diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,9492 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.995008319467554,
+  "eval_steps": 500,
+  "global_step": 1350,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0022185246810870773,
+      "grad_norm": 1.2858673334121704,
+      "learning_rate": 2.222222222222222e-06,
+      "loss": 10.6426,
+      "step": 1
+    },
+    {
+      "epoch": 0.004437049362174155,
+      "grad_norm": 1.3133524656295776,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 10.6201,
+      "step": 2
+    },
+    {
+      "epoch": 0.0066555740432612314,
+      "grad_norm": 1.3102678060531616,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 10.8057,
+      "step": 3
+    },
+    {
+      "epoch": 0.00887409872434831,
+      "grad_norm": 1.2896019220352173,
+      "learning_rate": 8.888888888888888e-06,
+      "loss": 10.5284,
+      "step": 4
+    },
+    {
+      "epoch": 0.011092623405435386,
+      "grad_norm": 1.2783793210983276,
+      "learning_rate": 1.111111111111111e-05,
+      "loss": 10.6651,
+      "step": 5
+    },
+    {
+      "epoch": 0.013311148086522463,
+      "grad_norm": 1.3131320476531982,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 10.6014,
+      "step": 6
+    },
+    {
+      "epoch": 0.01552967276760954,
+      "grad_norm": 1.300071120262146,
+      "learning_rate": 1.5555555555555555e-05,
+      "loss": 10.7266,
+      "step": 7
+    },
+    {
+      "epoch": 0.01774819744869662,
+      "grad_norm": 1.2465472221374512,
+      "learning_rate": 1.7777777777777777e-05,
+      "loss": 10.7058,
+      "step": 8
+    },
+    {
+      "epoch": 0.019966722129783693,
+      "grad_norm": 1.3223122358322144,
+      "learning_rate": 1.9999999999999998e-05,
+      "loss": 10.5937,
+      "step": 9
+    },
+    {
+      "epoch": 0.022185246810870772,
+      "grad_norm": 1.2338709831237793,
+      "learning_rate": 2.222222222222222e-05,
+      "loss": 10.5325,
+      "step": 10
+    },
+    {
+      "epoch": 0.024403771491957847,
+      "grad_norm": 1.3407450914382935,
+      "learning_rate": 2.4444444444444445e-05,
+      "loss": 10.8228,
+      "step": 11
+    },
+    {
+      "epoch": 0.026622296173044926,
+      "grad_norm": 1.2660531997680664,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 10.4437,
+      "step": 12
+    },
+    {
+      "epoch": 0.028840820854132,
+      "grad_norm": 1.2972372770309448,
+      "learning_rate": 2.8888888888888885e-05,
+      "loss": 10.6301,
+      "step": 13
+    },
+    {
+      "epoch": 0.03105934553521908,
+      "grad_norm": 1.2349822521209717,
+      "learning_rate": 3.111111111111111e-05,
+      "loss": 10.4986,
+      "step": 14
+    },
+    {
+      "epoch": 0.033277870216306155,
+      "grad_norm": 1.2688781023025513,
+      "learning_rate": 3.333333333333333e-05,
+      "loss": 10.8125,
+      "step": 15
+    },
+    {
+      "epoch": 0.03549639489739324,
+      "grad_norm": 1.2572486400604248,
+      "learning_rate": 3.555555555555555e-05,
+      "loss": 10.7632,
+      "step": 16
+    },
+    {
+      "epoch": 0.03771491957848031,
+      "grad_norm": 1.2791802883148193,
+      "learning_rate": 3.777777777777777e-05,
+      "loss": 10.7394,
+      "step": 17
+    },
+    {
+      "epoch": 0.03993344425956739,
+      "grad_norm": 1.2469416856765747,
+      "learning_rate": 3.9999999999999996e-05,
+      "loss": 10.5939,
+      "step": 18
+    },
+    {
+      "epoch": 0.04215196894065446,
+      "grad_norm": 1.2112987041473389,
+      "learning_rate": 4.222222222222222e-05,
+      "loss": 10.4755,
+      "step": 19
+    },
+    {
+      "epoch": 0.044370493621741544,
+      "grad_norm": 1.2535514831542969,
+      "learning_rate": 4.444444444444444e-05,
+      "loss": 10.6218,
+      "step": 20
+    },
+    {
+      "epoch": 0.04658901830282862,
+      "grad_norm": 1.218755841255188,
+      "learning_rate": 4.6666666666666665e-05,
+      "loss": 10.4233,
+      "step": 21
+    },
+    {
+      "epoch": 0.048807542983915694,
+      "grad_norm": 1.2221254110336304,
+      "learning_rate": 4.888888888888889e-05,
+      "loss": 10.4401,
+      "step": 22
+    },
+    {
+      "epoch": 0.051026067665002776,
+      "grad_norm": 1.150526523590088,
+      "learning_rate": 5.111111111111111e-05,
+      "loss": 10.3223,
+      "step": 23
+    },
+    {
+      "epoch": 0.05324459234608985,
+      "grad_norm": 1.1968152523040771,
+      "learning_rate": 5.333333333333333e-05,
+      "loss": 10.4616,
+      "step": 24
+    },
+    {
+      "epoch": 0.05546311702717693,
+      "grad_norm": 1.2215632200241089,
+      "learning_rate": 5.5555555555555545e-05,
+      "loss": 10.4273,
+      "step": 25
+    },
+    {
+      "epoch": 0.057681641708264,
+      "grad_norm": 1.2177790403366089,
+      "learning_rate": 5.777777777777777e-05,
+      "loss": 10.5054,
+      "step": 26
+    },
+    {
+      "epoch": 0.059900166389351084,
+      "grad_norm": 1.2023818492889404,
+      "learning_rate": 5.9999999999999995e-05,
+      "loss": 10.469,
+      "step": 27
+    },
+    {
+      "epoch": 0.06211869107043816,
+      "grad_norm": 1.2354178428649902,
+      "learning_rate": 6.222222222222222e-05,
+      "loss": 10.5258,
+      "step": 28
+    },
+    {
+      "epoch": 0.06433721575152523,
+      "grad_norm": 1.1676080226898193,
+      "learning_rate": 6.444444444444444e-05,
+      "loss": 10.6443,
+      "step": 29
+    },
+    {
+      "epoch": 0.06655574043261231,
+      "grad_norm": 1.174275279045105,
+      "learning_rate": 6.666666666666666e-05,
+      "loss": 10.4425,
+      "step": 30
+    },
+    {
+      "epoch": 0.06877426511369938,
+      "grad_norm": 1.1797528266906738,
+      "learning_rate": 6.888888888888888e-05,
+      "loss": 10.3409,
+      "step": 31
+    },
+    {
+      "epoch": 0.07099278979478647,
+      "grad_norm": 1.1630054712295532,
+      "learning_rate": 7.11111111111111e-05,
+      "loss": 10.2665,
+      "step": 32
+    },
+    {
+      "epoch": 0.07321131447587355,
+      "grad_norm": 1.166619062423706,
+      "learning_rate": 7.333333333333332e-05,
+      "loss": 10.4629,
+      "step": 33
+    },
+    {
+      "epoch": 0.07542983915696062,
+      "grad_norm": 1.1770161390304565,
+      "learning_rate": 7.555555555555554e-05,
+      "loss": 10.5795,
+      "step": 34
+    },
+    {
+      "epoch": 0.0776483638380477,
+      "grad_norm": 1.1769590377807617,
+      "learning_rate": 7.777777777777777e-05,
+      "loss": 10.0589,
+      "step": 35
+    },
+    {
+      "epoch": 0.07986688851913477,
+      "grad_norm": 1.1110798120498657,
+      "learning_rate": 7.999999999999999e-05,
+      "loss": 10.2854,
+      "step": 36
+    },
+    {
+      "epoch": 0.08208541320022185,
+      "grad_norm": 1.0831031799316406,
+      "learning_rate": 8.222222222222222e-05,
+      "loss": 10.3889,
+      "step": 37
+    },
+    {
+      "epoch": 0.08430393788130892,
+      "grad_norm": 1.1288526058197021,
+      "learning_rate": 8.444444444444444e-05,
+      "loss": 10.4924,
+      "step": 38
+    },
+    {
+      "epoch": 0.08652246256239601,
+      "grad_norm": 1.1578062772750854,
+      "learning_rate": 8.666666666666665e-05,
+      "loss": 10.3146,
+      "step": 39
+    },
+    {
+      "epoch": 0.08874098724348309,
+      "grad_norm": 1.119649887084961,
+      "learning_rate": 8.888888888888888e-05,
+      "loss": 10.3406,
+      "step": 40
+    },
+    {
+      "epoch": 0.09095951192457016,
+      "grad_norm": 1.0542603731155396,
+      "learning_rate": 9.11111111111111e-05,
+      "loss": 10.3539,
+      "step": 41
+    },
+    {
+      "epoch": 0.09317803660565724,
+      "grad_norm": 1.1087148189544678,
+      "learning_rate": 9.333333333333333e-05,
+      "loss": 10.3679,
+      "step": 42
+    },
+    {
+      "epoch": 0.09539656128674431,
+      "grad_norm": 1.0719577074050903,
+      "learning_rate": 9.555555555555555e-05,
+      "loss": 10.0828,
+      "step": 43
+    },
+    {
+      "epoch": 0.09761508596783139,
+      "grad_norm": 1.066102147102356,
+      "learning_rate": 9.777777777777778e-05,
+      "loss": 10.0571,
+      "step": 44
+    },
+    {
+      "epoch": 0.09983361064891846,
+      "grad_norm": 1.1067050695419312,
+      "learning_rate": 9.999999999999999e-05,
+      "loss": 10.2943,
+      "step": 45
+    },
+    {
+      "epoch": 0.10205213533000555,
+      "grad_norm": 1.0722019672393799,
+      "learning_rate": 0.00010222222222222222,
+      "loss": 10.2303,
+      "step": 46
+    },
+    {
+      "epoch": 0.10427066001109263,
+      "grad_norm": 1.0838395357131958,
+      "learning_rate": 0.00010444444444444444,
+      "loss": 10.0256,
+      "step": 47
+    },
+    {
+      "epoch": 0.1064891846921797,
+      "grad_norm": 1.0551950931549072,
+      "learning_rate": 0.00010666666666666667,
+      "loss": 9.976,
+      "step": 48
+    },
+    {
+      "epoch": 0.10870770937326678,
+      "grad_norm": 1.0216065645217896,
+      "learning_rate": 0.00010888888888888889,
+      "loss": 10.1802,
+      "step": 49
+    },
+    {
+      "epoch": 0.11092623405435385,
+      "grad_norm": 1.0394731760025024,
+      "learning_rate": 0.00011111111111111109,
+      "loss": 10.0419,
+      "step": 50
+    },
+    {
+      "epoch": 0.11314475873544093,
+      "grad_norm": 1.0701547861099243,
+      "learning_rate": 0.00011333333333333331,
+      "loss": 10.0181,
+      "step": 51
+    },
+    {
+      "epoch": 0.115363283416528,
+      "grad_norm": 0.9857007265090942,
+      "learning_rate": 0.00011555555555555554,
+      "loss": 9.7166,
+      "step": 52
+    },
+    {
+      "epoch": 0.11758180809761509,
+      "grad_norm": 0.9681416749954224,
+      "learning_rate": 0.00011777777777777776,
+      "loss": 9.8842,
+      "step": 53
+    },
+    {
+      "epoch": 0.11980033277870217,
+      "grad_norm": 0.9958339929580688,
+      "learning_rate": 0.00011999999999999999,
+      "loss": 9.8908,
+      "step": 54
+    },
+    {
+      "epoch": 0.12201885745978924,
+      "grad_norm": 1.0187242031097412,
+      "learning_rate": 0.00012222222222222221,
+      "loss": 9.7298,
+      "step": 55
+    },
+    {
+      "epoch": 0.12423738214087632,
+      "grad_norm": 1.0376415252685547,
+      "learning_rate": 0.00012444444444444444,
+      "loss": 10.0917,
+      "step": 56
+    },
+    {
+      "epoch": 0.1264559068219634,
+      "grad_norm": 0.9841012954711914,
+      "learning_rate": 0.00012666666666666666,
+      "loss": 9.8048,
+      "step": 57
+    },
+    {
+      "epoch": 0.12867443150305047,
+      "grad_norm": 0.9756758213043213,
+      "learning_rate": 0.0001288888888888889,
+      "loss": 9.6653,
+      "step": 58
+    },
+    {
+      "epoch": 0.13089295618413754,
+      "grad_norm": 0.9935914874076843,
+      "learning_rate": 0.00013111111111111111,
+      "loss": 9.4994,
+      "step": 59
+    },
+    {
+      "epoch": 0.13311148086522462,
+      "grad_norm": 0.9595178365707397,
+      "learning_rate": 0.0001333333333333333,
+      "loss": 9.3914,
+      "step": 60
+    },
+    {
+      "epoch": 0.1353300055463117,
+      "grad_norm": 0.9374932646751404,
+      "learning_rate": 0.00013555555555555554,
+      "loss": 9.7966,
+      "step": 61
+    },
+    {
+      "epoch": 0.13754853022739877,
+      "grad_norm": 0.9411095380783081,
+      "learning_rate": 0.00013777777777777776,
+      "loss": 9.5214,
+      "step": 62
+    },
+    {
+      "epoch": 0.13976705490848584,
+      "grad_norm": 0.9274992346763611,
+      "learning_rate": 0.00014,
+      "loss": 9.8231,
+      "step": 63
+    },
+    {
+      "epoch": 0.14198557958957295,
+      "grad_norm": 0.9439871907234192,
+      "learning_rate": 0.0001422222222222222,
+      "loss": 9.6171,
+      "step": 64
+    },
+    {
+      "epoch": 0.14420410427066002,
+      "grad_norm": 0.9327756762504578,
+      "learning_rate": 0.0001444444444444444,
+      "loss": 9.508,
+      "step": 65
+    },
+    {
+      "epoch": 0.1464226289517471,
+      "grad_norm": 0.9180635213851929,
+      "learning_rate": 0.00014666666666666664,
+      "loss": 9.7139,
+      "step": 66
+    },
+    {
+      "epoch": 0.14864115363283417,
+      "grad_norm": 0.9007446765899658,
+      "learning_rate": 0.00014888888888888886,
+      "loss": 9.567,
+      "step": 67
+    },
+    {
+      "epoch": 0.15085967831392125,
+      "grad_norm": 0.9101786613464355,
+      "learning_rate": 0.00015111111111111109,
+      "loss": 9.3885,
+      "step": 68
+    },
+    {
+      "epoch": 0.15307820299500832,
+      "grad_norm": 0.8953559398651123,
+      "learning_rate": 0.0001533333333333333,
+      "loss": 9.2774,
+      "step": 69
+    },
+    {
+      "epoch": 0.1552967276760954,
+      "grad_norm": 0.9264717698097229,
+      "learning_rate": 0.00015555555555555554,
+      "loss": 9.5706,
+      "step": 70
+    },
+    {
+      "epoch": 0.15751525235718247,
+      "grad_norm": 0.908782422542572,
+      "learning_rate": 0.00015777777777777776,
+      "loss": 9.6184,
+      "step": 71
+    },
+    {
+      "epoch": 0.15973377703826955,
+      "grad_norm": 0.9294350147247314,
+      "learning_rate": 0.00015999999999999999,
+      "loss": 9.6407,
+      "step": 72
+    },
+    {
+      "epoch": 0.16195230171935662,
+      "grad_norm": 0.8775391578674316,
+      "learning_rate": 0.0001622222222222222,
+      "loss": 9.306,
+      "step": 73
+    },
+    {
+      "epoch": 0.1641708264004437,
+      "grad_norm": 0.8257573843002319,
+      "learning_rate": 0.00016444444444444444,
+      "loss": 9.3125,
+      "step": 74
+    },
+    {
+      "epoch": 0.16638935108153077,
+      "grad_norm": 0.9011499285697937,
+      "learning_rate": 0.00016666666666666666,
+      "loss": 9.2683,
+      "step": 75
+    },
+    {
+      "epoch": 0.16860787576261785,
+      "grad_norm": 0.7745787501335144,
+      "learning_rate": 0.00016888888888888889,
+      "loss": 9.3802,
+      "step": 76
+    },
+    {
+      "epoch": 0.17082640044370492,
+      "grad_norm": 0.8384570479393005,
+      "learning_rate": 0.0001711111111111111,
+      "loss": 9.2846,
+      "step": 77
+    },
+    {
+      "epoch": 0.17304492512479203,
+      "grad_norm": 0.8339052796363831,
+      "learning_rate": 0.0001733333333333333,
+      "loss": 9.1193,
+      "step": 78
+    },
+    {
+      "epoch": 0.1752634498058791,
+      "grad_norm": 0.7853884100914001,
+      "learning_rate": 0.00017555555555555553,
+      "loss": 9.147,
+      "step": 79
+    },
+    {
+      "epoch": 0.17748197448696618,
+      "grad_norm": 0.8202596306800842,
+      "learning_rate": 0.00017777777777777776,
+      "loss": 9.3111,
+      "step": 80
+    },
+    {
+      "epoch": 0.17970049916805325,
+      "grad_norm": 0.8197675943374634,
+      "learning_rate": 0.00017999999999999998,
+      "loss": 9.2848,
+      "step": 81
+    },
+    {
+      "epoch": 0.18191902384914033,
+      "grad_norm": 0.7439857721328735,
+      "learning_rate": 0.0001822222222222222,
+      "loss": 9.0122,
+      "step": 82
+    },
+    {
+      "epoch": 0.1841375485302274,
+      "grad_norm": 0.8088521957397461,
+      "learning_rate": 0.00018444444444444443,
+      "loss": 9.0406,
+      "step": 83
+    },
+    {
+      "epoch": 0.18635607321131448,
+      "grad_norm": 0.7830685377120972,
+      "learning_rate": 0.00018666666666666666,
+      "loss": 9.0335,
+      "step": 84
+    },
+    {
+      "epoch": 0.18857459789240155,
+      "grad_norm": 0.802688479423523,
+      "learning_rate": 0.00018888888888888888,
+      "loss": 9.0237,
+      "step": 85
+    },
+    {
+      "epoch": 0.19079312257348863,
+      "grad_norm": 0.7901186347007751,
+      "learning_rate": 0.0001911111111111111,
+      "loss": 8.7781,
+      "step": 86
+    },
+    {
+      "epoch": 0.1930116472545757,
+      "grad_norm": 0.7717525362968445,
+      "learning_rate": 0.00019333333333333333,
+      "loss": 8.9623,
+      "step": 87
+    },
+    {
+      "epoch": 0.19523017193566278,
+      "grad_norm": 0.7694629430770874,
+      "learning_rate": 0.00019555555555555556,
+      "loss": 9.0221,
+      "step": 88
+    },
+    {
+      "epoch": 0.19744869661674985,
+      "grad_norm": 0.761546790599823,
+      "learning_rate": 0.00019777777777777776,
+      "loss": 9.0615,
+      "step": 89
+    },
+    {
+      "epoch": 0.19966722129783693,
+      "grad_norm": 0.7163148522377014,
+      "learning_rate": 0.00019999999999999998,
+      "loss": 8.7093,
+      "step": 90
+    },
+    {
+      "epoch": 0.201885745978924,
+      "grad_norm": 0.7290346622467041,
+      "learning_rate": 0.0002022222222222222,
+      "loss": 8.9644,
+      "step": 91
+    },
+    {
+      "epoch": 0.2041042706600111,
+      "grad_norm": 0.718863844871521,
+      "learning_rate": 0.00020444444444444443,
+      "loss": 8.579,
+      "step": 92
+    },
+    {
+      "epoch": 0.20632279534109818,
+      "grad_norm": 0.7100619673728943,
+      "learning_rate": 0.00020666666666666666,
+      "loss": 8.7074,
+      "step": 93
+    },
+    {
+      "epoch": 0.20854132002218526,
+      "grad_norm": 0.7021188735961914,
+      "learning_rate": 0.00020888888888888888,
+      "loss": 8.9422,
+      "step": 94
+    },
+    {
+      "epoch": 0.21075984470327233,
+      "grad_norm": 0.6996264457702637,
+      "learning_rate": 0.0002111111111111111,
+      "loss": 8.6309,
+      "step": 95
+    },
+    {
+      "epoch": 0.2129783693843594,
+      "grad_norm": 0.6723406314849854,
+      "learning_rate": 0.00021333333333333333,
+      "loss": 8.7309,
+      "step": 96
+    },
+    {
+      "epoch": 0.21519689406544648,
+      "grad_norm": 0.6969107985496521,
+      "learning_rate": 0.00021555555555555556,
+      "loss": 8.7646,
+      "step": 97
+    },
+    {
+      "epoch": 0.21741541874653356,
+      "grad_norm": 0.6710256338119507,
+      "learning_rate": 0.00021777777777777778,
+      "loss": 8.7508,
+      "step": 98
+    },
+    {
+      "epoch": 0.21963394342762063,
+      "grad_norm": 0.6454452276229858,
+      "learning_rate": 0.00021999999999999995,
+      "loss": 8.7111,
+      "step": 99
+    },
+    {
+      "epoch": 0.2218524681087077,
+      "grad_norm": 0.6765632629394531,
+      "learning_rate": 0.00022222222222222218,
+      "loss": 8.389,
+      "step": 100
+    },
+    {
+      "epoch": 0.22407099278979478,
+      "grad_norm": 0.7494005560874939,
+      "learning_rate": 0.0002244444444444444,
+      "loss": 8.1889,
+      "step": 101
+    },
+    {
+      "epoch": 0.22628951747088186,
+      "grad_norm": 0.6766241192817688,
+      "learning_rate": 0.00022666666666666663,
+      "loss": 8.4914,
+      "step": 102
+    },
+    {
+      "epoch": 0.22850804215196893,
+      "grad_norm": 0.6485863924026489,
+      "learning_rate": 0.00022888888888888885,
+      "loss": 8.5982,
+      "step": 103
+    },
+    {
+      "epoch": 0.230726566833056,
+      "grad_norm": 0.7525724768638611,
+      "learning_rate": 0.00023111111111111108,
+      "loss": 8.2318,
+      "step": 104
+    },
+    {
+      "epoch": 0.23294509151414308,
+      "grad_norm": 0.6623886823654175,
+      "learning_rate": 0.0002333333333333333,
+      "loss": 8.6508,
+      "step": 105
+    },
+    {
+      "epoch": 0.23516361619523019,
+      "grad_norm": 0.6299893260002136,
+      "learning_rate": 0.00023555555555555553,
+      "loss": 8.3167,
+      "step": 106
+    },
+    {
+      "epoch": 0.23738214087631726,
+      "grad_norm": 0.6460957527160645,
+      "learning_rate": 0.00023777777777777775,
+      "loss": 8.3357,
+      "step": 107
+    },
+    {
+      "epoch": 0.23960066555740434,
+      "grad_norm": 0.6030427813529968,
+      "learning_rate": 0.00023999999999999998,
+      "loss": 8.4493,
+      "step": 108
+    },
+    {
+      "epoch": 0.2418191902384914,
+      "grad_norm": 0.5774738788604736,
+      "learning_rate": 0.0002422222222222222,
+      "loss": 8.2635,
+      "step": 109
+    },
+    {
+      "epoch": 0.24403771491957849,
+      "grad_norm": 0.5799251794815063,
+      "learning_rate": 0.00024444444444444443,
+      "loss": 8.4008,
+      "step": 110
+    },
+    {
+      "epoch": 0.24625623960066556,
+      "grad_norm": 0.5859992504119873,
+      "learning_rate": 0.0002466666666666666,
+      "loss": 8.2664,
+      "step": 111
+    },
+    {
+      "epoch": 0.24847476428175264,
+      "grad_norm": 0.6388427019119263,
+      "learning_rate": 0.0002488888888888889,
+      "loss": 8.1833,
+      "step": 112
+    },
+    {
+      "epoch": 0.2506932889628397,
+      "grad_norm": 0.6277703046798706,
+      "learning_rate": 0.0002511111111111111,
+      "loss": 8.0622,
+      "step": 113
+    },
+    {
+      "epoch": 0.2529118136439268,
+      "grad_norm": 0.5670188665390015,
+      "learning_rate": 0.00025333333333333333,
+      "loss": 8.3236,
+      "step": 114
+    },
+    {
+      "epoch": 0.25513033832501386,
+      "grad_norm": 0.5631948113441467,
+      "learning_rate": 0.00025555555555555553,
+      "loss": 8.1217,
+      "step": 115
+    },
+    {
+      "epoch": 0.25734886300610094,
+      "grad_norm": 0.5848596096038818,
+      "learning_rate": 0.0002577777777777778,
+      "loss": 7.9796,
+      "step": 116
+    },
+    {
+      "epoch": 0.259567387687188,
+      "grad_norm": 0.6274029612541199,
+      "learning_rate": 0.00026,
+      "loss": 8.2633,
+      "step": 117
+    },
+    {
+      "epoch": 0.2617859123682751,
+      "grad_norm": 0.5669984817504883,
+      "learning_rate": 0.00026222222222222223,
+      "loss": 8.0929,
+      "step": 118
+    },
+    {
+      "epoch": 0.26400443704936216,
+      "grad_norm": 0.5368912220001221,
+      "learning_rate": 0.00026444444444444443,
+      "loss": 8.0291,
+      "step": 119
+    },
+    {
+      "epoch": 0.26622296173044924,
+      "grad_norm": 0.5468351244926453,
+      "learning_rate": 0.0002666666666666666,
+      "loss": 8.1378,
+      "step": 120
+    },
+    {
+      "epoch": 0.2684414864115363,
+      "grad_norm": 0.5614748597145081,
+      "learning_rate": 0.0002688888888888889,
+      "loss": 7.9155,
+      "step": 121
+    },
+    {
+      "epoch": 0.2706600110926234,
+      "grad_norm": 0.5161176323890686,
+      "learning_rate": 0.0002711111111111111,
+      "loss": 7.8517,
+      "step": 122
+    },
+    {
+      "epoch": 0.27287853577371046,
+      "grad_norm": 0.5486504435539246,
+      "learning_rate": 0.00027333333333333333,
+      "loss": 7.9138,
+      "step": 123
+    },
+    {
+      "epoch": 0.27509706045479754,
+      "grad_norm": 0.5163204669952393,
+      "learning_rate": 0.0002755555555555555,
+      "loss": 8.0059,
+      "step": 124
+    },
+    {
+      "epoch": 0.2773155851358846,
+      "grad_norm": 0.45901772379875183,
+      "learning_rate": 0.0002777777777777778,
+      "loss": 7.9699,
+      "step": 125
+    },
+    {
+      "epoch": 0.2795341098169717,
+      "grad_norm": 0.4993106424808502,
+      "learning_rate": 0.00028,
+      "loss": 7.9057,
+      "step": 126
+    },
+    {
+      "epoch": 0.2817526344980588,
+      "grad_norm": 0.4832811653614044,
+      "learning_rate": 0.00028222222222222223,
+      "loss": 7.6726,
+      "step": 127
+    },
+    {
+      "epoch": 0.2839711591791459,
+      "grad_norm": 0.4817212224006653,
+      "learning_rate": 0.0002844444444444444,
+      "loss": 8.1137,
+      "step": 128
+    },
+    {
+      "epoch": 0.28618968386023297,
+      "grad_norm": 0.49282753467559814,
+      "learning_rate": 0.0002866666666666667,
+      "loss": 7.9598,
+      "step": 129
+    },
+    {
+      "epoch": 0.28840820854132004,
+      "grad_norm": 0.4494469165802002,
+      "learning_rate": 0.0002888888888888888,
+      "loss": 8.0565,
+      "step": 130
+    },
+    {
+      "epoch": 0.2906267332224071,
+      "grad_norm": 0.4509477913379669,
+      "learning_rate": 0.0002911111111111111,
+      "loss": 7.9788,
+      "step": 131
+    },
+    {
+      "epoch": 0.2928452579034942,
+      "grad_norm": 0.5167444944381714,
+      "learning_rate": 0.00029333333333333327,
+      "loss": 7.8186,
+      "step": 132
+    },
+    {
+      "epoch": 0.29506378258458127,
+      "grad_norm": 0.4326598644256592,
+      "learning_rate": 0.0002955555555555555,
+      "loss": 7.8381,
+      "step": 133
+    },
+    {
+      "epoch": 0.29728230726566834,
+      "grad_norm": 0.43255409598350525,
+      "learning_rate": 0.0002977777777777777,
+      "loss": 7.688,
+      "step": 134
+    },
+    {
+      "epoch": 0.2995008319467554,
+      "grad_norm": 0.4602767527103424,
+      "learning_rate": 0.0003,
+      "loss": 7.6396,
+      "step": 135
+    },
+    {
+      "epoch": 0.3017193566278425,
+      "grad_norm": 0.46446493268013,
+      "learning_rate": 0.00029975308641975307,
+      "loss": 7.5687,
+      "step": 136
+    },
+    {
+      "epoch": 0.30393788130892957,
+      "grad_norm": 0.4175952076911926,
+      "learning_rate": 0.00029950617283950616,
+      "loss": 7.7182,
+      "step": 137
+    },
+    {
+      "epoch": 0.30615640599001664,
+      "grad_norm": 0.4082953631877899,
+      "learning_rate": 0.00029925925925925926,
+      "loss": 7.6621,
+      "step": 138
+    },
+    {
+      "epoch": 0.3083749306711037,
+      "grad_norm": 0.4042870104312897,
+      "learning_rate": 0.0002990123456790123,
+      "loss": 7.4184,
+      "step": 139
+    },
+    {
+      "epoch": 0.3105934553521908,
+      "grad_norm": 0.4017655849456787,
+      "learning_rate": 0.0002987654320987654,
+      "loss": 7.5842,
+      "step": 140
+    },
+    {
+      "epoch": 0.31281198003327787,
+      "grad_norm": 0.42101818323135376,
+      "learning_rate": 0.0002985185185185185,
+      "loss": 7.4431,
+      "step": 141
+    },
+    {
+      "epoch": 0.31503050471436495,
+      "grad_norm": 0.3846524953842163,
+      "learning_rate": 0.0002982716049382716,
+      "loss": 7.4123,
+      "step": 142
+    },
+    {
+      "epoch": 0.317249029395452,
+      "grad_norm": 0.39397579431533813,
+      "learning_rate": 0.0002980246913580247,
+      "loss": 7.394,
+      "step": 143
+    },
+    {
+      "epoch": 0.3194675540765391,
+      "grad_norm": 0.34467625617980957,
+      "learning_rate": 0.0002977777777777777,
+      "loss": 7.6047,
+      "step": 144
+    },
+    {
+      "epoch": 0.32168607875762617,
+      "grad_norm": 0.40106266736984253,
+      "learning_rate": 0.00029753086419753087,
+      "loss": 7.6535,
+      "step": 145
+    },
+    {
+      "epoch": 0.32390460343871325,
+      "grad_norm": 0.36764979362487793,
+      "learning_rate": 0.0002972839506172839,
+      "loss": 7.5045,
+      "step": 146
+    },
+    {
+      "epoch": 0.3261231281198003,
+      "grad_norm": 0.40877142548561096,
+      "learning_rate": 0.000297037037037037,
+      "loss": 7.4408,
+      "step": 147
+    },
+    {
+      "epoch": 0.3283416528008874,
+      "grad_norm": 0.37857964634895325,
+      "learning_rate": 0.0002967901234567901,
+      "loss": 7.7624,
+      "step": 148
+    },
+    {
+      "epoch": 0.33056017748197447,
+      "grad_norm": 0.39117729663848877,
+      "learning_rate": 0.0002965432098765432,
+      "loss": 7.6961,
+      "step": 149
+    },
+    {
+      "epoch": 0.33277870216306155,
+      "grad_norm": 0.36084800958633423,
+      "learning_rate": 0.00029629629629629624,
+      "loss": 7.3881,
+      "step": 150
+    },
+    {
+      "epoch": 0.3349972268441486,
+      "grad_norm": 0.379761278629303,
+      "learning_rate": 0.00029604938271604933,
+      "loss": 7.4768,
+      "step": 151
+    },
+    {
+      "epoch": 0.3372157515252357,
+      "grad_norm": 0.34341010451316833,
+      "learning_rate": 0.0002958024691358025,
+      "loss": 7.4803,
+      "step": 152
+    },
+    {
+      "epoch": 0.33943427620632277,
+      "grad_norm": 0.36923789978027344,
+      "learning_rate": 0.0002955555555555555,
+      "loss": 7.613,
+      "step": 153
+    },
+    {
+      "epoch": 0.34165280088740985,
+      "grad_norm": 0.3406626284122467,
+      "learning_rate": 0.0002953086419753086,
+      "loss": 7.4491,
+      "step": 154
+    },
+    {
+      "epoch": 0.343871325568497,
+      "grad_norm": 0.3167622685432434,
+      "learning_rate": 0.0002950617283950617,
+      "loss": 7.4884,
+      "step": 155
+    },
+    {
+      "epoch": 0.34608985024958405,
+      "grad_norm": 0.3429216146469116,
+      "learning_rate": 0.0002948148148148148,
+      "loss": 7.4174,
+      "step": 156
+    },
+    {
+      "epoch": 0.3483083749306711,
+      "grad_norm": 0.34287068247795105,
+      "learning_rate": 0.00029456790123456785,
+      "loss": 7.4636,
+      "step": 157
+    },
+    {
+      "epoch": 0.3505268996117582,
+      "grad_norm": 0.3350711166858673,
+      "learning_rate": 0.00029432098765432095,
+      "loss": 7.5935,
+      "step": 158
+    },
+    {
+      "epoch": 0.3527454242928453,
+      "grad_norm": 0.3687296211719513,
+      "learning_rate": 0.00029407407407407404,
+      "loss": 7.2136,
+      "step": 159
+    },
+    {
+      "epoch": 0.35496394897393235,
+      "grad_norm": 0.3349868357181549,
+      "learning_rate": 0.00029382716049382714,
+      "loss": 7.3784,
+      "step": 160
+    },
+    {
+      "epoch": 0.35718247365501943,
+      "grad_norm": 0.35006824135780334,
+      "learning_rate": 0.00029358024691358023,
+      "loss": 7.629,
+      "step": 161
+    },
+    {
+      "epoch": 0.3594009983361065,
+      "grad_norm": 0.3273570239543915,
+      "learning_rate": 0.00029333333333333327,
+      "loss": 7.6591,
+      "step": 162
+    },
+    {
+      "epoch": 0.3616195230171936,
+      "grad_norm": 0.3436873257160187,
+      "learning_rate": 0.0002930864197530864,
+      "loss": 7.464,
+      "step": 163
+    },
+    {
+      "epoch": 0.36383804769828065,
+      "grad_norm": 0.3078007698059082,
+      "learning_rate": 0.00029283950617283946,
+      "loss": 7.3546,
+      "step": 164
+    },
+    {
+      "epoch": 0.36605657237936773,
+      "grad_norm": 0.3399384617805481,
+      "learning_rate": 0.00029259259259259256,
+      "loss": 7.5016,
+      "step": 165
+    },
+    {
+      "epoch": 0.3682750970604548,
+      "grad_norm": 0.3513259291648865,
+      "learning_rate": 0.00029234567901234565,
+      "loss": 7.5932,
+      "step": 166
+    },
+    {
+      "epoch": 0.3704936217415419,
+      "grad_norm": 0.31468260288238525,
+      "learning_rate": 0.00029209876543209875,
+      "loss": 7.1392,
+      "step": 167
+    },
+    {
+      "epoch": 0.37271214642262895,
+      "grad_norm": 0.3267529606819153,
+      "learning_rate": 0.00029185185185185184,
+      "loss": 7.2745,
+      "step": 168
+    },
+    {
+      "epoch": 0.37493067110371603,
+      "grad_norm": 0.35863837599754333,
+      "learning_rate": 0.0002916049382716049,
+      "loss": 7.1919,
+      "step": 169
+    },
+    {
+      "epoch": 0.3771491957848031,
+      "grad_norm": 0.35205790400505066,
+      "learning_rate": 0.000291358024691358,
+      "loss": 7.1959,
+      "step": 170
+    },
+    {
+      "epoch": 0.3793677204658902,
+      "grad_norm": 0.29836606979370117,
+      "learning_rate": 0.0002911111111111111,
+      "loss": 7.3426,
+      "step": 171
+    },
+    {
+      "epoch": 0.38158624514697725,
+      "grad_norm": 0.3301176428794861,
+      "learning_rate": 0.00029086419753086417,
+      "loss": 7.0225,
+      "step": 172
+    },
+    {
+      "epoch": 0.38380476982806433,
+      "grad_norm": 0.36162078380584717,
+      "learning_rate": 0.00029061728395061726,
+      "loss": 6.9085,
+      "step": 173
+    },
+    {
+      "epoch": 0.3860232945091514,
+      "grad_norm": 0.3385748565196991,
+      "learning_rate": 0.00029037037037037036,
+      "loss": 7.2266,
+      "step": 174
+    },
+    {
+      "epoch": 0.3882418191902385,
+      "grad_norm": 0.31260550022125244,
+      "learning_rate": 0.00029012345679012345,
+      "loss": 7.337,
+      "step": 175
+    },
+    {
+      "epoch": 0.39046034387132555,
+      "grad_norm": 0.30554333329200745,
+      "learning_rate": 0.0002898765432098765,
+      "loss": 7.2557,
+      "step": 176
+    },
+    {
+      "epoch": 0.39267886855241263,
+      "grad_norm": 0.3228949010372162,
+      "learning_rate": 0.0002896296296296296,
+      "loss": 7.0266,
+      "step": 177
+    },
+    {
+      "epoch": 0.3948973932334997,
+      "grad_norm": 0.32274696230888367,
+      "learning_rate": 0.0002893827160493827,
+      "loss": 7.1147,
+      "step": 178
+    },
+    {
+      "epoch": 0.3971159179145868,
+      "grad_norm": 0.2970581352710724,
+      "learning_rate": 0.0002891358024691358,
+      "loss": 7.3087,
+      "step": 179
+    },
+    {
+      "epoch": 0.39933444259567386,
+      "grad_norm": 0.3574936091899872,
+      "learning_rate": 0.0002888888888888888,
+      "loss": 7.3228,
+      "step": 180
+    },
+    {
+      "epoch": 0.40155296727676093,
+      "grad_norm": 0.3356855809688568,
+      "learning_rate": 0.00028864197530864197,
+      "loss": 7.1229,
+      "step": 181
+    },
+    {
+      "epoch": 0.403771491957848,
+      "grad_norm": 0.3054789900779724,
+      "learning_rate": 0.00028839506172839507,
+      "loss": 7.1922,
+      "step": 182
+    },
+    {
+      "epoch": 0.4059900166389351,
+      "grad_norm": 0.3243318200111389,
+      "learning_rate": 0.0002881481481481481,
+      "loss": 6.9791,
+      "step": 183
+    },
+    {
+      "epoch": 0.4082085413200222,
+      "grad_norm": 0.30905842781066895,
+      "learning_rate": 0.0002879012345679012,
+      "loss": 7.2552,
+      "step": 184
+    },
+    {
+      "epoch": 0.4104270660011093,
+      "grad_norm": 0.32617422938346863,
+      "learning_rate": 0.0002876543209876543,
+      "loss": 7.1709,
+      "step": 185
+    },
+    {
+      "epoch": 0.41264559068219636,
+      "grad_norm": 0.3043220043182373,
+      "learning_rate": 0.0002874074074074074,
+      "loss": 6.9716,
+      "step": 186
+    },
+    {
+      "epoch": 0.41486411536328344,
+      "grad_norm": 0.3036252558231354,
+      "learning_rate": 0.00028716049382716043,
+      "loss": 7.0305,
+      "step": 187
+    },
+    {
+      "epoch": 0.4170826400443705,
+      "grad_norm": 0.34982895851135254,
+      "learning_rate": 0.00028691358024691353,
+      "loss": 7.3356,
+      "step": 188
+    },
+    {
+      "epoch": 0.4193011647254576,
+      "grad_norm": 0.2808395326137543,
+      "learning_rate": 0.0002866666666666667,
+      "loss": 6.9606,
+      "step": 189
+    },
+    {
+      "epoch": 0.42151968940654466,
+      "grad_norm": 0.28308701515197754,
+      "learning_rate": 0.0002864197530864197,
+      "loss": 7.0282,
+      "step": 190
+    },
+    {
+      "epoch": 0.42373821408763174,
+      "grad_norm": 0.3088284730911255,
+      "learning_rate": 0.0002861728395061728,
+      "loss": 7.1575,
+      "step": 191
+    },
+    {
+      "epoch": 0.4259567387687188,
+      "grad_norm": 0.2815094292163849,
+      "learning_rate": 0.0002859259259259259,
+      "loss": 7.017,
+      "step": 192
+    },
+    {
+      "epoch": 0.4281752634498059,
+      "grad_norm": 0.27697208523750305,
+      "learning_rate": 0.000285679012345679,
+      "loss": 7.1683,
+      "step": 193
+    },
+    {
+      "epoch": 0.43039378813089296,
+      "grad_norm": 0.30706897377967834,
+      "learning_rate": 0.00028543209876543204,
+      "loss": 7.0879,
+      "step": 194
+    },
+    {
+      "epoch": 0.43261231281198004,
+      "grad_norm": 0.2846527397632599,
+      "learning_rate": 0.00028518518518518514,
+      "loss": 7.052,
+      "step": 195
+    },
+    {
+      "epoch": 0.4348308374930671,
+      "grad_norm": 0.28693267703056335,
+      "learning_rate": 0.00028493827160493824,
+      "loss": 7.0918,
+      "step": 196
+    },
+    {
+      "epoch": 0.4370493621741542,
+      "grad_norm": 0.264326274394989,
+      "learning_rate": 0.00028469135802469133,
+      "loss": 7.1305,
+      "step": 197
+    },
+    {
+      "epoch": 0.43926788685524126,
+      "grad_norm": 0.2811662256717682,
+      "learning_rate": 0.0002844444444444444,
+      "loss": 7.2153,
+      "step": 198
+    },
+    {
+      "epoch": 0.44148641153632834,
+      "grad_norm": 0.2833110988140106,
+      "learning_rate": 0.0002841975308641975,
+      "loss": 7.1177,
+      "step": 199
+    },
+    {
+      "epoch": 0.4437049362174154,
+      "grad_norm": 0.29461929202079773,
+      "learning_rate": 0.0002839506172839506,
+      "loss": 7.1325,
+      "step": 200
+    },
+    {
+      "epoch": 0.4459234608985025,
+      "grad_norm": 0.3090573847293854,
+      "learning_rate": 0.00028370370370370366,
+      "loss": 7.0066,
+      "step": 201
+    },
+    {
+      "epoch": 0.44814198557958956,
+      "grad_norm": 0.3218264579772949,
+      "learning_rate": 0.00028345679012345675,
+      "loss": 7.3131,
+      "step": 202
+    },
+    {
+      "epoch": 0.45036051026067664,
+      "grad_norm": 0.2607736885547638,
+      "learning_rate": 0.00028320987654320985,
+      "loss": 7.1063,
+      "step": 203
+    },
+    {
+      "epoch": 0.4525790349417637,
+      "grad_norm": 0.28818807005882263,
+      "learning_rate": 0.00028296296296296294,
+      "loss": 7.2934,
+      "step": 204
+    },
+    {
+      "epoch": 0.4547975596228508,
+      "grad_norm": 0.31554779410362244,
+      "learning_rate": 0.00028271604938271604,
+      "loss": 7.1074,
+      "step": 205
+    },
+    {
+      "epoch": 0.45701608430393786,
+      "grad_norm": 0.2890768051147461,
+      "learning_rate": 0.0002824691358024691,
+      "loss": 7.013,
+      "step": 206
+    },
+    {
+      "epoch": 0.45923460898502494,
+      "grad_norm": 0.25389760732650757,
+      "learning_rate": 0.00028222222222222223,
+      "loss": 6.9485,
+      "step": 207
+    },
+    {
+      "epoch": 0.461453133666112,
+      "grad_norm": 0.28469231724739075,
+      "learning_rate": 0.00028197530864197527,
+      "loss": 6.6512,
+      "step": 208
+    },
+    {
+      "epoch": 0.4636716583471991,
+      "grad_norm": 0.260659784078598,
+      "learning_rate": 0.00028172839506172836,
+      "loss": 6.9209,
+      "step": 209
+    },
+    {
+      "epoch": 0.46589018302828616,
+      "grad_norm": 0.2575090527534485,
+      "learning_rate": 0.00028148148148148146,
+      "loss": 6.9054,
+      "step": 210
+    },
+    {
+      "epoch": 0.46810870770937324,
+      "grad_norm": 0.25472602248191833,
+      "learning_rate": 0.00028123456790123455,
+      "loss": 6.9883,
+      "step": 211
+    },
+    {
+      "epoch": 0.47032723239046037,
+      "grad_norm": 0.2650412321090698,
+      "learning_rate": 0.00028098765432098765,
+      "loss": 6.5799,
+      "step": 212
+    },
+    {
+      "epoch": 0.47254575707154745,
+      "grad_norm": 0.28965824842453003,
+      "learning_rate": 0.0002807407407407407,
+      "loss": 7.0815,
+      "step": 213
+    },
+    {
+      "epoch": 0.4747642817526345,
+      "grad_norm": 0.2619493305683136,
+      "learning_rate": 0.0002804938271604938,
+      "loss": 6.6163,
+      "step": 214
+    },
+    {
+      "epoch": 0.4769828064337216,
+      "grad_norm": 0.2571360468864441,
+      "learning_rate": 0.0002802469135802469,
+      "loss": 7.0078,
+      "step": 215
+    },
+    {
+      "epoch": 0.47920133111480867,
+      "grad_norm": 0.2628685534000397,
+      "learning_rate": 0.00028,
+      "loss": 6.7173,
+      "step": 216
+    },
+    {
+      "epoch": 0.48141985579589575,
+      "grad_norm": 0.29474902153015137,
+      "learning_rate": 0.00027975308641975307,
+      "loss": 6.9638,
+      "step": 217
+    },
+    {
+      "epoch": 0.4836383804769828,
+      "grad_norm": 0.24751241505146027,
+      "learning_rate": 0.00027950617283950617,
+      "loss": 6.582,
+      "step": 218
+    },
+    {
+      "epoch": 0.4858569051580699,
+      "grad_norm": 0.2683435082435608,
+      "learning_rate": 0.00027925925925925926,
+      "loss": 6.7422,
+      "step": 219
+    },
+    {
+      "epoch": 0.48807542983915697,
+      "grad_norm": 0.27542054653167725,
+      "learning_rate": 0.0002790123456790123,
+      "loss": 7.0102,
+      "step": 220
+    },
+    {
+      "epoch": 0.49029395452024405,
+      "grad_norm": 0.25449153780937195,
+      "learning_rate": 0.0002787654320987654,
+      "loss": 7.1574,
+      "step": 221
+    },
+    {
+      "epoch": 0.4925124792013311,
+      "grad_norm": 0.2633727192878723,
+      "learning_rate": 0.0002785185185185185,
+      "loss": 6.9003,
+      "step": 222
+    },
+    {
+      "epoch": 0.4947310038824182,
+      "grad_norm": 0.26743316650390625,
+      "learning_rate": 0.0002782716049382716,
+      "loss": 7.1016,
+      "step": 223
+    },
+    {
+      "epoch": 0.49694952856350527,
+      "grad_norm": 0.24339991807937622,
+      "learning_rate": 0.00027802469135802463,
+      "loss": 6.7837,
+      "step": 224
+    },
+    {
+      "epoch": 0.49916805324459235,
+      "grad_norm": 0.2439383864402771,
+      "learning_rate": 0.0002777777777777778,
+      "loss": 6.8307,
+      "step": 225
+    },
+    {
+      "epoch": 0.5013865779256794,
+      "grad_norm": 0.2492426186800003,
+      "learning_rate": 0.00027753086419753087,
+      "loss": 6.8307,
+      "step": 226
+    },
+    {
+      "epoch": 0.5036051026067665,
+      "grad_norm": 0.26650193333625793,
+      "learning_rate": 0.0002772839506172839,
+      "loss": 6.9375,
+      "step": 227
+    },
+    {
+      "epoch": 0.5058236272878536,
+      "grad_norm": 0.2751457393169403,
+      "learning_rate": 0.000277037037037037,
+      "loss": 6.9539,
+      "step": 228
+    },
+    {
+      "epoch": 0.5080421519689406,
+      "grad_norm": 0.23804768919944763,
+      "learning_rate": 0.0002767901234567901,
+      "loss": 6.6881,
+      "step": 229
+    },
+    {
+      "epoch": 0.5102606766500277,
+      "grad_norm": 0.23854023218154907,
+      "learning_rate": 0.0002765432098765432,
+      "loss": 6.8586,
+      "step": 230
+    },
+    {
+      "epoch": 0.5124792013311148,
+      "grad_norm": 0.24612320959568024,
+      "learning_rate": 0.00027629629629629624,
+      "loss": 6.7817,
+      "step": 231
+    },
+    {
+      "epoch": 0.5146977260122019,
+      "grad_norm": 0.2650507986545563,
+      "learning_rate": 0.00027604938271604934,
+      "loss": 6.7761,
+      "step": 232
+    },
+    {
+      "epoch": 0.516916250693289,
+      "grad_norm": 0.23029375076293945,
+      "learning_rate": 0.00027580246913580243,
+      "loss": 6.7838,
+      "step": 233
+    },
+    {
+      "epoch": 0.519134775374376,
+      "grad_norm": 0.27077022194862366,
+      "learning_rate": 0.0002755555555555555,
+      "loss": 6.7498,
+      "step": 234
+    },
+    {
+      "epoch": 0.5213533000554631,
+      "grad_norm": 0.3083948791027069,
+      "learning_rate": 0.0002753086419753086,
+      "loss": 6.5948,
+      "step": 235
+    },
+    {
+      "epoch": 0.5235718247365502,
+      "grad_norm": 0.22145842015743256,
+      "learning_rate": 0.0002750617283950617,
+      "loss": 6.7718,
+      "step": 236
+    },
+    {
+      "epoch": 0.5257903494176372,
+      "grad_norm": 0.25570163130760193,
+      "learning_rate": 0.0002748148148148148,
+      "loss": 6.9161,
+      "step": 237
+    },
+    {
+      "epoch": 0.5280088740987243,
+      "grad_norm": 0.2459215670824051,
+      "learning_rate": 0.00027456790123456785,
+      "loss": 6.548,
+      "step": 238
+    },
+    {
+      "epoch": 0.5302273987798114,
+      "grad_norm": 0.26374247670173645,
+      "learning_rate": 0.00027432098765432095,
+      "loss": 6.6719,
+      "step": 239
+    },
+    {
+      "epoch": 0.5324459234608985,
+      "grad_norm": 0.262940376996994,
+      "learning_rate": 0.00027407407407407404,
+      "loss": 6.811,
+      "step": 240
+    },
+    {
+      "epoch": 0.5346644481419855,
+      "grad_norm": 0.27558085322380066,
+      "learning_rate": 0.00027382716049382714,
+      "loss": 6.4132,
+      "step": 241
+    },
+    {
+      "epoch": 0.5368829728230726,
+      "grad_norm": 0.24056218564510345,
+      "learning_rate": 0.00027358024691358023,
+      "loss": 6.6795,
+      "step": 242
+    },
+    {
+      "epoch": 0.5391014975041597,
+      "grad_norm": 0.238752543926239,
+      "learning_rate": 0.00027333333333333333,
+      "loss": 6.7221,
+      "step": 243
+    },
+    {
+      "epoch": 0.5413200221852468,
+      "grad_norm": 0.2519296705722809,
+      "learning_rate": 0.0002730864197530864,
+      "loss": 6.7385,
+      "step": 244
+    },
+    {
+      "epoch": 0.5435385468663338,
+      "grad_norm": 0.27987128496170044,
+      "learning_rate": 0.00027283950617283946,
+      "loss": 6.5771,
+      "step": 245
+    },
+    {
+      "epoch": 0.5457570715474209,
+      "grad_norm": 0.2742338478565216,
+      "learning_rate": 0.00027259259259259256,
+      "loss": 6.7621,
+      "step": 246
+    },
+    {
+      "epoch": 0.547975596228508,
+      "grad_norm": 0.2317933738231659,
+      "learning_rate": 0.00027234567901234565,
+      "loss": 6.6533,
+      "step": 247
+    },
+    {
+      "epoch": 0.5501941209095951,
+      "grad_norm": 0.2501550614833832,
+      "learning_rate": 0.00027209876543209875,
+      "loss": 6.8292,
+      "step": 248
+    },
+    {
+      "epoch": 0.5524126455906821,
+      "grad_norm": 0.23890024423599243,
+      "learning_rate": 0.00027185185185185184,
+      "loss": 6.6285,
+      "step": 249
+    },
+    {
+      "epoch": 0.5546311702717692,
+      "grad_norm": 0.24808567762374878,
+      "learning_rate": 0.0002716049382716049,
+      "loss": 6.6668,
+      "step": 250
+    },
+    {
+      "epoch": 0.5568496949528563,
+      "grad_norm": 0.23725511133670807,
+      "learning_rate": 0.00027135802469135803,
+      "loss": 6.6848,
+      "step": 251
+    },
+    {
+      "epoch": 0.5590682196339434,
+      "grad_norm": 0.2412443310022354,
+      "learning_rate": 0.0002711111111111111,
+      "loss": 6.6939,
+      "step": 252
+    },
+    {
+      "epoch": 0.5612867443150305,
+      "grad_norm": 0.23535434901714325,
+      "learning_rate": 0.00027086419753086417,
+      "loss": 6.8225,
+      "step": 253
+    },
+    {
+      "epoch": 0.5635052689961176,
+      "grad_norm": 0.23818330466747284,
+      "learning_rate": 0.00027061728395061727,
+      "loss": 6.6261,
+      "step": 254
+    },
+    {
+      "epoch": 0.5657237936772047,
+      "grad_norm": 0.22107459604740143,
+      "learning_rate": 0.00027037037037037036,
+      "loss": 6.6456,
+      "step": 255
+    },
+    {
+      "epoch": 0.5679423183582918,
+      "grad_norm": 0.2538255751132965,
+      "learning_rate": 0.00027012345679012346,
+      "loss": 6.6327,
+      "step": 256
+    },
+    {
+      "epoch": 0.5701608430393789,
+      "grad_norm": 0.2374805510044098,
+      "learning_rate": 0.0002698765432098765,
+      "loss": 6.5746,
+      "step": 257
+    },
+    {
+      "epoch": 0.5723793677204659,
+      "grad_norm": 0.20528048276901245,
+      "learning_rate": 0.0002696296296296296,
+      "loss": 6.598,
+      "step": 258
+    },
+    {
+      "epoch": 0.574597892401553,
+      "grad_norm": 0.2272138148546219,
+      "learning_rate": 0.0002693827160493827,
+      "loss": 6.7462,
+      "step": 259
+    },
+    {
+      "epoch": 0.5768164170826401,
+      "grad_norm": 0.23767343163490295,
+      "learning_rate": 0.0002691358024691358,
+      "loss": 6.8301,
+      "step": 260
+    },
+    {
+      "epoch": 0.5790349417637272,
+      "grad_norm": 0.23041395843029022,
+      "learning_rate": 0.0002688888888888889,
+      "loss": 6.7038,
+      "step": 261
+    },
+    {
+      "epoch": 0.5812534664448142,
+      "grad_norm": 0.2564769685268402,
+      "learning_rate": 0.00026864197530864197,
+      "loss": 6.2089,
+      "step": 262
+    },
+    {
+      "epoch": 0.5834719911259013,
+      "grad_norm": 0.25524330139160156,
+      "learning_rate": 0.000268395061728395,
+      "loss": 6.5608,
+      "step": 263
+    },
+    {
+      "epoch": 0.5856905158069884,
+      "grad_norm": 0.228645920753479,
+      "learning_rate": 0.0002681481481481481,
+      "loss": 6.5278,
+      "step": 264
+    },
+    {
+      "epoch": 0.5879090404880755,
+      "grad_norm": 0.2703951597213745,
+      "learning_rate": 0.0002679012345679012,
+      "loss": 6.824,
+      "step": 265
+    },
+    {
+      "epoch": 0.5901275651691625,
+      "grad_norm": 0.2671278715133667,
+      "learning_rate": 0.0002676543209876543,
+      "loss": 6.6325,
+      "step": 266
+    },
+    {
+      "epoch": 0.5923460898502496,
+      "grad_norm": 0.23450027406215668,
+      "learning_rate": 0.0002674074074074074,
+      "loss": 6.6665,
+      "step": 267
+    },
+    {
+      "epoch": 0.5945646145313367,
+      "grad_norm": 0.2256949096918106,
+      "learning_rate": 0.00026716049382716043,
+      "loss": 6.8691,
+      "step": 268
+    },
+    {
+      "epoch": 0.5967831392124238,
+      "grad_norm": 0.24632082879543304,
+      "learning_rate": 0.0002669135802469136,
+      "loss": 6.8657,
+      "step": 269
+    },
+    {
+      "epoch": 0.5990016638935108,
+      "grad_norm": 0.25359541177749634,
+      "learning_rate": 0.0002666666666666666,
+      "loss": 6.6262,
+      "step": 270
+    },
+    {
+      "epoch": 0.6012201885745979,
+      "grad_norm": 0.25616469979286194,
+      "learning_rate": 0.0002664197530864197,
+      "loss": 6.6533,
+      "step": 271
+    },
+    {
+      "epoch": 0.603438713255685,
+      "grad_norm": 0.22981275618076324,
+      "learning_rate": 0.0002661728395061728,
+      "loss": 6.5217,
+      "step": 272
+    },
+    {
+      "epoch": 0.6056572379367721,
+      "grad_norm": 0.22860075533390045,
+      "learning_rate": 0.0002659259259259259,
+      "loss": 6.5787,
+      "step": 273
+    },
+    {
+      "epoch": 0.6078757626178591,
+      "grad_norm": 0.2413710653781891,
+      "learning_rate": 0.000265679012345679,
+      "loss": 6.5743,
+      "step": 274
+    },
+    {
+      "epoch": 0.6100942872989462,
+      "grad_norm": 0.24007168412208557,
+      "learning_rate": 0.00026543209876543205,
+      "loss": 6.7246,
+      "step": 275
+    },
+    {
+      "epoch": 0.6123128119800333,
+      "grad_norm": 0.23446588218212128,
+      "learning_rate": 0.00026518518518518514,
+      "loss": 6.817,
+      "step": 276
+    },
+    {
+      "epoch": 0.6145313366611204,
+      "grad_norm": 0.21462015807628632,
+      "learning_rate": 0.00026493827160493824,
+      "loss": 6.783,
+      "step": 277
+    },
+    {
+      "epoch": 0.6167498613422074,
+      "grad_norm": 0.23096944391727448,
+      "learning_rate": 0.00026469135802469133,
+      "loss": 6.4108,
+      "step": 278
+    },
+    {
+      "epoch": 0.6189683860232945,
+      "grad_norm": 0.23907549679279327,
+      "learning_rate": 0.00026444444444444443,
+      "loss": 6.7468,
+      "step": 279
+    },
+    {
+      "epoch": 0.6211869107043816,
+      "grad_norm": 0.20310968160629272,
+      "learning_rate": 0.0002641975308641975,
+      "loss": 6.6065,
+      "step": 280
+    },
+    {
+      "epoch": 0.6234054353854687,
+      "grad_norm": 0.2030031681060791,
+      "learning_rate": 0.0002639506172839506,
+      "loss": 6.5283,
+      "step": 281
+    },
+    {
+      "epoch": 0.6256239600665557,
+      "grad_norm": 0.26847827434539795,
+      "learning_rate": 0.00026370370370370366,
+      "loss": 6.0551,
+      "step": 282
+    },
+    {
+      "epoch": 0.6278424847476428,
+      "grad_norm": 0.24113918840885162,
+      "learning_rate": 0.00026345679012345675,
+      "loss": 6.8271,
+      "step": 283
+    },
+    {
+      "epoch": 0.6300610094287299,
+      "grad_norm": 0.2643338441848755,
+      "learning_rate": 0.00026320987654320985,
+      "loss": 6.6905,
+      "step": 284
+    },
+    {
+      "epoch": 0.632279534109817,
+      "grad_norm": 0.21447430551052094,
+      "learning_rate": 0.00026296296296296294,
+      "loss": 6.6154,
+      "step": 285
+    },
+    {
+      "epoch": 0.634498058790904,
+      "grad_norm": 0.1953769028186798,
+      "learning_rate": 0.00026271604938271604,
+      "loss": 6.5985,
+      "step": 286
+    },
+    {
+      "epoch": 0.6367165834719911,
+      "grad_norm": 0.20745937526226044,
+      "learning_rate": 0.00026246913580246913,
+      "loss": 6.5352,
+      "step": 287
+    },
+    {
+      "epoch": 0.6389351081530782,
+      "grad_norm": 0.3284569978713989,
+      "learning_rate": 0.00026222222222222223,
+      "loss": 6.2003,
+      "step": 288
+    },
+    {
+      "epoch": 0.6411536328341653,
+      "grad_norm": 0.2083611935377121,
+      "learning_rate": 0.00026197530864197527,
+      "loss": 6.5822,
+      "step": 289
+    },
+    {
+      "epoch": 0.6433721575152523,
+      "grad_norm": 0.3313221335411072,
+      "learning_rate": 0.00026172839506172837,
+      "loss": 6.1594,
+      "step": 290
+    },
+    {
+      "epoch": 0.6455906821963394,
+      "grad_norm": 0.21450215578079224,
+      "learning_rate": 0.00026148148148148146,
+      "loss": 6.6236,
+      "step": 291
+    },
+    {
+      "epoch": 0.6478092068774265,
+      "grad_norm": 0.22877761721611023,
+      "learning_rate": 0.00026123456790123456,
+      "loss": 6.7431,
+      "step": 292
+    },
+    {
+      "epoch": 0.6500277315585136,
+      "grad_norm": 0.24853159487247467,
+      "learning_rate": 0.0002609876543209876,
+      "loss": 6.7258,
+      "step": 293
+    },
+    {
+      "epoch": 0.6522462562396006,
+      "grad_norm": 0.24076354503631592,
+      "learning_rate": 0.0002607407407407407,
+      "loss": 6.5921,
+      "step": 294
+    },
+    {
+      "epoch": 0.6544647809206877,
+      "grad_norm": 0.21901164948940277,
+      "learning_rate": 0.00026049382716049384,
+      "loss": 6.4178,
+      "step": 295
+    },
+    {
+      "epoch": 0.6566833056017748,
+      "grad_norm": 0.25942957401275635,
+      "learning_rate": 0.0002602469135802469,
+      "loss": 6.6554,
+      "step": 296
+    },
+    {
+      "epoch": 0.6589018302828619,
+      "grad_norm": 0.2387097030878067,
+      "learning_rate": 0.00026,
+      "loss": 6.4346,
+      "step": 297
+    },
+    {
+      "epoch": 0.6611203549639489,
+      "grad_norm": 0.21059872210025787,
+      "learning_rate": 0.00025975308641975307,
+      "loss": 6.4939,
+      "step": 298
+    },
+    {
+      "epoch": 0.663338879645036,
+      "grad_norm": 0.20095428824424744,
+      "learning_rate": 0.00025950617283950617,
+      "loss": 6.8089,
+      "step": 299
+    },
+    {
+      "epoch": 0.6655574043261231,
+      "grad_norm": 0.20134761929512024,
+      "learning_rate": 0.0002592592592592592,
+      "loss": 6.7093,
+      "step": 300
+    },
+    {
+      "epoch": 0.6677759290072102,
+      "grad_norm": 0.21925686299800873,
+      "learning_rate": 0.0002590123456790123,
+      "loss": 6.3138,
+      "step": 301
+    },
+    {
+      "epoch": 0.6699944536882972,
+      "grad_norm": 0.18013806641101837,
+      "learning_rate": 0.0002587654320987654,
+      "loss": 6.4776,
+      "step": 302
+    },
+    {
+      "epoch": 0.6722129783693843,
+      "grad_norm": 0.2511918246746063,
+      "learning_rate": 0.0002585185185185185,
+      "loss": 6.5667,
+      "step": 303
+    },
+    {
+      "epoch": 0.6744315030504714,
+      "grad_norm": 0.22312134504318237,
+      "learning_rate": 0.0002582716049382716,
+      "loss": 6.6706,
+      "step": 304
+    },
+    {
+      "epoch": 0.6766500277315585,
+      "grad_norm": 0.22215072810649872,
+      "learning_rate": 0.0002580246913580247,
+      "loss": 6.4513,
+      "step": 305
+    },
+    {
+      "epoch": 0.6788685524126455,
+      "grad_norm": 0.25114405155181885,
+      "learning_rate": 0.0002577777777777778,
+      "loss": 6.4055,
+      "step": 306
+    },
+    {
+      "epoch": 0.6810870770937326,
+      "grad_norm": 0.2036086767911911,
+      "learning_rate": 0.0002575308641975308,
+      "loss": 6.6007,
+      "step": 307
+    },
+    {
+      "epoch": 0.6833056017748197,
+      "grad_norm": 0.21653708815574646,
+      "learning_rate": 0.0002572839506172839,
+      "loss": 6.5207,
+      "step": 308
+    },
+    {
+      "epoch": 0.6855241264559068,
+      "grad_norm": 0.21439294517040253,
+      "learning_rate": 0.000257037037037037,
+      "loss": 6.5844,
+      "step": 309
+    },
+    {
+      "epoch": 0.687742651136994,
+      "grad_norm": 0.23314706981182098,
+      "learning_rate": 0.0002567901234567901,
+      "loss": 6.3977,
+      "step": 310
+    },
+    {
+      "epoch": 0.689961175818081,
+      "grad_norm": 0.19121813774108887,
+      "learning_rate": 0.0002565432098765432,
+      "loss": 6.6241,
+      "step": 311
+    },
+    {
+      "epoch": 0.6921797004991681,
+      "grad_norm": 0.19185547530651093,
+      "learning_rate": 0.00025629629629629624,
+      "loss": 6.5957,
+      "step": 312
+    },
+    {
+      "epoch": 0.6943982251802552,
+      "grad_norm": 0.2640933096408844,
+      "learning_rate": 0.0002560493827160494,
+      "loss": 6.453,
+      "step": 313
+    },
+    {
+      "epoch": 0.6966167498613423,
+      "grad_norm": 0.1978221982717514,
+      "learning_rate": 0.00025580246913580243,
+      "loss": 6.4967,
+      "step": 314
+    },
+    {
+      "epoch": 0.6988352745424293,
+      "grad_norm": 0.24565671384334564,
+      "learning_rate": 0.00025555555555555553,
+      "loss": 6.3337,
+      "step": 315
+    },
+    {
+      "epoch": 0.7010537992235164,
+      "grad_norm": 0.24592889845371246,
+      "learning_rate": 0.0002553086419753086,
+      "loss": 6.4784,
+      "step": 316
+    },
+    {
+      "epoch": 0.7032723239046035,
+      "grad_norm": 0.19284385442733765,
+      "learning_rate": 0.0002550617283950617,
+      "loss": 6.4602,
+      "step": 317
+    },
+    {
+      "epoch": 0.7054908485856906,
+      "grad_norm": 0.2490207403898239,
+      "learning_rate": 0.0002548148148148148,
+      "loss": 6.3905,
+      "step": 318
+    },
+    {
+      "epoch": 0.7077093732667776,
+      "grad_norm": 0.1993972510099411,
+      "learning_rate": 0.00025456790123456785,
+      "loss": 6.3628,
+      "step": 319
+    },
+    {
+      "epoch": 0.7099278979478647,
+      "grad_norm": 0.22228364646434784,
+      "learning_rate": 0.00025432098765432095,
+      "loss": 6.4056,
+      "step": 320
+    },
+    {
+      "epoch": 0.7121464226289518,
+      "grad_norm": 0.17104414105415344,
+      "learning_rate": 0.00025407407407407404,
+      "loss": 6.4197,
+      "step": 321
+    },
+    {
+      "epoch": 0.7143649473100389,
+      "grad_norm": 0.2422782927751541,
+      "learning_rate": 0.00025382716049382714,
+      "loss": 6.4768,
+      "step": 322
+    },
+    {
+      "epoch": 0.7165834719911259,
+      "grad_norm": 0.22366926074028015,
+      "learning_rate": 0.00025358024691358023,
+      "loss": 6.3662,
+      "step": 323
+    },
+    {
+      "epoch": 0.718801996672213,
+      "grad_norm": 0.19414354860782623,
+      "learning_rate": 0.00025333333333333333,
+      "loss": 6.4239,
+      "step": 324
+    },
+    {
+      "epoch": 0.7210205213533001,
+      "grad_norm": 0.21946461498737335,
+      "learning_rate": 0.0002530864197530864,
+      "loss": 6.2966,
+      "step": 325
+    },
+    {
+      "epoch": 0.7232390460343872,
+      "grad_norm": 0.2561095058917999,
+      "learning_rate": 0.00025283950617283947,
+      "loss": 6.515,
+      "step": 326
+    },
+    {
+      "epoch": 0.7254575707154742,
+      "grad_norm": 0.22377291321754456,
+      "learning_rate": 0.00025259259259259256,
+      "loss": 6.6789,
+      "step": 327
+    },
+    {
+      "epoch": 0.7276760953965613,
+      "grad_norm": 0.24095018208026886,
+      "learning_rate": 0.00025234567901234566,
+      "loss": 6.5115,
+      "step": 328
+    },
+    {
+      "epoch": 0.7298946200776484,
+      "grad_norm": 0.23663397133350372,
+      "learning_rate": 0.00025209876543209875,
+      "loss": 6.5176,
+      "step": 329
+    },
+    {
+      "epoch": 0.7321131447587355,
+      "grad_norm": 0.18908272683620453,
+      "learning_rate": 0.0002518518518518518,
+      "loss": 6.4015,
+      "step": 330
+    },
+    {
+      "epoch": 0.7343316694398225,
+      "grad_norm": 0.2379632145166397,
+      "learning_rate": 0.00025160493827160494,
+      "loss": 6.5332,
+      "step": 331
+    },
+    {
+      "epoch": 0.7365501941209096,
+      "grad_norm": 0.25492769479751587,
+      "learning_rate": 0.00025135802469135804,
+      "loss": 6.6514,
+      "step": 332
+    },
+    {
+      "epoch": 0.7387687188019967,
+      "grad_norm": 0.17972701787948608,
+      "learning_rate": 0.0002511111111111111,
+      "loss": 6.3011,
+      "step": 333
+    },
+    {
+      "epoch": 0.7409872434830838,
+      "grad_norm": 0.21416546404361725,
+      "learning_rate": 0.00025086419753086417,
+      "loss": 6.4311,
+      "step": 334
+    },
+    {
+      "epoch": 0.7432057681641708,
+      "grad_norm": 0.21596059203147888,
+      "learning_rate": 0.00025061728395061727,
+      "loss": 6.3778,
+      "step": 335
+    },
+    {
+      "epoch": 0.7454242928452579,
+      "grad_norm": 0.20073531568050385,
+      "learning_rate": 0.00025037037037037036,
+      "loss": 6.4671,
+      "step": 336
+    },
+    {
+      "epoch": 0.747642817526345,
+      "grad_norm": 0.2114843726158142,
+      "learning_rate": 0.0002501234567901234,
+      "loss": 6.5777,
+      "step": 337
+    },
+    {
+      "epoch": 0.7498613422074321,
+      "grad_norm": 0.22219784557819366,
+      "learning_rate": 0.0002498765432098765,
+      "loss": 6.6169,
+      "step": 338
+    },
+    {
+      "epoch": 0.7520798668885191,
+      "grad_norm": 0.1996789127588272,
+      "learning_rate": 0.00024962962962962965,
+      "loss": 6.4218,
+      "step": 339
+    },
+    {
+      "epoch": 0.7542983915696062,
+      "grad_norm": 0.1850207895040512,
+      "learning_rate": 0.0002493827160493827,
+      "loss": 6.5524,
+      "step": 340
+    },
+    {
+      "epoch": 0.7565169162506933,
+      "grad_norm": 0.21482406556606293,
+      "learning_rate": 0.0002491358024691358,
+      "loss": 6.1304,
+      "step": 341
+    },
+    {
+      "epoch": 0.7587354409317804,
+      "grad_norm": 0.2280534952878952,
+      "learning_rate": 0.0002488888888888889,
+      "loss": 6.2226,
+      "step": 342
+    },
+    {
+      "epoch": 0.7609539656128674,
+      "grad_norm": 0.19665175676345825,
+      "learning_rate": 0.000248641975308642,
+      "loss": 6.6564,
+      "step": 343
+    },
+    {
+      "epoch": 0.7631724902939545,
+      "grad_norm": 0.2260160744190216,
+      "learning_rate": 0.000248395061728395,
+      "loss": 6.5268,
+      "step": 344
+    },
+    {
+      "epoch": 0.7653910149750416,
+      "grad_norm": 0.2247382551431656,
+      "learning_rate": 0.0002481481481481481,
+      "loss": 6.6323,
+      "step": 345
+    },
+    {
+      "epoch": 0.7676095396561287,
+      "grad_norm": 0.1932763010263443,
+      "learning_rate": 0.0002479012345679012,
+      "loss": 6.4234,
+      "step": 346
+    },
+    {
+      "epoch": 0.7698280643372157,
+      "grad_norm": 0.27263736724853516,
+      "learning_rate": 0.0002476543209876543,
+      "loss": 6.4279,
+      "step": 347
+    },
+    {
+      "epoch": 0.7720465890183028,
+      "grad_norm": 0.20110246539115906,
+      "learning_rate": 0.0002474074074074074,
+      "loss": 6.5224,
+      "step": 348
+    },
+    {
+      "epoch": 0.7742651136993899,
+      "grad_norm": 0.19617919623851776,
+      "learning_rate": 0.0002471604938271605,
+      "loss": 6.4968,
+      "step": 349
+    },
+    {
+      "epoch": 0.776483638380477,
+      "grad_norm": 0.1829862743616104,
+      "learning_rate": 0.0002469135802469136,
+      "loss": 6.4124,
+      "step": 350
+    },
+    {
+      "epoch": 0.778702163061564,
+      "grad_norm": 0.19442309439182281,
+      "learning_rate": 0.0002466666666666666,
+      "loss": 6.4776,
+      "step": 351
+    },
+    {
+      "epoch": 0.7809206877426511,
+      "grad_norm": 0.21634119749069214,
+      "learning_rate": 0.0002464197530864197,
+      "loss": 6.4124,
+      "step": 352
+    },
+    {
+      "epoch": 0.7831392124237382,
+      "grad_norm": 0.2233932763338089,
+      "learning_rate": 0.0002461728395061728,
+      "loss": 6.6008,
+      "step": 353
+    },
+    {
+      "epoch": 0.7853577371048253,
+      "grad_norm": 0.2523881196975708,
+      "learning_rate": 0.0002459259259259259,
+      "loss": 6.4921,
+      "step": 354
+    },
+    {
+      "epoch": 0.7875762617859123,
+      "grad_norm": 0.2476198524236679,
+      "learning_rate": 0.000245679012345679,
+      "loss": 6.3321,
+      "step": 355
+    },
+    {
+      "epoch": 0.7897947864669994,
+      "grad_norm": 0.21258245408535004,
+      "learning_rate": 0.00024543209876543205,
+      "loss": 6.5338,
+      "step": 356
+    },
+    {
+      "epoch": 0.7920133111480865,
+      "grad_norm": 0.2177080363035202,
+      "learning_rate": 0.0002451851851851852,
+      "loss": 6.3615,
+      "step": 357
+    },
+    {
+      "epoch": 0.7942318358291736,
+      "grad_norm": 0.21250708401203156,
+      "learning_rate": 0.00024493827160493824,
+      "loss": 6.3665,
+      "step": 358
+    },
+    {
+      "epoch": 0.7964503605102606,
+      "grad_norm": 0.20654664933681488,
+      "learning_rate": 0.00024469135802469133,
+      "loss": 6.2509,
+      "step": 359
+    },
+    {
+      "epoch": 0.7986688851913477,
+      "grad_norm": 0.18696270883083344,
+      "learning_rate": 0.00024444444444444443,
+      "loss": 6.3309,
+      "step": 360
+    },
+    {
+      "epoch": 0.8008874098724348,
+      "grad_norm": 0.2428227961063385,
+      "learning_rate": 0.0002441975308641975,
+      "loss": 6.2911,
+      "step": 361
+    },
+    {
+      "epoch": 0.8031059345535219,
+      "grad_norm": 0.20401546359062195,
+      "learning_rate": 0.00024395061728395062,
+      "loss": 6.4071,
+      "step": 362
+    },
+    {
+      "epoch": 0.8053244592346089,
+      "grad_norm": 0.1921582967042923,
+      "learning_rate": 0.0002437037037037037,
+      "loss": 6.5119,
+      "step": 363
+    },
+    {
+      "epoch": 0.807542983915696,
+      "grad_norm": 0.2018212080001831,
+      "learning_rate": 0.00024345679012345678,
+      "loss": 6.5049,
+      "step": 364
+    },
+    {
+      "epoch": 0.8097615085967831,
+      "grad_norm": 0.26928919553756714,
+      "learning_rate": 0.00024320987654320985,
+      "loss": 6.4017,
+      "step": 365
+    },
+    {
+      "epoch": 0.8119800332778702,
+      "grad_norm": 0.19453440606594086,
+      "learning_rate": 0.00024296296296296295,
+      "loss": 6.3087,
+      "step": 366
+    },
+    {
+      "epoch": 0.8141985579589573,
+      "grad_norm": 0.21426989138126373,
+      "learning_rate": 0.00024271604938271601,
+      "loss": 6.5557,
+      "step": 367
+    },
+    {
+      "epoch": 0.8164170826400444,
+      "grad_norm": 0.20990267395973206,
+      "learning_rate": 0.0002424691358024691,
+      "loss": 6.3143,
+      "step": 368
+    },
+    {
+      "epoch": 0.8186356073211315,
+      "grad_norm": 0.2221631556749344,
+      "learning_rate": 0.0002422222222222222,
+      "loss": 6.6324,
+      "step": 369
+    },
+    {
+      "epoch": 0.8208541320022186,
+      "grad_norm": 0.1980641782283783,
+      "learning_rate": 0.00024197530864197527,
+      "loss": 6.3113,
+      "step": 370
+    },
+    {
+      "epoch": 0.8230726566833056,
+      "grad_norm": 0.18872343003749847,
+      "learning_rate": 0.0002417283950617284,
+      "loss": 6.2736,
+      "step": 371
+    },
+    {
+      "epoch": 0.8252911813643927,
+      "grad_norm": 0.22105860710144043,
+      "learning_rate": 0.00024148148148148146,
+      "loss": 6.512,
+      "step": 372
+    },
+    {
+      "epoch": 0.8275097060454798,
+      "grad_norm": 0.23365360498428345,
+      "learning_rate": 0.00024123456790123456,
+      "loss": 6.3363,
+      "step": 373
+    },
+    {
+      "epoch": 0.8297282307265669,
+      "grad_norm": 0.20883749425411224,
+      "learning_rate": 0.00024098765432098763,
+      "loss": 6.2316,
+      "step": 374
+    },
+    {
+      "epoch": 0.831946755407654,
+      "grad_norm": 0.20335957407951355,
+      "learning_rate": 0.00024074074074074072,
+      "loss": 6.4047,
+      "step": 375
+    },
+    {
+      "epoch": 0.834165280088741,
+      "grad_norm": 0.21041427552700043,
+      "learning_rate": 0.0002404938271604938,
+      "loss": 6.6132,
+      "step": 376
+    },
+    {
+      "epoch": 0.8363838047698281,
+      "grad_norm": 0.20531633496284485,
+      "learning_rate": 0.00024024691358024688,
+      "loss": 6.5406,
+      "step": 377
+    },
+    {
+      "epoch": 0.8386023294509152,
+      "grad_norm": 0.2081940919160843,
+      "learning_rate": 0.00023999999999999998,
+      "loss": 6.4886,
+      "step": 378
+    },
+    {
+      "epoch": 0.8408208541320022,
+      "grad_norm": 0.19883885979652405,
+      "learning_rate": 0.00023975308641975305,
+      "loss": 6.4257,
+      "step": 379
+    },
+    {
+      "epoch": 0.8430393788130893,
+      "grad_norm": 0.2279159277677536,
+      "learning_rate": 0.00023950617283950617,
+      "loss": 6.4102,
+      "step": 380
+    },
+    {
+      "epoch": 0.8452579034941764,
+      "grad_norm": 0.24276921153068542,
+      "learning_rate": 0.00023925925925925924,
+      "loss": 6.2122,
+      "step": 381
+    },
+    {
+      "epoch": 0.8474764281752635,
+      "grad_norm": 0.1931747943162918,
+      "learning_rate": 0.00023901234567901233,
+      "loss": 6.0709,
+      "step": 382
+    },
+    {
+      "epoch": 0.8496949528563505,
+      "grad_norm": 0.24734923243522644,
+      "learning_rate": 0.0002387654320987654,
+      "loss": 6.3005,
+      "step": 383
+    },
+    {
+      "epoch": 0.8519134775374376,
+      "grad_norm": 0.2190587967634201,
+      "learning_rate": 0.0002385185185185185,
+      "loss": 6.191,
+      "step": 384
+    },
+    {
+      "epoch": 0.8541320022185247,
+      "grad_norm": 0.18894650042057037,
+      "learning_rate": 0.0002382716049382716,
+      "loss": 6.2207,
+      "step": 385
+    },
+    {
+      "epoch": 0.8563505268996118,
+      "grad_norm": 0.2331104725599289,
+      "learning_rate": 0.00023802469135802466,
+      "loss": 6.1855,
+      "step": 386
+    },
+    {
+      "epoch": 0.8585690515806988,
+      "grad_norm": 0.18451757729053497,
+      "learning_rate": 0.00023777777777777775,
+      "loss": 6.4521,
+      "step": 387
+    },
+    {
+      "epoch": 0.8607875762617859,
+      "grad_norm": 0.2656842768192291,
+      "learning_rate": 0.00023753086419753082,
+      "loss": 6.4572,
+      "step": 388
+    },
+    {
+      "epoch": 0.863006100942873,
+      "grad_norm": 0.22807979583740234,
+      "learning_rate": 0.00023728395061728394,
+      "loss": 6.2131,
+      "step": 389
+    },
+    {
+      "epoch": 0.8652246256239601,
+      "grad_norm": 0.23443302512168884,
+      "learning_rate": 0.000237037037037037,
+      "loss": 6.4335,
+      "step": 390
+    },
+    {
+      "epoch": 0.8674431503050472,
+      "grad_norm": 0.2197420746088028,
+      "learning_rate": 0.0002367901234567901,
+      "loss": 6.3759,
+      "step": 391
+    },
+    {
+      "epoch": 0.8696616749861342,
+      "grad_norm": 0.21451890468597412,
+      "learning_rate": 0.0002365432098765432,
+      "loss": 6.3253,
+      "step": 392
+    },
+    {
+      "epoch": 0.8718801996672213,
+      "grad_norm": 0.1962742954492569,
+      "learning_rate": 0.00023629629629629627,
+      "loss": 6.4191,
+      "step": 393
+    },
+    {
+      "epoch": 0.8740987243483084,
+      "grad_norm": 0.2340901643037796,
+      "learning_rate": 0.00023604938271604937,
+      "loss": 6.636,
+      "step": 394
+    },
+    {
+      "epoch": 0.8763172490293955,
+      "grad_norm": 0.17590107023715973,
+      "learning_rate": 0.00023580246913580243,
+      "loss": 6.1462,
+      "step": 395
+    },
+    {
+      "epoch": 0.8785357737104825,
+      "grad_norm": 0.22130917012691498,
+      "learning_rate": 0.00023555555555555553,
+      "loss": 6.3235,
+      "step": 396
+    },
+    {
+      "epoch": 0.8807542983915696,
+      "grad_norm": 0.1897369772195816,
+      "learning_rate": 0.0002353086419753086,
+      "loss": 6.4629,
+      "step": 397
+    },
+    {
+      "epoch": 0.8829728230726567,
+      "grad_norm": 0.24583031237125397,
+      "learning_rate": 0.00023506172839506172,
+      "loss": 6.3964,
+      "step": 398
+    },
+    {
+      "epoch": 0.8851913477537438,
+      "grad_norm": 0.2024451643228531,
+      "learning_rate": 0.0002348148148148148,
+      "loss": 6.4598,
+      "step": 399
+    },
+    {
+      "epoch": 0.8874098724348308,
+      "grad_norm": 0.23341435194015503,
+      "learning_rate": 0.00023456790123456788,
+      "loss": 6.4017,
+      "step": 400
+    },
+    {
+      "epoch": 0.8896283971159179,
+      "grad_norm": 0.2509920299053192,
+      "learning_rate": 0.00023432098765432098,
+      "loss": 5.8893,
+      "step": 401
+    },
+    {
+      "epoch": 0.891846921797005,
+      "grad_norm": 0.21451552212238312,
+      "learning_rate": 0.00023407407407407405,
+      "loss": 6.545,
+      "step": 402
+    },
+    {
+      "epoch": 0.894065446478092,
+      "grad_norm": 0.17398886382579803,
+      "learning_rate": 0.00023382716049382714,
+      "loss": 6.3675,
+      "step": 403
+    },
+    {
+      "epoch": 0.8962839711591791,
+      "grad_norm": 0.18728318810462952,
+      "learning_rate": 0.0002335802469135802,
+      "loss": 6.3444,
+      "step": 404
+    },
+    {
+      "epoch": 0.8985024958402662,
+      "grad_norm": 0.26972535252571106,
+      "learning_rate": 0.0002333333333333333,
+      "loss": 6.4963,
+      "step": 405
+    },
+    {
+      "epoch": 0.9007210205213533,
+      "grad_norm": 0.21416422724723816,
+      "learning_rate": 0.00023308641975308637,
+      "loss": 6.2244,
+      "step": 406
+    },
+    {
+      "epoch": 0.9029395452024404,
+      "grad_norm": 0.24076814949512482,
+      "learning_rate": 0.0002328395061728395,
+      "loss": 6.4137,
+      "step": 407
+    },
+    {
+      "epoch": 0.9051580698835274,
+      "grad_norm": 0.2561902105808258,
+      "learning_rate": 0.0002325925925925926,
+      "loss": 6.5358,
+      "step": 408
+    },
+    {
+      "epoch": 0.9073765945646145,
+      "grad_norm": 0.220223069190979,
+      "learning_rate": 0.00023234567901234566,
+      "loss": 6.3734,
+      "step": 409
+    },
+    {
+      "epoch": 0.9095951192457016,
+      "grad_norm": 0.19265079498291016,
+      "learning_rate": 0.00023209876543209875,
+      "loss": 6.2085,
+      "step": 410
+    },
+    {
+      "epoch": 0.9118136439267887,
+      "grad_norm": 0.25441083312034607,
+      "learning_rate": 0.00023185185185185182,
+      "loss": 6.2423,
+      "step": 411
+    },
+    {
+      "epoch": 0.9140321686078757,
+      "grad_norm": 0.22279766201972961,
+      "learning_rate": 0.00023160493827160492,
+      "loss": 6.3342,
+      "step": 412
+    },
+    {
+      "epoch": 0.9162506932889628,
+      "grad_norm": 0.2314429134130478,
+      "learning_rate": 0.00023135802469135798,
+      "loss": 6.1962,
+      "step": 413
+    },
+    {
+      "epoch": 0.9184692179700499,
+      "grad_norm": 0.21110893785953522,
+      "learning_rate": 0.00023111111111111108,
+      "loss": 6.3292,
+      "step": 414
+    },
+    {
+      "epoch": 0.920687742651137,
+      "grad_norm": 0.2102600336074829,
+      "learning_rate": 0.0002308641975308642,
+      "loss": 6.3331,
+      "step": 415
+    },
+    {
+      "epoch": 0.922906267332224,
+      "grad_norm": 0.20350497961044312,
+      "learning_rate": 0.00023061728395061727,
+      "loss": 6.1111,
+      "step": 416
+    },
+    {
+      "epoch": 0.9251247920133111,
+      "grad_norm": 0.2515314519405365,
+      "learning_rate": 0.00023037037037037036,
+      "loss": 6.0827,
+      "step": 417
+    },
+    {
+      "epoch": 0.9273433166943982,
+      "grad_norm": 0.18929435312747955,
+      "learning_rate": 0.00023012345679012343,
+      "loss": 6.2409,
+      "step": 418
+    },
+    {
+      "epoch": 0.9295618413754853,
+      "grad_norm": 0.2151847779750824,
+      "learning_rate": 0.00022987654320987653,
+      "loss": 6.3051,
+      "step": 419
+    },
+    {
+      "epoch": 0.9317803660565723,
+      "grad_norm": 0.17267687618732452,
+      "learning_rate": 0.0002296296296296296,
+      "loss": 6.3074,
+      "step": 420
+    },
+    {
+      "epoch": 0.9339988907376594,
+      "grad_norm": 0.21381689608097076,
+      "learning_rate": 0.0002293827160493827,
+      "loss": 6.2951,
+      "step": 421
+    },
+    {
+      "epoch": 0.9362174154187465,
+      "grad_norm": 0.21304073929786682,
+      "learning_rate": 0.00022913580246913579,
+      "loss": 6.368,
+      "step": 422
+    },
+    {
+      "epoch": 0.9384359400998337,
+      "grad_norm": 0.18792788684368134,
+      "learning_rate": 0.00022888888888888885,
+      "loss": 6.3617,
+      "step": 423
+    },
+    {
+      "epoch": 0.9406544647809207,
+      "grad_norm": 0.22208838164806366,
+      "learning_rate": 0.00022864197530864198,
+      "loss": 6.2685,
+      "step": 424
+    },
+    {
+      "epoch": 0.9428729894620078,
+      "grad_norm": 0.20574726164340973,
+      "learning_rate": 0.00022839506172839504,
+      "loss": 6.24,
+      "step": 425
+    },
+    {
+      "epoch": 0.9450915141430949,
+      "grad_norm": 0.2772735059261322,
+      "learning_rate": 0.00022814814814814814,
+      "loss": 6.2549,
+      "step": 426
+    },
+    {
+      "epoch": 0.947310038824182,
+      "grad_norm": 0.28948745131492615,
+      "learning_rate": 0.0002279012345679012,
+      "loss": 6.27,
+      "step": 427
+    },
+    {
+      "epoch": 0.949528563505269,
+      "grad_norm": 0.2673797905445099,
+      "learning_rate": 0.0002276543209876543,
+      "loss": 6.1938,
+      "step": 428
+    },
+    {
+      "epoch": 0.9517470881863561,
+      "grad_norm": 0.2008470892906189,
+      "learning_rate": 0.00022740740740740737,
+      "loss": 6.2245,
+      "step": 429
+    },
+    {
+      "epoch": 0.9539656128674432,
+      "grad_norm": 0.30518054962158203,
+      "learning_rate": 0.00022716049382716047,
+      "loss": 6.2861,
+      "step": 430
+    },
+    {
+      "epoch": 0.9561841375485303,
+      "grad_norm": 0.21313375234603882,
+      "learning_rate": 0.00022691358024691356,
+      "loss": 6.2727,
+      "step": 431
+    },
+    {
+      "epoch": 0.9584026622296173,
+      "grad_norm": 0.17316500842571259,
+      "learning_rate": 0.00022666666666666663,
+      "loss": 6.2104,
+      "step": 432
+    },
+    {
+      "epoch": 0.9606211869107044,
+      "grad_norm": 0.21675948798656464,
+      "learning_rate": 0.00022641975308641975,
+      "loss": 6.4053,
+      "step": 433
+    },
+    {
+      "epoch": 0.9628397115917915,
+      "grad_norm": 0.2081632763147354,
+      "learning_rate": 0.00022617283950617282,
+      "loss": 6.2353,
+      "step": 434
+    },
+    {
+      "epoch": 0.9650582362728786,
+      "grad_norm": 0.3015894591808319,
+      "learning_rate": 0.00022592592592592591,
+      "loss": 5.9931,
+      "step": 435
+    },
+    {
+      "epoch": 0.9672767609539656,
+      "grad_norm": 0.2265383005142212,
+      "learning_rate": 0.00022567901234567898,
+      "loss": 6.3502,
+      "step": 436
+    },
+    {
+      "epoch": 0.9694952856350527,
+      "grad_norm": 0.21558386087417603,
+      "learning_rate": 0.00022543209876543208,
+      "loss": 6.314,
+      "step": 437
+    },
+    {
+      "epoch": 0.9717138103161398,
+      "grad_norm": 0.2379695177078247,
+      "learning_rate": 0.00022518518518518517,
+      "loss": 6.2756,
+      "step": 438
+    },
+    {
+      "epoch": 0.9739323349972269,
+      "grad_norm": 0.21380138397216797,
+      "learning_rate": 0.00022493827160493824,
+      "loss": 6.2949,
+      "step": 439
+    },
+    {
+      "epoch": 0.9761508596783139,
+      "grad_norm": 0.203367680311203,
+      "learning_rate": 0.00022469135802469134,
+      "loss": 6.2218,
+      "step": 440
+    },
+    {
+      "epoch": 0.978369384359401,
+      "grad_norm": 0.22846561670303345,
+      "learning_rate": 0.0002244444444444444,
+      "loss": 6.0624,
+      "step": 441
+    },
+    {
+      "epoch": 0.9805879090404881,
+      "grad_norm": 0.19860203564167023,
+      "learning_rate": 0.00022419753086419753,
+      "loss": 6.158,
+      "step": 442
+    },
+    {
+      "epoch": 0.9828064337215752,
+      "grad_norm": 0.21622541546821594,
+      "learning_rate": 0.0002239506172839506,
+      "loss": 6.5092,
+      "step": 443
+    },
+    {
+      "epoch": 0.9850249584026622,
+      "grad_norm": 0.2241537719964981,
+      "learning_rate": 0.0002237037037037037,
+      "loss": 6.164,
+      "step": 444
+    },
+    {
+      "epoch": 0.9872434830837493,
+      "grad_norm": 0.20553408563137054,
+      "learning_rate": 0.00022345679012345678,
+      "loss": 6.0991,
+      "step": 445
+    },
+    {
+      "epoch": 0.9894620077648364,
+      "grad_norm": 0.2322665899991989,
+      "learning_rate": 0.00022320987654320985,
+      "loss": 6.1063,
+      "step": 446
+    },
+    {
+      "epoch": 0.9916805324459235,
+      "grad_norm": 0.2767675817012787,
+      "learning_rate": 0.00022296296296296295,
+      "loss": 5.9006,
+      "step": 447
+    },
+    {
+      "epoch": 0.9938990571270105,
+      "grad_norm": 0.20842640101909637,
+      "learning_rate": 0.00022271604938271602,
+      "loss": 6.3451,
+      "step": 448
+    },
+    {
+      "epoch": 0.9961175818080976,
+      "grad_norm": 0.20310746133327484,
+      "learning_rate": 0.0002224691358024691,
+      "loss": 5.9764,
+      "step": 449
+    },
+    {
+      "epoch": 0.9983361064891847,
+      "grad_norm": 0.28204262256622314,
+      "learning_rate": 0.00022222222222222218,
+      "loss": 6.1048,
+      "step": 450
+    },
+    {
+      "epoch": 1.0005546311702718,
+      "grad_norm": 0.27614691853523254,
+      "learning_rate": 0.0002219753086419753,
+      "loss": 6.4778,
+      "step": 451
+    },
+    {
+      "epoch": 1.0027731558513588,
+      "grad_norm": 0.2524718642234802,
+      "learning_rate": 0.0002217283950617284,
+      "loss": 6.3772,
+      "step": 452
+    },
+    {
+      "epoch": 1.004991680532446,
+      "grad_norm": 0.18646380305290222,
+      "learning_rate": 0.00022148148148148146,
+      "loss": 6.2951,
+      "step": 453
+    },
+    {
+      "epoch": 1.007210205213533,
+      "grad_norm": 0.24464640021324158,
+      "learning_rate": 0.00022123456790123456,
+      "loss": 6.154,
+      "step": 454
+    },
+    {
+      "epoch": 1.00942872989462,
+      "grad_norm": 0.19661787152290344,
+      "learning_rate": 0.00022098765432098763,
+      "loss": 6.1065,
+      "step": 455
+    },
+    {
+      "epoch": 1.0116472545757071,
+      "grad_norm": 0.21767693758010864,
+      "learning_rate": 0.00022074074074074072,
+      "loss": 6.1708,
+      "step": 456
+    },
+    {
+      "epoch": 1.0138657792567942,
+      "grad_norm": 0.2927335798740387,
+      "learning_rate": 0.0002204938271604938,
+      "loss": 5.978,
+      "step": 457
+    },
+    {
+      "epoch": 1.0160843039378813,
+      "grad_norm": 0.1915467530488968,
+      "learning_rate": 0.00022024691358024689,
+      "loss": 6.0331,
+      "step": 458
+    },
+    {
+      "epoch": 1.0183028286189684,
+      "grad_norm": 0.2059955894947052,
+      "learning_rate": 0.00021999999999999995,
+      "loss": 6.1056,
+      "step": 459
+    },
+    {
+      "epoch": 1.0205213533000554,
+      "grad_norm": 0.2013431042432785,
+      "learning_rate": 0.00021975308641975308,
+      "loss": 6.0417,
+      "step": 460
+    },
+    {
+      "epoch": 1.0227398779811425,
+      "grad_norm": 0.18286067247390747,
+      "learning_rate": 0.00021950617283950617,
+      "loss": 5.969,
+      "step": 461
+    },
+    {
+      "epoch": 1.0249584026622296,
+      "grad_norm": 0.19190767407417297,
+      "learning_rate": 0.00021925925925925924,
+      "loss": 6.2809,
+      "step": 462
+    },
+    {
+      "epoch": 1.0271769273433167,
+      "grad_norm": 0.26955410838127136,
+      "learning_rate": 0.00021901234567901233,
+      "loss": 6.3073,
+      "step": 463
+    },
+    {
+      "epoch": 1.0293954520244037,
+      "grad_norm": 0.20343808829784393,
+      "learning_rate": 0.0002187654320987654,
+      "loss": 6.1999,
+      "step": 464
+    },
+    {
+      "epoch": 1.0316139767054908,
+      "grad_norm": 0.2162073403596878,
+      "learning_rate": 0.0002185185185185185,
+      "loss": 6.3658,
+      "step": 465
+    },
+    {
+      "epoch": 1.033832501386578,
+      "grad_norm": 0.22319117188453674,
+      "learning_rate": 0.00021827160493827157,
+      "loss": 6.0411,
+      "step": 466
+    },
+    {
+      "epoch": 1.036051026067665,
+      "grad_norm": 0.1942998617887497,
+      "learning_rate": 0.00021802469135802466,
+      "loss": 6.1315,
+      "step": 467
+    },
+    {
+      "epoch": 1.038269550748752,
+      "grad_norm": 0.21322256326675415,
+      "learning_rate": 0.00021777777777777778,
+      "loss": 5.9546,
+      "step": 468
+    },
+    {
+      "epoch": 1.0404880754298391,
+      "grad_norm": 0.2036537081003189,
+      "learning_rate": 0.00021753086419753085,
+      "loss": 6.1129,
+      "step": 469
+    },
+    {
+      "epoch": 1.0427066001109262,
+      "grad_norm": 0.20604239404201508,
+      "learning_rate": 0.00021728395061728395,
+      "loss": 5.993,
+      "step": 470
+    },
+    {
+      "epoch": 1.0449251247920133,
+      "grad_norm": 0.1988423615694046,
+      "learning_rate": 0.00021703703703703701,
+      "loss": 5.9197,
+      "step": 471
+    },
+    {
+      "epoch": 1.0471436494731003,
+      "grad_norm": 0.2102552056312561,
+      "learning_rate": 0.0002167901234567901,
+      "loss": 6.164,
+      "step": 472
+    },
+    {
+      "epoch": 1.0493621741541874,
+      "grad_norm": 0.20812729001045227,
+      "learning_rate": 0.00021654320987654318,
+      "loss": 6.0933,
+      "step": 473
+    },
+    {
+      "epoch": 1.0515806988352745,
+      "grad_norm": 0.20782533288002014,
+      "learning_rate": 0.00021629629629629627,
+      "loss": 6.3353,
+      "step": 474
+    },
+    {
+      "epoch": 1.0537992235163616,
+      "grad_norm": 0.20195087790489197,
+      "learning_rate": 0.00021604938271604937,
+      "loss": 6.2814,
+      "step": 475
+    },
+    {
+      "epoch": 1.0560177481974486,
+      "grad_norm": 0.2682838439941406,
+      "learning_rate": 0.00021580246913580244,
+      "loss": 6.1996,
+      "step": 476
+    },
+    {
+      "epoch": 1.0582362728785357,
+      "grad_norm": 0.26408272981643677,
+      "learning_rate": 0.00021555555555555556,
+      "loss": 5.7039,
+      "step": 477
+    },
+    {
+      "epoch": 1.0604547975596228,
+      "grad_norm": 0.19157156348228455,
+      "learning_rate": 0.00021530864197530863,
+      "loss": 6.153,
+      "step": 478
+    },
+    {
+      "epoch": 1.0626733222407099,
+      "grad_norm": 0.1978674829006195,
+      "learning_rate": 0.00021506172839506172,
+      "loss": 6.189,
+      "step": 479
+    },
+    {
+      "epoch": 1.064891846921797,
+      "grad_norm": 0.26279377937316895,
+      "learning_rate": 0.0002148148148148148,
+      "loss": 6.3449,
+      "step": 480
+    },
+    {
+      "epoch": 1.067110371602884,
+      "grad_norm": 0.18888740241527557,
+      "learning_rate": 0.00021456790123456788,
+      "loss": 5.904,
+      "step": 481
+    },
+    {
+      "epoch": 1.069328896283971,
+      "grad_norm": 0.226004496216774,
+      "learning_rate": 0.00021432098765432095,
+      "loss": 6.1329,
+      "step": 482
+    },
+    {
+      "epoch": 1.0715474209650582,
+      "grad_norm": 0.2254335582256317,
+      "learning_rate": 0.00021407407407407405,
+      "loss": 6.2669,
+      "step": 483
+    },
+    {
+      "epoch": 1.0737659456461452,
+      "grad_norm": 0.2689749598503113,
+      "learning_rate": 0.00021382716049382714,
+      "loss": 6.3659,
+      "step": 484
+    },
+    {
+      "epoch": 1.0759844703272323,
+      "grad_norm": 0.21105527877807617,
+      "learning_rate": 0.0002135802469135802,
+      "loss": 5.9176,
+      "step": 485
+    },
+    {
+      "epoch": 1.0782029950083194,
+      "grad_norm": 0.22421985864639282,
+      "learning_rate": 0.00021333333333333333,
+      "loss": 6.4442,
+      "step": 486
+    },
+    {
+      "epoch": 1.0804215196894065,
+      "grad_norm": 0.25460535287857056,
+      "learning_rate": 0.0002130864197530864,
+      "loss": 6.2099,
+      "step": 487
+    },
+    {
+      "epoch": 1.0826400443704935,
+      "grad_norm": 0.2072068452835083,
+      "learning_rate": 0.0002128395061728395,
+      "loss": 6.1441,
+      "step": 488
+    },
+    {
+      "epoch": 1.0848585690515806,
+      "grad_norm": 0.24068021774291992,
+      "learning_rate": 0.00021259259259259256,
+      "loss": 6.3545,
+      "step": 489
+    },
+    {
+      "epoch": 1.0870770937326677,
+      "grad_norm": 0.18164944648742676,
+      "learning_rate": 0.00021234567901234566,
+      "loss": 6.2874,
+      "step": 490
+    },
+    {
+      "epoch": 1.0892956184137548,
+      "grad_norm": 0.31560930609703064,
+      "learning_rate": 0.00021209876543209875,
+      "loss": 6.0878,
+      "step": 491
+    },
+    {
+      "epoch": 1.0915141430948418,
+      "grad_norm": 0.2550957202911377,
+      "learning_rate": 0.00021185185185185182,
+      "loss": 6.2911,
+      "step": 492
+    },
+    {
+      "epoch": 1.093732667775929,
+      "grad_norm": 0.24728110432624817,
+      "learning_rate": 0.00021160493827160492,
+      "loss": 6.4203,
+      "step": 493
+    },
+    {
+      "epoch": 1.095951192457016,
+      "grad_norm": 0.20083194971084595,
+      "learning_rate": 0.00021135802469135799,
+      "loss": 5.9348,
+      "step": 494
+    },
+    {
+      "epoch": 1.098169717138103,
+      "grad_norm": 0.20149637758731842,
+      "learning_rate": 0.0002111111111111111,
+      "loss": 6.0973,
+      "step": 495
+    },
+    {
+      "epoch": 1.1003882418191901,
+      "grad_norm": 0.20398758351802826,
+      "learning_rate": 0.00021086419753086418,
+      "loss": 6.1436,
+      "step": 496
+    },
+    {
+      "epoch": 1.1026067665002772,
+      "grad_norm": 0.23233990371227264,
+      "learning_rate": 0.00021061728395061727,
+      "loss": 6.3519,
+      "step": 497
+    },
+    {
+      "epoch": 1.1048252911813643,
+      "grad_norm": 0.2554318606853485,
+      "learning_rate": 0.00021037037037037037,
+      "loss": 6.1499,
+      "step": 498
+    },
+    {
+      "epoch": 1.1070438158624514,
+      "grad_norm": 0.2432209700345993,
+      "learning_rate": 0.00021012345679012343,
+      "loss": 6.1209,
+      "step": 499
+    },
+    {
+      "epoch": 1.1092623405435384,
+      "grad_norm": 0.21404734253883362,
+      "learning_rate": 0.00020987654320987653,
+      "loss": 6.1482,
+      "step": 500
+    },
+    {
+      "epoch": 1.1114808652246255,
+      "grad_norm": 0.21774731576442719,
+      "learning_rate": 0.0002096296296296296,
+      "loss": 6.0395,
+      "step": 501
+    },
+    {
+      "epoch": 1.1136993899057126,
+      "grad_norm": 0.21658863127231598,
+      "learning_rate": 0.0002093827160493827,
+      "loss": 6.3416,
+      "step": 502
+    },
+    {
+      "epoch": 1.1159179145867997,
+      "grad_norm": 0.23965108394622803,
+      "learning_rate": 0.00020913580246913576,
+      "loss": 6.4375,
+      "step": 503
+    },
+    {
+      "epoch": 1.1181364392678868,
+      "grad_norm": 0.21659262478351593,
+      "learning_rate": 0.00020888888888888888,
+      "loss": 5.9621,
+      "step": 504
+    },
+    {
+      "epoch": 1.1203549639489738,
+      "grad_norm": 0.19859051704406738,
+      "learning_rate": 0.00020864197530864198,
+      "loss": 6.1204,
+      "step": 505
+    },
+    {
+      "epoch": 1.122573488630061,
+      "grad_norm": 0.18197152018547058,
+      "learning_rate": 0.00020839506172839505,
+      "loss": 6.0275,
+      "step": 506
+    },
+    {
+      "epoch": 1.124792013311148,
+      "grad_norm": 0.2217610627412796,
+      "learning_rate": 0.00020814814814814814,
+      "loss": 6.3442,
+      "step": 507
+    },
+    {
+      "epoch": 1.127010537992235,
+      "grad_norm": 0.2464493215084076,
+      "learning_rate": 0.0002079012345679012,
+      "loss": 5.7869,
+      "step": 508
+    },
+    {
+      "epoch": 1.1292290626733221,
+      "grad_norm": 0.2446204274892807,
+      "learning_rate": 0.0002076543209876543,
+      "loss": 6.1157,
+      "step": 509
+    },
+    {
+      "epoch": 1.1314475873544092,
+      "grad_norm": 0.18952661752700806,
+      "learning_rate": 0.00020740740740740737,
+      "loss": 6.1021,
+      "step": 510
+    },
+    {
+      "epoch": 1.1336661120354963,
+      "grad_norm": 0.27472108602523804,
+      "learning_rate": 0.00020716049382716047,
+      "loss": 6.4953,
+      "step": 511
+    },
+    {
+      "epoch": 1.1358846367165834,
+      "grad_norm": 0.26027482748031616,
+      "learning_rate": 0.00020691358024691354,
+      "loss": 6.0857,
+      "step": 512
+    },
+    {
+      "epoch": 1.1381031613976704,
+      "grad_norm": 0.23206189274787903,
+      "learning_rate": 0.00020666666666666666,
+      "loss": 6.0705,
+      "step": 513
+    },
+    {
+      "epoch": 1.1403216860787575,
+      "grad_norm": 0.2878470718860626,
+      "learning_rate": 0.00020641975308641975,
+      "loss": 5.6963,
+      "step": 514
+    },
+    {
+      "epoch": 1.1425402107598448,
+      "grad_norm": 0.28168272972106934,
+      "learning_rate": 0.00020617283950617282,
+      "loss": 6.0005,
+      "step": 515
+    },
+    {
+      "epoch": 1.1447587354409319,
+      "grad_norm": 0.2529776990413666,
+      "learning_rate": 0.00020592592592592592,
+      "loss": 6.5515,
+      "step": 516
+    },
+    {
+      "epoch": 1.146977260122019,
+      "grad_norm": 0.24619634449481964,
+      "learning_rate": 0.00020567901234567898,
+      "loss": 6.2221,
+      "step": 517
+    },
+    {
+      "epoch": 1.149195784803106,
+      "grad_norm": 0.2685058116912842,
+      "learning_rate": 0.00020543209876543208,
+      "loss": 6.0284,
+      "step": 518
+    },
+    {
+      "epoch": 1.151414309484193,
+      "grad_norm": 0.19598890841007233,
+      "learning_rate": 0.00020518518518518515,
+      "loss": 6.1202,
+      "step": 519
+    },
+    {
+      "epoch": 1.1536328341652802,
+      "grad_norm": 0.21533448994159698,
+      "learning_rate": 0.00020493827160493824,
+      "loss": 5.8771,
+      "step": 520
+    },
+    {
+      "epoch": 1.1558513588463672,
+      "grad_norm": 0.2333855926990509,
+      "learning_rate": 0.00020469135802469136,
+      "loss": 6.0801,
+      "step": 521
+    },
+    {
+      "epoch": 1.1580698835274543,
+      "grad_norm": 0.2213953584432602,
+      "learning_rate": 0.00020444444444444443,
+      "loss": 6.0442,
+      "step": 522
+    },
+    {
+      "epoch": 1.1602884082085414,
+      "grad_norm": 0.25825896859169006,
+      "learning_rate": 0.00020419753086419753,
+      "loss": 6.0617,
+      "step": 523
+    },
+    {
+      "epoch": 1.1625069328896285,
+      "grad_norm": 0.21634967625141144,
+      "learning_rate": 0.0002039506172839506,
+      "loss": 6.1176,
+      "step": 524
+    },
+    {
+      "epoch": 1.1647254575707155,
+      "grad_norm": 0.24216841161251068,
+      "learning_rate": 0.0002037037037037037,
+      "loss": 6.0726,
+      "step": 525
+    },
+    {
+      "epoch": 1.1669439822518026,
+      "grad_norm": 0.22673873603343964,
+      "learning_rate": 0.00020345679012345676,
+      "loss": 6.0141,
+      "step": 526
+    },
+    {
+      "epoch": 1.1691625069328897,
+      "grad_norm": 0.2144697904586792,
+      "learning_rate": 0.00020320987654320985,
+      "loss": 5.8762,
+      "step": 527
+    },
+    {
+      "epoch": 1.1713810316139768,
+      "grad_norm": 0.292371928691864,
+      "learning_rate": 0.00020296296296296295,
+      "loss": 6.3492,
+      "step": 528
+    },
+    {
+      "epoch": 1.1735995562950639,
+      "grad_norm": 0.2608895003795624,
+      "learning_rate": 0.00020271604938271602,
+      "loss": 6.0569,
+      "step": 529
+    },
+    {
+      "epoch": 1.175818080976151,
+      "grad_norm": 0.2051520049571991,
+      "learning_rate": 0.00020246913580246914,
+      "loss": 6.1281,
+      "step": 530
+    },
+    {
+      "epoch": 1.178036605657238,
+      "grad_norm": 0.24821048974990845,
+      "learning_rate": 0.0002022222222222222,
+      "loss": 5.9842,
+      "step": 531
+    },
+    {
+      "epoch": 1.180255130338325,
+      "grad_norm": 0.2552020251750946,
+      "learning_rate": 0.0002019753086419753,
+      "loss": 6.1525,
+      "step": 532
+    },
+    {
+      "epoch": 1.1824736550194122,
+      "grad_norm": 0.2671736478805542,
+      "learning_rate": 0.00020172839506172837,
+      "loss": 5.6541,
+      "step": 533
+    },
+    {
+      "epoch": 1.1846921797004992,
+      "grad_norm": 0.21748094260692596,
+      "learning_rate": 0.00020148148148148147,
+      "loss": 5.8334,
+      "step": 534
+    },
+    {
+      "epoch": 1.1869107043815863,
+      "grad_norm": 0.27664902806282043,
+      "learning_rate": 0.00020123456790123456,
+      "loss": 6.1332,
+      "step": 535
+    },
+    {
+      "epoch": 1.1891292290626734,
+      "grad_norm": 0.20520375669002533,
+      "learning_rate": 0.00020098765432098763,
+      "loss": 5.7822,
+      "step": 536
+    },
+    {
+      "epoch": 1.1913477537437605,
+      "grad_norm": 0.22293919324874878,
+      "learning_rate": 0.00020074074074074072,
+      "loss": 5.932,
+      "step": 537
+    },
+    {
+      "epoch": 1.1935662784248475,
+      "grad_norm": 0.20420442521572113,
+      "learning_rate": 0.0002004938271604938,
+      "loss": 6.0208,
+      "step": 538
+    },
+    {
+      "epoch": 1.1957848031059346,
+      "grad_norm": 0.22204145789146423,
+      "learning_rate": 0.00020024691358024691,
+      "loss": 6.2548,
+      "step": 539
+    },
+    {
+      "epoch": 1.1980033277870217,
+      "grad_norm": 0.236099973320961,
+      "learning_rate": 0.00019999999999999998,
+      "loss": 5.8519,
+      "step": 540
+    },
+    {
+      "epoch": 1.2002218524681088,
+      "grad_norm": 0.3160380721092224,
+      "learning_rate": 0.00019975308641975308,
+      "loss": 6.04,
+      "step": 541
+    },
+    {
+      "epoch": 1.2024403771491958,
+      "grad_norm": 0.21357060968875885,
+      "learning_rate": 0.00019950617283950615,
+      "loss": 6.154,
+      "step": 542
+    },
+    {
+      "epoch": 1.204658901830283,
+      "grad_norm": 0.2152855545282364,
+      "learning_rate": 0.00019925925925925924,
+      "loss": 6.0944,
+      "step": 543
+    },
+    {
+      "epoch": 1.20687742651137,
+      "grad_norm": 0.2307923138141632,
+      "learning_rate": 0.00019901234567901234,
+      "loss": 6.0277,
+      "step": 544
+    },
+    {
+      "epoch": 1.209095951192457,
+      "grad_norm": 0.31007638573646545,
+      "learning_rate": 0.0001987654320987654,
+      "loss": 5.8482,
+      "step": 545
+    },
+    {
+      "epoch": 1.2113144758735441,
+      "grad_norm": 0.2275347113609314,
+      "learning_rate": 0.0001985185185185185,
+      "loss": 5.8623,
+      "step": 546
+    },
+    {
+      "epoch": 1.2135330005546312,
+      "grad_norm": 0.216464102268219,
+      "learning_rate": 0.00019827160493827157,
+      "loss": 5.7501,
+      "step": 547
+    },
+    {
+      "epoch": 1.2157515252357183,
+      "grad_norm": 0.31815075874328613,
+      "learning_rate": 0.0001980246913580247,
+      "loss": 5.8252,
+      "step": 548
+    },
+    {
+      "epoch": 1.2179700499168054,
+      "grad_norm": 0.2610233724117279,
+      "learning_rate": 0.00019777777777777776,
+      "loss": 6.2653,
+      "step": 549
+    },
+    {
+      "epoch": 1.2201885745978924,
+      "grad_norm": 0.2641243040561676,
+      "learning_rate": 0.00019753086419753085,
+      "loss": 5.8609,
+      "step": 550
+    },
+    {
+      "epoch": 1.2224070992789795,
+      "grad_norm": 0.20771335065364838,
+      "learning_rate": 0.00019728395061728395,
+      "loss": 6.0489,
+      "step": 551
+    },
+    {
+      "epoch": 1.2246256239600666,
+      "grad_norm": 0.3027734160423279,
+      "learning_rate": 0.00019703703703703702,
+      "loss": 6.0033,
+      "step": 552
+    },
+    {
+      "epoch": 1.2268441486411537,
+      "grad_norm": 0.2504933476448059,
+      "learning_rate": 0.0001967901234567901,
+      "loss": 5.9858,
+      "step": 553
+    },
+    {
+      "epoch": 1.2290626733222407,
+      "grad_norm": 0.2636372148990631,
+      "learning_rate": 0.00019654320987654318,
+      "loss": 6.0741,
+      "step": 554
+    },
+    {
+      "epoch": 1.2312811980033278,
+      "grad_norm": 0.189767524600029,
+      "learning_rate": 0.00019629629629629627,
+      "loss": 5.9111,
+      "step": 555
+    },
+    {
+      "epoch": 1.2334997226844149,
+      "grad_norm": 0.2583034336566925,
+      "learning_rate": 0.00019604938271604934,
+      "loss": 5.9604,
+      "step": 556
+    },
+    {
+      "epoch": 1.235718247365502,
+      "grad_norm": 0.2803707718849182,
+      "learning_rate": 0.00019580246913580246,
+      "loss": 6.0876,
+      "step": 557
+    },
+    {
+      "epoch": 1.237936772046589,
+      "grad_norm": 0.20788563787937164,
+      "learning_rate": 0.00019555555555555556,
+      "loss": 5.6963,
+      "step": 558
+    },
+    {
+      "epoch": 1.240155296727676,
+      "grad_norm": 0.22030335664749146,
+      "learning_rate": 0.00019530864197530863,
+      "loss": 6.0868,
+      "step": 559
+    },
+    {
+      "epoch": 1.2423738214087632,
+      "grad_norm": 0.22802013158798218,
+      "learning_rate": 0.00019506172839506172,
+      "loss": 5.96,
+      "step": 560
+    },
+    {
+      "epoch": 1.2445923460898503,
+      "grad_norm": 0.22124886512756348,
+      "learning_rate": 0.0001948148148148148,
+      "loss": 5.5719,
+      "step": 561
+    },
+    {
+      "epoch": 1.2468108707709373,
+      "grad_norm": 0.2793128192424774,
+      "learning_rate": 0.00019456790123456789,
+      "loss": 6.083,
+      "step": 562
+    },
+    {
+      "epoch": 1.2490293954520244,
+      "grad_norm": 0.2232375293970108,
+      "learning_rate": 0.00019432098765432095,
+      "loss": 5.9478,
+      "step": 563
+    },
+    {
+      "epoch": 1.2512479201331115,
+      "grad_norm": 0.21672700345516205,
+      "learning_rate": 0.00019407407407407405,
+      "loss": 5.9364,
+      "step": 564
+    },
+    {
+      "epoch": 1.2534664448141986,
+      "grad_norm": 0.20991209149360657,
+      "learning_rate": 0.00019382716049382712,
+      "loss": 6.0498,
+      "step": 565
+    },
+    {
+      "epoch": 1.2556849694952856,
+      "grad_norm": 0.23595640063285828,
+      "learning_rate": 0.00019358024691358024,
+      "loss": 6.1763,
+      "step": 566
+    },
+    {
+      "epoch": 1.2579034941763727,
+      "grad_norm": 0.20590583980083466,
+      "learning_rate": 0.00019333333333333333,
+      "loss": 6.1026,
+      "step": 567
+    },
+    {
+      "epoch": 1.2601220188574598,
+      "grad_norm": 0.24334143102169037,
+      "learning_rate": 0.0001930864197530864,
+      "loss": 5.8401,
+      "step": 568
+    },
+    {
+      "epoch": 1.2623405435385469,
+      "grad_norm": 0.204694002866745,
+      "learning_rate": 0.0001928395061728395,
+      "loss": 5.9312,
+      "step": 569
+    },
+    {
+      "epoch": 1.264559068219634,
+      "grad_norm": 0.20970626175403595,
+      "learning_rate": 0.00019259259259259257,
+      "loss": 5.8068,
+      "step": 570
+    },
+    {
+      "epoch": 1.266777592900721,
+      "grad_norm": 0.24134007096290588,
+      "learning_rate": 0.00019234567901234566,
+      "loss": 5.6104,
+      "step": 571
+    },
+    {
+      "epoch": 1.268996117581808,
+      "grad_norm": 0.20575691759586334,
+      "learning_rate": 0.00019209876543209873,
+      "loss": 6.0975,
+      "step": 572
+    },
+    {
+      "epoch": 1.2712146422628952,
+      "grad_norm": 0.2341838926076889,
+      "learning_rate": 0.00019185185185185182,
+      "loss": 5.9888,
+      "step": 573
+    },
+    {
+      "epoch": 1.2734331669439822,
+      "grad_norm": 0.26650622487068176,
+      "learning_rate": 0.00019160493827160495,
+      "loss": 5.8473,
+      "step": 574
+    },
+    {
+      "epoch": 1.2756516916250693,
+      "grad_norm": 0.2761963903903961,
+      "learning_rate": 0.00019135802469135801,
+      "loss": 6.2065,
+      "step": 575
+    },
+    {
+      "epoch": 1.2778702163061564,
+      "grad_norm": 0.24920348823070526,
+      "learning_rate": 0.0001911111111111111,
+      "loss": 5.547,
+      "step": 576
+    },
+    {
+      "epoch": 1.2800887409872435,
+      "grad_norm": 0.2553333044052124,
+      "learning_rate": 0.00019086419753086418,
+      "loss": 6.2321,
+      "step": 577
+    },
+    {
+      "epoch": 1.2823072656683305,
+      "grad_norm": 0.20263828337192535,
+      "learning_rate": 0.00019061728395061727,
+      "loss": 6.0229,
+      "step": 578
+    },
+    {
+      "epoch": 1.2845257903494176,
+      "grad_norm": 0.2541116774082184,
+      "learning_rate": 0.00019037037037037034,
+      "loss": 5.7055,
+      "step": 579
+    },
+    {
+      "epoch": 1.2867443150305047,
+      "grad_norm": 0.291706383228302,
+      "learning_rate": 0.00019012345679012344,
+      "loss": 5.7731,
+      "step": 580
+    },
+    {
+      "epoch": 1.2889628397115918,
+      "grad_norm": 0.2801738381385803,
+      "learning_rate": 0.00018987654320987653,
+      "loss": 6.0944,
+      "step": 581
+    },
+    {
+      "epoch": 1.2911813643926788,
+      "grad_norm": 0.17551471292972565,
+      "learning_rate": 0.0001896296296296296,
+      "loss": 5.9055,
+      "step": 582
+    },
+    {
+      "epoch": 1.293399889073766,
+      "grad_norm": 0.24934734404087067,
+      "learning_rate": 0.00018938271604938272,
+      "loss": 5.5106,
+      "step": 583
+    },
+    {
+      "epoch": 1.295618413754853,
+      "grad_norm": 0.23498648405075073,
+      "learning_rate": 0.0001891358024691358,
+      "loss": 5.8927,
+      "step": 584
+    },
+    {
+      "epoch": 1.29783693843594,
+      "grad_norm": 0.1973291039466858,
+      "learning_rate": 0.00018888888888888888,
+      "loss": 6.0334,
+      "step": 585
+    },
+    {
+      "epoch": 1.3000554631170271,
+      "grad_norm": 0.22123892605304718,
+      "learning_rate": 0.00018864197530864195,
+      "loss": 5.8937,
+      "step": 586
+    },
+    {
+      "epoch": 1.3022739877981142,
+      "grad_norm": 0.20009583234786987,
+      "learning_rate": 0.00018839506172839505,
+      "loss": 6.0609,
+      "step": 587
+    },
+    {
+      "epoch": 1.3044925124792013,
+      "grad_norm": 0.20208057761192322,
+      "learning_rate": 0.00018814814814814814,
+      "loss": 6.0692,
+      "step": 588
+    },
+    {
+      "epoch": 1.3067110371602884,
+      "grad_norm": 0.2744377553462982,
+      "learning_rate": 0.0001879012345679012,
+      "loss": 6.0541,
+      "step": 589
+    },
+    {
+      "epoch": 1.3089295618413754,
+      "grad_norm": 0.2372061163187027,
+      "learning_rate": 0.0001876543209876543,
+      "loss": 5.7593,
+      "step": 590
+    },
+    {
+      "epoch": 1.3111480865224625,
+      "grad_norm": 0.23008209466934204,
+      "learning_rate": 0.00018740740740740737,
+      "loss": 6.0837,
+      "step": 591
+    },
+    {
+      "epoch": 1.3133666112035496,
+      "grad_norm": 0.27167603373527527,
+      "learning_rate": 0.0001871604938271605,
+      "loss": 5.7787,
+      "step": 592
+    },
+    {
+      "epoch": 1.3155851358846367,
+      "grad_norm": 0.24951684474945068,
+      "learning_rate": 0.00018691358024691356,
+      "loss": 5.9249,
+      "step": 593
+    },
+    {
+      "epoch": 1.3178036605657237,
+      "grad_norm": 0.2166278064250946,
+      "learning_rate": 0.00018666666666666666,
+      "loss": 6.15,
+      "step": 594
+    },
+    {
+      "epoch": 1.3200221852468108,
+      "grad_norm": 0.2808057367801666,
+      "learning_rate": 0.00018641975308641973,
+      "loss": 6.0342,
+      "step": 595
+    },
+    {
+      "epoch": 1.3222407099278979,
+      "grad_norm": 0.23195438086986542,
+      "learning_rate": 0.00018617283950617282,
+      "loss": 5.9864,
+      "step": 596
+    },
+    {
+      "epoch": 1.324459234608985,
+      "grad_norm": 0.22163689136505127,
+      "learning_rate": 0.00018592592592592592,
+      "loss": 5.6793,
+      "step": 597
+    },
+    {
+      "epoch": 1.326677759290072,
+      "grad_norm": 0.3342806100845337,
+      "learning_rate": 0.00018567901234567899,
+      "loss": 5.7509,
+      "step": 598
+    },
+    {
+      "epoch": 1.328896283971159,
+      "grad_norm": 0.25519901514053345,
+      "learning_rate": 0.00018543209876543208,
+      "loss": 5.7018,
+      "step": 599
+    },
+    {
+      "epoch": 1.3311148086522462,
+      "grad_norm": 0.220490500330925,
+      "learning_rate": 0.00018518518518518515,
+      "loss": 5.797,
+      "step": 600
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 0.2441834956407547,
+      "learning_rate": 0.00018493827160493827,
+      "loss": 5.8496,
+      "step": 601
+    },
+    {
+      "epoch": 1.3355518580144203,
+      "grad_norm": 0.21132062375545502,
+      "learning_rate": 0.00018469135802469134,
+      "loss": 6.0636,
+      "step": 602
+    },
+    {
+      "epoch": 1.3377703826955074,
+      "grad_norm": 0.1960887461900711,
+      "learning_rate": 0.00018444444444444443,
+      "loss": 5.8444,
+      "step": 603
+    },
+    {
+      "epoch": 1.3399889073765945,
+      "grad_norm": 0.27217400074005127,
+      "learning_rate": 0.00018419753086419753,
+      "loss": 6.4093,
+      "step": 604
+    },
+    {
+      "epoch": 1.3422074320576816,
+      "grad_norm": 0.21061958372592926,
+      "learning_rate": 0.0001839506172839506,
+      "loss": 6.0336,
+      "step": 605
+    },
+    {
+      "epoch": 1.3444259567387689,
+      "grad_norm": 0.28332802653312683,
+      "learning_rate": 0.0001837037037037037,
+      "loss": 6.0702,
+      "step": 606
+    },
+    {
+      "epoch": 1.346644481419856,
+      "grad_norm": 0.25394898653030396,
+      "learning_rate": 0.00018345679012345676,
+      "loss": 6.1692,
+      "step": 607
+    },
+    {
+      "epoch": 1.348863006100943,
+      "grad_norm": 0.24195651710033417,
+      "learning_rate": 0.00018320987654320986,
+      "loss": 5.715,
+      "step": 608
+    },
+    {
+      "epoch": 1.35108153078203,
+      "grad_norm": 0.24449539184570312,
+      "learning_rate": 0.00018296296296296292,
+      "loss": 6.0665,
+      "step": 609
+    },
+    {
+      "epoch": 1.3533000554631172,
+      "grad_norm": 0.26209670305252075,
+      "learning_rate": 0.00018271604938271605,
+      "loss": 5.6949,
+      "step": 610
+    },
+    {
+      "epoch": 1.3555185801442042,
+      "grad_norm": 0.22789019346237183,
+      "learning_rate": 0.00018246913580246914,
+      "loss": 5.8878,
+      "step": 611
+    },
+    {
+      "epoch": 1.3577371048252913,
+      "grad_norm": 0.3141170144081116,
+      "learning_rate": 0.0001822222222222222,
+      "loss": 5.9607,
+      "step": 612
+    },
+    {
+      "epoch": 1.3599556295063784,
+      "grad_norm": 0.24164436757564545,
+      "learning_rate": 0.0001819753086419753,
+      "loss": 5.656,
+      "step": 613
+    },
+    {
+      "epoch": 1.3621741541874655,
+      "grad_norm": 0.2335316389799118,
+      "learning_rate": 0.00018172839506172837,
+      "loss": 5.9465,
+      "step": 614
+    },
+    {
+      "epoch": 1.3643926788685525,
+      "grad_norm": 0.33491250872612,
+      "learning_rate": 0.00018148148148148147,
+      "loss": 5.9443,
+      "step": 615
+    },
+    {
+      "epoch": 1.3666112035496396,
+      "grad_norm": 0.2319815754890442,
+      "learning_rate": 0.00018123456790123454,
+      "loss": 6.0288,
+      "step": 616
+    },
+    {
+      "epoch": 1.3688297282307267,
+      "grad_norm": 0.28581365942955017,
+      "learning_rate": 0.00018098765432098763,
+      "loss": 5.9475,
+      "step": 617
+    },
+    {
+      "epoch": 1.3710482529118138,
+      "grad_norm": 0.22758707404136658,
+      "learning_rate": 0.00018074074074074075,
+      "loss": 6.0384,
+      "step": 618
+    },
+    {
+      "epoch": 1.3732667775929008,
+      "grad_norm": 0.27819007635116577,
+      "learning_rate": 0.00018049382716049382,
+      "loss": 5.7041,
+      "step": 619
+    },
+    {
+      "epoch": 1.375485302273988,
+      "grad_norm": 0.21772634983062744,
+      "learning_rate": 0.00018024691358024692,
+      "loss": 5.8114,
+      "step": 620
+    },
+    {
+      "epoch": 1.377703826955075,
+      "grad_norm": 0.23647594451904297,
+      "learning_rate": 0.00017999999999999998,
+      "loss": 6.0323,
+      "step": 621
+    },
+    {
+      "epoch": 1.379922351636162,
+      "grad_norm": 0.321921169757843,
+      "learning_rate": 0.00017975308641975308,
+      "loss": 5.7722,
+      "step": 622
+    },
+    {
+      "epoch": 1.3821408763172491,
+      "grad_norm": 0.2796967923641205,
+      "learning_rate": 0.00017950617283950615,
+      "loss": 5.976,
+      "step": 623
+    },
+    {
+      "epoch": 1.3843594009983362,
+      "grad_norm": 0.31009525060653687,
+      "learning_rate": 0.00017925925925925924,
+      "loss": 5.7344,
+      "step": 624
+    },
+    {
+      "epoch": 1.3865779256794233,
+      "grad_norm": 0.25374799966812134,
+      "learning_rate": 0.0001790123456790123,
+      "loss": 5.901,
+      "step": 625
+    },
+    {
+      "epoch": 1.3887964503605104,
+      "grad_norm": 0.24280957877635956,
+      "learning_rate": 0.0001787654320987654,
+      "loss": 5.5368,
+      "step": 626
+    },
+    {
+      "epoch": 1.3910149750415974,
+      "grad_norm": 0.26173558831214905,
+      "learning_rate": 0.00017851851851851853,
+      "loss": 5.5704,
+      "step": 627
+    },
+    {
+      "epoch": 1.3932334997226845,
+      "grad_norm": 0.26110395789146423,
+      "learning_rate": 0.0001782716049382716,
+      "loss": 5.9234,
+      "step": 628
+    },
+    {
+      "epoch": 1.3954520244037716,
+      "grad_norm": 0.33118900656700134,
+      "learning_rate": 0.0001780246913580247,
+      "loss": 5.4649,
+      "step": 629
+    },
+    {
+      "epoch": 1.3976705490848587,
+      "grad_norm": 0.23458221554756165,
+      "learning_rate": 0.00017777777777777776,
+      "loss": 5.8983,
+      "step": 630
+    },
+    {
+      "epoch": 1.3998890737659457,
+      "grad_norm": 0.21687628328800201,
+      "learning_rate": 0.00017753086419753085,
+      "loss": 5.8511,
+      "step": 631
+    },
+    {
+      "epoch": 1.4021075984470328,
+      "grad_norm": 0.25080716609954834,
+      "learning_rate": 0.00017728395061728392,
+      "loss": 5.7728,
+      "step": 632
+    },
+    {
+      "epoch": 1.4043261231281199,
+      "grad_norm": 0.24884070456027985,
+      "learning_rate": 0.00017703703703703702,
+      "loss": 5.7798,
+      "step": 633
+    },
+    {
+      "epoch": 1.406544647809207,
+      "grad_norm": 0.23810969293117523,
+      "learning_rate": 0.0001767901234567901,
+      "loss": 5.6819,
+      "step": 634
+    },
+    {
+      "epoch": 1.408763172490294,
+      "grad_norm": 0.2846151888370514,
+      "learning_rate": 0.00017654320987654318,
+      "loss": 5.8561,
+      "step": 635
+    },
+    {
+      "epoch": 1.410981697171381,
+      "grad_norm": 0.27155426144599915,
+      "learning_rate": 0.0001762962962962963,
+      "loss": 5.7175,
+      "step": 636
+    },
+    {
+      "epoch": 1.4132002218524682,
+      "grad_norm": 0.2543492019176483,
+      "learning_rate": 0.00017604938271604937,
+      "loss": 5.7313,
+      "step": 637
+    },
+    {
+      "epoch": 1.4154187465335553,
+      "grad_norm": 0.27509814500808716,
+      "learning_rate": 0.00017580246913580247,
+      "loss": 6.0105,
+      "step": 638
+    },
+    {
+      "epoch": 1.4176372712146423,
+      "grad_norm": 0.22599969804286957,
+      "learning_rate": 0.00017555555555555553,
+      "loss": 5.8201,
+      "step": 639
+    },
+    {
+      "epoch": 1.4198557958957294,
+      "grad_norm": 0.3042387366294861,
+      "learning_rate": 0.00017530864197530863,
+      "loss": 5.8365,
+      "step": 640
+    },
+    {
+      "epoch": 1.4220743205768165,
+      "grad_norm": 0.2555197775363922,
+      "learning_rate": 0.00017506172839506172,
+      "loss": 5.9536,
+      "step": 641
+    },
+    {
+      "epoch": 1.4242928452579036,
+      "grad_norm": 0.3498784899711609,
+      "learning_rate": 0.0001748148148148148,
+      "loss": 5.673,
+      "step": 642
+    },
+    {
+      "epoch": 1.4265113699389906,
+      "grad_norm": 0.28615400195121765,
+      "learning_rate": 0.0001745679012345679,
+      "loss": 5.7292,
+      "step": 643
+    },
+    {
+      "epoch": 1.4287298946200777,
+      "grad_norm": 0.20748300850391388,
+      "learning_rate": 0.00017432098765432096,
+      "loss": 5.8704,
+      "step": 644
+    },
+    {
+      "epoch": 1.4309484193011648,
+      "grad_norm": 0.41079720854759216,
+      "learning_rate": 0.00017407407407407408,
+      "loss": 5.9406,
+      "step": 645
+    },
+    {
+      "epoch": 1.4331669439822519,
+      "grad_norm": 0.30695968866348267,
+      "learning_rate": 0.00017382716049382715,
+      "loss": 5.8726,
+      "step": 646
+    },
+    {
+      "epoch": 1.435385468663339,
+      "grad_norm": 0.2608798146247864,
+      "learning_rate": 0.00017358024691358024,
+      "loss": 5.9359,
+      "step": 647
+    },
+    {
+      "epoch": 1.437603993344426,
+      "grad_norm": 0.26002204418182373,
+      "learning_rate": 0.0001733333333333333,
+      "loss": 5.7162,
+      "step": 648
+    },
+    {
+      "epoch": 1.439822518025513,
+      "grad_norm": 0.2675764262676239,
+      "learning_rate": 0.0001730864197530864,
+      "loss": 5.9239,
+      "step": 649
+    },
+    {
+      "epoch": 1.4420410427066002,
+      "grad_norm": 0.2525325417518616,
+      "learning_rate": 0.0001728395061728395,
+      "loss": 5.8598,
+      "step": 650
+    },
+    {
+      "epoch": 1.4442595673876872,
+      "grad_norm": 0.2752697765827179,
+      "learning_rate": 0.00017259259259259257,
+      "loss": 6.0397,
+      "step": 651
+    },
+    {
+      "epoch": 1.4464780920687743,
+      "grad_norm": 0.405160129070282,
+      "learning_rate": 0.00017234567901234566,
+      "loss": 5.6497,
+      "step": 652
+    },
+    {
+      "epoch": 1.4486966167498614,
+      "grad_norm": 0.2556341588497162,
+      "learning_rate": 0.00017209876543209873,
+      "loss": 5.7416,
+      "step": 653
+    },
+    {
+      "epoch": 1.4509151414309485,
+      "grad_norm": 0.328778475522995,
+      "learning_rate": 0.00017185185185185185,
+      "loss": 5.6667,
+      "step": 654
+    },
+    {
+      "epoch": 1.4531336661120355,
+      "grad_norm": 0.28422367572784424,
+      "learning_rate": 0.00017160493827160492,
+      "loss": 5.6271,
+      "step": 655
+    },
+    {
+      "epoch": 1.4553521907931226,
+      "grad_norm": 0.2538285553455353,
+      "learning_rate": 0.00017135802469135802,
+      "loss": 5.6001,
+      "step": 656
+    },
+    {
+      "epoch": 1.4575707154742097,
+      "grad_norm": 0.2799350321292877,
+      "learning_rate": 0.0001711111111111111,
+      "loss": 5.5445,
+      "step": 657
+    },
+    {
+      "epoch": 1.4597892401552968,
+      "grad_norm": 0.2248498648405075,
+      "learning_rate": 0.00017086419753086418,
+      "loss": 5.8009,
+      "step": 658
+    },
+    {
+      "epoch": 1.4620077648363838,
+      "grad_norm": 0.28957173228263855,
+      "learning_rate": 0.00017061728395061727,
+      "loss": 6.0301,
+      "step": 659
+    },
+    {
+      "epoch": 1.464226289517471,
+      "grad_norm": 0.28417953848838806,
+      "learning_rate": 0.00017037037037037034,
+      "loss": 5.5013,
+      "step": 660
+    },
+    {
+      "epoch": 1.466444814198558,
+      "grad_norm": 0.23794357478618622,
+      "learning_rate": 0.00017012345679012344,
+      "loss": 5.8855,
+      "step": 661
+    },
+    {
+      "epoch": 1.468663338879645,
+      "grad_norm": 0.28338712453842163,
+      "learning_rate": 0.0001698765432098765,
+      "loss": 5.6995,
+      "step": 662
+    },
+    {
+      "epoch": 1.4708818635607321,
+      "grad_norm": 0.4475978910923004,
+      "learning_rate": 0.00016962962962962963,
+      "loss": 6.0014,
+      "step": 663
+    },
+    {
+      "epoch": 1.4731003882418192,
+      "grad_norm": 0.2130076140165329,
+      "learning_rate": 0.00016938271604938272,
+      "loss": 5.7099,
+      "step": 664
+    },
+    {
+      "epoch": 1.4753189129229063,
+      "grad_norm": 0.4895320534706116,
+      "learning_rate": 0.0001691358024691358,
+      "loss": 5.9294,
+      "step": 665
+    },
+    {
+      "epoch": 1.4775374376039934,
+      "grad_norm": 0.38744571805000305,
+      "learning_rate": 0.00016888888888888889,
+      "loss": 6.0661,
+      "step": 666
+    },
+    {
+      "epoch": 1.4797559622850804,
+      "grad_norm": 0.27259376645088196,
+      "learning_rate": 0.00016864197530864195,
+      "loss": 5.9862,
+      "step": 667
+    },
+    {
+      "epoch": 1.4819744869661675,
+      "grad_norm": 0.25079137086868286,
+      "learning_rate": 0.00016839506172839505,
+      "loss": 5.7392,
+      "step": 668
+    },
+    {
+      "epoch": 1.4841930116472546,
+      "grad_norm": 0.24778392910957336,
+      "learning_rate": 0.00016814814814814812,
+      "loss": 5.5795,
+      "step": 669
+    },
+    {
+      "epoch": 1.4864115363283417,
+      "grad_norm": 0.2550187110900879,
+      "learning_rate": 0.0001679012345679012,
+      "loss": 5.6606,
+      "step": 670
+    },
+    {
+      "epoch": 1.4886300610094287,
+      "grad_norm": 0.24944689869880676,
+      "learning_rate": 0.00016765432098765433,
+      "loss": 5.7174,
+      "step": 671
+    },
+    {
+      "epoch": 1.4908485856905158,
+      "grad_norm": 0.29522982239723206,
+      "learning_rate": 0.0001674074074074074,
+      "loss": 5.7471,
+      "step": 672
+    },
+    {
+      "epoch": 1.493067110371603,
+      "grad_norm": 0.2295566350221634,
+      "learning_rate": 0.0001671604938271605,
+      "loss": 5.8346,
+      "step": 673
+    },
+    {
+      "epoch": 1.49528563505269,
+      "grad_norm": 0.2597060799598694,
+      "learning_rate": 0.00016691358024691357,
+      "loss": 5.8357,
+      "step": 674
+    },
+    {
+      "epoch": 1.497504159733777,
+      "grad_norm": 0.224629208445549,
+      "learning_rate": 0.00016666666666666666,
+      "loss": 5.766,
+      "step": 675
+    },
+    {
+      "epoch": 1.4997226844148641,
+      "grad_norm": 0.28246065974235535,
+      "learning_rate": 0.00016641975308641973,
+      "loss": 5.9547,
+      "step": 676
+    },
+    {
+      "epoch": 1.5019412090959512,
+      "grad_norm": 0.2420993447303772,
+      "learning_rate": 0.00016617283950617282,
+      "loss": 5.9778,
+      "step": 677
+    },
+    {
+      "epoch": 1.5041597337770383,
+      "grad_norm": 0.229862779378891,
+      "learning_rate": 0.0001659259259259259,
+      "loss": 5.4661,
+      "step": 678
+    },
+    {
+      "epoch": 1.5063782584581253,
+      "grad_norm": 0.2690744698047638,
+      "learning_rate": 0.000165679012345679,
+      "loss": 5.6395,
+      "step": 679
+    },
+    {
+      "epoch": 1.5085967831392124,
+      "grad_norm": 0.2714030146598816,
+      "learning_rate": 0.0001654320987654321,
+      "loss": 5.9289,
+      "step": 680
+    },
+    {
+      "epoch": 1.5108153078202995,
+      "grad_norm": 0.292216956615448,
+      "learning_rate": 0.00016518518518518518,
+      "loss": 6.0026,
+      "step": 681
+    },
+    {
+      "epoch": 1.5130338325013866,
+      "grad_norm": 0.3228589594364166,
+      "learning_rate": 0.00016493827160493827,
+      "loss": 5.6148,
+      "step": 682
+    },
+    {
+      "epoch": 1.5152523571824736,
+      "grad_norm": 0.2871633768081665,
+      "learning_rate": 0.00016469135802469134,
+      "loss": 5.7366,
+      "step": 683
+    },
+    {
+      "epoch": 1.5174708818635607,
+      "grad_norm": 0.2201642245054245,
+      "learning_rate": 0.00016444444444444444,
+      "loss": 5.74,
+      "step": 684
+    },
+    {
+      "epoch": 1.5196894065446478,
+      "grad_norm": 0.24324072897434235,
+      "learning_rate": 0.0001641975308641975,
+      "loss": 5.6705,
+      "step": 685
+    },
+    {
+      "epoch": 1.5219079312257349,
+      "grad_norm": 0.30100396275520325,
+      "learning_rate": 0.0001639506172839506,
+      "loss": 5.8527,
+      "step": 686
+    },
+    {
+      "epoch": 1.524126455906822,
+      "grad_norm": 0.3281867802143097,
+      "learning_rate": 0.0001637037037037037,
+      "loss": 5.6384,
+      "step": 687
+    },
+    {
+      "epoch": 1.526344980587909,
+      "grad_norm": 0.2666301131248474,
+      "learning_rate": 0.00016345679012345676,
+      "loss": 5.9446,
+      "step": 688
+    },
+    {
+      "epoch": 1.528563505268996,
+      "grad_norm": 0.3064371943473816,
+      "learning_rate": 0.00016320987654320988,
+      "loss": 5.6311,
+      "step": 689
+    },
+    {
+      "epoch": 1.5307820299500832,
+      "grad_norm": 0.2843113839626312,
+      "learning_rate": 0.00016296296296296295,
+      "loss": 5.8221,
+      "step": 690
+    },
+    {
+      "epoch": 1.5330005546311702,
+      "grad_norm": 0.33836203813552856,
+      "learning_rate": 0.00016271604938271605,
+      "loss": 5.8814,
+      "step": 691
+    },
+    {
+      "epoch": 1.5352190793122573,
+      "grad_norm": 0.23588187992572784,
+      "learning_rate": 0.00016246913580246912,
+      "loss": 5.8729,
+      "step": 692
+    },
+    {
+      "epoch": 1.5374376039933444,
+      "grad_norm": 0.37101081013679504,
+      "learning_rate": 0.0001622222222222222,
+      "loss": 5.9488,
+      "step": 693
+    },
+    {
+      "epoch": 1.5396561286744315,
+      "grad_norm": 0.2393975704908371,
+      "learning_rate": 0.0001619753086419753,
+      "loss": 5.5923,
+      "step": 694
+    },
+    {
+      "epoch": 1.5418746533555185,
+      "grad_norm": 0.2354043424129486,
+      "learning_rate": 0.00016172839506172837,
+      "loss": 5.5373,
+      "step": 695
+    },
+    {
+      "epoch": 1.5440931780366056,
+      "grad_norm": 0.2429513782262802,
+      "learning_rate": 0.00016148148148148147,
+      "loss": 5.7939,
+      "step": 696
+    },
+    {
+      "epoch": 1.5463117027176927,
+      "grad_norm": 0.2338414490222931,
+      "learning_rate": 0.00016123456790123454,
+      "loss": 5.6767,
+      "step": 697
+    },
+    {
+      "epoch": 1.5485302273987798,
+      "grad_norm": 0.23638315498828888,
+      "learning_rate": 0.00016098765432098766,
+      "loss": 5.6714,
+      "step": 698
+    },
+    {
+      "epoch": 1.5507487520798668,
+      "grad_norm": 0.2488742619752884,
+      "learning_rate": 0.00016074074074074073,
+      "loss": 5.5853,
+      "step": 699
+    },
+    {
+      "epoch": 1.552967276760954,
+      "grad_norm": 0.23063461482524872,
+      "learning_rate": 0.00016049382716049382,
+      "loss": 5.7143,
+      "step": 700
+    },
+    {
+      "epoch": 1.555185801442041,
+      "grad_norm": 0.33041372895240784,
+      "learning_rate": 0.00016024691358024692,
+      "loss": 5.5362,
+      "step": 701
+    },
+    {
+      "epoch": 1.557404326123128,
+      "grad_norm": 0.2807255685329437,
+      "learning_rate": 0.00015999999999999999,
+      "loss": 5.8051,
+      "step": 702
+    },
+    {
+      "epoch": 1.5596228508042151,
+      "grad_norm": 0.30456334352493286,
+      "learning_rate": 0.00015975308641975308,
+      "loss": 6.1592,
+      "step": 703
+    },
+    {
+      "epoch": 1.5618413754853022,
+      "grad_norm": 0.2742255628108978,
+      "learning_rate": 0.00015950617283950615,
+      "loss": 5.8741,
+      "step": 704
+    },
+    {
+      "epoch": 1.5640599001663893,
+      "grad_norm": 0.3346208930015564,
+      "learning_rate": 0.00015925925925925924,
+      "loss": 5.3566,
+      "step": 705
+    },
+    {
+      "epoch": 1.5662784248474764,
+      "grad_norm": 0.29471608996391296,
+      "learning_rate": 0.0001590123456790123,
+      "loss": 5.6887,
+      "step": 706
+    },
+    {
+      "epoch": 1.5684969495285634,
+      "grad_norm": 0.2504662275314331,
+      "learning_rate": 0.00015876543209876543,
+      "loss": 5.5752,
+      "step": 707
+    },
+    {
+      "epoch": 1.5707154742096505,
+      "grad_norm": 0.2609747052192688,
+      "learning_rate": 0.0001585185185185185,
+      "loss": 5.7176,
+      "step": 708
+    },
+    {
+      "epoch": 1.5729339988907376,
+      "grad_norm": 0.2682695984840393,
+      "learning_rate": 0.0001582716049382716,
+      "loss": 5.8101,
+      "step": 709
+    },
+    {
+      "epoch": 1.5751525235718247,
+      "grad_norm": 0.2635984420776367,
+      "learning_rate": 0.0001580246913580247,
+      "loss": 5.7725,
+      "step": 710
+    },
+    {
+      "epoch": 1.5773710482529117,
+      "grad_norm": 0.2730768620967865,
+      "learning_rate": 0.00015777777777777776,
+      "loss": 5.8942,
+      "step": 711
+    },
+    {
+      "epoch": 1.5795895729339988,
+      "grad_norm": 0.29341018199920654,
+      "learning_rate": 0.00015753086419753086,
+      "loss": 5.7957,
+      "step": 712
+    },
+    {
+      "epoch": 1.581808097615086,
+      "grad_norm": 0.24498017132282257,
+      "learning_rate": 0.00015728395061728392,
+      "loss": 5.6089,
+      "step": 713
+    },
+    {
+      "epoch": 1.584026622296173,
+      "grad_norm": 0.1924431025981903,
+      "learning_rate": 0.00015703703703703702,
+      "loss": 5.716,
+      "step": 714
+    },
+    {
+      "epoch": 1.58624514697726,
+      "grad_norm": 0.2945975065231323,
+      "learning_rate": 0.0001567901234567901,
+      "loss": 5.7599,
+      "step": 715
+    },
+    {
+      "epoch": 1.5884636716583471,
+      "grad_norm": 0.2768838107585907,
+      "learning_rate": 0.0001565432098765432,
+      "loss": 5.8354,
+      "step": 716
+    },
+    {
+      "epoch": 1.5906821963394342,
+      "grad_norm": 0.28372520208358765,
+      "learning_rate": 0.0001562962962962963,
+      "loss": 5.8025,
+      "step": 717
+    },
+    {
+      "epoch": 1.5929007210205213,
+      "grad_norm": 0.2542419135570526,
+      "learning_rate": 0.00015604938271604937,
+      "loss": 5.6808,
+      "step": 718
+    },
+    {
+      "epoch": 1.5951192457016083,
+      "grad_norm": 0.2888288199901581,
+      "learning_rate": 0.00015580246913580247,
+      "loss": 5.4955,
+      "step": 719
+    },
+    {
+      "epoch": 1.5973377703826954,
+      "grad_norm": 0.2929094135761261,
+      "learning_rate": 0.00015555555555555554,
+      "loss": 5.9274,
+      "step": 720
+    },
+    {
+      "epoch": 1.5995562950637825,
+      "grad_norm": 0.30099013447761536,
+      "learning_rate": 0.00015530864197530863,
+      "loss": 5.7329,
+      "step": 721
+    },
+    {
+      "epoch": 1.6017748197448696,
+      "grad_norm": 0.2588217556476593,
+      "learning_rate": 0.0001550617283950617,
+      "loss": 5.8093,
+      "step": 722
+    },
+    {
+      "epoch": 1.6039933444259566,
+      "grad_norm": 0.3084394335746765,
+      "learning_rate": 0.0001548148148148148,
+      "loss": 5.378,
+      "step": 723
+    },
+    {
+      "epoch": 1.6062118691070437,
+      "grad_norm": 0.3200731873512268,
+      "learning_rate": 0.00015456790123456792,
+      "loss": 6.0228,
+      "step": 724
+    },
+    {
+      "epoch": 1.6084303937881308,
+      "grad_norm": 0.32377877831459045,
+      "learning_rate": 0.00015432098765432098,
+      "loss": 5.4862,
+      "step": 725
+    },
+    {
+      "epoch": 1.6106489184692179,
+      "grad_norm": 0.29225271940231323,
+      "learning_rate": 0.00015407407407407408,
+      "loss": 5.6643,
+      "step": 726
+    },
+    {
+      "epoch": 1.612867443150305,
+      "grad_norm": 0.27575793862342834,
+      "learning_rate": 0.00015382716049382715,
+      "loss": 5.4817,
+      "step": 727
+    },
+    {
+      "epoch": 1.615085967831392,
+      "grad_norm": 0.24983204901218414,
+      "learning_rate": 0.00015358024691358024,
+      "loss": 5.6399,
+      "step": 728
+    },
+    {
+      "epoch": 1.617304492512479,
+      "grad_norm": 0.2839362919330597,
+      "learning_rate": 0.0001533333333333333,
+      "loss": 5.8557,
+      "step": 729
+    },
+    {
+      "epoch": 1.6195230171935662,
+      "grad_norm": 0.2826875150203705,
+      "learning_rate": 0.0001530864197530864,
+      "loss": 5.5113,
+      "step": 730
+    },
+    {
+      "epoch": 1.6217415418746532,
+      "grad_norm": 0.27290812134742737,
+      "learning_rate": 0.00015283950617283947,
+      "loss": 5.7378,
+      "step": 731
+    },
+    {
+      "epoch": 1.6239600665557403,
+      "grad_norm": 0.3488001227378845,
+      "learning_rate": 0.00015259259259259257,
+      "loss": 5.403,
+      "step": 732
+    },
+    {
+      "epoch": 1.6261785912368274,
+      "grad_norm": 0.37526747584342957,
+      "learning_rate": 0.0001523456790123457,
+      "loss": 5.5884,
+      "step": 733
+    },
+    {
+      "epoch": 1.6283971159179145,
+      "grad_norm": 0.3014378249645233,
+      "learning_rate": 0.00015209876543209873,
+      "loss": 5.5719,
+      "step": 734
+    },
+    {
+      "epoch": 1.6306156405990015,
+      "grad_norm": 0.24637694656848907,
+      "learning_rate": 0.00015185185185185185,
+      "loss": 5.9287,
+      "step": 735
+    },
+    {
+      "epoch": 1.6328341652800886,
+      "grad_norm": 0.2303023338317871,
+      "learning_rate": 0.00015160493827160492,
+      "loss": 5.6075,
+      "step": 736
+    },
+    {
+      "epoch": 1.6350526899611757,
+      "grad_norm": 0.2608555257320404,
+      "learning_rate": 0.00015135802469135802,
+      "loss": 5.7664,
+      "step": 737
+    },
+    {
+      "epoch": 1.6372712146422628,
+      "grad_norm": 0.27438387274742126,
+      "learning_rate": 0.00015111111111111109,
+      "loss": 5.2631,
+      "step": 738
+    },
+    {
+      "epoch": 1.6394897393233498,
+      "grad_norm": 0.28841111063957214,
+      "learning_rate": 0.00015086419753086418,
+      "loss": 5.7166,
+      "step": 739
+    },
+    {
+      "epoch": 1.641708264004437,
+      "grad_norm": 0.44722285866737366,
+      "learning_rate": 0.00015061728395061728,
+      "loss": 5.7997,
+      "step": 740
+    },
+    {
+      "epoch": 1.643926788685524,
+      "grad_norm": 0.22557969391345978,
+      "learning_rate": 0.00015037037037037034,
+      "loss": 5.8273,
+      "step": 741
+    },
+    {
+      "epoch": 1.646145313366611,
+      "grad_norm": 0.3122265338897705,
+      "learning_rate": 0.00015012345679012347,
+      "loss": 5.8026,
+      "step": 742
+    },
+    {
+      "epoch": 1.6483638380476981,
+      "grad_norm": 0.2234640121459961,
+      "learning_rate": 0.00014987654320987653,
+      "loss": 5.8234,
+      "step": 743
+    },
+    {
+      "epoch": 1.6505823627287852,
+      "grad_norm": 0.235027015209198,
+      "learning_rate": 0.00014962962962962963,
+      "loss": 5.7361,
+      "step": 744
+    },
+    {
+      "epoch": 1.6528008874098723,
+      "grad_norm": 0.37154093384742737,
+      "learning_rate": 0.0001493827160493827,
+      "loss": 5.6142,
+      "step": 745
+    },
+    {
+      "epoch": 1.6550194120909594,
+      "grad_norm": 0.42254334688186646,
+      "learning_rate": 0.0001491358024691358,
+      "loss": 5.8301,
+      "step": 746
+    },
+    {
+      "epoch": 1.6572379367720464,
+      "grad_norm": 0.2580878734588623,
+      "learning_rate": 0.00014888888888888886,
+      "loss": 5.5943,
+      "step": 747
+    },
+    {
+      "epoch": 1.6594564614531335,
+      "grad_norm": 0.44888368248939514,
+      "learning_rate": 0.00014864197530864196,
+      "loss": 5.8275,
+      "step": 748
+    },
+    {
+      "epoch": 1.6616749861342206,
+      "grad_norm": 0.31153783202171326,
+      "learning_rate": 0.00014839506172839505,
+      "loss": 5.6785,
+      "step": 749
+    },
+    {
+      "epoch": 1.6638935108153077,
+      "grad_norm": 0.4075663089752197,
+      "learning_rate": 0.00014814814814814812,
+      "loss": 5.6897,
+      "step": 750
+    },
+    {
+      "epoch": 1.6661120354963947,
+      "grad_norm": 0.34542617201805115,
+      "learning_rate": 0.00014790123456790124,
+      "loss": 5.5867,
+      "step": 751
+    },
+    {
+      "epoch": 1.6683305601774818,
+      "grad_norm": 0.3305094242095947,
+      "learning_rate": 0.0001476543209876543,
+      "loss": 5.529,
+      "step": 752
+    },
+    {
+      "epoch": 1.670549084858569,
+      "grad_norm": 0.2836512625217438,
+      "learning_rate": 0.0001474074074074074,
+      "loss": 5.5726,
+      "step": 753
+    },
+    {
+      "epoch": 1.672767609539656,
+      "grad_norm": 0.4211433529853821,
+      "learning_rate": 0.00014716049382716047,
+      "loss": 5.7545,
+      "step": 754
+    },
+    {
+      "epoch": 1.674986134220743,
+      "grad_norm": 0.29735034704208374,
+      "learning_rate": 0.00014691358024691357,
+      "loss": 5.8194,
+      "step": 755
+    },
+    {
+      "epoch": 1.6772046589018301,
+      "grad_norm": 0.2671893835067749,
+      "learning_rate": 0.00014666666666666664,
+      "loss": 5.9629,
+      "step": 756
+    },
+    {
+      "epoch": 1.6794231835829172,
+      "grad_norm": 0.3316991329193115,
+      "learning_rate": 0.00014641975308641973,
+      "loss": 5.5824,
+      "step": 757
+    },
+    {
+      "epoch": 1.6816417082640043,
+      "grad_norm": 0.2453877329826355,
+      "learning_rate": 0.00014617283950617283,
+      "loss": 5.6438,
+      "step": 758
+    },
+    {
+      "epoch": 1.6838602329450914,
+      "grad_norm": 0.2457333505153656,
+      "learning_rate": 0.00014592592592592592,
+      "loss": 5.1229,
+      "step": 759
+    },
+    {
+      "epoch": 1.6860787576261784,
+      "grad_norm": 0.3393521308898926,
+      "learning_rate": 0.000145679012345679,
+      "loss": 5.5722,
+      "step": 760
+    },
+    {
+      "epoch": 1.6882972823072657,
+      "grad_norm": 0.2524617612361908,
+      "learning_rate": 0.00014543209876543208,
+      "loss": 5.4699,
+      "step": 761
+    },
+    {
+      "epoch": 1.6905158069883528,
+      "grad_norm": 0.23664480447769165,
+      "learning_rate": 0.00014518518518518518,
+      "loss": 5.4724,
+      "step": 762
+    },
+    {
+      "epoch": 1.6927343316694399,
+      "grad_norm": 0.3203359842300415,
+      "learning_rate": 0.00014493827160493825,
+      "loss": 5.6344,
+      "step": 763
+    },
+    {
+      "epoch": 1.694952856350527,
+      "grad_norm": 0.27076730132102966,
+      "learning_rate": 0.00014469135802469134,
+      "loss": 5.5308,
+      "step": 764
+    },
+    {
+      "epoch": 1.697171381031614,
+      "grad_norm": 0.2773015797138214,
+      "learning_rate": 0.0001444444444444444,
+      "loss": 5.8604,
+      "step": 765
+    },
+    {
+      "epoch": 1.699389905712701,
+      "grad_norm": 0.3945075571537018,
+      "learning_rate": 0.00014419753086419753,
+      "loss": 5.4246,
+      "step": 766
+    },
+    {
+      "epoch": 1.7016084303937882,
+      "grad_norm": 0.2466626763343811,
+      "learning_rate": 0.0001439506172839506,
+      "loss": 5.9182,
+      "step": 767
+    },
+    {
+      "epoch": 1.7038269550748752,
+      "grad_norm": 0.3174384534358978,
+      "learning_rate": 0.0001437037037037037,
+      "loss": 5.5657,
+      "step": 768
+    },
+    {
+      "epoch": 1.7060454797559623,
+      "grad_norm": 0.259668231010437,
+      "learning_rate": 0.00014345679012345676,
+      "loss": 5.3777,
+      "step": 769
+    },
+    {
+      "epoch": 1.7082640044370494,
+      "grad_norm": 0.3013722598552704,
+      "learning_rate": 0.00014320987654320986,
+      "loss": 5.64,
+      "step": 770
+    },
+    {
+      "epoch": 1.7104825291181365,
+      "grad_norm": 0.2857159674167633,
+      "learning_rate": 0.00014296296296296295,
+      "loss": 5.7783,
+      "step": 771
+    },
+    {
+      "epoch": 1.7127010537992235,
+      "grad_norm": 0.30150118470191956,
+      "learning_rate": 0.00014271604938271602,
+      "loss": 5.5701,
+      "step": 772
+    },
+    {
+      "epoch": 1.7149195784803106,
+      "grad_norm": 0.31380176544189453,
+      "learning_rate": 0.00014246913580246912,
+      "loss": 5.7915,
+      "step": 773
+    },
+    {
+      "epoch": 1.7171381031613977,
+      "grad_norm": 0.2531150281429291,
+      "learning_rate": 0.0001422222222222222,
+      "loss": 5.5103,
+      "step": 774
+    },
+    {
+      "epoch": 1.7193566278424848,
+      "grad_norm": 0.3764350712299347,
+      "learning_rate": 0.0001419753086419753,
+      "loss": 5.5791,
+      "step": 775
+    },
+    {
+      "epoch": 1.7215751525235718,
+      "grad_norm": 0.25026488304138184,
+      "learning_rate": 0.00014172839506172838,
+      "loss": 5.5473,
+      "step": 776
+    },
+    {
+      "epoch": 1.723793677204659,
+      "grad_norm": 0.3245653212070465,
+      "learning_rate": 0.00014148148148148147,
+      "loss": 5.6907,
+      "step": 777
+    },
+    {
+      "epoch": 1.726012201885746,
+      "grad_norm": 0.3314420282840729,
+      "learning_rate": 0.00014123456790123454,
+      "loss": 5.7701,
+      "step": 778
+    },
+    {
+      "epoch": 1.728230726566833,
+      "grad_norm": 0.2817484736442566,
+      "learning_rate": 0.00014098765432098763,
+      "loss": 5.8482,
+      "step": 779
+    },
+    {
+      "epoch": 1.7304492512479202,
+      "grad_norm": 0.31235364079475403,
+      "learning_rate": 0.00014074074074074073,
+      "loss": 5.8475,
+      "step": 780
+    },
+    {
+      "epoch": 1.7326677759290072,
+      "grad_norm": 0.31663402915000916,
+      "learning_rate": 0.00014049382716049382,
+      "loss": 5.3052,
+      "step": 781
+    },
+    {
+      "epoch": 1.7348863006100943,
+      "grad_norm": 0.23209644854068756,
+      "learning_rate": 0.0001402469135802469,
+      "loss": 5.4722,
+      "step": 782
+    },
+    {
+      "epoch": 1.7371048252911814,
+      "grad_norm": 0.3214818835258484,
+      "learning_rate": 0.00014,
+      "loss": 5.5745,
+      "step": 783
+    },
+    {
+      "epoch": 1.7393233499722685,
+      "grad_norm": 0.36744388937950134,
+      "learning_rate": 0.00013975308641975308,
+      "loss": 5.5678,
+      "step": 784
+    },
+    {
+      "epoch": 1.7415418746533555,
+      "grad_norm": 0.28808900713920593,
+      "learning_rate": 0.00013950617283950615,
+      "loss": 5.6295,
+      "step": 785
+    },
+    {
+      "epoch": 1.7437603993344426,
+      "grad_norm": 0.31135958433151245,
+      "learning_rate": 0.00013925925925925925,
+      "loss": 5.4995,
+      "step": 786
+    },
+    {
+      "epoch": 1.7459789240155297,
+      "grad_norm": 0.24643488228321075,
+      "learning_rate": 0.00013901234567901231,
+      "loss": 5.3226,
+      "step": 787
+    },
+    {
+      "epoch": 1.7481974486966168,
+      "grad_norm": 0.2972148656845093,
+      "learning_rate": 0.00013876543209876544,
+      "loss": 5.5552,
+      "step": 788
+    },
+    {
+      "epoch": 1.7504159733777038,
+      "grad_norm": 0.2605087161064148,
+      "learning_rate": 0.0001385185185185185,
+      "loss": 5.4354,
+      "step": 789
+    },
+    {
+      "epoch": 1.752634498058791,
+      "grad_norm": 0.33900362253189087,
+      "learning_rate": 0.0001382716049382716,
+      "loss": 5.5413,
+      "step": 790
+    },
+    {
+      "epoch": 1.754853022739878,
+      "grad_norm": 0.29835039377212524,
+      "learning_rate": 0.00013802469135802467,
+      "loss": 5.6261,
+      "step": 791
+    },
+    {
+      "epoch": 1.757071547420965,
+      "grad_norm": 0.4668044447898865,
+      "learning_rate": 0.00013777777777777776,
+      "loss": 5.7129,
+      "step": 792
+    },
+    {
+      "epoch": 1.7592900721020521,
+      "grad_norm": 0.280445396900177,
+      "learning_rate": 0.00013753086419753086,
+      "loss": 5.4943,
+      "step": 793
+    },
+    {
+      "epoch": 1.7615085967831392,
+      "grad_norm": 0.34586605429649353,
+      "learning_rate": 0.00013728395061728393,
+      "loss": 5.4107,
+      "step": 794
+    },
+    {
+      "epoch": 1.7637271214642263,
+      "grad_norm": 0.218814879655838,
+      "learning_rate": 0.00013703703703703702,
+      "loss": 5.4616,
+      "step": 795
+    },
+    {
+      "epoch": 1.7659456461453134,
+      "grad_norm": 0.3199771046638489,
+      "learning_rate": 0.00013679012345679012,
+      "loss": 5.7538,
+      "step": 796
+    },
+    {
+      "epoch": 1.7681641708264004,
+      "grad_norm": 0.34684082865715027,
+      "learning_rate": 0.0001365432098765432,
+      "loss": 5.6568,
+      "step": 797
+    },
+    {
+      "epoch": 1.7703826955074875,
+      "grad_norm": 0.29375481605529785,
+      "learning_rate": 0.00013629629629629628,
+      "loss": 5.6501,
+      "step": 798
+    },
+    {
+      "epoch": 1.7726012201885746,
+      "grad_norm": 0.21283216774463654,
+      "learning_rate": 0.00013604938271604937,
+      "loss": 5.4571,
+      "step": 799
+    },
+    {
+      "epoch": 1.7748197448696617,
+      "grad_norm": 0.286857932806015,
+      "learning_rate": 0.00013580246913580244,
+      "loss": 5.3268,
+      "step": 800
+    },
+    {
+      "epoch": 1.7770382695507487,
+      "grad_norm": 0.27948445081710815,
+      "learning_rate": 0.00013555555555555554,
+      "loss": 5.9547,
+      "step": 801
+    },
+    {
+      "epoch": 1.7792567942318358,
+      "grad_norm": 0.286746621131897,
+      "learning_rate": 0.00013530864197530863,
+      "loss": 5.7483,
+      "step": 802
+    },
+    {
+      "epoch": 1.7814753189129229,
+      "grad_norm": 0.2571532428264618,
+      "learning_rate": 0.00013506172839506173,
+      "loss": 5.6259,
+      "step": 803
+    },
+    {
+      "epoch": 1.78369384359401,
+      "grad_norm": 0.3977102041244507,
+      "learning_rate": 0.0001348148148148148,
+      "loss": 5.4452,
+      "step": 804
+    },
+    {
+      "epoch": 1.785912368275097,
+      "grad_norm": 0.2919734716415405,
+      "learning_rate": 0.0001345679012345679,
+      "loss": 5.8466,
+      "step": 805
+    },
+    {
+      "epoch": 1.788130892956184,
+      "grad_norm": 0.3611086905002594,
+      "learning_rate": 0.00013432098765432099,
+      "loss": 5.4713,
+      "step": 806
+    },
+    {
+      "epoch": 1.7903494176372712,
+      "grad_norm": 0.33190035820007324,
+      "learning_rate": 0.00013407407407407405,
+      "loss": 5.3175,
+      "step": 807
+    },
+    {
+      "epoch": 1.7925679423183583,
+      "grad_norm": 0.21127909421920776,
+      "learning_rate": 0.00013382716049382715,
+      "loss": 5.5959,
+      "step": 808
+    },
+    {
+      "epoch": 1.7947864669994453,
+      "grad_norm": 0.23990701138973236,
+      "learning_rate": 0.00013358024691358022,
+      "loss": 5.4977,
+      "step": 809
+    },
+    {
+      "epoch": 1.7970049916805324,
+      "grad_norm": 0.2841923236846924,
+      "learning_rate": 0.0001333333333333333,
+      "loss": 5.6646,
+      "step": 810
+    },
+    {
+      "epoch": 1.7992235163616195,
+      "grad_norm": 0.24061940610408783,
+      "learning_rate": 0.0001330864197530864,
+      "loss": 5.7169,
+      "step": 811
+    },
+    {
+      "epoch": 1.8014420410427066,
+      "grad_norm": 0.3535049259662628,
+      "learning_rate": 0.0001328395061728395,
+      "loss": 5.5092,
+      "step": 812
+    },
+    {
+      "epoch": 1.8036605657237936,
+      "grad_norm": 0.39811021089553833,
+      "learning_rate": 0.00013259259259259257,
+      "loss": 5.6383,
+      "step": 813
+    },
+    {
+      "epoch": 1.8058790904048807,
+      "grad_norm": 0.28438103199005127,
+      "learning_rate": 0.00013234567901234567,
+      "loss": 5.432,
+      "step": 814
+    },
+    {
+      "epoch": 1.8080976150859678,
+      "grad_norm": 0.24104441702365875,
+      "learning_rate": 0.00013209876543209876,
+      "loss": 5.3935,
+      "step": 815
+    },
+    {
+      "epoch": 1.8103161397670549,
+      "grad_norm": 0.3507119417190552,
+      "learning_rate": 0.00013185185185185183,
+      "loss": 5.7662,
+      "step": 816
+    },
+    {
+      "epoch": 1.8125346644481422,
+      "grad_norm": 0.3717477023601532,
+      "learning_rate": 0.00013160493827160492,
+      "loss": 5.7403,
+      "step": 817
+    },
+    {
+      "epoch": 1.8147531891292292,
+      "grad_norm": 0.2917785942554474,
+      "learning_rate": 0.00013135802469135802,
+      "loss": 5.4035,
+      "step": 818
+    },
+    {
+      "epoch": 1.8169717138103163,
+      "grad_norm": 0.4126758575439453,
+      "learning_rate": 0.00013111111111111111,
+      "loss": 5.2851,
+      "step": 819
+    },
+    {
+      "epoch": 1.8191902384914034,
+      "grad_norm": 0.31985747814178467,
+      "learning_rate": 0.00013086419753086418,
+      "loss": 5.5209,
+      "step": 820
+    },
+    {
+      "epoch": 1.8214087631724905,
+      "grad_norm": 0.232477605342865,
+      "learning_rate": 0.00013061728395061728,
+      "loss": 5.6132,
+      "step": 821
+    },
+    {
+      "epoch": 1.8236272878535775,
+      "grad_norm": 0.33025696873664856,
+      "learning_rate": 0.00013037037037037035,
+      "loss": 5.6514,
+      "step": 822
+    },
+    {
+      "epoch": 1.8258458125346646,
+      "grad_norm": 0.2652396261692047,
+      "learning_rate": 0.00013012345679012344,
+      "loss": 5.4131,
+      "step": 823
+    },
+    {
+      "epoch": 1.8280643372157517,
+      "grad_norm": 0.28160303831100464,
+      "learning_rate": 0.00012987654320987654,
+      "loss": 5.4205,
+      "step": 824
+    },
+    {
+      "epoch": 1.8302828618968388,
+      "grad_norm": 0.3324123024940491,
+      "learning_rate": 0.0001296296296296296,
+      "loss": 5.7134,
+      "step": 825
+    },
+    {
+      "epoch": 1.8325013865779258,
+      "grad_norm": 0.40539151430130005,
+      "learning_rate": 0.0001293827160493827,
+      "loss": 5.3623,
+      "step": 826
+    },
+    {
+      "epoch": 1.834719911259013,
+      "grad_norm": 0.2936537265777588,
+      "learning_rate": 0.0001291358024691358,
+      "loss": 5.9396,
+      "step": 827
+    },
+    {
+      "epoch": 1.8369384359401,
+      "grad_norm": 0.35287320613861084,
+      "learning_rate": 0.0001288888888888889,
+      "loss": 5.2313,
+      "step": 828
+    },
+    {
+      "epoch": 1.839156960621187,
+      "grad_norm": 0.3678433299064636,
+      "learning_rate": 0.00012864197530864196,
+      "loss": 5.4955,
+      "step": 829
+    },
+    {
+      "epoch": 1.8413754853022741,
+      "grad_norm": 0.2675914168357849,
+      "learning_rate": 0.00012839506172839505,
+      "loss": 5.3342,
+      "step": 830
+    },
+    {
+      "epoch": 1.8435940099833612,
+      "grad_norm": 0.37433922290802,
+      "learning_rate": 0.00012814814814814812,
+      "loss": 5.6718,
+      "step": 831
+    },
+    {
+      "epoch": 1.8458125346644483,
+      "grad_norm": 0.3871198892593384,
+      "learning_rate": 0.00012790123456790122,
+      "loss": 5.7977,
+      "step": 832
+    },
+    {
+      "epoch": 1.8480310593455354,
+      "grad_norm": 0.27952104806900024,
+      "learning_rate": 0.0001276543209876543,
+      "loss": 5.4088,
+      "step": 833
+    },
+    {
+      "epoch": 1.8502495840266224,
+      "grad_norm": 0.3553330898284912,
+      "learning_rate": 0.0001274074074074074,
+      "loss": 5.5197,
+      "step": 834
+    },
+    {
+      "epoch": 1.8524681087077095,
+      "grad_norm": 0.22942374646663666,
+      "learning_rate": 0.00012716049382716047,
+      "loss": 5.3981,
+      "step": 835
+    },
+    {
+      "epoch": 1.8546866333887966,
+      "grad_norm": 0.30643031001091003,
+      "learning_rate": 0.00012691358024691357,
+      "loss": 5.8412,
+      "step": 836
+    },
+    {
+      "epoch": 1.8569051580698837,
+      "grad_norm": 0.2597436308860779,
+      "learning_rate": 0.00012666666666666666,
+      "loss": 5.5821,
+      "step": 837
+    },
+    {
+      "epoch": 1.8591236827509707,
+      "grad_norm": 0.32818323373794556,
+      "learning_rate": 0.00012641975308641973,
+      "loss": 5.5244,
+      "step": 838
+    },
+    {
+      "epoch": 1.8613422074320578,
+      "grad_norm": 0.3747442066669464,
+      "learning_rate": 0.00012617283950617283,
+      "loss": 5.369,
+      "step": 839
+    },
+    {
+      "epoch": 1.8635607321131449,
+      "grad_norm": 0.2921963632106781,
+      "learning_rate": 0.0001259259259259259,
+      "loss": 5.4996,
+      "step": 840
+    },
+    {
+      "epoch": 1.865779256794232,
+      "grad_norm": 0.3889302611351013,
+      "learning_rate": 0.00012567901234567902,
+      "loss": 5.5765,
+      "step": 841
+    },
+    {
+      "epoch": 1.867997781475319,
+      "grad_norm": 0.396106094121933,
+      "learning_rate": 0.00012543209876543209,
+      "loss": 5.5827,
+      "step": 842
+    },
+    {
+      "epoch": 1.870216306156406,
+      "grad_norm": 0.2868570387363434,
+      "learning_rate": 0.00012518518518518518,
+      "loss": 5.5043,
+      "step": 843
+    },
+    {
+      "epoch": 1.8724348308374932,
+      "grad_norm": 0.34068024158477783,
+      "learning_rate": 0.00012493827160493825,
+      "loss": 5.6453,
+      "step": 844
+    },
+    {
+      "epoch": 1.8746533555185803,
+      "grad_norm": 0.2496734857559204,
+      "learning_rate": 0.00012469135802469134,
+      "loss": 5.6836,
+      "step": 845
+    },
+    {
+      "epoch": 1.8768718801996673,
+      "grad_norm": 0.2308034598827362,
+      "learning_rate": 0.00012444444444444444,
+      "loss": 5.417,
+      "step": 846
+    },
+    {
+      "epoch": 1.8790904048807544,
+      "grad_norm": 0.28422677516937256,
+      "learning_rate": 0.0001241975308641975,
+      "loss": 5.4762,
+      "step": 847
+    },
+    {
+      "epoch": 1.8813089295618415,
+      "grad_norm": 0.3898235559463501,
+      "learning_rate": 0.0001239506172839506,
+      "loss": 5.293,
+      "step": 848
+    },
+    {
+      "epoch": 1.8835274542429286,
+      "grad_norm": 0.3652763068675995,
+      "learning_rate": 0.0001237037037037037,
+      "loss": 5.5566,
+      "step": 849
+    },
+    {
+      "epoch": 1.8857459789240156,
+      "grad_norm": 0.21545055508613586,
+      "learning_rate": 0.0001234567901234568,
+      "loss": 5.7777,
+      "step": 850
+    },
+    {
+      "epoch": 1.8879645036051027,
+      "grad_norm": 0.32686519622802734,
+      "learning_rate": 0.00012320987654320986,
+      "loss": 5.7727,
+      "step": 851
+    },
+    {
+      "epoch": 1.8901830282861898,
+      "grad_norm": 0.2869747579097748,
+      "learning_rate": 0.00012296296296296296,
+      "loss": 5.445,
+      "step": 852
+    },
+    {
+      "epoch": 1.8924015529672769,
+      "grad_norm": 0.29152712225914,
+      "learning_rate": 0.00012271604938271602,
+      "loss": 5.5237,
+      "step": 853
+    },
+    {
+      "epoch": 1.894620077648364,
+      "grad_norm": 0.2569763660430908,
+      "learning_rate": 0.00012246913580246912,
+      "loss": 5.5895,
+      "step": 854
+    },
+    {
+      "epoch": 1.896838602329451,
+      "grad_norm": 0.37539026141166687,
+      "learning_rate": 0.00012222222222222221,
+      "loss": 5.8043,
+      "step": 855
+    },
+    {
+      "epoch": 1.899057127010538,
+      "grad_norm": 0.2599134147167206,
+      "learning_rate": 0.00012197530864197531,
+      "loss": 5.5276,
+      "step": 856
+    },
+    {
+      "epoch": 1.9012756516916252,
+      "grad_norm": 0.3916271924972534,
+      "learning_rate": 0.00012172839506172839,
+      "loss": 5.1583,
+      "step": 857
+    },
+    {
+      "epoch": 1.9034941763727122,
+      "grad_norm": 0.35661837458610535,
+      "learning_rate": 0.00012148148148148147,
+      "loss": 5.5743,
+      "step": 858
+    },
+    {
+      "epoch": 1.9057127010537993,
+      "grad_norm": 0.3635735809803009,
+      "learning_rate": 0.00012123456790123455,
+      "loss": 5.6192,
+      "step": 859
+    },
+    {
+      "epoch": 1.9079312257348864,
+      "grad_norm": 0.38036492466926575,
+      "learning_rate": 0.00012098765432098764,
+      "loss": 5.6458,
+      "step": 860
+    },
+    {
+      "epoch": 1.9101497504159735,
+      "grad_norm": 0.5092012882232666,
+      "learning_rate": 0.00012074074074074073,
+      "loss": 5.9004,
+      "step": 861
+    },
+    {
+      "epoch": 1.9123682750970605,
+      "grad_norm": 0.22783057391643524,
+      "learning_rate": 0.00012049382716049381,
+      "loss": 5.1797,
+      "step": 862
+    },
+    {
+      "epoch": 1.9145867997781476,
+      "grad_norm": 0.2916918694972992,
+      "learning_rate": 0.0001202469135802469,
+      "loss": 5.3149,
+      "step": 863
+    },
+    {
+      "epoch": 1.9168053244592347,
+      "grad_norm": 0.24761202931404114,
+      "learning_rate": 0.00011999999999999999,
+      "loss": 5.2396,
+      "step": 864
+    },
+    {
+      "epoch": 1.9190238491403218,
+      "grad_norm": 0.29561570286750793,
+      "learning_rate": 0.00011975308641975308,
+      "loss": 5.7087,
+      "step": 865
+    },
+    {
+      "epoch": 1.9212423738214088,
+      "grad_norm": 0.47623154520988464,
+      "learning_rate": 0.00011950617283950617,
+      "loss": 5.5606,
+      "step": 866
+    },
+    {
+      "epoch": 1.923460898502496,
+      "grad_norm": 0.40492722392082214,
+      "learning_rate": 0.00011925925925925925,
+      "loss": 5.3069,
+      "step": 867
+    },
+    {
+      "epoch": 1.925679423183583,
+      "grad_norm": 0.23254160583019257,
+      "learning_rate": 0.00011901234567901233,
+      "loss": 5.7075,
+      "step": 868
+    },
+    {
+      "epoch": 1.92789794786467,
+      "grad_norm": 0.3111009895801544,
+      "learning_rate": 0.00011876543209876541,
+      "loss": 5.565,
+      "step": 869
+    },
+    {
+      "epoch": 1.9301164725457571,
+      "grad_norm": 0.362149715423584,
+      "learning_rate": 0.0001185185185185185,
+      "loss": 5.6842,
+      "step": 870
+    },
+    {
+      "epoch": 1.9323349972268442,
+      "grad_norm": 0.24925507605075836,
+      "learning_rate": 0.0001182716049382716,
+      "loss": 5.5599,
+      "step": 871
+    },
+    {
+      "epoch": 1.9345535219079313,
+      "grad_norm": 0.2599934935569763,
+      "learning_rate": 0.00011802469135802468,
+      "loss": 5.3396,
+      "step": 872
+    },
+    {
+      "epoch": 1.9367720465890184,
+      "grad_norm": 0.45159056782722473,
+      "learning_rate": 0.00011777777777777776,
+      "loss": 5.5162,
+      "step": 873
+    },
+    {
+      "epoch": 1.9389905712701054,
+      "grad_norm": 0.2707643210887909,
+      "learning_rate": 0.00011753086419753086,
+      "loss": 5.4196,
+      "step": 874
+    },
+    {
+      "epoch": 1.9412090959511925,
+      "grad_norm": 0.3296864628791809,
+      "learning_rate": 0.00011728395061728394,
+      "loss": 5.7971,
+      "step": 875
+    },
+    {
+      "epoch": 1.9434276206322796,
+      "grad_norm": 0.40003880858421326,
+      "learning_rate": 0.00011703703703703702,
+      "loss": 5.3223,
+      "step": 876
+    },
+    {
+      "epoch": 1.9456461453133667,
+      "grad_norm": 0.2516932487487793,
+      "learning_rate": 0.0001167901234567901,
+      "loss": 5.415,
+      "step": 877
+    },
+    {
+      "epoch": 1.9478646699944537,
+      "grad_norm": 0.30859947204589844,
+      "learning_rate": 0.00011654320987654319,
+      "loss": 5.3281,
+      "step": 878
+    },
+    {
+      "epoch": 1.9500831946755408,
+      "grad_norm": 0.30502399802207947,
+      "learning_rate": 0.0001162962962962963,
+      "loss": 5.6042,
+      "step": 879
+    },
+    {
+      "epoch": 1.9523017193566279,
+      "grad_norm": 0.39825645089149475,
+      "learning_rate": 0.00011604938271604938,
+      "loss": 5.4689,
+      "step": 880
+    },
+    {
+      "epoch": 1.954520244037715,
+      "grad_norm": 0.30410581827163696,
+      "learning_rate": 0.00011580246913580246,
+      "loss": 5.4682,
+      "step": 881
+    },
+    {
+      "epoch": 1.956738768718802,
+      "grad_norm": 0.41302332282066345,
+      "learning_rate": 0.00011555555555555554,
+      "loss": 5.6471,
+      "step": 882
+    },
+    {
+      "epoch": 1.958957293399889,
+      "grad_norm": 0.23752185702323914,
+      "learning_rate": 0.00011530864197530863,
+      "loss": 5.6215,
+      "step": 883
+    },
+    {
+      "epoch": 1.9611758180809762,
+      "grad_norm": 0.2683930993080139,
+      "learning_rate": 0.00011506172839506172,
+      "loss": 5.3983,
+      "step": 884
+    },
+    {
+      "epoch": 1.9633943427620633,
+      "grad_norm": 0.30169442296028137,
+      "learning_rate": 0.0001148148148148148,
+      "loss": 5.1849,
+      "step": 885
+    },
+    {
+      "epoch": 1.9656128674431503,
+      "grad_norm": 0.2745766043663025,
+      "learning_rate": 0.00011456790123456789,
+      "loss": 5.4173,
+      "step": 886
+    },
+    {
+      "epoch": 1.9678313921242374,
+      "grad_norm": 0.3758724629878998,
+      "learning_rate": 0.00011432098765432099,
+      "loss": 5.8986,
+      "step": 887
+    },
+    {
+      "epoch": 1.9700499168053245,
+      "grad_norm": 0.30434733629226685,
+      "learning_rate": 0.00011407407407407407,
+      "loss": 5.6249,
+      "step": 888
+    },
+    {
+      "epoch": 1.9722684414864116,
+      "grad_norm": 0.4520629048347473,
+      "learning_rate": 0.00011382716049382715,
+      "loss": 5.5995,
+      "step": 889
+    },
+    {
+      "epoch": 1.9744869661674986,
+      "grad_norm": 0.39764168858528137,
+      "learning_rate": 0.00011358024691358023,
+      "loss": 5.2474,
+      "step": 890
+    },
+    {
+      "epoch": 1.9767054908485857,
+      "grad_norm": 0.48245230317115784,
+      "learning_rate": 0.00011333333333333331,
+      "loss": 5.5742,
+      "step": 891
+    },
+    {
+      "epoch": 1.9789240155296728,
+      "grad_norm": 0.2739104628562927,
+      "learning_rate": 0.00011308641975308641,
+      "loss": 5.4699,
+      "step": 892
+    },
+    {
+      "epoch": 1.9811425402107599,
+      "grad_norm": 0.29253920912742615,
+      "learning_rate": 0.00011283950617283949,
+      "loss": 5.3611,
+      "step": 893
+    },
+    {
+      "epoch": 1.983361064891847,
+      "grad_norm": 0.32551905512809753,
+      "learning_rate": 0.00011259259259259259,
+      "loss": 5.5264,
+      "step": 894
+    },
+    {
+      "epoch": 1.985579589572934,
+      "grad_norm": 0.4280093312263489,
+      "learning_rate": 0.00011234567901234567,
+      "loss": 5.6746,
+      "step": 895
+    },
+    {
+      "epoch": 1.987798114254021,
+      "grad_norm": 0.2495141625404358,
+      "learning_rate": 0.00011209876543209876,
+      "loss": 5.3924,
+      "step": 896
+    },
+    {
+      "epoch": 1.9900166389351082,
+      "grad_norm": 0.3221922814846039,
+      "learning_rate": 0.00011185185185185184,
+      "loss": 5.2494,
+      "step": 897
+    },
+    {
+      "epoch": 1.9922351636161952,
+      "grad_norm": 0.26265108585357666,
+      "learning_rate": 0.00011160493827160493,
+      "loss": 5.0683,
+      "step": 898
+    },
+    {
+      "epoch": 1.9944536882972823,
+      "grad_norm": 0.22699572145938873,
+      "learning_rate": 0.00011135802469135801,
+      "loss": 5.3254,
+      "step": 899
+    },
+    {
+      "epoch": 1.9966722129783694,
+      "grad_norm": 0.2500637173652649,
+      "learning_rate": 0.00011111111111111109,
+      "loss": 5.3929,
+      "step": 900
+    },
+    {
+      "epoch": 1.9988907376594565,
+      "grad_norm": 0.35316479206085205,
+      "learning_rate": 0.0001108641975308642,
+      "loss": 5.3596,
+      "step": 901
+    },
+    {
+      "epoch": 2.0011092623405435,
+      "grad_norm": 0.4560403823852539,
+      "learning_rate": 0.00011061728395061728,
+      "loss": 5.4005,
+      "step": 902
+    },
+    {
+      "epoch": 2.0033277870216306,
+      "grad_norm": 0.3310658633708954,
+      "learning_rate": 0.00011037037037037036,
+      "loss": 5.2508,
+      "step": 903
+    },
+    {
+      "epoch": 2.0055463117027177,
+      "grad_norm": 0.38432323932647705,
+      "learning_rate": 0.00011012345679012344,
+      "loss": 5.3983,
+      "step": 904
+    },
+    {
+      "epoch": 2.0077648363838048,
+      "grad_norm": 0.3098140060901642,
+      "learning_rate": 0.00010987654320987654,
+      "loss": 5.7085,
+      "step": 905
+    },
+    {
+      "epoch": 2.009983361064892,
+      "grad_norm": 0.36812424659729004,
+      "learning_rate": 0.00010962962962962962,
+      "loss": 5.4507,
+      "step": 906
+    },
+    {
+      "epoch": 2.012201885745979,
+      "grad_norm": 0.3451833426952362,
+      "learning_rate": 0.0001093827160493827,
+      "loss": 5.4351,
+      "step": 907
+    },
+    {
+      "epoch": 2.014420410427066,
+      "grad_norm": 0.23451659083366394,
+      "learning_rate": 0.00010913580246913578,
+      "loss": 5.5118,
+      "step": 908
+    },
+    {
+      "epoch": 2.016638935108153,
+      "grad_norm": 0.2787967026233673,
+      "learning_rate": 0.00010888888888888889,
+      "loss": 5.5314,
+      "step": 909
+    },
+    {
+      "epoch": 2.01885745978924,
+      "grad_norm": 0.3139168918132782,
+      "learning_rate": 0.00010864197530864197,
+      "loss": 5.3767,
+      "step": 910
+    },
+    {
+      "epoch": 2.021075984470327,
+      "grad_norm": 0.275867760181427,
+      "learning_rate": 0.00010839506172839505,
+      "loss": 5.4445,
+      "step": 911
+    },
+    {
+      "epoch": 2.0232945091514143,
+      "grad_norm": 0.5114609599113464,
+      "learning_rate": 0.00010814814814814814,
+      "loss": 5.2227,
+      "step": 912
+    },
+    {
+      "epoch": 2.0255130338325014,
+      "grad_norm": 0.2581237256526947,
+      "learning_rate": 0.00010790123456790122,
+      "loss": 5.305,
+      "step": 913
+    },
+    {
+      "epoch": 2.0277315585135884,
+      "grad_norm": 0.23800574243068695,
+      "learning_rate": 0.00010765432098765431,
+      "loss": 5.6333,
+      "step": 914
+    },
+    {
+      "epoch": 2.0299500831946755,
+      "grad_norm": 0.4039807617664337,
+      "learning_rate": 0.0001074074074074074,
+      "loss": 5.2987,
+      "step": 915
+    },
+    {
+      "epoch": 2.0321686078757626,
+      "grad_norm": 0.26278218626976013,
+      "learning_rate": 0.00010716049382716048,
+      "loss": 5.649,
+      "step": 916
+    },
+    {
+      "epoch": 2.0343871325568497,
+      "grad_norm": 0.4739043116569519,
+      "learning_rate": 0.00010691358024691357,
+      "loss": 5.2798,
+      "step": 917
+    },
+    {
+      "epoch": 2.0366056572379367,
+      "grad_norm": 0.26767653226852417,
+      "learning_rate": 0.00010666666666666667,
+      "loss": 5.256,
+      "step": 918
+    },
+    {
+      "epoch": 2.038824181919024,
+      "grad_norm": 0.3919017016887665,
+      "learning_rate": 0.00010641975308641975,
+      "loss": 5.4002,
+      "step": 919
+    },
+    {
+      "epoch": 2.041042706600111,
+      "grad_norm": 0.4538804292678833,
+      "learning_rate": 0.00010617283950617283,
+      "loss": 5.4704,
+      "step": 920
+    },
+    {
+      "epoch": 2.043261231281198,
+      "grad_norm": 0.28279614448547363,
+      "learning_rate": 0.00010592592592592591,
+      "loss": 5.4061,
+      "step": 921
+    },
+    {
+      "epoch": 2.045479755962285,
+      "grad_norm": 0.2386205941438675,
+      "learning_rate": 0.00010567901234567899,
+      "loss": 5.4342,
+      "step": 922
+    },
+    {
+      "epoch": 2.047698280643372,
+      "grad_norm": 0.38964003324508667,
+      "learning_rate": 0.00010543209876543209,
+      "loss": 5.5685,
+      "step": 923
+    },
+    {
+      "epoch": 2.049916805324459,
+      "grad_norm": 0.4006921648979187,
+      "learning_rate": 0.00010518518518518518,
+      "loss": 5.2665,
+      "step": 924
+    },
+    {
+      "epoch": 2.0521353300055463,
+      "grad_norm": 0.3002506494522095,
+      "learning_rate": 0.00010493827160493826,
+      "loss": 5.4148,
+      "step": 925
+    },
+    {
+      "epoch": 2.0543538546866333,
+      "grad_norm": 0.26828068494796753,
+      "learning_rate": 0.00010469135802469135,
+      "loss": 5.4952,
+      "step": 926
+    },
+    {
+      "epoch": 2.0565723793677204,
+      "grad_norm": 0.3504580557346344,
+      "learning_rate": 0.00010444444444444444,
+      "loss": 5.5801,
+      "step": 927
+    },
+    {
+      "epoch": 2.0587909040488075,
+      "grad_norm": 0.41068488359451294,
+      "learning_rate": 0.00010419753086419752,
+      "loss": 5.2722,
+      "step": 928
+    },
+    {
+      "epoch": 2.0610094287298946,
+      "grad_norm": 0.3777349889278412,
+      "learning_rate": 0.0001039506172839506,
+      "loss": 5.4635,
+      "step": 929
+    },
+    {
+      "epoch": 2.0632279534109816,
+      "grad_norm": 0.2353261411190033,
+      "learning_rate": 0.00010370370370370369,
+      "loss": 5.3688,
+      "step": 930
+    },
+    {
+      "epoch": 2.0654464780920687,
+      "grad_norm": 0.32856759428977966,
+      "learning_rate": 0.00010345679012345677,
+      "loss": 5.5283,
+      "step": 931
+    },
+    {
+      "epoch": 2.067665002773156,
+      "grad_norm": 0.3015420436859131,
+      "learning_rate": 0.00010320987654320988,
+      "loss": 5.0904,
+      "step": 932
+    },
+    {
+      "epoch": 2.069883527454243,
+      "grad_norm": 0.24403491616249084,
+      "learning_rate": 0.00010296296296296296,
+      "loss": 5.4115,
+      "step": 933
+    },
+    {
+      "epoch": 2.07210205213533,
+      "grad_norm": 0.27562209963798523,
+      "learning_rate": 0.00010271604938271604,
+      "loss": 5.6052,
+      "step": 934
+    },
+    {
+      "epoch": 2.074320576816417,
+      "grad_norm": 0.343569815158844,
+      "learning_rate": 0.00010246913580246912,
+      "loss": 5.2049,
+      "step": 935
+    },
+    {
+      "epoch": 2.076539101497504,
+      "grad_norm": 0.27486303448677063,
+      "learning_rate": 0.00010222222222222222,
+      "loss": 5.5385,
+      "step": 936
+    },
+    {
+      "epoch": 2.078757626178591,
+      "grad_norm": 0.3153358995914459,
+      "learning_rate": 0.0001019753086419753,
+      "loss": 5.1744,
+      "step": 937
+    },
+    {
+      "epoch": 2.0809761508596782,
+      "grad_norm": 0.3876603841781616,
+      "learning_rate": 0.00010172839506172838,
+      "loss": 5.3729,
+      "step": 938
+    },
+    {
+      "epoch": 2.0831946755407653,
+      "grad_norm": 0.4374476373195648,
+      "learning_rate": 0.00010148148148148147,
+      "loss": 5.1481,
+      "step": 939
+    },
+    {
+      "epoch": 2.0854132002218524,
+      "grad_norm": 0.5372658371925354,
+      "learning_rate": 0.00010123456790123457,
+      "loss": 5.4333,
+      "step": 940
+    },
+    {
+      "epoch": 2.0876317249029395,
+      "grad_norm": 0.3700786232948303,
+      "learning_rate": 0.00010098765432098765,
+      "loss": 5.352,
+      "step": 941
+    },
+    {
+      "epoch": 2.0898502495840265,
+      "grad_norm": 0.307565301656723,
+      "learning_rate": 0.00010074074074074073,
+      "loss": 5.4442,
+      "step": 942
+    },
+    {
+      "epoch": 2.0920687742651136,
+      "grad_norm": 0.2938573658466339,
+      "learning_rate": 0.00010049382716049381,
+      "loss": 5.1108,
+      "step": 943
+    },
+    {
+      "epoch": 2.0942872989462007,
+      "grad_norm": 0.29402637481689453,
+      "learning_rate": 0.0001002469135802469,
+      "loss": 5.445,
+      "step": 944
+    },
+    {
+      "epoch": 2.0965058236272878,
+      "grad_norm": 0.4891315698623657,
+      "learning_rate": 9.999999999999999e-05,
+      "loss": 5.5637,
+      "step": 945
+    },
+    {
+      "epoch": 2.098724348308375,
+      "grad_norm": 0.2771971821784973,
+      "learning_rate": 9.975308641975307e-05,
+      "loss": 5.3702,
+      "step": 946
+    },
+    {
+      "epoch": 2.100942872989462,
+      "grad_norm": 0.2870939075946808,
+      "learning_rate": 9.950617283950617e-05,
+      "loss": 5.4332,
+      "step": 947
+    },
+    {
+      "epoch": 2.103161397670549,
+      "grad_norm": 0.26091375946998596,
+      "learning_rate": 9.925925925925925e-05,
+      "loss": 5.3093,
+      "step": 948
+    },
+    {
+      "epoch": 2.105379922351636,
+      "grad_norm": 0.36121365427970886,
+      "learning_rate": 9.901234567901234e-05,
+      "loss": 5.475,
+      "step": 949
+    },
+    {
+      "epoch": 2.107598447032723,
+      "grad_norm": 0.24968165159225464,
+      "learning_rate": 9.876543209876543e-05,
+      "loss": 5.2578,
+      "step": 950
+    },
+    {
+      "epoch": 2.10981697171381,
+      "grad_norm": 0.25511908531188965,
+      "learning_rate": 9.851851851851851e-05,
+      "loss": 5.2613,
+      "step": 951
+    },
+    {
+      "epoch": 2.1120354963948973,
+      "grad_norm": 0.26556840538978577,
+      "learning_rate": 9.827160493827159e-05,
+      "loss": 5.4068,
+      "step": 952
+    },
+    {
+      "epoch": 2.1142540210759844,
+      "grad_norm": 0.23080062866210938,
+      "learning_rate": 9.802469135802467e-05,
+      "loss": 5.305,
+      "step": 953
+    },
+    {
+      "epoch": 2.1164725457570714,
+      "grad_norm": 0.22475649416446686,
+      "learning_rate": 9.777777777777778e-05,
+      "loss": 5.4491,
+      "step": 954
+    },
+    {
+      "epoch": 2.1186910704381585,
+      "grad_norm": 0.35835081338882446,
+      "learning_rate": 9.753086419753086e-05,
+      "loss": 5.6167,
+      "step": 955
+    },
+    {
+      "epoch": 2.1209095951192456,
+      "grad_norm": 0.30577215552330017,
+      "learning_rate": 9.728395061728394e-05,
+      "loss": 5.6379,
+      "step": 956
+    },
+    {
+      "epoch": 2.1231281198003327,
+      "grad_norm": 0.32588234543800354,
+      "learning_rate": 9.703703703703702e-05,
+      "loss": 5.8424,
+      "step": 957
+    },
+    {
+      "epoch": 2.1253466444814197,
+      "grad_norm": 0.3204634189605713,
+      "learning_rate": 9.679012345679012e-05,
+      "loss": 5.5117,
+      "step": 958
+    },
+    {
+      "epoch": 2.127565169162507,
+      "grad_norm": 0.336595743894577,
+      "learning_rate": 9.65432098765432e-05,
+      "loss": 5.4104,
+      "step": 959
+    },
+    {
+      "epoch": 2.129783693843594,
+      "grad_norm": 0.3437650203704834,
+      "learning_rate": 9.629629629629628e-05,
+      "loss": 5.7698,
+      "step": 960
+    },
+    {
+      "epoch": 2.132002218524681,
+      "grad_norm": 0.45331552624702454,
+      "learning_rate": 9.604938271604936e-05,
+      "loss": 5.2775,
+      "step": 961
+    },
+    {
+      "epoch": 2.134220743205768,
+      "grad_norm": 0.2846590578556061,
+      "learning_rate": 9.580246913580247e-05,
+      "loss": 5.5112,
+      "step": 962
+    },
+    {
+      "epoch": 2.136439267886855,
+      "grad_norm": 0.5604024529457092,
+      "learning_rate": 9.555555555555555e-05,
+      "loss": 5.666,
+      "step": 963
+    },
+    {
+      "epoch": 2.138657792567942,
+      "grad_norm": 0.3234822750091553,
+      "learning_rate": 9.530864197530864e-05,
+      "loss": 5.7524,
+      "step": 964
+    },
+    {
+      "epoch": 2.1408763172490293,
+      "grad_norm": 0.3267075717449188,
+      "learning_rate": 9.506172839506172e-05,
+      "loss": 5.6162,
+      "step": 965
+    },
+    {
+      "epoch": 2.1430948419301163,
+      "grad_norm": 0.34921547770500183,
+      "learning_rate": 9.48148148148148e-05,
+      "loss": 5.3594,
+      "step": 966
+    },
+    {
+      "epoch": 2.1453133666112034,
+      "grad_norm": 0.3202040493488312,
+      "learning_rate": 9.45679012345679e-05,
+      "loss": 5.4458,
+      "step": 967
+    },
+    {
+      "epoch": 2.1475318912922905,
+      "grad_norm": 0.3196270763874054,
+      "learning_rate": 9.432098765432098e-05,
+      "loss": 5.6855,
+      "step": 968
+    },
+    {
+      "epoch": 2.1497504159733776,
+      "grad_norm": 0.2972642183303833,
+      "learning_rate": 9.407407407407407e-05,
+      "loss": 5.4165,
+      "step": 969
+    },
+    {
+      "epoch": 2.1519689406544646,
+      "grad_norm": 0.26860707998275757,
+      "learning_rate": 9.382716049382715e-05,
+      "loss": 5.3555,
+      "step": 970
+    },
+    {
+      "epoch": 2.1541874653355517,
+      "grad_norm": 0.42726168036460876,
+      "learning_rate": 9.358024691358025e-05,
+      "loss": 5.4482,
+      "step": 971
+    },
+    {
+      "epoch": 2.156405990016639,
+      "grad_norm": 0.27418166399002075,
+      "learning_rate": 9.333333333333333e-05,
+      "loss": 5.2568,
+      "step": 972
+    },
+    {
+      "epoch": 2.158624514697726,
+      "grad_norm": 0.37415987253189087,
+      "learning_rate": 9.308641975308641e-05,
+      "loss": 5.5456,
+      "step": 973
+    },
+    {
+      "epoch": 2.160843039378813,
+      "grad_norm": 0.23949208855628967,
+      "learning_rate": 9.283950617283949e-05,
+      "loss": 5.3831,
+      "step": 974
+    },
+    {
+      "epoch": 2.1630615640599,
+      "grad_norm": 0.2556234896183014,
+      "learning_rate": 9.259259259259257e-05,
+      "loss": 5.3842,
+      "step": 975
+    },
+    {
+      "epoch": 2.165280088740987,
+      "grad_norm": 0.28209635615348816,
+      "learning_rate": 9.234567901234567e-05,
+      "loss": 5.5707,
+      "step": 976
+    },
+    {
+      "epoch": 2.167498613422074,
+      "grad_norm": 0.382671982049942,
+      "learning_rate": 9.209876543209876e-05,
+      "loss": 5.3816,
+      "step": 977
+    },
+    {
+      "epoch": 2.1697171381031612,
+      "grad_norm": 0.3056615889072418,
+      "learning_rate": 9.185185185185185e-05,
+      "loss": 5.3155,
+      "step": 978
+    },
+    {
+      "epoch": 2.1719356627842483,
+      "grad_norm": 0.40775182843208313,
+      "learning_rate": 9.160493827160493e-05,
+      "loss": 5.6908,
+      "step": 979
+    },
+    {
+      "epoch": 2.1741541874653354,
+      "grad_norm": 0.34975379705429077,
+      "learning_rate": 9.135802469135802e-05,
+      "loss": 5.3669,
+      "step": 980
+    },
+    {
+      "epoch": 2.1763727121464225,
+      "grad_norm": 0.2527734339237213,
+      "learning_rate": 9.11111111111111e-05,
+      "loss": 5.3869,
+      "step": 981
+    },
+    {
+      "epoch": 2.1785912368275095,
+      "grad_norm": 0.3068617582321167,
+      "learning_rate": 9.086419753086419e-05,
+      "loss": 5.1575,
+      "step": 982
+    },
+    {
+      "epoch": 2.1808097615085966,
+      "grad_norm": 0.440679132938385,
+      "learning_rate": 9.061728395061727e-05,
+      "loss": 5.9541,
+      "step": 983
+    },
+    {
+      "epoch": 2.1830282861896837,
+      "grad_norm": 0.29510989785194397,
+      "learning_rate": 9.037037037037038e-05,
+      "loss": 5.2021,
+      "step": 984
+    },
+    {
+      "epoch": 2.1852468108707708,
+      "grad_norm": 0.3276568651199341,
+      "learning_rate": 9.012345679012346e-05,
+      "loss": 5.6885,
+      "step": 985
+    },
+    {
+      "epoch": 2.187465335551858,
+      "grad_norm": 0.3074929714202881,
+      "learning_rate": 8.987654320987654e-05,
+      "loss": 5.613,
+      "step": 986
+    },
+    {
+      "epoch": 2.189683860232945,
+      "grad_norm": 0.2834450900554657,
+      "learning_rate": 8.962962962962962e-05,
+      "loss": 5.0579,
+      "step": 987
+    },
+    {
+      "epoch": 2.191902384914032,
+      "grad_norm": 0.2908533811569214,
+      "learning_rate": 8.93827160493827e-05,
+      "loss": 5.0217,
+      "step": 988
+    },
+    {
+      "epoch": 2.194120909595119,
+      "grad_norm": 0.253155380487442,
+      "learning_rate": 8.91358024691358e-05,
+      "loss": 5.4454,
+      "step": 989
+    },
+    {
+      "epoch": 2.196339434276206,
+      "grad_norm": 0.28673794865608215,
+      "learning_rate": 8.888888888888888e-05,
+      "loss": 5.1854,
+      "step": 990
+    },
+    {
+      "epoch": 2.1985579589572932,
+      "grad_norm": 0.31364136934280396,
+      "learning_rate": 8.864197530864196e-05,
+      "loss": 5.2835,
+      "step": 991
+    },
+    {
+      "epoch": 2.2007764836383803,
+      "grad_norm": 0.28010228276252747,
+      "learning_rate": 8.839506172839506e-05,
+      "loss": 5.4829,
+      "step": 992
+    },
+    {
+      "epoch": 2.2029950083194674,
+      "grad_norm": 0.4050219655036926,
+      "learning_rate": 8.814814814814815e-05,
+      "loss": 5.74,
+      "step": 993
+    },
+    {
+      "epoch": 2.2052135330005544,
+      "grad_norm": 0.34561365842819214,
+      "learning_rate": 8.790123456790123e-05,
+      "loss": 5.6314,
+      "step": 994
+    },
+    {
+      "epoch": 2.2074320576816415,
+      "grad_norm": 0.28431236743927,
+      "learning_rate": 8.765432098765431e-05,
+      "loss": 5.3429,
+      "step": 995
+    },
+    {
+      "epoch": 2.2096505823627286,
+      "grad_norm": 0.27258390188217163,
+      "learning_rate": 8.74074074074074e-05,
+      "loss": 5.1526,
+      "step": 996
+    },
+    {
+      "epoch": 2.2118691070438157,
+      "grad_norm": 0.2811555862426758,
+      "learning_rate": 8.716049382716048e-05,
+      "loss": 5.4106,
+      "step": 997
+    },
+    {
+      "epoch": 2.2140876317249027,
+      "grad_norm": 0.36345386505126953,
+      "learning_rate": 8.691358024691357e-05,
+      "loss": 5.4389,
+      "step": 998
+    },
+    {
+      "epoch": 2.21630615640599,
+      "grad_norm": 0.35687047243118286,
+      "learning_rate": 8.666666666666665e-05,
+      "loss": 4.9865,
+      "step": 999
+    },
+    {
+      "epoch": 2.218524681087077,
+      "grad_norm": 0.24369649589061737,
+      "learning_rate": 8.641975308641975e-05,
+      "loss": 5.1904,
+      "step": 1000
+    },
+    {
+      "epoch": 2.220743205768164,
+      "grad_norm": 0.4945223927497864,
+      "learning_rate": 8.617283950617283e-05,
+      "loss": 5.2792,
+      "step": 1001
+    },
+    {
+      "epoch": 2.222961730449251,
+      "grad_norm": 0.3147221803665161,
+      "learning_rate": 8.592592592592593e-05,
+      "loss": 5.5771,
+      "step": 1002
+    },
+    {
+      "epoch": 2.225180255130338,
+      "grad_norm": 0.22104422748088837,
+      "learning_rate": 8.567901234567901e-05,
+      "loss": 5.2272,
+      "step": 1003
+    },
+    {
+      "epoch": 2.227398779811425,
+      "grad_norm": 0.29240623116493225,
+      "learning_rate": 8.543209876543209e-05,
+      "loss": 5.2446,
+      "step": 1004
+    },
+    {
+      "epoch": 2.2296173044925123,
+      "grad_norm": 0.2969723641872406,
+      "learning_rate": 8.518518518518517e-05,
+      "loss": 5.3392,
+      "step": 1005
+    },
+    {
+      "epoch": 2.2318358291735994,
+      "grad_norm": 0.3051271140575409,
+      "learning_rate": 8.493827160493825e-05,
+      "loss": 5.35,
+      "step": 1006
+    },
+    {
+      "epoch": 2.2340543538546864,
+      "grad_norm": 0.4393060803413391,
+      "learning_rate": 8.469135802469136e-05,
+      "loss": 5.7605,
+      "step": 1007
+    },
+    {
+      "epoch": 2.2362728785357735,
+      "grad_norm": 0.2566257119178772,
+      "learning_rate": 8.444444444444444e-05,
+      "loss": 5.6132,
+      "step": 1008
+    },
+    {
+      "epoch": 2.2384914032168606,
+      "grad_norm": 0.3133380711078644,
+      "learning_rate": 8.419753086419752e-05,
+      "loss": 5.7128,
+      "step": 1009
+    },
+    {
+      "epoch": 2.2407099278979477,
+      "grad_norm": 0.3654865622520447,
+      "learning_rate": 8.39506172839506e-05,
+      "loss": 5.5414,
+      "step": 1010
+    },
+    {
+      "epoch": 2.2429284525790347,
+      "grad_norm": 0.25325027108192444,
+      "learning_rate": 8.37037037037037e-05,
+      "loss": 5.2434,
+      "step": 1011
+    },
+    {
+      "epoch": 2.245146977260122,
+      "grad_norm": 0.2352600246667862,
+      "learning_rate": 8.345679012345678e-05,
+      "loss": 5.2277,
+      "step": 1012
+    },
+    {
+      "epoch": 2.247365501941209,
+      "grad_norm": 0.3061700761318207,
+      "learning_rate": 8.320987654320986e-05,
+      "loss": 5.1402,
+      "step": 1013
+    },
+    {
+      "epoch": 2.249584026622296,
+      "grad_norm": 0.27736759185791016,
+      "learning_rate": 8.296296296296295e-05,
+      "loss": 5.6654,
+      "step": 1014
+    },
+    {
+      "epoch": 2.251802551303383,
+      "grad_norm": 0.36114227771759033,
+      "learning_rate": 8.271604938271605e-05,
+      "loss": 5.6282,
+      "step": 1015
+    },
+    {
+      "epoch": 2.25402107598447,
+      "grad_norm": 0.3595137894153595,
+      "learning_rate": 8.246913580246914e-05,
+      "loss": 5.6064,
+      "step": 1016
+    },
+    {
+      "epoch": 2.256239600665557,
+      "grad_norm": 0.4248819351196289,
+      "learning_rate": 8.222222222222222e-05,
+      "loss": 5.3839,
+      "step": 1017
+    },
+    {
+      "epoch": 2.2584581253466443,
+      "grad_norm": 0.35670825839042664,
+      "learning_rate": 8.19753086419753e-05,
+      "loss": 5.6902,
+      "step": 1018
+    },
+    {
+      "epoch": 2.2606766500277313,
+      "grad_norm": 0.23827995359897614,
+      "learning_rate": 8.172839506172838e-05,
+      "loss": 5.4254,
+      "step": 1019
+    },
+    {
+      "epoch": 2.2628951747088184,
+      "grad_norm": 0.35613736510276794,
+      "learning_rate": 8.148148148148148e-05,
+      "loss": 5.4952,
+      "step": 1020
+    },
+    {
+      "epoch": 2.2651136993899055,
+      "grad_norm": 0.32223621010780334,
+      "learning_rate": 8.123456790123456e-05,
+      "loss": 5.441,
+      "step": 1021
+    },
+    {
+      "epoch": 2.2673322240709926,
+      "grad_norm": 0.29039034247398376,
+      "learning_rate": 8.098765432098765e-05,
+      "loss": 5.1411,
+      "step": 1022
+    },
+    {
+      "epoch": 2.2695507487520796,
+      "grad_norm": 0.29579341411590576,
+      "learning_rate": 8.074074074074073e-05,
+      "loss": 5.3576,
+      "step": 1023
+    },
+    {
+      "epoch": 2.2717692734331667,
+      "grad_norm": 0.33683279156684875,
+      "learning_rate": 8.049382716049383e-05,
+      "loss": 5.378,
+      "step": 1024
+    },
+    {
+      "epoch": 2.2739877981142538,
+      "grad_norm": 0.3006892204284668,
+      "learning_rate": 8.024691358024691e-05,
+      "loss": 5.5976,
+      "step": 1025
+    },
+    {
+      "epoch": 2.276206322795341,
+      "grad_norm": 0.256975382566452,
+      "learning_rate": 7.999999999999999e-05,
+      "loss": 5.5569,
+      "step": 1026
+    },
+    {
+      "epoch": 2.278424847476428,
+      "grad_norm": 0.35778412222862244,
+      "learning_rate": 7.975308641975307e-05,
+      "loss": 5.1325,
+      "step": 1027
+    },
+    {
+      "epoch": 2.280643372157515,
+      "grad_norm": 0.274276465177536,
+      "learning_rate": 7.950617283950616e-05,
+      "loss": 5.6109,
+      "step": 1028
+    },
+    {
+      "epoch": 2.2828618968386025,
+      "grad_norm": 0.3542903959751129,
+      "learning_rate": 7.925925925925925e-05,
+      "loss": 5.1813,
+      "step": 1029
+    },
+    {
+      "epoch": 2.2850804215196896,
+      "grad_norm": 0.43924480676651,
+      "learning_rate": 7.901234567901235e-05,
+      "loss": 5.5082,
+      "step": 1030
+    },
+    {
+      "epoch": 2.2872989462007767,
+      "grad_norm": 0.277004212141037,
+      "learning_rate": 7.876543209876543e-05,
+      "loss": 5.4471,
+      "step": 1031
+    },
+    {
+      "epoch": 2.2895174708818637,
+      "grad_norm": 0.28540557622909546,
+      "learning_rate": 7.851851851851851e-05,
+      "loss": 5.5756,
+      "step": 1032
+    },
+    {
+      "epoch": 2.291735995562951,
+      "grad_norm": 0.27395644783973694,
+      "learning_rate": 7.82716049382716e-05,
+      "loss": 5.2599,
+      "step": 1033
+    },
+    {
+      "epoch": 2.293954520244038,
+      "grad_norm": 0.4076818525791168,
+      "learning_rate": 7.802469135802469e-05,
+      "loss": 5.4432,
+      "step": 1034
+    },
+    {
+      "epoch": 2.296173044925125,
+      "grad_norm": 0.49046868085861206,
+      "learning_rate": 7.777777777777777e-05,
+      "loss": 5.6251,
+      "step": 1035
+    },
+    {
+      "epoch": 2.298391569606212,
+      "grad_norm": 0.35189691185951233,
+      "learning_rate": 7.753086419753085e-05,
+      "loss": 5.5599,
+      "step": 1036
+    },
+    {
+      "epoch": 2.300610094287299,
+      "grad_norm": 0.2997983396053314,
+      "learning_rate": 7.728395061728396e-05,
+      "loss": 5.1934,
+      "step": 1037
+    },
+    {
+      "epoch": 2.302828618968386,
+      "grad_norm": 0.4447542130947113,
+      "learning_rate": 7.703703703703704e-05,
+      "loss": 5.5211,
+      "step": 1038
+    },
+    {
+      "epoch": 2.3050471436494733,
+      "grad_norm": 0.45167839527130127,
+      "learning_rate": 7.679012345679012e-05,
+      "loss": 4.99,
+      "step": 1039
+    },
+    {
+      "epoch": 2.3072656683305603,
+      "grad_norm": 0.30541661381721497,
+      "learning_rate": 7.65432098765432e-05,
+      "loss": 5.414,
+      "step": 1040
+    },
+    {
+      "epoch": 2.3094841930116474,
+      "grad_norm": 0.25259336829185486,
+      "learning_rate": 7.629629629629628e-05,
+      "loss": 5.6279,
+      "step": 1041
+    },
+    {
+      "epoch": 2.3117027176927345,
+      "grad_norm": 0.24241523444652557,
+      "learning_rate": 7.604938271604937e-05,
+      "loss": 5.3981,
+      "step": 1042
+    },
+    {
+      "epoch": 2.3139212423738216,
+      "grad_norm": 0.27996739745140076,
+      "learning_rate": 7.580246913580246e-05,
+      "loss": 5.1404,
+      "step": 1043
+    },
+    {
+      "epoch": 2.3161397670549086,
+      "grad_norm": 0.2647869288921356,
+      "learning_rate": 7.555555555555554e-05,
+      "loss": 5.2036,
+      "step": 1044
+    },
+    {
+      "epoch": 2.3183582917359957,
+      "grad_norm": 0.40196937322616577,
+      "learning_rate": 7.530864197530864e-05,
+      "loss": 5.5158,
+      "step": 1045
+    },
+    {
+      "epoch": 2.320576816417083,
+      "grad_norm": 0.3019854724407196,
+      "learning_rate": 7.506172839506173e-05,
+      "loss": 5.3546,
+      "step": 1046
+    },
+    {
+      "epoch": 2.32279534109817,
+      "grad_norm": 0.32454583048820496,
+      "learning_rate": 7.481481481481481e-05,
+      "loss": 5.2748,
+      "step": 1047
+    },
+    {
+      "epoch": 2.325013865779257,
+      "grad_norm": 0.2833717465400696,
+      "learning_rate": 7.45679012345679e-05,
+      "loss": 5.2753,
+      "step": 1048
+    },
+    {
+      "epoch": 2.327232390460344,
+      "grad_norm": 0.4223242700099945,
+      "learning_rate": 7.432098765432098e-05,
+      "loss": 4.8844,
+      "step": 1049
+    },
+    {
+      "epoch": 2.329450915141431,
+      "grad_norm": 0.3175380527973175,
+      "learning_rate": 7.407407407407406e-05,
+      "loss": 5.1475,
+      "step": 1050
+    },
+    {
+      "epoch": 2.331669439822518,
+      "grad_norm": 0.19656610488891602,
+      "learning_rate": 7.382716049382715e-05,
+      "loss": 5.2584,
+      "step": 1051
+    },
+    {
+      "epoch": 2.3338879645036053,
+      "grad_norm": 0.24176956713199615,
+      "learning_rate": 7.358024691358024e-05,
+      "loss": 5.2488,
+      "step": 1052
+    },
+    {
+      "epoch": 2.3361064891846923,
+      "grad_norm": 0.3703049421310425,
+      "learning_rate": 7.333333333333332e-05,
+      "loss": 5.5718,
+      "step": 1053
+    },
+    {
+      "epoch": 2.3383250138657794,
+      "grad_norm": 0.3057176470756531,
+      "learning_rate": 7.308641975308641e-05,
+      "loss": 5.3893,
+      "step": 1054
+    },
+    {
+      "epoch": 2.3405435385468665,
+      "grad_norm": 0.2758178114891052,
+      "learning_rate": 7.28395061728395e-05,
+      "loss": 5.1639,
+      "step": 1055
+    },
+    {
+      "epoch": 2.3427620632279536,
+      "grad_norm": 0.2883493900299072,
+      "learning_rate": 7.259259259259259e-05,
+      "loss": 5.402,
+      "step": 1056
+    },
+    {
+      "epoch": 2.3449805879090406,
+      "grad_norm": 0.33940428495407104,
+      "learning_rate": 7.234567901234567e-05,
+      "loss": 5.2558,
+      "step": 1057
+    },
+    {
+      "epoch": 2.3471991125901277,
+      "grad_norm": 0.2206302136182785,
+      "learning_rate": 7.209876543209877e-05,
+      "loss": 5.4368,
+      "step": 1058
+    },
+    {
+      "epoch": 2.3494176372712148,
+      "grad_norm": 0.25819826126098633,
+      "learning_rate": 7.185185185185185e-05,
+      "loss": 5.338,
+      "step": 1059
+    },
+    {
+      "epoch": 2.351636161952302,
+      "grad_norm": 0.2388448417186737,
+      "learning_rate": 7.160493827160493e-05,
+      "loss": 5.1028,
+      "step": 1060
+    },
+    {
+      "epoch": 2.353854686633389,
+      "grad_norm": 0.31586071848869324,
+      "learning_rate": 7.135802469135801e-05,
+      "loss": 5.4035,
+      "step": 1061
+    },
+    {
+      "epoch": 2.356073211314476,
+      "grad_norm": 0.44885945320129395,
+      "learning_rate": 7.11111111111111e-05,
+      "loss": 5.5176,
+      "step": 1062
+    },
+    {
+      "epoch": 2.358291735995563,
+      "grad_norm": 0.3284108340740204,
+      "learning_rate": 7.086419753086419e-05,
+      "loss": 5.4281,
+      "step": 1063
+    },
+    {
+      "epoch": 2.36051026067665,
+      "grad_norm": 0.3629972040653229,
+      "learning_rate": 7.061728395061727e-05,
+      "loss": 5.188,
+      "step": 1064
+    },
+    {
+      "epoch": 2.3627287853577372,
+      "grad_norm": 0.34551528096199036,
+      "learning_rate": 7.037037037037036e-05,
+      "loss": 5.0855,
+      "step": 1065
+    },
+    {
+      "epoch": 2.3649473100388243,
+      "grad_norm": 0.24513523280620575,
+      "learning_rate": 7.012345679012345e-05,
+      "loss": 5.1504,
+      "step": 1066
+    },
+    {
+      "epoch": 2.3671658347199114,
+      "grad_norm": 0.4057064950466156,
+      "learning_rate": 6.987654320987654e-05,
+      "loss": 5.1597,
+      "step": 1067
+    },
+    {
+      "epoch": 2.3693843594009985,
+      "grad_norm": 0.3843954801559448,
+      "learning_rate": 6.962962962962962e-05,
+      "loss": 5.3139,
+      "step": 1068
+    },
+    {
+      "epoch": 2.3716028840820855,
+      "grad_norm": 0.32409077882766724,
+      "learning_rate": 6.938271604938272e-05,
+      "loss": 5.3873,
+      "step": 1069
+    },
+    {
+      "epoch": 2.3738214087631726,
+      "grad_norm": 0.25862130522727966,
+      "learning_rate": 6.91358024691358e-05,
+      "loss": 5.1908,
+      "step": 1070
+    },
+    {
+      "epoch": 2.3760399334442597,
+      "grad_norm": 0.33141621947288513,
+      "learning_rate": 6.888888888888888e-05,
+      "loss": 5.6388,
+      "step": 1071
+    },
+    {
+      "epoch": 2.3782584581253468,
+      "grad_norm": 0.40585803985595703,
+      "learning_rate": 6.864197530864196e-05,
+      "loss": 5.5227,
+      "step": 1072
+    },
+    {
+      "epoch": 2.380476982806434,
+      "grad_norm": 0.2878483235836029,
+      "learning_rate": 6.839506172839506e-05,
+      "loss": 5.4587,
+      "step": 1073
+    },
+    {
+      "epoch": 2.382695507487521,
+      "grad_norm": 0.26604005694389343,
+      "learning_rate": 6.814814814814814e-05,
+      "loss": 5.2873,
+      "step": 1074
+    },
+    {
+      "epoch": 2.384914032168608,
+      "grad_norm": 0.3085010349750519,
+      "learning_rate": 6.790123456790122e-05,
+      "loss": 5.2197,
+      "step": 1075
+    },
+    {
+      "epoch": 2.387132556849695,
+      "grad_norm": 0.32559192180633545,
+      "learning_rate": 6.765432098765432e-05,
+      "loss": 5.1899,
+      "step": 1076
+    },
+    {
+      "epoch": 2.389351081530782,
+      "grad_norm": 0.38467463850975037,
+      "learning_rate": 6.74074074074074e-05,
+      "loss": 5.7358,
+      "step": 1077
+    },
+    {
+      "epoch": 2.391569606211869,
+      "grad_norm": 0.3152386248111725,
+      "learning_rate": 6.716049382716049e-05,
+      "loss": 5.5747,
+      "step": 1078
+    },
+    {
+      "epoch": 2.3937881308929563,
+      "grad_norm": 0.4028296172618866,
+      "learning_rate": 6.691358024691357e-05,
+      "loss": 5.2894,
+      "step": 1079
+    },
+    {
+      "epoch": 2.3960066555740434,
+      "grad_norm": 0.3209373354911804,
+      "learning_rate": 6.666666666666666e-05,
+      "loss": 5.4461,
+      "step": 1080
+    },
+    {
+      "epoch": 2.3982251802551304,
+      "grad_norm": 0.44152015447616577,
+      "learning_rate": 6.641975308641975e-05,
+      "loss": 5.2171,
+      "step": 1081
+    },
+    {
+      "epoch": 2.4004437049362175,
+      "grad_norm": 0.23774267733097076,
+      "learning_rate": 6.617283950617283e-05,
+      "loss": 5.3656,
+      "step": 1082
+    },
+    {
+      "epoch": 2.4026622296173046,
+      "grad_norm": 0.34255844354629517,
+      "learning_rate": 6.592592592592591e-05,
+      "loss": 5.4982,
+      "step": 1083
+    },
+    {
+      "epoch": 2.4048807542983917,
+      "grad_norm": 0.216425359249115,
+      "learning_rate": 6.567901234567901e-05,
+      "loss": 5.2975,
+      "step": 1084
+    },
+    {
+      "epoch": 2.4070992789794787,
+      "grad_norm": 0.22249232232570648,
+      "learning_rate": 6.543209876543209e-05,
+      "loss": 5.3426,
+      "step": 1085
+    },
+    {
+      "epoch": 2.409317803660566,
+      "grad_norm": 0.38628220558166504,
+      "learning_rate": 6.518518518518517e-05,
+      "loss": 5.2367,
+      "step": 1086
+    },
+    {
+      "epoch": 2.411536328341653,
+      "grad_norm": 0.29897111654281616,
+      "learning_rate": 6.493827160493827e-05,
+      "loss": 5.2144,
+      "step": 1087
+    },
+    {
+      "epoch": 2.41375485302274,
+      "grad_norm": 0.4442440867424011,
+      "learning_rate": 6.469135802469135e-05,
+      "loss": 5.1179,
+      "step": 1088
+    },
+    {
+      "epoch": 2.415973377703827,
+      "grad_norm": 0.2641626298427582,
+      "learning_rate": 6.444444444444444e-05,
+      "loss": 5.3902,
+      "step": 1089
+    },
+    {
+      "epoch": 2.418191902384914,
+      "grad_norm": 0.3379814624786377,
+      "learning_rate": 6.419753086419753e-05,
+      "loss": 5.3515,
+      "step": 1090
+    },
+    {
+      "epoch": 2.420410427066001,
+      "grad_norm": 0.3071598410606384,
+      "learning_rate": 6.395061728395061e-05,
+      "loss": 5.4766,
+      "step": 1091
+    },
+    {
+      "epoch": 2.4226289517470883,
+      "grad_norm": 0.23948736488819122,
+      "learning_rate": 6.37037037037037e-05,
+      "loss": 5.2652,
+      "step": 1092
+    },
+    {
+      "epoch": 2.4248474764281753,
+      "grad_norm": 0.2697477638721466,
+      "learning_rate": 6.345679012345678e-05,
+      "loss": 5.3718,
+      "step": 1093
+    },
+    {
+      "epoch": 2.4270660011092624,
+      "grad_norm": 0.2818484902381897,
+      "learning_rate": 6.320987654320987e-05,
+      "loss": 5.3712,
+      "step": 1094
+    },
+    {
+      "epoch": 2.4292845257903495,
+      "grad_norm": 0.2793472409248352,
+      "learning_rate": 6.296296296296295e-05,
+      "loss": 5.4039,
+      "step": 1095
+    },
+    {
+      "epoch": 2.4315030504714366,
+      "grad_norm": 0.2116321623325348,
+      "learning_rate": 6.271604938271604e-05,
+      "loss": 5.3526,
+      "step": 1096
+    },
+    {
+      "epoch": 2.4337215751525236,
+      "grad_norm": 0.2560115456581116,
+      "learning_rate": 6.246913580246912e-05,
+      "loss": 5.3044,
+      "step": 1097
+    },
+    {
+      "epoch": 2.4359400998336107,
+      "grad_norm": 0.31316524744033813,
+      "learning_rate": 6.222222222222222e-05,
+      "loss": 5.3194,
+      "step": 1098
+    },
+    {
+      "epoch": 2.438158624514698,
+      "grad_norm": 0.3625088036060333,
+      "learning_rate": 6.19753086419753e-05,
+      "loss": 5.1519,
+      "step": 1099
+    },
+    {
+      "epoch": 2.440377149195785,
+      "grad_norm": 0.34598520398139954,
+      "learning_rate": 6.17283950617284e-05,
+      "loss": 5.2085,
+      "step": 1100
+    },
+    {
+      "epoch": 2.442595673876872,
+      "grad_norm": 0.28281474113464355,
+      "learning_rate": 6.148148148148148e-05,
+      "loss": 5.4146,
+      "step": 1101
+    },
+    {
+      "epoch": 2.444814198557959,
+      "grad_norm": 0.32686668634414673,
+      "learning_rate": 6.123456790123456e-05,
+      "loss": 5.2631,
+      "step": 1102
+    },
+    {
+      "epoch": 2.447032723239046,
+      "grad_norm": 0.2726743817329407,
+      "learning_rate": 6.0987654320987655e-05,
+      "loss": 5.2797,
+      "step": 1103
+    },
+    {
+      "epoch": 2.449251247920133,
+      "grad_norm": 0.26691094040870667,
+      "learning_rate": 6.0740740740740736e-05,
+      "loss": 5.1877,
+      "step": 1104
+    },
+    {
+      "epoch": 2.4514697726012202,
+      "grad_norm": 0.2542896568775177,
+      "learning_rate": 6.049382716049382e-05,
+      "loss": 5.2457,
+      "step": 1105
+    },
+    {
+      "epoch": 2.4536882972823073,
+      "grad_norm": 0.3773442804813385,
+      "learning_rate": 6.0246913580246906e-05,
+      "loss": 5.1967,
+      "step": 1106
+    },
+    {
+      "epoch": 2.4559068219633944,
+      "grad_norm": 0.4310261905193329,
+      "learning_rate": 5.9999999999999995e-05,
+      "loss": 5.292,
+      "step": 1107
+    },
+    {
+      "epoch": 2.4581253466444815,
+      "grad_norm": 0.29710909724235535,
+      "learning_rate": 5.975308641975308e-05,
+      "loss": 5.2269,
+      "step": 1108
+    },
+    {
+      "epoch": 2.4603438713255685,
+      "grad_norm": 0.44203415513038635,
+      "learning_rate": 5.9506172839506165e-05,
+      "loss": 4.8899,
+      "step": 1109
+    },
+    {
+      "epoch": 2.4625623960066556,
+      "grad_norm": 0.1960148811340332,
+      "learning_rate": 5.925925925925925e-05,
+      "loss": 5.35,
+      "step": 1110
+    },
+    {
+      "epoch": 2.4647809206877427,
+      "grad_norm": 0.30657729506492615,
+      "learning_rate": 5.901234567901234e-05,
+      "loss": 5.3244,
+      "step": 1111
+    },
+    {
+      "epoch": 2.4669994453688298,
+      "grad_norm": 0.2113593965768814,
+      "learning_rate": 5.876543209876543e-05,
+      "loss": 5.1399,
+      "step": 1112
+    },
+    {
+      "epoch": 2.469217970049917,
+      "grad_norm": 0.4424862563610077,
+      "learning_rate": 5.851851851851851e-05,
+      "loss": 5.0886,
+      "step": 1113
+    },
+    {
+      "epoch": 2.471436494731004,
+      "grad_norm": 0.32460764050483704,
+      "learning_rate": 5.827160493827159e-05,
+      "loss": 5.5093,
+      "step": 1114
+    },
+    {
+      "epoch": 2.473655019412091,
+      "grad_norm": 0.4945247769355774,
+      "learning_rate": 5.802469135802469e-05,
+      "loss": 4.9839,
+      "step": 1115
+    },
+    {
+      "epoch": 2.475873544093178,
+      "grad_norm": 0.29972872138023376,
+      "learning_rate": 5.777777777777777e-05,
+      "loss": 5.4372,
+      "step": 1116
+    },
+    {
+      "epoch": 2.478092068774265,
+      "grad_norm": 0.3412848711013794,
+      "learning_rate": 5.753086419753086e-05,
+      "loss": 5.4982,
+      "step": 1117
+    },
+    {
+      "epoch": 2.480310593455352,
+      "grad_norm": 0.2872641384601593,
+      "learning_rate": 5.7283950617283946e-05,
+      "loss": 5.1745,
+      "step": 1118
+    },
+    {
+      "epoch": 2.4825291181364393,
+      "grad_norm": 0.3014221489429474,
+      "learning_rate": 5.7037037037037035e-05,
+      "loss": 5.4885,
+      "step": 1119
+    },
+    {
+      "epoch": 2.4847476428175264,
+      "grad_norm": 0.3246877193450928,
+      "learning_rate": 5.6790123456790116e-05,
+      "loss": 5.4656,
+      "step": 1120
+    },
+    {
+      "epoch": 2.4869661674986134,
+      "grad_norm": 0.2587421238422394,
+      "learning_rate": 5.6543209876543205e-05,
+      "loss": 5.2698,
+      "step": 1121
+    },
+    {
+      "epoch": 2.4891846921797005,
+      "grad_norm": 0.32959452271461487,
+      "learning_rate": 5.629629629629629e-05,
+      "loss": 5.3694,
+      "step": 1122
+    },
+    {
+      "epoch": 2.4914032168607876,
+      "grad_norm": 0.3469100892543793,
+      "learning_rate": 5.604938271604938e-05,
+      "loss": 5.0788,
+      "step": 1123
+    },
+    {
+      "epoch": 2.4936217415418747,
+      "grad_norm": 0.2824125587940216,
+      "learning_rate": 5.580246913580246e-05,
+      "loss": 5.2047,
+      "step": 1124
+    },
+    {
+      "epoch": 2.4958402662229617,
+      "grad_norm": 0.35667991638183594,
+      "learning_rate": 5.5555555555555545e-05,
+      "loss": 5.9037,
+      "step": 1125
+    },
+    {
+      "epoch": 2.498058790904049,
+      "grad_norm": 0.21590334177017212,
+      "learning_rate": 5.530864197530864e-05,
+      "loss": 5.6078,
+      "step": 1126
+    },
+    {
+      "epoch": 2.500277315585136,
+      "grad_norm": 0.3106406331062317,
+      "learning_rate": 5.506172839506172e-05,
+      "loss": 5.1875,
+      "step": 1127
+    },
+    {
+      "epoch": 2.502495840266223,
+      "grad_norm": 0.4681743383407593,
+      "learning_rate": 5.481481481481481e-05,
+      "loss": 5.4678,
+      "step": 1128
+    },
+    {
+      "epoch": 2.50471436494731,
+      "grad_norm": 0.28938785195350647,
+      "learning_rate": 5.456790123456789e-05,
+      "loss": 5.5124,
+      "step": 1129
+    },
+    {
+      "epoch": 2.506932889628397,
+      "grad_norm": 0.2987858057022095,
+      "learning_rate": 5.4320987654320986e-05,
+      "loss": 5.3042,
+      "step": 1130
+    },
+    {
+      "epoch": 2.509151414309484,
+      "grad_norm": 0.24429476261138916,
+      "learning_rate": 5.407407407407407e-05,
+      "loss": 5.301,
+      "step": 1131
+    },
+    {
+      "epoch": 2.5113699389905713,
+      "grad_norm": 0.3310264050960541,
+      "learning_rate": 5.3827160493827156e-05,
+      "loss": 5.3185,
+      "step": 1132
+    },
+    {
+      "epoch": 2.5135884636716583,
+      "grad_norm": 0.2519128620624542,
+      "learning_rate": 5.358024691358024e-05,
+      "loss": 5.2714,
+      "step": 1133
+    },
+    {
+      "epoch": 2.5158069883527454,
+      "grad_norm": 0.3855387568473816,
+      "learning_rate": 5.333333333333333e-05,
+      "loss": 5.4052,
+      "step": 1134
+    },
+    {
+      "epoch": 2.5180255130338325,
+      "grad_norm": 0.2603036165237427,
+      "learning_rate": 5.3086419753086415e-05,
+      "loss": 5.2747,
+      "step": 1135
+    },
+    {
+      "epoch": 2.5202440377149196,
+      "grad_norm": 0.3584638237953186,
+      "learning_rate": 5.2839506172839496e-05,
+      "loss": 5.3027,
+      "step": 1136
+    },
+    {
+      "epoch": 2.5224625623960066,
+      "grad_norm": 0.3059612512588501,
+      "learning_rate": 5.259259259259259e-05,
+      "loss": 5.2702,
+      "step": 1137
+    },
+    {
+      "epoch": 2.5246810870770937,
+      "grad_norm": 0.253773033618927,
+      "learning_rate": 5.234567901234567e-05,
+      "loss": 5.5383,
+      "step": 1138
+    },
+    {
+      "epoch": 2.526899611758181,
+      "grad_norm": 0.286532998085022,
+      "learning_rate": 5.209876543209876e-05,
+      "loss": 5.5625,
+      "step": 1139
+    },
+    {
+      "epoch": 2.529118136439268,
+      "grad_norm": 0.491559237241745,
+      "learning_rate": 5.185185185185184e-05,
+      "loss": 5.2429,
+      "step": 1140
+    },
+    {
+      "epoch": 2.531336661120355,
+      "grad_norm": 0.36755648255348206,
+      "learning_rate": 5.160493827160494e-05,
+      "loss": 5.2742,
+      "step": 1141
+    },
+    {
+      "epoch": 2.533555185801442,
+      "grad_norm": 0.2779437005519867,
+      "learning_rate": 5.135802469135802e-05,
+      "loss": 5.2372,
+      "step": 1142
+    },
+    {
+      "epoch": 2.535773710482529,
+      "grad_norm": 0.27602648735046387,
+      "learning_rate": 5.111111111111111e-05,
+      "loss": 5.3142,
+      "step": 1143
+    },
+    {
+      "epoch": 2.537992235163616,
+      "grad_norm": 0.27551576495170593,
+      "learning_rate": 5.086419753086419e-05,
+      "loss": 5.6086,
+      "step": 1144
+    },
+    {
+      "epoch": 2.5402107598447032,
+      "grad_norm": 0.29893508553504944,
+      "learning_rate": 5.0617283950617285e-05,
+      "loss": 5.2827,
+      "step": 1145
+    },
+    {
+      "epoch": 2.5424292845257903,
+      "grad_norm": 0.3390686810016632,
+      "learning_rate": 5.0370370370370366e-05,
+      "loss": 5.3852,
+      "step": 1146
+    },
+    {
+      "epoch": 2.5446478092068774,
+      "grad_norm": 0.33556997776031494,
+      "learning_rate": 5.012345679012345e-05,
+      "loss": 5.4553,
+      "step": 1147
+    },
+    {
+      "epoch": 2.5468663338879645,
+      "grad_norm": 0.2784704566001892,
+      "learning_rate": 4.9876543209876536e-05,
+      "loss": 5.4165,
+      "step": 1148
+    },
+    {
+      "epoch": 2.5490848585690515,
+      "grad_norm": 0.3496081531047821,
+      "learning_rate": 4.9629629629629625e-05,
+      "loss": 5.3527,
+      "step": 1149
+    },
+    {
+      "epoch": 2.5513033832501386,
+      "grad_norm": 0.34020179510116577,
+      "learning_rate": 4.938271604938271e-05,
+      "loss": 5.4007,
+      "step": 1150
+    },
+    {
+      "epoch": 2.5535219079312257,
+      "grad_norm": 0.26129862666130066,
+      "learning_rate": 4.9135802469135795e-05,
+      "loss": 5.283,
+      "step": 1151
+    },
+    {
+      "epoch": 2.5557404326123128,
+      "grad_norm": 0.26866698265075684,
+      "learning_rate": 4.888888888888889e-05,
+      "loss": 5.1608,
+      "step": 1152
+    },
+    {
+      "epoch": 2.5579589572934,
+      "grad_norm": 0.37572357058525085,
+      "learning_rate": 4.864197530864197e-05,
+      "loss": 5.4889,
+      "step": 1153
+    },
+    {
+      "epoch": 2.560177481974487,
+      "grad_norm": 0.2943112254142761,
+      "learning_rate": 4.839506172839506e-05,
+      "loss": 5.5124,
+      "step": 1154
+    },
+    {
+      "epoch": 2.562396006655574,
+      "grad_norm": 0.24817174673080444,
+      "learning_rate": 4.814814814814814e-05,
+      "loss": 5.3109,
+      "step": 1155
+    },
+    {
+      "epoch": 2.564614531336661,
+      "grad_norm": 0.5732212662696838,
+      "learning_rate": 4.7901234567901237e-05,
+      "loss": 5.1937,
+      "step": 1156
+    },
+    {
+      "epoch": 2.566833056017748,
+      "grad_norm": 0.2725611627101898,
+      "learning_rate": 4.765432098765432e-05,
+      "loss": 5.3224,
+      "step": 1157
+    },
+    {
+      "epoch": 2.569051580698835,
+      "grad_norm": 0.3332749009132385,
+      "learning_rate": 4.74074074074074e-05,
+      "loss": 5.641,
+      "step": 1158
+    },
+    {
+      "epoch": 2.5712701053799223,
+      "grad_norm": 0.29672205448150635,
+      "learning_rate": 4.716049382716049e-05,
+      "loss": 5.1794,
+      "step": 1159
+    },
+    {
+      "epoch": 2.5734886300610094,
+      "grad_norm": 0.24692672491073608,
+      "learning_rate": 4.6913580246913576e-05,
+      "loss": 5.0788,
+      "step": 1160
+    },
+    {
+      "epoch": 2.5757071547420964,
+      "grad_norm": 0.2875199615955353,
+      "learning_rate": 4.6666666666666665e-05,
+      "loss": 5.4423,
+      "step": 1161
+    },
+    {
+      "epoch": 2.5779256794231835,
+      "grad_norm": 0.3328399062156677,
+      "learning_rate": 4.6419753086419746e-05,
+      "loss": 5.0678,
+      "step": 1162
+    },
+    {
+      "epoch": 2.5801442041042706,
+      "grad_norm": 0.22561120986938477,
+      "learning_rate": 4.6172839506172835e-05,
+      "loss": 5.1625,
+      "step": 1163
+    },
+    {
+      "epoch": 2.5823627287853577,
+      "grad_norm": 0.30483946204185486,
+      "learning_rate": 4.592592592592592e-05,
+      "loss": 5.4013,
+      "step": 1164
+    },
+    {
+      "epoch": 2.5845812534664447,
+      "grad_norm": 0.23932000994682312,
+      "learning_rate": 4.567901234567901e-05,
+      "loss": 5.4177,
+      "step": 1165
+    },
+    {
+      "epoch": 2.586799778147532,
+      "grad_norm": 0.368051677942276,
+      "learning_rate": 4.543209876543209e-05,
+      "loss": 5.4202,
+      "step": 1166
+    },
+    {
+      "epoch": 2.589018302828619,
+      "grad_norm": 0.30605459213256836,
+      "learning_rate": 4.518518518518519e-05,
+      "loss": 5.447,
+      "step": 1167
+    },
+    {
+      "epoch": 2.591236827509706,
+      "grad_norm": 0.2281208336353302,
+      "learning_rate": 4.493827160493827e-05,
+      "loss": 5.2527,
+      "step": 1168
+    },
+    {
+      "epoch": 2.593455352190793,
+      "grad_norm": 0.36720696091651917,
+      "learning_rate": 4.469135802469135e-05,
+      "loss": 5.1909,
+      "step": 1169
+    },
+    {
+      "epoch": 2.59567387687188,
+      "grad_norm": 0.3549429476261139,
+      "learning_rate": 4.444444444444444e-05,
+      "loss": 5.4724,
+      "step": 1170
+    },
+    {
+      "epoch": 2.597892401552967,
+      "grad_norm": 0.21177859604358673,
+      "learning_rate": 4.419753086419753e-05,
+      "loss": 5.2581,
+      "step": 1171
+    },
+    {
+      "epoch": 2.6001109262340543,
+      "grad_norm": 0.2845475971698761,
+      "learning_rate": 4.3950617283950617e-05,
+      "loss": 5.5322,
+      "step": 1172
+    },
+    {
+      "epoch": 2.6023294509151413,
+      "grad_norm": 0.27170529961586,
+      "learning_rate": 4.37037037037037e-05,
+      "loss": 5.2308,
+      "step": 1173
+    },
+    {
+      "epoch": 2.6045479755962284,
+      "grad_norm": 0.2638813555240631,
+      "learning_rate": 4.3456790123456786e-05,
+      "loss": 5.3142,
+      "step": 1174
+    },
+    {
+      "epoch": 2.6067665002773155,
+      "grad_norm": 0.23649510741233826,
+      "learning_rate": 4.3209876543209875e-05,
+      "loss": 5.1707,
+      "step": 1175
+    },
+    {
+      "epoch": 2.6089850249584026,
+      "grad_norm": 0.2635962665081024,
+      "learning_rate": 4.296296296296296e-05,
+      "loss": 5.5224,
+      "step": 1176
+    },
+    {
+      "epoch": 2.6112035496394896,
+      "grad_norm": 0.2735772132873535,
+      "learning_rate": 4.2716049382716045e-05,
+      "loss": 5.5504,
+      "step": 1177
+    },
+    {
+      "epoch": 2.6134220743205767,
+      "grad_norm": 0.3349439799785614,
+      "learning_rate": 4.2469135802469126e-05,
+      "loss": 5.1301,
+      "step": 1178
+    },
+    {
+      "epoch": 2.615640599001664,
+      "grad_norm": 0.3301205635070801,
+      "learning_rate": 4.222222222222222e-05,
+      "loss": 5.6948,
+      "step": 1179
+    },
+    {
+      "epoch": 2.617859123682751,
+      "grad_norm": 0.3731234669685364,
+      "learning_rate": 4.19753086419753e-05,
+      "loss": 5.4289,
+      "step": 1180
+    },
+    {
+      "epoch": 2.620077648363838,
+      "grad_norm": 0.5486841201782227,
+      "learning_rate": 4.172839506172839e-05,
+      "loss": 5.4626,
+      "step": 1181
+    },
+    {
+      "epoch": 2.622296173044925,
+      "grad_norm": 0.3334302604198456,
+      "learning_rate": 4.148148148148147e-05,
+      "loss": 5.7461,
+      "step": 1182
+    },
+    {
+      "epoch": 2.624514697726012,
+      "grad_norm": 0.35488757491111755,
+      "learning_rate": 4.123456790123457e-05,
+      "loss": 5.071,
+      "step": 1183
+    },
+    {
+      "epoch": 2.626733222407099,
+      "grad_norm": 0.3549345135688782,
+      "learning_rate": 4.098765432098765e-05,
+      "loss": 5.305,
+      "step": 1184
+    },
+    {
+      "epoch": 2.6289517470881862,
+      "grad_norm": 0.3602257966995239,
+      "learning_rate": 4.074074074074074e-05,
+      "loss": 5.4473,
+      "step": 1185
+    },
+    {
+      "epoch": 2.6311702717692733,
+      "grad_norm": 0.3078634738922119,
+      "learning_rate": 4.0493827160493827e-05,
+      "loss": 5.3166,
+      "step": 1186
+    },
+    {
+      "epoch": 2.6333887964503604,
+      "grad_norm": 0.230717271566391,
+      "learning_rate": 4.0246913580246915e-05,
+      "loss": 5.4567,
+      "step": 1187
+    },
+    {
+      "epoch": 2.6356073211314475,
+      "grad_norm": 0.367397278547287,
+      "learning_rate": 3.9999999999999996e-05,
+      "loss": 5.3924,
+      "step": 1188
+    },
+    {
+      "epoch": 2.6378258458125345,
+      "grad_norm": 0.25954845547676086,
+      "learning_rate": 3.975308641975308e-05,
+      "loss": 5.1509,
+      "step": 1189
+    },
+    {
+      "epoch": 2.6400443704936216,
+      "grad_norm": 0.33391278982162476,
+      "learning_rate": 3.950617283950617e-05,
+      "loss": 5.4269,
+      "step": 1190
+    },
+    {
+      "epoch": 2.6422628951747087,
+      "grad_norm": 0.5180467963218689,
+      "learning_rate": 3.9259259259259255e-05,
+      "loss": 5.0661,
+      "step": 1191
+    },
+    {
+      "epoch": 2.6444814198557958,
+      "grad_norm": 0.22108538448810577,
+      "learning_rate": 3.901234567901234e-05,
+      "loss": 5.2844,
+      "step": 1192
+    },
+    {
+      "epoch": 2.646699944536883,
+      "grad_norm": 0.40545639395713806,
+      "learning_rate": 3.8765432098765425e-05,
+      "loss": 5.7328,
+      "step": 1193
+    },
+    {
+      "epoch": 2.64891846921797,
+      "grad_norm": 0.3981581926345825,
+      "learning_rate": 3.851851851851852e-05,
+      "loss": 5.3984,
+      "step": 1194
+    },
+    {
+      "epoch": 2.651136993899057,
+      "grad_norm": 0.3435943126678467,
+      "learning_rate": 3.82716049382716e-05,
+      "loss": 5.1652,
+      "step": 1195
+    },
+    {
+      "epoch": 2.653355518580144,
+      "grad_norm": 0.3751172125339508,
+      "learning_rate": 3.802469135802468e-05,
+      "loss": 5.1027,
+      "step": 1196
+    },
+    {
+      "epoch": 2.655574043261231,
+      "grad_norm": 0.28855839371681213,
+      "learning_rate": 3.777777777777777e-05,
+      "loss": 5.1177,
+      "step": 1197
+    },
+    {
+      "epoch": 2.657792567942318,
+      "grad_norm": 0.265729159116745,
+      "learning_rate": 3.7530864197530867e-05,
+      "loss": 5.3031,
+      "step": 1198
+    },
+    {
+      "epoch": 2.6600110926234053,
+      "grad_norm": 0.2024485319852829,
+      "learning_rate": 3.728395061728395e-05,
+      "loss": 5.3172,
+      "step": 1199
+    },
+    {
+      "epoch": 2.6622296173044924,
+      "grad_norm": 0.32965338230133057,
+      "learning_rate": 3.703703703703703e-05,
+      "loss": 5.2398,
+      "step": 1200
+    },
+    {
+      "epoch": 2.6644481419855794,
+      "grad_norm": 0.2866857349872589,
+      "learning_rate": 3.679012345679012e-05,
+      "loss": 5.2808,
+      "step": 1201
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 0.28987565636634827,
+      "learning_rate": 3.6543209876543207e-05,
+      "loss": 5.6073,
+      "step": 1202
+    },
+    {
+      "epoch": 2.6688851913477536,
+      "grad_norm": 0.2434692233800888,
+      "learning_rate": 3.6296296296296295e-05,
+      "loss": 5.2961,
+      "step": 1203
+    },
+    {
+      "epoch": 2.6711037160288407,
+      "grad_norm": 0.4112333357334137,
+      "learning_rate": 3.604938271604938e-05,
+      "loss": 5.6926,
+      "step": 1204
+    },
+    {
+      "epoch": 2.6733222407099277,
+      "grad_norm": 0.29736217856407166,
+      "learning_rate": 3.5802469135802465e-05,
+      "loss": 5.3853,
+      "step": 1205
+    },
+    {
+      "epoch": 2.675540765391015,
+      "grad_norm": 0.26408255100250244,
+      "learning_rate": 3.555555555555555e-05,
+      "loss": 5.2637,
+      "step": 1206
+    },
+    {
+      "epoch": 2.677759290072102,
+      "grad_norm": 0.30152517557144165,
+      "learning_rate": 3.5308641975308635e-05,
+      "loss": 5.4466,
+      "step": 1207
+    },
+    {
+      "epoch": 2.679977814753189,
+      "grad_norm": 0.29022765159606934,
+      "learning_rate": 3.506172839506172e-05,
+      "loss": 5.4397,
+      "step": 1208
+    },
+    {
+      "epoch": 2.682196339434276,
+      "grad_norm": 0.3077150285243988,
+      "learning_rate": 3.481481481481481e-05,
+      "loss": 5.2388,
+      "step": 1209
+    },
+    {
+      "epoch": 2.684414864115363,
+      "grad_norm": 0.3032444417476654,
+      "learning_rate": 3.45679012345679e-05,
+      "loss": 5.2435,
+      "step": 1210
+    },
+    {
+      "epoch": 2.68663338879645,
+      "grad_norm": 0.3223309814929962,
+      "learning_rate": 3.432098765432098e-05,
+      "loss": 5.3049,
+      "step": 1211
+    },
+    {
+      "epoch": 2.6888519134775377,
+      "grad_norm": 0.22808535397052765,
+      "learning_rate": 3.407407407407407e-05,
+      "loss": 5.6471,
+      "step": 1212
+    },
+    {
+      "epoch": 2.691070438158625,
+      "grad_norm": 0.34905630350112915,
+      "learning_rate": 3.382716049382716e-05,
+      "loss": 5.5567,
+      "step": 1213
+    },
+    {
+      "epoch": 2.693288962839712,
+      "grad_norm": 0.2083464413881302,
+      "learning_rate": 3.3580246913580247e-05,
+      "loss": 5.1654,
+      "step": 1214
+    },
+    {
+      "epoch": 2.695507487520799,
+      "grad_norm": 0.26644986867904663,
+      "learning_rate": 3.333333333333333e-05,
+      "loss": 5.2419,
+      "step": 1215
+    },
+    {
+      "epoch": 2.697726012201886,
+      "grad_norm": 0.2860036790370941,
+      "learning_rate": 3.3086419753086417e-05,
+      "loss": 5.1953,
+      "step": 1216
+    },
+    {
+      "epoch": 2.699944536882973,
+      "grad_norm": 0.37523239850997925,
+      "learning_rate": 3.2839506172839505e-05,
+      "loss": 5.3084,
+      "step": 1217
+    },
+    {
+      "epoch": 2.70216306156406,
+      "grad_norm": 0.29500287771224976,
+      "learning_rate": 3.2592592592592586e-05,
+      "loss": 5.5608,
+      "step": 1218
+    },
+    {
+      "epoch": 2.7043815862451472,
+      "grad_norm": 0.33670473098754883,
+      "learning_rate": 3.2345679012345675e-05,
+      "loss": 5.3564,
+      "step": 1219
+    },
+    {
+      "epoch": 2.7066001109262343,
+      "grad_norm": 0.2998238205909729,
+      "learning_rate": 3.209876543209876e-05,
+      "loss": 5.4191,
+      "step": 1220
+    },
+    {
+      "epoch": 2.7088186356073214,
+      "grad_norm": 0.35395529866218567,
+      "learning_rate": 3.185185185185185e-05,
+      "loss": 5.335,
+      "step": 1221
+    },
+    {
+      "epoch": 2.7110371602884085,
+      "grad_norm": 0.23540151119232178,
+      "learning_rate": 3.160493827160493e-05,
+      "loss": 5.3694,
+      "step": 1222
+    },
+    {
+      "epoch": 2.7132556849694955,
+      "grad_norm": 0.3453834652900696,
+      "learning_rate": 3.135802469135802e-05,
+      "loss": 5.4462,
+      "step": 1223
+    },
+    {
+      "epoch": 2.7154742096505826,
+      "grad_norm": 0.47278037667274475,
+      "learning_rate": 3.111111111111111e-05,
+      "loss": 5.531,
+      "step": 1224
+    },
+    {
+      "epoch": 2.7176927343316697,
+      "grad_norm": 0.30973273515701294,
+      "learning_rate": 3.08641975308642e-05,
+      "loss": 5.2872,
+      "step": 1225
+    },
+    {
+      "epoch": 2.7199112590127568,
+      "grad_norm": 0.264349102973938,
+      "learning_rate": 3.061728395061728e-05,
+      "loss": 5.0501,
+      "step": 1226
+    },
+    {
+      "epoch": 2.722129783693844,
+      "grad_norm": 0.32654666900634766,
+      "learning_rate": 3.0370370370370368e-05,
+      "loss": 5.247,
+      "step": 1227
+    },
+    {
+      "epoch": 2.724348308374931,
+      "grad_norm": 0.231369286775589,
+      "learning_rate": 3.0123456790123453e-05,
+      "loss": 5.4922,
+      "step": 1228
+    },
+    {
+      "epoch": 2.726566833056018,
+      "grad_norm": 0.3915281295776367,
+      "learning_rate": 2.987654320987654e-05,
+      "loss": 5.4201,
+      "step": 1229
+    },
+    {
+      "epoch": 2.728785357737105,
+      "grad_norm": 0.25998198986053467,
+      "learning_rate": 2.9629629629629627e-05,
+      "loss": 5.3079,
+      "step": 1230
+    },
+    {
+      "epoch": 2.731003882418192,
+      "grad_norm": 0.251535564661026,
+      "learning_rate": 2.9382716049382715e-05,
+      "loss": 5.6018,
+      "step": 1231
+    },
+    {
+      "epoch": 2.733222407099279,
+      "grad_norm": 0.450369268655777,
+      "learning_rate": 2.9135802469135796e-05,
+      "loss": 5.1633,
+      "step": 1232
+    },
+    {
+      "epoch": 2.7354409317803663,
+      "grad_norm": 0.3516184091567993,
+      "learning_rate": 2.8888888888888885e-05,
+      "loss": 5.1506,
+      "step": 1233
+    },
+    {
+      "epoch": 2.7376594564614534,
+      "grad_norm": 0.36717933416366577,
+      "learning_rate": 2.8641975308641973e-05,
+      "loss": 5.432,
+      "step": 1234
+    },
+    {
+      "epoch": 2.7398779811425404,
+      "grad_norm": 0.5635153651237488,
+      "learning_rate": 2.8395061728395058e-05,
+      "loss": 5.6064,
+      "step": 1235
+    },
+    {
+      "epoch": 2.7420965058236275,
+      "grad_norm": 0.28254491090774536,
+      "learning_rate": 2.8148148148148147e-05,
+      "loss": 5.2926,
+      "step": 1236
+    },
+    {
+      "epoch": 2.7443150305047146,
+      "grad_norm": 0.27664676308631897,
+      "learning_rate": 2.790123456790123e-05,
+      "loss": 5.3576,
+      "step": 1237
+    },
+    {
+      "epoch": 2.7465335551858017,
+      "grad_norm": 0.2542949616909027,
+      "learning_rate": 2.765432098765432e-05,
+      "loss": 5.3545,
+      "step": 1238
+    },
+    {
+      "epoch": 2.7487520798668887,
+      "grad_norm": 0.36289048194885254,
+      "learning_rate": 2.7407407407407405e-05,
+      "loss": 5.4027,
+      "step": 1239
+    },
+    {
+      "epoch": 2.750970604547976,
+      "grad_norm": 0.3777407705783844,
+      "learning_rate": 2.7160493827160493e-05,
+      "loss": 5.1518,
+      "step": 1240
+    },
+    {
+      "epoch": 2.753189129229063,
+      "grad_norm": 0.3103990852832794,
+      "learning_rate": 2.6913580246913578e-05,
+      "loss": 5.3512,
+      "step": 1241
+    },
+    {
+      "epoch": 2.75540765391015,
+      "grad_norm": 0.2951963543891907,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 5.6455,
+      "step": 1242
+    },
+    {
+      "epoch": 2.757626178591237,
+      "grad_norm": 0.31276029348373413,
+      "learning_rate": 2.6419753086419748e-05,
+      "loss": 5.1746,
+      "step": 1243
+    },
+    {
+      "epoch": 2.759844703272324,
+      "grad_norm": 0.2910645306110382,
+      "learning_rate": 2.6172839506172837e-05,
+      "loss": 5.4686,
+      "step": 1244
+    },
+    {
+      "epoch": 2.762063227953411,
+      "grad_norm": 0.33922600746154785,
+      "learning_rate": 2.592592592592592e-05,
+      "loss": 5.5508,
+      "step": 1245
+    },
+    {
+      "epoch": 2.7642817526344983,
+      "grad_norm": 0.2820761203765869,
+      "learning_rate": 2.567901234567901e-05,
+      "loss": 5.4248,
+      "step": 1246
+    },
+    {
+      "epoch": 2.7665002773155853,
+      "grad_norm": 0.31524378061294556,
+      "learning_rate": 2.5432098765432095e-05,
+      "loss": 5.3756,
+      "step": 1247
+    },
+    {
+      "epoch": 2.7687188019966724,
+      "grad_norm": 0.2558484673500061,
+      "learning_rate": 2.5185185185185183e-05,
+      "loss": 5.047,
+      "step": 1248
+    },
+    {
+      "epoch": 2.7709373266777595,
+      "grad_norm": 0.3089102804660797,
+      "learning_rate": 2.4938271604938268e-05,
+      "loss": 5.2901,
+      "step": 1249
+    },
+    {
+      "epoch": 2.7731558513588466,
+      "grad_norm": 0.2552051842212677,
+      "learning_rate": 2.4691358024691357e-05,
+      "loss": 5.1945,
+      "step": 1250
+    },
+    {
+      "epoch": 2.7753743760399336,
+      "grad_norm": 0.24549280107021332,
+      "learning_rate": 2.4444444444444445e-05,
+      "loss": 5.5403,
+      "step": 1251
+    },
+    {
+      "epoch": 2.7775929007210207,
+      "grad_norm": 0.28649234771728516,
+      "learning_rate": 2.419753086419753e-05,
+      "loss": 5.1108,
+      "step": 1252
+    },
+    {
+      "epoch": 2.779811425402108,
+      "grad_norm": 0.27364397048950195,
+      "learning_rate": 2.3950617283950618e-05,
+      "loss": 5.4951,
+      "step": 1253
+    },
+    {
+      "epoch": 2.782029950083195,
+      "grad_norm": 0.3319871127605438,
+      "learning_rate": 2.37037037037037e-05,
+      "loss": 5.3115,
+      "step": 1254
+    },
+    {
+      "epoch": 2.784248474764282,
+      "grad_norm": 0.3046015202999115,
+      "learning_rate": 2.3456790123456788e-05,
+      "loss": 5.5167,
+      "step": 1255
+    },
+    {
+      "epoch": 2.786466999445369,
+      "grad_norm": 0.2389862835407257,
+      "learning_rate": 2.3209876543209873e-05,
+      "loss": 5.1978,
+      "step": 1256
+    },
+    {
+      "epoch": 2.788685524126456,
+      "grad_norm": 0.2544116973876953,
+      "learning_rate": 2.296296296296296e-05,
+      "loss": 5.5738,
+      "step": 1257
+    },
+    {
+      "epoch": 2.790904048807543,
+      "grad_norm": 0.31389614939689636,
+      "learning_rate": 2.2716049382716047e-05,
+      "loss": 5.2303,
+      "step": 1258
+    },
+    {
+      "epoch": 2.7931225734886302,
+      "grad_norm": 0.29155874252319336,
+      "learning_rate": 2.2469135802469135e-05,
+      "loss": 5.4313,
+      "step": 1259
+    },
+    {
+      "epoch": 2.7953410981697173,
+      "grad_norm": 0.25418829917907715,
+      "learning_rate": 2.222222222222222e-05,
+      "loss": 5.3877,
+      "step": 1260
+    },
+    {
+      "epoch": 2.7975596228508044,
+      "grad_norm": 0.3132433593273163,
+      "learning_rate": 2.1975308641975308e-05,
+      "loss": 5.0854,
+      "step": 1261
+    },
+    {
+      "epoch": 2.7997781475318915,
+      "grad_norm": 0.36064741015434265,
+      "learning_rate": 2.1728395061728393e-05,
+      "loss": 5.1396,
+      "step": 1262
+    },
+    {
+      "epoch": 2.8019966722129785,
+      "grad_norm": 0.2926681637763977,
+      "learning_rate": 2.148148148148148e-05,
+      "loss": 5.1442,
+      "step": 1263
+    },
+    {
+      "epoch": 2.8042151968940656,
+      "grad_norm": 0.26540645956993103,
+      "learning_rate": 2.1234567901234563e-05,
+      "loss": 5.1844,
+      "step": 1264
+    },
+    {
+      "epoch": 2.8064337215751527,
+      "grad_norm": 0.26886364817619324,
+      "learning_rate": 2.098765432098765e-05,
+      "loss": 5.3585,
+      "step": 1265
+    },
+    {
+      "epoch": 2.8086522462562398,
+      "grad_norm": 0.31341826915740967,
+      "learning_rate": 2.0740740740740737e-05,
+      "loss": 5.1397,
+      "step": 1266
+    },
+    {
+      "epoch": 2.810870770937327,
+      "grad_norm": 0.34874504804611206,
+      "learning_rate": 2.0493827160493825e-05,
+      "loss": 5.482,
+      "step": 1267
+    },
+    {
+      "epoch": 2.813089295618414,
+      "grad_norm": 0.26580578088760376,
+      "learning_rate": 2.0246913580246913e-05,
+      "loss": 5.2398,
+      "step": 1268
+    },
+    {
+      "epoch": 2.815307820299501,
+      "grad_norm": 0.31819334626197815,
+      "learning_rate": 1.9999999999999998e-05,
+      "loss": 5.6104,
+      "step": 1269
+    },
+    {
+      "epoch": 2.817526344980588,
+      "grad_norm": 0.24842265248298645,
+      "learning_rate": 1.9753086419753087e-05,
+      "loss": 5.2521,
+      "step": 1270
+    },
+    {
+      "epoch": 2.819744869661675,
+      "grad_norm": 0.3569657802581787,
+      "learning_rate": 1.950617283950617e-05,
+      "loss": 5.2093,
+      "step": 1271
+    },
+    {
+      "epoch": 2.821963394342762,
+      "grad_norm": 0.29460278153419495,
+      "learning_rate": 1.925925925925926e-05,
+      "loss": 5.1674,
+      "step": 1272
+    },
+    {
+      "epoch": 2.8241819190238493,
+      "grad_norm": 0.2850482761859894,
+      "learning_rate": 1.901234567901234e-05,
+      "loss": 5.3629,
+      "step": 1273
+    },
+    {
+      "epoch": 2.8264004437049364,
+      "grad_norm": 0.3233969807624817,
+      "learning_rate": 1.8765432098765433e-05,
+      "loss": 5.149,
+      "step": 1274
+    },
+    {
+      "epoch": 2.8286189683860234,
+      "grad_norm": 0.24255545437335968,
+      "learning_rate": 1.8518518518518515e-05,
+      "loss": 5.3325,
+      "step": 1275
+    },
+    {
+      "epoch": 2.8308374930671105,
+      "grad_norm": 0.29345574975013733,
+      "learning_rate": 1.8271604938271603e-05,
+      "loss": 5.2511,
+      "step": 1276
+    },
+    {
+      "epoch": 2.8330560177481976,
+      "grad_norm": 0.30346032977104187,
+      "learning_rate": 1.802469135802469e-05,
+      "loss": 5.3355,
+      "step": 1277
+    },
+    {
+      "epoch": 2.8352745424292847,
+      "grad_norm": 0.31460854411125183,
+      "learning_rate": 1.7777777777777777e-05,
+      "loss": 5.502,
+      "step": 1278
+    },
+    {
+      "epoch": 2.8374930671103717,
+      "grad_norm": 0.27682924270629883,
+      "learning_rate": 1.753086419753086e-05,
+      "loss": 5.1538,
+      "step": 1279
+    },
+    {
+      "epoch": 2.839711591791459,
+      "grad_norm": 0.30318590998649597,
+      "learning_rate": 1.728395061728395e-05,
+      "loss": 5.4292,
+      "step": 1280
+    },
+    {
+      "epoch": 2.841930116472546,
+      "grad_norm": 0.2632366120815277,
+      "learning_rate": 1.7037037037037035e-05,
+      "loss": 5.0421,
+      "step": 1281
+    },
+    {
+      "epoch": 2.844148641153633,
+      "grad_norm": 0.32092002034187317,
+      "learning_rate": 1.6790123456790123e-05,
+      "loss": 5.5032,
+      "step": 1282
+    },
+    {
+      "epoch": 2.84636716583472,
+      "grad_norm": 0.40137261152267456,
+      "learning_rate": 1.6543209876543208e-05,
+      "loss": 5.1807,
+      "step": 1283
+    },
+    {
+      "epoch": 2.848585690515807,
+      "grad_norm": 0.2960865795612335,
+      "learning_rate": 1.6296296296296293e-05,
+      "loss": 5.4206,
+      "step": 1284
+    },
+    {
+      "epoch": 2.850804215196894,
+      "grad_norm": 0.24926912784576416,
+      "learning_rate": 1.604938271604938e-05,
+      "loss": 5.1226,
+      "step": 1285
+    },
+    {
+      "epoch": 2.8530227398779813,
+      "grad_norm": 0.2585188150405884,
+      "learning_rate": 1.5802469135802467e-05,
+      "loss": 5.1559,
+      "step": 1286
+    },
+    {
+      "epoch": 2.8552412645590683,
+      "grad_norm": 0.47920915484428406,
+      "learning_rate": 1.5555555555555555e-05,
+      "loss": 4.7971,
+      "step": 1287
+    },
+    {
+      "epoch": 2.8574597892401554,
+      "grad_norm": 0.2381104975938797,
+      "learning_rate": 1.530864197530864e-05,
+      "loss": 5.4434,
+      "step": 1288
+    },
+    {
+      "epoch": 2.8596783139212425,
+      "grad_norm": 0.423104852437973,
+      "learning_rate": 1.5061728395061727e-05,
+      "loss": 5.1869,
+      "step": 1289
+    },
+    {
+      "epoch": 2.8618968386023296,
+      "grad_norm": 0.229563906788826,
+      "learning_rate": 1.4814814814814813e-05,
+      "loss": 5.1568,
+      "step": 1290
+    },
+    {
+      "epoch": 2.8641153632834166,
+      "grad_norm": 0.3002553880214691,
+      "learning_rate": 1.4567901234567898e-05,
+      "loss": 5.1969,
+      "step": 1291
+    },
+    {
+      "epoch": 2.8663338879645037,
+      "grad_norm": 0.3251195251941681,
+      "learning_rate": 1.4320987654320987e-05,
+      "loss": 5.0613,
+      "step": 1292
+    },
+    {
+      "epoch": 2.868552412645591,
+      "grad_norm": 0.2975667715072632,
+      "learning_rate": 1.4074074074074073e-05,
+      "loss": 5.2185,
+      "step": 1293
+    },
+    {
+      "epoch": 2.870770937326678,
+      "grad_norm": 0.5068483352661133,
+      "learning_rate": 1.382716049382716e-05,
+      "loss": 5.4351,
+      "step": 1294
+    },
+    {
+      "epoch": 2.872989462007765,
+      "grad_norm": 0.2540512681007385,
+      "learning_rate": 1.3580246913580247e-05,
+      "loss": 5.3066,
+      "step": 1295
+    },
+    {
+      "epoch": 2.875207986688852,
+      "grad_norm": 0.24906310439109802,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 5.4611,
+      "step": 1296
+    },
+    {
+      "epoch": 2.877426511369939,
+      "grad_norm": 0.4097791016101837,
+      "learning_rate": 1.3086419753086418e-05,
+      "loss": 5.0192,
+      "step": 1297
+    },
+    {
+      "epoch": 2.879645036051026,
+      "grad_norm": 0.306248277425766,
+      "learning_rate": 1.2839506172839505e-05,
+      "loss": 5.3281,
+      "step": 1298
+    },
+    {
+      "epoch": 2.8818635607321132,
+      "grad_norm": 0.268237829208374,
+      "learning_rate": 1.2592592592592592e-05,
+      "loss": 5.4258,
+      "step": 1299
+    },
+    {
+      "epoch": 2.8840820854132003,
+      "grad_norm": 0.3661245107650757,
+      "learning_rate": 1.2345679012345678e-05,
+      "loss": 5.2189,
+      "step": 1300
+    },
+    {
+      "epoch": 2.8863006100942874,
+      "grad_norm": 0.25602731108665466,
+      "learning_rate": 1.2098765432098765e-05,
+      "loss": 5.1307,
+      "step": 1301
+    },
+    {
+      "epoch": 2.8885191347753745,
+      "grad_norm": 0.3084339201450348,
+      "learning_rate": 1.185185185185185e-05,
+      "loss": 5.3281,
+      "step": 1302
+    },
+    {
+      "epoch": 2.8907376594564616,
+      "grad_norm": 0.26393452286720276,
+      "learning_rate": 1.1604938271604937e-05,
+      "loss": 5.0966,
+      "step": 1303
+    },
+    {
+      "epoch": 2.8929561841375486,
+      "grad_norm": 0.32150304317474365,
+      "learning_rate": 1.1358024691358023e-05,
+      "loss": 5.116,
+      "step": 1304
+    },
+    {
+      "epoch": 2.8951747088186357,
+      "grad_norm": 0.25514769554138184,
+      "learning_rate": 1.111111111111111e-05,
+      "loss": 5.2178,
+      "step": 1305
+    },
+    {
+      "epoch": 2.8973932334997228,
+      "grad_norm": 0.24598628282546997,
+      "learning_rate": 1.0864197530864197e-05,
+      "loss": 5.3293,
+      "step": 1306
+    },
+    {
+      "epoch": 2.89961175818081,
+      "grad_norm": 0.3140576481819153,
+      "learning_rate": 1.0617283950617282e-05,
+      "loss": 5.3597,
+      "step": 1307
+    },
+    {
+      "epoch": 2.901830282861897,
+      "grad_norm": 0.2074941247701645,
+      "learning_rate": 1.0370370370370368e-05,
+      "loss": 5.436,
+      "step": 1308
+    },
+    {
+      "epoch": 2.904048807542984,
+      "grad_norm": 0.29329895973205566,
+      "learning_rate": 1.0123456790123457e-05,
+      "loss": 5.1038,
+      "step": 1309
+    },
+    {
+      "epoch": 2.906267332224071,
+      "grad_norm": 0.5422505140304565,
+      "learning_rate": 9.876543209876543e-06,
+      "loss": 5.0648,
+      "step": 1310
+    },
+    {
+      "epoch": 2.908485856905158,
+      "grad_norm": 0.36102378368377686,
+      "learning_rate": 9.62962962962963e-06,
+      "loss": 5.3377,
+      "step": 1311
+    },
+    {
+      "epoch": 2.9107043815862452,
+      "grad_norm": 0.24145841598510742,
+      "learning_rate": 9.382716049382717e-06,
+      "loss": 5.1834,
+      "step": 1312
+    },
+    {
+      "epoch": 2.9129229062673323,
+      "grad_norm": 0.29608622193336487,
+      "learning_rate": 9.135802469135802e-06,
+      "loss": 5.0764,
+      "step": 1313
+    },
+    {
+      "epoch": 2.9151414309484194,
+      "grad_norm": 0.3211316466331482,
+      "learning_rate": 8.888888888888888e-06,
+      "loss": 5.1018,
+      "step": 1314
+    },
+    {
+      "epoch": 2.9173599556295065,
+      "grad_norm": 0.31115204095840454,
+      "learning_rate": 8.641975308641975e-06,
+      "loss": 5.2415,
+      "step": 1315
+    },
+    {
+      "epoch": 2.9195784803105935,
+      "grad_norm": 0.3673313856124878,
+      "learning_rate": 8.395061728395062e-06,
+      "loss": 5.5613,
+      "step": 1316
+    },
+    {
+      "epoch": 2.9217970049916806,
+      "grad_norm": 0.2981352210044861,
+      "learning_rate": 8.148148148148147e-06,
+      "loss": 5.1247,
+      "step": 1317
+    },
+    {
+      "epoch": 2.9240155296727677,
+      "grad_norm": 0.46185657382011414,
+      "learning_rate": 7.901234567901233e-06,
+      "loss": 5.4439,
+      "step": 1318
+    },
+    {
+      "epoch": 2.9262340543538548,
+      "grad_norm": 0.2616991102695465,
+      "learning_rate": 7.65432098765432e-06,
+      "loss": 5.4004,
+      "step": 1319
+    },
+    {
+      "epoch": 2.928452579034942,
+      "grad_norm": 0.40473881363868713,
+      "learning_rate": 7.407407407407407e-06,
+      "loss": 5.619,
+      "step": 1320
+    },
+    {
+      "epoch": 2.930671103716029,
+      "grad_norm": 0.257402241230011,
+      "learning_rate": 7.160493827160493e-06,
+      "loss": 5.3289,
+      "step": 1321
+    },
+    {
+      "epoch": 2.932889628397116,
+      "grad_norm": 0.2515251934528351,
+      "learning_rate": 6.91358024691358e-06,
+      "loss": 5.3676,
+      "step": 1322
+    },
+    {
+      "epoch": 2.935108153078203,
+      "grad_norm": 0.358523428440094,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 5.2094,
+      "step": 1323
+    },
+    {
+      "epoch": 2.93732667775929,
+      "grad_norm": 0.2683858275413513,
+      "learning_rate": 6.4197530864197525e-06,
+      "loss": 5.3003,
+      "step": 1324
+    },
+    {
+      "epoch": 2.939545202440377,
+      "grad_norm": 0.4365321099758148,
+      "learning_rate": 6.172839506172839e-06,
+      "loss": 5.5619,
+      "step": 1325
+    },
+    {
+      "epoch": 2.9417637271214643,
+      "grad_norm": 0.34674835205078125,
+      "learning_rate": 5.925925925925925e-06,
+      "loss": 5.4672,
+      "step": 1326
+    },
+    {
+      "epoch": 2.9439822518025514,
+      "grad_norm": 0.27708345651626587,
+      "learning_rate": 5.679012345679012e-06,
+      "loss": 5.2673,
+      "step": 1327
+    },
+    {
+      "epoch": 2.9462007764836384,
+      "grad_norm": 0.28084462881088257,
+      "learning_rate": 5.432098765432098e-06,
+      "loss": 5.5228,
+      "step": 1328
+    },
+    {
+      "epoch": 2.9484193011647255,
+      "grad_norm": 0.25741901993751526,
+      "learning_rate": 5.185185185185184e-06,
+      "loss": 5.1345,
+      "step": 1329
+    },
+    {
+      "epoch": 2.9506378258458126,
+      "grad_norm": 0.2840319871902466,
+      "learning_rate": 4.938271604938272e-06,
+      "loss": 5.404,
+      "step": 1330
+    },
+    {
+      "epoch": 2.9528563505268997,
+      "grad_norm": 0.30760452151298523,
+      "learning_rate": 4.691358024691358e-06,
+      "loss": 5.7036,
+      "step": 1331
+    },
+    {
+      "epoch": 2.9550748752079867,
+      "grad_norm": 0.34554052352905273,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 5.4521,
+      "step": 1332
+    },
+    {
+      "epoch": 2.957293399889074,
+      "grad_norm": 0.45375654101371765,
+      "learning_rate": 4.197530864197531e-06,
+      "loss": 5.1218,
+      "step": 1333
+    },
+    {
+      "epoch": 2.959511924570161,
+      "grad_norm": 0.5373749136924744,
+      "learning_rate": 3.950617283950617e-06,
+      "loss": 5.3386,
+      "step": 1334
+    },
+    {
+      "epoch": 2.961730449251248,
+      "grad_norm": 0.3119126260280609,
+      "learning_rate": 3.7037037037037033e-06,
+      "loss": 5.2403,
+      "step": 1335
+    },
+    {
+      "epoch": 2.963948973932335,
+      "grad_norm": 0.3123949468135834,
+      "learning_rate": 3.45679012345679e-06,
+      "loss": 5.3343,
+      "step": 1336
+    },
+    {
+      "epoch": 2.966167498613422,
+      "grad_norm": 0.2617880403995514,
+      "learning_rate": 3.2098765432098762e-06,
+      "loss": 5.1726,
+      "step": 1337
+    },
+    {
+      "epoch": 2.968386023294509,
+      "grad_norm": 0.4033494293689728,
+      "learning_rate": 2.9629629629629625e-06,
+      "loss": 5.4485,
+      "step": 1338
+    },
+    {
+      "epoch": 2.9706045479755963,
+      "grad_norm": 0.34001246094703674,
+      "learning_rate": 2.716049382716049e-06,
+      "loss": 5.5537,
+      "step": 1339
+    },
+    {
+      "epoch": 2.9728230726566833,
+      "grad_norm": 0.2771886885166168,
+      "learning_rate": 2.469135802469136e-06,
+      "loss": 5.2143,
+      "step": 1340
+    },
+    {
+      "epoch": 2.9750415973377704,
+      "grad_norm": 0.5141960978507996,
+      "learning_rate": 2.222222222222222e-06,
+      "loss": 5.1099,
+      "step": 1341
+    },
+    {
+      "epoch": 2.9772601220188575,
+      "grad_norm": 0.32522207498550415,
+      "learning_rate": 1.9753086419753083e-06,
+      "loss": 5.3053,
+      "step": 1342
+    },
+    {
+      "epoch": 2.9794786466999446,
+      "grad_norm": 0.2551051080226898,
+      "learning_rate": 1.728395061728395e-06,
+      "loss": 5.4307,
+      "step": 1343
+    },
+    {
+      "epoch": 2.9816971713810316,
+      "grad_norm": 0.36061891913414,
+      "learning_rate": 1.4814814814814812e-06,
+      "loss": 5.4128,
+      "step": 1344
+    },
+    {
+      "epoch": 2.9839156960621187,
+      "grad_norm": 0.3414318561553955,
+      "learning_rate": 1.234567901234568e-06,
+      "loss": 5.5602,
+      "step": 1345
+    },
+    {
+      "epoch": 2.986134220743206,
+      "grad_norm": 0.5490456819534302,
+      "learning_rate": 9.876543209876542e-07,
+      "loss": 5.4759,
+      "step": 1346
+    },
+    {
+      "epoch": 2.988352745424293,
+      "grad_norm": 0.27826812863349915,
+      "learning_rate": 7.407407407407406e-07,
+      "loss": 5.2321,
+      "step": 1347
+    },
+    {
+      "epoch": 2.99057127010538,
+      "grad_norm": 0.2671157717704773,
+      "learning_rate": 4.938271604938271e-07,
+      "loss": 5.4813,
+      "step": 1348
+    },
+    {
+      "epoch": 2.992789794786467,
+      "grad_norm": 0.32477816939353943,
+      "learning_rate": 2.4691358024691354e-07,
+      "loss": 5.3214,
+      "step": 1349
+    },
+    {
+      "epoch": 2.995008319467554,
+      "grad_norm": 0.37242522835731506,
+      "learning_rate": 0.0,
+      "loss": 5.3868,
+      "step": 1350
+    },
+    {
+      "epoch": 2.995008319467554,
+      "step": 1350,
+      "total_flos": 2.621662698626089e+17,
+      "train_loss": 6.214993870699847,
+      "train_runtime": 2348.8903,
+      "train_samples_per_second": 9.21,
+      "train_steps_per_second": 0.575
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1350,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.621662698626089e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}