diff --git "a/outputs/qlora-out/checkpoint-2655/trainer_state.json" "b/outputs/qlora-out/checkpoint-2655/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/outputs/qlora-out/checkpoint-2655/trainer_state.json"
@@ -0,0 +1,18698 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.994915254237288,
+  "eval_steps": 295,
+  "global_step": 2655,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0011296243998870376,
+      "grad_norm": 0.03433802351355553,
+      "learning_rate": 2e-05,
+      "loss": 1.1396,
+      "step": 1
+    },
+    {
+      "epoch": 0.0011296243998870376,
+      "eval_loss": 1.1312694549560547,
+      "eval_runtime": 554.2279,
+      "eval_samples_per_second": 17.652,
+      "eval_steps_per_second": 8.827,
+      "step": 1
+    },
+    {
+      "epoch": 0.0022592487997740753,
+      "grad_norm": 0.03576268255710602,
+      "learning_rate": 4e-05,
+      "loss": 1.1188,
+      "step": 2
+    },
+    {
+      "epoch": 0.003388873199661113,
+      "grad_norm": 0.03207174688577652,
+      "learning_rate": 6e-05,
+      "loss": 1.2194,
+      "step": 3
+    },
+    {
+      "epoch": 0.0045184975995481505,
+      "grad_norm": 0.03231927007436752,
+      "learning_rate": 8e-05,
+      "loss": 1.1735,
+      "step": 4
+    },
+    {
+      "epoch": 0.005648121999435188,
+      "grad_norm": 0.03306754678487778,
+      "learning_rate": 0.0001,
+      "loss": 1.1689,
+      "step": 5
+    },
+    {
+      "epoch": 0.006777746399322226,
+      "grad_norm": 0.035009413957595825,
+      "learning_rate": 0.00012,
+      "loss": 1.1808,
+      "step": 6
+    },
+    {
+      "epoch": 0.007907370799209263,
+      "grad_norm": 0.035900842398405075,
+      "learning_rate": 0.00014,
+      "loss": 1.0441,
+      "step": 7
+    },
+    {
+      "epoch": 0.009036995199096301,
+      "grad_norm": 0.028419604524970055,
+      "learning_rate": 0.00016,
+      "loss": 1.0858,
+      "step": 8
+    },
+    {
+      "epoch": 0.010166619598983339,
+      "grad_norm": 0.024826928973197937,
+      "learning_rate": 0.00018,
+      "loss": 1.1937,
+      "step": 9
+    },
+    {
+      "epoch": 0.011296243998870376,
+      "grad_norm": 0.026519587263464928,
+      "learning_rate": 0.0002,
+      "loss": 1.0726,
+      "step": 10
+    },
+    {
+      "epoch": 0.012425868398757414,
+      "grad_norm": 0.024713166058063507,
+      "learning_rate": 0.00019999992946277893,
+      "loss": 1.1191,
+      "step": 11
+    },
+    {
+      "epoch": 0.013555492798644452,
+      "grad_norm": 0.02494538575410843,
+      "learning_rate": 0.00019999971785121523,
+      "loss": 1.1751,
+      "step": 12
+    },
+    {
+      "epoch": 0.014685117198531489,
+      "grad_norm": 0.023202786222100258,
+      "learning_rate": 0.00019999936516560744,
+      "loss": 1.1801,
+      "step": 13
+    },
+    {
+      "epoch": 0.015814741598418527,
+      "grad_norm": 0.021448194980621338,
+      "learning_rate": 0.00019999887140645308,
+      "loss": 1.1676,
+      "step": 14
+    },
+    {
+      "epoch": 0.016944365998305563,
+      "grad_norm": 0.021449347957968712,
+      "learning_rate": 0.00019999823657444873,
+      "loss": 1.1224,
+      "step": 15
+    },
+    {
+      "epoch": 0.018073990398192602,
+      "grad_norm": 0.021438075229525566,
+      "learning_rate": 0.00019999746067049,
+      "loss": 1.0713,
+      "step": 16
+    },
+    {
+      "epoch": 0.019203614798079638,
+      "grad_norm": 0.020727725699543953,
+      "learning_rate": 0.00019999654369567147,
+      "loss": 1.1317,
+      "step": 17
+    },
+    {
+      "epoch": 0.020333239197966677,
+      "grad_norm": 0.023839153349399567,
+      "learning_rate": 0.00019999548565128678,
+      "loss": 0.994,
+      "step": 18
+    },
+    {
+      "epoch": 0.021462863597853713,
+      "grad_norm": 0.020854901522397995,
+      "learning_rate": 0.0001999942865388285,
+      "loss": 1.0086,
+      "step": 19
+    },
+    {
+      "epoch": 0.022592487997740753,
+      "grad_norm": 0.019373737275600433,
+      "learning_rate": 0.00019999294635998833,
+      "loss": 1.0284,
+      "step": 20
+    },
+    {
+      "epoch": 0.02372211239762779,
+      "grad_norm": 0.020383458584547043,
+      "learning_rate": 0.00019999146511665692,
+      "loss": 1.0926,
+      "step": 21
+    },
+    {
+      "epoch": 0.024851736797514828,
+      "grad_norm": 0.018995020538568497,
+      "learning_rate": 0.0001999898428109239,
+      "loss": 1.0333,
+      "step": 22
+    },
+    {
+      "epoch": 0.025981361197401864,
+      "grad_norm": 0.018192732706665993,
+      "learning_rate": 0.00019998807944507791,
+      "loss": 1.1471,
+      "step": 23
+    },
+    {
+      "epoch": 0.027110985597288903,
+      "grad_norm": 0.017383620142936707,
+      "learning_rate": 0.00019998617502160664,
+      "loss": 1.0417,
+      "step": 24
+    },
+    {
+      "epoch": 0.02824060999717594,
+      "grad_norm": 0.0169441569596529,
+      "learning_rate": 0.00019998412954319675,
+      "loss": 1.0375,
+      "step": 25
+    },
+    {
+      "epoch": 0.029370234397062978,
+      "grad_norm": 0.017483294010162354,
+      "learning_rate": 0.0001999819430127339,
+      "loss": 1.1357,
+      "step": 26
+    },
+    {
+      "epoch": 0.030499858796950014,
+      "grad_norm": 0.016769247129559517,
+      "learning_rate": 0.00019997961543330269,
+      "loss": 1.0641,
+      "step": 27
+    },
+    {
+      "epoch": 0.031629483196837054,
+      "grad_norm": 0.017466655001044273,
+      "learning_rate": 0.00019997714680818673,
+      "loss": 1.0919,
+      "step": 28
+    },
+    {
+      "epoch": 0.03275910759672409,
+      "grad_norm": 0.01684914343059063,
+      "learning_rate": 0.00019997453714086866,
+      "loss": 1.1458,
+      "step": 29
+    },
+    {
+      "epoch": 0.033888731996611125,
+      "grad_norm": 0.017144447192549706,
+      "learning_rate": 0.00019997178643503004,
+      "loss": 1.0499,
+      "step": 30
+    },
+    {
+      "epoch": 0.03501835639649816,
+      "grad_norm": 0.01914755254983902,
+      "learning_rate": 0.0001999688946945514,
+      "loss": 1.03,
+      "step": 31
+    },
+    {
+      "epoch": 0.036147980796385204,
+      "grad_norm": 0.017046675086021423,
+      "learning_rate": 0.00019996586192351225,
+      "loss": 1.0747,
+      "step": 32
+    },
+    {
+      "epoch": 0.03727760519627224,
+      "grad_norm": 0.01652969978749752,
+      "learning_rate": 0.00019996268812619107,
+      "loss": 1.1694,
+      "step": 33
+    },
+    {
+      "epoch": 0.038407229596159276,
+      "grad_norm": 0.016702843829989433,
+      "learning_rate": 0.00019995937330706526,
+      "loss": 1.0339,
+      "step": 34
+    },
+    {
+      "epoch": 0.03953685399604631,
+      "grad_norm": 0.016058262437582016,
+      "learning_rate": 0.00019995591747081122,
+      "loss": 1.1002,
+      "step": 35
+    },
+    {
+      "epoch": 0.040666478395933355,
+      "grad_norm": 0.01609026826918125,
+      "learning_rate": 0.0001999523206223042,
+      "loss": 1.0934,
+      "step": 36
+    },
+    {
+      "epoch": 0.04179610279582039,
+      "grad_norm": 0.017004741355776787,
+      "learning_rate": 0.00019994858276661844,
+      "loss": 1.0748,
+      "step": 37
+    },
+    {
+      "epoch": 0.042925727195707426,
+      "grad_norm": 0.017046496272087097,
+      "learning_rate": 0.00019994470390902712,
+      "loss": 1.0851,
+      "step": 38
+    },
+    {
+      "epoch": 0.04405535159559446,
+      "grad_norm": 0.016196010634303093,
+      "learning_rate": 0.0001999406840550023,
+      "loss": 0.9796,
+      "step": 39
+    },
+    {
+      "epoch": 0.045184975995481505,
+      "grad_norm": 0.01703856885433197,
+      "learning_rate": 0.000199936523210215,
+      "loss": 1.0511,
+      "step": 40
+    },
+    {
+      "epoch": 0.04631460039536854,
+      "grad_norm": 0.017249640077352524,
+      "learning_rate": 0.00019993222138053507,
+      "loss": 0.9521,
+      "step": 41
+    },
+    {
+      "epoch": 0.04744422479525558,
+      "grad_norm": 0.01665697991847992,
+      "learning_rate": 0.0001999277785720313,
+      "loss": 1.0732,
+      "step": 42
+    },
+    {
+      "epoch": 0.04857384919514261,
+      "grad_norm": 0.016691800206899643,
+      "learning_rate": 0.0001999231947909714,
+      "loss": 1.1504,
+      "step": 43
+    },
+    {
+      "epoch": 0.049703473595029656,
+      "grad_norm": 0.016521582379937172,
+      "learning_rate": 0.00019991847004382186,
+      "loss": 1.1022,
+      "step": 44
+    },
+    {
+      "epoch": 0.05083309799491669,
+      "grad_norm": 0.017150631174445152,
+      "learning_rate": 0.00019991360433724813,
+      "loss": 1.0843,
+      "step": 45
+    },
+    {
+      "epoch": 0.05196272239480373,
+      "grad_norm": 0.016666896641254425,
+      "learning_rate": 0.00019990859767811444,
+      "loss": 1.0328,
+      "step": 46
+    },
+    {
+      "epoch": 0.05309234679469076,
+      "grad_norm": 0.016831668093800545,
+      "learning_rate": 0.0001999034500734839,
+      "loss": 1.0285,
+      "step": 47
+    },
+    {
+      "epoch": 0.054221971194577806,
+      "grad_norm": 0.01656479388475418,
+      "learning_rate": 0.00019989816153061853,
+      "loss": 0.9825,
+      "step": 48
+    },
+    {
+      "epoch": 0.05535159559446484,
+      "grad_norm": 0.016812896355986595,
+      "learning_rate": 0.00019989273205697904,
+      "loss": 1.0912,
+      "step": 49
+    },
+    {
+      "epoch": 0.05648121999435188,
+      "grad_norm": 0.01740356534719467,
+      "learning_rate": 0.0001998871616602251,
+      "loss": 1.0601,
+      "step": 50
+    },
+    {
+      "epoch": 0.057610844394238914,
+      "grad_norm": 0.017337938770651817,
+      "learning_rate": 0.00019988145034821502,
+      "loss": 1.1482,
+      "step": 51
+    },
+    {
+      "epoch": 0.058740468794125957,
+      "grad_norm": 0.017546923831105232,
+      "learning_rate": 0.0001998755981290061,
+      "loss": 1.0472,
+      "step": 52
+    },
+    {
+      "epoch": 0.05987009319401299,
+      "grad_norm": 0.017129387706518173,
+      "learning_rate": 0.00019986960501085428,
+      "loss": 1.1083,
+      "step": 53
+    },
+    {
+      "epoch": 0.06099971759390003,
+      "grad_norm": 0.01688031479716301,
+      "learning_rate": 0.00019986347100221433,
+      "loss": 1.0589,
+      "step": 54
+    },
+    {
+      "epoch": 0.062129341993787064,
+      "grad_norm": 0.017208745703101158,
+      "learning_rate": 0.00019985719611173973,
+      "loss": 1.1292,
+      "step": 55
+    },
+    {
+      "epoch": 0.06325896639367411,
+      "grad_norm": 0.01749996654689312,
+      "learning_rate": 0.0001998507803482828,
+      "loss": 1.1252,
+      "step": 56
+    },
+    {
+      "epoch": 0.06438859079356114,
+      "grad_norm": 0.017861831933259964,
+      "learning_rate": 0.00019984422372089453,
+      "loss": 1.104,
+      "step": 57
+    },
+    {
+      "epoch": 0.06551821519344818,
+      "grad_norm": 0.01797177456319332,
+      "learning_rate": 0.00019983752623882462,
+      "loss": 1.0569,
+      "step": 58
+    },
+    {
+      "epoch": 0.06664783959333521,
+      "grad_norm": 0.01702985167503357,
+      "learning_rate": 0.00019983068791152152,
+      "loss": 1.0238,
+      "step": 59
+    },
+    {
+      "epoch": 0.06777746399322225,
+      "grad_norm": 0.01766786351799965,
+      "learning_rate": 0.00019982370874863236,
+      "loss": 1.0509,
+      "step": 60
+    },
+    {
+      "epoch": 0.06890708839310929,
+      "grad_norm": 0.017592614516615868,
+      "learning_rate": 0.00019981658876000298,
+      "loss": 1.0613,
+      "step": 61
+    },
+    {
+      "epoch": 0.07003671279299632,
+      "grad_norm": 0.018118126317858696,
+      "learning_rate": 0.00019980932795567782,
+      "loss": 1.1727,
+      "step": 62
+    },
+    {
+      "epoch": 0.07116633719288337,
+      "grad_norm": 0.017846032977104187,
+      "learning_rate": 0.00019980192634590007,
+      "loss": 1.0042,
+      "step": 63
+    },
+    {
+      "epoch": 0.07229596159277041,
+      "grad_norm": 0.018457984551787376,
+      "learning_rate": 0.00019979438394111145,
+      "loss": 1.0648,
+      "step": 64
+    },
+    {
+      "epoch": 0.07342558599265744,
+      "grad_norm": 0.019480090588331223,
+      "learning_rate": 0.0001997867007519524,
+      "loss": 1.0307,
+      "step": 65
+    },
+    {
+      "epoch": 0.07455521039254448,
+      "grad_norm": 0.018102938309311867,
+      "learning_rate": 0.00019977887678926195,
+      "loss": 1.1129,
+      "step": 66
+    },
+    {
+      "epoch": 0.07568483479243152,
+      "grad_norm": 0.017858000472187996,
+      "learning_rate": 0.00019977091206407768,
+      "loss": 1.1574,
+      "step": 67
+    },
+    {
+      "epoch": 0.07681445919231855,
+      "grad_norm": 0.019296329468488693,
+      "learning_rate": 0.0001997628065876358,
+      "loss": 1.046,
+      "step": 68
+    },
+    {
+      "epoch": 0.07794408359220559,
+      "grad_norm": 0.017915885895490646,
+      "learning_rate": 0.0001997545603713711,
+      "loss": 1.0357,
+      "step": 69
+    },
+    {
+      "epoch": 0.07907370799209262,
+      "grad_norm": 0.01840789057314396,
+      "learning_rate": 0.00019974617342691678,
+      "loss": 1.0383,
+      "step": 70
+    },
+    {
+      "epoch": 0.08020333239197967,
+      "grad_norm": 0.019900545477867126,
+      "learning_rate": 0.00019973764576610478,
+      "loss": 0.9994,
+      "step": 71
+    },
+    {
+      "epoch": 0.08133295679186671,
+      "grad_norm": 0.0189906544983387,
+      "learning_rate": 0.0001997289774009654,
+      "loss": 0.9418,
+      "step": 72
+    },
+    {
+      "epoch": 0.08246258119175374,
+      "grad_norm": 0.01873377151787281,
+      "learning_rate": 0.00019972016834372749,
+      "loss": 0.9937,
+      "step": 73
+    },
+    {
+      "epoch": 0.08359220559164078,
+      "grad_norm": 0.019596470519900322,
+      "learning_rate": 0.0001997112186068184,
+      "loss": 1.0887,
+      "step": 74
+    },
+    {
+      "epoch": 0.08472182999152782,
+      "grad_norm": 0.020303891971707344,
+      "learning_rate": 0.00019970212820286394,
+      "loss": 1.0142,
+      "step": 75
+    },
+    {
+      "epoch": 0.08585145439141485,
+      "grad_norm": 0.019804317504167557,
+      "learning_rate": 0.00019969289714468825,
+      "loss": 1.0394,
+      "step": 76
+    },
+    {
+      "epoch": 0.08698107879130189,
+      "grad_norm": 0.019536610692739487,
+      "learning_rate": 0.0001996835254453141,
+      "loss": 0.9576,
+      "step": 77
+    },
+    {
+      "epoch": 0.08811070319118892,
+      "grad_norm": 0.019902685657143593,
+      "learning_rate": 0.0001996740131179625,
+      "loss": 0.9835,
+      "step": 78
+    },
+    {
+      "epoch": 0.08924032759107597,
+      "grad_norm": 0.01986609399318695,
+      "learning_rate": 0.00019966436017605297,
+      "loss": 1.0597,
+      "step": 79
+    },
+    {
+      "epoch": 0.09036995199096301,
+      "grad_norm": 0.019435487687587738,
+      "learning_rate": 0.00019965456663320329,
+      "loss": 1.0863,
+      "step": 80
+    },
+    {
+      "epoch": 0.09149957639085005,
+      "grad_norm": 0.019000260159373283,
+      "learning_rate": 0.00019964463250322966,
+      "loss": 1.0935,
+      "step": 81
+    },
+    {
+      "epoch": 0.09262920079073708,
+      "grad_norm": 0.018888210877776146,
+      "learning_rate": 0.0001996345578001466,
+      "loss": 1.0399,
+      "step": 82
+    },
+    {
+      "epoch": 0.09375882519062412,
+      "grad_norm": 0.019765490666031837,
+      "learning_rate": 0.00019962434253816694,
+      "loss": 1.0265,
+      "step": 83
+    },
+    {
+      "epoch": 0.09488844959051115,
+      "grad_norm": 0.01926722563803196,
+      "learning_rate": 0.00019961398673170181,
+      "loss": 1.0307,
+      "step": 84
+    },
+    {
+      "epoch": 0.09601807399039819,
+      "grad_norm": 0.019572502002120018,
+      "learning_rate": 0.00019960349039536062,
+      "loss": 1.0217,
+      "step": 85
+    },
+    {
+      "epoch": 0.09714769839028523,
+      "grad_norm": 0.024138784036040306,
+      "learning_rate": 0.000199592853543951,
+      "loss": 1.1376,
+      "step": 86
+    },
+    {
+      "epoch": 0.09827732279017226,
+      "grad_norm": 0.03155818581581116,
+      "learning_rate": 0.0001995820761924788,
+      "loss": 1.1029,
+      "step": 87
+    },
+    {
+      "epoch": 0.09940694719005931,
+      "grad_norm": 0.019589390605688095,
+      "learning_rate": 0.00019957115835614816,
+      "loss": 1.0353,
+      "step": 88
+    },
+    {
+      "epoch": 0.10053657158994635,
+      "grad_norm": 0.020419439300894737,
+      "learning_rate": 0.00019956010005036133,
+      "loss": 1.0228,
+      "step": 89
+    },
+    {
+      "epoch": 0.10166619598983338,
+      "grad_norm": 0.02149847149848938,
+      "learning_rate": 0.00019954890129071876,
+      "loss": 1.1214,
+      "step": 90
+    },
+    {
+      "epoch": 0.10279582038972042,
+      "grad_norm": 0.01982825994491577,
+      "learning_rate": 0.00019953756209301903,
+      "loss": 1.0302,
+      "step": 91
+    },
+    {
+      "epoch": 0.10392544478960745,
+      "grad_norm": 0.01985330507159233,
+      "learning_rate": 0.00019952608247325885,
+      "loss": 1.0674,
+      "step": 92
+    },
+    {
+      "epoch": 0.10505506918949449,
+      "grad_norm": 0.020196454599499702,
+      "learning_rate": 0.00019951446244763309,
+      "loss": 1.0113,
+      "step": 93
+    },
+    {
+      "epoch": 0.10618469358938153,
+      "grad_norm": 0.020652327686548233,
+      "learning_rate": 0.00019950270203253454,
+      "loss": 1.0635,
+      "step": 94
+    },
+    {
+      "epoch": 0.10731431798926856,
+      "grad_norm": 0.020714478567242622,
+      "learning_rate": 0.00019949080124455416,
+      "loss": 1.0226,
+      "step": 95
+    },
+    {
+      "epoch": 0.10844394238915561,
+      "grad_norm": 0.021647842600941658,
+      "learning_rate": 0.000199478760100481,
+      "loss": 1.0575,
+      "step": 96
+    },
+    {
+      "epoch": 0.10957356678904265,
+      "grad_norm": 0.02076675370335579,
+      "learning_rate": 0.00019946657861730194,
+      "loss": 1.1146,
+      "step": 97
+    },
+    {
+      "epoch": 0.11070319118892968,
+      "grad_norm": 0.02103651687502861,
+      "learning_rate": 0.000199454256812202,
+      "loss": 0.9953,
+      "step": 98
+    },
+    {
+      "epoch": 0.11183281558881672,
+      "grad_norm": 0.02276523970067501,
+      "learning_rate": 0.00019944179470256405,
+      "loss": 1.021,
+      "step": 99
+    },
+    {
+      "epoch": 0.11296243998870376,
+      "grad_norm": 0.020957166329026222,
+      "learning_rate": 0.00019942919230596896,
+      "loss": 0.9838,
+      "step": 100
+    },
+    {
+      "epoch": 0.11409206438859079,
+      "grad_norm": 0.022394055500626564,
+      "learning_rate": 0.00019941644964019552,
+      "loss": 1.0169,
+      "step": 101
+    },
+    {
+      "epoch": 0.11522168878847783,
+      "grad_norm": 0.02139163948595524,
+      "learning_rate": 0.00019940356672322037,
+      "loss": 1.0788,
+      "step": 102
+    },
+    {
+      "epoch": 0.11635131318836486,
+      "grad_norm": 0.021381577476859093,
+      "learning_rate": 0.00019939054357321799,
+      "loss": 1.0669,
+      "step": 103
+    },
+    {
+      "epoch": 0.11748093758825191,
+      "grad_norm": 0.02302641049027443,
+      "learning_rate": 0.00019937738020856072,
+      "loss": 1.0122,
+      "step": 104
+    },
+    {
+      "epoch": 0.11861056198813895,
+      "grad_norm": 0.021372724324464798,
+      "learning_rate": 0.00019936407664781868,
+      "loss": 1.0974,
+      "step": 105
+    },
+    {
+      "epoch": 0.11974018638802598,
+      "grad_norm": 0.021260784938931465,
+      "learning_rate": 0.00019935063290975986,
+      "loss": 0.9996,
+      "step": 106
+    },
+    {
+      "epoch": 0.12086981078791302,
+      "grad_norm": 0.021557705476880074,
+      "learning_rate": 0.0001993370490133499,
+      "loss": 1.0215,
+      "step": 107
+    },
+    {
+      "epoch": 0.12199943518780006,
+      "grad_norm": 0.023252975195646286,
+      "learning_rate": 0.00019932332497775215,
+      "loss": 1.0908,
+      "step": 108
+    },
+    {
+      "epoch": 0.12312905958768709,
+      "grad_norm": 0.02185026742517948,
+      "learning_rate": 0.00019930946082232783,
+      "loss": 1.0751,
+      "step": 109
+    },
+    {
+      "epoch": 0.12425868398757413,
+      "grad_norm": 0.022223595529794693,
+      "learning_rate": 0.00019929545656663562,
+      "loss": 0.9737,
+      "step": 110
+    },
+    {
+      "epoch": 0.12538830838746118,
+      "grad_norm": 0.021415019407868385,
+      "learning_rate": 0.000199281312230432,
+      "loss": 1.0864,
+      "step": 111
+    },
+    {
+      "epoch": 0.12651793278734821,
+      "grad_norm": 0.02144046686589718,
+      "learning_rate": 0.000199267027833671,
+      "loss": 0.9984,
+      "step": 112
+    },
+    {
+      "epoch": 0.12764755718723525,
+      "grad_norm": 0.0225879717618227,
+      "learning_rate": 0.00019925260339650428,
+      "loss": 1.0685,
+      "step": 113
+    },
+    {
+      "epoch": 0.12877718158712229,
+      "grad_norm": 0.022809404879808426,
+      "learning_rate": 0.000199238038939281,
+      "loss": 1.0733,
+      "step": 114
+    },
+    {
+      "epoch": 0.12990680598700932,
+      "grad_norm": 0.023381488397717476,
+      "learning_rate": 0.00019922333448254786,
+      "loss": 1.0107,
+      "step": 115
+    },
+    {
+      "epoch": 0.13103643038689636,
+      "grad_norm": 0.022633766755461693,
+      "learning_rate": 0.00019920849004704914,
+      "loss": 0.9885,
+      "step": 116
+    },
+    {
+      "epoch": 0.1321660547867834,
+      "grad_norm": 0.02235741913318634,
+      "learning_rate": 0.00019919350565372656,
+      "loss": 1.0714,
+      "step": 117
+    },
+    {
+      "epoch": 0.13329567918667043,
+      "grad_norm": 0.02206304483115673,
+      "learning_rate": 0.00019917838132371923,
+      "loss": 1.0749,
+      "step": 118
+    },
+    {
+      "epoch": 0.13442530358655747,
+      "grad_norm": 0.022310512140393257,
+      "learning_rate": 0.0001991631170783637,
+      "loss": 1.0437,
+      "step": 119
+    },
+    {
+      "epoch": 0.1355549279864445,
+      "grad_norm": 0.021498341113328934,
+      "learning_rate": 0.00019914771293919395,
+      "loss": 1.0317,
+      "step": 120
+    },
+    {
+      "epoch": 0.13668455238633154,
+      "grad_norm": 0.021773051470518112,
+      "learning_rate": 0.0001991321689279413,
+      "loss": 1.0489,
+      "step": 121
+    },
+    {
+      "epoch": 0.13781417678621857,
+      "grad_norm": 0.021639568731188774,
+      "learning_rate": 0.0001991164850665343,
+      "loss": 0.9893,
+      "step": 122
+    },
+    {
+      "epoch": 0.1389438011861056,
+      "grad_norm": 0.022304847836494446,
+      "learning_rate": 0.00019910066137709896,
+      "loss": 1.0542,
+      "step": 123
+    },
+    {
+      "epoch": 0.14007342558599264,
+      "grad_norm": 0.022173380479216576,
+      "learning_rate": 0.0001990846978819584,
+      "loss": 1.0776,
+      "step": 124
+    },
+    {
+      "epoch": 0.1412030499858797,
+      "grad_norm": 0.023623231798410416,
+      "learning_rate": 0.00019906859460363307,
+      "loss": 1.1279,
+      "step": 125
+    },
+    {
+      "epoch": 0.14233267438576674,
+      "grad_norm": 0.022697214037179947,
+      "learning_rate": 0.0001990523515648406,
+      "loss": 0.9973,
+      "step": 126
+    },
+    {
+      "epoch": 0.14346229878565378,
+      "grad_norm": 0.02267601527273655,
+      "learning_rate": 0.00019903596878849568,
+      "loss": 1.1131,
+      "step": 127
+    },
+    {
+      "epoch": 0.14459192318554082,
+      "grad_norm": 0.02244328148663044,
+      "learning_rate": 0.0001990194462977103,
+      "loss": 1.0157,
+      "step": 128
+    },
+    {
+      "epoch": 0.14572154758542785,
+      "grad_norm": 0.02371121197938919,
+      "learning_rate": 0.00019900278411579344,
+      "loss": 0.9888,
+      "step": 129
+    },
+    {
+      "epoch": 0.1468511719853149,
+      "grad_norm": 0.0227819811552763,
+      "learning_rate": 0.00019898598226625119,
+      "loss": 1.0003,
+      "step": 130
+    },
+    {
+      "epoch": 0.14798079638520192,
+      "grad_norm": 0.02316221408545971,
+      "learning_rate": 0.00019896904077278663,
+      "loss": 1.0181,
+      "step": 131
+    },
+    {
+      "epoch": 0.14911042078508896,
+      "grad_norm": 0.022808320820331573,
+      "learning_rate": 0.00019895195965929994,
+      "loss": 1.0546,
+      "step": 132
+    },
+    {
+      "epoch": 0.150240045184976,
+      "grad_norm": 0.022865859791636467,
+      "learning_rate": 0.00019893473894988815,
+      "loss": 1.1513,
+      "step": 133
+    },
+    {
+      "epoch": 0.15136966958486303,
+      "grad_norm": 0.024973077699542046,
+      "learning_rate": 0.0001989173786688453,
+      "loss": 1.0109,
+      "step": 134
+    },
+    {
+      "epoch": 0.15249929398475007,
+      "grad_norm": 0.02237241342663765,
+      "learning_rate": 0.00019889987884066237,
+      "loss": 1.0991,
+      "step": 135
+    },
+    {
+      "epoch": 0.1536289183846371,
+      "grad_norm": 0.023280750960111618,
+      "learning_rate": 0.000198882239490027,
+      "loss": 1.0501,
+      "step": 136
+    },
+    {
+      "epoch": 0.15475854278452414,
+      "grad_norm": 0.022803954780101776,
+      "learning_rate": 0.00019886446064182396,
+      "loss": 1.0033,
+      "step": 137
+    },
+    {
+      "epoch": 0.15588816718441117,
+      "grad_norm": 0.02270156517624855,
+      "learning_rate": 0.0001988465423211346,
+      "loss": 1.0715,
+      "step": 138
+    },
+    {
+      "epoch": 0.1570177915842982,
+      "grad_norm": 0.023484455421566963,
+      "learning_rate": 0.00019882848455323704,
+      "loss": 1.1598,
+      "step": 139
+    },
+    {
+      "epoch": 0.15814741598418525,
+      "grad_norm": 0.02300065942108631,
+      "learning_rate": 0.00019881028736360622,
+      "loss": 1.0813,
+      "step": 140
+    },
+    {
+      "epoch": 0.15927704038407228,
+      "grad_norm": 0.02320142462849617,
+      "learning_rate": 0.00019879195077791376,
+      "loss": 1.0169,
+      "step": 141
+    },
+    {
+      "epoch": 0.16040666478395935,
+      "grad_norm": 0.02317328006029129,
+      "learning_rate": 0.00019877347482202785,
+      "loss": 1.0301,
+      "step": 142
+    },
+    {
+      "epoch": 0.16153628918384638,
+      "grad_norm": 0.02378895878791809,
+      "learning_rate": 0.0001987548595220133,
+      "loss": 0.977,
+      "step": 143
+    },
+    {
+      "epoch": 0.16266591358373342,
+      "grad_norm": 0.023976027965545654,
+      "learning_rate": 0.00019873610490413166,
+      "loss": 1.0859,
+      "step": 144
+    },
+    {
+      "epoch": 0.16379553798362045,
+      "grad_norm": 0.022843923419713974,
+      "learning_rate": 0.0001987172109948408,
+      "loss": 1.0103,
+      "step": 145
+    },
+    {
+      "epoch": 0.1649251623835075,
+      "grad_norm": 0.023723525926470757,
+      "learning_rate": 0.00019869817782079525,
+      "loss": 1.0704,
+      "step": 146
+    },
+    {
+      "epoch": 0.16605478678339453,
+      "grad_norm": 0.02391059510409832,
+      "learning_rate": 0.00019867900540884592,
+      "loss": 1.058,
+      "step": 147
+    },
+    {
+      "epoch": 0.16718441118328156,
+      "grad_norm": 0.023995989933609962,
+      "learning_rate": 0.0001986596937860402,
+      "loss": 1.0034,
+      "step": 148
+    },
+    {
+      "epoch": 0.1683140355831686,
+      "grad_norm": 0.024091636762022972,
+      "learning_rate": 0.00019864024297962186,
+      "loss": 1.1214,
+      "step": 149
+    },
+    {
+      "epoch": 0.16944365998305563,
+      "grad_norm": 0.024035949259996414,
+      "learning_rate": 0.000198620653017031,
+      "loss": 1.0219,
+      "step": 150
+    },
+    {
+      "epoch": 0.17057328438294267,
+      "grad_norm": 0.02359904535114765,
+      "learning_rate": 0.00019860092392590408,
+      "loss": 0.9627,
+      "step": 151
+    },
+    {
+      "epoch": 0.1717029087828297,
+      "grad_norm": 0.023622050881385803,
+      "learning_rate": 0.00019858105573407377,
+      "loss": 1.0582,
+      "step": 152
+    },
+    {
+      "epoch": 0.17283253318271674,
+      "grad_norm": 0.02392633818089962,
+      "learning_rate": 0.00019856104846956906,
+      "loss": 1.0089,
+      "step": 153
+    },
+    {
+      "epoch": 0.17396215758260378,
+      "grad_norm": 0.024305053055286407,
+      "learning_rate": 0.00019854090216061502,
+      "loss": 1.0222,
+      "step": 154
+    },
+    {
+      "epoch": 0.1750917819824908,
+      "grad_norm": 0.024212822318077087,
+      "learning_rate": 0.00019852061683563296,
+      "loss": 1.0429,
+      "step": 155
+    },
+    {
+      "epoch": 0.17622140638237785,
+      "grad_norm": 0.024261048063635826,
+      "learning_rate": 0.00019850019252324032,
+      "loss": 1.0506,
+      "step": 156
+    },
+    {
+      "epoch": 0.17735103078226488,
+      "grad_norm": 0.022689295932650566,
+      "learning_rate": 0.0001984796292522506,
+      "loss": 0.9999,
+      "step": 157
+    },
+    {
+      "epoch": 0.17848065518215195,
+      "grad_norm": 0.023288823664188385,
+      "learning_rate": 0.00019845892705167324,
+      "loss": 1.0242,
+      "step": 158
+    },
+    {
+      "epoch": 0.17961027958203898,
+      "grad_norm": 0.030173135921359062,
+      "learning_rate": 0.00019843808595071383,
+      "loss": 1.0641,
+      "step": 159
+    },
+    {
+      "epoch": 0.18073990398192602,
+      "grad_norm": 0.024562738835811615,
+      "learning_rate": 0.00019841710597877382,
+      "loss": 0.9781,
+      "step": 160
+    },
+    {
+      "epoch": 0.18186952838181306,
+      "grad_norm": 0.024897055700421333,
+      "learning_rate": 0.00019839598716545057,
+      "loss": 1.1015,
+      "step": 161
+    },
+    {
+      "epoch": 0.1829991527817001,
+      "grad_norm": 0.023950692266225815,
+      "learning_rate": 0.00019837472954053732,
+      "loss": 1.125,
+      "step": 162
+    },
+    {
+      "epoch": 0.18412877718158713,
+      "grad_norm": 0.025099674239754677,
+      "learning_rate": 0.00019835333313402318,
+      "loss": 1.0359,
+      "step": 163
+    },
+    {
+      "epoch": 0.18525840158147416,
+      "grad_norm": 0.025351393967866898,
+      "learning_rate": 0.000198331797976093,
+      "loss": 1.0281,
+      "step": 164
+    },
+    {
+      "epoch": 0.1863880259813612,
+      "grad_norm": 0.024693114683032036,
+      "learning_rate": 0.00019831012409712737,
+      "loss": 1.1521,
+      "step": 165
+    },
+    {
+      "epoch": 0.18751765038124824,
+      "grad_norm": 0.024255136027932167,
+      "learning_rate": 0.0001982883115277026,
+      "loss": 1.0842,
+      "step": 166
+    },
+    {
+      "epoch": 0.18864727478113527,
+      "grad_norm": 0.02501499280333519,
+      "learning_rate": 0.00019826636029859066,
+      "loss": 0.9975,
+      "step": 167
+    },
+    {
+      "epoch": 0.1897768991810223,
+      "grad_norm": 0.025276506319642067,
+      "learning_rate": 0.00019824427044075912,
+      "loss": 1.0119,
+      "step": 168
+    },
+    {
+      "epoch": 0.19090652358090934,
+      "grad_norm": 0.024857770651578903,
+      "learning_rate": 0.0001982220419853711,
+      "loss": 0.9733,
+      "step": 169
+    },
+    {
+      "epoch": 0.19203614798079638,
+      "grad_norm": 0.02459135465323925,
+      "learning_rate": 0.0001981996749637853,
+      "loss": 1.0791,
+      "step": 170
+    },
+    {
+      "epoch": 0.19316577238068341,
+      "grad_norm": 0.026056725531816483,
+      "learning_rate": 0.00019817716940755586,
+      "loss": 1.0849,
+      "step": 171
+    },
+    {
+      "epoch": 0.19429539678057045,
+      "grad_norm": 0.024211106821894646,
+      "learning_rate": 0.0001981545253484324,
+      "loss": 1.0253,
+      "step": 172
+    },
+    {
+      "epoch": 0.1954250211804575,
+      "grad_norm": 0.024450423195958138,
+      "learning_rate": 0.00019813174281835982,
+      "loss": 1.1101,
+      "step": 173
+    },
+    {
+      "epoch": 0.19655464558034452,
+      "grad_norm": 0.02433086559176445,
+      "learning_rate": 0.0001981088218494785,
+      "loss": 0.9887,
+      "step": 174
+    },
+    {
+      "epoch": 0.1976842699802316,
+      "grad_norm": 0.02424442023038864,
+      "learning_rate": 0.0001980857624741241,
+      "loss": 1.074,
+      "step": 175
+    },
+    {
+      "epoch": 0.19881389438011862,
+      "grad_norm": 0.02318243682384491,
+      "learning_rate": 0.00019806256472482744,
+      "loss": 1.1045,
+      "step": 176
+    },
+    {
+      "epoch": 0.19994351878000566,
+      "grad_norm": 0.02407553791999817,
+      "learning_rate": 0.00019803922863431467,
+      "loss": 1.0062,
+      "step": 177
+    },
+    {
+      "epoch": 0.2010731431798927,
+      "grad_norm": 0.02463892102241516,
+      "learning_rate": 0.000198015754235507,
+      "loss": 1.0689,
+      "step": 178
+    },
+    {
+      "epoch": 0.20220276757977973,
+      "grad_norm": 0.023701028898358345,
+      "learning_rate": 0.00019799214156152083,
+      "loss": 1.0672,
+      "step": 179
+    },
+    {
+      "epoch": 0.20333239197966677,
+      "grad_norm": 0.02471453696489334,
+      "learning_rate": 0.00019796839064566761,
+      "loss": 1.033,
+      "step": 180
+    },
+    {
+      "epoch": 0.2044620163795538,
+      "grad_norm": 0.02426736056804657,
+      "learning_rate": 0.00019794450152145382,
+      "loss": 1.0831,
+      "step": 181
+    },
+    {
+      "epoch": 0.20559164077944084,
+      "grad_norm": 0.0243529062718153,
+      "learning_rate": 0.0001979204742225809,
+      "loss": 1.0815,
+      "step": 182
+    },
+    {
+      "epoch": 0.20672126517932787,
+      "grad_norm": 0.0243973471224308,
+      "learning_rate": 0.00019789630878294526,
+      "loss": 1.0541,
+      "step": 183
+    },
+    {
+      "epoch": 0.2078508895792149,
+      "grad_norm": 0.02461186796426773,
+      "learning_rate": 0.0001978720052366381,
+      "loss": 1.1203,
+      "step": 184
+    },
+    {
+      "epoch": 0.20898051397910195,
+      "grad_norm": 0.02479882724583149,
+      "learning_rate": 0.00019784756361794555,
+      "loss": 1.078,
+      "step": 185
+    },
+    {
+      "epoch": 0.21011013837898898,
+      "grad_norm": 0.02605288103222847,
+      "learning_rate": 0.00019782298396134844,
+      "loss": 1.01,
+      "step": 186
+    },
+    {
+      "epoch": 0.21123976277887602,
+      "grad_norm": 0.025911834090948105,
+      "learning_rate": 0.00019779826630152245,
+      "loss": 1.1173,
+      "step": 187
+    },
+    {
+      "epoch": 0.21236938717876305,
+      "grad_norm": 0.024420902132987976,
+      "learning_rate": 0.00019777341067333786,
+      "loss": 1.0023,
+      "step": 188
+    },
+    {
+      "epoch": 0.2134990115786501,
+      "grad_norm": 0.024010393768548965,
+      "learning_rate": 0.0001977484171118596,
+      "loss": 1.1382,
+      "step": 189
+    },
+    {
+      "epoch": 0.21462863597853712,
+      "grad_norm": 0.024915101006627083,
+      "learning_rate": 0.00019772328565234717,
+      "loss": 1.0734,
+      "step": 190
+    },
+    {
+      "epoch": 0.21575826037842416,
+      "grad_norm": 0.025032367557287216,
+      "learning_rate": 0.0001976980163302547,
+      "loss": 0.9585,
+      "step": 191
+    },
+    {
+      "epoch": 0.21688788477831122,
+      "grad_norm": 0.024727528914809227,
+      "learning_rate": 0.0001976726091812307,
+      "loss": 1.0731,
+      "step": 192
+    },
+    {
+      "epoch": 0.21801750917819826,
+      "grad_norm": 0.024914614856243134,
+      "learning_rate": 0.00019764706424111816,
+      "loss": 0.9522,
+      "step": 193
+    },
+    {
+      "epoch": 0.2191471335780853,
+      "grad_norm": 0.024750174954533577,
+      "learning_rate": 0.00019762138154595446,
+      "loss": 0.9646,
+      "step": 194
+    },
+    {
+      "epoch": 0.22027675797797233,
+      "grad_norm": 0.02512511797249317,
+      "learning_rate": 0.00019759556113197135,
+      "loss": 1.0643,
+      "step": 195
+    },
+    {
+      "epoch": 0.22140638237785937,
+      "grad_norm": 0.026546582579612732,
+      "learning_rate": 0.00019756960303559483,
+      "loss": 1.1158,
+      "step": 196
+    },
+    {
+      "epoch": 0.2225360067777464,
+      "grad_norm": 0.02506748028099537,
+      "learning_rate": 0.0001975435072934451,
+      "loss": 1.0261,
+      "step": 197
+    },
+    {
+      "epoch": 0.22366563117763344,
+      "grad_norm": 0.024585796520113945,
+      "learning_rate": 0.00019751727394233667,
+      "loss": 1.017,
+      "step": 198
+    },
+    {
+      "epoch": 0.22479525557752048,
+      "grad_norm": 0.02528531290590763,
+      "learning_rate": 0.00019749090301927796,
+      "loss": 1.042,
+      "step": 199
+    },
+    {
+      "epoch": 0.2259248799774075,
+      "grad_norm": 0.025023646652698517,
+      "learning_rate": 0.00019746439456147172,
+      "loss": 0.9618,
+      "step": 200
+    },
+    {
+      "epoch": 0.22705450437729455,
+      "grad_norm": 0.025859549641609192,
+      "learning_rate": 0.00019743774860631457,
+      "loss": 0.9982,
+      "step": 201
+    },
+    {
+      "epoch": 0.22818412877718158,
+      "grad_norm": 0.026021264493465424,
+      "learning_rate": 0.00019741096519139713,
+      "loss": 1.0131,
+      "step": 202
+    },
+    {
+      "epoch": 0.22931375317706862,
+      "grad_norm": 0.025675011798739433,
+      "learning_rate": 0.00019738404435450395,
+      "loss": 1.0186,
+      "step": 203
+    },
+    {
+      "epoch": 0.23044337757695565,
+      "grad_norm": 0.025758078321814537,
+      "learning_rate": 0.00019735698613361347,
+      "loss": 1.0869,
+      "step": 204
+    },
+    {
+      "epoch": 0.2315730019768427,
+      "grad_norm": 0.02666814811527729,
+      "learning_rate": 0.00019732979056689794,
+      "loss": 1.0894,
+      "step": 205
+    },
+    {
+      "epoch": 0.23270262637672973,
+      "grad_norm": 0.024690723046660423,
+      "learning_rate": 0.0001973024576927233,
+      "loss": 1.0898,
+      "step": 206
+    },
+    {
+      "epoch": 0.23383225077661676,
+      "grad_norm": 0.025678694248199463,
+      "learning_rate": 0.00019727498754964928,
+      "loss": 1.091,
+      "step": 207
+    },
+    {
+      "epoch": 0.23496187517650383,
+      "grad_norm": 0.025275958701968193,
+      "learning_rate": 0.00019724738017642924,
+      "loss": 1.089,
+      "step": 208
+    },
+    {
+      "epoch": 0.23609149957639086,
+      "grad_norm": 0.02560093067586422,
+      "learning_rate": 0.00019721963561201012,
+      "loss": 0.9755,
+      "step": 209
+    },
+    {
+      "epoch": 0.2372211239762779,
+      "grad_norm": 0.026244761422276497,
+      "learning_rate": 0.00019719175389553242,
+      "loss": 1.0696,
+      "step": 210
+    },
+    {
+      "epoch": 0.23835074837616493,
+      "grad_norm": 0.025443457067012787,
+      "learning_rate": 0.0001971637350663301,
+      "loss": 1.0032,
+      "step": 211
+    },
+    {
+      "epoch": 0.23948037277605197,
+      "grad_norm": 0.027356769889593124,
+      "learning_rate": 0.00019713557916393058,
+      "loss": 1.0393,
+      "step": 212
+    },
+    {
+      "epoch": 0.240609997175939,
+      "grad_norm": 0.025765880942344666,
+      "learning_rate": 0.0001971072862280546,
+      "loss": 1.015,
+      "step": 213
+    },
+    {
+      "epoch": 0.24173962157582604,
+      "grad_norm": 0.025718411430716515,
+      "learning_rate": 0.00019707885629861632,
+      "loss": 1.0343,
+      "step": 214
+    },
+    {
+      "epoch": 0.24286924597571308,
+      "grad_norm": 0.026691369712352753,
+      "learning_rate": 0.00019705028941572307,
+      "loss": 1.0896,
+      "step": 215
+    },
+    {
+      "epoch": 0.2439988703756001,
+      "grad_norm": 0.025440771132707596,
+      "learning_rate": 0.00019702158561967544,
+      "loss": 0.9986,
+      "step": 216
+    },
+    {
+      "epoch": 0.24512849477548715,
+      "grad_norm": 0.02483600750565529,
+      "learning_rate": 0.00019699274495096712,
+      "loss": 1.0287,
+      "step": 217
+    },
+    {
+      "epoch": 0.24625811917537418,
+      "grad_norm": 0.027423838153481483,
+      "learning_rate": 0.00019696376745028497,
+      "loss": 1.0626,
+      "step": 218
+    },
+    {
+      "epoch": 0.24738774357526122,
+      "grad_norm": 0.026005201041698456,
+      "learning_rate": 0.0001969346531585088,
+      "loss": 1.0203,
+      "step": 219
+    },
+    {
+      "epoch": 0.24851736797514826,
+      "grad_norm": 0.026350049301981926,
+      "learning_rate": 0.00019690540211671144,
+      "loss": 1.0482,
+      "step": 220
+    },
+    {
+      "epoch": 0.2496469923750353,
+      "grad_norm": 0.026930196210741997,
+      "learning_rate": 0.00019687601436615864,
+      "loss": 1.0258,
+      "step": 221
+    },
+    {
+      "epoch": 0.25077661677492236,
+      "grad_norm": 0.025890439748764038,
+      "learning_rate": 0.00019684648994830903,
+      "loss": 1.0886,
+      "step": 222
+    },
+    {
+      "epoch": 0.25190624117480936,
+      "grad_norm": 0.025864360854029655,
+      "learning_rate": 0.00019681682890481398,
+      "loss": 0.976,
+      "step": 223
+    },
+    {
+      "epoch": 0.25303586557469643,
+      "grad_norm": 0.025524241849780083,
+      "learning_rate": 0.00019678703127751763,
+      "loss": 1.0251,
+      "step": 224
+    },
+    {
+      "epoch": 0.25416548997458344,
+      "grad_norm": 0.02650127001106739,
+      "learning_rate": 0.00019675709710845687,
+      "loss": 1.0435,
+      "step": 225
+    },
+    {
+      "epoch": 0.2552951143744705,
+      "grad_norm": 0.025557860732078552,
+      "learning_rate": 0.00019672702643986113,
+      "loss": 1.0555,
+      "step": 226
+    },
+    {
+      "epoch": 0.2564247387743575,
+      "grad_norm": 0.027075499296188354,
+      "learning_rate": 0.0001966968193141524,
+      "loss": 0.9965,
+      "step": 227
+    },
+    {
+      "epoch": 0.25755436317424457,
+      "grad_norm": 0.025682270526885986,
+      "learning_rate": 0.00019666647577394527,
+      "loss": 1.0151,
+      "step": 228
+    },
+    {
+      "epoch": 0.2586839875741316,
+      "grad_norm": 0.026663288474082947,
+      "learning_rate": 0.00019663599586204673,
+      "loss": 1.0354,
+      "step": 229
+    },
+    {
+      "epoch": 0.25981361197401864,
+      "grad_norm": 0.026434747502207756,
+      "learning_rate": 0.0001966053796214561,
+      "loss": 1.0551,
+      "step": 230
+    },
+    {
+      "epoch": 0.26094323637390565,
+      "grad_norm": 0.025536926463246346,
+      "learning_rate": 0.0001965746270953651,
+      "loss": 0.9731,
+      "step": 231
+    },
+    {
+      "epoch": 0.2620728607737927,
+      "grad_norm": 0.07522192597389221,
+      "learning_rate": 0.0001965437383271577,
+      "loss": 0.9796,
+      "step": 232
+    },
+    {
+      "epoch": 0.2632024851736798,
+      "grad_norm": 0.027285447344183922,
+      "learning_rate": 0.00019651271336040997,
+      "loss": 1.011,
+      "step": 233
+    },
+    {
+      "epoch": 0.2643321095735668,
+      "grad_norm": 0.026399778202176094,
+      "learning_rate": 0.0001964815522388903,
+      "loss": 1.0199,
+      "step": 234
+    },
+    {
+      "epoch": 0.26546173397345385,
+      "grad_norm": 0.026532689109444618,
+      "learning_rate": 0.00019645025500655906,
+      "loss": 0.9918,
+      "step": 235
+    },
+    {
+      "epoch": 0.26659135837334086,
+      "grad_norm": 0.025576921179890633,
+      "learning_rate": 0.00019641882170756862,
+      "loss": 1.0198,
+      "step": 236
+    },
+    {
+      "epoch": 0.2677209827732279,
+      "grad_norm": 0.026158379390835762,
+      "learning_rate": 0.00019638725238626335,
+      "loss": 1.0204,
+      "step": 237
+    },
+    {
+      "epoch": 0.26885060717311493,
+      "grad_norm": 0.025530420243740082,
+      "learning_rate": 0.00019635554708717946,
+      "loss": 1.0885,
+      "step": 238
+    },
+    {
+      "epoch": 0.269980231573002,
+      "grad_norm": 0.02707337960600853,
+      "learning_rate": 0.00019632370585504502,
+      "loss": 1.0649,
+      "step": 239
+    },
+    {
+      "epoch": 0.271109855972889,
+      "grad_norm": 0.027028286829590797,
+      "learning_rate": 0.00019629172873477995,
+      "loss": 1.0544,
+      "step": 240
+    },
+    {
+      "epoch": 0.27223948037277607,
+      "grad_norm": 0.02564058266580105,
+      "learning_rate": 0.0001962596157714957,
+      "loss": 1.0481,
+      "step": 241
+    },
+    {
+      "epoch": 0.2733691047726631,
+      "grad_norm": 0.026479296386241913,
+      "learning_rate": 0.0001962273670104955,
+      "loss": 1.0413,
+      "step": 242
+    },
+    {
+      "epoch": 0.27449872917255014,
+      "grad_norm": 0.0330955870449543,
+      "learning_rate": 0.00019619498249727412,
+      "loss": 1.0292,
+      "step": 243
+    },
+    {
+      "epoch": 0.27562835357243715,
+      "grad_norm": 0.02611500211060047,
+      "learning_rate": 0.0001961624622775178,
+      "loss": 1.009,
+      "step": 244
+    },
+    {
+      "epoch": 0.2767579779723242,
+      "grad_norm": 0.026876097545027733,
+      "learning_rate": 0.00019612980639710428,
+      "loss": 0.9854,
+      "step": 245
+    },
+    {
+      "epoch": 0.2778876023722112,
+      "grad_norm": 0.02685077115893364,
+      "learning_rate": 0.00019609701490210264,
+      "loss": 1.0282,
+      "step": 246
+    },
+    {
+      "epoch": 0.2790172267720983,
+      "grad_norm": 0.026131028309464455,
+      "learning_rate": 0.00019606408783877334,
+      "loss": 1.0673,
+      "step": 247
+    },
+    {
+      "epoch": 0.2801468511719853,
+      "grad_norm": 0.02628222666680813,
+      "learning_rate": 0.00019603102525356798,
+      "loss": 1.0659,
+      "step": 248
+    },
+    {
+      "epoch": 0.28127647557187235,
+      "grad_norm": 0.027401477098464966,
+      "learning_rate": 0.00019599782719312948,
+      "loss": 0.9942,
+      "step": 249
+    },
+    {
+      "epoch": 0.2824060999717594,
+      "grad_norm": 0.02594529278576374,
+      "learning_rate": 0.00019596449370429183,
+      "loss": 1.0091,
+      "step": 250
+    },
+    {
+      "epoch": 0.2835357243716464,
+      "grad_norm": 0.028301890939474106,
+      "learning_rate": 0.00019593102483408,
+      "loss": 1.0083,
+      "step": 251
+    },
+    {
+      "epoch": 0.2846653487715335,
+      "grad_norm": 0.02808901108801365,
+      "learning_rate": 0.00019589742062971007,
+      "loss": 1.071,
+      "step": 252
+    },
+    {
+      "epoch": 0.2857949731714205,
+      "grad_norm": 0.02654552273452282,
+      "learning_rate": 0.00019586368113858892,
+      "loss": 1.0865,
+      "step": 253
+    },
+    {
+      "epoch": 0.28692459757130756,
+      "grad_norm": 0.02610975131392479,
+      "learning_rate": 0.00019582980640831443,
+      "loss": 1.1093,
+      "step": 254
+    },
+    {
+      "epoch": 0.28805422197119457,
+      "grad_norm": 0.027240293100476265,
+      "learning_rate": 0.0001957957964866751,
+      "loss": 1.0822,
+      "step": 255
+    },
+    {
+      "epoch": 0.28918384637108163,
+      "grad_norm": 0.027821950614452362,
+      "learning_rate": 0.00019576165142165032,
+      "loss": 1.0371,
+      "step": 256
+    },
+    {
+      "epoch": 0.29031347077096864,
+      "grad_norm": 0.02755453623831272,
+      "learning_rate": 0.00019572737126141002,
+      "loss": 1.0752,
+      "step": 257
+    },
+    {
+      "epoch": 0.2914430951708557,
+      "grad_norm": 0.02676587551832199,
+      "learning_rate": 0.0001956929560543147,
+      "loss": 1.0599,
+      "step": 258
+    },
+    {
+      "epoch": 0.2925727195707427,
+      "grad_norm": 0.02904544584453106,
+      "learning_rate": 0.00019565840584891549,
+      "loss": 1.0568,
+      "step": 259
+    },
+    {
+      "epoch": 0.2937023439706298,
+      "grad_norm": 0.027289781719446182,
+      "learning_rate": 0.00019562372069395384,
+      "loss": 1.0671,
+      "step": 260
+    },
+    {
+      "epoch": 0.2948319683705168,
+      "grad_norm": 0.025955747812986374,
+      "learning_rate": 0.00019558890063836167,
+      "loss": 0.9118,
+      "step": 261
+    },
+    {
+      "epoch": 0.29596159277040385,
+      "grad_norm": 0.028641648590564728,
+      "learning_rate": 0.00019555394573126118,
+      "loss": 1.0498,
+      "step": 262
+    },
+    {
+      "epoch": 0.29709121717029086,
+      "grad_norm": 0.028356773778796196,
+      "learning_rate": 0.0001955188560219648,
+      "loss": 1.0238,
+      "step": 263
+    },
+    {
+      "epoch": 0.2982208415701779,
+      "grad_norm": 0.02746075950562954,
+      "learning_rate": 0.00019548363155997517,
+      "loss": 1.0741,
+      "step": 264
+    },
+    {
+      "epoch": 0.2993504659700649,
+      "grad_norm": 0.02712567336857319,
+      "learning_rate": 0.000195448272394985,
+      "loss": 1.0861,
+      "step": 265
+    },
+    {
+      "epoch": 0.300480090369952,
+      "grad_norm": 0.026709580793976784,
+      "learning_rate": 0.00019541277857687694,
+      "loss": 1.0024,
+      "step": 266
+    },
+    {
+      "epoch": 0.30160971476983905,
+      "grad_norm": 0.027716003358364105,
+      "learning_rate": 0.00019537715015572382,
+      "loss": 1.0406,
+      "step": 267
+    },
+    {
+      "epoch": 0.30273933916972606,
+      "grad_norm": 0.02704858034849167,
+      "learning_rate": 0.00019534138718178818,
+      "loss": 1.0088,
+      "step": 268
+    },
+    {
+      "epoch": 0.3038689635696131,
+      "grad_norm": 0.026793915778398514,
+      "learning_rate": 0.00019530548970552247,
+      "loss": 1.0556,
+      "step": 269
+    },
+    {
+      "epoch": 0.30499858796950013,
+      "grad_norm": 0.028323287144303322,
+      "learning_rate": 0.00019526945777756879,
+      "loss": 1.057,
+      "step": 270
+    },
+    {
+      "epoch": 0.3061282123693872,
+      "grad_norm": 0.0279136560857296,
+      "learning_rate": 0.00019523329144875904,
+      "loss": 1.0654,
+      "step": 271
+    },
+    {
+      "epoch": 0.3072578367692742,
+      "grad_norm": 0.02878638356924057,
+      "learning_rate": 0.00019519699077011465,
+      "loss": 1.0357,
+      "step": 272
+    },
+    {
+      "epoch": 0.30838746116916127,
+      "grad_norm": 0.026021145284175873,
+      "learning_rate": 0.00019516055579284658,
+      "loss": 1.092,
+      "step": 273
+    },
+    {
+      "epoch": 0.3095170855690483,
+      "grad_norm": 0.0282638818025589,
+      "learning_rate": 0.00019512398656835528,
+      "loss": 1.0242,
+      "step": 274
+    },
+    {
+      "epoch": 0.31064670996893534,
+      "grad_norm": 0.0277785062789917,
+      "learning_rate": 0.00019508728314823062,
+      "loss": 1.0922,
+      "step": 275
+    },
+    {
+      "epoch": 0.31177633436882235,
+      "grad_norm": 0.027666205540299416,
+      "learning_rate": 0.00019505044558425168,
+      "loss": 1.0434,
+      "step": 276
+    },
+    {
+      "epoch": 0.3129059587687094,
+      "grad_norm": 0.02734490856528282,
+      "learning_rate": 0.0001950134739283869,
+      "loss": 1.0726,
+      "step": 277
+    },
+    {
+      "epoch": 0.3140355831685964,
+      "grad_norm": 0.026907166466116905,
+      "learning_rate": 0.0001949763682327938,
+      "loss": 1.0807,
+      "step": 278
+    },
+    {
+      "epoch": 0.3151652075684835,
+      "grad_norm": 0.02773541398346424,
+      "learning_rate": 0.00019493912854981905,
+      "loss": 1.0941,
+      "step": 279
+    },
+    {
+      "epoch": 0.3162948319683705,
+      "grad_norm": 0.027467425912618637,
+      "learning_rate": 0.00019490175493199833,
+      "loss": 1.031,
+      "step": 280
+    },
+    {
+      "epoch": 0.31742445636825756,
+      "grad_norm": 0.02712651528418064,
+      "learning_rate": 0.00019486424743205626,
+      "loss": 1.0015,
+      "step": 281
+    },
+    {
+      "epoch": 0.31855408076814457,
+      "grad_norm": 0.026572776958346367,
+      "learning_rate": 0.00019482660610290636,
+      "loss": 0.9459,
+      "step": 282
+    },
+    {
+      "epoch": 0.31968370516803163,
+      "grad_norm": 0.02701294608414173,
+      "learning_rate": 0.00019478883099765086,
+      "loss": 1.0652,
+      "step": 283
+    },
+    {
+      "epoch": 0.3208133295679187,
+      "grad_norm": 0.02713761292397976,
+      "learning_rate": 0.0001947509221695808,
+      "loss": 1.0455,
+      "step": 284
+    },
+    {
+      "epoch": 0.3219429539678057,
+      "grad_norm": 0.028251413255929947,
+      "learning_rate": 0.00019471287967217594,
+      "loss": 0.9885,
+      "step": 285
+    },
+    {
+      "epoch": 0.32307257836769276,
+      "grad_norm": 0.028362903743982315,
+      "learning_rate": 0.00019467470355910438,
+      "loss": 1.0896,
+      "step": 286
+    },
+    {
+      "epoch": 0.3242022027675798,
+      "grad_norm": 0.027835773304104805,
+      "learning_rate": 0.00019463639388422297,
+      "loss": 0.9381,
+      "step": 287
+    },
+    {
+      "epoch": 0.32533182716746684,
+      "grad_norm": 0.026659086346626282,
+      "learning_rate": 0.0001945979507015768,
+      "loss": 0.9987,
+      "step": 288
+    },
+    {
+      "epoch": 0.32646145156735384,
+      "grad_norm": 0.028285473585128784,
+      "learning_rate": 0.0001945593740653994,
+      "loss": 1.0055,
+      "step": 289
+    },
+    {
+      "epoch": 0.3275910759672409,
+      "grad_norm": 0.027459239587187767,
+      "learning_rate": 0.00019452066403011253,
+      "loss": 1.0468,
+      "step": 290
+    },
+    {
+      "epoch": 0.3287207003671279,
+      "grad_norm": 0.028836321085691452,
+      "learning_rate": 0.00019448182065032621,
+      "loss": 1.0855,
+      "step": 291
+    },
+    {
+      "epoch": 0.329850324767015,
+      "grad_norm": 0.029597043991088867,
+      "learning_rate": 0.00019444284398083847,
+      "loss": 1.1135,
+      "step": 292
+    },
+    {
+      "epoch": 0.330979949166902,
+      "grad_norm": 0.029845820739865303,
+      "learning_rate": 0.00019440373407663542,
+      "loss": 1.0117,
+      "step": 293
+    },
+    {
+      "epoch": 0.33210957356678905,
+      "grad_norm": 0.027042267844080925,
+      "learning_rate": 0.00019436449099289119,
+      "loss": 1.0173,
+      "step": 294
+    },
+    {
+      "epoch": 0.33323919796667606,
+      "grad_norm": 0.027646934613585472,
+      "learning_rate": 0.00019432511478496768,
+      "loss": 1.0777,
+      "step": 295
+    },
+    {
+      "epoch": 0.33323919796667606,
+      "eval_loss": 1.0277949571609497,
+      "eval_runtime": 565.1236,
+      "eval_samples_per_second": 17.311,
+      "eval_steps_per_second": 8.657,
+      "step": 295
+    },
+    {
+      "epoch": 0.3343688223665631,
+      "grad_norm": 0.026499278843402863,
+      "learning_rate": 0.00019428560550841472,
+      "loss": 0.9618,
+      "step": 296
+    },
+    {
+      "epoch": 0.33549844676645013,
+      "grad_norm": 0.027500445023179054,
+      "learning_rate": 0.00019424596321896976,
+      "loss": 0.9794,
+      "step": 297
+    },
+    {
+      "epoch": 0.3366280711663372,
+      "grad_norm": 0.027349818497896194,
+      "learning_rate": 0.00019420618797255795,
+      "loss": 1.1008,
+      "step": 298
+    },
+    {
+      "epoch": 0.3377576955662242,
+      "grad_norm": 0.027657683938741684,
+      "learning_rate": 0.000194166279825292,
+      "loss": 1.0801,
+      "step": 299
+    },
+    {
+      "epoch": 0.33888731996611127,
+      "grad_norm": 0.027384718880057335,
+      "learning_rate": 0.00019412623883347207,
+      "loss": 1.038,
+      "step": 300
+    },
+    {
+      "epoch": 0.34001694436599833,
+      "grad_norm": 0.026920663192868233,
+      "learning_rate": 0.00019408606505358583,
+      "loss": 0.9868,
+      "step": 301
+    },
+    {
+      "epoch": 0.34114656876588534,
+      "grad_norm": 0.028844624757766724,
+      "learning_rate": 0.00019404575854230818,
+      "loss": 1.0293,
+      "step": 302
+    },
+    {
+      "epoch": 0.3422761931657724,
+      "grad_norm": 0.02755833975970745,
+      "learning_rate": 0.00019400531935650128,
+      "loss": 1.0087,
+      "step": 303
+    },
+    {
+      "epoch": 0.3434058175656594,
+      "grad_norm": 0.027301400899887085,
+      "learning_rate": 0.00019396474755321456,
+      "loss": 1.0318,
+      "step": 304
+    },
+    {
+      "epoch": 0.3445354419655465,
+      "grad_norm": 0.02760390006005764,
+      "learning_rate": 0.0001939240431896844,
+      "loss": 0.9421,
+      "step": 305
+    },
+    {
+      "epoch": 0.3456650663654335,
+      "grad_norm": 0.027442464604973793,
+      "learning_rate": 0.00019388320632333429,
+      "loss": 1.0801,
+      "step": 306
+    },
+    {
+      "epoch": 0.34679469076532055,
+      "grad_norm": 0.027593247592449188,
+      "learning_rate": 0.00019384223701177455,
+      "loss": 1.0607,
+      "step": 307
+    },
+    {
+      "epoch": 0.34792431516520755,
+      "grad_norm": 0.028117630630731583,
+      "learning_rate": 0.00019380113531280245,
+      "loss": 1.054,
+      "step": 308
+    },
+    {
+      "epoch": 0.3490539395650946,
+      "grad_norm": 0.029217706993222237,
+      "learning_rate": 0.00019375990128440204,
+      "loss": 1.0997,
+      "step": 309
+    },
+    {
+      "epoch": 0.3501835639649816,
+      "grad_norm": 0.027274932712316513,
+      "learning_rate": 0.0001937185349847439,
+      "loss": 1.0051,
+      "step": 310
+    },
+    {
+      "epoch": 0.3513131883648687,
+      "grad_norm": 0.03279178589582443,
+      "learning_rate": 0.0001936770364721854,
+      "loss": 1.0293,
+      "step": 311
+    },
+    {
+      "epoch": 0.3524428127647557,
+      "grad_norm": 0.026957320049405098,
+      "learning_rate": 0.00019363540580527025,
+      "loss": 1.0358,
+      "step": 312
+    },
+    {
+      "epoch": 0.35357243716464276,
+      "grad_norm": 0.029469158500432968,
+      "learning_rate": 0.0001935936430427287,
+      "loss": 1.1446,
+      "step": 313
+    },
+    {
+      "epoch": 0.35470206156452977,
+      "grad_norm": 0.03025597333908081,
+      "learning_rate": 0.00019355174824347735,
+      "loss": 1.0722,
+      "step": 314
+    },
+    {
+      "epoch": 0.35583168596441683,
+      "grad_norm": 0.02727232687175274,
+      "learning_rate": 0.00019350972146661905,
+      "loss": 1.0592,
+      "step": 315
+    },
+    {
+      "epoch": 0.3569613103643039,
+      "grad_norm": 0.028911981731653214,
+      "learning_rate": 0.00019346756277144285,
+      "loss": 1.1644,
+      "step": 316
+    },
+    {
+      "epoch": 0.3580909347641909,
+      "grad_norm": 0.02783570997416973,
+      "learning_rate": 0.0001934252722174239,
+      "loss": 0.9406,
+      "step": 317
+    },
+    {
+      "epoch": 0.35922055916407797,
+      "grad_norm": 0.02677338756620884,
+      "learning_rate": 0.00019338284986422335,
+      "loss": 0.9287,
+      "step": 318
+    },
+    {
+      "epoch": 0.360350183563965,
+      "grad_norm": 0.027951853349804878,
+      "learning_rate": 0.00019334029577168827,
+      "loss": 0.9541,
+      "step": 319
+    },
+    {
+      "epoch": 0.36147980796385204,
+      "grad_norm": 0.028323214501142502,
+      "learning_rate": 0.00019329760999985167,
+      "loss": 1.1566,
+      "step": 320
+    },
+    {
+      "epoch": 0.36260943236373905,
+      "grad_norm": 0.027881423011422157,
+      "learning_rate": 0.00019325479260893223,
+      "loss": 1.0662,
+      "step": 321
+    },
+    {
+      "epoch": 0.3637390567636261,
+      "grad_norm": 0.02717737667262554,
+      "learning_rate": 0.00019321184365933433,
+      "loss": 1.0317,
+      "step": 322
+    },
+    {
+      "epoch": 0.3648686811635131,
+      "grad_norm": 0.028628146275877953,
+      "learning_rate": 0.00019316876321164798,
+      "loss": 1.0503,
+      "step": 323
+    },
+    {
+      "epoch": 0.3659983055634002,
+      "grad_norm": 0.02851051092147827,
+      "learning_rate": 0.0001931255513266487,
+      "loss": 1.0565,
+      "step": 324
+    },
+    {
+      "epoch": 0.3671279299632872,
+      "grad_norm": 0.02863175794482231,
+      "learning_rate": 0.00019308220806529738,
+      "loss": 1.0243,
+      "step": 325
+    },
+    {
+      "epoch": 0.36825755436317426,
+      "grad_norm": 0.03015504591166973,
+      "learning_rate": 0.0001930387334887403,
+      "loss": 1.0208,
+      "step": 326
+    },
+    {
+      "epoch": 0.36938717876306126,
+      "grad_norm": 0.02771030366420746,
+      "learning_rate": 0.00019299512765830895,
+      "loss": 1.0094,
+      "step": 327
+    },
+    {
+      "epoch": 0.3705168031629483,
+      "grad_norm": 0.027864158153533936,
+      "learning_rate": 0.00019295139063552007,
+      "loss": 0.9863,
+      "step": 328
+    },
+    {
+      "epoch": 0.37164642756283534,
+      "grad_norm": 0.028755534440279007,
+      "learning_rate": 0.00019290752248207537,
+      "loss": 1.0542,
+      "step": 329
+    },
+    {
+      "epoch": 0.3727760519627224,
+      "grad_norm": 0.029860056936740875,
+      "learning_rate": 0.00019286352325986164,
+      "loss": 1.0006,
+      "step": 330
+    },
+    {
+      "epoch": 0.3739056763626094,
+      "grad_norm": 0.027963971719145775,
+      "learning_rate": 0.0001928193930309505,
+      "loss": 0.9609,
+      "step": 331
+    },
+    {
+      "epoch": 0.37503530076249647,
+      "grad_norm": 0.02750619500875473,
+      "learning_rate": 0.00019277513185759844,
+      "loss": 1.0076,
+      "step": 332
+    },
+    {
+      "epoch": 0.37616492516238353,
+      "grad_norm": 0.02815542183816433,
+      "learning_rate": 0.0001927307398022467,
+      "loss": 1.04,
+      "step": 333
+    },
+    {
+      "epoch": 0.37729454956227054,
+      "grad_norm": 0.028742128983139992,
+      "learning_rate": 0.00019268621692752108,
+      "loss": 0.9947,
+      "step": 334
+    },
+    {
+      "epoch": 0.3784241739621576,
+      "grad_norm": 0.027735736221075058,
+      "learning_rate": 0.00019264156329623197,
+      "loss": 1.0265,
+      "step": 335
+    },
+    {
+      "epoch": 0.3795537983620446,
+      "grad_norm": 0.02745204232633114,
+      "learning_rate": 0.00019259677897137426,
+      "loss": 1.0308,
+      "step": 336
+    },
+    {
+      "epoch": 0.3806834227619317,
+      "grad_norm": 0.028459064662456512,
+      "learning_rate": 0.00019255186401612718,
+      "loss": 1.0069,
+      "step": 337
+    },
+    {
+      "epoch": 0.3818130471618187,
+      "grad_norm": 0.028107335790991783,
+      "learning_rate": 0.00019250681849385424,
+      "loss": 1.0812,
+      "step": 338
+    },
+    {
+      "epoch": 0.38294267156170575,
+      "grad_norm": 0.029490889981389046,
+      "learning_rate": 0.00019246164246810316,
+      "loss": 1.0247,
+      "step": 339
+    },
+    {
+      "epoch": 0.38407229596159276,
+      "grad_norm": 0.027926163747906685,
+      "learning_rate": 0.00019241633600260578,
+      "loss": 0.9761,
+      "step": 340
+    },
+    {
+      "epoch": 0.3852019203614798,
+      "grad_norm": 0.02847837097942829,
+      "learning_rate": 0.00019237089916127793,
+      "loss": 1.0841,
+      "step": 341
+    },
+    {
+      "epoch": 0.38633154476136683,
+      "grad_norm": 0.027178598567843437,
+      "learning_rate": 0.00019232533200821942,
+      "loss": 1.1123,
+      "step": 342
+    },
+    {
+      "epoch": 0.3874611691612539,
+      "grad_norm": 0.027773573994636536,
+      "learning_rate": 0.00019227963460771377,
+      "loss": 0.9871,
+      "step": 343
+    },
+    {
+      "epoch": 0.3885907935611409,
+      "grad_norm": 0.027409275993704796,
+      "learning_rate": 0.00019223380702422844,
+      "loss": 1.0916,
+      "step": 344
+    },
+    {
+      "epoch": 0.38972041796102797,
+      "grad_norm": 0.028152553364634514,
+      "learning_rate": 0.00019218784932241434,
+      "loss": 1.0301,
+      "step": 345
+    },
+    {
+      "epoch": 0.390850042360915,
+      "grad_norm": 0.028817711398005486,
+      "learning_rate": 0.00019214176156710612,
+      "loss": 1.0203,
+      "step": 346
+    },
+    {
+      "epoch": 0.39197966676080204,
+      "grad_norm": 0.02772883139550686,
+      "learning_rate": 0.0001920955438233218,
+      "loss": 0.9991,
+      "step": 347
+    },
+    {
+      "epoch": 0.39310929116068904,
+      "grad_norm": 0.028133943676948547,
+      "learning_rate": 0.00019204919615626275,
+      "loss": 0.9834,
+      "step": 348
+    },
+    {
+      "epoch": 0.3942389155605761,
+      "grad_norm": 0.02936532348394394,
+      "learning_rate": 0.00019200271863131375,
+      "loss": 1.0227,
+      "step": 349
+    },
+    {
+      "epoch": 0.3953685399604632,
+      "grad_norm": 0.028890248388051987,
+      "learning_rate": 0.0001919561113140427,
+      "loss": 0.9551,
+      "step": 350
+    },
+    {
+      "epoch": 0.3964981643603502,
+      "grad_norm": 0.02820666879415512,
+      "learning_rate": 0.0001919093742702006,
+      "loss": 1.0343,
+      "step": 351
+    },
+    {
+      "epoch": 0.39762778876023724,
+      "grad_norm": 0.029474567621946335,
+      "learning_rate": 0.00019186250756572144,
+      "loss": 0.9853,
+      "step": 352
+    },
+    {
+      "epoch": 0.39875741316012425,
+      "grad_norm": 0.02914329618215561,
+      "learning_rate": 0.0001918155112667222,
+      "loss": 0.9542,
+      "step": 353
+    },
+    {
+      "epoch": 0.3998870375600113,
+      "grad_norm": 0.028036657720804214,
+      "learning_rate": 0.00019176838543950267,
+      "loss": 0.945,
+      "step": 354
+    },
+    {
+      "epoch": 0.4010166619598983,
+      "grad_norm": 0.027309326454997063,
+      "learning_rate": 0.00019172113015054532,
+      "loss": 0.977,
+      "step": 355
+    },
+    {
+      "epoch": 0.4021462863597854,
+      "grad_norm": 0.027427159249782562,
+      "learning_rate": 0.00019167374546651526,
+      "loss": 1.0505,
+      "step": 356
+    },
+    {
+      "epoch": 0.4032759107596724,
+      "grad_norm": 0.03023376129567623,
+      "learning_rate": 0.0001916262314542602,
+      "loss": 1.1378,
+      "step": 357
+    },
+    {
+      "epoch": 0.40440553515955946,
+      "grad_norm": 0.027807191014289856,
+      "learning_rate": 0.00019157858818081026,
+      "loss": 1.0516,
+      "step": 358
+    },
+    {
+      "epoch": 0.40553515955944647,
+      "grad_norm": 0.028308499604463577,
+      "learning_rate": 0.00019153081571337795,
+      "loss": 1.0673,
+      "step": 359
+    },
+    {
+      "epoch": 0.40666478395933353,
+      "grad_norm": 0.028541473671793938,
+      "learning_rate": 0.00019148291411935796,
+      "loss": 1.0567,
+      "step": 360
+    },
+    {
+      "epoch": 0.40779440835922054,
+      "grad_norm": 0.027455326169729233,
+      "learning_rate": 0.00019143488346632723,
+      "loss": 1.0078,
+      "step": 361
+    },
+    {
+      "epoch": 0.4089240327591076,
+      "grad_norm": 0.02952658385038376,
+      "learning_rate": 0.00019138672382204471,
+      "loss": 1.0686,
+      "step": 362
+    },
+    {
+      "epoch": 0.4100536571589946,
+      "grad_norm": 0.028435127809643745,
+      "learning_rate": 0.0001913384352544514,
+      "loss": 0.9846,
+      "step": 363
+    },
+    {
+      "epoch": 0.4111832815588817,
+      "grad_norm": 0.028838949277997017,
+      "learning_rate": 0.00019129001783167005,
+      "loss": 1.0602,
+      "step": 364
+    },
+    {
+      "epoch": 0.4123129059587687,
+      "grad_norm": 0.029650872573256493,
+      "learning_rate": 0.00019124147162200535,
+      "loss": 0.9967,
+      "step": 365
+    },
+    {
+      "epoch": 0.41344253035865575,
+      "grad_norm": 0.028792966157197952,
+      "learning_rate": 0.00019119279669394353,
+      "loss": 1.0562,
+      "step": 366
+    },
+    {
+      "epoch": 0.4145721547585428,
+      "grad_norm": 0.029962720349431038,
+      "learning_rate": 0.00019114399311615253,
+      "loss": 1.0016,
+      "step": 367
+    },
+    {
+      "epoch": 0.4157017791584298,
+      "grad_norm": 0.029513955116271973,
+      "learning_rate": 0.00019109506095748167,
+      "loss": 1.007,
+      "step": 368
+    },
+    {
+      "epoch": 0.4168314035583169,
+      "grad_norm": 0.028869032859802246,
+      "learning_rate": 0.00019104600028696175,
+      "loss": 1.033,
+      "step": 369
+    },
+    {
+      "epoch": 0.4179610279582039,
+      "grad_norm": 0.02818440832197666,
+      "learning_rate": 0.00019099681117380486,
+      "loss": 0.9947,
+      "step": 370
+    },
+    {
+      "epoch": 0.41909065235809095,
+      "grad_norm": 0.030735397711396217,
+      "learning_rate": 0.00019094749368740423,
+      "loss": 1.031,
+      "step": 371
+    },
+    {
+      "epoch": 0.42022027675797796,
+      "grad_norm": 0.029516831040382385,
+      "learning_rate": 0.00019089804789733424,
+      "loss": 1.1093,
+      "step": 372
+    },
+    {
+      "epoch": 0.421349901157865,
+      "grad_norm": 0.028589509427547455,
+      "learning_rate": 0.00019084847387335025,
+      "loss": 1.0524,
+      "step": 373
+    },
+    {
+      "epoch": 0.42247952555775203,
+      "grad_norm": 0.029599323868751526,
+      "learning_rate": 0.00019079877168538855,
+      "loss": 1.0867,
+      "step": 374
+    },
+    {
+      "epoch": 0.4236091499576391,
+      "grad_norm": 0.029633615165948868,
+      "learning_rate": 0.00019074894140356624,
+      "loss": 1.0187,
+      "step": 375
+    },
+    {
+      "epoch": 0.4247387743575261,
+      "grad_norm": 0.029569542035460472,
+      "learning_rate": 0.00019069898309818106,
+      "loss": 1.0172,
+      "step": 376
+    },
+    {
+      "epoch": 0.42586839875741317,
+      "grad_norm": 0.02864873595535755,
+      "learning_rate": 0.00019064889683971149,
+      "loss": 1.0408,
+      "step": 377
+    },
+    {
+      "epoch": 0.4269980231573002,
+      "grad_norm": 0.02849559485912323,
+      "learning_rate": 0.0001905986826988164,
+      "loss": 1.0513,
+      "step": 378
+    },
+    {
+      "epoch": 0.42812764755718724,
+      "grad_norm": 0.028202759101986885,
+      "learning_rate": 0.00019054834074633506,
+      "loss": 1.0536,
+      "step": 379
+    },
+    {
+      "epoch": 0.42925727195707425,
+      "grad_norm": 0.02983192540705204,
+      "learning_rate": 0.00019049787105328715,
+      "loss": 1.0294,
+      "step": 380
+    },
+    {
+      "epoch": 0.4303868963569613,
+      "grad_norm": 0.028043275699019432,
+      "learning_rate": 0.0001904472736908725,
+      "loss": 0.9645,
+      "step": 381
+    },
+    {
+      "epoch": 0.4315165207568483,
+      "grad_norm": 0.02895670384168625,
+      "learning_rate": 0.0001903965487304711,
+      "loss": 1.154,
+      "step": 382
+    },
+    {
+      "epoch": 0.4326461451567354,
+      "grad_norm": 0.02832162007689476,
+      "learning_rate": 0.0001903456962436428,
+      "loss": 1.0332,
+      "step": 383
+    },
+    {
+      "epoch": 0.43377576955662245,
+      "grad_norm": 0.029863545671105385,
+      "learning_rate": 0.00019029471630212762,
+      "loss": 1.0002,
+      "step": 384
+    },
+    {
+      "epoch": 0.43490539395650946,
+      "grad_norm": 0.02890811115503311,
+      "learning_rate": 0.00019024360897784508,
+      "loss": 1.0644,
+      "step": 385
+    },
+    {
+      "epoch": 0.4360350183563965,
+      "grad_norm": 0.03050493635237217,
+      "learning_rate": 0.0001901923743428946,
+      "loss": 1.0324,
+      "step": 386
+    },
+    {
+      "epoch": 0.43716464275628353,
+      "grad_norm": 0.029246153309941292,
+      "learning_rate": 0.00019014101246955515,
+      "loss": 1.0591,
+      "step": 387
+    },
+    {
+      "epoch": 0.4382942671561706,
+      "grad_norm": 0.02876698225736618,
+      "learning_rate": 0.00019008952343028526,
+      "loss": 0.9519,
+      "step": 388
+    },
+    {
+      "epoch": 0.4394238915560576,
+      "grad_norm": 0.029059743508696556,
+      "learning_rate": 0.00019003790729772273,
+      "loss": 1.0165,
+      "step": 389
+    },
+    {
+      "epoch": 0.44055351595594466,
+      "grad_norm": 0.02885555475950241,
+      "learning_rate": 0.00018998616414468478,
+      "loss": 1.004,
+      "step": 390
+    },
+    {
+      "epoch": 0.44168314035583167,
+      "grad_norm": 0.02809917740523815,
+      "learning_rate": 0.00018993429404416773,
+      "loss": 0.9685,
+      "step": 391
+    },
+    {
+      "epoch": 0.44281276475571874,
+      "grad_norm": 0.028004605323076248,
+      "learning_rate": 0.0001898822970693471,
+      "loss": 0.9923,
+      "step": 392
+    },
+    {
+      "epoch": 0.44394238915560574,
+      "grad_norm": 0.029958872124552727,
+      "learning_rate": 0.00018983017329357729,
+      "loss": 1.0468,
+      "step": 393
+    },
+    {
+      "epoch": 0.4450720135554928,
+      "grad_norm": 0.03032870590686798,
+      "learning_rate": 0.00018977792279039162,
+      "loss": 0.9573,
+      "step": 394
+    },
+    {
+      "epoch": 0.4462016379553798,
+      "grad_norm": 0.029365211725234985,
+      "learning_rate": 0.0001897255456335022,
+      "loss": 0.9673,
+      "step": 395
+    },
+    {
+      "epoch": 0.4473312623552669,
+      "grad_norm": 0.03092394582927227,
+      "learning_rate": 0.00018967304189679984,
+      "loss": 1.1468,
+      "step": 396
+    },
+    {
+      "epoch": 0.4484608867551539,
+      "grad_norm": 0.029345886781811714,
+      "learning_rate": 0.00018962041165435388,
+      "loss": 1.1213,
+      "step": 397
+    },
+    {
+      "epoch": 0.44959051115504095,
+      "grad_norm": 0.029504388570785522,
+      "learning_rate": 0.0001895676549804121,
+      "loss": 1.0483,
+      "step": 398
+    },
+    {
+      "epoch": 0.450720135554928,
+      "grad_norm": 0.029384993016719818,
+      "learning_rate": 0.00018951477194940075,
+      "loss": 0.9973,
+      "step": 399
+    },
+    {
+      "epoch": 0.451849759954815,
+      "grad_norm": 0.02798447571694851,
+      "learning_rate": 0.0001894617626359242,
+      "loss": 1.0041,
+      "step": 400
+    },
+    {
+      "epoch": 0.4529793843547021,
+      "grad_norm": 0.028576720505952835,
+      "learning_rate": 0.00018940862711476513,
+      "loss": 1.0699,
+      "step": 401
+    },
+    {
+      "epoch": 0.4541090087545891,
+      "grad_norm": 0.029531830921769142,
+      "learning_rate": 0.0001893553654608841,
+      "loss": 1.0396,
+      "step": 402
+    },
+    {
+      "epoch": 0.45523863315447616,
+      "grad_norm": 0.02875913865864277,
+      "learning_rate": 0.00018930197774941974,
+      "loss": 1.0302,
+      "step": 403
+    },
+    {
+      "epoch": 0.45636825755436317,
+      "grad_norm": 0.02790944278240204,
+      "learning_rate": 0.00018924846405568845,
+      "loss": 1.1243,
+      "step": 404
+    },
+    {
+      "epoch": 0.45749788195425023,
+      "grad_norm": 0.02811037190258503,
+      "learning_rate": 0.00018919482445518436,
+      "loss": 1.0377,
+      "step": 405
+    },
+    {
+      "epoch": 0.45862750635413724,
+      "grad_norm": 0.029786163941025734,
+      "learning_rate": 0.00018914105902357925,
+      "loss": 0.9825,
+      "step": 406
+    },
+    {
+      "epoch": 0.4597571307540243,
+      "grad_norm": 0.028242526575922966,
+      "learning_rate": 0.0001890871678367224,
+      "loss": 1.0738,
+      "step": 407
+    },
+    {
+      "epoch": 0.4608867551539113,
+      "grad_norm": 0.028527051210403442,
+      "learning_rate": 0.00018903315097064055,
+      "loss": 1.0024,
+      "step": 408
+    },
+    {
+      "epoch": 0.4620163795537984,
+      "grad_norm": 0.02773975394666195,
+      "learning_rate": 0.0001889790085015376,
+      "loss": 1.0042,
+      "step": 409
+    },
+    {
+      "epoch": 0.4631460039536854,
+      "grad_norm": 0.028500793501734734,
+      "learning_rate": 0.0001889247405057948,
+      "loss": 1.0938,
+      "step": 410
+    },
+    {
+      "epoch": 0.46427562835357244,
+      "grad_norm": 0.028347400948405266,
+      "learning_rate": 0.0001888703470599704,
+      "loss": 0.9892,
+      "step": 411
+    },
+    {
+      "epoch": 0.46540525275345945,
+      "grad_norm": 0.030584534630179405,
+      "learning_rate": 0.00018881582824079965,
+      "loss": 0.9977,
+      "step": 412
+    },
+    {
+      "epoch": 0.4665348771533465,
+      "grad_norm": 0.030196473002433777,
+      "learning_rate": 0.0001887611841251947,
+      "loss": 1.0442,
+      "step": 413
+    },
+    {
+      "epoch": 0.4676645015532335,
+      "grad_norm": 0.02942134439945221,
+      "learning_rate": 0.00018870641479024438,
+      "loss": 1.0096,
+      "step": 414
+    },
+    {
+      "epoch": 0.4687941259531206,
+      "grad_norm": 0.0283603947609663,
+      "learning_rate": 0.00018865152031321427,
+      "loss": 1.1341,
+      "step": 415
+    },
+    {
+      "epoch": 0.46992375035300765,
+      "grad_norm": 0.02936590276658535,
+      "learning_rate": 0.0001885965007715464,
+      "loss": 1.0823,
+      "step": 416
+    },
+    {
+      "epoch": 0.47105337475289466,
+      "grad_norm": 0.029375478625297546,
+      "learning_rate": 0.00018854135624285935,
+      "loss": 1.1148,
+      "step": 417
+    },
+    {
+      "epoch": 0.4721829991527817,
+      "grad_norm": 0.02892325632274151,
+      "learning_rate": 0.00018848608680494788,
+      "loss": 1.0905,
+      "step": 418
+    },
+    {
+      "epoch": 0.47331262355266873,
+      "grad_norm": 0.028916003182530403,
+      "learning_rate": 0.00018843069253578312,
+      "loss": 1.0133,
+      "step": 419
+    },
+    {
+      "epoch": 0.4744422479525558,
+      "grad_norm": 0.03031068667769432,
+      "learning_rate": 0.00018837517351351214,
+      "loss": 0.9835,
+      "step": 420
+    },
+    {
+      "epoch": 0.4755718723524428,
+      "grad_norm": 0.02931569144129753,
+      "learning_rate": 0.00018831952981645817,
+      "loss": 0.9664,
+      "step": 421
+    },
+    {
+      "epoch": 0.47670149675232987,
+      "grad_norm": 0.029150547459721565,
+      "learning_rate": 0.0001882637615231202,
+      "loss": 0.9604,
+      "step": 422
+    },
+    {
+      "epoch": 0.4778311211522169,
+      "grad_norm": 0.03003125637769699,
+      "learning_rate": 0.00018820786871217305,
+      "loss": 1.0735,
+      "step": 423
+    },
+    {
+      "epoch": 0.47896074555210394,
+      "grad_norm": 0.030021261423826218,
+      "learning_rate": 0.00018815185146246716,
+      "loss": 1.0005,
+      "step": 424
+    },
+    {
+      "epoch": 0.48009036995199095,
+      "grad_norm": 0.029816657304763794,
+      "learning_rate": 0.00018809570985302862,
+      "loss": 0.9366,
+      "step": 425
+    },
+    {
+      "epoch": 0.481219994351878,
+      "grad_norm": 0.02971251681447029,
+      "learning_rate": 0.00018803944396305884,
+      "loss": 1.0121,
+      "step": 426
+    },
+    {
+      "epoch": 0.482349618751765,
+      "grad_norm": 0.03110647387802601,
+      "learning_rate": 0.00018798305387193463,
+      "loss": 1.0021,
+      "step": 427
+    },
+    {
+      "epoch": 0.4834792431516521,
+      "grad_norm": 0.030216267332434654,
+      "learning_rate": 0.000187926539659208,
+      "loss": 0.9594,
+      "step": 428
+    },
+    {
+      "epoch": 0.4846088675515391,
+      "grad_norm": 0.030311699956655502,
+      "learning_rate": 0.000187869901404606,
+      "loss": 1.0478,
+      "step": 429
+    },
+    {
+      "epoch": 0.48573849195142615,
+      "grad_norm": 0.028579862788319588,
+      "learning_rate": 0.00018781313918803086,
+      "loss": 0.9539,
+      "step": 430
+    },
+    {
+      "epoch": 0.48686811635131316,
+      "grad_norm": 0.03003637120127678,
+      "learning_rate": 0.00018775625308955942,
+      "loss": 1.0172,
+      "step": 431
+    },
+    {
+      "epoch": 0.4879977407512002,
+      "grad_norm": 0.03043578751385212,
+      "learning_rate": 0.0001876992431894435,
+      "loss": 0.9997,
+      "step": 432
+    },
+    {
+      "epoch": 0.4891273651510873,
+      "grad_norm": 0.03140099346637726,
+      "learning_rate": 0.0001876421095681095,
+      "loss": 1.0307,
+      "step": 433
+    },
+    {
+      "epoch": 0.4902569895509743,
+      "grad_norm": 0.03060254082083702,
+      "learning_rate": 0.00018758485230615837,
+      "loss": 0.9873,
+      "step": 434
+    },
+    {
+      "epoch": 0.49138661395086136,
+      "grad_norm": 0.030223416164517403,
+      "learning_rate": 0.00018752747148436543,
+      "loss": 1.0629,
+      "step": 435
+    },
+    {
+      "epoch": 0.49251623835074837,
+      "grad_norm": 0.030368085950613022,
+      "learning_rate": 0.00018746996718368037,
+      "loss": 0.9692,
+      "step": 436
+    },
+    {
+      "epoch": 0.49364586275063543,
+      "grad_norm": 0.03002486564218998,
+      "learning_rate": 0.00018741233948522707,
+      "loss": 1.0334,
+      "step": 437
+    },
+    {
+      "epoch": 0.49477548715052244,
+      "grad_norm": 0.029050812125205994,
+      "learning_rate": 0.0001873545884703035,
+      "loss": 0.9861,
+      "step": 438
+    },
+    {
+      "epoch": 0.4959051115504095,
+      "grad_norm": 0.030488910153508186,
+      "learning_rate": 0.0001872967142203815,
+      "loss": 1.1141,
+      "step": 439
+    },
+    {
+      "epoch": 0.4970347359502965,
+      "grad_norm": 0.029405072331428528,
+      "learning_rate": 0.00018723871681710697,
+      "loss": 1.0318,
+      "step": 440
+    },
+    {
+      "epoch": 0.4981643603501836,
+      "grad_norm": 0.030446210876107216,
+      "learning_rate": 0.0001871805963422993,
+      "loss": 0.9895,
+      "step": 441
+    },
+    {
+      "epoch": 0.4992939847500706,
+      "grad_norm": 0.029718847945332527,
+      "learning_rate": 0.00018712235287795176,
+      "loss": 1.1104,
+      "step": 442
+    },
+    {
+      "epoch": 0.5004236091499576,
+      "grad_norm": 0.03045968897640705,
+      "learning_rate": 0.00018706398650623088,
+      "loss": 0.9305,
+      "step": 443
+    },
+    {
+      "epoch": 0.5015532335498447,
+      "grad_norm": 0.030085409060120583,
+      "learning_rate": 0.0001870054973094767,
+      "loss": 1.0243,
+      "step": 444
+    },
+    {
+      "epoch": 0.5026828579497317,
+      "grad_norm": 0.030122725293040276,
+      "learning_rate": 0.0001869468853702026,
+      "loss": 1.0977,
+      "step": 445
+    },
+    {
+      "epoch": 0.5038124823496187,
+      "grad_norm": 0.03070569783449173,
+      "learning_rate": 0.00018688815077109498,
+      "loss": 1.0352,
+      "step": 446
+    },
+    {
+      "epoch": 0.5049421067495058,
+      "grad_norm": 0.029172202572226524,
+      "learning_rate": 0.00018682929359501338,
+      "loss": 1.0018,
+      "step": 447
+    },
+    {
+      "epoch": 0.5060717311493929,
+      "grad_norm": 0.02992609702050686,
+      "learning_rate": 0.00018677031392499023,
+      "loss": 1.0543,
+      "step": 448
+    },
+    {
+      "epoch": 0.5072013555492799,
+      "grad_norm": 0.03060738928616047,
+      "learning_rate": 0.00018671121184423076,
+      "loss": 0.9548,
+      "step": 449
+    },
+    {
+      "epoch": 0.5083309799491669,
+      "grad_norm": 0.03061763569712639,
+      "learning_rate": 0.0001866519874361129,
+      "loss": 1.0017,
+      "step": 450
+    },
+    {
+      "epoch": 0.5094606043490539,
+      "grad_norm": 0.031224450096488,
+      "learning_rate": 0.00018659264078418718,
+      "loss": 1.0203,
+      "step": 451
+    },
+    {
+      "epoch": 0.510590228748941,
+      "grad_norm": 0.028874509036540985,
+      "learning_rate": 0.00018653317197217653,
+      "loss": 1.0266,
+      "step": 452
+    },
+    {
+      "epoch": 0.5117198531488281,
+      "grad_norm": 0.029967116191983223,
+      "learning_rate": 0.00018647358108397625,
+      "loss": 1.0335,
+      "step": 453
+    },
+    {
+      "epoch": 0.512849477548715,
+      "grad_norm": 0.030794909223914146,
+      "learning_rate": 0.00018641386820365385,
+      "loss": 1.0284,
+      "step": 454
+    },
+    {
+      "epoch": 0.5139791019486021,
+      "grad_norm": 0.031100483611226082,
+      "learning_rate": 0.000186354033415449,
+      "loss": 1.0486,
+      "step": 455
+    },
+    {
+      "epoch": 0.5151087263484891,
+      "grad_norm": 0.030945099890232086,
+      "learning_rate": 0.00018629407680377318,
+      "loss": 1.0685,
+      "step": 456
+    },
+    {
+      "epoch": 0.5162383507483762,
+      "grad_norm": 0.030694004148244858,
+      "learning_rate": 0.00018623399845320993,
+      "loss": 0.9765,
+      "step": 457
+    },
+    {
+      "epoch": 0.5173679751482632,
+      "grad_norm": 0.03131450340151787,
+      "learning_rate": 0.00018617379844851443,
+      "loss": 1.0927,
+      "step": 458
+    },
+    {
+      "epoch": 0.5184975995481502,
+      "grad_norm": 0.030793707817792892,
+      "learning_rate": 0.00018611347687461349,
+      "loss": 0.9999,
+      "step": 459
+    },
+    {
+      "epoch": 0.5196272239480373,
+      "grad_norm": 0.029182102531194687,
+      "learning_rate": 0.00018605303381660543,
+      "loss": 0.967,
+      "step": 460
+    },
+    {
+      "epoch": 0.5207568483479244,
+      "grad_norm": 0.030693160369992256,
+      "learning_rate": 0.00018599246935976,
+      "loss": 1.084,
+      "step": 461
+    },
+    {
+      "epoch": 0.5218864727478113,
+      "grad_norm": 0.030196724459528923,
+      "learning_rate": 0.0001859317835895181,
+      "loss": 1.024,
+      "step": 462
+    },
+    {
+      "epoch": 0.5230160971476984,
+      "grad_norm": 0.029934274032711983,
+      "learning_rate": 0.0001858709765914919,
+      "loss": 1.0975,
+      "step": 463
+    },
+    {
+      "epoch": 0.5241457215475854,
+      "grad_norm": 0.030209926888346672,
+      "learning_rate": 0.00018581004845146453,
+      "loss": 1.0485,
+      "step": 464
+    },
+    {
+      "epoch": 0.5252753459474725,
+      "grad_norm": 0.0305222999304533,
+      "learning_rate": 0.00018574899925538998,
+      "loss": 1.0272,
+      "step": 465
+    },
+    {
+      "epoch": 0.5264049703473596,
+      "grad_norm": 0.029943542554974556,
+      "learning_rate": 0.00018568782908939309,
+      "loss": 1.0122,
+      "step": 466
+    },
+    {
+      "epoch": 0.5275345947472465,
+      "grad_norm": 0.02910439483821392,
+      "learning_rate": 0.00018562653803976936,
+      "loss": 0.8831,
+      "step": 467
+    },
+    {
+      "epoch": 0.5286642191471336,
+      "grad_norm": 0.030156375840306282,
+      "learning_rate": 0.00018556512619298472,
+      "loss": 1.0245,
+      "step": 468
+    },
+    {
+      "epoch": 0.5297938435470206,
+      "grad_norm": 0.029457733035087585,
+      "learning_rate": 0.00018550359363567567,
+      "loss": 0.9933,
+      "step": 469
+    },
+    {
+      "epoch": 0.5309234679469077,
+      "grad_norm": 0.03006352297961712,
+      "learning_rate": 0.00018544194045464886,
+      "loss": 0.9978,
+      "step": 470
+    },
+    {
+      "epoch": 0.5320530923467947,
+      "grad_norm": 0.03152355179190636,
+      "learning_rate": 0.0001853801667368812,
+      "loss": 0.9832,
+      "step": 471
+    },
+    {
+      "epoch": 0.5331827167466817,
+      "grad_norm": 0.02921919897198677,
+      "learning_rate": 0.00018531827256951962,
+      "loss": 0.9178,
+      "step": 472
+    },
+    {
+      "epoch": 0.5343123411465688,
+      "grad_norm": 0.031064407899975777,
+      "learning_rate": 0.00018525625803988104,
+      "loss": 1.0384,
+      "step": 473
+    },
+    {
+      "epoch": 0.5354419655464558,
+      "grad_norm": 0.029859617352485657,
+      "learning_rate": 0.00018519412323545194,
+      "loss": 0.9886,
+      "step": 474
+    },
+    {
+      "epoch": 0.5365715899463428,
+      "grad_norm": 0.030883649364113808,
+      "learning_rate": 0.00018513186824388879,
+      "loss": 1.1247,
+      "step": 475
+    },
+    {
+      "epoch": 0.5377012143462299,
+      "grad_norm": 0.030706819146871567,
+      "learning_rate": 0.00018506949315301742,
+      "loss": 0.9923,
+      "step": 476
+    },
+    {
+      "epoch": 0.5388308387461169,
+      "grad_norm": 0.02973487228155136,
+      "learning_rate": 0.00018500699805083318,
+      "loss": 0.9388,
+      "step": 477
+    },
+    {
+      "epoch": 0.539960463146004,
+      "grad_norm": 0.03165286406874657,
+      "learning_rate": 0.00018494438302550062,
+      "loss": 1.0297,
+      "step": 478
+    },
+    {
+      "epoch": 0.5410900875458909,
+      "grad_norm": 0.0324639268219471,
+      "learning_rate": 0.0001848816481653536,
+      "loss": 1.0399,
+      "step": 479
+    },
+    {
+      "epoch": 0.542219711945778,
+      "grad_norm": 0.03156152740120888,
+      "learning_rate": 0.00018481879355889495,
+      "loss": 0.9528,
+      "step": 480
+    },
+    {
+      "epoch": 0.5433493363456651,
+      "grad_norm": 0.030102282762527466,
+      "learning_rate": 0.00018475581929479646,
+      "loss": 0.9972,
+      "step": 481
+    },
+    {
+      "epoch": 0.5444789607455521,
+      "grad_norm": 0.03062708117067814,
+      "learning_rate": 0.0001846927254618987,
+      "loss": 0.9629,
+      "step": 482
+    },
+    {
+      "epoch": 0.5456085851454392,
+      "grad_norm": 0.02973772957921028,
+      "learning_rate": 0.000184629512149211,
+      "loss": 1.1059,
+      "step": 483
+    },
+    {
+      "epoch": 0.5467382095453261,
+      "grad_norm": 0.030491316691040993,
+      "learning_rate": 0.00018456617944591111,
+      "loss": 1.093,
+      "step": 484
+    },
+    {
+      "epoch": 0.5478678339452132,
+      "grad_norm": 0.029982471838593483,
+      "learning_rate": 0.00018450272744134532,
+      "loss": 1.0719,
+      "step": 485
+    },
+    {
+      "epoch": 0.5489974583451003,
+      "grad_norm": 0.03204856067895889,
+      "learning_rate": 0.00018443915622502822,
+      "loss": 1.0136,
+      "step": 486
+    },
+    {
+      "epoch": 0.5501270827449873,
+      "grad_norm": 0.030183738097548485,
+      "learning_rate": 0.00018437546588664252,
+      "loss": 1.0613,
+      "step": 487
+    },
+    {
+      "epoch": 0.5512567071448743,
+      "grad_norm": 0.03049345500767231,
+      "learning_rate": 0.00018431165651603903,
+      "loss": 0.9428,
+      "step": 488
+    },
+    {
+      "epoch": 0.5523863315447614,
+      "grad_norm": 0.030976206064224243,
+      "learning_rate": 0.00018424772820323644,
+      "loss": 0.9908,
+      "step": 489
+    },
+    {
+      "epoch": 0.5535159559446484,
+      "grad_norm": 0.030059922486543655,
+      "learning_rate": 0.00018418368103842125,
+      "loss": 0.9546,
+      "step": 490
+    },
+    {
+      "epoch": 0.5546455803445355,
+      "grad_norm": 0.029848681762814522,
+      "learning_rate": 0.0001841195151119477,
+      "loss": 1.0269,
+      "step": 491
+    },
+    {
+      "epoch": 0.5557752047444224,
+      "grad_norm": 0.03216058760881424,
+      "learning_rate": 0.00018405523051433743,
+      "loss": 0.9717,
+      "step": 492
+    },
+    {
+      "epoch": 0.5569048291443095,
+      "grad_norm": 0.030524935573339462,
+      "learning_rate": 0.00018399082733627965,
+      "loss": 1.0208,
+      "step": 493
+    },
+    {
+      "epoch": 0.5580344535441966,
+      "grad_norm": 0.03152266517281532,
+      "learning_rate": 0.00018392630566863076,
+      "loss": 1.0353,
+      "step": 494
+    },
+    {
+      "epoch": 0.5591640779440836,
+      "grad_norm": 0.03233015537261963,
+      "learning_rate": 0.00018386166560241434,
+      "loss": 1.1238,
+      "step": 495
+    },
+    {
+      "epoch": 0.5602937023439706,
+      "grad_norm": 0.031183136627078056,
+      "learning_rate": 0.000183796907228821,
+      "loss": 1.0266,
+      "step": 496
+    },
+    {
+      "epoch": 0.5614233267438576,
+      "grad_norm": 0.030228251591324806,
+      "learning_rate": 0.00018373203063920822,
+      "loss": 1.0074,
+      "step": 497
+    },
+    {
+      "epoch": 0.5625529511437447,
+      "grad_norm": 0.031268905848264694,
+      "learning_rate": 0.00018366703592510034,
+      "loss": 1.0106,
+      "step": 498
+    },
+    {
+      "epoch": 0.5636825755436318,
+      "grad_norm": 0.031185952946543694,
+      "learning_rate": 0.0001836019231781883,
+      "loss": 1.0476,
+      "step": 499
+    },
+    {
+      "epoch": 0.5648121999435188,
+      "grad_norm": 0.03026709146797657,
+      "learning_rate": 0.0001835366924903295,
+      "loss": 1.0619,
+      "step": 500
+    },
+    {
+      "epoch": 0.5659418243434058,
+      "grad_norm": 0.029817136004567146,
+      "learning_rate": 0.00018347134395354776,
+      "loss": 1.0016,
+      "step": 501
+    },
+    {
+      "epoch": 0.5670714487432928,
+      "grad_norm": 0.030526304617524147,
+      "learning_rate": 0.00018340587766003323,
+      "loss": 1.0559,
+      "step": 502
+    },
+    {
+      "epoch": 0.5682010731431799,
+      "grad_norm": 0.03136800602078438,
+      "learning_rate": 0.00018334029370214208,
+      "loss": 0.9867,
+      "step": 503
+    },
+    {
+      "epoch": 0.569330697543067,
+      "grad_norm": 0.030273810029029846,
+      "learning_rate": 0.0001832745921723965,
+      "loss": 0.9358,
+      "step": 504
+    },
+    {
+      "epoch": 0.5704603219429539,
+      "grad_norm": 0.02991536259651184,
+      "learning_rate": 0.00018320877316348454,
+      "loss": 0.9964,
+      "step": 505
+    },
+    {
+      "epoch": 0.571589946342841,
+      "grad_norm": 0.031318966299295425,
+      "learning_rate": 0.00018314283676826009,
+      "loss": 0.9946,
+      "step": 506
+    },
+    {
+      "epoch": 0.5727195707427281,
+      "grad_norm": 0.030620397999882698,
+      "learning_rate": 0.00018307678307974241,
+      "loss": 1.0597,
+      "step": 507
+    },
+    {
+      "epoch": 0.5738491951426151,
+      "grad_norm": 0.03023059107363224,
+      "learning_rate": 0.0001830106121911165,
+      "loss": 0.9825,
+      "step": 508
+    },
+    {
+      "epoch": 0.5749788195425021,
+      "grad_norm": 0.03067387081682682,
+      "learning_rate": 0.0001829443241957325,
+      "loss": 0.9863,
+      "step": 509
+    },
+    {
+      "epoch": 0.5761084439423891,
+      "grad_norm": 0.03259598836302757,
+      "learning_rate": 0.00018287791918710587,
+      "loss": 1.0366,
+      "step": 510
+    },
+    {
+      "epoch": 0.5772380683422762,
+      "grad_norm": 0.03081597201526165,
+      "learning_rate": 0.00018281139725891707,
+      "loss": 1.144,
+      "step": 511
+    },
+    {
+      "epoch": 0.5783676927421633,
+      "grad_norm": 0.03100423514842987,
+      "learning_rate": 0.00018274475850501158,
+      "loss": 1.011,
+      "step": 512
+    },
+    {
+      "epoch": 0.5794973171420502,
+      "grad_norm": 0.030796082690358162,
+      "learning_rate": 0.00018267800301939965,
+      "loss": 0.8843,
+      "step": 513
+    },
+    {
+      "epoch": 0.5806269415419373,
+      "grad_norm": 0.030977580696344376,
+      "learning_rate": 0.00018261113089625613,
+      "loss": 1.0606,
+      "step": 514
+    },
+    {
+      "epoch": 0.5817565659418243,
+      "grad_norm": 0.03037908300757408,
+      "learning_rate": 0.0001825441422299206,
+      "loss": 0.9751,
+      "step": 515
+    },
+    {
+      "epoch": 0.5828861903417114,
+      "grad_norm": 0.03079284355044365,
+      "learning_rate": 0.00018247703711489686,
+      "loss": 1.0062,
+      "step": 516
+    },
+    {
+      "epoch": 0.5840158147415985,
+      "grad_norm": 0.031534090638160706,
+      "learning_rate": 0.00018240981564585313,
+      "loss": 0.949,
+      "step": 517
+    },
+    {
+      "epoch": 0.5851454391414854,
+      "grad_norm": 0.03137180209159851,
+      "learning_rate": 0.0001823424779176217,
+      "loss": 1.0799,
+      "step": 518
+    },
+    {
+      "epoch": 0.5862750635413725,
+      "grad_norm": 0.0305685643106699,
+      "learning_rate": 0.00018227502402519893,
+      "loss": 1.0609,
+      "step": 519
+    },
+    {
+      "epoch": 0.5874046879412596,
+      "grad_norm": 0.02950458414852619,
+      "learning_rate": 0.00018220745406374498,
+      "loss": 0.9671,
+      "step": 520
+    },
+    {
+      "epoch": 0.5885343123411466,
+      "grad_norm": 0.030199820175766945,
+      "learning_rate": 0.00018213976812858382,
+      "loss": 1.0684,
+      "step": 521
+    },
+    {
+      "epoch": 0.5896639367410336,
+      "grad_norm": 0.031708989292383194,
+      "learning_rate": 0.00018207196631520297,
+      "loss": 0.9994,
+      "step": 522
+    },
+    {
+      "epoch": 0.5907935611409206,
+      "grad_norm": 0.03120891936123371,
+      "learning_rate": 0.00018200404871925353,
+      "loss": 1.001,
+      "step": 523
+    },
+    {
+      "epoch": 0.5919231855408077,
+      "grad_norm": 0.033152077347040176,
+      "learning_rate": 0.0001819360154365498,
+      "loss": 1.0489,
+      "step": 524
+    },
+    {
+      "epoch": 0.5930528099406948,
+      "grad_norm": 0.03135927394032478,
+      "learning_rate": 0.00018186786656306935,
+      "loss": 1.1065,
+      "step": 525
+    },
+    {
+      "epoch": 0.5941824343405817,
+      "grad_norm": 0.030605459585785866,
+      "learning_rate": 0.0001817996021949529,
+      "loss": 1.0116,
+      "step": 526
+    },
+    {
+      "epoch": 0.5953120587404688,
+      "grad_norm": 0.031958550214767456,
+      "learning_rate": 0.00018173122242850397,
+      "loss": 1.0113,
+      "step": 527
+    },
+    {
+      "epoch": 0.5964416831403558,
+      "grad_norm": 0.033079009503126144,
+      "learning_rate": 0.00018166272736018895,
+      "loss": 0.9531,
+      "step": 528
+    },
+    {
+      "epoch": 0.5975713075402429,
+      "grad_norm": 0.0316440686583519,
+      "learning_rate": 0.00018159411708663684,
+      "loss": 0.9916,
+      "step": 529
+    },
+    {
+      "epoch": 0.5987009319401299,
+      "grad_norm": 0.030489858239889145,
+      "learning_rate": 0.00018152539170463925,
+      "loss": 0.995,
+      "step": 530
+    },
+    {
+      "epoch": 0.5998305563400169,
+      "grad_norm": 0.0322355218231678,
+      "learning_rate": 0.00018145655131115009,
+      "loss": 1.0784,
+      "step": 531
+    },
+    {
+      "epoch": 0.600960180739904,
+      "grad_norm": 0.03130833059549332,
+      "learning_rate": 0.00018138759600328563,
+      "loss": 1.0537,
+      "step": 532
+    },
+    {
+      "epoch": 0.602089805139791,
+      "grad_norm": 0.031001951545476913,
+      "learning_rate": 0.0001813185258783241,
+      "loss": 0.9956,
+      "step": 533
+    },
+    {
+      "epoch": 0.6032194295396781,
+      "grad_norm": 0.03067929483950138,
+      "learning_rate": 0.0001812493410337058,
+      "loss": 1.0148,
+      "step": 534
+    },
+    {
+      "epoch": 0.6043490539395651,
+      "grad_norm": 0.03192298486828804,
+      "learning_rate": 0.00018118004156703296,
+      "loss": 0.9635,
+      "step": 535
+    },
+    {
+      "epoch": 0.6054786783394521,
+      "grad_norm": 0.031253885477781296,
+      "learning_rate": 0.00018111062757606932,
+      "loss": 0.9987,
+      "step": 536
+    },
+    {
+      "epoch": 0.6066083027393392,
+      "grad_norm": 0.031125715002417564,
+      "learning_rate": 0.0001810410991587403,
+      "loss": 0.9915,
+      "step": 537
+    },
+    {
+      "epoch": 0.6077379271392263,
+      "grad_norm": 0.03175501897931099,
+      "learning_rate": 0.00018097145641313272,
+      "loss": 1.0357,
+      "step": 538
+    },
+    {
+      "epoch": 0.6088675515391132,
+      "grad_norm": 0.031910236924886703,
+      "learning_rate": 0.00018090169943749476,
+      "loss": 1.0679,
+      "step": 539
+    },
+    {
+      "epoch": 0.6099971759390003,
+      "grad_norm": 0.03214259445667267,
+      "learning_rate": 0.00018083182833023562,
+      "loss": 1.0173,
+      "step": 540
+    },
+    {
+      "epoch": 0.6111268003388873,
+      "grad_norm": 0.03169810026884079,
+      "learning_rate": 0.00018076184318992558,
+      "loss": 1.0428,
+      "step": 541
+    },
+    {
+      "epoch": 0.6122564247387744,
+      "grad_norm": 0.03129338473081589,
+      "learning_rate": 0.00018069174411529577,
+      "loss": 1.0236,
+      "step": 542
+    },
+    {
+      "epoch": 0.6133860491386613,
+      "grad_norm": 0.03245764225721359,
+      "learning_rate": 0.0001806215312052381,
+      "loss": 1.0081,
+      "step": 543
+    },
+    {
+      "epoch": 0.6145156735385484,
+      "grad_norm": 0.030435949563980103,
+      "learning_rate": 0.0001805512045588051,
+      "loss": 1.0731,
+      "step": 544
+    },
+    {
+      "epoch": 0.6156452979384355,
+      "grad_norm": 0.030730856582522392,
+      "learning_rate": 0.0001804807642752096,
+      "loss": 1.0793,
+      "step": 545
+    },
+    {
+      "epoch": 0.6167749223383225,
+      "grad_norm": 0.02937515825033188,
+      "learning_rate": 0.00018041021045382485,
+      "loss": 1.0123,
+      "step": 546
+    },
+    {
+      "epoch": 0.6179045467382095,
+      "grad_norm": 0.03019302524626255,
+      "learning_rate": 0.0001803395431941843,
+      "loss": 1.0232,
+      "step": 547
+    },
+    {
+      "epoch": 0.6190341711380966,
+      "grad_norm": 0.04123188927769661,
+      "learning_rate": 0.00018026876259598135,
+      "loss": 1.0309,
+      "step": 548
+    },
+    {
+      "epoch": 0.6201637955379836,
+      "grad_norm": 0.03046722523868084,
+      "learning_rate": 0.00018019786875906935,
+      "loss": 0.9721,
+      "step": 549
+    },
+    {
+      "epoch": 0.6212934199378707,
+      "grad_norm": 0.033260468393564224,
+      "learning_rate": 0.00018012686178346142,
+      "loss": 1.0726,
+      "step": 550
+    },
+    {
+      "epoch": 0.6224230443377577,
+      "grad_norm": 0.03144606575369835,
+      "learning_rate": 0.0001800557417693302,
+      "loss": 0.947,
+      "step": 551
+    },
+    {
+      "epoch": 0.6235526687376447,
+      "grad_norm": 0.03095083311200142,
+      "learning_rate": 0.00017998450881700787,
+      "loss": 0.9937,
+      "step": 552
+    },
+    {
+      "epoch": 0.6246822931375318,
+      "grad_norm": 0.03133854269981384,
+      "learning_rate": 0.00017991316302698595,
+      "loss": 0.9502,
+      "step": 553
+    },
+    {
+      "epoch": 0.6258119175374188,
+      "grad_norm": 0.03148304298520088,
+      "learning_rate": 0.00017984170449991506,
+      "loss": 1.1628,
+      "step": 554
+    },
+    {
+      "epoch": 0.6269415419373059,
+      "grad_norm": 0.03164827451109886,
+      "learning_rate": 0.000179770133336605,
+      "loss": 0.8814,
+      "step": 555
+    },
+    {
+      "epoch": 0.6280711663371928,
+      "grad_norm": 0.03083074651658535,
+      "learning_rate": 0.0001796984496380243,
+      "loss": 0.9999,
+      "step": 556
+    },
+    {
+      "epoch": 0.6292007907370799,
+      "grad_norm": 0.03223288804292679,
+      "learning_rate": 0.0001796266535053004,
+      "loss": 1.0819,
+      "step": 557
+    },
+    {
+      "epoch": 0.630330415136967,
+      "grad_norm": 0.03053288348019123,
+      "learning_rate": 0.00017955474503971925,
+      "loss": 1.1077,
+      "step": 558
+    },
+    {
+      "epoch": 0.631460039536854,
+      "grad_norm": 0.03127776086330414,
+      "learning_rate": 0.00017948272434272535,
+      "loss": 1.041,
+      "step": 559
+    },
+    {
+      "epoch": 0.632589663936741,
+      "grad_norm": 0.03209880739450455,
+      "learning_rate": 0.00017941059151592147,
+      "loss": 1.0081,
+      "step": 560
+    },
+    {
+      "epoch": 0.633719288336628,
+      "grad_norm": 0.02959609404206276,
+      "learning_rate": 0.00017933834666106864,
+      "loss": 0.9875,
+      "step": 561
+    },
+    {
+      "epoch": 0.6348489127365151,
+      "grad_norm": 0.03344092145562172,
+      "learning_rate": 0.00017926598988008582,
+      "loss": 0.9677,
+      "step": 562
+    },
+    {
+      "epoch": 0.6359785371364022,
+      "grad_norm": 0.03260407596826553,
+      "learning_rate": 0.00017919352127505,
+      "loss": 1.0449,
+      "step": 563
+    },
+    {
+      "epoch": 0.6371081615362891,
+      "grad_norm": 0.031249945983290672,
+      "learning_rate": 0.0001791209409481958,
+      "loss": 1.0662,
+      "step": 564
+    },
+    {
+      "epoch": 0.6382377859361762,
+      "grad_norm": 0.031923823058605194,
+      "learning_rate": 0.00017904824900191556,
+      "loss": 1.0379,
+      "step": 565
+    },
+    {
+      "epoch": 0.6393674103360633,
+      "grad_norm": 0.030242929235100746,
+      "learning_rate": 0.00017897544553875902,
+      "loss": 1.0257,
+      "step": 566
+    },
+    {
+      "epoch": 0.6404970347359503,
+      "grad_norm": 0.032716382294893265,
+      "learning_rate": 0.00017890253066143324,
+      "loss": 0.9987,
+      "step": 567
+    },
+    {
+      "epoch": 0.6416266591358374,
+      "grad_norm": 0.03140626102685928,
+      "learning_rate": 0.0001788295044728025,
+      "loss": 1.0162,
+      "step": 568
+    },
+    {
+      "epoch": 0.6427562835357243,
+      "grad_norm": 0.029912738129496574,
+      "learning_rate": 0.0001787563670758881,
+      "loss": 1.0318,
+      "step": 569
+    },
+    {
+      "epoch": 0.6438859079356114,
+      "grad_norm": 0.03130066394805908,
+      "learning_rate": 0.0001786831185738682,
+      "loss": 1.0026,
+      "step": 570
+    },
+    {
+      "epoch": 0.6450155323354985,
+      "grad_norm": 0.033079057931900024,
+      "learning_rate": 0.00017860975907007772,
+      "loss": 1.0262,
+      "step": 571
+    },
+    {
+      "epoch": 0.6461451567353855,
+      "grad_norm": 0.03027520515024662,
+      "learning_rate": 0.00017853628866800812,
+      "loss": 1.0075,
+      "step": 572
+    },
+    {
+      "epoch": 0.6472747811352725,
+      "grad_norm": 0.03166157007217407,
+      "learning_rate": 0.00017846270747130742,
+      "loss": 1.0858,
+      "step": 573
+    },
+    {
+      "epoch": 0.6484044055351595,
+      "grad_norm": 0.03081650286912918,
+      "learning_rate": 0.00017838901558377986,
+      "loss": 1.0215,
+      "step": 574
+    },
+    {
+      "epoch": 0.6495340299350466,
+      "grad_norm": 0.03256387263536453,
+      "learning_rate": 0.0001783152131093859,
+      "loss": 0.973,
+      "step": 575
+    },
+    {
+      "epoch": 0.6506636543349337,
+      "grad_norm": 0.030604898929595947,
+      "learning_rate": 0.00017824130015224192,
+      "loss": 1.057,
+      "step": 576
+    },
+    {
+      "epoch": 0.6517932787348206,
+      "grad_norm": 0.030695458874106407,
+      "learning_rate": 0.00017816727681662023,
+      "loss": 1.0804,
+      "step": 577
+    },
+    {
+      "epoch": 0.6529229031347077,
+      "grad_norm": 0.031340498477220535,
+      "learning_rate": 0.0001780931432069488,
+      "loss": 0.9722,
+      "step": 578
+    },
+    {
+      "epoch": 0.6540525275345948,
+      "grad_norm": 0.03206819295883179,
+      "learning_rate": 0.00017801889942781126,
+      "loss": 1.0593,
+      "step": 579
+    },
+    {
+      "epoch": 0.6551821519344818,
+      "grad_norm": 0.030380915850400925,
+      "learning_rate": 0.00017794454558394657,
+      "loss": 0.9263,
+      "step": 580
+    },
+    {
+      "epoch": 0.6563117763343688,
+      "grad_norm": 0.03320132568478584,
+      "learning_rate": 0.00017787008178024905,
+      "loss": 1.0798,
+      "step": 581
+    },
+    {
+      "epoch": 0.6574414007342558,
+      "grad_norm": 0.0311865396797657,
+      "learning_rate": 0.00017779550812176806,
+      "loss": 0.9205,
+      "step": 582
+    },
+    {
+      "epoch": 0.6585710251341429,
+      "grad_norm": 0.032210152596235275,
+      "learning_rate": 0.00017772082471370797,
+      "loss": 1.0411,
+      "step": 583
+    },
+    {
+      "epoch": 0.65970064953403,
+      "grad_norm": 0.03178109973669052,
+      "learning_rate": 0.00017764603166142798,
+      "loss": 1.0502,
+      "step": 584
+    },
+    {
+      "epoch": 0.660830273933917,
+      "grad_norm": 0.0323721244931221,
+      "learning_rate": 0.000177571129070442,
+      "loss": 1.0372,
+      "step": 585
+    },
+    {
+      "epoch": 0.661959898333804,
+      "grad_norm": 0.031241752207279205,
+      "learning_rate": 0.0001774961170464184,
+      "loss": 0.9741,
+      "step": 586
+    },
+    {
+      "epoch": 0.663089522733691,
+      "grad_norm": 0.03263148292899132,
+      "learning_rate": 0.00017742099569518,
+      "loss": 1.0956,
+      "step": 587
+    },
+    {
+      "epoch": 0.6642191471335781,
+      "grad_norm": 0.031760070472955704,
+      "learning_rate": 0.00017734576512270383,
+      "loss": 0.9795,
+      "step": 588
+    },
+    {
+      "epoch": 0.6653487715334652,
+      "grad_norm": 0.03184381127357483,
+      "learning_rate": 0.00017727042543512099,
+      "loss": 0.9054,
+      "step": 589
+    },
+    {
+      "epoch": 0.6664783959333521,
+      "grad_norm": 0.03145081177353859,
+      "learning_rate": 0.00017719497673871653,
+      "loss": 1.0219,
+      "step": 590
+    },
+    {
+      "epoch": 0.6664783959333521,
+      "eval_loss": 1.0118999481201172,
+      "eval_runtime": 547.41,
+      "eval_samples_per_second": 17.871,
+      "eval_steps_per_second": 8.937,
+      "step": 590
+    },
+    {
+      "epoch": 0.6676080203332392,
+      "grad_norm": 0.03253559768199921,
+      "learning_rate": 0.00017711941913992928,
+      "loss": 0.9635,
+      "step": 591
+    },
+    {
+      "epoch": 0.6687376447331262,
+      "grad_norm": 0.03181855380535126,
+      "learning_rate": 0.00017704375274535167,
+      "loss": 0.8852,
+      "step": 592
+    },
+    {
+      "epoch": 0.6698672691330133,
+      "grad_norm": 0.03165988251566887,
+      "learning_rate": 0.0001769679776617297,
+      "loss": 1.0201,
+      "step": 593
+    },
+    {
+      "epoch": 0.6709968935329003,
+      "grad_norm": 0.03077312745153904,
+      "learning_rate": 0.00017689209399596257,
+      "loss": 1.0307,
+      "step": 594
+    },
+    {
+      "epoch": 0.6721265179327873,
+      "grad_norm": 0.032232630997896194,
+      "learning_rate": 0.00017681610185510285,
+      "loss": 1.0121,
+      "step": 595
+    },
+    {
+      "epoch": 0.6732561423326744,
+      "grad_norm": 0.03249699994921684,
+      "learning_rate": 0.0001767400013463559,
+      "loss": 0.9288,
+      "step": 596
+    },
+    {
+      "epoch": 0.6743857667325615,
+      "grad_norm": 0.03216133266687393,
+      "learning_rate": 0.0001766637925770802,
+      "loss": 0.9665,
+      "step": 597
+    },
+    {
+      "epoch": 0.6755153911324484,
+      "grad_norm": 0.03151794150471687,
+      "learning_rate": 0.00017658747565478677,
+      "loss": 1.0497,
+      "step": 598
+    },
+    {
+      "epoch": 0.6766450155323355,
+      "grad_norm": 0.03118024580180645,
+      "learning_rate": 0.00017651105068713935,
+      "loss": 0.9403,
+      "step": 599
+    },
+    {
+      "epoch": 0.6777746399322225,
+      "grad_norm": 0.030804403126239777,
+      "learning_rate": 0.00017643451778195395,
+      "loss": 1.0011,
+      "step": 600
+    },
+    {
+      "epoch": 0.6789042643321096,
+      "grad_norm": 0.03352154418826103,
+      "learning_rate": 0.000176357877047199,
+      "loss": 1.0294,
+      "step": 601
+    },
+    {
+      "epoch": 0.6800338887319967,
+      "grad_norm": 0.03205511346459389,
+      "learning_rate": 0.00017628112859099498,
+      "loss": 1.0487,
+      "step": 602
+    },
+    {
+      "epoch": 0.6811635131318836,
+      "grad_norm": 0.031228026375174522,
+      "learning_rate": 0.00017620427252161433,
+      "loss": 0.9319,
+      "step": 603
+    },
+    {
+      "epoch": 0.6822931375317707,
+      "grad_norm": 0.031973280012607574,
+      "learning_rate": 0.00017612730894748136,
+      "loss": 1.0829,
+      "step": 604
+    },
+    {
+      "epoch": 0.6834227619316577,
+      "grad_norm": 0.03331442177295685,
+      "learning_rate": 0.00017605023797717195,
+      "loss": 1.0669,
+      "step": 605
+    },
+    {
+      "epoch": 0.6845523863315448,
+      "grad_norm": 0.0336139053106308,
+      "learning_rate": 0.00017597305971941358,
+      "loss": 1.0722,
+      "step": 606
+    },
+    {
+      "epoch": 0.6856820107314318,
+      "grad_norm": 0.03086121752858162,
+      "learning_rate": 0.00017589577428308502,
+      "loss": 1.092,
+      "step": 607
+    },
+    {
+      "epoch": 0.6868116351313188,
+      "grad_norm": 0.032204341143369675,
+      "learning_rate": 0.0001758183817772163,
+      "loss": 0.9483,
+      "step": 608
+    },
+    {
+      "epoch": 0.6879412595312059,
+      "grad_norm": 0.03183162584900856,
+      "learning_rate": 0.00017574088231098843,
+      "loss": 1.0029,
+      "step": 609
+    },
+    {
+      "epoch": 0.689070883931093,
+      "grad_norm": 0.031096026301383972,
+      "learning_rate": 0.00017566327599373338,
+      "loss": 1.0094,
+      "step": 610
+    },
+    {
+      "epoch": 0.6902005083309799,
+      "grad_norm": 0.032303981482982635,
+      "learning_rate": 0.0001755855629349338,
+      "loss": 0.976,
+      "step": 611
+    },
+    {
+      "epoch": 0.691330132730867,
+      "grad_norm": 0.03237254545092583,
+      "learning_rate": 0.00017550774324422296,
+      "loss": 0.9472,
+      "step": 612
+    },
+    {
+      "epoch": 0.692459757130754,
+      "grad_norm": 0.03161952272057533,
+      "learning_rate": 0.0001754298170313846,
+      "loss": 0.995,
+      "step": 613
+    },
+    {
+      "epoch": 0.6935893815306411,
+      "grad_norm": 0.032882727682590485,
+      "learning_rate": 0.00017535178440635264,
+      "loss": 0.9078,
+      "step": 614
+    },
+    {
+      "epoch": 0.694719005930528,
+      "grad_norm": 0.030476143583655357,
+      "learning_rate": 0.0001752736454792112,
+      "loss": 0.9488,
+      "step": 615
+    },
+    {
+      "epoch": 0.6958486303304151,
+      "grad_norm": 0.032640308141708374,
+      "learning_rate": 0.00017519540036019428,
+      "loss": 0.9968,
+      "step": 616
+    },
+    {
+      "epoch": 0.6969782547303022,
+      "grad_norm": 0.03207506611943245,
+      "learning_rate": 0.00017511704915968581,
+      "loss": 1.0598,
+      "step": 617
+    },
+    {
+      "epoch": 0.6981078791301892,
+      "grad_norm": 0.04379906877875328,
+      "learning_rate": 0.0001750385919882193,
+      "loss": 1.0801,
+      "step": 618
+    },
+    {
+      "epoch": 0.6992375035300763,
+      "grad_norm": 0.03258811682462692,
+      "learning_rate": 0.00017496002895647775,
+      "loss": 1.0901,
+      "step": 619
+    },
+    {
+      "epoch": 0.7003671279299633,
+      "grad_norm": 0.034346289932727814,
+      "learning_rate": 0.0001748813601752935,
+      "loss": 0.9995,
+      "step": 620
+    },
+    {
+      "epoch": 0.7014967523298503,
+      "grad_norm": 0.0332537442445755,
+      "learning_rate": 0.0001748025857556481,
+      "loss": 1.0382,
+      "step": 621
+    },
+    {
+      "epoch": 0.7026263767297374,
+      "grad_norm": 0.031845249235630035,
+      "learning_rate": 0.0001747237058086722,
+      "loss": 1.1217,
+      "step": 622
+    },
+    {
+      "epoch": 0.7037560011296244,
+      "grad_norm": 0.032574612647295,
+      "learning_rate": 0.00017464472044564512,
+      "loss": 0.8765,
+      "step": 623
+    },
+    {
+      "epoch": 0.7048856255295114,
+      "grad_norm": 0.031227873638272285,
+      "learning_rate": 0.00017456562977799514,
+      "loss": 0.9676,
+      "step": 624
+    },
+    {
+      "epoch": 0.7060152499293985,
+      "grad_norm": 0.032643262296915054,
+      "learning_rate": 0.00017448643391729888,
+      "loss": 0.9842,
+      "step": 625
+    },
+    {
+      "epoch": 0.7071448743292855,
+      "grad_norm": 0.031137650832533836,
+      "learning_rate": 0.00017440713297528154,
+      "loss": 0.9877,
+      "step": 626
+    },
+    {
+      "epoch": 0.7082744987291726,
+      "grad_norm": 0.030961019918322563,
+      "learning_rate": 0.0001743277270638164,
+      "loss": 1.017,
+      "step": 627
+    },
+    {
+      "epoch": 0.7094041231290595,
+      "grad_norm": 0.032677747309207916,
+      "learning_rate": 0.00017424821629492495,
+      "loss": 1.0023,
+      "step": 628
+    },
+    {
+      "epoch": 0.7105337475289466,
+      "grad_norm": 0.032716501504182816,
+      "learning_rate": 0.00017416860078077657,
+      "loss": 0.9893,
+      "step": 629
+    },
+    {
+      "epoch": 0.7116633719288337,
+      "grad_norm": 0.03239135444164276,
+      "learning_rate": 0.0001740888806336884,
+      "loss": 0.9949,
+      "step": 630
+    },
+    {
+      "epoch": 0.7127929963287207,
+      "grad_norm": 0.03276536986231804,
+      "learning_rate": 0.0001740090559661252,
+      "loss": 1.0778,
+      "step": 631
+    },
+    {
+      "epoch": 0.7139226207286078,
+      "grad_norm": 0.030909627676010132,
+      "learning_rate": 0.00017392912689069917,
+      "loss": 1.0098,
+      "step": 632
+    },
+    {
+      "epoch": 0.7150522451284947,
+      "grad_norm": 0.032094262540340424,
+      "learning_rate": 0.00017384909352016975,
+      "loss": 0.9703,
+      "step": 633
+    },
+    {
+      "epoch": 0.7161818695283818,
+      "grad_norm": 0.03388513997197151,
+      "learning_rate": 0.00017376895596744367,
+      "loss": 1.0014,
+      "step": 634
+    },
+    {
+      "epoch": 0.7173114939282689,
+      "grad_norm": 0.03186871111392975,
+      "learning_rate": 0.00017368871434557447,
+      "loss": 1.0076,
+      "step": 635
+    },
+    {
+      "epoch": 0.7184411183281559,
+      "grad_norm": 0.03189585730433464,
+      "learning_rate": 0.00017360836876776256,
+      "loss": 0.9721,
+      "step": 636
+    },
+    {
+      "epoch": 0.7195707427280429,
+      "grad_norm": 0.033433668315410614,
+      "learning_rate": 0.0001735279193473551,
+      "loss": 0.9798,
+      "step": 637
+    },
+    {
+      "epoch": 0.72070036712793,
+      "grad_norm": 0.031073307618498802,
+      "learning_rate": 0.00017344736619784553,
+      "loss": 0.9629,
+      "step": 638
+    },
+    {
+      "epoch": 0.721829991527817,
+      "grad_norm": 0.030326619744300842,
+      "learning_rate": 0.00017336670943287388,
+      "loss": 1.0727,
+      "step": 639
+    },
+    {
+      "epoch": 0.7229596159277041,
+      "grad_norm": 0.03189557045698166,
+      "learning_rate": 0.00017328594916622616,
+      "loss": 1.0175,
+      "step": 640
+    },
+    {
+      "epoch": 0.724089240327591,
+      "grad_norm": 0.03241390734910965,
+      "learning_rate": 0.00017320508551183446,
+      "loss": 1.1313,
+      "step": 641
+    },
+    {
+      "epoch": 0.7252188647274781,
+      "grad_norm": 0.0323265865445137,
+      "learning_rate": 0.0001731241185837768,
+      "loss": 1.0418,
+      "step": 642
+    },
+    {
+      "epoch": 0.7263484891273652,
+      "grad_norm": 0.032344575971364975,
+      "learning_rate": 0.00017304304849627677,
+      "loss": 1.0882,
+      "step": 643
+    },
+    {
+      "epoch": 0.7274781135272522,
+      "grad_norm": 0.032916001975536346,
+      "learning_rate": 0.00017296187536370355,
+      "loss": 0.9596,
+      "step": 644
+    },
+    {
+      "epoch": 0.7286077379271392,
+      "grad_norm": 0.031346168369054794,
+      "learning_rate": 0.00017288059930057166,
+      "loss": 0.9729,
+      "step": 645
+    },
+    {
+      "epoch": 0.7297373623270262,
+      "grad_norm": 0.032094355672597885,
+      "learning_rate": 0.00017279922042154092,
+      "loss": 1.0331,
+      "step": 646
+    },
+    {
+      "epoch": 0.7308669867269133,
+      "grad_norm": 0.03289850801229477,
+      "learning_rate": 0.00017271773884141607,
+      "loss": 1.0411,
+      "step": 647
+    },
+    {
+      "epoch": 0.7319966111268004,
+      "grad_norm": 0.03297988697886467,
+      "learning_rate": 0.0001726361546751468,
+      "loss": 0.9847,
+      "step": 648
+    },
+    {
+      "epoch": 0.7331262355266874,
+      "grad_norm": 0.03368350863456726,
+      "learning_rate": 0.00017255446803782754,
+      "loss": 0.9978,
+      "step": 649
+    },
+    {
+      "epoch": 0.7342558599265744,
+      "grad_norm": 0.03280177712440491,
+      "learning_rate": 0.00017247267904469725,
+      "loss": 1.0363,
+      "step": 650
+    },
+    {
+      "epoch": 0.7353854843264614,
+      "grad_norm": 0.031202217563986778,
+      "learning_rate": 0.00017239078781113926,
+      "loss": 1.025,
+      "step": 651
+    },
+    {
+      "epoch": 0.7365151087263485,
+      "grad_norm": 0.03219619765877724,
+      "learning_rate": 0.00017230879445268124,
+      "loss": 0.9878,
+      "step": 652
+    },
+    {
+      "epoch": 0.7376447331262356,
+      "grad_norm": 0.033005617558956146,
+      "learning_rate": 0.00017222669908499482,
+      "loss": 1.0223,
+      "step": 653
+    },
+    {
+      "epoch": 0.7387743575261225,
+      "grad_norm": 0.03350326791405678,
+      "learning_rate": 0.00017214450182389559,
+      "loss": 1.0173,
+      "step": 654
+    },
+    {
+      "epoch": 0.7399039819260096,
+      "grad_norm": 0.031389541923999786,
+      "learning_rate": 0.00017206220278534286,
+      "loss": 1.0458,
+      "step": 655
+    },
+    {
+      "epoch": 0.7410336063258967,
+      "grad_norm": 0.031541019678115845,
+      "learning_rate": 0.00017197980208543954,
+      "loss": 0.9489,
+      "step": 656
+    },
+    {
+      "epoch": 0.7421632307257837,
+      "grad_norm": 0.03202977776527405,
+      "learning_rate": 0.00017189729984043204,
+      "loss": 1.0364,
+      "step": 657
+    },
+    {
+      "epoch": 0.7432928551256707,
+      "grad_norm": 0.03152487054467201,
+      "learning_rate": 0.00017181469616670984,
+      "loss": 0.9827,
+      "step": 658
+    },
+    {
+      "epoch": 0.7444224795255577,
+      "grad_norm": 0.03225429356098175,
+      "learning_rate": 0.00017173199118080564,
+      "loss": 1.0996,
+      "step": 659
+    },
+    {
+      "epoch": 0.7455521039254448,
+      "grad_norm": 0.03217494860291481,
+      "learning_rate": 0.00017164918499939504,
+      "loss": 0.9355,
+      "step": 660
+    },
+    {
+      "epoch": 0.7466817283253319,
+      "grad_norm": 0.032104648649692535,
+      "learning_rate": 0.00017156627773929644,
+      "loss": 1.0552,
+      "step": 661
+    },
+    {
+      "epoch": 0.7478113527252188,
+      "grad_norm": 0.03186746686697006,
+      "learning_rate": 0.0001714832695174707,
+      "loss": 1.071,
+      "step": 662
+    },
+    {
+      "epoch": 0.7489409771251059,
+      "grad_norm": 0.03182530775666237,
+      "learning_rate": 0.00017140016045102133,
+      "loss": 1.0688,
+      "step": 663
+    },
+    {
+      "epoch": 0.7500706015249929,
+      "grad_norm": 0.03153397887945175,
+      "learning_rate": 0.00017131695065719386,
+      "loss": 0.9624,
+      "step": 664
+    },
+    {
+      "epoch": 0.75120022592488,
+      "grad_norm": 0.03226126730442047,
+      "learning_rate": 0.0001712336402533761,
+      "loss": 1.1134,
+      "step": 665
+    },
+    {
+      "epoch": 0.7523298503247671,
+      "grad_norm": 0.031511638313531876,
+      "learning_rate": 0.00017115022935709778,
+      "loss": 1.0753,
+      "step": 666
+    },
+    {
+      "epoch": 0.753459474724654,
+      "grad_norm": 0.03331499546766281,
+      "learning_rate": 0.00017106671808603027,
+      "loss": 0.9709,
+      "step": 667
+    },
+    {
+      "epoch": 0.7545890991245411,
+      "grad_norm": 0.032829850912094116,
+      "learning_rate": 0.0001709831065579867,
+      "loss": 0.9839,
+      "step": 668
+    },
+    {
+      "epoch": 0.7557187235244281,
+      "grad_norm": 0.032828208059072495,
+      "learning_rate": 0.00017089939489092152,
+      "loss": 0.9924,
+      "step": 669
+    },
+    {
+      "epoch": 0.7568483479243152,
+      "grad_norm": 0.0320126973092556,
+      "learning_rate": 0.00017081558320293055,
+      "loss": 0.9649,
+      "step": 670
+    },
+    {
+      "epoch": 0.7579779723242022,
+      "grad_norm": 0.03252957761287689,
+      "learning_rate": 0.0001707316716122506,
+      "loss": 0.9643,
+      "step": 671
+    },
+    {
+      "epoch": 0.7591075967240892,
+      "grad_norm": 0.032323576509952545,
+      "learning_rate": 0.00017064766023725948,
+      "loss": 0.9962,
+      "step": 672
+    },
+    {
+      "epoch": 0.7602372211239763,
+      "grad_norm": 0.03305547684431076,
+      "learning_rate": 0.00017056354919647583,
+      "loss": 1.0864,
+      "step": 673
+    },
+    {
+      "epoch": 0.7613668455238634,
+      "grad_norm": 0.0321505106985569,
+      "learning_rate": 0.0001704793386085588,
+      "loss": 1.01,
+      "step": 674
+    },
+    {
+      "epoch": 0.7624964699237503,
+      "grad_norm": 0.03243474289774895,
+      "learning_rate": 0.000170395028592308,
+      "loss": 1.1192,
+      "step": 675
+    },
+    {
+      "epoch": 0.7636260943236374,
+      "grad_norm": 0.03369235247373581,
+      "learning_rate": 0.00017031061926666333,
+      "loss": 0.9846,
+      "step": 676
+    },
+    {
+      "epoch": 0.7647557187235244,
+      "grad_norm": 0.03172389790415764,
+      "learning_rate": 0.00017022611075070474,
+      "loss": 1.0406,
+      "step": 677
+    },
+    {
+      "epoch": 0.7658853431234115,
+      "grad_norm": 0.03241589665412903,
+      "learning_rate": 0.00017014150316365216,
+      "loss": 0.9235,
+      "step": 678
+    },
+    {
+      "epoch": 0.7670149675232985,
+      "grad_norm": 0.03271762281656265,
+      "learning_rate": 0.0001700567966248653,
+      "loss": 0.9516,
+      "step": 679
+    },
+    {
+      "epoch": 0.7681445919231855,
+      "grad_norm": 0.032931018620729446,
+      "learning_rate": 0.00016997199125384343,
+      "loss": 1.0315,
+      "step": 680
+    },
+    {
+      "epoch": 0.7692742163230726,
+      "grad_norm": 0.032814498990774155,
+      "learning_rate": 0.00016988708717022522,
+      "loss": 0.941,
+      "step": 681
+    },
+    {
+      "epoch": 0.7704038407229596,
+      "grad_norm": 0.031445201486349106,
+      "learning_rate": 0.00016980208449378866,
+      "loss": 1.0588,
+      "step": 682
+    },
+    {
+      "epoch": 0.7715334651228467,
+      "grad_norm": 0.033274564892053604,
+      "learning_rate": 0.0001697169833444508,
+      "loss": 0.9968,
+      "step": 683
+    },
+    {
+      "epoch": 0.7726630895227337,
+      "grad_norm": 0.03313668072223663,
+      "learning_rate": 0.00016963178384226763,
+      "loss": 1.0308,
+      "step": 684
+    },
+    {
+      "epoch": 0.7737927139226207,
+      "grad_norm": 0.032828278839588165,
+      "learning_rate": 0.00016954648610743384,
+      "loss": 1.0245,
+      "step": 685
+    },
+    {
+      "epoch": 0.7749223383225078,
+      "grad_norm": 0.03268923610448837,
+      "learning_rate": 0.00016946109026028274,
+      "loss": 1.0515,
+      "step": 686
+    },
+    {
+      "epoch": 0.7760519627223949,
+      "grad_norm": 0.03162987902760506,
+      "learning_rate": 0.00016937559642128604,
+      "loss": 0.9649,
+      "step": 687
+    },
+    {
+      "epoch": 0.7771815871222818,
+      "grad_norm": 0.03206837549805641,
+      "learning_rate": 0.0001692900047110537,
+      "loss": 1.0174,
+      "step": 688
+    },
+    {
+      "epoch": 0.7783112115221689,
+      "grad_norm": 0.03194599226117134,
+      "learning_rate": 0.0001692043152503338,
+      "loss": 0.9872,
+      "step": 689
+    },
+    {
+      "epoch": 0.7794408359220559,
+      "grad_norm": 0.032261595129966736,
+      "learning_rate": 0.0001691185281600122,
+      "loss": 1.0046,
+      "step": 690
+    },
+    {
+      "epoch": 0.780570460321943,
+      "grad_norm": 0.032003022730350494,
+      "learning_rate": 0.00016903264356111258,
+      "loss": 1.0223,
+      "step": 691
+    },
+    {
+      "epoch": 0.78170008472183,
+      "grad_norm": 0.03204648569226265,
+      "learning_rate": 0.00016894666157479614,
+      "loss": 0.9402,
+      "step": 692
+    },
+    {
+      "epoch": 0.782829709121717,
+      "grad_norm": 0.03307194262742996,
+      "learning_rate": 0.00016886058232236156,
+      "loss": 0.977,
+      "step": 693
+    },
+    {
+      "epoch": 0.7839593335216041,
+      "grad_norm": 0.03305744007229805,
+      "learning_rate": 0.00016877440592524457,
+      "loss": 1.0158,
+      "step": 694
+    },
+    {
+      "epoch": 0.7850889579214911,
+      "grad_norm": 0.03376347944140434,
+      "learning_rate": 0.0001686881325050181,
+      "loss": 0.9266,
+      "step": 695
+    },
+    {
+      "epoch": 0.7862185823213781,
+      "grad_norm": 0.031977638602256775,
+      "learning_rate": 0.0001686017621833919,
+      "loss": 0.9966,
+      "step": 696
+    },
+    {
+      "epoch": 0.7873482067212652,
+      "grad_norm": 0.02983999252319336,
+      "learning_rate": 0.00016851529508221235,
+      "loss": 1.0418,
+      "step": 697
+    },
+    {
+      "epoch": 0.7884778311211522,
+      "grad_norm": 0.032275013625621796,
+      "learning_rate": 0.00016842873132346252,
+      "loss": 1.0745,
+      "step": 698
+    },
+    {
+      "epoch": 0.7896074555210393,
+      "grad_norm": 0.031730227172374725,
+      "learning_rate": 0.0001683420710292617,
+      "loss": 1.0505,
+      "step": 699
+    },
+    {
+      "epoch": 0.7907370799209263,
+      "grad_norm": 0.03272555023431778,
+      "learning_rate": 0.00016825531432186543,
+      "loss": 0.9826,
+      "step": 700
+    },
+    {
+      "epoch": 0.7918667043208133,
+      "grad_norm": 0.03270925581455231,
+      "learning_rate": 0.00016816846132366523,
+      "loss": 0.9954,
+      "step": 701
+    },
+    {
+      "epoch": 0.7929963287207004,
+      "grad_norm": 0.03216206654906273,
+      "learning_rate": 0.00016808151215718853,
+      "loss": 1.0266,
+      "step": 702
+    },
+    {
+      "epoch": 0.7941259531205874,
+      "grad_norm": 0.035289812833070755,
+      "learning_rate": 0.00016799446694509834,
+      "loss": 0.9776,
+      "step": 703
+    },
+    {
+      "epoch": 0.7952555775204745,
+      "grad_norm": 0.03199274092912674,
+      "learning_rate": 0.00016790732581019321,
+      "loss": 1.1088,
+      "step": 704
+    },
+    {
+      "epoch": 0.7963852019203614,
+      "grad_norm": 0.032748714089393616,
+      "learning_rate": 0.00016782008887540704,
+      "loss": 1.0957,
+      "step": 705
+    },
+    {
+      "epoch": 0.7975148263202485,
+      "grad_norm": 0.03328438848257065,
+      "learning_rate": 0.00016773275626380882,
+      "loss": 1.033,
+      "step": 706
+    },
+    {
+      "epoch": 0.7986444507201356,
+      "grad_norm": 0.031454868614673615,
+      "learning_rate": 0.00016764532809860255,
+      "loss": 0.9854,
+      "step": 707
+    },
+    {
+      "epoch": 0.7997740751200226,
+      "grad_norm": 0.032634053379297256,
+      "learning_rate": 0.00016755780450312705,
+      "loss": 0.9914,
+      "step": 708
+    },
+    {
+      "epoch": 0.8009036995199096,
+      "grad_norm": 0.03341520577669144,
+      "learning_rate": 0.00016747018560085572,
+      "loss": 0.9696,
+      "step": 709
+    },
+    {
+      "epoch": 0.8020333239197966,
+      "grad_norm": 0.03382014483213425,
+      "learning_rate": 0.00016738247151539643,
+      "loss": 1.0987,
+      "step": 710
+    },
+    {
+      "epoch": 0.8031629483196837,
+      "grad_norm": 0.03288474678993225,
+      "learning_rate": 0.00016729466237049137,
+      "loss": 1.0378,
+      "step": 711
+    },
+    {
+      "epoch": 0.8042925727195708,
+      "grad_norm": 0.03332820534706116,
+      "learning_rate": 0.00016720675829001675,
+      "loss": 1.0544,
+      "step": 712
+    },
+    {
+      "epoch": 0.8054221971194577,
+      "grad_norm": 0.033210329711437225,
+      "learning_rate": 0.0001671187593979828,
+      "loss": 0.9805,
+      "step": 713
+    },
+    {
+      "epoch": 0.8065518215193448,
+      "grad_norm": 0.03217016160488129,
+      "learning_rate": 0.00016703066581853345,
+      "loss": 1.0576,
+      "step": 714
+    },
+    {
+      "epoch": 0.8076814459192319,
+      "grad_norm": 0.03476932644844055,
+      "learning_rate": 0.00016694247767594624,
+      "loss": 1.0224,
+      "step": 715
+    },
+    {
+      "epoch": 0.8088110703191189,
+      "grad_norm": 0.03346959874033928,
+      "learning_rate": 0.00016685419509463213,
+      "loss": 1.0332,
+      "step": 716
+    },
+    {
+      "epoch": 0.809940694719006,
+      "grad_norm": 0.03398541361093521,
+      "learning_rate": 0.00016676581819913516,
+      "loss": 0.8649,
+      "step": 717
+    },
+    {
+      "epoch": 0.8110703191188929,
+      "grad_norm": 0.033196430653333664,
+      "learning_rate": 0.0001666773471141327,
+      "loss": 0.9434,
+      "step": 718
+    },
+    {
+      "epoch": 0.81219994351878,
+      "grad_norm": 0.03290561959147453,
+      "learning_rate": 0.00016658878196443476,
+      "loss": 0.9993,
+      "step": 719
+    },
+    {
+      "epoch": 0.8133295679186671,
+      "grad_norm": 0.03241690620779991,
+      "learning_rate": 0.00016650012287498412,
+      "loss": 1.0136,
+      "step": 720
+    },
+    {
+      "epoch": 0.8144591923185541,
+      "grad_norm": 0.03312429040670395,
+      "learning_rate": 0.00016641136997085608,
+      "loss": 1.0292,
+      "step": 721
+    },
+    {
+      "epoch": 0.8155888167184411,
+      "grad_norm": 0.03105839341878891,
+      "learning_rate": 0.0001663225233772584,
+      "loss": 0.9277,
+      "step": 722
+    },
+    {
+      "epoch": 0.8167184411183281,
+      "grad_norm": 0.0325755774974823,
+      "learning_rate": 0.00016623358321953078,
+      "loss": 1.0722,
+      "step": 723
+    },
+    {
+      "epoch": 0.8178480655182152,
+      "grad_norm": 0.033952005207538605,
+      "learning_rate": 0.00016614454962314516,
+      "loss": 1.0253,
+      "step": 724
+    },
+    {
+      "epoch": 0.8189776899181023,
+      "grad_norm": 0.0334470197558403,
+      "learning_rate": 0.00016605542271370513,
+      "loss": 1.0267,
+      "step": 725
+    },
+    {
+      "epoch": 0.8201073143179892,
+      "grad_norm": 0.03237008675932884,
+      "learning_rate": 0.00016596620261694604,
+      "loss": 1.0669,
+      "step": 726
+    },
+    {
+      "epoch": 0.8212369387178763,
+      "grad_norm": 0.03195658326148987,
+      "learning_rate": 0.00016587688945873458,
+      "loss": 0.9879,
+      "step": 727
+    },
+    {
+      "epoch": 0.8223665631177633,
+      "grad_norm": 0.03366916999220848,
+      "learning_rate": 0.0001657874833650688,
+      "loss": 0.9801,
+      "step": 728
+    },
+    {
+      "epoch": 0.8234961875176504,
+      "grad_norm": 0.03287327662110329,
+      "learning_rate": 0.0001656979844620779,
+      "loss": 0.9283,
+      "step": 729
+    },
+    {
+      "epoch": 0.8246258119175374,
+      "grad_norm": 0.03366275876760483,
+      "learning_rate": 0.00016560839287602192,
+      "loss": 1.0678,
+      "step": 730
+    },
+    {
+      "epoch": 0.8257554363174244,
+      "grad_norm": 0.03446445241570473,
+      "learning_rate": 0.00016551870873329167,
+      "loss": 0.9899,
+      "step": 731
+    },
+    {
+      "epoch": 0.8268850607173115,
+      "grad_norm": 0.03468972072005272,
+      "learning_rate": 0.0001654289321604086,
+      "loss": 1.0614,
+      "step": 732
+    },
+    {
+      "epoch": 0.8280146851171986,
+      "grad_norm": 0.03443734720349312,
+      "learning_rate": 0.00016533906328402448,
+      "loss": 1.0321,
+      "step": 733
+    },
+    {
+      "epoch": 0.8291443095170856,
+      "grad_norm": 0.03276367112994194,
+      "learning_rate": 0.0001652491022309213,
+      "loss": 0.9848,
+      "step": 734
+    },
+    {
+      "epoch": 0.8302739339169726,
+      "grad_norm": 0.03289159759879112,
+      "learning_rate": 0.00016515904912801118,
+      "loss": 1.0121,
+      "step": 735
+    },
+    {
+      "epoch": 0.8314035583168596,
+      "grad_norm": 0.034025318920612335,
+      "learning_rate": 0.000165068904102336,
+      "loss": 1.0589,
+      "step": 736
+    },
+    {
+      "epoch": 0.8325331827167467,
+      "grad_norm": 0.03421149030327797,
+      "learning_rate": 0.00016497866728106735,
+      "loss": 1.0138,
+      "step": 737
+    },
+    {
+      "epoch": 0.8336628071166338,
+      "grad_norm": 0.03334156796336174,
+      "learning_rate": 0.0001648883387915063,
+      "loss": 1.0337,
+      "step": 738
+    },
+    {
+      "epoch": 0.8347924315165207,
+      "grad_norm": 0.03213927149772644,
+      "learning_rate": 0.0001647979187610833,
+      "loss": 1.0248,
+      "step": 739
+    },
+    {
+      "epoch": 0.8359220559164078,
+      "grad_norm": 0.03407248482108116,
+      "learning_rate": 0.00016470740731735787,
+      "loss": 0.9995,
+      "step": 740
+    },
+    {
+      "epoch": 0.8370516803162948,
+      "grad_norm": 0.03234965354204178,
+      "learning_rate": 0.00016461680458801858,
+      "loss": 1.0526,
+      "step": 741
+    },
+    {
+      "epoch": 0.8381813047161819,
+      "grad_norm": 0.03325793519616127,
+      "learning_rate": 0.0001645261107008827,
+      "loss": 0.9461,
+      "step": 742
+    },
+    {
+      "epoch": 0.8393109291160689,
+      "grad_norm": 0.034206606447696686,
+      "learning_rate": 0.00016443532578389606,
+      "loss": 0.9095,
+      "step": 743
+    },
+    {
+      "epoch": 0.8404405535159559,
+      "grad_norm": 0.03346103057265282,
+      "learning_rate": 0.00016434444996513305,
+      "loss": 1.0337,
+      "step": 744
+    },
+    {
+      "epoch": 0.841570177915843,
+      "grad_norm": 0.03360540792346001,
+      "learning_rate": 0.0001642534833727962,
+      "loss": 0.9532,
+      "step": 745
+    },
+    {
+      "epoch": 0.84269980231573,
+      "grad_norm": 0.03263968229293823,
+      "learning_rate": 0.0001641624261352161,
+      "loss": 1.0579,
+      "step": 746
+    },
+    {
+      "epoch": 0.843829426715617,
+      "grad_norm": 0.033077508211135864,
+      "learning_rate": 0.0001640712783808513,
+      "loss": 0.9993,
+      "step": 747
+    },
+    {
+      "epoch": 0.8449590511155041,
+      "grad_norm": 0.03186168894171715,
+      "learning_rate": 0.00016398004023828797,
+      "loss": 0.9576,
+      "step": 748
+    },
+    {
+      "epoch": 0.8460886755153911,
+      "grad_norm": 0.032343216240406036,
+      "learning_rate": 0.00016388871183623977,
+      "loss": 1.0693,
+      "step": 749
+    },
+    {
+      "epoch": 0.8472182999152782,
+      "grad_norm": 0.03365077078342438,
+      "learning_rate": 0.00016379729330354774,
+      "loss": 0.9867,
+      "step": 750
+    },
+    {
+      "epoch": 0.8483479243151653,
+      "grad_norm": 0.03302355110645294,
+      "learning_rate": 0.00016370578476918008,
+      "loss": 1.002,
+      "step": 751
+    },
+    {
+      "epoch": 0.8494775487150522,
+      "grad_norm": 0.034719739109277725,
+      "learning_rate": 0.00016361418636223198,
+      "loss": 0.9621,
+      "step": 752
+    },
+    {
+      "epoch": 0.8506071731149393,
+      "grad_norm": 0.03225456923246384,
+      "learning_rate": 0.0001635224982119253,
+      "loss": 1.0285,
+      "step": 753
+    },
+    {
+      "epoch": 0.8517367975148263,
+      "grad_norm": 0.03379584476351738,
+      "learning_rate": 0.0001634307204476087,
+      "loss": 1.0787,
+      "step": 754
+    },
+    {
+      "epoch": 0.8528664219147134,
+      "grad_norm": 0.03374066203832626,
+      "learning_rate": 0.00016333885319875702,
+      "loss": 1.0322,
+      "step": 755
+    },
+    {
+      "epoch": 0.8539960463146004,
+      "grad_norm": 0.03438407927751541,
+      "learning_rate": 0.00016324689659497155,
+      "loss": 1.0204,
+      "step": 756
+    },
+    {
+      "epoch": 0.8551256707144874,
+      "grad_norm": 0.03300711140036583,
+      "learning_rate": 0.00016315485076597957,
+      "loss": 1.0088,
+      "step": 757
+    },
+    {
+      "epoch": 0.8562552951143745,
+      "grad_norm": 0.032439880073070526,
+      "learning_rate": 0.00016306271584163416,
+      "loss": 1.0198,
+      "step": 758
+    },
+    {
+      "epoch": 0.8573849195142615,
+      "grad_norm": 0.0341016985476017,
+      "learning_rate": 0.00016297049195191415,
+      "loss": 1.0242,
+      "step": 759
+    },
+    {
+      "epoch": 0.8585145439141485,
+      "grad_norm": 0.032917320728302,
+      "learning_rate": 0.00016287817922692395,
+      "loss": 1.0012,
+      "step": 760
+    },
+    {
+      "epoch": 0.8596441683140356,
+      "grad_norm": 0.03229722008109093,
+      "learning_rate": 0.00016278577779689314,
+      "loss": 0.9944,
+      "step": 761
+    },
+    {
+      "epoch": 0.8607737927139226,
+      "grad_norm": 0.0344838984310627,
+      "learning_rate": 0.0001626932877921766,
+      "loss": 0.9813,
+      "step": 762
+    },
+    {
+      "epoch": 0.8619034171138097,
+      "grad_norm": 0.033522870391607285,
+      "learning_rate": 0.00016260070934325402,
+      "loss": 1.0256,
+      "step": 763
+    },
+    {
+      "epoch": 0.8630330415136966,
+      "grad_norm": 0.03514671325683594,
+      "learning_rate": 0.00016250804258072997,
+      "loss": 0.9543,
+      "step": 764
+    },
+    {
+      "epoch": 0.8641626659135837,
+      "grad_norm": 0.03211130201816559,
+      "learning_rate": 0.00016241528763533353,
+      "loss": 1.0009,
+      "step": 765
+    },
+    {
+      "epoch": 0.8652922903134708,
+      "grad_norm": 0.033048368990421295,
+      "learning_rate": 0.00016232244463791826,
+      "loss": 1.0042,
+      "step": 766
+    },
+    {
+      "epoch": 0.8664219147133578,
+      "grad_norm": 0.031953100115060806,
+      "learning_rate": 0.00016222951371946192,
+      "loss": 1.0096,
+      "step": 767
+    },
+    {
+      "epoch": 0.8675515391132449,
+      "grad_norm": 0.03293442353606224,
+      "learning_rate": 0.00016213649501106622,
+      "loss": 0.9987,
+      "step": 768
+    },
+    {
+      "epoch": 0.8686811635131318,
+      "grad_norm": 0.033335424959659576,
+      "learning_rate": 0.00016204338864395684,
+      "loss": 1.0035,
+      "step": 769
+    },
+    {
+      "epoch": 0.8698107879130189,
+      "grad_norm": 0.04050195962190628,
+      "learning_rate": 0.00016195019474948299,
+      "loss": 1.0326,
+      "step": 770
+    },
+    {
+      "epoch": 0.870940412312906,
+      "grad_norm": 0.03311360627412796,
+      "learning_rate": 0.00016185691345911755,
+      "loss": 1.0184,
+      "step": 771
+    },
+    {
+      "epoch": 0.872070036712793,
+      "grad_norm": 0.03323720395565033,
+      "learning_rate": 0.0001617635449044565,
+      "loss": 0.9625,
+      "step": 772
+    },
+    {
+      "epoch": 0.87319966111268,
+      "grad_norm": 0.03422234579920769,
+      "learning_rate": 0.00016167008921721902,
+      "loss": 1.0654,
+      "step": 773
+    },
+    {
+      "epoch": 0.8743292855125671,
+      "grad_norm": 0.034163184463977814,
+      "learning_rate": 0.00016157654652924723,
+      "loss": 0.9953,
+      "step": 774
+    },
+    {
+      "epoch": 0.8754589099124541,
+      "grad_norm": 0.03320545703172684,
+      "learning_rate": 0.00016148291697250594,
+      "loss": 0.9766,
+      "step": 775
+    },
+    {
+      "epoch": 0.8765885343123412,
+      "grad_norm": 0.03346817195415497,
+      "learning_rate": 0.0001613892006790825,
+      "loss": 0.9201,
+      "step": 776
+    },
+    {
+      "epoch": 0.8777181587122281,
+      "grad_norm": 0.03284529596567154,
+      "learning_rate": 0.00016129539778118667,
+      "loss": 0.9284,
+      "step": 777
+    },
+    {
+      "epoch": 0.8788477831121152,
+      "grad_norm": 0.032990384846925735,
+      "learning_rate": 0.00016120150841115037,
+      "loss": 1.0058,
+      "step": 778
+    },
+    {
+      "epoch": 0.8799774075120023,
+      "grad_norm": 0.03396923094987869,
+      "learning_rate": 0.0001611075327014275,
+      "loss": 1.0831,
+      "step": 779
+    },
+    {
+      "epoch": 0.8811070319118893,
+      "grad_norm": 0.03224315121769905,
+      "learning_rate": 0.00016101347078459373,
+      "loss": 0.9318,
+      "step": 780
+    },
+    {
+      "epoch": 0.8822366563117763,
+      "grad_norm": 0.031812380999326706,
+      "learning_rate": 0.00016091932279334645,
+      "loss": 1.0566,
+      "step": 781
+    },
+    {
+      "epoch": 0.8833662807116633,
+      "grad_norm": 0.033773597329854965,
+      "learning_rate": 0.00016082508886050437,
+      "loss": 0.9349,
+      "step": 782
+    },
+    {
+      "epoch": 0.8844959051115504,
+      "grad_norm": 0.03375870734453201,
+      "learning_rate": 0.00016073076911900754,
+      "loss": 0.9875,
+      "step": 783
+    },
+    {
+      "epoch": 0.8856255295114375,
+      "grad_norm": 0.03443336486816406,
+      "learning_rate": 0.00016063636370191692,
+      "loss": 1.0604,
+      "step": 784
+    },
+    {
+      "epoch": 0.8867551539113245,
+      "grad_norm": 0.032187797129154205,
+      "learning_rate": 0.0001605418727424145,
+      "loss": 0.986,
+      "step": 785
+    },
+    {
+      "epoch": 0.8878847783112115,
+      "grad_norm": 0.03341427072882652,
+      "learning_rate": 0.00016044729637380284,
+      "loss": 0.9184,
+      "step": 786
+    },
+    {
+      "epoch": 0.8890144027110986,
+      "grad_norm": 0.03245866298675537,
+      "learning_rate": 0.000160352634729505,
+      "loss": 1.1511,
+      "step": 787
+    },
+    {
+      "epoch": 0.8901440271109856,
+      "grad_norm": 0.032569848001003265,
+      "learning_rate": 0.00016025788794306442,
+      "loss": 1.0948,
+      "step": 788
+    },
+    {
+      "epoch": 0.8912736515108727,
+      "grad_norm": 0.03429558128118515,
+      "learning_rate": 0.0001601630561481446,
+      "loss": 0.9379,
+      "step": 789
+    },
+    {
+      "epoch": 0.8924032759107596,
+      "grad_norm": 0.033720601350069046,
+      "learning_rate": 0.00016006813947852893,
+      "loss": 0.9845,
+      "step": 790
+    },
+    {
+      "epoch": 0.8935329003106467,
+      "grad_norm": 0.033525846898555756,
+      "learning_rate": 0.00015997313806812057,
+      "loss": 1.0279,
+      "step": 791
+    },
+    {
+      "epoch": 0.8946625247105338,
+      "grad_norm": 0.03577594459056854,
+      "learning_rate": 0.00015987805205094227,
+      "loss": 0.9654,
+      "step": 792
+    },
+    {
+      "epoch": 0.8957921491104208,
+      "grad_norm": 0.03572090342640877,
+      "learning_rate": 0.00015978288156113604,
+      "loss": 1.0292,
+      "step": 793
+    },
+    {
+      "epoch": 0.8969217735103078,
+      "grad_norm": 0.0330742709338665,
+      "learning_rate": 0.00015968762673296318,
+      "loss": 1.0898,
+      "step": 794
+    },
+    {
+      "epoch": 0.8980513979101948,
+      "grad_norm": 0.03374762088060379,
+      "learning_rate": 0.0001595922877008039,
+      "loss": 1.0368,
+      "step": 795
+    },
+    {
+      "epoch": 0.8991810223100819,
+      "grad_norm": 0.035000476986169815,
+      "learning_rate": 0.00015949686459915715,
+      "loss": 1.0531,
+      "step": 796
+    },
+    {
+      "epoch": 0.900310646709969,
+      "grad_norm": 0.03325015306472778,
+      "learning_rate": 0.00015940135756264062,
+      "loss": 1.0199,
+      "step": 797
+    },
+    {
+      "epoch": 0.901440271109856,
+      "grad_norm": 0.03547768294811249,
+      "learning_rate": 0.0001593057667259902,
+      "loss": 0.9988,
+      "step": 798
+    },
+    {
+      "epoch": 0.902569895509743,
+      "grad_norm": 0.03294992819428444,
+      "learning_rate": 0.0001592100922240603,
+      "loss": 0.9943,
+      "step": 799
+    },
+    {
+      "epoch": 0.90369951990963,
+      "grad_norm": 0.03369821235537529,
+      "learning_rate": 0.00015911433419182305,
+      "loss": 1.0186,
+      "step": 800
+    },
+    {
+      "epoch": 0.9048291443095171,
+      "grad_norm": 0.03317281976342201,
+      "learning_rate": 0.00015901849276436862,
+      "loss": 0.9601,
+      "step": 801
+    },
+    {
+      "epoch": 0.9059587687094042,
+      "grad_norm": 0.036661259829998016,
+      "learning_rate": 0.00015892256807690478,
+      "loss": 1.0847,
+      "step": 802
+    },
+    {
+      "epoch": 0.9070883931092911,
+      "grad_norm": 0.0334974080324173,
+      "learning_rate": 0.00015882656026475672,
+      "loss": 1.0264,
+      "step": 803
+    },
+    {
+      "epoch": 0.9082180175091782,
+      "grad_norm": 0.03364727646112442,
+      "learning_rate": 0.00015873046946336694,
+      "loss": 0.9768,
+      "step": 804
+    },
+    {
+      "epoch": 0.9093476419090653,
+      "grad_norm": 0.03534623235464096,
+      "learning_rate": 0.000158634295808295,
+      "loss": 1.0705,
+      "step": 805
+    },
+    {
+      "epoch": 0.9104772663089523,
+      "grad_norm": 0.032764844596385956,
+      "learning_rate": 0.00015853803943521733,
+      "loss": 0.9543,
+      "step": 806
+    },
+    {
+      "epoch": 0.9116068907088393,
+      "grad_norm": 0.03310185670852661,
+      "learning_rate": 0.00015844170047992712,
+      "loss": 1.0077,
+      "step": 807
+    },
+    {
+      "epoch": 0.9127365151087263,
+      "grad_norm": 0.0327795036137104,
+      "learning_rate": 0.00015834527907833396,
+      "loss": 0.9765,
+      "step": 808
+    },
+    {
+      "epoch": 0.9138661395086134,
+      "grad_norm": 0.03351445123553276,
+      "learning_rate": 0.00015824877536646382,
+      "loss": 1.0634,
+      "step": 809
+    },
+    {
+      "epoch": 0.9149957639085005,
+      "grad_norm": 0.03497536852955818,
+      "learning_rate": 0.00015815218948045878,
+      "loss": 0.9211,
+      "step": 810
+    },
+    {
+      "epoch": 0.9161253883083874,
+      "grad_norm": 0.03262564167380333,
+      "learning_rate": 0.00015805552155657683,
+      "loss": 0.9841,
+      "step": 811
+    },
+    {
+      "epoch": 0.9172550127082745,
+      "grad_norm": 0.03305838629603386,
+      "learning_rate": 0.00015795877173119176,
+      "loss": 0.9968,
+      "step": 812
+    },
+    {
+      "epoch": 0.9183846371081615,
+      "grad_norm": 0.03393985703587532,
+      "learning_rate": 0.00015786194014079274,
+      "loss": 1.0257,
+      "step": 813
+    },
+    {
+      "epoch": 0.9195142615080486,
+      "grad_norm": 0.03377285972237587,
+      "learning_rate": 0.00015776502692198448,
+      "loss": 0.979,
+      "step": 814
+    },
+    {
+      "epoch": 0.9206438859079357,
+      "grad_norm": 0.03390325978398323,
+      "learning_rate": 0.00015766803221148673,
+      "loss": 1.0935,
+      "step": 815
+    },
+    {
+      "epoch": 0.9217735103078226,
+      "grad_norm": 0.034586288034915924,
+      "learning_rate": 0.00015757095614613427,
+      "loss": 1.0286,
+      "step": 816
+    },
+    {
+      "epoch": 0.9229031347077097,
+      "grad_norm": 0.034462425857782364,
+      "learning_rate": 0.00015747379886287655,
+      "loss": 0.9826,
+      "step": 817
+    },
+    {
+      "epoch": 0.9240327591075967,
+      "grad_norm": 0.03412788733839989,
+      "learning_rate": 0.0001573765604987777,
+      "loss": 1.0391,
+      "step": 818
+    },
+    {
+      "epoch": 0.9251623835074838,
+      "grad_norm": 0.03411950543522835,
+      "learning_rate": 0.0001572792411910162,
+      "loss": 1.014,
+      "step": 819
+    },
+    {
+      "epoch": 0.9262920079073708,
+      "grad_norm": 0.03366335481405258,
+      "learning_rate": 0.0001571818410768848,
+      "loss": 1.0191,
+      "step": 820
+    },
+    {
+      "epoch": 0.9274216323072578,
+      "grad_norm": 0.033515483140945435,
+      "learning_rate": 0.00015708436029379004,
+      "loss": 1.0072,
+      "step": 821
+    },
+    {
+      "epoch": 0.9285512567071449,
+      "grad_norm": 0.033421795815229416,
+      "learning_rate": 0.0001569867989792525,
+      "loss": 1.0311,
+      "step": 822
+    },
+    {
+      "epoch": 0.929680881107032,
+      "grad_norm": 0.032961517572402954,
+      "learning_rate": 0.00015688915727090613,
+      "loss": 1.0476,
+      "step": 823
+    },
+    {
+      "epoch": 0.9308105055069189,
+      "grad_norm": 0.03382313251495361,
+      "learning_rate": 0.00015679143530649854,
+      "loss": 0.9863,
+      "step": 824
+    },
+    {
+      "epoch": 0.931940129906806,
+      "grad_norm": 0.03453601896762848,
+      "learning_rate": 0.0001566936332238904,
+      "loss": 0.981,
+      "step": 825
+    },
+    {
+      "epoch": 0.933069754306693,
+      "grad_norm": 0.03426108881831169,
+      "learning_rate": 0.00015659575116105544,
+      "loss": 1.0615,
+      "step": 826
+    },
+    {
+      "epoch": 0.9341993787065801,
+      "grad_norm": 0.03343765065073967,
+      "learning_rate": 0.0001564977892560803,
+      "loss": 1.0745,
+      "step": 827
+    },
+    {
+      "epoch": 0.935329003106467,
+      "grad_norm": 0.03495456650853157,
+      "learning_rate": 0.00015639974764716414,
+      "loss": 0.9985,
+      "step": 828
+    },
+    {
+      "epoch": 0.9364586275063541,
+      "grad_norm": 0.033679116517305374,
+      "learning_rate": 0.0001563016264726186,
+      "loss": 1.0216,
+      "step": 829
+    },
+    {
+      "epoch": 0.9375882519062412,
+      "grad_norm": 0.03362250700592995,
+      "learning_rate": 0.0001562034258708676,
+      "loss": 1.0337,
+      "step": 830
+    },
+    {
+      "epoch": 0.9387178763061282,
+      "grad_norm": 0.034377310425043106,
+      "learning_rate": 0.00015610514598044707,
+      "loss": 1.0583,
+      "step": 831
+    },
+    {
+      "epoch": 0.9398475007060153,
+      "grad_norm": 0.033647313714027405,
+      "learning_rate": 0.00015600678694000487,
+      "loss": 1.0126,
+      "step": 832
+    },
+    {
+      "epoch": 0.9409771251059023,
+      "grad_norm": 0.03457539901137352,
+      "learning_rate": 0.0001559083488883004,
+      "loss": 1.1528,
+      "step": 833
+    },
+    {
+      "epoch": 0.9421067495057893,
+      "grad_norm": 0.03426367789506912,
+      "learning_rate": 0.00015580983196420464,
+      "loss": 0.9055,
+      "step": 834
+    },
+    {
+      "epoch": 0.9432363739056764,
+      "grad_norm": 0.03347745165228844,
+      "learning_rate": 0.0001557112363066998,
+      "loss": 0.9978,
+      "step": 835
+    },
+    {
+      "epoch": 0.9443659983055634,
+      "grad_norm": 0.03355059772729874,
+      "learning_rate": 0.00015561256205487908,
+      "loss": 0.9844,
+      "step": 836
+    },
+    {
+      "epoch": 0.9454956227054504,
+      "grad_norm": 0.032837532460689545,
+      "learning_rate": 0.0001555138093479467,
+      "loss": 0.932,
+      "step": 837
+    },
+    {
+      "epoch": 0.9466252471053375,
+      "grad_norm": 0.03441225364804268,
+      "learning_rate": 0.0001554149783252175,
+      "loss": 0.9975,
+      "step": 838
+    },
+    {
+      "epoch": 0.9477548715052245,
+      "grad_norm": 0.033451907336711884,
+      "learning_rate": 0.00015531606912611674,
+      "loss": 0.9707,
+      "step": 839
+    },
+    {
+      "epoch": 0.9488844959051116,
+      "grad_norm": 0.03538847342133522,
+      "learning_rate": 0.00015521708189018005,
+      "loss": 1.0129,
+      "step": 840
+    },
+    {
+      "epoch": 0.9500141203049985,
+      "grad_norm": 0.033600080758333206,
+      "learning_rate": 0.00015511801675705312,
+      "loss": 1.0403,
+      "step": 841
+    },
+    {
+      "epoch": 0.9511437447048856,
+      "grad_norm": 0.03426308557391167,
+      "learning_rate": 0.00015501887386649155,
+      "loss": 0.9879,
+      "step": 842
+    },
+    {
+      "epoch": 0.9522733691047727,
+      "grad_norm": 0.033120229840278625,
+      "learning_rate": 0.00015491965335836055,
+      "loss": 1.0627,
+      "step": 843
+    },
+    {
+      "epoch": 0.9534029935046597,
+      "grad_norm": 0.0343567430973053,
+      "learning_rate": 0.00015482035537263498,
+      "loss": 1.0308,
+      "step": 844
+    },
+    {
+      "epoch": 0.9545326179045467,
+      "grad_norm": 0.033301327377557755,
+      "learning_rate": 0.00015472098004939888,
+      "loss": 1.0106,
+      "step": 845
+    },
+    {
+      "epoch": 0.9556622423044338,
+      "grad_norm": 0.03342900052666664,
+      "learning_rate": 0.00015462152752884544,
+      "loss": 1.0261,
+      "step": 846
+    },
+    {
+      "epoch": 0.9567918667043208,
+      "grad_norm": 0.032714009284973145,
+      "learning_rate": 0.00015452199795127678,
+      "loss": 0.8953,
+      "step": 847
+    },
+    {
+      "epoch": 0.9579214911042079,
+      "grad_norm": 0.0333135612308979,
+      "learning_rate": 0.00015442239145710364,
+      "loss": 1.0105,
+      "step": 848
+    },
+    {
+      "epoch": 0.9590511155040949,
+      "grad_norm": 0.03534407541155815,
+      "learning_rate": 0.00015432270818684532,
+      "loss": 0.9325,
+      "step": 849
+    },
+    {
+      "epoch": 0.9601807399039819,
+      "grad_norm": 0.03319082036614418,
+      "learning_rate": 0.00015422294828112954,
+      "loss": 0.9187,
+      "step": 850
+    },
+    {
+      "epoch": 0.961310364303869,
+      "grad_norm": 0.03402223438024521,
+      "learning_rate": 0.00015412311188069193,
+      "loss": 0.9523,
+      "step": 851
+    },
+    {
+      "epoch": 0.962439988703756,
+      "grad_norm": 0.038419678807258606,
+      "learning_rate": 0.00015402319912637613,
+      "loss": 1.0135,
+      "step": 852
+    },
+    {
+      "epoch": 0.9635696131036431,
+      "grad_norm": 0.03462392836809158,
+      "learning_rate": 0.00015392321015913357,
+      "loss": 1.0811,
+      "step": 853
+    },
+    {
+      "epoch": 0.96469923750353,
+      "grad_norm": 0.033567875623703,
+      "learning_rate": 0.0001538231451200231,
+      "loss": 1.0052,
+      "step": 854
+    },
+    {
+      "epoch": 0.9658288619034171,
+      "grad_norm": 0.03398734703660011,
+      "learning_rate": 0.00015372300415021091,
+      "loss": 0.9939,
+      "step": 855
+    },
+    {
+      "epoch": 0.9669584863033042,
+      "grad_norm": 0.03315124288201332,
+      "learning_rate": 0.00015362278739097026,
+      "loss": 1.0515,
+      "step": 856
+    },
+    {
+      "epoch": 0.9680881107031912,
+      "grad_norm": 0.03387816995382309,
+      "learning_rate": 0.0001535224949836815,
+      "loss": 1.0906,
+      "step": 857
+    },
+    {
+      "epoch": 0.9692177351030782,
+      "grad_norm": 0.033208638429641724,
+      "learning_rate": 0.00015342212706983153,
+      "loss": 0.9542,
+      "step": 858
+    },
+    {
+      "epoch": 0.9703473595029652,
+      "grad_norm": 0.0338163860142231,
+      "learning_rate": 0.00015332168379101377,
+      "loss": 0.9892,
+      "step": 859
+    },
+    {
+      "epoch": 0.9714769839028523,
+      "grad_norm": 0.033496033400297165,
+      "learning_rate": 0.00015322116528892807,
+      "loss": 1.0253,
+      "step": 860
+    },
+    {
+      "epoch": 0.9726066083027394,
+      "grad_norm": 0.034597545862197876,
+      "learning_rate": 0.00015312057170538035,
+      "loss": 1.0102,
+      "step": 861
+    },
+    {
+      "epoch": 0.9737362327026263,
+      "grad_norm": 0.03476065397262573,
+      "learning_rate": 0.00015301990318228244,
+      "loss": 0.938,
+      "step": 862
+    },
+    {
+      "epoch": 0.9748658571025134,
+      "grad_norm": 0.036271460354328156,
+      "learning_rate": 0.00015291915986165186,
+      "loss": 0.9072,
+      "step": 863
+    },
+    {
+      "epoch": 0.9759954815024005,
+      "grad_norm": 0.032739460468292236,
+      "learning_rate": 0.00015281834188561174,
+      "loss": 0.9955,
+      "step": 864
+    },
+    {
+      "epoch": 0.9771251059022875,
+      "grad_norm": 0.03603595495223999,
+      "learning_rate": 0.0001527174493963905,
+      "loss": 0.978,
+      "step": 865
+    },
+    {
+      "epoch": 0.9782547303021746,
+      "grad_norm": 0.03469686582684517,
+      "learning_rate": 0.00015261648253632156,
+      "loss": 1.0928,
+      "step": 866
+    },
+    {
+      "epoch": 0.9793843547020615,
+      "grad_norm": 0.03487220034003258,
+      "learning_rate": 0.0001525154414478434,
+      "loss": 1.0144,
+      "step": 867
+    },
+    {
+      "epoch": 0.9805139791019486,
+      "grad_norm": 0.03308931365609169,
+      "learning_rate": 0.00015241432627349918,
+      "loss": 0.9912,
+      "step": 868
+    },
+    {
+      "epoch": 0.9816436035018357,
+      "grad_norm": 0.0350349023938179,
+      "learning_rate": 0.00015231313715593662,
+      "loss": 1.0209,
+      "step": 869
+    },
+    {
+      "epoch": 0.9827732279017227,
+      "grad_norm": 0.034897249191999435,
+      "learning_rate": 0.0001522118742379076,
+      "loss": 0.9873,
+      "step": 870
+    },
+    {
+      "epoch": 0.9839028523016097,
+      "grad_norm": 0.03427942842245102,
+      "learning_rate": 0.00015211053766226828,
+      "loss": 0.9497,
+      "step": 871
+    },
+    {
+      "epoch": 0.9850324767014967,
+      "grad_norm": 0.0339798741042614,
+      "learning_rate": 0.00015200912757197868,
+      "loss": 0.9741,
+      "step": 872
+    },
+    {
+      "epoch": 0.9861621011013838,
+      "grad_norm": 0.03557536378502846,
+      "learning_rate": 0.00015190764411010247,
+      "loss": 0.9747,
+      "step": 873
+    },
+    {
+      "epoch": 0.9872917255012709,
+      "grad_norm": 0.036786146461963654,
+      "learning_rate": 0.00015180608741980692,
+      "loss": 1.0296,
+      "step": 874
+    },
+    {
+      "epoch": 0.9884213499011578,
+      "grad_norm": 0.03306087478995323,
+      "learning_rate": 0.00015170445764436252,
+      "loss": 1.0559,
+      "step": 875
+    },
+    {
+      "epoch": 0.9895509743010449,
+      "grad_norm": 0.03436678647994995,
+      "learning_rate": 0.00015160275492714296,
+      "loss": 0.9572,
+      "step": 876
+    },
+    {
+      "epoch": 0.990680598700932,
+      "grad_norm": 0.03426647186279297,
+      "learning_rate": 0.00015150097941162474,
+      "loss": 0.999,
+      "step": 877
+    },
+    {
+      "epoch": 0.991810223100819,
+      "grad_norm": 0.03366367891430855,
+      "learning_rate": 0.00015139913124138715,
+      "loss": 1.0365,
+      "step": 878
+    },
+    {
+      "epoch": 0.992939847500706,
+      "grad_norm": 0.034058500081300735,
+      "learning_rate": 0.00015129721056011185,
+      "loss": 0.9835,
+      "step": 879
+    },
+    {
+      "epoch": 0.994069471900593,
+      "grad_norm": 0.03479884937405586,
+      "learning_rate": 0.00015119521751158296,
+      "loss": 1.0604,
+      "step": 880
+    },
+    {
+      "epoch": 0.9951990963004801,
+      "grad_norm": 0.03426951542496681,
+      "learning_rate": 0.00015109315223968655,
+      "loss": 1.0344,
+      "step": 881
+    },
+    {
+      "epoch": 0.9963287207003672,
+      "grad_norm": 0.034726936370134354,
+      "learning_rate": 0.0001509910148884106,
+      "loss": 0.927,
+      "step": 882
+    },
+    {
+      "epoch": 0.9974583451002542,
+      "grad_norm": 0.03522869199514389,
+      "learning_rate": 0.00015088880560184493,
+      "loss": 1.035,
+      "step": 883
+    },
+    {
+      "epoch": 0.9985879695001412,
+      "grad_norm": 0.03507549315690994,
+      "learning_rate": 0.00015078652452418063,
+      "loss": 0.952,
+      "step": 884
+    },
+    {
+      "epoch": 0.9997175939000282,
+      "grad_norm": 0.03401617333292961,
+      "learning_rate": 0.00015068417179971014,
+      "loss": 1.0006,
+      "step": 885
+    },
+    {
+      "epoch": 0.9997175939000282,
+      "eval_loss": 1.0020042657852173,
+      "eval_runtime": 552.3244,
+      "eval_samples_per_second": 17.712,
+      "eval_steps_per_second": 8.857,
+      "step": 885
+    },
+    {
+      "epoch": 1.0008472182999153,
+      "grad_norm": 0.03515418618917465,
+      "learning_rate": 0.00015058174757282705,
+      "loss": 0.9925,
+      "step": 886
+    },
+    {
+      "epoch": 1.0019768426998024,
+      "grad_norm": 0.03516097739338875,
+      "learning_rate": 0.00015047925198802574,
+      "loss": 0.9803,
+      "step": 887
+    },
+    {
+      "epoch": 1.0008472182999153,
+      "grad_norm": 0.03710052743554115,
+      "learning_rate": 0.0001503766851899013,
+      "loss": 0.9624,
+      "step": 888
+    },
+    {
+      "epoch": 1.0019768426998024,
+      "grad_norm": 0.03626672551035881,
+      "learning_rate": 0.00015027404732314922,
+      "loss": 1.0015,
+      "step": 889
+    },
+    {
+      "epoch": 1.0031064670996894,
+      "grad_norm": 0.03808250650763512,
+      "learning_rate": 0.00015017133853256537,
+      "loss": 0.935,
+      "step": 890
+    },
+    {
+      "epoch": 1.0042360914995765,
+      "grad_norm": 0.035625848919153214,
+      "learning_rate": 0.00015006855896304558,
+      "loss": 0.9196,
+      "step": 891
+    },
+    {
+      "epoch": 1.0053657158994633,
+      "grad_norm": 0.03489432483911514,
+      "learning_rate": 0.00014996570875958553,
+      "loss": 0.9544,
+      "step": 892
+    },
+    {
+      "epoch": 1.0064953402993504,
+      "grad_norm": 0.034372955560684204,
+      "learning_rate": 0.0001498627880672806,
+      "loss": 1.011,
+      "step": 893
+    },
+    {
+      "epoch": 1.0076249646992375,
+      "grad_norm": 0.034961700439453125,
+      "learning_rate": 0.00014975979703132556,
+      "loss": 0.9769,
+      "step": 894
+    },
+    {
+      "epoch": 1.0087545890991245,
+      "grad_norm": 0.03485623002052307,
+      "learning_rate": 0.00014965673579701445,
+      "loss": 0.9856,
+      "step": 895
+    },
+    {
+      "epoch": 1.0098842134990116,
+      "grad_norm": 0.03519703075289726,
+      "learning_rate": 0.0001495536045097403,
+      "loss": 0.9178,
+      "step": 896
+    },
+    {
+      "epoch": 1.0110138378988986,
+      "grad_norm": 0.03748798742890358,
+      "learning_rate": 0.00014945040331499504,
+      "loss": 1.0493,
+      "step": 897
+    },
+    {
+      "epoch": 1.0121434622987857,
+      "grad_norm": 0.03519289940595627,
+      "learning_rate": 0.0001493471323583692,
+      "loss": 0.9539,
+      "step": 898
+    },
+    {
+      "epoch": 1.0132730866986728,
+      "grad_norm": 0.035702627152204514,
+      "learning_rate": 0.00014924379178555167,
+      "loss": 1.016,
+      "step": 899
+    },
+    {
+      "epoch": 1.0144027110985596,
+      "grad_norm": 0.03649812936782837,
+      "learning_rate": 0.00014914038174232956,
+      "loss": 1.0289,
+      "step": 900
+    },
+    {
+      "epoch": 1.0155323354984467,
+      "grad_norm": 0.03464585542678833,
+      "learning_rate": 0.00014903690237458802,
+      "loss": 1.0565,
+      "step": 901
+    },
+    {
+      "epoch": 1.0166619598983337,
+      "grad_norm": 0.035987745970487595,
+      "learning_rate": 0.00014893335382831007,
+      "loss": 1.0472,
+      "step": 902
+    },
+    {
+      "epoch": 1.0177915842982208,
+      "grad_norm": 0.035208627581596375,
+      "learning_rate": 0.00014882973624957615,
+      "loss": 1.0795,
+      "step": 903
+    },
+    {
+      "epoch": 1.0189212086981079,
+      "grad_norm": 0.035057902336120605,
+      "learning_rate": 0.0001487260497845642,
+      "loss": 1.0426,
+      "step": 904
+    },
+    {
+      "epoch": 1.020050833097995,
+      "grad_norm": 0.035428449511528015,
+      "learning_rate": 0.0001486222945795494,
+      "loss": 0.9672,
+      "step": 905
+    },
+    {
+      "epoch": 1.021180457497882,
+      "grad_norm": 0.034387003630399704,
+      "learning_rate": 0.0001485184707809037,
+      "loss": 0.9816,
+      "step": 906
+    },
+    {
+      "epoch": 1.022310081897769,
+      "grad_norm": 0.03481113165616989,
+      "learning_rate": 0.00014841457853509606,
+      "loss": 0.998,
+      "step": 907
+    },
+    {
+      "epoch": 1.0234397062976561,
+      "grad_norm": 0.03739047050476074,
+      "learning_rate": 0.00014831061798869182,
+      "loss": 1.0395,
+      "step": 908
+    },
+    {
+      "epoch": 1.024569330697543,
+      "grad_norm": 0.03472413867712021,
+      "learning_rate": 0.00014820658928835277,
+      "loss": 1.0398,
+      "step": 909
+    },
+    {
+      "epoch": 1.02569895509743,
+      "grad_norm": 0.03586706519126892,
+      "learning_rate": 0.00014810249258083677,
+      "loss": 1.0135,
+      "step": 910
+    },
+    {
+      "epoch": 1.026828579497317,
+      "grad_norm": 0.037143245339393616,
+      "learning_rate": 0.00014799832801299775,
+      "loss": 0.965,
+      "step": 911
+    },
+    {
+      "epoch": 1.0279582038972042,
+      "grad_norm": 0.03444663807749748,
+      "learning_rate": 0.00014789409573178521,
+      "loss": 1.001,
+      "step": 912
+    },
+    {
+      "epoch": 1.0290878282970912,
+      "grad_norm": 0.0362403504550457,
+      "learning_rate": 0.00014778979588424428,
+      "loss": 1.03,
+      "step": 913
+    },
+    {
+      "epoch": 1.0302174526969783,
+      "grad_norm": 0.03418262302875519,
+      "learning_rate": 0.0001476854286175155,
+      "loss": 0.9705,
+      "step": 914
+    },
+    {
+      "epoch": 1.0313470770968654,
+      "grad_norm": 0.0341293029487133,
+      "learning_rate": 0.00014758099407883422,
+      "loss": 1.0715,
+      "step": 915
+    },
+    {
+      "epoch": 1.0324767014967524,
+      "grad_norm": 0.03373926132917404,
+      "learning_rate": 0.00014747649241553102,
+      "loss": 0.9609,
+      "step": 916
+    },
+    {
+      "epoch": 1.0336063258966393,
+      "grad_norm": 0.0360686220228672,
+      "learning_rate": 0.00014737192377503098,
+      "loss": 1.0252,
+      "step": 917
+    },
+    {
+      "epoch": 1.0347359502965263,
+      "grad_norm": 0.036107342690229416,
+      "learning_rate": 0.00014726728830485376,
+      "loss": 0.8894,
+      "step": 918
+    },
+    {
+      "epoch": 1.0358655746964134,
+      "grad_norm": 0.03535093739628792,
+      "learning_rate": 0.00014716258615261323,
+      "loss": 0.9827,
+      "step": 919
+    },
+    {
+      "epoch": 1.0369951990963004,
+      "grad_norm": 0.03566374629735947,
+      "learning_rate": 0.0001470578174660174,
+      "loss": 0.9728,
+      "step": 920
+    },
+    {
+      "epoch": 1.0381248234961875,
+      "grad_norm": 0.035979773849248886,
+      "learning_rate": 0.0001469529823928681,
+      "loss": 0.9051,
+      "step": 921
+    },
+    {
+      "epoch": 1.0392544478960746,
+      "grad_norm": 0.03736504539847374,
+      "learning_rate": 0.0001468480810810608,
+      "loss": 1.0889,
+      "step": 922
+    },
+    {
+      "epoch": 1.0403840722959616,
+      "grad_norm": 0.0376417301595211,
+      "learning_rate": 0.0001467431136785845,
+      "loss": 1.0897,
+      "step": 923
+    },
+    {
+      "epoch": 1.0415136966958487,
+      "grad_norm": 0.034662578254938126,
+      "learning_rate": 0.00014663808033352132,
+      "loss": 0.9442,
+      "step": 924
+    },
+    {
+      "epoch": 1.0426433210957358,
+      "grad_norm": 0.03605816140770912,
+      "learning_rate": 0.00014653298119404645,
+      "loss": 0.9179,
+      "step": 925
+    },
+    {
+      "epoch": 1.0437729454956226,
+      "grad_norm": 0.03609084710478783,
+      "learning_rate": 0.000146427816408428,
+      "loss": 0.9713,
+      "step": 926
+    },
+    {
+      "epoch": 1.0449025698955097,
+      "grad_norm": 0.036889102309942245,
+      "learning_rate": 0.0001463225861250265,
+      "loss": 1.049,
+      "step": 927
+    },
+    {
+      "epoch": 1.0460321942953967,
+      "grad_norm": 0.035482168197631836,
+      "learning_rate": 0.00014621729049229507,
+      "loss": 1.0385,
+      "step": 928
+    },
+    {
+      "epoch": 1.0471618186952838,
+      "grad_norm": 0.035426728427410126,
+      "learning_rate": 0.00014611192965877892,
+      "loss": 0.9582,
+      "step": 929
+    },
+    {
+      "epoch": 1.0482914430951709,
+      "grad_norm": 0.037651337683200836,
+      "learning_rate": 0.00014600650377311522,
+      "loss": 0.9866,
+      "step": 930
+    },
+    {
+      "epoch": 1.049421067495058,
+      "grad_norm": 0.037454381585121155,
+      "learning_rate": 0.00014590101298403297,
+      "loss": 1.0368,
+      "step": 931
+    },
+    {
+      "epoch": 1.050550691894945,
+      "grad_norm": 0.037140969187021255,
+      "learning_rate": 0.0001457954574403527,
+      "loss": 1.0504,
+      "step": 932
+    },
+    {
+      "epoch": 1.051680316294832,
+      "grad_norm": 0.037180185317993164,
+      "learning_rate": 0.0001456898372909864,
+      "loss": 0.9409,
+      "step": 933
+    },
+    {
+      "epoch": 1.052809940694719,
+      "grad_norm": 0.03617197647690773,
+      "learning_rate": 0.00014558415268493694,
+      "loss": 1.0344,
+      "step": 934
+    },
+    {
+      "epoch": 1.053939565094606,
+      "grad_norm": 0.037174291908741,
+      "learning_rate": 0.00014547840377129842,
+      "loss": 1.0079,
+      "step": 935
+    },
+    {
+      "epoch": 1.055069189494493,
+      "grad_norm": 0.03483375906944275,
+      "learning_rate": 0.0001453725906992555,
+      "loss": 0.961,
+      "step": 936
+    },
+    {
+      "epoch": 1.05619881389438,
+      "grad_norm": 0.03393753245472908,
+      "learning_rate": 0.00014526671361808331,
+      "loss": 0.9776,
+      "step": 937
+    },
+    {
+      "epoch": 1.0573284382942671,
+      "grad_norm": 0.03698374330997467,
+      "learning_rate": 0.00014516077267714744,
+      "loss": 1.0631,
+      "step": 938
+    },
+    {
+      "epoch": 1.0584580626941542,
+      "grad_norm": 0.03589322045445442,
+      "learning_rate": 0.00014505476802590344,
+      "loss": 1.0516,
+      "step": 939
+    },
+    {
+      "epoch": 1.0595876870940413,
+      "grad_norm": 0.03666504845023155,
+      "learning_rate": 0.0001449486998138968,
+      "loss": 1.0051,
+      "step": 940
+    },
+    {
+      "epoch": 1.0607173114939283,
+      "grad_norm": 0.03527549281716347,
+      "learning_rate": 0.00014484256819076265,
+      "loss": 0.9629,
+      "step": 941
+    },
+    {
+      "epoch": 1.0618469358938154,
+      "grad_norm": 0.03527344390749931,
+      "learning_rate": 0.00014473637330622558,
+      "loss": 1.0052,
+      "step": 942
+    },
+    {
+      "epoch": 1.0629765602937022,
+      "grad_norm": 0.03708739951252937,
+      "learning_rate": 0.00014463011531009942,
+      "loss": 0.9415,
+      "step": 943
+    },
+    {
+      "epoch": 1.0641061846935893,
+      "grad_norm": 0.03615740314126015,
+      "learning_rate": 0.0001445237943522871,
+      "loss": 1.0204,
+      "step": 944
+    },
+    {
+      "epoch": 1.0652358090934764,
+      "grad_norm": 0.036477502435445786,
+      "learning_rate": 0.00014441741058278024,
+      "loss": 0.9572,
+      "step": 945
+    },
+    {
+      "epoch": 1.0663654334933634,
+      "grad_norm": 0.03599262610077858,
+      "learning_rate": 0.00014431096415165918,
+      "loss": 0.9827,
+      "step": 946
+    },
+    {
+      "epoch": 1.0674950578932505,
+      "grad_norm": 0.03469396010041237,
+      "learning_rate": 0.00014420445520909266,
+      "loss": 0.9757,
+      "step": 947
+    },
+    {
+      "epoch": 1.0686246822931376,
+      "grad_norm": 0.03587264567613602,
+      "learning_rate": 0.00014409788390533753,
+      "loss": 0.8875,
+      "step": 948
+    },
+    {
+      "epoch": 1.0697543066930246,
+      "grad_norm": 0.03640183061361313,
+      "learning_rate": 0.0001439912503907387,
+      "loss": 1.0261,
+      "step": 949
+    },
+    {
+      "epoch": 1.0708839310929117,
+      "grad_norm": 0.03576480597257614,
+      "learning_rate": 0.0001438845548157288,
+      "loss": 1.0378,
+      "step": 950
+    },
+    {
+      "epoch": 1.0720135554927985,
+      "grad_norm": 0.03689443692564964,
+      "learning_rate": 0.00014377779733082798,
+      "loss": 0.9868,
+      "step": 951
+    },
+    {
+      "epoch": 1.0731431798926856,
+      "grad_norm": 0.036124926060438156,
+      "learning_rate": 0.00014367097808664383,
+      "loss": 1.0565,
+      "step": 952
+    },
+    {
+      "epoch": 1.0742728042925727,
+      "grad_norm": 0.036695219576358795,
+      "learning_rate": 0.0001435640972338709,
+      "loss": 1.0738,
+      "step": 953
+    },
+    {
+      "epoch": 1.0754024286924597,
+      "grad_norm": 0.03690825775265694,
+      "learning_rate": 0.0001434571549232909,
+      "loss": 0.9761,
+      "step": 954
+    },
+    {
+      "epoch": 1.0765320530923468,
+      "grad_norm": 0.03635778650641441,
+      "learning_rate": 0.000143350151305772,
+      "loss": 0.947,
+      "step": 955
+    },
+    {
+      "epoch": 1.0776616774922338,
+      "grad_norm": 0.036799509078264236,
+      "learning_rate": 0.00014324308653226896,
+      "loss": 0.9861,
+      "step": 956
+    },
+    {
+      "epoch": 1.078791301892121,
+      "grad_norm": 0.03723785653710365,
+      "learning_rate": 0.00014313596075382284,
+      "loss": 1.0015,
+      "step": 957
+    },
+    {
+      "epoch": 1.079920926292008,
+      "grad_norm": 0.03606920316815376,
+      "learning_rate": 0.0001430287741215607,
+      "loss": 1.0458,
+      "step": 958
+    },
+    {
+      "epoch": 1.081050550691895,
+      "grad_norm": 0.03719672933220863,
+      "learning_rate": 0.00014292152678669557,
+      "loss": 0.9969,
+      "step": 959
+    },
+    {
+      "epoch": 1.0821801750917819,
+      "grad_norm": 0.03698967769742012,
+      "learning_rate": 0.0001428142189005259,
+      "loss": 1.0023,
+      "step": 960
+    },
+    {
+      "epoch": 1.083309799491669,
+      "grad_norm": 0.035385068506002426,
+      "learning_rate": 0.0001427068506144358,
+      "loss": 1.0768,
+      "step": 961
+    },
+    {
+      "epoch": 1.084439423891556,
+      "grad_norm": 0.03606635332107544,
+      "learning_rate": 0.00014259942207989443,
+      "loss": 1.0263,
+      "step": 962
+    },
+    {
+      "epoch": 1.085569048291443,
+      "grad_norm": 0.03499152511358261,
+      "learning_rate": 0.00014249193344845602,
+      "loss": 1.0146,
+      "step": 963
+    },
+    {
+      "epoch": 1.0866986726913301,
+      "grad_norm": 0.0371355339884758,
+      "learning_rate": 0.0001423843848717595,
+      "loss": 0.9832,
+      "step": 964
+    },
+    {
+      "epoch": 1.0878282970912172,
+      "grad_norm": 0.036203138530254364,
+      "learning_rate": 0.0001422767765015285,
+      "loss": 1.042,
+      "step": 965
+    },
+    {
+      "epoch": 1.0889579214911043,
+      "grad_norm": 0.03647124394774437,
+      "learning_rate": 0.00014216910848957088,
+      "loss": 0.9835,
+      "step": 966
+    },
+    {
+      "epoch": 1.0900875458909913,
+      "grad_norm": 0.03794392943382263,
+      "learning_rate": 0.0001420613809877787,
+      "loss": 0.974,
+      "step": 967
+    },
+    {
+      "epoch": 1.0912171702908782,
+      "grad_norm": 0.03726119548082352,
+      "learning_rate": 0.00014195359414812797,
+      "loss": 0.9381,
+      "step": 968
+    },
+    {
+      "epoch": 1.0923467946907652,
+      "grad_norm": 0.03955681249499321,
+      "learning_rate": 0.0001418457481226783,
+      "loss": 1.0461,
+      "step": 969
+    },
+    {
+      "epoch": 1.0934764190906523,
+      "grad_norm": 0.03713424876332283,
+      "learning_rate": 0.0001417378430635729,
+      "loss": 0.9963,
+      "step": 970
+    },
+    {
+      "epoch": 1.0946060434905394,
+      "grad_norm": 0.035701628774404526,
+      "learning_rate": 0.00014162987912303828,
+      "loss": 1.023,
+      "step": 971
+    },
+    {
+      "epoch": 1.0957356678904264,
+      "grad_norm": 0.037078987807035446,
+      "learning_rate": 0.00014152185645338388,
+      "loss": 0.9829,
+      "step": 972
+    },
+    {
+      "epoch": 1.0968652922903135,
+      "grad_norm": 0.037014082074165344,
+      "learning_rate": 0.00014141377520700214,
+      "loss": 1.1029,
+      "step": 973
+    },
+    {
+      "epoch": 1.0979949166902006,
+      "grad_norm": 0.03914770111441612,
+      "learning_rate": 0.00014130563553636807,
+      "loss": 0.9921,
+      "step": 974
+    },
+    {
+      "epoch": 1.0991245410900876,
+      "grad_norm": 0.03586632385849953,
+      "learning_rate": 0.00014119743759403907,
+      "loss": 0.9454,
+      "step": 975
+    },
+    {
+      "epoch": 1.1002541654899747,
+      "grad_norm": 0.03491047024726868,
+      "learning_rate": 0.00014108918153265485,
+      "loss": 0.8762,
+      "step": 976
+    },
+    {
+      "epoch": 1.1013837898898615,
+      "grad_norm": 0.036289557814598083,
+      "learning_rate": 0.00014098086750493694,
+      "loss": 1.0473,
+      "step": 977
+    },
+    {
+      "epoch": 1.1025134142897486,
+      "grad_norm": 0.036362823098897934,
+      "learning_rate": 0.0001408724956636889,
+      "loss": 0.9767,
+      "step": 978
+    },
+    {
+      "epoch": 1.1036430386896356,
+      "grad_norm": 0.03629835322499275,
+      "learning_rate": 0.0001407640661617955,
+      "loss": 0.9525,
+      "step": 979
+    },
+    {
+      "epoch": 1.1047726630895227,
+      "grad_norm": 0.03653542697429657,
+      "learning_rate": 0.00014065557915222322,
+      "loss": 0.9991,
+      "step": 980
+    },
+    {
+      "epoch": 1.1059022874894098,
+      "grad_norm": 0.03811173886060715,
+      "learning_rate": 0.00014054703478801948,
+      "loss": 0.9329,
+      "step": 981
+    },
+    {
+      "epoch": 1.1070319118892968,
+      "grad_norm": 0.037317369133234024,
+      "learning_rate": 0.00014043843322231257,
+      "loss": 1.022,
+      "step": 982
+    },
+    {
+      "epoch": 1.108161536289184,
+      "grad_norm": 0.03802327439188957,
+      "learning_rate": 0.00014032977460831162,
+      "loss": 0.9009,
+      "step": 983
+    },
+    {
+      "epoch": 1.109291160689071,
+      "grad_norm": 0.03663121536374092,
+      "learning_rate": 0.0001402210590993061,
+      "loss": 1.0042,
+      "step": 984
+    },
+    {
+      "epoch": 1.1104207850889578,
+      "grad_norm": 0.03598729893565178,
+      "learning_rate": 0.00014011228684866582,
+      "loss": 0.9048,
+      "step": 985
+    },
+    {
+      "epoch": 1.1115504094888449,
+      "grad_norm": 0.036852069199085236,
+      "learning_rate": 0.00014000345800984065,
+      "loss": 1.0355,
+      "step": 986
+    },
+    {
+      "epoch": 1.112680033888732,
+      "grad_norm": 0.03889515623450279,
+      "learning_rate": 0.00013989457273636026,
+      "loss": 1.0336,
+      "step": 987
+    },
+    {
+      "epoch": 1.113809658288619,
+      "grad_norm": 0.037266094237565994,
+      "learning_rate": 0.00013978563118183395,
+      "loss": 0.9659,
+      "step": 988
+    },
+    {
+      "epoch": 1.114939282688506,
+      "grad_norm": 0.035660672932863235,
+      "learning_rate": 0.00013967663349995037,
+      "loss": 1.0061,
+      "step": 989
+    },
+    {
+      "epoch": 1.1160689070883931,
+      "grad_norm": 0.035659607499837875,
+      "learning_rate": 0.00013956757984447745,
+      "loss": 0.9296,
+      "step": 990
+    },
+    {
+      "epoch": 1.1171985314882802,
+      "grad_norm": 0.03629400581121445,
+      "learning_rate": 0.00013945847036926195,
+      "loss": 1.0935,
+      "step": 991
+    },
+    {
+      "epoch": 1.1183281558881673,
+      "grad_norm": 0.03713002800941467,
+      "learning_rate": 0.00013934930522822953,
+      "loss": 0.9994,
+      "step": 992
+    },
+    {
+      "epoch": 1.1194577802880543,
+      "grad_norm": 0.03690033778548241,
+      "learning_rate": 0.0001392400845753843,
+      "loss": 0.9517,
+      "step": 993
+    },
+    {
+      "epoch": 1.1205874046879412,
+      "grad_norm": 0.03637135028839111,
+      "learning_rate": 0.0001391308085648086,
+      "loss": 0.9312,
+      "step": 994
+    },
+    {
+      "epoch": 1.1217170290878282,
+      "grad_norm": 0.03516946732997894,
+      "learning_rate": 0.00013902147735066306,
+      "loss": 0.8822,
+      "step": 995
+    },
+    {
+      "epoch": 1.1228466534877153,
+      "grad_norm": 0.03652889281511307,
+      "learning_rate": 0.00013891209108718599,
+      "loss": 0.9571,
+      "step": 996
+    },
+    {
+      "epoch": 1.1239762778876023,
+      "grad_norm": 0.036057133227586746,
+      "learning_rate": 0.00013880264992869354,
+      "loss": 0.9589,
+      "step": 997
+    },
+    {
+      "epoch": 1.1251059022874894,
+      "grad_norm": 0.03780772536993027,
+      "learning_rate": 0.00013869315402957914,
+      "loss": 1.0094,
+      "step": 998
+    },
+    {
+      "epoch": 1.1262355266873765,
+      "grad_norm": 0.03670530021190643,
+      "learning_rate": 0.00013858360354431355,
+      "loss": 0.9969,
+      "step": 999
+    },
+    {
+      "epoch": 1.1273651510872635,
+      "grad_norm": 0.037929147481918335,
+      "learning_rate": 0.0001384739986274445,
+      "loss": 0.9614,
+      "step": 1000
+    },
+    {
+      "epoch": 1.1284947754871506,
+      "grad_norm": 0.037020258605480194,
+      "learning_rate": 0.00013836433943359652,
+      "loss": 0.9567,
+      "step": 1001
+    },
+    {
+      "epoch": 1.1296243998870374,
+      "grad_norm": 0.036838166415691376,
+      "learning_rate": 0.0001382546261174707,
+      "loss": 0.9537,
+      "step": 1002
+    },
+    {
+      "epoch": 1.1307540242869245,
+      "grad_norm": 0.036922816187143326,
+      "learning_rate": 0.0001381448588338445,
+      "loss": 1.0113,
+      "step": 1003
+    },
+    {
+      "epoch": 1.1318836486868116,
+      "grad_norm": 0.036879848688840866,
+      "learning_rate": 0.00013803503773757148,
+      "loss": 1.0202,
+      "step": 1004
+    },
+    {
+      "epoch": 1.1330132730866986,
+      "grad_norm": 0.03841286897659302,
+      "learning_rate": 0.00013792516298358114,
+      "loss": 1.0836,
+      "step": 1005
+    },
+    {
+      "epoch": 1.1341428974865857,
+      "grad_norm": 0.03743935376405716,
+      "learning_rate": 0.00013781523472687873,
+      "loss": 1.006,
+      "step": 1006
+    },
+    {
+      "epoch": 1.1352725218864728,
+      "grad_norm": 0.038434430956840515,
+      "learning_rate": 0.0001377052531225449,
+      "loss": 0.9719,
+      "step": 1007
+    },
+    {
+      "epoch": 1.1364021462863598,
+      "grad_norm": 0.03810707479715347,
+      "learning_rate": 0.0001375952183257355,
+      "loss": 0.9137,
+      "step": 1008
+    },
+    {
+      "epoch": 1.137531770686247,
+      "grad_norm": 0.03748295083642006,
+      "learning_rate": 0.00013748513049168163,
+      "loss": 0.9542,
+      "step": 1009
+    },
+    {
+      "epoch": 1.138661395086134,
+      "grad_norm": 0.03579265996813774,
+      "learning_rate": 0.000137374989775689,
+      "loss": 0.9137,
+      "step": 1010
+    },
+    {
+      "epoch": 1.139791019486021,
+      "grad_norm": 0.037356823682785034,
+      "learning_rate": 0.00013726479633313808,
+      "loss": 1.037,
+      "step": 1011
+    },
+    {
+      "epoch": 1.1409206438859079,
+      "grad_norm": 0.03657485544681549,
+      "learning_rate": 0.0001371545503194836,
+      "loss": 0.9773,
+      "step": 1012
+    },
+    {
+      "epoch": 1.142050268285795,
+      "grad_norm": 0.03649429604411125,
+      "learning_rate": 0.00013704425189025452,
+      "loss": 0.9728,
+      "step": 1013
+    },
+    {
+      "epoch": 1.143179892685682,
+      "grad_norm": 0.036418452858924866,
+      "learning_rate": 0.00013693390120105375,
+      "loss": 0.9354,
+      "step": 1014
+    },
+    {
+      "epoch": 1.144309517085569,
+      "grad_norm": 0.036211349070072174,
+      "learning_rate": 0.00013682349840755785,
+      "loss": 0.9981,
+      "step": 1015
+    },
+    {
+      "epoch": 1.1454391414854561,
+      "grad_norm": 0.037094857543706894,
+      "learning_rate": 0.00013671304366551706,
+      "loss": 1.0344,
+      "step": 1016
+    },
+    {
+      "epoch": 1.1465687658853432,
+      "grad_norm": 0.03698652982711792,
+      "learning_rate": 0.00013660253713075471,
+      "loss": 1.0336,
+      "step": 1017
+    },
+    {
+      "epoch": 1.1476983902852302,
+      "grad_norm": 0.038060273975133896,
+      "learning_rate": 0.0001364919789591673,
+      "loss": 1.0393,
+      "step": 1018
+    },
+    {
+      "epoch": 1.148828014685117,
+      "grad_norm": 0.03657998517155647,
+      "learning_rate": 0.0001363813693067241,
+      "loss": 0.9779,
+      "step": 1019
+    },
+    {
+      "epoch": 1.1499576390850041,
+      "grad_norm": 0.037362392991781235,
+      "learning_rate": 0.00013627070832946718,
+      "loss": 1.0074,
+      "step": 1020
+    },
+    {
+      "epoch": 1.1510872634848912,
+      "grad_norm": 0.03672691062092781,
+      "learning_rate": 0.00013615999618351077,
+      "loss": 0.9936,
+      "step": 1021
+    },
+    {
+      "epoch": 1.1522168878847783,
+      "grad_norm": 0.03584654629230499,
+      "learning_rate": 0.00013604923302504147,
+      "loss": 1.0992,
+      "step": 1022
+    },
+    {
+      "epoch": 1.1533465122846653,
+      "grad_norm": 0.03621530905365944,
+      "learning_rate": 0.00013593841901031781,
+      "loss": 0.9189,
+      "step": 1023
+    },
+    {
+      "epoch": 1.1544761366845524,
+      "grad_norm": 0.03851611912250519,
+      "learning_rate": 0.00013582755429567,
+      "loss": 1.0137,
+      "step": 1024
+    },
+    {
+      "epoch": 1.1556057610844395,
+      "grad_norm": 0.03611522167921066,
+      "learning_rate": 0.00013571663903749984,
+      "loss": 0.9942,
+      "step": 1025
+    },
+    {
+      "epoch": 1.1567353854843265,
+      "grad_norm": 0.03798006474971771,
+      "learning_rate": 0.00013560567339228037,
+      "loss": 1.0745,
+      "step": 1026
+    },
+    {
+      "epoch": 1.1578650098842136,
+      "grad_norm": 0.03875018656253815,
+      "learning_rate": 0.00013549465751655585,
+      "loss": 0.9921,
+      "step": 1027
+    },
+    {
+      "epoch": 1.1589946342841007,
+      "grad_norm": 0.03791335970163345,
+      "learning_rate": 0.00013538359156694125,
+      "loss": 1.0061,
+      "step": 1028
+    },
+    {
+      "epoch": 1.1601242586839875,
+      "grad_norm": 0.03835611790418625,
+      "learning_rate": 0.00013527247570012217,
+      "loss": 0.922,
+      "step": 1029
+    },
+    {
+      "epoch": 1.1612538830838746,
+      "grad_norm": 0.03751290589570999,
+      "learning_rate": 0.00013516131007285483,
+      "loss": 0.9814,
+      "step": 1030
+    },
+    {
+      "epoch": 1.1623835074837616,
+      "grad_norm": 0.037721749395132065,
+      "learning_rate": 0.00013505009484196542,
+      "loss": 0.921,
+      "step": 1031
+    },
+    {
+      "epoch": 1.1635131318836487,
+      "grad_norm": 0.03777645528316498,
+      "learning_rate": 0.00013493883016435024,
+      "loss": 0.9686,
+      "step": 1032
+    },
+    {
+      "epoch": 1.1646427562835358,
+      "grad_norm": 0.037747886031866074,
+      "learning_rate": 0.00013482751619697534,
+      "loss": 1.0293,
+      "step": 1033
+    },
+    {
+      "epoch": 1.1657723806834228,
+      "grad_norm": 0.035930532962083817,
+      "learning_rate": 0.00013471615309687624,
+      "loss": 0.9538,
+      "step": 1034
+    },
+    {
+      "epoch": 1.1669020050833099,
+      "grad_norm": 0.036953311413526535,
+      "learning_rate": 0.00013460474102115785,
+      "loss": 0.9935,
+      "step": 1035
+    },
+    {
+      "epoch": 1.1680316294831967,
+      "grad_norm": 0.03994593396782875,
+      "learning_rate": 0.00013449328012699407,
+      "loss": 1.1051,
+      "step": 1036
+    },
+    {
+      "epoch": 1.1691612538830838,
+      "grad_norm": 0.03813205286860466,
+      "learning_rate": 0.0001343817705716278,
+      "loss": 0.9401,
+      "step": 1037
+    },
+    {
+      "epoch": 1.1702908782829708,
+      "grad_norm": 0.03704039007425308,
+      "learning_rate": 0.0001342702125123705,
+      "loss": 0.9015,
+      "step": 1038
+    },
+    {
+      "epoch": 1.171420502682858,
+      "grad_norm": 0.037067439407110214,
+      "learning_rate": 0.00013415860610660208,
+      "loss": 0.9702,
+      "step": 1039
+    },
+    {
+      "epoch": 1.172550127082745,
+      "grad_norm": 0.03761277720332146,
+      "learning_rate": 0.0001340469515117706,
+      "loss": 0.9302,
+      "step": 1040
+    },
+    {
+      "epoch": 1.173679751482632,
+      "grad_norm": 0.0372072272002697,
+      "learning_rate": 0.00013393524888539226,
+      "loss": 0.9595,
+      "step": 1041
+    },
+    {
+      "epoch": 1.174809375882519,
+      "grad_norm": 0.037263937294483185,
+      "learning_rate": 0.00013382349838505088,
+      "loss": 0.9506,
+      "step": 1042
+    },
+    {
+      "epoch": 1.1759390002824062,
+      "grad_norm": 0.03690359368920326,
+      "learning_rate": 0.0001337117001683978,
+      "loss": 0.9424,
+      "step": 1043
+    },
+    {
+      "epoch": 1.1770686246822932,
+      "grad_norm": 0.03702644258737564,
+      "learning_rate": 0.0001335998543931518,
+      "loss": 0.9763,
+      "step": 1044
+    },
+    {
+      "epoch": 1.1781982490821803,
+      "grad_norm": 0.03696833550930023,
+      "learning_rate": 0.00013348796121709862,
+      "loss": 0.9366,
+      "step": 1045
+    },
+    {
+      "epoch": 1.1793278734820671,
+      "grad_norm": 0.03843120485544205,
+      "learning_rate": 0.00013337602079809102,
+      "loss": 1.0288,
+      "step": 1046
+    },
+    {
+      "epoch": 1.1804574978819542,
+      "grad_norm": 0.03919968008995056,
+      "learning_rate": 0.0001332640332940482,
+      "loss": 1.0994,
+      "step": 1047
+    },
+    {
+      "epoch": 1.1815871222818413,
+      "grad_norm": 0.038000430911779404,
+      "learning_rate": 0.00013315199886295602,
+      "loss": 0.9932,
+      "step": 1048
+    },
+    {
+      "epoch": 1.1827167466817283,
+      "grad_norm": 0.037957947701215744,
+      "learning_rate": 0.0001330399176628664,
+      "loss": 1.0122,
+      "step": 1049
+    },
+    {
+      "epoch": 1.1838463710816154,
+      "grad_norm": 0.037756092846393585,
+      "learning_rate": 0.00013292778985189724,
+      "loss": 0.9397,
+      "step": 1050
+    },
+    {
+      "epoch": 1.1849759954815025,
+      "grad_norm": 0.03866475820541382,
+      "learning_rate": 0.00013281561558823224,
+      "loss": 0.9598,
+      "step": 1051
+    },
+    {
+      "epoch": 1.1861056198813895,
+      "grad_norm": 0.037318069487810135,
+      "learning_rate": 0.0001327033950301206,
+      "loss": 0.9921,
+      "step": 1052
+    },
+    {
+      "epoch": 1.1872352442812764,
+      "grad_norm": 0.039401739835739136,
+      "learning_rate": 0.00013259112833587686,
+      "loss": 1.0142,
+      "step": 1053
+    },
+    {
+      "epoch": 1.1883648686811634,
+      "grad_norm": 0.039969827979803085,
+      "learning_rate": 0.00013247881566388062,
+      "loss": 1.0763,
+      "step": 1054
+    },
+    {
+      "epoch": 1.1894944930810505,
+      "grad_norm": 0.038948189467191696,
+      "learning_rate": 0.0001323664571725764,
+      "loss": 0.9663,
+      "step": 1055
+    },
+    {
+      "epoch": 1.1906241174809375,
+      "grad_norm": 0.03894684091210365,
+      "learning_rate": 0.00013225405302047326,
+      "loss": 1.0565,
+      "step": 1056
+    },
+    {
+      "epoch": 1.1917537418808246,
+      "grad_norm": 0.03931012749671936,
+      "learning_rate": 0.00013214160336614476,
+      "loss": 0.9607,
+      "step": 1057
+    },
+    {
+      "epoch": 1.1928833662807117,
+      "grad_norm": 0.03936886042356491,
+      "learning_rate": 0.00013202910836822865,
+      "loss": 0.9651,
+      "step": 1058
+    },
+    {
+      "epoch": 1.1940129906805987,
+      "grad_norm": 0.03898691013455391,
+      "learning_rate": 0.00013191656818542654,
+      "loss": 0.9504,
+      "step": 1059
+    },
+    {
+      "epoch": 1.1951426150804858,
+      "grad_norm": 0.03756885230541229,
+      "learning_rate": 0.00013180398297650393,
+      "loss": 0.9724,
+      "step": 1060
+    },
+    {
+      "epoch": 1.1962722394803729,
+      "grad_norm": 0.03658022731542587,
+      "learning_rate": 0.00013169135290028974,
+      "loss": 1.0255,
+      "step": 1061
+    },
+    {
+      "epoch": 1.19740186388026,
+      "grad_norm": 0.03740183636546135,
+      "learning_rate": 0.00013157867811567628,
+      "loss": 1.0126,
+      "step": 1062
+    },
+    {
+      "epoch": 1.1985314882801468,
+      "grad_norm": 0.03823421150445938,
+      "learning_rate": 0.0001314659587816188,
+      "loss": 0.8985,
+      "step": 1063
+    },
+    {
+      "epoch": 1.1996611126800338,
+      "grad_norm": 0.039322640746831894,
+      "learning_rate": 0.0001313531950571355,
+      "loss": 1.0385,
+      "step": 1064
+    },
+    {
+      "epoch": 1.200790737079921,
+      "grad_norm": 0.03806902840733528,
+      "learning_rate": 0.00013124038710130722,
+      "loss": 1.0221,
+      "step": 1065
+    },
+    {
+      "epoch": 1.201920361479808,
+      "grad_norm": 0.037562061101198196,
+      "learning_rate": 0.00013112753507327706,
+      "loss": 0.9412,
+      "step": 1066
+    },
+    {
+      "epoch": 1.203049985879695,
+      "grad_norm": 0.03723952919244766,
+      "learning_rate": 0.00013101463913225048,
+      "loss": 1.0065,
+      "step": 1067
+    },
+    {
+      "epoch": 1.204179610279582,
+      "grad_norm": 0.03826143220067024,
+      "learning_rate": 0.00013090169943749476,
+      "loss": 0.9315,
+      "step": 1068
+    },
+    {
+      "epoch": 1.2053092346794692,
+      "grad_norm": 0.03751199692487717,
+      "learning_rate": 0.00013078871614833894,
+      "loss": 0.9418,
+      "step": 1069
+    },
+    {
+      "epoch": 1.206438859079356,
+      "grad_norm": 0.03741021826863289,
+      "learning_rate": 0.00013067568942417356,
+      "loss": 1.0147,
+      "step": 1070
+    },
+    {
+      "epoch": 1.207568483479243,
+      "grad_norm": 0.0373196117579937,
+      "learning_rate": 0.00013056261942445045,
+      "loss": 0.9944,
+      "step": 1071
+    },
+    {
+      "epoch": 1.2086981078791301,
+      "grad_norm": 0.038884326815605164,
+      "learning_rate": 0.0001304495063086825,
+      "loss": 1.064,
+      "step": 1072
+    },
+    {
+      "epoch": 1.2098277322790172,
+      "grad_norm": 0.037447500973939896,
+      "learning_rate": 0.00013033635023644338,
+      "loss": 1.0027,
+      "step": 1073
+    },
+    {
+      "epoch": 1.2109573566789043,
+      "grad_norm": 0.0381641611456871,
+      "learning_rate": 0.00013022315136736743,
+      "loss": 1.0136,
+      "step": 1074
+    },
+    {
+      "epoch": 1.2120869810787913,
+      "grad_norm": 0.03642114996910095,
+      "learning_rate": 0.00013010990986114926,
+      "loss": 0.9936,
+      "step": 1075
+    },
+    {
+      "epoch": 1.2132166054786784,
+      "grad_norm": 0.03828851506114006,
+      "learning_rate": 0.0001299966258775437,
+      "loss": 0.9752,
+      "step": 1076
+    },
+    {
+      "epoch": 1.2143462298785654,
+      "grad_norm": 0.03775739669799805,
+      "learning_rate": 0.00012988329957636558,
+      "loss": 1.0207,
+      "step": 1077
+    },
+    {
+      "epoch": 1.2154758542784525,
+      "grad_norm": 0.036843057721853256,
+      "learning_rate": 0.00012976993111748922,
+      "loss": 0.995,
+      "step": 1078
+    },
+    {
+      "epoch": 1.2166054786783396,
+      "grad_norm": 0.03826352581381798,
+      "learning_rate": 0.00012965652066084866,
+      "loss": 0.9746,
+      "step": 1079
+    },
+    {
+      "epoch": 1.2177351030782264,
+      "grad_norm": 0.03939265385270119,
+      "learning_rate": 0.00012954306836643703,
+      "loss": 0.9897,
+      "step": 1080
+    },
+    {
+      "epoch": 1.2188647274781135,
+      "grad_norm": 0.038330599665641785,
+      "learning_rate": 0.00012942957439430648,
+      "loss": 1.0044,
+      "step": 1081
+    },
+    {
+      "epoch": 1.2199943518780005,
+      "grad_norm": 0.03781147673726082,
+      "learning_rate": 0.00012931603890456805,
+      "loss": 1.0505,
+      "step": 1082
+    },
+    {
+      "epoch": 1.2211239762778876,
+      "grad_norm": 0.03792247548699379,
+      "learning_rate": 0.00012920246205739124,
+      "loss": 1.0327,
+      "step": 1083
+    },
+    {
+      "epoch": 1.2222536006777747,
+      "grad_norm": 0.03832760825753212,
+      "learning_rate": 0.00012908884401300404,
+      "loss": 1.0185,
+      "step": 1084
+    },
+    {
+      "epoch": 1.2233832250776617,
+      "grad_norm": 0.038666147738695145,
+      "learning_rate": 0.0001289751849316924,
+      "loss": 0.9639,
+      "step": 1085
+    },
+    {
+      "epoch": 1.2245128494775488,
+      "grad_norm": 0.037469346076250076,
+      "learning_rate": 0.00012886148497380024,
+      "loss": 0.9962,
+      "step": 1086
+    },
+    {
+      "epoch": 1.2256424738774356,
+      "grad_norm": 0.0382404662668705,
+      "learning_rate": 0.00012874774429972918,
+      "loss": 1.0073,
+      "step": 1087
+    },
+    {
+      "epoch": 1.2267720982773227,
+      "grad_norm": 0.037142958492040634,
+      "learning_rate": 0.0001286339630699382,
+      "loss": 0.9954,
+      "step": 1088
+    },
+    {
+      "epoch": 1.2279017226772098,
+      "grad_norm": 0.03799648955464363,
+      "learning_rate": 0.0001285201414449436,
+      "loss": 0.9685,
+      "step": 1089
+    },
+    {
+      "epoch": 1.2290313470770968,
+      "grad_norm": 0.03858242928981781,
+      "learning_rate": 0.0001284062795853185,
+      "loss": 0.9869,
+      "step": 1090
+    },
+    {
+      "epoch": 1.2301609714769839,
+      "grad_norm": 0.03654414415359497,
+      "learning_rate": 0.000128292377651693,
+      "loss": 1.0773,
+      "step": 1091
+    },
+    {
+      "epoch": 1.231290595876871,
+      "grad_norm": 0.037349898368120193,
+      "learning_rate": 0.00012817843580475354,
+      "loss": 1.0794,
+      "step": 1092
+    },
+    {
+      "epoch": 1.232420220276758,
+      "grad_norm": 0.03809039667248726,
+      "learning_rate": 0.00012806445420524297,
+      "loss": 1.0409,
+      "step": 1093
+    },
+    {
+      "epoch": 1.233549844676645,
+      "grad_norm": 0.03639010712504387,
+      "learning_rate": 0.00012795043301396016,
+      "loss": 1.0052,
+      "step": 1094
+    },
+    {
+      "epoch": 1.2346794690765321,
+      "grad_norm": 0.03701767325401306,
+      "learning_rate": 0.00012783637239175994,
+      "loss": 0.9447,
+      "step": 1095
+    },
+    {
+      "epoch": 1.2358090934764192,
+      "grad_norm": 0.03904819115996361,
+      "learning_rate": 0.00012772227249955263,
+      "loss": 0.9801,
+      "step": 1096
+    },
+    {
+      "epoch": 1.236938717876306,
+      "grad_norm": 0.037295203655958176,
+      "learning_rate": 0.00012760813349830405,
+      "loss": 0.9683,
+      "step": 1097
+    },
+    {
+      "epoch": 1.2380683422761931,
+      "grad_norm": 0.037878576666116714,
+      "learning_rate": 0.00012749395554903514,
+      "loss": 1.0504,
+      "step": 1098
+    },
+    {
+      "epoch": 1.2391979666760802,
+      "grad_norm": 0.03729383647441864,
+      "learning_rate": 0.00012737973881282185,
+      "loss": 0.9637,
+      "step": 1099
+    },
+    {
+      "epoch": 1.2403275910759672,
+      "grad_norm": 0.038625191897153854,
+      "learning_rate": 0.00012726548345079475,
+      "loss": 1.0177,
+      "step": 1100
+    },
+    {
+      "epoch": 1.2414572154758543,
+      "grad_norm": 0.03700273111462593,
+      "learning_rate": 0.00012715118962413894,
+      "loss": 1.064,
+      "step": 1101
+    },
+    {
+      "epoch": 1.2425868398757414,
+      "grad_norm": 0.03757200762629509,
+      "learning_rate": 0.00012703685749409387,
+      "loss": 0.9652,
+      "step": 1102
+    },
+    {
+      "epoch": 1.2437164642756284,
+      "grad_norm": 0.037905555218458176,
+      "learning_rate": 0.00012692248722195293,
+      "loss": 1.0696,
+      "step": 1103
+    },
+    {
+      "epoch": 1.2448460886755153,
+      "grad_norm": 0.03769846633076668,
+      "learning_rate": 0.00012680807896906329,
+      "loss": 1.0299,
+      "step": 1104
+    },
+    {
+      "epoch": 1.2459757130754023,
+      "grad_norm": 0.0391731932759285,
+      "learning_rate": 0.00012669363289682584,
+      "loss": 0.9584,
+      "step": 1105
+    },
+    {
+      "epoch": 1.2471053374752894,
+      "grad_norm": 0.03786834701895714,
+      "learning_rate": 0.00012657914916669466,
+      "loss": 0.9974,
+      "step": 1106
+    },
+    {
+      "epoch": 1.2482349618751765,
+      "grad_norm": 0.037806760519742966,
+      "learning_rate": 0.00012646462794017708,
+      "loss": 0.9201,
+      "step": 1107
+    },
+    {
+      "epoch": 1.2493645862750635,
+      "grad_norm": 0.04048847034573555,
+      "learning_rate": 0.00012635006937883328,
+      "loss": 0.9848,
+      "step": 1108
+    },
+    {
+      "epoch": 1.2504942106749506,
+      "grad_norm": 0.03775718808174133,
+      "learning_rate": 0.00012623547364427605,
+      "loss": 0.9275,
+      "step": 1109
+    },
+    {
+      "epoch": 1.2516238350748377,
+      "grad_norm": 0.03765375167131424,
+      "learning_rate": 0.0001261208408981708,
+      "loss": 0.9887,
+      "step": 1110
+    },
+    {
+      "epoch": 1.2527534594747247,
+      "grad_norm": 0.03869464248418808,
+      "learning_rate": 0.00012600617130223494,
+      "loss": 0.9137,
+      "step": 1111
+    },
+    {
+      "epoch": 1.2538830838746118,
+      "grad_norm": 0.038646209985017776,
+      "learning_rate": 0.000125891465018238,
+      "loss": 0.9535,
+      "step": 1112
+    },
+    {
+      "epoch": 1.2550127082744988,
+      "grad_norm": 0.038947030901908875,
+      "learning_rate": 0.00012577672220800122,
+      "loss": 0.988,
+      "step": 1113
+    },
+    {
+      "epoch": 1.2561423326743857,
+      "grad_norm": 0.039047569036483765,
+      "learning_rate": 0.00012566194303339739,
+      "loss": 1.0917,
+      "step": 1114
+    },
+    {
+      "epoch": 1.2572719570742728,
+      "grad_norm": 0.03860335052013397,
+      "learning_rate": 0.00012554712765635057,
+      "loss": 1.0186,
+      "step": 1115
+    },
+    {
+      "epoch": 1.2584015814741598,
+      "grad_norm": 0.03737378120422363,
+      "learning_rate": 0.0001254322762388359,
+      "loss": 0.9183,
+      "step": 1116
+    },
+    {
+      "epoch": 1.2595312058740469,
+      "grad_norm": 0.038941189646720886,
+      "learning_rate": 0.00012531738894287946,
+      "loss": 1.0043,
+      "step": 1117
+    },
+    {
+      "epoch": 1.260660830273934,
+      "grad_norm": 0.038606077432632446,
+      "learning_rate": 0.00012520246593055777,
+      "loss": 0.9232,
+      "step": 1118
+    },
+    {
+      "epoch": 1.261790454673821,
+      "grad_norm": 0.03935806453227997,
+      "learning_rate": 0.0001250875073639979,
+      "loss": 1.0401,
+      "step": 1119
+    },
+    {
+      "epoch": 1.262920079073708,
+      "grad_norm": 0.03954112157225609,
+      "learning_rate": 0.0001249725134053769,
+      "loss": 1.0234,
+      "step": 1120
+    },
+    {
+      "epoch": 1.264049703473595,
+      "grad_norm": 0.03769481182098389,
+      "learning_rate": 0.00012485748421692196,
+      "loss": 1.1324,
+      "step": 1121
+    },
+    {
+      "epoch": 1.265179327873482,
+      "grad_norm": 0.037308190017938614,
+      "learning_rate": 0.00012474241996090986,
+      "loss": 0.983,
+      "step": 1122
+    },
+    {
+      "epoch": 1.266308952273369,
+      "grad_norm": 0.03750931844115257,
+      "learning_rate": 0.00012462732079966676,
+      "loss": 0.9654,
+      "step": 1123
+    },
+    {
+      "epoch": 1.267438576673256,
+      "grad_norm": 0.03773000091314316,
+      "learning_rate": 0.0001245121868955683,
+      "loss": 0.9537,
+      "step": 1124
+    },
+    {
+      "epoch": 1.2685682010731432,
+      "grad_norm": 0.03808850795030594,
+      "learning_rate": 0.00012439701841103888,
+      "loss": 1.0058,
+      "step": 1125
+    },
+    {
+      "epoch": 1.2696978254730302,
+      "grad_norm": 0.03719565272331238,
+      "learning_rate": 0.00012428181550855184,
+      "loss": 1.047,
+      "step": 1126
+    },
+    {
+      "epoch": 1.2708274498729173,
+      "grad_norm": 0.03920544311404228,
+      "learning_rate": 0.00012416657835062907,
+      "loss": 0.9937,
+      "step": 1127
+    },
+    {
+      "epoch": 1.2719570742728044,
+      "grad_norm": 0.0384603887796402,
+      "learning_rate": 0.00012405130709984064,
+      "loss": 1.0042,
+      "step": 1128
+    },
+    {
+      "epoch": 1.2730866986726914,
+      "grad_norm": 0.039956919848918915,
+      "learning_rate": 0.00012393600191880493,
+      "loss": 1.0264,
+      "step": 1129
+    },
+    {
+      "epoch": 1.2742163230725785,
+      "grad_norm": 0.03926371783018112,
+      "learning_rate": 0.00012382066297018804,
+      "loss": 1.0665,
+      "step": 1130
+    },
+    {
+      "epoch": 1.2753459474724653,
+      "grad_norm": 0.03931189700961113,
+      "learning_rate": 0.00012370529041670376,
+      "loss": 1.004,
+      "step": 1131
+    },
+    {
+      "epoch": 1.2764755718723524,
+      "grad_norm": 0.038920726627111435,
+      "learning_rate": 0.00012358988442111326,
+      "loss": 0.9131,
+      "step": 1132
+    },
+    {
+      "epoch": 1.2776051962722395,
+      "grad_norm": 0.039931900799274445,
+      "learning_rate": 0.0001234744451462249,
+      "loss": 0.9556,
+      "step": 1133
+    },
+    {
+      "epoch": 1.2787348206721265,
+      "grad_norm": 0.03920763358473778,
+      "learning_rate": 0.000123358972754894,
+      "loss": 1.0534,
+      "step": 1134
+    },
+    {
+      "epoch": 1.2798644450720136,
+      "grad_norm": 0.03894329071044922,
+      "learning_rate": 0.0001232434674100226,
+      "loss": 0.9777,
+      "step": 1135
+    },
+    {
+      "epoch": 1.2809940694719006,
+      "grad_norm": 0.039821792393922806,
+      "learning_rate": 0.00012312792927455924,
+      "loss": 1.0461,
+      "step": 1136
+    },
+    {
+      "epoch": 1.2821236938717877,
+      "grad_norm": 0.041238732635974884,
+      "learning_rate": 0.00012301235851149865,
+      "loss": 0.9593,
+      "step": 1137
+    },
+    {
+      "epoch": 1.2832533182716745,
+      "grad_norm": 0.039833102375268936,
+      "learning_rate": 0.00012289675528388169,
+      "loss": 1.0378,
+      "step": 1138
+    },
+    {
+      "epoch": 1.2843829426715616,
+      "grad_norm": 0.03923288360238075,
+      "learning_rate": 0.00012278111975479492,
+      "loss": 0.9528,
+      "step": 1139
+    },
+    {
+      "epoch": 1.2855125670714487,
+      "grad_norm": 0.03867553547024727,
+      "learning_rate": 0.00012266545208737054,
+      "loss": 1.0494,
+      "step": 1140
+    },
+    {
+      "epoch": 1.2866421914713357,
+      "grad_norm": 0.03829865902662277,
+      "learning_rate": 0.0001225497524447861,
+      "loss": 0.889,
+      "step": 1141
+    },
+    {
+      "epoch": 1.2877718158712228,
+      "grad_norm": 0.039233699440956116,
+      "learning_rate": 0.00012243402099026417,
+      "loss": 0.8935,
+      "step": 1142
+    },
+    {
+      "epoch": 1.2889014402711099,
+      "grad_norm": 0.03830961883068085,
+      "learning_rate": 0.0001223182578870723,
+      "loss": 0.9697,
+      "step": 1143
+    },
+    {
+      "epoch": 1.290031064670997,
+      "grad_norm": 0.03787538781762123,
+      "learning_rate": 0.0001222024632985226,
+      "loss": 0.9645,
+      "step": 1144
+    },
+    {
+      "epoch": 1.291160689070884,
+      "grad_norm": 0.03876281529664993,
+      "learning_rate": 0.00012208663738797165,
+      "loss": 0.9705,
+      "step": 1145
+    },
+    {
+      "epoch": 1.292290313470771,
+      "grad_norm": 0.03762456402182579,
+      "learning_rate": 0.00012197078031882026,
+      "loss": 1.012,
+      "step": 1146
+    },
+    {
+      "epoch": 1.2934199378706581,
+      "grad_norm": 0.037920914590358734,
+      "learning_rate": 0.00012185489225451305,
+      "loss": 0.8949,
+      "step": 1147
+    },
+    {
+      "epoch": 1.294549562270545,
+      "grad_norm": 0.03920237347483635,
+      "learning_rate": 0.00012173897335853856,
+      "loss": 0.959,
+      "step": 1148
+    },
+    {
+      "epoch": 1.295679186670432,
+      "grad_norm": 0.03853004053235054,
+      "learning_rate": 0.00012162302379442865,
+      "loss": 1.011,
+      "step": 1149
+    },
+    {
+      "epoch": 1.296808811070319,
+      "grad_norm": 0.039866186678409576,
+      "learning_rate": 0.00012150704372575854,
+      "loss": 1.0368,
+      "step": 1150
+    },
+    {
+      "epoch": 1.2979384354702062,
+      "grad_norm": 0.040773652493953705,
+      "learning_rate": 0.00012139103331614649,
+      "loss": 1.0744,
+      "step": 1151
+    },
+    {
+      "epoch": 1.2990680598700932,
+      "grad_norm": 0.038420483469963074,
+      "learning_rate": 0.00012127499272925352,
+      "loss": 0.993,
+      "step": 1152
+    },
+    {
+      "epoch": 1.3001976842699803,
+      "grad_norm": 0.038056086748838425,
+      "learning_rate": 0.00012115892212878324,
+      "loss": 0.9368,
+      "step": 1153
+    },
+    {
+      "epoch": 1.3013273086698673,
+      "grad_norm": 0.04063399136066437,
+      "learning_rate": 0.00012104282167848157,
+      "loss": 0.8962,
+      "step": 1154
+    },
+    {
+      "epoch": 1.3024569330697542,
+      "grad_norm": 0.038728486746549606,
+      "learning_rate": 0.00012092669154213665,
+      "loss": 0.9737,
+      "step": 1155
+    },
+    {
+      "epoch": 1.3035865574696412,
+      "grad_norm": 0.03924306482076645,
+      "learning_rate": 0.00012081053188357838,
+      "loss": 1.0224,
+      "step": 1156
+    },
+    {
+      "epoch": 1.3047161818695283,
+      "grad_norm": 0.03962641581892967,
+      "learning_rate": 0.00012069434286667833,
+      "loss": 0.9779,
+      "step": 1157
+    },
+    {
+      "epoch": 1.3058458062694154,
+      "grad_norm": 0.03946220874786377,
+      "learning_rate": 0.00012057812465534953,
+      "loss": 1.0043,
+      "step": 1158
+    },
+    {
+      "epoch": 1.3069754306693024,
+      "grad_norm": 0.038344915956258774,
+      "learning_rate": 0.00012046187741354618,
+      "loss": 1.0668,
+      "step": 1159
+    },
+    {
+      "epoch": 1.3081050550691895,
+      "grad_norm": 0.03802265226840973,
+      "learning_rate": 0.0001203456013052634,
+      "loss": 0.9928,
+      "step": 1160
+    },
+    {
+      "epoch": 1.3092346794690766,
+      "grad_norm": 0.03755849227309227,
+      "learning_rate": 0.00012022929649453707,
+      "loss": 0.9103,
+      "step": 1161
+    },
+    {
+      "epoch": 1.3103643038689636,
+      "grad_norm": 0.03990541025996208,
+      "learning_rate": 0.00012011296314544359,
+      "loss": 0.9837,
+      "step": 1162
+    },
+    {
+      "epoch": 1.3114939282688507,
+      "grad_norm": 0.03864932060241699,
+      "learning_rate": 0.00011999660142209953,
+      "loss": 0.99,
+      "step": 1163
+    },
+    {
+      "epoch": 1.3126235526687378,
+      "grad_norm": 0.03857451304793358,
+      "learning_rate": 0.00011988021148866158,
+      "loss": 1.0028,
+      "step": 1164
+    },
+    {
+      "epoch": 1.3137531770686246,
+      "grad_norm": 0.03803737089037895,
+      "learning_rate": 0.00011976379350932618,
+      "loss": 0.9621,
+      "step": 1165
+    },
+    {
+      "epoch": 1.3148828014685117,
+      "grad_norm": 0.037970781326293945,
+      "learning_rate": 0.00011964734764832933,
+      "loss": 0.942,
+      "step": 1166
+    },
+    {
+      "epoch": 1.3160124258683987,
+      "grad_norm": 0.03878822922706604,
+      "learning_rate": 0.0001195308740699464,
+      "loss": 0.9238,
+      "step": 1167
+    },
+    {
+      "epoch": 1.3171420502682858,
+      "grad_norm": 0.03999705985188484,
+      "learning_rate": 0.00011941437293849181,
+      "loss": 1.0493,
+      "step": 1168
+    },
+    {
+      "epoch": 1.3182716746681729,
+      "grad_norm": 0.04126083478331566,
+      "learning_rate": 0.00011929784441831893,
+      "loss": 0.9518,
+      "step": 1169
+    },
+    {
+      "epoch": 1.31940129906806,
+      "grad_norm": 0.03893827274441719,
+      "learning_rate": 0.00011918128867381965,
+      "loss": 0.8951,
+      "step": 1170
+    },
+    {
+      "epoch": 1.320530923467947,
+      "grad_norm": 0.039558008313179016,
+      "learning_rate": 0.00011906470586942445,
+      "loss": 0.9446,
+      "step": 1171
+    },
+    {
+      "epoch": 1.3216605478678338,
+      "grad_norm": 0.03813496232032776,
+      "learning_rate": 0.00011894809616960174,
+      "loss": 0.9837,
+      "step": 1172
+    },
+    {
+      "epoch": 1.3227901722677209,
+      "grad_norm": 0.03912735357880592,
+      "learning_rate": 0.00011883145973885805,
+      "loss": 1.0211,
+      "step": 1173
+    },
+    {
+      "epoch": 1.323919796667608,
+      "grad_norm": 0.0382121279835701,
+      "learning_rate": 0.00011871479674173762,
+      "loss": 1.0054,
+      "step": 1174
+    },
+    {
+      "epoch": 1.325049421067495,
+      "grad_norm": 0.03812730684876442,
+      "learning_rate": 0.00011859810734282208,
+      "loss": 0.9412,
+      "step": 1175
+    },
+    {
+      "epoch": 1.326179045467382,
+      "grad_norm": 0.04326564446091652,
+      "learning_rate": 0.00011848139170673036,
+      "loss": 1.0046,
+      "step": 1176
+    },
+    {
+      "epoch": 1.3273086698672691,
+      "grad_norm": 0.038262803107500076,
+      "learning_rate": 0.00011836464999811838,
+      "loss": 0.936,
+      "step": 1177
+    },
+    {
+      "epoch": 1.3284382942671562,
+      "grad_norm": 0.03915071487426758,
+      "learning_rate": 0.0001182478823816789,
+      "loss": 0.9397,
+      "step": 1178
+    },
+    {
+      "epoch": 1.3295679186670433,
+      "grad_norm": 0.0392167903482914,
+      "learning_rate": 0.00011813108902214111,
+      "loss": 1.0413,
+      "step": 1179
+    },
+    {
+      "epoch": 1.3306975430669303,
+      "grad_norm": 0.04144560545682907,
+      "learning_rate": 0.00011801427008427063,
+      "loss": 1.0385,
+      "step": 1180
+    },
+    {
+      "epoch": 1.3306975430669303,
+      "eval_loss": 0.995391845703125,
+      "eval_runtime": 559.0238,
+      "eval_samples_per_second": 17.5,
+      "eval_steps_per_second": 8.751,
+      "step": 1180
+    },
+    {
+      "epoch": 1.3318271674668174,
+      "grad_norm": 0.039459336549043655,
+      "learning_rate": 0.00011789742573286915,
+      "loss": 0.9662,
+      "step": 1181
+    },
+    {
+      "epoch": 1.3329567918667042,
+      "grad_norm": 0.03919503092765808,
+      "learning_rate": 0.00011778055613277415,
+      "loss": 0.9673,
+      "step": 1182
+    },
+    {
+      "epoch": 1.3340864162665913,
+      "grad_norm": 0.03915826976299286,
+      "learning_rate": 0.00011766366144885877,
+      "loss": 0.9992,
+      "step": 1183
+    },
+    {
+      "epoch": 1.3352160406664784,
+      "grad_norm": 0.038948994129896164,
+      "learning_rate": 0.00011754674184603149,
+      "loss": 0.959,
+      "step": 1184
+    },
+    {
+      "epoch": 1.3363456650663654,
+      "grad_norm": 0.03835888206958771,
+      "learning_rate": 0.00011742979748923611,
+      "loss": 1.0011,
+      "step": 1185
+    },
+    {
+      "epoch": 1.3374752894662525,
+      "grad_norm": 0.03993493691086769,
+      "learning_rate": 0.00011731282854345114,
+      "loss": 1.0694,
+      "step": 1186
+    },
+    {
+      "epoch": 1.3386049138661396,
+      "grad_norm": 0.03888334706425667,
+      "learning_rate": 0.00011719583517368985,
+      "loss": 1.0382,
+      "step": 1187
+    },
+    {
+      "epoch": 1.3397345382660266,
+      "grad_norm": 0.03841853514313698,
+      "learning_rate": 0.00011707881754500005,
+      "loss": 0.9161,
+      "step": 1188
+    },
+    {
+      "epoch": 1.3408641626659135,
+      "grad_norm": 0.038549747318029404,
+      "learning_rate": 0.00011696177582246367,
+      "loss": 0.9674,
+      "step": 1189
+    },
+    {
+      "epoch": 1.3419937870658005,
+      "grad_norm": 0.04010612517595291,
+      "learning_rate": 0.00011684471017119667,
+      "loss": 1.0048,
+      "step": 1190
+    },
+    {
+      "epoch": 1.3431234114656876,
+      "grad_norm": 0.03912900388240814,
+      "learning_rate": 0.00011672762075634873,
+      "loss": 0.9314,
+      "step": 1191
+    },
+    {
+      "epoch": 1.3442530358655747,
+      "grad_norm": 0.041035935282707214,
+      "learning_rate": 0.00011661050774310319,
+      "loss": 0.9267,
+      "step": 1192
+    },
+    {
+      "epoch": 1.3453826602654617,
+      "grad_norm": 0.03944677114486694,
+      "learning_rate": 0.00011649337129667648,
+      "loss": 0.9811,
+      "step": 1193
+    },
+    {
+      "epoch": 1.3465122846653488,
+      "grad_norm": 0.040055934339761734,
+      "learning_rate": 0.00011637621158231822,
+      "loss": 0.9618,
+      "step": 1194
+    },
+    {
+      "epoch": 1.3476419090652358,
+      "grad_norm": 0.038965675979852676,
+      "learning_rate": 0.00011625902876531082,
+      "loss": 0.9501,
+      "step": 1195
+    },
+    {
+      "epoch": 1.348771533465123,
+      "grad_norm": 0.038453359156847,
+      "learning_rate": 0.00011614182301096927,
+      "loss": 0.9526,
+      "step": 1196
+    },
+    {
+      "epoch": 1.34990115786501,
+      "grad_norm": 0.038837965577840805,
+      "learning_rate": 0.00011602459448464102,
+      "loss": 0.9861,
+      "step": 1197
+    },
+    {
+      "epoch": 1.351030782264897,
+      "grad_norm": 0.03882889077067375,
+      "learning_rate": 0.00011590734335170545,
+      "loss": 0.9874,
+      "step": 1198
+    },
+    {
+      "epoch": 1.352160406664784,
+      "grad_norm": 0.0396343395113945,
+      "learning_rate": 0.000115790069777574,
+      "loss": 0.9412,
+      "step": 1199
+    },
+    {
+      "epoch": 1.353290031064671,
+      "grad_norm": 0.03892532363533974,
+      "learning_rate": 0.00011567277392768972,
+      "loss": 0.8851,
+      "step": 1200
+    },
+    {
+      "epoch": 1.354419655464558,
+      "grad_norm": 0.0408422015607357,
+      "learning_rate": 0.00011555545596752701,
+      "loss": 0.9849,
+      "step": 1201
+    },
+    {
+      "epoch": 1.355549279864445,
+      "grad_norm": 0.040196966379880905,
+      "learning_rate": 0.0001154381160625916,
+      "loss": 0.9962,
+      "step": 1202
+    },
+    {
+      "epoch": 1.3566789042643321,
+      "grad_norm": 0.04056261479854584,
+      "learning_rate": 0.00011532075437842004,
+      "loss": 1.0322,
+      "step": 1203
+    },
+    {
+      "epoch": 1.3578085286642192,
+      "grad_norm": 0.03838193044066429,
+      "learning_rate": 0.00011520337108057974,
+      "loss": 1.061,
+      "step": 1204
+    },
+    {
+      "epoch": 1.3589381530641063,
+      "grad_norm": 0.03917081281542778,
+      "learning_rate": 0.00011508596633466852,
+      "loss": 1.0052,
+      "step": 1205
+    },
+    {
+      "epoch": 1.360067777463993,
+      "grad_norm": 0.040417592972517014,
+      "learning_rate": 0.00011496854030631443,
+      "loss": 0.9492,
+      "step": 1206
+    },
+    {
+      "epoch": 1.3611974018638802,
+      "grad_norm": 0.03877858445048332,
+      "learning_rate": 0.00011485109316117562,
+      "loss": 0.9382,
+      "step": 1207
+    },
+    {
+      "epoch": 1.3623270262637672,
+      "grad_norm": 0.03815083205699921,
+      "learning_rate": 0.00011473362506493997,
+      "loss": 0.9688,
+      "step": 1208
+    },
+    {
+      "epoch": 1.3634566506636543,
+      "grad_norm": 0.038873858749866486,
+      "learning_rate": 0.00011461613618332497,
+      "loss": 0.8748,
+      "step": 1209
+    },
+    {
+      "epoch": 1.3645862750635414,
+      "grad_norm": 0.03795653581619263,
+      "learning_rate": 0.00011449862668207734,
+      "loss": 0.9784,
+      "step": 1210
+    },
+    {
+      "epoch": 1.3657158994634284,
+      "grad_norm": 0.04054610803723335,
+      "learning_rate": 0.00011438109672697303,
+      "loss": 0.9856,
+      "step": 1211
+    },
+    {
+      "epoch": 1.3668455238633155,
+      "grad_norm": 0.039193637669086456,
+      "learning_rate": 0.00011426354648381674,
+      "loss": 0.9638,
+      "step": 1212
+    },
+    {
+      "epoch": 1.3679751482632025,
+      "grad_norm": 0.038261231034994125,
+      "learning_rate": 0.00011414597611844183,
+      "loss": 1.0003,
+      "step": 1213
+    },
+    {
+      "epoch": 1.3691047726630896,
+      "grad_norm": 0.039814114570617676,
+      "learning_rate": 0.00011402838579671001,
+      "loss": 0.9288,
+      "step": 1214
+    },
+    {
+      "epoch": 1.3702343970629767,
+      "grad_norm": 0.04093167930841446,
+      "learning_rate": 0.00011391077568451116,
+      "loss": 1.0528,
+      "step": 1215
+    },
+    {
+      "epoch": 1.3713640214628637,
+      "grad_norm": 0.040623877197504044,
+      "learning_rate": 0.00011379314594776315,
+      "loss": 0.9221,
+      "step": 1216
+    },
+    {
+      "epoch": 1.3724936458627506,
+      "grad_norm": 0.039268527179956436,
+      "learning_rate": 0.00011367549675241142,
+      "loss": 0.9739,
+      "step": 1217
+    },
+    {
+      "epoch": 1.3736232702626376,
+      "grad_norm": 0.038862016052007675,
+      "learning_rate": 0.00011355782826442895,
+      "loss": 0.9871,
+      "step": 1218
+    },
+    {
+      "epoch": 1.3747528946625247,
+      "grad_norm": 0.03837895020842552,
+      "learning_rate": 0.0001134401406498159,
+      "loss": 0.9517,
+      "step": 1219
+    },
+    {
+      "epoch": 1.3758825190624118,
+      "grad_norm": 0.0400700680911541,
+      "learning_rate": 0.0001133224340745994,
+      "loss": 1.0339,
+      "step": 1220
+    },
+    {
+      "epoch": 1.3770121434622988,
+      "grad_norm": 0.040071625262498856,
+      "learning_rate": 0.00011320470870483335,
+      "loss": 0.9464,
+      "step": 1221
+    },
+    {
+      "epoch": 1.378141767862186,
+      "grad_norm": 0.039051372557878494,
+      "learning_rate": 0.00011308696470659814,
+      "loss": 0.9558,
+      "step": 1222
+    },
+    {
+      "epoch": 1.3792713922620727,
+      "grad_norm": 0.04017410799860954,
+      "learning_rate": 0.0001129692022460005,
+      "loss": 0.9779,
+      "step": 1223
+    },
+    {
+      "epoch": 1.3804010166619598,
+      "grad_norm": 0.039144422858953476,
+      "learning_rate": 0.00011285142148917312,
+      "loss": 1.0111,
+      "step": 1224
+    },
+    {
+      "epoch": 1.3815306410618469,
+      "grad_norm": 0.04037921130657196,
+      "learning_rate": 0.00011273362260227458,
+      "loss": 0.9653,
+      "step": 1225
+    },
+    {
+      "epoch": 1.382660265461734,
+      "grad_norm": 0.03725403547286987,
+      "learning_rate": 0.000112615805751489,
+      "loss": 1.0813,
+      "step": 1226
+    },
+    {
+      "epoch": 1.383789889861621,
+      "grad_norm": 0.03872475028038025,
+      "learning_rate": 0.00011249797110302582,
+      "loss": 0.9006,
+      "step": 1227
+    },
+    {
+      "epoch": 1.384919514261508,
+      "grad_norm": 0.03926214575767517,
+      "learning_rate": 0.00011238011882311963,
+      "loss": 1.0214,
+      "step": 1228
+    },
+    {
+      "epoch": 1.3860491386613951,
+      "grad_norm": 0.03871994838118553,
+      "learning_rate": 0.00011226224907802985,
+      "loss": 1.0218,
+      "step": 1229
+    },
+    {
+      "epoch": 1.3871787630612822,
+      "grad_norm": 0.03772478923201561,
+      "learning_rate": 0.00011214436203404062,
+      "loss": 0.918,
+      "step": 1230
+    },
+    {
+      "epoch": 1.3883083874611692,
+      "grad_norm": 0.038930702954530716,
+      "learning_rate": 0.0001120264578574604,
+      "loss": 0.9678,
+      "step": 1231
+    },
+    {
+      "epoch": 1.3894380118610563,
+      "grad_norm": 0.03879974037408829,
+      "learning_rate": 0.00011190853671462183,
+      "loss": 0.9241,
+      "step": 1232
+    },
+    {
+      "epoch": 1.3905676362609434,
+      "grad_norm": 0.04013784974813461,
+      "learning_rate": 0.00011179059877188154,
+      "loss": 1.0302,
+      "step": 1233
+    },
+    {
+      "epoch": 1.3916972606608302,
+      "grad_norm": 0.039637017995119095,
+      "learning_rate": 0.00011167264419561978,
+      "loss": 1.0615,
+      "step": 1234
+    },
+    {
+      "epoch": 1.3928268850607173,
+      "grad_norm": 0.03899321332573891,
+      "learning_rate": 0.00011155467315224038,
+      "loss": 0.963,
+      "step": 1235
+    },
+    {
+      "epoch": 1.3939565094606043,
+      "grad_norm": 0.03954652324318886,
+      "learning_rate": 0.00011143668580817026,
+      "loss": 1.081,
+      "step": 1236
+    },
+    {
+      "epoch": 1.3950861338604914,
+      "grad_norm": 0.03980173170566559,
+      "learning_rate": 0.00011131868232985945,
+      "loss": 1.0521,
+      "step": 1237
+    },
+    {
+      "epoch": 1.3962157582603785,
+      "grad_norm": 0.04101087152957916,
+      "learning_rate": 0.00011120066288378066,
+      "loss": 1.0194,
+      "step": 1238
+    },
+    {
+      "epoch": 1.3973453826602655,
+      "grad_norm": 0.03934963047504425,
+      "learning_rate": 0.00011108262763642923,
+      "loss": 1.0247,
+      "step": 1239
+    },
+    {
+      "epoch": 1.3984750070601524,
+      "grad_norm": 0.04077817499637604,
+      "learning_rate": 0.00011096457675432266,
+      "loss": 1.0232,
+      "step": 1240
+    },
+    {
+      "epoch": 1.3996046314600394,
+      "grad_norm": 0.040299609303474426,
+      "learning_rate": 0.00011084651040400057,
+      "loss": 1.0593,
+      "step": 1241
+    },
+    {
+      "epoch": 1.4007342558599265,
+      "grad_norm": 0.03928808122873306,
+      "learning_rate": 0.00011072842875202451,
+      "loss": 1.0006,
+      "step": 1242
+    },
+    {
+      "epoch": 1.4018638802598136,
+      "grad_norm": 0.03891727328300476,
+      "learning_rate": 0.00011061033196497738,
+      "loss": 0.9564,
+      "step": 1243
+    },
+    {
+      "epoch": 1.4029935046597006,
+      "grad_norm": 0.03931816667318344,
+      "learning_rate": 0.00011049222020946366,
+      "loss": 0.9873,
+      "step": 1244
+    },
+    {
+      "epoch": 1.4041231290595877,
+      "grad_norm": 0.039088763296604156,
+      "learning_rate": 0.0001103740936521088,
+      "loss": 1.0766,
+      "step": 1245
+    },
+    {
+      "epoch": 1.4052527534594748,
+      "grad_norm": 0.03867960348725319,
+      "learning_rate": 0.00011025595245955917,
+      "loss": 0.9637,
+      "step": 1246
+    },
+    {
+      "epoch": 1.4063823778593618,
+      "grad_norm": 0.03779779002070427,
+      "learning_rate": 0.00011013779679848184,
+      "loss": 0.927,
+      "step": 1247
+    },
+    {
+      "epoch": 1.4075120022592489,
+      "grad_norm": 0.04010738059878349,
+      "learning_rate": 0.00011001962683556421,
+      "loss": 0.8963,
+      "step": 1248
+    },
+    {
+      "epoch": 1.408641626659136,
+      "grad_norm": 0.03821851313114166,
+      "learning_rate": 0.00010990144273751394,
+      "loss": 0.9803,
+      "step": 1249
+    },
+    {
+      "epoch": 1.409771251059023,
+      "grad_norm": 0.03944707289338112,
+      "learning_rate": 0.00010978324467105858,
+      "loss": 0.9464,
+      "step": 1250
+    },
+    {
+      "epoch": 1.4109008754589099,
+      "grad_norm": 0.04029305279254913,
+      "learning_rate": 0.00010966503280294535,
+      "loss": 1.007,
+      "step": 1251
+    },
+    {
+      "epoch": 1.412030499858797,
+      "grad_norm": 0.039815753698349,
+      "learning_rate": 0.00010954680729994102,
+      "loss": 0.9242,
+      "step": 1252
+    },
+    {
+      "epoch": 1.413160124258684,
+      "grad_norm": 0.03888542577624321,
+      "learning_rate": 0.00010942856832883153,
+      "loss": 1.0182,
+      "step": 1253
+    },
+    {
+      "epoch": 1.414289748658571,
+      "grad_norm": 0.040413666516542435,
+      "learning_rate": 0.00010931031605642193,
+      "loss": 0.9702,
+      "step": 1254
+    },
+    {
+      "epoch": 1.415419373058458,
+      "grad_norm": 0.04115651547908783,
+      "learning_rate": 0.00010919205064953582,
+      "loss": 1.0048,
+      "step": 1255
+    },
+    {
+      "epoch": 1.4165489974583452,
+      "grad_norm": 0.03888091444969177,
+      "learning_rate": 0.00010907377227501556,
+      "loss": 1.0556,
+      "step": 1256
+    },
+    {
+      "epoch": 1.417678621858232,
+      "grad_norm": 0.03853550925850868,
+      "learning_rate": 0.0001089554810997217,
+      "loss": 0.9906,
+      "step": 1257
+    },
+    {
+      "epoch": 1.418808246258119,
+      "grad_norm": 0.04010673239827156,
+      "learning_rate": 0.0001088371772905328,
+      "loss": 0.9924,
+      "step": 1258
+    },
+    {
+      "epoch": 1.4199378706580061,
+      "grad_norm": 0.04074999690055847,
+      "learning_rate": 0.00010871886101434536,
+      "loss": 0.9516,
+      "step": 1259
+    },
+    {
+      "epoch": 1.4210674950578932,
+      "grad_norm": 0.03997982293367386,
+      "learning_rate": 0.00010860053243807338,
+      "loss": 1.0788,
+      "step": 1260
+    },
+    {
+      "epoch": 1.4221971194577803,
+      "grad_norm": 0.040215082466602325,
+      "learning_rate": 0.00010848219172864827,
+      "loss": 0.9652,
+      "step": 1261
+    },
+    {
+      "epoch": 1.4233267438576673,
+      "grad_norm": 0.04047797992825508,
+      "learning_rate": 0.00010836383905301844,
+      "loss": 0.9426,
+      "step": 1262
+    },
+    {
+      "epoch": 1.4244563682575544,
+      "grad_norm": 0.039171457290649414,
+      "learning_rate": 0.00010824547457814935,
+      "loss": 0.9483,
+      "step": 1263
+    },
+    {
+      "epoch": 1.4255859926574415,
+      "grad_norm": 0.039281874895095825,
+      "learning_rate": 0.000108127098471023,
+      "loss": 0.9718,
+      "step": 1264
+    },
+    {
+      "epoch": 1.4267156170573285,
+      "grad_norm": 0.03949813172221184,
+      "learning_rate": 0.00010800871089863785,
+      "loss": 1.0681,
+      "step": 1265
+    },
+    {
+      "epoch": 1.4278452414572156,
+      "grad_norm": 0.047373417764902115,
+      "learning_rate": 0.00010789031202800844,
+      "loss": 1.0289,
+      "step": 1266
+    },
+    {
+      "epoch": 1.4289748658571026,
+      "grad_norm": 0.04009164124727249,
+      "learning_rate": 0.00010777190202616535,
+      "loss": 1.0028,
+      "step": 1267
+    },
+    {
+      "epoch": 1.4301044902569895,
+      "grad_norm": 0.039989493787288666,
+      "learning_rate": 0.00010765348106015483,
+      "loss": 0.9449,
+      "step": 1268
+    },
+    {
+      "epoch": 1.4312341146568766,
+      "grad_norm": 0.039491135627031326,
+      "learning_rate": 0.00010753504929703861,
+      "loss": 1.0107,
+      "step": 1269
+    },
+    {
+      "epoch": 1.4323637390567636,
+      "grad_norm": 0.04091602563858032,
+      "learning_rate": 0.00010741660690389365,
+      "loss": 0.9778,
+      "step": 1270
+    },
+    {
+      "epoch": 1.4334933634566507,
+      "grad_norm": 0.03951510414481163,
+      "learning_rate": 0.00010729815404781188,
+      "loss": 0.9379,
+      "step": 1271
+    },
+    {
+      "epoch": 1.4346229878565377,
+      "grad_norm": 0.04040463641285896,
+      "learning_rate": 0.00010717969089589994,
+      "loss": 0.9273,
+      "step": 1272
+    },
+    {
+      "epoch": 1.4357526122564248,
+      "grad_norm": 0.038381997495889664,
+      "learning_rate": 0.00010706121761527919,
+      "loss": 1.1048,
+      "step": 1273
+    },
+    {
+      "epoch": 1.4368822366563117,
+      "grad_norm": 0.04138065129518509,
+      "learning_rate": 0.00010694273437308501,
+      "loss": 0.9641,
+      "step": 1274
+    },
+    {
+      "epoch": 1.4380118610561987,
+      "grad_norm": 0.03996184840798378,
+      "learning_rate": 0.0001068242413364671,
+      "loss": 0.9644,
+      "step": 1275
+    },
+    {
+      "epoch": 1.4391414854560858,
+      "grad_norm": 0.040266938507556915,
+      "learning_rate": 0.0001067057386725888,
+      "loss": 0.9741,
+      "step": 1276
+    },
+    {
+      "epoch": 1.4402711098559728,
+      "grad_norm": 0.03877119719982147,
+      "learning_rate": 0.00010658722654862705,
+      "loss": 1.0218,
+      "step": 1277
+    },
+    {
+      "epoch": 1.44140073425586,
+      "grad_norm": 0.041445109993219376,
+      "learning_rate": 0.00010646870513177221,
+      "loss": 1.0361,
+      "step": 1278
+    },
+    {
+      "epoch": 1.442530358655747,
+      "grad_norm": 0.03916264697909355,
+      "learning_rate": 0.00010635017458922765,
+      "loss": 0.9704,
+      "step": 1279
+    },
+    {
+      "epoch": 1.443659983055634,
+      "grad_norm": 0.03879700228571892,
+      "learning_rate": 0.00010623163508820977,
+      "loss": 0.9553,
+      "step": 1280
+    },
+    {
+      "epoch": 1.444789607455521,
+      "grad_norm": 0.0390193834900856,
+      "learning_rate": 0.0001061130867959474,
+      "loss": 0.9379,
+      "step": 1281
+    },
+    {
+      "epoch": 1.4459192318554082,
+      "grad_norm": 0.040195297449827194,
+      "learning_rate": 0.00010599452987968196,
+      "loss": 0.9409,
+      "step": 1282
+    },
+    {
+      "epoch": 1.4470488562552952,
+      "grad_norm": 0.040100473910570145,
+      "learning_rate": 0.00010587596450666694,
+      "loss": 1.0221,
+      "step": 1283
+    },
+    {
+      "epoch": 1.4481784806551823,
+      "grad_norm": 0.03950706124305725,
+      "learning_rate": 0.00010575739084416778,
+      "loss": 0.9796,
+      "step": 1284
+    },
+    {
+      "epoch": 1.4493081050550691,
+      "grad_norm": 0.03975612670183182,
+      "learning_rate": 0.00010563880905946159,
+      "loss": 0.9528,
+      "step": 1285
+    },
+    {
+      "epoch": 1.4504377294549562,
+      "grad_norm": 0.039496973156929016,
+      "learning_rate": 0.00010552021931983696,
+      "loss": 0.9168,
+      "step": 1286
+    },
+    {
+      "epoch": 1.4515673538548433,
+      "grad_norm": 0.039589397609233856,
+      "learning_rate": 0.00010540162179259373,
+      "loss": 0.9511,
+      "step": 1287
+    },
+    {
+      "epoch": 1.4526969782547303,
+      "grad_norm": 0.03889545425772667,
+      "learning_rate": 0.00010528301664504273,
+      "loss": 0.9859,
+      "step": 1288
+    },
+    {
+      "epoch": 1.4538266026546174,
+      "grad_norm": 0.04123200848698616,
+      "learning_rate": 0.00010516440404450543,
+      "loss": 1.0383,
+      "step": 1289
+    },
+    {
+      "epoch": 1.4549562270545044,
+      "grad_norm": 0.03866191580891609,
+      "learning_rate": 0.00010504578415831395,
+      "loss": 0.9974,
+      "step": 1290
+    },
+    {
+      "epoch": 1.4560858514543913,
+      "grad_norm": 0.03936515003442764,
+      "learning_rate": 0.00010492715715381059,
+      "loss": 0.9545,
+      "step": 1291
+    },
+    {
+      "epoch": 1.4572154758542784,
+      "grad_norm": 0.04141482710838318,
+      "learning_rate": 0.0001048085231983478,
+      "loss": 1.035,
+      "step": 1292
+    },
+    {
+      "epoch": 1.4583451002541654,
+      "grad_norm": 0.03921012952923775,
+      "learning_rate": 0.00010468988245928768,
+      "loss": 0.8901,
+      "step": 1293
+    },
+    {
+      "epoch": 1.4594747246540525,
+      "grad_norm": 0.038768839091062546,
+      "learning_rate": 0.00010457123510400209,
+      "loss": 0.9591,
+      "step": 1294
+    },
+    {
+      "epoch": 1.4606043490539395,
+      "grad_norm": 0.04044971987605095,
+      "learning_rate": 0.00010445258129987206,
+      "loss": 0.9327,
+      "step": 1295
+    },
+    {
+      "epoch": 1.4617339734538266,
+      "grad_norm": 0.04001708701252937,
+      "learning_rate": 0.0001043339212142878,
+      "loss": 0.9784,
+      "step": 1296
+    },
+    {
+      "epoch": 1.4628635978537137,
+      "grad_norm": 0.03915560990571976,
+      "learning_rate": 0.00010421525501464837,
+      "loss": 0.9788,
+      "step": 1297
+    },
+    {
+      "epoch": 1.4639932222536007,
+      "grad_norm": 0.04014548286795616,
+      "learning_rate": 0.00010409658286836143,
+      "loss": 0.95,
+      "step": 1298
+    },
+    {
+      "epoch": 1.4651228466534878,
+      "grad_norm": 0.039338886737823486,
+      "learning_rate": 0.0001039779049428431,
+      "loss": 1.0238,
+      "step": 1299
+    },
+    {
+      "epoch": 1.4662524710533749,
+      "grad_norm": 0.0389542393386364,
+      "learning_rate": 0.00010385922140551752,
+      "loss": 1.0861,
+      "step": 1300
+    },
+    {
+      "epoch": 1.467382095453262,
+      "grad_norm": 0.038751065731048584,
+      "learning_rate": 0.00010374053242381687,
+      "loss": 1.0137,
+      "step": 1301
+    },
+    {
+      "epoch": 1.4685117198531488,
+      "grad_norm": 0.041603147983551025,
+      "learning_rate": 0.00010362183816518101,
+      "loss": 1.0291,
+      "step": 1302
+    },
+    {
+      "epoch": 1.4696413442530358,
+      "grad_norm": 0.04023897647857666,
+      "learning_rate": 0.00010350313879705713,
+      "loss": 0.9787,
+      "step": 1303
+    },
+    {
+      "epoch": 1.470770968652923,
+      "grad_norm": 0.041538018733263016,
+      "learning_rate": 0.00010338443448689976,
+      "loss": 1.0501,
+      "step": 1304
+    },
+    {
+      "epoch": 1.47190059305281,
+      "grad_norm": 0.039176035672426224,
+      "learning_rate": 0.00010326572540217028,
+      "loss": 1.012,
+      "step": 1305
+    },
+    {
+      "epoch": 1.473030217452697,
+      "grad_norm": 0.03936734050512314,
+      "learning_rate": 0.00010314701171033694,
+      "loss": 1.0412,
+      "step": 1306
+    },
+    {
+      "epoch": 1.474159841852584,
+      "grad_norm": 0.04016166925430298,
+      "learning_rate": 0.00010302829357887438,
+      "loss": 0.9524,
+      "step": 1307
+    },
+    {
+      "epoch": 1.475289466252471,
+      "grad_norm": 0.04301619902253151,
+      "learning_rate": 0.00010290957117526354,
+      "loss": 1.0209,
+      "step": 1308
+    },
+    {
+      "epoch": 1.476419090652358,
+      "grad_norm": 0.03983006253838539,
+      "learning_rate": 0.0001027908446669914,
+      "loss": 0.9807,
+      "step": 1309
+    },
+    {
+      "epoch": 1.477548715052245,
+      "grad_norm": 0.0387161448597908,
+      "learning_rate": 0.00010267211422155072,
+      "loss": 1.0256,
+      "step": 1310
+    },
+    {
+      "epoch": 1.4786783394521321,
+      "grad_norm": 0.040488190948963165,
+      "learning_rate": 0.00010255338000643979,
+      "loss": 0.9816,
+      "step": 1311
+    },
+    {
+      "epoch": 1.4798079638520192,
+      "grad_norm": 0.04232305288314819,
+      "learning_rate": 0.00010243464218916226,
+      "loss": 1.0266,
+      "step": 1312
+    },
+    {
+      "epoch": 1.4809375882519062,
+      "grad_norm": 0.039844729006290436,
+      "learning_rate": 0.00010231590093722687,
+      "loss": 0.9535,
+      "step": 1313
+    },
+    {
+      "epoch": 1.4820672126517933,
+      "grad_norm": 0.0414462611079216,
+      "learning_rate": 0.00010219715641814714,
+      "loss": 1.0776,
+      "step": 1314
+    },
+    {
+      "epoch": 1.4831968370516804,
+      "grad_norm": 0.04016523063182831,
+      "learning_rate": 0.00010207840879944123,
+      "loss": 0.9812,
+      "step": 1315
+    },
+    {
+      "epoch": 1.4843264614515674,
+      "grad_norm": 0.039963267743587494,
+      "learning_rate": 0.00010195965824863174,
+      "loss": 0.92,
+      "step": 1316
+    },
+    {
+      "epoch": 1.4854560858514545,
+      "grad_norm": 0.040194012224674225,
+      "learning_rate": 0.00010184090493324527,
+      "loss": 0.9259,
+      "step": 1317
+    },
+    {
+      "epoch": 1.4865857102513416,
+      "grad_norm": 0.03941455855965614,
+      "learning_rate": 0.00010172214902081247,
+      "loss": 1.0022,
+      "step": 1318
+    },
+    {
+      "epoch": 1.4877153346512284,
+      "grad_norm": 0.03942098468542099,
+      "learning_rate": 0.00010160339067886751,
+      "loss": 0.9878,
+      "step": 1319
+    },
+    {
+      "epoch": 1.4888449590511155,
+      "grad_norm": 0.04082900658249855,
+      "learning_rate": 0.0001014846300749481,
+      "loss": 1.0775,
+      "step": 1320
+    },
+    {
+      "epoch": 1.4899745834510025,
+      "grad_norm": 0.040975864976644516,
+      "learning_rate": 0.0001013658673765951,
+      "loss": 0.884,
+      "step": 1321
+    },
+    {
+      "epoch": 1.4911042078508896,
+      "grad_norm": 0.037688951939344406,
+      "learning_rate": 0.00010124710275135236,
+      "loss": 0.9211,
+      "step": 1322
+    },
+    {
+      "epoch": 1.4922338322507767,
+      "grad_norm": 0.038889314979314804,
+      "learning_rate": 0.00010112833636676633,
+      "loss": 0.9202,
+      "step": 1323
+    },
+    {
+      "epoch": 1.4933634566506637,
+      "grad_norm": 0.040478792041540146,
+      "learning_rate": 0.00010100956839038605,
+      "loss": 0.9496,
+      "step": 1324
+    },
+    {
+      "epoch": 1.4944930810505506,
+      "grad_norm": 0.040962960571050644,
+      "learning_rate": 0.00010089079898976284,
+      "loss": 0.9801,
+      "step": 1325
+    },
+    {
+      "epoch": 1.4956227054504376,
+      "grad_norm": 0.04065919294953346,
+      "learning_rate": 0.00010077202833244987,
+      "loss": 0.9985,
+      "step": 1326
+    },
+    {
+      "epoch": 1.4967523298503247,
+      "grad_norm": 0.039901912212371826,
+      "learning_rate": 0.00010065325658600228,
+      "loss": 0.9965,
+      "step": 1327
+    },
+    {
+      "epoch": 1.4978819542502118,
+      "grad_norm": 0.040795568376779556,
+      "learning_rate": 0.00010053448391797657,
+      "loss": 0.9618,
+      "step": 1328
+    },
+    {
+      "epoch": 1.4990115786500988,
+      "grad_norm": 0.03914114087820053,
+      "learning_rate": 0.0001004157104959307,
+      "loss": 0.9579,
+      "step": 1329
+    },
+    {
+      "epoch": 1.5001412030499859,
+      "grad_norm": 0.0392233207821846,
+      "learning_rate": 0.00010029693648742355,
+      "loss": 1.0439,
+      "step": 1330
+    },
+    {
+      "epoch": 1.501270827449873,
+      "grad_norm": 0.04035951942205429,
+      "learning_rate": 0.00010017816206001489,
+      "loss": 0.9729,
+      "step": 1331
+    },
+    {
+      "epoch": 1.50240045184976,
+      "grad_norm": 0.04052640497684479,
+      "learning_rate": 0.0001000593873812651,
+      "loss": 1.0254,
+      "step": 1332
+    },
+    {
+      "epoch": 1.503530076249647,
+      "grad_norm": 0.03904925286769867,
+      "learning_rate": 9.994061261873491e-05,
+      "loss": 1.017,
+      "step": 1333
+    },
+    {
+      "epoch": 1.5046597006495341,
+      "grad_norm": 0.04079846292734146,
+      "learning_rate": 9.98218379399851e-05,
+      "loss": 0.9723,
+      "step": 1334
+    },
+    {
+      "epoch": 1.5057893250494212,
+      "grad_norm": 0.04014943167567253,
+      "learning_rate": 9.970306351257647e-05,
+      "loss": 1.0047,
+      "step": 1335
+    },
+    {
+      "epoch": 1.5069189494493083,
+      "grad_norm": 0.040499597787857056,
+      "learning_rate": 9.958428950406935e-05,
+      "loss": 0.981,
+      "step": 1336
+    },
+    {
+      "epoch": 1.508048573849195,
+      "grad_norm": 0.039357781410217285,
+      "learning_rate": 9.946551608202343e-05,
+      "loss": 1.0248,
+      "step": 1337
+    },
+    {
+      "epoch": 1.5091781982490822,
+      "grad_norm": 0.0400700643658638,
+      "learning_rate": 9.934674341399773e-05,
+      "loss": 0.8825,
+      "step": 1338
+    },
+    {
+      "epoch": 1.5103078226489692,
+      "grad_norm": 0.04014676809310913,
+      "learning_rate": 9.922797166755016e-05,
+      "loss": 0.9863,
+      "step": 1339
+    },
+    {
+      "epoch": 1.5114374470488563,
+      "grad_norm": 0.03986269235610962,
+      "learning_rate": 9.910920101023717e-05,
+      "loss": 0.9944,
+      "step": 1340
+    },
+    {
+      "epoch": 1.5125670714487431,
+      "grad_norm": 0.04043012112379074,
+      "learning_rate": 9.899043160961396e-05,
+      "loss": 1.0018,
+      "step": 1341
+    },
+    {
+      "epoch": 1.5136966958486302,
+      "grad_norm": 0.04049236699938774,
+      "learning_rate": 9.887166363323372e-05,
+      "loss": 1.0049,
+      "step": 1342
+    },
+    {
+      "epoch": 1.5148263202485173,
+      "grad_norm": 0.03971569612622261,
+      "learning_rate": 9.875289724864767e-05,
+      "loss": 0.9351,
+      "step": 1343
+    },
+    {
+      "epoch": 1.5159559446484043,
+      "grad_norm": 0.039951734244823456,
+      "learning_rate": 9.863413262340491e-05,
+      "loss": 0.8937,
+      "step": 1344
+    },
+    {
+      "epoch": 1.5170855690482914,
+      "grad_norm": 0.04067753627896309,
+      "learning_rate": 9.851536992505188e-05,
+      "loss": 0.9653,
+      "step": 1345
+    },
+    {
+      "epoch": 1.5182151934481785,
+      "grad_norm": 0.03986869379878044,
+      "learning_rate": 9.839660932113252e-05,
+      "loss": 0.9987,
+      "step": 1346
+    },
+    {
+      "epoch": 1.5193448178480655,
+      "grad_norm": 0.039932139217853546,
+      "learning_rate": 9.827785097918757e-05,
+      "loss": 0.9733,
+      "step": 1347
+    },
+    {
+      "epoch": 1.5204744422479526,
+      "grad_norm": 0.039806436747312546,
+      "learning_rate": 9.815909506675474e-05,
+      "loss": 0.9847,
+      "step": 1348
+    },
+    {
+      "epoch": 1.5216040666478396,
+      "grad_norm": 0.039268314838409424,
+      "learning_rate": 9.804034175136828e-05,
+      "loss": 1.0022,
+      "step": 1349
+    },
+    {
+      "epoch": 1.5227336910477267,
+      "grad_norm": 0.04103934392333031,
+      "learning_rate": 9.79215912005588e-05,
+      "loss": 0.9476,
+      "step": 1350
+    },
+    {
+      "epoch": 1.5238633154476138,
+      "grad_norm": 0.040989138185977936,
+      "learning_rate": 9.780284358185289e-05,
+      "loss": 0.9352,
+      "step": 1351
+    },
+    {
+      "epoch": 1.5249929398475008,
+      "grad_norm": 0.04021565616130829,
+      "learning_rate": 9.768409906277316e-05,
+      "loss": 0.9293,
+      "step": 1352
+    },
+    {
+      "epoch": 1.526122564247388,
+      "grad_norm": 0.040516503155231476,
+      "learning_rate": 9.756535781083773e-05,
+      "loss": 0.9935,
+      "step": 1353
+    },
+    {
+      "epoch": 1.5272521886472747,
+      "grad_norm": 0.03917067497968674,
+      "learning_rate": 9.744661999356022e-05,
+      "loss": 0.9802,
+      "step": 1354
+    },
+    {
+      "epoch": 1.5283818130471618,
+      "grad_norm": 0.04264604300260544,
+      "learning_rate": 9.732788577844934e-05,
+      "loss": 0.9984,
+      "step": 1355
+    },
+    {
+      "epoch": 1.5295114374470489,
+      "grad_norm": 0.04073306545615196,
+      "learning_rate": 9.720915533300863e-05,
+      "loss": 0.987,
+      "step": 1356
+    },
+    {
+      "epoch": 1.530641061846936,
+      "grad_norm": 0.040012288838624954,
+      "learning_rate": 9.70904288247365e-05,
+      "loss": 0.9858,
+      "step": 1357
+    },
+    {
+      "epoch": 1.5317706862468228,
+      "grad_norm": 0.03969816863536835,
+      "learning_rate": 9.697170642112567e-05,
+      "loss": 0.8985,
+      "step": 1358
+    },
+    {
+      "epoch": 1.5329003106467098,
+      "grad_norm": 0.04010999575257301,
+      "learning_rate": 9.685298828966307e-05,
+      "loss": 1.04,
+      "step": 1359
+    },
+    {
+      "epoch": 1.534029935046597,
+      "grad_norm": 0.0400669127702713,
+      "learning_rate": 9.673427459782974e-05,
+      "loss": 1.0496,
+      "step": 1360
+    },
+    {
+      "epoch": 1.535159559446484,
+      "grad_norm": 0.03961464762687683,
+      "learning_rate": 9.661556551310029e-05,
+      "loss": 1.0093,
+      "step": 1361
+    },
+    {
+      "epoch": 1.536289183846371,
+      "grad_norm": 0.04031101614236832,
+      "learning_rate": 9.64968612029429e-05,
+      "loss": 0.9953,
+      "step": 1362
+    },
+    {
+      "epoch": 1.537418808246258,
+      "grad_norm": 0.0386832058429718,
+      "learning_rate": 9.637816183481903e-05,
+      "loss": 0.9974,
+      "step": 1363
+    },
+    {
+      "epoch": 1.5385484326461452,
+      "grad_norm": 0.04059210419654846,
+      "learning_rate": 9.625946757618311e-05,
+      "loss": 1.0859,
+      "step": 1364
+    },
+    {
+      "epoch": 1.5396780570460322,
+      "grad_norm": 0.039874449372291565,
+      "learning_rate": 9.61407785944825e-05,
+      "loss": 0.9252,
+      "step": 1365
+    },
+    {
+      "epoch": 1.5408076814459193,
+      "grad_norm": 0.040245506912469864,
+      "learning_rate": 9.602209505715695e-05,
+      "loss": 0.9719,
+      "step": 1366
+    },
+    {
+      "epoch": 1.5419373058458063,
+      "grad_norm": 0.0418454185128212,
+      "learning_rate": 9.590341713163858e-05,
+      "loss": 0.9523,
+      "step": 1367
+    },
+    {
+      "epoch": 1.5430669302456934,
+      "grad_norm": 0.041850898414850235,
+      "learning_rate": 9.578474498535165e-05,
+      "loss": 0.9707,
+      "step": 1368
+    },
+    {
+      "epoch": 1.5441965546455805,
+      "grad_norm": 0.04082093760371208,
+      "learning_rate": 9.566607878571225e-05,
+      "loss": 0.9638,
+      "step": 1369
+    },
+    {
+      "epoch": 1.5453261790454675,
+      "grad_norm": 0.040829937905073166,
+      "learning_rate": 9.554741870012797e-05,
+      "loss": 0.9991,
+      "step": 1370
+    },
+    {
+      "epoch": 1.5464558034453544,
+      "grad_norm": 0.04339444637298584,
+      "learning_rate": 9.542876489599795e-05,
+      "loss": 1.0947,
+      "step": 1371
+    },
+    {
+      "epoch": 1.5475854278452414,
+      "grad_norm": 0.040259361267089844,
+      "learning_rate": 9.531011754071232e-05,
+      "loss": 0.9616,
+      "step": 1372
+    },
+    {
+      "epoch": 1.5487150522451285,
+      "grad_norm": 0.04023364931344986,
+      "learning_rate": 9.519147680165224e-05,
+      "loss": 0.9364,
+      "step": 1373
+    },
+    {
+      "epoch": 1.5498446766450156,
+      "grad_norm": 0.03999504819512367,
+      "learning_rate": 9.507284284618945e-05,
+      "loss": 0.9235,
+      "step": 1374
+    },
+    {
+      "epoch": 1.5509743010449024,
+      "grad_norm": 0.03931557759642601,
+      "learning_rate": 9.495421584168609e-05,
+      "loss": 0.9417,
+      "step": 1375
+    },
+    {
+      "epoch": 1.5521039254447895,
+      "grad_norm": 0.040559954941272736,
+      "learning_rate": 9.483559595549461e-05,
+      "loss": 1.0114,
+      "step": 1376
+    },
+    {
+      "epoch": 1.5532335498446765,
+      "grad_norm": 0.040581028908491135,
+      "learning_rate": 9.471698335495731e-05,
+      "loss": 1.0169,
+      "step": 1377
+    },
+    {
+      "epoch": 1.5543631742445636,
+      "grad_norm": 0.04158001020550728,
+      "learning_rate": 9.459837820740626e-05,
+      "loss": 0.9635,
+      "step": 1378
+    },
+    {
+      "epoch": 1.5554927986444507,
+      "grad_norm": 0.0392904132604599,
+      "learning_rate": 9.447978068016307e-05,
+      "loss": 0.9105,
+      "step": 1379
+    },
+    {
+      "epoch": 1.5566224230443377,
+      "grad_norm": 0.04135749489068985,
+      "learning_rate": 9.436119094053846e-05,
+      "loss": 1.004,
+      "step": 1380
+    },
+    {
+      "epoch": 1.5577520474442248,
+      "grad_norm": 0.040026694536209106,
+      "learning_rate": 9.424260915583225e-05,
+      "loss": 1.0355,
+      "step": 1381
+    },
+    {
+      "epoch": 1.5588816718441119,
+      "grad_norm": 0.0401214100420475,
+      "learning_rate": 9.412403549333307e-05,
+      "loss": 0.9886,
+      "step": 1382
+    },
+    {
+      "epoch": 1.560011296243999,
+      "grad_norm": 0.039697665721178055,
+      "learning_rate": 9.400547012031803e-05,
+      "loss": 0.9499,
+      "step": 1383
+    },
+    {
+      "epoch": 1.561140920643886,
+      "grad_norm": 0.03938675299286842,
+      "learning_rate": 9.388691320405262e-05,
+      "loss": 0.9735,
+      "step": 1384
+    },
+    {
+      "epoch": 1.562270545043773,
+      "grad_norm": 0.041321761906147,
+      "learning_rate": 9.376836491179028e-05,
+      "loss": 1.0457,
+      "step": 1385
+    },
+    {
+      "epoch": 1.5634001694436601,
+      "grad_norm": 0.04004902020096779,
+      "learning_rate": 9.364982541077235e-05,
+      "loss": 0.9871,
+      "step": 1386
+    },
+    {
+      "epoch": 1.5645297938435472,
+      "grad_norm": 0.04067892208695412,
+      "learning_rate": 9.353129486822783e-05,
+      "loss": 0.9761,
+      "step": 1387
+    },
+    {
+      "epoch": 1.565659418243434,
+      "grad_norm": 0.04096290096640587,
+      "learning_rate": 9.341277345137301e-05,
+      "loss": 1.023,
+      "step": 1388
+    },
+    {
+      "epoch": 1.566789042643321,
+      "grad_norm": 0.04284733161330223,
+      "learning_rate": 9.329426132741124e-05,
+      "loss": 1.0182,
+      "step": 1389
+    },
+    {
+      "epoch": 1.5679186670432081,
+      "grad_norm": 0.04189145565032959,
+      "learning_rate": 9.317575866353292e-05,
+      "loss": 1.044,
+      "step": 1390
+    },
+    {
+      "epoch": 1.5690482914430952,
+      "grad_norm": 0.04177414998412132,
+      "learning_rate": 9.3057265626915e-05,
+      "loss": 1.0317,
+      "step": 1391
+    },
+    {
+      "epoch": 1.570177915842982,
+      "grad_norm": 0.04260309785604477,
+      "learning_rate": 9.293878238472084e-05,
+      "loss": 0.9835,
+      "step": 1392
+    },
+    {
+      "epoch": 1.5713075402428691,
+      "grad_norm": 0.04053134843707085,
+      "learning_rate": 9.282030910410007e-05,
+      "loss": 1.0568,
+      "step": 1393
+    },
+    {
+      "epoch": 1.5724371646427562,
+      "grad_norm": 0.039543889462947845,
+      "learning_rate": 9.270184595218816e-05,
+      "loss": 0.9351,
+      "step": 1394
+    },
+    {
+      "epoch": 1.5735667890426432,
+      "grad_norm": 0.04078560322523117,
+      "learning_rate": 9.258339309610637e-05,
+      "loss": 0.9573,
+      "step": 1395
+    },
+    {
+      "epoch": 1.5746964134425303,
+      "grad_norm": 0.04066834971308708,
+      "learning_rate": 9.24649507029614e-05,
+      "loss": 0.9305,
+      "step": 1396
+    },
+    {
+      "epoch": 1.5758260378424174,
+      "grad_norm": 0.039552509784698486,
+      "learning_rate": 9.234651893984516e-05,
+      "loss": 0.8974,
+      "step": 1397
+    },
+    {
+      "epoch": 1.5769556622423044,
+      "grad_norm": 0.04099886491894722,
+      "learning_rate": 9.222809797383468e-05,
+      "loss": 0.9938,
+      "step": 1398
+    },
+    {
+      "epoch": 1.5780852866421915,
+      "grad_norm": 0.041135065257549286,
+      "learning_rate": 9.210968797199161e-05,
+      "loss": 0.9801,
+      "step": 1399
+    },
+    {
+      "epoch": 1.5792149110420786,
+      "grad_norm": 0.041974954307079315,
+      "learning_rate": 9.199128910136219e-05,
+      "loss": 0.9009,
+      "step": 1400
+    },
+    {
+      "epoch": 1.5803445354419656,
+      "grad_norm": 0.04135981947183609,
+      "learning_rate": 9.1872901528977e-05,
+      "loss": 0.9265,
+      "step": 1401
+    },
+    {
+      "epoch": 1.5814741598418527,
+      "grad_norm": 0.041727129369974136,
+      "learning_rate": 9.175452542185064e-05,
+      "loss": 1.0046,
+      "step": 1402
+    },
+    {
+      "epoch": 1.5826037842417398,
+      "grad_norm": 0.040945183485746384,
+      "learning_rate": 9.163616094698159e-05,
+      "loss": 1.0032,
+      "step": 1403
+    },
+    {
+      "epoch": 1.5837334086416268,
+      "grad_norm": 0.04023808240890503,
+      "learning_rate": 9.151780827135178e-05,
+      "loss": 1.1407,
+      "step": 1404
+    },
+    {
+      "epoch": 1.5848630330415137,
+      "grad_norm": 0.04090253263711929,
+      "learning_rate": 9.139946756192663e-05,
+      "loss": 0.9573,
+      "step": 1405
+    },
+    {
+      "epoch": 1.5859926574414007,
+      "grad_norm": 0.04100187495350838,
+      "learning_rate": 9.128113898565465e-05,
+      "loss": 0.9929,
+      "step": 1406
+    },
+    {
+      "epoch": 1.5871222818412878,
+      "grad_norm": 0.039680980145931244,
+      "learning_rate": 9.116282270946724e-05,
+      "loss": 1.127,
+      "step": 1407
+    },
+    {
+      "epoch": 1.5882519062411748,
+      "grad_norm": 0.0428316704928875,
+      "learning_rate": 9.104451890027834e-05,
+      "loss": 0.9507,
+      "step": 1408
+    },
+    {
+      "epoch": 1.5893815306410617,
+      "grad_norm": 0.03888211399316788,
+      "learning_rate": 9.092622772498448e-05,
+      "loss": 0.9702,
+      "step": 1409
+    },
+    {
+      "epoch": 1.5905111550409488,
+      "grad_norm": 0.04164845868945122,
+      "learning_rate": 9.080794935046421e-05,
+      "loss": 0.9594,
+      "step": 1410
+    },
+    {
+      "epoch": 1.5916407794408358,
+      "grad_norm": 0.04320342466235161,
+      "learning_rate": 9.068968394357811e-05,
+      "loss": 1.0014,
+      "step": 1411
+    },
+    {
+      "epoch": 1.5927704038407229,
+      "grad_norm": 0.039522089064121246,
+      "learning_rate": 9.057143167116848e-05,
+      "loss": 0.9515,
+      "step": 1412
+    },
+    {
+      "epoch": 1.59390002824061,
+      "grad_norm": 0.03963286057114601,
+      "learning_rate": 9.0453192700059e-05,
+      "loss": 0.9353,
+      "step": 1413
+    },
+    {
+      "epoch": 1.595029652640497,
+      "grad_norm": 0.03974980115890503,
+      "learning_rate": 9.033496719705467e-05,
+      "loss": 0.9351,
+      "step": 1414
+    },
+    {
+      "epoch": 1.596159277040384,
+      "grad_norm": 0.03890387713909149,
+      "learning_rate": 9.021675532894145e-05,
+      "loss": 1.0577,
+      "step": 1415
+    },
+    {
+      "epoch": 1.5972889014402711,
+      "grad_norm": 0.04113437607884407,
+      "learning_rate": 9.009855726248605e-05,
+      "loss": 1.0971,
+      "step": 1416
+    },
+    {
+      "epoch": 1.5984185258401582,
+      "grad_norm": 0.04086640477180481,
+      "learning_rate": 8.99803731644358e-05,
+      "loss": 0.979,
+      "step": 1417
+    },
+    {
+      "epoch": 1.5995481502400453,
+      "grad_norm": 0.04131541773676872,
+      "learning_rate": 8.98622032015182e-05,
+      "loss": 0.9165,
+      "step": 1418
+    },
+    {
+      "epoch": 1.6006777746399323,
+      "grad_norm": 0.04054603353142738,
+      "learning_rate": 8.974404754044085e-05,
+      "loss": 1.0683,
+      "step": 1419
+    },
+    {
+      "epoch": 1.6018073990398194,
+      "grad_norm": 0.039694368839263916,
+      "learning_rate": 8.962590634789123e-05,
+      "loss": 0.9606,
+      "step": 1420
+    },
+    {
+      "epoch": 1.6029370234397065,
+      "grad_norm": 0.0408555269241333,
+      "learning_rate": 8.950777979053635e-05,
+      "loss": 0.8758,
+      "step": 1421
+    },
+    {
+      "epoch": 1.6040666478395933,
+      "grad_norm": 0.04044231399893761,
+      "learning_rate": 8.938966803502264e-05,
+      "loss": 1.0119,
+      "step": 1422
+    },
+    {
+      "epoch": 1.6051962722394804,
+      "grad_norm": 0.03972948342561722,
+      "learning_rate": 8.927157124797552e-05,
+      "loss": 1.0015,
+      "step": 1423
+    },
+    {
+      "epoch": 1.6063258966393674,
+      "grad_norm": 0.04121416062116623,
+      "learning_rate": 8.915348959599941e-05,
+      "loss": 0.93,
+      "step": 1424
+    },
+    {
+      "epoch": 1.6074555210392545,
+      "grad_norm": 0.03931145742535591,
+      "learning_rate": 8.903542324567736e-05,
+      "loss": 1.0265,
+      "step": 1425
+    },
+    {
+      "epoch": 1.6085851454391413,
+      "grad_norm": 0.03970059007406235,
+      "learning_rate": 8.891737236357083e-05,
+      "loss": 1.034,
+      "step": 1426
+    },
+    {
+      "epoch": 1.6097147698390284,
+      "grad_norm": 0.039576053619384766,
+      "learning_rate": 8.879933711621935e-05,
+      "loss": 0.9432,
+      "step": 1427
+    },
+    {
+      "epoch": 1.6108443942389155,
+      "grad_norm": 0.04142745956778526,
+      "learning_rate": 8.868131767014059e-05,
+      "loss": 0.9162,
+      "step": 1428
+    },
+    {
+      "epoch": 1.6119740186388025,
+      "grad_norm": 0.03949185460805893,
+      "learning_rate": 8.856331419182977e-05,
+      "loss": 1.0063,
+      "step": 1429
+    },
+    {
+      "epoch": 1.6131036430386896,
+      "grad_norm": 0.04371139034628868,
+      "learning_rate": 8.844532684775963e-05,
+      "loss": 1.0026,
+      "step": 1430
+    },
+    {
+      "epoch": 1.6142332674385766,
+      "grad_norm": 0.039744243025779724,
+      "learning_rate": 8.832735580438025e-05,
+      "loss": 0.9801,
+      "step": 1431
+    },
+    {
+      "epoch": 1.6153628918384637,
+      "grad_norm": 0.041075777262449265,
+      "learning_rate": 8.820940122811849e-05,
+      "loss": 1.0475,
+      "step": 1432
+    },
+    {
+      "epoch": 1.6164925162383508,
+      "grad_norm": 0.039307381957769394,
+      "learning_rate": 8.809146328537818e-05,
+      "loss": 0.9662,
+      "step": 1433
+    },
+    {
+      "epoch": 1.6176221406382378,
+      "grad_norm": 0.04150167852640152,
+      "learning_rate": 8.797354214253963e-05,
+      "loss": 0.941,
+      "step": 1434
+    },
+    {
+      "epoch": 1.618751765038125,
+      "grad_norm": 0.04133110120892525,
+      "learning_rate": 8.785563796595938e-05,
+      "loss": 0.9079,
+      "step": 1435
+    },
+    {
+      "epoch": 1.619881389438012,
+      "grad_norm": 0.04027198255062103,
+      "learning_rate": 8.773775092197017e-05,
+      "loss": 0.9626,
+      "step": 1436
+    },
+    {
+      "epoch": 1.621011013837899,
+      "grad_norm": 0.04010196402668953,
+      "learning_rate": 8.761988117688041e-05,
+      "loss": 0.9184,
+      "step": 1437
+    },
+    {
+      "epoch": 1.622140638237786,
+      "grad_norm": 0.042807456105947495,
+      "learning_rate": 8.750202889697421e-05,
+      "loss": 1.0578,
+      "step": 1438
+    },
+    {
+      "epoch": 1.623270262637673,
+      "grad_norm": 0.03901375085115433,
+      "learning_rate": 8.738419424851104e-05,
+      "loss": 0.9621,
+      "step": 1439
+    },
+    {
+      "epoch": 1.62439988703756,
+      "grad_norm": 0.04136718809604645,
+      "learning_rate": 8.726637739772542e-05,
+      "loss": 1.0693,
+      "step": 1440
+    },
+    {
+      "epoch": 1.625529511437447,
+      "grad_norm": 0.04196292906999588,
+      "learning_rate": 8.71485785108269e-05,
+      "loss": 1.0302,
+      "step": 1441
+    },
+    {
+      "epoch": 1.6266591358373341,
+      "grad_norm": 0.03940456733107567,
+      "learning_rate": 8.703079775399954e-05,
+      "loss": 0.9193,
+      "step": 1442
+    },
+    {
+      "epoch": 1.627788760237221,
+      "grad_norm": 0.04063562676310539,
+      "learning_rate": 8.691303529340187e-05,
+      "loss": 1.0028,
+      "step": 1443
+    },
+    {
+      "epoch": 1.628918384637108,
+      "grad_norm": 0.041786015033721924,
+      "learning_rate": 8.679529129516667e-05,
+      "loss": 0.9887,
+      "step": 1444
+    },
+    {
+      "epoch": 1.630048009036995,
+      "grad_norm": 0.04099738225340843,
+      "learning_rate": 8.667756592540064e-05,
+      "loss": 0.9782,
+      "step": 1445
+    },
+    {
+      "epoch": 1.6311776334368822,
+      "grad_norm": 0.040208104997873306,
+      "learning_rate": 8.655985935018411e-05,
+      "loss": 0.8863,
+      "step": 1446
+    },
+    {
+      "epoch": 1.6323072578367692,
+      "grad_norm": 0.0400582030415535,
+      "learning_rate": 8.644217173557106e-05,
+      "loss": 1.0338,
+      "step": 1447
+    },
+    {
+      "epoch": 1.6334368822366563,
+      "grad_norm": 0.04026160016655922,
+      "learning_rate": 8.632450324758859e-05,
+      "loss": 0.924,
+      "step": 1448
+    },
+    {
+      "epoch": 1.6345665066365433,
+      "grad_norm": 0.04190199449658394,
+      "learning_rate": 8.620685405223686e-05,
+      "loss": 0.9588,
+      "step": 1449
+    },
+    {
+      "epoch": 1.6356961310364304,
+      "grad_norm": 0.040600717067718506,
+      "learning_rate": 8.608922431548887e-05,
+      "loss": 0.988,
+      "step": 1450
+    },
+    {
+      "epoch": 1.6368257554363175,
+      "grad_norm": 0.03968328237533569,
+      "learning_rate": 8.597161420329e-05,
+      "loss": 1.0247,
+      "step": 1451
+    },
+    {
+      "epoch": 1.6379553798362045,
+      "grad_norm": 0.03925072401762009,
+      "learning_rate": 8.585402388155821e-05,
+      "loss": 0.9461,
+      "step": 1452
+    },
+    {
+      "epoch": 1.6390850042360916,
+      "grad_norm": 0.04031050205230713,
+      "learning_rate": 8.573645351618329e-05,
+      "loss": 0.9915,
+      "step": 1453
+    },
+    {
+      "epoch": 1.6402146286359787,
+      "grad_norm": 0.04113445058465004,
+      "learning_rate": 8.561890327302698e-05,
+      "loss": 0.9443,
+      "step": 1454
+    },
+    {
+      "epoch": 1.6413442530358657,
+      "grad_norm": 0.04077060893177986,
+      "learning_rate": 8.55013733179227e-05,
+      "loss": 1.0212,
+      "step": 1455
+    },
+    {
+      "epoch": 1.6424738774357526,
+      "grad_norm": 0.04069150611758232,
+      "learning_rate": 8.53838638166751e-05,
+      "loss": 1.0852,
+      "step": 1456
+    },
+    {
+      "epoch": 1.6436035018356396,
+      "grad_norm": 0.04216321185231209,
+      "learning_rate": 8.526637493506006e-05,
+      "loss": 1.0119,
+      "step": 1457
+    },
+    {
+      "epoch": 1.6447331262355267,
+      "grad_norm": 0.04014064744114876,
+      "learning_rate": 8.51489068388244e-05,
+      "loss": 1.0884,
+      "step": 1458
+    },
+    {
+      "epoch": 1.6458627506354138,
+      "grad_norm": 0.04238475114107132,
+      "learning_rate": 8.503145969368562e-05,
+      "loss": 1.0737,
+      "step": 1459
+    },
+    {
+      "epoch": 1.6469923750353006,
+      "grad_norm": 0.04255363717675209,
+      "learning_rate": 8.49140336653315e-05,
+      "loss": 0.9817,
+      "step": 1460
+    },
+    {
+      "epoch": 1.6481219994351877,
+      "grad_norm": 0.0400969423353672,
+      "learning_rate": 8.479662891942026e-05,
+      "loss": 0.9089,
+      "step": 1461
+    },
+    {
+      "epoch": 1.6492516238350747,
+      "grad_norm": 0.03944752737879753,
+      "learning_rate": 8.467924562157994e-05,
+      "loss": 1.0311,
+      "step": 1462
+    },
+    {
+      "epoch": 1.6503812482349618,
+      "grad_norm": 0.04170246049761772,
+      "learning_rate": 8.456188393740841e-05,
+      "loss": 1.0078,
+      "step": 1463
+    },
+    {
+      "epoch": 1.6515108726348489,
+      "grad_norm": 0.04236338660120964,
+      "learning_rate": 8.444454403247302e-05,
+      "loss": 0.9292,
+      "step": 1464
+    },
+    {
+      "epoch": 1.652640497034736,
+      "grad_norm": 0.04006924107670784,
+      "learning_rate": 8.43272260723103e-05,
+      "loss": 0.9691,
+      "step": 1465
+    },
+    {
+      "epoch": 1.653770121434623,
+      "grad_norm": 0.03962864354252815,
+      "learning_rate": 8.420993022242602e-05,
+      "loss": 1.025,
+      "step": 1466
+    },
+    {
+      "epoch": 1.65489974583451,
+      "grad_norm": 0.04008259251713753,
+      "learning_rate": 8.409265664829457e-05,
+      "loss": 0.9579,
+      "step": 1467
+    },
+    {
+      "epoch": 1.6560293702343971,
+      "grad_norm": 0.04024090990424156,
+      "learning_rate": 8.3975405515359e-05,
+      "loss": 1.0279,
+      "step": 1468
+    },
+    {
+      "epoch": 1.6571589946342842,
+      "grad_norm": 0.042322732508182526,
+      "learning_rate": 8.385817698903074e-05,
+      "loss": 0.9587,
+      "step": 1469
+    },
+    {
+      "epoch": 1.6582886190341712,
+      "grad_norm": 0.04091969132423401,
+      "learning_rate": 8.374097123468918e-05,
+      "loss": 1.0587,
+      "step": 1470
+    },
+    {
+      "epoch": 1.6594182434340583,
+      "grad_norm": 0.04016717150807381,
+      "learning_rate": 8.362378841768182e-05,
+      "loss": 0.9552,
+      "step": 1471
+    },
+    {
+      "epoch": 1.6605478678339454,
+      "grad_norm": 0.04016115888953209,
+      "learning_rate": 8.350662870332356e-05,
+      "loss": 1.0046,
+      "step": 1472
+    },
+    {
+      "epoch": 1.6616774922338322,
+      "grad_norm": 0.04118018224835396,
+      "learning_rate": 8.338949225689682e-05,
+      "loss": 0.9951,
+      "step": 1473
+    },
+    {
+      "epoch": 1.6628071166337193,
+      "grad_norm": 0.041833024471998215,
+      "learning_rate": 8.327237924365128e-05,
+      "loss": 0.9437,
+      "step": 1474
+    },
+    {
+      "epoch": 1.6639367410336063,
+      "grad_norm": 0.03964128717780113,
+      "learning_rate": 8.315528982880337e-05,
+      "loss": 0.9405,
+      "step": 1475
+    },
+    {
+      "epoch": 1.6639367410336063,
+      "eval_loss": 0.9902071356773376,
+      "eval_runtime": 554.2501,
+      "eval_samples_per_second": 17.651,
+      "eval_steps_per_second": 8.826,
+      "step": 1475
+    },
+    {
+      "epoch": 1.6650663654334934,
+      "grad_norm": 0.04184534028172493,
+      "learning_rate": 8.303822417753634e-05,
+      "loss": 0.9693,
+      "step": 1476
+    },
+    {
+      "epoch": 1.6661959898333802,
+      "grad_norm": 0.040639445185661316,
+      "learning_rate": 8.292118245499996e-05,
+      "loss": 1.026,
+      "step": 1477
+    },
+    {
+      "epoch": 1.6673256142332673,
+      "grad_norm": 0.041504424065351486,
+      "learning_rate": 8.280416482631018e-05,
+      "loss": 1.0356,
+      "step": 1478
+    },
+    {
+      "epoch": 1.6684552386331544,
+      "grad_norm": 0.041965845972299576,
+      "learning_rate": 8.268717145654887e-05,
+      "loss": 0.9386,
+      "step": 1479
+    },
+    {
+      "epoch": 1.6695848630330414,
+      "grad_norm": 0.04126819968223572,
+      "learning_rate": 8.257020251076393e-05,
+      "loss": 0.9731,
+      "step": 1480
+    },
+    {
+      "epoch": 1.6707144874329285,
+      "grad_norm": 0.03976297006011009,
+      "learning_rate": 8.245325815396849e-05,
+      "loss": 0.9571,
+      "step": 1481
+    },
+    {
+      "epoch": 1.6718441118328156,
+      "grad_norm": 0.0409400649368763,
+      "learning_rate": 8.233633855114127e-05,
+      "loss": 0.9836,
+      "step": 1482
+    },
+    {
+      "epoch": 1.6729737362327026,
+      "grad_norm": 0.04177910462021828,
+      "learning_rate": 8.221944386722591e-05,
+      "loss": 0.9716,
+      "step": 1483
+    },
+    {
+      "epoch": 1.6741033606325897,
+      "grad_norm": 0.041480690240859985,
+      "learning_rate": 8.210257426713086e-05,
+      "loss": 0.9171,
+      "step": 1484
+    },
+    {
+      "epoch": 1.6752329850324768,
+      "grad_norm": 0.0406128354370594,
+      "learning_rate": 8.198572991572939e-05,
+      "loss": 0.9579,
+      "step": 1485
+    },
+    {
+      "epoch": 1.6763626094323638,
+      "grad_norm": 0.04042517766356468,
+      "learning_rate": 8.186891097785891e-05,
+      "loss": 0.9316,
+      "step": 1486
+    },
+    {
+      "epoch": 1.6774922338322509,
+      "grad_norm": 0.042964909225702286,
+      "learning_rate": 8.175211761832113e-05,
+      "loss": 0.9642,
+      "step": 1487
+    },
+    {
+      "epoch": 1.678621858232138,
+      "grad_norm": 0.04327230900526047,
+      "learning_rate": 8.163535000188164e-05,
+      "loss": 1.0398,
+      "step": 1488
+    },
+    {
+      "epoch": 1.679751482632025,
+      "grad_norm": 0.04038830101490021,
+      "learning_rate": 8.151860829326962e-05,
+      "loss": 0.9255,
+      "step": 1489
+    },
+    {
+      "epoch": 1.6808811070319118,
+      "grad_norm": 0.04081222414970398,
+      "learning_rate": 8.140189265717794e-05,
+      "loss": 0.9471,
+      "step": 1490
+    },
+    {
+      "epoch": 1.682010731431799,
+      "grad_norm": 0.04002978280186653,
+      "learning_rate": 8.12852032582624e-05,
+      "loss": 0.9539,
+      "step": 1491
+    },
+    {
+      "epoch": 1.683140355831686,
+      "grad_norm": 0.04034026339650154,
+      "learning_rate": 8.116854026114194e-05,
+      "loss": 1.0661,
+      "step": 1492
+    },
+    {
+      "epoch": 1.684269980231573,
+      "grad_norm": 0.0421736054122448,
+      "learning_rate": 8.105190383039828e-05,
+      "loss": 0.9198,
+      "step": 1493
+    },
+    {
+      "epoch": 1.6853996046314599,
+      "grad_norm": 0.04237702116370201,
+      "learning_rate": 8.09352941305756e-05,
+      "loss": 0.9406,
+      "step": 1494
+    },
+    {
+      "epoch": 1.686529229031347,
+      "grad_norm": 0.04037948697805405,
+      "learning_rate": 8.081871132618036e-05,
+      "loss": 0.9974,
+      "step": 1495
+    },
+    {
+      "epoch": 1.687658853431234,
+      "grad_norm": 0.04098232835531235,
+      "learning_rate": 8.070215558168111e-05,
+      "loss": 0.9271,
+      "step": 1496
+    },
+    {
+      "epoch": 1.688788477831121,
+      "grad_norm": 0.04053972661495209,
+      "learning_rate": 8.058562706150823e-05,
+      "loss": 0.9823,
+      "step": 1497
+    },
+    {
+      "epoch": 1.6899181022310081,
+      "grad_norm": 0.041380010545253754,
+      "learning_rate": 8.046912593005361e-05,
+      "loss": 1.078,
+      "step": 1498
+    },
+    {
+      "epoch": 1.6910477266308952,
+      "grad_norm": 0.04158872738480568,
+      "learning_rate": 8.035265235167071e-05,
+      "loss": 0.9958,
+      "step": 1499
+    },
+    {
+      "epoch": 1.6921773510307823,
+      "grad_norm": 0.04147128760814667,
+      "learning_rate": 8.023620649067384e-05,
+      "loss": 0.9634,
+      "step": 1500
+    },
+    {
+      "epoch": 1.6933069754306693,
+      "grad_norm": 0.04079499468207359,
+      "learning_rate": 8.011978851133843e-05,
+      "loss": 0.9581,
+      "step": 1501
+    },
+    {
+      "epoch": 1.6944365998305564,
+      "grad_norm": 0.04090631753206253,
+      "learning_rate": 8.000339857790052e-05,
+      "loss": 0.9927,
+      "step": 1502
+    },
+    {
+      "epoch": 1.6955662242304435,
+      "grad_norm": 0.03968152403831482,
+      "learning_rate": 7.988703685455642e-05,
+      "loss": 0.8988,
+      "step": 1503
+    },
+    {
+      "epoch": 1.6966958486303305,
+      "grad_norm": 0.04115234315395355,
+      "learning_rate": 7.977070350546295e-05,
+      "loss": 1.0064,
+      "step": 1504
+    },
+    {
+      "epoch": 1.6978254730302176,
+      "grad_norm": 0.04148809239268303,
+      "learning_rate": 7.965439869473664e-05,
+      "loss": 1.0416,
+      "step": 1505
+    },
+    {
+      "epoch": 1.6989550974301046,
+      "grad_norm": 0.040575575083494186,
+      "learning_rate": 7.953812258645384e-05,
+      "loss": 1.0512,
+      "step": 1506
+    },
+    {
+      "epoch": 1.7000847218299915,
+      "grad_norm": 0.042429015040397644,
+      "learning_rate": 7.942187534465048e-05,
+      "loss": 0.9684,
+      "step": 1507
+    },
+    {
+      "epoch": 1.7012143462298785,
+      "grad_norm": 0.04083799198269844,
+      "learning_rate": 7.930565713332172e-05,
+      "loss": 0.9619,
+      "step": 1508
+    },
+    {
+      "epoch": 1.7023439706297656,
+      "grad_norm": 0.041108790785074234,
+      "learning_rate": 7.918946811642165e-05,
+      "loss": 1.0179,
+      "step": 1509
+    },
+    {
+      "epoch": 1.7034735950296527,
+      "grad_norm": 0.040565114468336105,
+      "learning_rate": 7.907330845786337e-05,
+      "loss": 1.0207,
+      "step": 1510
+    },
+    {
+      "epoch": 1.7046032194295395,
+      "grad_norm": 0.04313211888074875,
+      "learning_rate": 7.895717832151842e-05,
+      "loss": 0.9691,
+      "step": 1511
+    },
+    {
+      "epoch": 1.7057328438294266,
+      "grad_norm": 0.04209309443831444,
+      "learning_rate": 7.884107787121678e-05,
+      "loss": 1.0266,
+      "step": 1512
+    },
+    {
+      "epoch": 1.7068624682293136,
+      "grad_norm": 0.04235268756747246,
+      "learning_rate": 7.872500727074652e-05,
+      "loss": 1.0672,
+      "step": 1513
+    },
+    {
+      "epoch": 1.7079920926292007,
+      "grad_norm": 0.04129282012581825,
+      "learning_rate": 7.860896668385353e-05,
+      "loss": 1.0046,
+      "step": 1514
+    },
+    {
+      "epoch": 1.7091217170290878,
+      "grad_norm": 0.040298108011484146,
+      "learning_rate": 7.849295627424148e-05,
+      "loss": 1.0131,
+      "step": 1515
+    },
+    {
+      "epoch": 1.7102513414289748,
+      "grad_norm": 0.041110679507255554,
+      "learning_rate": 7.837697620557141e-05,
+      "loss": 0.8495,
+      "step": 1516
+    },
+    {
+      "epoch": 1.711380965828862,
+      "grad_norm": 0.04193076491355896,
+      "learning_rate": 7.826102664146146e-05,
+      "loss": 1.0323,
+      "step": 1517
+    },
+    {
+      "epoch": 1.712510590228749,
+      "grad_norm": 0.041734110563993454,
+      "learning_rate": 7.814510774548696e-05,
+      "loss": 0.9111,
+      "step": 1518
+    },
+    {
+      "epoch": 1.713640214628636,
+      "grad_norm": 0.04231907054781914,
+      "learning_rate": 7.802921968117976e-05,
+      "loss": 1.1051,
+      "step": 1519
+    },
+    {
+      "epoch": 1.714769839028523,
+      "grad_norm": 0.04157987982034683,
+      "learning_rate": 7.791336261202835e-05,
+      "loss": 0.9876,
+      "step": 1520
+    },
+    {
+      "epoch": 1.7158994634284102,
+      "grad_norm": 0.04158562049269676,
+      "learning_rate": 7.779753670147742e-05,
+      "loss": 1.0402,
+      "step": 1521
+    },
+    {
+      "epoch": 1.7170290878282972,
+      "grad_norm": 0.042495615780353546,
+      "learning_rate": 7.768174211292771e-05,
+      "loss": 0.9308,
+      "step": 1522
+    },
+    {
+      "epoch": 1.7181587122281843,
+      "grad_norm": 0.041278909891843796,
+      "learning_rate": 7.756597900973586e-05,
+      "loss": 1.0045,
+      "step": 1523
+    },
+    {
+      "epoch": 1.7192883366280711,
+      "grad_norm": 0.04124930873513222,
+      "learning_rate": 7.745024755521393e-05,
+      "loss": 0.9869,
+      "step": 1524
+    },
+    {
+      "epoch": 1.7204179610279582,
+      "grad_norm": 0.04180390387773514,
+      "learning_rate": 7.733454791262947e-05,
+      "loss": 0.9802,
+      "step": 1525
+    },
+    {
+      "epoch": 1.7215475854278453,
+      "grad_norm": 0.0410812608897686,
+      "learning_rate": 7.72188802452051e-05,
+      "loss": 1.0069,
+      "step": 1526
+    },
+    {
+      "epoch": 1.7226772098277323,
+      "grad_norm": 0.04038868471980095,
+      "learning_rate": 7.710324471611837e-05,
+      "loss": 0.92,
+      "step": 1527
+    },
+    {
+      "epoch": 1.7238068342276192,
+      "grad_norm": 0.04037635400891304,
+      "learning_rate": 7.698764148850137e-05,
+      "loss": 0.9254,
+      "step": 1528
+    },
+    {
+      "epoch": 1.7249364586275062,
+      "grad_norm": 0.041685521602630615,
+      "learning_rate": 7.68720707254408e-05,
+      "loss": 0.9573,
+      "step": 1529
+    },
+    {
+      "epoch": 1.7260660830273933,
+      "grad_norm": 0.040962155908346176,
+      "learning_rate": 7.67565325899774e-05,
+      "loss": 1.0302,
+      "step": 1530
+    },
+    {
+      "epoch": 1.7271957074272803,
+      "grad_norm": 0.04144284501671791,
+      "learning_rate": 7.6641027245106e-05,
+      "loss": 0.9257,
+      "step": 1531
+    },
+    {
+      "epoch": 1.7283253318271674,
+      "grad_norm": 0.04156067594885826,
+      "learning_rate": 7.652555485377515e-05,
+      "loss": 1.0532,
+      "step": 1532
+    },
+    {
+      "epoch": 1.7294549562270545,
+      "grad_norm": 0.041655533015728,
+      "learning_rate": 7.641011557888677e-05,
+      "loss": 1.0718,
+      "step": 1533
+    },
+    {
+      "epoch": 1.7305845806269415,
+      "grad_norm": 0.042078595608472824,
+      "learning_rate": 7.629470958329628e-05,
+      "loss": 0.9245,
+      "step": 1534
+    },
+    {
+      "epoch": 1.7317142050268286,
+      "grad_norm": 0.04095086082816124,
+      "learning_rate": 7.617933702981198e-05,
+      "loss": 1.0378,
+      "step": 1535
+    },
+    {
+      "epoch": 1.7328438294267157,
+      "grad_norm": 0.0403096005320549,
+      "learning_rate": 7.606399808119506e-05,
+      "loss": 1.0323,
+      "step": 1536
+    },
+    {
+      "epoch": 1.7339734538266027,
+      "grad_norm": 0.04328161105513573,
+      "learning_rate": 7.594869290015938e-05,
+      "loss": 0.9129,
+      "step": 1537
+    },
+    {
+      "epoch": 1.7351030782264898,
+      "grad_norm": 0.043508730828762054,
+      "learning_rate": 7.583342164937097e-05,
+      "loss": 1.0068,
+      "step": 1538
+    },
+    {
+      "epoch": 1.7362327026263769,
+      "grad_norm": 0.040654052048921585,
+      "learning_rate": 7.571818449144817e-05,
+      "loss": 1.0229,
+      "step": 1539
+    },
+    {
+      "epoch": 1.737362327026264,
+      "grad_norm": 0.04089539498090744,
+      "learning_rate": 7.560298158896114e-05,
+      "loss": 0.9388,
+      "step": 1540
+    },
+    {
+      "epoch": 1.7384919514261508,
+      "grad_norm": 0.04065493494272232,
+      "learning_rate": 7.548781310443172e-05,
+      "loss": 1.0679,
+      "step": 1541
+    },
+    {
+      "epoch": 1.7396215758260378,
+      "grad_norm": 0.041994400322437286,
+      "learning_rate": 7.537267920033325e-05,
+      "loss": 0.8963,
+      "step": 1542
+    },
+    {
+      "epoch": 1.7407512002259249,
+      "grad_norm": 0.04161880537867546,
+      "learning_rate": 7.525758003909019e-05,
+      "loss": 0.893,
+      "step": 1543
+    },
+    {
+      "epoch": 1.741880824625812,
+      "grad_norm": 0.04373447597026825,
+      "learning_rate": 7.514251578307805e-05,
+      "loss": 1.0542,
+      "step": 1544
+    },
+    {
+      "epoch": 1.7430104490256988,
+      "grad_norm": 0.04112675040960312,
+      "learning_rate": 7.502748659462311e-05,
+      "loss": 0.9247,
+      "step": 1545
+    },
+    {
+      "epoch": 1.7441400734255859,
+      "grad_norm": 0.04262956976890564,
+      "learning_rate": 7.491249263600217e-05,
+      "loss": 0.9339,
+      "step": 1546
+    },
+    {
+      "epoch": 1.745269697825473,
+      "grad_norm": 0.04103526473045349,
+      "learning_rate": 7.479753406944226e-05,
+      "loss": 0.9179,
+      "step": 1547
+    },
+    {
+      "epoch": 1.74639932222536,
+      "grad_norm": 0.040054481476545334,
+      "learning_rate": 7.468261105712058e-05,
+      "loss": 0.9594,
+      "step": 1548
+    },
+    {
+      "epoch": 1.747528946625247,
+      "grad_norm": 0.04210900515317917,
+      "learning_rate": 7.456772376116408e-05,
+      "loss": 0.996,
+      "step": 1549
+    },
+    {
+      "epoch": 1.7486585710251341,
+      "grad_norm": 0.040741559118032455,
+      "learning_rate": 7.445287234364946e-05,
+      "loss": 1.0583,
+      "step": 1550
+    },
+    {
+      "epoch": 1.7497881954250212,
+      "grad_norm": 0.04023589938879013,
+      "learning_rate": 7.433805696660266e-05,
+      "loss": 1.0283,
+      "step": 1551
+    },
+    {
+      "epoch": 1.7509178198249082,
+      "grad_norm": 0.04167277738451958,
+      "learning_rate": 7.42232777919988e-05,
+      "loss": 0.9618,
+      "step": 1552
+    },
+    {
+      "epoch": 1.7520474442247953,
+      "grad_norm": 0.0415283665060997,
+      "learning_rate": 7.410853498176202e-05,
+      "loss": 0.9279,
+      "step": 1553
+    },
+    {
+      "epoch": 1.7531770686246824,
+      "grad_norm": 0.04055981710553169,
+      "learning_rate": 7.399382869776508e-05,
+      "loss": 0.9431,
+      "step": 1554
+    },
+    {
+      "epoch": 1.7543066930245694,
+      "grad_norm": 0.04017677158117294,
+      "learning_rate": 7.38791591018292e-05,
+      "loss": 0.9168,
+      "step": 1555
+    },
+    {
+      "epoch": 1.7554363174244565,
+      "grad_norm": 0.04040129482746124,
+      "learning_rate": 7.376452635572395e-05,
+      "loss": 1.0269,
+      "step": 1556
+    },
+    {
+      "epoch": 1.7565659418243436,
+      "grad_norm": 0.04052029550075531,
+      "learning_rate": 7.364993062116674e-05,
+      "loss": 0.9945,
+      "step": 1557
+    },
+    {
+      "epoch": 1.7576955662242304,
+      "grad_norm": 0.04180474951863289,
+      "learning_rate": 7.353537205982294e-05,
+      "loss": 0.9983,
+      "step": 1558
+    },
+    {
+      "epoch": 1.7588251906241175,
+      "grad_norm": 0.04247574508190155,
+      "learning_rate": 7.342085083330537e-05,
+      "loss": 1.051,
+      "step": 1559
+    },
+    {
+      "epoch": 1.7599548150240045,
+      "grad_norm": 0.04135637730360031,
+      "learning_rate": 7.330636710317417e-05,
+      "loss": 0.9702,
+      "step": 1560
+    },
+    {
+      "epoch": 1.7610844394238916,
+      "grad_norm": 0.04159768298268318,
+      "learning_rate": 7.319192103093672e-05,
+      "loss": 1.0098,
+      "step": 1561
+    },
+    {
+      "epoch": 1.7622140638237784,
+      "grad_norm": 0.04054262861609459,
+      "learning_rate": 7.307751277804711e-05,
+      "loss": 0.9221,
+      "step": 1562
+    },
+    {
+      "epoch": 1.7633436882236655,
+      "grad_norm": 0.042798902839422226,
+      "learning_rate": 7.296314250590615e-05,
+      "loss": 0.9744,
+      "step": 1563
+    },
+    {
+      "epoch": 1.7644733126235526,
+      "grad_norm": 0.04235600307583809,
+      "learning_rate": 7.284881037586107e-05,
+      "loss": 0.9744,
+      "step": 1564
+    },
+    {
+      "epoch": 1.7656029370234396,
+      "grad_norm": 0.040483228862285614,
+      "learning_rate": 7.273451654920532e-05,
+      "loss": 0.9765,
+      "step": 1565
+    },
+    {
+      "epoch": 1.7667325614233267,
+      "grad_norm": 0.04099193215370178,
+      "learning_rate": 7.26202611871782e-05,
+      "loss": 1.0124,
+      "step": 1566
+    },
+    {
+      "epoch": 1.7678621858232137,
+      "grad_norm": 0.04022025689482689,
+      "learning_rate": 7.250604445096487e-05,
+      "loss": 0.9559,
+      "step": 1567
+    },
+    {
+      "epoch": 1.7689918102231008,
+      "grad_norm": 0.040389448404312134,
+      "learning_rate": 7.239186650169596e-05,
+      "loss": 0.9435,
+      "step": 1568
+    },
+    {
+      "epoch": 1.7701214346229879,
+      "grad_norm": 0.04212417080998421,
+      "learning_rate": 7.227772750044739e-05,
+      "loss": 0.9733,
+      "step": 1569
+    },
+    {
+      "epoch": 1.771251059022875,
+      "grad_norm": 0.03963673487305641,
+      "learning_rate": 7.21636276082401e-05,
+      "loss": 1.0231,
+      "step": 1570
+    },
+    {
+      "epoch": 1.772380683422762,
+      "grad_norm": 0.04181777685880661,
+      "learning_rate": 7.204956698603984e-05,
+      "loss": 0.9407,
+      "step": 1571
+    },
+    {
+      "epoch": 1.773510307822649,
+      "grad_norm": 0.04021856188774109,
+      "learning_rate": 7.193554579475705e-05,
+      "loss": 0.9671,
+      "step": 1572
+    },
+    {
+      "epoch": 1.7746399322225361,
+      "grad_norm": 0.044537365436553955,
+      "learning_rate": 7.182156419524649e-05,
+      "loss": 1.0034,
+      "step": 1573
+    },
+    {
+      "epoch": 1.7757695566224232,
+      "grad_norm": 0.04181935638189316,
+      "learning_rate": 7.170762234830699e-05,
+      "loss": 0.9827,
+      "step": 1574
+    },
+    {
+      "epoch": 1.77689918102231,
+      "grad_norm": 0.04236932843923569,
+      "learning_rate": 7.15937204146815e-05,
+      "loss": 0.9793,
+      "step": 1575
+    },
+    {
+      "epoch": 1.778028805422197,
+      "grad_norm": 0.041638486087322235,
+      "learning_rate": 7.147985855505643e-05,
+      "loss": 0.9354,
+      "step": 1576
+    },
+    {
+      "epoch": 1.7791584298220842,
+      "grad_norm": 0.04149574786424637,
+      "learning_rate": 7.136603693006179e-05,
+      "loss": 0.9866,
+      "step": 1577
+    },
+    {
+      "epoch": 1.7802880542219712,
+      "grad_norm": 0.04303409904241562,
+      "learning_rate": 7.125225570027083e-05,
+      "loss": 0.9862,
+      "step": 1578
+    },
+    {
+      "epoch": 1.781417678621858,
+      "grad_norm": 0.04238145053386688,
+      "learning_rate": 7.113851502619974e-05,
+      "loss": 0.9455,
+      "step": 1579
+    },
+    {
+      "epoch": 1.7825473030217451,
+      "grad_norm": 0.04182044416666031,
+      "learning_rate": 7.102481506830764e-05,
+      "loss": 0.9714,
+      "step": 1580
+    },
+    {
+      "epoch": 1.7836769274216322,
+      "grad_norm": 0.04323654994368553,
+      "learning_rate": 7.0911155986996e-05,
+      "loss": 1.0505,
+      "step": 1581
+    },
+    {
+      "epoch": 1.7848065518215193,
+      "grad_norm": 0.04246308654546738,
+      "learning_rate": 7.079753794260876e-05,
+      "loss": 0.9668,
+      "step": 1582
+    },
+    {
+      "epoch": 1.7859361762214063,
+      "grad_norm": 0.042159292846918106,
+      "learning_rate": 7.068396109543199e-05,
+      "loss": 1.0844,
+      "step": 1583
+    },
+    {
+      "epoch": 1.7870658006212934,
+      "grad_norm": 0.04247213155031204,
+      "learning_rate": 7.057042560569356e-05,
+      "loss": 1.0371,
+      "step": 1584
+    },
+    {
+      "epoch": 1.7881954250211805,
+      "grad_norm": 0.04189344495534897,
+      "learning_rate": 7.0456931633563e-05,
+      "loss": 1.0026,
+      "step": 1585
+    },
+    {
+      "epoch": 1.7893250494210675,
+      "grad_norm": 0.040689561516046524,
+      "learning_rate": 7.034347933915135e-05,
+      "loss": 0.9506,
+      "step": 1586
+    },
+    {
+      "epoch": 1.7904546738209546,
+      "grad_norm": 0.04134576767683029,
+      "learning_rate": 7.023006888251076e-05,
+      "loss": 0.96,
+      "step": 1587
+    },
+    {
+      "epoch": 1.7915842982208416,
+      "grad_norm": 0.03989570587873459,
+      "learning_rate": 7.011670042363445e-05,
+      "loss": 0.9603,
+      "step": 1588
+    },
+    {
+      "epoch": 1.7927139226207287,
+      "grad_norm": 0.04224357753992081,
+      "learning_rate": 7.000337412245632e-05,
+      "loss": 0.9133,
+      "step": 1589
+    },
+    {
+      "epoch": 1.7938435470206158,
+      "grad_norm": 0.04070102795958519,
+      "learning_rate": 6.989009013885077e-05,
+      "loss": 1.0841,
+      "step": 1590
+    },
+    {
+      "epoch": 1.7949731714205028,
+      "grad_norm": 0.04164205491542816,
+      "learning_rate": 6.977684863263261e-05,
+      "loss": 0.9859,
+      "step": 1591
+    },
+    {
+      "epoch": 1.7961027958203897,
+      "grad_norm": 0.040442369878292084,
+      "learning_rate": 6.966364976355664e-05,
+      "loss": 1.0549,
+      "step": 1592
+    },
+    {
+      "epoch": 1.7972324202202767,
+      "grad_norm": 0.04000772163271904,
+      "learning_rate": 6.955049369131749e-05,
+      "loss": 1.0889,
+      "step": 1593
+    },
+    {
+      "epoch": 1.7983620446201638,
+      "grad_norm": 0.0411187969148159,
+      "learning_rate": 6.943738057554957e-05,
+      "loss": 1.0049,
+      "step": 1594
+    },
+    {
+      "epoch": 1.7994916690200509,
+      "grad_norm": 0.041261740028858185,
+      "learning_rate": 6.932431057582647e-05,
+      "loss": 0.9334,
+      "step": 1595
+    },
+    {
+      "epoch": 1.8006212934199377,
+      "grad_norm": 0.041718561202287674,
+      "learning_rate": 6.921128385166109e-05,
+      "loss": 1.0152,
+      "step": 1596
+    },
+    {
+      "epoch": 1.8017509178198248,
+      "grad_norm": 0.042475439608097076,
+      "learning_rate": 6.909830056250527e-05,
+      "loss": 0.923,
+      "step": 1597
+    },
+    {
+      "epoch": 1.8028805422197118,
+      "grad_norm": 0.040802422910928726,
+      "learning_rate": 6.898536086774952e-05,
+      "loss": 0.9905,
+      "step": 1598
+    },
+    {
+      "epoch": 1.804010166619599,
+      "grad_norm": 0.04284251853823662,
+      "learning_rate": 6.887246492672296e-05,
+      "loss": 0.9836,
+      "step": 1599
+    },
+    {
+      "epoch": 1.805139791019486,
+      "grad_norm": 0.04458894580602646,
+      "learning_rate": 6.875961289869283e-05,
+      "loss": 0.9887,
+      "step": 1600
+    },
+    {
+      "epoch": 1.806269415419373,
+      "grad_norm": 0.04009930416941643,
+      "learning_rate": 6.864680494286451e-05,
+      "loss": 0.9395,
+      "step": 1601
+    },
+    {
+      "epoch": 1.80739903981926,
+      "grad_norm": 0.0409211665391922,
+      "learning_rate": 6.853404121838121e-05,
+      "loss": 0.954,
+      "step": 1602
+    },
+    {
+      "epoch": 1.8085286642191472,
+      "grad_norm": 0.04242323338985443,
+      "learning_rate": 6.842132188432376e-05,
+      "loss": 0.9966,
+      "step": 1603
+    },
+    {
+      "epoch": 1.8096582886190342,
+      "grad_norm": 0.04248444736003876,
+      "learning_rate": 6.830864709971027e-05,
+      "loss": 0.9439,
+      "step": 1604
+    },
+    {
+      "epoch": 1.8107879130189213,
+      "grad_norm": 0.041173093020915985,
+      "learning_rate": 6.819601702349609e-05,
+      "loss": 0.8787,
+      "step": 1605
+    },
+    {
+      "epoch": 1.8119175374188083,
+      "grad_norm": 0.042021602392196655,
+      "learning_rate": 6.808343181457346e-05,
+      "loss": 1.0753,
+      "step": 1606
+    },
+    {
+      "epoch": 1.8130471618186954,
+      "grad_norm": 0.04070396348834038,
+      "learning_rate": 6.797089163177137e-05,
+      "loss": 0.9971,
+      "step": 1607
+    },
+    {
+      "epoch": 1.8141767862185825,
+      "grad_norm": 0.04318329691886902,
+      "learning_rate": 6.785839663385526e-05,
+      "loss": 1.0268,
+      "step": 1608
+    },
+    {
+      "epoch": 1.8153064106184693,
+      "grad_norm": 0.04180993512272835,
+      "learning_rate": 6.774594697952672e-05,
+      "loss": 0.9524,
+      "step": 1609
+    },
+    {
+      "epoch": 1.8164360350183564,
+      "grad_norm": 0.042706914246082306,
+      "learning_rate": 6.763354282742363e-05,
+      "loss": 1.0231,
+      "step": 1610
+    },
+    {
+      "epoch": 1.8175656594182434,
+      "grad_norm": 0.03994883596897125,
+      "learning_rate": 6.752118433611939e-05,
+      "loss": 0.9749,
+      "step": 1611
+    },
+    {
+      "epoch": 1.8186952838181305,
+      "grad_norm": 0.042355459183454514,
+      "learning_rate": 6.740887166412315e-05,
+      "loss": 1.0237,
+      "step": 1612
+    },
+    {
+      "epoch": 1.8198249082180173,
+      "grad_norm": 0.04314768686890602,
+      "learning_rate": 6.729660496987944e-05,
+      "loss": 1.0333,
+      "step": 1613
+    },
+    {
+      "epoch": 1.8209545326179044,
+      "grad_norm": 0.043471548706293106,
+      "learning_rate": 6.718438441176781e-05,
+      "loss": 1.0369,
+      "step": 1614
+    },
+    {
+      "epoch": 1.8220841570177915,
+      "grad_norm": 0.040395159274339676,
+      "learning_rate": 6.707221014810279e-05,
+      "loss": 0.9598,
+      "step": 1615
+    },
+    {
+      "epoch": 1.8232137814176785,
+      "grad_norm": 0.04183076322078705,
+      "learning_rate": 6.696008233713362e-05,
+      "loss": 1.0045,
+      "step": 1616
+    },
+    {
+      "epoch": 1.8243434058175656,
+      "grad_norm": 0.04115899279713631,
+      "learning_rate": 6.684800113704397e-05,
+      "loss": 0.9235,
+      "step": 1617
+    },
+    {
+      "epoch": 1.8254730302174527,
+      "grad_norm": 0.04163803160190582,
+      "learning_rate": 6.673596670595181e-05,
+      "loss": 1.0222,
+      "step": 1618
+    },
+    {
+      "epoch": 1.8266026546173397,
+      "grad_norm": 0.042601440101861954,
+      "learning_rate": 6.662397920190904e-05,
+      "loss": 1.0039,
+      "step": 1619
+    },
+    {
+      "epoch": 1.8277322790172268,
+      "grad_norm": 0.04220454394817352,
+      "learning_rate": 6.651203878290139e-05,
+      "loss": 0.9655,
+      "step": 1620
+    },
+    {
+      "epoch": 1.8288619034171139,
+      "grad_norm": 0.04085693880915642,
+      "learning_rate": 6.640014560684824e-05,
+      "loss": 1.0043,
+      "step": 1621
+    },
+    {
+      "epoch": 1.829991527817001,
+      "grad_norm": 0.04273395612835884,
+      "learning_rate": 6.628829983160225e-05,
+      "loss": 0.9698,
+      "step": 1622
+    },
+    {
+      "epoch": 1.831121152216888,
+      "grad_norm": 0.04279292747378349,
+      "learning_rate": 6.617650161494915e-05,
+      "loss": 0.8682,
+      "step": 1623
+    },
+    {
+      "epoch": 1.832250776616775,
+      "grad_norm": 0.04293496534228325,
+      "learning_rate": 6.606475111460776e-05,
+      "loss": 1.0443,
+      "step": 1624
+    },
+    {
+      "epoch": 1.833380401016662,
+      "grad_norm": 0.042040567845106125,
+      "learning_rate": 6.59530484882294e-05,
+      "loss": 0.9817,
+      "step": 1625
+    },
+    {
+      "epoch": 1.834510025416549,
+      "grad_norm": 0.04191702604293823,
+      "learning_rate": 6.584139389339796e-05,
+      "loss": 1.0594,
+      "step": 1626
+    },
+    {
+      "epoch": 1.835639649816436,
+      "grad_norm": 0.0418466292321682,
+      "learning_rate": 6.572978748762954e-05,
+      "loss": 0.9879,
+      "step": 1627
+    },
+    {
+      "epoch": 1.836769274216323,
+      "grad_norm": 0.0425787977874279,
+      "learning_rate": 6.56182294283722e-05,
+      "loss": 0.9546,
+      "step": 1628
+    },
+    {
+      "epoch": 1.8378988986162101,
+      "grad_norm": 0.041348233819007874,
+      "learning_rate": 6.550671987300594e-05,
+      "loss": 0.964,
+      "step": 1629
+    },
+    {
+      "epoch": 1.839028523016097,
+      "grad_norm": 0.04128558188676834,
+      "learning_rate": 6.539525897884219e-05,
+      "loss": 0.9743,
+      "step": 1630
+    },
+    {
+      "epoch": 1.840158147415984,
+      "grad_norm": 0.042008284479379654,
+      "learning_rate": 6.528384690312375e-05,
+      "loss": 0.9439,
+      "step": 1631
+    },
+    {
+      "epoch": 1.841287771815871,
+      "grad_norm": 0.042078789323568344,
+      "learning_rate": 6.517248380302469e-05,
+      "loss": 0.9519,
+      "step": 1632
+    },
+    {
+      "epoch": 1.8424173962157582,
+      "grad_norm": 0.04227921739220619,
+      "learning_rate": 6.506116983564978e-05,
+      "loss": 0.992,
+      "step": 1633
+    },
+    {
+      "epoch": 1.8435470206156452,
+      "grad_norm": 0.04202863574028015,
+      "learning_rate": 6.49499051580346e-05,
+      "loss": 0.9948,
+      "step": 1634
+    },
+    {
+      "epoch": 1.8446766450155323,
+      "grad_norm": 0.04132537916302681,
+      "learning_rate": 6.48386899271452e-05,
+      "loss": 1.0068,
+      "step": 1635
+    },
+    {
+      "epoch": 1.8458062694154194,
+      "grad_norm": 0.04138810560107231,
+      "learning_rate": 6.472752429987782e-05,
+      "loss": 0.9869,
+      "step": 1636
+    },
+    {
+      "epoch": 1.8469358938153064,
+      "grad_norm": 0.04405840113759041,
+      "learning_rate": 6.461640843305878e-05,
+      "loss": 0.9432,
+      "step": 1637
+    },
+    {
+      "epoch": 1.8480655182151935,
+      "grad_norm": 0.04127390310168266,
+      "learning_rate": 6.450534248344417e-05,
+      "loss": 0.9554,
+      "step": 1638
+    },
+    {
+      "epoch": 1.8491951426150806,
+      "grad_norm": 0.04151296615600586,
+      "learning_rate": 6.439432660771962e-05,
+      "loss": 0.9998,
+      "step": 1639
+    },
+    {
+      "epoch": 1.8503247670149676,
+      "grad_norm": 0.04165178909897804,
+      "learning_rate": 6.428336096250019e-05,
+      "loss": 0.9282,
+      "step": 1640
+    },
+    {
+      "epoch": 1.8514543914148547,
+      "grad_norm": 0.04152395576238632,
+      "learning_rate": 6.417244570433005e-05,
+      "loss": 0.9607,
+      "step": 1641
+    },
+    {
+      "epoch": 1.8525840158147417,
+      "grad_norm": 0.041522059589624405,
+      "learning_rate": 6.406158098968221e-05,
+      "loss": 1.0155,
+      "step": 1642
+    },
+    {
+      "epoch": 1.8537136402146286,
+      "grad_norm": 0.041412562131881714,
+      "learning_rate": 6.395076697495854e-05,
+      "loss": 0.9229,
+      "step": 1643
+    },
+    {
+      "epoch": 1.8548432646145157,
+      "grad_norm": 0.042269017547369,
+      "learning_rate": 6.384000381648926e-05,
+      "loss": 0.9884,
+      "step": 1644
+    },
+    {
+      "epoch": 1.8559728890144027,
+      "grad_norm": 0.04206148535013199,
+      "learning_rate": 6.372929167053286e-05,
+      "loss": 0.9337,
+      "step": 1645
+    },
+    {
+      "epoch": 1.8571025134142898,
+      "grad_norm": 0.04192545637488365,
+      "learning_rate": 6.361863069327591e-05,
+      "loss": 1.0505,
+      "step": 1646
+    },
+    {
+      "epoch": 1.8582321378141766,
+      "grad_norm": 0.04114922136068344,
+      "learning_rate": 6.350802104083271e-05,
+      "loss": 0.9875,
+      "step": 1647
+    },
+    {
+      "epoch": 1.8593617622140637,
+      "grad_norm": 0.04190947860479355,
+      "learning_rate": 6.339746286924531e-05,
+      "loss": 0.9939,
+      "step": 1648
+    },
+    {
+      "epoch": 1.8604913866139507,
+      "grad_norm": 0.043601252138614655,
+      "learning_rate": 6.328695633448296e-05,
+      "loss": 1.061,
+      "step": 1649
+    },
+    {
+      "epoch": 1.8616210110138378,
+      "grad_norm": 0.042812932282686234,
+      "learning_rate": 6.317650159244212e-05,
+      "loss": 0.9783,
+      "step": 1650
+    },
+    {
+      "epoch": 1.8627506354137249,
+      "grad_norm": 0.04309513419866562,
+      "learning_rate": 6.306609879894627e-05,
+      "loss": 0.9543,
+      "step": 1651
+    },
+    {
+      "epoch": 1.863880259813612,
+      "grad_norm": 0.04289943352341652,
+      "learning_rate": 6.295574810974552e-05,
+      "loss": 0.9788,
+      "step": 1652
+    },
+    {
+      "epoch": 1.865009884213499,
+      "grad_norm": 0.04318351671099663,
+      "learning_rate": 6.284544968051643e-05,
+      "loss": 0.9669,
+      "step": 1653
+    },
+    {
+      "epoch": 1.866139508613386,
+      "grad_norm": 0.04084068536758423,
+      "learning_rate": 6.273520366686195e-05,
+      "loss": 0.905,
+      "step": 1654
+    },
+    {
+      "epoch": 1.8672691330132731,
+      "grad_norm": 0.04186054319143295,
+      "learning_rate": 6.2625010224311e-05,
+      "loss": 0.9222,
+      "step": 1655
+    },
+    {
+      "epoch": 1.8683987574131602,
+      "grad_norm": 0.04201950505375862,
+      "learning_rate": 6.251486950831838e-05,
+      "loss": 0.9293,
+      "step": 1656
+    },
+    {
+      "epoch": 1.8695283818130473,
+      "grad_norm": 0.043176308274269104,
+      "learning_rate": 6.240478167426451e-05,
+      "loss": 1.0731,
+      "step": 1657
+    },
+    {
+      "epoch": 1.8706580062129343,
+      "grad_norm": 0.042063016444444656,
+      "learning_rate": 6.229474687745513e-05,
+      "loss": 1.0344,
+      "step": 1658
+    },
+    {
+      "epoch": 1.8717876306128214,
+      "grad_norm": 0.041596777737140656,
+      "learning_rate": 6.218476527312127e-05,
+      "loss": 0.9725,
+      "step": 1659
+    },
+    {
+      "epoch": 1.8729172550127082,
+      "grad_norm": 0.04136804863810539,
+      "learning_rate": 6.207483701641888e-05,
+      "loss": 0.9089,
+      "step": 1660
+    },
+    {
+      "epoch": 1.8740468794125953,
+      "grad_norm": 0.04120111092925072,
+      "learning_rate": 6.196496226242852e-05,
+      "loss": 1.0282,
+      "step": 1661
+    },
+    {
+      "epoch": 1.8751765038124824,
+      "grad_norm": 0.04079049080610275,
+      "learning_rate": 6.185514116615553e-05,
+      "loss": 0.9121,
+      "step": 1662
+    },
+    {
+      "epoch": 1.8763061282123694,
+      "grad_norm": 0.041485901921987534,
+      "learning_rate": 6.174537388252932e-05,
+      "loss": 1.017,
+      "step": 1663
+    },
+    {
+      "epoch": 1.8774357526122563,
+      "grad_norm": 0.04155506566166878,
+      "learning_rate": 6.163566056640349e-05,
+      "loss": 1.0028,
+      "step": 1664
+    },
+    {
+      "epoch": 1.8785653770121433,
+      "grad_norm": 0.041246477514505386,
+      "learning_rate": 6.15260013725555e-05,
+      "loss": 1.0494,
+      "step": 1665
+    },
+    {
+      "epoch": 1.8796950014120304,
+      "grad_norm": 0.04152505099773407,
+      "learning_rate": 6.141639645568646e-05,
+      "loss": 0.9382,
+      "step": 1666
+    },
+    {
+      "epoch": 1.8808246258119174,
+      "grad_norm": 0.03967185318470001,
+      "learning_rate": 6.130684597042088e-05,
+      "loss": 0.9703,
+      "step": 1667
+    },
+    {
+      "epoch": 1.8819542502118045,
+      "grad_norm": 0.04061749577522278,
+      "learning_rate": 6.119735007130649e-05,
+      "loss": 0.9481,
+      "step": 1668
+    },
+    {
+      "epoch": 1.8830838746116916,
+      "grad_norm": 0.04188203811645508,
+      "learning_rate": 6.1087908912814e-05,
+      "loss": 0.9199,
+      "step": 1669
+    },
+    {
+      "epoch": 1.8842134990115786,
+      "grad_norm": 0.041013531386852264,
+      "learning_rate": 6.097852264933697e-05,
+      "loss": 0.8728,
+      "step": 1670
+    },
+    {
+      "epoch": 1.8853431234114657,
+      "grad_norm": 0.04213571920990944,
+      "learning_rate": 6.086919143519143e-05,
+      "loss": 0.9471,
+      "step": 1671
+    },
+    {
+      "epoch": 1.8864727478113528,
+      "grad_norm": 0.04296332970261574,
+      "learning_rate": 6.075991542461574e-05,
+      "loss": 0.9888,
+      "step": 1672
+    },
+    {
+      "epoch": 1.8876023722112398,
+      "grad_norm": 0.040672652423381805,
+      "learning_rate": 6.065069477177048e-05,
+      "loss": 0.9785,
+      "step": 1673
+    },
+    {
+      "epoch": 1.888731996611127,
+      "grad_norm": 0.041439056396484375,
+      "learning_rate": 6.054152963073809e-05,
+      "loss": 0.9904,
+      "step": 1674
+    },
+    {
+      "epoch": 1.889861621011014,
+      "grad_norm": 0.04170332849025726,
+      "learning_rate": 6.043242015552258e-05,
+      "loss": 0.9502,
+      "step": 1675
+    },
+    {
+      "epoch": 1.890991245410901,
+      "grad_norm": 0.043845389038324356,
+      "learning_rate": 6.032336650004966e-05,
+      "loss": 0.9832,
+      "step": 1676
+    },
+    {
+      "epoch": 1.8921208698107879,
+      "grad_norm": 0.042602553963661194,
+      "learning_rate": 6.021436881816608e-05,
+      "loss": 1.0506,
+      "step": 1677
+    },
+    {
+      "epoch": 1.893250494210675,
+      "grad_norm": 0.04084152728319168,
+      "learning_rate": 6.010542726363976e-05,
+      "loss": 0.9826,
+      "step": 1678
+    },
+    {
+      "epoch": 1.894380118610562,
+      "grad_norm": 0.04469555988907814,
+      "learning_rate": 5.9996541990159383e-05,
+      "loss": 1.0022,
+      "step": 1679
+    },
+    {
+      "epoch": 1.895509743010449,
+      "grad_norm": 0.04405542463064194,
+      "learning_rate": 5.988771315133418e-05,
+      "loss": 0.9563,
+      "step": 1680
+    },
+    {
+      "epoch": 1.896639367410336,
+      "grad_norm": 0.041179101914167404,
+      "learning_rate": 5.9778940900693935e-05,
+      "loss": 0.971,
+      "step": 1681
+    },
+    {
+      "epoch": 1.897768991810223,
+      "grad_norm": 0.04140183702111244,
+      "learning_rate": 5.967022539168843e-05,
+      "loss": 0.8642,
+      "step": 1682
+    },
+    {
+      "epoch": 1.89889861621011,
+      "grad_norm": 0.04198272526264191,
+      "learning_rate": 5.9561566777687427e-05,
+      "loss": 0.9716,
+      "step": 1683
+    },
+    {
+      "epoch": 1.900028240609997,
+      "grad_norm": 0.042881209403276443,
+      "learning_rate": 5.945296521198054e-05,
+      "loss": 0.9638,
+      "step": 1684
+    },
+    {
+      "epoch": 1.9011578650098842,
+      "grad_norm": 0.042486660182476044,
+      "learning_rate": 5.934442084777676e-05,
+      "loss": 0.9335,
+      "step": 1685
+    },
+    {
+      "epoch": 1.9022874894097712,
+      "grad_norm": 0.04221144691109657,
+      "learning_rate": 5.9235933838204516e-05,
+      "loss": 0.9281,
+      "step": 1686
+    },
+    {
+      "epoch": 1.9034171138096583,
+      "grad_norm": 0.04329012706875801,
+      "learning_rate": 5.9127504336311176e-05,
+      "loss": 1.0478,
+      "step": 1687
+    },
+    {
+      "epoch": 1.9045467382095453,
+      "grad_norm": 0.04217381402850151,
+      "learning_rate": 5.9019132495063056e-05,
+      "loss": 1.0038,
+      "step": 1688
+    },
+    {
+      "epoch": 1.9056763626094324,
+      "grad_norm": 0.042386867105960846,
+      "learning_rate": 5.8910818467345185e-05,
+      "loss": 1.0279,
+      "step": 1689
+    },
+    {
+      "epoch": 1.9068059870093195,
+      "grad_norm": 0.04081982001662254,
+      "learning_rate": 5.880256240596096e-05,
+      "loss": 1.0205,
+      "step": 1690
+    },
+    {
+      "epoch": 1.9079356114092065,
+      "grad_norm": 0.043334148824214935,
+      "learning_rate": 5.869436446363195e-05,
+      "loss": 0.9938,
+      "step": 1691
+    },
+    {
+      "epoch": 1.9090652358090936,
+      "grad_norm": 0.04122162610292435,
+      "learning_rate": 5.858622479299787e-05,
+      "loss": 0.9001,
+      "step": 1692
+    },
+    {
+      "epoch": 1.9101948602089807,
+      "grad_norm": 0.041972994804382324,
+      "learning_rate": 5.847814354661616e-05,
+      "loss": 0.9105,
+      "step": 1693
+    },
+    {
+      "epoch": 1.9113244846088675,
+      "grad_norm": 0.04239688068628311,
+      "learning_rate": 5.8370120876961745e-05,
+      "loss": 0.9657,
+      "step": 1694
+    },
+    {
+      "epoch": 1.9124541090087546,
+      "grad_norm": 0.043491121381521225,
+      "learning_rate": 5.8262156936427096e-05,
+      "loss": 1.0711,
+      "step": 1695
+    },
+    {
+      "epoch": 1.9135837334086416,
+      "grad_norm": 0.04262327775359154,
+      "learning_rate": 5.8154251877321706e-05,
+      "loss": 0.8719,
+      "step": 1696
+    },
+    {
+      "epoch": 1.9147133578085287,
+      "grad_norm": 0.04023859649896622,
+      "learning_rate": 5.804640585187207e-05,
+      "loss": 1.0929,
+      "step": 1697
+    },
+    {
+      "epoch": 1.9158429822084158,
+      "grad_norm": 0.04104539379477501,
+      "learning_rate": 5.793861901222131e-05,
+      "loss": 0.8602,
+      "step": 1698
+    },
+    {
+      "epoch": 1.9169726066083026,
+      "grad_norm": 0.04198300093412399,
+      "learning_rate": 5.783089151042914e-05,
+      "loss": 0.9534,
+      "step": 1699
+    },
+    {
+      "epoch": 1.9181022310081897,
+      "grad_norm": 0.04305735230445862,
+      "learning_rate": 5.772322349847154e-05,
+      "loss": 0.948,
+      "step": 1700
+    },
+    {
+      "epoch": 1.9192318554080767,
+      "grad_norm": 0.04169195145368576,
+      "learning_rate": 5.761561512824053e-05,
+      "loss": 0.9201,
+      "step": 1701
+    },
+    {
+      "epoch": 1.9203614798079638,
+      "grad_norm": 0.041344236582517624,
+      "learning_rate": 5.750806655154399e-05,
+      "loss": 0.996,
+      "step": 1702
+    },
+    {
+      "epoch": 1.9214911042078509,
+      "grad_norm": 0.04161575064063072,
+      "learning_rate": 5.740057792010562e-05,
+      "loss": 0.98,
+      "step": 1703
+    },
+    {
+      "epoch": 1.922620728607738,
+      "grad_norm": 0.04355592653155327,
+      "learning_rate": 5.72931493855642e-05,
+      "loss": 0.9855,
+      "step": 1704
+    },
+    {
+      "epoch": 1.923750353007625,
+      "grad_norm": 0.043292030692100525,
+      "learning_rate": 5.71857810994741e-05,
+      "loss": 1.0197,
+      "step": 1705
+    },
+    {
+      "epoch": 1.924879977407512,
+      "grad_norm": 0.04138379916548729,
+      "learning_rate": 5.70784732133045e-05,
+      "loss": 0.9655,
+      "step": 1706
+    },
+    {
+      "epoch": 1.926009601807399,
+      "grad_norm": 0.04593029245734215,
+      "learning_rate": 5.6971225878439285e-05,
+      "loss": 1.0195,
+      "step": 1707
+    },
+    {
+      "epoch": 1.9271392262072862,
+      "grad_norm": 0.042191557586193085,
+      "learning_rate": 5.686403924617718e-05,
+      "loss": 1.0543,
+      "step": 1708
+    },
+    {
+      "epoch": 1.9282688506071732,
+      "grad_norm": 0.041936252266168594,
+      "learning_rate": 5.6756913467731066e-05,
+      "loss": 0.9987,
+      "step": 1709
+    },
+    {
+      "epoch": 1.9293984750070603,
+      "grad_norm": 0.042339473962783813,
+      "learning_rate": 5.6649848694228026e-05,
+      "loss": 0.8652,
+      "step": 1710
+    },
+    {
+      "epoch": 1.9305280994069471,
+      "grad_norm": 0.04215146601200104,
+      "learning_rate": 5.6542845076709126e-05,
+      "loss": 0.9681,
+      "step": 1711
+    },
+    {
+      "epoch": 1.9316577238068342,
+      "grad_norm": 0.042300038039684296,
+      "learning_rate": 5.643590276612909e-05,
+      "loss": 0.9267,
+      "step": 1712
+    },
+    {
+      "epoch": 1.9327873482067213,
+      "grad_norm": 0.04266273230314255,
+      "learning_rate": 5.6329021913356216e-05,
+      "loss": 1.0148,
+      "step": 1713
+    },
+    {
+      "epoch": 1.9339169726066083,
+      "grad_norm": 0.04163605719804764,
+      "learning_rate": 5.6222202669172045e-05,
+      "loss": 1.1034,
+      "step": 1714
+    },
+    {
+      "epoch": 1.9350465970064954,
+      "grad_norm": 0.044020287692546844,
+      "learning_rate": 5.611544518427121e-05,
+      "loss": 1.0142,
+      "step": 1715
+    },
+    {
+      "epoch": 1.9361762214063822,
+      "grad_norm": 0.041142720729112625,
+      "learning_rate": 5.6008749609261304e-05,
+      "loss": 1.0388,
+      "step": 1716
+    },
+    {
+      "epoch": 1.9373058458062693,
+      "grad_norm": 0.04347783327102661,
+      "learning_rate": 5.59021160946625e-05,
+      "loss": 1.0435,
+      "step": 1717
+    },
+    {
+      "epoch": 1.9384354702061564,
+      "grad_norm": 0.04174618050456047,
+      "learning_rate": 5.579554479090735e-05,
+      "loss": 0.9933,
+      "step": 1718
+    },
+    {
+      "epoch": 1.9395650946060434,
+      "grad_norm": 0.042470984160900116,
+      "learning_rate": 5.568903584834082e-05,
+      "loss": 1.0287,
+      "step": 1719
+    },
+    {
+      "epoch": 1.9406947190059305,
+      "grad_norm": 0.04115034639835358,
+      "learning_rate": 5.558258941721982e-05,
+      "loss": 0.9838,
+      "step": 1720
+    },
+    {
+      "epoch": 1.9418243434058176,
+      "grad_norm": 0.04163762181997299,
+      "learning_rate": 5.547620564771293e-05,
+      "loss": 0.9575,
+      "step": 1721
+    },
+    {
+      "epoch": 1.9429539678057046,
+      "grad_norm": 0.04199086129665375,
+      "learning_rate": 5.5369884689900586e-05,
+      "loss": 1.0274,
+      "step": 1722
+    },
+    {
+      "epoch": 1.9440835922055917,
+      "grad_norm": 0.04397021606564522,
+      "learning_rate": 5.526362669377444e-05,
+      "loss": 1.0434,
+      "step": 1723
+    },
+    {
+      "epoch": 1.9452132166054787,
+      "grad_norm": 0.0407724566757679,
+      "learning_rate": 5.515743180923737e-05,
+      "loss": 1.1053,
+      "step": 1724
+    },
+    {
+      "epoch": 1.9463428410053658,
+      "grad_norm": 0.040555331856012344,
+      "learning_rate": 5.5051300186103214e-05,
+      "loss": 0.9554,
+      "step": 1725
+    },
+    {
+      "epoch": 1.9474724654052529,
+      "grad_norm": 0.043255507946014404,
+      "learning_rate": 5.494523197409653e-05,
+      "loss": 0.9669,
+      "step": 1726
+    },
+    {
+      "epoch": 1.94860208980514,
+      "grad_norm": 0.04308564215898514,
+      "learning_rate": 5.483922732285258e-05,
+      "loss": 0.9083,
+      "step": 1727
+    },
+    {
+      "epoch": 1.9497317142050268,
+      "grad_norm": 0.042100612074136734,
+      "learning_rate": 5.4733286381916715e-05,
+      "loss": 0.9747,
+      "step": 1728
+    },
+    {
+      "epoch": 1.9508613386049138,
+      "grad_norm": 0.042345307767391205,
+      "learning_rate": 5.4627409300744526e-05,
+      "loss": 0.9809,
+      "step": 1729
+    },
+    {
+      "epoch": 1.951990963004801,
+      "grad_norm": 0.041533615440130234,
+      "learning_rate": 5.452159622870158e-05,
+      "loss": 0.9315,
+      "step": 1730
+    },
+    {
+      "epoch": 1.953120587404688,
+      "grad_norm": 0.041634928435087204,
+      "learning_rate": 5.4415847315063083e-05,
+      "loss": 0.9254,
+      "step": 1731
+    },
+    {
+      "epoch": 1.954250211804575,
+      "grad_norm": 0.04313870519399643,
+      "learning_rate": 5.431016270901362e-05,
+      "loss": 0.926,
+      "step": 1732
+    },
+    {
+      "epoch": 1.9553798362044619,
+      "grad_norm": 0.04185828939080238,
+      "learning_rate": 5.4204542559647266e-05,
+      "loss": 0.8805,
+      "step": 1733
+    },
+    {
+      "epoch": 1.956509460604349,
+      "grad_norm": 0.04114462062716484,
+      "learning_rate": 5.409898701596703e-05,
+      "loss": 1.0163,
+      "step": 1734
+    },
+    {
+      "epoch": 1.957639085004236,
+      "grad_norm": 0.04308726638555527,
+      "learning_rate": 5.399349622688479e-05,
+      "loss": 1.0412,
+      "step": 1735
+    },
+    {
+      "epoch": 1.958768709404123,
+      "grad_norm": 0.04099753499031067,
+      "learning_rate": 5.388807034122111e-05,
+      "loss": 0.8463,
+      "step": 1736
+    },
+    {
+      "epoch": 1.9598983338040101,
+      "grad_norm": 0.04242704063653946,
+      "learning_rate": 5.378270950770494e-05,
+      "loss": 0.9864,
+      "step": 1737
+    },
+    {
+      "epoch": 1.9610279582038972,
+      "grad_norm": 0.04200004041194916,
+      "learning_rate": 5.367741387497351e-05,
+      "loss": 0.9477,
+      "step": 1738
+    },
+    {
+      "epoch": 1.9621575826037843,
+      "grad_norm": 0.04250415042042732,
+      "learning_rate": 5.3572183591572054e-05,
+      "loss": 0.9758,
+      "step": 1739
+    },
+    {
+      "epoch": 1.9632872070036713,
+      "grad_norm": 0.04299885779619217,
+      "learning_rate": 5.346701880595354e-05,
+      "loss": 1.0595,
+      "step": 1740
+    },
+    {
+      "epoch": 1.9644168314035584,
+      "grad_norm": 0.04244587942957878,
+      "learning_rate": 5.336191966647874e-05,
+      "loss": 1.0024,
+      "step": 1741
+    },
+    {
+      "epoch": 1.9655464558034454,
+      "grad_norm": 0.04144721105694771,
+      "learning_rate": 5.325688632141555e-05,
+      "loss": 0.986,
+      "step": 1742
+    },
+    {
+      "epoch": 1.9666760802033325,
+      "grad_norm": 0.045342057943344116,
+      "learning_rate": 5.315191891893919e-05,
+      "loss": 1.0682,
+      "step": 1743
+    },
+    {
+      "epoch": 1.9678057046032196,
+      "grad_norm": 0.043371427804231644,
+      "learning_rate": 5.3047017607131955e-05,
+      "loss": 0.9772,
+      "step": 1744
+    },
+    {
+      "epoch": 1.9689353290031064,
+      "grad_norm": 0.04567921906709671,
+      "learning_rate": 5.29421825339826e-05,
+      "loss": 0.9572,
+      "step": 1745
+    },
+    {
+      "epoch": 1.9700649534029935,
+      "grad_norm": 0.040609635412693024,
+      "learning_rate": 5.2837413847386776e-05,
+      "loss": 0.9193,
+      "step": 1746
+    },
+    {
+      "epoch": 1.9711945778028805,
+      "grad_norm": 0.04302290827035904,
+      "learning_rate": 5.2732711695146266e-05,
+      "loss": 0.9864,
+      "step": 1747
+    },
+    {
+      "epoch": 1.9723242022027676,
+      "grad_norm": 0.04100389406085014,
+      "learning_rate": 5.2628076224969036e-05,
+      "loss": 0.9455,
+      "step": 1748
+    },
+    {
+      "epoch": 1.9734538266026547,
+      "grad_norm": 0.04347492754459381,
+      "learning_rate": 5.2523507584469e-05,
+      "loss": 0.8589,
+      "step": 1749
+    },
+    {
+      "epoch": 1.9745834510025415,
+      "grad_norm": 0.041986506432294846,
+      "learning_rate": 5.24190059211658e-05,
+      "loss": 0.9477,
+      "step": 1750
+    },
+    {
+      "epoch": 1.9757130754024286,
+      "grad_norm": 0.041839152574539185,
+      "learning_rate": 5.231457138248457e-05,
+      "loss": 1.0711,
+      "step": 1751
+    },
+    {
+      "epoch": 1.9768426998023156,
+      "grad_norm": 0.04245884716510773,
+      "learning_rate": 5.221020411575572e-05,
+      "loss": 1.0331,
+      "step": 1752
+    },
+    {
+      "epoch": 1.9779723242022027,
+      "grad_norm": 0.04223821312189102,
+      "learning_rate": 5.210590426821479e-05,
+      "loss": 1.0202,
+      "step": 1753
+    },
+    {
+      "epoch": 1.9791019486020898,
+      "grad_norm": 0.0426202192902565,
+      "learning_rate": 5.200167198700228e-05,
+      "loss": 0.9441,
+      "step": 1754
+    },
+    {
+      "epoch": 1.9802315730019768,
+      "grad_norm": 0.04572174325585365,
+      "learning_rate": 5.189750741916326e-05,
+      "loss": 0.9966,
+      "step": 1755
+    },
+    {
+      "epoch": 1.981361197401864,
+      "grad_norm": 0.04126838594675064,
+      "learning_rate": 5.179341071164725e-05,
+      "loss": 1.0346,
+      "step": 1756
+    },
+    {
+      "epoch": 1.982490821801751,
+      "grad_norm": 0.041993822902441025,
+      "learning_rate": 5.16893820113082e-05,
+      "loss": 0.8974,
+      "step": 1757
+    },
+    {
+      "epoch": 1.983620446201638,
+      "grad_norm": 0.042118530720472336,
+      "learning_rate": 5.1585421464903994e-05,
+      "loss": 1.0124,
+      "step": 1758
+    },
+    {
+      "epoch": 1.984750070601525,
+      "grad_norm": 0.04224333539605141,
+      "learning_rate": 5.1481529219096304e-05,
+      "loss": 1.0564,
+      "step": 1759
+    },
+    {
+      "epoch": 1.9858796950014121,
+      "grad_norm": 0.04151271656155586,
+      "learning_rate": 5.137770542045063e-05,
+      "loss": 0.9412,
+      "step": 1760
+    },
+    {
+      "epoch": 1.9870093194012992,
+      "grad_norm": 0.04101775959134102,
+      "learning_rate": 5.12739502154358e-05,
+      "loss": 0.9164,
+      "step": 1761
+    },
+    {
+      "epoch": 1.988138943801186,
+      "grad_norm": 0.04164358228445053,
+      "learning_rate": 5.117026375042387e-05,
+      "loss": 0.9745,
+      "step": 1762
+    },
+    {
+      "epoch": 1.9892685682010731,
+      "grad_norm": 0.04084084928035736,
+      "learning_rate": 5.106664617168997e-05,
+      "loss": 0.9126,
+      "step": 1763
+    },
+    {
+      "epoch": 1.9903981926009602,
+      "grad_norm": 0.042791228741407394,
+      "learning_rate": 5.096309762541196e-05,
+      "loss": 0.9558,
+      "step": 1764
+    },
+    {
+      "epoch": 1.9915278170008472,
+      "grad_norm": 0.042839985340833664,
+      "learning_rate": 5.085961825767049e-05,
+      "loss": 0.9901,
+      "step": 1765
+    },
+    {
+      "epoch": 1.9926574414007343,
+      "grad_norm": 0.043768156319856644,
+      "learning_rate": 5.075620821444839e-05,
+      "loss": 0.9564,
+      "step": 1766
+    },
+    {
+      "epoch": 1.9937870658006211,
+      "grad_norm": 0.042571209371089935,
+      "learning_rate": 5.06528676416308e-05,
+      "loss": 0.9687,
+      "step": 1767
+    },
+    {
+      "epoch": 1.9949166902005082,
+      "grad_norm": 0.04314889758825302,
+      "learning_rate": 5.054959668500494e-05,
+      "loss": 1.1016,
+      "step": 1768
+    },
+    {
+      "epoch": 1.9960463146003953,
+      "grad_norm": 0.04145493358373642,
+      "learning_rate": 5.0446395490259734e-05,
+      "loss": 0.8943,
+      "step": 1769
+    },
+    {
+      "epoch": 1.9971759390002823,
+      "grad_norm": 0.04249756410717964,
+      "learning_rate": 5.0343264202985575e-05,
+      "loss": 0.9249,
+      "step": 1770
+    },
+    {
+      "epoch": 1.9971759390002823,
+      "eval_loss": 0.98674476146698,
+      "eval_runtime": 550.5161,
+      "eval_samples_per_second": 17.771,
+      "eval_steps_per_second": 8.886,
+      "step": 1770
+    },
+    {
+      "epoch": 1.9983055634001694,
+      "grad_norm": 0.04155099019408226,
+      "learning_rate": 5.024020296867447e-05,
+      "loss": 0.9206,
+      "step": 1771
+    },
+    {
+      "epoch": 1.9994351878000565,
+      "grad_norm": 0.04307400807738304,
+      "learning_rate": 5.013721193271943e-05,
+      "loss": 0.9068,
+      "step": 1772
+    },
+    {
+      "epoch": 2.0005648121999435,
+      "grad_norm": 0.04108371585607529,
+      "learning_rate": 5.003429124041448e-05,
+      "loss": 1.0198,
+      "step": 1773
+    },
+    {
+      "epoch": 2.0016944365998306,
+      "grad_norm": 0.04241110011935234,
+      "learning_rate": 4.993144103695444e-05,
+      "loss": 1.0401,
+      "step": 1774
+    },
+    {
+      "epoch": 2.0005649717514125,
+      "grad_norm": 0.04497067630290985,
+      "learning_rate": 4.9828661467434644e-05,
+      "loss": 0.8991,
+      "step": 1775
+    },
+    {
+      "epoch": 2.0016949152542374,
+      "grad_norm": 0.04190199822187424,
+      "learning_rate": 4.9725952676850794e-05,
+      "loss": 1.0002,
+      "step": 1776
+    },
+    {
+      "epoch": 2.0028248587570623,
+      "grad_norm": 0.04270040616393089,
+      "learning_rate": 4.9623314810098755e-05,
+      "loss": 0.9177,
+      "step": 1777
+    },
+    {
+      "epoch": 2.003954802259887,
+      "grad_norm": 0.042360976338386536,
+      "learning_rate": 4.952074801197426e-05,
+      "loss": 0.9497,
+      "step": 1778
+    },
+    {
+      "epoch": 2.0050847457627117,
+      "grad_norm": 0.04216707870364189,
+      "learning_rate": 4.9418252427172996e-05,
+      "loss": 0.9471,
+      "step": 1779
+    },
+    {
+      "epoch": 2.0062146892655366,
+      "grad_norm": 0.04215894639492035,
+      "learning_rate": 4.93158282002899e-05,
+      "loss": 0.956,
+      "step": 1780
+    },
+    {
+      "epoch": 2.0073446327683615,
+      "grad_norm": 0.041472285985946655,
+      "learning_rate": 4.921347547581939e-05,
+      "loss": 1.0184,
+      "step": 1781
+    },
+    {
+      "epoch": 2.0084745762711864,
+      "grad_norm": 0.041694026440382004,
+      "learning_rate": 4.911119439815508e-05,
+      "loss": 0.9715,
+      "step": 1782
+    },
+    {
+      "epoch": 2.0096045197740113,
+      "grad_norm": 0.041800353676080704,
+      "learning_rate": 4.900898511158938e-05,
+      "loss": 0.9482,
+      "step": 1783
+    },
+    {
+      "epoch": 2.0107344632768362,
+      "grad_norm": 0.04290417581796646,
+      "learning_rate": 4.890684776031347e-05,
+      "loss": 0.9799,
+      "step": 1784
+    },
+    {
+      "epoch": 2.011864406779661,
+      "grad_norm": 0.04424067586660385,
+      "learning_rate": 4.8804782488417054e-05,
+      "loss": 0.9554,
+      "step": 1785
+    },
+    {
+      "epoch": 2.012994350282486,
+      "grad_norm": 0.04423059523105621,
+      "learning_rate": 4.870278943988815e-05,
+      "loss": 0.9836,
+      "step": 1786
+    },
+    {
+      "epoch": 2.0141242937853105,
+      "grad_norm": 0.041760075837373734,
+      "learning_rate": 4.860086875861288e-05,
+      "loss": 0.9582,
+      "step": 1787
+    },
+    {
+      "epoch": 2.0152542372881355,
+      "grad_norm": 0.04206801578402519,
+      "learning_rate": 4.8499020588375274e-05,
+      "loss": 1.0098,
+      "step": 1788
+    },
+    {
+      "epoch": 2.0163841807909604,
+      "grad_norm": 0.041129548102617264,
+      "learning_rate": 4.8397245072857066e-05,
+      "loss": 0.999,
+      "step": 1789
+    },
+    {
+      "epoch": 2.0175141242937853,
+      "grad_norm": 0.042199425399303436,
+      "learning_rate": 4.82955423556375e-05,
+      "loss": 1.0089,
+      "step": 1790
+    },
+    {
+      "epoch": 2.01864406779661,
+      "grad_norm": 0.04368184879422188,
+      "learning_rate": 4.8193912580193126e-05,
+      "loss": 0.9612,
+      "step": 1791
+    },
+    {
+      "epoch": 2.019774011299435,
+      "grad_norm": 0.04354587942361832,
+      "learning_rate": 4.8092355889897535e-05,
+      "loss": 0.9033,
+      "step": 1792
+    },
+    {
+      "epoch": 2.02090395480226,
+      "grad_norm": 0.04332384094595909,
+      "learning_rate": 4.799087242802137e-05,
+      "loss": 0.965,
+      "step": 1793
+    },
+    {
+      "epoch": 2.022033898305085,
+      "grad_norm": 0.04101352021098137,
+      "learning_rate": 4.788946233773171e-05,
+      "loss": 0.9246,
+      "step": 1794
+    },
+    {
+      "epoch": 2.0231638418079094,
+      "grad_norm": 0.0444595068693161,
+      "learning_rate": 4.778812576209241e-05,
+      "loss": 0.9641,
+      "step": 1795
+    },
+    {
+      "epoch": 2.0242937853107343,
+      "grad_norm": 0.04191228747367859,
+      "learning_rate": 4.768686284406341e-05,
+      "loss": 0.9286,
+      "step": 1796
+    },
+    {
+      "epoch": 2.0254237288135593,
+      "grad_norm": 0.042017534375190735,
+      "learning_rate": 4.758567372650081e-05,
+      "loss": 1.0086,
+      "step": 1797
+    },
+    {
+      "epoch": 2.026553672316384,
+      "grad_norm": 0.0429287888109684,
+      "learning_rate": 4.748455855215661e-05,
+      "loss": 0.9385,
+      "step": 1798
+    },
+    {
+      "epoch": 2.027683615819209,
+      "grad_norm": 0.04178991913795471,
+      "learning_rate": 4.738351746367847e-05,
+      "loss": 0.8944,
+      "step": 1799
+    },
+    {
+      "epoch": 2.028813559322034,
+      "grad_norm": 0.04435624182224274,
+      "learning_rate": 4.728255060360955e-05,
+      "loss": 0.9517,
+      "step": 1800
+    },
+    {
+      "epoch": 2.029943502824859,
+      "grad_norm": 0.04260590299963951,
+      "learning_rate": 4.718165811438827e-05,
+      "loss": 0.868,
+      "step": 1801
+    },
+    {
+      "epoch": 2.031073446327684,
+      "grad_norm": 0.04216260462999344,
+      "learning_rate": 4.708084013834813e-05,
+      "loss": 1.0083,
+      "step": 1802
+    },
+    {
+      "epoch": 2.0322033898305083,
+      "grad_norm": 0.04152260348200798,
+      "learning_rate": 4.69800968177176e-05,
+      "loss": 1.0892,
+      "step": 1803
+    },
+    {
+      "epoch": 2.033333333333333,
+      "grad_norm": 0.0426262728869915,
+      "learning_rate": 4.687942829461969e-05,
+      "loss": 0.939,
+      "step": 1804
+    },
+    {
+      "epoch": 2.034463276836158,
+      "grad_norm": 0.042451802641153336,
+      "learning_rate": 4.677883471107193e-05,
+      "loss": 0.9649,
+      "step": 1805
+    },
+    {
+      "epoch": 2.035593220338983,
+      "grad_norm": 0.04238414391875267,
+      "learning_rate": 4.667831620898624e-05,
+      "loss": 1.0071,
+      "step": 1806
+    },
+    {
+      "epoch": 2.036723163841808,
+      "grad_norm": 0.04201498627662659,
+      "learning_rate": 4.657787293016854e-05,
+      "loss": 1.0189,
+      "step": 1807
+    },
+    {
+      "epoch": 2.037853107344633,
+      "grad_norm": 0.042049601674079895,
+      "learning_rate": 4.64775050163185e-05,
+      "loss": 0.9891,
+      "step": 1808
+    },
+    {
+      "epoch": 2.038983050847458,
+      "grad_norm": 0.04302692413330078,
+      "learning_rate": 4.6377212609029727e-05,
+      "loss": 0.928,
+      "step": 1809
+    },
+    {
+      "epoch": 2.0401129943502827,
+      "grad_norm": 0.04353903979063034,
+      "learning_rate": 4.6276995849789115e-05,
+      "loss": 0.9651,
+      "step": 1810
+    },
+    {
+      "epoch": 2.041242937853107,
+      "grad_norm": 0.04473048448562622,
+      "learning_rate": 4.617685487997693e-05,
+      "loss": 1.036,
+      "step": 1811
+    },
+    {
+      "epoch": 2.042372881355932,
+      "grad_norm": 0.043298009783029556,
+      "learning_rate": 4.607678984086644e-05,
+      "loss": 0.9935,
+      "step": 1812
+    },
+    {
+      "epoch": 2.043502824858757,
+      "grad_norm": 0.04168165847659111,
+      "learning_rate": 4.597680087362388e-05,
+      "loss": 0.9986,
+      "step": 1813
+    },
+    {
+      "epoch": 2.044632768361582,
+      "grad_norm": 0.04360107332468033,
+      "learning_rate": 4.5876888119308116e-05,
+      "loss": 1.041,
+      "step": 1814
+    },
+    {
+      "epoch": 2.045762711864407,
+      "grad_norm": 0.0433119460940361,
+      "learning_rate": 4.57770517188705e-05,
+      "loss": 0.9559,
+      "step": 1815
+    },
+    {
+      "epoch": 2.0468926553672318,
+      "grad_norm": 0.04289395362138748,
+      "learning_rate": 4.567729181315466e-05,
+      "loss": 0.9049,
+      "step": 1816
+    },
+    {
+      "epoch": 2.0480225988700567,
+      "grad_norm": 0.04195884242653847,
+      "learning_rate": 4.5577608542896414e-05,
+      "loss": 0.9174,
+      "step": 1817
+    },
+    {
+      "epoch": 2.049152542372881,
+      "grad_norm": 0.045005012303590775,
+      "learning_rate": 4.547800204872328e-05,
+      "loss": 1.0006,
+      "step": 1818
+    },
+    {
+      "epoch": 2.050282485875706,
+      "grad_norm": 0.042177747935056686,
+      "learning_rate": 4.537847247115455e-05,
+      "loss": 0.8667,
+      "step": 1819
+    },
+    {
+      "epoch": 2.051412429378531,
+      "grad_norm": 0.04145647957921028,
+      "learning_rate": 4.527901995060113e-05,
+      "loss": 0.9482,
+      "step": 1820
+    },
+    {
+      "epoch": 2.052542372881356,
+      "grad_norm": 0.04456596449017525,
+      "learning_rate": 4.517964462736504e-05,
+      "loss": 1.0383,
+      "step": 1821
+    },
+    {
+      "epoch": 2.053672316384181,
+      "grad_norm": 0.04184206575155258,
+      "learning_rate": 4.5080346641639474e-05,
+      "loss": 0.8904,
+      "step": 1822
+    },
+    {
+      "epoch": 2.0548022598870057,
+      "grad_norm": 0.04329001158475876,
+      "learning_rate": 4.49811261335085e-05,
+      "loss": 1.0087,
+      "step": 1823
+    },
+    {
+      "epoch": 2.0559322033898306,
+      "grad_norm": 0.043258413672447205,
+      "learning_rate": 4.488198324294691e-05,
+      "loss": 0.9363,
+      "step": 1824
+    },
+    {
+      "epoch": 2.0570621468926555,
+      "grad_norm": 0.04341473430395126,
+      "learning_rate": 4.478291810981998e-05,
+      "loss": 0.9583,
+      "step": 1825
+    },
+    {
+      "epoch": 2.05819209039548,
+      "grad_norm": 0.042815063148736954,
+      "learning_rate": 4.46839308738833e-05,
+      "loss": 0.9449,
+      "step": 1826
+    },
+    {
+      "epoch": 2.059322033898305,
+      "grad_norm": 0.04257015511393547,
+      "learning_rate": 4.4585021674782534e-05,
+      "loss": 0.961,
+      "step": 1827
+    },
+    {
+      "epoch": 2.06045197740113,
+      "grad_norm": 0.043428368866443634,
+      "learning_rate": 4.4486190652053315e-05,
+      "loss": 0.9658,
+      "step": 1828
+    },
+    {
+      "epoch": 2.0615819209039548,
+      "grad_norm": 0.042757805436849594,
+      "learning_rate": 4.438743794512096e-05,
+      "loss": 1.0266,
+      "step": 1829
+    },
+    {
+      "epoch": 2.0627118644067797,
+      "grad_norm": 0.0427665188908577,
+      "learning_rate": 4.428876369330023e-05,
+      "loss": 1.0276,
+      "step": 1830
+    },
+    {
+      "epoch": 2.0638418079096046,
+      "grad_norm": 0.04279656335711479,
+      "learning_rate": 4.41901680357954e-05,
+      "loss": 0.8914,
+      "step": 1831
+    },
+    {
+      "epoch": 2.0649717514124295,
+      "grad_norm": 0.04068029671907425,
+      "learning_rate": 4.4091651111699606e-05,
+      "loss": 0.9335,
+      "step": 1832
+    },
+    {
+      "epoch": 2.0661016949152544,
+      "grad_norm": 0.043668482452631,
+      "learning_rate": 4.3993213059995154e-05,
+      "loss": 1.0198,
+      "step": 1833
+    },
+    {
+      "epoch": 2.067231638418079,
+      "grad_norm": 0.04225129261612892,
+      "learning_rate": 4.389485401955294e-05,
+      "loss": 1.0887,
+      "step": 1834
+    },
+    {
+      "epoch": 2.068361581920904,
+      "grad_norm": 0.042218949645757675,
+      "learning_rate": 4.379657412913243e-05,
+      "loss": 0.9673,
+      "step": 1835
+    },
+    {
+      "epoch": 2.0694915254237287,
+      "grad_norm": 0.044487837702035904,
+      "learning_rate": 4.369837352738143e-05,
+      "loss": 1.0627,
+      "step": 1836
+    },
+    {
+      "epoch": 2.0706214689265536,
+      "grad_norm": 0.043943922966718674,
+      "learning_rate": 4.360025235283589e-05,
+      "loss": 1.028,
+      "step": 1837
+    },
+    {
+      "epoch": 2.0717514124293785,
+      "grad_norm": 0.04223185032606125,
+      "learning_rate": 4.3502210743919716e-05,
+      "loss": 0.982,
+      "step": 1838
+    },
+    {
+      "epoch": 2.0728813559322035,
+      "grad_norm": 0.04249482974410057,
+      "learning_rate": 4.340424883894456e-05,
+      "loss": 0.9398,
+      "step": 1839
+    },
+    {
+      "epoch": 2.0740112994350284,
+      "grad_norm": 0.04417464882135391,
+      "learning_rate": 4.330636677610962e-05,
+      "loss": 0.9966,
+      "step": 1840
+    },
+    {
+      "epoch": 2.0751412429378533,
+      "grad_norm": 0.04311511293053627,
+      "learning_rate": 4.320856469350148e-05,
+      "loss": 0.942,
+      "step": 1841
+    },
+    {
+      "epoch": 2.0762711864406778,
+      "grad_norm": 0.04165413975715637,
+      "learning_rate": 4.3110842729093905e-05,
+      "loss": 0.9925,
+      "step": 1842
+    },
+    {
+      "epoch": 2.0774011299435027,
+      "grad_norm": 0.04295306280255318,
+      "learning_rate": 4.301320102074753e-05,
+      "loss": 1.0347,
+      "step": 1843
+    },
+    {
+      "epoch": 2.0785310734463276,
+      "grad_norm": 0.042987748980522156,
+      "learning_rate": 4.291563970620995e-05,
+      "loss": 1.0159,
+      "step": 1844
+    },
+    {
+      "epoch": 2.0796610169491525,
+      "grad_norm": 0.04220904782414436,
+      "learning_rate": 4.281815892311525e-05,
+      "loss": 0.8988,
+      "step": 1845
+    },
+    {
+      "epoch": 2.0807909604519774,
+      "grad_norm": 0.04398202523589134,
+      "learning_rate": 4.2720758808983764e-05,
+      "loss": 0.8488,
+      "step": 1846
+    },
+    {
+      "epoch": 2.0819209039548023,
+      "grad_norm": 0.042306363582611084,
+      "learning_rate": 4.262343950122231e-05,
+      "loss": 0.9823,
+      "step": 1847
+    },
+    {
+      "epoch": 2.0830508474576273,
+      "grad_norm": 0.04340597242116928,
+      "learning_rate": 4.2526201137123476e-05,
+      "loss": 0.9914,
+      "step": 1848
+    },
+    {
+      "epoch": 2.084180790960452,
+      "grad_norm": 0.04187840595841408,
+      "learning_rate": 4.242904385386578e-05,
+      "loss": 0.9613,
+      "step": 1849
+    },
+    {
+      "epoch": 2.0853107344632766,
+      "grad_norm": 0.04024931788444519,
+      "learning_rate": 4.2331967788513295e-05,
+      "loss": 0.8752,
+      "step": 1850
+    },
+    {
+      "epoch": 2.0864406779661016,
+      "grad_norm": 0.04166257008910179,
+      "learning_rate": 4.223497307801551e-05,
+      "loss": 0.9149,
+      "step": 1851
+    },
+    {
+      "epoch": 2.0875706214689265,
+      "grad_norm": 0.04291348159313202,
+      "learning_rate": 4.213805985920729e-05,
+      "loss": 0.9848,
+      "step": 1852
+    },
+    {
+      "epoch": 2.0887005649717514,
+      "grad_norm": 0.043250638991594315,
+      "learning_rate": 4.2041228268808294e-05,
+      "loss": 0.9481,
+      "step": 1853
+    },
+    {
+      "epoch": 2.0898305084745763,
+      "grad_norm": 0.04172235727310181,
+      "learning_rate": 4.1944478443423155e-05,
+      "loss": 0.9443,
+      "step": 1854
+    },
+    {
+      "epoch": 2.090960451977401,
+      "grad_norm": 0.04241979867219925,
+      "learning_rate": 4.1847810519541255e-05,
+      "loss": 0.9697,
+      "step": 1855
+    },
+    {
+      "epoch": 2.092090395480226,
+      "grad_norm": 0.042877666652202606,
+      "learning_rate": 4.175122463353621e-05,
+      "loss": 0.9557,
+      "step": 1856
+    },
+    {
+      "epoch": 2.093220338983051,
+      "grad_norm": 0.04124724119901657,
+      "learning_rate": 4.1654720921666044e-05,
+      "loss": 0.9078,
+      "step": 1857
+    },
+    {
+      "epoch": 2.0943502824858755,
+      "grad_norm": 0.04259972274303436,
+      "learning_rate": 4.1558299520072885e-05,
+      "loss": 0.9561,
+      "step": 1858
+    },
+    {
+      "epoch": 2.0954802259887004,
+      "grad_norm": 0.04378335177898407,
+      "learning_rate": 4.14619605647827e-05,
+      "loss": 0.9466,
+      "step": 1859
+    },
+    {
+      "epoch": 2.0966101694915253,
+      "grad_norm": 0.04312341660261154,
+      "learning_rate": 4.136570419170501e-05,
+      "loss": 0.9631,
+      "step": 1860
+    },
+    {
+      "epoch": 2.0977401129943503,
+      "grad_norm": 0.042929697781801224,
+      "learning_rate": 4.126953053663307e-05,
+      "loss": 0.8175,
+      "step": 1861
+    },
+    {
+      "epoch": 2.098870056497175,
+      "grad_norm": 0.0430251844227314,
+      "learning_rate": 4.117343973524329e-05,
+      "loss": 0.9864,
+      "step": 1862
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.04333322122693062,
+      "learning_rate": 4.1077431923095244e-05,
+      "loss": 0.9386,
+      "step": 1863
+    },
+    {
+      "epoch": 2.101129943502825,
+      "grad_norm": 0.04226114600896835,
+      "learning_rate": 4.09815072356314e-05,
+      "loss": 1.0237,
+      "step": 1864
+    },
+    {
+      "epoch": 2.10225988700565,
+      "grad_norm": 0.0435493066906929,
+      "learning_rate": 4.088566580817694e-05,
+      "loss": 0.9987,
+      "step": 1865
+    },
+    {
+      "epoch": 2.1033898305084744,
+      "grad_norm": 0.04193365201354027,
+      "learning_rate": 4.078990777593975e-05,
+      "loss": 1.0508,
+      "step": 1866
+    },
+    {
+      "epoch": 2.1045197740112993,
+      "grad_norm": 0.04185868427157402,
+      "learning_rate": 4.0694233274009816e-05,
+      "loss": 0.9143,
+      "step": 1867
+    },
+    {
+      "epoch": 2.105649717514124,
+      "grad_norm": 0.042803600430488586,
+      "learning_rate": 4.0598642437359414e-05,
+      "loss": 0.9395,
+      "step": 1868
+    },
+    {
+      "epoch": 2.106779661016949,
+      "grad_norm": 0.042586468160152435,
+      "learning_rate": 4.05031354008429e-05,
+      "loss": 1.0041,
+      "step": 1869
+    },
+    {
+      "epoch": 2.107909604519774,
+      "grad_norm": 0.04308640584349632,
+      "learning_rate": 4.040771229919612e-05,
+      "loss": 0.95,
+      "step": 1870
+    },
+    {
+      "epoch": 2.109039548022599,
+      "grad_norm": 0.04416211321949959,
+      "learning_rate": 4.0312373267036816e-05,
+      "loss": 0.8787,
+      "step": 1871
+    },
+    {
+      "epoch": 2.110169491525424,
+      "grad_norm": 0.04484262317419052,
+      "learning_rate": 4.021711843886395e-05,
+      "loss": 0.9828,
+      "step": 1872
+    },
+    {
+      "epoch": 2.111299435028249,
+      "grad_norm": 0.04606883227825165,
+      "learning_rate": 4.012194794905775e-05,
+      "loss": 1.0205,
+      "step": 1873
+    },
+    {
+      "epoch": 2.1124293785310733,
+      "grad_norm": 0.043447453528642654,
+      "learning_rate": 4.0026861931879446e-05,
+      "loss": 0.9687,
+      "step": 1874
+    },
+    {
+      "epoch": 2.113559322033898,
+      "grad_norm": 0.042595911771059036,
+      "learning_rate": 3.99318605214711e-05,
+      "loss": 0.9269,
+      "step": 1875
+    },
+    {
+      "epoch": 2.114689265536723,
+      "grad_norm": 0.04446543753147125,
+      "learning_rate": 3.983694385185543e-05,
+      "loss": 0.937,
+      "step": 1876
+    },
+    {
+      "epoch": 2.115819209039548,
+      "grad_norm": 0.04304026812314987,
+      "learning_rate": 3.974211205693559e-05,
+      "loss": 0.8803,
+      "step": 1877
+    },
+    {
+      "epoch": 2.116949152542373,
+      "grad_norm": 0.041843317449092865,
+      "learning_rate": 3.964736527049502e-05,
+      "loss": 0.9684,
+      "step": 1878
+    },
+    {
+      "epoch": 2.118079096045198,
+      "grad_norm": 0.04251433536410332,
+      "learning_rate": 3.955270362619717e-05,
+      "loss": 0.9508,
+      "step": 1879
+    },
+    {
+      "epoch": 2.1192090395480228,
+      "grad_norm": 0.04271414875984192,
+      "learning_rate": 3.945812725758554e-05,
+      "loss": 0.9956,
+      "step": 1880
+    },
+    {
+      "epoch": 2.1203389830508477,
+      "grad_norm": 0.04318327456712723,
+      "learning_rate": 3.936363629808309e-05,
+      "loss": 0.981,
+      "step": 1881
+    },
+    {
+      "epoch": 2.121468926553672,
+      "grad_norm": 0.04361184686422348,
+      "learning_rate": 3.926923088099248e-05,
+      "loss": 0.9366,
+      "step": 1882
+    },
+    {
+      "epoch": 2.122598870056497,
+      "grad_norm": 0.043581850826740265,
+      "learning_rate": 3.917491113949566e-05,
+      "loss": 0.907,
+      "step": 1883
+    },
+    {
+      "epoch": 2.123728813559322,
+      "grad_norm": 0.04366597160696983,
+      "learning_rate": 3.908067720665356e-05,
+      "loss": 0.9216,
+      "step": 1884
+    },
+    {
+      "epoch": 2.124858757062147,
+      "grad_norm": 0.043010421097278595,
+      "learning_rate": 3.8986529215406275e-05,
+      "loss": 0.9093,
+      "step": 1885
+    },
+    {
+      "epoch": 2.125988700564972,
+      "grad_norm": 0.044093113392591476,
+      "learning_rate": 3.889246729857253e-05,
+      "loss": 0.967,
+      "step": 1886
+    },
+    {
+      "epoch": 2.1271186440677967,
+      "grad_norm": 0.043733999133110046,
+      "learning_rate": 3.879849158884966e-05,
+      "loss": 0.9415,
+      "step": 1887
+    },
+    {
+      "epoch": 2.1282485875706216,
+      "grad_norm": 0.04293812811374664,
+      "learning_rate": 3.870460221881336e-05,
+      "loss": 1.055,
+      "step": 1888
+    },
+    {
+      "epoch": 2.1293785310734465,
+      "grad_norm": 0.04446000978350639,
+      "learning_rate": 3.861079932091749e-05,
+      "loss": 0.9815,
+      "step": 1889
+    },
+    {
+      "epoch": 2.130508474576271,
+      "grad_norm": 0.04384174942970276,
+      "learning_rate": 3.851708302749409e-05,
+      "loss": 1.0459,
+      "step": 1890
+    },
+    {
+      "epoch": 2.131638418079096,
+      "grad_norm": 0.04370885714888573,
+      "learning_rate": 3.8423453470752805e-05,
+      "loss": 0.9833,
+      "step": 1891
+    },
+    {
+      "epoch": 2.132768361581921,
+      "grad_norm": 0.046146828681230545,
+      "learning_rate": 3.8329910782780966e-05,
+      "loss": 0.9669,
+      "step": 1892
+    },
+    {
+      "epoch": 2.1338983050847458,
+      "grad_norm": 0.04334993287920952,
+      "learning_rate": 3.82364550955435e-05,
+      "loss": 0.9578,
+      "step": 1893
+    },
+    {
+      "epoch": 2.1350282485875707,
+      "grad_norm": 0.04254217445850372,
+      "learning_rate": 3.81430865408825e-05,
+      "loss": 0.9224,
+      "step": 1894
+    },
+    {
+      "epoch": 2.1361581920903956,
+      "grad_norm": 0.04299743473529816,
+      "learning_rate": 3.8049805250517004e-05,
+      "loss": 0.9902,
+      "step": 1895
+    },
+    {
+      "epoch": 2.1372881355932205,
+      "grad_norm": 0.04331592097878456,
+      "learning_rate": 3.795661135604319e-05,
+      "loss": 0.8947,
+      "step": 1896
+    },
+    {
+      "epoch": 2.138418079096045,
+      "grad_norm": 0.04248592630028725,
+      "learning_rate": 3.786350498893384e-05,
+      "loss": 0.9523,
+      "step": 1897
+    },
+    {
+      "epoch": 2.13954802259887,
+      "grad_norm": 0.04246117174625397,
+      "learning_rate": 3.777048628053812e-05,
+      "loss": 0.9214,
+      "step": 1898
+    },
+    {
+      "epoch": 2.140677966101695,
+      "grad_norm": 0.043325964361429214,
+      "learning_rate": 3.7677555362081753e-05,
+      "loss": 0.9706,
+      "step": 1899
+    },
+    {
+      "epoch": 2.1418079096045197,
+      "grad_norm": 0.043188970535993576,
+      "learning_rate": 3.7584712364666494e-05,
+      "loss": 0.9794,
+      "step": 1900
+    },
+    {
+      "epoch": 2.1429378531073446,
+      "grad_norm": 0.045290421694517136,
+      "learning_rate": 3.7491957419270076e-05,
+      "loss": 0.9523,
+      "step": 1901
+    },
+    {
+      "epoch": 2.1440677966101696,
+      "grad_norm": 0.042237911373376846,
+      "learning_rate": 3.7399290656746025e-05,
+      "loss": 0.9501,
+      "step": 1902
+    },
+    {
+      "epoch": 2.1451977401129945,
+      "grad_norm": 0.045046236366033554,
+      "learning_rate": 3.7306712207823416e-05,
+      "loss": 1.0151,
+      "step": 1903
+    },
+    {
+      "epoch": 2.1463276836158194,
+      "grad_norm": 0.04322218894958496,
+      "learning_rate": 3.721422220310689e-05,
+      "loss": 0.9431,
+      "step": 1904
+    },
+    {
+      "epoch": 2.1474576271186443,
+      "grad_norm": 0.04495770111680031,
+      "learning_rate": 3.71218207730761e-05,
+      "loss": 0.9045,
+      "step": 1905
+    },
+    {
+      "epoch": 2.1485875706214688,
+      "grad_norm": 0.04479131102561951,
+      "learning_rate": 3.702950804808585e-05,
+      "loss": 1.0059,
+      "step": 1906
+    },
+    {
+      "epoch": 2.1497175141242937,
+      "grad_norm": 0.043446823954582214,
+      "learning_rate": 3.693728415836586e-05,
+      "loss": 0.9376,
+      "step": 1907
+    },
+    {
+      "epoch": 2.1508474576271186,
+      "grad_norm": 0.04486696794629097,
+      "learning_rate": 3.684514923402049e-05,
+      "loss": 0.9882,
+      "step": 1908
+    },
+    {
+      "epoch": 2.1519774011299435,
+      "grad_norm": 0.04472246393561363,
+      "learning_rate": 3.6753103405028455e-05,
+      "loss": 1.0832,
+      "step": 1909
+    },
+    {
+      "epoch": 2.1531073446327684,
+      "grad_norm": 0.04336708039045334,
+      "learning_rate": 3.666114680124298e-05,
+      "loss": 1.0151,
+      "step": 1910
+    },
+    {
+      "epoch": 2.1542372881355933,
+      "grad_norm": 0.04265212640166283,
+      "learning_rate": 3.6569279552391335e-05,
+      "loss": 0.9912,
+      "step": 1911
+    },
+    {
+      "epoch": 2.1553672316384183,
+      "grad_norm": 0.04515647888183594,
+      "learning_rate": 3.64775017880747e-05,
+      "loss": 0.9827,
+      "step": 1912
+    },
+    {
+      "epoch": 2.1564971751412427,
+      "grad_norm": 0.043546319007873535,
+      "learning_rate": 3.638581363776805e-05,
+      "loss": 0.9497,
+      "step": 1913
+    },
+    {
+      "epoch": 2.1576271186440676,
+      "grad_norm": 0.04316361993551254,
+      "learning_rate": 3.629421523081993e-05,
+      "loss": 0.9203,
+      "step": 1914
+    },
+    {
+      "epoch": 2.1587570621468926,
+      "grad_norm": 0.04629300907254219,
+      "learning_rate": 3.620270669645228e-05,
+      "loss": 0.9532,
+      "step": 1915
+    },
+    {
+      "epoch": 2.1598870056497175,
+      "grad_norm": 0.04403712972998619,
+      "learning_rate": 3.611128816376027e-05,
+      "loss": 0.9163,
+      "step": 1916
+    },
+    {
+      "epoch": 2.1610169491525424,
+      "grad_norm": 0.04466381296515465,
+      "learning_rate": 3.601995976171204e-05,
+      "loss": 0.9866,
+      "step": 1917
+    },
+    {
+      "epoch": 2.1621468926553673,
+      "grad_norm": 0.043207425624132156,
+      "learning_rate": 3.592872161914872e-05,
+      "loss": 0.984,
+      "step": 1918
+    },
+    {
+      "epoch": 2.163276836158192,
+      "grad_norm": 0.043047528713941574,
+      "learning_rate": 3.583757386478389e-05,
+      "loss": 0.9795,
+      "step": 1919
+    },
+    {
+      "epoch": 2.164406779661017,
+      "grad_norm": 0.04385092481970787,
+      "learning_rate": 3.574651662720382e-05,
+      "loss": 1.0078,
+      "step": 1920
+    },
+    {
+      "epoch": 2.1655367231638416,
+      "grad_norm": 0.04322254657745361,
+      "learning_rate": 3.565555003486697e-05,
+      "loss": 1.0193,
+      "step": 1921
+    },
+    {
+      "epoch": 2.1666666666666665,
+      "grad_norm": 0.0432460755109787,
+      "learning_rate": 3.556467421610397e-05,
+      "loss": 1.0581,
+      "step": 1922
+    },
+    {
+      "epoch": 2.1677966101694914,
+      "grad_norm": 0.0425468273460865,
+      "learning_rate": 3.547388929911735e-05,
+      "loss": 0.9893,
+      "step": 1923
+    },
+    {
+      "epoch": 2.1689265536723163,
+      "grad_norm": 0.04245666787028313,
+      "learning_rate": 3.538319541198144e-05,
+      "loss": 0.949,
+      "step": 1924
+    },
+    {
+      "epoch": 2.1700564971751413,
+      "grad_norm": 0.04199038818478584,
+      "learning_rate": 3.5292592682642134e-05,
+      "loss": 0.9138,
+      "step": 1925
+    },
+    {
+      "epoch": 2.171186440677966,
+      "grad_norm": 0.04395987465977669,
+      "learning_rate": 3.5202081238916716e-05,
+      "loss": 0.9457,
+      "step": 1926
+    },
+    {
+      "epoch": 2.172316384180791,
+      "grad_norm": 0.044708240777254105,
+      "learning_rate": 3.511166120849373e-05,
+      "loss": 1.0507,
+      "step": 1927
+    },
+    {
+      "epoch": 2.173446327683616,
+      "grad_norm": 0.04187402501702309,
+      "learning_rate": 3.502133271893269e-05,
+      "loss": 0.9955,
+      "step": 1928
+    },
+    {
+      "epoch": 2.1745762711864405,
+      "grad_norm": 0.04379774630069733,
+      "learning_rate": 3.493109589766403e-05,
+      "loss": 0.9605,
+      "step": 1929
+    },
+    {
+      "epoch": 2.1757062146892654,
+      "grad_norm": 0.04420413821935654,
+      "learning_rate": 3.484095087198881e-05,
+      "loss": 0.909,
+      "step": 1930
+    },
+    {
+      "epoch": 2.1768361581920903,
+      "grad_norm": 0.045495763421058655,
+      "learning_rate": 3.475089776907868e-05,
+      "loss": 0.9914,
+      "step": 1931
+    },
+    {
+      "epoch": 2.1779661016949152,
+      "grad_norm": 0.04469761997461319,
+      "learning_rate": 3.466093671597557e-05,
+      "loss": 0.9338,
+      "step": 1932
+    },
+    {
+      "epoch": 2.17909604519774,
+      "grad_norm": 0.0445980429649353,
+      "learning_rate": 3.457106783959141e-05,
+      "loss": 0.9165,
+      "step": 1933
+    },
+    {
+      "epoch": 2.180225988700565,
+      "grad_norm": 0.04591384530067444,
+      "learning_rate": 3.448129126670834e-05,
+      "loss": 0.9446,
+      "step": 1934
+    },
+    {
+      "epoch": 2.18135593220339,
+      "grad_norm": 0.042553581297397614,
+      "learning_rate": 3.4391607123978095e-05,
+      "loss": 0.9666,
+      "step": 1935
+    },
+    {
+      "epoch": 2.182485875706215,
+      "grad_norm": 0.042708031833171844,
+      "learning_rate": 3.43020155379221e-05,
+      "loss": 1.0542,
+      "step": 1936
+    },
+    {
+      "epoch": 2.1836158192090394,
+      "grad_norm": 0.04311813414096832,
+      "learning_rate": 3.42125166349312e-05,
+      "loss": 1.0275,
+      "step": 1937
+    },
+    {
+      "epoch": 2.1847457627118643,
+      "grad_norm": 0.04380349814891815,
+      "learning_rate": 3.4123110541265443e-05,
+      "loss": 0.9017,
+      "step": 1938
+    },
+    {
+      "epoch": 2.185875706214689,
+      "grad_norm": 0.04285889118909836,
+      "learning_rate": 3.403379738305399e-05,
+      "loss": 0.966,
+      "step": 1939
+    },
+    {
+      "epoch": 2.187005649717514,
+      "grad_norm": 0.04314585402607918,
+      "learning_rate": 3.394457728629489e-05,
+      "loss": 0.9751,
+      "step": 1940
+    },
+    {
+      "epoch": 2.188135593220339,
+      "grad_norm": 0.04352666810154915,
+      "learning_rate": 3.385545037685485e-05,
+      "loss": 0.8714,
+      "step": 1941
+    },
+    {
+      "epoch": 2.189265536723164,
+      "grad_norm": 0.04633624106645584,
+      "learning_rate": 3.3766416780469256e-05,
+      "loss": 1.0029,
+      "step": 1942
+    },
+    {
+      "epoch": 2.190395480225989,
+      "grad_norm": 0.043088801205158234,
+      "learning_rate": 3.3677476622741664e-05,
+      "loss": 1.0206,
+      "step": 1943
+    },
+    {
+      "epoch": 2.1915254237288138,
+      "grad_norm": 0.044619057327508926,
+      "learning_rate": 3.358863002914392e-05,
+      "loss": 1.0072,
+      "step": 1944
+    },
+    {
+      "epoch": 2.1926553672316382,
+      "grad_norm": 0.044183578342199326,
+      "learning_rate": 3.349987712501591e-05,
+      "loss": 0.9604,
+      "step": 1945
+    },
+    {
+      "epoch": 2.193785310734463,
+      "grad_norm": 0.04243571683764458,
+      "learning_rate": 3.341121803556529e-05,
+      "loss": 0.9624,
+      "step": 1946
+    },
+    {
+      "epoch": 2.194915254237288,
+      "grad_norm": 0.04398589953780174,
+      "learning_rate": 3.3322652885867315e-05,
+      "loss": 1.0083,
+      "step": 1947
+    },
+    {
+      "epoch": 2.196045197740113,
+      "grad_norm": 0.04456605017185211,
+      "learning_rate": 3.323418180086483e-05,
+      "loss": 0.9609,
+      "step": 1948
+    },
+    {
+      "epoch": 2.197175141242938,
+      "grad_norm": 0.0444977842271328,
+      "learning_rate": 3.3145804905367904e-05,
+      "loss": 0.9846,
+      "step": 1949
+    },
+    {
+      "epoch": 2.198305084745763,
+      "grad_norm": 0.04405605047941208,
+      "learning_rate": 3.305752232405377e-05,
+      "loss": 0.9386,
+      "step": 1950
+    },
+    {
+      "epoch": 2.1994350282485877,
+      "grad_norm": 0.0432204008102417,
+      "learning_rate": 3.296933418146656e-05,
+      "loss": 0.959,
+      "step": 1951
+    },
+    {
+      "epoch": 2.2005649717514126,
+      "grad_norm": 0.04267331585288048,
+      "learning_rate": 3.2881240602017224e-05,
+      "loss": 0.9566,
+      "step": 1952
+    },
+    {
+      "epoch": 2.201694915254237,
+      "grad_norm": 0.0441194623708725,
+      "learning_rate": 3.279324170998328e-05,
+      "loss": 0.9098,
+      "step": 1953
+    },
+    {
+      "epoch": 2.202824858757062,
+      "grad_norm": 0.043350979685783386,
+      "learning_rate": 3.270533762950868e-05,
+      "loss": 0.9884,
+      "step": 1954
+    },
+    {
+      "epoch": 2.203954802259887,
+      "grad_norm": 0.04236207902431488,
+      "learning_rate": 3.2617528484603576e-05,
+      "loss": 0.9565,
+      "step": 1955
+    },
+    {
+      "epoch": 2.205084745762712,
+      "grad_norm": 0.043681688606739044,
+      "learning_rate": 3.252981439914432e-05,
+      "loss": 0.9966,
+      "step": 1956
+    },
+    {
+      "epoch": 2.2062146892655368,
+      "grad_norm": 0.044432997703552246,
+      "learning_rate": 3.244219549687298e-05,
+      "loss": 0.9039,
+      "step": 1957
+    },
+    {
+      "epoch": 2.2073446327683617,
+      "grad_norm": 0.0422402061522007,
+      "learning_rate": 3.235467190139744e-05,
+      "loss": 1.036,
+      "step": 1958
+    },
+    {
+      "epoch": 2.2084745762711866,
+      "grad_norm": 0.044290442019701004,
+      "learning_rate": 3.226724373619118e-05,
+      "loss": 1.0285,
+      "step": 1959
+    },
+    {
+      "epoch": 2.209604519774011,
+      "grad_norm": 0.04244611784815788,
+      "learning_rate": 3.2179911124592966e-05,
+      "loss": 0.954,
+      "step": 1960
+    },
+    {
+      "epoch": 2.210734463276836,
+      "grad_norm": 0.04315735027194023,
+      "learning_rate": 3.2092674189806796e-05,
+      "loss": 0.9831,
+      "step": 1961
+    },
+    {
+      "epoch": 2.211864406779661,
+      "grad_norm": 0.04285259544849396,
+      "learning_rate": 3.200553305490168e-05,
+      "loss": 1.0086,
+      "step": 1962
+    },
+    {
+      "epoch": 2.212994350282486,
+      "grad_norm": 0.043484896421432495,
+      "learning_rate": 3.19184878428115e-05,
+      "loss": 0.9422,
+      "step": 1963
+    },
+    {
+      "epoch": 2.2141242937853107,
+      "grad_norm": 0.0440831333398819,
+      "learning_rate": 3.183153867633478e-05,
+      "loss": 0.9518,
+      "step": 1964
+    },
+    {
+      "epoch": 2.2152542372881356,
+      "grad_norm": 0.044822998344898224,
+      "learning_rate": 3.174468567813461e-05,
+      "loss": 0.9831,
+      "step": 1965
+    },
+    {
+      "epoch": 2.2163841807909606,
+      "grad_norm": 0.04435247182846069,
+      "learning_rate": 3.165792897073834e-05,
+      "loss": 0.9201,
+      "step": 1966
+    },
+    {
+      "epoch": 2.2175141242937855,
+      "grad_norm": 0.04330061003565788,
+      "learning_rate": 3.157126867653753e-05,
+      "loss": 0.9692,
+      "step": 1967
+    },
+    {
+      "epoch": 2.2186440677966104,
+      "grad_norm": 0.043659813702106476,
+      "learning_rate": 3.1484704917787654e-05,
+      "loss": 0.9948,
+      "step": 1968
+    },
+    {
+      "epoch": 2.219774011299435,
+      "grad_norm": 0.044164128601551056,
+      "learning_rate": 3.1398237816608135e-05,
+      "loss": 0.9184,
+      "step": 1969
+    },
+    {
+      "epoch": 2.2209039548022598,
+      "grad_norm": 0.04370899870991707,
+      "learning_rate": 3.131186749498195e-05,
+      "loss": 0.9378,
+      "step": 1970
+    },
+    {
+      "epoch": 2.2220338983050847,
+      "grad_norm": 0.04361080005764961,
+      "learning_rate": 3.122559407475545e-05,
+      "loss": 0.9835,
+      "step": 1971
+    },
+    {
+      "epoch": 2.2231638418079096,
+      "grad_norm": 0.04388871416449547,
+      "learning_rate": 3.113941767763847e-05,
+      "loss": 0.9748,
+      "step": 1972
+    },
+    {
+      "epoch": 2.2242937853107345,
+      "grad_norm": 0.04267182946205139,
+      "learning_rate": 3.105333842520386e-05,
+      "loss": 0.9749,
+      "step": 1973
+    },
+    {
+      "epoch": 2.2254237288135594,
+      "grad_norm": 0.043973181396722794,
+      "learning_rate": 3.096735643888744e-05,
+      "loss": 1.0114,
+      "step": 1974
+    },
+    {
+      "epoch": 2.2265536723163843,
+      "grad_norm": 0.04484979808330536,
+      "learning_rate": 3.088147183998782e-05,
+      "loss": 0.96,
+      "step": 1975
+    },
+    {
+      "epoch": 2.227683615819209,
+      "grad_norm": 0.04428961127996445,
+      "learning_rate": 3.079568474966622e-05,
+      "loss": 0.9723,
+      "step": 1976
+    },
+    {
+      "epoch": 2.2288135593220337,
+      "grad_norm": 0.0446840338408947,
+      "learning_rate": 3.070999528894629e-05,
+      "loss": 0.9633,
+      "step": 1977
+    },
+    {
+      "epoch": 2.2299435028248586,
+      "grad_norm": 0.04326969385147095,
+      "learning_rate": 3.0624403578713976e-05,
+      "loss": 0.934,
+      "step": 1978
+    },
+    {
+      "epoch": 2.2310734463276836,
+      "grad_norm": 0.04401664808392525,
+      "learning_rate": 3.053890973971726e-05,
+      "loss": 1.055,
+      "step": 1979
+    },
+    {
+      "epoch": 2.2322033898305085,
+      "grad_norm": 0.04194195196032524,
+      "learning_rate": 3.0453513892566197e-05,
+      "loss": 1.0163,
+      "step": 1980
+    },
+    {
+      "epoch": 2.2333333333333334,
+      "grad_norm": 0.04289555549621582,
+      "learning_rate": 3.0368216157732397e-05,
+      "loss": 1.0094,
+      "step": 1981
+    },
+    {
+      "epoch": 2.2344632768361583,
+      "grad_norm": 0.04160468652844429,
+      "learning_rate": 3.028301665554919e-05,
+      "loss": 0.9903,
+      "step": 1982
+    },
+    {
+      "epoch": 2.2355932203389832,
+      "grad_norm": 0.04424525424838066,
+      "learning_rate": 3.0197915506211337e-05,
+      "loss": 0.8995,
+      "step": 1983
+    },
+    {
+      "epoch": 2.236723163841808,
+      "grad_norm": 0.04448120296001434,
+      "learning_rate": 3.011291282977482e-05,
+      "loss": 0.9845,
+      "step": 1984
+    },
+    {
+      "epoch": 2.2378531073446326,
+      "grad_norm": 0.04262761026620865,
+      "learning_rate": 3.0028008746156588e-05,
+      "loss": 1.0196,
+      "step": 1985
+    },
+    {
+      "epoch": 2.2389830508474575,
+      "grad_norm": 0.042251285165548325,
+      "learning_rate": 2.994320337513471e-05,
+      "loss": 1.0227,
+      "step": 1986
+    },
+    {
+      "epoch": 2.2401129943502824,
+      "grad_norm": 0.04287414252758026,
+      "learning_rate": 2.9858496836347848e-05,
+      "loss": 0.956,
+      "step": 1987
+    },
+    {
+      "epoch": 2.2412429378531074,
+      "grad_norm": 0.04530943185091019,
+      "learning_rate": 2.9773889249295294e-05,
+      "loss": 1.0611,
+      "step": 1988
+    },
+    {
+      "epoch": 2.2423728813559323,
+      "grad_norm": 0.044923651963472366,
+      "learning_rate": 2.9689380733336714e-05,
+      "loss": 0.996,
+      "step": 1989
+    },
+    {
+      "epoch": 2.243502824858757,
+      "grad_norm": 0.044317737221717834,
+      "learning_rate": 2.9604971407692027e-05,
+      "loss": 1.0236,
+      "step": 1990
+    },
+    {
+      "epoch": 2.244632768361582,
+      "grad_norm": 0.04512808844447136,
+      "learning_rate": 2.9520661391441216e-05,
+      "loss": 1.011,
+      "step": 1991
+    },
+    {
+      "epoch": 2.2457627118644066,
+      "grad_norm": 0.04485659301280975,
+      "learning_rate": 2.9436450803524183e-05,
+      "loss": 0.8739,
+      "step": 1992
+    },
+    {
+      "epoch": 2.2468926553672315,
+      "grad_norm": 0.04554317891597748,
+      "learning_rate": 2.93523397627405e-05,
+      "loss": 0.9842,
+      "step": 1993
+    },
+    {
+      "epoch": 2.2480225988700564,
+      "grad_norm": 0.044335950165987015,
+      "learning_rate": 2.9268328387749442e-05,
+      "loss": 0.9293,
+      "step": 1994
+    },
+    {
+      "epoch": 2.2491525423728813,
+      "grad_norm": 0.0435885488986969,
+      "learning_rate": 2.918441679706949e-05,
+      "loss": 0.9581,
+      "step": 1995
+    },
+    {
+      "epoch": 2.2502824858757062,
+      "grad_norm": 0.0456608422100544,
+      "learning_rate": 2.910060510907847e-05,
+      "loss": 0.9955,
+      "step": 1996
+    },
+    {
+      "epoch": 2.251412429378531,
+      "grad_norm": 0.04341673105955124,
+      "learning_rate": 2.9016893442013306e-05,
+      "loss": 1.0348,
+      "step": 1997
+    },
+    {
+      "epoch": 2.252542372881356,
+      "grad_norm": 0.04355182126164436,
+      "learning_rate": 2.8933281913969734e-05,
+      "loss": 0.9885,
+      "step": 1998
+    },
+    {
+      "epoch": 2.253672316384181,
+      "grad_norm": 0.04326220974326134,
+      "learning_rate": 2.884977064290224e-05,
+      "loss": 0.895,
+      "step": 1999
+    },
+    {
+      "epoch": 2.254802259887006,
+      "grad_norm": 0.044238653033971786,
+      "learning_rate": 2.8766359746623894e-05,
+      "loss": 0.8844,
+      "step": 2000
+    },
+    {
+      "epoch": 2.2559322033898304,
+      "grad_norm": 0.043896935880184174,
+      "learning_rate": 2.8683049342806157e-05,
+      "loss": 0.9641,
+      "step": 2001
+    },
+    {
+      "epoch": 2.2570621468926553,
+      "grad_norm": 0.04326968267560005,
+      "learning_rate": 2.8599839548978713e-05,
+      "loss": 0.9907,
+      "step": 2002
+    },
+    {
+      "epoch": 2.25819209039548,
+      "grad_norm": 0.04462304711341858,
+      "learning_rate": 2.851673048252931e-05,
+      "loss": 1.0387,
+      "step": 2003
+    },
+    {
+      "epoch": 2.259322033898305,
+      "grad_norm": 0.04370822757482529,
+      "learning_rate": 2.843372226070361e-05,
+      "loss": 1.0079,
+      "step": 2004
+    },
+    {
+      "epoch": 2.26045197740113,
+      "grad_norm": 0.04261447489261627,
+      "learning_rate": 2.835081500060498e-05,
+      "loss": 0.9493,
+      "step": 2005
+    },
+    {
+      "epoch": 2.261581920903955,
+      "grad_norm": 0.044322796165943146,
+      "learning_rate": 2.826800881919437e-05,
+      "loss": 0.9265,
+      "step": 2006
+    },
+    {
+      "epoch": 2.26271186440678,
+      "grad_norm": 0.043250150978565216,
+      "learning_rate": 2.8185303833290177e-05,
+      "loss": 0.9663,
+      "step": 2007
+    },
+    {
+      "epoch": 2.2638418079096043,
+      "grad_norm": 0.043140899389982224,
+      "learning_rate": 2.8102700159568008e-05,
+      "loss": 0.9805,
+      "step": 2008
+    },
+    {
+      "epoch": 2.2649717514124292,
+      "grad_norm": 0.044603534042835236,
+      "learning_rate": 2.802019791456044e-05,
+      "loss": 1.0215,
+      "step": 2009
+    },
+    {
+      "epoch": 2.266101694915254,
+      "grad_norm": 0.04410766065120697,
+      "learning_rate": 2.7937797214657147e-05,
+      "loss": 1.0059,
+      "step": 2010
+    },
+    {
+      "epoch": 2.267231638418079,
+      "grad_norm": 0.043483491986989975,
+      "learning_rate": 2.7855498176104434e-05,
+      "loss": 0.9487,
+      "step": 2011
+    },
+    {
+      "epoch": 2.268361581920904,
+      "grad_norm": 0.042907100170850754,
+      "learning_rate": 2.7773300915005207e-05,
+      "loss": 0.84,
+      "step": 2012
+    },
+    {
+      "epoch": 2.269491525423729,
+      "grad_norm": 0.044326696544885635,
+      "learning_rate": 2.7691205547318776e-05,
+      "loss": 1.0064,
+      "step": 2013
+    },
+    {
+      "epoch": 2.270621468926554,
+      "grad_norm": 0.04372712969779968,
+      "learning_rate": 2.7609212188860757e-05,
+      "loss": 0.9424,
+      "step": 2014
+    },
+    {
+      "epoch": 2.2717514124293787,
+      "grad_norm": 0.04335528612136841,
+      "learning_rate": 2.7527320955302794e-05,
+      "loss": 0.9816,
+      "step": 2015
+    },
+    {
+      "epoch": 2.272881355932203,
+      "grad_norm": 0.04416688159108162,
+      "learning_rate": 2.74455319621725e-05,
+      "loss": 0.9326,
+      "step": 2016
+    },
+    {
+      "epoch": 2.274011299435028,
+      "grad_norm": 0.04589290916919708,
+      "learning_rate": 2.736384532485321e-05,
+      "loss": 0.9499,
+      "step": 2017
+    },
+    {
+      "epoch": 2.275141242937853,
+      "grad_norm": 0.0437130331993103,
+      "learning_rate": 2.7282261158583976e-05,
+      "loss": 0.9826,
+      "step": 2018
+    },
+    {
+      "epoch": 2.276271186440678,
+      "grad_norm": 0.04344571754336357,
+      "learning_rate": 2.7200779578459123e-05,
+      "loss": 0.9085,
+      "step": 2019
+    },
+    {
+      "epoch": 2.277401129943503,
+      "grad_norm": 0.04497040435671806,
+      "learning_rate": 2.7119400699428332e-05,
+      "loss": 0.937,
+      "step": 2020
+    },
+    {
+      "epoch": 2.2785310734463278,
+      "grad_norm": 0.04277808219194412,
+      "learning_rate": 2.703812463629646e-05,
+      "loss": 0.9305,
+      "step": 2021
+    },
+    {
+      "epoch": 2.2796610169491527,
+      "grad_norm": 0.04330938309431076,
+      "learning_rate": 2.6956951503723272e-05,
+      "loss": 0.9323,
+      "step": 2022
+    },
+    {
+      "epoch": 2.280790960451977,
+      "grad_norm": 0.04527467489242554,
+      "learning_rate": 2.6875881416223204e-05,
+      "loss": 1.0554,
+      "step": 2023
+    },
+    {
+      "epoch": 2.281920903954802,
+      "grad_norm": 0.04678976908326149,
+      "learning_rate": 2.6794914488165533e-05,
+      "loss": 0.9549,
+      "step": 2024
+    },
+    {
+      "epoch": 2.283050847457627,
+      "grad_norm": 0.04412659630179405,
+      "learning_rate": 2.671405083377386e-05,
+      "loss": 1.0194,
+      "step": 2025
+    },
+    {
+      "epoch": 2.284180790960452,
+      "grad_norm": 0.04408615082502365,
+      "learning_rate": 2.6633290567126157e-05,
+      "loss": 0.9893,
+      "step": 2026
+    },
+    {
+      "epoch": 2.285310734463277,
+      "grad_norm": 0.04321965202689171,
+      "learning_rate": 2.6552633802154493e-05,
+      "loss": 0.9805,
+      "step": 2027
+    },
+    {
+      "epoch": 2.2864406779661017,
+      "grad_norm": 0.04430380091071129,
+      "learning_rate": 2.6472080652644926e-05,
+      "loss": 0.9462,
+      "step": 2028
+    },
+    {
+      "epoch": 2.2875706214689266,
+      "grad_norm": 0.04322722926735878,
+      "learning_rate": 2.639163123223747e-05,
+      "loss": 0.9606,
+      "step": 2029
+    },
+    {
+      "epoch": 2.2887005649717516,
+      "grad_norm": 0.043180808424949646,
+      "learning_rate": 2.6311285654425575e-05,
+      "loss": 0.9322,
+      "step": 2030
+    },
+    {
+      "epoch": 2.2898305084745765,
+      "grad_norm": 0.043794550001621246,
+      "learning_rate": 2.623104403255634e-05,
+      "loss": 0.9762,
+      "step": 2031
+    },
+    {
+      "epoch": 2.290960451977401,
+      "grad_norm": 0.044038672000169754,
+      "learning_rate": 2.6150906479830274e-05,
+      "loss": 0.9987,
+      "step": 2032
+    },
+    {
+      "epoch": 2.292090395480226,
+      "grad_norm": 0.0434204563498497,
+      "learning_rate": 2.6070873109300885e-05,
+      "loss": 0.9565,
+      "step": 2033
+    },
+    {
+      "epoch": 2.2932203389830508,
+      "grad_norm": 0.04463200643658638,
+      "learning_rate": 2.599094403387481e-05,
+      "loss": 0.9487,
+      "step": 2034
+    },
+    {
+      "epoch": 2.2943502824858757,
+      "grad_norm": 0.04487144574522972,
+      "learning_rate": 2.5911119366311597e-05,
+      "loss": 0.9826,
+      "step": 2035
+    },
+    {
+      "epoch": 2.2954802259887006,
+      "grad_norm": 0.044315457344055176,
+      "learning_rate": 2.5831399219223428e-05,
+      "loss": 1.0061,
+      "step": 2036
+    },
+    {
+      "epoch": 2.2966101694915255,
+      "grad_norm": 0.045690327882766724,
+      "learning_rate": 2.575178370507506e-05,
+      "loss": 0.9714,
+      "step": 2037
+    },
+    {
+      "epoch": 2.2977401129943504,
+      "grad_norm": 0.042950958013534546,
+      "learning_rate": 2.5672272936183627e-05,
+      "loss": 0.9542,
+      "step": 2038
+    },
+    {
+      "epoch": 2.298870056497175,
+      "grad_norm": 0.04238516837358475,
+      "learning_rate": 2.55928670247185e-05,
+      "loss": 0.9595,
+      "step": 2039
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.04378442093729973,
+      "learning_rate": 2.5513566082701135e-05,
+      "loss": 0.9148,
+      "step": 2040
+    },
+    {
+      "epoch": 2.3011299435028247,
+      "grad_norm": 0.043305352330207825,
+      "learning_rate": 2.543437022200489e-05,
+      "loss": 0.9211,
+      "step": 2041
+    },
+    {
+      "epoch": 2.3022598870056497,
+      "grad_norm": 0.04488563537597656,
+      "learning_rate": 2.535527955435486e-05,
+      "loss": 0.9947,
+      "step": 2042
+    },
+    {
+      "epoch": 2.3033898305084746,
+      "grad_norm": 0.04428262263536453,
+      "learning_rate": 2.527629419132783e-05,
+      "loss": 1.0371,
+      "step": 2043
+    },
+    {
+      "epoch": 2.3045197740112995,
+      "grad_norm": 0.04419199004769325,
+      "learning_rate": 2.5197414244351904e-05,
+      "loss": 0.927,
+      "step": 2044
+    },
+    {
+      "epoch": 2.3056497175141244,
+      "grad_norm": 0.04307791590690613,
+      "learning_rate": 2.51186398247065e-05,
+      "loss": 1.0249,
+      "step": 2045
+    },
+    {
+      "epoch": 2.3067796610169493,
+      "grad_norm": 0.04518159106373787,
+      "learning_rate": 2.5039971043522292e-05,
+      "loss": 0.9505,
+      "step": 2046
+    },
+    {
+      "epoch": 2.3079096045197742,
+      "grad_norm": 0.04357128217816353,
+      "learning_rate": 2.4961408011780707e-05,
+      "loss": 1.0103,
+      "step": 2047
+    },
+    {
+      "epoch": 2.3090395480225987,
+      "grad_norm": 0.0436529703438282,
+      "learning_rate": 2.488295084031419e-05,
+      "loss": 1.0638,
+      "step": 2048
+    },
+    {
+      "epoch": 2.3101694915254236,
+      "grad_norm": 0.04202603921294212,
+      "learning_rate": 2.4804599639805714e-05,
+      "loss": 0.9998,
+      "step": 2049
+    },
+    {
+      "epoch": 2.3112994350282485,
+      "grad_norm": 0.04418287053704262,
+      "learning_rate": 2.472635452078883e-05,
+      "loss": 0.9245,
+      "step": 2050
+    },
+    {
+      "epoch": 2.3124293785310734,
+      "grad_norm": 0.04419273883104324,
+      "learning_rate": 2.464821559364737e-05,
+      "loss": 0.9414,
+      "step": 2051
+    },
+    {
+      "epoch": 2.3135593220338984,
+      "grad_norm": 0.043836288154125214,
+      "learning_rate": 2.4570182968615418e-05,
+      "loss": 0.9468,
+      "step": 2052
+    },
+    {
+      "epoch": 2.3146892655367233,
+      "grad_norm": 0.04379745200276375,
+      "learning_rate": 2.4492256755777042e-05,
+      "loss": 0.9917,
+      "step": 2053
+    },
+    {
+      "epoch": 2.315819209039548,
+      "grad_norm": 0.04468725621700287,
+      "learning_rate": 2.441443706506623e-05,
+      "loss": 0.9677,
+      "step": 2054
+    },
+    {
+      "epoch": 2.3169491525423727,
+      "grad_norm": 0.04374300688505173,
+      "learning_rate": 2.433672400626663e-05,
+      "loss": 1.0079,
+      "step": 2055
+    },
+    {
+      "epoch": 2.3180790960451976,
+      "grad_norm": 0.045238886028528214,
+      "learning_rate": 2.4259117689011578e-05,
+      "loss": 0.9881,
+      "step": 2056
+    },
+    {
+      "epoch": 2.3192090395480225,
+      "grad_norm": 0.043223727494478226,
+      "learning_rate": 2.418161822278374e-05,
+      "loss": 1.0842,
+      "step": 2057
+    },
+    {
+      "epoch": 2.3203389830508474,
+      "grad_norm": 0.04424876347184181,
+      "learning_rate": 2.4104225716914985e-05,
+      "loss": 0.9624,
+      "step": 2058
+    },
+    {
+      "epoch": 2.3214689265536723,
+      "grad_norm": 0.04261065274477005,
+      "learning_rate": 2.402694028058644e-05,
+      "loss": 0.9402,
+      "step": 2059
+    },
+    {
+      "epoch": 2.3225988700564972,
+      "grad_norm": 0.042224884033203125,
+      "learning_rate": 2.3949762022828092e-05,
+      "loss": 0.9993,
+      "step": 2060
+    },
+    {
+      "epoch": 2.323728813559322,
+      "grad_norm": 0.0436599999666214,
+      "learning_rate": 2.387269105251866e-05,
+      "loss": 1.0189,
+      "step": 2061
+    },
+    {
+      "epoch": 2.324858757062147,
+      "grad_norm": 0.04297950863838196,
+      "learning_rate": 2.3795727478385676e-05,
+      "loss": 1.1014,
+      "step": 2062
+    },
+    {
+      "epoch": 2.325988700564972,
+      "grad_norm": 0.04437912628054619,
+      "learning_rate": 2.371887140900504e-05,
+      "loss": 0.943,
+      "step": 2063
+    },
+    {
+      "epoch": 2.3271186440677964,
+      "grad_norm": 0.043281424790620804,
+      "learning_rate": 2.3642122952801015e-05,
+      "loss": 0.9319,
+      "step": 2064
+    },
+    {
+      "epoch": 2.3282485875706214,
+      "grad_norm": 0.04431607574224472,
+      "learning_rate": 2.3565482218046075e-05,
+      "loss": 0.9951,
+      "step": 2065
+    },
+    {
+      "epoch": 2.3282485875706214,
+      "eval_loss": 0.9855674505233765,
+      "eval_runtime": 554.8838,
+      "eval_samples_per_second": 17.631,
+      "eval_steps_per_second": 8.816,
+      "step": 2065
+    },
+    {
+      "epoch": 2.3293785310734463,
+      "grad_norm": 0.044959913939237595,
+      "learning_rate": 2.3488949312860664e-05,
+      "loss": 0.8822,
+      "step": 2066
+    },
+    {
+      "epoch": 2.330508474576271,
+      "grad_norm": 0.044671010226011276,
+      "learning_rate": 2.341252434521325e-05,
+      "loss": 0.9366,
+      "step": 2067
+    },
+    {
+      "epoch": 2.331638418079096,
+      "grad_norm": 0.044197868555784225,
+      "learning_rate": 2.333620742291983e-05,
+      "loss": 0.9288,
+      "step": 2068
+    },
+    {
+      "epoch": 2.332768361581921,
+      "grad_norm": 0.045790866017341614,
+      "learning_rate": 2.325999865364409e-05,
+      "loss": 0.953,
+      "step": 2069
+    },
+    {
+      "epoch": 2.333898305084746,
+      "grad_norm": 0.043006639927625656,
+      "learning_rate": 2.3183898144897177e-05,
+      "loss": 0.9384,
+      "step": 2070
+    },
+    {
+      "epoch": 2.3350282485875704,
+      "grad_norm": 0.04369287192821503,
+      "learning_rate": 2.310790600403745e-05,
+      "loss": 0.9972,
+      "step": 2071
+    },
+    {
+      "epoch": 2.3361581920903953,
+      "grad_norm": 0.044209450483322144,
+      "learning_rate": 2.303202233827033e-05,
+      "loss": 0.9869,
+      "step": 2072
+    },
+    {
+      "epoch": 2.3372881355932202,
+      "grad_norm": 0.04336726665496826,
+      "learning_rate": 2.2956247254648344e-05,
+      "loss": 0.9897,
+      "step": 2073
+    },
+    {
+      "epoch": 2.338418079096045,
+      "grad_norm": 0.04320811852812767,
+      "learning_rate": 2.2880580860070778e-05,
+      "loss": 1.0209,
+      "step": 2074
+    },
+    {
+      "epoch": 2.33954802259887,
+      "grad_norm": 0.04383328929543495,
+      "learning_rate": 2.2805023261283497e-05,
+      "loss": 0.9673,
+      "step": 2075
+    },
+    {
+      "epoch": 2.340677966101695,
+      "grad_norm": 0.04465322569012642,
+      "learning_rate": 2.272957456487903e-05,
+      "loss": 1.0069,
+      "step": 2076
+    },
+    {
+      "epoch": 2.34180790960452,
+      "grad_norm": 0.04262759909033775,
+      "learning_rate": 2.2654234877296198e-05,
+      "loss": 0.8755,
+      "step": 2077
+    },
+    {
+      "epoch": 2.342937853107345,
+      "grad_norm": 0.0429883673787117,
+      "learning_rate": 2.257900430482003e-05,
+      "loss": 0.9556,
+      "step": 2078
+    },
+    {
+      "epoch": 2.3440677966101697,
+      "grad_norm": 0.0438074916601181,
+      "learning_rate": 2.2503882953581622e-05,
+      "loss": 0.9628,
+      "step": 2079
+    },
+    {
+      "epoch": 2.345197740112994,
+      "grad_norm": 0.0434563122689724,
+      "learning_rate": 2.242887092955801e-05,
+      "loss": 0.8773,
+      "step": 2080
+    },
+    {
+      "epoch": 2.346327683615819,
+      "grad_norm": 0.0433909073472023,
+      "learning_rate": 2.235396833857204e-05,
+      "loss": 1.0392,
+      "step": 2081
+    },
+    {
+      "epoch": 2.347457627118644,
+      "grad_norm": 0.04353194683790207,
+      "learning_rate": 2.2279175286292064e-05,
+      "loss": 0.9412,
+      "step": 2082
+    },
+    {
+      "epoch": 2.348587570621469,
+      "grad_norm": 0.04275263100862503,
+      "learning_rate": 2.2204491878231948e-05,
+      "loss": 0.9089,
+      "step": 2083
+    },
+    {
+      "epoch": 2.349717514124294,
+      "grad_norm": 0.0452563501894474,
+      "learning_rate": 2.2129918219750944e-05,
+      "loss": 1.0463,
+      "step": 2084
+    },
+    {
+      "epoch": 2.3508474576271188,
+      "grad_norm": 0.04346321523189545,
+      "learning_rate": 2.2055454416053422e-05,
+      "loss": 0.9901,
+      "step": 2085
+    },
+    {
+      "epoch": 2.3519774011299437,
+      "grad_norm": 0.04487701877951622,
+      "learning_rate": 2.1981100572188752e-05,
+      "loss": 0.981,
+      "step": 2086
+    },
+    {
+      "epoch": 2.353107344632768,
+      "grad_norm": 0.04522349685430527,
+      "learning_rate": 2.190685679305121e-05,
+      "loss": 0.9812,
+      "step": 2087
+    },
+    {
+      "epoch": 2.354237288135593,
+      "grad_norm": 0.04697408527135849,
+      "learning_rate": 2.1832723183379812e-05,
+      "loss": 0.9467,
+      "step": 2088
+    },
+    {
+      "epoch": 2.355367231638418,
+      "grad_norm": 0.042166102677583694,
+      "learning_rate": 2.1758699847758112e-05,
+      "loss": 1.0908,
+      "step": 2089
+    },
+    {
+      "epoch": 2.356497175141243,
+      "grad_norm": 0.04440850764513016,
+      "learning_rate": 2.168478689061413e-05,
+      "loss": 0.9584,
+      "step": 2090
+    },
+    {
+      "epoch": 2.357627118644068,
+      "grad_norm": 0.044397566467523575,
+      "learning_rate": 2.161098441622014e-05,
+      "loss": 1.0119,
+      "step": 2091
+    },
+    {
+      "epoch": 2.3587570621468927,
+      "grad_norm": 0.0443037785589695,
+      "learning_rate": 2.1537292528692598e-05,
+      "loss": 0.9441,
+      "step": 2092
+    },
+    {
+      "epoch": 2.3598870056497177,
+      "grad_norm": 0.043571796268224716,
+      "learning_rate": 2.146371133199191e-05,
+      "loss": 0.9358,
+      "step": 2093
+    },
+    {
+      "epoch": 2.3610169491525426,
+      "grad_norm": 0.045840222388505936,
+      "learning_rate": 2.13902409299223e-05,
+      "loss": 0.93,
+      "step": 2094
+    },
+    {
+      "epoch": 2.3621468926553675,
+      "grad_norm": 0.043795760720968246,
+      "learning_rate": 2.131688142613183e-05,
+      "loss": 1.0515,
+      "step": 2095
+    },
+    {
+      "epoch": 2.363276836158192,
+      "grad_norm": 0.04440205171704292,
+      "learning_rate": 2.124363292411189e-05,
+      "loss": 0.9891,
+      "step": 2096
+    },
+    {
+      "epoch": 2.364406779661017,
+      "grad_norm": 0.04491677135229111,
+      "learning_rate": 2.1170495527197487e-05,
+      "loss": 0.9825,
+      "step": 2097
+    },
+    {
+      "epoch": 2.365536723163842,
+      "grad_norm": 0.0445803739130497,
+      "learning_rate": 2.1097469338566754e-05,
+      "loss": 0.9716,
+      "step": 2098
+    },
+    {
+      "epoch": 2.3666666666666667,
+      "grad_norm": 0.04519984871149063,
+      "learning_rate": 2.1024554461240986e-05,
+      "loss": 0.9865,
+      "step": 2099
+    },
+    {
+      "epoch": 2.3677966101694916,
+      "grad_norm": 0.04491204023361206,
+      "learning_rate": 2.095175099808444e-05,
+      "loss": 0.9552,
+      "step": 2100
+    },
+    {
+      "epoch": 2.3689265536723165,
+      "grad_norm": 0.04308044910430908,
+      "learning_rate": 2.0879059051804206e-05,
+      "loss": 1.0304,
+      "step": 2101
+    },
+    {
+      "epoch": 2.370056497175141,
+      "grad_norm": 0.04669945314526558,
+      "learning_rate": 2.080647872495002e-05,
+      "loss": 1.0882,
+      "step": 2102
+    },
+    {
+      "epoch": 2.371186440677966,
+      "grad_norm": 0.04366466403007507,
+      "learning_rate": 2.0734010119914192e-05,
+      "loss": 1.0511,
+      "step": 2103
+    },
+    {
+      "epoch": 2.372316384180791,
+      "grad_norm": 0.04505183920264244,
+      "learning_rate": 2.0661653338931364e-05,
+      "loss": 0.9223,
+      "step": 2104
+    },
+    {
+      "epoch": 2.3734463276836157,
+      "grad_norm": 0.043640442192554474,
+      "learning_rate": 2.058940848407854e-05,
+      "loss": 0.9174,
+      "step": 2105
+    },
+    {
+      "epoch": 2.3745762711864407,
+      "grad_norm": 0.044932059943675995,
+      "learning_rate": 2.0517275657274683e-05,
+      "loss": 0.9584,
+      "step": 2106
+    },
+    {
+      "epoch": 2.3757062146892656,
+      "grad_norm": 0.04342171177268028,
+      "learning_rate": 2.0445254960280757e-05,
+      "loss": 0.9948,
+      "step": 2107
+    },
+    {
+      "epoch": 2.3768361581920905,
+      "grad_norm": 0.042078595608472824,
+      "learning_rate": 2.0373346494699618e-05,
+      "loss": 1.0606,
+      "step": 2108
+    },
+    {
+      "epoch": 2.3779661016949154,
+      "grad_norm": 0.04417315125465393,
+      "learning_rate": 2.030155036197574e-05,
+      "loss": 0.9545,
+      "step": 2109
+    },
+    {
+      "epoch": 2.3790960451977403,
+      "grad_norm": 0.04296877607703209,
+      "learning_rate": 2.0229866663395026e-05,
+      "loss": 0.9642,
+      "step": 2110
+    },
+    {
+      "epoch": 2.380225988700565,
+      "grad_norm": 0.04403414949774742,
+      "learning_rate": 2.0158295500084945e-05,
+      "loss": 0.9322,
+      "step": 2111
+    },
+    {
+      "epoch": 2.3813559322033897,
+      "grad_norm": 0.044930048286914825,
+      "learning_rate": 2.008683697301408e-05,
+      "loss": 0.9121,
+      "step": 2112
+    },
+    {
+      "epoch": 2.3824858757062146,
+      "grad_norm": 0.04404081776738167,
+      "learning_rate": 2.001549118299214e-05,
+      "loss": 0.9026,
+      "step": 2113
+    },
+    {
+      "epoch": 2.3836158192090395,
+      "grad_norm": 0.044441040605306625,
+      "learning_rate": 1.9944258230669833e-05,
+      "loss": 0.9232,
+      "step": 2114
+    },
+    {
+      "epoch": 2.3847457627118644,
+      "grad_norm": 0.04437883198261261,
+      "learning_rate": 1.987313821653861e-05,
+      "loss": 1.0162,
+      "step": 2115
+    },
+    {
+      "epoch": 2.3858757062146894,
+      "grad_norm": 0.04486614465713501,
+      "learning_rate": 1.9802131240930664e-05,
+      "loss": 0.9181,
+      "step": 2116
+    },
+    {
+      "epoch": 2.3870056497175143,
+      "grad_norm": 0.04392695426940918,
+      "learning_rate": 1.9731237404018676e-05,
+      "loss": 0.8908,
+      "step": 2117
+    },
+    {
+      "epoch": 2.3881355932203387,
+      "grad_norm": 0.047182969748973846,
+      "learning_rate": 1.966045680581572e-05,
+      "loss": 0.9931,
+      "step": 2118
+    },
+    {
+      "epoch": 2.3892655367231637,
+      "grad_norm": 0.04336957260966301,
+      "learning_rate": 1.9589789546175176e-05,
+      "loss": 0.9822,
+      "step": 2119
+    },
+    {
+      "epoch": 2.3903954802259886,
+      "grad_norm": 0.04483437165617943,
+      "learning_rate": 1.951923572479044e-05,
+      "loss": 1.024,
+      "step": 2120
+    },
+    {
+      "epoch": 2.3915254237288135,
+      "grad_norm": 0.045611586421728134,
+      "learning_rate": 1.944879544119491e-05,
+      "loss": 0.9401,
+      "step": 2121
+    },
+    {
+      "epoch": 2.3926553672316384,
+      "grad_norm": 0.04340572655200958,
+      "learning_rate": 1.9378468794761873e-05,
+      "loss": 1.0403,
+      "step": 2122
+    },
+    {
+      "epoch": 2.3937853107344633,
+      "grad_norm": 0.046446606516838074,
+      "learning_rate": 1.9308255884704253e-05,
+      "loss": 0.9327,
+      "step": 2123
+    },
+    {
+      "epoch": 2.3949152542372882,
+      "grad_norm": 0.0440828800201416,
+      "learning_rate": 1.9238156810074448e-05,
+      "loss": 0.9262,
+      "step": 2124
+    },
+    {
+      "epoch": 2.396045197740113,
+      "grad_norm": 0.04405950754880905,
+      "learning_rate": 1.9168171669764413e-05,
+      "loss": 0.9528,
+      "step": 2125
+    },
+    {
+      "epoch": 2.397175141242938,
+      "grad_norm": 0.044946879148483276,
+      "learning_rate": 1.9098300562505266e-05,
+      "loss": 0.9574,
+      "step": 2126
+    },
+    {
+      "epoch": 2.3983050847457625,
+      "grad_norm": 0.04468640685081482,
+      "learning_rate": 1.9028543586867276e-05,
+      "loss": 0.9745,
+      "step": 2127
+    },
+    {
+      "epoch": 2.3994350282485875,
+      "grad_norm": 0.04496829956769943,
+      "learning_rate": 1.895890084125973e-05,
+      "loss": 0.99,
+      "step": 2128
+    },
+    {
+      "epoch": 2.4005649717514124,
+      "grad_norm": 0.04439869895577431,
+      "learning_rate": 1.888937242393072e-05,
+      "loss": 0.9164,
+      "step": 2129
+    },
+    {
+      "epoch": 2.4016949152542373,
+      "grad_norm": 0.04531821236014366,
+      "learning_rate": 1.881995843296708e-05,
+      "loss": 1.0018,
+      "step": 2130
+    },
+    {
+      "epoch": 2.402824858757062,
+      "grad_norm": 0.043409690260887146,
+      "learning_rate": 1.87506589662942e-05,
+      "loss": 1.0898,
+      "step": 2131
+    },
+    {
+      "epoch": 2.403954802259887,
+      "grad_norm": 0.04595141485333443,
+      "learning_rate": 1.8681474121675913e-05,
+      "loss": 0.9994,
+      "step": 2132
+    },
+    {
+      "epoch": 2.405084745762712,
+      "grad_norm": 0.04307853803038597,
+      "learning_rate": 1.8612403996714413e-05,
+      "loss": 1.0001,
+      "step": 2133
+    },
+    {
+      "epoch": 2.4062146892655365,
+      "grad_norm": 0.045866724103689194,
+      "learning_rate": 1.8543448688849897e-05,
+      "loss": 0.9808,
+      "step": 2134
+    },
+    {
+      "epoch": 2.4073446327683614,
+      "grad_norm": 0.04307013005018234,
+      "learning_rate": 1.847460829536075e-05,
+      "loss": 0.8761,
+      "step": 2135
+    },
+    {
+      "epoch": 2.4084745762711863,
+      "grad_norm": 0.04371463879942894,
+      "learning_rate": 1.840588291336317e-05,
+      "loss": 0.9833,
+      "step": 2136
+    },
+    {
+      "epoch": 2.4096045197740112,
+      "grad_norm": 0.04334999993443489,
+      "learning_rate": 1.8337272639811075e-05,
+      "loss": 0.936,
+      "step": 2137
+    },
+    {
+      "epoch": 2.410734463276836,
+      "grad_norm": 0.04375695809721947,
+      "learning_rate": 1.8268777571496044e-05,
+      "loss": 0.9321,
+      "step": 2138
+    },
+    {
+      "epoch": 2.411864406779661,
+      "grad_norm": 0.044900234788656235,
+      "learning_rate": 1.8200397805047108e-05,
+      "loss": 0.8979,
+      "step": 2139
+    },
+    {
+      "epoch": 2.412994350282486,
+      "grad_norm": 0.04461583495140076,
+      "learning_rate": 1.8132133436930642e-05,
+      "loss": 1.0154,
+      "step": 2140
+    },
+    {
+      "epoch": 2.414124293785311,
+      "grad_norm": 0.043774548918008804,
+      "learning_rate": 1.806398456345023e-05,
+      "loss": 0.9681,
+      "step": 2141
+    },
+    {
+      "epoch": 2.415254237288136,
+      "grad_norm": 0.044340088963508606,
+      "learning_rate": 1.7995951280746503e-05,
+      "loss": 0.9608,
+      "step": 2142
+    },
+    {
+      "epoch": 2.4163841807909603,
+      "grad_norm": 0.043895475566387177,
+      "learning_rate": 1.792803368479704e-05,
+      "loss": 0.9723,
+      "step": 2143
+    },
+    {
+      "epoch": 2.417514124293785,
+      "grad_norm": 0.04527180269360542,
+      "learning_rate": 1.786023187141621e-05,
+      "loss": 0.9095,
+      "step": 2144
+    },
+    {
+      "epoch": 2.41864406779661,
+      "grad_norm": 0.04495590180158615,
+      "learning_rate": 1.7792545936255013e-05,
+      "loss": 0.9953,
+      "step": 2145
+    },
+    {
+      "epoch": 2.419774011299435,
+      "grad_norm": 0.04520578309893608,
+      "learning_rate": 1.7724975974801073e-05,
+      "loss": 0.9458,
+      "step": 2146
+    },
+    {
+      "epoch": 2.42090395480226,
+      "grad_norm": 0.04260654002428055,
+      "learning_rate": 1.765752208237831e-05,
+      "loss": 0.9465,
+      "step": 2147
+    },
+    {
+      "epoch": 2.422033898305085,
+      "grad_norm": 0.04469961300492287,
+      "learning_rate": 1.7590184354146867e-05,
+      "loss": 0.9411,
+      "step": 2148
+    },
+    {
+      "epoch": 2.42316384180791,
+      "grad_norm": 0.0442170612514019,
+      "learning_rate": 1.7522962885103145e-05,
+      "loss": 0.8851,
+      "step": 2149
+    },
+    {
+      "epoch": 2.4242937853107343,
+      "grad_norm": 0.04404238983988762,
+      "learning_rate": 1.745585777007943e-05,
+      "loss": 0.9593,
+      "step": 2150
+    },
+    {
+      "epoch": 2.425423728813559,
+      "grad_norm": 0.04603072255849838,
+      "learning_rate": 1.7388869103743878e-05,
+      "loss": 1.013,
+      "step": 2151
+    },
+    {
+      "epoch": 2.426553672316384,
+      "grad_norm": 0.04436330869793892,
+      "learning_rate": 1.7321996980600387e-05,
+      "loss": 0.9281,
+      "step": 2152
+    },
+    {
+      "epoch": 2.427683615819209,
+      "grad_norm": 0.043367523699998856,
+      "learning_rate": 1.7255241494988405e-05,
+      "loss": 1.0137,
+      "step": 2153
+    },
+    {
+      "epoch": 2.428813559322034,
+      "grad_norm": 0.046013541519641876,
+      "learning_rate": 1.7188602741082938e-05,
+      "loss": 0.964,
+      "step": 2154
+    },
+    {
+      "epoch": 2.429943502824859,
+      "grad_norm": 0.04216546192765236,
+      "learning_rate": 1.7122080812894147e-05,
+      "loss": 0.8503,
+      "step": 2155
+    },
+    {
+      "epoch": 2.4310734463276837,
+      "grad_norm": 0.04499104246497154,
+      "learning_rate": 1.705567580426749e-05,
+      "loss": 0.9901,
+      "step": 2156
+    },
+    {
+      "epoch": 2.4322033898305087,
+      "grad_norm": 0.04435711354017258,
+      "learning_rate": 1.698938780888354e-05,
+      "loss": 0.9676,
+      "step": 2157
+    },
+    {
+      "epoch": 2.4333333333333336,
+      "grad_norm": 0.04303577542304993,
+      "learning_rate": 1.6923216920257612e-05,
+      "loss": 0.9955,
+      "step": 2158
+    },
+    {
+      "epoch": 2.434463276836158,
+      "grad_norm": 0.04677002876996994,
+      "learning_rate": 1.6857163231739948e-05,
+      "loss": 0.9851,
+      "step": 2159
+    },
+    {
+      "epoch": 2.435593220338983,
+      "grad_norm": 0.045381397008895874,
+      "learning_rate": 1.679122683651546e-05,
+      "loss": 1.0507,
+      "step": 2160
+    },
+    {
+      "epoch": 2.436723163841808,
+      "grad_norm": 0.044448669999837875,
+      "learning_rate": 1.6725407827603546e-05,
+      "loss": 1.0459,
+      "step": 2161
+    },
+    {
+      "epoch": 2.437853107344633,
+      "grad_norm": 0.0439726784825325,
+      "learning_rate": 1.6659706297857945e-05,
+      "loss": 0.9604,
+      "step": 2162
+    },
+    {
+      "epoch": 2.4389830508474577,
+      "grad_norm": 0.04220819100737572,
+      "learning_rate": 1.6594122339966778e-05,
+      "loss": 0.9844,
+      "step": 2163
+    },
+    {
+      "epoch": 2.4401129943502826,
+      "grad_norm": 0.044570211321115494,
+      "learning_rate": 1.6528656046452228e-05,
+      "loss": 0.9174,
+      "step": 2164
+    },
+    {
+      "epoch": 2.4412429378531075,
+      "grad_norm": 0.04373630881309509,
+      "learning_rate": 1.6463307509670524e-05,
+      "loss": 0.9263,
+      "step": 2165
+    },
+    {
+      "epoch": 2.442372881355932,
+      "grad_norm": 0.043320853263139725,
+      "learning_rate": 1.6398076821811724e-05,
+      "loss": 0.9403,
+      "step": 2166
+    },
+    {
+      "epoch": 2.443502824858757,
+      "grad_norm": 0.04534811154007912,
+      "learning_rate": 1.6332964074899636e-05,
+      "loss": 1.05,
+      "step": 2167
+    },
+    {
+      "epoch": 2.444632768361582,
+      "grad_norm": 0.04526575654745102,
+      "learning_rate": 1.626796936079179e-05,
+      "loss": 0.9709,
+      "step": 2168
+    },
+    {
+      "epoch": 2.4457627118644067,
+      "grad_norm": 0.04434860870242119,
+      "learning_rate": 1.620309277117904e-05,
+      "loss": 0.9629,
+      "step": 2169
+    },
+    {
+      "epoch": 2.4468926553672317,
+      "grad_norm": 0.04492155835032463,
+      "learning_rate": 1.6138334397585675e-05,
+      "loss": 0.9906,
+      "step": 2170
+    },
+    {
+      "epoch": 2.4480225988700566,
+      "grad_norm": 0.04469876363873482,
+      "learning_rate": 1.6073694331369272e-05,
+      "loss": 1.0313,
+      "step": 2171
+    },
+    {
+      "epoch": 2.4491525423728815,
+      "grad_norm": 0.045252226293087006,
+      "learning_rate": 1.600917266372035e-05,
+      "loss": 0.908,
+      "step": 2172
+    },
+    {
+      "epoch": 2.4502824858757064,
+      "grad_norm": 0.044812172651290894,
+      "learning_rate": 1.5944769485662568e-05,
+      "loss": 0.9541,
+      "step": 2173
+    },
+    {
+      "epoch": 2.4514124293785313,
+      "grad_norm": 0.04414849728345871,
+      "learning_rate": 1.5880484888052328e-05,
+      "loss": 0.9989,
+      "step": 2174
+    },
+    {
+      "epoch": 2.452542372881356,
+      "grad_norm": 0.04734036698937416,
+      "learning_rate": 1.5816318961578757e-05,
+      "loss": 0.9075,
+      "step": 2175
+    },
+    {
+      "epoch": 2.4536723163841807,
+      "grad_norm": 0.044426627457141876,
+      "learning_rate": 1.5752271796763584e-05,
+      "loss": 0.9566,
+      "step": 2176
+    },
+    {
+      "epoch": 2.4548022598870056,
+      "grad_norm": 0.04688658565282822,
+      "learning_rate": 1.5688343483961e-05,
+      "loss": 1.0372,
+      "step": 2177
+    },
+    {
+      "epoch": 2.4559322033898305,
+      "grad_norm": 0.04371168091893196,
+      "learning_rate": 1.5624534113357493e-05,
+      "loss": 0.9158,
+      "step": 2178
+    },
+    {
+      "epoch": 2.4570621468926555,
+      "grad_norm": 0.045110173523426056,
+      "learning_rate": 1.55608437749718e-05,
+      "loss": 0.9342,
+      "step": 2179
+    },
+    {
+      "epoch": 2.4581920903954804,
+      "grad_norm": 0.04536379873752594,
+      "learning_rate": 1.5497272558654697e-05,
+      "loss": 0.9582,
+      "step": 2180
+    },
+    {
+      "epoch": 2.459322033898305,
+      "grad_norm": 0.043405305594205856,
+      "learning_rate": 1.5433820554088895e-05,
+      "loss": 0.9627,
+      "step": 2181
+    },
+    {
+      "epoch": 2.4604519774011298,
+      "grad_norm": 0.0462508387863636,
+      "learning_rate": 1.537048785078905e-05,
+      "loss": 0.9083,
+      "step": 2182
+    },
+    {
+      "epoch": 2.4615819209039547,
+      "grad_norm": 0.04306486248970032,
+      "learning_rate": 1.5307274538101292e-05,
+      "loss": 0.9715,
+      "step": 2183
+    },
+    {
+      "epoch": 2.4627118644067796,
+      "grad_norm": 0.04309242591261864,
+      "learning_rate": 1.5244180705203547e-05,
+      "loss": 0.946,
+      "step": 2184
+    },
+    {
+      "epoch": 2.4638418079096045,
+      "grad_norm": 0.04370222985744476,
+      "learning_rate": 1.5181206441105078e-05,
+      "loss": 0.9582,
+      "step": 2185
+    },
+    {
+      "epoch": 2.4649717514124294,
+      "grad_norm": 0.04389223828911781,
+      "learning_rate": 1.5118351834646405e-05,
+      "loss": 0.916,
+      "step": 2186
+    },
+    {
+      "epoch": 2.4661016949152543,
+      "grad_norm": 0.04450360685586929,
+      "learning_rate": 1.5055616974499375e-05,
+      "loss": 0.9326,
+      "step": 2187
+    },
+    {
+      "epoch": 2.4672316384180792,
+      "grad_norm": 0.04638772830367088,
+      "learning_rate": 1.499300194916684e-05,
+      "loss": 1.0533,
+      "step": 2188
+    },
+    {
+      "epoch": 2.468361581920904,
+      "grad_norm": 0.04407142847776413,
+      "learning_rate": 1.4930506846982585e-05,
+      "loss": 0.9369,
+      "step": 2189
+    },
+    {
+      "epoch": 2.4694915254237286,
+      "grad_norm": 0.04272819682955742,
+      "learning_rate": 1.4868131756111225e-05,
+      "loss": 1.03,
+      "step": 2190
+    },
+    {
+      "epoch": 2.4706214689265535,
+      "grad_norm": 0.044162772595882416,
+      "learning_rate": 1.4805876764548077e-05,
+      "loss": 0.9255,
+      "step": 2191
+    },
+    {
+      "epoch": 2.4717514124293785,
+      "grad_norm": 0.04563014954328537,
+      "learning_rate": 1.4743741960119018e-05,
+      "loss": 0.9713,
+      "step": 2192
+    },
+    {
+      "epoch": 2.4728813559322034,
+      "grad_norm": 0.04250924661755562,
+      "learning_rate": 1.4681727430480375e-05,
+      "loss": 1.0386,
+      "step": 2193
+    },
+    {
+      "epoch": 2.4740112994350283,
+      "grad_norm": 0.044044073671102524,
+      "learning_rate": 1.4619833263118788e-05,
+      "loss": 0.9318,
+      "step": 2194
+    },
+    {
+      "epoch": 2.475141242937853,
+      "grad_norm": 0.04529934003949165,
+      "learning_rate": 1.4558059545351143e-05,
+      "loss": 1.1511,
+      "step": 2195
+    },
+    {
+      "epoch": 2.476271186440678,
+      "grad_norm": 0.04386807605624199,
+      "learning_rate": 1.4496406364324367e-05,
+      "loss": 0.9201,
+      "step": 2196
+    },
+    {
+      "epoch": 2.4774011299435026,
+      "grad_norm": 0.04460158571600914,
+      "learning_rate": 1.4434873807015281e-05,
+      "loss": 1.0432,
+      "step": 2197
+    },
+    {
+      "epoch": 2.4785310734463275,
+      "grad_norm": 0.04447031766176224,
+      "learning_rate": 1.4373461960230671e-05,
+      "loss": 0.872,
+      "step": 2198
+    },
+    {
+      "epoch": 2.4796610169491524,
+      "grad_norm": 0.04573468118906021,
+      "learning_rate": 1.4312170910606937e-05,
+      "loss": 0.9578,
+      "step": 2199
+    },
+    {
+      "epoch": 2.4807909604519773,
+      "grad_norm": 0.04439990594983101,
+      "learning_rate": 1.4251000744610033e-05,
+      "loss": 0.9185,
+      "step": 2200
+    },
+    {
+      "epoch": 2.4819209039548022,
+      "grad_norm": 0.045020122081041336,
+      "learning_rate": 1.4189951548535496e-05,
+      "loss": 0.9954,
+      "step": 2201
+    },
+    {
+      "epoch": 2.483050847457627,
+      "grad_norm": 0.04263998568058014,
+      "learning_rate": 1.4129023408508101e-05,
+      "loss": 0.9367,
+      "step": 2202
+    },
+    {
+      "epoch": 2.484180790960452,
+      "grad_norm": 0.04334124177694321,
+      "learning_rate": 1.4068216410481904e-05,
+      "loss": 0.9856,
+      "step": 2203
+    },
+    {
+      "epoch": 2.485310734463277,
+      "grad_norm": 0.04389631003141403,
+      "learning_rate": 1.4007530640240029e-05,
+      "loss": 0.8721,
+      "step": 2204
+    },
+    {
+      "epoch": 2.486440677966102,
+      "grad_norm": 0.046165741980075836,
+      "learning_rate": 1.394696618339456e-05,
+      "loss": 1.0324,
+      "step": 2205
+    },
+    {
+      "epoch": 2.4875706214689264,
+      "grad_norm": 0.04491398110985756,
+      "learning_rate": 1.388652312538653e-05,
+      "loss": 0.9486,
+      "step": 2206
+    },
+    {
+      "epoch": 2.4887005649717513,
+      "grad_norm": 0.043603088706731796,
+      "learning_rate": 1.3826201551485596e-05,
+      "loss": 0.9196,
+      "step": 2207
+    },
+    {
+      "epoch": 2.489830508474576,
+      "grad_norm": 0.04453226923942566,
+      "learning_rate": 1.3766001546790075e-05,
+      "loss": 0.9218,
+      "step": 2208
+    },
+    {
+      "epoch": 2.490960451977401,
+      "grad_norm": 0.04467688500881195,
+      "learning_rate": 1.3705923196226832e-05,
+      "loss": 0.8989,
+      "step": 2209
+    },
+    {
+      "epoch": 2.492090395480226,
+      "grad_norm": 0.04468687251210213,
+      "learning_rate": 1.364596658455105e-05,
+      "loss": 0.894,
+      "step": 2210
+    },
+    {
+      "epoch": 2.493220338983051,
+      "grad_norm": 0.043691717088222504,
+      "learning_rate": 1.3586131796346147e-05,
+      "loss": 0.9544,
+      "step": 2211
+    },
+    {
+      "epoch": 2.494350282485876,
+      "grad_norm": 0.04409053549170494,
+      "learning_rate": 1.3526418916023753e-05,
+      "loss": 0.9463,
+      "step": 2212
+    },
+    {
+      "epoch": 2.4954802259887003,
+      "grad_norm": 0.0438973568379879,
+      "learning_rate": 1.3466828027823475e-05,
+      "loss": 1.0117,
+      "step": 2213
+    },
+    {
+      "epoch": 2.4966101694915253,
+      "grad_norm": 0.04333541914820671,
+      "learning_rate": 1.3407359215812832e-05,
+      "loss": 1.0121,
+      "step": 2214
+    },
+    {
+      "epoch": 2.49774011299435,
+      "grad_norm": 0.04311882331967354,
+      "learning_rate": 1.3348012563887102e-05,
+      "loss": 0.9785,
+      "step": 2215
+    },
+    {
+      "epoch": 2.498870056497175,
+      "grad_norm": 0.044175051152706146,
+      "learning_rate": 1.328878815576926e-05,
+      "loss": 1.0375,
+      "step": 2216
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.04385006055235863,
+      "learning_rate": 1.3229686075009795e-05,
+      "loss": 0.9311,
+      "step": 2217
+    },
+    {
+      "epoch": 2.501129943502825,
+      "grad_norm": 0.044461339712142944,
+      "learning_rate": 1.3170706404986644e-05,
+      "loss": 0.9383,
+      "step": 2218
+    },
+    {
+      "epoch": 2.50225988700565,
+      "grad_norm": 0.042694512754678726,
+      "learning_rate": 1.3111849228905026e-05,
+      "loss": 0.9092,
+      "step": 2219
+    },
+    {
+      "epoch": 2.5033898305084747,
+      "grad_norm": 0.04385200887918472,
+      "learning_rate": 1.3053114629797437e-05,
+      "loss": 0.9969,
+      "step": 2220
+    },
+    {
+      "epoch": 2.5045197740112997,
+      "grad_norm": 0.04622391611337662,
+      "learning_rate": 1.2994502690523303e-05,
+      "loss": 0.9167,
+      "step": 2221
+    },
+    {
+      "epoch": 2.505649717514124,
+      "grad_norm": 0.043433066457509995,
+      "learning_rate": 1.293601349376915e-05,
+      "loss": 1.0352,
+      "step": 2222
+    },
+    {
+      "epoch": 2.506779661016949,
+      "grad_norm": 0.04414517059922218,
+      "learning_rate": 1.2877647122048265e-05,
+      "loss": 1.0007,
+      "step": 2223
+    },
+    {
+      "epoch": 2.507909604519774,
+      "grad_norm": 0.04432892054319382,
+      "learning_rate": 1.281940365770068e-05,
+      "loss": 0.9849,
+      "step": 2224
+    },
+    {
+      "epoch": 2.509039548022599,
+      "grad_norm": 0.04392955079674721,
+      "learning_rate": 1.2761283182893047e-05,
+      "loss": 0.9475,
+      "step": 2225
+    },
+    {
+      "epoch": 2.510169491525424,
+      "grad_norm": 0.045272983610630035,
+      "learning_rate": 1.2703285779618491e-05,
+      "loss": 0.9326,
+      "step": 2226
+    },
+    {
+      "epoch": 2.5112994350282487,
+      "grad_norm": 0.04357200860977173,
+      "learning_rate": 1.2645411529696527e-05,
+      "loss": 1.004,
+      "step": 2227
+    },
+    {
+      "epoch": 2.512429378531073,
+      "grad_norm": 0.04428897425532341,
+      "learning_rate": 1.2587660514772936e-05,
+      "loss": 1.0016,
+      "step": 2228
+    },
+    {
+      "epoch": 2.513559322033898,
+      "grad_norm": 0.042956896126270294,
+      "learning_rate": 1.2530032816319637e-05,
+      "loss": 0.9741,
+      "step": 2229
+    },
+    {
+      "epoch": 2.514689265536723,
+      "grad_norm": 0.04535180702805519,
+      "learning_rate": 1.2472528515634584e-05,
+      "loss": 0.9554,
+      "step": 2230
+    },
+    {
+      "epoch": 2.515819209039548,
+      "grad_norm": 0.044575612992048264,
+      "learning_rate": 1.2415147693841644e-05,
+      "loss": 1.0013,
+      "step": 2231
+    },
+    {
+      "epoch": 2.516949152542373,
+      "grad_norm": 0.04531654715538025,
+      "learning_rate": 1.2357890431890473e-05,
+      "loss": 0.9653,
+      "step": 2232
+    },
+    {
+      "epoch": 2.5180790960451978,
+      "grad_norm": 0.044301070272922516,
+      "learning_rate": 1.230075681055648e-05,
+      "loss": 0.9754,
+      "step": 2233
+    },
+    {
+      "epoch": 2.5192090395480227,
+      "grad_norm": 0.04337549954652786,
+      "learning_rate": 1.22437469104406e-05,
+      "loss": 0.9641,
+      "step": 2234
+    },
+    {
+      "epoch": 2.5203389830508476,
+      "grad_norm": 0.044852569699287415,
+      "learning_rate": 1.218686081196917e-05,
+      "loss": 0.9933,
+      "step": 2235
+    },
+    {
+      "epoch": 2.5214689265536725,
+      "grad_norm": 0.0434139259159565,
+      "learning_rate": 1.213009859539399e-05,
+      "loss": 0.8976,
+      "step": 2236
+    },
+    {
+      "epoch": 2.5225988700564974,
+      "grad_norm": 0.043631184846162796,
+      "learning_rate": 1.207346034079203e-05,
+      "loss": 0.9398,
+      "step": 2237
+    },
+    {
+      "epoch": 2.523728813559322,
+      "grad_norm": 0.0428677536547184,
+      "learning_rate": 1.2016946128065387e-05,
+      "loss": 0.9076,
+      "step": 2238
+    },
+    {
+      "epoch": 2.524858757062147,
+      "grad_norm": 0.04406916722655296,
+      "learning_rate": 1.1960556036941174e-05,
+      "loss": 0.9515,
+      "step": 2239
+    },
+    {
+      "epoch": 2.5259887005649717,
+      "grad_norm": 0.04295162111520767,
+      "learning_rate": 1.1904290146971397e-05,
+      "loss": 0.9814,
+      "step": 2240
+    },
+    {
+      "epoch": 2.5271186440677966,
+      "grad_norm": 0.043710824102163315,
+      "learning_rate": 1.1848148537532843e-05,
+      "loss": 0.9396,
+      "step": 2241
+    },
+    {
+      "epoch": 2.5282485875706215,
+      "grad_norm": 0.044418223202228546,
+      "learning_rate": 1.179213128782698e-05,
+      "loss": 0.9687,
+      "step": 2242
+    },
+    {
+      "epoch": 2.5293785310734465,
+      "grad_norm": 0.043215930461883545,
+      "learning_rate": 1.1736238476879802e-05,
+      "loss": 0.9498,
+      "step": 2243
+    },
+    {
+      "epoch": 2.530508474576271,
+      "grad_norm": 0.04229172319173813,
+      "learning_rate": 1.1680470183541847e-05,
+      "loss": 0.9091,
+      "step": 2244
+    },
+    {
+      "epoch": 2.531638418079096,
+      "grad_norm": 0.04341178387403488,
+      "learning_rate": 1.1624826486487871e-05,
+      "loss": 1.0278,
+      "step": 2245
+    },
+    {
+      "epoch": 2.5327683615819208,
+      "grad_norm": 0.046501412987709045,
+      "learning_rate": 1.1569307464216895e-05,
+      "loss": 1.0282,
+      "step": 2246
+    },
+    {
+      "epoch": 2.5338983050847457,
+      "grad_norm": 0.04544202238321304,
+      "learning_rate": 1.1513913195052107e-05,
+      "loss": 0.9748,
+      "step": 2247
+    },
+    {
+      "epoch": 2.5350282485875706,
+      "grad_norm": 0.045019011944532394,
+      "learning_rate": 1.1458643757140674e-05,
+      "loss": 0.9622,
+      "step": 2248
+    },
+    {
+      "epoch": 2.5361581920903955,
+      "grad_norm": 0.04544651135802269,
+      "learning_rate": 1.1403499228453596e-05,
+      "loss": 0.9654,
+      "step": 2249
+    },
+    {
+      "epoch": 2.5372881355932204,
+      "grad_norm": 0.04545415937900543,
+      "learning_rate": 1.1348479686785751e-05,
+      "loss": 0.8964,
+      "step": 2250
+    },
+    {
+      "epoch": 2.5384180790960453,
+      "grad_norm": 0.0433647446334362,
+      "learning_rate": 1.1293585209755631e-05,
+      "loss": 0.8951,
+      "step": 2251
+    },
+    {
+      "epoch": 2.5395480225988702,
+      "grad_norm": 0.045199718326330185,
+      "learning_rate": 1.123881587480533e-05,
+      "loss": 0.9357,
+      "step": 2252
+    },
+    {
+      "epoch": 2.540677966101695,
+      "grad_norm": 0.04812634736299515,
+      "learning_rate": 1.1184171759200368e-05,
+      "loss": 0.8831,
+      "step": 2253
+    },
+    {
+      "epoch": 2.5418079096045196,
+      "grad_norm": 0.04364898055791855,
+      "learning_rate": 1.1129652940029623e-05,
+      "loss": 0.9079,
+      "step": 2254
+    },
+    {
+      "epoch": 2.5429378531073445,
+      "grad_norm": 0.044137731194496155,
+      "learning_rate": 1.1075259494205225e-05,
+      "loss": 1.0158,
+      "step": 2255
+    },
+    {
+      "epoch": 2.5440677966101695,
+      "grad_norm": 0.044673506170511246,
+      "learning_rate": 1.102099149846243e-05,
+      "loss": 0.9106,
+      "step": 2256
+    },
+    {
+      "epoch": 2.5451977401129944,
+      "grad_norm": 0.04392934590578079,
+      "learning_rate": 1.0966849029359472e-05,
+      "loss": 0.9462,
+      "step": 2257
+    },
+    {
+      "epoch": 2.5463276836158193,
+      "grad_norm": 0.044603992253541946,
+      "learning_rate": 1.0912832163277609e-05,
+      "loss": 1.0413,
+      "step": 2258
+    },
+    {
+      "epoch": 2.547457627118644,
+      "grad_norm": 0.043722860515117645,
+      "learning_rate": 1.0858940976420772e-05,
+      "loss": 1.0259,
+      "step": 2259
+    },
+    {
+      "epoch": 2.5485875706214687,
+      "grad_norm": 0.04358895495533943,
+      "learning_rate": 1.0805175544815648e-05,
+      "loss": 0.907,
+      "step": 2260
+    },
+    {
+      "epoch": 2.5497175141242936,
+      "grad_norm": 0.0456516295671463,
+      "learning_rate": 1.0751535944311564e-05,
+      "loss": 0.9535,
+      "step": 2261
+    },
+    {
+      "epoch": 2.5508474576271185,
+      "grad_norm": 0.046546079218387604,
+      "learning_rate": 1.0698022250580264e-05,
+      "loss": 1.042,
+      "step": 2262
+    },
+    {
+      "epoch": 2.5519774011299434,
+      "grad_norm": 0.043882519006729126,
+      "learning_rate": 1.0644634539115906e-05,
+      "loss": 0.883,
+      "step": 2263
+    },
+    {
+      "epoch": 2.5531073446327683,
+      "grad_norm": 0.04427448660135269,
+      "learning_rate": 1.0591372885234885e-05,
+      "loss": 0.9493,
+      "step": 2264
+    },
+    {
+      "epoch": 2.5542372881355933,
+      "grad_norm": 0.04398886859416962,
+      "learning_rate": 1.0538237364075787e-05,
+      "loss": 1.0124,
+      "step": 2265
+    },
+    {
+      "epoch": 2.555367231638418,
+      "grad_norm": 0.04535575583577156,
+      "learning_rate": 1.0485228050599272e-05,
+      "loss": 0.9521,
+      "step": 2266
+    },
+    {
+      "epoch": 2.556497175141243,
+      "grad_norm": 0.04494888707995415,
+      "learning_rate": 1.0432345019587908e-05,
+      "loss": 0.9673,
+      "step": 2267
+    },
+    {
+      "epoch": 2.557627118644068,
+      "grad_norm": 0.04512500762939453,
+      "learning_rate": 1.0379588345646152e-05,
+      "loss": 0.9744,
+      "step": 2268
+    },
+    {
+      "epoch": 2.558757062146893,
+      "grad_norm": 0.045277807861566544,
+      "learning_rate": 1.032695810320018e-05,
+      "loss": 0.9401,
+      "step": 2269
+    },
+    {
+      "epoch": 2.5598870056497174,
+      "grad_norm": 0.044249020516872406,
+      "learning_rate": 1.0274454366497787e-05,
+      "loss": 1.0551,
+      "step": 2270
+    },
+    {
+      "epoch": 2.5610169491525423,
+      "grad_norm": 0.04531550407409668,
+      "learning_rate": 1.0222077209608383e-05,
+      "loss": 0.9485,
+      "step": 2271
+    },
+    {
+      "epoch": 2.562146892655367,
+      "grad_norm": 0.044957417994737625,
+      "learning_rate": 1.0169826706422735e-05,
+      "loss": 1.0354,
+      "step": 2272
+    },
+    {
+      "epoch": 2.563276836158192,
+      "grad_norm": 0.04343345761299133,
+      "learning_rate": 1.0117702930652906e-05,
+      "loss": 0.9992,
+      "step": 2273
+    },
+    {
+      "epoch": 2.564406779661017,
+      "grad_norm": 0.0441376157104969,
+      "learning_rate": 1.006570595583226e-05,
+      "loss": 0.8785,
+      "step": 2274
+    },
+    {
+      "epoch": 2.565536723163842,
+      "grad_norm": 0.04558190703392029,
+      "learning_rate": 1.0013835855315235e-05,
+      "loss": 1.0488,
+      "step": 2275
+    },
+    {
+      "epoch": 2.5666666666666664,
+      "grad_norm": 0.04628458246588707,
+      "learning_rate": 9.962092702277282e-06,
+      "loss": 0.9738,
+      "step": 2276
+    },
+    {
+      "epoch": 2.5677966101694913,
+      "grad_norm": 0.046162836253643036,
+      "learning_rate": 9.910476569714766e-06,
+      "loss": 1.0152,
+      "step": 2277
+    },
+    {
+      "epoch": 2.5689265536723163,
+      "grad_norm": 0.0438779816031456,
+      "learning_rate": 9.858987530444852e-06,
+      "loss": 0.9672,
+      "step": 2278
+    },
+    {
+      "epoch": 2.570056497175141,
+      "grad_norm": 0.0449378527700901,
+      "learning_rate": 9.807625657105424e-06,
+      "loss": 0.8919,
+      "step": 2279
+    },
+    {
+      "epoch": 2.571186440677966,
+      "grad_norm": 0.04468364268541336,
+      "learning_rate": 9.756391022154954e-06,
+      "loss": 0.9767,
+      "step": 2280
+    },
+    {
+      "epoch": 2.572316384180791,
+      "grad_norm": 0.05266987159848213,
+      "learning_rate": 9.705283697872402e-06,
+      "loss": 1.0026,
+      "step": 2281
+    },
+    {
+      "epoch": 2.573446327683616,
+      "grad_norm": 0.044424451887607574,
+      "learning_rate": 9.654303756357186e-06,
+      "loss": 0.9913,
+      "step": 2282
+    },
+    {
+      "epoch": 2.574576271186441,
+      "grad_norm": 0.045492034405469894,
+      "learning_rate": 9.60345126952893e-06,
+      "loss": 0.9934,
+      "step": 2283
+    },
+    {
+      "epoch": 2.5757062146892657,
+      "grad_norm": 0.04521951451897621,
+      "learning_rate": 9.552726309127492e-06,
+      "loss": 0.9527,
+      "step": 2284
+    },
+    {
+      "epoch": 2.5768361581920907,
+      "grad_norm": 0.04271269962191582,
+      "learning_rate": 9.502128946712862e-06,
+      "loss": 0.9101,
+      "step": 2285
+    },
+    {
+      "epoch": 2.577966101694915,
+      "grad_norm": 0.043572571128606796,
+      "learning_rate": 9.45165925366498e-06,
+      "loss": 0.9369,
+      "step": 2286
+    },
+    {
+      "epoch": 2.57909604519774,
+      "grad_norm": 0.04333474859595299,
+      "learning_rate": 9.401317301183655e-06,
+      "loss": 1.0792,
+      "step": 2287
+    },
+    {
+      "epoch": 2.580225988700565,
+      "grad_norm": 0.044597212225198746,
+      "learning_rate": 9.351103160288533e-06,
+      "loss": 0.9374,
+      "step": 2288
+    },
+    {
+      "epoch": 2.58135593220339,
+      "grad_norm": 0.04439055919647217,
+      "learning_rate": 9.301016901818948e-06,
+      "loss": 0.869,
+      "step": 2289
+    },
+    {
+      "epoch": 2.582485875706215,
+      "grad_norm": 0.04525582492351532,
+      "learning_rate": 9.251058596433793e-06,
+      "loss": 1.0324,
+      "step": 2290
+    },
+    {
+      "epoch": 2.5836158192090397,
+      "grad_norm": 0.04282711073756218,
+      "learning_rate": 9.201228314611477e-06,
+      "loss": 0.9317,
+      "step": 2291
+    },
+    {
+      "epoch": 2.584745762711864,
+      "grad_norm": 0.044124454259872437,
+      "learning_rate": 9.151526126649779e-06,
+      "loss": 1.0026,
+      "step": 2292
+    },
+    {
+      "epoch": 2.585875706214689,
+      "grad_norm": 0.043038588017225266,
+      "learning_rate": 9.10195210266579e-06,
+      "loss": 0.9802,
+      "step": 2293
+    },
+    {
+      "epoch": 2.587005649717514,
+      "grad_norm": 0.04550214856863022,
+      "learning_rate": 9.052506312595798e-06,
+      "loss": 1.0226,
+      "step": 2294
+    },
+    {
+      "epoch": 2.588135593220339,
+      "grad_norm": 0.0450584851205349,
+      "learning_rate": 9.003188826195142e-06,
+      "loss": 1.0131,
+      "step": 2295
+    },
+    {
+      "epoch": 2.589265536723164,
+      "grad_norm": 0.0439411997795105,
+      "learning_rate": 8.953999713038252e-06,
+      "loss": 1.002,
+      "step": 2296
+    },
+    {
+      "epoch": 2.5903954802259888,
+      "grad_norm": 0.04387841746211052,
+      "learning_rate": 8.904939042518345e-06,
+      "loss": 0.993,
+      "step": 2297
+    },
+    {
+      "epoch": 2.5915254237288137,
+      "grad_norm": 0.046317506581544876,
+      "learning_rate": 8.856006883847479e-06,
+      "loss": 0.9878,
+      "step": 2298
+    },
+    {
+      "epoch": 2.5926553672316386,
+      "grad_norm": 0.04398587346076965,
+      "learning_rate": 8.807203306056466e-06,
+      "loss": 0.9364,
+      "step": 2299
+    },
+    {
+      "epoch": 2.5937853107344635,
+      "grad_norm": 0.044241469353437424,
+      "learning_rate": 8.758528377994667e-06,
+      "loss": 0.9447,
+      "step": 2300
+    },
+    {
+      "epoch": 2.594915254237288,
+      "grad_norm": 0.045419175177812576,
+      "learning_rate": 8.709982168329955e-06,
+      "loss": 0.9441,
+      "step": 2301
+    },
+    {
+      "epoch": 2.596045197740113,
+      "grad_norm": 0.04503266513347626,
+      "learning_rate": 8.661564745548634e-06,
+      "loss": 0.9268,
+      "step": 2302
+    },
+    {
+      "epoch": 2.597175141242938,
+      "grad_norm": 0.04366137087345123,
+      "learning_rate": 8.613276177955309e-06,
+      "loss": 0.9875,
+      "step": 2303
+    },
+    {
+      "epoch": 2.5983050847457627,
+      "grad_norm": 0.045485030859708786,
+      "learning_rate": 8.565116533672801e-06,
+      "loss": 0.9896,
+      "step": 2304
+    },
+    {
+      "epoch": 2.5994350282485876,
+      "grad_norm": 0.04749319702386856,
+      "learning_rate": 8.517085880642062e-06,
+      "loss": 0.9509,
+      "step": 2305
+    },
+    {
+      "epoch": 2.6005649717514125,
+      "grad_norm": 0.044025685638189316,
+      "learning_rate": 8.469184286622078e-06,
+      "loss": 0.9259,
+      "step": 2306
+    },
+    {
+      "epoch": 2.601694915254237,
+      "grad_norm": 0.04714653640985489,
+      "learning_rate": 8.42141181918975e-06,
+      "loss": 0.986,
+      "step": 2307
+    },
+    {
+      "epoch": 2.602824858757062,
+      "grad_norm": 0.04400424286723137,
+      "learning_rate": 8.373768545739813e-06,
+      "loss": 0.9512,
+      "step": 2308
+    },
+    {
+      "epoch": 2.603954802259887,
+      "grad_norm": 0.0437365286052227,
+      "learning_rate": 8.326254533484745e-06,
+      "loss": 1.0059,
+      "step": 2309
+    },
+    {
+      "epoch": 2.6050847457627118,
+      "grad_norm": 0.045271988958120346,
+      "learning_rate": 8.278869849454718e-06,
+      "loss": 1.0108,
+      "step": 2310
+    },
+    {
+      "epoch": 2.6062146892655367,
+      "grad_norm": 0.04391276836395264,
+      "learning_rate": 8.231614560497337e-06,
+      "loss": 0.9316,
+      "step": 2311
+    },
+    {
+      "epoch": 2.6073446327683616,
+      "grad_norm": 0.04322647675871849,
+      "learning_rate": 8.184488733277796e-06,
+      "loss": 0.954,
+      "step": 2312
+    },
+    {
+      "epoch": 2.6084745762711865,
+      "grad_norm": 0.04381094500422478,
+      "learning_rate": 8.13749243427857e-06,
+      "loss": 0.996,
+      "step": 2313
+    },
+    {
+      "epoch": 2.6096045197740114,
+      "grad_norm": 0.04461012780666351,
+      "learning_rate": 8.090625729799429e-06,
+      "loss": 0.9191,
+      "step": 2314
+    },
+    {
+      "epoch": 2.6107344632768363,
+      "grad_norm": 0.04649262875318527,
+      "learning_rate": 8.043888685957313e-06,
+      "loss": 0.9661,
+      "step": 2315
+    },
+    {
+      "epoch": 2.6118644067796613,
+      "grad_norm": 0.043844398111104965,
+      "learning_rate": 7.99728136868626e-06,
+      "loss": 1.0126,
+      "step": 2316
+    },
+    {
+      "epoch": 2.6129943502824857,
+      "grad_norm": 0.04373704642057419,
+      "learning_rate": 7.950803843737264e-06,
+      "loss": 1.0559,
+      "step": 2317
+    },
+    {
+      "epoch": 2.6141242937853106,
+      "grad_norm": 0.04947759956121445,
+      "learning_rate": 7.904456176678232e-06,
+      "loss": 0.8689,
+      "step": 2318
+    },
+    {
+      "epoch": 2.6152542372881356,
+      "grad_norm": 0.04527619481086731,
+      "learning_rate": 7.858238432893883e-06,
+      "loss": 0.9564,
+      "step": 2319
+    },
+    {
+      "epoch": 2.6163841807909605,
+      "grad_norm": 0.04394813999533653,
+      "learning_rate": 7.812150677585673e-06,
+      "loss": 0.8771,
+      "step": 2320
+    },
+    {
+      "epoch": 2.6175141242937854,
+      "grad_norm": 0.04304569587111473,
+      "learning_rate": 7.766192975771591e-06,
+      "loss": 0.9332,
+      "step": 2321
+    },
+    {
+      "epoch": 2.6186440677966103,
+      "grad_norm": 0.04365582764148712,
+      "learning_rate": 7.720365392286222e-06,
+      "loss": 1.0163,
+      "step": 2322
+    },
+    {
+      "epoch": 2.6197740112994348,
+      "grad_norm": 0.045340195298194885,
+      "learning_rate": 7.674667991780604e-06,
+      "loss": 0.9353,
+      "step": 2323
+    },
+    {
+      "epoch": 2.6209039548022597,
+      "grad_norm": 0.045364297926425934,
+      "learning_rate": 7.6291008387220916e-06,
+      "loss": 0.9099,
+      "step": 2324
+    },
+    {
+      "epoch": 2.6220338983050846,
+      "grad_norm": 0.04610748961567879,
+      "learning_rate": 7.583663997394241e-06,
+      "loss": 0.9208,
+      "step": 2325
+    },
+    {
+      "epoch": 2.6231638418079095,
+      "grad_norm": 0.0443345345556736,
+      "learning_rate": 7.538357531896856e-06,
+      "loss": 0.9391,
+      "step": 2326
+    },
+    {
+      "epoch": 2.6242937853107344,
+      "grad_norm": 0.04437210038304329,
+      "learning_rate": 7.4931815061457855e-06,
+      "loss": 1.0321,
+      "step": 2327
+    },
+    {
+      "epoch": 2.6254237288135593,
+      "grad_norm": 0.04455529898405075,
+      "learning_rate": 7.44813598387285e-06,
+      "loss": 0.9334,
+      "step": 2328
+    },
+    {
+      "epoch": 2.6265536723163843,
+      "grad_norm": 0.04353608191013336,
+      "learning_rate": 7.403221028625762e-06,
+      "loss": 0.9266,
+      "step": 2329
+    },
+    {
+      "epoch": 2.627683615819209,
+      "grad_norm": 0.04302187263965607,
+      "learning_rate": 7.358436703768035e-06,
+      "loss": 0.9863,
+      "step": 2330
+    },
+    {
+      "epoch": 2.628813559322034,
+      "grad_norm": 0.04428388550877571,
+      "learning_rate": 7.313783072478953e-06,
+      "loss": 1.0419,
+      "step": 2331
+    },
+    {
+      "epoch": 2.629943502824859,
+      "grad_norm": 0.04457690566778183,
+      "learning_rate": 7.269260197753325e-06,
+      "loss": 0.9549,
+      "step": 2332
+    },
+    {
+      "epoch": 2.6310734463276835,
+      "grad_norm": 0.04523680359125137,
+      "learning_rate": 7.224868142401542e-06,
+      "loss": 1.0456,
+      "step": 2333
+    },
+    {
+      "epoch": 2.6322033898305084,
+      "grad_norm": 0.04431130737066269,
+      "learning_rate": 7.180606969049519e-06,
+      "loss": 0.8987,
+      "step": 2334
+    },
+    {
+      "epoch": 2.6333333333333333,
+      "grad_norm": 0.045462481677532196,
+      "learning_rate": 7.136476740138387e-06,
+      "loss": 0.9716,
+      "step": 2335
+    },
+    {
+      "epoch": 2.634463276836158,
+      "grad_norm": 0.044277604669332504,
+      "learning_rate": 7.092477517924634e-06,
+      "loss": 0.9404,
+      "step": 2336
+    },
+    {
+      "epoch": 2.635593220338983,
+      "grad_norm": 0.045289479196071625,
+      "learning_rate": 7.048609364479941e-06,
+      "loss": 0.9956,
+      "step": 2337
+    },
+    {
+      "epoch": 2.636723163841808,
+      "grad_norm": 0.04256889224052429,
+      "learning_rate": 7.00487234169106e-06,
+      "loss": 0.9515,
+      "step": 2338
+    },
+    {
+      "epoch": 2.6378531073446325,
+      "grad_norm": 0.04596678167581558,
+      "learning_rate": 6.961266511259734e-06,
+      "loss": 0.8839,
+      "step": 2339
+    },
+    {
+      "epoch": 2.6389830508474574,
+      "grad_norm": 0.04532042145729065,
+      "learning_rate": 6.917791934702655e-06,
+      "loss": 1.0426,
+      "step": 2340
+    },
+    {
+      "epoch": 2.6401129943502823,
+      "grad_norm": 0.04318855702877045,
+      "learning_rate": 6.87444867335133e-06,
+      "loss": 1.0885,
+      "step": 2341
+    },
+    {
+      "epoch": 2.6412429378531073,
+      "grad_norm": 0.0447513610124588,
+      "learning_rate": 6.831236788352035e-06,
+      "loss": 0.9046,
+      "step": 2342
+    },
+    {
+      "epoch": 2.642372881355932,
+      "grad_norm": 0.04536350071430206,
+      "learning_rate": 6.788156340665697e-06,
+      "loss": 0.9783,
+      "step": 2343
+    },
+    {
+      "epoch": 2.643502824858757,
+      "grad_norm": 0.04435262084007263,
+      "learning_rate": 6.745207391067787e-06,
+      "loss": 0.9362,
+      "step": 2344
+    },
+    {
+      "epoch": 2.644632768361582,
+      "grad_norm": 0.04777863621711731,
+      "learning_rate": 6.702390000148351e-06,
+      "loss": 0.9676,
+      "step": 2345
+    },
+    {
+      "epoch": 2.645762711864407,
+      "grad_norm": 0.04534154385328293,
+      "learning_rate": 6.6597042283117365e-06,
+      "loss": 0.9768,
+      "step": 2346
+    },
+    {
+      "epoch": 2.646892655367232,
+      "grad_norm": 0.046897709369659424,
+      "learning_rate": 6.617150135776662e-06,
+      "loss": 0.9542,
+      "step": 2347
+    },
+    {
+      "epoch": 2.6480225988700568,
+      "grad_norm": 0.04535379633307457,
+      "learning_rate": 6.574727782576129e-06,
+      "loss": 0.9005,
+      "step": 2348
+    },
+    {
+      "epoch": 2.6491525423728812,
+      "grad_norm": 0.04338109865784645,
+      "learning_rate": 6.532437228557153e-06,
+      "loss": 0.9573,
+      "step": 2349
+    },
+    {
+      "epoch": 2.650282485875706,
+      "grad_norm": 0.043659161776304245,
+      "learning_rate": 6.490278533380956e-06,
+      "loss": 1.0269,
+      "step": 2350
+    },
+    {
+      "epoch": 2.651412429378531,
+      "grad_norm": 0.04302853345870972,
+      "learning_rate": 6.4482517565226715e-06,
+      "loss": 0.8879,
+      "step": 2351
+    },
+    {
+      "epoch": 2.652542372881356,
+      "grad_norm": 0.04258056730031967,
+      "learning_rate": 6.406356957271331e-06,
+      "loss": 0.8815,
+      "step": 2352
+    },
+    {
+      "epoch": 2.653672316384181,
+      "grad_norm": 0.04558559134602547,
+      "learning_rate": 6.364594194729789e-06,
+      "loss": 0.9809,
+      "step": 2353
+    },
+    {
+      "epoch": 2.654802259887006,
+      "grad_norm": 0.04610477015376091,
+      "learning_rate": 6.3229635278146295e-06,
+      "loss": 1.0469,
+      "step": 2354
+    },
+    {
+      "epoch": 2.6559322033898303,
+      "grad_norm": 0.044572293758392334,
+      "learning_rate": 6.281465015256094e-06,
+      "loss": 1.0038,
+      "step": 2355
+    },
+    {
+      "epoch": 2.657062146892655,
+      "grad_norm": 0.043487418442964554,
+      "learning_rate": 6.240098715597975e-06,
+      "loss": 0.9886,
+      "step": 2356
+    },
+    {
+      "epoch": 2.65819209039548,
+      "grad_norm": 0.044853370636701584,
+      "learning_rate": 6.198864687197536e-06,
+      "loss": 1.0096,
+      "step": 2357
+    },
+    {
+      "epoch": 2.659322033898305,
+      "grad_norm": 0.044973455369472504,
+      "learning_rate": 6.157762988225457e-06,
+      "loss": 0.9626,
+      "step": 2358
+    },
+    {
+      "epoch": 2.66045197740113,
+      "grad_norm": 0.04432562738656998,
+      "learning_rate": 6.116793676665755e-06,
+      "loss": 1.0168,
+      "step": 2359
+    },
+    {
+      "epoch": 2.661581920903955,
+      "grad_norm": 0.046326398849487305,
+      "learning_rate": 6.0759568103156195e-06,
+      "loss": 0.9713,
+      "step": 2360
+    },
+    {
+      "epoch": 2.661581920903955,
+      "eval_loss": 0.9848312735557556,
+      "eval_runtime": 556.2884,
+      "eval_samples_per_second": 17.586,
+      "eval_steps_per_second": 8.794,
+      "step": 2360
+    },
+    {
+      "epoch": 2.6627118644067798,
+      "grad_norm": 0.04329252615571022,
+      "learning_rate": 6.0352524467854555e-06,
+      "loss": 0.9532,
+      "step": 2361
+    },
+    {
+      "epoch": 2.6638418079096047,
+      "grad_norm": 0.04437597095966339,
+      "learning_rate": 5.994680643498729e-06,
+      "loss": 0.8682,
+      "step": 2362
+    },
+    {
+      "epoch": 2.6649717514124296,
+      "grad_norm": 0.04312526434659958,
+      "learning_rate": 5.954241457691834e-06,
+      "loss": 1.0423,
+      "step": 2363
+    },
+    {
+      "epoch": 2.6661016949152545,
+      "grad_norm": 0.04437880218029022,
+      "learning_rate": 5.913934946414179e-06,
+      "loss": 0.9964,
+      "step": 2364
+    },
+    {
+      "epoch": 2.667231638418079,
+      "grad_norm": 0.04453466460108757,
+      "learning_rate": 5.873761166527936e-06,
+      "loss": 0.9702,
+      "step": 2365
+    },
+    {
+      "epoch": 2.668361581920904,
+      "grad_norm": 0.04452778771519661,
+      "learning_rate": 5.833720174708024e-06,
+      "loss": 0.9637,
+      "step": 2366
+    },
+    {
+      "epoch": 2.669491525423729,
+      "grad_norm": 0.04603665694594383,
+      "learning_rate": 5.793812027442069e-06,
+      "loss": 1.0437,
+      "step": 2367
+    },
+    {
+      "epoch": 2.6706214689265537,
+      "grad_norm": 0.04555994272232056,
+      "learning_rate": 5.754036781030247e-06,
+      "loss": 0.9752,
+      "step": 2368
+    },
+    {
+      "epoch": 2.6717514124293786,
+      "grad_norm": 0.04573677107691765,
+      "learning_rate": 5.714394491585295e-06,
+      "loss": 0.94,
+      "step": 2369
+    },
+    {
+      "epoch": 2.6728813559322036,
+      "grad_norm": 0.04386160522699356,
+      "learning_rate": 5.674885215032322e-06,
+      "loss": 0.9949,
+      "step": 2370
+    },
+    {
+      "epoch": 2.674011299435028,
+      "grad_norm": 0.04481654241681099,
+      "learning_rate": 5.635509007108819e-06,
+      "loss": 1.0066,
+      "step": 2371
+    },
+    {
+      "epoch": 2.675141242937853,
+      "grad_norm": 0.04394556209445,
+      "learning_rate": 5.596265923364585e-06,
+      "loss": 0.9715,
+      "step": 2372
+    },
+    {
+      "epoch": 2.676271186440678,
+      "grad_norm": 0.0456613190472126,
+      "learning_rate": 5.557156019161558e-06,
+      "loss": 0.9375,
+      "step": 2373
+    },
+    {
+      "epoch": 2.6774011299435028,
+      "grad_norm": 0.04373668506741524,
+      "learning_rate": 5.518179349673802e-06,
+      "loss": 0.9866,
+      "step": 2374
+    },
+    {
+      "epoch": 2.6785310734463277,
+      "grad_norm": 0.04538121819496155,
+      "learning_rate": 5.479335969887467e-06,
+      "loss": 0.9699,
+      "step": 2375
+    },
+    {
+      "epoch": 2.6796610169491526,
+      "grad_norm": 0.04715288430452347,
+      "learning_rate": 5.44062593460063e-06,
+      "loss": 0.9896,
+      "step": 2376
+    },
+    {
+      "epoch": 2.6807909604519775,
+      "grad_norm": 0.044993508607149124,
+      "learning_rate": 5.4020492984232175e-06,
+      "loss": 0.9059,
+      "step": 2377
+    },
+    {
+      "epoch": 2.6819209039548024,
+      "grad_norm": 0.04353228956460953,
+      "learning_rate": 5.363606115777054e-06,
+      "loss": 0.9714,
+      "step": 2378
+    },
+    {
+      "epoch": 2.6830508474576273,
+      "grad_norm": 0.04626986011862755,
+      "learning_rate": 5.325296440895622e-06,
+      "loss": 1.0475,
+      "step": 2379
+    },
+    {
+      "epoch": 2.684180790960452,
+      "grad_norm": 0.04400772973895073,
+      "learning_rate": 5.287120327824091e-06,
+      "loss": 0.9536,
+      "step": 2380
+    },
+    {
+      "epoch": 2.6853107344632767,
+      "grad_norm": 0.04464118182659149,
+      "learning_rate": 5.249077830419191e-06,
+      "loss": 0.9327,
+      "step": 2381
+    },
+    {
+      "epoch": 2.6864406779661016,
+      "grad_norm": 0.044131822884082794,
+      "learning_rate": 5.211169002349148e-06,
+      "loss": 0.8624,
+      "step": 2382
+    },
+    {
+      "epoch": 2.6875706214689266,
+      "grad_norm": 0.04314563795924187,
+      "learning_rate": 5.1733938970936705e-06,
+      "loss": 1.057,
+      "step": 2383
+    },
+    {
+      "epoch": 2.6887005649717515,
+      "grad_norm": 0.0449894517660141,
+      "learning_rate": 5.135752567943753e-06,
+      "loss": 1.0021,
+      "step": 2384
+    },
+    {
+      "epoch": 2.6898305084745764,
+      "grad_norm": 0.04238582029938698,
+      "learning_rate": 5.098245068001661e-06,
+      "loss": 0.9477,
+      "step": 2385
+    },
+    {
+      "epoch": 2.690960451977401,
+      "grad_norm": 0.04517286270856857,
+      "learning_rate": 5.060871450180949e-06,
+      "loss": 0.9401,
+      "step": 2386
+    },
+    {
+      "epoch": 2.6920903954802258,
+      "grad_norm": 0.0464588962495327,
+      "learning_rate": 5.023631767206205e-06,
+      "loss": 0.9651,
+      "step": 2387
+    },
+    {
+      "epoch": 2.6932203389830507,
+      "grad_norm": 0.04352294281125069,
+      "learning_rate": 4.986526071613118e-06,
+      "loss": 0.9318,
+      "step": 2388
+    },
+    {
+      "epoch": 2.6943502824858756,
+      "grad_norm": 0.04457921162247658,
+      "learning_rate": 4.949554415748325e-06,
+      "loss": 0.9933,
+      "step": 2389
+    },
+    {
+      "epoch": 2.6954802259887005,
+      "grad_norm": 0.04590234532952309,
+      "learning_rate": 4.9127168517693946e-06,
+      "loss": 0.9513,
+      "step": 2390
+    },
+    {
+      "epoch": 2.6966101694915254,
+      "grad_norm": 0.04296498000621796,
+      "learning_rate": 4.876013431644721e-06,
+      "loss": 0.9084,
+      "step": 2391
+    },
+    {
+      "epoch": 2.6977401129943503,
+      "grad_norm": 0.044914163649082184,
+      "learning_rate": 4.839444207153432e-06,
+      "loss": 0.9736,
+      "step": 2392
+    },
+    {
+      "epoch": 2.6988700564971753,
+      "grad_norm": 0.04357539862394333,
+      "learning_rate": 4.803009229885369e-06,
+      "loss": 1.0393,
+      "step": 2393
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.04416859522461891,
+      "learning_rate": 4.766708551240972e-06,
+      "loss": 1.0616,
+      "step": 2394
+    },
+    {
+      "epoch": 2.701129943502825,
+      "grad_norm": 0.046145226806402206,
+      "learning_rate": 4.730542222431223e-06,
+      "loss": 0.9901,
+      "step": 2395
+    },
+    {
+      "epoch": 2.7022598870056496,
+      "grad_norm": 0.04451952874660492,
+      "learning_rate": 4.6945102944775385e-06,
+      "loss": 0.9736,
+      "step": 2396
+    },
+    {
+      "epoch": 2.7033898305084745,
+      "grad_norm": 0.04450220614671707,
+      "learning_rate": 4.658612818211827e-06,
+      "loss": 0.8922,
+      "step": 2397
+    },
+    {
+      "epoch": 2.7045197740112994,
+      "grad_norm": 0.043102215975522995,
+      "learning_rate": 4.6228498442761785e-06,
+      "loss": 1.0498,
+      "step": 2398
+    },
+    {
+      "epoch": 2.7056497175141243,
+      "grad_norm": 0.043673399835824966,
+      "learning_rate": 4.587221423123056e-06,
+      "loss": 1.0152,
+      "step": 2399
+    },
+    {
+      "epoch": 2.7067796610169492,
+      "grad_norm": 0.04494861513376236,
+      "learning_rate": 4.551727605015032e-06,
+      "loss": 0.9538,
+      "step": 2400
+    },
+    {
+      "epoch": 2.707909604519774,
+      "grad_norm": 0.04306402429938316,
+      "learning_rate": 4.516368440024832e-06,
+      "loss": 0.9531,
+      "step": 2401
+    },
+    {
+      "epoch": 2.7090395480225986,
+      "grad_norm": 0.04380949214100838,
+      "learning_rate": 4.481143978035196e-06,
+      "loss": 0.9118,
+      "step": 2402
+    },
+    {
+      "epoch": 2.7101694915254235,
+      "grad_norm": 0.04429548606276512,
+      "learning_rate": 4.44605426873882e-06,
+      "loss": 0.9699,
+      "step": 2403
+    },
+    {
+      "epoch": 2.7112994350282484,
+      "grad_norm": 0.04391038790345192,
+      "learning_rate": 4.411099361638338e-06,
+      "loss": 0.9136,
+      "step": 2404
+    },
+    {
+      "epoch": 2.7124293785310734,
+      "grad_norm": 0.04415955767035484,
+      "learning_rate": 4.376279306046183e-06,
+      "loss": 0.9088,
+      "step": 2405
+    },
+    {
+      "epoch": 2.7135593220338983,
+      "grad_norm": 0.04572036862373352,
+      "learning_rate": 4.341594151084538e-06,
+      "loss": 0.9603,
+      "step": 2406
+    },
+    {
+      "epoch": 2.714689265536723,
+      "grad_norm": 0.044412463903427124,
+      "learning_rate": 4.307043945685318e-06,
+      "loss": 0.8958,
+      "step": 2407
+    },
+    {
+      "epoch": 2.715819209039548,
+      "grad_norm": 0.044557929039001465,
+      "learning_rate": 4.272628738590012e-06,
+      "loss": 0.9025,
+      "step": 2408
+    },
+    {
+      "epoch": 2.716949152542373,
+      "grad_norm": 0.044220615178346634,
+      "learning_rate": 4.238348578349683e-06,
+      "loss": 0.9579,
+      "step": 2409
+    },
+    {
+      "epoch": 2.718079096045198,
+      "grad_norm": 0.04436076059937477,
+      "learning_rate": 4.2042035133248895e-06,
+      "loss": 0.9139,
+      "step": 2410
+    },
+    {
+      "epoch": 2.719209039548023,
+      "grad_norm": 0.04490151256322861,
+      "learning_rate": 4.170193591685601e-06,
+      "loss": 0.9488,
+      "step": 2411
+    },
+    {
+      "epoch": 2.7203389830508473,
+      "grad_norm": 0.0449664331972599,
+      "learning_rate": 4.136318861411081e-06,
+      "loss": 0.9821,
+      "step": 2412
+    },
+    {
+      "epoch": 2.7214689265536722,
+      "grad_norm": 0.04529043659567833,
+      "learning_rate": 4.10257937028996e-06,
+      "loss": 0.9616,
+      "step": 2413
+    },
+    {
+      "epoch": 2.722598870056497,
+      "grad_norm": 0.04390520602464676,
+      "learning_rate": 4.068975165920008e-06,
+      "loss": 0.9357,
+      "step": 2414
+    },
+    {
+      "epoch": 2.723728813559322,
+      "grad_norm": 0.04567769914865494,
+      "learning_rate": 4.035506295708191e-06,
+      "loss": 0.9597,
+      "step": 2415
+    },
+    {
+      "epoch": 2.724858757062147,
+      "grad_norm": 0.044814545661211014,
+      "learning_rate": 4.002172806870519e-06,
+      "loss": 0.9892,
+      "step": 2416
+    },
+    {
+      "epoch": 2.725988700564972,
+      "grad_norm": 0.044856686145067215,
+      "learning_rate": 3.96897474643203e-06,
+      "loss": 1.0405,
+      "step": 2417
+    },
+    {
+      "epoch": 2.7271186440677964,
+      "grad_norm": 0.04615604504942894,
+      "learning_rate": 3.935912161226696e-06,
+      "loss": 1.0903,
+      "step": 2418
+    },
+    {
+      "epoch": 2.7282485875706213,
+      "grad_norm": 0.043409042060375214,
+      "learning_rate": 3.9029850978973715e-06,
+      "loss": 0.9417,
+      "step": 2419
+    },
+    {
+      "epoch": 2.729378531073446,
+      "grad_norm": 0.04470163583755493,
+      "learning_rate": 3.870193602895733e-06,
+      "loss": 0.945,
+      "step": 2420
+    },
+    {
+      "epoch": 2.730508474576271,
+      "grad_norm": 0.04446692019701004,
+      "learning_rate": 3.8375377224822204e-06,
+      "loss": 1.0123,
+      "step": 2421
+    },
+    {
+      "epoch": 2.731638418079096,
+      "grad_norm": 0.04441095516085625,
+      "learning_rate": 3.805017502725905e-06,
+      "loss": 1.002,
+      "step": 2422
+    },
+    {
+      "epoch": 2.732768361581921,
+      "grad_norm": 0.04440312832593918,
+      "learning_rate": 3.7726329895044986e-06,
+      "loss": 0.9104,
+      "step": 2423
+    },
+    {
+      "epoch": 2.733898305084746,
+      "grad_norm": 0.04501642286777496,
+      "learning_rate": 3.740384228504312e-06,
+      "loss": 0.9695,
+      "step": 2424
+    },
+    {
+      "epoch": 2.7350282485875708,
+      "grad_norm": 0.04526914656162262,
+      "learning_rate": 3.7082712652200867e-06,
+      "loss": 0.9805,
+      "step": 2425
+    },
+    {
+      "epoch": 2.7361581920903957,
+      "grad_norm": 0.04381269961595535,
+      "learning_rate": 3.6762941449549727e-06,
+      "loss": 0.9831,
+      "step": 2426
+    },
+    {
+      "epoch": 2.7372881355932206,
+      "grad_norm": 0.04696391150355339,
+      "learning_rate": 3.6444529128205618e-06,
+      "loss": 0.9726,
+      "step": 2427
+    },
+    {
+      "epoch": 2.738418079096045,
+      "grad_norm": 0.04461664706468582,
+      "learning_rate": 3.6127476137366777e-06,
+      "loss": 1.0091,
+      "step": 2428
+    },
+    {
+      "epoch": 2.73954802259887,
+      "grad_norm": 0.04383271187543869,
+      "learning_rate": 3.5811782924313864e-06,
+      "loss": 1.0026,
+      "step": 2429
+    },
+    {
+      "epoch": 2.740677966101695,
+      "grad_norm": 0.04371367394924164,
+      "learning_rate": 3.54974499344094e-06,
+      "loss": 0.9416,
+      "step": 2430
+    },
+    {
+      "epoch": 2.74180790960452,
+      "grad_norm": 0.044015347957611084,
+      "learning_rate": 3.518447761109689e-06,
+      "loss": 0.9705,
+      "step": 2431
+    },
+    {
+      "epoch": 2.7429378531073447,
+      "grad_norm": 0.04504389315843582,
+      "learning_rate": 3.487286639590026e-06,
+      "loss": 0.9769,
+      "step": 2432
+    },
+    {
+      "epoch": 2.7440677966101696,
+      "grad_norm": 0.04418431222438812,
+      "learning_rate": 3.4562616728423426e-06,
+      "loss": 0.8872,
+      "step": 2433
+    },
+    {
+      "epoch": 2.745197740112994,
+      "grad_norm": 0.042971495538949966,
+      "learning_rate": 3.425372904634905e-06,
+      "loss": 0.9208,
+      "step": 2434
+    },
+    {
+      "epoch": 2.746327683615819,
+      "grad_norm": 0.04686740040779114,
+      "learning_rate": 3.3946203785439113e-06,
+      "loss": 1.0176,
+      "step": 2435
+    },
+    {
+      "epoch": 2.747457627118644,
+      "grad_norm": 0.04372727870941162,
+      "learning_rate": 3.3640041379532805e-06,
+      "loss": 0.8971,
+      "step": 2436
+    },
+    {
+      "epoch": 2.748587570621469,
+      "grad_norm": 0.04424417391419411,
+      "learning_rate": 3.3335242260547293e-06,
+      "loss": 1.0166,
+      "step": 2437
+    },
+    {
+      "epoch": 2.7497175141242938,
+      "grad_norm": 0.04451939836144447,
+      "learning_rate": 3.3031806858476065e-06,
+      "loss": 1.0402,
+      "step": 2438
+    },
+    {
+      "epoch": 2.7508474576271187,
+      "grad_norm": 0.04643287882208824,
+      "learning_rate": 3.2729735601389143e-06,
+      "loss": 0.8732,
+      "step": 2439
+    },
+    {
+      "epoch": 2.7519774011299436,
+      "grad_norm": 0.043419696390628815,
+      "learning_rate": 3.2429028915431534e-06,
+      "loss": 1.0057,
+      "step": 2440
+    },
+    {
+      "epoch": 2.7531073446327685,
+      "grad_norm": 0.04414771869778633,
+      "learning_rate": 3.21296872248239e-06,
+      "loss": 1.0273,
+      "step": 2441
+    },
+    {
+      "epoch": 2.7542372881355934,
+      "grad_norm": 0.04677413031458855,
+      "learning_rate": 3.1831710951860547e-06,
+      "loss": 0.8554,
+      "step": 2442
+    },
+    {
+      "epoch": 2.7553672316384183,
+      "grad_norm": 0.044694963842630386,
+      "learning_rate": 3.153510051690989e-06,
+      "loss": 1.0362,
+      "step": 2443
+    },
+    {
+      "epoch": 2.756497175141243,
+      "grad_norm": 0.043344851583242416,
+      "learning_rate": 3.123985633841364e-06,
+      "loss": 0.9214,
+      "step": 2444
+    },
+    {
+      "epoch": 2.7576271186440677,
+      "grad_norm": 0.04490336403250694,
+      "learning_rate": 3.094597883288575e-06,
+      "loss": 0.9133,
+      "step": 2445
+    },
+    {
+      "epoch": 2.7587570621468926,
+      "grad_norm": 0.045138705521821976,
+      "learning_rate": 3.0653468414912123e-06,
+      "loss": 0.9534,
+      "step": 2446
+    },
+    {
+      "epoch": 2.7598870056497176,
+      "grad_norm": 0.04454415664076805,
+      "learning_rate": 3.0362325497150348e-06,
+      "loss": 1.0111,
+      "step": 2447
+    },
+    {
+      "epoch": 2.7610169491525425,
+      "grad_norm": 0.04441482201218605,
+      "learning_rate": 3.0072550490328753e-06,
+      "loss": 0.9419,
+      "step": 2448
+    },
+    {
+      "epoch": 2.7621468926553674,
+      "grad_norm": 0.0444534532725811,
+      "learning_rate": 2.978414380324579e-06,
+      "loss": 0.8905,
+      "step": 2449
+    },
+    {
+      "epoch": 2.763276836158192,
+      "grad_norm": 0.045210570096969604,
+      "learning_rate": 2.9497105842769435e-06,
+      "loss": 0.9332,
+      "step": 2450
+    },
+    {
+      "epoch": 2.7644067796610168,
+      "grad_norm": 0.04406363144516945,
+      "learning_rate": 2.9211437013836995e-06,
+      "loss": 0.8547,
+      "step": 2451
+    },
+    {
+      "epoch": 2.7655367231638417,
+      "grad_norm": 0.043664876371622086,
+      "learning_rate": 2.8927137719454103e-06,
+      "loss": 0.8963,
+      "step": 2452
+    },
+    {
+      "epoch": 2.7666666666666666,
+      "grad_norm": 0.04488592594861984,
+      "learning_rate": 2.864420836069459e-06,
+      "loss": 0.9805,
+      "step": 2453
+    },
+    {
+      "epoch": 2.7677966101694915,
+      "grad_norm": 0.04462852701544762,
+      "learning_rate": 2.836264933669919e-06,
+      "loss": 1.0324,
+      "step": 2454
+    },
+    {
+      "epoch": 2.7689265536723164,
+      "grad_norm": 0.04454018175601959,
+      "learning_rate": 2.808246104467582e-06,
+      "loss": 0.9059,
+      "step": 2455
+    },
+    {
+      "epoch": 2.7700564971751414,
+      "grad_norm": 0.04549877718091011,
+      "learning_rate": 2.7803643879898865e-06,
+      "loss": 0.8686,
+      "step": 2456
+    },
+    {
+      "epoch": 2.7711864406779663,
+      "grad_norm": 0.04532479867339134,
+      "learning_rate": 2.7526198235707678e-06,
+      "loss": 1.0447,
+      "step": 2457
+    },
+    {
+      "epoch": 2.772316384180791,
+      "grad_norm": 0.04685727506875992,
+      "learning_rate": 2.7250124503507168e-06,
+      "loss": 0.9719,
+      "step": 2458
+    },
+    {
+      "epoch": 2.7734463276836157,
+      "grad_norm": 0.04455072060227394,
+      "learning_rate": 2.697542307276724e-06,
+      "loss": 1.0226,
+      "step": 2459
+    },
+    {
+      "epoch": 2.7745762711864406,
+      "grad_norm": 0.04388142749667168,
+      "learning_rate": 2.6702094331020887e-06,
+      "loss": 1.0189,
+      "step": 2460
+    },
+    {
+      "epoch": 2.7757062146892655,
+      "grad_norm": 0.04537327587604523,
+      "learning_rate": 2.6430138663865346e-06,
+      "loss": 0.9486,
+      "step": 2461
+    },
+    {
+      "epoch": 2.7768361581920904,
+      "grad_norm": 0.04427479952573776,
+      "learning_rate": 2.6159556454960487e-06,
+      "loss": 0.9353,
+      "step": 2462
+    },
+    {
+      "epoch": 2.7779661016949153,
+      "grad_norm": 0.046056341379880905,
+      "learning_rate": 2.589034808602897e-06,
+      "loss": 0.9175,
+      "step": 2463
+    },
+    {
+      "epoch": 2.7790960451977402,
+      "grad_norm": 0.043814852833747864,
+      "learning_rate": 2.5622513936854442e-06,
+      "loss": 0.9208,
+      "step": 2464
+    },
+    {
+      "epoch": 2.7802259887005647,
+      "grad_norm": 0.04527292400598526,
+      "learning_rate": 2.5356054385282766e-06,
+      "loss": 0.975,
+      "step": 2465
+    },
+    {
+      "epoch": 2.7813559322033896,
+      "grad_norm": 0.04537815973162651,
+      "learning_rate": 2.5090969807220366e-06,
+      "loss": 0.8724,
+      "step": 2466
+    },
+    {
+      "epoch": 2.7824858757062145,
+      "grad_norm": 0.04719525948166847,
+      "learning_rate": 2.482726057663365e-06,
+      "loss": 0.8909,
+      "step": 2467
+    },
+    {
+      "epoch": 2.7836158192090394,
+      "grad_norm": 0.04599473252892494,
+      "learning_rate": 2.4564927065548914e-06,
+      "loss": 0.9732,
+      "step": 2468
+    },
+    {
+      "epoch": 2.7847457627118644,
+      "grad_norm": 0.04588894918560982,
+      "learning_rate": 2.430396964405168e-06,
+      "loss": 1.0697,
+      "step": 2469
+    },
+    {
+      "epoch": 2.7858757062146893,
+      "grad_norm": 0.045891888439655304,
+      "learning_rate": 2.404438868028658e-06,
+      "loss": 0.9133,
+      "step": 2470
+    },
+    {
+      "epoch": 2.787005649717514,
+      "grad_norm": 0.04246148839592934,
+      "learning_rate": 2.3786184540455448e-06,
+      "loss": 0.9665,
+      "step": 2471
+    },
+    {
+      "epoch": 2.788135593220339,
+      "grad_norm": 0.042736880481243134,
+      "learning_rate": 2.3529357588818577e-06,
+      "loss": 0.8699,
+      "step": 2472
+    },
+    {
+      "epoch": 2.789265536723164,
+      "grad_norm": 0.04353400319814682,
+      "learning_rate": 2.327390818769337e-06,
+      "loss": 0.984,
+      "step": 2473
+    },
+    {
+      "epoch": 2.790395480225989,
+      "grad_norm": 0.04404514282941818,
+      "learning_rate": 2.301983669745322e-06,
+      "loss": 0.8818,
+      "step": 2474
+    },
+    {
+      "epoch": 2.7915254237288134,
+      "grad_norm": 0.044174037873744965,
+      "learning_rate": 2.276714347652831e-06,
+      "loss": 1.0032,
+      "step": 2475
+    },
+    {
+      "epoch": 2.7926553672316383,
+      "grad_norm": 0.044073160737752914,
+      "learning_rate": 2.2515828881404134e-06,
+      "loss": 0.9107,
+      "step": 2476
+    },
+    {
+      "epoch": 2.7937853107344632,
+      "grad_norm": 0.0445500873029232,
+      "learning_rate": 2.226589326662143e-06,
+      "loss": 0.9533,
+      "step": 2477
+    },
+    {
+      "epoch": 2.794915254237288,
+      "grad_norm": 0.044462572783231735,
+      "learning_rate": 2.2017336984775484e-06,
+      "loss": 0.9722,
+      "step": 2478
+    },
+    {
+      "epoch": 2.796045197740113,
+      "grad_norm": 0.04642799124121666,
+      "learning_rate": 2.177016038651558e-06,
+      "loss": 1.0028,
+      "step": 2479
+    },
+    {
+      "epoch": 2.797175141242938,
+      "grad_norm": 0.04265080764889717,
+      "learning_rate": 2.152436382054479e-06,
+      "loss": 0.9624,
+      "step": 2480
+    },
+    {
+      "epoch": 2.7983050847457624,
+      "grad_norm": 0.04329387843608856,
+      "learning_rate": 2.127994763361918e-06,
+      "loss": 0.934,
+      "step": 2481
+    },
+    {
+      "epoch": 2.7994350282485874,
+      "grad_norm": 0.04579446464776993,
+      "learning_rate": 2.1036912170547595e-06,
+      "loss": 0.8815,
+      "step": 2482
+    },
+    {
+      "epoch": 2.8005649717514123,
+      "grad_norm": 0.04465530812740326,
+      "learning_rate": 2.079525777419078e-06,
+      "loss": 0.9744,
+      "step": 2483
+    },
+    {
+      "epoch": 2.801694915254237,
+      "grad_norm": 0.04475056752562523,
+      "learning_rate": 2.0554984785461806e-06,
+      "loss": 0.9676,
+      "step": 2484
+    },
+    {
+      "epoch": 2.802824858757062,
+      "grad_norm": 0.045168641954660416,
+      "learning_rate": 2.0316093543323757e-06,
+      "loss": 0.9999,
+      "step": 2485
+    },
+    {
+      "epoch": 2.803954802259887,
+      "grad_norm": 0.04566594585776329,
+      "learning_rate": 2.007858438479171e-06,
+      "loss": 1.0069,
+      "step": 2486
+    },
+    {
+      "epoch": 2.805084745762712,
+      "grad_norm": 0.04546283185482025,
+      "learning_rate": 1.98424576449302e-06,
+      "loss": 0.9726,
+      "step": 2487
+    },
+    {
+      "epoch": 2.806214689265537,
+      "grad_norm": 0.043535858392715454,
+      "learning_rate": 1.9607713656853544e-06,
+      "loss": 0.9872,
+      "step": 2488
+    },
+    {
+      "epoch": 2.8073446327683618,
+      "grad_norm": 0.04416786879301071,
+      "learning_rate": 1.937435275172572e-06,
+      "loss": 0.9428,
+      "step": 2489
+    },
+    {
+      "epoch": 2.8084745762711867,
+      "grad_norm": 0.04551788419485092,
+      "learning_rate": 1.914237525875917e-06,
+      "loss": 0.9504,
+      "step": 2490
+    },
+    {
+      "epoch": 2.809604519774011,
+      "grad_norm": 0.046153049916028976,
+      "learning_rate": 1.8911781505215e-06,
+      "loss": 0.9752,
+      "step": 2491
+    },
+    {
+      "epoch": 2.810734463276836,
+      "grad_norm": 0.04438695311546326,
+      "learning_rate": 1.8682571816401983e-06,
+      "loss": 0.9908,
+      "step": 2492
+    },
+    {
+      "epoch": 2.811864406779661,
+      "grad_norm": 0.04513789713382721,
+      "learning_rate": 1.8454746515676357e-06,
+      "loss": 0.9173,
+      "step": 2493
+    },
+    {
+      "epoch": 2.812994350282486,
+      "grad_norm": 0.04486611858010292,
+      "learning_rate": 1.822830592444147e-06,
+      "loss": 0.986,
+      "step": 2494
+    },
+    {
+      "epoch": 2.814124293785311,
+      "grad_norm": 0.04494575038552284,
+      "learning_rate": 1.8003250362147005e-06,
+      "loss": 0.9892,
+      "step": 2495
+    },
+    {
+      "epoch": 2.8152542372881357,
+      "grad_norm": 0.04305053502321243,
+      "learning_rate": 1.7779580146288999e-06,
+      "loss": 0.8947,
+      "step": 2496
+    },
+    {
+      "epoch": 2.81638418079096,
+      "grad_norm": 0.041880443692207336,
+      "learning_rate": 1.755729559240893e-06,
+      "loss": 0.919,
+      "step": 2497
+    },
+    {
+      "epoch": 2.817514124293785,
+      "grad_norm": 0.0441063791513443,
+      "learning_rate": 1.7336397014093508e-06,
+      "loss": 0.9521,
+      "step": 2498
+    },
+    {
+      "epoch": 2.81864406779661,
+      "grad_norm": 0.043986279517412186,
+      "learning_rate": 1.7116884722974013e-06,
+      "loss": 1.0188,
+      "step": 2499
+    },
+    {
+      "epoch": 2.819774011299435,
+      "grad_norm": 0.04386138543486595,
+      "learning_rate": 1.6898759028726285e-06,
+      "loss": 0.9757,
+      "step": 2500
+    },
+    {
+      "epoch": 2.82090395480226,
+      "grad_norm": 0.04467868059873581,
+      "learning_rate": 1.6682020239070173e-06,
+      "loss": 1.0435,
+      "step": 2501
+    },
+    {
+      "epoch": 2.8220338983050848,
+      "grad_norm": 0.04562905803322792,
+      "learning_rate": 1.6466668659768202e-06,
+      "loss": 1.0047,
+      "step": 2502
+    },
+    {
+      "epoch": 2.8231638418079097,
+      "grad_norm": 0.04512863978743553,
+      "learning_rate": 1.62527045946268e-06,
+      "loss": 1.0126,
+      "step": 2503
+    },
+    {
+      "epoch": 2.8242937853107346,
+      "grad_norm": 0.045594554394483566,
+      "learning_rate": 1.6040128345494398e-06,
+      "loss": 0.9089,
+      "step": 2504
+    },
+    {
+      "epoch": 2.8254237288135595,
+      "grad_norm": 0.044841185212135315,
+      "learning_rate": 1.5828940212261889e-06,
+      "loss": 0.9067,
+      "step": 2505
+    },
+    {
+      "epoch": 2.8265536723163844,
+      "grad_norm": 0.043986570090055466,
+      "learning_rate": 1.561914049286173e-06,
+      "loss": 0.9603,
+      "step": 2506
+    },
+    {
+      "epoch": 2.827683615819209,
+      "grad_norm": 0.044366851449012756,
+      "learning_rate": 1.5410729483267606e-06,
+      "loss": 1.0132,
+      "step": 2507
+    },
+    {
+      "epoch": 2.828813559322034,
+      "grad_norm": 0.04770446941256523,
+      "learning_rate": 1.5203707477494333e-06,
+      "loss": 0.9823,
+      "step": 2508
+    },
+    {
+      "epoch": 2.8299435028248587,
+      "grad_norm": 0.04479978233575821,
+      "learning_rate": 1.4998074767596848e-06,
+      "loss": 0.9858,
+      "step": 2509
+    },
+    {
+      "epoch": 2.8310734463276837,
+      "grad_norm": 0.044626832008361816,
+      "learning_rate": 1.479383164367043e-06,
+      "loss": 0.9863,
+      "step": 2510
+    },
+    {
+      "epoch": 2.8322033898305086,
+      "grad_norm": 0.04503661394119263,
+      "learning_rate": 1.459097839385004e-06,
+      "loss": 0.9606,
+      "step": 2511
+    },
+    {
+      "epoch": 2.8333333333333335,
+      "grad_norm": 0.043242525309324265,
+      "learning_rate": 1.4389515304309763e-06,
+      "loss": 0.9008,
+      "step": 2512
+    },
+    {
+      "epoch": 2.834463276836158,
+      "grad_norm": 0.04532679542899132,
+      "learning_rate": 1.4189442659262365e-06,
+      "loss": 0.9146,
+      "step": 2513
+    },
+    {
+      "epoch": 2.835593220338983,
+      "grad_norm": 0.04340112954378128,
+      "learning_rate": 1.3990760740959285e-06,
+      "loss": 1.0163,
+      "step": 2514
+    },
+    {
+      "epoch": 2.836723163841808,
+      "grad_norm": 0.045547567307949066,
+      "learning_rate": 1.3793469829689987e-06,
+      "loss": 0.9454,
+      "step": 2515
+    },
+    {
+      "epoch": 2.8378531073446327,
+      "grad_norm": 0.04489200562238693,
+      "learning_rate": 1.3597570203781496e-06,
+      "loss": 0.9677,
+      "step": 2516
+    },
+    {
+      "epoch": 2.8389830508474576,
+      "grad_norm": 0.0449584536254406,
+      "learning_rate": 1.3403062139598076e-06,
+      "loss": 0.8906,
+      "step": 2517
+    },
+    {
+      "epoch": 2.8401129943502825,
+      "grad_norm": 0.04351430386304855,
+      "learning_rate": 1.32099459115409e-06,
+      "loss": 0.9033,
+      "step": 2518
+    },
+    {
+      "epoch": 2.8412429378531074,
+      "grad_norm": 0.04647156223654747,
+      "learning_rate": 1.3018221792047703e-06,
+      "loss": 1.0343,
+      "step": 2519
+    },
+    {
+      "epoch": 2.8423728813559324,
+      "grad_norm": 0.04549736902117729,
+      "learning_rate": 1.2827890051592128e-06,
+      "loss": 0.9813,
+      "step": 2520
+    },
+    {
+      "epoch": 2.8435028248587573,
+      "grad_norm": 0.04418494924902916,
+      "learning_rate": 1.2638950958683504e-06,
+      "loss": 0.9703,
+      "step": 2521
+    },
+    {
+      "epoch": 2.844632768361582,
+      "grad_norm": 0.04414811357855797,
+      "learning_rate": 1.2451404779866948e-06,
+      "loss": 1.0038,
+      "step": 2522
+    },
+    {
+      "epoch": 2.8457627118644067,
+      "grad_norm": 0.04323131591081619,
+      "learning_rate": 1.226525177972182e-06,
+      "loss": 1.0242,
+      "step": 2523
+    },
+    {
+      "epoch": 2.8468926553672316,
+      "grad_norm": 0.04595525190234184,
+      "learning_rate": 1.2080492220862605e-06,
+      "loss": 0.9779,
+      "step": 2524
+    },
+    {
+      "epoch": 2.8480225988700565,
+      "grad_norm": 0.04375181347131729,
+      "learning_rate": 1.1897126363937804e-06,
+      "loss": 0.9557,
+      "step": 2525
+    },
+    {
+      "epoch": 2.8491525423728814,
+      "grad_norm": 0.04520711675286293,
+      "learning_rate": 1.171515446762972e-06,
+      "loss": 0.9467,
+      "step": 2526
+    },
+    {
+      "epoch": 2.8502824858757063,
+      "grad_norm": 0.04446513205766678,
+      "learning_rate": 1.1534576788654327e-06,
+      "loss": 0.8768,
+      "step": 2527
+    },
+    {
+      "epoch": 2.8514124293785312,
+      "grad_norm": 0.044064104557037354,
+      "learning_rate": 1.135539358176041e-06,
+      "loss": 0.9144,
+      "step": 2528
+    },
+    {
+      "epoch": 2.8525423728813557,
+      "grad_norm": 0.04525630176067352,
+      "learning_rate": 1.117760509972987e-06,
+      "loss": 0.9452,
+      "step": 2529
+    },
+    {
+      "epoch": 2.8536723163841806,
+      "grad_norm": 0.042912956327199936,
+      "learning_rate": 1.1001211593376526e-06,
+      "loss": 1.0189,
+      "step": 2530
+    },
+    {
+      "epoch": 2.8548022598870055,
+      "grad_norm": 0.04408475011587143,
+      "learning_rate": 1.0826213311546873e-06,
+      "loss": 0.9009,
+      "step": 2531
+    },
+    {
+      "epoch": 2.8559322033898304,
+      "grad_norm": 0.04350871592760086,
+      "learning_rate": 1.0652610501118542e-06,
+      "loss": 1.0334,
+      "step": 2532
+    },
+    {
+      "epoch": 2.8570621468926554,
+      "grad_norm": 0.04493604972958565,
+      "learning_rate": 1.0480403407000738e-06,
+      "loss": 0.945,
+      "step": 2533
+    },
+    {
+      "epoch": 2.8581920903954803,
+      "grad_norm": 0.04436785727739334,
+      "learning_rate": 1.0309592272133684e-06,
+      "loss": 1.0529,
+      "step": 2534
+    },
+    {
+      "epoch": 2.859322033898305,
+      "grad_norm": 0.04460444301366806,
+      "learning_rate": 1.0140177337488288e-06,
+      "loss": 0.9559,
+      "step": 2535
+    },
+    {
+      "epoch": 2.86045197740113,
+      "grad_norm": 0.04335474595427513,
+      "learning_rate": 9.972158842065816e-07,
+      "loss": 0.9376,
+      "step": 2536
+    },
+    {
+      "epoch": 2.861581920903955,
+      "grad_norm": 0.04430084675550461,
+      "learning_rate": 9.805537022897104e-07,
+      "loss": 0.9212,
+      "step": 2537
+    },
+    {
+      "epoch": 2.86271186440678,
+      "grad_norm": 0.0432843454182148,
+      "learning_rate": 9.640312115043237e-07,
+      "loss": 0.9804,
+      "step": 2538
+    },
+    {
+      "epoch": 2.8638418079096044,
+      "grad_norm": 0.04469630494713783,
+      "learning_rate": 9.476484351594317e-07,
+      "loss": 0.9615,
+      "step": 2539
+    },
+    {
+      "epoch": 2.8649717514124293,
+      "grad_norm": 0.046468090265989304,
+      "learning_rate": 9.314053963669245e-07,
+      "loss": 0.9386,
+      "step": 2540
+    },
+    {
+      "epoch": 2.8661016949152542,
+      "grad_norm": 0.044023171067237854,
+      "learning_rate": 9.153021180415944e-07,
+      "loss": 0.9795,
+      "step": 2541
+    },
+    {
+      "epoch": 2.867231638418079,
+      "grad_norm": 0.045041874051094055,
+      "learning_rate": 8.993386229010581e-07,
+      "loss": 0.9741,
+      "step": 2542
+    },
+    {
+      "epoch": 2.868361581920904,
+      "grad_norm": 0.04462430626153946,
+      "learning_rate": 8.835149334657122e-07,
+      "loss": 0.9663,
+      "step": 2543
+    },
+    {
+      "epoch": 2.8694915254237285,
+      "grad_norm": 0.04438344016671181,
+      "learning_rate": 8.678310720587335e-07,
+      "loss": 0.9934,
+      "step": 2544
+    },
+    {
+      "epoch": 2.8706214689265535,
+      "grad_norm": 0.0462101474404335,
+      "learning_rate": 8.522870608060563e-07,
+      "loss": 1.1079,
+      "step": 2545
+    },
+    {
+      "epoch": 2.8717514124293784,
+      "grad_norm": 0.04502633213996887,
+      "learning_rate": 8.368829216363172e-07,
+      "loss": 0.9364,
+      "step": 2546
+    },
+    {
+      "epoch": 2.8728813559322033,
+      "grad_norm": 0.04489148035645485,
+      "learning_rate": 8.216186762807998e-07,
+      "loss": 1.0012,
+      "step": 2547
+    },
+    {
+      "epoch": 2.874011299435028,
+      "grad_norm": 0.04346054419875145,
+      "learning_rate": 8.064943462734454e-07,
+      "loss": 0.9731,
+      "step": 2548
+    },
+    {
+      "epoch": 2.875141242937853,
+      "grad_norm": 0.04357984662055969,
+      "learning_rate": 7.91509952950853e-07,
+      "loss": 1.0232,
+      "step": 2549
+    },
+    {
+      "epoch": 2.876271186440678,
+      "grad_norm": 0.04210647568106651,
+      "learning_rate": 7.766655174521465e-07,
+      "loss": 1.039,
+      "step": 2550
+    },
+    {
+      "epoch": 2.877401129943503,
+      "grad_norm": 0.044638440012931824,
+      "learning_rate": 7.619610607190186e-07,
+      "loss": 0.8978,
+      "step": 2551
+    },
+    {
+      "epoch": 2.878531073446328,
+      "grad_norm": 0.04344021528959274,
+      "learning_rate": 7.473966034957313e-07,
+      "loss": 0.9152,
+      "step": 2552
+    },
+    {
+      "epoch": 2.8796610169491528,
+      "grad_norm": 0.04684427008032799,
+      "learning_rate": 7.329721663289935e-07,
+      "loss": 0.9515,
+      "step": 2553
+    },
+    {
+      "epoch": 2.8807909604519772,
+      "grad_norm": 0.04457107558846474,
+      "learning_rate": 7.186877695679939e-07,
+      "loss": 0.9952,
+      "step": 2554
+    },
+    {
+      "epoch": 2.881920903954802,
+      "grad_norm": 0.043536800891160965,
+      "learning_rate": 7.045434333643797e-07,
+      "loss": 0.9389,
+      "step": 2555
+    },
+    {
+      "epoch": 2.883050847457627,
+      "grad_norm": 0.04445602744817734,
+      "learning_rate": 6.905391776721893e-07,
+      "loss": 0.9033,
+      "step": 2556
+    },
+    {
+      "epoch": 2.884180790960452,
+      "grad_norm": 0.042980775237083435,
+      "learning_rate": 6.766750222478524e-07,
+      "loss": 0.849,
+      "step": 2557
+    },
+    {
+      "epoch": 2.885310734463277,
+      "grad_norm": 0.045031286776065826,
+      "learning_rate": 6.629509866501349e-07,
+      "loss": 0.9813,
+      "step": 2558
+    },
+    {
+      "epoch": 2.886440677966102,
+      "grad_norm": 0.0436282679438591,
+      "learning_rate": 6.493670902401494e-07,
+      "loss": 0.9479,
+      "step": 2559
+    },
+    {
+      "epoch": 2.8875706214689263,
+      "grad_norm": 0.045009516179561615,
+      "learning_rate": 6.359233521813224e-07,
+      "loss": 0.8999,
+      "step": 2560
+    },
+    {
+      "epoch": 2.888700564971751,
+      "grad_norm": 0.04531530290842056,
+      "learning_rate": 6.22619791439305e-07,
+      "loss": 1.0032,
+      "step": 2561
+    },
+    {
+      "epoch": 2.889830508474576,
+      "grad_norm": 0.04437597468495369,
+      "learning_rate": 6.094564267820291e-07,
+      "loss": 1.0318,
+      "step": 2562
+    },
+    {
+      "epoch": 2.890960451977401,
+      "grad_norm": 0.04388117045164108,
+      "learning_rate": 5.964332767796399e-07,
+      "loss": 0.9455,
+      "step": 2563
+    },
+    {
+      "epoch": 2.892090395480226,
+      "grad_norm": 0.04577458277344704,
+      "learning_rate": 5.835503598044745e-07,
+      "loss": 1.0662,
+      "step": 2564
+    },
+    {
+      "epoch": 2.893220338983051,
+      "grad_norm": 0.04454333335161209,
+      "learning_rate": 5.70807694031028e-07,
+      "loss": 1.0782,
+      "step": 2565
+    },
+    {
+      "epoch": 2.894350282485876,
+      "grad_norm": 0.045315299183130264,
+      "learning_rate": 5.582052974359653e-07,
+      "loss": 0.9799,
+      "step": 2566
+    },
+    {
+      "epoch": 2.8954802259887007,
+      "grad_norm": 0.044649627059698105,
+      "learning_rate": 5.457431877980312e-07,
+      "loss": 1.0305,
+      "step": 2567
+    },
+    {
+      "epoch": 2.8966101694915256,
+      "grad_norm": 0.044768646359443665,
+      "learning_rate": 5.334213826980738e-07,
+      "loss": 1.0419,
+      "step": 2568
+    },
+    {
+      "epoch": 2.8977401129943505,
+      "grad_norm": 0.04542625695466995,
+      "learning_rate": 5.212398995190215e-07,
+      "loss": 0.9791,
+      "step": 2569
+    },
+    {
+      "epoch": 2.898870056497175,
+      "grad_norm": 0.04401678591966629,
+      "learning_rate": 5.091987554458388e-07,
+      "loss": 0.9531,
+      "step": 2570
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.04501966014504433,
+      "learning_rate": 4.972979674654821e-07,
+      "loss": 0.9026,
+      "step": 2571
+    },
+    {
+      "epoch": 2.901129943502825,
+      "grad_norm": 0.04327463358640671,
+      "learning_rate": 4.855375523669326e-07,
+      "loss": 0.957,
+      "step": 2572
+    },
+    {
+      "epoch": 2.9022598870056497,
+      "grad_norm": 0.043885570019483566,
+      "learning_rate": 4.7391752674113e-07,
+      "loss": 0.9422,
+      "step": 2573
+    },
+    {
+      "epoch": 2.9033898305084747,
+      "grad_norm": 0.043965715914964676,
+      "learning_rate": 4.6243790698097255e-07,
+      "loss": 0.9477,
+      "step": 2574
+    },
+    {
+      "epoch": 2.9045197740112996,
+      "grad_norm": 0.04610178992152214,
+      "learning_rate": 4.510987092812502e-07,
+      "loss": 1.0274,
+      "step": 2575
+    },
+    {
+      "epoch": 2.905649717514124,
+      "grad_norm": 0.04622510448098183,
+      "learning_rate": 4.398999496386891e-07,
+      "loss": 0.9179,
+      "step": 2576
+    },
+    {
+      "epoch": 2.906779661016949,
+      "grad_norm": 0.04459752142429352,
+      "learning_rate": 4.288416438518628e-07,
+      "loss": 1.0092,
+      "step": 2577
+    },
+    {
+      "epoch": 2.907909604519774,
+      "grad_norm": 0.043733566999435425,
+      "learning_rate": 4.179238075212144e-07,
+      "loss": 0.9825,
+      "step": 2578
+    },
+    {
+      "epoch": 2.909039548022599,
+      "grad_norm": 0.04439612850546837,
+      "learning_rate": 4.071464560490346e-07,
+      "loss": 1.0026,
+      "step": 2579
+    },
+    {
+      "epoch": 2.9101694915254237,
+      "grad_norm": 0.044625185430049896,
+      "learning_rate": 3.965096046394057e-07,
+      "loss": 0.9278,
+      "step": 2580
+    },
+    {
+      "epoch": 2.9112994350282486,
+      "grad_norm": 0.04455350339412689,
+      "learning_rate": 3.860132682982021e-07,
+      "loss": 1.0179,
+      "step": 2581
+    },
+    {
+      "epoch": 2.9124293785310735,
+      "grad_norm": 0.043388571590185165,
+      "learning_rate": 3.7565746183307884e-07,
+      "loss": 0.9418,
+      "step": 2582
+    },
+    {
+      "epoch": 2.9135593220338984,
+      "grad_norm": 0.04536639526486397,
+      "learning_rate": 3.654421998534163e-07,
+      "loss": 1.0083,
+      "step": 2583
+    },
+    {
+      "epoch": 2.9146892655367234,
+      "grad_norm": 0.04366553574800491,
+      "learning_rate": 3.553674967703646e-07,
+      "loss": 1.0002,
+      "step": 2584
+    },
+    {
+      "epoch": 2.9158192090395483,
+      "grad_norm": 0.04244455695152283,
+      "learning_rate": 3.4543336679673245e-07,
+      "loss": 1.0485,
+      "step": 2585
+    },
+    {
+      "epoch": 2.9169491525423727,
+      "grad_norm": 0.04259354993700981,
+      "learning_rate": 3.3563982394704266e-07,
+      "loss": 0.9429,
+      "step": 2586
+    },
+    {
+      "epoch": 2.9180790960451977,
+      "grad_norm": 0.04567457735538483,
+      "learning_rate": 3.259868820374878e-07,
+      "loss": 1.0307,
+      "step": 2587
+    },
+    {
+      "epoch": 2.9192090395480226,
+      "grad_norm": 0.044698674231767654,
+      "learning_rate": 3.1647455468590825e-07,
+      "loss": 0.9137,
+      "step": 2588
+    },
+    {
+      "epoch": 2.9203389830508475,
+      "grad_norm": 0.04482744634151459,
+      "learning_rate": 3.071028553117472e-07,
+      "loss": 0.9424,
+      "step": 2589
+    },
+    {
+      "epoch": 2.9214689265536724,
+      "grad_norm": 0.043734245002269745,
+      "learning_rate": 2.978717971360956e-07,
+      "loss": 0.8577,
+      "step": 2590
+    },
+    {
+      "epoch": 2.9225988700564973,
+      "grad_norm": 0.04413614049553871,
+      "learning_rate": 2.88781393181603e-07,
+      "loss": 0.9462,
+      "step": 2591
+    },
+    {
+      "epoch": 2.923728813559322,
+      "grad_norm": 0.044499874114990234,
+      "learning_rate": 2.7983165627251116e-07,
+      "loss": 1.0218,
+      "step": 2592
+    },
+    {
+      "epoch": 2.9248587570621467,
+      "grad_norm": 0.042965903878211975,
+      "learning_rate": 2.7102259903460935e-07,
+      "loss": 0.9094,
+      "step": 2593
+    },
+    {
+      "epoch": 2.9259887005649716,
+      "grad_norm": 0.04347725212574005,
+      "learning_rate": 2.623542338952345e-07,
+      "loss": 1.0104,
+      "step": 2594
+    },
+    {
+      "epoch": 2.9271186440677965,
+      "grad_norm": 0.044730983674526215,
+      "learning_rate": 2.5382657308322676e-07,
+      "loss": 0.943,
+      "step": 2595
+    },
+    {
+      "epoch": 2.9282485875706215,
+      "grad_norm": 0.04504537582397461,
+      "learning_rate": 2.4543962862894063e-07,
+      "loss": 0.9281,
+      "step": 2596
+    },
+    {
+      "epoch": 2.9293785310734464,
+      "grad_norm": 0.04293621703982353,
+      "learning_rate": 2.3719341236420057e-07,
+      "loss": 0.9524,
+      "step": 2597
+    },
+    {
+      "epoch": 2.9305084745762713,
+      "grad_norm": 0.04390338063240051,
+      "learning_rate": 2.2908793592232303e-07,
+      "loss": 0.9315,
+      "step": 2598
+    },
+    {
+      "epoch": 2.931638418079096,
+      "grad_norm": 0.04320131614804268,
+      "learning_rate": 2.211232107380612e-07,
+      "loss": 0.9744,
+      "step": 2599
+    },
+    {
+      "epoch": 2.932768361581921,
+      "grad_norm": 0.04456019774079323,
+      "learning_rate": 2.1329924804760482e-07,
+      "loss": 1.0404,
+      "step": 2600
+    },
+    {
+      "epoch": 2.933898305084746,
+      "grad_norm": 0.04545271396636963,
+      "learning_rate": 2.0561605888855805e-07,
+      "loss": 0.9803,
+      "step": 2601
+    },
+    {
+      "epoch": 2.9350282485875705,
+      "grad_norm": 0.04348572716116905,
+      "learning_rate": 1.980736540999506e-07,
+      "loss": 0.9718,
+      "step": 2602
+    },
+    {
+      "epoch": 2.9361581920903954,
+      "grad_norm": 0.04441867396235466,
+      "learning_rate": 1.906720443221821e-07,
+      "loss": 0.988,
+      "step": 2603
+    },
+    {
+      "epoch": 2.9372881355932203,
+      "grad_norm": 0.04416580870747566,
+      "learning_rate": 1.8341123999703337e-07,
+      "loss": 1.0091,
+      "step": 2604
+    },
+    {
+      "epoch": 2.9384180790960452,
+      "grad_norm": 0.04366505518555641,
+      "learning_rate": 1.7629125136764402e-07,
+      "loss": 0.9855,
+      "step": 2605
+    },
+    {
+      "epoch": 2.93954802259887,
+      "grad_norm": 0.04411231353878975,
+      "learning_rate": 1.6931208847849045e-07,
+      "loss": 0.9712,
+      "step": 2606
+    },
+    {
+      "epoch": 2.940677966101695,
+      "grad_norm": 0.04340221732854843,
+      "learning_rate": 1.6247376117539682e-07,
+      "loss": 1.0107,
+      "step": 2607
+    },
+    {
+      "epoch": 2.9418079096045195,
+      "grad_norm": 0.044296473264694214,
+      "learning_rate": 1.5577627910549063e-07,
+      "loss": 1.0654,
+      "step": 2608
+    },
+    {
+      "epoch": 2.9429378531073445,
+      "grad_norm": 0.04423755407333374,
+      "learning_rate": 1.4921965171720287e-07,
+      "loss": 0.872,
+      "step": 2609
+    },
+    {
+      "epoch": 2.9440677966101694,
+      "grad_norm": 0.044309426099061966,
+      "learning_rate": 1.4280388826026782e-07,
+      "loss": 0.8893,
+      "step": 2610
+    },
+    {
+      "epoch": 2.9451977401129943,
+      "grad_norm": 0.04572597146034241,
+      "learning_rate": 1.3652899778568985e-07,
+      "loss": 1.0088,
+      "step": 2611
+    },
+    {
+      "epoch": 2.946327683615819,
+      "grad_norm": 0.04595278576016426,
+      "learning_rate": 1.3039498914573235e-07,
+      "loss": 0.9908,
+      "step": 2612
+    },
+    {
+      "epoch": 2.947457627118644,
+      "grad_norm": 0.04614327847957611,
+      "learning_rate": 1.2440187099390654e-07,
+      "loss": 0.9492,
+      "step": 2613
+    },
+    {
+      "epoch": 2.948587570621469,
+      "grad_norm": 0.04578055068850517,
+      "learning_rate": 1.1854965178497158e-07,
+      "loss": 0.996,
+      "step": 2614
+    },
+    {
+      "epoch": 2.949717514124294,
+      "grad_norm": 0.04251642897725105,
+      "learning_rate": 1.128383397749233e-07,
+      "loss": 1.0487,
+      "step": 2615
+    },
+    {
+      "epoch": 2.950847457627119,
+      "grad_norm": 0.0450914241373539,
+      "learning_rate": 1.072679430209611e-07,
+      "loss": 0.8909,
+      "step": 2616
+    },
+    {
+      "epoch": 2.951977401129944,
+      "grad_norm": 0.04336519166827202,
+      "learning_rate": 1.0183846938148778e-07,
+      "loss": 1.069,
+      "step": 2617
+    },
+    {
+      "epoch": 2.9531073446327682,
+      "grad_norm": 0.045954346656799316,
+      "learning_rate": 9.654992651609851e-08,
+      "loss": 0.9546,
+      "step": 2618
+    },
+    {
+      "epoch": 2.954237288135593,
+      "grad_norm": 0.04354023560881615,
+      "learning_rate": 9.140232188558085e-08,
+      "loss": 0.9622,
+      "step": 2619
+    },
+    {
+      "epoch": 2.955367231638418,
+      "grad_norm": 0.044812485575675964,
+      "learning_rate": 8.639566275189248e-08,
+      "loss": 0.905,
+      "step": 2620
+    },
+    {
+      "epoch": 2.956497175141243,
+      "grad_norm": 0.044489793479442596,
+      "learning_rate": 8.152995617815018e-08,
+      "loss": 1.0048,
+      "step": 2621
+    },
+    {
+      "epoch": 2.957627118644068,
+      "grad_norm": 0.0438714399933815,
+      "learning_rate": 7.680520902860755e-08,
+      "loss": 0.9843,
+      "step": 2622
+    },
+    {
+      "epoch": 2.9587570621468924,
+      "grad_norm": 0.04459919407963753,
+      "learning_rate": 7.222142796868835e-08,
+      "loss": 0.9293,
+      "step": 2623
+    },
+    {
+      "epoch": 2.9598870056497173,
+      "grad_norm": 0.0430004857480526,
+      "learning_rate": 6.777861946493102e-08,
+      "loss": 0.9283,
+      "step": 2624
+    },
+    {
+      "epoch": 2.961016949152542,
+      "grad_norm": 0.04521413892507553,
+      "learning_rate": 6.347678978501082e-08,
+      "loss": 0.9789,
+      "step": 2625
+    },
+    {
+      "epoch": 2.962146892655367,
+      "grad_norm": 0.04452219605445862,
+      "learning_rate": 5.931594499770654e-08,
+      "loss": 0.9629,
+      "step": 2626
+    },
+    {
+      "epoch": 2.963276836158192,
+      "grad_norm": 0.04484087601304054,
+      "learning_rate": 5.529609097290056e-08,
+      "loss": 0.9989,
+      "step": 2627
+    },
+    {
+      "epoch": 2.964406779661017,
+      "grad_norm": 0.04512220248579979,
+      "learning_rate": 5.1417233381578775e-08,
+      "loss": 0.9498,
+      "step": 2628
+    },
+    {
+      "epoch": 2.965536723163842,
+      "grad_norm": 0.04488738253712654,
+      "learning_rate": 4.767937769583064e-08,
+      "loss": 0.9256,
+      "step": 2629
+    },
+    {
+      "epoch": 2.966666666666667,
+      "grad_norm": 0.04371628910303116,
+      "learning_rate": 4.408252918880473e-08,
+      "loss": 1.0065,
+      "step": 2630
+    },
+    {
+      "epoch": 2.9677966101694917,
+      "grad_norm": 0.04563480243086815,
+      "learning_rate": 4.062669293474208e-08,
+      "loss": 0.9899,
+      "step": 2631
+    },
+    {
+      "epoch": 2.9689265536723166,
+      "grad_norm": 0.04385736212134361,
+      "learning_rate": 3.731187380893175e-08,
+      "loss": 0.9837,
+      "step": 2632
+    },
+    {
+      "epoch": 2.970056497175141,
+      "grad_norm": 0.04431382194161415,
+      "learning_rate": 3.413807648775524e-08,
+      "loss": 0.9099,
+      "step": 2633
+    },
+    {
+      "epoch": 2.971186440677966,
+      "grad_norm": 0.045138463377952576,
+      "learning_rate": 3.110530544860879e-08,
+      "loss": 0.9418,
+      "step": 2634
+    },
+    {
+      "epoch": 2.972316384180791,
+      "grad_norm": 0.04519420862197876,
+      "learning_rate": 2.8213564969969963e-08,
+      "loss": 0.9792,
+      "step": 2635
+    },
+    {
+      "epoch": 2.973446327683616,
+      "grad_norm": 0.04272979870438576,
+      "learning_rate": 2.5462859131353266e-08,
+      "loss": 0.9315,
+      "step": 2636
+    },
+    {
+      "epoch": 2.9745762711864407,
+      "grad_norm": 0.043719951063394547,
+      "learning_rate": 2.2853191813276832e-08,
+      "loss": 0.8662,
+      "step": 2637
+    },
+    {
+      "epoch": 2.9757062146892657,
+      "grad_norm": 0.0462685190141201,
+      "learning_rate": 2.0384566697329023e-08,
+      "loss": 0.9797,
+      "step": 2638
+    },
+    {
+      "epoch": 2.97683615819209,
+      "grad_norm": 0.04535675793886185,
+      "learning_rate": 1.8056987266112933e-08,
+      "loss": 1.0174,
+      "step": 2639
+    },
+    {
+      "epoch": 2.977966101694915,
+      "grad_norm": 0.042910125106573105,
+      "learning_rate": 1.5870456803246392e-08,
+      "loss": 0.927,
+      "step": 2640
+    },
+    {
+      "epoch": 2.97909604519774,
+      "grad_norm": 0.044517967849969864,
+      "learning_rate": 1.3824978393361943e-08,
+      "loss": 0.9859,
+      "step": 2641
+    },
+    {
+      "epoch": 2.980225988700565,
+      "grad_norm": 0.044496312737464905,
+      "learning_rate": 1.1920554922106864e-08,
+      "loss": 1.0117,
+      "step": 2642
+    },
+    {
+      "epoch": 2.98135593220339,
+      "grad_norm": 0.044253502041101456,
+      "learning_rate": 1.0157189076132056e-08,
+      "loss": 0.957,
+      "step": 2643
+    },
+    {
+      "epoch": 2.9824858757062147,
+      "grad_norm": 0.04357994347810745,
+      "learning_rate": 8.53488334310315e-09,
+      "loss": 0.8748,
+      "step": 2644
+    },
+    {
+      "epoch": 2.9836158192090396,
+      "grad_norm": 0.043963946402072906,
+      "learning_rate": 7.053640011678297e-09,
+      "loss": 0.9425,
+      "step": 2645
+    },
+    {
+      "epoch": 2.9847457627118645,
+      "grad_norm": 0.04307138919830322,
+      "learning_rate": 5.713461171508172e-09,
+      "loss": 1.022,
+      "step": 2646
+    },
+    {
+      "epoch": 2.9858757062146895,
+      "grad_norm": 0.04411528259515762,
+      "learning_rate": 4.514348713247074e-09,
+      "loss": 0.9617,
+      "step": 2647
+    },
+    {
+      "epoch": 2.9870056497175144,
+      "grad_norm": 0.044936712831258774,
+      "learning_rate": 3.4563043285418264e-09,
+      "loss": 0.9227,
+      "step": 2648
+    },
+    {
+      "epoch": 2.988135593220339,
+      "grad_norm": 0.0439041331410408,
+      "learning_rate": 2.5393295100095695e-09,
+      "loss": 1.0105,
+      "step": 2649
+    },
+    {
+      "epoch": 2.9892655367231638,
+      "grad_norm": 0.04542997479438782,
+      "learning_rate": 1.7634255512710695e-09,
+      "loss": 0.983,
+      "step": 2650
+    },
+    {
+      "epoch": 2.9903954802259887,
+      "grad_norm": 0.044694844633340836,
+      "learning_rate": 1.1285935469285136e-09,
+      "loss": 0.8849,
+      "step": 2651
+    },
+    {
+      "epoch": 2.9915254237288136,
+      "grad_norm": 0.043964337557554245,
+      "learning_rate": 6.348343925766109e-10,
+      "loss": 1.0035,
+      "step": 2652
+    },
+    {
+      "epoch": 2.9926553672316385,
+      "grad_norm": 0.04477889835834503,
+      "learning_rate": 2.821487847692872e-10,
+      "loss": 1.0285,
+      "step": 2653
+    },
+    {
+      "epoch": 2.9937853107344634,
+      "grad_norm": 0.044497277587652206,
+      "learning_rate": 7.053722107519533e-11,
+      "loss": 0.9643,
+      "step": 2654
+    },
+    {
+      "epoch": 2.994915254237288,
+      "grad_norm": 0.04465208947658539,
+      "learning_rate": 0.0,
+      "loss": 0.9576,
+      "step": 2655
+    },
+    {
+      "epoch": 2.994915254237288,
+      "eval_loss": 0.984717071056366,
+      "eval_runtime": 562.7269,
+      "eval_samples_per_second": 17.385,
+      "eval_steps_per_second": 8.693,
+      "step": 2655
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2655,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 885,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.202210731046994e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}