515613190687,
+      "learning_rate": 7.704134739182942e-07,
+      "loss": 0.9781,
+      "step": 7092
+    },
+    {
+      "epoch": 0.7459739440230323,
+      "grad_norm": 3.123876981837614,
+      "learning_rate": 7.698107839461111e-07,
+      "loss": 0.9893,
+      "step": 7093
+    },
+    {
+      "epoch": 0.7460791144648796,
+      "grad_norm": 3.5591005303250935,
+      "learning_rate": 7.692082869034662e-07,
+      "loss": 0.995,
+      "step": 7094
+    },
+    {
+      "epoch": 0.746184284906727,
+      "grad_norm": 2.346896318567527,
+      "learning_rate": 7.686059828575423e-07,
+      "loss": 0.9834,
+      "step": 7095
+    },
+    {
+      "epoch": 0.7462894553485743,
+      "grad_norm": 2.6908060947387304,
+      "learning_rate": 7.680038718755023e-07,
+      "loss": 0.9847,
+      "step": 7096
+    },
+    {
+      "epoch": 0.7463946257904216,
+      "grad_norm": 2.4966779519732865,
+      "learning_rate": 7.674019540244843e-07,
+      "loss": 0.9635,
+      "step": 7097
+    },
+    {
+      "epoch": 0.7464997962322689,
+      "grad_norm": 2.173016936111415,
+      "learning_rate": 7.668002293716084e-07,
+      "loss": 1.0019,
+      "step": 7098
+    },
+    {
+      "epoch": 0.7466049666741162,
+      "grad_norm": 2.4007122420388285,
+      "learning_rate": 7.661986979839703e-07,
+      "loss": 0.996,
+      "step": 7099
+    },
+    {
+      "epoch": 0.7467101371159636,
+      "grad_norm": 2.305372137742398,
+      "learning_rate": 7.655973599286459e-07,
+      "loss": 1.0022,
+      "step": 7100
+    },
+    {
+      "epoch": 0.7468153075578109,
+      "grad_norm": 2.7387496498746966,
+      "learning_rate": 7.649962152726903e-07,
+      "loss": 1.0304,
+      "step": 7101
+    },
+    {
+      "epoch": 0.7469204779996582,
+      "grad_norm": 3.7124819307244246,
+      "learning_rate": 7.643952640831334e-07,
+      "loss": 0.9823,
+      "step": 7102
+    },
+    {
+      "epoch": 0.7470256484415055,
+      "grad_norm": 2.3773703882173183,
+      "learning_rate": 7.637945064269883e-07,
+      "loss": 1.0051,
+      "step": 7103
+    },
+    {
+      "epoch": 0.7471308188833529,
+      "grad_norm": 2.3915425949886466,
+      "learning_rate": 7.631939423712414e-07,
+      "loss": 0.9314,
+      "step": 7104
+    },
+    {
+      "epoch": 0.7472359893252002,
+      "grad_norm": 2.7826299688394003,
+      "learning_rate": 7.625935719828633e-07,
+      "loss": 0.9793,
+      "step": 7105
+    },
+    {
+      "epoch": 0.7473411597670475,
+      "grad_norm": 2.1770658600493884,
+      "learning_rate": 7.619933953287989e-07,
+      "loss": 0.989,
+      "step": 7106
+    },
+    {
+      "epoch": 0.7474463302088948,
+      "grad_norm": 2.877859070061451,
+      "learning_rate": 7.613934124759712e-07,
+      "loss": 1.0117,
+      "step": 7107
+    },
+    {
+      "epoch": 0.7475515006507422,
+      "grad_norm": 2.1132339489401253,
+      "learning_rate": 7.607936234912841e-07,
+      "loss": 1.0075,
+      "step": 7108
+    },
+    {
+      "epoch": 0.7476566710925894,
+      "grad_norm": 2.04198700411287,
+      "learning_rate": 7.601940284416187e-07,
+      "loss": 0.9941,
+      "step": 7109
+    },
+    {
+      "epoch": 0.7477618415344367,
+      "grad_norm": 3.077969718435496,
+      "learning_rate": 7.595946273938348e-07,
+      "loss": 0.9855,
+      "step": 7110
+    },
+    {
+      "epoch": 0.747867011976284,
+      "grad_norm": 1.851022408153244,
+      "learning_rate": 7.589954204147696e-07,
+      "loss": 0.9606,
+      "step": 7111
+    },
+    {
+      "epoch": 0.7479721824181313,
+      "grad_norm": 2.181580709946769,
+      "learning_rate": 7.583964075712402e-07,
+      "loss": 1.0158,
+      "step": 7112
+    },
+    {
+      "epoch": 0.7480773528599787,
+      "grad_norm": 2.5644893957012824,
+      "learning_rate": 7.577975889300395e-07,
+      "loss": 1.0342,
+      "step": 7113
+    },
+    {
+      "epoch": 0.748182523301826,
+      "grad_norm": 2.117573151952007,
+      "learning_rate": 7.57198964557942e-07,
+      "loss": 0.9742,
+      "step": 7114
+    },
+    {
+      "epoch": 0.7482876937436733,
+      "grad_norm": 2.645246462063244,
+      "learning_rate": 7.566005345216993e-07,
+      "loss": 1.0054,
+      "step": 7115
+    },
+    {
+      "epoch": 0.7483928641855206,
+      "grad_norm": 2.997740803186767,
+      "learning_rate": 7.560022988880392e-07,
+      "loss": 1.0511,
+      "step": 7116
+    },
+    {
+      "epoch": 0.748498034627368,
+      "grad_norm": 2.925719317892139,
+      "learning_rate": 7.554042577236706e-07,
+      "loss": 1.0065,
+      "step": 7117
+    },
+    {
+      "epoch": 0.7486032050692153,
+      "grad_norm": 1.9177386221309367,
+      "learning_rate": 7.548064110952799e-07,
+      "loss": 0.9757,
+      "step": 7118
+    },
+    {
+      "epoch": 0.7487083755110626,
+      "grad_norm": 2.088839253519746,
+      "learning_rate": 7.542087590695321e-07,
+      "loss": 0.9794,
+      "step": 7119
+    },
+    {
+      "epoch": 0.7488135459529099,
+      "grad_norm": 2.5378066740840493,
+      "learning_rate": 7.536113017130686e-07,
+      "loss": 0.9974,
+      "step": 7120
+    },
+    {
+      "epoch": 0.7489187163947573,
+      "grad_norm": 2.3323792119173956,
+      "learning_rate": 7.530140390925125e-07,
+      "loss": 0.9962,
+      "step": 7121
+    },
+    {
+      "epoch": 0.7490238868366046,
+      "grad_norm": 3.319157282262614,
+      "learning_rate": 7.524169712744612e-07,
+      "loss": 1.0356,
+      "step": 7122
+    },
+    {
+      "epoch": 0.7491290572784519,
+      "grad_norm": 3.4503166532322234,
+      "learning_rate": 7.518200983254931e-07,
+      "loss": 0.9966,
+      "step": 7123
+    },
+    {
+      "epoch": 0.7492342277202992,
+      "grad_norm": 2.23731079204491,
+      "learning_rate": 7.512234203121655e-07,
+      "loss": 1.0404,
+      "step": 7124
+    },
+    {
+      "epoch": 0.7493393981621466,
+      "grad_norm": 2.889672775952245,
+      "learning_rate": 7.506269373010106e-07,
+      "loss": 0.9698,
+      "step": 7125
+    },
+    {
+      "epoch": 0.7494445686039939,
+      "grad_norm": 2.3947512239243682,
+      "learning_rate": 7.500306493585424e-07,
+      "loss": 0.9817,
+      "step": 7126
+    },
+    {
+      "epoch": 0.7495497390458412,
+      "grad_norm": 2.1520151953932376,
+      "learning_rate": 7.494345565512504e-07,
+      "loss": 0.973,
+      "step": 7127
+    },
+    {
+      "epoch": 0.7496549094876885,
+      "grad_norm": 3.2505704587799995,
+      "learning_rate": 7.488386589456043e-07,
+      "loss": 1.011,
+      "step": 7128
+    },
+    {
+      "epoch": 0.7497600799295357,
+      "grad_norm": 2.6688857111662054,
+      "learning_rate": 7.482429566080518e-07,
+      "loss": 0.99,
+      "step": 7129
+    },
+    {
+      "epoch": 0.7498652503713831,
+      "grad_norm": 2.2536394568527727,
+      "learning_rate": 7.47647449605017e-07,
+      "loss": 0.9788,
+      "step": 7130
+    },
+    {
+      "epoch": 0.7499704208132304,
+      "grad_norm": 2.3527982867557764,
+      "learning_rate": 7.470521380029044e-07,
+      "loss": 1.0023,
+      "step": 7131
+    },
+    {
+      "epoch": 0.7500755912550777,
+      "grad_norm": 2.2238317421368268,
+      "learning_rate": 7.464570218680958e-07,
+      "loss": 0.9577,
+      "step": 7132
+    },
+    {
+      "epoch": 0.750180761696925,
+      "grad_norm": 2.500971462662424,
+      "learning_rate": 7.45862101266952e-07,
+      "loss": 0.9753,
+      "step": 7133
+    },
+    {
+      "epoch": 0.7502859321387724,
+      "grad_norm": 2.404293358312979,
+      "learning_rate": 7.452673762658096e-07,
+      "loss": 0.9654,
+      "step": 7134
+    },
+    {
+      "epoch": 0.7503911025806197,
+      "grad_norm": 2.0933883750611204,
+      "learning_rate": 7.446728469309872e-07,
+      "loss": 0.9753,
+      "step": 7135
+    },
+    {
+      "epoch": 0.750496273022467,
+      "grad_norm": 2.2681195797049076,
+      "learning_rate": 7.440785133287773e-07,
+      "loss": 0.9794,
+      "step": 7136
+    },
+    {
+      "epoch": 0.7506014434643143,
+      "grad_norm": 2.422935286133933,
+      "learning_rate": 7.434843755254534e-07,
+      "loss": 1.0004,
+      "step": 7137
+    },
+    {
+      "epoch": 0.7507066139061617,
+      "grad_norm": 3.881491057142409,
+      "learning_rate": 7.428904335872675e-07,
+      "loss": 0.9864,
+      "step": 7138
+    },
+    {
+      "epoch": 0.750811784348009,
+      "grad_norm": 2.966377670151308,
+      "learning_rate": 7.422966875804475e-07,
+      "loss": 1.0794,
+      "step": 7139
+    },
+    {
+      "epoch": 0.7509169547898563,
+      "grad_norm": 2.859330892149561,
+      "learning_rate": 7.417031375712009e-07,
+      "loss": 1.0196,
+      "step": 7140
+    },
+    {
+      "epoch": 0.7510221252317036,
+      "grad_norm": 2.8843383108973084,
+      "learning_rate": 7.411097836257141e-07,
+      "loss": 0.9402,
+      "step": 7141
+    },
+    {
+      "epoch": 0.751127295673551,
+      "grad_norm": 2.1645139983935056,
+      "learning_rate": 7.405166258101495e-07,
+      "loss": 0.998,
+      "step": 7142
+    },
+    {
+      "epoch": 0.7512324661153983,
+      "grad_norm": 3.3119860219502337,
+      "learning_rate": 7.399236641906498e-07,
+      "loss": 0.9658,
+      "step": 7143
+    },
+    {
+      "epoch": 0.7513376365572456,
+      "grad_norm": 2.0595918641126327,
+      "learning_rate": 7.393308988333337e-07,
+      "loss": 0.9997,
+      "step": 7144
+    },
+    {
+      "epoch": 0.7514428069990929,
+      "grad_norm": 3.1286563898373543,
+      "learning_rate": 7.387383298042994e-07,
+      "loss": 0.9831,
+      "step": 7145
+    },
+    {
+      "epoch": 0.7515479774409403,
+      "grad_norm": 2.8018989951253417,
+      "learning_rate": 7.381459571696237e-07,
+      "loss": 0.9272,
+      "step": 7146
+    },
+    {
+      "epoch": 0.7516531478827876,
+      "grad_norm": 2.7502028796110176,
+      "learning_rate": 7.375537809953609e-07,
+      "loss": 0.9867,
+      "step": 7147
+    },
+    {
+      "epoch": 0.7517583183246349,
+      "grad_norm": 2.678162761432206,
+      "learning_rate": 7.369618013475419e-07,
+      "loss": 1.0405,
+      "step": 7148
+    },
+    {
+      "epoch": 0.7518634887664822,
+      "grad_norm": 2.3470597444469794,
+      "learning_rate": 7.363700182921784e-07,
+      "loss": 1.0009,
+      "step": 7149
+    },
+    {
+      "epoch": 0.7519686592083294,
+      "grad_norm": 2.5934942008660897,
+      "learning_rate": 7.357784318952579e-07,
+      "loss": 1.0031,
+      "step": 7150
+    },
+    {
+      "epoch": 0.7520738296501768,
+      "grad_norm": 3.16545323400702,
+      "learning_rate": 7.35187042222747e-07,
+      "loss": 1.0078,
+      "step": 7151
+    },
+    {
+      "epoch": 0.7521790000920241,
+      "grad_norm": 2.409118323522591,
+      "learning_rate": 7.345958493405911e-07,
+      "loss": 0.9592,
+      "step": 7152
+    },
+    {
+      "epoch": 0.7522841705338714,
+      "grad_norm": 2.8526257467266967,
+      "learning_rate": 7.340048533147112e-07,
+      "loss": 0.9807,
+      "step": 7153
+    },
+    {
+      "epoch": 0.7523893409757187,
+      "grad_norm": 2.4382304350488035,
+      "learning_rate": 7.334140542110093e-07,
+      "loss": 0.9787,
+      "step": 7154
+    },
+    {
+      "epoch": 0.7524945114175661,
+      "grad_norm": 1.893802244461524,
+      "learning_rate": 7.328234520953634e-07,
+      "loss": 1.0078,
+      "step": 7155
+    },
+    {
+      "epoch": 0.7525996818594134,
+      "grad_norm": 2.3356044875747672,
+      "learning_rate": 7.322330470336314e-07,
+      "loss": 0.98,
+      "step": 7156
+    },
+    {
+      "epoch": 0.7527048523012607,
+      "grad_norm": 2.6357389878275543,
+      "learning_rate": 7.316428390916471e-07,
+      "loss": 0.976,
+      "step": 7157
+    },
+    {
+      "epoch": 0.752810022743108,
+      "grad_norm": 2.0816090851410043,
+      "learning_rate": 7.310528283352225e-07,
+      "loss": 0.9998,
+      "step": 7158
+    },
+    {
+      "epoch": 0.7529151931849554,
+      "grad_norm": 2.432089883393319,
+      "learning_rate": 7.304630148301495e-07,
+      "loss": 1.0173,
+      "step": 7159
+    },
+    {
+      "epoch": 0.7530203636268027,
+      "grad_norm": 3.2576438811204267,
+      "learning_rate": 7.298733986421963e-07,
+      "loss": 0.9316,
+      "step": 7160
+    },
+    {
+      "epoch": 0.75312553406865,
+      "grad_norm": 2.5192616299402535,
+      "learning_rate": 7.292839798371107e-07,
+      "loss": 0.9752,
+      "step": 7161
+    },
+    {
+      "epoch": 0.7532307045104973,
+      "grad_norm": 2.0662627309135466,
+      "learning_rate": 7.286947584806162e-07,
+      "loss": 1.0003,
+      "step": 7162
+    },
+    {
+      "epoch": 0.7533358749523447,
+      "grad_norm": 2.3318286805896697,
+      "learning_rate": 7.28105734638416e-07,
+      "loss": 0.9491,
+      "step": 7163
+    },
+    {
+      "epoch": 0.753441045394192,
+      "grad_norm": 1.7589980888158547,
+      "learning_rate": 7.275169083761915e-07,
+      "loss": 1.0146,
+      "step": 7164
+    },
+    {
+      "epoch": 0.7535462158360393,
+      "grad_norm": 2.3273013030064167,
+      "learning_rate": 7.269282797596003e-07,
+      "loss": 0.9642,
+      "step": 7165
+    },
+    {
+      "epoch": 0.7536513862778866,
+      "grad_norm": 2.912414660527191,
+      "learning_rate": 7.263398488542805e-07,
+      "loss": 1.0203,
+      "step": 7166
+    },
+    {
+      "epoch": 0.753756556719734,
+      "grad_norm": 2.82234921943586,
+      "learning_rate": 7.257516157258448e-07,
+      "loss": 0.9563,
+      "step": 7167
+    },
+    {
+      "epoch": 0.7538617271615813,
+      "grad_norm": 2.5257753971825565,
+      "learning_rate": 7.251635804398871e-07,
+      "loss": 0.9375,
+      "step": 7168
+    },
+    {
+      "epoch": 0.7539668976034286,
+      "grad_norm": 2.2636444621395486,
+      "learning_rate": 7.245757430619774e-07,
+      "loss": 0.9048,
+      "step": 7169
+    },
+    {
+      "epoch": 0.7540720680452758,
+      "grad_norm": 1.7968514845268797,
+      "learning_rate": 7.239881036576652e-07,
+      "loss": 1.0071,
+      "step": 7170
+    },
+    {
+      "epoch": 0.7541772384871231,
+      "grad_norm": 2.5407406591367714,
+      "learning_rate": 7.23400662292475e-07,
+      "loss": 0.9604,
+      "step": 7171
+    },
+    {
+      "epoch": 0.7542824089289705,
+      "grad_norm": 2.8060825489630976,
+      "learning_rate": 7.228134190319131e-07,
+      "loss": 1.016,
+      "step": 7172
+    },
+    {
+      "epoch": 0.7543875793708178,
+      "grad_norm": 2.7592153012337985,
+      "learning_rate": 7.222263739414595e-07,
+      "loss": 0.9704,
+      "step": 7173
+    },
+    {
+      "epoch": 0.7544927498126651,
+      "grad_norm": 1.982882504910971,
+      "learning_rate": 7.216395270865759e-07,
+      "loss": 0.9705,
+      "step": 7174
+    },
+    {
+      "epoch": 0.7545979202545124,
+      "grad_norm": 2.3971758450955254,
+      "learning_rate": 7.210528785327001e-07,
+      "loss": 0.9987,
+      "step": 7175
+    },
+    {
+      "epoch": 0.7547030906963598,
+      "grad_norm": 2.0404726366158235,
+      "learning_rate": 7.204664283452472e-07,
+      "loss": 0.9884,
+      "step": 7176
+    },
+    {
+      "epoch": 0.7548082611382071,
+      "grad_norm": 2.169832032755285,
+      "learning_rate": 7.198801765896115e-07,
+      "loss": 0.9897,
+      "step": 7177
+    },
+    {
+      "epoch": 0.7549134315800544,
+      "grad_norm": 3.109999050865016,
+      "learning_rate": 7.192941233311651e-07,
+      "loss": 1.0375,
+      "step": 7178
+    },
+    {
+      "epoch": 0.7550186020219017,
+      "grad_norm": 2.4498043657390474,
+      "learning_rate": 7.187082686352564e-07,
+      "loss": 0.9986,
+      "step": 7179
+    },
+    {
+      "epoch": 0.7551237724637491,
+      "grad_norm": 2.045678192773114,
+      "learning_rate": 7.181226125672142e-07,
+      "loss": 0.9928,
+      "step": 7180
+    },
+    {
+      "epoch": 0.7552289429055964,
+      "grad_norm": 2.506873399635673,
+      "learning_rate": 7.175371551923418e-07,
+      "loss": 0.9825,
+      "step": 7181
+    },
+    {
+      "epoch": 0.7553341133474437,
+      "grad_norm": 2.679384773475905,
+      "learning_rate": 7.169518965759231e-07,
+      "loss": 1.0099,
+      "step": 7182
+    },
+    {
+      "epoch": 0.755439283789291,
+      "grad_norm": 2.6893589546797467,
+      "learning_rate": 7.163668367832194e-07,
+      "loss": 0.9387,
+      "step": 7183
+    },
+    {
+      "epoch": 0.7555444542311384,
+      "grad_norm": 2.9496315141406906,
+      "learning_rate": 7.157819758794699e-07,
+      "loss": 0.9685,
+      "step": 7184
+    },
+    {
+      "epoch": 0.7556496246729857,
+      "grad_norm": 2.8192968825983615,
+      "learning_rate": 7.151973139298895e-07,
+      "loss": 0.9811,
+      "step": 7185
+    },
+    {
+      "epoch": 0.755754795114833,
+      "grad_norm": 2.823604895336445,
+      "learning_rate": 7.146128509996736e-07,
+      "loss": 1.0068,
+      "step": 7186
+    },
+    {
+      "epoch": 0.7558599655566803,
+      "grad_norm": 2.6359789895073855,
+      "learning_rate": 7.140285871539948e-07,
+      "loss": 0.9866,
+      "step": 7187
+    },
+    {
+      "epoch": 0.7559651359985277,
+      "grad_norm": 2.1868328887950144,
+      "learning_rate": 7.134445224580017e-07,
+      "loss": 0.9832,
+      "step": 7188
+    },
+    {
+      "epoch": 0.756070306440375,
+      "grad_norm": 2.2297442915008046,
+      "learning_rate": 7.128606569768237e-07,
+      "loss": 0.9415,
+      "step": 7189
+    },
+    {
+      "epoch": 0.7561754768822222,
+      "grad_norm": 2.157623716865805,
+      "learning_rate": 7.122769907755644e-07,
+      "loss": 1.0098,
+      "step": 7190
+    },
+    {
+      "epoch": 0.7562806473240695,
+      "grad_norm": 2.393849877655623,
+      "learning_rate": 7.116935239193085e-07,
+      "loss": 0.9886,
+      "step": 7191
+    },
+    {
+      "epoch": 0.7563858177659168,
+      "grad_norm": 2.5597033186090457,
+      "learning_rate": 7.111102564731173e-07,
+      "loss": 0.9664,
+      "step": 7192
+    },
+    {
+      "epoch": 0.7564909882077642,
+      "grad_norm": 3.176157747900133,
+      "learning_rate": 7.105271885020281e-07,
+      "loss": 1.0258,
+      "step": 7193
+    },
+    {
+      "epoch": 0.7565961586496115,
+      "grad_norm": 2.9605515801527513,
+      "learning_rate": 7.099443200710595e-07,
+      "loss": 0.9994,
+      "step": 7194
+    },
+    {
+      "epoch": 0.7567013290914588,
+      "grad_norm": 2.0428204214114,
+      "learning_rate": 7.093616512452042e-07,
+      "loss": 0.9946,
+      "step": 7195
+    },
+    {
+      "epoch": 0.7568064995333061,
+      "grad_norm": 2.5922763139539353,
+      "learning_rate": 7.087791820894349e-07,
+      "loss": 0.9879,
+      "step": 7196
+    },
+    {
+      "epoch": 0.7569116699751535,
+      "grad_norm": 2.2625784519484355,
+      "learning_rate": 7.081969126687014e-07,
+      "loss": 0.9551,
+      "step": 7197
+    },
+    {
+      "epoch": 0.7570168404170008,
+      "grad_norm": 2.7498123014483973,
+      "learning_rate": 7.076148430479321e-07,
+      "loss": 0.9919,
+      "step": 7198
+    },
+    {
+      "epoch": 0.7571220108588481,
+      "grad_norm": 2.2860922030180015,
+      "learning_rate": 7.070329732920308e-07,
+      "loss": 0.9593,
+      "step": 7199
+    },
+    {
+      "epoch": 0.7572271813006954,
+      "grad_norm": 2.1018686765589782,
+      "learning_rate": 7.064513034658812e-07,
+      "loss": 0.9519,
+      "step": 7200
+    },
+    {
+      "epoch": 0.7573323517425428,
+      "grad_norm": 2.0969982207355655,
+      "learning_rate": 7.05869833634345e-07,
+      "loss": 0.964,
+      "step": 7201
+    },
+    {
+      "epoch": 0.7574375221843901,
+      "grad_norm": 2.2200649084471045,
+      "learning_rate": 7.052885638622586e-07,
+      "loss": 0.949,
+      "step": 7202
+    },
+    {
+      "epoch": 0.7575426926262374,
+      "grad_norm": 2.5288138123773556,
+      "learning_rate": 7.047074942144399e-07,
+      "loss": 1.0022,
+      "step": 7203
+    },
+    {
+      "epoch": 0.7576478630680847,
+      "grad_norm": 2.914980411654761,
+      "learning_rate": 7.041266247556814e-07,
+      "loss": 0.9717,
+      "step": 7204
+    },
+    {
+      "epoch": 0.7577530335099321,
+      "grad_norm": 1.7833952010134386,
+      "learning_rate": 7.035459555507549e-07,
+      "loss": 0.9748,
+      "step": 7205
+    },
+    {
+      "epoch": 0.7578582039517794,
+      "grad_norm": 2.49777934486207,
+      "learning_rate": 7.029654866644098e-07,
+      "loss": 0.9616,
+      "step": 7206
+    },
+    {
+      "epoch": 0.7579633743936267,
+      "grad_norm": 2.748297431592962,
+      "learning_rate": 7.023852181613735e-07,
+      "loss": 0.99,
+      "step": 7207
+    },
+    {
+      "epoch": 0.758068544835474,
+      "grad_norm": 2.6942600142402537,
+      "learning_rate": 7.018051501063497e-07,
+      "loss": 1.026,
+      "step": 7208
+    },
+    {
+      "epoch": 0.7581737152773214,
+      "grad_norm": 3.1894240209205713,
+      "learning_rate": 7.01225282564019e-07,
+      "loss": 1.0089,
+      "step": 7209
+    },
+    {
+      "epoch": 0.7582788857191687,
+      "grad_norm": 1.8391440626240132,
+      "learning_rate": 7.006456155990444e-07,
+      "loss": 0.978,
+      "step": 7210
+    },
+    {
+      "epoch": 0.7583840561610159,
+      "grad_norm": 2.4573368871580317,
+      "learning_rate": 7.000661492760605e-07,
+      "loss": 0.9747,
+      "step": 7211
+    },
+    {
+      "epoch": 0.7584892266028632,
+      "grad_norm": 2.0727123712660758,
+      "learning_rate": 6.994868836596841e-07,
+      "loss": 1.0151,
+      "step": 7212
+    },
+    {
+      "epoch": 0.7585943970447105,
+      "grad_norm": 1.8852340383663915,
+      "learning_rate": 6.989078188145065e-07,
+      "loss": 0.9984,
+      "step": 7213
+    },
+    {
+      "epoch": 0.7586995674865579,
+      "grad_norm": 1.9223909811199098,
+      "learning_rate": 6.983289548050984e-07,
+      "loss": 1.0037,
+      "step": 7214
+    },
+    {
+      "epoch": 0.7588047379284052,
+      "grad_norm": 2.520783501726942,
+      "learning_rate": 6.977502916960083e-07,
+      "loss": 0.9742,
+      "step": 7215
+    },
+    {
+      "epoch": 0.7589099083702525,
+      "grad_norm": 2.838885326430099,
+      "learning_rate": 6.971718295517604e-07,
+      "loss": 1.0104,
+      "step": 7216
+    },
+    {
+      "epoch": 0.7590150788120998,
+      "grad_norm": 2.9156056508616257,
+      "learning_rate": 6.965935684368591e-07,
+      "loss": 1.0002,
+      "step": 7217
+    },
+    {
+      "epoch": 0.7591202492539472,
+      "grad_norm": 2.3167176341509514,
+      "learning_rate": 6.960155084157835e-07,
+      "loss": 0.9942,
+      "step": 7218
+    },
+    {
+      "epoch": 0.7592254196957945,
+      "grad_norm": 2.154619542983804,
+      "learning_rate": 6.954376495529927e-07,
+      "loss": 0.9635,
+      "step": 7219
+    },
+    {
+      "epoch": 0.7593305901376418,
+      "grad_norm": 2.708144396111361,
+      "learning_rate": 6.94859991912922e-07,
+      "loss": 0.9492,
+      "step": 7220
+    },
+    {
+      "epoch": 0.7594357605794891,
+      "grad_norm": 2.869899495540543,
+      "learning_rate": 6.942825355599861e-07,
+      "loss": 0.9659,
+      "step": 7221
+    },
+    {
+      "epoch": 0.7595409310213365,
+      "grad_norm": 2.7681751925235063,
+      "learning_rate": 6.93705280558574e-07,
+      "loss": 0.9965,
+      "step": 7222
+    },
+    {
+      "epoch": 0.7596461014631838,
+      "grad_norm": 2.454449329821474,
+      "learning_rate": 6.931282269730547e-07,
+      "loss": 0.9723,
+      "step": 7223
+    },
+    {
+      "epoch": 0.7597512719050311,
+      "grad_norm": 2.2182925652533085,
+      "learning_rate": 6.925513748677754e-07,
+      "loss": 0.9866,
+      "step": 7224
+    },
+    {
+      "epoch": 0.7598564423468784,
+      "grad_norm": 2.3307398988105255,
+      "learning_rate": 6.919747243070576e-07,
+      "loss": 0.9802,
+      "step": 7225
+    },
+    {
+      "epoch": 0.7599616127887258,
+      "grad_norm": 2.9729325369975705,
+      "learning_rate": 6.91398275355204e-07,
+      "loss": 0.9946,
+      "step": 7226
+    },
+    {
+      "epoch": 0.7600667832305731,
+      "grad_norm": 2.7311855204096185,
+      "learning_rate": 6.908220280764918e-07,
+      "loss": 0.9488,
+      "step": 7227
+    },
+    {
+      "epoch": 0.7601719536724204,
+      "grad_norm": 2.68897996658099,
+      "learning_rate": 6.902459825351776e-07,
+      "loss": 0.9728,
+      "step": 7228
+    },
+    {
+      "epoch": 0.7602771241142677,
+      "grad_norm": 2.3536132874792934,
+      "learning_rate": 6.896701387954957e-07,
+      "loss": 0.9418,
+      "step": 7229
+    },
+    {
+      "epoch": 0.7603822945561151,
+      "grad_norm": 2.604781814575496,
+      "learning_rate": 6.890944969216556e-07,
+      "loss": 0.9645,
+      "step": 7230
+    },
+    {
+      "epoch": 0.7604874649979623,
+      "grad_norm": 2.703365663091066,
+      "learning_rate": 6.885190569778477e-07,
+      "loss": 1.0083,
+      "step": 7231
+    },
+    {
+      "epoch": 0.7605926354398096,
+      "grad_norm": 2.995617326069344,
+      "learning_rate": 6.879438190282354e-07,
+      "loss": 0.9152,
+      "step": 7232
+    },
+    {
+      "epoch": 0.7606978058816569,
+      "grad_norm": 2.4869310918420573,
+      "learning_rate": 6.873687831369655e-07,
+      "loss": 1.0181,
+      "step": 7233
+    },
+    {
+      "epoch": 0.7608029763235042,
+      "grad_norm": 3.3720251417185545,
+      "learning_rate": 6.867939493681563e-07,
+      "loss": 0.9915,
+      "step": 7234
+    },
+    {
+      "epoch": 0.7609081467653516,
+      "grad_norm": 2.751154189351907,
+      "learning_rate": 6.862193177859081e-07,
+      "loss": 1.0399,
+      "step": 7235
+    },
+    {
+      "epoch": 0.7610133172071989,
+      "grad_norm": 2.9074664793566565,
+      "learning_rate": 6.856448884542951e-07,
+      "loss": 0.9934,
+      "step": 7236
+    },
+    {
+      "epoch": 0.7611184876490462,
+      "grad_norm": 1.9259825570766822,
+      "learning_rate": 6.850706614373715e-07,
+      "loss": 1.0186,
+      "step": 7237
+    },
+    {
+      "epoch": 0.7612236580908935,
+      "grad_norm": 2.2646689754710816,
+      "learning_rate": 6.844966367991688e-07,
+      "loss": 0.9753,
+      "step": 7238
+    },
+    {
+      "epoch": 0.7613288285327409,
+      "grad_norm": 1.9865318190535692,
+      "learning_rate": 6.839228146036936e-07,
+      "loss": 0.9906,
+      "step": 7239
+    },
+    {
+      "epoch": 0.7614339989745882,
+      "grad_norm": 2.101934374400583,
+      "learning_rate": 6.833491949149329e-07,
+      "loss": 0.9754,
+      "step": 7240
+    },
+    {
+      "epoch": 0.7615391694164355,
+      "grad_norm": 2.9805412783118554,
+      "learning_rate": 6.82775777796848e-07,
+      "loss": 0.9668,
+      "step": 7241
+    },
+    {
+      "epoch": 0.7616443398582828,
+      "grad_norm": 2.1434579040583226,
+      "learning_rate": 6.82202563313382e-07,
+      "loss": 1.009,
+      "step": 7242
+    },
+    {
+      "epoch": 0.7617495103001302,
+      "grad_norm": 2.5155188921882763,
+      "learning_rate": 6.816295515284513e-07,
+      "loss": 0.9747,
+      "step": 7243
+    },
+    {
+      "epoch": 0.7618546807419775,
+      "grad_norm": 2.351626885137821,
+      "learning_rate": 6.810567425059506e-07,
+      "loss": 0.9613,
+      "step": 7244
+    },
+    {
+      "epoch": 0.7619598511838248,
+      "grad_norm": 2.499487860831982,
+      "learning_rate": 6.804841363097536e-07,
+      "loss": 0.998,
+      "step": 7245
+    },
+    {
+      "epoch": 0.7620650216256721,
+      "grad_norm": 3.5082989671199454,
+      "learning_rate": 6.799117330037086e-07,
+      "loss": 1.0091,
+      "step": 7246
+    },
+    {
+      "epoch": 0.7621701920675195,
+      "grad_norm": 2.5497617555706222,
+      "learning_rate": 6.793395326516458e-07,
+      "loss": 1.0019,
+      "step": 7247
+    },
+    {
+      "epoch": 0.7622753625093668,
+      "grad_norm": 2.9850232501231315,
+      "learning_rate": 6.787675353173675e-07,
+      "loss": 0.9852,
+      "step": 7248
+    },
+    {
+      "epoch": 0.7623805329512141,
+      "grad_norm": 2.369868434119237,
+      "learning_rate": 6.781957410646581e-07,
+      "loss": 0.9482,
+      "step": 7249
+    },
+    {
+      "epoch": 0.7624857033930614,
+      "grad_norm": 2.2851427983587325,
+      "learning_rate": 6.776241499572747e-07,
+      "loss": 0.997,
+      "step": 7250
+    },
+    {
+      "epoch": 0.7625908738349086,
+      "grad_norm": 2.566339810963556,
+      "learning_rate": 6.770527620589554e-07,
+      "loss": 0.9698,
+      "step": 7251
+    },
+    {
+      "epoch": 0.762696044276756,
+      "grad_norm": 2.5226398298922685,
+      "learning_rate": 6.764815774334149e-07,
+      "loss": 1.0029,
+      "step": 7252
+    },
+    {
+      "epoch": 0.7628012147186033,
+      "grad_norm": 2.0900487583661564,
+      "learning_rate": 6.759105961443435e-07,
+      "loss": 0.9886,
+      "step": 7253
+    },
+    {
+      "epoch": 0.7629063851604506,
+      "grad_norm": 2.8035948742537626,
+      "learning_rate": 6.753398182554116e-07,
+      "loss": 0.954,
+      "step": 7254
+    },
+    {
+      "epoch": 0.7630115556022979,
+      "grad_norm": 2.2764400788498347,
+      "learning_rate": 6.747692438302624e-07,
+      "loss": 1.0132,
+      "step": 7255
+    },
+    {
+      "epoch": 0.7631167260441453,
+      "grad_norm": 3.4744512830475545,
+      "learning_rate": 6.741988729325235e-07,
+      "loss": 1.0112,
+      "step": 7256
+    },
+    {
+      "epoch": 0.7632218964859926,
+      "grad_norm": 3.0967876034792012,
+      "learning_rate": 6.736287056257925e-07,
+      "loss": 0.9576,
+      "step": 7257
+    },
+    {
+      "epoch": 0.7633270669278399,
+      "grad_norm": 2.346841697317024,
+      "learning_rate": 6.730587419736492e-07,
+      "loss": 0.9708,
+      "step": 7258
+    },
+    {
+      "epoch": 0.7634322373696872,
+      "grad_norm": 2.201886938588659,
+      "learning_rate": 6.724889820396488e-07,
+      "loss": 0.993,
+      "step": 7259
+    },
+    {
+      "epoch": 0.7635374078115346,
+      "grad_norm": 2.8480079520191044,
+      "learning_rate": 6.719194258873216e-07,
+      "loss": 0.9501,
+      "step": 7260
+    },
+    {
+      "epoch": 0.7636425782533819,
+      "grad_norm": 2.560782072345373,
+      "learning_rate": 6.713500735801811e-07,
+      "loss": 0.8903,
+      "step": 7261
+    },
+    {
+      "epoch": 0.7637477486952292,
+      "grad_norm": 2.159045835231064,
+      "learning_rate": 6.707809251817121e-07,
+      "loss": 0.9807,
+      "step": 7262
+    },
+    {
+      "epoch": 0.7638529191370765,
+      "grad_norm": 1.9003145255033675,
+      "learning_rate": 6.702119807553806e-07,
+      "loss": 0.9616,
+      "step": 7263
+    },
+    {
+      "epoch": 0.7639580895789239,
+      "grad_norm": 2.7151279520190976,
+      "learning_rate": 6.696432403646267e-07,
+      "loss": 0.9819,
+      "step": 7264
+    },
+    {
+      "epoch": 0.7640632600207712,
+      "grad_norm": 2.657753683243195,
+      "learning_rate": 6.690747040728702e-07,
+      "loss": 0.9873,
+      "step": 7265
+    },
+    {
+      "epoch": 0.7641684304626185,
+      "grad_norm": 1.8721650852607126,
+      "learning_rate": 6.685063719435081e-07,
+      "loss": 0.9638,
+      "step": 7266
+    },
+    {
+      "epoch": 0.7642736009044658,
+      "grad_norm": 2.385971269188057,
+      "learning_rate": 6.679382440399121e-07,
+      "loss": 0.9701,
+      "step": 7267
+    },
+    {
+      "epoch": 0.7643787713463132,
+      "grad_norm": 1.6808229698056358,
+      "learning_rate": 6.673703204254348e-07,
+      "loss": 0.9355,
+      "step": 7268
+    },
+    {
+      "epoch": 0.7644839417881605,
+      "grad_norm": 2.741069590738767,
+      "learning_rate": 6.668026011634019e-07,
+      "loss": 0.9926,
+      "step": 7269
+    },
+    {
+      "epoch": 0.7645891122300078,
+      "grad_norm": 2.117414364110104,
+      "learning_rate": 6.662350863171207e-07,
+      "loss": 0.9928,
+      "step": 7270
+    },
+    {
+      "epoch": 0.7646942826718551,
+      "grad_norm": 3.3484101267491737,
+      "learning_rate": 6.656677759498722e-07,
+      "loss": 0.9656,
+      "step": 7271
+    },
+    {
+      "epoch": 0.7647994531137023,
+      "grad_norm": 1.9706890021853318,
+      "learning_rate": 6.651006701249168e-07,
+      "loss": 1.0416,
+      "step": 7272
+    },
+    {
+      "epoch": 0.7649046235555497,
+      "grad_norm": 3.0413485520399663,
+      "learning_rate": 6.645337689054901e-07,
+      "loss": 0.9993,
+      "step": 7273
+    },
+    {
+      "epoch": 0.765009793997397,
+      "grad_norm": 2.0490262655425626,
+      "learning_rate": 6.639670723548066e-07,
+      "loss": 1.0095,
+      "step": 7274
+    },
+    {
+      "epoch": 0.7651149644392443,
+      "grad_norm": 2.8013420666362183,
+      "learning_rate": 6.63400580536058e-07,
+      "loss": 0.9992,
+      "step": 7275
+    },
+    {
+      "epoch": 0.7652201348810916,
+      "grad_norm": 2.047516643390938,
+      "learning_rate": 6.628342935124113e-07,
+      "loss": 1.0431,
+      "step": 7276
+    },
+    {
+      "epoch": 0.765325305322939,
+      "grad_norm": 1.8990163407770566,
+      "learning_rate": 6.622682113470131e-07,
+      "loss": 0.9764,
+      "step": 7277
+    },
+    {
+      "epoch": 0.7654304757647863,
+      "grad_norm": 2.450404087097505,
+      "learning_rate": 6.617023341029835e-07,
+      "loss": 0.9911,
+      "step": 7278
+    },
+    {
+      "epoch": 0.7655356462066336,
+      "grad_norm": 2.3743125938259735,
+      "learning_rate": 6.611366618434259e-07,
+      "loss": 0.9901,
+      "step": 7279
+    },
+    {
+      "epoch": 0.7656408166484809,
+      "grad_norm": 2.6744820836183436,
+      "learning_rate": 6.605711946314153e-07,
+      "loss": 0.9975,
+      "step": 7280
+    },
+    {
+      "epoch": 0.7657459870903283,
+      "grad_norm": 2.2318926191610093,
+      "learning_rate": 6.600059325300049e-07,
+      "loss": 0.9983,
+      "step": 7281
+    },
+    {
+      "epoch": 0.7658511575321756,
+      "grad_norm": 2.6271251940737734,
+      "learning_rate": 6.594408756022272e-07,
+      "loss": 0.9516,
+      "step": 7282
+    },
+    {
+      "epoch": 0.7659563279740229,
+      "grad_norm": 2.686237334138938,
+      "learning_rate": 6.588760239110887e-07,
+      "loss": 1.0418,
+      "step": 7283
+    },
+    {
+      "epoch": 0.7660614984158702,
+      "grad_norm": 2.7534033695073665,
+      "learning_rate": 6.583113775195771e-07,
+      "loss": 1.0173,
+      "step": 7284
+    },
+    {
+      "epoch": 0.7661666688577176,
+      "grad_norm": 2.2831574538057744,
+      "learning_rate": 6.577469364906527e-07,
+      "loss": 0.9971,
+      "step": 7285
+    },
+    {
+      "epoch": 0.7662718392995649,
+      "grad_norm": 2.7528580601431263,
+      "learning_rate": 6.571827008872572e-07,
+      "loss": 0.9858,
+      "step": 7286
+    },
+    {
+      "epoch": 0.7663770097414122,
+      "grad_norm": 2.305246939209575,
+      "learning_rate": 6.566186707723049e-07,
+      "loss": 0.9331,
+      "step": 7287
+    },
+    {
+      "epoch": 0.7664821801832595,
+      "grad_norm": 2.1919285177867747,
+      "learning_rate": 6.560548462086911e-07,
+      "loss": 0.9929,
+      "step": 7288
+    },
+    {
+      "epoch": 0.7665873506251069,
+      "grad_norm": 2.617341622706207,
+      "learning_rate": 6.554912272592867e-07,
+      "loss": 0.9793,
+      "step": 7289
+    },
+    {
+      "epoch": 0.7666925210669542,
+      "grad_norm": 3.2075296541728506,
+      "learning_rate": 6.549278139869383e-07,
+      "loss": 0.963,
+      "step": 7290
+    },
+    {
+      "epoch": 0.7667976915088015,
+      "grad_norm": 2.4692337859546227,
+      "learning_rate": 6.543646064544725e-07,
+      "loss": 0.9813,
+      "step": 7291
+    },
+    {
+      "epoch": 0.7669028619506487,
+      "grad_norm": 2.161864541098391,
+      "learning_rate": 6.53801604724689e-07,
+      "loss": 0.9615,
+      "step": 7292
+    },
+    {
+      "epoch": 0.767008032392496,
+      "grad_norm": 2.2773528909260428,
+      "learning_rate": 6.5323880886037e-07,
+      "loss": 0.9771,
+      "step": 7293
+    },
+    {
+      "epoch": 0.7671132028343434,
+      "grad_norm": 2.9832947081070955,
+      "learning_rate": 6.526762189242692e-07,
+      "loss": 1.0007,
+      "step": 7294
+    },
+    {
+      "epoch": 0.7672183732761907,
+      "grad_norm": 2.715135683153078,
+      "learning_rate": 6.521138349791209e-07,
+      "loss": 0.9939,
+      "step": 7295
+    },
+    {
+      "epoch": 0.767323543718038,
+      "grad_norm": 2.470749049910447,
+      "learning_rate": 6.515516570876351e-07,
+      "loss": 0.9985,
+      "step": 7296
+    },
+    {
+      "epoch": 0.7674287141598853,
+      "grad_norm": 2.2143297282737655,
+      "learning_rate": 6.509896853124972e-07,
+      "loss": 0.9607,
+      "step": 7297
+    },
+    {
+      "epoch": 0.7675338846017327,
+      "grad_norm": 2.887958915857869,
+      "learning_rate": 6.504279197163746e-07,
+      "loss": 0.9331,
+      "step": 7298
+    },
+    {
+      "epoch": 0.76763905504358,
+      "grad_norm": 3.027736095235375,
+      "learning_rate": 6.498663603619062e-07,
+      "loss": 1.005,
+      "step": 7299
+    },
+    {
+      "epoch": 0.7677442254854273,
+      "grad_norm": 2.6040148356675403,
+      "learning_rate": 6.493050073117115e-07,
+      "loss": 1.0007,
+      "step": 7300
+    },
+    {
+      "epoch": 0.7678493959272746,
+      "grad_norm": 2.4651062503117127,
+      "learning_rate": 6.487438606283847e-07,
+      "loss": 0.9931,
+      "step": 7301
+    },
+    {
+      "epoch": 0.767954566369122,
+      "grad_norm": 2.1995832711178593,
+      "learning_rate": 6.481829203744985e-07,
+      "loss": 0.9499,
+      "step": 7302
+    },
+    {
+      "epoch": 0.7680597368109693,
+      "grad_norm": 2.7534359404754913,
+      "learning_rate": 6.476221866126029e-07,
+      "loss": 0.9385,
+      "step": 7303
+    },
+    {
+      "epoch": 0.7681649072528166,
+      "grad_norm": 2.0684676604556387,
+      "learning_rate": 6.470616594052223e-07,
+      "loss": 1.0037,
+      "step": 7304
+    },
+    {
+      "epoch": 0.7682700776946639,
+      "grad_norm": 2.474552308274405,
+      "learning_rate": 6.465013388148616e-07,
+      "loss": 0.9923,
+      "step": 7305
+    },
+    {
+      "epoch": 0.7683752481365113,
+      "grad_norm": 1.7973959275988587,
+      "learning_rate": 6.459412249039987e-07,
+      "loss": 0.9635,
+      "step": 7306
+    },
+    {
+      "epoch": 0.7684804185783586,
+      "grad_norm": 2.12766206020307,
+      "learning_rate": 6.453813177350934e-07,
+      "loss": 0.9721,
+      "step": 7307
+    },
+    {
+      "epoch": 0.7685855890202059,
+      "grad_norm": 2.371873219751724,
+      "learning_rate": 6.448216173705777e-07,
+      "loss": 0.9784,
+      "step": 7308
+    },
+    {
+      "epoch": 0.7686907594620532,
+      "grad_norm": 2.0694569481455987,
+      "learning_rate": 6.442621238728639e-07,
+      "loss": 0.9855,
+      "step": 7309
+    },
+    {
+      "epoch": 0.7687959299039006,
+      "grad_norm": 3.206943381862623,
+      "learning_rate": 6.437028373043386e-07,
+      "loss": 1.0048,
+      "step": 7310
+    },
+    {
+      "epoch": 0.7689011003457479,
+      "grad_norm": 3.365367181152907,
+      "learning_rate": 6.431437577273669e-07,
+      "loss": 0.9555,
+      "step": 7311
+    },
+    {
+      "epoch": 0.7690062707875951,
+      "grad_norm": 2.8202002768816925,
+      "learning_rate": 6.425848852042918e-07,
+      "loss": 1.0238,
+      "step": 7312
+    },
+    {
+      "epoch": 0.7691114412294424,
+      "grad_norm": 2.47701840199492,
+      "learning_rate": 6.420262197974297e-07,
+      "loss": 0.991,
+      "step": 7313
+    },
+    {
+      "epoch": 0.7692166116712897,
+      "grad_norm": 2.3044685577607513,
+      "learning_rate": 6.414677615690784e-07,
+      "loss": 0.9681,
+      "step": 7314
+    },
+    {
+      "epoch": 0.7693217821131371,
+      "grad_norm": 3.4475573948156413,
+      "learning_rate": 6.409095105815086e-07,
+      "loss": 0.9781,
+      "step": 7315
+    },
+    {
+      "epoch": 0.7694269525549844,
+      "grad_norm": 1.8908622557377632,
+      "learning_rate": 6.403514668969702e-07,
+      "loss": 0.976,
+      "step": 7316
+    },
+    {
+      "epoch": 0.7695321229968317,
+      "grad_norm": 2.1550644787890785,
+      "learning_rate": 6.397936305776903e-07,
+      "loss": 0.9988,
+      "step": 7317
+    },
+    {
+      "epoch": 0.769637293438679,
+      "grad_norm": 3.0769836395456025,
+      "learning_rate": 6.392360016858703e-07,
+      "loss": 0.9766,
+      "step": 7318
+    },
+    {
+      "epoch": 0.7697424638805264,
+      "grad_norm": 2.5329631219295625,
+      "learning_rate": 6.38678580283692e-07,
+      "loss": 1.0409,
+      "step": 7319
+    },
+    {
+      "epoch": 0.7698476343223737,
+      "grad_norm": 3.1252877879621637,
+      "learning_rate": 6.381213664333096e-07,
+      "loss": 0.9786,
+      "step": 7320
+    },
+    {
+      "epoch": 0.769952804764221,
+      "grad_norm": 2.477880542696044,
+      "learning_rate": 6.375643601968598e-07,
+      "loss": 0.985,
+      "step": 7321
+    },
+    {
+      "epoch": 0.7700579752060683,
+      "grad_norm": 2.8094356003054846,
+      "learning_rate": 6.370075616364512e-07,
+      "loss": 1.0039,
+      "step": 7322
+    },
+    {
+      "epoch": 0.7701631456479157,
+      "grad_norm": 2.8079925876475476,
+      "learning_rate": 6.364509708141725e-07,
+      "loss": 1.0002,
+      "step": 7323
+    },
+    {
+      "epoch": 0.770268316089763,
+      "grad_norm": 2.3156139981672035,
+      "learning_rate": 6.358945877920861e-07,
+      "loss": 0.9701,
+      "step": 7324
+    },
+    {
+      "epoch": 0.7703734865316103,
+      "grad_norm": 2.3629511594914994,
+      "learning_rate": 6.353384126322343e-07,
+      "loss": 1.0154,
+      "step": 7325
+    },
+    {
+      "epoch": 0.7704786569734576,
+      "grad_norm": 2.898189649288246,
+      "learning_rate": 6.347824453966354e-07,
+      "loss": 0.9892,
+      "step": 7326
+    },
+    {
+      "epoch": 0.770583827415305,
+      "grad_norm": 2.122895447793783,
+      "learning_rate": 6.342266861472823e-07,
+      "loss": 0.9902,
+      "step": 7327
+    },
+    {
+      "epoch": 0.7706889978571523,
+      "grad_norm": 2.7711699801067486,
+      "learning_rate": 6.336711349461486e-07,
+      "loss": 0.9745,
+      "step": 7328
+    },
+    {
+      "epoch": 0.7707941682989996,
+      "grad_norm": 2.5654011499268354,
+      "learning_rate": 6.331157918551801e-07,
+      "loss": 1.033,
+      "step": 7329
+    },
+    {
+      "epoch": 0.7708993387408469,
+      "grad_norm": 1.7248857187294304,
+      "learning_rate": 6.325606569363044e-07,
+      "loss": 0.9897,
+      "step": 7330
+    },
+    {
+      "epoch": 0.7710045091826943,
+      "grad_norm": 2.540837264737888,
+      "learning_rate": 6.320057302514223e-07,
+      "loss": 0.9716,
+      "step": 7331
+    },
+    {
+      "epoch": 0.7711096796245416,
+      "grad_norm": 2.1994727634563467,
+      "learning_rate": 6.314510118624121e-07,
+      "loss": 1.0289,
+      "step": 7332
+    },
+    {
+      "epoch": 0.7712148500663888,
+      "grad_norm": 3.2728653021918515,
+      "learning_rate": 6.308965018311289e-07,
+      "loss": 0.9727,
+      "step": 7333
+    },
+    {
+      "epoch": 0.7713200205082361,
+      "grad_norm": 2.323117451826077,
+      "learning_rate": 6.303422002194057e-07,
+      "loss": 1.0584,
+      "step": 7334
+    },
+    {
+      "epoch": 0.7714251909500834,
+      "grad_norm": 2.991321743338717,
+      "learning_rate": 6.297881070890519e-07,
+      "loss": 0.9433,
+      "step": 7335
+    },
+    {
+      "epoch": 0.7715303613919308,
+      "grad_norm": 1.953839287290166,
+      "learning_rate": 6.292342225018517e-07,
+      "loss": 1.0006,
+      "step": 7336
+    },
+    {
+      "epoch": 0.7716355318337781,
+      "grad_norm": 2.0866972087128044,
+      "learning_rate": 6.286805465195691e-07,
+      "loss": 0.964,
+      "step": 7337
+    },
+    {
+      "epoch": 0.7717407022756254,
+      "grad_norm": 2.647572811142748,
+      "learning_rate": 6.281270792039418e-07,
+      "loss": 1.0017,
+      "step": 7338
+    },
+    {
+      "epoch": 0.7718458727174727,
+      "grad_norm": 2.3592122810731553,
+      "learning_rate": 6.275738206166862e-07,
+      "loss": 0.9775,
+      "step": 7339
+    },
+    {
+      "epoch": 0.7719510431593201,
+      "grad_norm": 1.9493966901790556,
+      "learning_rate": 6.270207708194959e-07,
+      "loss": 1.0169,
+      "step": 7340
+    },
+    {
+      "epoch": 0.7720562136011674,
+      "grad_norm": 2.4294207839002384,
+      "learning_rate": 6.264679298740389e-07,
+      "loss": 0.9461,
+      "step": 7341
+    },
+    {
+      "epoch": 0.7721613840430147,
+      "grad_norm": 2.1711248655387347,
+      "learning_rate": 6.259152978419625e-07,
+      "loss": 0.9646,
+      "step": 7342
+    },
+    {
+      "epoch": 0.772266554484862,
+      "grad_norm": 2.526321431390834,
+      "learning_rate": 6.253628747848872e-07,
+      "loss": 0.9981,
+      "step": 7343
+    },
+    {
+      "epoch": 0.7723717249267094,
+      "grad_norm": 2.4845350540111806,
+      "learning_rate": 6.248106607644155e-07,
+      "loss": 0.9788,
+      "step": 7344
+    },
+    {
+      "epoch": 0.7724768953685567,
+      "grad_norm": 2.3629873655101923,
+      "learning_rate": 6.242586558421216e-07,
+      "loss": 0.9314,
+      "step": 7345
+    },
+    {
+      "epoch": 0.772582065810404,
+      "grad_norm": 3.0934894712871963,
+      "learning_rate": 6.237068600795593e-07,
+      "loss": 0.9113,
+      "step": 7346
+    },
+    {
+      "epoch": 0.7726872362522513,
+      "grad_norm": 2.022489507710269,
+      "learning_rate": 6.23155273538257e-07,
+      "loss": 1.0347,
+      "step": 7347
+    },
+    {
+      "epoch": 0.7727924066940987,
+      "grad_norm": 1.9206629909055462,
+      "learning_rate": 6.226038962797218e-07,
+      "loss": 0.9649,
+      "step": 7348
+    },
+    {
+      "epoch": 0.772897577135946,
+      "grad_norm": 2.4285553532799167,
+      "learning_rate": 6.220527283654368e-07,
+      "loss": 0.9841,
+      "step": 7349
+    },
+    {
+      "epoch": 0.7730027475777933,
+      "grad_norm": 2.2432647116261912,
+      "learning_rate": 6.215017698568604e-07,
+      "loss": 0.9608,
+      "step": 7350
+    },
+    {
+      "epoch": 0.7731079180196406,
+      "grad_norm": 2.2176480767044664,
+      "learning_rate": 6.209510208154299e-07,
+      "loss": 0.9774,
+      "step": 7351
+    },
+    {
+      "epoch": 0.773213088461488,
+      "grad_norm": 2.250001090894873,
+      "learning_rate": 6.204004813025569e-07,
+      "loss": 1.0104,
+      "step": 7352
+    },
+    {
+      "epoch": 0.7733182589033352,
+      "grad_norm": 1.5454534677927962,
+      "learning_rate": 6.198501513796315e-07,
+      "loss": 0.9279,
+      "step": 7353
+    },
+    {
+      "epoch": 0.7734234293451825,
+      "grad_norm": 2.839261743707263,
+      "learning_rate": 6.193000311080203e-07,
+      "loss": 0.967,
+      "step": 7354
+    },
+    {
+      "epoch": 0.7735285997870298,
+      "grad_norm": 2.6135585390560903,
+      "learning_rate": 6.187501205490648e-07,
+      "loss": 0.9666,
+      "step": 7355
+    },
+    {
+      "epoch": 0.7736337702288771,
+      "grad_norm": 2.441585421023636,
+      "learning_rate": 6.18200419764085e-07,
+      "loss": 0.9487,
+      "step": 7356
+    },
+    {
+      "epoch": 0.7737389406707245,
+      "grad_norm": 2.154135654667443,
+      "learning_rate": 6.176509288143768e-07,
+      "loss": 0.9512,
+      "step": 7357
+    },
+    {
+      "epoch": 0.7738441111125718,
+      "grad_norm": 2.0470908979748317,
+      "learning_rate": 6.17101647761213e-07,
+      "loss": 1.009,
+      "step": 7358
+    },
+    {
+      "epoch": 0.7739492815544191,
+      "grad_norm": 2.30034891783103,
+      "learning_rate": 6.16552576665842e-07,
+      "loss": 0.9784,
+      "step": 7359
+    },
+    {
+      "epoch": 0.7740544519962664,
+      "grad_norm": 2.796522111094815,
+      "learning_rate": 6.160037155894902e-07,
+      "loss": 0.9799,
+      "step": 7360
+    },
+    {
+      "epoch": 0.7741596224381138,
+      "grad_norm": 2.635375796645631,
+      "learning_rate": 6.154550645933591e-07,
+      "loss": 0.9638,
+      "step": 7361
+    },
+    {
+      "epoch": 0.7742647928799611,
+      "grad_norm": 2.5343887605480924,
+      "learning_rate": 6.149066237386278e-07,
+      "loss": 1.0155,
+      "step": 7362
+    },
+    {
+      "epoch": 0.7743699633218084,
+      "grad_norm": 2.3060590549923874,
+      "learning_rate": 6.143583930864527e-07,
+      "loss": 0.9765,
+      "step": 7363
+    },
+    {
+      "epoch": 0.7744751337636557,
+      "grad_norm": 2.196968553332674,
+      "learning_rate": 6.138103726979641e-07,
+      "loss": 0.9767,
+      "step": 7364
+    },
+    {
+      "epoch": 0.7745803042055031,
+      "grad_norm": 1.8053420148777697,
+      "learning_rate": 6.132625626342723e-07,
+      "loss": 1.0004,
+      "step": 7365
+    },
+    {
+      "epoch": 0.7746854746473504,
+      "grad_norm": 1.974966755215082,
+      "learning_rate": 6.127149629564605e-07,
+      "loss": 1.0041,
+      "step": 7366
+    },
+    {
+      "epoch": 0.7747906450891977,
+      "grad_norm": 2.4074095725370777,
+      "learning_rate": 6.121675737255913e-07,
+      "loss": 0.9991,
+      "step": 7367
+    },
+    {
+      "epoch": 0.774895815531045,
+      "grad_norm": 2.771485579995371,
+      "learning_rate": 6.116203950027036e-07,
+      "loss": 0.9738,
+      "step": 7368
+    },
+    {
+      "epoch": 0.7750009859728924,
+      "grad_norm": 2.5495785766223857,
+      "learning_rate": 6.110734268488106e-07,
+      "loss": 1.0109,
+      "step": 7369
+    },
+    {
+      "epoch": 0.7751061564147397,
+      "grad_norm": 2.665289353085339,
+      "learning_rate": 6.105266693249043e-07,
+      "loss": 0.9922,
+      "step": 7370
+    },
+    {
+      "epoch": 0.775211326856587,
+      "grad_norm": 3.2766595461201926,
+      "learning_rate": 6.099801224919522e-07,
+      "loss": 1.0238,
+      "step": 7371
+    },
+    {
+      "epoch": 0.7753164972984343,
+      "grad_norm": 2.810829062797779,
+      "learning_rate": 6.094337864108993e-07,
+      "loss": 0.9732,
+      "step": 7372
+    },
+    {
+      "epoch": 0.7754216677402815,
+      "grad_norm": 2.2147571794146486,
+      "learning_rate": 6.08887661142665e-07,
+      "loss": 0.9908,
+      "step": 7373
+    },
+    {
+      "epoch": 0.7755268381821289,
+      "grad_norm": 2.1162563329579482,
+      "learning_rate": 6.083417467481479e-07,
+      "loss": 0.9771,
+      "step": 7374
+    },
+    {
+      "epoch": 0.7756320086239762,
+      "grad_norm": 3.0165229052813265,
+      "learning_rate": 6.077960432882202e-07,
+      "loss": 0.9558,
+      "step": 7375
+    },
+    {
+      "epoch": 0.7757371790658235,
+      "grad_norm": 2.780788058142792,
+      "learning_rate": 6.072505508237328e-07,
+      "loss": 0.9393,
+      "step": 7376
+    },
+    {
+      "epoch": 0.7758423495076708,
+      "grad_norm": 2.746344323741756,
+      "learning_rate": 6.067052694155132e-07,
+      "loss": 0.9889,
+      "step": 7377
+    },
+    {
+      "epoch": 0.7759475199495182,
+      "grad_norm": 2.8304408738719853,
+      "learning_rate": 6.061601991243629e-07,
+      "loss": 0.976,
+      "step": 7378
+    },
+    {
+      "epoch": 0.7760526903913655,
+      "grad_norm": 2.7760091448267428,
+      "learning_rate": 6.056153400110623e-07,
+      "loss": 0.9813,
+      "step": 7379
+    },
+    {
+      "epoch": 0.7761578608332128,
+      "grad_norm": 2.5521937340727234,
+      "learning_rate": 6.050706921363672e-07,
+      "loss": 0.9967,
+      "step": 7380
+    },
+    {
+      "epoch": 0.7762630312750601,
+      "grad_norm": 2.179859514373383,
+      "learning_rate": 6.045262555610113e-07,
+      "loss": 0.9829,
+      "step": 7381
+    },
+    {
+      "epoch": 0.7763682017169075,
+      "grad_norm": 2.7632582697479187,
+      "learning_rate": 6.039820303457022e-07,
+      "loss": 1.0111,
+      "step": 7382
+    },
+    {
+      "epoch": 0.7764733721587548,
+      "grad_norm": 2.822582215401159,
+      "learning_rate": 6.034380165511247e-07,
+      "loss": 0.996,
+      "step": 7383
+    },
+    {
+      "epoch": 0.7765785426006021,
+      "grad_norm": 2.8025819156968517,
+      "learning_rate": 6.028942142379416e-07,
+      "loss": 0.9985,
+      "step": 7384
+    },
+    {
+      "epoch": 0.7766837130424494,
+      "grad_norm": 2.311116198164196,
+      "learning_rate": 6.023506234667908e-07,
+      "loss": 0.9911,
+      "step": 7385
+    },
+    {
+      "epoch": 0.7767888834842968,
+      "grad_norm": 2.7484844931535504,
+      "learning_rate": 6.018072442982875e-07,
+      "loss": 1.0146,
+      "step": 7386
+    },
+    {
+      "epoch": 0.7768940539261441,
+      "grad_norm": 2.3912901717058896,
+      "learning_rate": 6.012640767930217e-07,
+      "loss": 0.9586,
+      "step": 7387
+    },
+    {
+      "epoch": 0.7769992243679914,
+      "grad_norm": 2.7928691824138614,
+      "learning_rate": 6.00721121011561e-07,
+      "loss": 0.9506,
+      "step": 7388
+    },
+    {
+      "epoch": 0.7771043948098387,
+      "grad_norm": 2.0128303006970305,
+      "learning_rate": 6.001783770144504e-07,
+      "loss": 1.0171,
+      "step": 7389
+    },
+    {
+      "epoch": 0.777209565251686,
+      "grad_norm": 2.240404598690575,
+      "learning_rate": 5.99635844862208e-07,
+      "loss": 0.9801,
+      "step": 7390
+    },
+    {
+      "epoch": 0.7773147356935334,
+      "grad_norm": 2.001832281945611,
+      "learning_rate": 5.990935246153326e-07,
+      "loss": 0.9841,
+      "step": 7391
+    },
+    {
+      "epoch": 0.7774199061353807,
+      "grad_norm": 2.653987884349642,
+      "learning_rate": 5.985514163342948e-07,
+      "loss": 0.9892,
+      "step": 7392
+    },
+    {
+      "epoch": 0.777525076577228,
+      "grad_norm": 2.5723150950012834,
+      "learning_rate": 5.980095200795452e-07,
+      "loss": 0.9935,
+      "step": 7393
+    },
+    {
+      "epoch": 0.7776302470190752,
+      "grad_norm": 2.0286439737482875,
+      "learning_rate": 5.974678359115094e-07,
+      "loss": 0.9336,
+      "step": 7394
+    },
+    {
+      "epoch": 0.7777354174609226,
+      "grad_norm": 2.672262654260771,
+      "learning_rate": 5.9692636389059e-07,
+      "loss": 0.9862,
+      "step": 7395
+    },
+    {
+      "epoch": 0.7778405879027699,
+      "grad_norm": 2.9017249311317173,
+      "learning_rate": 5.963851040771639e-07,
+      "loss": 0.9836,
+      "step": 7396
+    },
+    {
+      "epoch": 0.7779457583446172,
+      "grad_norm": 3.0804120297128468,
+      "learning_rate": 5.958440565315871e-07,
+      "loss": 1.0163,
+      "step": 7397
+    },
+    {
+      "epoch": 0.7780509287864645,
+      "grad_norm": 2.6802032536538176,
+      "learning_rate": 5.953032213141894e-07,
+      "loss": 0.9765,
+      "step": 7398
+    },
+    {
+      "epoch": 0.7781560992283119,
+      "grad_norm": 2.051038288544476,
+      "learning_rate": 5.947625984852787e-07,
+      "loss": 0.9461,
+      "step": 7399
+    },
+    {
+      "epoch": 0.7782612696701592,
+      "grad_norm": 2.1836813883193362,
+      "learning_rate": 5.942221881051394e-07,
+      "loss": 0.9772,
+      "step": 7400
+    },
+    {
+      "epoch": 0.7783664401120065,
+      "grad_norm": 2.6840419851820956,
+      "learning_rate": 5.936819902340299e-07,
+      "loss": 1.0152,
+      "step": 7401
+    },
+    {
+      "epoch": 0.7784716105538538,
+      "grad_norm": 2.431122357081175,
+      "learning_rate": 5.931420049321873e-07,
+      "loss": 0.9785,
+      "step": 7402
+    },
+    {
+      "epoch": 0.7785767809957012,
+      "grad_norm": 2.439690169026877,
+      "learning_rate": 5.926022322598249e-07,
+      "loss": 0.9506,
+      "step": 7403
+    },
+    {
+      "epoch": 0.7786819514375485,
+      "grad_norm": 3.004232477345118,
+      "learning_rate": 5.920626722771303e-07,
+      "loss": 0.9238,
+      "step": 7404
+    },
+    {
+      "epoch": 0.7787871218793958,
+      "grad_norm": 2.7134097989656207,
+      "learning_rate": 5.915233250442695e-07,
+      "loss": 0.9726,
+      "step": 7405
+    },
+    {
+      "epoch": 0.7788922923212431,
+      "grad_norm": 3.4146505334812702,
+      "learning_rate": 5.909841906213828e-07,
+      "loss": 1.0107,
+      "step": 7406
+    },
+    {
+      "epoch": 0.7789974627630905,
+      "grad_norm": 1.9095501012682825,
+      "learning_rate": 5.904452690685888e-07,
+      "loss": 1.0127,
+      "step": 7407
+    },
+    {
+      "epoch": 0.7791026332049378,
+      "grad_norm": 2.5125920673209254,
+      "learning_rate": 5.899065604459814e-07,
+      "loss": 0.9826,
+      "step": 7408
+    },
+    {
+      "epoch": 0.7792078036467851,
+      "grad_norm": 2.028561314466869,
+      "learning_rate": 5.893680648136311e-07,
+      "loss": 0.9689,
+      "step": 7409
+    },
+    {
+      "epoch": 0.7793129740886324,
+      "grad_norm": 2.20112168268948,
+      "learning_rate": 5.888297822315831e-07,
+      "loss": 0.9506,
+      "step": 7410
+    },
+    {
+      "epoch": 0.7794181445304798,
+      "grad_norm": 1.8066767860491197,
+      "learning_rate": 5.882917127598608e-07,
+      "loss": 0.9993,
+      "step": 7411
+    },
+    {
+      "epoch": 0.7795233149723271,
+      "grad_norm": 2.975379972499003,
+      "learning_rate": 5.877538564584642e-07,
+      "loss": 0.9631,
+      "step": 7412
+    },
+    {
+      "epoch": 0.7796284854141744,
+      "grad_norm": 2.533926267530022,
+      "learning_rate": 5.872162133873666e-07,
+      "loss": 0.9686,
+      "step": 7413
+    },
+    {
+      "epoch": 0.7797336558560216,
+      "grad_norm": 2.5144691976081175,
+      "learning_rate": 5.866787836065211e-07,
+      "loss": 1.0199,
+      "step": 7414
+    },
+    {
+      "epoch": 0.7798388262978689,
+      "grad_norm": 1.9435003292682747,
+      "learning_rate": 5.861415671758536e-07,
+      "loss": 0.977,
+      "step": 7415
+    },
+    {
+      "epoch": 0.7799439967397163,
+      "grad_norm": 2.516714132211873,
+      "learning_rate": 5.856045641552685e-07,
+      "loss": 1.0082,
+      "step": 7416
+    },
+    {
+      "epoch": 0.7800491671815636,
+      "grad_norm": 2.283896919126085,
+      "learning_rate": 5.850677746046471e-07,
+      "loss": 0.9556,
+      "step": 7417
+    },
+    {
+      "epoch": 0.7801543376234109,
+      "grad_norm": 2.396317543406843,
+      "learning_rate": 5.845311985838437e-07,
+      "loss": 1.0118,
+      "step": 7418
+    },
+    {
+      "epoch": 0.7802595080652582,
+      "grad_norm": 1.8411662172460828,
+      "learning_rate": 5.83994836152692e-07,
+      "loss": 1.0051,
+      "step": 7419
+    },
+    {
+      "epoch": 0.7803646785071056,
+      "grad_norm": 2.1120279732981038,
+      "learning_rate": 5.83458687371e-07,
+      "loss": 0.9711,
+      "step": 7420
+    },
+    {
+      "epoch": 0.7804698489489529,
+      "grad_norm": 2.5159122284997815,
+      "learning_rate": 5.82922752298552e-07,
+      "loss": 1.0153,
+      "step": 7421
+    },
+    {
+      "epoch": 0.7805750193908002,
+      "grad_norm": 3.0017293005335217,
+      "learning_rate": 5.823870309951096e-07,
+      "loss": 0.9498,
+      "step": 7422
+    },
+    {
+      "epoch": 0.7806801898326475,
+      "grad_norm": 2.7311817018256264,
+      "learning_rate": 5.818515235204109e-07,
+      "loss": 1.0429,
+      "step": 7423
+    },
+    {
+      "epoch": 0.7807853602744949,
+      "grad_norm": 2.1998485383151873,
+      "learning_rate": 5.813162299341669e-07,
+      "loss": 0.971,
+      "step": 7424
+    },
+    {
+      "epoch": 0.7808905307163422,
+      "grad_norm": 2.364551392600312,
+      "learning_rate": 5.807811502960683e-07,
+      "loss": 0.9979,
+      "step": 7425
+    },
+    {
+      "epoch": 0.7809957011581895,
+      "grad_norm": 2.574638685273489,
+      "learning_rate": 5.802462846657811e-07,
+      "loss": 1.0097,
+      "step": 7426
+    },
+    {
+      "epoch": 0.7811008716000368,
+      "grad_norm": 2.1856284743115335,
+      "learning_rate": 5.797116331029456e-07,
+      "loss": 0.9878,
+      "step": 7427
+    },
+    {
+      "epoch": 0.7812060420418842,
+      "grad_norm": 2.1441801091958776,
+      "learning_rate": 5.79177195667181e-07,
+      "loss": 1.005,
+      "step": 7428
+    },
+    {
+      "epoch": 0.7813112124837315,
+      "grad_norm": 2.3913829746681805,
+      "learning_rate": 5.786429724180801e-07,
+      "loss": 0.9599,
+      "step": 7429
+    },
+    {
+      "epoch": 0.7814163829255788,
+      "grad_norm": 2.4163311751474006,
+      "learning_rate": 5.781089634152131e-07,
+      "loss": 0.9963,
+      "step": 7430
+    },
+    {
+      "epoch": 0.7815215533674261,
+      "grad_norm": 3.0068250817206343,
+      "learning_rate": 5.775751687181266e-07,
+      "loss": 0.9787,
+      "step": 7431
+    },
+    {
+      "epoch": 0.7816267238092734,
+      "grad_norm": 2.896952363328142,
+      "learning_rate": 5.770415883863434e-07,
+      "loss": 1.0102,
+      "step": 7432
+    },
+    {
+      "epoch": 0.7817318942511208,
+      "grad_norm": 2.342426916784213,
+      "learning_rate": 5.765082224793611e-07,
+      "loss": 0.9588,
+      "step": 7433
+    },
+    {
+      "epoch": 0.781837064692968,
+      "grad_norm": 2.322954163018337,
+      "learning_rate": 5.75975071056653e-07,
+      "loss": 0.9995,
+      "step": 7434
+    },
+    {
+      "epoch": 0.7819422351348153,
+      "grad_norm": 2.7124852787883476,
+      "learning_rate": 5.75442134177672e-07,
+      "loss": 0.9848,
+      "step": 7435
+    },
+    {
+      "epoch": 0.7820474055766626,
+      "grad_norm": 2.262898311145278,
+      "learning_rate": 5.749094119018431e-07,
+      "loss": 1.0258,
+      "step": 7436
+    },
+    {
+      "epoch": 0.78215257601851,
+      "grad_norm": 2.4401944041421793,
+      "learning_rate": 5.743769042885697e-07,
+      "loss": 0.9959,
+      "step": 7437
+    },
+    {
+      "epoch": 0.7822577464603573,
+      "grad_norm": 3.501822098234867,
+      "learning_rate": 5.738446113972298e-07,
+      "loss": 1.0054,
+      "step": 7438
+    },
+    {
+      "epoch": 0.7823629169022046,
+      "grad_norm": 2.4007429912831073,
+      "learning_rate": 5.733125332871786e-07,
+      "loss": 0.9699,
+      "step": 7439
+    },
+    {
+      "epoch": 0.7824680873440519,
+      "grad_norm": 1.5286404234678088,
+      "learning_rate": 5.727806700177477e-07,
+      "loss": 0.9517,
+      "step": 7440
+    },
+    {
+      "epoch": 0.7825732577858993,
+      "grad_norm": 1.9106204956021087,
+      "learning_rate": 5.722490216482424e-07,
+      "loss": 0.9594,
+      "step": 7441
+    },
+    {
+      "epoch": 0.7826784282277466,
+      "grad_norm": 2.392573929678364,
+      "learning_rate": 5.717175882379473e-07,
+      "loss": 0.9796,
+      "step": 7442
+    },
+    {
+      "epoch": 0.7827835986695939,
+      "grad_norm": 2.4076467512658484,
+      "learning_rate": 5.711863698461198e-07,
+      "loss": 0.9668,
+      "step": 7443
+    },
+    {
+      "epoch": 0.7828887691114412,
+      "grad_norm": 2.9227877491691956,
+      "learning_rate": 5.706553665319955e-07,
+      "loss": 1.0166,
+      "step": 7444
+    },
+    {
+      "epoch": 0.7829939395532886,
+      "grad_norm": 1.7837281364826976,
+      "learning_rate": 5.701245783547856e-07,
+      "loss": 0.9527,
+      "step": 7445
+    },
+    {
+      "epoch": 0.7830991099951359,
+      "grad_norm": 3.8675285298144995,
+      "learning_rate": 5.695940053736779e-07,
+      "loss": 0.9783,
+      "step": 7446
+    },
+    {
+      "epoch": 0.7832042804369832,
+      "grad_norm": 2.6749508025236657,
+      "learning_rate": 5.690636476478337e-07,
+      "loss": 0.9893,
+      "step": 7447
+    },
+    {
+      "epoch": 0.7833094508788305,
+      "grad_norm": 2.344089841049213,
+      "learning_rate": 5.685335052363927e-07,
+      "loss": 0.9719,
+      "step": 7448
+    },
+    {
+      "epoch": 0.7834146213206779,
+      "grad_norm": 2.6937529790986607,
+      "learning_rate": 5.680035781984708e-07,
+      "loss": 0.9855,
+      "step": 7449
+    },
+    {
+      "epoch": 0.7835197917625252,
+      "grad_norm": 2.309732894716535,
+      "learning_rate": 5.674738665931575e-07,
+      "loss": 0.9556,
+      "step": 7450
+    },
+    {
+      "epoch": 0.7836249622043725,
+      "grad_norm": 2.380015885121929,
+      "learning_rate": 5.669443704795214e-07,
+      "loss": 1.007,
+      "step": 7451
+    },
+    {
+      "epoch": 0.7837301326462198,
+      "grad_norm": 2.175460160357494,
+      "learning_rate": 5.664150899166035e-07,
+      "loss": 0.9653,
+      "step": 7452
+    },
+    {
+      "epoch": 0.7838353030880671,
+      "grad_norm": 2.493955497046477,
+      "learning_rate": 5.658860249634237e-07,
+      "loss": 0.9901,
+      "step": 7453
+    },
+    {
+      "epoch": 0.7839404735299145,
+      "grad_norm": 3.0014222312856185,
+      "learning_rate": 5.653571756789777e-07,
+      "loss": 0.9816,
+      "step": 7454
+    },
+    {
+      "epoch": 0.7840456439717617,
+      "grad_norm": 1.879178780187409,
+      "learning_rate": 5.648285421222347e-07,
+      "loss": 0.9732,
+      "step": 7455
+    },
+    {
+      "epoch": 0.784150814413609,
+      "grad_norm": 2.4428944524881455,
+      "learning_rate": 5.643001243521429e-07,
+      "loss": 0.9955,
+      "step": 7456
+    },
+    {
+      "epoch": 0.7842559848554563,
+      "grad_norm": 2.4832015579932363,
+      "learning_rate": 5.637719224276228e-07,
+      "loss": 0.982,
+      "step": 7457
+    },
+    {
+      "epoch": 0.7843611552973037,
+      "grad_norm": 2.7676659547524247,
+      "learning_rate": 5.632439364075759e-07,
+      "loss": 1.0221,
+      "step": 7458
+    },
+    {
+      "epoch": 0.784466325739151,
+      "grad_norm": 2.900996870688416,
+      "learning_rate": 5.627161663508745e-07,
+      "loss": 0.9675,
+      "step": 7459
+    },
+    {
+      "epoch": 0.7845714961809983,
+      "grad_norm": 2.6087512206617434,
+      "learning_rate": 5.621886123163708e-07,
+      "loss": 1.0026,
+      "step": 7460
+    },
+    {
+      "epoch": 0.7846766666228456,
+      "grad_norm": 2.075036647944145,
+      "learning_rate": 5.616612743628896e-07,
+      "loss": 0.9838,
+      "step": 7461
+    },
+    {
+      "epoch": 0.784781837064693,
+      "grad_norm": 2.300780190568228,
+      "learning_rate": 5.611341525492337e-07,
+      "loss": 0.9434,
+      "step": 7462
+    },
+    {
+      "epoch": 0.7848870075065403,
+      "grad_norm": 1.8042294075183316,
+      "learning_rate": 5.606072469341823e-07,
+      "loss": 0.9684,
+      "step": 7463
+    },
+    {
+      "epoch": 0.7849921779483876,
+      "grad_norm": 2.5965873366492382,
+      "learning_rate": 5.600805575764878e-07,
+      "loss": 1.0404,
+      "step": 7464
+    },
+    {
+      "epoch": 0.7850973483902349,
+      "grad_norm": 2.5263383523718552,
+      "learning_rate": 5.595540845348815e-07,
+      "loss": 0.96,
+      "step": 7465
+    },
+    {
+      "epoch": 0.7852025188320823,
+      "grad_norm": 2.5277784570322916,
+      "learning_rate": 5.590278278680683e-07,
+      "loss": 1.0546,
+      "step": 7466
+    },
+    {
+      "epoch": 0.7853076892739296,
+      "grad_norm": 2.7356237220044166,
+      "learning_rate": 5.585017876347301e-07,
+      "loss": 0.9886,
+      "step": 7467
+    },
+    {
+      "epoch": 0.7854128597157769,
+      "grad_norm": 2.8495412546660472,
+      "learning_rate": 5.579759638935254e-07,
+      "loss": 1.0331,
+      "step": 7468
+    },
+    {
+      "epoch": 0.7855180301576242,
+      "grad_norm": 2.053798417338012,
+      "learning_rate": 5.574503567030861e-07,
+      "loss": 0.9651,
+      "step": 7469
+    },
+    {
+      "epoch": 0.7856232005994715,
+      "grad_norm": 3.106394812966373,
+      "learning_rate": 5.569249661220233e-07,
+      "loss": 0.9802,
+      "step": 7470
+    },
+    {
+      "epoch": 0.7857283710413189,
+      "grad_norm": 2.8187679442298954,
+      "learning_rate": 5.563997922089193e-07,
+      "loss": 1.0118,
+      "step": 7471
+    },
+    {
+      "epoch": 0.7858335414831662,
+      "grad_norm": 2.5263959039316113,
+      "learning_rate": 5.558748350223386e-07,
+      "loss": 0.9958,
+      "step": 7472
+    },
+    {
+      "epoch": 0.7859387119250135,
+      "grad_norm": 2.6283622778437747,
+      "learning_rate": 5.553500946208154e-07,
+      "loss": 0.9912,
+      "step": 7473
+    },
+    {
+      "epoch": 0.7860438823668608,
+      "grad_norm": 1.9388197155123486,
+      "learning_rate": 5.548255710628639e-07,
+      "loss": 1.0043,
+      "step": 7474
+    },
+    {
+      "epoch": 0.7861490528087081,
+      "grad_norm": 3.725787391804805,
+      "learning_rate": 5.543012644069714e-07,
+      "loss": 0.9957,
+      "step": 7475
+    },
+    {
+      "epoch": 0.7862542232505554,
+      "grad_norm": 2.8279644242846778,
+      "learning_rate": 5.537771747116024e-07,
+      "loss": 1.0032,
+      "step": 7476
+    },
+    {
+      "epoch": 0.7863593936924027,
+      "grad_norm": 2.05494665096958,
+      "learning_rate": 5.53253302035198e-07,
+      "loss": 0.9707,
+      "step": 7477
+    },
+    {
+      "epoch": 0.78646456413425,
+      "grad_norm": 2.2246914275505922,
+      "learning_rate": 5.527296464361725e-07,
+      "loss": 0.9912,
+      "step": 7478
+    },
+    {
+      "epoch": 0.7865697345760974,
+      "grad_norm": 2.390526330629458,
+      "learning_rate": 5.522062079729192e-07,
+      "loss": 1.0125,
+      "step": 7479
+    },
+    {
+      "epoch": 0.7866749050179447,
+      "grad_norm": 2.3621903521012877,
+      "learning_rate": 5.516829867038034e-07,
+      "loss": 0.9723,
+      "step": 7480
+    },
+    {
+      "epoch": 0.786780075459792,
+      "grad_norm": 2.432774028158043,
+      "learning_rate": 5.511599826871708e-07,
+      "loss": 0.9861,
+      "step": 7481
+    },
+    {
+      "epoch": 0.7868852459016393,
+      "grad_norm": 3.122179435897935,
+      "learning_rate": 5.506371959813386e-07,
+      "loss": 0.9478,
+      "step": 7482
+    },
+    {
+      "epoch": 0.7869904163434867,
+      "grad_norm": 2.2616352167488705,
+      "learning_rate": 5.501146266446031e-07,
+      "loss": 0.9887,
+      "step": 7483
+    },
+    {
+      "epoch": 0.787095586785334,
+      "grad_norm": 1.7783208579319671,
+      "learning_rate": 5.495922747352336e-07,
+      "loss": 0.9767,
+      "step": 7484
+    },
+    {
+      "epoch": 0.7872007572271813,
+      "grad_norm": 2.7785258155561667,
+      "learning_rate": 5.490701403114757e-07,
+      "loss": 1.0049,
+      "step": 7485
+    },
+    {
+      "epoch": 0.7873059276690286,
+      "grad_norm": 2.063969890861771,
+      "learning_rate": 5.485482234315537e-07,
+      "loss": 0.994,
+      "step": 7486
+    },
+    {
+      "epoch": 0.787411098110876,
+      "grad_norm": 1.9172938948485974,
+      "learning_rate": 5.480265241536636e-07,
+      "loss": 0.981,
+      "step": 7487
+    },
+    {
+      "epoch": 0.7875162685527233,
+      "grad_norm": 2.3568236845146515,
+      "learning_rate": 5.475050425359805e-07,
+      "loss": 0.9963,
+      "step": 7488
+    },
+    {
+      "epoch": 0.7876214389945706,
+      "grad_norm": 2.365770278271118,
+      "learning_rate": 5.469837786366514e-07,
+      "loss": 0.9908,
+      "step": 7489
+    },
+    {
+      "epoch": 0.7877266094364179,
+      "grad_norm": 2.0233878116931843,
+      "learning_rate": 5.464627325138031e-07,
+      "loss": 0.955,
+      "step": 7490
+    },
+    {
+      "epoch": 0.7878317798782652,
+      "grad_norm": 2.8116469349236874,
+      "learning_rate": 5.45941904225536e-07,
+      "loss": 0.9791,
+      "step": 7491
+    },
+    {
+      "epoch": 0.7879369503201126,
+      "grad_norm": 2.389189229680465,
+      "learning_rate": 5.454212938299256e-07,
+      "loss": 0.9532,
+      "step": 7492
+    },
+    {
+      "epoch": 0.7880421207619599,
+      "grad_norm": 3.2244042201754755,
+      "learning_rate": 5.449009013850253e-07,
+      "loss": 1.0245,
+      "step": 7493
+    },
+    {
+      "epoch": 0.7881472912038072,
+      "grad_norm": 2.350726006073612,
+      "learning_rate": 5.443807269488613e-07,
+      "loss": 0.9812,
+      "step": 7494
+    },
+    {
+      "epoch": 0.7882524616456544,
+      "grad_norm": 2.33734989566254,
+      "learning_rate": 5.438607705794391e-07,
+      "loss": 0.977,
+      "step": 7495
+    },
+    {
+      "epoch": 0.7883576320875018,
+      "grad_norm": 3.1617106808359474,
+      "learning_rate": 5.43341032334736e-07,
+      "loss": 1.0034,
+      "step": 7496
+    },
+    {
+      "epoch": 0.7884628025293491,
+      "grad_norm": 2.062664190865852,
+      "learning_rate": 5.428215122727084e-07,
+      "loss": 0.9338,
+      "step": 7497
+    },
+    {
+      "epoch": 0.7885679729711964,
+      "grad_norm": 2.541915976209122,
+      "learning_rate": 5.423022104512854e-07,
+      "loss": 1.01,
+      "step": 7498
+    },
+    {
+      "epoch": 0.7886731434130437,
+      "grad_norm": 2.6573006964220665,
+      "learning_rate": 5.417831269283741e-07,
+      "loss": 0.9934,
+      "step": 7499
+    },
+    {
+      "epoch": 0.7887783138548911,
+      "grad_norm": 3.667198982465556,
+      "learning_rate": 5.412642617618565e-07,
+      "loss": 1.0216,
+      "step": 7500
+    },
+    {
+      "epoch": 0.7888834842967384,
+      "grad_norm": 2.450330102335447,
+      "learning_rate": 5.407456150095891e-07,
+      "loss": 0.9705,
+      "step": 7501
+    },
+    {
+      "epoch": 0.7889886547385857,
+      "grad_norm": 2.6639634437563466,
+      "learning_rate": 5.402271867294062e-07,
+      "loss": 1.0375,
+      "step": 7502
+    },
+    {
+      "epoch": 0.789093825180433,
+      "grad_norm": 2.215568756488248,
+      "learning_rate": 5.39708976979115e-07,
+      "loss": 0.9998,
+      "step": 7503
+    },
+    {
+      "epoch": 0.7891989956222804,
+      "grad_norm": 2.7328182673914556,
+      "learning_rate": 5.39190985816502e-07,
+      "loss": 1.0026,
+      "step": 7504
+    },
+    {
+      "epoch": 0.7893041660641277,
+      "grad_norm": 2.3163502581110684,
+      "learning_rate": 5.386732132993264e-07,
+      "loss": 0.9868,
+      "step": 7505
+    },
+    {
+      "epoch": 0.789409336505975,
+      "grad_norm": 2.0522924122717083,
+      "learning_rate": 5.381556594853226e-07,
+      "loss": 0.9908,
+      "step": 7506
+    },
+    {
+      "epoch": 0.7895145069478223,
+      "grad_norm": 2.6138338131436214,
+      "learning_rate": 5.376383244322039e-07,
+      "loss": 0.9801,
+      "step": 7507
+    },
+    {
+      "epoch": 0.7896196773896696,
+      "grad_norm": 1.6691242712576417,
+      "learning_rate": 5.371212081976548e-07,
+      "loss": 0.9766,
+      "step": 7508
+    },
+    {
+      "epoch": 0.789724847831517,
+      "grad_norm": 2.8741546215777496,
+      "learning_rate": 5.366043108393407e-07,
+      "loss": 0.9829,
+      "step": 7509
+    },
+    {
+      "epoch": 0.7898300182733643,
+      "grad_norm": 2.2165275941322284,
+      "learning_rate": 5.360876324148972e-07,
+      "loss": 0.9629,
+      "step": 7510
+    },
+    {
+      "epoch": 0.7899351887152116,
+      "grad_norm": 1.605507267413963,
+      "learning_rate": 5.355711729819396e-07,
+      "loss": 0.9566,
+      "step": 7511
+    },
+    {
+      "epoch": 0.790040359157059,
+      "grad_norm": 2.2184467134191035,
+      "learning_rate": 5.350549325980558e-07,
+      "loss": 0.9624,
+      "step": 7512
+    },
+    {
+      "epoch": 0.7901455295989063,
+      "grad_norm": 2.840349144152613,
+      "learning_rate": 5.34538911320811e-07,
+      "loss": 0.9807,
+      "step": 7513
+    },
+    {
+      "epoch": 0.7902507000407536,
+      "grad_norm": 2.6440663804972333,
+      "learning_rate": 5.340231092077469e-07,
+      "loss": 0.998,
+      "step": 7514
+    },
+    {
+      "epoch": 0.7903558704826009,
+      "grad_norm": 2.9746470518481423,
+      "learning_rate": 5.335075263163774e-07,
+      "loss": 0.9896,
+      "step": 7515
+    },
+    {
+      "epoch": 0.7904610409244481,
+      "grad_norm": 2.217261654951149,
+      "learning_rate": 5.329921627041959e-07,
+      "loss": 0.9776,
+      "step": 7516
+    },
+    {
+      "epoch": 0.7905662113662955,
+      "grad_norm": 2.275129661652413,
+      "learning_rate": 5.324770184286668e-07,
+      "loss": 0.9754,
+      "step": 7517
+    },
+    {
+      "epoch": 0.7906713818081428,
+      "grad_norm": 2.125756351189819,
+      "learning_rate": 5.319620935472361e-07,
+      "loss": 0.9576,
+      "step": 7518
+    },
+    {
+      "epoch": 0.7907765522499901,
+      "grad_norm": 1.8544116319858905,
+      "learning_rate": 5.314473881173193e-07,
+      "loss": 1.0268,
+      "step": 7519
+    },
+    {
+      "epoch": 0.7908817226918374,
+      "grad_norm": 2.7586623648851947,
+      "learning_rate": 5.309329021963116e-07,
+      "loss": 0.9541,
+      "step": 7520
+    },
+    {
+      "epoch": 0.7909868931336848,
+      "grad_norm": 2.0077432117735503,
+      "learning_rate": 5.30418635841582e-07,
+      "loss": 1.0106,
+      "step": 7521
+    },
+    {
+      "epoch": 0.7910920635755321,
+      "grad_norm": 2.211322527248184,
+      "learning_rate": 5.299045891104729e-07,
+      "loss": 0.9729,
+      "step": 7522
+    },
+    {
+      "epoch": 0.7911972340173794,
+      "grad_norm": 3.561907418437129,
+      "learning_rate": 5.293907620603081e-07,
+      "loss": 1.0103,
+      "step": 7523
+    },
+    {
+      "epoch": 0.7913024044592267,
+      "grad_norm": 2.919228767233691,
+      "learning_rate": 5.288771547483807e-07,
+      "loss": 1.0016,
+      "step": 7524
+    },
+    {
+      "epoch": 0.791407574901074,
+      "grad_norm": 2.4514739629485764,
+      "learning_rate": 5.283637672319633e-07,
+      "loss": 1.0114,
+      "step": 7525
+    },
+    {
+      "epoch": 0.7915127453429214,
+      "grad_norm": 2.248230346805148,
+      "learning_rate": 5.278505995683014e-07,
+      "loss": 0.9975,
+      "step": 7526
+    },
+    {
+      "epoch": 0.7916179157847687,
+      "grad_norm": 2.299965654128683,
+      "learning_rate": 5.273376518146181e-07,
+      "loss": 0.9858,
+      "step": 7527
+    },
+    {
+      "epoch": 0.791723086226616,
+      "grad_norm": 2.631945993060661,
+      "learning_rate": 5.268249240281112e-07,
+      "loss": 1.053,
+      "step": 7528
+    },
+    {
+      "epoch": 0.7918282566684633,
+      "grad_norm": 2.965446830432594,
+      "learning_rate": 5.263124162659531e-07,
+      "loss": 1.0042,
+      "step": 7529
+    },
+    {
+      "epoch": 0.7919334271103107,
+      "grad_norm": 2.8523687341385866,
+      "learning_rate": 5.258001285852931e-07,
+      "loss": 0.998,
+      "step": 7530
+    },
+    {
+      "epoch": 0.792038597552158,
+      "grad_norm": 2.8763350701028076,
+      "learning_rate": 5.252880610432537e-07,
+      "loss": 0.9963,
+      "step": 7531
+    },
+    {
+      "epoch": 0.7921437679940053,
+      "grad_norm": 2.7401215806024384,
+      "learning_rate": 5.247762136969367e-07,
+      "loss": 0.9984,
+      "step": 7532
+    },
+    {
+      "epoch": 0.7922489384358526,
+      "grad_norm": 2.381855054501238,
+      "learning_rate": 5.242645866034154e-07,
+      "loss": 0.9807,
+      "step": 7533
+    },
+    {
+      "epoch": 0.7923541088777,
+      "grad_norm": 2.2923915086079303,
+      "learning_rate": 5.237531798197415e-07,
+      "loss": 0.9697,
+      "step": 7534
+    },
+    {
+      "epoch": 0.7924592793195473,
+      "grad_norm": 2.5095749130378695,
+      "learning_rate": 5.232419934029395e-07,
+      "loss": 0.9681,
+      "step": 7535
+    },
+    {
+      "epoch": 0.7925644497613945,
+      "grad_norm": 2.3453279983832793,
+      "learning_rate": 5.227310274100112e-07,
+      "loss": 1.0093,
+      "step": 7536
+    },
+    {
+      "epoch": 0.7926696202032418,
+      "grad_norm": 2.140545124978573,
+      "learning_rate": 5.222202818979338e-07,
+      "loss": 0.9813,
+      "step": 7537
+    },
+    {
+      "epoch": 0.7927747906450892,
+      "grad_norm": 2.4063998911718634,
+      "learning_rate": 5.217097569236581e-07,
+      "loss": 0.9952,
+      "step": 7538
+    },
+    {
+      "epoch": 0.7928799610869365,
+      "grad_norm": 3.1685790964036076,
+      "learning_rate": 5.211994525441133e-07,
+      "loss": 0.9982,
+      "step": 7539
+    },
+    {
+      "epoch": 0.7929851315287838,
+      "grad_norm": 2.789219170600396,
+      "learning_rate": 5.20689368816201e-07,
+      "loss": 1.0258,
+      "step": 7540
+    },
+    {
+      "epoch": 0.7930903019706311,
+      "grad_norm": 2.2210430744775196,
+      "learning_rate": 5.201795057967995e-07,
+      "loss": 1.0254,
+      "step": 7541
+    },
+    {
+      "epoch": 0.7931954724124785,
+      "grad_norm": 2.7466488855873314,
+      "learning_rate": 5.196698635427636e-07,
+      "loss": 0.9965,
+      "step": 7542
+    },
+    {
+      "epoch": 0.7933006428543258,
+      "grad_norm": 2.225180678588441,
+      "learning_rate": 5.191604421109211e-07,
+      "loss": 0.986,
+      "step": 7543
+    },
+    {
+      "epoch": 0.7934058132961731,
+      "grad_norm": 2.553086727079915,
+      "learning_rate": 5.186512415580778e-07,
+      "loss": 0.971,
+      "step": 7544
+    },
+    {
+      "epoch": 0.7935109837380204,
+      "grad_norm": 2.382433854163622,
+      "learning_rate": 5.181422619410109e-07,
+      "loss": 0.9843,
+      "step": 7545
+    },
+    {
+      "epoch": 0.7936161541798677,
+      "grad_norm": 2.3343492155076127,
+      "learning_rate": 5.176335033164793e-07,
+      "loss": 1.0111,
+      "step": 7546
+    },
+    {
+      "epoch": 0.7937213246217151,
+      "grad_norm": 2.256309064903577,
+      "learning_rate": 5.171249657412106e-07,
+      "loss": 0.9924,
+      "step": 7547
+    },
+    {
+      "epoch": 0.7938264950635624,
+      "grad_norm": 2.8424568000686192,
+      "learning_rate": 5.166166492719124e-07,
+      "loss": 0.9672,
+      "step": 7548
+    },
+    {
+      "epoch": 0.7939316655054097,
+      "grad_norm": 2.563574424214626,
+      "learning_rate": 5.161085539652649e-07,
+      "loss": 0.9811,
+      "step": 7549
+    },
+    {
+      "epoch": 0.794036835947257,
+      "grad_norm": 1.8740387198796324,
+      "learning_rate": 5.156006798779248e-07,
+      "loss": 1.0023,
+      "step": 7550
+    },
+    {
+      "epoch": 0.7941420063891044,
+      "grad_norm": 2.7294780666990612,
+      "learning_rate": 5.15093027066525e-07,
+      "loss": 0.9843,
+      "step": 7551
+    },
+    {
+      "epoch": 0.7942471768309517,
+      "grad_norm": 2.4042826855611787,
+      "learning_rate": 5.145855955876713e-07,
+      "loss": 1.0067,
+      "step": 7552
+    },
+    {
+      "epoch": 0.794352347272799,
+      "grad_norm": 2.6315789434241643,
+      "learning_rate": 5.140783854979475e-07,
+      "loss": 0.9392,
+      "step": 7553
+    },
+    {
+      "epoch": 0.7944575177146463,
+      "grad_norm": 2.3247912098399364,
+      "learning_rate": 5.135713968539096e-07,
+      "loss": 1.0035,
+      "step": 7554
+    },
+    {
+      "epoch": 0.7945626881564937,
+      "grad_norm": 2.311734008689543,
+      "learning_rate": 5.130646297120936e-07,
+      "loss": 1.0073,
+      "step": 7555
+    },
+    {
+      "epoch": 0.7946678585983409,
+      "grad_norm": 2.771031149106594,
+      "learning_rate": 5.125580841290063e-07,
+      "loss": 0.9839,
+      "step": 7556
+    },
+    {
+      "epoch": 0.7947730290401882,
+      "grad_norm": 2.5613460622634263,
+      "learning_rate": 5.12051760161131e-07,
+      "loss": 0.964,
+      "step": 7557
+    },
+    {
+      "epoch": 0.7948781994820355,
+      "grad_norm": 2.9699017410914097,
+      "learning_rate": 5.115456578649272e-07,
+      "loss": 0.9706,
+      "step": 7558
+    },
+    {
+      "epoch": 0.7949833699238829,
+      "grad_norm": 2.8383038959737035,
+      "learning_rate": 5.110397772968295e-07,
+      "loss": 1.0041,
+      "step": 7559
+    },
+    {
+      "epoch": 0.7950885403657302,
+      "grad_norm": 2.4690199798689925,
+      "learning_rate": 5.105341185132481e-07,
+      "loss": 1.0097,
+      "step": 7560
+    },
+    {
+      "epoch": 0.7951937108075775,
+      "grad_norm": 2.253149181816096,
+      "learning_rate": 5.100286815705665e-07,
+      "loss": 0.973,
+      "step": 7561
+    },
+    {
+      "epoch": 0.7952988812494248,
+      "grad_norm": 2.6862303113933934,
+      "learning_rate": 5.095234665251464e-07,
+      "loss": 0.9876,
+      "step": 7562
+    },
+    {
+      "epoch": 0.7954040516912722,
+      "grad_norm": 2.7435662802894734,
+      "learning_rate": 5.090184734333214e-07,
+      "loss": 0.9361,
+      "step": 7563
+    },
+    {
+      "epoch": 0.7955092221331195,
+      "grad_norm": 2.149989633339049,
+      "learning_rate": 5.085137023514034e-07,
+      "loss": 1.0072,
+      "step": 7564
+    },
+    {
+      "epoch": 0.7956143925749668,
+      "grad_norm": 2.5740579082453494,
+      "learning_rate": 5.080091533356784e-07,
+      "loss": 0.9908,
+      "step": 7565
+    },
+    {
+      "epoch": 0.7957195630168141,
+      "grad_norm": 2.3471609710728822,
+      "learning_rate": 5.075048264424065e-07,
+      "loss": 1.0258,
+      "step": 7566
+    },
+    {
+      "epoch": 0.7958247334586614,
+      "grad_norm": 2.406035264757779,
+      "learning_rate": 5.070007217278253e-07,
+      "loss": 0.9846,
+      "step": 7567
+    },
+    {
+      "epoch": 0.7959299039005088,
+      "grad_norm": 2.117858329168068,
+      "learning_rate": 5.064968392481448e-07,
+      "loss": 0.9419,
+      "step": 7568
+    },
+    {
+      "epoch": 0.7960350743423561,
+      "grad_norm": 2.193661332028775,
+      "learning_rate": 5.059931790595538e-07,
+      "loss": 0.9726,
+      "step": 7569
+    },
+    {
+      "epoch": 0.7961402447842034,
+      "grad_norm": 3.4342436943341283,
+      "learning_rate": 5.054897412182128e-07,
+      "loss": 0.9513,
+      "step": 7570
+    },
+    {
+      "epoch": 0.7962454152260507,
+      "grad_norm": 2.6107851702888367,
+      "learning_rate": 5.049865257802599e-07,
+      "loss": 1.0154,
+      "step": 7571
+    },
+    {
+      "epoch": 0.7963505856678981,
+      "grad_norm": 2.593692528488154,
+      "learning_rate": 5.044835328018069e-07,
+      "loss": 0.9833,
+      "step": 7572
+    },
+    {
+      "epoch": 0.7964557561097454,
+      "grad_norm": 2.374728480707458,
+      "learning_rate": 5.039807623389418e-07,
+      "loss": 0.9846,
+      "step": 7573
+    },
+    {
+      "epoch": 0.7965609265515927,
+      "grad_norm": 3.1091749055796667,
+      "learning_rate": 5.034782144477276e-07,
+      "loss": 1.0286,
+      "step": 7574
+    },
+    {
+      "epoch": 0.79666609699344,
+      "grad_norm": 2.3409299896756965,
+      "learning_rate": 5.029758891842015e-07,
+      "loss": 1.0111,
+      "step": 7575
+    },
+    {
+      "epoch": 0.7967712674352874,
+      "grad_norm": 2.659765654383437,
+      "learning_rate": 5.02473786604378e-07,
+      "loss": 0.9769,
+      "step": 7576
+    },
+    {
+      "epoch": 0.7968764378771346,
+      "grad_norm": 2.6745888226219923,
+      "learning_rate": 5.019719067642437e-07,
+      "loss": 0.9902,
+      "step": 7577
+    },
+    {
+      "epoch": 0.7969816083189819,
+      "grad_norm": 2.987835654743405,
+      "learning_rate": 5.014702497197632e-07,
+      "loss": 1.028,
+      "step": 7578
+    },
+    {
+      "epoch": 0.7970867787608292,
+      "grad_norm": 2.401381290760075,
+      "learning_rate": 5.009688155268755e-07,
+      "loss": 0.9726,
+      "step": 7579
+    },
+    {
+      "epoch": 0.7971919492026766,
+      "grad_norm": 2.120903536373421,
+      "learning_rate": 5.004676042414936e-07,
+      "loss": 0.9738,
+      "step": 7580
+    },
+    {
+      "epoch": 0.7972971196445239,
+      "grad_norm": 2.2322743664947065,
+      "learning_rate": 4.999666159195063e-07,
+      "loss": 0.9493,
+      "step": 7581
+    },
+    {
+      "epoch": 0.7974022900863712,
+      "grad_norm": 3.0099693723923004,
+      "learning_rate": 4.994658506167782e-07,
+      "loss": 1.0182,
+      "step": 7582
+    },
+    {
+      "epoch": 0.7975074605282185,
+      "grad_norm": 1.656585929583154,
+      "learning_rate": 4.989653083891493e-07,
+      "loss": 0.9718,
+      "step": 7583
+    },
+    {
+      "epoch": 0.7976126309700659,
+      "grad_norm": 2.5313550286466056,
+      "learning_rate": 4.984649892924323e-07,
+      "loss": 0.9698,
+      "step": 7584
+    },
+    {
+      "epoch": 0.7977178014119132,
+      "grad_norm": 3.027950770126478,
+      "learning_rate": 4.979648933824183e-07,
+      "loss": 1.0244,
+      "step": 7585
+    },
+    {
+      "epoch": 0.7978229718537605,
+      "grad_norm": 3.251073308685086,
+      "learning_rate": 4.974650207148701e-07,
+      "loss": 0.9823,
+      "step": 7586
+    },
+    {
+      "epoch": 0.7979281422956078,
+      "grad_norm": 2.2190156660054363,
+      "learning_rate": 4.969653713455286e-07,
+      "loss": 0.979,
+      "step": 7587
+    },
+    {
+      "epoch": 0.7980333127374551,
+      "grad_norm": 2.1132623922709732,
+      "learning_rate": 4.964659453301088e-07,
+      "loss": 0.9428,
+      "step": 7588
+    },
+    {
+      "epoch": 0.7981384831793025,
+      "grad_norm": 2.2951743727023306,
+      "learning_rate": 4.959667427242995e-07,
+      "loss": 0.9855,
+      "step": 7589
+    },
+    {
+      "epoch": 0.7982436536211498,
+      "grad_norm": 2.2665071186589705,
+      "learning_rate": 4.954677635837668e-07,
+      "loss": 0.9737,
+      "step": 7590
+    },
+    {
+      "epoch": 0.7983488240629971,
+      "grad_norm": 2.327325568834645,
+      "learning_rate": 4.949690079641498e-07,
+      "loss": 0.9589,
+      "step": 7591
+    },
+    {
+      "epoch": 0.7984539945048444,
+      "grad_norm": 2.8604374797701353,
+      "learning_rate": 4.944704759210637e-07,
+      "loss": 0.9786,
+      "step": 7592
+    },
+    {
+      "epoch": 0.7985591649466918,
+      "grad_norm": 2.2699713114872684,
+      "learning_rate": 4.939721675101001e-07,
+      "loss": 0.9744,
+      "step": 7593
+    },
+    {
+      "epoch": 0.7986643353885391,
+      "grad_norm": 2.1867243451965,
+      "learning_rate": 4.934740827868221e-07,
+      "loss": 0.9416,
+      "step": 7594
+    },
+    {
+      "epoch": 0.7987695058303864,
+      "grad_norm": 2.032577080022669,
+      "learning_rate": 4.929762218067713e-07,
+      "loss": 0.9277,
+      "step": 7595
+    },
+    {
+      "epoch": 0.7988746762722337,
+      "grad_norm": 2.469836920824304,
+      "learning_rate": 4.924785846254629e-07,
+      "loss": 1.0052,
+      "step": 7596
+    },
+    {
+      "epoch": 0.798979846714081,
+      "grad_norm": 2.2285999066069544,
+      "learning_rate": 4.91981171298388e-07,
+      "loss": 1.0035,
+      "step": 7597
+    },
+    {
+      "epoch": 0.7990850171559283,
+      "grad_norm": 2.7998486974210626,
+      "learning_rate": 4.914839818810105e-07,
+      "loss": 0.9873,
+      "step": 7598
+    },
+    {
+      "epoch": 0.7991901875977756,
+      "grad_norm": 2.2023794298511308,
+      "learning_rate": 4.909870164287725e-07,
+      "loss": 0.9966,
+      "step": 7599
+    },
+    {
+      "epoch": 0.7992953580396229,
+      "grad_norm": 2.5707030773678423,
+      "learning_rate": 4.904902749970883e-07,
+      "loss": 0.9916,
+      "step": 7600
+    },
+    {
+      "epoch": 0.7994005284814703,
+      "grad_norm": 2.3180485019623442,
+      "learning_rate": 4.899937576413486e-07,
+      "loss": 0.961,
+      "step": 7601
+    },
+    {
+      "epoch": 0.7995056989233176,
+      "grad_norm": 2.3791731807882335,
+      "learning_rate": 4.894974644169198e-07,
+      "loss": 0.9891,
+      "step": 7602
+    },
+    {
+      "epoch": 0.7996108693651649,
+      "grad_norm": 2.479486951922807,
+      "learning_rate": 4.890013953791411e-07,
+      "loss": 0.9645,
+      "step": 7603
+    },
+    {
+      "epoch": 0.7997160398070122,
+      "grad_norm": 2.633404751177737,
+      "learning_rate": 4.885055505833291e-07,
+      "loss": 0.9565,
+      "step": 7604
+    },
+    {
+      "epoch": 0.7998212102488595,
+      "grad_norm": 2.684117417959632,
+      "learning_rate": 4.880099300847741e-07,
+      "loss": 0.9838,
+      "step": 7605
+    },
+    {
+      "epoch": 0.7999263806907069,
+      "grad_norm": 2.8746409642803243,
+      "learning_rate": 4.875145339387418e-07,
+      "loss": 1.036,
+      "step": 7606
+    },
+    {
+      "epoch": 0.8000315511325542,
+      "grad_norm": 2.0807055982824276,
+      "learning_rate": 4.870193622004729e-07,
+      "loss": 0.9486,
+      "step": 7607
+    },
+    {
+      "epoch": 0.8001367215744015,
+      "grad_norm": 2.471988925752685,
+      "learning_rate": 4.865244149251815e-07,
+      "loss": 0.9727,
+      "step": 7608
+    },
+    {
+      "epoch": 0.8002418920162488,
+      "grad_norm": 1.9412511005330855,
+      "learning_rate": 4.860296921680593e-07,
+      "loss": 0.9654,
+      "step": 7609
+    },
+    {
+      "epoch": 0.8003470624580962,
+      "grad_norm": 1.9790419108013764,
+      "learning_rate": 4.855351939842717e-07,
+      "loss": 0.9808,
+      "step": 7610
+    },
+    {
+      "epoch": 0.8004522328999435,
+      "grad_norm": 2.4043654013707254,
+      "learning_rate": 4.850409204289591e-07,
+      "loss": 0.9727,
+      "step": 7611
+    },
+    {
+      "epoch": 0.8005574033417908,
+      "grad_norm": 2.7262779181762666,
+      "learning_rate": 4.845468715572363e-07,
+      "loss": 0.9834,
+      "step": 7612
+    },
+    {
+      "epoch": 0.8006625737836381,
+      "grad_norm": 2.2391899912644715,
+      "learning_rate": 4.840530474241944e-07,
+      "loss": 0.9767,
+      "step": 7613
+    },
+    {
+      "epoch": 0.8007677442254855,
+      "grad_norm": 2.4174326455832977,
+      "learning_rate": 4.835594480848974e-07,
+      "loss": 0.9769,
+      "step": 7614
+    },
+    {
+      "epoch": 0.8008729146673328,
+      "grad_norm": 2.6250830234736307,
+      "learning_rate": 4.830660735943865e-07,
+      "loss": 0.9464,
+      "step": 7615
+    },
+    {
+      "epoch": 0.8009780851091801,
+      "grad_norm": 2.34149423966112,
+      "learning_rate": 4.825729240076768e-07,
+      "loss": 1.0037,
+      "step": 7616
+    },
+    {
+      "epoch": 0.8010832555510273,
+      "grad_norm": 1.8300856800483398,
+      "learning_rate": 4.820799993797576e-07,
+      "loss": 0.9735,
+      "step": 7617
+    },
+    {
+      "epoch": 0.8011884259928747,
+      "grad_norm": 1.764340560176855,
+      "learning_rate": 4.815872997655941e-07,
+      "loss": 0.9249,
+      "step": 7618
+    },
+    {
+      "epoch": 0.801293596434722,
+      "grad_norm": 2.320272163563018,
+      "learning_rate": 4.810948252201261e-07,
+      "loss": 0.9664,
+      "step": 7619
+    },
+    {
+      "epoch": 0.8013987668765693,
+      "grad_norm": 3.0912542656654733,
+      "learning_rate": 4.806025757982694e-07,
+      "loss": 0.9396,
+      "step": 7620
+    },
+    {
+      "epoch": 0.8015039373184166,
+      "grad_norm": 2.2930284148085316,
+      "learning_rate": 4.80110551554912e-07,
+      "loss": 0.9745,
+      "step": 7621
+    },
+    {
+      "epoch": 0.801609107760264,
+      "grad_norm": 2.3498868797024106,
+      "learning_rate": 4.796187525449198e-07,
+      "loss": 1.0354,
+      "step": 7622
+    },
+    {
+      "epoch": 0.8017142782021113,
+      "grad_norm": 2.2536909100182534,
+      "learning_rate": 4.791271788231309e-07,
+      "loss": 1.011,
+      "step": 7623
+    },
+    {
+      "epoch": 0.8018194486439586,
+      "grad_norm": 2.5584982606240603,
+      "learning_rate": 4.786358304443603e-07,
+      "loss": 1.0019,
+      "step": 7624
+    },
+    {
+      "epoch": 0.8019246190858059,
+      "grad_norm": 2.356806815511279,
+      "learning_rate": 4.781447074633977e-07,
+      "loss": 0.9796,
+      "step": 7625
+    },
+    {
+      "epoch": 0.8020297895276532,
+      "grad_norm": 2.654243031206716,
+      "learning_rate": 4.776538099350058e-07,
+      "loss": 0.966,
+      "step": 7626
+    },
+    {
+      "epoch": 0.8021349599695006,
+      "grad_norm": 2.7182415623648843,
+      "learning_rate": 4.771631379139244e-07,
+      "loss": 0.9779,
+      "step": 7627
+    },
+    {
+      "epoch": 0.8022401304113479,
+      "grad_norm": 2.5630762496199044,
+      "learning_rate": 4.7667269145486777e-07,
+      "loss": 0.9753,
+      "step": 7628
+    },
+    {
+      "epoch": 0.8023453008531952,
+      "grad_norm": 2.485238287825221,
+      "learning_rate": 4.7618247061252286e-07,
+      "loss": 0.9542,
+      "step": 7629
+    },
+    {
+      "epoch": 0.8024504712950425,
+      "grad_norm": 3.0101852738819375,
+      "learning_rate": 4.7569247544155486e-07,
+      "loss": 1.0098,
+      "step": 7630
+    },
+    {
+      "epoch": 0.8025556417368899,
+      "grad_norm": 2.4862972563714694,
+      "learning_rate": 4.7520270599660053e-07,
+      "loss": 0.907,
+      "step": 7631
+    },
+    {
+      "epoch": 0.8026608121787372,
+      "grad_norm": 3.043564384568565,
+      "learning_rate": 4.747131623322737e-07,
+      "loss": 1.0138,
+      "step": 7632
+    },
+    {
+      "epoch": 0.8027659826205845,
+      "grad_norm": 2.1240214301887583,
+      "learning_rate": 4.7422384450316244e-07,
+      "loss": 1.0008,
+      "step": 7633
+    },
+    {
+      "epoch": 0.8028711530624318,
+      "grad_norm": 2.6774258034215537,
+      "learning_rate": 4.7373475256382973e-07,
+      "loss": 0.9516,
+      "step": 7634
+    },
+    {
+      "epoch": 0.8029763235042792,
+      "grad_norm": 2.5417333491458285,
+      "learning_rate": 4.7324588656881194e-07,
+      "loss": 1.0021,
+      "step": 7635
+    },
+    {
+      "epoch": 0.8030814939461265,
+      "grad_norm": 2.8488514697489786,
+      "learning_rate": 4.7275724657262293e-07,
+      "loss": 0.9775,
+      "step": 7636
+    },
+    {
+      "epoch": 0.8031866643879738,
+      "grad_norm": 2.3577591104181437,
+      "learning_rate": 4.7226883262974865e-07,
+      "loss": 1.0096,
+      "step": 7637
+    },
+    {
+      "epoch": 0.803291834829821,
+      "grad_norm": 2.6074240401396542,
+      "learning_rate": 4.717806447946513e-07,
+      "loss": 0.9777,
+      "step": 7638
+    },
+    {
+      "epoch": 0.8033970052716684,
+      "grad_norm": 3.3809444206445995,
+      "learning_rate": 4.712926831217685e-07,
+      "loss": 0.9335,
+      "step": 7639
+    },
+    {
+      "epoch": 0.8035021757135157,
+      "grad_norm": 2.0108425607516263,
+      "learning_rate": 4.708049476655105e-07,
+      "loss": 0.9809,
+      "step": 7640
+    },
+    {
+      "epoch": 0.803607346155363,
+      "grad_norm": 2.6521203726038514,
+      "learning_rate": 4.703174384802639e-07,
+      "loss": 1.0248,
+      "step": 7641
+    },
+    {
+      "epoch": 0.8037125165972103,
+      "grad_norm": 2.848291027442207,
+      "learning_rate": 4.69830155620391e-07,
+      "loss": 0.9835,
+      "step": 7642
+    },
+    {
+      "epoch": 0.8038176870390576,
+      "grad_norm": 2.475573839743065,
+      "learning_rate": 4.693430991402259e-07,
+      "loss": 0.9945,
+      "step": 7643
+    },
+    {
+      "epoch": 0.803922857480905,
+      "grad_norm": 2.345453581854106,
+      "learning_rate": 4.6885626909408037e-07,
+      "loss": 0.9622,
+      "step": 7644
+    },
+    {
+      "epoch": 0.8040280279227523,
+      "grad_norm": 2.7478312834618808,
+      "learning_rate": 4.683696655362391e-07,
+      "loss": 0.9866,
+      "step": 7645
+    },
+    {
+      "epoch": 0.8041331983645996,
+      "grad_norm": 1.9800930016394065,
+      "learning_rate": 4.678832885209622e-07,
+      "loss": 1.0,
+      "step": 7646
+    },
+    {
+      "epoch": 0.804238368806447,
+      "grad_norm": 1.6413329542669401,
+      "learning_rate": 4.6739713810248443e-07,
+      "loss": 0.9856,
+      "step": 7647
+    },
+    {
+      "epoch": 0.8043435392482943,
+      "grad_norm": 2.414821706263618,
+      "learning_rate": 4.6691121433501657e-07,
+      "loss": 0.9904,
+      "step": 7648
+    },
+    {
+      "epoch": 0.8044487096901416,
+      "grad_norm": 2.4848547284531506,
+      "learning_rate": 4.664255172727411e-07,
+      "loss": 0.9819,
+      "step": 7649
+    },
+    {
+      "epoch": 0.8045538801319889,
+      "grad_norm": 1.9804328938918532,
+      "learning_rate": 4.6594004696981774e-07,
+      "loss": 0.9987,
+      "step": 7650
+    },
+    {
+      "epoch": 0.8046590505738362,
+      "grad_norm": 2.5014058610541983,
+      "learning_rate": 4.65454803480381e-07,
+      "loss": 0.9909,
+      "step": 7651
+    },
+    {
+      "epoch": 0.8047642210156836,
+      "grad_norm": 2.202903476738561,
+      "learning_rate": 4.649697868585379e-07,
+      "loss": 0.9658,
+      "step": 7652
+    },
+    {
+      "epoch": 0.8048693914575309,
+      "grad_norm": 2.380362604444181,
+      "learning_rate": 4.644849971583729e-07,
+      "loss": 1.0007,
+      "step": 7653
+    },
+    {
+      "epoch": 0.8049745618993782,
+      "grad_norm": 6.351090296949297,
+      "learning_rate": 4.6400043443394246e-07,
+      "loss": 1.0105,
+      "step": 7654
+    },
+    {
+      "epoch": 0.8050797323412255,
+      "grad_norm": 2.8797515711713615,
+      "learning_rate": 4.6351609873927976e-07,
+      "loss": 1.03,
+      "step": 7655
+    },
+    {
+      "epoch": 0.8051849027830729,
+      "grad_norm": 3.2146357197391846,
+      "learning_rate": 4.6303199012839225e-07,
+      "loss": 1.0333,
+      "step": 7656
+    },
+    {
+      "epoch": 0.8052900732249202,
+      "grad_norm": 2.079836576250059,
+      "learning_rate": 4.6254810865526187e-07,
+      "loss": 0.9962,
+      "step": 7657
+    },
+    {
+      "epoch": 0.8053952436667674,
+      "grad_norm": 1.649972670577275,
+      "learning_rate": 4.6206445437384504e-07,
+      "loss": 0.9932,
+      "step": 7658
+    },
+    {
+      "epoch": 0.8055004141086147,
+      "grad_norm": 1.9272657097173842,
+      "learning_rate": 4.615810273380722e-07,
+      "loss": 0.9981,
+      "step": 7659
+    },
+    {
+      "epoch": 0.805605584550462,
+      "grad_norm": 2.8001761793669577,
+      "learning_rate": 4.610978276018496e-07,
+      "loss": 0.9994,
+      "step": 7660
+    },
+    {
+      "epoch": 0.8057107549923094,
+      "grad_norm": 2.0720005349560995,
+      "learning_rate": 4.6061485521905815e-07,
+      "loss": 0.9708,
+      "step": 7661
+    },
+    {
+      "epoch": 0.8058159254341567,
+      "grad_norm": 1.9105799009272426,
+      "learning_rate": 4.6013211024355353e-07,
+      "loss": 1.0271,
+      "step": 7662
+    },
+    {
+      "epoch": 0.805921095876004,
+      "grad_norm": 1.9282586380528615,
+      "learning_rate": 4.5964959272916415e-07,
+      "loss": 1.0037,
+      "step": 7663
+    },
+    {
+      "epoch": 0.8060262663178513,
+      "grad_norm": 1.9696455842256768,
+      "learning_rate": 4.59167302729695e-07,
+      "loss": 1.0094,
+      "step": 7664
+    },
+    {
+      "epoch": 0.8061314367596987,
+      "grad_norm": 1.9870486254495214,
+      "learning_rate": 4.5868524029892616e-07,
+      "loss": 0.9653,
+      "step": 7665
+    },
+    {
+      "epoch": 0.806236607201546,
+      "grad_norm": 2.631029480979719,
+      "learning_rate": 4.582034054906098e-07,
+      "loss": 1.0113,
+      "step": 7666
+    },
+    {
+      "epoch": 0.8063417776433933,
+      "grad_norm": 1.9761375577657758,
+      "learning_rate": 4.5772179835847584e-07,
+      "loss": 0.9523,
+      "step": 7667
+    },
+    {
+      "epoch": 0.8064469480852406,
+      "grad_norm": 2.7997469746623267,
+      "learning_rate": 4.5724041895622546e-07,
+      "loss": 0.9985,
+      "step": 7668
+    },
+    {
+      "epoch": 0.806552118527088,
+      "grad_norm": 2.581890446856423,
+      "learning_rate": 4.567592673375371e-07,
+      "loss": 0.9612,
+      "step": 7669
+    },
+    {
+      "epoch": 0.8066572889689353,
+      "grad_norm": 2.1149257021470946,
+      "learning_rate": 4.56278343556063e-07,
+      "loss": 0.9749,
+      "step": 7670
+    },
+    {
+      "epoch": 0.8067624594107826,
+      "grad_norm": 2.547543034627653,
+      "learning_rate": 4.5579764766543023e-07,
+      "loss": 1.0053,
+      "step": 7671
+    },
+    {
+      "epoch": 0.8068676298526299,
+      "grad_norm": 2.369288975718212,
+      "learning_rate": 4.5531717971923904e-07,
+      "loss": 0.9468,
+      "step": 7672
+    },
+    {
+      "epoch": 0.8069728002944773,
+      "grad_norm": 1.9398001873861406,
+      "learning_rate": 4.5483693977106614e-07,
+      "loss": 0.9719,
+      "step": 7673
+    },
+    {
+      "epoch": 0.8070779707363246,
+      "grad_norm": 1.7047845543366074,
+      "learning_rate": 4.543569278744625e-07,
+      "loss": 0.9695,
+      "step": 7674
+    },
+    {
+      "epoch": 0.8071831411781719,
+      "grad_norm": 2.311255391335769,
+      "learning_rate": 4.5387714408295174e-07,
+      "loss": 1.021,
+      "step": 7675
+    },
+    {
+      "epoch": 0.8072883116200192,
+      "grad_norm": 2.5665068402737297,
+      "learning_rate": 4.5339758845003497e-07,
+      "loss": 0.9434,
+      "step": 7676
+    },
+    {
+      "epoch": 0.8073934820618666,
+      "grad_norm": 1.987419007348485,
+      "learning_rate": 4.529182610291849e-07,
+      "loss": 1.0189,
+      "step": 7677
+    },
+    {
+      "epoch": 0.8074986525037138,
+      "grad_norm": 2.7148221725671258,
+      "learning_rate": 4.5243916187385104e-07,
+      "loss": 1.0317,
+      "step": 7678
+    },
+    {
+      "epoch": 0.8076038229455611,
+      "grad_norm": 1.9771244563101253,
+      "learning_rate": 4.5196029103745736e-07,
+      "loss": 0.9574,
+      "step": 7679
+    },
+    {
+      "epoch": 0.8077089933874084,
+      "grad_norm": 2.5643391230827577,
+      "learning_rate": 4.514816485734003e-07,
+      "loss": 0.9949,
+      "step": 7680
+    },
+    {
+      "epoch": 0.8078141638292557,
+      "grad_norm": 2.6365658218139947,
+      "learning_rate": 4.5100323453505344e-07,
+      "loss": 0.991,
+      "step": 7681
+    },
+    {
+      "epoch": 0.8079193342711031,
+      "grad_norm": 2.0213099184862746,
+      "learning_rate": 4.5052504897576177e-07,
+      "loss": 1.0065,
+      "step": 7682
+    },
+    {
+      "epoch": 0.8080245047129504,
+      "grad_norm": 2.430948178647297,
+      "learning_rate": 4.500470919488495e-07,
+      "loss": 0.9754,
+      "step": 7683
+    },
+    {
+      "epoch": 0.8081296751547977,
+      "grad_norm": 2.432965011287217,
+      "learning_rate": 4.495693635076101e-07,
+      "loss": 0.9688,
+      "step": 7684
+    },
+    {
+      "epoch": 0.808234845596645,
+      "grad_norm": 1.9832928166525534,
+      "learning_rate": 4.49091863705316e-07,
+      "loss": 0.967,
+      "step": 7685
+    },
+    {
+      "epoch": 0.8083400160384924,
+      "grad_norm": 2.718332781399688,
+      "learning_rate": 4.4861459259521013e-07,
+      "loss": 0.9604,
+      "step": 7686
+    },
+    {
+      "epoch": 0.8084451864803397,
+      "grad_norm": 2.5869970801279707,
+      "learning_rate": 4.4813755023051323e-07,
+      "loss": 1.0265,
+      "step": 7687
+    },
+    {
+      "epoch": 0.808550356922187,
+      "grad_norm": 2.249567127749935,
+      "learning_rate": 4.476607366644192e-07,
+      "loss": 1.0034,
+      "step": 7688
+    },
+    {
+      "epoch": 0.8086555273640343,
+      "grad_norm": 1.8946857360784728,
+      "learning_rate": 4.4718415195009577e-07,
+      "loss": 1.009,
+      "step": 7689
+    },
+    {
+      "epoch": 0.8087606978058817,
+      "grad_norm": 2.775009327722139,
+      "learning_rate": 4.4670779614068683e-07,
+      "loss": 0.9728,
+      "step": 7690
+    },
+    {
+      "epoch": 0.808865868247729,
+      "grad_norm": 2.3078323532076626,
+      "learning_rate": 4.4623166928930846e-07,
+      "loss": 0.9838,
+      "step": 7691
+    },
+    {
+      "epoch": 0.8089710386895763,
+      "grad_norm": 2.454669259509621,
+      "learning_rate": 4.457557714490532e-07,
+      "loss": 1.0128,
+      "step": 7692
+    },
+    {
+      "epoch": 0.8090762091314236,
+      "grad_norm": 2.256343596048476,
+      "learning_rate": 4.4528010267298803e-07,
+      "loss": 1.0084,
+      "step": 7693
+    },
+    {
+      "epoch": 0.809181379573271,
+      "grad_norm": 2.0591184791276738,
+      "learning_rate": 4.4480466301415253e-07,
+      "loss": 0.9828,
+      "step": 7694
+    },
+    {
+      "epoch": 0.8092865500151183,
+      "grad_norm": 2.8039610013628145,
+      "learning_rate": 4.4432945252556285e-07,
+      "loss": 1.0063,
+      "step": 7695
+    },
+    {
+      "epoch": 0.8093917204569656,
+      "grad_norm": 1.8867274438204686,
+      "learning_rate": 4.4385447126020695e-07,
+      "loss": 0.979,
+      "step": 7696
+    },
+    {
+      "epoch": 0.8094968908988129,
+      "grad_norm": 2.0342847166171265,
+      "learning_rate": 4.433797192710515e-07,
+      "loss": 0.959,
+      "step": 7697
+    },
+    {
+      "epoch": 0.8096020613406603,
+      "grad_norm": 2.143958088883331,
+      "learning_rate": 4.4290519661103313e-07,
+      "loss": 0.9682,
+      "step": 7698
+    },
+    {
+      "epoch": 0.8097072317825075,
+      "grad_norm": 2.307795459628221,
+      "learning_rate": 4.424309033330662e-07,
+      "loss": 0.9781,
+      "step": 7699
+    },
+    {
+      "epoch": 0.8098124022243548,
+      "grad_norm": 2.6737394534107755,
+      "learning_rate": 4.4195683949003645e-07,
+      "loss": 0.9879,
+      "step": 7700
+    },
+    {
+      "epoch": 0.8099175726662021,
+      "grad_norm": 2.468361380665413,
+      "learning_rate": 4.414830051348068e-07,
+      "loss": 0.9823,
+      "step": 7701
+    },
+    {
+      "epoch": 0.8100227431080494,
+      "grad_norm": 2.464813507317941,
+      "learning_rate": 4.410094003202134e-07,
+      "loss": 0.9437,
+      "step": 7702
+    },
+    {
+      "epoch": 0.8101279135498968,
+      "grad_norm": 2.439270737619183,
+      "learning_rate": 4.4053602509906647e-07,
+      "loss": 0.9409,
+      "step": 7703
+    },
+    {
+      "epoch": 0.8102330839917441,
+      "grad_norm": 2.2253177582591266,
+      "learning_rate": 4.4006287952415156e-07,
+      "loss": 1.0017,
+      "step": 7704
+    },
+    {
+      "epoch": 0.8103382544335914,
+      "grad_norm": 1.8670735694432252,
+      "learning_rate": 4.395899636482265e-07,
+      "loss": 0.9638,
+      "step": 7705
+    },
+    {
+      "epoch": 0.8104434248754387,
+      "grad_norm": 2.537791466303488,
+      "learning_rate": 4.391172775240277e-07,
+      "loss": 0.9782,
+      "step": 7706
+    },
+    {
+      "epoch": 0.8105485953172861,
+      "grad_norm": 2.2360740386296953,
+      "learning_rate": 4.386448212042613e-07,
+      "loss": 0.9666,
+      "step": 7707
+    },
+    {
+      "epoch": 0.8106537657591334,
+      "grad_norm": 1.9347825652393957,
+      "learning_rate": 4.38172594741611e-07,
+      "loss": 1.0115,
+      "step": 7708
+    },
+    {
+      "epoch": 0.8107589362009807,
+      "grad_norm": 2.644131866061192,
+      "learning_rate": 4.3770059818873335e-07,
+      "loss": 0.9877,
+      "step": 7709
+    },
+    {
+      "epoch": 0.810864106642828,
+      "grad_norm": 2.7175911960332906,
+      "learning_rate": 4.372288315982584e-07,
+      "loss": 0.9678,
+      "step": 7710
+    },
+    {
+      "epoch": 0.8109692770846754,
+      "grad_norm": 2.336165314669402,
+      "learning_rate": 4.3675729502279385e-07,
+      "loss": 1.0202,
+      "step": 7711
+    },
+    {
+      "epoch": 0.8110744475265227,
+      "grad_norm": 2.5282003663888952,
+      "learning_rate": 4.3628598851491816e-07,
+      "loss": 0.9948,
+      "step": 7712
+    },
+    {
+      "epoch": 0.81117961796837,
+      "grad_norm": 2.6721797231956863,
+      "learning_rate": 4.358149121271871e-07,
+      "loss": 1.0091,
+      "step": 7713
+    },
+    {
+      "epoch": 0.8112847884102173,
+      "grad_norm": 2.155263569819944,
+      "learning_rate": 4.3534406591212775e-07,
+      "loss": 0.9918,
+      "step": 7714
+    },
+    {
+      "epoch": 0.8113899588520647,
+      "grad_norm": 2.878491927577863,
+      "learning_rate": 4.3487344992224375e-07,
+      "loss": 0.9162,
+      "step": 7715
+    },
+    {
+      "epoch": 0.811495129293912,
+      "grad_norm": 2.6975923680738703,
+      "learning_rate": 4.344030642100133e-07,
+      "loss": 1.0168,
+      "step": 7716
+    },
+    {
+      "epoch": 0.8116002997357593,
+      "grad_norm": 2.24948546017479,
+      "learning_rate": 4.3393290882788654e-07,
+      "loss": 0.979,
+      "step": 7717
+    },
+    {
+      "epoch": 0.8117054701776066,
+      "grad_norm": 2.2782812764649836,
+      "learning_rate": 4.3346298382829085e-07,
+      "loss": 0.9802,
+      "step": 7718
+    },
+    {
+      "epoch": 0.8118106406194538,
+      "grad_norm": 2.369431861546599,
+      "learning_rate": 4.329932892636249e-07,
+      "loss": 0.9852,
+      "step": 7719
+    },
+    {
+      "epoch": 0.8119158110613012,
+      "grad_norm": 2.819628728763317,
+      "learning_rate": 4.325238251862651e-07,
+      "loss": 1.007,
+      "step": 7720
+    },
+    {
+      "epoch": 0.8120209815031485,
+      "grad_norm": 2.480164594518704,
+      "learning_rate": 4.3205459164855914e-07,
+      "loss": 0.9727,
+      "step": 7721
+    },
+    {
+      "epoch": 0.8121261519449958,
+      "grad_norm": 2.3940648557433524,
+      "learning_rate": 4.315855887028314e-07,
+      "loss": 0.9746,
+      "step": 7722
+    },
+    {
+      "epoch": 0.8122313223868431,
+      "grad_norm": 3.393286161834988,
+      "learning_rate": 4.3111681640137763e-07,
+      "loss": 0.9979,
+      "step": 7723
+    },
+    {
+      "epoch": 0.8123364928286905,
+      "grad_norm": 2.154801256417696,
+      "learning_rate": 4.306482747964705e-07,
+      "loss": 1.0169,
+      "step": 7724
+    },
+    {
+      "epoch": 0.8124416632705378,
+      "grad_norm": 2.6404030152506657,
+      "learning_rate": 4.3017996394035674e-07,
+      "loss": 0.9906,
+      "step": 7725
+    },
+    {
+      "epoch": 0.8125468337123851,
+      "grad_norm": 2.921067892801514,
+      "learning_rate": 4.2971188388525515e-07,
+      "loss": 1.0098,
+      "step": 7726
+    },
+    {
+      "epoch": 0.8126520041542324,
+      "grad_norm": 2.151845511245677,
+      "learning_rate": 4.292440346833618e-07,
+      "loss": 0.9851,
+      "step": 7727
+    },
+    {
+      "epoch": 0.8127571745960798,
+      "grad_norm": 2.1131586999724665,
+      "learning_rate": 4.287764163868441e-07,
+      "loss": 0.9912,
+      "step": 7728
+    },
+    {
+      "epoch": 0.8128623450379271,
+      "grad_norm": 2.540790376974774,
+      "learning_rate": 4.283090290478459e-07,
+      "loss": 0.9808,
+      "step": 7729
+    },
+    {
+      "epoch": 0.8129675154797744,
+      "grad_norm": 2.5175938968858476,
+      "learning_rate": 4.278418727184849e-07,
+      "loss": 0.9565,
+      "step": 7730
+    },
+    {
+      "epoch": 0.8130726859216217,
+      "grad_norm": 3.2981499359448265,
+      "learning_rate": 4.273749474508515e-07,
+      "loss": 0.9686,
+      "step": 7731
+    },
+    {
+      "epoch": 0.8131778563634691,
+      "grad_norm": 2.9415949773488514,
+      "learning_rate": 4.2690825329701313e-07,
+      "loss": 0.9655,
+      "step": 7732
+    },
+    {
+      "epoch": 0.8132830268053164,
+      "grad_norm": 2.7939627395608104,
+      "learning_rate": 4.2644179030900734e-07,
+      "loss": 0.9893,
+      "step": 7733
+    },
+    {
+      "epoch": 0.8133881972471637,
+      "grad_norm": 2.668181131671682,
+      "learning_rate": 4.2597555853885117e-07,
+      "loss": 1.0038,
+      "step": 7734
+    },
+    {
+      "epoch": 0.813493367689011,
+      "grad_norm": 2.065827171733609,
+      "learning_rate": 4.2550955803853125e-07,
+      "loss": 0.9932,
+      "step": 7735
+    },
+    {
+      "epoch": 0.8135985381308584,
+      "grad_norm": 2.4982865299615513,
+      "learning_rate": 4.250437888600115e-07,
+      "loss": 0.9941,
+      "step": 7736
+    },
+    {
+      "epoch": 0.8137037085727057,
+      "grad_norm": 2.9611235171473536,
+      "learning_rate": 4.2457825105522736e-07,
+      "loss": 0.9584,
+      "step": 7737
+    },
+    {
+      "epoch": 0.813808879014553,
+      "grad_norm": 1.856983927521474,
+      "learning_rate": 4.241129446760905e-07,
+      "loss": 0.9412,
+      "step": 7738
+    },
+    {
+      "epoch": 0.8139140494564002,
+      "grad_norm": 2.298057826980122,
+      "learning_rate": 4.2364786977448723e-07,
+      "loss": 0.977,
+      "step": 7739
+    },
+    {
+      "epoch": 0.8140192198982475,
+      "grad_norm": 2.857504158227023,
+      "learning_rate": 4.231830264022757e-07,
+      "loss": 0.9809,
+      "step": 7740
+    },
+    {
+      "epoch": 0.8141243903400949,
+      "grad_norm": 1.9125168283750302,
+      "learning_rate": 4.227184146112903e-07,
+      "loss": 0.9547,
+      "step": 7741
+    },
+    {
+      "epoch": 0.8142295607819422,
+      "grad_norm": 2.3292724206858115,
+      "learning_rate": 4.222540344533374e-07,
+      "loss": 0.9752,
+      "step": 7742
+    },
+    {
+      "epoch": 0.8143347312237895,
+      "grad_norm": 2.8514614953686768,
+      "learning_rate": 4.2178988598020164e-07,
+      "loss": 0.9877,
+      "step": 7743
+    },
+    {
+      "epoch": 0.8144399016656368,
+      "grad_norm": 2.3444264375718555,
+      "learning_rate": 4.2132596924363666e-07,
+      "loss": 0.9912,
+      "step": 7744
+    },
+    {
+      "epoch": 0.8145450721074842,
+      "grad_norm": 2.51703936814917,
+      "learning_rate": 4.208622842953747e-07,
+      "loss": 0.9761,
+      "step": 7745
+    },
+    {
+      "epoch": 0.8146502425493315,
+      "grad_norm": 2.401729234637294,
+      "learning_rate": 4.2039883118711935e-07,
+      "loss": 0.9613,
+      "step": 7746
+    },
+    {
+      "epoch": 0.8147554129911788,
+      "grad_norm": 2.10443613373023,
+      "learning_rate": 4.1993560997054787e-07,
+      "loss": 0.9799,
+      "step": 7747
+    },
+    {
+      "epoch": 0.8148605834330261,
+      "grad_norm": 2.7838525850839613,
+      "learning_rate": 4.194726206973157e-07,
+      "loss": 0.958,
+      "step": 7748
+    },
+    {
+      "epoch": 0.8149657538748735,
+      "grad_norm": 2.471550286019982,
+      "learning_rate": 4.1900986341904787e-07,
+      "loss": 0.957,
+      "step": 7749
+    },
+    {
+      "epoch": 0.8150709243167208,
+      "grad_norm": 2.6262073326290087,
+      "learning_rate": 4.1854733818734646e-07,
+      "loss": 0.9636,
+      "step": 7750
+    },
+    {
+      "epoch": 0.8151760947585681,
+      "grad_norm": 1.7860526138744137,
+      "learning_rate": 4.180850450537857e-07,
+      "loss": 0.9505,
+      "step": 7751
+    },
+    {
+      "epoch": 0.8152812652004154,
+      "grad_norm": 2.3407264097231866,
+      "learning_rate": 4.1762298406991503e-07,
+      "loss": 0.9565,
+      "step": 7752
+    },
+    {
+      "epoch": 0.8153864356422628,
+      "grad_norm": 2.150623638303471,
+      "learning_rate": 4.1716115528725896e-07,
+      "loss": 0.9697,
+      "step": 7753
+    },
+    {
+      "epoch": 0.8154916060841101,
+      "grad_norm": 3.2525410184474173,
+      "learning_rate": 4.166995587573133e-07,
+      "loss": 1.0229,
+      "step": 7754
+    },
+    {
+      "epoch": 0.8155967765259574,
+      "grad_norm": 2.1671249003463577,
+      "learning_rate": 4.1623819453155133e-07,
+      "loss": 0.9757,
+      "step": 7755
+    },
+    {
+      "epoch": 0.8157019469678047,
+      "grad_norm": 2.097060159238962,
+      "learning_rate": 4.157770626614166e-07,
+      "loss": 0.9595,
+      "step": 7756
+    },
+    {
+      "epoch": 0.8158071174096521,
+      "grad_norm": 2.5745070957454947,
+      "learning_rate": 4.153161631983313e-07,
+      "loss": 0.9547,
+      "step": 7757
+    },
+    {
+      "epoch": 0.8159122878514994,
+      "grad_norm": 2.6795685146216854,
+      "learning_rate": 4.1485549619368765e-07,
+      "loss": 0.9458,
+      "step": 7758
+    },
+    {
+      "epoch": 0.8160174582933467,
+      "grad_norm": 2.462379782717814,
+      "learning_rate": 4.1439506169885484e-07,
+      "loss": 0.9906,
+      "step": 7759
+    },
+    {
+      "epoch": 0.8161226287351939,
+      "grad_norm": 2.4584203736359918,
+      "learning_rate": 4.139348597651735e-07,
+      "loss": 0.9678,
+      "step": 7760
+    },
+    {
+      "epoch": 0.8162277991770412,
+      "grad_norm": 2.373737093242701,
+      "learning_rate": 4.134748904439603e-07,
+      "loss": 0.9874,
+      "step": 7761
+    },
+    {
+      "epoch": 0.8163329696188886,
+      "grad_norm": 2.682893399793693,
+      "learning_rate": 4.1301515378650654e-07,
+      "loss": 0.9744,
+      "step": 7762
+    },
+    {
+      "epoch": 0.8164381400607359,
+      "grad_norm": 2.6158066777998443,
+      "learning_rate": 4.1255564984407445e-07,
+      "loss": 0.9914,
+      "step": 7763
+    },
+    {
+      "epoch": 0.8165433105025832,
+      "grad_norm": 2.320463099049456,
+      "learning_rate": 4.120963786679039e-07,
+      "loss": 0.9913,
+      "step": 7764
+    },
+    {
+      "epoch": 0.8166484809444305,
+      "grad_norm": 2.584830907970578,
+      "learning_rate": 4.116373403092058e-07,
+      "loss": 0.9571,
+      "step": 7765
+    },
+    {
+      "epoch": 0.8167536513862779,
+      "grad_norm": 2.3221349508899425,
+      "learning_rate": 4.111785348191674e-07,
+      "loss": 0.9629,
+      "step": 7766
+    },
+    {
+      "epoch": 0.8168588218281252,
+      "grad_norm": 2.3555780902934207,
+      "learning_rate": 4.1071996224894937e-07,
+      "loss": 0.9926,
+      "step": 7767
+    },
+    {
+      "epoch": 0.8169639922699725,
+      "grad_norm": 2.4447457069951763,
+      "learning_rate": 4.1026162264968526e-07,
+      "loss": 0.9714,
+      "step": 7768
+    },
+    {
+      "epoch": 0.8170691627118198,
+      "grad_norm": 2.5930949291002143,
+      "learning_rate": 4.098035160724842e-07,
+      "loss": 0.9894,
+      "step": 7769
+    },
+    {
+      "epoch": 0.8171743331536672,
+      "grad_norm": 2.4978191323631975,
+      "learning_rate": 4.093456425684275e-07,
+      "loss": 0.9675,
+      "step": 7770
+    },
+    {
+      "epoch": 0.8172795035955145,
+      "grad_norm": 1.763876023991041,
+      "learning_rate": 4.0888800218857326e-07,
+      "loss": 0.9674,
+      "step": 7771
+    },
+    {
+      "epoch": 0.8173846740373618,
+      "grad_norm": 2.562886748389969,
+      "learning_rate": 4.0843059498395063e-07,
+      "loss": 0.9959,
+      "step": 7772
+    },
+    {
+      "epoch": 0.8174898444792091,
+      "grad_norm": 3.542366051507965,
+      "learning_rate": 4.0797342100556526e-07,
+      "loss": 0.9939,
+      "step": 7773
+    },
+    {
+      "epoch": 0.8175950149210565,
+      "grad_norm": 2.978184338970861,
+      "learning_rate": 4.075164803043941e-07,
+      "loss": 1.0054,
+      "step": 7774
+    },
+    {
+      "epoch": 0.8177001853629038,
+      "grad_norm": 2.0685694903463263,
+      "learning_rate": 4.070597729313905e-07,
+      "loss": 0.9819,
+      "step": 7775
+    },
+    {
+      "epoch": 0.8178053558047511,
+      "grad_norm": 2.4513268899881755,
+      "learning_rate": 4.066032989374813e-07,
+      "loss": 0.9578,
+      "step": 7776
+    },
+    {
+      "epoch": 0.8179105262465984,
+      "grad_norm": 2.789567940068473,
+      "learning_rate": 4.061470583735655e-07,
+      "loss": 0.9826,
+      "step": 7777
+    },
+    {
+      "epoch": 0.8180156966884458,
+      "grad_norm": 2.2071024352563495,
+      "learning_rate": 4.056910512905193e-07,
+      "loss": 0.9408,
+      "step": 7778
+    },
+    {
+      "epoch": 0.8181208671302931,
+      "grad_norm": 2.462265054582973,
+      "learning_rate": 4.052352777391885e-07,
+      "loss": 0.9993,
+      "step": 7779
+    },
+    {
+      "epoch": 0.8182260375721403,
+      "grad_norm": 2.1507474468996537,
+      "learning_rate": 4.0477973777039854e-07,
+      "loss": 0.9428,
+      "step": 7780
+    },
+    {
+      "epoch": 0.8183312080139876,
+      "grad_norm": 1.9943322945072193,
+      "learning_rate": 4.0432443143494373e-07,
+      "loss": 0.9267,
+      "step": 7781
+    },
+    {
+      "epoch": 0.818436378455835,
+      "grad_norm": 2.4651273882986757,
+      "learning_rate": 4.03869358783594e-07,
+      "loss": 0.9777,
+      "step": 7782
+    },
+    {
+      "epoch": 0.8185415488976823,
+      "grad_norm": 2.3629386718998173,
+      "learning_rate": 4.0341451986709494e-07,
+      "loss": 1.0024,
+      "step": 7783
+    },
+    {
+      "epoch": 0.8186467193395296,
+      "grad_norm": 1.98682520235926,
+      "learning_rate": 4.029599147361624e-07,
+      "loss": 1.0093,
+      "step": 7784
+    },
+    {
+      "epoch": 0.8187518897813769,
+      "grad_norm": 2.6681064116513826,
+      "learning_rate": 4.0250554344149094e-07,
+      "loss": 0.998,
+      "step": 7785
+    },
+    {
+      "epoch": 0.8188570602232242,
+      "grad_norm": 2.0258952155927568,
+      "learning_rate": 4.020514060337447e-07,
+      "loss": 0.9847,
+      "step": 7786
+    },
+    {
+      "epoch": 0.8189622306650716,
+      "grad_norm": 2.2616741774571074,
+      "learning_rate": 4.015975025635646e-07,
+      "loss": 0.9739,
+      "step": 7787
+    },
+    {
+      "epoch": 0.8190674011069189,
+      "grad_norm": 2.946241928430217,
+      "learning_rate": 4.011438330815634e-07,
+      "loss": 0.9951,
+      "step": 7788
+    },
+    {
+      "epoch": 0.8191725715487662,
+      "grad_norm": 2.6286525561939342,
+      "learning_rate": 4.0069039763832946e-07,
+      "loss": 0.9628,
+      "step": 7789
+    },
+    {
+      "epoch": 0.8192777419906135,
+      "grad_norm": 2.3544960061523943,
+      "learning_rate": 4.0023719628442454e-07,
+      "loss": 0.9692,
+      "step": 7790
+    },
+    {
+      "epoch": 0.8193829124324609,
+      "grad_norm": 2.447301762694223,
+      "learning_rate": 3.997842290703835e-07,
+      "loss": 0.988,
+      "step": 7791
+    },
+    {
+      "epoch": 0.8194880828743082,
+      "grad_norm": 2.0937992451075864,
+      "learning_rate": 3.9933149604671624e-07,
+      "loss": 0.9527,
+      "step": 7792
+    },
+    {
+      "epoch": 0.8195932533161555,
+      "grad_norm": 2.5017080295162653,
+      "learning_rate": 3.9887899726390455e-07,
+      "loss": 0.9585,
+      "step": 7793
+    },
+    {
+      "epoch": 0.8196984237580028,
+      "grad_norm": 2.1071679855154173,
+      "learning_rate": 3.9842673277240813e-07,
+      "loss": 0.9637,
+      "step": 7794
+    },
+    {
+      "epoch": 0.8198035941998502,
+      "grad_norm": 2.6444030746408793,
+      "learning_rate": 3.97974702622656e-07,
+      "loss": 0.986,
+      "step": 7795
+    },
+    {
+      "epoch": 0.8199087646416975,
+      "grad_norm": 2.2415822506715366,
+      "learning_rate": 3.975229068650541e-07,
+      "loss": 0.9998,
+      "step": 7796
+    },
+    {
+      "epoch": 0.8200139350835448,
+      "grad_norm": 2.8744646198567714,
+      "learning_rate": 3.970713455499803e-07,
+      "loss": 0.9492,
+      "step": 7797
+    },
+    {
+      "epoch": 0.8201191055253921,
+      "grad_norm": 2.0835911020378837,
+      "learning_rate": 3.9662001872778746e-07,
+      "loss": 0.9694,
+      "step": 7798
+    },
+    {
+      "epoch": 0.8202242759672395,
+      "grad_norm": 2.0735661240104832,
+      "learning_rate": 3.961689264488025e-07,
+      "loss": 0.9942,
+      "step": 7799
+    },
+    {
+      "epoch": 0.8203294464090867,
+      "grad_norm": 2.132869834888772,
+      "learning_rate": 3.9571806876332527e-07,
+      "loss": 0.9882,
+      "step": 7800
+    },
+    {
+      "epoch": 0.820434616850934,
+      "grad_norm": 3.1197931626041218,
+      "learning_rate": 3.9526744572163016e-07,
+      "loss": 0.9942,
+      "step": 7801
+    },
+    {
+      "epoch": 0.8205397872927813,
+      "grad_norm": 2.4680929826487352,
+      "learning_rate": 3.9481705737396457e-07,
+      "loss": 1.01,
+      "step": 7802
+    },
+    {
+      "epoch": 0.8206449577346286,
+      "grad_norm": 3.1575237316281837,
+      "learning_rate": 3.943669037705508e-07,
+      "loss": 1.0134,
+      "step": 7803
+    },
+    {
+      "epoch": 0.820750128176476,
+      "grad_norm": 2.5835775583665943,
+      "learning_rate": 3.939169849615851e-07,
+      "loss": 1.0241,
+      "step": 7804
+    },
+    {
+      "epoch": 0.8208552986183233,
+      "grad_norm": 2.231742487138655,
+      "learning_rate": 3.9346730099723533e-07,
+      "loss": 0.9715,
+      "step": 7805
+    },
+    {
+      "epoch": 0.8209604690601706,
+      "grad_norm": 2.6521036190376783,
+      "learning_rate": 3.930178519276462e-07,
+      "loss": 1.0162,
+      "step": 7806
+    },
+    {
+      "epoch": 0.8210656395020179,
+      "grad_norm": 2.867026937723473,
+      "learning_rate": 3.925686378029331e-07,
+      "loss": 0.9818,
+      "step": 7807
+    },
+    {
+      "epoch": 0.8211708099438653,
+      "grad_norm": 2.263917199281129,
+      "learning_rate": 3.921196586731893e-07,
+      "loss": 0.9731,
+      "step": 7808
+    },
+    {
+      "epoch": 0.8212759803857126,
+      "grad_norm": 2.3946757029431684,
+      "learning_rate": 3.916709145884773e-07,
+      "loss": 0.9887,
+      "step": 7809
+    },
+    {
+      "epoch": 0.8213811508275599,
+      "grad_norm": 2.280792937113479,
+      "learning_rate": 3.912224055988373e-07,
+      "loss": 0.9711,
+      "step": 7810
+    },
+    {
+      "epoch": 0.8214863212694072,
+      "grad_norm": 2.3686260861759547,
+      "learning_rate": 3.907741317542799e-07,
+      "loss": 0.9954,
+      "step": 7811
+    },
+    {
+      "epoch": 0.8215914917112546,
+      "grad_norm": 2.6674673001111167,
+      "learning_rate": 3.903260931047917e-07,
+      "loss": 0.9966,
+      "step": 7812
+    },
+    {
+      "epoch": 0.8216966621531019,
+      "grad_norm": 2.7598548320790517,
+      "learning_rate": 3.898782897003334e-07,
+      "loss": 0.9814,
+      "step": 7813
+    },
+    {
+      "epoch": 0.8218018325949492,
+      "grad_norm": 2.464756135941767,
+      "learning_rate": 3.894307215908372e-07,
+      "loss": 1.0448,
+      "step": 7814
+    },
+    {
+      "epoch": 0.8219070030367965,
+      "grad_norm": 1.6827932013253648,
+      "learning_rate": 3.889833888262115e-07,
+      "loss": 0.9651,
+      "step": 7815
+    },
+    {
+      "epoch": 0.8220121734786439,
+      "grad_norm": 2.3519562279418014,
+      "learning_rate": 3.8853629145633646e-07,
+      "loss": 1.0091,
+      "step": 7816
+    },
+    {
+      "epoch": 0.8221173439204912,
+      "grad_norm": 1.9766869386337151,
+      "learning_rate": 3.880894295310672e-07,
+      "loss": 0.9603,
+      "step": 7817
+    },
+    {
+      "epoch": 0.8222225143623385,
+      "grad_norm": 2.3412319875825585,
+      "learning_rate": 3.87642803100233e-07,
+      "loss": 0.987,
+      "step": 7818
+    },
+    {
+      "epoch": 0.8223276848041858,
+      "grad_norm": 2.2093606730393156,
+      "learning_rate": 3.8719641221363514e-07,
+      "loss": 0.9679,
+      "step": 7819
+    },
+    {
+      "epoch": 0.8224328552460332,
+      "grad_norm": 3.16030150590197,
+      "learning_rate": 3.8675025692105015e-07,
+      "loss": 0.9895,
+      "step": 7820
+    },
+    {
+      "epoch": 0.8225380256878804,
+      "grad_norm": 2.3349525577022585,
+      "learning_rate": 3.863043372722278e-07,
+      "loss": 1.0192,
+      "step": 7821
+    },
+    {
+      "epoch": 0.8226431961297277,
+      "grad_norm": 2.6371286764996618,
+      "learning_rate": 3.858586533168923e-07,
+      "loss": 0.96,
+      "step": 7822
+    },
+    {
+      "epoch": 0.822748366571575,
+      "grad_norm": 2.30573441792884,
+      "learning_rate": 3.8541320510473963e-07,
+      "loss": 0.9927,
+      "step": 7823
+    },
+    {
+      "epoch": 0.8228535370134223,
+      "grad_norm": 2.873334351698845,
+      "learning_rate": 3.849679926854416e-07,
+      "loss": 0.9672,
+      "step": 7824
+    },
+    {
+      "epoch": 0.8229587074552697,
+      "grad_norm": 2.4655038088603978,
+      "learning_rate": 3.845230161086422e-07,
+      "loss": 1.0114,
+      "step": 7825
+    },
+    {
+      "epoch": 0.823063877897117,
+      "grad_norm": 2.277321669840928,
+      "learning_rate": 3.840782754239597e-07,
+      "loss": 0.9895,
+      "step": 7826
+    },
+    {
+      "epoch": 0.8231690483389643,
+      "grad_norm": 2.4885441680443154,
+      "learning_rate": 3.8363377068098756e-07,
+      "loss": 1.0021,
+      "step": 7827
+    },
+    {
+      "epoch": 0.8232742187808116,
+      "grad_norm": 2.581402189491797,
+      "learning_rate": 3.831895019292897e-07,
+      "loss": 0.9841,
+      "step": 7828
+    },
+    {
+      "epoch": 0.823379389222659,
+      "grad_norm": 2.6225815215802952,
+      "learning_rate": 3.8274546921840653e-07,
+      "loss": 0.98,
+      "step": 7829
+    },
+    {
+      "epoch": 0.8234845596645063,
+      "grad_norm": 2.672257972613925,
+      "learning_rate": 3.823016725978507e-07,
+      "loss": 0.917,
+      "step": 7830
+    },
+    {
+      "epoch": 0.8235897301063536,
+      "grad_norm": 2.6941147355501207,
+      "learning_rate": 3.8185811211710997e-07,
+      "loss": 0.9917,
+      "step": 7831
+    },
+    {
+      "epoch": 0.8236949005482009,
+      "grad_norm": 2.5181042464817778,
+      "learning_rate": 3.814147878256441e-07,
+      "loss": 0.9903,
+      "step": 7832
+    },
+    {
+      "epoch": 0.8238000709900483,
+      "grad_norm": 2.982863876704575,
+      "learning_rate": 3.809716997728863e-07,
+      "loss": 0.9598,
+      "step": 7833
+    },
+    {
+      "epoch": 0.8239052414318956,
+      "grad_norm": 2.4829472094226075,
+      "learning_rate": 3.805288480082453e-07,
+      "loss": 1.0105,
+      "step": 7834
+    },
+    {
+      "epoch": 0.8240104118737429,
+      "grad_norm": 2.262790895769715,
+      "learning_rate": 3.800862325811025e-07,
+      "loss": 0.9802,
+      "step": 7835
+    },
+    {
+      "epoch": 0.8241155823155902,
+      "grad_norm": 2.3779766898281287,
+      "learning_rate": 3.796438535408134e-07,
+      "loss": 0.9596,
+      "step": 7836
+    },
+    {
+      "epoch": 0.8242207527574376,
+      "grad_norm": 2.5486165785150594,
+      "learning_rate": 3.792017109367052e-07,
+      "loss": 0.9497,
+      "step": 7837
+    },
+    {
+      "epoch": 0.8243259231992849,
+      "grad_norm": 2.2377493045016306,
+      "learning_rate": 3.787598048180821e-07,
+      "loss": 0.9521,
+      "step": 7838
+    },
+    {
+      "epoch": 0.8244310936411322,
+      "grad_norm": 2.560942533796768,
+      "learning_rate": 3.7831813523421856e-07,
+      "loss": 0.9648,
+      "step": 7839
+    },
+    {
+      "epoch": 0.8245362640829795,
+      "grad_norm": 3.155199120538444,
+      "learning_rate": 3.7787670223436464e-07,
+      "loss": 1.0093,
+      "step": 7840
+    },
+    {
+      "epoch": 0.8246414345248267,
+      "grad_norm": 2.5558594549066047,
+      "learning_rate": 3.7743550586774425e-07,
+      "loss": 0.9772,
+      "step": 7841
+    },
+    {
+      "epoch": 0.8247466049666741,
+      "grad_norm": 3.286085851170872,
+      "learning_rate": 3.7699454618355306e-07,
+      "loss": 1.0098,
+      "step": 7842
+    },
+    {
+      "epoch": 0.8248517754085214,
+      "grad_norm": 3.071919558180755,
+      "learning_rate": 3.765538232309621e-07,
+      "loss": 0.978,
+      "step": 7843
+    },
+    {
+      "epoch": 0.8249569458503687,
+      "grad_norm": 1.9611542629163186,
+      "learning_rate": 3.7611333705911526e-07,
+      "loss": 0.9318,
+      "step": 7844
+    },
+    {
+      "epoch": 0.825062116292216,
+      "grad_norm": 2.668955460154382,
+      "learning_rate": 3.756730877171308e-07,
+      "loss": 0.9825,
+      "step": 7845
+    },
+    {
+      "epoch": 0.8251672867340634,
+      "grad_norm": 2.097041197652859,
+      "learning_rate": 3.752330752540989e-07,
+      "loss": 0.951,
+      "step": 7846
+    },
+    {
+      "epoch": 0.8252724571759107,
+      "grad_norm": 2.4381455759333552,
+      "learning_rate": 3.747932997190853e-07,
+      "loss": 0.9731,
+      "step": 7847
+    },
+    {
+      "epoch": 0.825377627617758,
+      "grad_norm": 1.9389944794723009,
+      "learning_rate": 3.7435376116112753e-07,
+      "loss": 0.9945,
+      "step": 7848
+    },
+    {
+      "epoch": 0.8254827980596053,
+      "grad_norm": 2.6544698505058437,
+      "learning_rate": 3.7391445962923824e-07,
+      "loss": 0.9866,
+      "step": 7849
+    },
+    {
+      "epoch": 0.8255879685014527,
+      "grad_norm": 2.205561586287741,
+      "learning_rate": 3.734753951724029e-07,
+      "loss": 0.9731,
+      "step": 7850
+    },
+    {
+      "epoch": 0.8256931389433,
+      "grad_norm": 2.4950986903703165,
+      "learning_rate": 3.7303656783958e-07,
+      "loss": 0.9696,
+      "step": 7851
+    },
+    {
+      "epoch": 0.8257983093851473,
+      "grad_norm": 2.824630018964161,
+      "learning_rate": 3.7259797767970255e-07,
+      "loss": 1.0108,
+      "step": 7852
+    },
+    {
+      "epoch": 0.8259034798269946,
+      "grad_norm": 1.5885466979887306,
+      "learning_rate": 3.7215962474167754e-07,
+      "loss": 0.9861,
+      "step": 7853
+    },
+    {
+      "epoch": 0.826008650268842,
+      "grad_norm": 2.407273379719963,
+      "learning_rate": 3.717215090743834e-07,
+      "loss": 0.9186,
+      "step": 7854
+    },
+    {
+      "epoch": 0.8261138207106893,
+      "grad_norm": 2.108109969570331,
+      "learning_rate": 3.7128363072667455e-07,
+      "loss": 0.9826,
+      "step": 7855
+    },
+    {
+      "epoch": 0.8262189911525366,
+      "grad_norm": 2.125810712110948,
+      "learning_rate": 3.708459897473768e-07,
+      "loss": 0.9589,
+      "step": 7856
+    },
+    {
+      "epoch": 0.8263241615943839,
+      "grad_norm": 2.276739627965222,
+      "learning_rate": 3.70408586185291e-07,
+      "loss": 0.9668,
+      "step": 7857
+    },
+    {
+      "epoch": 0.8264293320362313,
+      "grad_norm": 1.9368576617583315,
+      "learning_rate": 3.69971420089191e-07,
+      "loss": 1.0004,
+      "step": 7858
+    },
+    {
+      "epoch": 0.8265345024780786,
+      "grad_norm": 2.122279799459209,
+      "learning_rate": 3.69534491507825e-07,
+      "loss": 0.9629,
+      "step": 7859
+    },
+    {
+      "epoch": 0.8266396729199259,
+      "grad_norm": 3.094014832454307,
+      "learning_rate": 3.6909780048991264e-07,
+      "loss": 0.9908,
+      "step": 7860
+    },
+    {
+      "epoch": 0.8267448433617731,
+      "grad_norm": 2.759129120062527,
+      "learning_rate": 3.6866134708414964e-07,
+      "loss": 0.98,
+      "step": 7861
+    },
+    {
+      "epoch": 0.8268500138036204,
+      "grad_norm": 2.276400624904361,
+      "learning_rate": 3.682251313392024e-07,
+      "loss": 1.0001,
+      "step": 7862
+    },
+    {
+      "epoch": 0.8269551842454678,
+      "grad_norm": 1.7834147878732247,
+      "learning_rate": 3.677891533037134e-07,
+      "loss": 0.9976,
+      "step": 7863
+    },
+    {
+      "epoch": 0.8270603546873151,
+      "grad_norm": 2.167298852018403,
+      "learning_rate": 3.6735341302629816e-07,
+      "loss": 0.991,
+      "step": 7864
+    },
+    {
+      "epoch": 0.8271655251291624,
+      "grad_norm": 2.480960772681218,
+      "learning_rate": 3.6691791055554344e-07,
+      "loss": 0.9629,
+      "step": 7865
+    },
+    {
+      "epoch": 0.8272706955710097,
+      "grad_norm": 2.448579622721506,
+      "learning_rate": 3.664826459400123e-07,
+      "loss": 0.9664,
+      "step": 7866
+    },
+    {
+      "epoch": 0.8273758660128571,
+      "grad_norm": 2.2157167037720105,
+      "learning_rate": 3.660476192282403e-07,
+      "loss": 0.9307,
+      "step": 7867
+    },
+    {
+      "epoch": 0.8274810364547044,
+      "grad_norm": 2.4536996567992246,
+      "learning_rate": 3.656128304687354e-07,
+      "loss": 0.9936,
+      "step": 7868
+    },
+    {
+      "epoch": 0.8275862068965517,
+      "grad_norm": 3.368456460878063,
+      "learning_rate": 3.6517827970998095e-07,
+      "loss": 0.9781,
+      "step": 7869
+    },
+    {
+      "epoch": 0.827691377338399,
+      "grad_norm": 1.71414901522867,
+      "learning_rate": 3.647439670004316e-07,
+      "loss": 0.9807,
+      "step": 7870
+    },
+    {
+      "epoch": 0.8277965477802464,
+      "grad_norm": 2.97285183356878,
+      "learning_rate": 3.6430989238851717e-07,
+      "loss": 0.9555,
+      "step": 7871
+    },
+    {
+      "epoch": 0.8279017182220937,
+      "grad_norm": 2.96702239531014,
+      "learning_rate": 3.6387605592264036e-07,
+      "loss": 0.9843,
+      "step": 7872
+    },
+    {
+      "epoch": 0.828006888663941,
+      "grad_norm": 2.166596883395088,
+      "learning_rate": 3.63442457651178e-07,
+      "loss": 0.9585,
+      "step": 7873
+    },
+    {
+      "epoch": 0.8281120591057883,
+      "grad_norm": 2.5899405106290123,
+      "learning_rate": 3.6300909762247813e-07,
+      "loss": 0.9518,
+      "step": 7874
+    },
+    {
+      "epoch": 0.8282172295476357,
+      "grad_norm": 2.2108133293110157,
+      "learning_rate": 3.625759758848649e-07,
+      "loss": 0.9702,
+      "step": 7875
+    },
+    {
+      "epoch": 0.828322399989483,
+      "grad_norm": 1.7475000460222763,
+      "learning_rate": 3.621430924866348e-07,
+      "loss": 0.9557,
+      "step": 7876
+    },
+    {
+      "epoch": 0.8284275704313303,
+      "grad_norm": 2.9751242328091165,
+      "learning_rate": 3.617104474760569e-07,
+      "loss": 0.9627,
+      "step": 7877
+    },
+    {
+      "epoch": 0.8285327408731776,
+      "grad_norm": 2.2824598602628066,
+      "learning_rate": 3.612780409013755e-07,
+      "loss": 0.9557,
+      "step": 7878
+    },
+    {
+      "epoch": 0.828637911315025,
+      "grad_norm": 2.560107410245565,
+      "learning_rate": 3.6084587281080605e-07,
+      "loss": 0.9479,
+      "step": 7879
+    },
+    {
+      "epoch": 0.8287430817568723,
+      "grad_norm": 2.6404332069080954,
+      "learning_rate": 3.604139432525394e-07,
+      "loss": 1.0084,
+      "step": 7880
+    },
+    {
+      "epoch": 0.8288482521987196,
+      "grad_norm": 3.5468570151691994,
+      "learning_rate": 3.59982252274739e-07,
+      "loss": 0.9756,
+      "step": 7881
+    },
+    {
+      "epoch": 0.8289534226405668,
+      "grad_norm": 3.1714421863078326,
+      "learning_rate": 3.595507999255421e-07,
+      "loss": 0.9722,
+      "step": 7882
+    },
+    {
+      "epoch": 0.8290585930824141,
+      "grad_norm": 1.7670191648623403,
+      "learning_rate": 3.591195862530586e-07,
+      "loss": 0.9664,
+      "step": 7883
+    },
+    {
+      "epoch": 0.8291637635242615,
+      "grad_norm": 2.8058619365312394,
+      "learning_rate": 3.586886113053717e-07,
+      "loss": 1.0019,
+      "step": 7884
+    },
+    {
+      "epoch": 0.8292689339661088,
+      "grad_norm": 2.274264569970648,
+      "learning_rate": 3.5825787513053905e-07,
+      "loss": 1.001,
+      "step": 7885
+    },
+    {
+      "epoch": 0.8293741044079561,
+      "grad_norm": 2.1691351608255953,
+      "learning_rate": 3.5782737777659086e-07,
+      "loss": 0.9753,
+      "step": 7886
+    },
+    {
+      "epoch": 0.8294792748498034,
+      "grad_norm": 3.0588797917567683,
+      "learning_rate": 3.573971192915315e-07,
+      "loss": 0.9733,
+      "step": 7887
+    },
+    {
+      "epoch": 0.8295844452916508,
+      "grad_norm": 2.2148340995812124,
+      "learning_rate": 3.569670997233371e-07,
+      "loss": 0.9621,
+      "step": 7888
+    },
+    {
+      "epoch": 0.8296896157334981,
+      "grad_norm": 3.2864299842355895,
+      "learning_rate": 3.5653731911995893e-07,
+      "loss": 1.0221,
+      "step": 7889
+    },
+    {
+      "epoch": 0.8297947861753454,
+      "grad_norm": 2.4206938864595036,
+      "learning_rate": 3.5610777752932117e-07,
+      "loss": 0.9424,
+      "step": 7890
+    },
+    {
+      "epoch": 0.8298999566171927,
+      "grad_norm": 2.524736186816006,
+      "learning_rate": 3.5567847499932004e-07,
+      "loss": 1.0228,
+      "step": 7891
+    },
+    {
+      "epoch": 0.8300051270590401,
+      "grad_norm": 2.8928862446277224,
+      "learning_rate": 3.552494115778271e-07,
+      "loss": 0.9781,
+      "step": 7892
+    },
+    {
+      "epoch": 0.8301102975008874,
+      "grad_norm": 1.951995142147348,
+      "learning_rate": 3.5482058731268503e-07,
+      "loss": 0.9185,
+      "step": 7893
+    },
+    {
+      "epoch": 0.8302154679427347,
+      "grad_norm": 1.5618142840999194,
+      "learning_rate": 3.5439200225171193e-07,
+      "loss": 0.8677,
+      "step": 7894
+    },
+    {
+      "epoch": 0.830320638384582,
+      "grad_norm": 2.567017305726929,
+      "learning_rate": 3.539636564426985e-07,
+      "loss": 1.0362,
+      "step": 7895
+    },
+    {
+      "epoch": 0.8304258088264294,
+      "grad_norm": 2.357574361477999,
+      "learning_rate": 3.535355499334087e-07,
+      "loss": 1.0022,
+      "step": 7896
+    },
+    {
+      "epoch": 0.8305309792682767,
+      "grad_norm": 2.3855821109988904,
+      "learning_rate": 3.5310768277157875e-07,
+      "loss": 0.9882,
+      "step": 7897
+    },
+    {
+      "epoch": 0.830636149710124,
+      "grad_norm": 2.4659785940655357,
+      "learning_rate": 3.5268005500492004e-07,
+      "loss": 0.9991,
+      "step": 7898
+    },
+    {
+      "epoch": 0.8307413201519713,
+      "grad_norm": 2.8955739385288086,
+      "learning_rate": 3.522526666811166e-07,
+      "loss": 0.9721,
+      "step": 7899
+    },
+    {
+      "epoch": 0.8308464905938187,
+      "grad_norm": 2.2199143818637275,
+      "learning_rate": 3.518255178478247e-07,
+      "loss": 1.028,
+      "step": 7900
+    },
+    {
+      "epoch": 0.830951661035666,
+      "grad_norm": 2.198553962750598,
+      "learning_rate": 3.513986085526755e-07,
+      "loss": 0.949,
+      "step": 7901
+    },
+    {
+      "epoch": 0.8310568314775132,
+      "grad_norm": 2.6375318358953828,
+      "learning_rate": 3.50971938843272e-07,
+      "loss": 1.0045,
+      "step": 7902
+    },
+    {
+      "epoch": 0.8311620019193605,
+      "grad_norm": 2.2539176960571585,
+      "learning_rate": 3.5054550876719134e-07,
+      "loss": 0.9797,
+      "step": 7903
+    },
+    {
+      "epoch": 0.8312671723612078,
+      "grad_norm": 2.621497832230979,
+      "learning_rate": 3.5011931837198476e-07,
+      "loss": 0.9706,
+      "step": 7904
+    },
+    {
+      "epoch": 0.8313723428030552,
+      "grad_norm": 2.408630576835979,
+      "learning_rate": 3.496933677051745e-07,
+      "loss": 1.0101,
+      "step": 7905
+    },
+    {
+      "epoch": 0.8314775132449025,
+      "grad_norm": 2.5560399097177635,
+      "learning_rate": 3.4926765681425835e-07,
+      "loss": 1.0304,
+      "step": 7906
+    },
+    {
+      "epoch": 0.8315826836867498,
+      "grad_norm": 2.6823556900283565,
+      "learning_rate": 3.488421857467053e-07,
+      "loss": 0.9543,
+      "step": 7907
+    },
+    {
+      "epoch": 0.8316878541285971,
+      "grad_norm": 1.7262174828332262,
+      "learning_rate": 3.484169545499594e-07,
+      "loss": 1.0048,
+      "step": 7908
+    },
+    {
+      "epoch": 0.8317930245704445,
+      "grad_norm": 2.243625820617893,
+      "learning_rate": 3.479919632714374e-07,
+      "loss": 0.9714,
+      "step": 7909
+    },
+    {
+      "epoch": 0.8318981950122918,
+      "grad_norm": 2.2906586964812985,
+      "learning_rate": 3.475672119585291e-07,
+      "loss": 0.9659,
+      "step": 7910
+    },
+    {
+      "epoch": 0.8320033654541391,
+      "grad_norm": 2.6417368186284964,
+      "learning_rate": 3.4714270065859673e-07,
+      "loss": 0.9472,
+      "step": 7911
+    },
+    {
+      "epoch": 0.8321085358959864,
+      "grad_norm": 3.0260483510274017,
+      "learning_rate": 3.4671842941897764e-07,
+      "loss": 1.0046,
+      "step": 7912
+    },
+    {
+      "epoch": 0.8322137063378338,
+      "grad_norm": 2.59174736426387,
+      "learning_rate": 3.4629439828698125e-07,
+      "loss": 1.0117,
+      "step": 7913
+    },
+    {
+      "epoch": 0.8323188767796811,
+      "grad_norm": 2.855621058386072,
+      "learning_rate": 3.4587060730988943e-07,
+      "loss": 0.9702,
+      "step": 7914
+    },
+    {
+      "epoch": 0.8324240472215284,
+      "grad_norm": 2.3861443308093295,
+      "learning_rate": 3.4544705653495936e-07,
+      "loss": 0.9683,
+      "step": 7915
+    },
+    {
+      "epoch": 0.8325292176633757,
+      "grad_norm": 2.956455393722912,
+      "learning_rate": 3.4502374600941915e-07,
+      "loss": 0.9796,
+      "step": 7916
+    },
+    {
+      "epoch": 0.8326343881052231,
+      "grad_norm": 3.3272393954310275,
+      "learning_rate": 3.446006757804715e-07,
+      "loss": 0.9626,
+      "step": 7917
+    },
+    {
+      "epoch": 0.8327395585470704,
+      "grad_norm": 3.117936160153246,
+      "learning_rate": 3.44177845895293e-07,
+      "loss": 1.0031,
+      "step": 7918
+    },
+    {
+      "epoch": 0.8328447289889177,
+      "grad_norm": 2.3984852683425406,
+      "learning_rate": 3.4375525640103103e-07,
+      "loss": 0.9611,
+      "step": 7919
+    },
+    {
+      "epoch": 0.832949899430765,
+      "grad_norm": 2.548645604067526,
+      "learning_rate": 3.4333290734480885e-07,
+      "loss": 0.9952,
+      "step": 7920
+    },
+    {
+      "epoch": 0.8330550698726124,
+      "grad_norm": 2.5500483837731815,
+      "learning_rate": 3.4291079877371984e-07,
+      "loss": 0.9917,
+      "step": 7921
+    },
+    {
+      "epoch": 0.8331602403144596,
+      "grad_norm": 2.514529585621814,
+      "learning_rate": 3.4248893073483504e-07,
+      "loss": 1.0036,
+      "step": 7922
+    },
+    {
+      "epoch": 0.8332654107563069,
+      "grad_norm": 2.025207298660727,
+      "learning_rate": 3.4206730327519365e-07,
+      "loss": 0.978,
+      "step": 7923
+    },
+    {
+      "epoch": 0.8333705811981542,
+      "grad_norm": 2.5083140341213355,
+      "learning_rate": 3.4164591644181233e-07,
+      "loss": 0.9394,
+      "step": 7924
+    },
+    {
+      "epoch": 0.8334757516400015,
+      "grad_norm": 3.419460327667554,
+      "learning_rate": 3.412247702816771e-07,
+      "loss": 0.9656,
+      "step": 7925
+    },
+    {
+      "epoch": 0.8335809220818489,
+      "grad_norm": 2.5984729823260144,
+      "learning_rate": 3.408038648417503e-07,
+      "loss": 0.9887,
+      "step": 7926
+    },
+    {
+      "epoch": 0.8336860925236962,
+      "grad_norm": 2.3818846798670767,
+      "learning_rate": 3.403832001689661e-07,
+      "loss": 0.9715,
+      "step": 7927
+    },
+    {
+      "epoch": 0.8337912629655435,
+      "grad_norm": 2.210634104658908,
+      "learning_rate": 3.399627763102309e-07,
+      "loss": 0.978,
+      "step": 7928
+    },
+    {
+      "epoch": 0.8338964334073908,
+      "grad_norm": 2.5371862459046146,
+      "learning_rate": 3.3954259331242664e-07,
+      "loss": 0.9618,
+      "step": 7929
+    },
+    {
+      "epoch": 0.8340016038492382,
+      "grad_norm": 2.480566157647304,
+      "learning_rate": 3.391226512224055e-07,
+      "loss": 0.9725,
+      "step": 7930
+    },
+    {
+      "epoch": 0.8341067742910855,
+      "grad_norm": 2.1229236675007845,
+      "learning_rate": 3.387029500869954e-07,
+      "loss": 0.9619,
+      "step": 7931
+    },
+    {
+      "epoch": 0.8342119447329328,
+      "grad_norm": 1.9698108801024228,
+      "learning_rate": 3.3828348995299575e-07,
+      "loss": 0.9741,
+      "step": 7932
+    },
+    {
+      "epoch": 0.8343171151747801,
+      "grad_norm": 1.9418318681400226,
+      "learning_rate": 3.3786427086718006e-07,
+      "loss": 0.9577,
+      "step": 7933
+    },
+    {
+      "epoch": 0.8344222856166275,
+      "grad_norm": 3.3957679090439017,
+      "learning_rate": 3.3744529287629446e-07,
+      "loss": 0.9471,
+      "step": 7934
+    },
+    {
+      "epoch": 0.8345274560584748,
+      "grad_norm": 2.4124644360616436,
+      "learning_rate": 3.3702655602705693e-07,
+      "loss": 0.983,
+      "step": 7935
+    },
+    {
+      "epoch": 0.8346326265003221,
+      "grad_norm": 2.7101113991825825,
+      "learning_rate": 3.36608060366162e-07,
+      "loss": 0.9811,
+      "step": 7936
+    },
+    {
+      "epoch": 0.8347377969421694,
+      "grad_norm": 2.2798654842255157,
+      "learning_rate": 3.361898059402738e-07,
+      "loss": 0.9592,
+      "step": 7937
+    },
+    {
+      "epoch": 0.8348429673840168,
+      "grad_norm": 2.982597973091621,
+      "learning_rate": 3.357717927960316e-07,
+      "loss": 1.0268,
+      "step": 7938
+    },
+    {
+      "epoch": 0.8349481378258641,
+      "grad_norm": 2.298141064600936,
+      "learning_rate": 3.3535402098004626e-07,
+      "loss": 0.9903,
+      "step": 7939
+    },
+    {
+      "epoch": 0.8350533082677114,
+      "grad_norm": 2.3752357601389704,
+      "learning_rate": 3.3493649053890325e-07,
+      "loss": 0.9704,
+      "step": 7940
+    },
+    {
+      "epoch": 0.8351584787095587,
+      "grad_norm": 2.763407955749695,
+      "learning_rate": 3.345192015191609e-07,
+      "loss": 0.9829,
+      "step": 7941
+    },
+    {
+      "epoch": 0.835263649151406,
+      "grad_norm": 2.828981552542588,
+      "learning_rate": 3.341021539673492e-07,
+      "loss": 1.0321,
+      "step": 7942
+    },
+    {
+      "epoch": 0.8353688195932533,
+      "grad_norm": 2.2262635053414574,
+      "learning_rate": 3.3368534792997285e-07,
+      "loss": 0.9639,
+      "step": 7943
+    },
+    {
+      "epoch": 0.8354739900351006,
+      "grad_norm": 2.446688699369652,
+      "learning_rate": 3.3326878345350784e-07,
+      "loss": 1.0052,
+      "step": 7944
+    },
+    {
+      "epoch": 0.8355791604769479,
+      "grad_norm": 2.7455178000846363,
+      "learning_rate": 3.3285246058440635e-07,
+      "loss": 0.9826,
+      "step": 7945
+    },
+    {
+      "epoch": 0.8356843309187952,
+      "grad_norm": 2.464176366963785,
+      "learning_rate": 3.324363793690899e-07,
+      "loss": 0.9878,
+      "step": 7946
+    },
+    {
+      "epoch": 0.8357895013606426,
+      "grad_norm": 2.2726195863266705,
+      "learning_rate": 3.320205398539561e-07,
+      "loss": 0.9455,
+      "step": 7947
+    },
+    {
+      "epoch": 0.8358946718024899,
+      "grad_norm": 2.2989720657926402,
+      "learning_rate": 3.316049420853729e-07,
+      "loss": 0.9815,
+      "step": 7948
+    },
+    {
+      "epoch": 0.8359998422443372,
+      "grad_norm": 2.1279969880508705,
+      "learning_rate": 3.311895861096831e-07,
+      "loss": 0.9454,
+      "step": 7949
+    },
+    {
+      "epoch": 0.8361050126861845,
+      "grad_norm": 2.6446950816118195,
+      "learning_rate": 3.307744719732031e-07,
+      "loss": 0.9773,
+      "step": 7950
+    },
+    {
+      "epoch": 0.8362101831280319,
+      "grad_norm": 1.955699911337364,
+      "learning_rate": 3.3035959972222e-07,
+      "loss": 0.9484,
+      "step": 7951
+    },
+    {
+      "epoch": 0.8363153535698792,
+      "grad_norm": 2.332225298540404,
+      "learning_rate": 3.299449694029966e-07,
+      "loss": 0.9963,
+      "step": 7952
+    },
+    {
+      "epoch": 0.8364205240117265,
+      "grad_norm": 4.216667098594006,
+      "learning_rate": 3.2953058106176607e-07,
+      "loss": 1.0338,
+      "step": 7953
+    },
+    {
+      "epoch": 0.8365256944535738,
+      "grad_norm": 2.0765665881975752,
+      "learning_rate": 3.2911643474473647e-07,
+      "loss": 0.9991,
+      "step": 7954
+    },
+    {
+      "epoch": 0.8366308648954212,
+      "grad_norm": 2.044749035524921,
+      "learning_rate": 3.28702530498089e-07,
+      "loss": 0.9589,
+      "step": 7955
+    },
+    {
+      "epoch": 0.8367360353372685,
+      "grad_norm": 3.061177661810104,
+      "learning_rate": 3.282888683679758e-07,
+      "loss": 1.0022,
+      "step": 7956
+    },
+    {
+      "epoch": 0.8368412057791158,
+      "grad_norm": 1.852367172280801,
+      "learning_rate": 3.2787544840052503e-07,
+      "loss": 0.9729,
+      "step": 7957
+    },
+    {
+      "epoch": 0.8369463762209631,
+      "grad_norm": 2.4129977271544436,
+      "learning_rate": 3.274622706418346e-07,
+      "loss": 0.9655,
+      "step": 7958
+    },
+    {
+      "epoch": 0.8370515466628105,
+      "grad_norm": 2.467914364505314,
+      "learning_rate": 3.270493351379786e-07,
+      "loss": 0.9492,
+      "step": 7959
+    },
+    {
+      "epoch": 0.8371567171046578,
+      "grad_norm": 2.311681107608195,
+      "learning_rate": 3.2663664193500114e-07,
+      "loss": 0.9855,
+      "step": 7960
+    },
+    {
+      "epoch": 0.8372618875465051,
+      "grad_norm": 2.670758248675138,
+      "learning_rate": 3.262241910789221e-07,
+      "loss": 0.9683,
+      "step": 7961
+    },
+    {
+      "epoch": 0.8373670579883524,
+      "grad_norm": 2.1454789719559284,
+      "learning_rate": 3.258119826157319e-07,
+      "loss": 0.9359,
+      "step": 7962
+    },
+    {
+      "epoch": 0.8374722284301996,
+      "grad_norm": 2.4776693808188446,
+      "learning_rate": 3.2540001659139506e-07,
+      "loss": 1.0157,
+      "step": 7963
+    },
+    {
+      "epoch": 0.837577398872047,
+      "grad_norm": 1.8011863229194303,
+      "learning_rate": 3.249882930518497e-07,
+      "loss": 0.9874,
+      "step": 7964
+    },
+    {
+      "epoch": 0.8376825693138943,
+      "grad_norm": 2.168737945078986,
+      "learning_rate": 3.245768120430054e-07,
+      "loss": 0.993,
+      "step": 7965
+    },
+    {
+      "epoch": 0.8377877397557416,
+      "grad_norm": 3.0895348956728617,
+      "learning_rate": 3.241655736107466e-07,
+      "loss": 0.9992,
+      "step": 7966
+    },
+    {
+      "epoch": 0.8378929101975889,
+      "grad_norm": 3.3010610656877093,
+      "learning_rate": 3.237545778009274e-07,
+      "loss": 0.9793,
+      "step": 7967
+    },
+    {
+      "epoch": 0.8379980806394363,
+      "grad_norm": 2.7672284112761956,
+      "learning_rate": 3.233438246593795e-07,
+      "loss": 0.9919,
+      "step": 7968
+    },
+    {
+      "epoch": 0.8381032510812836,
+      "grad_norm": 2.0517422870765625,
+      "learning_rate": 3.229333142319041e-07,
+      "loss": 0.991,
+      "step": 7969
+    },
+    {
+      "epoch": 0.8382084215231309,
+      "grad_norm": 2.0009951842118143,
+      "learning_rate": 3.2252304656427566e-07,
+      "loss": 0.9795,
+      "step": 7970
+    },
+    {
+      "epoch": 0.8383135919649782,
+      "grad_norm": 2.4005492702244977,
+      "learning_rate": 3.221130217022433e-07,
+      "loss": 0.9675,
+      "step": 7971
+    },
+    {
+      "epoch": 0.8384187624068256,
+      "grad_norm": 2.1804042366396232,
+      "learning_rate": 3.217032396915265e-07,
+      "loss": 0.972,
+      "step": 7972
+    },
+    {
+      "epoch": 0.8385239328486729,
+      "grad_norm": 2.973961335735684,
+      "learning_rate": 3.2129370057782106e-07,
+      "loss": 0.9951,
+      "step": 7973
+    },
+    {
+      "epoch": 0.8386291032905202,
+      "grad_norm": 2.025266752429975,
+      "learning_rate": 3.208844044067921e-07,
+      "loss": 1.0002,
+      "step": 7974
+    },
+    {
+      "epoch": 0.8387342737323675,
+      "grad_norm": 2.4059729700613457,
+      "learning_rate": 3.2047535122408076e-07,
+      "loss": 0.9795,
+      "step": 7975
+    },
+    {
+      "epoch": 0.8388394441742149,
+      "grad_norm": 2.753932435517075,
+      "learning_rate": 3.200665410752982e-07,
+      "loss": 0.9798,
+      "step": 7976
+    },
+    {
+      "epoch": 0.8389446146160622,
+      "grad_norm": 2.412042721936347,
+      "learning_rate": 3.196579740060307e-07,
+      "loss": 0.9764,
+      "step": 7977
+    },
+    {
+      "epoch": 0.8390497850579095,
+      "grad_norm": 2.922557537170381,
+      "learning_rate": 3.192496500618372e-07,
+      "loss": 0.9586,
+      "step": 7978
+    },
+    {
+      "epoch": 0.8391549554997568,
+      "grad_norm": 2.2610207215000173,
+      "learning_rate": 3.188415692882477e-07,
+      "loss": 0.9687,
+      "step": 7979
+    },
+    {
+      "epoch": 0.8392601259416042,
+      "grad_norm": 2.6933915853361006,
+      "learning_rate": 3.1843373173076784e-07,
+      "loss": 1.0122,
+      "step": 7980
+    },
+    {
+      "epoch": 0.8393652963834515,
+      "grad_norm": 1.9203482516232904,
+      "learning_rate": 3.1802613743487255e-07,
+      "loss": 0.9712,
+      "step": 7981
+    },
+    {
+      "epoch": 0.8394704668252988,
+      "grad_norm": 2.8523465380370348,
+      "learning_rate": 3.1761878644601425e-07,
+      "loss": 0.9694,
+      "step": 7982
+    },
+    {
+      "epoch": 0.839575637267146,
+      "grad_norm": 2.2911244103217774,
+      "learning_rate": 3.172116788096141e-07,
+      "loss": 0.9529,
+      "step": 7983
+    },
+    {
+      "epoch": 0.8396808077089933,
+      "grad_norm": 2.715895442199244,
+      "learning_rate": 3.1680481457106895e-07,
+      "loss": 0.9757,
+      "step": 7984
+    },
+    {
+      "epoch": 0.8397859781508407,
+      "grad_norm": 2.250183946803774,
+      "learning_rate": 3.1639819377574624e-07,
+      "loss": 0.967,
+      "step": 7985
+    },
+    {
+      "epoch": 0.839891148592688,
+      "grad_norm": 2.107649904873707,
+      "learning_rate": 3.159918164689871e-07,
+      "loss": 1.0024,
+      "step": 7986
+    },
+    {
+      "epoch": 0.8399963190345353,
+      "grad_norm": 2.094200221460402,
+      "learning_rate": 3.1558568269610707e-07,
+      "loss": 0.9948,
+      "step": 7987
+    },
+    {
+      "epoch": 0.8401014894763826,
+      "grad_norm": 2.144487336078061,
+      "learning_rate": 3.151797925023922e-07,
+      "loss": 1.03,
+      "step": 7988
+    },
+    {
+      "epoch": 0.84020665991823,
+      "grad_norm": 2.264741520119065,
+      "learning_rate": 3.147741459331033e-07,
+      "loss": 0.9903,
+      "step": 7989
+    },
+    {
+      "epoch": 0.8403118303600773,
+      "grad_norm": 2.3375622944629764,
+      "learning_rate": 3.1436874303347185e-07,
+      "loss": 0.9826,
+      "step": 7990
+    },
+    {
+      "epoch": 0.8404170008019246,
+      "grad_norm": 2.5671078673407473,
+      "learning_rate": 3.1396358384870426e-07,
+      "loss": 0.9646,
+      "step": 7991
+    },
+    {
+      "epoch": 0.8405221712437719,
+      "grad_norm": 2.3619671652091667,
+      "learning_rate": 3.135586684239794e-07,
+      "loss": 0.9667,
+      "step": 7992
+    },
+    {
+      "epoch": 0.8406273416856193,
+      "grad_norm": 2.416335511787064,
+      "learning_rate": 3.1315399680444727e-07,
+      "loss": 0.9691,
+      "step": 7993
+    },
+    {
+      "epoch": 0.8407325121274666,
+      "grad_norm": 1.9661025621449095,
+      "learning_rate": 3.127495690352331e-07,
+      "loss": 0.9958,
+      "step": 7994
+    },
+    {
+      "epoch": 0.8408376825693139,
+      "grad_norm": 3.035263773083995,
+      "learning_rate": 3.1234538516143194e-07,
+      "loss": 0.9768,
+      "step": 7995
+    },
+    {
+      "epoch": 0.8409428530111612,
+      "grad_norm": 2.0867961026565207,
+      "learning_rate": 3.119414452281158e-07,
+      "loss": 1.0095,
+      "step": 7996
+    },
+    {
+      "epoch": 0.8410480234530086,
+      "grad_norm": 2.7704288005888826,
+      "learning_rate": 3.1153774928032564e-07,
+      "loss": 0.9915,
+      "step": 7997
+    },
+    {
+      "epoch": 0.8411531938948559,
+      "grad_norm": 2.6501249494588413,
+      "learning_rate": 3.111342973630771e-07,
+      "loss": 0.9808,
+      "step": 7998
+    },
+    {
+      "epoch": 0.8412583643367032,
+      "grad_norm": 2.438311465297352,
+      "learning_rate": 3.107310895213578e-07,
+      "loss": 0.962,
+      "step": 7999
+    },
+    {
+      "epoch": 0.8413635347785505,
+      "grad_norm": 2.345077898806914,
+      "learning_rate": 3.1032812580012895e-07,
+      "loss": 0.9818,
+      "step": 8000
+    },
+    {
+      "epoch": 0.8414687052203979,
+      "grad_norm": 2.891411545914225,
+      "learning_rate": 3.0992540624432436e-07,
+      "loss": 0.9883,
+      "step": 8001
+    },
+    {
+      "epoch": 0.8415738756622452,
+      "grad_norm": 2.415834378113018,
+      "learning_rate": 3.0952293089884984e-07,
+      "loss": 0.9521,
+      "step": 8002
+    },
+    {
+      "epoch": 0.8416790461040925,
+      "grad_norm": 2.8617855981045426,
+      "learning_rate": 3.0912069980858496e-07,
+      "loss": 0.9902,
+      "step": 8003
+    },
+    {
+      "epoch": 0.8417842165459397,
+      "grad_norm": 2.8879149599658422,
+      "learning_rate": 3.0871871301838053e-07,
+      "loss": 1.0216,
+      "step": 8004
+    },
+    {
+      "epoch": 0.841889386987787,
+      "grad_norm": 1.8612027892835419,
+      "learning_rate": 3.083169705730629e-07,
+      "loss": 0.96,
+      "step": 8005
+    },
+    {
+      "epoch": 0.8419945574296344,
+      "grad_norm": 2.6330298233689122,
+      "learning_rate": 3.079154725174288e-07,
+      "loss": 0.9465,
+      "step": 8006
+    },
+    {
+      "epoch": 0.8420997278714817,
+      "grad_norm": 2.138955256278022,
+      "learning_rate": 3.0751421889624766e-07,
+      "loss": 0.947,
+      "step": 8007
+    },
+    {
+      "epoch": 0.842204898313329,
+      "grad_norm": 2.616727462598564,
+      "learning_rate": 3.0711320975426367e-07,
+      "loss": 0.9434,
+      "step": 8008
+    },
+    {
+      "epoch": 0.8423100687551763,
+      "grad_norm": 2.4895885340978543,
+      "learning_rate": 3.067124451361905e-07,
+      "loss": 1.0038,
+      "step": 8009
+    },
+    {
+      "epoch": 0.8424152391970237,
+      "grad_norm": 3.4008930173947873,
+      "learning_rate": 3.063119250867186e-07,
+      "loss": 0.9394,
+      "step": 8010
+    },
+    {
+      "epoch": 0.842520409638871,
+      "grad_norm": 2.851466967075379,
+      "learning_rate": 3.05911649650508e-07,
+      "loss": 0.954,
+      "step": 8011
+    },
+    {
+      "epoch": 0.8426255800807183,
+      "grad_norm": 2.9185957846098027,
+      "learning_rate": 3.055116188721932e-07,
+      "loss": 1.0138,
+      "step": 8012
+    },
+    {
+      "epoch": 0.8427307505225656,
+      "grad_norm": 2.260085530350411,
+      "learning_rate": 3.051118327963798e-07,
+      "loss": 0.9917,
+      "step": 8013
+    },
+    {
+      "epoch": 0.842835920964413,
+      "grad_norm": 2.458882313705006,
+      "learning_rate": 3.047122914676473e-07,
+      "loss": 1.0258,
+      "step": 8014
+    },
+    {
+      "epoch": 0.8429410914062603,
+      "grad_norm": 3.1014214186637883,
+      "learning_rate": 3.043129949305487e-07,
+      "loss": 0.9228,
+      "step": 8015
+    },
+    {
+      "epoch": 0.8430462618481076,
+      "grad_norm": 1.6614260018195441,
+      "learning_rate": 3.0391394322960706e-07,
+      "loss": 0.9825,
+      "step": 8016
+    },
+    {
+      "epoch": 0.8431514322899549,
+      "grad_norm": 2.0667501456498605,
+      "learning_rate": 3.035151364093214e-07,
+      "loss": 1.0093,
+      "step": 8017
+    },
+    {
+      "epoch": 0.8432566027318023,
+      "grad_norm": 2.598733278916,
+      "learning_rate": 3.031165745141595e-07,
+      "loss": 0.9792,
+      "step": 8018
+    },
+    {
+      "epoch": 0.8433617731736496,
+      "grad_norm": 2.5060894728850913,
+      "learning_rate": 3.0271825758856675e-07,
+      "loss": 0.977,
+      "step": 8019
+    },
+    {
+      "epoch": 0.8434669436154969,
+      "grad_norm": 2.55384223599344,
+      "learning_rate": 3.02320185676957e-07,
+      "loss": 1.0313,
+      "step": 8020
+    },
+    {
+      "epoch": 0.8435721140573442,
+      "grad_norm": 3.1720208128466916,
+      "learning_rate": 3.0192235882371913e-07,
+      "loss": 0.9646,
+      "step": 8021
+    },
+    {
+      "epoch": 0.8436772844991915,
+      "grad_norm": 2.813164544453831,
+      "learning_rate": 3.01524777073213e-07,
+      "loss": 0.9844,
+      "step": 8022
+    },
+    {
+      "epoch": 0.8437824549410389,
+      "grad_norm": 1.7681715170092602,
+      "learning_rate": 3.0112744046977255e-07,
+      "loss": 0.9852,
+      "step": 8023
+    },
+    {
+      "epoch": 0.8438876253828861,
+      "grad_norm": 2.8437496154777326,
+      "learning_rate": 3.007303490577043e-07,
+      "loss": 1.0048,
+      "step": 8024
+    },
+    {
+      "epoch": 0.8439927958247334,
+      "grad_norm": 2.7174670709863884,
+      "learning_rate": 3.003335028812862e-07,
+      "loss": 0.978,
+      "step": 8025
+    },
+    {
+      "epoch": 0.8440979662665807,
+      "grad_norm": 3.067825714603444,
+      "learning_rate": 2.999369019847709e-07,
+      "loss": 0.9515,
+      "step": 8026
+    },
+    {
+      "epoch": 0.8442031367084281,
+      "grad_norm": 2.8915929181982105,
+      "learning_rate": 2.99540546412381e-07,
+      "loss": 1.0105,
+      "step": 8027
+    },
+    {
+      "epoch": 0.8443083071502754,
+      "grad_norm": 1.8900985976770452,
+      "learning_rate": 2.99144436208314e-07,
+      "loss": 0.9914,
+      "step": 8028
+    },
+    {
+      "epoch": 0.8444134775921227,
+      "grad_norm": 2.2877820769502213,
+      "learning_rate": 2.987485714167396e-07,
+      "loss": 0.9966,
+      "step": 8029
+    },
+    {
+      "epoch": 0.84451864803397,
+      "grad_norm": 2.196650528892029,
+      "learning_rate": 2.9835295208179907e-07,
+      "loss": 0.9936,
+      "step": 8030
+    },
+    {
+      "epoch": 0.8446238184758174,
+      "grad_norm": 3.2894078035994663,
+      "learning_rate": 2.9795757824760796e-07,
+      "loss": 1.0142,
+      "step": 8031
+    },
+    {
+      "epoch": 0.8447289889176647,
+      "grad_norm": 2.297997513581351,
+      "learning_rate": 2.975624499582516e-07,
+      "loss": 0.911,
+      "step": 8032
+    },
+    {
+      "epoch": 0.844834159359512,
+      "grad_norm": 2.647899001008078,
+      "learning_rate": 2.9716756725779254e-07,
+      "loss": 0.9509,
+      "step": 8033
+    },
+    {
+      "epoch": 0.8449393298013593,
+      "grad_norm": 2.360019920855114,
+      "learning_rate": 2.967729301902611e-07,
+      "loss": 1.0027,
+      "step": 8034
+    },
+    {
+      "epoch": 0.8450445002432067,
+      "grad_norm": 2.999008966583943,
+      "learning_rate": 2.96378538799664e-07,
+      "loss": 1.0052,
+      "step": 8035
+    },
+    {
+      "epoch": 0.845149670685054,
+      "grad_norm": 2.6618078141442987,
+      "learning_rate": 2.9598439312997745e-07,
+      "loss": 0.9906,
+      "step": 8036
+    },
+    {
+      "epoch": 0.8452548411269013,
+      "grad_norm": 2.4720173368158074,
+      "learning_rate": 2.955904932251527e-07,
+      "loss": 0.9599,
+      "step": 8037
+    },
+    {
+      "epoch": 0.8453600115687486,
+      "grad_norm": 2.8992055015069678,
+      "learning_rate": 2.9519683912911267e-07,
+      "loss": 1.0055,
+      "step": 8038
+    },
+    {
+      "epoch": 0.845465182010596,
+      "grad_norm": 2.265207100310217,
+      "learning_rate": 2.9480343088575227e-07,
+      "loss": 1.0136,
+      "step": 8039
+    },
+    {
+      "epoch": 0.8455703524524433,
+      "grad_norm": 2.4872651865839615,
+      "learning_rate": 2.9441026853894024e-07,
+      "loss": 0.9741,
+      "step": 8040
+    },
+    {
+      "epoch": 0.8456755228942906,
+      "grad_norm": 2.3337883217309714,
+      "learning_rate": 2.940173521325165e-07,
+      "loss": 0.9959,
+      "step": 8041
+    },
+    {
+      "epoch": 0.8457806933361379,
+      "grad_norm": 2.4394701294128054,
+      "learning_rate": 2.936246817102947e-07,
+      "loss": 0.9584,
+      "step": 8042
+    },
+    {
+      "epoch": 0.8458858637779852,
+      "grad_norm": 1.8882443397504602,
+      "learning_rate": 2.9323225731606094e-07,
+      "loss": 0.9236,
+      "step": 8043
+    },
+    {
+      "epoch": 0.8459910342198325,
+      "grad_norm": 2.2667033507578913,
+      "learning_rate": 2.9284007899357286e-07,
+      "loss": 0.9221,
+      "step": 8044
+    },
+    {
+      "epoch": 0.8460962046616798,
+      "grad_norm": 2.370795500055156,
+      "learning_rate": 2.924481467865617e-07,
+      "loss": 0.9892,
+      "step": 8045
+    },
+    {
+      "epoch": 0.8462013751035271,
+      "grad_norm": 2.6229927783843827,
+      "learning_rate": 2.920564607387313e-07,
+      "loss": 1.0033,
+      "step": 8046
+    },
+    {
+      "epoch": 0.8463065455453744,
+      "grad_norm": 2.675563045044748,
+      "learning_rate": 2.916650208937577e-07,
+      "loss": 0.9703,
+      "step": 8047
+    },
+    {
+      "epoch": 0.8464117159872218,
+      "grad_norm": 2.7466454975694754,
+      "learning_rate": 2.912738272952889e-07,
+      "loss": 0.9759,
+      "step": 8048
+    },
+    {
+      "epoch": 0.8465168864290691,
+      "grad_norm": 2.1964656610410445,
+      "learning_rate": 2.9088287998694673e-07,
+      "loss": 0.9888,
+      "step": 8049
+    },
+    {
+      "epoch": 0.8466220568709164,
+      "grad_norm": 2.479272749557706,
+      "learning_rate": 2.9049217901232387e-07,
+      "loss": 0.9398,
+      "step": 8050
+    },
+    {
+      "epoch": 0.8467272273127637,
+      "grad_norm": 2.787185292911654,
+      "learning_rate": 2.901017244149873e-07,
+      "loss": 0.9828,
+      "step": 8051
+    },
+    {
+      "epoch": 0.846832397754611,
+      "grad_norm": 1.98715699218698,
+      "learning_rate": 2.897115162384759e-07,
+      "loss": 1.0073,
+      "step": 8052
+    },
+    {
+      "epoch": 0.8469375681964584,
+      "grad_norm": 2.17127153048729,
+      "learning_rate": 2.893215545263001e-07,
+      "loss": 0.9981,
+      "step": 8053
+    },
+    {
+      "epoch": 0.8470427386383057,
+      "grad_norm": 2.3336424406843115,
+      "learning_rate": 2.889318393219445e-07,
+      "loss": 0.9829,
+      "step": 8054
+    },
+    {
+      "epoch": 0.847147909080153,
+      "grad_norm": 2.250375988178359,
+      "learning_rate": 2.885423706688642e-07,
+      "loss": 0.9889,
+      "step": 8055
+    },
+    {
+      "epoch": 0.8472530795220004,
+      "grad_norm": 2.1354421736272293,
+      "learning_rate": 2.8815314861048966e-07,
+      "loss": 1.0016,
+      "step": 8056
+    },
+    {
+      "epoch": 0.8473582499638477,
+      "grad_norm": 2.331711007943775,
+      "learning_rate": 2.8776417319022145e-07,
+      "loss": 0.9307,
+      "step": 8057
+    },
+    {
+      "epoch": 0.847463420405695,
+      "grad_norm": 2.8514494357851214,
+      "learning_rate": 2.8737544445143263e-07,
+      "loss": 0.9425,
+      "step": 8058
+    },
+    {
+      "epoch": 0.8475685908475423,
+      "grad_norm": 2.6329555820560557,
+      "learning_rate": 2.869869624374699e-07,
+      "loss": 0.9704,
+      "step": 8059
+    },
+    {
+      "epoch": 0.8476737612893896,
+      "grad_norm": 2.129162011010815,
+      "learning_rate": 2.8659872719165203e-07,
+      "loss": 0.9776,
+      "step": 8060
+    },
+    {
+      "epoch": 0.847778931731237,
+      "grad_norm": 2.334892721506637,
+      "learning_rate": 2.86210738757271e-07,
+      "loss": 0.9834,
+      "step": 8061
+    },
+    {
+      "epoch": 0.8478841021730843,
+      "grad_norm": 2.3641357364728273,
+      "learning_rate": 2.858229971775894e-07,
+      "loss": 0.9409,
+      "step": 8062
+    },
+    {
+      "epoch": 0.8479892726149316,
+      "grad_norm": 2.1169833059337764,
+      "learning_rate": 2.8543550249584465e-07,
+      "loss": 0.9716,
+      "step": 8063
+    },
+    {
+      "epoch": 0.848094443056779,
+      "grad_norm": 2.2632523362465875,
+      "learning_rate": 2.850482547552441e-07,
+      "loss": 0.9724,
+      "step": 8064
+    },
+    {
+      "epoch": 0.8481996134986262,
+      "grad_norm": 2.508398111461374,
+      "learning_rate": 2.846612539989693e-07,
+      "loss": 0.9645,
+      "step": 8065
+    },
+    {
+      "epoch": 0.8483047839404735,
+      "grad_norm": 3.107750551526337,
+      "learning_rate": 2.8427450027017493e-07,
+      "loss": 0.9812,
+      "step": 8066
+    },
+    {
+      "epoch": 0.8484099543823208,
+      "grad_norm": 2.1641973776169845,
+      "learning_rate": 2.838879936119854e-07,
+      "loss": 0.9508,
+      "step": 8067
+    },
+    {
+      "epoch": 0.8485151248241681,
+      "grad_norm": 2.465015127853074,
+      "learning_rate": 2.8350173406749975e-07,
+      "loss": 0.9792,
+      "step": 8068
+    },
+    {
+      "epoch": 0.8486202952660155,
+      "grad_norm": 1.9790492113518683,
+      "learning_rate": 2.8311572167978947e-07,
+      "loss": 0.9586,
+      "step": 8069
+    },
+    {
+      "epoch": 0.8487254657078628,
+      "grad_norm": 3.114675329966703,
+      "learning_rate": 2.827299564918978e-07,
+      "loss": 0.9591,
+      "step": 8070
+    },
+    {
+      "epoch": 0.8488306361497101,
+      "grad_norm": 2.733280124937022,
+      "learning_rate": 2.823444385468399e-07,
+      "loss": 1.0032,
+      "step": 8071
+    },
+    {
+      "epoch": 0.8489358065915574,
+      "grad_norm": 2.0042271783468766,
+      "learning_rate": 2.819591678876052e-07,
+      "loss": 0.9599,
+      "step": 8072
+    },
+    {
+      "epoch": 0.8490409770334048,
+      "grad_norm": 2.8190894222776337,
+      "learning_rate": 2.815741445571529e-07,
+      "loss": 0.9586,
+      "step": 8073
+    },
+    {
+      "epoch": 0.8491461474752521,
+      "grad_norm": 2.8042120225394753,
+      "learning_rate": 2.8118936859841684e-07,
+      "loss": 0.9776,
+      "step": 8074
+    },
+    {
+      "epoch": 0.8492513179170994,
+      "grad_norm": 2.0986367427377446,
+      "learning_rate": 2.8080484005430313e-07,
+      "loss": 0.9594,
+      "step": 8075
+    },
+    {
+      "epoch": 0.8493564883589467,
+      "grad_norm": 2.8316742922493017,
+      "learning_rate": 2.804205589676884e-07,
+      "loss": 0.9596,
+      "step": 8076
+    },
+    {
+      "epoch": 0.849461658800794,
+      "grad_norm": 3.077204387316349,
+      "learning_rate": 2.8003652538142413e-07,
+      "loss": 0.96,
+      "step": 8077
+    },
+    {
+      "epoch": 0.8495668292426414,
+      "grad_norm": 2.310308677624084,
+      "learning_rate": 2.796527393383322e-07,
+      "loss": 0.9743,
+      "step": 8078
+    },
+    {
+      "epoch": 0.8496719996844887,
+      "grad_norm": 2.1833465701082884,
+      "learning_rate": 2.7926920088120786e-07,
+      "loss": 0.9564,
+      "step": 8079
+    },
+    {
+      "epoch": 0.849777170126336,
+      "grad_norm": 2.3941525202307274,
+      "learning_rate": 2.788859100528196e-07,
+      "loss": 0.9678,
+      "step": 8080
+    },
+    {
+      "epoch": 0.8498823405681833,
+      "grad_norm": 2.6311798448564607,
+      "learning_rate": 2.785028668959061e-07,
+      "loss": 0.9633,
+      "step": 8081
+    },
+    {
+      "epoch": 0.8499875110100307,
+      "grad_norm": 2.963362741185363,
+      "learning_rate": 2.781200714531801e-07,
+      "loss": 0.9707,
+      "step": 8082
+    },
+    {
+      "epoch": 0.850092681451878,
+      "grad_norm": 3.184965341727443,
+      "learning_rate": 2.7773752376732605e-07,
+      "loss": 1.0092,
+      "step": 8083
+    },
+    {
+      "epoch": 0.8501978518937253,
+      "grad_norm": 1.8532486139994901,
+      "learning_rate": 2.7735522388100206e-07,
+      "loss": 0.9832,
+      "step": 8084
+    },
+    {
+      "epoch": 0.8503030223355725,
+      "grad_norm": 2.1835915901309075,
+      "learning_rate": 2.76973171836836e-07,
+      "loss": 0.9377,
+      "step": 8085
+    },
+    {
+      "epoch": 0.8504081927774199,
+      "grad_norm": 3.5734845337817935,
+      "learning_rate": 2.7659136767743096e-07,
+      "loss": 1.0057,
+      "step": 8086
+    },
+    {
+      "epoch": 0.8505133632192672,
+      "grad_norm": 2.1809470260774644,
+      "learning_rate": 2.7620981144536014e-07,
+      "loss": 0.943,
+      "step": 8087
+    },
+    {
+      "epoch": 0.8506185336611145,
+      "grad_norm": 2.348459690619445,
+      "learning_rate": 2.7582850318317005e-07,
+      "loss": 0.929,
+      "step": 8088
+    },
+    {
+      "epoch": 0.8507237041029618,
+      "grad_norm": 2.4740076632830146,
+      "learning_rate": 2.7544744293338063e-07,
+      "loss": 0.9618,
+      "step": 8089
+    },
+    {
+      "epoch": 0.8508288745448092,
+      "grad_norm": 3.4907303797398517,
+      "learning_rate": 2.750666307384814e-07,
+      "loss": 1.0089,
+      "step": 8090
+    },
+    {
+      "epoch": 0.8509340449866565,
+      "grad_norm": 3.116098339723679,
+      "learning_rate": 2.746860666409371e-07,
+      "loss": 0.9985,
+      "step": 8091
+    },
+    {
+      "epoch": 0.8510392154285038,
+      "grad_norm": 2.302135132288717,
+      "learning_rate": 2.743057506831834e-07,
+      "loss": 0.9955,
+      "step": 8092
+    },
+    {
+      "epoch": 0.8511443858703511,
+      "grad_norm": 2.454597566905971,
+      "learning_rate": 2.7392568290762786e-07,
+      "loss": 0.9345,
+      "step": 8093
+    },
+    {
+      "epoch": 0.8512495563121985,
+      "grad_norm": 2.297569677220409,
+      "learning_rate": 2.7354586335665205e-07,
+      "loss": 0.9634,
+      "step": 8094
+    },
+    {
+      "epoch": 0.8513547267540458,
+      "grad_norm": 2.5670899366081614,
+      "learning_rate": 2.7316629207260745e-07,
+      "loss": 0.9547,
+      "step": 8095
+    },
+    {
+      "epoch": 0.8514598971958931,
+      "grad_norm": 2.111670089542108,
+      "learning_rate": 2.727869690978202e-07,
+      "loss": 0.9991,
+      "step": 8096
+    },
+    {
+      "epoch": 0.8515650676377404,
+      "grad_norm": 2.1740992859773454,
+      "learning_rate": 2.7240789447458756e-07,
+      "loss": 0.9991,
+      "step": 8097
+    },
+    {
+      "epoch": 0.8516702380795877,
+      "grad_norm": 3.134509753421687,
+      "learning_rate": 2.7202906824517955e-07,
+      "loss": 0.9657,
+      "step": 8098
+    },
+    {
+      "epoch": 0.8517754085214351,
+      "grad_norm": 2.375342092275538,
+      "learning_rate": 2.716504904518372e-07,
+      "loss": 0.9857,
+      "step": 8099
+    },
+    {
+      "epoch": 0.8518805789632824,
+      "grad_norm": 2.1830836767759045,
+      "learning_rate": 2.7127216113677636e-07,
+      "loss": 0.9715,
+      "step": 8100
+    },
+    {
+      "epoch": 0.8519857494051297,
+      "grad_norm": 3.471010207274205,
+      "learning_rate": 2.708940803421825e-07,
+      "loss": 1.0784,
+      "step": 8101
+    },
+    {
+      "epoch": 0.852090919846977,
+      "grad_norm": 2.6989538225629897,
+      "learning_rate": 2.705162481102147e-07,
+      "loss": 0.9515,
+      "step": 8102
+    },
+    {
+      "epoch": 0.8521960902888244,
+      "grad_norm": 2.428855683370294,
+      "learning_rate": 2.7013866448300506e-07,
+      "loss": 1.0098,
+      "step": 8103
+    },
+    {
+      "epoch": 0.8523012607306717,
+      "grad_norm": 1.9294068184039133,
+      "learning_rate": 2.697613295026563e-07,
+      "loss": 0.9569,
+      "step": 8104
+    },
+    {
+      "epoch": 0.8524064311725189,
+      "grad_norm": 2.703791788987313,
+      "learning_rate": 2.6938424321124396e-07,
+      "loss": 1.0116,
+      "step": 8105
+    },
+    {
+      "epoch": 0.8525116016143662,
+      "grad_norm": 1.8334006288217113,
+      "learning_rate": 2.690074056508168e-07,
+      "loss": 1.0143,
+      "step": 8106
+    },
+    {
+      "epoch": 0.8526167720562136,
+      "grad_norm": 2.086771645052246,
+      "learning_rate": 2.686308168633953e-07,
+      "loss": 0.9975,
+      "step": 8107
+    },
+    {
+      "epoch": 0.8527219424980609,
+      "grad_norm": 2.064759245549448,
+      "learning_rate": 2.6825447689097174e-07,
+      "loss": 0.9949,
+      "step": 8108
+    },
+    {
+      "epoch": 0.8528271129399082,
+      "grad_norm": 2.5706290935033627,
+      "learning_rate": 2.678783857755102e-07,
+      "loss": 0.9576,
+      "step": 8109
+    },
+    {
+      "epoch": 0.8529322833817555,
+      "grad_norm": 2.7378677889966974,
+      "learning_rate": 2.675025435589482e-07,
+      "loss": 0.9744,
+      "step": 8110
+    },
+    {
+      "epoch": 0.8530374538236029,
+      "grad_norm": 2.8537643482308424,
+      "learning_rate": 2.671269502831955e-07,
+      "loss": 0.9979,
+      "step": 8111
+    },
+    {
+      "epoch": 0.8531426242654502,
+      "grad_norm": 2.18393636096386,
+      "learning_rate": 2.6675160599013374e-07,
+      "loss": 0.9741,
+      "step": 8112
+    },
+    {
+      "epoch": 0.8532477947072975,
+      "grad_norm": 2.101881287973468,
+      "learning_rate": 2.66376510721616e-07,
+      "loss": 0.9905,
+      "step": 8113
+    },
+    {
+      "epoch": 0.8533529651491448,
+      "grad_norm": 2.1876358871831973,
+      "learning_rate": 2.660016645194688e-07,
+      "loss": 0.9924,
+      "step": 8114
+    },
+    {
+      "epoch": 0.8534581355909922,
+      "grad_norm": 3.243463727452169,
+      "learning_rate": 2.6562706742549073e-07,
+      "loss": 0.9812,
+      "step": 8115
+    },
+    {
+      "epoch": 0.8535633060328395,
+      "grad_norm": 3.2673747014657084,
+      "learning_rate": 2.652527194814511e-07,
+      "loss": 1.0241,
+      "step": 8116
+    },
+    {
+      "epoch": 0.8536684764746868,
+      "grad_norm": 2.4360167551039074,
+      "learning_rate": 2.6487862072909404e-07,
+      "loss": 0.9474,
+      "step": 8117
+    },
+    {
+      "epoch": 0.8537736469165341,
+      "grad_norm": 2.431103723048994,
+      "learning_rate": 2.645047712101334e-07,
+      "loss": 1.0166,
+      "step": 8118
+    },
+    {
+      "epoch": 0.8538788173583814,
+      "grad_norm": 2.598232947614198,
+      "learning_rate": 2.641311709662567e-07,
+      "loss": 0.9566,
+      "step": 8119
+    },
+    {
+      "epoch": 0.8539839878002288,
+      "grad_norm": 2.5875433885989123,
+      "learning_rate": 2.637578200391233e-07,
+      "loss": 0.9922,
+      "step": 8120
+    },
+    {
+      "epoch": 0.8540891582420761,
+      "grad_norm": 2.3388693300392327,
+      "learning_rate": 2.63384718470365e-07,
+      "loss": 0.9561,
+      "step": 8121
+    },
+    {
+      "epoch": 0.8541943286839234,
+      "grad_norm": 2.754827845758967,
+      "learning_rate": 2.630118663015849e-07,
+      "loss": 0.989,
+      "step": 8122
+    },
+    {
+      "epoch": 0.8542994991257707,
+      "grad_norm": 2.1192941339795683,
+      "learning_rate": 2.626392635743594e-07,
+      "loss": 0.9967,
+      "step": 8123
+    },
+    {
+      "epoch": 0.8544046695676181,
+      "grad_norm": 2.616739686808066,
+      "learning_rate": 2.6226691033023665e-07,
+      "loss": 0.9751,
+      "step": 8124
+    },
+    {
+      "epoch": 0.8545098400094654,
+      "grad_norm": 2.0640093068786705,
+      "learning_rate": 2.618948066107363e-07,
+      "loss": 0.9815,
+      "step": 8125
+    },
+    {
+      "epoch": 0.8546150104513126,
+      "grad_norm": 3.0722152608836364,
+      "learning_rate": 2.615229524573518e-07,
+      "loss": 0.9867,
+      "step": 8126
+    },
+    {
+      "epoch": 0.8547201808931599,
+      "grad_norm": 2.697664599244337,
+      "learning_rate": 2.6115134791154657e-07,
+      "loss": 0.9973,
+      "step": 8127
+    },
+    {
+      "epoch": 0.8548253513350073,
+      "grad_norm": 2.1491230648206012,
+      "learning_rate": 2.607799930147581e-07,
+      "loss": 0.9659,
+      "step": 8128
+    },
+    {
+      "epoch": 0.8549305217768546,
+      "grad_norm": 2.091866541025715,
+      "learning_rate": 2.604088878083957e-07,
+      "loss": 0.9658,
+      "step": 8129
+    },
+    {
+      "epoch": 0.8550356922187019,
+      "grad_norm": 2.065310881673391,
+      "learning_rate": 2.600380323338397e-07,
+      "loss": 0.981,
+      "step": 8130
+    },
+    {
+      "epoch": 0.8551408626605492,
+      "grad_norm": 2.637190411853234,
+      "learning_rate": 2.596674266324442e-07,
+      "loss": 0.9403,
+      "step": 8131
+    },
+    {
+      "epoch": 0.8552460331023966,
+      "grad_norm": 2.2549585723194756,
+      "learning_rate": 2.5929707074553364e-07,
+      "loss": 0.9679,
+      "step": 8132
+    },
+    {
+      "epoch": 0.8553512035442439,
+      "grad_norm": 2.1466692322213095,
+      "learning_rate": 2.58926964714406e-07,
+      "loss": 0.947,
+      "step": 8133
+    },
+    {
+      "epoch": 0.8554563739860912,
+      "grad_norm": 3.13947290894236,
+      "learning_rate": 2.5855710858033096e-07,
+      "loss": 0.9944,
+      "step": 8134
+    },
+    {
+      "epoch": 0.8555615444279385,
+      "grad_norm": 2.2159814940956912,
+      "learning_rate": 2.581875023845512e-07,
+      "loss": 0.9693,
+      "step": 8135
+    },
+    {
+      "epoch": 0.8556667148697858,
+      "grad_norm": 2.261359755971855,
+      "learning_rate": 2.5781814616827936e-07,
+      "loss": 0.9711,
+      "step": 8136
+    },
+    {
+      "epoch": 0.8557718853116332,
+      "grad_norm": 2.1664017570010468,
+      "learning_rate": 2.57449039972702e-07,
+      "loss": 0.9714,
+      "step": 8137
+    },
+    {
+      "epoch": 0.8558770557534805,
+      "grad_norm": 2.2579137649584418,
+      "learning_rate": 2.5708018383897803e-07,
+      "loss": 0.9975,
+      "step": 8138
+    },
+    {
+      "epoch": 0.8559822261953278,
+      "grad_norm": 3.238109092541349,
+      "learning_rate": 2.567115778082366e-07,
+      "loss": 0.9784,
+      "step": 8139
+    },
+    {
+      "epoch": 0.8560873966371751,
+      "grad_norm": 2.4097012019064357,
+      "learning_rate": 2.563432219215814e-07,
+      "loss": 0.9779,
+      "step": 8140
+    },
+    {
+      "epoch": 0.8561925670790225,
+      "grad_norm": 2.2467448008644575,
+      "learning_rate": 2.559751162200855e-07,
+      "loss": 0.9641,
+      "step": 8141
+    },
+    {
+      "epoch": 0.8562977375208698,
+      "grad_norm": 2.3268805950761577,
+      "learning_rate": 2.556072607447965e-07,
+      "loss": 1.0105,
+      "step": 8142
+    },
+    {
+      "epoch": 0.8564029079627171,
+      "grad_norm": 2.887233445189877,
+      "learning_rate": 2.552396555367334e-07,
+      "loss": 0.9784,
+      "step": 8143
+    },
+    {
+      "epoch": 0.8565080784045644,
+      "grad_norm": 2.2652184394813246,
+      "learning_rate": 2.548723006368864e-07,
+      "loss": 0.9439,
+      "step": 8144
+    },
+    {
+      "epoch": 0.8566132488464118,
+      "grad_norm": 3.416084310322756,
+      "learning_rate": 2.545051960862188e-07,
+      "loss": 1.0174,
+      "step": 8145
+    },
+    {
+      "epoch": 0.856718419288259,
+      "grad_norm": 2.2619883051129888,
+      "learning_rate": 2.541383419256646e-07,
+      "loss": 0.9155,
+      "step": 8146
+    },
+    {
+      "epoch": 0.8568235897301063,
+      "grad_norm": 3.2034627263132034,
+      "learning_rate": 2.5377173819613274e-07,
+      "loss": 0.953,
+      "step": 8147
+    },
+    {
+      "epoch": 0.8569287601719536,
+      "grad_norm": 2.3308420908834213,
+      "learning_rate": 2.53405384938501e-07,
+      "loss": 0.977,
+      "step": 8148
+    },
+    {
+      "epoch": 0.857033930613801,
+      "grad_norm": 2.549797506834715,
+      "learning_rate": 2.530392821936212e-07,
+      "loss": 0.9436,
+      "step": 8149
+    },
+    {
+      "epoch": 0.8571391010556483,
+      "grad_norm": 2.001168695818676,
+      "learning_rate": 2.526734300023162e-07,
+      "loss": 0.965,
+      "step": 8150
+    },
+    {
+      "epoch": 0.8572442714974956,
+      "grad_norm": 2.9838680004948293,
+      "learning_rate": 2.5230782840538147e-07,
+      "loss": 0.9686,
+      "step": 8151
+    },
+    {
+      "epoch": 0.8573494419393429,
+      "grad_norm": 2.0859980624244487,
+      "learning_rate": 2.5194247744358523e-07,
+      "loss": 0.9598,
+      "step": 8152
+    },
+    {
+      "epoch": 0.8574546123811903,
+      "grad_norm": 1.6033395922396991,
+      "learning_rate": 2.5157737715766544e-07,
+      "loss": 0.9206,
+      "step": 8153
+    },
+    {
+      "epoch": 0.8575597828230376,
+      "grad_norm": 2.47393713284398,
+      "learning_rate": 2.5121252758833537e-07,
+      "loss": 0.9946,
+      "step": 8154
+    },
+    {
+      "epoch": 0.8576649532648849,
+      "grad_norm": 2.375844708577336,
+      "learning_rate": 2.508479287762769e-07,
+      "loss": 0.9725,
+      "step": 8155
+    },
+    {
+      "epoch": 0.8577701237067322,
+      "grad_norm": 1.907812917306332,
+      "learning_rate": 2.504835807621464e-07,
+      "loss": 0.9562,
+      "step": 8156
+    },
+    {
+      "epoch": 0.8578752941485795,
+      "grad_norm": 1.967540450625011,
+      "learning_rate": 2.501194835865717e-07,
+      "loss": 0.9612,
+      "step": 8157
+    },
+    {
+      "epoch": 0.8579804645904269,
+      "grad_norm": 2.9450000337243147,
+      "learning_rate": 2.4975563729015244e-07,
+      "loss": 0.9839,
+      "step": 8158
+    },
+    {
+      "epoch": 0.8580856350322742,
+      "grad_norm": 2.85718562948114,
+      "learning_rate": 2.493920419134604e-07,
+      "loss": 0.9857,
+      "step": 8159
+    },
+    {
+      "epoch": 0.8581908054741215,
+      "grad_norm": 2.384457833137985,
+      "learning_rate": 2.4902869749703797e-07,
+      "loss": 0.9708,
+      "step": 8160
+    },
+    {
+      "epoch": 0.8582959759159688,
+      "grad_norm": 3.4120765102011985,
+      "learning_rate": 2.4866560408140284e-07,
+      "loss": 0.9713,
+      "step": 8161
+    },
+    {
+      "epoch": 0.8584011463578162,
+      "grad_norm": 2.19415436252959,
+      "learning_rate": 2.483027617070413e-07,
+      "loss": 0.953,
+      "step": 8162
+    },
+    {
+      "epoch": 0.8585063167996635,
+      "grad_norm": 2.8855737475987264,
+      "learning_rate": 2.479401704144144e-07,
+      "loss": 0.9333,
+      "step": 8163
+    },
+    {
+      "epoch": 0.8586114872415108,
+      "grad_norm": 2.3336198104728374,
+      "learning_rate": 2.4757783024395244e-07,
+      "loss": 0.9902,
+      "step": 8164
+    },
+    {
+      "epoch": 0.8587166576833581,
+      "grad_norm": 2.645052818170446,
+      "learning_rate": 2.472157412360596e-07,
+      "loss": 0.9993,
+      "step": 8165
+    },
+    {
+      "epoch": 0.8588218281252055,
+      "grad_norm": 2.332866176403049,
+      "learning_rate": 2.4685390343111265e-07,
+      "loss": 0.9832,
+      "step": 8166
+    },
+    {
+      "epoch": 0.8589269985670527,
+      "grad_norm": 2.8089915163327266,
+      "learning_rate": 2.46492316869458e-07,
+      "loss": 0.9792,
+      "step": 8167
+    },
+    {
+      "epoch": 0.8590321690089,
+      "grad_norm": 2.459421993014172,
+      "learning_rate": 2.461309815914162e-07,
+      "loss": 1.0708,
+      "step": 8168
+    },
+    {
+      "epoch": 0.8591373394507473,
+      "grad_norm": 2.3521339437040734,
+      "learning_rate": 2.45769897637278e-07,
+      "loss": 1.0167,
+      "step": 8169
+    },
+    {
+      "epoch": 0.8592425098925947,
+      "grad_norm": 3.0044324465369336,
+      "learning_rate": 2.4540906504730814e-07,
+      "loss": 0.9617,
+      "step": 8170
+    },
+    {
+      "epoch": 0.859347680334442,
+      "grad_norm": 2.554374282868245,
+      "learning_rate": 2.450484838617417e-07,
+      "loss": 0.9495,
+      "step": 8171
+    },
+    {
+      "epoch": 0.8594528507762893,
+      "grad_norm": 2.387412844879945,
+      "learning_rate": 2.446881541207868e-07,
+      "loss": 0.9826,
+      "step": 8172
+    },
+    {
+      "epoch": 0.8595580212181366,
+      "grad_norm": 1.942667199406407,
+      "learning_rate": 2.4432807586462214e-07,
+      "loss": 0.9746,
+      "step": 8173
+    },
+    {
+      "epoch": 0.859663191659984,
+      "grad_norm": 2.590230625555086,
+      "learning_rate": 2.4396824913339946e-07,
+      "loss": 0.9675,
+      "step": 8174
+    },
+    {
+      "epoch": 0.8597683621018313,
+      "grad_norm": 2.544725569198336,
+      "learning_rate": 2.436086739672433e-07,
+      "loss": 0.9573,
+      "step": 8175
+    },
+    {
+      "epoch": 0.8598735325436786,
+      "grad_norm": 2.875401976015427,
+      "learning_rate": 2.4324935040624747e-07,
+      "loss": 0.9746,
+      "step": 8176
+    },
+    {
+      "epoch": 0.8599787029855259,
+      "grad_norm": 2.330210767863903,
+      "learning_rate": 2.4289027849048076e-07,
+      "loss": 0.9509,
+      "step": 8177
+    },
+    {
+      "epoch": 0.8600838734273732,
+      "grad_norm": 3.8731187149712456,
+      "learning_rate": 2.4253145825998134e-07,
+      "loss": 0.972,
+      "step": 8178
+    },
+    {
+      "epoch": 0.8601890438692206,
+      "grad_norm": 2.4303181220029746,
+      "learning_rate": 2.4217288975476094e-07,
+      "loss": 0.9954,
+      "step": 8179
+    },
+    {
+      "epoch": 0.8602942143110679,
+      "grad_norm": 1.9710335180041265,
+      "learning_rate": 2.418145730148033e-07,
+      "loss": 0.957,
+      "step": 8180
+    },
+    {
+      "epoch": 0.8603993847529152,
+      "grad_norm": 2.7525905486987288,
+      "learning_rate": 2.414565080800624e-07,
+      "loss": 0.9754,
+      "step": 8181
+    },
+    {
+      "epoch": 0.8605045551947625,
+      "grad_norm": 3.292704941422295,
+      "learning_rate": 2.4109869499046647e-07,
+      "loss": 0.9751,
+      "step": 8182
+    },
+    {
+      "epoch": 0.8606097256366099,
+      "grad_norm": 2.165266902747974,
+      "learning_rate": 2.4074113378591254e-07,
+      "loss": 0.9757,
+      "step": 8183
+    },
+    {
+      "epoch": 0.8607148960784572,
+      "grad_norm": 2.4020231317592806,
+      "learning_rate": 2.4038382450627396e-07,
+      "loss": 0.9993,
+      "step": 8184
+    },
+    {
+      "epoch": 0.8608200665203045,
+      "grad_norm": 2.5604039867867816,
+      "learning_rate": 2.400267671913917e-07,
+      "loss": 1.0048,
+      "step": 8185
+    },
+    {
+      "epoch": 0.8609252369621518,
+      "grad_norm": 1.8354917739415633,
+      "learning_rate": 2.3966996188108133e-07,
+      "loss": 0.9127,
+      "step": 8186
+    },
+    {
+      "epoch": 0.861030407403999,
+      "grad_norm": 2.865868216648736,
+      "learning_rate": 2.3931340861512884e-07,
+      "loss": 0.9865,
+      "step": 8187
+    },
+    {
+      "epoch": 0.8611355778458464,
+      "grad_norm": 1.9746436738041568,
+      "learning_rate": 2.389571074332928e-07,
+      "loss": 0.9466,
+      "step": 8188
+    },
+    {
+      "epoch": 0.8612407482876937,
+      "grad_norm": 3.3252842390412796,
+      "learning_rate": 2.38601058375304e-07,
+      "loss": 1.0068,
+      "step": 8189
+    },
+    {
+      "epoch": 0.861345918729541,
+      "grad_norm": 2.847887785313545,
+      "learning_rate": 2.382452614808642e-07,
+      "loss": 0.999,
+      "step": 8190
+    },
+    {
+      "epoch": 0.8614510891713884,
+      "grad_norm": 1.955383199190147,
+      "learning_rate": 2.3788971678964794e-07,
+      "loss": 0.9989,
+      "step": 8191
+    },
+    {
+      "epoch": 0.8615562596132357,
+      "grad_norm": 2.285791537624384,
+      "learning_rate": 2.3753442434129998e-07,
+      "loss": 0.9703,
+      "step": 8192
+    },
+    {
+      "epoch": 0.861661430055083,
+      "grad_norm": 2.6866877107259026,
+      "learning_rate": 2.3717938417543995e-07,
+      "loss": 0.9215,
+      "step": 8193
+    },
+    {
+      "epoch": 0.8617666004969303,
+      "grad_norm": 2.7588675656238677,
+      "learning_rate": 2.3682459633165704e-07,
+      "loss": 0.974,
+      "step": 8194
+    },
+    {
+      "epoch": 0.8618717709387776,
+      "grad_norm": 3.3460239543274843,
+      "learning_rate": 2.3647006084951214e-07,
+      "loss": 0.9426,
+      "step": 8195
+    },
+    {
+      "epoch": 0.861976941380625,
+      "grad_norm": 2.3650104551290685,
+      "learning_rate": 2.3611577776853965e-07,
+      "loss": 0.9484,
+      "step": 8196
+    },
+    {
+      "epoch": 0.8620821118224723,
+      "grad_norm": 2.834027735071727,
+      "learning_rate": 2.3576174712824335e-07,
+      "loss": 0.9879,
+      "step": 8197
+    },
+    {
+      "epoch": 0.8621872822643196,
+      "grad_norm": 2.2266254740819966,
+      "learning_rate": 2.354079689681024e-07,
+      "loss": 1.0312,
+      "step": 8198
+    },
+    {
+      "epoch": 0.862292452706167,
+      "grad_norm": 2.550002403776596,
+      "learning_rate": 2.350544433275645e-07,
+      "loss": 0.9572,
+      "step": 8199
+    },
+    {
+      "epoch": 0.8623976231480143,
+      "grad_norm": 3.3029388049198825,
+      "learning_rate": 2.3470117024605138e-07,
+      "loss": 0.9695,
+      "step": 8200
+    },
+    {
+      "epoch": 0.8625027935898616,
+      "grad_norm": 2.7620071406830635,
+      "learning_rate": 2.3434814976295462e-07,
+      "loss": 0.9729,
+      "step": 8201
+    },
+    {
+      "epoch": 0.8626079640317089,
+      "grad_norm": 2.279717835201499,
+      "learning_rate": 2.3399538191763937e-07,
+      "loss": 1.0069,
+      "step": 8202
+    },
+    {
+      "epoch": 0.8627131344735562,
+      "grad_norm": 2.509494166741847,
+      "learning_rate": 2.3364286674944254e-07,
+      "loss": 0.9409,
+      "step": 8203
+    },
+    {
+      "epoch": 0.8628183049154036,
+      "grad_norm": 2.7206541859104267,
+      "learning_rate": 2.3329060429767124e-07,
+      "loss": 0.9442,
+      "step": 8204
+    },
+    {
+      "epoch": 0.8629234753572509,
+      "grad_norm": 2.372820235661989,
+      "learning_rate": 2.329385946016066e-07,
+      "loss": 0.9734,
+      "step": 8205
+    },
+    {
+      "epoch": 0.8630286457990982,
+      "grad_norm": 2.6935312672829874,
+      "learning_rate": 2.325868377004986e-07,
+      "loss": 0.9754,
+      "step": 8206
+    },
+    {
+      "epoch": 0.8631338162409454,
+      "grad_norm": 2.1552030244555587,
+      "learning_rate": 2.3223533363357308e-07,
+      "loss": 0.9789,
+      "step": 8207
+    },
+    {
+      "epoch": 0.8632389866827928,
+      "grad_norm": 2.412292284565908,
+      "learning_rate": 2.3188408244002424e-07,
+      "loss": 0.9809,
+      "step": 8208
+    },
+    {
+      "epoch": 0.8633441571246401,
+      "grad_norm": 2.088983007031558,
+      "learning_rate": 2.3153308415901993e-07,
+      "loss": 1.0039,
+      "step": 8209
+    },
+    {
+      "epoch": 0.8634493275664874,
+      "grad_norm": 2.4663761465549547,
+      "learning_rate": 2.3118233882969854e-07,
+      "loss": 1.0072,
+      "step": 8210
+    },
+    {
+      "epoch": 0.8635544980083347,
+      "grad_norm": 2.147148470340049,
+      "learning_rate": 2.3083184649117046e-07,
+      "loss": 0.9939,
+      "step": 8211
+    },
+    {
+      "epoch": 0.863659668450182,
+      "grad_norm": 2.709658500142506,
+      "learning_rate": 2.3048160718252e-07,
+      "loss": 0.9834,
+      "step": 8212
+    },
+    {
+      "epoch": 0.8637648388920294,
+      "grad_norm": 2.7515549275388804,
+      "learning_rate": 2.3013162094279977e-07,
+      "loss": 1.009,
+      "step": 8213
+    },
+    {
+      "epoch": 0.8638700093338767,
+      "grad_norm": 2.999999695153477,
+      "learning_rate": 2.297818878110375e-07,
+      "loss": 0.918,
+      "step": 8214
+    },
+    {
+      "epoch": 0.863975179775724,
+      "grad_norm": 2.8176326824155358,
+      "learning_rate": 2.2943240782623e-07,
+      "loss": 0.9514,
+      "step": 8215
+    },
+    {
+      "epoch": 0.8640803502175713,
+      "grad_norm": 2.673202502494635,
+      "learning_rate": 2.2908318102734724e-07,
+      "loss": 1.003,
+      "step": 8216
+    },
+    {
+      "epoch": 0.8641855206594187,
+      "grad_norm": 2.9697524493518253,
+      "learning_rate": 2.2873420745333163e-07,
+      "loss": 0.9965,
+      "step": 8217
+    },
+    {
+      "epoch": 0.864290691101266,
+      "grad_norm": 2.3065886709900365,
+      "learning_rate": 2.283854871430949e-07,
+      "loss": 0.9499,
+      "step": 8218
+    },
+    {
+      "epoch": 0.8643958615431133,
+      "grad_norm": 2.3908258377099565,
+      "learning_rate": 2.2803702013552364e-07,
+      "loss": 0.9679,
+      "step": 8219
+    },
+    {
+      "epoch": 0.8645010319849606,
+      "grad_norm": 2.586564468311502,
+      "learning_rate": 2.2768880646947268e-07,
+      "loss": 0.9946,
+      "step": 8220
+    },
+    {
+      "epoch": 0.864606202426808,
+      "grad_norm": 3.179651943448657,
+      "learning_rate": 2.2734084618377285e-07,
+      "loss": 1.0029,
+      "step": 8221
+    },
+    {
+      "epoch": 0.8647113728686553,
+      "grad_norm": 3.5059554764194156,
+      "learning_rate": 2.2699313931722284e-07,
+      "loss": 1.0056,
+      "step": 8222
+    },
+    {
+      "epoch": 0.8648165433105026,
+      "grad_norm": 2.3482968719594823,
+      "learning_rate": 2.2664568590859525e-07,
+      "loss": 0.9858,
+      "step": 8223
+    },
+    {
+      "epoch": 0.8649217137523499,
+      "grad_norm": 2.786091616547652,
+      "learning_rate": 2.2629848599663357e-07,
+      "loss": 0.9675,
+      "step": 8224
+    },
+    {
+      "epoch": 0.8650268841941973,
+      "grad_norm": 2.151802230862855,
+      "learning_rate": 2.2595153962005345e-07,
+      "loss": 1.0044,
+      "step": 8225
+    },
+    {
+      "epoch": 0.8651320546360446,
+      "grad_norm": 2.2954772594860793,
+      "learning_rate": 2.2560484681754258e-07,
+      "loss": 1.0249,
+      "step": 8226
+    },
+    {
+      "epoch": 0.8652372250778919,
+      "grad_norm": 2.4848587884057065,
+      "learning_rate": 2.2525840762775863e-07,
+      "loss": 0.9688,
+      "step": 8227
+    },
+    {
+      "epoch": 0.8653423955197391,
+      "grad_norm": 2.825905090752021,
+      "learning_rate": 2.2491222208933377e-07,
+      "loss": 1.004,
+      "step": 8228
+    },
+    {
+      "epoch": 0.8654475659615865,
+      "grad_norm": 2.6469755844197635,
+      "learning_rate": 2.245662902408688e-07,
+      "loss": 0.999,
+      "step": 8229
+    },
+    {
+      "epoch": 0.8655527364034338,
+      "grad_norm": 3.240033136066609,
+      "learning_rate": 2.2422061212093947e-07,
+      "loss": 0.9899,
+      "step": 8230
+    },
+    {
+      "epoch": 0.8656579068452811,
+      "grad_norm": 2.8389209244754787,
+      "learning_rate": 2.238751877680906e-07,
+      "loss": 1.0278,
+      "step": 8231
+    },
+    {
+      "epoch": 0.8657630772871284,
+      "grad_norm": 2.6429507672792374,
+      "learning_rate": 2.2353001722083962e-07,
+      "loss": 0.986,
+      "step": 8232
+    },
+    {
+      "epoch": 0.8658682477289757,
+      "grad_norm": 2.0796610902282047,
+      "learning_rate": 2.2318510051767632e-07,
+      "loss": 0.9746,
+      "step": 8233
+    },
+    {
+      "epoch": 0.8659734181708231,
+      "grad_norm": 2.1854334963011137,
+      "learning_rate": 2.2284043769706026e-07,
+      "loss": 0.9798,
+      "step": 8234
+    },
+    {
+      "epoch": 0.8660785886126704,
+      "grad_norm": 2.154239959742983,
+      "learning_rate": 2.2249602879742594e-07,
+      "loss": 0.9868,
+      "step": 8235
+    },
+    {
+      "epoch": 0.8661837590545177,
+      "grad_norm": 2.6042428551850563,
+      "learning_rate": 2.2215187385717623e-07,
+      "loss": 1.0048,
+      "step": 8236
+    },
+    {
+      "epoch": 0.866288929496365,
+      "grad_norm": 2.3168200385005435,
+      "learning_rate": 2.2180797291468825e-07,
+      "loss": 0.9696,
+      "step": 8237
+    },
+    {
+      "epoch": 0.8663940999382124,
+      "grad_norm": 2.612002220024735,
+      "learning_rate": 2.2146432600830799e-07,
+      "loss": 0.9707,
+      "step": 8238
+    },
+    {
+      "epoch": 0.8664992703800597,
+      "grad_norm": 2.7310391889877494,
+      "learning_rate": 2.2112093317635587e-07,
+      "loss": 0.9626,
+      "step": 8239
+    },
+    {
+      "epoch": 0.866604440821907,
+      "grad_norm": 2.1973508769452152,
+      "learning_rate": 2.207777944571232e-07,
+      "loss": 0.9735,
+      "step": 8240
+    },
+    {
+      "epoch": 0.8667096112637543,
+      "grad_norm": 2.184444168526134,
+      "learning_rate": 2.2043490988887168e-07,
+      "loss": 0.9651,
+      "step": 8241
+    },
+    {
+      "epoch": 0.8668147817056017,
+      "grad_norm": 1.7840753769682516,
+      "learning_rate": 2.200922795098362e-07,
+      "loss": 0.9442,
+      "step": 8242
+    },
+    {
+      "epoch": 0.866919952147449,
+      "grad_norm": 2.612245040597428,
+      "learning_rate": 2.1974990335822179e-07,
+      "loss": 0.9973,
+      "step": 8243
+    },
+    {
+      "epoch": 0.8670251225892963,
+      "grad_norm": 2.7915387255241346,
+      "learning_rate": 2.1940778147220758e-07,
+      "loss": 0.9892,
+      "step": 8244
+    },
+    {
+      "epoch": 0.8671302930311436,
+      "grad_norm": 2.104198732577539,
+      "learning_rate": 2.1906591388994148e-07,
+      "loss": 0.9436,
+      "step": 8245
+    },
+    {
+      "epoch": 0.867235463472991,
+      "grad_norm": 2.381480939039506,
+      "learning_rate": 2.187243006495454e-07,
+      "loss": 1.0353,
+      "step": 8246
+    },
+    {
+      "epoch": 0.8673406339148383,
+      "grad_norm": 2.7229634312983917,
+      "learning_rate": 2.1838294178911146e-07,
+      "loss": 1.0069,
+      "step": 8247
+    },
+    {
+      "epoch": 0.8674458043566855,
+      "grad_norm": 2.3789881685811154,
+      "learning_rate": 2.1804183734670277e-07,
+      "loss": 0.9485,
+      "step": 8248
+    },
+    {
+      "epoch": 0.8675509747985328,
+      "grad_norm": 2.5490016088242125,
+      "learning_rate": 2.1770098736035694e-07,
+      "loss": 0.9878,
+      "step": 8249
+    },
+    {
+      "epoch": 0.8676561452403801,
+      "grad_norm": 2.8862833601914764,
+      "learning_rate": 2.1736039186808e-07,
+      "loss": 0.988,
+      "step": 8250
+    },
+    {
+      "epoch": 0.8677613156822275,
+      "grad_norm": 2.4021252306313747,
+      "learning_rate": 2.170200509078521e-07,
+      "loss": 0.9925,
+      "step": 8251
+    },
+    {
+      "epoch": 0.8678664861240748,
+      "grad_norm": 2.489969509776828,
+      "learning_rate": 2.166799645176229e-07,
+      "loss": 1.0052,
+      "step": 8252
+    },
+    {
+      "epoch": 0.8679716565659221,
+      "grad_norm": 1.9153232869677428,
+      "learning_rate": 2.1634013273531512e-07,
+      "loss": 0.9971,
+      "step": 8253
+    },
+    {
+      "epoch": 0.8680768270077694,
+      "grad_norm": 2.0476262837462156,
+      "learning_rate": 2.160005555988229e-07,
+      "loss": 0.9905,
+      "step": 8254
+    },
+    {
+      "epoch": 0.8681819974496168,
+      "grad_norm": 3.0949990892144665,
+      "learning_rate": 2.1566123314601118e-07,
+      "loss": 0.9642,
+      "step": 8255
+    },
+    {
+      "epoch": 0.8682871678914641,
+      "grad_norm": 2.359080168169369,
+      "learning_rate": 2.1532216541471778e-07,
+      "loss": 1.0135,
+      "step": 8256
+    },
+    {
+      "epoch": 0.8683923383333114,
+      "grad_norm": 1.66683761103825,
+      "learning_rate": 2.1498335244275e-07,
+      "loss": 0.9266,
+      "step": 8257
+    },
+    {
+      "epoch": 0.8684975087751587,
+      "grad_norm": 2.133569622507562,
+      "learning_rate": 2.1464479426789005e-07,
+      "loss": 0.961,
+      "step": 8258
+    },
+    {
+      "epoch": 0.8686026792170061,
+      "grad_norm": 2.588002312127146,
+      "learning_rate": 2.143064909278883e-07,
+      "loss": 0.9796,
+      "step": 8259
+    },
+    {
+      "epoch": 0.8687078496588534,
+      "grad_norm": 3.192126932345157,
+      "learning_rate": 2.1396844246046904e-07,
+      "loss": 1.006,
+      "step": 8260
+    },
+    {
+      "epoch": 0.8688130201007007,
+      "grad_norm": 2.4896171740508946,
+      "learning_rate": 2.1363064890332658e-07,
+      "loss": 0.9849,
+      "step": 8261
+    },
+    {
+      "epoch": 0.868918190542548,
+      "grad_norm": 2.7731731799976305,
+      "learning_rate": 2.1329311029412796e-07,
+      "loss": 0.9458,
+      "step": 8262
+    },
+    {
+      "epoch": 0.8690233609843954,
+      "grad_norm": 2.722330994924211,
+      "learning_rate": 2.1295582667051173e-07,
+      "loss": 1.0028,
+      "step": 8263
+    },
+    {
+      "epoch": 0.8691285314262427,
+      "grad_norm": 2.1961408410319008,
+      "learning_rate": 2.1261879807008667e-07,
+      "loss": 1.0216,
+      "step": 8264
+    },
+    {
+      "epoch": 0.86923370186809,
+      "grad_norm": 2.123750174696564,
+      "learning_rate": 2.1228202453043522e-07,
+      "loss": 1.01,
+      "step": 8265
+    },
+    {
+      "epoch": 0.8693388723099373,
+      "grad_norm": 2.6809762221376463,
+      "learning_rate": 2.1194550608910902e-07,
+      "loss": 1.0123,
+      "step": 8266
+    },
+    {
+      "epoch": 0.8694440427517847,
+      "grad_norm": 2.5457872775140986,
+      "learning_rate": 2.1160924278363333e-07,
+      "loss": 0.9877,
+      "step": 8267
+    },
+    {
+      "epoch": 0.8695492131936319,
+      "grad_norm": 2.085122359184593,
+      "learning_rate": 2.1127323465150422e-07,
+      "loss": 0.9743,
+      "step": 8268
+    },
+    {
+      "epoch": 0.8696543836354792,
+      "grad_norm": 2.146743192267201,
+      "learning_rate": 2.1093748173018846e-07,
+      "loss": 0.9686,
+      "step": 8269
+    },
+    {
+      "epoch": 0.8697595540773265,
+      "grad_norm": 2.3756277033536812,
+      "learning_rate": 2.106019840571255e-07,
+      "loss": 0.954,
+      "step": 8270
+    },
+    {
+      "epoch": 0.8698647245191738,
+      "grad_norm": 1.9419516360185134,
+      "learning_rate": 2.1026674166972627e-07,
+      "loss": 0.9669,
+      "step": 8271
+    },
+    {
+      "epoch": 0.8699698949610212,
+      "grad_norm": 3.0111258367329397,
+      "learning_rate": 2.099317546053728e-07,
+      "loss": 1.0088,
+      "step": 8272
+    },
+    {
+      "epoch": 0.8700750654028685,
+      "grad_norm": 2.3215678619891364,
+      "learning_rate": 2.09597022901418e-07,
+      "loss": 0.9994,
+      "step": 8273
+    },
+    {
+      "epoch": 0.8701802358447158,
+      "grad_norm": 2.866407727751618,
+      "learning_rate": 2.0926254659518835e-07,
+      "loss": 0.9804,
+      "step": 8274
+    },
+    {
+      "epoch": 0.8702854062865631,
+      "grad_norm": 2.189238826195812,
+      "learning_rate": 2.0892832572397935e-07,
+      "loss": 0.9907,
+      "step": 8275
+    },
+    {
+      "epoch": 0.8703905767284105,
+      "grad_norm": 1.9925032781875867,
+      "learning_rate": 2.0859436032505954e-07,
+      "loss": 1.0121,
+      "step": 8276
+    },
+    {
+      "epoch": 0.8704957471702578,
+      "grad_norm": 2.0387873295462744,
+      "learning_rate": 2.0826065043566935e-07,
+      "loss": 0.9812,
+      "step": 8277
+    },
+    {
+      "epoch": 0.8706009176121051,
+      "grad_norm": 1.838300067613648,
+      "learning_rate": 2.0792719609301904e-07,
+      "loss": 0.9273,
+      "step": 8278
+    },
+    {
+      "epoch": 0.8707060880539524,
+      "grad_norm": 2.587747678789586,
+      "learning_rate": 2.075939973342922e-07,
+      "loss": 0.9786,
+      "step": 8279
+    },
+    {
+      "epoch": 0.8708112584957998,
+      "grad_norm": 3.083618118123322,
+      "learning_rate": 2.0726105419664188e-07,
+      "loss": 0.9509,
+      "step": 8280
+    },
+    {
+      "epoch": 0.8709164289376471,
+      "grad_norm": 2.1215733170081488,
+      "learning_rate": 2.0692836671719536e-07,
+      "loss": 0.9764,
+      "step": 8281
+    },
+    {
+      "epoch": 0.8710215993794944,
+      "grad_norm": 2.305156580274503,
+      "learning_rate": 2.065959349330493e-07,
+      "loss": 1.0078,
+      "step": 8282
+    },
+    {
+      "epoch": 0.8711267698213417,
+      "grad_norm": 2.6860204494280153,
+      "learning_rate": 2.0626375888127187e-07,
+      "loss": 0.996,
+      "step": 8283
+    },
+    {
+      "epoch": 0.8712319402631891,
+      "grad_norm": 2.897921994981041,
+      "learning_rate": 2.0593183859890369e-07,
+      "loss": 0.9731,
+      "step": 8284
+    },
+    {
+      "epoch": 0.8713371107050364,
+      "grad_norm": 3.5559919783151974,
+      "learning_rate": 2.0560017412295658e-07,
+      "loss": 0.9663,
+      "step": 8285
+    },
+    {
+      "epoch": 0.8714422811468837,
+      "grad_norm": 2.4981822525310826,
+      "learning_rate": 2.0526876549041368e-07,
+      "loss": 0.9962,
+      "step": 8286
+    },
+    {
+      "epoch": 0.871547451588731,
+      "grad_norm": 2.1707524269042926,
+      "learning_rate": 2.0493761273822937e-07,
+      "loss": 1.0296,
+      "step": 8287
+    },
+    {
+      "epoch": 0.8716526220305784,
+      "grad_norm": 1.9671933144654712,
+      "learning_rate": 2.046067159033302e-07,
+      "loss": 1.0113,
+      "step": 8288
+    },
+    {
+      "epoch": 0.8717577924724256,
+      "grad_norm": 1.8601153938979094,
+      "learning_rate": 2.0427607502261303e-07,
+      "loss": 1.0016,
+      "step": 8289
+    },
+    {
+      "epoch": 0.8718629629142729,
+      "grad_norm": 3.164956063667044,
+      "learning_rate": 2.039456901329473e-07,
+      "loss": 1.0117,
+      "step": 8290
+    },
+    {
+      "epoch": 0.8719681333561202,
+      "grad_norm": 2.629745086363934,
+      "learning_rate": 2.0361556127117404e-07,
+      "loss": 0.9834,
+      "step": 8291
+    },
+    {
+      "epoch": 0.8720733037979675,
+      "grad_norm": 2.1561302946827907,
+      "learning_rate": 2.0328568847410413e-07,
+      "loss": 0.9296,
+      "step": 8292
+    },
+    {
+      "epoch": 0.8721784742398149,
+      "grad_norm": 2.281348519655876,
+      "learning_rate": 2.0295607177852146e-07,
+      "loss": 0.9825,
+      "step": 8293
+    },
+    {
+      "epoch": 0.8722836446816622,
+      "grad_norm": 3.1673280335921876,
+      "learning_rate": 2.0262671122118078e-07,
+      "loss": 1.0055,
+      "step": 8294
+    },
+    {
+      "epoch": 0.8723888151235095,
+      "grad_norm": 2.292148446422658,
+      "learning_rate": 2.0229760683880884e-07,
+      "loss": 0.9556,
+      "step": 8295
+    },
+    {
+      "epoch": 0.8724939855653568,
+      "grad_norm": 3.1357836739117175,
+      "learning_rate": 2.0196875866810266e-07,
+      "loss": 1.0117,
+      "step": 8296
+    },
+    {
+      "epoch": 0.8725991560072042,
+      "grad_norm": 2.519946732312392,
+      "learning_rate": 2.0164016674573185e-07,
+      "loss": 0.9937,
+      "step": 8297
+    },
+    {
+      "epoch": 0.8727043264490515,
+      "grad_norm": 2.564131547407527,
+      "learning_rate": 2.0131183110833646e-07,
+      "loss": 0.9503,
+      "step": 8298
+    },
+    {
+      "epoch": 0.8728094968908988,
+      "grad_norm": 2.280967440361249,
+      "learning_rate": 2.0098375179252867e-07,
+      "loss": 0.9791,
+      "step": 8299
+    },
+    {
+      "epoch": 0.8729146673327461,
+      "grad_norm": 2.354741913861808,
+      "learning_rate": 2.006559288348922e-07,
+      "loss": 0.9536,
+      "step": 8300
+    },
+    {
+      "epoch": 0.8730198377745935,
+      "grad_norm": 3.5165919413572926,
+      "learning_rate": 2.003283622719815e-07,
+      "loss": 0.9616,
+      "step": 8301
+    },
+    {
+      "epoch": 0.8731250082164408,
+      "grad_norm": 2.355749665234238,
+      "learning_rate": 2.0000105214032313e-07,
+      "loss": 0.9959,
+      "step": 8302
+    },
+    {
+      "epoch": 0.8732301786582881,
+      "grad_norm": 2.5396746122652822,
+      "learning_rate": 1.99673998476414e-07,
+      "loss": 0.9603,
+      "step": 8303
+    },
+    {
+      "epoch": 0.8733353491001354,
+      "grad_norm": 2.314335599258069,
+      "learning_rate": 1.9934720131672357e-07,
+      "loss": 0.9933,
+      "step": 8304
+    },
+    {
+      "epoch": 0.8734405195419828,
+      "grad_norm": 2.4455577788280523,
+      "learning_rate": 1.99020660697693e-07,
+      "loss": 1.0243,
+      "step": 8305
+    },
+    {
+      "epoch": 0.8735456899838301,
+      "grad_norm": 2.393852124530319,
+      "learning_rate": 1.9869437665573255e-07,
+      "loss": 0.9855,
+      "step": 8306
+    },
+    {
+      "epoch": 0.8736508604256774,
+      "grad_norm": 1.9162658676831326,
+      "learning_rate": 1.9836834922722648e-07,
+      "loss": 0.9762,
+      "step": 8307
+    },
+    {
+      "epoch": 0.8737560308675247,
+      "grad_norm": 2.6740161604492907,
+      "learning_rate": 1.980425784485293e-07,
+      "loss": 0.9674,
+      "step": 8308
+    },
+    {
+      "epoch": 0.873861201309372,
+      "grad_norm": 2.380114295557652,
+      "learning_rate": 1.97717064355967e-07,
+      "loss": 0.999,
+      "step": 8309
+    },
+    {
+      "epoch": 0.8739663717512193,
+      "grad_norm": 3.1166913813516883,
+      "learning_rate": 1.973918069858366e-07,
+      "loss": 0.9538,
+      "step": 8310
+    },
+    {
+      "epoch": 0.8740715421930666,
+      "grad_norm": 2.807493356835643,
+      "learning_rate": 1.9706680637440745e-07,
+      "loss": 0.998,
+      "step": 8311
+    },
+    {
+      "epoch": 0.8741767126349139,
+      "grad_norm": 1.8540453587981012,
+      "learning_rate": 1.9674206255791862e-07,
+      "loss": 0.9781,
+      "step": 8312
+    },
+    {
+      "epoch": 0.8742818830767612,
+      "grad_norm": 2.297351573084357,
+      "learning_rate": 1.9641757557258223e-07,
+      "loss": 1.0059,
+      "step": 8313
+    },
+    {
+      "epoch": 0.8743870535186086,
+      "grad_norm": 2.3795558043590797,
+      "learning_rate": 1.9609334545458132e-07,
+      "loss": 0.9783,
+      "step": 8314
+    },
+    {
+      "epoch": 0.8744922239604559,
+      "grad_norm": 2.108673183419314,
+      "learning_rate": 1.9576937224006942e-07,
+      "loss": 0.9516,
+      "step": 8315
+    },
+    {
+      "epoch": 0.8745973944023032,
+      "grad_norm": 1.9087942867413161,
+      "learning_rate": 1.954456559651724e-07,
+      "loss": 0.9661,
+      "step": 8316
+    },
+    {
+      "epoch": 0.8747025648441505,
+      "grad_norm": 2.3081851420364266,
+      "learning_rate": 1.9512219666598774e-07,
+      "loss": 0.9554,
+      "step": 8317
+    },
+    {
+      "epoch": 0.8748077352859979,
+      "grad_norm": 2.9514737071246038,
+      "learning_rate": 1.9479899437858246e-07,
+      "loss": 1.0265,
+      "step": 8318
+    },
+    {
+      "epoch": 0.8749129057278452,
+      "grad_norm": 2.74991764931712,
+      "learning_rate": 1.9447604913899715e-07,
+      "loss": 0.9832,
+      "step": 8319
+    },
+    {
+      "epoch": 0.8750180761696925,
+      "grad_norm": 3.2909066831788985,
+      "learning_rate": 1.941533609832419e-07,
+      "loss": 1.0034,
+      "step": 8320
+    },
+    {
+      "epoch": 0.8751232466115398,
+      "grad_norm": 2.9061285090024063,
+      "learning_rate": 1.9383092994729956e-07,
+      "loss": 0.9832,
+      "step": 8321
+    },
+    {
+      "epoch": 0.8752284170533872,
+      "grad_norm": 2.0846159205628747,
+      "learning_rate": 1.935087560671231e-07,
+      "loss": 0.9958,
+      "step": 8322
+    },
+    {
+      "epoch": 0.8753335874952345,
+      "grad_norm": 2.3654315207259167,
+      "learning_rate": 1.9318683937863846e-07,
+      "loss": 0.9666,
+      "step": 8323
+    },
+    {
+      "epoch": 0.8754387579370818,
+      "grad_norm": 2.3194539073632963,
+      "learning_rate": 1.9286517991774084e-07,
+      "loss": 0.9446,
+      "step": 8324
+    },
+    {
+      "epoch": 0.8755439283789291,
+      "grad_norm": 2.3802207813099314,
+      "learning_rate": 1.9254377772029847e-07,
+      "loss": 0.9544,
+      "step": 8325
+    },
+    {
+      "epoch": 0.8756490988207765,
+      "grad_norm": 1.8571233807580478,
+      "learning_rate": 1.9222263282214908e-07,
+      "loss": 0.9913,
+      "step": 8326
+    },
+    {
+      "epoch": 0.8757542692626238,
+      "grad_norm": 3.2909851431362447,
+      "learning_rate": 1.9190174525910404e-07,
+      "loss": 0.9587,
+      "step": 8327
+    },
+    {
+      "epoch": 0.8758594397044711,
+      "grad_norm": 2.4571652278790554,
+      "learning_rate": 1.9158111506694442e-07,
+      "loss": 0.9733,
+      "step": 8328
+    },
+    {
+      "epoch": 0.8759646101463183,
+      "grad_norm": 2.497778471844159,
+      "learning_rate": 1.9126074228142278e-07,
+      "loss": 0.9895,
+      "step": 8329
+    },
+    {
+      "epoch": 0.8760697805881656,
+      "grad_norm": 2.9691315953675703,
+      "learning_rate": 1.9094062693826298e-07,
+      "loss": 1.0181,
+      "step": 8330
+    },
+    {
+      "epoch": 0.876174951030013,
+      "grad_norm": 2.5598658628083664,
+      "learning_rate": 1.9062076907316097e-07,
+      "loss": 1.0124,
+      "step": 8331
+    },
+    {
+      "epoch": 0.8762801214718603,
+      "grad_norm": 3.447326481914661,
+      "learning_rate": 1.9030116872178317e-07,
+      "loss": 1.0096,
+      "step": 8332
+    },
+    {
+      "epoch": 0.8763852919137076,
+      "grad_norm": 2.4686568055463027,
+      "learning_rate": 1.8998182591976776e-07,
+      "loss": 1.0027,
+      "step": 8333
+    },
+    {
+      "epoch": 0.876490462355555,
+      "grad_norm": 3.8447359211448853,
+      "learning_rate": 1.8966274070272294e-07,
+      "loss": 1.0149,
+      "step": 8334
+    },
+    {
+      "epoch": 0.8765956327974023,
+      "grad_norm": 2.5490921354790927,
+      "learning_rate": 1.8934391310622996e-07,
+      "loss": 0.9944,
+      "step": 8335
+    },
+    {
+      "epoch": 0.8767008032392496,
+      "grad_norm": 3.1163782715259734,
+      "learning_rate": 1.8902534316584065e-07,
+      "loss": 0.9888,
+      "step": 8336
+    },
+    {
+      "epoch": 0.8768059736810969,
+      "grad_norm": 2.157509845767865,
+      "learning_rate": 1.88707030917078e-07,
+      "loss": 0.9813,
+      "step": 8337
+    },
+    {
+      "epoch": 0.8769111441229442,
+      "grad_norm": 2.3749513405972773,
+      "learning_rate": 1.883889763954358e-07,
+      "loss": 1.0061,
+      "step": 8338
+    },
+    {
+      "epoch": 0.8770163145647916,
+      "grad_norm": 3.006163202985406,
+      "learning_rate": 1.8807117963637988e-07,
+      "loss": 0.9441,
+      "step": 8339
+    },
+    {
+      "epoch": 0.8771214850066389,
+      "grad_norm": 2.940885641311102,
+      "learning_rate": 1.8775364067534773e-07,
+      "loss": 0.9773,
+      "step": 8340
+    },
+    {
+      "epoch": 0.8772266554484862,
+      "grad_norm": 2.95414648550892,
+      "learning_rate": 1.874363595477463e-07,
+      "loss": 1.0312,
+      "step": 8341
+    },
+    {
+      "epoch": 0.8773318258903335,
+      "grad_norm": 1.881654030719675,
+      "learning_rate": 1.871193362889559e-07,
+      "loss": 0.9565,
+      "step": 8342
+    },
+    {
+      "epoch": 0.8774369963321809,
+      "grad_norm": 2.7330908747232567,
+      "learning_rate": 1.8680257093432603e-07,
+      "loss": 0.9809,
+      "step": 8343
+    },
+    {
+      "epoch": 0.8775421667740282,
+      "grad_norm": 2.3063931397393067,
+      "learning_rate": 1.8648606351917925e-07,
+      "loss": 0.9502,
+      "step": 8344
+    },
+    {
+      "epoch": 0.8776473372158755,
+      "grad_norm": 2.6553680130796127,
+      "learning_rate": 1.8616981407880818e-07,
+      "loss": 1.0175,
+      "step": 8345
+    },
+    {
+      "epoch": 0.8777525076577228,
+      "grad_norm": 2.198765623782212,
+      "learning_rate": 1.8585382264847795e-07,
+      "loss": 0.9657,
+      "step": 8346
+    },
+    {
+      "epoch": 0.8778576780995702,
+      "grad_norm": 2.2406067363185103,
+      "learning_rate": 1.8553808926342286e-07,
+      "loss": 0.9648,
+      "step": 8347
+    },
+    {
+      "epoch": 0.8779628485414175,
+      "grad_norm": 2.6426330128760926,
+      "learning_rate": 1.852226139588506e-07,
+      "loss": 0.9899,
+      "step": 8348
+    },
+    {
+      "epoch": 0.8780680189832648,
+      "grad_norm": 2.241496860388617,
+      "learning_rate": 1.849073967699383e-07,
+      "loss": 1.0067,
+      "step": 8349
+    },
+    {
+      "epoch": 0.878173189425112,
+      "grad_norm": 2.481058599968573,
+      "learning_rate": 1.845924377318356e-07,
+      "loss": 0.9788,
+      "step": 8350
+    },
+    {
+      "epoch": 0.8782783598669593,
+      "grad_norm": 2.046724061828166,
+      "learning_rate": 1.8427773687966304e-07,
+      "loss": 1.0109,
+      "step": 8351
+    },
+    {
+      "epoch": 0.8783835303088067,
+      "grad_norm": 2.376223438068095,
+      "learning_rate": 1.8396329424851174e-07,
+      "loss": 0.9638,
+      "step": 8352
+    },
+    {
+      "epoch": 0.878488700750654,
+      "grad_norm": 2.2632885387151354,
+      "learning_rate": 1.836491098734447e-07,
+      "loss": 1.0035,
+      "step": 8353
+    },
+    {
+      "epoch": 0.8785938711925013,
+      "grad_norm": 2.655319309933802,
+      "learning_rate": 1.8333518378949617e-07,
+      "loss": 0.997,
+      "step": 8354
+    },
+    {
+      "epoch": 0.8786990416343486,
+      "grad_norm": 2.788664034342027,
+      "learning_rate": 1.830215160316709e-07,
+      "loss": 0.9811,
+      "step": 8355
+    },
+    {
+      "epoch": 0.878804212076196,
+      "grad_norm": 2.4388561721428292,
+      "learning_rate": 1.8270810663494591e-07,
+      "loss": 0.9727,
+      "step": 8356
+    },
+    {
+      "epoch": 0.8789093825180433,
+      "grad_norm": 3.2497459609934944,
+      "learning_rate": 1.8239495563426802e-07,
+      "loss": 0.9763,
+      "step": 8357
+    },
+    {
+      "epoch": 0.8790145529598906,
+      "grad_norm": 2.7302020542205385,
+      "learning_rate": 1.8208206306455616e-07,
+      "loss": 0.9581,
+      "step": 8358
+    },
+    {
+      "epoch": 0.8791197234017379,
+      "grad_norm": 2.96775947443148,
+      "learning_rate": 1.8176942896070083e-07,
+      "loss": 0.9633,
+      "step": 8359
+    },
+    {
+      "epoch": 0.8792248938435853,
+      "grad_norm": 2.3368503352333776,
+      "learning_rate": 1.8145705335756298e-07,
+      "loss": 1.0008,
+      "step": 8360
+    },
+    {
+      "epoch": 0.8793300642854326,
+      "grad_norm": 2.601467656711171,
+      "learning_rate": 1.8114493628997448e-07,
+      "loss": 0.9852,
+      "step": 8361
+    },
+    {
+      "epoch": 0.8794352347272799,
+      "grad_norm": 2.0109472927516876,
+      "learning_rate": 1.808330777927389e-07,
+      "loss": 0.9794,
+      "step": 8362
+    },
+    {
+      "epoch": 0.8795404051691272,
+      "grad_norm": 2.457219484542861,
+      "learning_rate": 1.8052147790063146e-07,
+      "loss": 0.9961,
+      "step": 8363
+    },
+    {
+      "epoch": 0.8796455756109746,
+      "grad_norm": 2.36559880573984,
+      "learning_rate": 1.8021013664839744e-07,
+      "loss": 1.0001,
+      "step": 8364
+    },
+    {
+      "epoch": 0.8797507460528219,
+      "grad_norm": 2.0670040655202695,
+      "learning_rate": 1.7989905407075404e-07,
+      "loss": 0.9953,
+      "step": 8365
+    },
+    {
+      "epoch": 0.8798559164946692,
+      "grad_norm": 2.4646404767978787,
+      "learning_rate": 1.795882302023891e-07,
+      "loss": 0.9685,
+      "step": 8366
+    },
+    {
+      "epoch": 0.8799610869365165,
+      "grad_norm": 2.0767215820236795,
+      "learning_rate": 1.7927766507796208e-07,
+      "loss": 0.9679,
+      "step": 8367
+    },
+    {
+      "epoch": 0.8800662573783639,
+      "grad_norm": 2.3891164659375628,
+      "learning_rate": 1.7896735873210364e-07,
+      "loss": 0.9736,
+      "step": 8368
+    },
+    {
+      "epoch": 0.8801714278202112,
+      "grad_norm": 3.110884943284728,
+      "learning_rate": 1.7865731119941498e-07,
+      "loss": 0.9882,
+      "step": 8369
+    },
+    {
+      "epoch": 0.8802765982620584,
+      "grad_norm": 2.5430378783802707,
+      "learning_rate": 1.78347522514469e-07,
+      "loss": 0.9486,
+      "step": 8370
+    },
+    {
+      "epoch": 0.8803817687039057,
+      "grad_norm": 2.3703437804084424,
+      "learning_rate": 1.780379927118095e-07,
+      "loss": 0.931,
+      "step": 8371
+    },
+    {
+      "epoch": 0.880486939145753,
+      "grad_norm": 2.2176388854921583,
+      "learning_rate": 1.7772872182595102e-07,
+      "loss": 0.961,
+      "step": 8372
+    },
+    {
+      "epoch": 0.8805921095876004,
+      "grad_norm": 2.930906441515779,
+      "learning_rate": 1.7741970989138046e-07,
+      "loss": 0.9894,
+      "step": 8373
+    },
+    {
+      "epoch": 0.8806972800294477,
+      "grad_norm": 2.376144556185852,
+      "learning_rate": 1.771109569425547e-07,
+      "loss": 0.9908,
+      "step": 8374
+    },
+    {
+      "epoch": 0.880802450471295,
+      "grad_norm": 2.3563178048249656,
+      "learning_rate": 1.7680246301390202e-07,
+      "loss": 0.9782,
+      "step": 8375
+    },
+    {
+      "epoch": 0.8809076209131423,
+      "grad_norm": 2.407079080129235,
+      "learning_rate": 1.7649422813982187e-07,
+      "loss": 1.0058,
+      "step": 8376
+    },
+    {
+      "epoch": 0.8810127913549897,
+      "grad_norm": 2.6815694041139504,
+      "learning_rate": 1.7618625235468507e-07,
+      "loss": 0.9664,
+      "step": 8377
+    },
+    {
+      "epoch": 0.881117961796837,
+      "grad_norm": 3.395370750987973,
+      "learning_rate": 1.7587853569283302e-07,
+      "loss": 1.0168,
+      "step": 8378
+    },
+    {
+      "epoch": 0.8812231322386843,
+      "grad_norm": 2.2069125097551283,
+      "learning_rate": 1.7557107818857889e-07,
+      "loss": 0.9825,
+      "step": 8379
+    },
+    {
+      "epoch": 0.8813283026805316,
+      "grad_norm": 3.2766113793720266,
+      "learning_rate": 1.7526387987620602e-07,
+      "loss": 1.0089,
+      "step": 8380
+    },
+    {
+      "epoch": 0.881433473122379,
+      "grad_norm": 2.811898086555635,
+      "learning_rate": 1.7495694078996984e-07,
+      "loss": 1.0126,
+      "step": 8381
+    },
+    {
+      "epoch": 0.8815386435642263,
+      "grad_norm": 3.267487035349017,
+      "learning_rate": 1.7465026096409598e-07,
+      "loss": 1.006,
+      "step": 8382
+    },
+    {
+      "epoch": 0.8816438140060736,
+      "grad_norm": 2.2437889546052934,
+      "learning_rate": 1.743438404327827e-07,
+      "loss": 0.978,
+      "step": 8383
+    },
+    {
+      "epoch": 0.8817489844479209,
+      "grad_norm": 2.969311191359739,
+      "learning_rate": 1.740376792301973e-07,
+      "loss": 0.986,
+      "step": 8384
+    },
+    {
+      "epoch": 0.8818541548897683,
+      "grad_norm": 2.937579661645254,
+      "learning_rate": 1.7373177739047898e-07,
+      "loss": 0.9688,
+      "step": 8385
+    },
+    {
+      "epoch": 0.8819593253316156,
+      "grad_norm": 2.0161300763505494,
+      "learning_rate": 1.7342613494773896e-07,
+      "loss": 0.9934,
+      "step": 8386
+    },
+    {
+      "epoch": 0.8820644957734629,
+      "grad_norm": 2.640765009401412,
+      "learning_rate": 1.731207519360581e-07,
+      "loss": 0.971,
+      "step": 8387
+    },
+    {
+      "epoch": 0.8821696662153102,
+      "grad_norm": 2.4341326910073557,
+      "learning_rate": 1.7281562838948968e-07,
+      "loss": 1.0054,
+      "step": 8388
+    },
+    {
+      "epoch": 0.8822748366571576,
+      "grad_norm": 2.611158264197989,
+      "learning_rate": 1.725107643420565e-07,
+      "loss": 1.0137,
+      "step": 8389
+    },
+    {
+      "epoch": 0.8823800070990048,
+      "grad_norm": 2.040666613213012,
+      "learning_rate": 1.7220615982775357e-07,
+      "loss": 0.9651,
+      "step": 8390
+    },
+    {
+      "epoch": 0.8824851775408521,
+      "grad_norm": 1.9483793255653934,
+      "learning_rate": 1.719018148805471e-07,
+      "loss": 0.9993,
+      "step": 8391
+    },
+    {
+      "epoch": 0.8825903479826994,
+      "grad_norm": 2.2723352862096635,
+      "learning_rate": 1.715977295343732e-07,
+      "loss": 0.9899,
+      "step": 8392
+    },
+    {
+      "epoch": 0.8826955184245467,
+      "grad_norm": 1.9036323549582663,
+      "learning_rate": 1.7129390382314065e-07,
+      "loss": 0.9872,
+      "step": 8393
+    },
+    {
+      "epoch": 0.8828006888663941,
+      "grad_norm": 2.229306978269511,
+      "learning_rate": 1.7099033778072732e-07,
+      "loss": 1.0015,
+      "step": 8394
+    },
+    {
+      "epoch": 0.8829058593082414,
+      "grad_norm": 2.508833961836707,
+      "learning_rate": 1.7068703144098365e-07,
+      "loss": 0.9562,
+      "step": 8395
+    },
+    {
+      "epoch": 0.8830110297500887,
+      "grad_norm": 2.5956332163694156,
+      "learning_rate": 1.7038398483773088e-07,
+      "loss": 0.9996,
+      "step": 8396
+    },
+    {
+      "epoch": 0.883116200191936,
+      "grad_norm": 1.9692024923776876,
+      "learning_rate": 1.700811980047612e-07,
+      "loss": 0.9557,
+      "step": 8397
+    },
+    {
+      "epoch": 0.8832213706337834,
+      "grad_norm": 2.411668195333863,
+      "learning_rate": 1.69778670975837e-07,
+      "loss": 1.0025,
+      "step": 8398
+    },
+    {
+      "epoch": 0.8833265410756307,
+      "grad_norm": 2.4902348906801257,
+      "learning_rate": 1.69476403784693e-07,
+      "loss": 1.0012,
+      "step": 8399
+    },
+    {
+      "epoch": 0.883431711517478,
+      "grad_norm": 2.818533940448939,
+      "learning_rate": 1.6917439646503415e-07,
+      "loss": 0.9662,
+      "step": 8400
+    },
+    {
+      "epoch": 0.8835368819593253,
+      "grad_norm": 3.015978971725827,
+      "learning_rate": 1.688726490505363e-07,
+      "loss": 0.9438,
+      "step": 8401
+    },
+    {
+      "epoch": 0.8836420524011727,
+      "grad_norm": 3.354485337789241,
+      "learning_rate": 1.6857116157484755e-07,
+      "loss": 0.9615,
+      "step": 8402
+    },
+    {
+      "epoch": 0.88374722284302,
+      "grad_norm": 2.4921532513409224,
+      "learning_rate": 1.6826993407158488e-07,
+      "loss": 0.9736,
+      "step": 8403
+    },
+    {
+      "epoch": 0.8838523932848673,
+      "grad_norm": 2.414353246940312,
+      "learning_rate": 1.679689665743381e-07,
+      "loss": 0.9845,
+      "step": 8404
+    },
+    {
+      "epoch": 0.8839575637267146,
+      "grad_norm": 3.235772053731161,
+      "learning_rate": 1.6766825911666757e-07,
+      "loss": 1.0116,
+      "step": 8405
+    },
+    {
+      "epoch": 0.884062734168562,
+      "grad_norm": 2.262414448014693,
+      "learning_rate": 1.6736781173210426e-07,
+      "loss": 0.9497,
+      "step": 8406
+    },
+    {
+      "epoch": 0.8841679046104093,
+      "grad_norm": 2.731997494521064,
+      "learning_rate": 1.670676244541508e-07,
+      "loss": 1.0178,
+      "step": 8407
+    },
+    {
+      "epoch": 0.8842730750522566,
+      "grad_norm": 2.2117783222892786,
+      "learning_rate": 1.6676769731627907e-07,
+      "loss": 0.9542,
+      "step": 8408
+    },
+    {
+      "epoch": 0.8843782454941039,
+      "grad_norm": 2.330534181027143,
+      "learning_rate": 1.6646803035193532e-07,
+      "loss": 0.9347,
+      "step": 8409
+    },
+    {
+      "epoch": 0.8844834159359513,
+      "grad_norm": 2.492206689262255,
+      "learning_rate": 1.661686235945331e-07,
+      "loss": 1.0167,
+      "step": 8410
+    },
+    {
+      "epoch": 0.8845885863777985,
+      "grad_norm": 2.63746278586202,
+      "learning_rate": 1.6586947707745965e-07,
+      "loss": 0.9653,
+      "step": 8411
+    },
+    {
+      "epoch": 0.8846937568196458,
+      "grad_norm": 1.9434162730085702,
+      "learning_rate": 1.655705908340713e-07,
+      "loss": 0.893,
+      "step": 8412
+    },
+    {
+      "epoch": 0.8847989272614931,
+      "grad_norm": 1.9752644333020086,
+      "learning_rate": 1.6527196489769664e-07,
+      "loss": 1.0008,
+      "step": 8413
+    },
+    {
+      "epoch": 0.8849040977033404,
+      "grad_norm": 2.39797801920839,
+      "learning_rate": 1.6497359930163492e-07,
+      "loss": 1.0064,
+      "step": 8414
+    },
+    {
+      "epoch": 0.8850092681451878,
+      "grad_norm": 3.023142379449854,
+      "learning_rate": 1.6467549407915563e-07,
+      "loss": 0.9955,
+      "step": 8415
+    },
+    {
+      "epoch": 0.8851144385870351,
+      "grad_norm": 2.4594124824433967,
+      "learning_rate": 1.6437764926350074e-07,
+      "loss": 0.9866,
+      "step": 8416
+    },
+    {
+      "epoch": 0.8852196090288824,
+      "grad_norm": 3.1186745683219406,
+      "learning_rate": 1.640800648878807e-07,
+      "loss": 1.015,
+      "step": 8417
+    },
+    {
+      "epoch": 0.8853247794707297,
+      "grad_norm": 2.013217130768628,
+      "learning_rate": 1.637827409854803e-07,
+      "loss": 0.9681,
+      "step": 8418
+    },
+    {
+      "epoch": 0.8854299499125771,
+      "grad_norm": 2.412821949942149,
+      "learning_rate": 1.6348567758945277e-07,
+      "loss": 1.0154,
+      "step": 8419
+    },
+    {
+      "epoch": 0.8855351203544244,
+      "grad_norm": 2.82262481468373,
+      "learning_rate": 1.6318887473292245e-07,
+      "loss": 0.9982,
+      "step": 8420
+    },
+    {
+      "epoch": 0.8856402907962717,
+      "grad_norm": 2.4389919981151498,
+      "learning_rate": 1.6289233244898616e-07,
+      "loss": 0.9898,
+      "step": 8421
+    },
+    {
+      "epoch": 0.885745461238119,
+      "grad_norm": 2.5329579890512885,
+      "learning_rate": 1.6259605077070888e-07,
+      "loss": 0.9611,
+      "step": 8422
+    },
+    {
+      "epoch": 0.8858506316799664,
+      "grad_norm": 2.242020126640096,
+      "learning_rate": 1.6230002973113056e-07,
+      "loss": 0.9727,
+      "step": 8423
+    },
+    {
+      "epoch": 0.8859558021218137,
+      "grad_norm": 2.678622402703502,
+      "learning_rate": 1.620042693632584e-07,
+      "loss": 0.973,
+      "step": 8424
+    },
+    {
+      "epoch": 0.886060972563661,
+      "grad_norm": 3.081880759439436,
+      "learning_rate": 1.61708769700073e-07,
+      "loss": 0.9799,
+      "step": 8425
+    },
+    {
+      "epoch": 0.8861661430055083,
+      "grad_norm": 2.3125089299768073,
+      "learning_rate": 1.614135307745235e-07,
+      "loss": 0.9988,
+      "step": 8426
+    },
+    {
+      "epoch": 0.8862713134473557,
+      "grad_norm": 1.7855830987719163,
+      "learning_rate": 1.611185526195322e-07,
+      "loss": 0.9772,
+      "step": 8427
+    },
+    {
+      "epoch": 0.886376483889203,
+      "grad_norm": 1.8724555459292305,
+      "learning_rate": 1.6082383526799196e-07,
+      "loss": 1.0256,
+      "step": 8428
+    },
+    {
+      "epoch": 0.8864816543310503,
+      "grad_norm": 3.5183308798130972,
+      "learning_rate": 1.6052937875276481e-07,
+      "loss": 0.9885,
+      "step": 8429
+    },
+    {
+      "epoch": 0.8865868247728976,
+      "grad_norm": 1.9475600577969459,
+      "learning_rate": 1.602351831066862e-07,
+      "loss": 0.9556,
+      "step": 8430
+    },
+    {
+      "epoch": 0.8866919952147448,
+      "grad_norm": 2.068576517697177,
+      "learning_rate": 1.5994124836255952e-07,
+      "loss": 0.9723,
+      "step": 8431
+    },
+    {
+      "epoch": 0.8867971656565922,
+      "grad_norm": 2.0504911635408307,
+      "learning_rate": 1.5964757455316282e-07,
+      "loss": 0.9624,
+      "step": 8432
+    },
+    {
+      "epoch": 0.8869023360984395,
+      "grad_norm": 2.384507371414114,
+      "learning_rate": 1.593541617112415e-07,
+      "loss": 0.9625,
+      "step": 8433
+    },
+    {
+      "epoch": 0.8870075065402868,
+      "grad_norm": 2.2608056805889936,
+      "learning_rate": 1.5906100986951445e-07,
+      "loss": 0.9405,
+      "step": 8434
+    },
+    {
+      "epoch": 0.8871126769821341,
+      "grad_norm": 2.3029169049564775,
+      "learning_rate": 1.5876811906066992e-07,
+      "loss": 1.013,
+      "step": 8435
+    },
+    {
+      "epoch": 0.8872178474239815,
+      "grad_norm": 3.240342256666061,
+      "learning_rate": 1.5847548931736678e-07,
+      "loss": 0.9972,
+      "step": 8436
+    },
+    {
+      "epoch": 0.8873230178658288,
+      "grad_norm": 2.0368840918231412,
+      "learning_rate": 1.5818312067223673e-07,
+      "loss": 1.0,
+      "step": 8437
+    },
+    {
+      "epoch": 0.8874281883076761,
+      "grad_norm": 2.4964437098615773,
+      "learning_rate": 1.5789101315788007e-07,
+      "loss": 0.9554,
+      "step": 8438
+    },
+    {
+      "epoch": 0.8875333587495234,
+      "grad_norm": 2.719371843674228,
+      "learning_rate": 1.5759916680687022e-07,
+      "loss": 1.0049,
+      "step": 8439
+    },
+    {
+      "epoch": 0.8876385291913708,
+      "grad_norm": 2.545862838080534,
+      "learning_rate": 1.5730758165174914e-07,
+      "loss": 0.979,
+      "step": 8440
+    },
+    {
+      "epoch": 0.8877436996332181,
+      "grad_norm": 4.491846091536825,
+      "learning_rate": 1.5701625772503142e-07,
+      "loss": 1.0328,
+      "step": 8441
+    },
+    {
+      "epoch": 0.8878488700750654,
+      "grad_norm": 2.9635865459634285,
+      "learning_rate": 1.5672519505920248e-07,
+      "loss": 1.0136,
+      "step": 8442
+    },
+    {
+      "epoch": 0.8879540405169127,
+      "grad_norm": 2.5871391440866836,
+      "learning_rate": 1.5643439368671688e-07,
+      "loss": 1.0381,
+      "step": 8443
+    },
+    {
+      "epoch": 0.8880592109587601,
+      "grad_norm": 2.367393248660268,
+      "learning_rate": 1.561438536400023e-07,
+      "loss": 0.9617,
+      "step": 8444
+    },
+    {
+      "epoch": 0.8881643814006074,
+      "grad_norm": 2.293140538401526,
+      "learning_rate": 1.5585357495145504e-07,
+      "loss": 0.9558,
+      "step": 8445
+    },
+    {
+      "epoch": 0.8882695518424547,
+      "grad_norm": 2.217902357821252,
+      "learning_rate": 1.5556355765344477e-07,
+      "loss": 0.9856,
+      "step": 8446
+    },
+    {
+      "epoch": 0.888374722284302,
+      "grad_norm": 3.1515334464477927,
+      "learning_rate": 1.552738017783098e-07,
+      "loss": 0.9647,
+      "step": 8447
+    },
+    {
+      "epoch": 0.8884798927261494,
+      "grad_norm": 1.9568682729973774,
+      "learning_rate": 1.5498430735836067e-07,
+      "loss": 0.9509,
+      "step": 8448
+    },
+    {
+      "epoch": 0.8885850631679967,
+      "grad_norm": 2.243562043893478,
+      "learning_rate": 1.5469507442587766e-07,
+      "loss": 0.9903,
+      "step": 8449
+    },
+    {
+      "epoch": 0.888690233609844,
+      "grad_norm": 2.514164405589009,
+      "learning_rate": 1.5440610301311277e-07,
+      "loss": 1.0134,
+      "step": 8450
+    },
+    {
+      "epoch": 0.8887954040516912,
+      "grad_norm": 3.3526553400519883,
+      "learning_rate": 1.5411739315228909e-07,
+      "loss": 0.9841,
+      "step": 8451
+    },
+    {
+      "epoch": 0.8889005744935385,
+      "grad_norm": 2.7744608538169317,
+      "learning_rate": 1.538289448755989e-07,
+      "loss": 1.0142,
+      "step": 8452
+    },
+    {
+      "epoch": 0.8890057449353859,
+      "grad_norm": 2.2860868670728114,
+      "learning_rate": 1.5354075821520765e-07,
+      "loss": 1.0062,
+      "step": 8453
+    },
+    {
+      "epoch": 0.8891109153772332,
+      "grad_norm": 3.007217663938393,
+      "learning_rate": 1.5325283320324903e-07,
+      "loss": 0.9644,
+      "step": 8454
+    },
+    {
+      "epoch": 0.8892160858190805,
+      "grad_norm": 2.9791595489559435,
+      "learning_rate": 1.5296516987183042e-07,
+      "loss": 0.9896,
+      "step": 8455
+    },
+    {
+      "epoch": 0.8893212562609278,
+      "grad_norm": 1.8583784300969213,
+      "learning_rate": 1.5267776825302782e-07,
+      "loss": 1.0152,
+      "step": 8456
+    },
+    {
+      "epoch": 0.8894264267027752,
+      "grad_norm": 2.729685552570694,
+      "learning_rate": 1.523906283788884e-07,
+      "loss": 1.0083,
+      "step": 8457
+    },
+    {
+      "epoch": 0.8895315971446225,
+      "grad_norm": 3.1558272502401556,
+      "learning_rate": 1.5210375028143097e-07,
+      "loss": 0.9668,
+      "step": 8458
+    },
+    {
+      "epoch": 0.8896367675864698,
+      "grad_norm": 2.1971892785045033,
+      "learning_rate": 1.518171339926436e-07,
+      "loss": 1.0038,
+      "step": 8459
+    },
+    {
+      "epoch": 0.8897419380283171,
+      "grad_norm": 2.841371147766442,
+      "learning_rate": 1.5153077954448813e-07,
+      "loss": 0.9885,
+      "step": 8460
+    },
+    {
+      "epoch": 0.8898471084701645,
+      "grad_norm": 2.691644935448725,
+      "learning_rate": 1.5124468696889383e-07,
+      "loss": 0.9512,
+      "step": 8461
+    },
+    {
+      "epoch": 0.8899522789120118,
+      "grad_norm": 3.5217017668869097,
+      "learning_rate": 1.5095885629776319e-07,
+      "loss": 0.9877,
+      "step": 8462
+    },
+    {
+      "epoch": 0.8900574493538591,
+      "grad_norm": 1.6034639625291485,
+      "learning_rate": 1.506732875629674e-07,
+      "loss": 1.0055,
+      "step": 8463
+    },
+    {
+      "epoch": 0.8901626197957064,
+      "grad_norm": 2.1047634595751026,
+      "learning_rate": 1.5038798079635043e-07,
+      "loss": 0.9644,
+      "step": 8464
+    },
+    {
+      "epoch": 0.8902677902375538,
+      "grad_norm": 2.1406693190498256,
+      "learning_rate": 1.5010293602972653e-07,
+      "loss": 0.974,
+      "step": 8465
+    },
+    {
+      "epoch": 0.8903729606794011,
+      "grad_norm": 2.042851624189361,
+      "learning_rate": 1.4981815329487948e-07,
+      "loss": 0.9174,
+      "step": 8466
+    },
+    {
+      "epoch": 0.8904781311212484,
+      "grad_norm": 2.5483924252229673,
+      "learning_rate": 1.4953363262356552e-07,
+      "loss": 0.9662,
+      "step": 8467
+    },
+    {
+      "epoch": 0.8905833015630957,
+      "grad_norm": 3.0624720151851617,
+      "learning_rate": 1.4924937404750954e-07,
+      "loss": 0.9856,
+      "step": 8468
+    },
+    {
+      "epoch": 0.890688472004943,
+      "grad_norm": 3.0903489981255654,
+      "learning_rate": 1.4896537759841067e-07,
+      "loss": 1.0083,
+      "step": 8469
+    },
+    {
+      "epoch": 0.8907936424467904,
+      "grad_norm": 2.1498114196551823,
+      "learning_rate": 1.486816433079352e-07,
+      "loss": 0.9812,
+      "step": 8470
+    },
+    {
+      "epoch": 0.8908988128886377,
+      "grad_norm": 2.15804642211936,
+      "learning_rate": 1.4839817120772288e-07,
+      "loss": 0.9824,
+      "step": 8471
+    },
+    {
+      "epoch": 0.8910039833304849,
+      "grad_norm": 2.069762197489603,
+      "learning_rate": 1.4811496132938196e-07,
+      "loss": 0.9699,
+      "step": 8472
+    },
+    {
+      "epoch": 0.8911091537723322,
+      "grad_norm": 2.3885493176951296,
+      "learning_rate": 1.4783201370449225e-07,
+      "loss": 0.9592,
+      "step": 8473
+    },
+    {
+      "epoch": 0.8912143242141796,
+      "grad_norm": 2.5287347361503936,
+      "learning_rate": 1.4754932836460622e-07,
+      "loss": 0.927,
+      "step": 8474
+    },
+    {
+      "epoch": 0.8913194946560269,
+      "grad_norm": 2.477400242672135,
+      "learning_rate": 1.47266905341244e-07,
+      "loss": 0.9984,
+      "step": 8475
+    },
+    {
+      "epoch": 0.8914246650978742,
+      "grad_norm": 2.8722700619640027,
+      "learning_rate": 1.4698474466589896e-07,
+      "loss": 0.9341,
+      "step": 8476
+    },
+    {
+      "epoch": 0.8915298355397215,
+      "grad_norm": 2.722448788439971,
+      "learning_rate": 1.4670284637003345e-07,
+      "loss": 0.9496,
+      "step": 8477
+    },
+    {
+      "epoch": 0.8916350059815689,
+      "grad_norm": 2.8289714203437892,
+      "learning_rate": 1.464212104850815e-07,
+      "loss": 0.9568,
+      "step": 8478
+    },
+    {
+      "epoch": 0.8917401764234162,
+      "grad_norm": 2.6306864237152623,
+      "learning_rate": 1.4613983704244826e-07,
+      "loss": 1.0054,
+      "step": 8479
+    },
+    {
+      "epoch": 0.8918453468652635,
+      "grad_norm": 2.018496472612239,
+      "learning_rate": 1.4585872607350837e-07,
+      "loss": 0.9911,
+      "step": 8480
+    },
+    {
+      "epoch": 0.8919505173071108,
+      "grad_norm": 2.6075018388842577,
+      "learning_rate": 1.4557787760960812e-07,
+      "loss": 0.9536,
+      "step": 8481
+    },
+    {
+      "epoch": 0.8920556877489582,
+      "grad_norm": 2.419257762481413,
+      "learning_rate": 1.4529729168206386e-07,
+      "loss": 0.9466,
+      "step": 8482
+    },
+    {
+      "epoch": 0.8921608581908055,
+      "grad_norm": 2.726364003607082,
+      "learning_rate": 1.4501696832216445e-07,
+      "loss": 0.9499,
+      "step": 8483
+    },
+    {
+      "epoch": 0.8922660286326528,
+      "grad_norm": 2.3855719640701807,
+      "learning_rate": 1.4473690756116654e-07,
+      "loss": 0.9907,
+      "step": 8484
+    },
+    {
+      "epoch": 0.8923711990745001,
+      "grad_norm": 2.4566580192018845,
+      "learning_rate": 1.4445710943030018e-07,
+      "loss": 0.9854,
+      "step": 8485
+    },
+    {
+      "epoch": 0.8924763695163475,
+      "grad_norm": 2.816644567674129,
+      "learning_rate": 1.4417757396076427e-07,
+      "loss": 1.0299,
+      "step": 8486
+    },
+    {
+      "epoch": 0.8925815399581948,
+      "grad_norm": 2.4296406347952884,
+      "learning_rate": 1.4389830118372972e-07,
+      "loss": 0.9775,
+      "step": 8487
+    },
+    {
+      "epoch": 0.8926867104000421,
+      "grad_norm": 1.952494054315651,
+      "learning_rate": 1.4361929113033747e-07,
+      "loss": 0.9887,
+      "step": 8488
+    },
+    {
+      "epoch": 0.8927918808418894,
+      "grad_norm": 2.203328059269768,
+      "learning_rate": 1.4334054383169897e-07,
+      "loss": 0.9499,
+      "step": 8489
+    },
+    {
+      "epoch": 0.8928970512837368,
+      "grad_norm": 1.8322340247823978,
+      "learning_rate": 1.430620593188975e-07,
+      "loss": 0.9407,
+      "step": 8490
+    },
+    {
+      "epoch": 0.8930022217255841,
+      "grad_norm": 2.480996419359068,
+      "learning_rate": 1.4278383762298536e-07,
+      "loss": 0.9938,
+      "step": 8491
+    },
+    {
+      "epoch": 0.8931073921674313,
+      "grad_norm": 3.1090653197790528,
+      "learning_rate": 1.4250587877498695e-07,
+      "loss": 0.9909,
+      "step": 8492
+    },
+    {
+      "epoch": 0.8932125626092786,
+      "grad_norm": 2.720781388178653,
+      "learning_rate": 1.422281828058969e-07,
+      "loss": 1.0105,
+      "step": 8493
+    },
+    {
+      "epoch": 0.8933177330511259,
+      "grad_norm": 2.753443908893908,
+      "learning_rate": 1.4195074974667989e-07,
+      "loss": 0.994,
+      "step": 8494
+    },
+    {
+      "epoch": 0.8934229034929733,
+      "grad_norm": 1.8341611634515465,
+      "learning_rate": 1.4167357962827283e-07,
+      "loss": 0.9763,
+      "step": 8495
+    },
+    {
+      "epoch": 0.8935280739348206,
+      "grad_norm": 2.9427961867698977,
+      "learning_rate": 1.4139667248158072e-07,
+      "loss": 0.9565,
+      "step": 8496
+    },
+    {
+      "epoch": 0.8936332443766679,
+      "grad_norm": 2.4027494539397103,
+      "learning_rate": 1.4112002833748279e-07,
+      "loss": 0.9674,
+      "step": 8497
+    },
+    {
+      "epoch": 0.8937384148185152,
+      "grad_norm": 3.598778669005095,
+      "learning_rate": 1.4084364722682598e-07,
+      "loss": 0.9917,
+      "step": 8498
+    },
+    {
+      "epoch": 0.8938435852603626,
+      "grad_norm": 3.346549402658093,
+      "learning_rate": 1.4056752918042898e-07,
+      "loss": 0.9366,
+      "step": 8499
+    },
+    {
+      "epoch": 0.8939487557022099,
+      "grad_norm": 2.5730479469223897,
+      "learning_rate": 1.4029167422908107e-07,
+      "loss": 0.9992,
+      "step": 8500
+    },
+    {
+      "epoch": 0.8940539261440572,
+      "grad_norm": 2.7809454904611313,
+      "learning_rate": 1.4001608240354236e-07,
+      "loss": 0.9954,
+      "step": 8501
+    },
+    {
+      "epoch": 0.8941590965859045,
+      "grad_norm": 2.063097454423752,
+      "learning_rate": 1.3974075373454383e-07,
+      "loss": 0.9769,
+      "step": 8502
+    },
+    {
+      "epoch": 0.8942642670277519,
+      "grad_norm": 1.8225435784652433,
+      "learning_rate": 1.3946568825278618e-07,
+      "loss": 0.9645,
+      "step": 8503
+    },
+    {
+      "epoch": 0.8943694374695992,
+      "grad_norm": 2.755707682683513,
+      "learning_rate": 1.3919088598894154e-07,
+      "loss": 0.9932,
+      "step": 8504
+    },
+    {
+      "epoch": 0.8944746079114465,
+      "grad_norm": 3.3674579224287773,
+      "learning_rate": 1.3891634697365207e-07,
+      "loss": 1.0107,
+      "step": 8505
+    },
+    {
+      "epoch": 0.8945797783532938,
+      "grad_norm": 3.130174869282355,
+      "learning_rate": 1.386420712375322e-07,
+      "loss": 1.0157,
+      "step": 8506
+    },
+    {
+      "epoch": 0.8946849487951412,
+      "grad_norm": 2.0725928871095154,
+      "learning_rate": 1.3836805881116495e-07,
+      "loss": 0.9675,
+      "step": 8507
+    },
+    {
+      "epoch": 0.8947901192369885,
+      "grad_norm": 2.345929950085412,
+      "learning_rate": 1.3809430972510446e-07,
+      "loss": 0.9821,
+      "step": 8508
+    },
+    {
+      "epoch": 0.8948952896788358,
+      "grad_norm": 2.7903627116773,
+      "learning_rate": 1.378208240098766e-07,
+      "loss": 0.9848,
+      "step": 8509
+    },
+    {
+      "epoch": 0.8950004601206831,
+      "grad_norm": 2.328892581734959,
+      "learning_rate": 1.3754760169597676e-07,
+      "loss": 1.0026,
+      "step": 8510
+    },
+    {
+      "epoch": 0.8951056305625305,
+      "grad_norm": 2.8229279376291934,
+      "learning_rate": 1.3727464281387186e-07,
+      "loss": 0.9352,
+      "step": 8511
+    },
+    {
+      "epoch": 0.8952108010043777,
+      "grad_norm": 2.8468987303756284,
+      "learning_rate": 1.3700194739399848e-07,
+      "loss": 1.0079,
+      "step": 8512
+    },
+    {
+      "epoch": 0.895315971446225,
+      "grad_norm": 2.015499855514806,
+      "learning_rate": 1.3672951546676444e-07,
+      "loss": 0.9678,
+      "step": 8513
+    },
+    {
+      "epoch": 0.8954211418880723,
+      "grad_norm": 1.976934703918547,
+      "learning_rate": 1.364573470625477e-07,
+      "loss": 0.9829,
+      "step": 8514
+    },
+    {
+      "epoch": 0.8955263123299196,
+      "grad_norm": 2.31984836197906,
+      "learning_rate": 1.3618544221169732e-07,
+      "loss": 0.9756,
+      "step": 8515
+    },
+    {
+      "epoch": 0.895631482771767,
+      "grad_norm": 2.665380278443687,
+      "learning_rate": 1.359138009445335e-07,
+      "loss": 0.9988,
+      "step": 8516
+    },
+    {
+      "epoch": 0.8957366532136143,
+      "grad_norm": 1.9964367122956426,
+      "learning_rate": 1.3564242329134502e-07,
+      "loss": 0.944,
+      "step": 8517
+    },
+    {
+      "epoch": 0.8958418236554616,
+      "grad_norm": 2.576231474046989,
+      "learning_rate": 1.3537130928239383e-07,
+      "loss": 0.9717,
+      "step": 8518
+    },
+    {
+      "epoch": 0.8959469940973089,
+      "grad_norm": 2.5238951016013367,
+      "learning_rate": 1.3510045894791018e-07,
+      "loss": 0.9593,
+      "step": 8519
+    },
+    {
+      "epoch": 0.8960521645391563,
+      "grad_norm": 3.0112910269228275,
+      "learning_rate": 1.3482987231809714e-07,
+      "loss": 1.0174,
+      "step": 8520
+    },
+    {
+      "epoch": 0.8961573349810036,
+      "grad_norm": 2.728647164800237,
+      "learning_rate": 1.345595494231261e-07,
+      "loss": 1.0167,
+      "step": 8521
+    },
+    {
+      "epoch": 0.8962625054228509,
+      "grad_norm": 3.110617555783042,
+      "learning_rate": 1.3428949029314132e-07,
+      "loss": 1.0145,
+      "step": 8522
+    },
+    {
+      "epoch": 0.8963676758646982,
+      "grad_norm": 2.1989572332927216,
+      "learning_rate": 1.3401969495825534e-07,
+      "loss": 0.9382,
+      "step": 8523
+    },
+    {
+      "epoch": 0.8964728463065456,
+      "grad_norm": 2.530350556427271,
+      "learning_rate": 1.3375016344855302e-07,
+      "loss": 1.0152,
+      "step": 8524
+    },
+    {
+      "epoch": 0.8965780167483929,
+      "grad_norm": 3.24507065070509,
+      "learning_rate": 1.3348089579408946e-07,
+      "loss": 1.0011,
+      "step": 8525
+    },
+    {
+      "epoch": 0.8966831871902402,
+      "grad_norm": 2.292129085968337,
+      "learning_rate": 1.3321189202488955e-07,
+      "loss": 0.9794,
+      "step": 8526
+    },
+    {
+      "epoch": 0.8967883576320875,
+      "grad_norm": 2.5251603922264785,
+      "learning_rate": 1.3294315217094956e-07,
+      "loss": 1.0009,
+      "step": 8527
+    },
+    {
+      "epoch": 0.8968935280739349,
+      "grad_norm": 1.9633155096515325,
+      "learning_rate": 1.3267467626223606e-07,
+      "loss": 0.959,
+      "step": 8528
+    },
+    {
+      "epoch": 0.8969986985157822,
+      "grad_norm": 3.11725878427488,
+      "learning_rate": 1.32406464328686e-07,
+      "loss": 1.0017,
+      "step": 8529
+    },
+    {
+      "epoch": 0.8971038689576295,
+      "grad_norm": 1.9422340766869772,
+      "learning_rate": 1.321385164002076e-07,
+      "loss": 0.9736,
+      "step": 8530
+    },
+    {
+      "epoch": 0.8972090393994768,
+      "grad_norm": 2.3021774888473225,
+      "learning_rate": 1.3187083250667865e-07,
+      "loss": 0.9732,
+      "step": 8531
+    },
+    {
+      "epoch": 0.8973142098413242,
+      "grad_norm": 2.6462765296135227,
+      "learning_rate": 1.3160341267794778e-07,
+      "loss": 0.9473,
+      "step": 8532
+    },
+    {
+      "epoch": 0.8974193802831714,
+      "grad_norm": 2.9968459492356376,
+      "learning_rate": 1.3133625694383506e-07,
+      "loss": 1.0007,
+      "step": 8533
+    },
+    {
+      "epoch": 0.8975245507250187,
+      "grad_norm": 2.8871273274281726,
+      "learning_rate": 1.310693653341305e-07,
+      "loss": 0.9901,
+      "step": 8534
+    },
+    {
+      "epoch": 0.897629721166866,
+      "grad_norm": 2.4909489473224817,
+      "learning_rate": 1.3080273787859367e-07,
+      "loss": 1.0059,
+      "step": 8535
+    },
+    {
+      "epoch": 0.8977348916087133,
+      "grad_norm": 2.775824788943004,
+      "learning_rate": 1.3053637460695655e-07,
+      "loss": 1.0172,
+      "step": 8536
+    },
+    {
+      "epoch": 0.8978400620505607,
+      "grad_norm": 2.405665330311999,
+      "learning_rate": 1.3027027554891992e-07,
+      "loss": 1.0053,
+      "step": 8537
+    },
+    {
+      "epoch": 0.897945232492408,
+      "grad_norm": 2.194682374380952,
+      "learning_rate": 1.3000444073415637e-07,
+      "loss": 0.9454,
+      "step": 8538
+    },
+    {
+      "epoch": 0.8980504029342553,
+      "grad_norm": 2.5244706470088323,
+      "learning_rate": 1.2973887019230885e-07,
+      "loss": 0.9712,
+      "step": 8539
+    },
+    {
+      "epoch": 0.8981555733761026,
+      "grad_norm": 2.9834146366006844,
+      "learning_rate": 1.2947356395298956e-07,
+      "loss": 0.9946,
+      "step": 8540
+    },
+    {
+      "epoch": 0.89826074381795,
+      "grad_norm": 2.246603088988245,
+      "learning_rate": 1.292085220457834e-07,
+      "loss": 0.9229,
+      "step": 8541
+    },
+    {
+      "epoch": 0.8983659142597973,
+      "grad_norm": 2.4711768931739004,
+      "learning_rate": 1.2894374450024338e-07,
+      "loss": 0.9644,
+      "step": 8542
+    },
+    {
+      "epoch": 0.8984710847016446,
+      "grad_norm": 1.99224760834784,
+      "learning_rate": 1.2867923134589483e-07,
+      "loss": 0.9821,
+      "step": 8543
+    },
+    {
+      "epoch": 0.8985762551434919,
+      "grad_norm": 1.6913963223277704,
+      "learning_rate": 1.2841498261223355e-07,
+      "loss": 0.9516,
+      "step": 8544
+    },
+    {
+      "epoch": 0.8986814255853393,
+      "grad_norm": 2.161369459321979,
+      "learning_rate": 1.2815099832872463e-07,
+      "loss": 0.9564,
+      "step": 8545
+    },
+    {
+      "epoch": 0.8987865960271866,
+      "grad_norm": 2.1550070381084114,
+      "learning_rate": 1.278872785248042e-07,
+      "loss": 0.9686,
+      "step": 8546
+    },
+    {
+      "epoch": 0.8988917664690339,
+      "grad_norm": 2.840197703849838,
+      "learning_rate": 1.276238232298796e-07,
+      "loss": 0.9632,
+      "step": 8547
+    },
+    {
+      "epoch": 0.8989969369108812,
+      "grad_norm": 2.5769335426004982,
+      "learning_rate": 1.273606324733284e-07,
+      "loss": 0.9816,
+      "step": 8548
+    },
+    {
+      "epoch": 0.8991021073527286,
+      "grad_norm": 2.6707662326181425,
+      "learning_rate": 1.2709770628449776e-07,
+      "loss": 0.9601,
+      "step": 8549
+    },
+    {
+      "epoch": 0.8992072777945759,
+      "grad_norm": 1.7310467132195133,
+      "learning_rate": 1.2683504469270636e-07,
+      "loss": 0.9591,
+      "step": 8550
+    },
+    {
+      "epoch": 0.8993124482364232,
+      "grad_norm": 2.548589763012564,
+      "learning_rate": 1.2657264772724247e-07,
+      "loss": 0.993,
+      "step": 8551
+    },
+    {
+      "epoch": 0.8994176186782705,
+      "grad_norm": 2.2542436406644684,
+      "learning_rate": 1.2631051541736578e-07,
+      "loss": 0.9368,
+      "step": 8552
+    },
+    {
+      "epoch": 0.8995227891201177,
+      "grad_norm": 2.5992564379457455,
+      "learning_rate": 1.2604864779230674e-07,
+      "loss": 0.9385,
+      "step": 8553
+    },
+    {
+      "epoch": 0.8996279595619651,
+      "grad_norm": 2.3697543945400343,
+      "learning_rate": 1.2578704488126426e-07,
+      "loss": 0.9866,
+      "step": 8554
+    },
+    {
+      "epoch": 0.8997331300038124,
+      "grad_norm": 2.31815816883965,
+      "learning_rate": 1.2552570671340998e-07,
+      "loss": 0.974,
+      "step": 8555
+    },
+    {
+      "epoch": 0.8998383004456597,
+      "grad_norm": 2.6932418190961913,
+      "learning_rate": 1.2526463331788503e-07,
+      "loss": 1.0144,
+      "step": 8556
+    },
+    {
+      "epoch": 0.899943470887507,
+      "grad_norm": 2.969325960064139,
+      "learning_rate": 1.250038247238014e-07,
+      "loss": 1.0272,
+      "step": 8557
+    },
+    {
+      "epoch": 0.9000486413293544,
+      "grad_norm": 2.781868699429922,
+      "learning_rate": 1.2474328096024086e-07,
+      "loss": 0.999,
+      "step": 8558
+    },
+    {
+      "epoch": 0.9001538117712017,
+      "grad_norm": 3.139713598126397,
+      "learning_rate": 1.244830020562557e-07,
+      "loss": 0.9944,
+      "step": 8559
+    },
+    {
+      "epoch": 0.900258982213049,
+      "grad_norm": 2.2804116496124767,
+      "learning_rate": 1.2422298804086963e-07,
+      "loss": 0.9445,
+      "step": 8560
+    },
+    {
+      "epoch": 0.9003641526548963,
+      "grad_norm": 2.1291441323106786,
+      "learning_rate": 1.2396323894307587e-07,
+      "loss": 0.9773,
+      "step": 8561
+    },
+    {
+      "epoch": 0.9004693230967437,
+      "grad_norm": 2.7682785401579455,
+      "learning_rate": 1.2370375479183905e-07,
+      "loss": 0.9852,
+      "step": 8562
+    },
+    {
+      "epoch": 0.900574493538591,
+      "grad_norm": 2.906960990437752,
+      "learning_rate": 1.234445356160932e-07,
+      "loss": 0.958,
+      "step": 8563
+    },
+    {
+      "epoch": 0.9006796639804383,
+      "grad_norm": 3.5815863699736807,
+      "learning_rate": 1.2318558144474303e-07,
+      "loss": 1.064,
+      "step": 8564
+    },
+    {
+      "epoch": 0.9007848344222856,
+      "grad_norm": 2.7121326512924866,
+      "learning_rate": 1.2292689230666482e-07,
+      "loss": 0.9298,
+      "step": 8565
+    },
+    {
+      "epoch": 0.900890004864133,
+      "grad_norm": 2.7103260334531956,
+      "learning_rate": 1.226684682307036e-07,
+      "loss": 0.9481,
+      "step": 8566
+    },
+    {
+      "epoch": 0.9009951753059803,
+      "grad_norm": 2.534376466134969,
+      "learning_rate": 1.2241030924567603e-07,
+      "loss": 1.0111,
+      "step": 8567
+    },
+    {
+      "epoch": 0.9011003457478276,
+      "grad_norm": 2.1693292560203092,
+      "learning_rate": 1.2215241538036853e-07,
+      "loss": 0.9746,
+      "step": 8568
+    },
+    {
+      "epoch": 0.9012055161896749,
+      "grad_norm": 1.8401531315355144,
+      "learning_rate": 1.2189478666353865e-07,
+      "loss": 1.0031,
+      "step": 8569
+    },
+    {
+      "epoch": 0.9013106866315223,
+      "grad_norm": 2.575023205233234,
+      "learning_rate": 1.2163742312391342e-07,
+      "loss": 0.9486,
+      "step": 8570
+    },
+    {
+      "epoch": 0.9014158570733696,
+      "grad_norm": 1.9758094923843876,
+      "learning_rate": 1.2138032479019206e-07,
+      "loss": 0.9974,
+      "step": 8571
+    },
+    {
+      "epoch": 0.9015210275152169,
+      "grad_norm": 2.2170223011196706,
+      "learning_rate": 1.2112349169104172e-07,
+      "loss": 0.9782,
+      "step": 8572
+    },
+    {
+      "epoch": 0.9016261979570641,
+      "grad_norm": 1.7838799688859874,
+      "learning_rate": 1.208669238551019e-07,
+      "loss": 0.9755,
+      "step": 8573
+    },
+    {
+      "epoch": 0.9017313683989114,
+      "grad_norm": 2.4567274019385765,
+      "learning_rate": 1.2061062131098174e-07,
+      "loss": 0.9455,
+      "step": 8574
+    },
+    {
+      "epoch": 0.9018365388407588,
+      "grad_norm": 1.722594664429003,
+      "learning_rate": 1.203545840872611e-07,
+      "loss": 0.9946,
+      "step": 8575
+    },
+    {
+      "epoch": 0.9019417092826061,
+      "grad_norm": 2.469843612260426,
+      "learning_rate": 1.2009881221249047e-07,
+      "loss": 0.9817,
+      "step": 8576
+    },
+    {
+      "epoch": 0.9020468797244534,
+      "grad_norm": 2.763903318605096,
+      "learning_rate": 1.1984330571518932e-07,
+      "loss": 0.96,
+      "step": 8577
+    },
+    {
+      "epoch": 0.9021520501663007,
+      "grad_norm": 2.4544321149500306,
+      "learning_rate": 1.1958806462384953e-07,
+      "loss": 0.9663,
+      "step": 8578
+    },
+    {
+      "epoch": 0.9022572206081481,
+      "grad_norm": 2.977813358898067,
+      "learning_rate": 1.1933308896693253e-07,
+      "loss": 1.0148,
+      "step": 8579
+    },
+    {
+      "epoch": 0.9023623910499954,
+      "grad_norm": 1.999477072638366,
+      "learning_rate": 1.1907837877286943e-07,
+      "loss": 0.9989,
+      "step": 8580
+    },
+    {
+      "epoch": 0.9024675614918427,
+      "grad_norm": 2.2185946512743513,
+      "learning_rate": 1.188239340700631e-07,
+      "loss": 0.9987,
+      "step": 8581
+    },
+    {
+      "epoch": 0.90257273193369,
+      "grad_norm": 2.5096878993721856,
+      "learning_rate": 1.1856975488688555e-07,
+      "loss": 0.9858,
+      "step": 8582
+    },
+    {
+      "epoch": 0.9026779023755374,
+      "grad_norm": 2.1735177550715403,
+      "learning_rate": 1.1831584125167966e-07,
+      "loss": 1.0257,
+      "step": 8583
+    },
+    {
+      "epoch": 0.9027830728173847,
+      "grad_norm": 1.7897935124112625,
+      "learning_rate": 1.1806219319275918e-07,
+      "loss": 0.9778,
+      "step": 8584
+    },
+    {
+      "epoch": 0.902888243259232,
+      "grad_norm": 2.15199039194613,
+      "learning_rate": 1.1780881073840816e-07,
+      "loss": 1.0001,
+      "step": 8585
+    },
+    {
+      "epoch": 0.9029934137010793,
+      "grad_norm": 3.833073505264171,
+      "learning_rate": 1.1755569391687954e-07,
+      "loss": 1.0042,
+      "step": 8586
+    },
+    {
+      "epoch": 0.9030985841429267,
+      "grad_norm": 2.682110463618559,
+      "learning_rate": 1.173028427563988e-07,
+      "loss": 1.0197,
+      "step": 8587
+    },
+    {
+      "epoch": 0.903203754584774,
+      "grad_norm": 1.9252316839391137,
+      "learning_rate": 1.1705025728516089e-07,
+      "loss": 0.9597,
+      "step": 8588
+    },
+    {
+      "epoch": 0.9033089250266213,
+      "grad_norm": 3.407849681083908,
+      "learning_rate": 1.1679793753133024e-07,
+      "loss": 0.9905,
+      "step": 8589
+    },
+    {
+      "epoch": 0.9034140954684686,
+      "grad_norm": 2.514269869679688,
+      "learning_rate": 1.1654588352304347e-07,
+      "loss": 1.0422,
+      "step": 8590
+    },
+    {
+      "epoch": 0.903519265910316,
+      "grad_norm": 2.9042596313289244,
+      "learning_rate": 1.1629409528840534e-07,
+      "loss": 0.9899,
+      "step": 8591
+    },
+    {
+      "epoch": 0.9036244363521633,
+      "grad_norm": 1.8985121884778924,
+      "learning_rate": 1.1604257285549314e-07,
+      "loss": 0.9963,
+      "step": 8592
+    },
+    {
+      "epoch": 0.9037296067940106,
+      "grad_norm": 2.5551184251028602,
+      "learning_rate": 1.1579131625235356e-07,
+      "loss": 0.9615,
+      "step": 8593
+    },
+    {
+      "epoch": 0.9038347772358578,
+      "grad_norm": 2.1413999858159354,
+      "learning_rate": 1.1554032550700284e-07,
+      "loss": 0.9406,
+      "step": 8594
+    },
+    {
+      "epoch": 0.9039399476777051,
+      "grad_norm": 2.6069475924003704,
+      "learning_rate": 1.1528960064742967e-07,
+      "loss": 0.9775,
+      "step": 8595
+    },
+    {
+      "epoch": 0.9040451181195525,
+      "grad_norm": 1.9692405646354523,
+      "learning_rate": 1.1503914170159058e-07,
+      "loss": 0.9734,
+      "step": 8596
+    },
+    {
+      "epoch": 0.9041502885613998,
+      "grad_norm": 3.182347862653507,
+      "learning_rate": 1.1478894869741409e-07,
+      "loss": 0.9695,
+      "step": 8597
+    },
+    {
+      "epoch": 0.9042554590032471,
+      "grad_norm": 2.393093848945507,
+      "learning_rate": 1.1453902166279895e-07,
+      "loss": 0.9901,
+      "step": 8598
+    },
+    {
+      "epoch": 0.9043606294450944,
+      "grad_norm": 1.8954299188375663,
+      "learning_rate": 1.1428936062561402e-07,
+      "loss": 1.0036,
+      "step": 8599
+    },
+    {
+      "epoch": 0.9044657998869418,
+      "grad_norm": 2.0529572247299,
+      "learning_rate": 1.1403996561369812e-07,
+      "loss": 0.972,
+      "step": 8600
+    },
+    {
+      "epoch": 0.9045709703287891,
+      "grad_norm": 2.3468453698266063,
+      "learning_rate": 1.1379083665486068e-07,
+      "loss": 1.004,
+      "step": 8601
+    },
+    {
+      "epoch": 0.9046761407706364,
+      "grad_norm": 1.9568847387652115,
+      "learning_rate": 1.1354197377688198e-07,
+      "loss": 0.9641,
+      "step": 8602
+    },
+    {
+      "epoch": 0.9047813112124837,
+      "grad_norm": 2.24292284847942,
+      "learning_rate": 1.1329337700751147e-07,
+      "loss": 0.9581,
+      "step": 8603
+    },
+    {
+      "epoch": 0.904886481654331,
+      "grad_norm": 2.5786412520948656,
+      "learning_rate": 1.1304504637447062e-07,
+      "loss": 0.9849,
+      "step": 8604
+    },
+    {
+      "epoch": 0.9049916520961784,
+      "grad_norm": 2.6766324984329586,
+      "learning_rate": 1.1279698190544918e-07,
+      "loss": 0.9928,
+      "step": 8605
+    },
+    {
+      "epoch": 0.9050968225380257,
+      "grad_norm": 2.4750268462318665,
+      "learning_rate": 1.125491836281084e-07,
+      "loss": 0.9605,
+      "step": 8606
+    },
+    {
+      "epoch": 0.905201992979873,
+      "grad_norm": 2.3781929663367936,
+      "learning_rate": 1.1230165157008033e-07,
+      "loss": 1.0048,
+      "step": 8607
+    },
+    {
+      "epoch": 0.9053071634217204,
+      "grad_norm": 2.2874545893594553,
+      "learning_rate": 1.1205438575896677e-07,
+      "loss": 0.9634,
+      "step": 8608
+    },
+    {
+      "epoch": 0.9054123338635677,
+      "grad_norm": 2.526387831515657,
+      "learning_rate": 1.1180738622233928e-07,
+      "loss": 0.9509,
+      "step": 8609
+    },
+    {
+      "epoch": 0.905517504305415,
+      "grad_norm": 2.0896230571312957,
+      "learning_rate": 1.1156065298773972e-07,
+      "loss": 0.9917,
+      "step": 8610
+    },
+    {
+      "epoch": 0.9056226747472623,
+      "grad_norm": 2.022491968484562,
+      "learning_rate": 1.113141860826819e-07,
+      "loss": 1.0339,
+      "step": 8611
+    },
+    {
+      "epoch": 0.9057278451891096,
+      "grad_norm": 2.1059992344770246,
+      "learning_rate": 1.1106798553464804e-07,
+      "loss": 0.9661,
+      "step": 8612
+    },
+    {
+      "epoch": 0.905833015630957,
+      "grad_norm": 2.730011822148298,
+      "learning_rate": 1.1082205137109225e-07,
+      "loss": 0.9576,
+      "step": 8613
+    },
+    {
+      "epoch": 0.9059381860728042,
+      "grad_norm": 2.761974033618745,
+      "learning_rate": 1.1057638361943679e-07,
+      "loss": 0.9832,
+      "step": 8614
+    },
+    {
+      "epoch": 0.9060433565146515,
+      "grad_norm": 2.9220028143550656,
+      "learning_rate": 1.1033098230707668e-07,
+      "loss": 1.0099,
+      "step": 8615
+    },
+    {
+      "epoch": 0.9061485269564988,
+      "grad_norm": 3.1531693192642467,
+      "learning_rate": 1.1008584746137558e-07,
+      "loss": 0.9964,
+      "step": 8616
+    },
+    {
+      "epoch": 0.9062536973983462,
+      "grad_norm": 2.190987178982146,
+      "learning_rate": 1.0984097910966802e-07,
+      "loss": 0.9959,
+      "step": 8617
+    },
+    {
+      "epoch": 0.9063588678401935,
+      "grad_norm": 3.111299811220905,
+      "learning_rate": 1.0959637727925881e-07,
+      "loss": 1.0023,
+      "step": 8618
+    },
+    {
+      "epoch": 0.9064640382820408,
+      "grad_norm": 3.157837969992092,
+      "learning_rate": 1.0935204199742255e-07,
+      "loss": 1.0217,
+      "step": 8619
+    },
+    {
+      "epoch": 0.9065692087238881,
+      "grad_norm": 2.4262394538613283,
+      "learning_rate": 1.0910797329140466e-07,
+      "loss": 1.0104,
+      "step": 8620
+    },
+    {
+      "epoch": 0.9066743791657355,
+      "grad_norm": 2.2045284669897396,
+      "learning_rate": 1.0886417118842113e-07,
+      "loss": 0.9879,
+      "step": 8621
+    },
+    {
+      "epoch": 0.9067795496075828,
+      "grad_norm": 3.4589297463267084,
+      "learning_rate": 1.0862063571565773e-07,
+      "loss": 0.9798,
+      "step": 8622
+    },
+    {
+      "epoch": 0.9068847200494301,
+      "grad_norm": 2.8785742456954004,
+      "learning_rate": 1.0837736690026996e-07,
+      "loss": 0.9865,
+      "step": 8623
+    },
+    {
+      "epoch": 0.9069898904912774,
+      "grad_norm": 2.194186906971404,
+      "learning_rate": 1.0813436476938444e-07,
+      "loss": 0.9963,
+      "step": 8624
+    },
+    {
+      "epoch": 0.9070950609331248,
+      "grad_norm": 2.6444058808730273,
+      "learning_rate": 1.0789162935009839e-07,
+      "loss": 0.991,
+      "step": 8625
+    },
+    {
+      "epoch": 0.9072002313749721,
+      "grad_norm": 2.6206618870443417,
+      "learning_rate": 1.0764916066947795e-07,
+      "loss": 0.9771,
+      "step": 8626
+    },
+    {
+      "epoch": 0.9073054018168194,
+      "grad_norm": 1.9213999370076098,
+      "learning_rate": 1.0740695875456064e-07,
+      "loss": 0.9315,
+      "step": 8627
+    },
+    {
+      "epoch": 0.9074105722586667,
+      "grad_norm": 2.700327806526417,
+      "learning_rate": 1.0716502363235348e-07,
+      "loss": 0.9627,
+      "step": 8628
+    },
+    {
+      "epoch": 0.907515742700514,
+      "grad_norm": 2.9427305067667624,
+      "learning_rate": 1.069233553298346e-07,
+      "loss": 0.9787,
+      "step": 8629
+    },
+    {
+      "epoch": 0.9076209131423614,
+      "grad_norm": 2.7356277181692445,
+      "learning_rate": 1.0668195387395164e-07,
+      "loss": 1.0438,
+      "step": 8630
+    },
+    {
+      "epoch": 0.9077260835842087,
+      "grad_norm": 2.3404693898022777,
+      "learning_rate": 1.0644081929162275e-07,
+      "loss": 0.9924,
+      "step": 8631
+    },
+    {
+      "epoch": 0.907831254026056,
+      "grad_norm": 2.1474982707381893,
+      "learning_rate": 1.0619995160973645e-07,
+      "loss": 0.9078,
+      "step": 8632
+    },
+    {
+      "epoch": 0.9079364244679033,
+      "grad_norm": 2.233221210407833,
+      "learning_rate": 1.0595935085515069e-07,
+      "loss": 1.0021,
+      "step": 8633
+    },
+    {
+      "epoch": 0.9080415949097506,
+      "grad_norm": 2.707217846445669,
+      "learning_rate": 1.0571901705469567e-07,
+      "loss": 1.0373,
+      "step": 8634
+    },
+    {
+      "epoch": 0.9081467653515979,
+      "grad_norm": 2.0945684851181987,
+      "learning_rate": 1.0547895023516913e-07,
+      "loss": 0.9887,
+      "step": 8635
+    },
+    {
+      "epoch": 0.9082519357934452,
+      "grad_norm": 2.818236365087763,
+      "learning_rate": 1.0523915042334132e-07,
+      "loss": 0.9345,
+      "step": 8636
+    },
+    {
+      "epoch": 0.9083571062352925,
+      "grad_norm": 2.724573085747091,
+      "learning_rate": 1.0499961764595112e-07,
+      "loss": 1.036,
+      "step": 8637
+    },
+    {
+      "epoch": 0.9084622766771399,
+      "grad_norm": 2.706854585553475,
+      "learning_rate": 1.0476035192970857e-07,
+      "loss": 0.953,
+      "step": 8638
+    },
+    {
+      "epoch": 0.9085674471189872,
+      "grad_norm": 1.9228244106745358,
+      "learning_rate": 1.045213533012937e-07,
+      "loss": 0.9746,
+      "step": 8639
+    },
+    {
+      "epoch": 0.9086726175608345,
+      "grad_norm": 2.5814153527028973,
+      "learning_rate": 1.042826217873566e-07,
+      "loss": 0.9523,
+      "step": 8640
+    },
+    {
+      "epoch": 0.9087777880026818,
+      "grad_norm": 2.9607847333673694,
+      "learning_rate": 1.0404415741451818e-07,
+      "loss": 0.9877,
+      "step": 8641
+    },
+    {
+      "epoch": 0.9088829584445292,
+      "grad_norm": 2.9016091841556984,
+      "learning_rate": 1.0380596020936801e-07,
+      "loss": 1.0021,
+      "step": 8642
+    },
+    {
+      "epoch": 0.9089881288863765,
+      "grad_norm": 2.633994800512432,
+      "learning_rate": 1.035680301984679e-07,
+      "loss": 0.9496,
+      "step": 8643
+    },
+    {
+      "epoch": 0.9090932993282238,
+      "grad_norm": 3.3617291767747357,
+      "learning_rate": 1.0333036740834857e-07,
+      "loss": 0.9403,
+      "step": 8644
+    },
+    {
+      "epoch": 0.9091984697700711,
+      "grad_norm": 2.677523368933197,
+      "learning_rate": 1.0309297186551131e-07,
+      "loss": 0.9678,
+      "step": 8645
+    },
+    {
+      "epoch": 0.9093036402119185,
+      "grad_norm": 2.373476234340472,
+      "learning_rate": 1.0285584359642747e-07,
+      "loss": 0.9627,
+      "step": 8646
+    },
+    {
+      "epoch": 0.9094088106537658,
+      "grad_norm": 2.0766733599038845,
+      "learning_rate": 1.0261898262753811e-07,
+      "loss": 0.9632,
+      "step": 8647
+    },
+    {
+      "epoch": 0.9095139810956131,
+      "grad_norm": 2.4397958236252553,
+      "learning_rate": 1.0238238898525654e-07,
+      "loss": 0.9787,
+      "step": 8648
+    },
+    {
+      "epoch": 0.9096191515374604,
+      "grad_norm": 2.429982084341817,
+      "learning_rate": 1.0214606269596361e-07,
+      "loss": 0.9983,
+      "step": 8649
+    },
+    {
+      "epoch": 0.9097243219793077,
+      "grad_norm": 1.7677248006677606,
+      "learning_rate": 1.0191000378601213e-07,
+      "loss": 0.9348,
+      "step": 8650
+    },
+    {
+      "epoch": 0.9098294924211551,
+      "grad_norm": 1.8993802694959563,
+      "learning_rate": 1.0167421228172381e-07,
+      "loss": 1.0176,
+      "step": 8651
+    },
+    {
+      "epoch": 0.9099346628630024,
+      "grad_norm": 2.6748517834984593,
+      "learning_rate": 1.0143868820939179e-07,
+      "loss": 0.986,
+      "step": 8652
+    },
+    {
+      "epoch": 0.9100398333048497,
+      "grad_norm": 2.622839246677593,
+      "learning_rate": 1.0120343159527923e-07,
+      "loss": 0.9792,
+      "step": 8653
+    },
+    {
+      "epoch": 0.910145003746697,
+      "grad_norm": 1.9757068387048435,
+      "learning_rate": 1.0096844246561794e-07,
+      "loss": 0.9901,
+      "step": 8654
+    },
+    {
+      "epoch": 0.9102501741885443,
+      "grad_norm": 2.095050670569624,
+      "learning_rate": 1.0073372084661193e-07,
+      "loss": 0.9605,
+      "step": 8655
+    },
+    {
+      "epoch": 0.9103553446303916,
+      "grad_norm": 2.750980361243624,
+      "learning_rate": 1.0049926676443361e-07,
+      "loss": 0.9974,
+      "step": 8656
+    },
+    {
+      "epoch": 0.9104605150722389,
+      "grad_norm": 2.961100634051163,
+      "learning_rate": 1.0026508024522791e-07,
+      "loss": 0.981,
+      "step": 8657
+    },
+    {
+      "epoch": 0.9105656855140862,
+      "grad_norm": 1.8783234007957286,
+      "learning_rate": 1.0003116131510698e-07,
+      "loss": 0.989,
+      "step": 8658
+    },
+    {
+      "epoch": 0.9106708559559336,
+      "grad_norm": 2.671085261106148,
+      "learning_rate": 9.979751000015552e-08,
+      "loss": 0.9828,
+      "step": 8659
+    },
+    {
+      "epoch": 0.9107760263977809,
+      "grad_norm": 3.0384313224763737,
+      "learning_rate": 9.956412632642715e-08,
+      "loss": 0.9908,
+      "step": 8660
+    },
+    {
+      "epoch": 0.9108811968396282,
+      "grad_norm": 2.5512974558431227,
+      "learning_rate": 9.933101031994547e-08,
+      "loss": 0.9434,
+      "step": 8661
+    },
+    {
+      "epoch": 0.9109863672814755,
+      "grad_norm": 1.737662256811027,
+      "learning_rate": 9.909816200670552e-08,
+      "loss": 0.9693,
+      "step": 8662
+    },
+    {
+      "epoch": 0.9110915377233229,
+      "grad_norm": 3.0073945279323833,
+      "learning_rate": 9.886558141267127e-08,
+      "loss": 0.9521,
+      "step": 8663
+    },
+    {
+      "epoch": 0.9111967081651702,
+      "grad_norm": 1.884812573575343,
+      "learning_rate": 9.863326856377753e-08,
+      "loss": 0.9782,
+      "step": 8664
+    },
+    {
+      "epoch": 0.9113018786070175,
+      "grad_norm": 2.722404955393905,
+      "learning_rate": 9.840122348592857e-08,
+      "loss": 0.9983,
+      "step": 8665
+    },
+    {
+      "epoch": 0.9114070490488648,
+      "grad_norm": 2.5101444576387344,
+      "learning_rate": 9.816944620499952e-08,
+      "loss": 0.9915,
+      "step": 8666
+    },
+    {
+      "epoch": 0.9115122194907122,
+      "grad_norm": 2.763819589917786,
+      "learning_rate": 9.793793674683555e-08,
+      "loss": 1.0128,
+      "step": 8667
+    },
+    {
+      "epoch": 0.9116173899325595,
+      "grad_norm": 2.690122259083069,
+      "learning_rate": 9.770669513725128e-08,
+      "loss": 1.0093,
+      "step": 8668
+    },
+    {
+      "epoch": 0.9117225603744068,
+      "grad_norm": 2.5984553040911003,
+      "learning_rate": 9.747572140203221e-08,
+      "loss": 0.9857,
+      "step": 8669
+    },
+    {
+      "epoch": 0.9118277308162541,
+      "grad_norm": 2.3984861025276194,
+      "learning_rate": 9.724501556693327e-08,
+      "loss": 1.0068,
+      "step": 8670
+    },
+    {
+      "epoch": 0.9119329012581014,
+      "grad_norm": 3.124841252252845,
+      "learning_rate": 9.701457765768113e-08,
+      "loss": 0.9785,
+      "step": 8671
+    },
+    {
+      "epoch": 0.9120380716999488,
+      "grad_norm": 2.377091759025528,
+      "learning_rate": 9.678440769996994e-08,
+      "loss": 0.9923,
+      "step": 8672
+    },
+    {
+      "epoch": 0.9121432421417961,
+      "grad_norm": 2.439191969177443,
+      "learning_rate": 9.655450571946667e-08,
+      "loss": 0.9754,
+      "step": 8673
+    },
+    {
+      "epoch": 0.9122484125836434,
+      "grad_norm": 2.3851721865392355,
+      "learning_rate": 9.632487174180638e-08,
+      "loss": 0.9982,
+      "step": 8674
+    },
+    {
+      "epoch": 0.9123535830254906,
+      "grad_norm": 2.172576222886981,
+      "learning_rate": 9.609550579259497e-08,
+      "loss": 0.9679,
+      "step": 8675
+    },
+    {
+      "epoch": 0.912458753467338,
+      "grad_norm": 2.301925417223173,
+      "learning_rate": 9.586640789740948e-08,
+      "loss": 0.9739,
+      "step": 8676
+    },
+    {
+      "epoch": 0.9125639239091853,
+      "grad_norm": 2.478887405016733,
+      "learning_rate": 9.563757808179502e-08,
+      "loss": 1.0417,
+      "step": 8677
+    },
+    {
+      "epoch": 0.9126690943510326,
+      "grad_norm": 2.7317447487137407,
+      "learning_rate": 9.54090163712687e-08,
+      "loss": 0.9742,
+      "step": 8678
+    },
+    {
+      "epoch": 0.9127742647928799,
+      "grad_norm": 2.7528530250768046,
+      "learning_rate": 9.518072279131596e-08,
+      "loss": 0.9546,
+      "step": 8679
+    },
+    {
+      "epoch": 0.9128794352347273,
+      "grad_norm": 2.128520981516656,
+      "learning_rate": 9.495269736739448e-08,
+      "loss": 0.9033,
+      "step": 8680
+    },
+    {
+      "epoch": 0.9129846056765746,
+      "grad_norm": 2.8260209548553346,
+      "learning_rate": 9.472494012493034e-08,
+      "loss": 0.9833,
+      "step": 8681
+    },
+    {
+      "epoch": 0.9130897761184219,
+      "grad_norm": 2.710474999545366,
+      "learning_rate": 9.449745108931985e-08,
+      "loss": 0.9626,
+      "step": 8682
+    },
+    {
+      "epoch": 0.9131949465602692,
+      "grad_norm": 2.4800110937909734,
+      "learning_rate": 9.427023028593051e-08,
+      "loss": 0.9755,
+      "step": 8683
+    },
+    {
+      "epoch": 0.9133001170021166,
+      "grad_norm": 2.3592614302731683,
+      "learning_rate": 9.404327774009819e-08,
+      "loss": 0.9557,
+      "step": 8684
+    },
+    {
+      "epoch": 0.9134052874439639,
+      "grad_norm": 2.9956075352373364,
+      "learning_rate": 9.381659347713123e-08,
+      "loss": 0.9975,
+      "step": 8685
+    },
+    {
+      "epoch": 0.9135104578858112,
+      "grad_norm": 2.1831281057025613,
+      "learning_rate": 9.359017752230582e-08,
+      "loss": 0.9784,
+      "step": 8686
+    },
+    {
+      "epoch": 0.9136156283276585,
+      "grad_norm": 2.2725775975518148,
+      "learning_rate": 9.336402990086924e-08,
+      "loss": 1.0201,
+      "step": 8687
+    },
+    {
+      "epoch": 0.9137207987695058,
+      "grad_norm": 2.794970120659367,
+      "learning_rate": 9.313815063803883e-08,
+      "loss": 0.927,
+      "step": 8688
+    },
+    {
+      "epoch": 0.9138259692113532,
+      "grad_norm": 2.447763388249118,
+      "learning_rate": 9.291253975900138e-08,
+      "loss": 0.9516,
+      "step": 8689
+    },
+    {
+      "epoch": 0.9139311396532005,
+      "grad_norm": 2.5939870031479773,
+      "learning_rate": 9.268719728891512e-08,
+      "loss": 1.0149,
+      "step": 8690
+    },
+    {
+      "epoch": 0.9140363100950478,
+      "grad_norm": 2.5305551362373864,
+      "learning_rate": 9.24621232529066e-08,
+      "loss": 1.0017,
+      "step": 8691
+    },
+    {
+      "epoch": 0.9141414805368951,
+      "grad_norm": 2.901219705340327,
+      "learning_rate": 9.223731767607436e-08,
+      "loss": 0.9446,
+      "step": 8692
+    },
+    {
+      "epoch": 0.9142466509787425,
+      "grad_norm": 2.353334916914352,
+      "learning_rate": 9.201278058348446e-08,
+      "loss": 0.9655,
+      "step": 8693
+    },
+    {
+      "epoch": 0.9143518214205898,
+      "grad_norm": 2.2889584109431094,
+      "learning_rate": 9.17885120001763e-08,
+      "loss": 1.0182,
+      "step": 8694
+    },
+    {
+      "epoch": 0.914456991862437,
+      "grad_norm": 2.5187641418281723,
+      "learning_rate": 9.156451195115601e-08,
+      "loss": 0.9578,
+      "step": 8695
+    },
+    {
+      "epoch": 0.9145621623042843,
+      "grad_norm": 2.466078538297683,
+      "learning_rate": 9.134078046140249e-08,
+      "loss": 0.949,
+      "step": 8696
+    },
+    {
+      "epoch": 0.9146673327461317,
+      "grad_norm": 1.969452495220898,
+      "learning_rate": 9.111731755586329e-08,
+      "loss": 0.9636,
+      "step": 8697
+    },
+    {
+      "epoch": 0.914772503187979,
+      "grad_norm": 2.2396101458216187,
+      "learning_rate": 9.089412325945513e-08,
+      "loss": 0.9892,
+      "step": 8698
+    },
+    {
+      "epoch": 0.9148776736298263,
+      "grad_norm": 2.3519886141123956,
+      "learning_rate": 9.067119759706755e-08,
+      "loss": 0.9746,
+      "step": 8699
+    },
+    {
+      "epoch": 0.9149828440716736,
+      "grad_norm": 2.4208396168966275,
+      "learning_rate": 9.044854059355763e-08,
+      "loss": 1.05,
+      "step": 8700
+    },
+    {
+      "epoch": 0.915088014513521,
+      "grad_norm": 3.9523038988472425,
+      "learning_rate": 9.022615227375353e-08,
+      "loss": 1.0266,
+      "step": 8701
+    },
+    {
+      "epoch": 0.9151931849553683,
+      "grad_norm": 2.002467211998927,
+      "learning_rate": 9.000403266245294e-08,
+      "loss": 0.9787,
+      "step": 8702
+    },
+    {
+      "epoch": 0.9152983553972156,
+      "grad_norm": 2.0671531516467776,
+      "learning_rate": 8.978218178442439e-08,
+      "loss": 0.9251,
+      "step": 8703
+    },
+    {
+      "epoch": 0.9154035258390629,
+      "grad_norm": 2.7026551909268046,
+      "learning_rate": 8.956059966440583e-08,
+      "loss": 1.0106,
+      "step": 8704
+    },
+    {
+      "epoch": 0.9155086962809103,
+      "grad_norm": 2.3171869376715213,
+      "learning_rate": 8.933928632710531e-08,
+      "loss": 0.9916,
+      "step": 8705
+    },
+    {
+      "epoch": 0.9156138667227576,
+      "grad_norm": 2.317933352612227,
+      "learning_rate": 8.911824179720113e-08,
+      "loss": 0.973,
+      "step": 8706
+    },
+    {
+      "epoch": 0.9157190371646049,
+      "grad_norm": 2.524849884723814,
+      "learning_rate": 8.889746609934108e-08,
+      "loss": 1.0246,
+      "step": 8707
+    },
+    {
+      "epoch": 0.9158242076064522,
+      "grad_norm": 2.8343370527261627,
+      "learning_rate": 8.867695925814407e-08,
+      "loss": 0.9862,
+      "step": 8708
+    },
+    {
+      "epoch": 0.9159293780482995,
+      "grad_norm": 2.2205891500846002,
+      "learning_rate": 8.845672129819766e-08,
+      "loss": 0.9454,
+      "step": 8709
+    },
+    {
+      "epoch": 0.9160345484901469,
+      "grad_norm": 2.624543266262543,
+      "learning_rate": 8.823675224406052e-08,
+      "loss": 1.0148,
+      "step": 8710
+    },
+    {
+      "epoch": 0.9161397189319942,
+      "grad_norm": 1.924274060643869,
+      "learning_rate": 8.801705212026058e-08,
+      "loss": 0.9628,
+      "step": 8711
+    },
+    {
+      "epoch": 0.9162448893738415,
+      "grad_norm": 2.6357976887412855,
+      "learning_rate": 8.779762095129623e-08,
+      "loss": 0.9868,
+      "step": 8712
+    },
+    {
+      "epoch": 0.9163500598156888,
+      "grad_norm": 2.201860694431399,
+      "learning_rate": 8.757845876163601e-08,
+      "loss": 0.9836,
+      "step": 8713
+    },
+    {
+      "epoch": 0.9164552302575362,
+      "grad_norm": 2.6034939502760923,
+      "learning_rate": 8.735956557571785e-08,
+      "loss": 0.9164,
+      "step": 8714
+    },
+    {
+      "epoch": 0.9165604006993835,
+      "grad_norm": 2.7455401077210997,
+      "learning_rate": 8.714094141795026e-08,
+      "loss": 1.0081,
+      "step": 8715
+    },
+    {
+      "epoch": 0.9166655711412307,
+      "grad_norm": 3.016771299173225,
+      "learning_rate": 8.692258631271127e-08,
+      "loss": 0.928,
+      "step": 8716
+    },
+    {
+      "epoch": 0.916770741583078,
+      "grad_norm": 2.072039917472285,
+      "learning_rate": 8.670450028434946e-08,
+      "loss": 0.9588,
+      "step": 8717
+    },
+    {
+      "epoch": 0.9168759120249254,
+      "grad_norm": 2.987620512435773,
+      "learning_rate": 8.648668335718313e-08,
+      "loss": 0.9895,
+      "step": 8718
+    },
+    {
+      "epoch": 0.9169810824667727,
+      "grad_norm": 2.5553475573583873,
+      "learning_rate": 8.626913555550009e-08,
+      "loss": 0.9904,
+      "step": 8719
+    },
+    {
+      "epoch": 0.91708625290862,
+      "grad_norm": 3.2548118763612157,
+      "learning_rate": 8.605185690355927e-08,
+      "loss": 1.0137,
+      "step": 8720
+    },
+    {
+      "epoch": 0.9171914233504673,
+      "grad_norm": 2.524991929264644,
+      "learning_rate": 8.583484742558823e-08,
+      "loss": 0.9857,
+      "step": 8721
+    },
+    {
+      "epoch": 0.9172965937923147,
+      "grad_norm": 3.326075876639707,
+      "learning_rate": 8.561810714578595e-08,
+      "loss": 0.9922,
+      "step": 8722
+    },
+    {
+      "epoch": 0.917401764234162,
+      "grad_norm": 2.06394387394277,
+      "learning_rate": 8.54016360883203e-08,
+      "loss": 0.9517,
+      "step": 8723
+    },
+    {
+      "epoch": 0.9175069346760093,
+      "grad_norm": 2.7063497472638707,
+      "learning_rate": 8.518543427732951e-08,
+      "loss": 1.0339,
+      "step": 8724
+    },
+    {
+      "epoch": 0.9176121051178566,
+      "grad_norm": 2.6198469236901722,
+      "learning_rate": 8.496950173692147e-08,
+      "loss": 0.9994,
+      "step": 8725
+    },
+    {
+      "epoch": 0.917717275559704,
+      "grad_norm": 2.0768494403671203,
+      "learning_rate": 8.475383849117474e-08,
+      "loss": 0.9819,
+      "step": 8726
+    },
+    {
+      "epoch": 0.9178224460015513,
+      "grad_norm": 2.0259311339586747,
+      "learning_rate": 8.453844456413729e-08,
+      "loss": 0.9839,
+      "step": 8727
+    },
+    {
+      "epoch": 0.9179276164433986,
+      "grad_norm": 3.4906371799996205,
+      "learning_rate": 8.43233199798274e-08,
+      "loss": 0.9509,
+      "step": 8728
+    },
+    {
+      "epoch": 0.9180327868852459,
+      "grad_norm": 2.233812559283946,
+      "learning_rate": 8.410846476223283e-08,
+      "loss": 1.001,
+      "step": 8729
+    },
+    {
+      "epoch": 0.9181379573270932,
+      "grad_norm": 2.680065549787725,
+      "learning_rate": 8.389387893531137e-08,
+      "loss": 0.9476,
+      "step": 8730
+    },
+    {
+      "epoch": 0.9182431277689406,
+      "grad_norm": 1.8569584887135309,
+      "learning_rate": 8.367956252299192e-08,
+      "loss": 0.9769,
+      "step": 8731
+    },
+    {
+      "epoch": 0.9183482982107879,
+      "grad_norm": 2.3494913181215695,
+      "learning_rate": 8.346551554917204e-08,
+      "loss": 0.9841,
+      "step": 8732
+    },
+    {
+      "epoch": 0.9184534686526352,
+      "grad_norm": 2.450100459990453,
+      "learning_rate": 8.325173803771902e-08,
+      "loss": 0.9682,
+      "step": 8733
+    },
+    {
+      "epoch": 0.9185586390944825,
+      "grad_norm": 2.3534886505356836,
+      "learning_rate": 8.303823001247102e-08,
+      "loss": 0.9905,
+      "step": 8734
+    },
+    {
+      "epoch": 0.9186638095363299,
+      "grad_norm": 2.0769040702020463,
+      "learning_rate": 8.282499149723622e-08,
+      "loss": 0.9871,
+      "step": 8735
+    },
+    {
+      "epoch": 0.9187689799781771,
+      "grad_norm": 2.328949116074991,
+      "learning_rate": 8.261202251579253e-08,
+      "loss": 0.9732,
+      "step": 8736
+    },
+    {
+      "epoch": 0.9188741504200244,
+      "grad_norm": 2.3004852069789754,
+      "learning_rate": 8.239932309188681e-08,
+      "loss": 0.9529,
+      "step": 8737
+    },
+    {
+      "epoch": 0.9189793208618717,
+      "grad_norm": 2.467224662817621,
+      "learning_rate": 8.218689324923729e-08,
+      "loss": 0.9588,
+      "step": 8738
+    },
+    {
+      "epoch": 0.919084491303719,
+      "grad_norm": 2.2727911833425076,
+      "learning_rate": 8.197473301153142e-08,
+      "loss": 1.0025,
+      "step": 8739
+    },
+    {
+      "epoch": 0.9191896617455664,
+      "grad_norm": 2.8021991978145455,
+      "learning_rate": 8.176284240242638e-08,
+      "loss": 1.0289,
+      "step": 8740
+    },
+    {
+      "epoch": 0.9192948321874137,
+      "grad_norm": 2.5953858991707492,
+      "learning_rate": 8.15512214455505e-08,
+      "loss": 0.9927,
+      "step": 8741
+    },
+    {
+      "epoch": 0.919400002629261,
+      "grad_norm": 2.374685117659884,
+      "learning_rate": 8.133987016450018e-08,
+      "loss": 0.9481,
+      "step": 8742
+    },
+    {
+      "epoch": 0.9195051730711084,
+      "grad_norm": 2.722786813359639,
+      "learning_rate": 8.112878858284351e-08,
+      "loss": 0.9962,
+      "step": 8743
+    },
+    {
+      "epoch": 0.9196103435129557,
+      "grad_norm": 2.843740539019621,
+      "learning_rate": 8.091797672411666e-08,
+      "loss": 0.9894,
+      "step": 8744
+    },
+    {
+      "epoch": 0.919715513954803,
+      "grad_norm": 2.462450080595821,
+      "learning_rate": 8.070743461182807e-08,
+      "loss": 0.9424,
+      "step": 8745
+    },
+    {
+      "epoch": 0.9198206843966503,
+      "grad_norm": 2.200892646883302,
+      "learning_rate": 8.04971622694542e-08,
+      "loss": 0.9761,
+      "step": 8746
+    },
+    {
+      "epoch": 0.9199258548384976,
+      "grad_norm": 1.8444502246648864,
+      "learning_rate": 8.028715972044216e-08,
+      "loss": 0.9718,
+      "step": 8747
+    },
+    {
+      "epoch": 0.920031025280345,
+      "grad_norm": 2.6368564963597954,
+      "learning_rate": 8.007742698820848e-08,
+      "loss": 0.9807,
+      "step": 8748
+    },
+    {
+      "epoch": 0.9201361957221923,
+      "grad_norm": 2.85801832795167,
+      "learning_rate": 7.986796409614028e-08,
+      "loss": 0.9885,
+      "step": 8749
+    },
+    {
+      "epoch": 0.9202413661640396,
+      "grad_norm": 2.2819001687005476,
+      "learning_rate": 7.965877106759473e-08,
+      "loss": 0.9686,
+      "step": 8750
+    },
+    {
+      "epoch": 0.920346536605887,
+      "grad_norm": 2.506401312891756,
+      "learning_rate": 7.944984792589788e-08,
+      "loss": 1.021,
+      "step": 8751
+    },
+    {
+      "epoch": 0.9204517070477343,
+      "grad_norm": 2.446263624804167,
+      "learning_rate": 7.924119469434666e-08,
+      "loss": 1.0275,
+      "step": 8752
+    },
+    {
+      "epoch": 0.9205568774895816,
+      "grad_norm": 2.557300434643533,
+      "learning_rate": 7.90328113962069e-08,
+      "loss": 0.9428,
+      "step": 8753
+    },
+    {
+      "epoch": 0.9206620479314289,
+      "grad_norm": 2.3857984916068538,
+      "learning_rate": 7.882469805471582e-08,
+      "loss": 1.0545,
+      "step": 8754
+    },
+    {
+      "epoch": 0.9207672183732762,
+      "grad_norm": 2.7202727116635312,
+      "learning_rate": 7.861685469307905e-08,
+      "loss": 1.0019,
+      "step": 8755
+    },
+    {
+      "epoch": 0.9208723888151235,
+      "grad_norm": 1.5934819593693805,
+      "learning_rate": 7.840928133447306e-08,
+      "loss": 0.9311,
+      "step": 8756
+    },
+    {
+      "epoch": 0.9209775592569708,
+      "grad_norm": 2.9845596688934655,
+      "learning_rate": 7.82019780020435e-08,
+      "loss": 1.0253,
+      "step": 8757
+    },
+    {
+      "epoch": 0.9210827296988181,
+      "grad_norm": 2.046747867064963,
+      "learning_rate": 7.799494471890684e-08,
+      "loss": 1.0192,
+      "step": 8758
+    },
+    {
+      "epoch": 0.9211879001406654,
+      "grad_norm": 2.5073020601329485,
+      "learning_rate": 7.778818150814854e-08,
+      "loss": 0.9559,
+      "step": 8759
+    },
+    {
+      "epoch": 0.9212930705825128,
+      "grad_norm": 2.5720686156196897,
+      "learning_rate": 7.75816883928246e-08,
+      "loss": 1.013,
+      "step": 8760
+    },
+    {
+      "epoch": 0.9213982410243601,
+      "grad_norm": 2.4862086127778675,
+      "learning_rate": 7.73754653959602e-08,
+      "loss": 0.9639,
+      "step": 8761
+    },
+    {
+      "epoch": 0.9215034114662074,
+      "grad_norm": 2.364000470048889,
+      "learning_rate": 7.716951254055111e-08,
+      "loss": 0.9796,
+      "step": 8762
+    },
+    {
+      "epoch": 0.9216085819080547,
+      "grad_norm": 1.8736154182919098,
+      "learning_rate": 7.69638298495623e-08,
+      "loss": 0.9932,
+      "step": 8763
+    },
+    {
+      "epoch": 0.921713752349902,
+      "grad_norm": 2.178955434285585,
+      "learning_rate": 7.675841734592987e-08,
+      "loss": 1.0271,
+      "step": 8764
+    },
+    {
+      "epoch": 0.9218189227917494,
+      "grad_norm": 3.1224740330325993,
+      "learning_rate": 7.655327505255772e-08,
+      "loss": 0.9938,
+      "step": 8765
+    },
+    {
+      "epoch": 0.9219240932335967,
+      "grad_norm": 2.6693925921641553,
+      "learning_rate": 7.634840299232171e-08,
+      "loss": 0.9828,
+      "step": 8766
+    },
+    {
+      "epoch": 0.922029263675444,
+      "grad_norm": 2.7013624068102207,
+      "learning_rate": 7.614380118806636e-08,
+      "loss": 0.9895,
+      "step": 8767
+    },
+    {
+      "epoch": 0.9221344341172913,
+      "grad_norm": 2.201528972225391,
+      "learning_rate": 7.593946966260618e-08,
+      "loss": 0.9644,
+      "step": 8768
+    },
+    {
+      "epoch": 0.9222396045591387,
+      "grad_norm": 2.3707667253047697,
+      "learning_rate": 7.573540843872602e-08,
+      "loss": 1.0226,
+      "step": 8769
+    },
+    {
+      "epoch": 0.922344775000986,
+      "grad_norm": 2.1492440827092714,
+      "learning_rate": 7.553161753918015e-08,
+      "loss": 0.9661,
+      "step": 8770
+    },
+    {
+      "epoch": 0.9224499454428333,
+      "grad_norm": 2.1183442817355154,
+      "learning_rate": 7.532809698669263e-08,
+      "loss": 0.997,
+      "step": 8771
+    },
+    {
+      "epoch": 0.9225551158846806,
+      "grad_norm": 2.0129347965512996,
+      "learning_rate": 7.512484680395782e-08,
+      "loss": 0.9551,
+      "step": 8772
+    },
+    {
+      "epoch": 0.922660286326528,
+      "grad_norm": 3.3447687257492644,
+      "learning_rate": 7.492186701364007e-08,
+      "loss": 0.9852,
+      "step": 8773
+    },
+    {
+      "epoch": 0.9227654567683753,
+      "grad_norm": 2.5874808124779007,
+      "learning_rate": 7.471915763837268e-08,
+      "loss": 0.9863,
+      "step": 8774
+    },
+    {
+      "epoch": 0.9228706272102226,
+      "grad_norm": 2.247477950163746,
+      "learning_rate": 7.451671870075949e-08,
+      "loss": 0.9511,
+      "step": 8775
+    },
+    {
+      "epoch": 0.9229757976520699,
+      "grad_norm": 2.349577908913971,
+      "learning_rate": 7.431455022337386e-08,
+      "loss": 0.9971,
+      "step": 8776
+    },
+    {
+      "epoch": 0.9230809680939172,
+      "grad_norm": 2.4880434952386934,
+      "learning_rate": 7.411265222875913e-08,
+      "loss": 0.9879,
+      "step": 8777
+    },
+    {
+      "epoch": 0.9231861385357645,
+      "grad_norm": 2.6428533811226704,
+      "learning_rate": 7.391102473942897e-08,
+      "loss": 0.9675,
+      "step": 8778
+    },
+    {
+      "epoch": 0.9232913089776118,
+      "grad_norm": 2.2938047586552366,
+      "learning_rate": 7.370966777786564e-08,
+      "loss": 0.9626,
+      "step": 8779
+    },
+    {
+      "epoch": 0.9233964794194591,
+      "grad_norm": 2.684986649020841,
+      "learning_rate": 7.350858136652262e-08,
+      "loss": 1.0519,
+      "step": 8780
+    },
+    {
+      "epoch": 0.9235016498613065,
+      "grad_norm": 3.0768759986996947,
+      "learning_rate": 7.330776552782248e-08,
+      "loss": 0.9541,
+      "step": 8781
+    },
+    {
+      "epoch": 0.9236068203031538,
+      "grad_norm": 2.480420721077424,
+      "learning_rate": 7.310722028415762e-08,
+      "loss": 0.9568,
+      "step": 8782
+    },
+    {
+      "epoch": 0.9237119907450011,
+      "grad_norm": 2.7358308672297107,
+      "learning_rate": 7.290694565789069e-08,
+      "loss": 0.9606,
+      "step": 8783
+    },
+    {
+      "epoch": 0.9238171611868484,
+      "grad_norm": 2.4196600463864177,
+      "learning_rate": 7.270694167135356e-08,
+      "loss": 0.9517,
+      "step": 8784
+    },
+    {
+      "epoch": 0.9239223316286957,
+      "grad_norm": 3.033022762994752,
+      "learning_rate": 7.25072083468481e-08,
+      "loss": 1.033,
+      "step": 8785
+    },
+    {
+      "epoch": 0.9240275020705431,
+      "grad_norm": 2.710152700629641,
+      "learning_rate": 7.230774570664623e-08,
+      "loss": 0.9857,
+      "step": 8786
+    },
+    {
+      "epoch": 0.9241326725123904,
+      "grad_norm": 2.4433124063205027,
+      "learning_rate": 7.210855377299014e-08,
+      "loss": 0.9722,
+      "step": 8787
+    },
+    {
+      "epoch": 0.9242378429542377,
+      "grad_norm": 3.0106631380990425,
+      "learning_rate": 7.190963256809069e-08,
+      "loss": 0.9721,
+      "step": 8788
+    },
+    {
+      "epoch": 0.924343013396085,
+      "grad_norm": 2.482926124872465,
+      "learning_rate": 7.17109821141293e-08,
+      "loss": 0.9866,
+      "step": 8789
+    },
+    {
+      "epoch": 0.9244481838379324,
+      "grad_norm": 2.4915008975300927,
+      "learning_rate": 7.151260243325686e-08,
+      "loss": 1.024,
+      "step": 8790
+    },
+    {
+      "epoch": 0.9245533542797797,
+      "grad_norm": 2.9578504303961055,
+      "learning_rate": 7.13144935475943e-08,
+      "loss": 0.9146,
+      "step": 8791
+    },
+    {
+      "epoch": 0.924658524721627,
+      "grad_norm": 3.1081419272442856,
+      "learning_rate": 7.111665547923252e-08,
+      "loss": 0.9885,
+      "step": 8792
+    },
+    {
+      "epoch": 0.9247636951634743,
+      "grad_norm": 2.3985246667372544,
+      "learning_rate": 7.091908825023197e-08,
+      "loss": 1.0025,
+      "step": 8793
+    },
+    {
+      "epoch": 0.9248688656053217,
+      "grad_norm": 2.428656822228575,
+      "learning_rate": 7.072179188262252e-08,
+      "loss": 0.9968,
+      "step": 8794
+    },
+    {
+      "epoch": 0.924974036047169,
+      "grad_norm": 2.255455390801362,
+      "learning_rate": 7.052476639840489e-08,
+      "loss": 0.9685,
+      "step": 8795
+    },
+    {
+      "epoch": 0.9250792064890163,
+      "grad_norm": 2.3473107950280894,
+      "learning_rate": 7.032801181954873e-08,
+      "loss": 0.9967,
+      "step": 8796
+    },
+    {
+      "epoch": 0.9251843769308635,
+      "grad_norm": 2.762465947655461,
+      "learning_rate": 7.013152816799317e-08,
+      "loss": 1.0137,
+      "step": 8797
+    },
+    {
+      "epoch": 0.9252895473727109,
+      "grad_norm": 2.082339717782587,
+      "learning_rate": 6.993531546564874e-08,
+      "loss": 0.9553,
+      "step": 8798
+    },
+    {
+      "epoch": 0.9253947178145582,
+      "grad_norm": 2.4355695449007193,
+      "learning_rate": 6.973937373439349e-08,
+      "loss": 0.9761,
+      "step": 8799
+    },
+    {
+      "epoch": 0.9254998882564055,
+      "grad_norm": 2.3323761744575453,
+      "learning_rate": 6.954370299607715e-08,
+      "loss": 0.9978,
+      "step": 8800
+    },
+    {
+      "epoch": 0.9256050586982528,
+      "grad_norm": 2.7276265273799636,
+      "learning_rate": 6.93483032725184e-08,
+      "loss": 1.014,
+      "step": 8801
+    },
+    {
+      "epoch": 0.9257102291401001,
+      "grad_norm": 2.451649200404242,
+      "learning_rate": 6.91531745855059e-08,
+      "loss": 0.9806,
+      "step": 8802
+    },
+    {
+      "epoch": 0.9258153995819475,
+      "grad_norm": 2.3782834691727186,
+      "learning_rate": 6.895831695679756e-08,
+      "loss": 0.9996,
+      "step": 8803
+    },
+    {
+      "epoch": 0.9259205700237948,
+      "grad_norm": 2.725803550156134,
+      "learning_rate": 6.876373040812234e-08,
+      "loss": 0.9837,
+      "step": 8804
+    },
+    {
+      "epoch": 0.9260257404656421,
+      "grad_norm": 1.692896666258162,
+      "learning_rate": 6.856941496117736e-08,
+      "loss": 0.9844,
+      "step": 8805
+    },
+    {
+      "epoch": 0.9261309109074894,
+      "grad_norm": 2.627960374778733,
+      "learning_rate": 6.837537063763083e-08,
+      "loss": 0.9603,
+      "step": 8806
+    },
+    {
+      "epoch": 0.9262360813493368,
+      "grad_norm": 2.4123212586639045,
+      "learning_rate": 6.818159745911989e-08,
+      "loss": 0.9272,
+      "step": 8807
+    },
+    {
+      "epoch": 0.9263412517911841,
+      "grad_norm": 2.672047655684633,
+      "learning_rate": 6.798809544725171e-08,
+      "loss": 0.9814,
+      "step": 8808
+    },
+    {
+      "epoch": 0.9264464222330314,
+      "grad_norm": 2.8051273113292434,
+      "learning_rate": 6.779486462360346e-08,
+      "loss": 0.9619,
+      "step": 8809
+    },
+    {
+      "epoch": 0.9265515926748787,
+      "grad_norm": 1.930413903321851,
+      "learning_rate": 6.760190500972208e-08,
+      "loss": 0.9405,
+      "step": 8810
+    },
+    {
+      "epoch": 0.9266567631167261,
+      "grad_norm": 2.0349253679776975,
+      "learning_rate": 6.740921662712368e-08,
+      "loss": 0.9864,
+      "step": 8811
+    },
+    {
+      "epoch": 0.9267619335585734,
+      "grad_norm": 2.6944705175063954,
+      "learning_rate": 6.721679949729499e-08,
+      "loss": 0.9921,
+      "step": 8812
+    },
+    {
+      "epoch": 0.9268671040004207,
+      "grad_norm": 2.118350315003092,
+      "learning_rate": 6.702465364169103e-08,
+      "loss": 0.9487,
+      "step": 8813
+    },
+    {
+      "epoch": 0.926972274442268,
+      "grad_norm": 2.6969741905193363,
+      "learning_rate": 6.683277908173858e-08,
+      "loss": 0.9865,
+      "step": 8814
+    },
+    {
+      "epoch": 0.9270774448841154,
+      "grad_norm": 2.557886102693444,
+      "learning_rate": 6.664117583883272e-08,
+      "loss": 1.017,
+      "step": 8815
+    },
+    {
+      "epoch": 0.9271826153259627,
+      "grad_norm": 2.0970620367491635,
+      "learning_rate": 6.64498439343389e-08,
+      "loss": 1.0267,
+      "step": 8816
+    },
+    {
+      "epoch": 0.9272877857678099,
+      "grad_norm": 2.4865425974749003,
+      "learning_rate": 6.625878338959168e-08,
+      "loss": 1.0033,
+      "step": 8817
+    },
+    {
+      "epoch": 0.9273929562096572,
+      "grad_norm": 1.8276518666416652,
+      "learning_rate": 6.606799422589627e-08,
+      "loss": 0.9588,
+      "step": 8818
+    },
+    {
+      "epoch": 0.9274981266515046,
+      "grad_norm": 2.3070835376793304,
+      "learning_rate": 6.587747646452675e-08,
+      "loss": 0.966,
+      "step": 8819
+    },
+    {
+      "epoch": 0.9276032970933519,
+      "grad_norm": 2.5299309056388752,
+      "learning_rate": 6.568723012672779e-08,
+      "loss": 0.9866,
+      "step": 8820
+    },
+    {
+      "epoch": 0.9277084675351992,
+      "grad_norm": 2.547941361140502,
+      "learning_rate": 6.549725523371298e-08,
+      "loss": 0.992,
+      "step": 8821
+    },
+    {
+      "epoch": 0.9278136379770465,
+      "grad_norm": 2.417261025425601,
+      "learning_rate": 6.530755180666593e-08,
+      "loss": 0.981,
+      "step": 8822
+    },
+    {
+      "epoch": 0.9279188084188938,
+      "grad_norm": 2.9686003899940565,
+      "learning_rate": 6.511811986674028e-08,
+      "loss": 1.0337,
+      "step": 8823
+    },
+    {
+      "epoch": 0.9280239788607412,
+      "grad_norm": 2.8340124141461978,
+      "learning_rate": 6.49289594350594e-08,
+      "loss": 0.9824,
+      "step": 8824
+    },
+    {
+      "epoch": 0.9281291493025885,
+      "grad_norm": 3.171176281484354,
+      "learning_rate": 6.47400705327153e-08,
+      "loss": 0.9947,
+      "step": 8825
+    },
+    {
+      "epoch": 0.9282343197444358,
+      "grad_norm": 2.564414628087072,
+      "learning_rate": 6.455145318077144e-08,
+      "loss": 0.9985,
+      "step": 8826
+    },
+    {
+      "epoch": 0.9283394901862831,
+      "grad_norm": 2.4567715942774337,
+      "learning_rate": 6.436310740025986e-08,
+      "loss": 0.9892,
+      "step": 8827
+    },
+    {
+      "epoch": 0.9284446606281305,
+      "grad_norm": 2.4634151576295746,
+      "learning_rate": 6.41750332121821e-08,
+      "loss": 0.9948,
+      "step": 8828
+    },
+    {
+      "epoch": 0.9285498310699778,
+      "grad_norm": 2.2133318934099724,
+      "learning_rate": 6.398723063751083e-08,
+      "loss": 0.9897,
+      "step": 8829
+    },
+    {
+      "epoch": 0.9286550015118251,
+      "grad_norm": 2.215396747584313,
+      "learning_rate": 6.379969969718653e-08,
+      "loss": 0.9914,
+      "step": 8830
+    },
+    {
+      "epoch": 0.9287601719536724,
+      "grad_norm": 2.459138737636483,
+      "learning_rate": 6.361244041212078e-08,
+      "loss": 1.0084,
+      "step": 8831
+    },
+    {
+      "epoch": 0.9288653423955198,
+      "grad_norm": 1.934406044306015,
+      "learning_rate": 6.342545280319468e-08,
+      "loss": 0.9425,
+      "step": 8832
+    },
+    {
+      "epoch": 0.9289705128373671,
+      "grad_norm": 2.496708383635269,
+      "learning_rate": 6.323873689125848e-08,
+      "loss": 0.9742,
+      "step": 8833
+    },
+    {
+      "epoch": 0.9290756832792144,
+      "grad_norm": 2.078495436451324,
+      "learning_rate": 6.305229269713276e-08,
+      "loss": 0.9499,
+      "step": 8834
+    },
+    {
+      "epoch": 0.9291808537210617,
+      "grad_norm": 2.0553106122159948,
+      "learning_rate": 6.286612024160699e-08,
+      "loss": 0.984,
+      "step": 8835
+    },
+    {
+      "epoch": 0.9292860241629091,
+      "grad_norm": 1.9572815789421085,
+      "learning_rate": 6.268021954544095e-08,
+      "loss": 0.9922,
+      "step": 8836
+    },
+    {
+      "epoch": 0.9293911946047564,
+      "grad_norm": 2.0721426471435618,
+      "learning_rate": 6.249459062936447e-08,
+      "loss": 0.9806,
+      "step": 8837
+    },
+    {
+      "epoch": 0.9294963650466036,
+      "grad_norm": 2.9623950844860585,
+      "learning_rate": 6.230923351407653e-08,
+      "loss": 1.0178,
+      "step": 8838
+    },
+    {
+      "epoch": 0.9296015354884509,
+      "grad_norm": 2.736503439361389,
+      "learning_rate": 6.212414822024532e-08,
+      "loss": 1.0119,
+      "step": 8839
+    },
+    {
+      "epoch": 0.9297067059302982,
+      "grad_norm": 1.6867869389616887,
+      "learning_rate": 6.193933476850961e-08,
+      "loss": 0.9662,
+      "step": 8840
+    },
+    {
+      "epoch": 0.9298118763721456,
+      "grad_norm": 2.572027064275728,
+      "learning_rate": 6.175479317947818e-08,
+      "loss": 0.981,
+      "step": 8841
+    },
+    {
+      "epoch": 0.9299170468139929,
+      "grad_norm": 2.7961316524981563,
+      "learning_rate": 6.157052347372767e-08,
+      "loss": 0.978,
+      "step": 8842
+    },
+    {
+      "epoch": 0.9300222172558402,
+      "grad_norm": 2.0569779619598885,
+      "learning_rate": 6.138652567180658e-08,
+      "loss": 0.9811,
+      "step": 8843
+    },
+    {
+      "epoch": 0.9301273876976875,
+      "grad_norm": 2.1285919020070114,
+      "learning_rate": 6.120279979423133e-08,
+      "loss": 1.0071,
+      "step": 8844
+    },
+    {
+      "epoch": 0.9302325581395349,
+      "grad_norm": 2.6903412011111505,
+      "learning_rate": 6.101934586148938e-08,
+      "loss": 0.9554,
+      "step": 8845
+    },
+    {
+      "epoch": 0.9303377285813822,
+      "grad_norm": 2.8879413392379063,
+      "learning_rate": 6.083616389403691e-08,
+      "loss": 0.9768,
+      "step": 8846
+    },
+    {
+      "epoch": 0.9304428990232295,
+      "grad_norm": 2.7703622961279124,
+      "learning_rate": 6.065325391230032e-08,
+      "loss": 1.0119,
+      "step": 8847
+    },
+    {
+      "epoch": 0.9305480694650768,
+      "grad_norm": 2.2151304541476757,
+      "learning_rate": 6.047061593667552e-08,
+      "loss": 1.0256,
+      "step": 8848
+    },
+    {
+      "epoch": 0.9306532399069242,
+      "grad_norm": 2.3412586834401186,
+      "learning_rate": 6.028824998752764e-08,
+      "loss": 0.9941,
+      "step": 8849
+    },
+    {
+      "epoch": 0.9307584103487715,
+      "grad_norm": 3.0538317688803205,
+      "learning_rate": 6.010615608519261e-08,
+      "loss": 1.0315,
+      "step": 8850
+    },
+    {
+      "epoch": 0.9308635807906188,
+      "grad_norm": 2.2047583894385627,
+      "learning_rate": 5.992433424997473e-08,
+      "loss": 0.989,
+      "step": 8851
+    },
+    {
+      "epoch": 0.9309687512324661,
+      "grad_norm": 1.922856138628748,
+      "learning_rate": 5.974278450214893e-08,
+      "loss": 0.991,
+      "step": 8852
+    },
+    {
+      "epoch": 0.9310739216743135,
+      "grad_norm": 3.0208302634204536,
+      "learning_rate": 5.9561506861958994e-08,
+      "loss": 0.9735,
+      "step": 8853
+    },
+    {
+      "epoch": 0.9311790921161608,
+      "grad_norm": 2.8134158677180237,
+      "learning_rate": 5.9380501349619034e-08,
+      "loss": 0.9672,
+      "step": 8854
+    },
+    {
+      "epoch": 0.9312842625580081,
+      "grad_norm": 2.048098684977348,
+      "learning_rate": 5.9199767985312905e-08,
+      "loss": 0.993,
+      "step": 8855
+    },
+    {
+      "epoch": 0.9313894329998554,
+      "grad_norm": 2.345542378976877,
+      "learning_rate": 5.9019306789193374e-08,
+      "loss": 0.9825,
+      "step": 8856
+    },
+    {
+      "epoch": 0.9314946034417028,
+      "grad_norm": 2.3536588929040962,
+      "learning_rate": 5.883911778138324e-08,
+      "loss": 1.0004,
+      "step": 8857
+    },
+    {
+      "epoch": 0.93159977388355,
+      "grad_norm": 2.2434011250068533,
+      "learning_rate": 5.865920098197475e-08,
+      "loss": 1.0042,
+      "step": 8858
+    },
+    {
+      "epoch": 0.9317049443253973,
+      "grad_norm": 2.5379253604796577,
+      "learning_rate": 5.847955641103076e-08,
+      "loss": 0.9675,
+      "step": 8859
+    },
+    {
+      "epoch": 0.9318101147672446,
+      "grad_norm": 2.1102591640587627,
+      "learning_rate": 5.830018408858246e-08,
+      "loss": 0.9685,
+      "step": 8860
+    },
+    {
+      "epoch": 0.931915285209092,
+      "grad_norm": 2.0332852212785144,
+      "learning_rate": 5.8121084034631625e-08,
+      "loss": 0.9865,
+      "step": 8861
+    },
+    {
+      "epoch": 0.9320204556509393,
+      "grad_norm": 2.243828323356822,
+      "learning_rate": 5.7942256269148675e-08,
+      "loss": 0.9901,
+      "step": 8862
+    },
+    {
+      "epoch": 0.9321256260927866,
+      "grad_norm": 3.0058206320010403,
+      "learning_rate": 5.776370081207516e-08,
+      "loss": 1.0299,
+      "step": 8863
+    },
+    {
+      "epoch": 0.9322307965346339,
+      "grad_norm": 2.3047446223648373,
+      "learning_rate": 5.758541768332071e-08,
+      "loss": 0.9945,
+      "step": 8864
+    },
+    {
+      "epoch": 0.9323359669764812,
+      "grad_norm": 2.4665165924323453,
+      "learning_rate": 5.740740690276553e-08,
+      "loss": 0.994,
+      "step": 8865
+    },
+    {
+      "epoch": 0.9324411374183286,
+      "grad_norm": 2.5834709263911715,
+      "learning_rate": 5.722966849025957e-08,
+      "loss": 1.0037,
+      "step": 8866
+    },
+    {
+      "epoch": 0.9325463078601759,
+      "grad_norm": 2.2524845705290133,
+      "learning_rate": 5.7052202465621434e-08,
+      "loss": 1.002,
+      "step": 8867
+    },
+    {
+      "epoch": 0.9326514783020232,
+      "grad_norm": 2.4353939058546237,
+      "learning_rate": 5.687500884864e-08,
+      "loss": 0.983,
+      "step": 8868
+    },
+    {
+      "epoch": 0.9327566487438705,
+      "grad_norm": 2.5223813989202197,
+      "learning_rate": 5.66980876590742e-08,
+      "loss": 0.9392,
+      "step": 8869
+    },
+    {
+      "epoch": 0.9328618191857179,
+      "grad_norm": 2.2960675605121583,
+      "learning_rate": 5.652143891665157e-08,
+      "loss": 0.9687,
+      "step": 8870
+    },
+    {
+      "epoch": 0.9329669896275652,
+      "grad_norm": 2.10511952467411,
+      "learning_rate": 5.634506264107054e-08,
+      "loss": 0.9637,
+      "step": 8871
+    },
+    {
+      "epoch": 0.9330721600694125,
+      "grad_norm": 2.4587695553336513,
+      "learning_rate": 5.616895885199758e-08,
+      "loss": 0.9774,
+      "step": 8872
+    },
+    {
+      "epoch": 0.9331773305112598,
+      "grad_norm": 1.965729243927272,
+      "learning_rate": 5.5993127569070325e-08,
+      "loss": 0.9482,
+      "step": 8873
+    },
+    {
+      "epoch": 0.9332825009531072,
+      "grad_norm": 3.8358216369631597,
+      "learning_rate": 5.5817568811894763e-08,
+      "loss": 0.9987,
+      "step": 8874
+    },
+    {
+      "epoch": 0.9333876713949545,
+      "grad_norm": 1.5484337642536643,
+      "learning_rate": 5.564228260004773e-08,
+      "loss": 0.937,
+      "step": 8875
+    },
+    {
+      "epoch": 0.9334928418368018,
+      "grad_norm": 4.054983699134379,
+      "learning_rate": 5.5467268953074414e-08,
+      "loss": 1.0385,
+      "step": 8876
+    },
+    {
+      "epoch": 0.9335980122786491,
+      "grad_norm": 2.3147943515098715,
+      "learning_rate": 5.529252789049033e-08,
+      "loss": 1.0191,
+      "step": 8877
+    },
+    {
+      "epoch": 0.9337031827204963,
+      "grad_norm": 2.76261922574638,
+      "learning_rate": 5.5118059431781e-08,
+      "loss": 0.9745,
+      "step": 8878
+    },
+    {
+      "epoch": 0.9338083531623437,
+      "grad_norm": 3.2491525384795863,
+      "learning_rate": 5.4943863596400026e-08,
+      "loss": 0.9695,
+      "step": 8879
+    },
+    {
+      "epoch": 0.933913523604191,
+      "grad_norm": 2.4349584193925042,
+      "learning_rate": 5.476994040377243e-08,
+      "loss": 0.9973,
+      "step": 8880
+    },
+    {
+      "epoch": 0.9340186940460383,
+      "grad_norm": 1.57930103970821,
+      "learning_rate": 5.45962898732913e-08,
+      "loss": 0.9405,
+      "step": 8881
+    },
+    {
+      "epoch": 0.9341238644878856,
+      "grad_norm": 2.8444833537713805,
+      "learning_rate": 5.442291202432087e-08,
+      "loss": 1.0168,
+      "step": 8882
+    },
+    {
+      "epoch": 0.934229034929733,
+      "grad_norm": 2.0406400774301776,
+      "learning_rate": 5.424980687619319e-08,
+      "loss": 0.9987,
+      "step": 8883
+    },
+    {
+      "epoch": 0.9343342053715803,
+      "grad_norm": 1.6448616663744222,
+      "learning_rate": 5.407697444821169e-08,
+      "loss": 0.9668,
+      "step": 8884
+    },
+    {
+      "epoch": 0.9344393758134276,
+      "grad_norm": 2.320476937675156,
+      "learning_rate": 5.3904414759648195e-08,
+      "loss": 0.9719,
+      "step": 8885
+    },
+    {
+      "epoch": 0.9345445462552749,
+      "grad_norm": 2.5020571289750873,
+      "learning_rate": 5.3732127829743964e-08,
+      "loss": 0.9816,
+      "step": 8886
+    },
+    {
+      "epoch": 0.9346497166971223,
+      "grad_norm": 2.240359223042951,
+      "learning_rate": 5.356011367771113e-08,
+      "loss": 0.9784,
+      "step": 8887
+    },
+    {
+      "epoch": 0.9347548871389696,
+      "grad_norm": 1.9605700811751219,
+      "learning_rate": 5.338837232272992e-08,
+      "loss": 1.0099,
+      "step": 8888
+    },
+    {
+      "epoch": 0.9348600575808169,
+      "grad_norm": 2.0597267796298486,
+      "learning_rate": 5.321690378395167e-08,
+      "loss": 0.9846,
+      "step": 8889
+    },
+    {
+      "epoch": 0.9349652280226642,
+      "grad_norm": 1.8547486595770128,
+      "learning_rate": 5.304570808049553e-08,
+      "loss": 0.9384,
+      "step": 8890
+    },
+    {
+      "epoch": 0.9350703984645116,
+      "grad_norm": 2.197878394312316,
+      "learning_rate": 5.287478523145151e-08,
+      "loss": 0.9896,
+      "step": 8891
+    },
+    {
+      "epoch": 0.9351755689063589,
+      "grad_norm": 2.240523235271297,
+      "learning_rate": 5.2704135255879085e-08,
+      "loss": 0.9526,
+      "step": 8892
+    },
+    {
+      "epoch": 0.9352807393482062,
+      "grad_norm": 2.8082105695500172,
+      "learning_rate": 5.253375817280665e-08,
+      "loss": 0.9864,
+      "step": 8893
+    },
+    {
+      "epoch": 0.9353859097900535,
+      "grad_norm": 2.8371694669918472,
+      "learning_rate": 5.236365400123289e-08,
+      "loss": 0.9613,
+      "step": 8894
+    },
+    {
+      "epoch": 0.9354910802319009,
+      "grad_norm": 1.7047869479150661,
+      "learning_rate": 5.219382276012514e-08,
+      "loss": 0.9818,
+      "step": 8895
+    },
+    {
+      "epoch": 0.9355962506737482,
+      "grad_norm": 2.6335657856305885,
+      "learning_rate": 5.2024264468422125e-08,
+      "loss": 0.9802,
+      "step": 8896
+    },
+    {
+      "epoch": 0.9357014211155955,
+      "grad_norm": 2.1047402552338745,
+      "learning_rate": 5.185497914502957e-08,
+      "loss": 0.9918,
+      "step": 8897
+    },
+    {
+      "epoch": 0.9358065915574428,
+      "grad_norm": 2.964283660935744,
+      "learning_rate": 5.168596680882515e-08,
+      "loss": 1.0045,
+      "step": 8898
+    },
+    {
+      "epoch": 0.93591176199929,
+      "grad_norm": 2.164852522853516,
+      "learning_rate": 5.151722747865434e-08,
+      "loss": 0.9464,
+      "step": 8899
+    },
+    {
+      "epoch": 0.9360169324411374,
+      "grad_norm": 2.1258441873852463,
+      "learning_rate": 5.134876117333321e-08,
+      "loss": 0.9473,
+      "step": 8900
+    },
+    {
+      "epoch": 0.9361221028829847,
+      "grad_norm": 2.587963254415204,
+      "learning_rate": 5.1180567911646994e-08,
+      "loss": 1.0034,
+      "step": 8901
+    },
+    {
+      "epoch": 0.936227273324832,
+      "grad_norm": 2.6791686319504753,
+      "learning_rate": 5.1012647712350425e-08,
+      "loss": 0.9854,
+      "step": 8902
+    },
+    {
+      "epoch": 0.9363324437666793,
+      "grad_norm": 2.4951824060899597,
+      "learning_rate": 5.084500059416852e-08,
+      "loss": 0.9324,
+      "step": 8903
+    },
+    {
+      "epoch": 0.9364376142085267,
+      "grad_norm": 2.2357393856839405,
+      "learning_rate": 5.067762657579412e-08,
+      "loss": 0.9492,
+      "step": 8904
+    },
+    {
+      "epoch": 0.936542784650374,
+      "grad_norm": 1.9973404560425712,
+      "learning_rate": 5.0510525675891706e-08,
+      "loss": 0.9965,
+      "step": 8905
+    },
+    {
+      "epoch": 0.9366479550922213,
+      "grad_norm": 2.64657389622434,
+      "learning_rate": 5.0343697913093904e-08,
+      "loss": 0.982,
+      "step": 8906
+    },
+    {
+      "epoch": 0.9367531255340686,
+      "grad_norm": 1.9223409424201854,
+      "learning_rate": 5.017714330600332e-08,
+      "loss": 1.0052,
+      "step": 8907
+    },
+    {
+      "epoch": 0.936858295975916,
+      "grad_norm": 2.203306276144651,
+      "learning_rate": 5.0010861873192596e-08,
+      "loss": 1.022,
+      "step": 8908
+    },
+    {
+      "epoch": 0.9369634664177633,
+      "grad_norm": 2.4829606009339282,
+      "learning_rate": 4.984485363320218e-08,
+      "loss": 0.9611,
+      "step": 8909
+    },
+    {
+      "epoch": 0.9370686368596106,
+      "grad_norm": 1.9321781349119374,
+      "learning_rate": 4.9679118604544496e-08,
+      "loss": 0.9633,
+      "step": 8910
+    },
+    {
+      "epoch": 0.9371738073014579,
+      "grad_norm": 2.2454619717290285,
+      "learning_rate": 4.951365680569975e-08,
+      "loss": 0.9923,
+      "step": 8911
+    },
+    {
+      "epoch": 0.9372789777433053,
+      "grad_norm": 2.61658321641514,
+      "learning_rate": 4.9348468255118465e-08,
+      "loss": 1.0092,
+      "step": 8912
+    },
+    {
+      "epoch": 0.9373841481851526,
+      "grad_norm": 3.2178127868301494,
+      "learning_rate": 4.918355297122035e-08,
+      "loss": 1.0305,
+      "step": 8913
+    },
+    {
+      "epoch": 0.9374893186269999,
+      "grad_norm": 2.5158970965320178,
+      "learning_rate": 4.901891097239431e-08,
+      "loss": 0.932,
+      "step": 8914
+    },
+    {
+      "epoch": 0.9375944890688472,
+      "grad_norm": 2.853474567380966,
+      "learning_rate": 4.885454227700009e-08,
+      "loss": 0.9817,
+      "step": 8915
+    },
+    {
+      "epoch": 0.9376996595106946,
+      "grad_norm": 2.8015618857324105,
+      "learning_rate": 4.8690446903365e-08,
+      "loss": 0.9716,
+      "step": 8916
+    },
+    {
+      "epoch": 0.9378048299525419,
+      "grad_norm": 2.801611751147078,
+      "learning_rate": 4.8526624869787985e-08,
+      "loss": 0.9855,
+      "step": 8917
+    },
+    {
+      "epoch": 0.9379100003943892,
+      "grad_norm": 2.1624199744457426,
+      "learning_rate": 4.836307619453556e-08,
+      "loss": 0.9891,
+      "step": 8918
+    },
+    {
+      "epoch": 0.9380151708362364,
+      "grad_norm": 2.9139791008885907,
+      "learning_rate": 4.819980089584564e-08,
+      "loss": 0.9789,
+      "step": 8919
+    },
+    {
+      "epoch": 0.9381203412780837,
+      "grad_norm": 2.923200016146771,
+      "learning_rate": 4.8036798991923925e-08,
+      "loss": 0.9436,
+      "step": 8920
+    },
+    {
+      "epoch": 0.9382255117199311,
+      "grad_norm": 2.295229636965682,
+      "learning_rate": 4.7874070500946725e-08,
+      "loss": 0.975,
+      "step": 8921
+    },
+    {
+      "epoch": 0.9383306821617784,
+      "grad_norm": 3.0300460033187515,
+      "learning_rate": 4.771161544105951e-08,
+      "loss": 0.9951,
+      "step": 8922
+    },
+    {
+      "epoch": 0.9384358526036257,
+      "grad_norm": 2.115432155833796,
+      "learning_rate": 4.754943383037669e-08,
+      "loss": 0.9423,
+      "step": 8923
+    },
+    {
+      "epoch": 0.938541023045473,
+      "grad_norm": 2.2142585640978525,
+      "learning_rate": 4.7387525686983793e-08,
+      "loss": 1.0046,
+      "step": 8924
+    },
+    {
+      "epoch": 0.9386461934873204,
+      "grad_norm": 2.2974730077916714,
+      "learning_rate": 4.722589102893416e-08,
+      "loss": 0.9384,
+      "step": 8925
+    },
+    {
+      "epoch": 0.9387513639291677,
+      "grad_norm": 2.4189018597780216,
+      "learning_rate": 4.70645298742517e-08,
+      "loss": 0.9834,
+      "step": 8926
+    },
+    {
+      "epoch": 0.938856534371015,
+      "grad_norm": 2.2427060135015315,
+      "learning_rate": 4.690344224092924e-08,
+      "loss": 0.9165,
+      "step": 8927
+    },
+    {
+      "epoch": 0.9389617048128623,
+      "grad_norm": 3.2157140769814667,
+      "learning_rate": 4.674262814692909e-08,
+      "loss": 0.9877,
+      "step": 8928
+    },
+    {
+      "epoch": 0.9390668752547097,
+      "grad_norm": 2.2854910606933556,
+      "learning_rate": 4.658208761018357e-08,
+      "loss": 0.9979,
+      "step": 8929
+    },
+    {
+      "epoch": 0.939172045696557,
+      "grad_norm": 2.3119573645611307,
+      "learning_rate": 4.6421820648593906e-08,
+      "loss": 0.9796,
+      "step": 8930
+    },
+    {
+      "epoch": 0.9392772161384043,
+      "grad_norm": 2.2403463122387914,
+      "learning_rate": 4.626182728003165e-08,
+      "loss": 0.9831,
+      "step": 8931
+    },
+    {
+      "epoch": 0.9393823865802516,
+      "grad_norm": 1.9681555317969763,
+      "learning_rate": 4.61021075223364e-08,
+      "loss": 0.9454,
+      "step": 8932
+    },
+    {
+      "epoch": 0.939487557022099,
+      "grad_norm": 2.0758545826298844,
+      "learning_rate": 4.594266139331921e-08,
+      "loss": 0.9768,
+      "step": 8933
+    },
+    {
+      "epoch": 0.9395927274639463,
+      "grad_norm": 1.836114474356616,
+      "learning_rate": 4.57834889107589e-08,
+      "loss": 1.0333,
+      "step": 8934
+    },
+    {
+      "epoch": 0.9396978979057936,
+      "grad_norm": 2.439188983743261,
+      "learning_rate": 4.5624590092404884e-08,
+      "loss": 0.9691,
+      "step": 8935
+    },
+    {
+      "epoch": 0.9398030683476409,
+      "grad_norm": 2.093740532951504,
+      "learning_rate": 4.546596495597494e-08,
+      "loss": 0.9777,
+      "step": 8936
+    },
+    {
+      "epoch": 0.9399082387894883,
+      "grad_norm": 2.877703863906485,
+      "learning_rate": 4.530761351915741e-08,
+      "loss": 0.9488,
+      "step": 8937
+    },
+    {
+      "epoch": 0.9400134092313356,
+      "grad_norm": 2.757214678581196,
+      "learning_rate": 4.5149535799610125e-08,
+      "loss": 1.0115,
+      "step": 8938
+    },
+    {
+      "epoch": 0.9401185796731828,
+      "grad_norm": 3.325004102364057,
+      "learning_rate": 4.499173181495897e-08,
+      "loss": 0.9751,
+      "step": 8939
+    },
+    {
+      "epoch": 0.9402237501150301,
+      "grad_norm": 2.3523829345278915,
+      "learning_rate": 4.4834201582801275e-08,
+      "loss": 0.9737,
+      "step": 8940
+    },
+    {
+      "epoch": 0.9403289205568774,
+      "grad_norm": 3.8621952023589445,
+      "learning_rate": 4.467694512070242e-08,
+      "loss": 1.0262,
+      "step": 8941
+    },
+    {
+      "epoch": 0.9404340909987248,
+      "grad_norm": 2.185909471977706,
+      "learning_rate": 4.451996244619755e-08,
+      "loss": 1.009,
+      "step": 8942
+    },
+    {
+      "epoch": 0.9405392614405721,
+      "grad_norm": 1.7242270309634,
+      "learning_rate": 4.436325357679211e-08,
+      "loss": 0.9852,
+      "step": 8943
+    },
+    {
+      "epoch": 0.9406444318824194,
+      "grad_norm": 2.12009999134058,
+      "learning_rate": 4.420681852995962e-08,
+      "loss": 0.9544,
+      "step": 8944
+    },
+    {
+      "epoch": 0.9407496023242667,
+      "grad_norm": 2.7180402269060737,
+      "learning_rate": 4.4050657323144454e-08,
+      "loss": 0.9866,
+      "step": 8945
+    },
+    {
+      "epoch": 0.9408547727661141,
+      "grad_norm": 2.3821757021611227,
+      "learning_rate": 4.3894769973759075e-08,
+      "loss": 0.9786,
+      "step": 8946
+    },
+    {
+      "epoch": 0.9409599432079614,
+      "grad_norm": 2.4406138087963023,
+      "learning_rate": 4.3739156499186806e-08,
+      "loss": 1.0249,
+      "step": 8947
+    },
+    {
+      "epoch": 0.9410651136498087,
+      "grad_norm": 1.8333812495807653,
+      "learning_rate": 4.358381691677932e-08,
+      "loss": 0.9849,
+      "step": 8948
+    },
+    {
+      "epoch": 0.941170284091656,
+      "grad_norm": 1.973400633068133,
+      "learning_rate": 4.342875124385859e-08,
+      "loss": 0.9627,
+      "step": 8949
+    },
+    {
+      "epoch": 0.9412754545335034,
+      "grad_norm": 2.2827790185249075,
+      "learning_rate": 4.3273959497715234e-08,
+      "loss": 0.9529,
+      "step": 8950
+    },
+    {
+      "epoch": 0.9413806249753507,
+      "grad_norm": 1.789449664350765,
+      "learning_rate": 4.311944169560989e-08,
+      "loss": 0.9506,
+      "step": 8951
+    },
+    {
+      "epoch": 0.941485795417198,
+      "grad_norm": 2.0245454190254955,
+      "learning_rate": 4.296519785477293e-08,
+      "loss": 0.9788,
+      "step": 8952
+    },
+    {
+      "epoch": 0.9415909658590453,
+      "grad_norm": 2.983790376093045,
+      "learning_rate": 4.2811227992402834e-08,
+      "loss": 0.9872,
+      "step": 8953
+    },
+    {
+      "epoch": 0.9416961363008927,
+      "grad_norm": 2.448225720897263,
+      "learning_rate": 4.2657532125669196e-08,
+      "loss": 0.9493,
+      "step": 8954
+    },
+    {
+      "epoch": 0.94180130674274,
+      "grad_norm": 3.0859410244770693,
+      "learning_rate": 4.25041102717097e-08,
+      "loss": 0.9438,
+      "step": 8955
+    },
+    {
+      "epoch": 0.9419064771845873,
+      "grad_norm": 2.3832499552025466,
+      "learning_rate": 4.2350962447632594e-08,
+      "loss": 0.9902,
+      "step": 8956
+    },
+    {
+      "epoch": 0.9420116476264346,
+      "grad_norm": 2.251093370418025,
+      "learning_rate": 4.219808867051506e-08,
+      "loss": 0.9439,
+      "step": 8957
+    },
+    {
+      "epoch": 0.942116818068282,
+      "grad_norm": 3.0337285007524,
+      "learning_rate": 4.204548895740346e-08,
+      "loss": 0.9715,
+      "step": 8958
+    },
+    {
+      "epoch": 0.9422219885101293,
+      "grad_norm": 2.067055271479491,
+      "learning_rate": 4.18931633253139e-08,
+      "loss": 0.9864,
+      "step": 8959
+    },
+    {
+      "epoch": 0.9423271589519765,
+      "grad_norm": 2.198046384198305,
+      "learning_rate": 4.174111179123141e-08,
+      "loss": 0.994,
+      "step": 8960
+    },
+    {
+      "epoch": 0.9424323293938238,
+      "grad_norm": 2.2696199460704225,
+      "learning_rate": 4.158933437211188e-08,
+      "loss": 0.9611,
+      "step": 8961
+    },
+    {
+      "epoch": 0.9425374998356711,
+      "grad_norm": 2.2636911949582124,
+      "learning_rate": 4.1437831084878974e-08,
+      "loss": 0.9585,
+      "step": 8962
+    },
+    {
+      "epoch": 0.9426426702775185,
+      "grad_norm": 2.490147217829843,
+      "learning_rate": 4.12866019464267e-08,
+      "loss": 0.9212,
+      "step": 8963
+    },
+    {
+      "epoch": 0.9427478407193658,
+      "grad_norm": 2.279237319888199,
+      "learning_rate": 4.1135646973618214e-08,
+      "loss": 0.9686,
+      "step": 8964
+    },
+    {
+      "epoch": 0.9428530111612131,
+      "grad_norm": 2.9443355297300187,
+      "learning_rate": 4.098496618328618e-08,
+      "loss": 1.008,
+      "step": 8965
+    },
+    {
+      "epoch": 0.9429581816030604,
+      "grad_norm": 2.4101695450847074,
+      "learning_rate": 4.083455959223298e-08,
+      "loss": 0.9955,
+      "step": 8966
+    },
+    {
+      "epoch": 0.9430633520449078,
+      "grad_norm": 3.2074191412814916,
+      "learning_rate": 4.068442721722965e-08,
+      "loss": 1.0099,
+      "step": 8967
+    },
+    {
+      "epoch": 0.9431685224867551,
+      "grad_norm": 3.174497962789582,
+      "learning_rate": 4.0534569075017516e-08,
+      "loss": 0.9797,
+      "step": 8968
+    },
+    {
+      "epoch": 0.9432736929286024,
+      "grad_norm": 2.5583454557550014,
+      "learning_rate": 4.038498518230627e-08,
+      "loss": 1.006,
+      "step": 8969
+    },
+    {
+      "epoch": 0.9433788633704497,
+      "grad_norm": 1.8038618775066488,
+      "learning_rate": 4.0235675555776734e-08,
+      "loss": 0.9574,
+      "step": 8970
+    },
+    {
+      "epoch": 0.9434840338122971,
+      "grad_norm": 2.514174840873185,
+      "learning_rate": 4.008664021207698e-08,
+      "loss": 0.993,
+      "step": 8971
+    },
+    {
+      "epoch": 0.9435892042541444,
+      "grad_norm": 2.1814124733410387,
+      "learning_rate": 3.993787916782649e-08,
+      "loss": 0.9489,
+      "step": 8972
+    },
+    {
+      "epoch": 0.9436943746959917,
+      "grad_norm": 2.5226893485355952,
+      "learning_rate": 3.978939243961283e-08,
+      "loss": 1.0078,
+      "step": 8973
+    },
+    {
+      "epoch": 0.943799545137839,
+      "grad_norm": 2.256815087313752,
+      "learning_rate": 3.964118004399331e-08,
+      "loss": 0.9775,
+      "step": 8974
+    },
+    {
+      "epoch": 0.9439047155796864,
+      "grad_norm": 2.8223010195407947,
+      "learning_rate": 3.949324199749527e-08,
+      "loss": 1.0249,
+      "step": 8975
+    },
+    {
+      "epoch": 0.9440098860215337,
+      "grad_norm": 1.391951568158817,
+      "learning_rate": 3.9345578316614396e-08,
+      "loss": 0.9615,
+      "step": 8976
+    },
+    {
+      "epoch": 0.944115056463381,
+      "grad_norm": 2.2474764496627717,
+      "learning_rate": 3.919818901781669e-08,
+      "loss": 0.9517,
+      "step": 8977
+    },
+    {
+      "epoch": 0.9442202269052283,
+      "grad_norm": 2.472061001047028,
+      "learning_rate": 3.905107411753678e-08,
+      "loss": 0.9883,
+      "step": 8978
+    },
+    {
+      "epoch": 0.9443253973470757,
+      "grad_norm": 1.9454273863676264,
+      "learning_rate": 3.89042336321796e-08,
+      "loss": 0.9953,
+      "step": 8979
+    },
+    {
+      "epoch": 0.9444305677889229,
+      "grad_norm": 2.5983891669051045,
+      "learning_rate": 3.8757667578119e-08,
+      "loss": 0.9775,
+      "step": 8980
+    },
+    {
+      "epoch": 0.9445357382307702,
+      "grad_norm": 2.052082559249859,
+      "learning_rate": 3.8611375971698004e-08,
+      "loss": 1.0009,
+      "step": 8981
+    },
+    {
+      "epoch": 0.9446409086726175,
+      "grad_norm": 2.2703611102031402,
+      "learning_rate": 3.8465358829229415e-08,
+      "loss": 0.9423,
+      "step": 8982
+    },
+    {
+      "epoch": 0.9447460791144648,
+      "grad_norm": 2.2134942047175814,
+      "learning_rate": 3.831961616699464e-08,
+      "loss": 0.9994,
+      "step": 8983
+    },
+    {
+      "epoch": 0.9448512495563122,
+      "grad_norm": 2.9036139692399834,
+      "learning_rate": 3.8174148001246246e-08,
+      "loss": 0.9657,
+      "step": 8984
+    },
+    {
+      "epoch": 0.9449564199981595,
+      "grad_norm": 2.1246606320654897,
+      "learning_rate": 3.802895434820431e-08,
+      "loss": 0.9848,
+      "step": 8985
+    },
+    {
+      "epoch": 0.9450615904400068,
+      "grad_norm": 2.4282287550373476,
+      "learning_rate": 3.78840352240592e-08,
+      "loss": 0.9852,
+      "step": 8986
+    },
+    {
+      "epoch": 0.9451667608818541,
+      "grad_norm": 3.214310427554045,
+      "learning_rate": 3.773939064497051e-08,
+      "loss": 0.9898,
+      "step": 8987
+    },
+    {
+      "epoch": 0.9452719313237015,
+      "grad_norm": 2.344091022554252,
+      "learning_rate": 3.759502062706727e-08,
+      "loss": 0.9549,
+      "step": 8988
+    },
+    {
+      "epoch": 0.9453771017655488,
+      "grad_norm": 2.4153973988675057,
+      "learning_rate": 3.745092518644827e-08,
+      "loss": 0.9341,
+      "step": 8989
+    },
+    {
+      "epoch": 0.9454822722073961,
+      "grad_norm": 2.203918057550135,
+      "learning_rate": 3.730710433918039e-08,
+      "loss": 1.0003,
+      "step": 8990
+    },
+    {
+      "epoch": 0.9455874426492434,
+      "grad_norm": 2.265252052511168,
+      "learning_rate": 3.716355810130135e-08,
+      "loss": 1.002,
+      "step": 8991
+    },
+    {
+      "epoch": 0.9456926130910908,
+      "grad_norm": 3.0195268989146338,
+      "learning_rate": 3.70202864888175e-08,
+      "loss": 0.9745,
+      "step": 8992
+    },
+    {
+      "epoch": 0.9457977835329381,
+      "grad_norm": 2.1813144759793874,
+      "learning_rate": 3.687728951770497e-08,
+      "loss": 0.9996,
+      "step": 8993
+    },
+    {
+      "epoch": 0.9459029539747854,
+      "grad_norm": 2.382527317064494,
+      "learning_rate": 3.673456720390878e-08,
+      "loss": 1.0224,
+      "step": 8994
+    },
+    {
+      "epoch": 0.9460081244166327,
+      "grad_norm": 2.888448496222066,
+      "learning_rate": 3.659211956334369e-08,
+      "loss": 0.9979,
+      "step": 8995
+    },
+    {
+      "epoch": 0.9461132948584801,
+      "grad_norm": 2.428825105062496,
+      "learning_rate": 3.644994661189366e-08,
+      "loss": 0.9911,
+      "step": 8996
+    },
+    {
+      "epoch": 0.9462184653003274,
+      "grad_norm": 2.947866269473164,
+      "learning_rate": 3.630804836541213e-08,
+      "loss": 0.9448,
+      "step": 8997
+    },
+    {
+      "epoch": 0.9463236357421747,
+      "grad_norm": 2.5597218827599693,
+      "learning_rate": 3.616642483972199e-08,
+      "loss": 0.9859,
+      "step": 8998
+    },
+    {
+      "epoch": 0.946428806184022,
+      "grad_norm": 2.158200307069964,
+      "learning_rate": 3.602507605061478e-08,
+      "loss": 0.9825,
+      "step": 8999
+    },
+    {
+      "epoch": 0.9465339766258692,
+      "grad_norm": 3.1515542288121883,
+      "learning_rate": 3.588400201385289e-08,
+      "loss": 1.016,
+      "step": 9000
+    },
+    {
+      "epoch": 0.9466391470677166,
+      "grad_norm": 2.5215905656201505,
+      "learning_rate": 3.574320274516652e-08,
+      "loss": 0.9547,
+      "step": 9001
+    },
+    {
+      "epoch": 0.9467443175095639,
+      "grad_norm": 2.2845409836197135,
+      "learning_rate": 3.560267826025588e-08,
+      "loss": 0.969,
+      "step": 9002
+    },
+    {
+      "epoch": 0.9468494879514112,
+      "grad_norm": 2.7215628504061633,
+      "learning_rate": 3.546242857479093e-08,
+      "loss": 0.9303,
+      "step": 9003
+    },
+    {
+      "epoch": 0.9469546583932585,
+      "grad_norm": 2.9588772470370475,
+      "learning_rate": 3.5322453704410286e-08,
+      "loss": 1.0371,
+      "step": 9004
+    },
+    {
+      "epoch": 0.9470598288351059,
+      "grad_norm": 2.2593667415441954,
+      "learning_rate": 3.518275366472229e-08,
+      "loss": 0.9698,
+      "step": 9005
+    },
+    {
+      "epoch": 0.9471649992769532,
+      "grad_norm": 2.042081877507642,
+      "learning_rate": 3.504332847130476e-08,
+      "loss": 1.0017,
+      "step": 9006
+    },
+    {
+      "epoch": 0.9472701697188005,
+      "grad_norm": 2.587961130036968,
+      "learning_rate": 3.49041781397047e-08,
+      "loss": 0.9826,
+      "step": 9007
+    },
+    {
+      "epoch": 0.9473753401606478,
+      "grad_norm": 3.102076174146032,
+      "learning_rate": 3.4765302685438315e-08,
+      "loss": 0.9855,
+      "step": 9008
+    },
+    {
+      "epoch": 0.9474805106024952,
+      "grad_norm": 2.544139468740267,
+      "learning_rate": 3.462670212399099e-08,
+      "loss": 0.9721,
+      "step": 9009
+    },
+    {
+      "epoch": 0.9475856810443425,
+      "grad_norm": 2.7727335358217053,
+      "learning_rate": 3.4488376470818153e-08,
+      "loss": 0.9728,
+      "step": 9010
+    },
+    {
+      "epoch": 0.9476908514861898,
+      "grad_norm": 2.5294138314705696,
+      "learning_rate": 3.4350325741344114e-08,
+      "loss": 0.9813,
+      "step": 9011
+    },
+    {
+      "epoch": 0.9477960219280371,
+      "grad_norm": 2.7089584804818907,
+      "learning_rate": 3.421254995096268e-08,
+      "loss": 1.0279,
+      "step": 9012
+    },
+    {
+      "epoch": 0.9479011923698845,
+      "grad_norm": 2.281684751376811,
+      "learning_rate": 3.407504911503684e-08,
+      "loss": 0.9687,
+      "step": 9013
+    },
+    {
+      "epoch": 0.9480063628117318,
+      "grad_norm": 2.138661404129903,
+      "learning_rate": 3.3937823248899046e-08,
+      "loss": 0.9265,
+      "step": 9014
+    },
+    {
+      "epoch": 0.9481115332535791,
+      "grad_norm": 3.39614476261138,
+      "learning_rate": 3.3800872367850956e-08,
+      "loss": 0.9766,
+      "step": 9015
+    },
+    {
+      "epoch": 0.9482167036954264,
+      "grad_norm": 2.2640896136810955,
+      "learning_rate": 3.366419648716368e-08,
+      "loss": 0.967,
+      "step": 9016
+    },
+    {
+      "epoch": 0.9483218741372738,
+      "grad_norm": 2.047442890013021,
+      "learning_rate": 3.352779562207753e-08,
+      "loss": 0.9985,
+      "step": 9017
+    },
+    {
+      "epoch": 0.9484270445791211,
+      "grad_norm": 1.9785722794844491,
+      "learning_rate": 3.339166978780256e-08,
+      "loss": 0.9326,
+      "step": 9018
+    },
+    {
+      "epoch": 0.9485322150209684,
+      "grad_norm": 2.1691668115058733,
+      "learning_rate": 3.3255818999517465e-08,
+      "loss": 1.017,
+      "step": 9019
+    },
+    {
+      "epoch": 0.9486373854628157,
+      "grad_norm": 2.4799530161132863,
+      "learning_rate": 3.3120243272371236e-08,
+      "loss": 0.9924,
+      "step": 9020
+    },
+    {
+      "epoch": 0.9487425559046629,
+      "grad_norm": 2.353650653525652,
+      "learning_rate": 3.298494262148122e-08,
+      "loss": 1.0099,
+      "step": 9021
+    },
+    {
+      "epoch": 0.9488477263465103,
+      "grad_norm": 2.2157690574353905,
+      "learning_rate": 3.2849917061934245e-08,
+      "loss": 1.0099,
+      "step": 9022
+    },
+    {
+      "epoch": 0.9489528967883576,
+      "grad_norm": 2.6170941163454726,
+      "learning_rate": 3.2715166608787426e-08,
+      "loss": 1.0147,
+      "step": 9023
+    },
+    {
+      "epoch": 0.9490580672302049,
+      "grad_norm": 2.252136933422275,
+      "learning_rate": 3.2580691277065704e-08,
+      "loss": 0.9959,
+      "step": 9024
+    },
+    {
+      "epoch": 0.9491632376720522,
+      "grad_norm": 2.0628624156474906,
+      "learning_rate": 3.2446491081764566e-08,
+      "loss": 0.9494,
+      "step": 9025
+    },
+    {
+      "epoch": 0.9492684081138996,
+      "grad_norm": 2.8179669861299055,
+      "learning_rate": 3.2312566037848437e-08,
+      "loss": 0.9782,
+      "step": 9026
+    },
+    {
+      "epoch": 0.9493735785557469,
+      "grad_norm": 2.4587844515129365,
+      "learning_rate": 3.217891616025065e-08,
+      "loss": 0.9678,
+      "step": 9027
+    },
+    {
+      "epoch": 0.9494787489975942,
+      "grad_norm": 2.836843623950099,
+      "learning_rate": 3.204554146387456e-08,
+      "loss": 1.0119,
+      "step": 9028
+    },
+    {
+      "epoch": 0.9495839194394415,
+      "grad_norm": 2.6786242065548675,
+      "learning_rate": 3.191244196359244e-08,
+      "loss": 0.9626,
+      "step": 9029
+    },
+    {
+      "epoch": 0.9496890898812889,
+      "grad_norm": 1.9157598537055311,
+      "learning_rate": 3.1779617674245754e-08,
+      "loss": 0.9927,
+      "step": 9030
+    },
+    {
+      "epoch": 0.9497942603231362,
+      "grad_norm": 2.8959657091729114,
+      "learning_rate": 3.1647068610645706e-08,
+      "loss": 1.0009,
+      "step": 9031
+    },
+    {
+      "epoch": 0.9498994307649835,
+      "grad_norm": 2.5929815782861603,
+      "learning_rate": 3.151479478757186e-08,
+      "loss": 1.0032,
+      "step": 9032
+    },
+    {
+      "epoch": 0.9500046012068308,
+      "grad_norm": 2.5834124688388633,
+      "learning_rate": 3.1382796219774634e-08,
+      "loss": 1.0217,
+      "step": 9033
+    },
+    {
+      "epoch": 0.9501097716486782,
+      "grad_norm": 3.2272054253254363,
+      "learning_rate": 3.125107292197227e-08,
+      "loss": 1.0221,
+      "step": 9034
+    },
+    {
+      "epoch": 0.9502149420905255,
+      "grad_norm": 2.6054764436911895,
+      "learning_rate": 3.1119624908853286e-08,
+      "loss": 0.9901,
+      "step": 9035
+    },
+    {
+      "epoch": 0.9503201125323728,
+      "grad_norm": 3.169144160863623,
+      "learning_rate": 3.0988452195075127e-08,
+      "loss": 0.9878,
+      "step": 9036
+    },
+    {
+      "epoch": 0.9504252829742201,
+      "grad_norm": 2.6138918264045765,
+      "learning_rate": 3.085755479526442e-08,
+      "loss": 1.035,
+      "step": 9037
+    },
+    {
+      "epoch": 0.9505304534160675,
+      "grad_norm": 2.2611828427142058,
+      "learning_rate": 3.072693272401756e-08,
+      "loss": 0.9684,
+      "step": 9038
+    },
+    {
+      "epoch": 0.9506356238579148,
+      "grad_norm": 2.9501126652696676,
+      "learning_rate": 3.059658599589926e-08,
+      "loss": 0.9749,
+      "step": 9039
+    },
+    {
+      "epoch": 0.9507407942997621,
+      "grad_norm": 2.1468682912165087,
+      "learning_rate": 3.046651462544487e-08,
+      "loss": 0.9761,
+      "step": 9040
+    },
+    {
+      "epoch": 0.9508459647416093,
+      "grad_norm": 2.4497313864823753,
+      "learning_rate": 3.0336718627158035e-08,
+      "loss": 0.9701,
+      "step": 9041
+    },
+    {
+      "epoch": 0.9509511351834566,
+      "grad_norm": 2.153394730671303,
+      "learning_rate": 3.0207198015512195e-08,
+      "loss": 0.9441,
+      "step": 9042
+    },
+    {
+      "epoch": 0.951056305625304,
+      "grad_norm": 3.031401492377512,
+      "learning_rate": 3.007795280494996e-08,
+      "loss": 1.0247,
+      "step": 9043
+    },
+    {
+      "epoch": 0.9511614760671513,
+      "grad_norm": 2.5059187840965604,
+      "learning_rate": 2.994898300988258e-08,
+      "loss": 0.9365,
+      "step": 9044
+    },
+    {
+      "epoch": 0.9512666465089986,
+      "grad_norm": 2.4460236039465615,
+      "learning_rate": 2.9820288644692166e-08,
+      "loss": 0.9831,
+      "step": 9045
+    },
+    {
+      "epoch": 0.9513718169508459,
+      "grad_norm": 2.608884100112733,
+      "learning_rate": 2.969186972372806e-08,
+      "loss": 1.0081,
+      "step": 9046
+    },
+    {
+      "epoch": 0.9514769873926933,
+      "grad_norm": 2.7708175506380286,
+      "learning_rate": 2.9563726261310767e-08,
+      "loss": 0.9921,
+      "step": 9047
+    },
+    {
+      "epoch": 0.9515821578345406,
+      "grad_norm": 2.9910305301759186,
+      "learning_rate": 2.9435858271728845e-08,
+      "loss": 0.9821,
+      "step": 9048
+    },
+    {
+      "epoch": 0.9516873282763879,
+      "grad_norm": 2.9234495857772553,
+      "learning_rate": 2.9308265769240894e-08,
+      "loss": 1.0157,
+      "step": 9049
+    },
+    {
+      "epoch": 0.9517924987182352,
+      "grad_norm": 1.8321103596533066,
+      "learning_rate": 2.9180948768074424e-08,
+      "loss": 0.9784,
+      "step": 9050
+    },
+    {
+      "epoch": 0.9518976691600826,
+      "grad_norm": 2.357717738638271,
+      "learning_rate": 2.905390728242585e-08,
+      "loss": 0.9425,
+      "step": 9051
+    },
+    {
+      "epoch": 0.9520028396019299,
+      "grad_norm": 3.189128618030638,
+      "learning_rate": 2.8927141326461903e-08,
+      "loss": 1.0047,
+      "step": 9052
+    },
+    {
+      "epoch": 0.9521080100437772,
+      "grad_norm": 2.256843607403256,
+      "learning_rate": 2.8800650914317385e-08,
+      "loss": 0.946,
+      "step": 9053
+    },
+    {
+      "epoch": 0.9522131804856245,
+      "grad_norm": 3.062457453580989,
+      "learning_rate": 2.867443606009768e-08,
+      "loss": 1.0034,
+      "step": 9054
+    },
+    {
+      "epoch": 0.9523183509274719,
+      "grad_norm": 3.165772898246199,
+      "learning_rate": 2.854849677787569e-08,
+      "loss": 1.0181,
+      "step": 9055
+    },
+    {
+      "epoch": 0.9524235213693192,
+      "grad_norm": 2.173606329897896,
+      "learning_rate": 2.8422833081695466e-08,
+      "loss": 1.0107,
+      "step": 9056
+    },
+    {
+      "epoch": 0.9525286918111665,
+      "grad_norm": 2.1255256334631976,
+      "learning_rate": 2.8297444985569412e-08,
+      "loss": 1.0023,
+      "step": 9057
+    },
+    {
+      "epoch": 0.9526338622530138,
+      "grad_norm": 2.077684783487426,
+      "learning_rate": 2.8172332503479116e-08,
+      "loss": 0.9621,
+      "step": 9058
+    },
+    {
+      "epoch": 0.9527390326948612,
+      "grad_norm": 2.1934129457768834,
+      "learning_rate": 2.8047495649375366e-08,
+      "loss": 0.9669,
+      "step": 9059
+    },
+    {
+      "epoch": 0.9528442031367085,
+      "grad_norm": 2.8101529667642753,
+      "learning_rate": 2.7922934437178695e-08,
+      "loss": 0.9975,
+      "step": 9060
+    },
+    {
+      "epoch": 0.9529493735785557,
+      "grad_norm": 1.9383323625550897,
+      "learning_rate": 2.7798648880778545e-08,
+      "loss": 0.9921,
+      "step": 9061
+    },
+    {
+      "epoch": 0.953054544020403,
+      "grad_norm": 1.8036336585586927,
+      "learning_rate": 2.767463899403383e-08,
+      "loss": 0.9686,
+      "step": 9062
+    },
+    {
+      "epoch": 0.9531597144622503,
+      "grad_norm": 2.8407059895479696,
+      "learning_rate": 2.755090479077266e-08,
+      "loss": 0.9795,
+      "step": 9063
+    },
+    {
+      "epoch": 0.9532648849040977,
+      "grad_norm": 2.758918511717141,
+      "learning_rate": 2.7427446284792324e-08,
+      "loss": 0.9974,
+      "step": 9064
+    },
+    {
+      "epoch": 0.953370055345945,
+      "grad_norm": 3.586171387728783,
+      "learning_rate": 2.730426348985904e-08,
+      "loss": 0.9778,
+      "step": 9065
+    },
+    {
+      "epoch": 0.9534752257877923,
+      "grad_norm": 2.4256569743524983,
+      "learning_rate": 2.7181356419709313e-08,
+      "loss": 0.9593,
+      "step": 9066
+    },
+    {
+      "epoch": 0.9535803962296396,
+      "grad_norm": 2.3740228072151393,
+      "learning_rate": 2.7058725088047466e-08,
+      "loss": 0.9844,
+      "step": 9067
+    },
+    {
+      "epoch": 0.953685566671487,
+      "grad_norm": 2.483387574850686,
+      "learning_rate": 2.6936369508548664e-08,
+      "loss": 0.9624,
+      "step": 9068
+    },
+    {
+      "epoch": 0.9537907371133343,
+      "grad_norm": 3.1395779842210954,
+      "learning_rate": 2.681428969485589e-08,
+      "loss": 1.0125,
+      "step": 9069
+    },
+    {
+      "epoch": 0.9538959075551816,
+      "grad_norm": 3.2864445331020185,
+      "learning_rate": 2.6692485660582133e-08,
+      "loss": 0.9238,
+      "step": 9070
+    },
+    {
+      "epoch": 0.9540010779970289,
+      "grad_norm": 2.251033657502406,
+      "learning_rate": 2.6570957419309595e-08,
+      "loss": 0.9286,
+      "step": 9071
+    },
+    {
+      "epoch": 0.9541062484388763,
+      "grad_norm": 2.9488738742860527,
+      "learning_rate": 2.6449704984589652e-08,
+      "loss": 1.0007,
+      "step": 9072
+    },
+    {
+      "epoch": 0.9542114188807236,
+      "grad_norm": 2.3323515832683945,
+      "learning_rate": 2.632872836994288e-08,
+      "loss": 1.0012,
+      "step": 9073
+    },
+    {
+      "epoch": 0.9543165893225709,
+      "grad_norm": 3.032634694644143,
+      "learning_rate": 2.6208027588858765e-08,
+      "loss": 0.9769,
+      "step": 9074
+    },
+    {
+      "epoch": 0.9544217597644182,
+      "grad_norm": 1.9718142374783398,
+      "learning_rate": 2.6087602654797097e-08,
+      "loss": 0.9451,
+      "step": 9075
+    },
+    {
+      "epoch": 0.9545269302062656,
+      "grad_norm": 2.261250300289189,
+      "learning_rate": 2.5967453581185187e-08,
+      "loss": 1.0058,
+      "step": 9076
+    },
+    {
+      "epoch": 0.9546321006481129,
+      "grad_norm": 2.202583197447805,
+      "learning_rate": 2.5847580381421768e-08,
+      "loss": 0.9913,
+      "step": 9077
+    },
+    {
+      "epoch": 0.9547372710899602,
+      "grad_norm": 3.0052761549204203,
+      "learning_rate": 2.5727983068872532e-08,
+      "loss": 0.979,
+      "step": 9078
+    },
+    {
+      "epoch": 0.9548424415318075,
+      "grad_norm": 2.235135902420106,
+      "learning_rate": 2.560866165687431e-08,
+      "loss": 0.9828,
+      "step": 9079
+    },
+    {
+      "epoch": 0.9549476119736549,
+      "grad_norm": 1.8821520458231678,
+      "learning_rate": 2.548961615873202e-08,
+      "loss": 0.9579,
+      "step": 9080
+    },
+    {
+      "epoch": 0.9550527824155022,
+      "grad_norm": 2.1799190509368263,
+      "learning_rate": 2.5370846587720044e-08,
+      "loss": 0.9611,
+      "step": 9081
+    },
+    {
+      "epoch": 0.9551579528573494,
+      "grad_norm": 2.2231469103632016,
+      "learning_rate": 2.5252352957082505e-08,
+      "loss": 0.9672,
+      "step": 9082
+    },
+    {
+      "epoch": 0.9552631232991967,
+      "grad_norm": 2.48946023734881,
+      "learning_rate": 2.513413528003189e-08,
+      "loss": 1.0058,
+      "step": 9083
+    },
+    {
+      "epoch": 0.955368293741044,
+      "grad_norm": 2.9722438755007383,
+      "learning_rate": 2.5016193569750712e-08,
+      "loss": 0.9313,
+      "step": 9084
+    },
+    {
+      "epoch": 0.9554734641828914,
+      "grad_norm": 2.5433251733042903,
+      "learning_rate": 2.4898527839390118e-08,
+      "loss": 1.0113,
+      "step": 9085
+    },
+    {
+      "epoch": 0.9555786346247387,
+      "grad_norm": 1.991419615066436,
+      "learning_rate": 2.4781138102071278e-08,
+      "loss": 0.9612,
+      "step": 9086
+    },
+    {
+      "epoch": 0.955683805066586,
+      "grad_norm": 2.4260785139851464,
+      "learning_rate": 2.4664024370883444e-08,
+      "loss": 0.9726,
+      "step": 9087
+    },
+    {
+      "epoch": 0.9557889755084333,
+      "grad_norm": 2.623866077425562,
+      "learning_rate": 2.454718665888589e-08,
+      "loss": 0.9465,
+      "step": 9088
+    },
+    {
+      "epoch": 0.9558941459502807,
+      "grad_norm": 2.6503139564233282,
+      "learning_rate": 2.4430624979107365e-08,
+      "loss": 1.0072,
+      "step": 9089
+    },
+    {
+      "epoch": 0.955999316392128,
+      "grad_norm": 4.298737564898903,
+      "learning_rate": 2.431433934454497e-08,
+      "loss": 0.9838,
+      "step": 9090
+    },
+    {
+      "epoch": 0.9561044868339753,
+      "grad_norm": 2.166453454009922,
+      "learning_rate": 2.4198329768165552e-08,
+      "loss": 0.9526,
+      "step": 9091
+    },
+    {
+      "epoch": 0.9562096572758226,
+      "grad_norm": 2.262782344611168,
+      "learning_rate": 2.4082596262904877e-08,
+      "loss": 0.909,
+      "step": 9092
+    },
+    {
+      "epoch": 0.95631482771767,
+      "grad_norm": 2.4483911651903343,
+      "learning_rate": 2.396713884166818e-08,
+      "loss": 0.9938,
+      "step": 9093
+    },
+    {
+      "epoch": 0.9564199981595173,
+      "grad_norm": 3.1826776450887686,
+      "learning_rate": 2.385195751733044e-08,
+      "loss": 0.9652,
+      "step": 9094
+    },
+    {
+      "epoch": 0.9565251686013646,
+      "grad_norm": 2.0215236542665114,
+      "learning_rate": 2.3737052302734432e-08,
+      "loss": 0.9577,
+      "step": 9095
+    },
+    {
+      "epoch": 0.9566303390432119,
+      "grad_norm": 2.988986889944041,
+      "learning_rate": 2.36224232106938e-08,
+      "loss": 0.9887,
+      "step": 9096
+    },
+    {
+      "epoch": 0.9567355094850593,
+      "grad_norm": 2.795418561696307,
+      "learning_rate": 2.3508070253989712e-08,
+      "loss": 0.9974,
+      "step": 9097
+    },
+    {
+      "epoch": 0.9568406799269066,
+      "grad_norm": 2.7766248044059876,
+      "learning_rate": 2.3393993445374187e-08,
+      "loss": 1.005,
+      "step": 9098
+    },
+    {
+      "epoch": 0.9569458503687539,
+      "grad_norm": 2.559873723889261,
+      "learning_rate": 2.3280192797567046e-08,
+      "loss": 0.9853,
+      "step": 9099
+    },
+    {
+      "epoch": 0.9570510208106012,
+      "grad_norm": 1.838309819852872,
+      "learning_rate": 2.31666683232587e-08,
+      "loss": 0.9774,
+      "step": 9100
+    },
+    {
+      "epoch": 0.9571561912524486,
+      "grad_norm": 2.3283682967266404,
+      "learning_rate": 2.3053420035107077e-08,
+      "loss": 0.9614,
+      "step": 9101
+    },
+    {
+      "epoch": 0.9572613616942958,
+      "grad_norm": 2.0531537931400696,
+      "learning_rate": 2.294044794574096e-08,
+      "loss": 0.996,
+      "step": 9102
+    },
+    {
+      "epoch": 0.9573665321361431,
+      "grad_norm": 2.738491776846182,
+      "learning_rate": 2.2827752067757224e-08,
+      "loss": 0.9989,
+      "step": 9103
+    },
+    {
+      "epoch": 0.9574717025779904,
+      "grad_norm": 2.8541624802569086,
+      "learning_rate": 2.271533241372248e-08,
+      "loss": 0.9925,
+      "step": 9104
+    },
+    {
+      "epoch": 0.9575768730198377,
+      "grad_norm": 2.561463355615611,
+      "learning_rate": 2.260318899617281e-08,
+      "loss": 0.995,
+      "step": 9105
+    },
+    {
+      "epoch": 0.9576820434616851,
+      "grad_norm": 2.4009062245111807,
+      "learning_rate": 2.24913218276121e-08,
+      "loss": 1.0175,
+      "step": 9106
+    },
+    {
+      "epoch": 0.9577872139035324,
+      "grad_norm": 1.9085583128594759,
+      "learning_rate": 2.2379730920515096e-08,
+      "loss": 0.966,
+      "step": 9107
+    },
+    {
+      "epoch": 0.9578923843453797,
+      "grad_norm": 2.3035781917937923,
+      "learning_rate": 2.2268416287325178e-08,
+      "loss": 0.9906,
+      "step": 9108
+    },
+    {
+      "epoch": 0.957997554787227,
+      "grad_norm": 2.9253676588000084,
+      "learning_rate": 2.215737794045436e-08,
+      "loss": 0.9913,
+      "step": 9109
+    },
+    {
+      "epoch": 0.9581027252290744,
+      "grad_norm": 2.251760542497009,
+      "learning_rate": 2.2046615892284685e-08,
+      "loss": 0.9787,
+      "step": 9110
+    },
+    {
+      "epoch": 0.9582078956709217,
+      "grad_norm": 2.6136064885904142,
+      "learning_rate": 2.193613015516627e-08,
+      "loss": 1.0012,
+      "step": 9111
+    },
+    {
+      "epoch": 0.958313066112769,
+      "grad_norm": 2.378051255621682,
+      "learning_rate": 2.1825920741420092e-08,
+      "loss": 0.9869,
+      "step": 9112
+    },
+    {
+      "epoch": 0.9584182365546163,
+      "grad_norm": 2.420924559361127,
+      "learning_rate": 2.171598766333466e-08,
+      "loss": 0.9531,
+      "step": 9113
+    },
+    {
+      "epoch": 0.9585234069964637,
+      "grad_norm": 2.1094674393987125,
+      "learning_rate": 2.1606330933168496e-08,
+      "loss": 0.9684,
+      "step": 9114
+    },
+    {
+      "epoch": 0.958628577438311,
+      "grad_norm": 1.7218971943565433,
+      "learning_rate": 2.149695056314932e-08,
+      "loss": 0.9608,
+      "step": 9115
+    },
+    {
+      "epoch": 0.9587337478801583,
+      "grad_norm": 2.6145718462197967,
+      "learning_rate": 2.1387846565474047e-08,
+      "loss": 0.9348,
+      "step": 9116
+    },
+    {
+      "epoch": 0.9588389183220056,
+      "grad_norm": 2.7778015979594595,
+      "learning_rate": 2.1279018952308218e-08,
+      "loss": 1.015,
+      "step": 9117
+    },
+    {
+      "epoch": 0.958944088763853,
+      "grad_norm": 3.616813798965242,
+      "learning_rate": 2.1170467735787124e-08,
+      "loss": 0.9358,
+      "step": 9118
+    },
+    {
+      "epoch": 0.9590492592057003,
+      "grad_norm": 3.0938864234470684,
+      "learning_rate": 2.1062192928015536e-08,
+      "loss": 0.9831,
+      "step": 9119
+    },
+    {
+      "epoch": 0.9591544296475476,
+      "grad_norm": 2.5032862302480696,
+      "learning_rate": 2.095419454106573e-08,
+      "loss": 1.0055,
+      "step": 9120
+    },
+    {
+      "epoch": 0.9592596000893949,
+      "grad_norm": 2.4155226421641234,
+      "learning_rate": 2.084647258698169e-08,
+      "loss": 0.9836,
+      "step": 9121
+    },
+    {
+      "epoch": 0.9593647705312421,
+      "grad_norm": 2.2456972979046212,
+      "learning_rate": 2.073902707777464e-08,
+      "loss": 1.0023,
+      "step": 9122
+    },
+    {
+      "epoch": 0.9594699409730895,
+      "grad_norm": 3.1418236010714846,
+      "learning_rate": 2.0631858025425554e-08,
+      "loss": 0.9944,
+      "step": 9123
+    },
+    {
+      "epoch": 0.9595751114149368,
+      "grad_norm": 2.512431786493916,
+      "learning_rate": 2.052496544188487e-08,
+      "loss": 0.9862,
+      "step": 9124
+    },
+    {
+      "epoch": 0.9596802818567841,
+      "grad_norm": 2.4317108069473514,
+      "learning_rate": 2.0418349339071385e-08,
+      "loss": 0.9816,
+      "step": 9125
+    },
+    {
+      "epoch": 0.9597854522986314,
+      "grad_norm": 2.5061823875565654,
+      "learning_rate": 2.031200972887448e-08,
+      "loss": 0.9703,
+      "step": 9126
+    },
+    {
+      "epoch": 0.9598906227404788,
+      "grad_norm": 1.8971400885405667,
+      "learning_rate": 2.0205946623151063e-08,
+      "loss": 0.9792,
+      "step": 9127
+    },
+    {
+      "epoch": 0.9599957931823261,
+      "grad_norm": 2.8512234421263516,
+      "learning_rate": 2.010016003372861e-08,
+      "loss": 0.9609,
+      "step": 9128
+    },
+    {
+      "epoch": 0.9601009636241734,
+      "grad_norm": 1.8398551494029969,
+      "learning_rate": 1.9994649972402415e-08,
+      "loss": 0.9747,
+      "step": 9129
+    },
+    {
+      "epoch": 0.9602061340660207,
+      "grad_norm": 2.4990349994310415,
+      "learning_rate": 1.9889416450938335e-08,
+      "loss": 0.9787,
+      "step": 9130
+    },
+    {
+      "epoch": 0.9603113045078681,
+      "grad_norm": 1.9550532334702702,
+      "learning_rate": 1.9784459481070607e-08,
+      "loss": 0.9957,
+      "step": 9131
+    },
+    {
+      "epoch": 0.9604164749497154,
+      "grad_norm": 3.2601942510164252,
+      "learning_rate": 1.9679779074502636e-08,
+      "loss": 0.9309,
+      "step": 9132
+    },
+    {
+      "epoch": 0.9605216453915627,
+      "grad_norm": 2.862772288610713,
+      "learning_rate": 1.9575375242907035e-08,
+      "loss": 1.0278,
+      "step": 9133
+    },
+    {
+      "epoch": 0.96062681583341,
+      "grad_norm": 2.6535531520322495,
+      "learning_rate": 1.9471247997925324e-08,
+      "loss": 0.9815,
+      "step": 9134
+    },
+    {
+      "epoch": 0.9607319862752574,
+      "grad_norm": 2.6376660990295164,
+      "learning_rate": 1.9367397351169326e-08,
+      "loss": 0.9858,
+      "step": 9135
+    },
+    {
+      "epoch": 0.9608371567171047,
+      "grad_norm": 3.208400601816702,
+      "learning_rate": 1.9263823314218667e-08,
+      "loss": 0.9999,
+      "step": 9136
+    },
+    {
+      "epoch": 0.960942327158952,
+      "grad_norm": 3.117055000265947,
+      "learning_rate": 1.9160525898622716e-08,
+      "loss": 1.0176,
+      "step": 9137
+    },
+    {
+      "epoch": 0.9610474976007993,
+      "grad_norm": 2.539642788145546,
+      "learning_rate": 1.9057505115900043e-08,
+      "loss": 0.9976,
+      "step": 9138
+    },
+    {
+      "epoch": 0.9611526680426467,
+      "grad_norm": 2.622034495146119,
+      "learning_rate": 1.8954760977538122e-08,
+      "loss": 0.9342,
+      "step": 9139
+    },
+    {
+      "epoch": 0.961257838484494,
+      "grad_norm": 2.0976876781379796,
+      "learning_rate": 1.88522934949939e-08,
+      "loss": 0.939,
+      "step": 9140
+    },
+    {
+      "epoch": 0.9613630089263413,
+      "grad_norm": 3.6252618212734244,
+      "learning_rate": 1.875010267969296e-08,
+      "loss": 1.0144,
+      "step": 9141
+    },
+    {
+      "epoch": 0.9614681793681886,
+      "grad_norm": 2.2004361692134093,
+      "learning_rate": 1.864818854303091e-08,
+      "loss": 0.9742,
+      "step": 9142
+    },
+    {
+      "epoch": 0.9615733498100358,
+      "grad_norm": 2.5659276984677013,
+      "learning_rate": 1.8546551096371157e-08,
+      "loss": 0.994,
+      "step": 9143
+    },
+    {
+      "epoch": 0.9616785202518832,
+      "grad_norm": 2.2606940570606304,
+      "learning_rate": 1.844519035104797e-08,
+      "loss": 0.9785,
+      "step": 9144
+    },
+    {
+      "epoch": 0.9617836906937305,
+      "grad_norm": 3.1995639013446353,
+      "learning_rate": 1.834410631836342e-08,
+      "loss": 1.0375,
+      "step": 9145
+    },
+    {
+      "epoch": 0.9618888611355778,
+      "grad_norm": 2.8120072161352883,
+      "learning_rate": 1.8243299009589044e-08,
+      "loss": 0.9765,
+      "step": 9146
+    },
+    {
+      "epoch": 0.9619940315774251,
+      "grad_norm": 2.9702132072664504,
+      "learning_rate": 1.814276843596585e-08,
+      "loss": 0.984,
+      "step": 9147
+    },
+    {
+      "epoch": 0.9620992020192725,
+      "grad_norm": 2.36435367996225,
+      "learning_rate": 1.8042514608703765e-08,
+      "loss": 0.9905,
+      "step": 9148
+    },
+    {
+      "epoch": 0.9622043724611198,
+      "grad_norm": 2.4402173766387647,
+      "learning_rate": 1.7942537538981618e-08,
+      "loss": 0.9907,
+      "step": 9149
+    },
+    {
+      "epoch": 0.9623095429029671,
+      "grad_norm": 2.3039341033452048,
+      "learning_rate": 1.7842837237947997e-08,
+      "loss": 0.9937,
+      "step": 9150
+    },
+    {
+      "epoch": 0.9624147133448144,
+      "grad_norm": 2.5814133426313397,
+      "learning_rate": 1.7743413716720394e-08,
+      "loss": 0.9995,
+      "step": 9151
+    },
+    {
+      "epoch": 0.9625198837866618,
+      "grad_norm": 3.0310519987593323,
+      "learning_rate": 1.764426698638466e-08,
+      "loss": 0.9418,
+      "step": 9152
+    },
+    {
+      "epoch": 0.9626250542285091,
+      "grad_norm": 2.414115329844662,
+      "learning_rate": 1.7545397057996683e-08,
+      "loss": 1.0066,
+      "step": 9153
+    },
+    {
+      "epoch": 0.9627302246703564,
+      "grad_norm": 2.5905538989812484,
+      "learning_rate": 1.7446803942581524e-08,
+      "loss": 0.9571,
+      "step": 9154
+    },
+    {
+      "epoch": 0.9628353951122037,
+      "grad_norm": 2.4397822721298588,
+      "learning_rate": 1.7348487651132895e-08,
+      "loss": 0.9677,
+      "step": 9155
+    },
+    {
+      "epoch": 0.962940565554051,
+      "grad_norm": 2.369844437844643,
+      "learning_rate": 1.725044819461369e-08,
+      "loss": 0.9505,
+      "step": 9156
+    },
+    {
+      "epoch": 0.9630457359958984,
+      "grad_norm": 2.4593061093551927,
+      "learning_rate": 1.7152685583955995e-08,
+      "loss": 0.9638,
+      "step": 9157
+    },
+    {
+      "epoch": 0.9631509064377457,
+      "grad_norm": 2.058831249668448,
+      "learning_rate": 1.7055199830061653e-08,
+      "loss": 0.9734,
+      "step": 9158
+    },
+    {
+      "epoch": 0.963256076879593,
+      "grad_norm": 2.5481002332248703,
+      "learning_rate": 1.6957990943800574e-08,
+      "loss": 0.9717,
+      "step": 9159
+    },
+    {
+      "epoch": 0.9633612473214404,
+      "grad_norm": 3.1600633484938503,
+      "learning_rate": 1.686105893601242e-08,
+      "loss": 1.0042,
+      "step": 9160
+    },
+    {
+      "epoch": 0.9634664177632877,
+      "grad_norm": 3.3311087274523823,
+      "learning_rate": 1.6764403817506047e-08,
+      "loss": 1.0101,
+      "step": 9161
+    },
+    {
+      "epoch": 0.963571588205135,
+      "grad_norm": 1.9545499141129272,
+      "learning_rate": 1.6668025599058945e-08,
+      "loss": 0.9977,
+      "step": 9162
+    },
+    {
+      "epoch": 0.9636767586469822,
+      "grad_norm": 2.787786032166771,
+      "learning_rate": 1.6571924291418072e-08,
+      "loss": 0.9881,
+      "step": 9163
+    },
+    {
+      "epoch": 0.9637819290888295,
+      "grad_norm": 2.7339024071717035,
+      "learning_rate": 1.6476099905299857e-08,
+      "loss": 0.9863,
+      "step": 9164
+    },
+    {
+      "epoch": 0.9638870995306769,
+      "grad_norm": 2.3241276868769507,
+      "learning_rate": 1.6380552451389088e-08,
+      "loss": 0.985,
+      "step": 9165
+    },
+    {
+      "epoch": 0.9639922699725242,
+      "grad_norm": 2.615775704959889,
+      "learning_rate": 1.6285281940340016e-08,
+      "loss": 1.0107,
+      "step": 9166
+    },
+    {
+      "epoch": 0.9640974404143715,
+      "grad_norm": 2.61782809887508,
+      "learning_rate": 1.6190288382776363e-08,
+      "loss": 0.9895,
+      "step": 9167
+    },
+    {
+      "epoch": 0.9642026108562188,
+      "grad_norm": 2.9697250302024987,
+      "learning_rate": 1.60955717892905e-08,
+      "loss": 0.9869,
+      "step": 9168
+    },
+    {
+      "epoch": 0.9643077812980662,
+      "grad_norm": 2.1659305791507024,
+      "learning_rate": 1.6001132170443968e-08,
+      "loss": 0.9508,
+      "step": 9169
+    },
+    {
+      "epoch": 0.9644129517399135,
+      "grad_norm": 2.76396965745972,
+      "learning_rate": 1.5906969536767513e-08,
+      "loss": 0.9662,
+      "step": 9170
+    },
+    {
+      "epoch": 0.9645181221817608,
+      "grad_norm": 2.1674061651202434,
+      "learning_rate": 1.5813083898760793e-08,
+      "loss": 0.961,
+      "step": 9171
+    },
+    {
+      "epoch": 0.9646232926236081,
+      "grad_norm": 3.3828957028419944,
+      "learning_rate": 1.571947526689349e-08,
+      "loss": 0.9774,
+      "step": 9172
+    },
+    {
+      "epoch": 0.9647284630654555,
+      "grad_norm": 2.9496229688779834,
+      "learning_rate": 1.5626143651603087e-08,
+      "loss": 1.0287,
+      "step": 9173
+    },
+    {
+      "epoch": 0.9648336335073028,
+      "grad_norm": 2.576235051854255,
+      "learning_rate": 1.553308906329709e-08,
+      "loss": 1.0086,
+      "step": 9174
+    },
+    {
+      "epoch": 0.9649388039491501,
+      "grad_norm": 2.6024764773615705,
+      "learning_rate": 1.5440311512351646e-08,
+      "loss": 0.9895,
+      "step": 9175
+    },
+    {
+      "epoch": 0.9650439743909974,
+      "grad_norm": 2.7226540157491175,
+      "learning_rate": 1.534781100911209e-08,
+      "loss": 0.9951,
+      "step": 9176
+    },
+    {
+      "epoch": 0.9651491448328448,
+      "grad_norm": 2.4235683696464925,
+      "learning_rate": 1.5255587563893227e-08,
+      "loss": 1.0094,
+      "step": 9177
+    },
+    {
+      "epoch": 0.9652543152746921,
+      "grad_norm": 1.8039244408062456,
+      "learning_rate": 1.5163641186978216e-08,
+      "loss": 0.9846,
+      "step": 9178
+    },
+    {
+      "epoch": 0.9653594857165394,
+      "grad_norm": 2.208004892548754,
+      "learning_rate": 1.507197188862053e-08,
+      "loss": 0.9821,
+      "step": 9179
+    },
+    {
+      "epoch": 0.9654646561583867,
+      "grad_norm": 2.686724949615566,
+      "learning_rate": 1.4980579679041153e-08,
+      "loss": 0.9814,
+      "step": 9180
+    },
+    {
+      "epoch": 0.965569826600234,
+      "grad_norm": 2.218279129899888,
+      "learning_rate": 1.4889464568431656e-08,
+      "loss": 0.9947,
+      "step": 9181
+    },
+    {
+      "epoch": 0.9656749970420814,
+      "grad_norm": 2.619200221829575,
+      "learning_rate": 1.4798626566951968e-08,
+      "loss": 0.9968,
+      "step": 9182
+    },
+    {
+      "epoch": 0.9657801674839286,
+      "grad_norm": 2.226366283605976,
+      "learning_rate": 1.4708065684730932e-08,
+      "loss": 0.9942,
+      "step": 9183
+    },
+    {
+      "epoch": 0.9658853379257759,
+      "grad_norm": 1.8711672309528635,
+      "learning_rate": 1.4617781931867137e-08,
+      "loss": 0.9866,
+      "step": 9184
+    },
+    {
+      "epoch": 0.9659905083676232,
+      "grad_norm": 1.7591848744904266,
+      "learning_rate": 1.4527775318427806e-08,
+      "loss": 0.9644,
+      "step": 9185
+    },
+    {
+      "epoch": 0.9660956788094706,
+      "grad_norm": 2.3772749891154312,
+      "learning_rate": 1.4438045854449357e-08,
+      "loss": 0.9752,
+      "step": 9186
+    },
+    {
+      "epoch": 0.9662008492513179,
+      "grad_norm": 2.79994419425291,
+      "learning_rate": 1.4348593549937118e-08,
+      "loss": 1.0253,
+      "step": 9187
+    },
+    {
+      "epoch": 0.9663060196931652,
+      "grad_norm": 2.9430470064229963,
+      "learning_rate": 1.4259418414866166e-08,
+      "loss": 0.9725,
+      "step": 9188
+    },
+    {
+      "epoch": 0.9664111901350125,
+      "grad_norm": 2.844447657650461,
+      "learning_rate": 1.417052045917966e-08,
+      "loss": 1.0171,
+      "step": 9189
+    },
+    {
+      "epoch": 0.9665163605768599,
+      "grad_norm": 3.2049646354495067,
+      "learning_rate": 1.4081899692791058e-08,
+      "loss": 0.9617,
+      "step": 9190
+    },
+    {
+      "epoch": 0.9666215310187072,
+      "grad_norm": 2.6622301953211487,
+      "learning_rate": 1.399355612558162e-08,
+      "loss": 0.9837,
+      "step": 9191
+    },
+    {
+      "epoch": 0.9667267014605545,
+      "grad_norm": 2.122985949724431,
+      "learning_rate": 1.3905489767402913e-08,
+      "loss": 0.9636,
+      "step": 9192
+    },
+    {
+      "epoch": 0.9668318719024018,
+      "grad_norm": 2.6769674621149533,
+      "learning_rate": 1.3817700628074582e-08,
+      "loss": 0.9456,
+      "step": 9193
+    },
+    {
+      "epoch": 0.9669370423442492,
+      "grad_norm": 2.365588178179128,
+      "learning_rate": 1.3730188717386016e-08,
+      "loss": 0.9796,
+      "step": 9194
+    },
+    {
+      "epoch": 0.9670422127860965,
+      "grad_norm": 1.8396720663564547,
+      "learning_rate": 1.3642954045095525e-08,
+      "loss": 0.981,
+      "step": 9195
+    },
+    {
+      "epoch": 0.9671473832279438,
+      "grad_norm": 2.3943971715102736,
+      "learning_rate": 1.3555996620930323e-08,
+      "loss": 1.0078,
+      "step": 9196
+    },
+    {
+      "epoch": 0.9672525536697911,
+      "grad_norm": 2.419224654846345,
+      "learning_rate": 1.3469316454586823e-08,
+      "loss": 0.9814,
+      "step": 9197
+    },
+    {
+      "epoch": 0.9673577241116385,
+      "grad_norm": 2.4530714452657625,
+      "learning_rate": 1.3382913555730626e-08,
+      "loss": 0.9688,
+      "step": 9198
+    },
+    {
+      "epoch": 0.9674628945534858,
+      "grad_norm": 2.2680814906299824,
+      "learning_rate": 1.3296787933996246e-08,
+      "loss": 0.9347,
+      "step": 9199
+    },
+    {
+      "epoch": 0.9675680649953331,
+      "grad_norm": 2.189230585914418,
+      "learning_rate": 1.3210939598987394e-08,
+      "loss": 0.9303,
+      "step": 9200
+    },
+    {
+      "epoch": 0.9676732354371804,
+      "grad_norm": 2.3582843375690214,
+      "learning_rate": 1.3125368560276686e-08,
+      "loss": 1.0175,
+      "step": 9201
+    },
+    {
+      "epoch": 0.9677784058790277,
+      "grad_norm": 2.268975158784598,
+      "learning_rate": 1.3040074827406491e-08,
+      "loss": 0.9531,
+      "step": 9202
+    },
+    {
+      "epoch": 0.9678835763208751,
+      "grad_norm": 2.5739483459832098,
+      "learning_rate": 1.2955058409886978e-08,
+      "loss": 0.9481,
+      "step": 9203
+    },
+    {
+      "epoch": 0.9679887467627223,
+      "grad_norm": 1.8224473421389638,
+      "learning_rate": 1.2870319317198621e-08,
+      "loss": 0.9507,
+      "step": 9204
+    },
+    {
+      "epoch": 0.9680939172045696,
+      "grad_norm": 2.8730850996320743,
+      "learning_rate": 1.2785857558790526e-08,
+      "loss": 0.9602,
+      "step": 9205
+    },
+    {
+      "epoch": 0.9681990876464169,
+      "grad_norm": 2.724396836974365,
+      "learning_rate": 1.270167314408044e-08,
+      "loss": 0.9638,
+      "step": 9206
+    },
+    {
+      "epoch": 0.9683042580882643,
+      "grad_norm": 2.1481024531914046,
+      "learning_rate": 1.261776608245585e-08,
+      "loss": 1.0108,
+      "step": 9207
+    },
+    {
+      "epoch": 0.9684094285301116,
+      "grad_norm": 2.528448588522008,
+      "learning_rate": 1.2534136383272888e-08,
+      "loss": 0.9794,
+      "step": 9208
+    },
+    {
+      "epoch": 0.9685145989719589,
+      "grad_norm": 3.49084002666298,
+      "learning_rate": 1.2450784055857145e-08,
+      "loss": 0.9712,
+      "step": 9209
+    },
+    {
+      "epoch": 0.9686197694138062,
+      "grad_norm": 2.619678363223223,
+      "learning_rate": 1.2367709109503134e-08,
+      "loss": 0.9621,
+      "step": 9210
+    },
+    {
+      "epoch": 0.9687249398556536,
+      "grad_norm": 2.8477945371212465,
+      "learning_rate": 1.2284911553474e-08,
+      "loss": 0.9999,
+      "step": 9211
+    },
+    {
+      "epoch": 0.9688301102975009,
+      "grad_norm": 2.5699605691953225,
+      "learning_rate": 1.2202391397002355e-08,
+      "loss": 0.9942,
+      "step": 9212
+    },
+    {
+      "epoch": 0.9689352807393482,
+      "grad_norm": 2.96718335182804,
+      "learning_rate": 1.2120148649290008e-08,
+      "loss": 0.9763,
+      "step": 9213
+    },
+    {
+      "epoch": 0.9690404511811955,
+      "grad_norm": 2.7754708820964806,
+      "learning_rate": 1.2038183319507957e-08,
+      "loss": 0.9626,
+      "step": 9214
+    },
+    {
+      "epoch": 0.9691456216230429,
+      "grad_norm": 2.3177867748746466,
+      "learning_rate": 1.1956495416795277e-08,
+      "loss": 1.0024,
+      "step": 9215
+    },
+    {
+      "epoch": 0.9692507920648902,
+      "grad_norm": 2.593173481686718,
+      "learning_rate": 1.1875084950261351e-08,
+      "loss": 1.0028,
+      "step": 9216
+    },
+    {
+      "epoch": 0.9693559625067375,
+      "grad_norm": 2.9154573832597452,
+      "learning_rate": 1.1793951928983639e-08,
+      "loss": 0.9496,
+      "step": 9217
+    },
+    {
+      "epoch": 0.9694611329485848,
+      "grad_norm": 2.4979489626751343,
+      "learning_rate": 1.1713096362009346e-08,
+      "loss": 0.9504,
+      "step": 9218
+    },
+    {
+      "epoch": 0.9695663033904321,
+      "grad_norm": 1.9672731194839046,
+      "learning_rate": 1.1632518258354875e-08,
+      "loss": 0.9085,
+      "step": 9219
+    },
+    {
+      "epoch": 0.9696714738322795,
+      "grad_norm": 2.481803674239574,
+      "learning_rate": 1.1552217627004426e-08,
+      "loss": 0.966,
+      "step": 9220
+    },
+    {
+      "epoch": 0.9697766442741268,
+      "grad_norm": 2.6552953440994282,
+      "learning_rate": 1.1472194476913057e-08,
+      "loss": 1.0244,
+      "step": 9221
+    },
+    {
+      "epoch": 0.9698818147159741,
+      "grad_norm": 2.7197091300670135,
+      "learning_rate": 1.1392448817003354e-08,
+      "loss": 0.9968,
+      "step": 9222
+    },
+    {
+      "epoch": 0.9699869851578214,
+      "grad_norm": 2.067177962113043,
+      "learning_rate": 1.1312980656167927e-08,
+      "loss": 0.9947,
+      "step": 9223
+    },
+    {
+      "epoch": 0.9700921555996687,
+      "grad_norm": 2.347224718073338,
+      "learning_rate": 1.1233790003267741e-08,
+      "loss": 0.9577,
+      "step": 9224
+    },
+    {
+      "epoch": 0.970197326041516,
+      "grad_norm": 5.618522310449893,
+      "learning_rate": 1.115487686713379e-08,
+      "loss": 1.0061,
+      "step": 9225
+    },
+    {
+      "epoch": 0.9703024964833633,
+      "grad_norm": 1.7054244073511104,
+      "learning_rate": 1.107624125656459e-08,
+      "loss": 0.9655,
+      "step": 9226
+    },
+    {
+      "epoch": 0.9704076669252106,
+      "grad_norm": 2.0088677460681734,
+      "learning_rate": 1.0997883180329515e-08,
+      "loss": 0.985,
+      "step": 9227
+    },
+    {
+      "epoch": 0.970512837367058,
+      "grad_norm": 3.0153539597861982,
+      "learning_rate": 1.0919802647165467e-08,
+      "loss": 0.9946,
+      "step": 9228
+    },
+    {
+      "epoch": 0.9706180078089053,
+      "grad_norm": 3.353254395720258,
+      "learning_rate": 1.0841999665779368e-08,
+      "loss": 1.0188,
+      "step": 9229
+    },
+    {
+      "epoch": 0.9707231782507526,
+      "grad_norm": 2.081090994792908,
+      "learning_rate": 1.0764474244846778e-08,
+      "loss": 0.9783,
+      "step": 9230
+    },
+    {
+      "epoch": 0.9708283486925999,
+      "grad_norm": 3.243396799607356,
+      "learning_rate": 1.0687226393012173e-08,
+      "loss": 0.9995,
+      "step": 9231
+    },
+    {
+      "epoch": 0.9709335191344473,
+      "grad_norm": 1.8699534084309162,
+      "learning_rate": 1.0610256118889772e-08,
+      "loss": 0.9539,
+      "step": 9232
+    },
+    {
+      "epoch": 0.9710386895762946,
+      "grad_norm": 2.480474240008781,
+      "learning_rate": 1.0533563431062154e-08,
+      "loss": 0.9863,
+      "step": 9233
+    },
+    {
+      "epoch": 0.9711438600181419,
+      "grad_norm": 2.002217087521726,
+      "learning_rate": 1.0457148338080813e-08,
+      "loss": 0.9804,
+      "step": 9234
+    },
+    {
+      "epoch": 0.9712490304599892,
+      "grad_norm": 1.5904141400379188,
+      "learning_rate": 1.0381010848466988e-08,
+      "loss": 0.9646,
+      "step": 9235
+    },
+    {
+      "epoch": 0.9713542009018366,
+      "grad_norm": 2.4366903231145454,
+      "learning_rate": 1.0305150970710276e-08,
+      "loss": 0.9668,
+      "step": 9236
+    },
+    {
+      "epoch": 0.9714593713436839,
+      "grad_norm": 2.822796227094747,
+      "learning_rate": 1.02295687132703e-08,
+      "loss": 1.0271,
+      "step": 9237
+    },
+    {
+      "epoch": 0.9715645417855312,
+      "grad_norm": 1.9912340131203807,
+      "learning_rate": 1.0154264084574206e-08,
+      "loss": 0.9428,
+      "step": 9238
+    },
+    {
+      "epoch": 0.9716697122273785,
+      "grad_norm": 1.955104338884243,
+      "learning_rate": 1.0079237093019722e-08,
+      "loss": 0.9747,
+      "step": 9239
+    },
+    {
+      "epoch": 0.9717748826692258,
+      "grad_norm": 3.2050011615950793,
+      "learning_rate": 1.0004487746972657e-08,
+      "loss": 0.9594,
+      "step": 9240
+    },
+    {
+      "epoch": 0.9718800531110732,
+      "grad_norm": 2.526175674743241,
+      "learning_rate": 9.930016054768011e-09,
+      "loss": 0.9817,
+      "step": 9241
+    },
+    {
+      "epoch": 0.9719852235529205,
+      "grad_norm": 2.425815712365738,
+      "learning_rate": 9.85582202471025e-09,
+      "loss": 1.0223,
+      "step": 9242
+    },
+    {
+      "epoch": 0.9720903939947678,
+      "grad_norm": 2.1478474630955557,
+      "learning_rate": 9.7819056650722e-09,
+      "loss": 0.9611,
+      "step": 9243
+    },
+    {
+      "epoch": 0.972195564436615,
+      "grad_norm": 2.63035991433718,
+      "learning_rate": 9.708266984096714e-09,
+      "loss": 0.9549,
+      "step": 9244
+    },
+    {
+      "epoch": 0.9723007348784624,
+      "grad_norm": 2.787607325293226,
+      "learning_rate": 9.634905989994448e-09,
+      "loss": 0.9642,
+      "step": 9245
+    },
+    {
+      "epoch": 0.9724059053203097,
+      "grad_norm": 2.7011576256397056,
+      "learning_rate": 9.561822690946076e-09,
+      "loss": 0.935,
+      "step": 9246
+    },
+    {
+      "epoch": 0.972511075762157,
+      "grad_norm": 3.3865296605768545,
+      "learning_rate": 9.489017095100916e-09,
+      "loss": 0.9845,
+      "step": 9247
+    },
+    {
+      "epoch": 0.9726162462040043,
+      "grad_norm": 2.141639825480503,
+      "learning_rate": 9.416489210577473e-09,
+      "loss": 0.9714,
+      "step": 9248
+    },
+    {
+      "epoch": 0.9727214166458517,
+      "grad_norm": 2.381480065424657,
+      "learning_rate": 9.344239045462888e-09,
+      "loss": 1.0016,
+      "step": 9249
+    },
+    {
+      "epoch": 0.972826587087699,
+      "grad_norm": 3.061957347212088,
+      "learning_rate": 9.272266607813774e-09,
+      "loss": 1.0217,
+      "step": 9250
+    },
+    {
+      "epoch": 0.9729317575295463,
+      "grad_norm": 2.5614131075584416,
+      "learning_rate": 9.200571905655376e-09,
+      "loss": 0.9773,
+      "step": 9251
+    },
+    {
+      "epoch": 0.9730369279713936,
+      "grad_norm": 2.6614512876627745,
+      "learning_rate": 9.129154946982687e-09,
+      "loss": 1.009,
+      "step": 9252
+    },
+    {
+      "epoch": 0.973142098413241,
+      "grad_norm": 2.5591923309907956,
+      "learning_rate": 9.058015739758786e-09,
+      "loss": 0.9382,
+      "step": 9253
+    },
+    {
+      "epoch": 0.9732472688550883,
+      "grad_norm": 3.4612772871088255,
+      "learning_rate": 8.98715429191649e-09,
+      "loss": 0.9797,
+      "step": 9254
+    },
+    {
+      "epoch": 0.9733524392969356,
+      "grad_norm": 2.161508443645297,
+      "learning_rate": 8.916570611357534e-09,
+      "loss": 1.0028,
+      "step": 9255
+    },
+    {
+      "epoch": 0.9734576097387829,
+      "grad_norm": 2.777809434537492,
+      "learning_rate": 8.84626470595229e-09,
+      "loss": 0.9781,
+      "step": 9256
+    },
+    {
+      "epoch": 0.9735627801806302,
+      "grad_norm": 3.256651493127484,
+      "learning_rate": 8.776236583540321e-09,
+      "loss": 0.9788,
+      "step": 9257
+    },
+    {
+      "epoch": 0.9736679506224776,
+      "grad_norm": 3.137061562152753,
+      "learning_rate": 8.706486251930657e-09,
+      "loss": 0.9969,
+      "step": 9258
+    },
+    {
+      "epoch": 0.9737731210643249,
+      "grad_norm": 1.6215241145993706,
+      "learning_rate": 8.637013718900689e-09,
+      "loss": 0.9675,
+      "step": 9259
+    },
+    {
+      "epoch": 0.9738782915061722,
+      "grad_norm": 2.1994977784123075,
+      "learning_rate": 8.567818992197274e-09,
+      "loss": 0.9876,
+      "step": 9260
+    },
+    {
+      "epoch": 0.9739834619480195,
+      "grad_norm": 2.4037382184819585,
+      "learning_rate": 8.498902079536186e-09,
+      "loss": 0.9774,
+      "step": 9261
+    },
+    {
+      "epoch": 0.9740886323898669,
+      "grad_norm": 2.824206149642182,
+      "learning_rate": 8.430262988602389e-09,
+      "loss": 1.0053,
+      "step": 9262
+    },
+    {
+      "epoch": 0.9741938028317142,
+      "grad_norm": 2.9847969505243213,
+      "learning_rate": 8.361901727049204e-09,
+      "loss": 0.9833,
+      "step": 9263
+    },
+    {
+      "epoch": 0.9742989732735615,
+      "grad_norm": 2.358889620889671,
+      "learning_rate": 8.293818302499701e-09,
+      "loss": 0.9803,
+      "step": 9264
+    },
+    {
+      "epoch": 0.9744041437154087,
+      "grad_norm": 2.5381745630996044,
+      "learning_rate": 8.226012722545863e-09,
+      "loss": 0.9937,
+      "step": 9265
+    },
+    {
+      "epoch": 0.9745093141572561,
+      "grad_norm": 2.6358498110575583,
+      "learning_rate": 8.158484994748306e-09,
+      "loss": 0.9967,
+      "step": 9266
+    },
+    {
+      "epoch": 0.9746144845991034,
+      "grad_norm": 3.2362925404197846,
+      "learning_rate": 8.09123512663712e-09,
+      "loss": 1.0093,
+      "step": 9267
+    },
+    {
+      "epoch": 0.9747196550409507,
+      "grad_norm": 2.1160461048243304,
+      "learning_rate": 8.024263125710751e-09,
+      "loss": 1.0194,
+      "step": 9268
+    },
+    {
+      "epoch": 0.974824825482798,
+      "grad_norm": 2.1609581262482456,
+      "learning_rate": 7.95756899943767e-09,
+      "loss": 0.9649,
+      "step": 9269
+    },
+    {
+      "epoch": 0.9749299959246454,
+      "grad_norm": 3.1491039134025276,
+      "learning_rate": 7.891152755254427e-09,
+      "loss": 0.9911,
+      "step": 9270
+    },
+    {
+      "epoch": 0.9750351663664927,
+      "grad_norm": 2.8428124781388378,
+      "learning_rate": 7.825014400567044e-09,
+      "loss": 0.9895,
+      "step": 9271
+    },
+    {
+      "epoch": 0.97514033680834,
+      "grad_norm": 2.2055413871514307,
+      "learning_rate": 7.759153942750174e-09,
+      "loss": 0.9915,
+      "step": 9272
+    },
+    {
+      "epoch": 0.9752455072501873,
+      "grad_norm": 2.1726712220112545,
+      "learning_rate": 7.693571389148224e-09,
+      "loss": 0.9897,
+      "step": 9273
+    },
+    {
+      "epoch": 0.9753506776920347,
+      "grad_norm": 2.720353151960456,
+      "learning_rate": 7.628266747074231e-09,
+      "loss": 0.9897,
+      "step": 9274
+    },
+    {
+      "epoch": 0.975455848133882,
+      "grad_norm": 2.7591952094504424,
+      "learning_rate": 7.563240023809593e-09,
+      "loss": 0.9615,
+      "step": 9275
+    },
+    {
+      "epoch": 0.9755610185757293,
+      "grad_norm": 2.737597454130178,
+      "learning_rate": 7.498491226605731e-09,
+      "loss": 0.9927,
+      "step": 9276
+    },
+    {
+      "epoch": 0.9756661890175766,
+      "grad_norm": 2.5948670808268184,
+      "learning_rate": 7.434020362682703e-09,
+      "loss": 0.9945,
+      "step": 9277
+    },
+    {
+      "epoch": 0.975771359459424,
+      "grad_norm": 2.784078394241727,
+      "learning_rate": 7.369827439228927e-09,
+      "loss": 0.96,
+      "step": 9278
+    },
+    {
+      "epoch": 0.9758765299012713,
+      "grad_norm": 3.0703548564961394,
+      "learning_rate": 7.30591246340312e-09,
+      "loss": 0.9259,
+      "step": 9279
+    },
+    {
+      "epoch": 0.9759817003431186,
+      "grad_norm": 2.8232566736257385,
+      "learning_rate": 7.242275442332081e-09,
+      "loss": 0.9477,
+      "step": 9280
+    },
+    {
+      "epoch": 0.9760868707849659,
+      "grad_norm": 2.0272196755850214,
+      "learning_rate": 7.178916383111523e-09,
+      "loss": 0.9936,
+      "step": 9281
+    },
+    {
+      "epoch": 0.9761920412268132,
+      "grad_norm": 2.292451442175712,
+      "learning_rate": 7.1158352928066276e-09,
+      "loss": 0.9719,
+      "step": 9282
+    },
+    {
+      "epoch": 0.9762972116686606,
+      "grad_norm": 2.830967058321669,
+      "learning_rate": 7.053032178451769e-09,
+      "loss": 0.9359,
+      "step": 9283
+    },
+    {
+      "epoch": 0.9764023821105079,
+      "grad_norm": 2.5809673977523286,
+      "learning_rate": 6.990507047049677e-09,
+      "loss": 0.9662,
+      "step": 9284
+    },
+    {
+      "epoch": 0.9765075525523551,
+      "grad_norm": 2.6394164418653476,
+      "learning_rate": 6.928259905572277e-09,
+      "loss": 1.0017,
+      "step": 9285
+    },
+    {
+      "epoch": 0.9766127229942024,
+      "grad_norm": 2.486791493687076,
+      "learning_rate": 6.866290760960959e-09,
+      "loss": 1.0181,
+      "step": 9286
+    },
+    {
+      "epoch": 0.9767178934360498,
+      "grad_norm": 2.2338709539918225,
+      "learning_rate": 6.804599620125196e-09,
+      "loss": 0.9647,
+      "step": 9287
+    },
+    {
+      "epoch": 0.9768230638778971,
+      "grad_norm": 2.1691387566361082,
+      "learning_rate": 6.743186489944764e-09,
+      "loss": 0.9912,
+      "step": 9288
+    },
+    {
+      "epoch": 0.9769282343197444,
+      "grad_norm": 3.245572611882046,
+      "learning_rate": 6.68205137726724e-09,
+      "loss": 0.9692,
+      "step": 9289
+    },
+    {
+      "epoch": 0.9770334047615917,
+      "grad_norm": 3.04733299754532,
+      "learning_rate": 6.621194288909671e-09,
+      "loss": 0.9791,
+      "step": 9290
+    },
+    {
+      "epoch": 0.977138575203439,
+      "grad_norm": 1.8878390554596858,
+      "learning_rate": 6.560615231658296e-09,
+      "loss": 0.9996,
+      "step": 9291
+    },
+    {
+      "epoch": 0.9772437456452864,
+      "grad_norm": 2.2627442103553017,
+      "learning_rate": 6.500314212267989e-09,
+      "loss": 1.0112,
+      "step": 9292
+    },
+    {
+      "epoch": 0.9773489160871337,
+      "grad_norm": 2.218078785102874,
+      "learning_rate": 6.440291237462815e-09,
+      "loss": 0.9641,
+      "step": 9293
+    },
+    {
+      "epoch": 0.977454086528981,
+      "grad_norm": 1.8552740152450433,
+      "learning_rate": 6.380546313935754e-09,
+      "loss": 0.9732,
+      "step": 9294
+    },
+    {
+      "epoch": 0.9775592569708283,
+      "grad_norm": 2.4608837759222735,
+      "learning_rate": 6.321079448348977e-09,
+      "loss": 0.9696,
+      "step": 9295
+    },
+    {
+      "epoch": 0.9776644274126757,
+      "grad_norm": 2.552592842090153,
+      "learning_rate": 6.261890647333568e-09,
+      "loss": 0.9417,
+      "step": 9296
+    },
+    {
+      "epoch": 0.977769597854523,
+      "grad_norm": 2.009057625011869,
+      "learning_rate": 6.202979917489249e-09,
+      "loss": 0.9384,
+      "step": 9297
+    },
+    {
+      "epoch": 0.9778747682963703,
+      "grad_norm": 1.9808092212897492,
+      "learning_rate": 6.144347265384931e-09,
+      "loss": 0.9631,
+      "step": 9298
+    },
+    {
+      "epoch": 0.9779799387382176,
+      "grad_norm": 3.045529761172713,
+      "learning_rate": 6.085992697559273e-09,
+      "loss": 0.9743,
+      "step": 9299
+    },
+    {
+      "epoch": 0.978085109180065,
+      "grad_norm": 3.059783975457296,
+      "learning_rate": 6.027916220518459e-09,
+      "loss": 0.9701,
+      "step": 9300
+    },
+    {
+      "epoch": 0.9781902796219123,
+      "grad_norm": 3.0504326853392993,
+      "learning_rate": 5.970117840738976e-09,
+      "loss": 0.9683,
+      "step": 9301
+    },
+    {
+      "epoch": 0.9782954500637596,
+      "grad_norm": 2.7901710880220065,
+      "learning_rate": 5.912597564665667e-09,
+      "loss": 1.0016,
+      "step": 9302
+    },
+    {
+      "epoch": 0.9784006205056069,
+      "grad_norm": 2.4240633625968497,
+      "learning_rate": 5.85535539871257e-09,
+      "loss": 1.0195,
+      "step": 9303
+    },
+    {
+      "epoch": 0.9785057909474543,
+      "grad_norm": 2.4233089687632283,
+      "learning_rate": 5.798391349262356e-09,
+      "loss": 0.9825,
+      "step": 9304
+    },
+    {
+      "epoch": 0.9786109613893015,
+      "grad_norm": 1.8446993064170987,
+      "learning_rate": 5.741705422667443e-09,
+      "loss": 1.0112,
+      "step": 9305
+    },
+    {
+      "epoch": 0.9787161318311488,
+      "grad_norm": 2.2841403028774048,
+      "learning_rate": 5.685297625248054e-09,
+      "loss": 0.9792,
+      "step": 9306
+    },
+    {
+      "epoch": 0.9788213022729961,
+      "grad_norm": 2.755338437254536,
+      "learning_rate": 5.62916796329499e-09,
+      "loss": 0.9768,
+      "step": 9307
+    },
+    {
+      "epoch": 0.9789264727148435,
+      "grad_norm": 2.7700700654544432,
+      "learning_rate": 5.573316443066301e-09,
+      "loss": 0.9787,
+      "step": 9308
+    },
+    {
+      "epoch": 0.9790316431566908,
+      "grad_norm": 2.8359479362123223,
+      "learning_rate": 5.517743070790061e-09,
+      "loss": 0.9891,
+      "step": 9309
+    },
+    {
+      "epoch": 0.9791368135985381,
+      "grad_norm": 2.7613280805123583,
+      "learning_rate": 5.462447852663532e-09,
+      "loss": 0.9876,
+      "step": 9310
+    },
+    {
+      "epoch": 0.9792419840403854,
+      "grad_norm": 2.4322910758293825,
+      "learning_rate": 5.407430794852342e-09,
+      "loss": 1.0221,
+      "step": 9311
+    },
+    {
+      "epoch": 0.9793471544822328,
+      "grad_norm": 1.9160133184659012,
+      "learning_rate": 5.352691903491303e-09,
+      "loss": 1.0014,
+      "step": 9312
+    },
+    {
+      "epoch": 0.9794523249240801,
+      "grad_norm": 2.3481712518499207,
+      "learning_rate": 5.2982311846841436e-09,
+      "loss": 0.9383,
+      "step": 9313
+    },
+    {
+      "epoch": 0.9795574953659274,
+      "grad_norm": 2.588135744749517,
+      "learning_rate": 5.2440486445037855e-09,
+      "loss": 0.9673,
+      "step": 9314
+    },
+    {
+      "epoch": 0.9796626658077747,
+      "grad_norm": 3.1082126273430664,
+      "learning_rate": 5.190144288991783e-09,
+      "loss": 1.0052,
+      "step": 9315
+    },
+    {
+      "epoch": 0.979767836249622,
+      "grad_norm": 2.1785968088651773,
+      "learning_rate": 5.136518124159162e-09,
+      "loss": 0.9681,
+      "step": 9316
+    },
+    {
+      "epoch": 0.9798730066914694,
+      "grad_norm": 2.5729066271800622,
+      "learning_rate": 5.0831701559855835e-09,
+      "loss": 1.0088,
+      "step": 9317
+    },
+    {
+      "epoch": 0.9799781771333167,
+      "grad_norm": 1.9530621342015528,
+      "learning_rate": 5.030100390419623e-09,
+      "loss": 0.9577,
+      "step": 9318
+    },
+    {
+      "epoch": 0.980083347575164,
+      "grad_norm": 2.267823738676506,
+      "learning_rate": 4.977308833379324e-09,
+      "loss": 1.018,
+      "step": 9319
+    },
+    {
+      "epoch": 0.9801885180170113,
+      "grad_norm": 2.2816893245073584,
+      "learning_rate": 4.924795490750811e-09,
+      "loss": 0.9672,
+      "step": 9320
+    },
+    {
+      "epoch": 0.9802936884588587,
+      "grad_norm": 1.9259466407396693,
+      "learning_rate": 4.872560368390233e-09,
+      "loss": 0.9453,
+      "step": 9321
+    },
+    {
+      "epoch": 0.980398858900706,
+      "grad_norm": 1.7974805667836606,
+      "learning_rate": 4.820603472121821e-09,
+      "loss": 0.9947,
+      "step": 9322
+    },
+    {
+      "epoch": 0.9805040293425533,
+      "grad_norm": 2.3791236198376646,
+      "learning_rate": 4.768924807739273e-09,
+      "loss": 0.967,
+      "step": 9323
+    },
+    {
+      "epoch": 0.9806091997844006,
+      "grad_norm": 2.4419036752163743,
+      "learning_rate": 4.717524381005478e-09,
+      "loss": 0.9996,
+      "step": 9324
+    },
+    {
+      "epoch": 0.980714370226248,
+      "grad_norm": 2.7332285105011347,
+      "learning_rate": 4.666402197651687e-09,
+      "loss": 0.9636,
+      "step": 9325
+    },
+    {
+      "epoch": 0.9808195406680952,
+      "grad_norm": 1.8101579312807496,
+      "learning_rate": 4.615558263378617e-09,
+      "loss": 1.0082,
+      "step": 9326
+    },
+    {
+      "epoch": 0.9809247111099425,
+      "grad_norm": 3.2434727994808683,
+      "learning_rate": 4.5649925838553435e-09,
+      "loss": 1.0134,
+      "step": 9327
+    },
+    {
+      "epoch": 0.9810298815517898,
+      "grad_norm": 2.391788541909677,
+      "learning_rate": 4.5147051647206895e-09,
+      "loss": 1.0044,
+      "step": 9328
+    },
+    {
+      "epoch": 0.9811350519936372,
+      "grad_norm": 2.25620752561833,
+      "learning_rate": 4.464696011582115e-09,
+      "loss": 0.9879,
+      "step": 9329
+    },
+    {
+      "epoch": 0.9812402224354845,
+      "grad_norm": 2.4842392384829406,
+      "learning_rate": 4.414965130015991e-09,
+      "loss": 1.0062,
+      "step": 9330
+    },
+    {
+      "epoch": 0.9813453928773318,
+      "grad_norm": 2.2203661202851284,
+      "learning_rate": 4.365512525567605e-09,
+      "loss": 0.9546,
+      "step": 9331
+    },
+    {
+      "epoch": 0.9814505633191791,
+      "grad_norm": 2.711772293584609,
+      "learning_rate": 4.316338203751158e-09,
+      "loss": 0.9978,
+      "step": 9332
+    },
+    {
+      "epoch": 0.9815557337610264,
+      "grad_norm": 2.7700162537251076,
+      "learning_rate": 4.267442170050318e-09,
+      "loss": 0.9552,
+      "step": 9333
+    },
+    {
+      "epoch": 0.9816609042028738,
+      "grad_norm": 2.2577456211849176,
+      "learning_rate": 4.2188244299171126e-09,
+      "loss": 0.9859,
+      "step": 9334
+    },
+    {
+      "epoch": 0.9817660746447211,
+      "grad_norm": 2.5568997931283657,
+      "learning_rate": 4.170484988773038e-09,
+      "loss": 0.9332,
+      "step": 9335
+    },
+    {
+      "epoch": 0.9818712450865684,
+      "grad_norm": 1.9925306382045842,
+      "learning_rate": 4.122423852007951e-09,
+      "loss": 0.9782,
+      "step": 9336
+    },
+    {
+      "epoch": 0.9819764155284157,
+      "grad_norm": 2.921254419402326,
+      "learning_rate": 4.0746410249814515e-09,
+      "loss": 0.981,
+      "step": 9337
+    },
+    {
+      "epoch": 0.9820815859702631,
+      "grad_norm": 2.267912103577595,
+      "learning_rate": 4.027136513021501e-09,
+      "loss": 0.9974,
+      "step": 9338
+    },
+    {
+      "epoch": 0.9821867564121104,
+      "grad_norm": 4.541966161576452,
+      "learning_rate": 3.979910321425251e-09,
+      "loss": 1.0016,
+      "step": 9339
+    },
+    {
+      "epoch": 0.9822919268539577,
+      "grad_norm": 3.2761474483878206,
+      "learning_rate": 3.932962455458489e-09,
+      "loss": 0.9835,
+      "step": 9340
+    },
+    {
+      "epoch": 0.982397097295805,
+      "grad_norm": 2.228104089247583,
+      "learning_rate": 3.886292920356749e-09,
+      "loss": 1.0091,
+      "step": 9341
+    },
+    {
+      "epoch": 0.9825022677376524,
+      "grad_norm": 5.889922296145182,
+      "learning_rate": 3.839901721323925e-09,
+      "loss": 0.9941,
+      "step": 9342
+    },
+    {
+      "epoch": 0.9826074381794997,
+      "grad_norm": 1.9237694450938765,
+      "learning_rate": 3.793788863532822e-09,
+      "loss": 0.9619,
+      "step": 9343
+    },
+    {
+      "epoch": 0.982712608621347,
+      "grad_norm": 2.5296195092756477,
+      "learning_rate": 3.747954352125438e-09,
+      "loss": 0.958,
+      "step": 9344
+    },
+    {
+      "epoch": 0.9828177790631943,
+      "grad_norm": 2.864791152152932,
+      "learning_rate": 3.702398192212686e-09,
+      "loss": 0.9767,
+      "step": 9345
+    },
+    {
+      "epoch": 0.9829229495050416,
+      "grad_norm": 2.4375684665326727,
+      "learning_rate": 3.6571203888746685e-09,
+      "loss": 0.9733,
+      "step": 9346
+    },
+    {
+      "epoch": 0.9830281199468889,
+      "grad_norm": 2.0742623744580695,
+      "learning_rate": 3.6121209471595697e-09,
+      "loss": 1.0078,
+      "step": 9347
+    },
+    {
+      "epoch": 0.9831332903887362,
+      "grad_norm": 2.4063986427828614,
+      "learning_rate": 3.567399872085875e-09,
+      "loss": 0.998,
+      "step": 9348
+    },
+    {
+      "epoch": 0.9832384608305835,
+      "grad_norm": 2.383631755468102,
+      "learning_rate": 3.522957168640151e-09,
+      "loss": 0.9822,
+      "step": 9349
+    },
+    {
+      "epoch": 0.9833436312724309,
+      "grad_norm": 2.393452033459606,
+      "learning_rate": 3.4787928417778783e-09,
+      "loss": 0.9491,
+      "step": 9350
+    },
+    {
+      "epoch": 0.9834488017142782,
+      "grad_norm": 2.1318023663228094,
+      "learning_rate": 3.4349068964240063e-09,
+      "loss": 1.0221,
+      "step": 9351
+    },
+    {
+      "epoch": 0.9835539721561255,
+      "grad_norm": 2.126532018705832,
+      "learning_rate": 3.391299337471843e-09,
+      "loss": 1.0536,
+      "step": 9352
+    },
+    {
+      "epoch": 0.9836591425979728,
+      "grad_norm": 2.4227691688428643,
+      "learning_rate": 3.3479701697841647e-09,
+      "loss": 0.9667,
+      "step": 9353
+    },
+    {
+      "epoch": 0.9837643130398201,
+      "grad_norm": 2.472893734369122,
+      "learning_rate": 3.304919398192663e-09,
+      "loss": 0.961,
+      "step": 9354
+    },
+    {
+      "epoch": 0.9838694834816675,
+      "grad_norm": 2.646191353766477,
+      "learning_rate": 3.2621470274973867e-09,
+      "loss": 1.0272,
+      "step": 9355
+    },
+    {
+      "epoch": 0.9839746539235148,
+      "grad_norm": 3.2371826662759817,
+      "learning_rate": 3.219653062468131e-09,
+      "loss": 0.9586,
+      "step": 9356
+    },
+    {
+      "epoch": 0.9840798243653621,
+      "grad_norm": 2.6252562394371273,
+      "learning_rate": 3.1774375078433284e-09,
+      "loss": 1.0218,
+      "step": 9357
+    },
+    {
+      "epoch": 0.9841849948072094,
+      "grad_norm": 2.883626529975225,
+      "learning_rate": 3.1355003683303243e-09,
+      "loss": 1.0251,
+      "step": 9358
+    },
+    {
+      "epoch": 0.9842901652490568,
+      "grad_norm": 2.101061184091713,
+      "learning_rate": 3.0938416486051004e-09,
+      "loss": 0.9871,
+      "step": 9359
+    },
+    {
+      "epoch": 0.9843953356909041,
+      "grad_norm": 2.825052483693705,
+      "learning_rate": 3.0524613533133853e-09,
+      "loss": 1.0467,
+      "step": 9360
+    },
+    {
+      "epoch": 0.9845005061327514,
+      "grad_norm": 2.1917761304189898,
+      "learning_rate": 3.011359487068988e-09,
+      "loss": 0.9531,
+      "step": 9361
+    },
+    {
+      "epoch": 0.9846056765745987,
+      "grad_norm": 2.087827643664742,
+      "learning_rate": 2.970536054455464e-09,
+      "loss": 0.9927,
+      "step": 9362
+    },
+    {
+      "epoch": 0.9847108470164461,
+      "grad_norm": 2.849510889632073,
+      "learning_rate": 2.9299910600247285e-09,
+      "loss": 0.9576,
+      "step": 9363
+    },
+    {
+      "epoch": 0.9848160174582934,
+      "grad_norm": 2.3562660790190324,
+      "learning_rate": 2.8897245082978863e-09,
+      "loss": 0.9732,
+      "step": 9364
+    },
+    {
+      "epoch": 0.9849211879001407,
+      "grad_norm": 2.3002496067135803,
+      "learning_rate": 2.849736403764958e-09,
+      "loss": 0.9836,
+      "step": 9365
+    },
+    {
+      "epoch": 0.985026358341988,
+      "grad_norm": 1.726852714144181,
+      "learning_rate": 2.810026750885153e-09,
+      "loss": 0.9782,
+      "step": 9366
+    },
+    {
+      "epoch": 0.9851315287838353,
+      "grad_norm": 2.796815170845847,
+      "learning_rate": 2.770595554086042e-09,
+      "loss": 0.9365,
+      "step": 9367
+    },
+    {
+      "epoch": 0.9852366992256826,
+      "grad_norm": 2.6158136518772763,
+      "learning_rate": 2.7314428177646622e-09,
+      "loss": 0.9473,
+      "step": 9368
+    },
+    {
+      "epoch": 0.9853418696675299,
+      "grad_norm": 2.2921022351525555,
+      "learning_rate": 2.6925685462869664e-09,
+      "loss": 0.9774,
+      "step": 9369
+    },
+    {
+      "epoch": 0.9854470401093772,
+      "grad_norm": 1.8665257532281978,
+      "learning_rate": 2.6539727439875427e-09,
+      "loss": 0.9579,
+      "step": 9370
+    },
+    {
+      "epoch": 0.9855522105512246,
+      "grad_norm": 2.0803248457969206,
+      "learning_rate": 2.6156554151704484e-09,
+      "loss": 0.9765,
+      "step": 9371
+    },
+    {
+      "epoch": 0.9856573809930719,
+      "grad_norm": 3.028285191313156,
+      "learning_rate": 2.5776165641080986e-09,
+      "loss": 0.9999,
+      "step": 9372
+    },
+    {
+      "epoch": 0.9857625514349192,
+      "grad_norm": 3.576679364872881,
+      "learning_rate": 2.539856195041823e-09,
+      "loss": 0.977,
+      "step": 9373
+    },
+    {
+      "epoch": 0.9858677218767665,
+      "grad_norm": 2.4305778370700826,
+      "learning_rate": 2.5023743121826983e-09,
+      "loss": 1.0111,
+      "step": 9374
+    },
+    {
+      "epoch": 0.9859728923186138,
+      "grad_norm": 2.47884229852146,
+      "learning_rate": 2.465170919710158e-09,
+      "loss": 0.9983,
+      "step": 9375
+    },
+    {
+      "epoch": 0.9860780627604612,
+      "grad_norm": 2.7349159526364994,
+      "learning_rate": 2.428246021772551e-09,
+      "loss": 0.991,
+      "step": 9376
+    },
+    {
+      "epoch": 0.9861832332023085,
+      "grad_norm": 2.637560167507425,
+      "learning_rate": 2.391599622487417e-09,
+      "loss": 0.9667,
+      "step": 9377
+    },
+    {
+      "epoch": 0.9862884036441558,
+      "grad_norm": 1.9300449879076964,
+      "learning_rate": 2.3552317259409317e-09,
+      "loss": 1.0041,
+      "step": 9378
+    },
+    {
+      "epoch": 0.9863935740860031,
+      "grad_norm": 2.6765740954561363,
+      "learning_rate": 2.3191423361884626e-09,
+      "loss": 1.0011,
+      "step": 9379
+    },
+    {
+      "epoch": 0.9864987445278505,
+      "grad_norm": 2.616883579256923,
+      "learning_rate": 2.28333145725429e-09,
+      "loss": 1.0176,
+      "step": 9380
+    },
+    {
+      "epoch": 0.9866039149696978,
+      "grad_norm": 1.8573269651299802,
+      "learning_rate": 2.247799093131886e-09,
+      "loss": 0.969,
+      "step": 9381
+    },
+    {
+      "epoch": 0.9867090854115451,
+      "grad_norm": 2.60301904370189,
+      "learning_rate": 2.2125452477828047e-09,
+      "loss": 0.9769,
+      "step": 9382
+    },
+    {
+      "epoch": 0.9868142558533924,
+      "grad_norm": 1.9682961895311704,
+      "learning_rate": 2.1775699251383455e-09,
+      "loss": 0.9771,
+      "step": 9383
+    },
+    {
+      "epoch": 0.9869194262952398,
+      "grad_norm": 3.191223466664238,
+      "learning_rate": 2.1428731290987215e-09,
+      "loss": 1.046,
+      "step": 9384
+    },
+    {
+      "epoch": 0.9870245967370871,
+      "grad_norm": 2.892416445308266,
+      "learning_rate": 2.1084548635327828e-09,
+      "loss": 0.9734,
+      "step": 9385
+    },
+    {
+      "epoch": 0.9871297671789344,
+      "grad_norm": 3.095226990473517,
+      "learning_rate": 2.0743151322785703e-09,
+      "loss": 0.9719,
+      "step": 9386
+    },
+    {
+      "epoch": 0.9872349376207816,
+      "grad_norm": 2.7327344515972545,
+      "learning_rate": 2.0404539391427614e-09,
+      "loss": 0.9671,
+      "step": 9387
+    },
+    {
+      "epoch": 0.987340108062629,
+      "grad_norm": 2.9017278864296263,
+      "learning_rate": 2.0068712879009468e-09,
+      "loss": 1.0231,
+      "step": 9388
+    },
+    {
+      "epoch": 0.9874452785044763,
+      "grad_norm": 3.392301107777531,
+      "learning_rate": 1.973567182298186e-09,
+      "loss": 0.9928,
+      "step": 9389
+    },
+    {
+      "epoch": 0.9875504489463236,
+      "grad_norm": 2.3644487612910856,
+      "learning_rate": 1.9405416260481757e-09,
+      "loss": 0.9683,
+      "step": 9390
+    },
+    {
+      "epoch": 0.9876556193881709,
+      "grad_norm": 2.6559023017178194,
+      "learning_rate": 1.907794622833248e-09,
+      "loss": 0.9537,
+      "step": 9391
+    },
+    {
+      "epoch": 0.9877607898300182,
+      "grad_norm": 2.6251794536087494,
+      "learning_rate": 1.875326176304926e-09,
+      "loss": 1.0005,
+      "step": 9392
+    },
+    {
+      "epoch": 0.9878659602718656,
+      "grad_norm": 2.1131946840820253,
+      "learning_rate": 1.8431362900839244e-09,
+      "loss": 0.9866,
+      "step": 9393
+    },
+    {
+      "epoch": 0.9879711307137129,
+      "grad_norm": 3.060384152042341,
+      "learning_rate": 1.8112249677598726e-09,
+      "loss": 1.0004,
+      "step": 9394
+    },
+    {
+      "epoch": 0.9880763011555602,
+      "grad_norm": 2.7092939576558366,
+      "learning_rate": 1.7795922128904796e-09,
+      "loss": 0.9733,
+      "step": 9395
+    },
+    {
+      "epoch": 0.9881814715974075,
+      "grad_norm": 3.15026244560097,
+      "learning_rate": 1.7482380290034795e-09,
+      "loss": 0.9553,
+      "step": 9396
+    },
+    {
+      "epoch": 0.9882866420392549,
+      "grad_norm": 2.3143889813143637,
+      "learning_rate": 1.7171624195952418e-09,
+      "loss": 0.9775,
+      "step": 9397
+    },
+    {
+      "epoch": 0.9883918124811022,
+      "grad_norm": 1.749676378937021,
+      "learning_rate": 1.6863653881307728e-09,
+      "loss": 1.0136,
+      "step": 9398
+    },
+    {
+      "epoch": 0.9884969829229495,
+      "grad_norm": 2.4561058588577636,
+      "learning_rate": 1.6558469380439923e-09,
+      "loss": 0.9778,
+      "step": 9399
+    },
+    {
+      "epoch": 0.9886021533647968,
+      "grad_norm": 2.869598988661229,
+      "learning_rate": 1.6256070727380114e-09,
+      "loss": 0.9961,
+      "step": 9400
+    },
+    {
+      "epoch": 0.9887073238066442,
+      "grad_norm": 1.3519472395346488,
+      "learning_rate": 1.5956457955848547e-09,
+      "loss": 0.9565,
+      "step": 9401
+    },
+    {
+      "epoch": 0.9888124942484915,
+      "grad_norm": 2.550759243468463,
+      "learning_rate": 1.5659631099257389e-09,
+      "loss": 0.9986,
+      "step": 9402
+    },
+    {
+      "epoch": 0.9889176646903388,
+      "grad_norm": 2.2780299064704668,
+      "learning_rate": 1.5365590190699608e-09,
+      "loss": 0.9818,
+      "step": 9403
+    },
+    {
+      "epoch": 0.9890228351321861,
+      "grad_norm": 2.5188559087529003,
+      "learning_rate": 1.507433526296842e-09,
+      "loss": 0.9911,
+      "step": 9404
+    },
+    {
+      "epoch": 0.9891280055740335,
+      "grad_norm": 2.188706310697837,
+      "learning_rate": 1.4785866348537848e-09,
+      "loss": 0.9693,
+      "step": 9405
+    },
+    {
+      "epoch": 0.9892331760158808,
+      "grad_norm": 1.809924536593071,
+      "learning_rate": 1.4500183479573825e-09,
+      "loss": 0.9738,
+      "step": 9406
+    },
+    {
+      "epoch": 0.989338346457728,
+      "grad_norm": 3.093352924429147,
+      "learning_rate": 1.4217286687936982e-09,
+      "loss": 1.0348,
+      "step": 9407
+    },
+    {
+      "epoch": 0.9894435168995753,
+      "grad_norm": 2.7413192203962407,
+      "learning_rate": 1.3937176005165975e-09,
+      "loss": 1.0094,
+      "step": 9408
+    },
+    {
+      "epoch": 0.9895486873414227,
+      "grad_norm": 1.9967362712782808,
+      "learning_rate": 1.3659851462499708e-09,
+      "loss": 0.9604,
+      "step": 9409
+    },
+    {
+      "epoch": 0.98965385778327,
+      "grad_norm": 3.023690617394417,
+      "learning_rate": 1.3385313090857888e-09,
+      "loss": 0.9499,
+      "step": 9410
+    },
+    {
+      "epoch": 0.9897590282251173,
+      "grad_norm": 3.0257951189785106,
+      "learning_rate": 1.3113560920860468e-09,
+      "loss": 0.9351,
+      "step": 9411
+    },
+    {
+      "epoch": 0.9898641986669646,
+      "grad_norm": 2.413833812119721,
+      "learning_rate": 1.284459498280266e-09,
+      "loss": 1.0105,
+      "step": 9412
+    },
+    {
+      "epoch": 0.989969369108812,
+      "grad_norm": 2.037001468545561,
+      "learning_rate": 1.257841530668269e-09,
+      "loss": 0.9782,
+      "step": 9413
+    },
+    {
+      "epoch": 0.9900745395506593,
+      "grad_norm": 1.6511342738858803,
+      "learning_rate": 1.2315021922176817e-09,
+      "loss": 0.9412,
+      "step": 9414
+    },
+    {
+      "epoch": 0.9901797099925066,
+      "grad_norm": 2.8902872997925804,
+      "learning_rate": 1.2054414858655995e-09,
+      "loss": 0.9708,
+      "step": 9415
+    },
+    {
+      "epoch": 0.9902848804343539,
+      "grad_norm": 4.027061186967938,
+      "learning_rate": 1.1796594145183081e-09,
+      "loss": 1.0044,
+      "step": 9416
+    },
+    {
+      "epoch": 0.9903900508762012,
+      "grad_norm": 2.0376936451035483,
+      "learning_rate": 1.1541559810504532e-09,
+      "loss": 0.9808,
+      "step": 9417
+    },
+    {
+      "epoch": 0.9904952213180486,
+      "grad_norm": 2.5382357729640037,
+      "learning_rate": 1.1289311883058707e-09,
+      "loss": 0.9738,
+      "step": 9418
+    },
+    {
+      "epoch": 0.9906003917598959,
+      "grad_norm": 1.7785026493838247,
+      "learning_rate": 1.1039850390973107e-09,
+      "loss": 0.9819,
+      "step": 9419
+    },
+    {
+      "epoch": 0.9907055622017432,
+      "grad_norm": 2.6469976920281453,
+      "learning_rate": 1.0793175362067143e-09,
+      "loss": 0.9929,
+      "step": 9420
+    },
+    {
+      "epoch": 0.9908107326435905,
+      "grad_norm": 2.0316413833430387,
+      "learning_rate": 1.054928682384382e-09,
+      "loss": 0.9638,
+      "step": 9421
+    },
+    {
+      "epoch": 0.9909159030854379,
+      "grad_norm": 2.6241976883897,
+      "learning_rate": 1.0308184803498046e-09,
+      "loss": 0.9819,
+      "step": 9422
+    },
+    {
+      "epoch": 0.9910210735272852,
+      "grad_norm": 2.229629028463766,
+      "learning_rate": 1.0069869327919424e-09,
+      "loss": 0.9947,
+      "step": 9423
+    },
+    {
+      "epoch": 0.9911262439691325,
+      "grad_norm": 2.788186432126167,
+      "learning_rate": 9.834340423678368e-10,
+      "loss": 0.9522,
+      "step": 9424
+    },
+    {
+      "epoch": 0.9912314144109798,
+      "grad_norm": 2.698863781368917,
+      "learning_rate": 9.601598117037202e-10,
+      "loss": 0.9553,
+      "step": 9425
+    },
+    {
+      "epoch": 0.9913365848528272,
+      "grad_norm": 2.3066739529849736,
+      "learning_rate": 9.371642433952943e-10,
+      "loss": 0.9726,
+      "step": 9426
+    },
+    {
+      "epoch": 0.9914417552946745,
+      "grad_norm": 2.7776673360748387,
+      "learning_rate": 9.144473400060639e-10,
+      "loss": 1.005,
+      "step": 9427
+    },
+    {
+      "epoch": 0.9915469257365217,
+      "grad_norm": 2.033776373402885,
+      "learning_rate": 8.920091040698353e-10,
+      "loss": 0.9883,
+      "step": 9428
+    },
+    {
+      "epoch": 0.991652096178369,
+      "grad_norm": 2.9626974448308045,
+      "learning_rate": 8.698495380882188e-10,
+      "loss": 1.0091,
+      "step": 9429
+    },
+    {
+      "epoch": 0.9917572666202163,
+      "grad_norm": 2.412674687043095,
+      "learning_rate": 8.479686445322932e-10,
+      "loss": 0.9696,
+      "step": 9430
+    },
+    {
+      "epoch": 0.9918624370620637,
+      "grad_norm": 2.280018692073066,
+      "learning_rate": 8.263664258420512e-10,
+      "loss": 0.952,
+      "step": 9431
+    },
+    {
+      "epoch": 0.991967607503911,
+      "grad_norm": 2.536989287684638,
+      "learning_rate": 8.050428844261215e-10,
+      "loss": 1.0399,
+      "step": 9432
+    },
+    {
+      "epoch": 0.9920727779457583,
+      "grad_norm": 2.4868353766120803,
+      "learning_rate": 7.839980226623245e-10,
+      "loss": 0.9822,
+      "step": 9433
+    },
+    {
+      "epoch": 0.9921779483876056,
+      "grad_norm": 2.2345599750935303,
+      "learning_rate": 7.632318428976715e-10,
+      "loss": 0.9763,
+      "step": 9434
+    },
+    {
+      "epoch": 0.992283118829453,
+      "grad_norm": 2.7320164282700006,
+      "learning_rate": 7.427443474469776e-10,
+      "loss": 0.9614,
+      "step": 9435
+    },
+    {
+      "epoch": 0.9923882892713003,
+      "grad_norm": 2.693307353444875,
+      "learning_rate": 7.225355385953592e-10,
+      "loss": 0.9834,
+      "step": 9436
+    },
+    {
+      "epoch": 0.9924934597131476,
+      "grad_norm": 2.7375269267620266,
+      "learning_rate": 7.02605418596014e-10,
+      "loss": 1.0151,
+      "step": 9437
+    },
+    {
+      "epoch": 0.9925986301549949,
+      "grad_norm": 1.9154701081163854,
+      "learning_rate": 6.829539896716087e-10,
+      "loss": 0.9614,
+      "step": 9438
+    },
+    {
+      "epoch": 0.9927038005968423,
+      "grad_norm": 2.7318506358078727,
+      "learning_rate": 6.635812540131681e-10,
+      "loss": 0.9716,
+      "step": 9439
+    },
+    {
+      "epoch": 0.9928089710386896,
+      "grad_norm": 2.35587299097919,
+      "learning_rate": 6.444872137806313e-10,
+      "loss": 0.9932,
+      "step": 9440
+    },
+    {
+      "epoch": 0.9929141414805369,
+      "grad_norm": 2.7986557506186425,
+      "learning_rate": 6.256718711036835e-10,
+      "loss": 0.9715,
+      "step": 9441
+    },
+    {
+      "epoch": 0.9930193119223842,
+      "grad_norm": 2.958571293660485,
+      "learning_rate": 6.071352280800912e-10,
+      "loss": 0.988,
+      "step": 9442
+    },
+    {
+      "epoch": 0.9931244823642316,
+      "grad_norm": 3.0504234064777522,
+      "learning_rate": 5.888772867770897e-10,
+      "loss": 0.9765,
+      "step": 9443
+    },
+    {
+      "epoch": 0.9932296528060789,
+      "grad_norm": 2.2184835377963266,
+      "learning_rate": 5.708980492302729e-10,
+      "loss": 0.9987,
+      "step": 9444
+    },
+    {
+      "epoch": 0.9933348232479262,
+      "grad_norm": 1.9829209945604436,
+      "learning_rate": 5.531975174444259e-10,
+      "loss": 0.9188,
+      "step": 9445
+    },
+    {
+      "epoch": 0.9934399936897735,
+      "grad_norm": 2.833298967270071,
+      "learning_rate": 5.357756933935254e-10,
+      "loss": 1.0,
+      "step": 9446
+    },
+    {
+      "epoch": 0.9935451641316209,
+      "grad_norm": 2.299238400231606,
+      "learning_rate": 5.186325790199065e-10,
+      "loss": 0.9838,
+      "step": 9447
+    },
+    {
+      "epoch": 0.9936503345734681,
+      "grad_norm": 2.1636055939503707,
+      "learning_rate": 5.017681762356507e-10,
+      "loss": 0.9973,
+      "step": 9448
+    },
+    {
+      "epoch": 0.9937555050153154,
+      "grad_norm": 2.607403264070489,
+      "learning_rate": 4.851824869211985e-10,
+      "loss": 1.0123,
+      "step": 9449
+    },
+    {
+      "epoch": 0.9938606754571627,
+      "grad_norm": 2.921227734664202,
+      "learning_rate": 4.688755129256262e-10,
+      "loss": 1.0149,
+      "step": 9450
+    },
+    {
+      "epoch": 0.99396584589901,
+      "grad_norm": 2.106489666904462,
+      "learning_rate": 4.5284725606747903e-10,
+      "loss": 0.9528,
+      "step": 9451
+    },
+    {
+      "epoch": 0.9940710163408574,
+      "grad_norm": 2.2948653311351146,
+      "learning_rate": 4.3709771813393864e-10,
+      "loss": 0.9819,
+      "step": 9452
+    },
+    {
+      "epoch": 0.9941761867827047,
+      "grad_norm": 2.5397350328464,
+      "learning_rate": 4.216269008813778e-10,
+      "loss": 1.0424,
+      "step": 9453
+    },
+    {
+      "epoch": 0.994281357224552,
+      "grad_norm": 2.6859348273107098,
+      "learning_rate": 4.0643480603480554e-10,
+      "loss": 0.9152,
+      "step": 9454
+    },
+    {
+      "epoch": 0.9943865276663993,
+      "grad_norm": 2.305716938487438,
+      "learning_rate": 3.9152143528842226e-10,
+      "loss": 0.9636,
+      "step": 9455
+    },
+    {
+      "epoch": 0.9944916981082467,
+      "grad_norm": 2.317768982172829,
+      "learning_rate": 3.768867903047868e-10,
+      "loss": 0.9895,
+      "step": 9456
+    },
+    {
+      "epoch": 0.994596868550094,
+      "grad_norm": 1.9307647936153862,
+      "learning_rate": 3.625308727162047e-10,
+      "loss": 0.9515,
+      "step": 9457
+    },
+    {
+      "epoch": 0.9947020389919413,
+      "grad_norm": 2.9486289730118234,
+      "learning_rate": 3.4845368412306235e-10,
+      "loss": 0.9608,
+      "step": 9458
+    },
+    {
+      "epoch": 0.9948072094337886,
+      "grad_norm": 2.821432702813321,
+      "learning_rate": 3.346552260954927e-10,
+      "loss": 1.0182,
+      "step": 9459
+    },
+    {
+      "epoch": 0.994912379875636,
+      "grad_norm": 2.0682478422118526,
+      "learning_rate": 3.2113550017198734e-10,
+      "loss": 0.9518,
+      "step": 9460
+    },
+    {
+      "epoch": 0.9950175503174833,
+      "grad_norm": 2.4469821699306205,
+      "learning_rate": 3.0789450785995155e-10,
+      "loss": 1.0303,
+      "step": 9461
+    },
+    {
+      "epoch": 0.9951227207593306,
+      "grad_norm": 2.318744276916468,
+      "learning_rate": 2.94932250635982e-10,
+      "loss": 0.9428,
+      "step": 9462
+    },
+    {
+      "epoch": 0.9952278912011779,
+      "grad_norm": 1.9099911442636568,
+      "learning_rate": 2.8224872994558895e-10,
+      "loss": 0.9649,
+      "step": 9463
+    },
+    {
+      "epoch": 0.9953330616430253,
+      "grad_norm": 2.244161598411751,
+      "learning_rate": 2.6984394720291904e-10,
+      "loss": 0.9282,
+      "step": 9464
+    },
+    {
+      "epoch": 0.9954382320848726,
+      "grad_norm": 2.0228783659285217,
+      "learning_rate": 2.5771790379103267e-10,
+      "loss": 1.0379,
+      "step": 9465
+    },
+    {
+      "epoch": 0.9955434025267199,
+      "grad_norm": 2.7663617676626044,
+      "learning_rate": 2.45870601062459e-10,
+      "loss": 0.9732,
+      "step": 9466
+    },
+    {
+      "epoch": 0.9956485729685672,
+      "grad_norm": 2.783584442780796,
+      "learning_rate": 2.3430204033808577e-10,
+      "loss": 0.981,
+      "step": 9467
+    },
+    {
+      "epoch": 0.9957537434104144,
+      "grad_norm": 2.2784586979725168,
+      "learning_rate": 2.2301222290771473e-10,
+      "loss": 0.9731,
+      "step": 9468
+    },
+    {
+      "epoch": 0.9958589138522618,
+      "grad_norm": 2.474615759545288,
+      "learning_rate": 2.1200115003061628e-10,
+      "loss": 0.9718,
+      "step": 9469
+    },
+    {
+      "epoch": 0.9959640842941091,
+      "grad_norm": 2.7375505869626546,
+      "learning_rate": 2.012688229344195e-10,
+      "loss": 0.9472,
+      "step": 9470
+    },
+    {
+      "epoch": 0.9960692547359564,
+      "grad_norm": 3.181469509796547,
+      "learning_rate": 1.9081524281566732e-10,
+      "loss": 0.9785,
+      "step": 9471
+    },
+    {
+      "epoch": 0.9961744251778037,
+      "grad_norm": 3.8615566778110924,
+      "learning_rate": 1.8064041084037143e-10,
+      "loss": 0.9676,
+      "step": 9472
+    },
+    {
+      "epoch": 0.9962795956196511,
+      "grad_norm": 3.381624272619375,
+      "learning_rate": 1.707443281429022e-10,
+      "loss": 1.0194,
+      "step": 9473
+    },
+    {
+      "epoch": 0.9963847660614984,
+      "grad_norm": 3.1413976288468852,
+      "learning_rate": 1.6112699582682135e-10,
+      "loss": 1.0142,
+      "step": 9474
+    },
+    {
+      "epoch": 0.9964899365033457,
+      "grad_norm": 2.6439768248020177,
+      "learning_rate": 1.517884149646043e-10,
+      "loss": 0.9825,
+      "step": 9475
+    },
+    {
+      "epoch": 0.996595106945193,
+      "grad_norm": 2.47459179696731,
+      "learning_rate": 1.4272858659736265e-10,
+      "loss": 0.9735,
+      "step": 9476
+    },
+    {
+      "epoch": 0.9967002773870404,
+      "grad_norm": 1.6572520116151759,
+      "learning_rate": 1.33947511735677e-10,
+      "loss": 0.967,
+      "step": 9477
+    },
+    {
+      "epoch": 0.9968054478288877,
+      "grad_norm": 2.1650133863954624,
+      "learning_rate": 1.2544519135820887e-10,
+      "loss": 0.9935,
+      "step": 9478
+    },
+    {
+      "epoch": 0.996910618270735,
+      "grad_norm": 2.6395311326751396,
+      "learning_rate": 1.1722162641336633e-10,
+      "loss": 0.9942,
+      "step": 9479
+    },
+    {
+      "epoch": 0.9970157887125823,
+      "grad_norm": 2.3369728830368817,
+      "learning_rate": 1.0927681781819354e-10,
+      "loss": 0.9496,
+      "step": 9480
+    },
+    {
+      "epoch": 0.9971209591544297,
+      "grad_norm": 2.9099781993657454,
+      "learning_rate": 1.0161076645837098e-10,
+      "loss": 0.9617,
+      "step": 9481
+    },
+    {
+      "epoch": 0.997226129596277,
+      "grad_norm": 1.9322457713584655,
+      "learning_rate": 9.422347318877034e-11,
+      "loss": 1.0036,
+      "step": 9482
+    },
+    {
+      "epoch": 0.9973313000381243,
+      "grad_norm": 3.2635257365367516,
+      "learning_rate": 8.711493883317712e-11,
+      "loss": 0.9886,
+      "step": 9483
+    },
+    {
+      "epoch": 0.9974364704799716,
+      "grad_norm": 2.220418457360787,
+      "learning_rate": 8.028516418456811e-11,
+      "loss": 0.9676,
+      "step": 9484
+    },
+    {
+      "epoch": 0.997541640921819,
+      "grad_norm": 2.3115695199846384,
+      "learning_rate": 7.37341500040012e-11,
+      "loss": 0.9517,
+      "step": 9485
+    },
+    {
+      "epoch": 0.9976468113636663,
+      "grad_norm": 2.013983684435446,
+      "learning_rate": 6.746189702228067e-11,
+      "loss": 0.9492,
+      "step": 9486
+    },
+    {
+      "epoch": 0.9977519818055136,
+      "grad_norm": 2.816746556589395,
+      "learning_rate": 6.14684059385695e-11,
+      "loss": 0.9404,
+      "step": 9487
+    },
+    {
+      "epoch": 0.9978571522473609,
+      "grad_norm": 2.0503969665852066,
+      "learning_rate": 5.575367742149951e-11,
+      "loss": 0.9634,
+      "step": 9488
+    },
+    {
+      "epoch": 0.9979623226892081,
+      "grad_norm": 1.724401911640861,
+      "learning_rate": 5.0317712108061135e-11,
+      "loss": 0.9739,
+      "step": 9489
+    },
+    {
+      "epoch": 0.9980674931310555,
+      "grad_norm": 2.1346980648205705,
+      "learning_rate": 4.5160510604436156e-11,
+      "loss": 0.9721,
+      "step": 9490
+    },
+    {
+      "epoch": 0.9981726635729028,
+      "grad_norm": 2.090678339510078,
+      "learning_rate": 4.028207348572011e-11,
+      "loss": 0.9147,
+      "step": 9491
+    },
+    {
+      "epoch": 0.9982778340147501,
+      "grad_norm": 2.392059438797523,
+      "learning_rate": 3.568240129619982e-11,
+      "loss": 1.0114,
+      "step": 9492
+    },
+    {
+      "epoch": 0.9983830044565974,
+      "grad_norm": 2.506048349048106,
+      "learning_rate": 3.1361494548243224e-11,
+      "loss": 0.9004,
+      "step": 9493
+    },
+    {
+      "epoch": 0.9984881748984448,
+      "grad_norm": 2.7270228934874545,
+      "learning_rate": 2.7319353723964657e-11,
+      "loss": 0.9964,
+      "step": 9494
+    },
+    {
+      "epoch": 0.9985933453402921,
+      "grad_norm": 2.3101031627179864,
+      "learning_rate": 2.355597927411468e-11,
+      "loss": 0.9873,
+      "step": 9495
+    },
+    {
+      "epoch": 0.9986985157821394,
+      "grad_norm": 1.8549205557474275,
+      "learning_rate": 2.0071371618080037e-11,
+      "loss": 1.0167,
+      "step": 9496
+    },
+    {
+      "epoch": 0.9988036862239867,
+      "grad_norm": 2.4247474783252194,
+      "learning_rate": 1.686553114471634e-11,
+      "loss": 1.0092,
+      "step": 9497
+    },
+    {
+      "epoch": 0.9989088566658341,
+      "grad_norm": 2.5711740916551227,
+      "learning_rate": 1.3938458211515404e-11,
+      "loss": 0.996,
+      "step": 9498
+    },
+    {
+      "epoch": 0.9990140271076814,
+      "grad_norm": 2.6142784134493415,
+      "learning_rate": 1.1290153144605243e-11,
+      "loss": 0.936,
+      "step": 9499
+    },
+    {
+      "epoch": 0.9991191975495287,
+      "grad_norm": 2.4638656000993873,
+      "learning_rate": 8.920616239582735e-12,
+      "loss": 0.9754,
+      "step": 9500
+    },
+    {
+      "epoch": 0.999224367991376,
+      "grad_norm": 2.5496610952882843,
+      "learning_rate": 6.829847760403408e-12,
+      "loss": 0.9836,
+      "step": 9501
+    },
+    {
+      "epoch": 0.9993295384332234,
+      "grad_norm": 2.000881646427515,
+      "learning_rate": 5.017847940214093e-12,
+      "loss": 0.988,
+      "step": 9502
+    },
+    {
+      "epoch": 0.9994347088750707,
+      "grad_norm": 2.4163049388267765,
+      "learning_rate": 3.4846169813529394e-12,
+      "loss": 0.9964,
+      "step": 9503
+    },
+    {
+      "epoch": 0.999539879316918,
+      "grad_norm": 2.5790803945058864,
+      "learning_rate": 2.230155054516736e-12,
+      "loss": 0.9758,
+      "step": 9504
+    },
+    {
+      "epoch": 0.9996450497587653,
+      "grad_norm": 2.555299051963278,
+      "learning_rate": 1.2544622998711398e-12,
+      "loss": 1.0216,
+      "step": 9505
+    },
+    {
+      "epoch": 0.9997502202006127,
+      "grad_norm": 2.506038816699829,
+      "learning_rate": 5.575388259404512e-13,
+      "loss": 0.9635,
+      "step": 9506
+    },
+    {
+      "epoch": 0.99985539064246,
+      "grad_norm": 2.9286325043640606,
+      "learning_rate": 1.3938471044028234e-13,
+      "loss": 0.9874,
+      "step": 9507
+    },
+    {
+      "epoch": 0.9999605610843073,
+      "grad_norm": 1.8297745470230433,
+      "learning_rate": 0.0,
+      "loss": 0.948,
+      "step": 9508
+    },
+    {
+      "epoch": 0.9999605610843073,
+      "step": 9508,
+      "total_flos": 7610494285643776.0,
+      "train_loss": 1.0062551115383658,
+      "train_runtime": 85567.0625,
+      "train_samples_per_second": 7.112,
+      "train_steps_per_second": 0.111
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 9508,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 7610494285643776.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}