diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,22639 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.997677659080353,
+  "eval_steps": 500,
+  "global_step": 3228,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0009289363678588017,
+      "grad_norm": 54.7475656620091,
+      "learning_rate": 0.0,
+      "loss": 11.0245,
+      "step": 1
+    },
+    {
+      "epoch": 0.0018578727357176034,
+      "grad_norm": 56.51890067533405,
+      "learning_rate": 1.5479876160990715e-07,
+      "loss": 10.9563,
+      "step": 2
+    },
+    {
+      "epoch": 0.002786809103576405,
+      "grad_norm": 52.84327216336746,
+      "learning_rate": 3.095975232198143e-07,
+      "loss": 11.1232,
+      "step": 3
+    },
+    {
+      "epoch": 0.0037157454714352067,
+      "grad_norm": 54.47240324631095,
+      "learning_rate": 4.6439628482972136e-07,
+      "loss": 11.0395,
+      "step": 4
+    },
+    {
+      "epoch": 0.004644681839294009,
+      "grad_norm": 56.06603126343985,
+      "learning_rate": 6.191950464396286e-07,
+      "loss": 10.9521,
+      "step": 5
+    },
+    {
+      "epoch": 0.00557361820715281,
+      "grad_norm": 56.653051495890196,
+      "learning_rate": 7.739938080495357e-07,
+      "loss": 10.9283,
+      "step": 6
+    },
+    {
+      "epoch": 0.006502554575011612,
+      "grad_norm": 57.19065708163452,
+      "learning_rate": 9.287925696594427e-07,
+      "loss": 10.9048,
+      "step": 7
+    },
+    {
+      "epoch": 0.0074314909428704135,
+      "grad_norm": 57.13309141244447,
+      "learning_rate": 1.08359133126935e-06,
+      "loss": 10.8692,
+      "step": 8
+    },
+    {
+      "epoch": 0.008360427310729215,
+      "grad_norm": 61.70021371119148,
+      "learning_rate": 1.2383900928792572e-06,
+      "loss": 10.7026,
+      "step": 9
+    },
+    {
+      "epoch": 0.009289363678588018,
+      "grad_norm": 58.64471121402323,
+      "learning_rate": 1.3931888544891641e-06,
+      "loss": 10.7762,
+      "step": 10
+    },
+    {
+      "epoch": 0.010218300046446818,
+      "grad_norm": 61.95985964008096,
+      "learning_rate": 1.5479876160990713e-06,
+      "loss": 10.6442,
+      "step": 11
+    },
+    {
+      "epoch": 0.01114723641430562,
+      "grad_norm": 81.74822353424084,
+      "learning_rate": 1.7027863777089783e-06,
+      "loss": 9.3569,
+      "step": 12
+    },
+    {
+      "epoch": 0.012076172782164421,
+      "grad_norm": 87.15614377730954,
+      "learning_rate": 1.8575851393188855e-06,
+      "loss": 9.3139,
+      "step": 13
+    },
+    {
+      "epoch": 0.013005109150023224,
+      "grad_norm": 90.69218149138536,
+      "learning_rate": 2.012383900928793e-06,
+      "loss": 8.9405,
+      "step": 14
+    },
+    {
+      "epoch": 0.013934045517882025,
+      "grad_norm": 96.84729927584988,
+      "learning_rate": 2.1671826625387e-06,
+      "loss": 8.6999,
+      "step": 15
+    },
+    {
+      "epoch": 0.014862981885740827,
+      "grad_norm": 65.07865556439408,
+      "learning_rate": 2.321981424148607e-06,
+      "loss": 3.7051,
+      "step": 16
+    },
+    {
+      "epoch": 0.01579191825359963,
+      "grad_norm": 62.52820085696381,
+      "learning_rate": 2.4767801857585144e-06,
+      "loss": 3.6981,
+      "step": 17
+    },
+    {
+      "epoch": 0.01672085462145843,
+      "grad_norm": 53.649988186848155,
+      "learning_rate": 2.631578947368421e-06,
+      "loss": 3.2385,
+      "step": 18
+    },
+    {
+      "epoch": 0.01764979098931723,
+      "grad_norm": 39.157669241392924,
+      "learning_rate": 2.7863777089783283e-06,
+      "loss": 2.6875,
+      "step": 19
+    },
+    {
+      "epoch": 0.018578727357176035,
+      "grad_norm": 34.53162931711357,
+      "learning_rate": 2.9411764705882355e-06,
+      "loss": 2.4641,
+      "step": 20
+    },
+    {
+      "epoch": 0.019507663725034836,
+      "grad_norm": 14.044732265304198,
+      "learning_rate": 3.0959752321981426e-06,
+      "loss": 1.7491,
+      "step": 21
+    },
+    {
+      "epoch": 0.020436600092893636,
+      "grad_norm": 6.143568036385673,
+      "learning_rate": 3.25077399380805e-06,
+      "loss": 1.2988,
+      "step": 22
+    },
+    {
+      "epoch": 0.021365536460752437,
+      "grad_norm": 5.2826561116659745,
+      "learning_rate": 3.4055727554179566e-06,
+      "loss": 1.3204,
+      "step": 23
+    },
+    {
+      "epoch": 0.02229447282861124,
+      "grad_norm": 4.030771783214909,
+      "learning_rate": 3.560371517027864e-06,
+      "loss": 1.2404,
+      "step": 24
+    },
+    {
+      "epoch": 0.023223409196470042,
+      "grad_norm": 3.426990839601739,
+      "learning_rate": 3.715170278637771e-06,
+      "loss": 1.1722,
+      "step": 25
+    },
+    {
+      "epoch": 0.024152345564328843,
+      "grad_norm": 2.663561421971141,
+      "learning_rate": 3.869969040247678e-06,
+      "loss": 1.132,
+      "step": 26
+    },
+    {
+      "epoch": 0.025081281932187643,
+      "grad_norm": 2.2095185737240914,
+      "learning_rate": 4.024767801857586e-06,
+      "loss": 1.0879,
+      "step": 27
+    },
+    {
+      "epoch": 0.026010218300046448,
+      "grad_norm": 1.8901365750210222,
+      "learning_rate": 4.1795665634674924e-06,
+      "loss": 1.0258,
+      "step": 28
+    },
+    {
+      "epoch": 0.02693915466790525,
+      "grad_norm": 1.3861813570839094,
+      "learning_rate": 4.3343653250774e-06,
+      "loss": 0.9735,
+      "step": 29
+    },
+    {
+      "epoch": 0.02786809103576405,
+      "grad_norm": 1.5525651218677432,
+      "learning_rate": 4.489164086687307e-06,
+      "loss": 0.9165,
+      "step": 30
+    },
+    {
+      "epoch": 0.028797027403622853,
+      "grad_norm": 4.777669808853715,
+      "learning_rate": 4.643962848297214e-06,
+      "loss": 0.8589,
+      "step": 31
+    },
+    {
+      "epoch": 0.029725963771481654,
+      "grad_norm": 1.9514011751823228,
+      "learning_rate": 4.798761609907121e-06,
+      "loss": 0.8463,
+      "step": 32
+    },
+    {
+      "epoch": 0.030654900139340455,
+      "grad_norm": 1.0398033405311462,
+      "learning_rate": 4.953560371517029e-06,
+      "loss": 0.8531,
+      "step": 33
+    },
+    {
+      "epoch": 0.03158383650719926,
+      "grad_norm": 0.9239326446187678,
+      "learning_rate": 5.1083591331269355e-06,
+      "loss": 0.8295,
+      "step": 34
+    },
+    {
+      "epoch": 0.032512772875058056,
+      "grad_norm": 0.8239281374239108,
+      "learning_rate": 5.263157894736842e-06,
+      "loss": 0.7953,
+      "step": 35
+    },
+    {
+      "epoch": 0.03344170924291686,
+      "grad_norm": 0.7456489091955973,
+      "learning_rate": 5.41795665634675e-06,
+      "loss": 0.8011,
+      "step": 36
+    },
+    {
+      "epoch": 0.034370645610775664,
+      "grad_norm": 0.7322576669859537,
+      "learning_rate": 5.5727554179566566e-06,
+      "loss": 0.7743,
+      "step": 37
+    },
+    {
+      "epoch": 0.03529958197863446,
+      "grad_norm": 0.6841211213339884,
+      "learning_rate": 5.727554179566564e-06,
+      "loss": 0.7502,
+      "step": 38
+    },
+    {
+      "epoch": 0.036228518346493266,
+      "grad_norm": 0.6392360133178303,
+      "learning_rate": 5.882352941176471e-06,
+      "loss": 0.7588,
+      "step": 39
+    },
+    {
+      "epoch": 0.03715745471435207,
+      "grad_norm": 0.5816712224571956,
+      "learning_rate": 6.0371517027863785e-06,
+      "loss": 0.763,
+      "step": 40
+    },
+    {
+      "epoch": 0.03808639108221087,
+      "grad_norm": 0.5551930154686257,
+      "learning_rate": 6.191950464396285e-06,
+      "loss": 0.7203,
+      "step": 41
+    },
+    {
+      "epoch": 0.03901532745006967,
+      "grad_norm": 0.627393609542681,
+      "learning_rate": 6.346749226006192e-06,
+      "loss": 0.7372,
+      "step": 42
+    },
+    {
+      "epoch": 0.03994426381792847,
+      "grad_norm": 0.5209393663178439,
+      "learning_rate": 6.5015479876161e-06,
+      "loss": 0.7142,
+      "step": 43
+    },
+    {
+      "epoch": 0.04087320018578727,
+      "grad_norm": 0.4311011906767075,
+      "learning_rate": 6.656346749226007e-06,
+      "loss": 0.6497,
+      "step": 44
+    },
+    {
+      "epoch": 0.04180213655364608,
+      "grad_norm": 0.44948642321517024,
+      "learning_rate": 6.811145510835913e-06,
+      "loss": 0.694,
+      "step": 45
+    },
+    {
+      "epoch": 0.042731072921504874,
+      "grad_norm": 0.4845127976887666,
+      "learning_rate": 6.965944272445821e-06,
+      "loss": 0.7111,
+      "step": 46
+    },
+    {
+      "epoch": 0.04366000928936368,
+      "grad_norm": 0.4680884482351343,
+      "learning_rate": 7.120743034055728e-06,
+      "loss": 0.6841,
+      "step": 47
+    },
+    {
+      "epoch": 0.04458894565722248,
+      "grad_norm": 0.4268577390410502,
+      "learning_rate": 7.275541795665634e-06,
+      "loss": 0.6861,
+      "step": 48
+    },
+    {
+      "epoch": 0.04551788202508128,
+      "grad_norm": 0.4221226832933112,
+      "learning_rate": 7.430340557275542e-06,
+      "loss": 0.6741,
+      "step": 49
+    },
+    {
+      "epoch": 0.046446818392940084,
+      "grad_norm": 0.5220141311400619,
+      "learning_rate": 7.585139318885449e-06,
+      "loss": 0.669,
+      "step": 50
+    },
+    {
+      "epoch": 0.04737575476079889,
+      "grad_norm": 0.42881334524580667,
+      "learning_rate": 7.739938080495356e-06,
+      "loss": 0.676,
+      "step": 51
+    },
+    {
+      "epoch": 0.048304691128657685,
+      "grad_norm": 0.3524885307145976,
+      "learning_rate": 7.894736842105263e-06,
+      "loss": 0.6381,
+      "step": 52
+    },
+    {
+      "epoch": 0.04923362749651649,
+      "grad_norm": 0.4114675079627211,
+      "learning_rate": 8.049535603715171e-06,
+      "loss": 0.6049,
+      "step": 53
+    },
+    {
+      "epoch": 0.05016256386437529,
+      "grad_norm": 0.40656103184868714,
+      "learning_rate": 8.204334365325078e-06,
+      "loss": 0.6202,
+      "step": 54
+    },
+    {
+      "epoch": 0.05109150023223409,
+      "grad_norm": 0.3732755562975793,
+      "learning_rate": 8.359133126934985e-06,
+      "loss": 0.6231,
+      "step": 55
+    },
+    {
+      "epoch": 0.052020436600092895,
+      "grad_norm": 0.3799659025653529,
+      "learning_rate": 8.513931888544892e-06,
+      "loss": 0.6167,
+      "step": 56
+    },
+    {
+      "epoch": 0.05294937296795169,
+      "grad_norm": 0.3548794527966943,
+      "learning_rate": 8.6687306501548e-06,
+      "loss": 0.6235,
+      "step": 57
+    },
+    {
+      "epoch": 0.0538783093358105,
+      "grad_norm": 0.3366392962881869,
+      "learning_rate": 8.823529411764707e-06,
+      "loss": 0.6037,
+      "step": 58
+    },
+    {
+      "epoch": 0.0548072457036693,
+      "grad_norm": 0.35622745848432813,
+      "learning_rate": 8.978328173374614e-06,
+      "loss": 0.6505,
+      "step": 59
+    },
+    {
+      "epoch": 0.0557361820715281,
+      "grad_norm": 0.33139910543786816,
+      "learning_rate": 9.13312693498452e-06,
+      "loss": 0.5907,
+      "step": 60
+    },
+    {
+      "epoch": 0.0566651184393869,
+      "grad_norm": 0.26984628140930006,
+      "learning_rate": 9.287925696594429e-06,
+      "loss": 0.6075,
+      "step": 61
+    },
+    {
+      "epoch": 0.057594054807245706,
+      "grad_norm": 0.30168457074265415,
+      "learning_rate": 9.442724458204334e-06,
+      "loss": 0.5929,
+      "step": 62
+    },
+    {
+      "epoch": 0.058522991175104504,
+      "grad_norm": 0.3514265855140124,
+      "learning_rate": 9.597523219814242e-06,
+      "loss": 0.5962,
+      "step": 63
+    },
+    {
+      "epoch": 0.05945192754296331,
+      "grad_norm": 0.3551734532081626,
+      "learning_rate": 9.752321981424149e-06,
+      "loss": 0.6197,
+      "step": 64
+    },
+    {
+      "epoch": 0.06038086391082211,
+      "grad_norm": 0.29947361957296,
+      "learning_rate": 9.907120743034057e-06,
+      "loss": 0.5796,
+      "step": 65
+    },
+    {
+      "epoch": 0.06130980027868091,
+      "grad_norm": 0.31337803543321063,
+      "learning_rate": 1.0061919504643963e-05,
+      "loss": 0.5948,
+      "step": 66
+    },
+    {
+      "epoch": 0.062238736646539713,
+      "grad_norm": 0.3215034540150923,
+      "learning_rate": 1.0216718266253871e-05,
+      "loss": 0.5481,
+      "step": 67
+    },
+    {
+      "epoch": 0.06316767301439852,
+      "grad_norm": 0.3102955136307414,
+      "learning_rate": 1.0371517027863778e-05,
+      "loss": 0.5789,
+      "step": 68
+    },
+    {
+      "epoch": 0.06409660938225732,
+      "grad_norm": 0.2573332075607664,
+      "learning_rate": 1.0526315789473684e-05,
+      "loss": 0.5656,
+      "step": 69
+    },
+    {
+      "epoch": 0.06502554575011611,
+      "grad_norm": 0.2537807753587459,
+      "learning_rate": 1.0681114551083591e-05,
+      "loss": 0.5678,
+      "step": 70
+    },
+    {
+      "epoch": 0.06595448211797492,
+      "grad_norm": 0.30023463727934896,
+      "learning_rate": 1.08359133126935e-05,
+      "loss": 0.572,
+      "step": 71
+    },
+    {
+      "epoch": 0.06688341848583372,
+      "grad_norm": 0.2705492110612055,
+      "learning_rate": 1.0990712074303406e-05,
+      "loss": 0.5956,
+      "step": 72
+    },
+    {
+      "epoch": 0.06781235485369252,
+      "grad_norm": 0.27304258159582684,
+      "learning_rate": 1.1145510835913313e-05,
+      "loss": 0.5943,
+      "step": 73
+    },
+    {
+      "epoch": 0.06874129122155133,
+      "grad_norm": 0.27460235847783515,
+      "learning_rate": 1.130030959752322e-05,
+      "loss": 0.5738,
+      "step": 74
+    },
+    {
+      "epoch": 0.06967022758941012,
+      "grad_norm": 0.24083051338698006,
+      "learning_rate": 1.1455108359133128e-05,
+      "loss": 0.5543,
+      "step": 75
+    },
+    {
+      "epoch": 0.07059916395726892,
+      "grad_norm": 0.2862680829656088,
+      "learning_rate": 1.1609907120743033e-05,
+      "loss": 0.5574,
+      "step": 76
+    },
+    {
+      "epoch": 0.07152810032512773,
+      "grad_norm": 0.2522441867822406,
+      "learning_rate": 1.1764705882352942e-05,
+      "loss": 0.5561,
+      "step": 77
+    },
+    {
+      "epoch": 0.07245703669298653,
+      "grad_norm": 0.2731655705611119,
+      "learning_rate": 1.1919504643962849e-05,
+      "loss": 0.5736,
+      "step": 78
+    },
+    {
+      "epoch": 0.07338597306084534,
+      "grad_norm": 0.2520420481239673,
+      "learning_rate": 1.2074303405572757e-05,
+      "loss": 0.5418,
+      "step": 79
+    },
+    {
+      "epoch": 0.07431490942870414,
+      "grad_norm": 0.22930537041421725,
+      "learning_rate": 1.2229102167182662e-05,
+      "loss": 0.5171,
+      "step": 80
+    },
+    {
+      "epoch": 0.07524384579656293,
+      "grad_norm": 0.273072858512794,
+      "learning_rate": 1.238390092879257e-05,
+      "loss": 0.5433,
+      "step": 81
+    },
+    {
+      "epoch": 0.07617278216442173,
+      "grad_norm": 0.2891845927667174,
+      "learning_rate": 1.2538699690402477e-05,
+      "loss": 0.5779,
+      "step": 82
+    },
+    {
+      "epoch": 0.07710171853228054,
+      "grad_norm": 0.2523018814544343,
+      "learning_rate": 1.2693498452012384e-05,
+      "loss": 0.5461,
+      "step": 83
+    },
+    {
+      "epoch": 0.07803065490013934,
+      "grad_norm": 0.3166725101519938,
+      "learning_rate": 1.2848297213622292e-05,
+      "loss": 0.5643,
+      "step": 84
+    },
+    {
+      "epoch": 0.07895959126799815,
+      "grad_norm": 0.2631824444127893,
+      "learning_rate": 1.30030959752322e-05,
+      "loss": 0.5138,
+      "step": 85
+    },
+    {
+      "epoch": 0.07988852763585694,
+      "grad_norm": 0.3004458758805224,
+      "learning_rate": 1.3157894736842106e-05,
+      "loss": 0.551,
+      "step": 86
+    },
+    {
+      "epoch": 0.08081746400371574,
+      "grad_norm": 0.27824957348028967,
+      "learning_rate": 1.3312693498452014e-05,
+      "loss": 0.5774,
+      "step": 87
+    },
+    {
+      "epoch": 0.08174640037157455,
+      "grad_norm": 0.30038729524349966,
+      "learning_rate": 1.346749226006192e-05,
+      "loss": 0.5255,
+      "step": 88
+    },
+    {
+      "epoch": 0.08267533673943335,
+      "grad_norm": 0.2942525266347019,
+      "learning_rate": 1.3622291021671826e-05,
+      "loss": 0.5569,
+      "step": 89
+    },
+    {
+      "epoch": 0.08360427310729215,
+      "grad_norm": 0.29120858799373767,
+      "learning_rate": 1.3777089783281735e-05,
+      "loss": 0.5224,
+      "step": 90
+    },
+    {
+      "epoch": 0.08453320947515096,
+      "grad_norm": 0.3596968247905661,
+      "learning_rate": 1.3931888544891641e-05,
+      "loss": 0.5199,
+      "step": 91
+    },
+    {
+      "epoch": 0.08546214584300975,
+      "grad_norm": 0.2812166695733365,
+      "learning_rate": 1.4086687306501548e-05,
+      "loss": 0.5363,
+      "step": 92
+    },
+    {
+      "epoch": 0.08639108221086855,
+      "grad_norm": 0.2992496470672875,
+      "learning_rate": 1.4241486068111457e-05,
+      "loss": 0.5634,
+      "step": 93
+    },
+    {
+      "epoch": 0.08732001857872736,
+      "grad_norm": 0.29257796169998596,
+      "learning_rate": 1.4396284829721363e-05,
+      "loss": 0.5527,
+      "step": 94
+    },
+    {
+      "epoch": 0.08824895494658616,
+      "grad_norm": 0.2732189071664008,
+      "learning_rate": 1.4551083591331268e-05,
+      "loss": 0.5026,
+      "step": 95
+    },
+    {
+      "epoch": 0.08917789131444497,
+      "grad_norm": 0.27378573341841284,
+      "learning_rate": 1.4705882352941177e-05,
+      "loss": 0.575,
+      "step": 96
+    },
+    {
+      "epoch": 0.09010682768230376,
+      "grad_norm": 0.2917134046408098,
+      "learning_rate": 1.4860681114551084e-05,
+      "loss": 0.5561,
+      "step": 97
+    },
+    {
+      "epoch": 0.09103576405016256,
+      "grad_norm": 0.26829458430398617,
+      "learning_rate": 1.5015479876160992e-05,
+      "loss": 0.5082,
+      "step": 98
+    },
+    {
+      "epoch": 0.09196470041802136,
+      "grad_norm": 0.34740553302312405,
+      "learning_rate": 1.5170278637770899e-05,
+      "loss": 0.5794,
+      "step": 99
+    },
+    {
+      "epoch": 0.09289363678588017,
+      "grad_norm": 0.2936917575237726,
+      "learning_rate": 1.5325077399380806e-05,
+      "loss": 0.548,
+      "step": 100
+    },
+    {
+      "epoch": 0.09382257315373897,
+      "grad_norm": 0.3117382936574086,
+      "learning_rate": 1.5479876160990712e-05,
+      "loss": 0.5368,
+      "step": 101
+    },
+    {
+      "epoch": 0.09475150952159778,
+      "grad_norm": 0.22828777606096362,
+      "learning_rate": 1.563467492260062e-05,
+      "loss": 0.5314,
+      "step": 102
+    },
+    {
+      "epoch": 0.09568044588945657,
+      "grad_norm": 0.2771765662389807,
+      "learning_rate": 1.5789473684210526e-05,
+      "loss": 0.5482,
+      "step": 103
+    },
+    {
+      "epoch": 0.09660938225731537,
+      "grad_norm": 0.2789355900210165,
+      "learning_rate": 1.5944272445820436e-05,
+      "loss": 0.5214,
+      "step": 104
+    },
+    {
+      "epoch": 0.09753831862517418,
+      "grad_norm": 0.29645303240245474,
+      "learning_rate": 1.6099071207430343e-05,
+      "loss": 0.5403,
+      "step": 105
+    },
+    {
+      "epoch": 0.09846725499303298,
+      "grad_norm": 0.2501658792361872,
+      "learning_rate": 1.6253869969040246e-05,
+      "loss": 0.5125,
+      "step": 106
+    },
+    {
+      "epoch": 0.09939619136089178,
+      "grad_norm": 0.27343053126969535,
+      "learning_rate": 1.6408668730650156e-05,
+      "loss": 0.4957,
+      "step": 107
+    },
+    {
+      "epoch": 0.10032512772875057,
+      "grad_norm": 0.2578817321890831,
+      "learning_rate": 1.6563467492260063e-05,
+      "loss": 0.5118,
+      "step": 108
+    },
+    {
+      "epoch": 0.10125406409660938,
+      "grad_norm": 0.26242875789808284,
+      "learning_rate": 1.671826625386997e-05,
+      "loss": 0.5009,
+      "step": 109
+    },
+    {
+      "epoch": 0.10218300046446818,
+      "grad_norm": 0.27788258841659896,
+      "learning_rate": 1.6873065015479876e-05,
+      "loss": 0.5152,
+      "step": 110
+    },
+    {
+      "epoch": 0.10311193683232699,
+      "grad_norm": 0.32503303320367943,
+      "learning_rate": 1.7027863777089783e-05,
+      "loss": 0.5675,
+      "step": 111
+    },
+    {
+      "epoch": 0.10404087320018579,
+      "grad_norm": 0.25485198676459764,
+      "learning_rate": 1.7182662538699693e-05,
+      "loss": 0.4957,
+      "step": 112
+    },
+    {
+      "epoch": 0.1049698095680446,
+      "grad_norm": 0.33134645178022537,
+      "learning_rate": 1.73374613003096e-05,
+      "loss": 0.5523,
+      "step": 113
+    },
+    {
+      "epoch": 0.10589874593590339,
+      "grad_norm": 0.29136775834871204,
+      "learning_rate": 1.7492260061919503e-05,
+      "loss": 0.5432,
+      "step": 114
+    },
+    {
+      "epoch": 0.10682768230376219,
+      "grad_norm": 0.29657828249266954,
+      "learning_rate": 1.7647058823529414e-05,
+      "loss": 0.5074,
+      "step": 115
+    },
+    {
+      "epoch": 0.107756618671621,
+      "grad_norm": 0.2694020146187824,
+      "learning_rate": 1.780185758513932e-05,
+      "loss": 0.5159,
+      "step": 116
+    },
+    {
+      "epoch": 0.1086855550394798,
+      "grad_norm": 0.32008332295676645,
+      "learning_rate": 1.7956656346749227e-05,
+      "loss": 0.5078,
+      "step": 117
+    },
+    {
+      "epoch": 0.1096144914073386,
+      "grad_norm": 0.30370374026876107,
+      "learning_rate": 1.8111455108359134e-05,
+      "loss": 0.5251,
+      "step": 118
+    },
+    {
+      "epoch": 0.11054342777519739,
+      "grad_norm": 0.2540384388820426,
+      "learning_rate": 1.826625386996904e-05,
+      "loss": 0.4889,
+      "step": 119
+    },
+    {
+      "epoch": 0.1114723641430562,
+      "grad_norm": 0.2817516173937527,
+      "learning_rate": 1.8421052631578947e-05,
+      "loss": 0.5096,
+      "step": 120
+    },
+    {
+      "epoch": 0.112401300510915,
+      "grad_norm": 0.28087162622672707,
+      "learning_rate": 1.8575851393188857e-05,
+      "loss": 0.5184,
+      "step": 121
+    },
+    {
+      "epoch": 0.1133302368787738,
+      "grad_norm": 0.2638926817448589,
+      "learning_rate": 1.873065015479876e-05,
+      "loss": 0.5073,
+      "step": 122
+    },
+    {
+      "epoch": 0.11425917324663261,
+      "grad_norm": 0.27924441224774316,
+      "learning_rate": 1.8885448916408668e-05,
+      "loss": 0.5045,
+      "step": 123
+    },
+    {
+      "epoch": 0.11518810961449141,
+      "grad_norm": 0.3038546535105753,
+      "learning_rate": 1.9040247678018578e-05,
+      "loss": 0.5647,
+      "step": 124
+    },
+    {
+      "epoch": 0.1161170459823502,
+      "grad_norm": 0.28072761267618945,
+      "learning_rate": 1.9195046439628485e-05,
+      "loss": 0.5041,
+      "step": 125
+    },
+    {
+      "epoch": 0.11704598235020901,
+      "grad_norm": 0.30934280983091644,
+      "learning_rate": 1.934984520123839e-05,
+      "loss": 0.5189,
+      "step": 126
+    },
+    {
+      "epoch": 0.11797491871806781,
+      "grad_norm": 0.3017588490881342,
+      "learning_rate": 1.9504643962848298e-05,
+      "loss": 0.5113,
+      "step": 127
+    },
+    {
+      "epoch": 0.11890385508592662,
+      "grad_norm": 0.2595245492733277,
+      "learning_rate": 1.9659442724458205e-05,
+      "loss": 0.5396,
+      "step": 128
+    },
+    {
+      "epoch": 0.11983279145378542,
+      "grad_norm": 0.3188023407278314,
+      "learning_rate": 1.9814241486068115e-05,
+      "loss": 0.4933,
+      "step": 129
+    },
+    {
+      "epoch": 0.12076172782164422,
+      "grad_norm": 0.2709603354646054,
+      "learning_rate": 1.9969040247678018e-05,
+      "loss": 0.5159,
+      "step": 130
+    },
+    {
+      "epoch": 0.12169066418950301,
+      "grad_norm": 0.28324770317715753,
+      "learning_rate": 2.0123839009287925e-05,
+      "loss": 0.4932,
+      "step": 131
+    },
+    {
+      "epoch": 0.12261960055736182,
+      "grad_norm": 0.27619503735171647,
+      "learning_rate": 2.0278637770897835e-05,
+      "loss": 0.5303,
+      "step": 132
+    },
+    {
+      "epoch": 0.12354853692522062,
+      "grad_norm": 0.26401711089649377,
+      "learning_rate": 2.0433436532507742e-05,
+      "loss": 0.4868,
+      "step": 133
+    },
+    {
+      "epoch": 0.12447747329307943,
+      "grad_norm": 0.2593772391982529,
+      "learning_rate": 2.058823529411765e-05,
+      "loss": 0.4929,
+      "step": 134
+    },
+    {
+      "epoch": 0.12540640966093822,
+      "grad_norm": 0.28933660519338583,
+      "learning_rate": 2.0743034055727555e-05,
+      "loss": 0.476,
+      "step": 135
+    },
+    {
+      "epoch": 0.12633534602879704,
+      "grad_norm": 0.23316118822530046,
+      "learning_rate": 2.0897832817337462e-05,
+      "loss": 0.4568,
+      "step": 136
+    },
+    {
+      "epoch": 0.12726428239665583,
+      "grad_norm": 0.31249950051428027,
+      "learning_rate": 2.105263157894737e-05,
+      "loss": 0.5187,
+      "step": 137
+    },
+    {
+      "epoch": 0.12819321876451464,
+      "grad_norm": 0.29165292379136437,
+      "learning_rate": 2.1207430340557276e-05,
+      "loss": 0.5117,
+      "step": 138
+    },
+    {
+      "epoch": 0.12912215513237343,
+      "grad_norm": 0.26649297211443057,
+      "learning_rate": 2.1362229102167182e-05,
+      "loss": 0.479,
+      "step": 139
+    },
+    {
+      "epoch": 0.13005109150023222,
+      "grad_norm": 0.3142784146098132,
+      "learning_rate": 2.151702786377709e-05,
+      "loss": 0.5216,
+      "step": 140
+    },
+    {
+      "epoch": 0.13098002786809104,
+      "grad_norm": 0.29423375787358585,
+      "learning_rate": 2.1671826625387e-05,
+      "loss": 0.4994,
+      "step": 141
+    },
+    {
+      "epoch": 0.13190896423594983,
+      "grad_norm": 0.28110784066712485,
+      "learning_rate": 2.1826625386996906e-05,
+      "loss": 0.4734,
+      "step": 142
+    },
+    {
+      "epoch": 0.13283790060380865,
+      "grad_norm": 0.29325368926120227,
+      "learning_rate": 2.1981424148606813e-05,
+      "loss": 0.4963,
+      "step": 143
+    },
+    {
+      "epoch": 0.13376683697166744,
+      "grad_norm": 0.29510748124918773,
+      "learning_rate": 2.213622291021672e-05,
+      "loss": 0.4982,
+      "step": 144
+    },
+    {
+      "epoch": 0.13469577333952623,
+      "grad_norm": 0.26461232115849287,
+      "learning_rate": 2.2291021671826626e-05,
+      "loss": 0.5166,
+      "step": 145
+    },
+    {
+      "epoch": 0.13562470970738505,
+      "grad_norm": 0.31236587334144944,
+      "learning_rate": 2.2445820433436533e-05,
+      "loss": 0.5312,
+      "step": 146
+    },
+    {
+      "epoch": 0.13655364607524384,
+      "grad_norm": 0.300081395956833,
+      "learning_rate": 2.260061919504644e-05,
+      "loss": 0.5129,
+      "step": 147
+    },
+    {
+      "epoch": 0.13748258244310266,
+      "grad_norm": 0.31610036230407396,
+      "learning_rate": 2.2755417956656347e-05,
+      "loss": 0.4776,
+      "step": 148
+    },
+    {
+      "epoch": 0.13841151881096145,
+      "grad_norm": 0.30145949638272806,
+      "learning_rate": 2.2910216718266257e-05,
+      "loss": 0.4902,
+      "step": 149
+    },
+    {
+      "epoch": 0.13934045517882024,
+      "grad_norm": 0.27339946306277607,
+      "learning_rate": 2.3065015479876163e-05,
+      "loss": 0.464,
+      "step": 150
+    },
+    {
+      "epoch": 0.14026939154667906,
+      "grad_norm": 0.3156892182016014,
+      "learning_rate": 2.3219814241486067e-05,
+      "loss": 0.495,
+      "step": 151
+    },
+    {
+      "epoch": 0.14119832791453785,
+      "grad_norm": 0.3687764130938456,
+      "learning_rate": 2.3374613003095977e-05,
+      "loss": 0.4868,
+      "step": 152
+    },
+    {
+      "epoch": 0.14212726428239666,
+      "grad_norm": 0.2800596635623465,
+      "learning_rate": 2.3529411764705884e-05,
+      "loss": 0.4739,
+      "step": 153
+    },
+    {
+      "epoch": 0.14305620065025546,
+      "grad_norm": 0.3992825928546232,
+      "learning_rate": 2.368421052631579e-05,
+      "loss": 0.4913,
+      "step": 154
+    },
+    {
+      "epoch": 0.14398513701811425,
+      "grad_norm": 0.3239223108616911,
+      "learning_rate": 2.3839009287925697e-05,
+      "loss": 0.4734,
+      "step": 155
+    },
+    {
+      "epoch": 0.14491407338597306,
+      "grad_norm": 0.3963740287758293,
+      "learning_rate": 2.3993808049535604e-05,
+      "loss": 0.5013,
+      "step": 156
+    },
+    {
+      "epoch": 0.14584300975383185,
+      "grad_norm": 0.36982949622797573,
+      "learning_rate": 2.4148606811145514e-05,
+      "loss": 0.5329,
+      "step": 157
+    },
+    {
+      "epoch": 0.14677194612169067,
+      "grad_norm": 0.3446342173260943,
+      "learning_rate": 2.430340557275542e-05,
+      "loss": 0.4617,
+      "step": 158
+    },
+    {
+      "epoch": 0.14770088248954946,
+      "grad_norm": 0.3726891488879446,
+      "learning_rate": 2.4458204334365324e-05,
+      "loss": 0.5158,
+      "step": 159
+    },
+    {
+      "epoch": 0.14862981885740828,
+      "grad_norm": 0.34306897903348194,
+      "learning_rate": 2.4613003095975234e-05,
+      "loss": 0.5011,
+      "step": 160
+    },
+    {
+      "epoch": 0.14955875522526707,
+      "grad_norm": 0.3509672474944589,
+      "learning_rate": 2.476780185758514e-05,
+      "loss": 0.5179,
+      "step": 161
+    },
+    {
+      "epoch": 0.15048769159312586,
+      "grad_norm": 0.36888921482695486,
+      "learning_rate": 2.4922600619195048e-05,
+      "loss": 0.488,
+      "step": 162
+    },
+    {
+      "epoch": 0.15141662796098468,
+      "grad_norm": 0.2879360129205803,
+      "learning_rate": 2.5077399380804955e-05,
+      "loss": 0.4818,
+      "step": 163
+    },
+    {
+      "epoch": 0.15234556432884347,
+      "grad_norm": 0.3541612368911695,
+      "learning_rate": 2.5232198142414865e-05,
+      "loss": 0.4645,
+      "step": 164
+    },
+    {
+      "epoch": 0.1532745006967023,
+      "grad_norm": 0.35885056781188374,
+      "learning_rate": 2.5386996904024768e-05,
+      "loss": 0.5074,
+      "step": 165
+    },
+    {
+      "epoch": 0.15420343706456108,
+      "grad_norm": 0.34753901489218775,
+      "learning_rate": 2.5541795665634678e-05,
+      "loss": 0.5209,
+      "step": 166
+    },
+    {
+      "epoch": 0.15513237343241987,
+      "grad_norm": 0.3353853516137484,
+      "learning_rate": 2.5696594427244585e-05,
+      "loss": 0.5021,
+      "step": 167
+    },
+    {
+      "epoch": 0.15606130980027869,
+      "grad_norm": 0.3298140566724545,
+      "learning_rate": 2.585139318885449e-05,
+      "loss": 0.5159,
+      "step": 168
+    },
+    {
+      "epoch": 0.15699024616813748,
+      "grad_norm": 0.3391997674833228,
+      "learning_rate": 2.60061919504644e-05,
+      "loss": 0.4754,
+      "step": 169
+    },
+    {
+      "epoch": 0.1579191825359963,
+      "grad_norm": 0.28704771216792646,
+      "learning_rate": 2.616099071207431e-05,
+      "loss": 0.4703,
+      "step": 170
+    },
+    {
+      "epoch": 0.15884811890385508,
+      "grad_norm": 0.37267201290204965,
+      "learning_rate": 2.6315789473684212e-05,
+      "loss": 0.5214,
+      "step": 171
+    },
+    {
+      "epoch": 0.15977705527171387,
+      "grad_norm": 0.27902165277762125,
+      "learning_rate": 2.647058823529412e-05,
+      "loss": 0.4699,
+      "step": 172
+    },
+    {
+      "epoch": 0.1607059916395727,
+      "grad_norm": 0.3398458423351264,
+      "learning_rate": 2.662538699690403e-05,
+      "loss": 0.5197,
+      "step": 173
+    },
+    {
+      "epoch": 0.16163492800743148,
+      "grad_norm": 0.3435324805081321,
+      "learning_rate": 2.6780185758513932e-05,
+      "loss": 0.4789,
+      "step": 174
+    },
+    {
+      "epoch": 0.1625638643752903,
+      "grad_norm": 0.4351648566726122,
+      "learning_rate": 2.693498452012384e-05,
+      "loss": 0.4987,
+      "step": 175
+    },
+    {
+      "epoch": 0.1634928007431491,
+      "grad_norm": 0.3198149348956976,
+      "learning_rate": 2.708978328173375e-05,
+      "loss": 0.475,
+      "step": 176
+    },
+    {
+      "epoch": 0.1644217371110079,
+      "grad_norm": 0.3950400431692381,
+      "learning_rate": 2.7244582043343652e-05,
+      "loss": 0.4744,
+      "step": 177
+    },
+    {
+      "epoch": 0.1653506734788667,
+      "grad_norm": 0.3760190425753378,
+      "learning_rate": 2.7399380804953563e-05,
+      "loss": 0.4938,
+      "step": 178
+    },
+    {
+      "epoch": 0.1662796098467255,
+      "grad_norm": 0.417176002219102,
+      "learning_rate": 2.755417956656347e-05,
+      "loss": 0.513,
+      "step": 179
+    },
+    {
+      "epoch": 0.1672085462145843,
+      "grad_norm": 0.3474000705492875,
+      "learning_rate": 2.7708978328173373e-05,
+      "loss": 0.464,
+      "step": 180
+    },
+    {
+      "epoch": 0.1681374825824431,
+      "grad_norm": 0.35375534821068394,
+      "learning_rate": 2.7863777089783283e-05,
+      "loss": 0.5123,
+      "step": 181
+    },
+    {
+      "epoch": 0.16906641895030192,
+      "grad_norm": 0.4153734241752173,
+      "learning_rate": 2.8018575851393193e-05,
+      "loss": 0.4786,
+      "step": 182
+    },
+    {
+      "epoch": 0.1699953553181607,
+      "grad_norm": 0.3633545185465792,
+      "learning_rate": 2.8173374613003096e-05,
+      "loss": 0.4881,
+      "step": 183
+    },
+    {
+      "epoch": 0.1709242916860195,
+      "grad_norm": 0.4177673269252791,
+      "learning_rate": 2.8328173374613003e-05,
+      "loss": 0.4851,
+      "step": 184
+    },
+    {
+      "epoch": 0.17185322805387832,
+      "grad_norm": 0.34872663223370376,
+      "learning_rate": 2.8482972136222913e-05,
+      "loss": 0.4562,
+      "step": 185
+    },
+    {
+      "epoch": 0.1727821644217371,
+      "grad_norm": 0.3826892295449655,
+      "learning_rate": 2.8637770897832817e-05,
+      "loss": 0.4704,
+      "step": 186
+    },
+    {
+      "epoch": 0.17371110078959592,
+      "grad_norm": 0.2911667819931114,
+      "learning_rate": 2.8792569659442727e-05,
+      "loss": 0.4647,
+      "step": 187
+    },
+    {
+      "epoch": 0.17464003715745471,
+      "grad_norm": 0.3577750615715225,
+      "learning_rate": 2.8947368421052634e-05,
+      "loss": 0.4758,
+      "step": 188
+    },
+    {
+      "epoch": 0.1755689735253135,
+      "grad_norm": 0.28796027674445784,
+      "learning_rate": 2.9102167182662537e-05,
+      "loss": 0.4934,
+      "step": 189
+    },
+    {
+      "epoch": 0.17649790989317232,
+      "grad_norm": 0.328868526631627,
+      "learning_rate": 2.9256965944272447e-05,
+      "loss": 0.4989,
+      "step": 190
+    },
+    {
+      "epoch": 0.1774268462610311,
+      "grad_norm": 0.30926168846748986,
+      "learning_rate": 2.9411764705882354e-05,
+      "loss": 0.4641,
+      "step": 191
+    },
+    {
+      "epoch": 0.17835578262888993,
+      "grad_norm": 0.3990926605295306,
+      "learning_rate": 2.9566563467492264e-05,
+      "loss": 0.5012,
+      "step": 192
+    },
+    {
+      "epoch": 0.17928471899674872,
+      "grad_norm": 0.2967167603656067,
+      "learning_rate": 2.9721362229102167e-05,
+      "loss": 0.4804,
+      "step": 193
+    },
+    {
+      "epoch": 0.1802136553646075,
+      "grad_norm": 0.334297436056706,
+      "learning_rate": 2.9876160990712077e-05,
+      "loss": 0.4589,
+      "step": 194
+    },
+    {
+      "epoch": 0.18114259173246633,
+      "grad_norm": 0.39015351822564853,
+      "learning_rate": 3.0030959752321984e-05,
+      "loss": 0.5029,
+      "step": 195
+    },
+    {
+      "epoch": 0.18207152810032512,
+      "grad_norm": 0.3166981562612517,
+      "learning_rate": 3.0185758513931888e-05,
+      "loss": 0.4826,
+      "step": 196
+    },
+    {
+      "epoch": 0.18300046446818394,
+      "grad_norm": 0.3697612896094445,
+      "learning_rate": 3.0340557275541798e-05,
+      "loss": 0.4834,
+      "step": 197
+    },
+    {
+      "epoch": 0.18392940083604273,
+      "grad_norm": 0.3931362669456308,
+      "learning_rate": 3.0495356037151708e-05,
+      "loss": 0.4827,
+      "step": 198
+    },
+    {
+      "epoch": 0.18485833720390155,
+      "grad_norm": 0.37542137369737677,
+      "learning_rate": 3.065015479876161e-05,
+      "loss": 0.5137,
+      "step": 199
+    },
+    {
+      "epoch": 0.18578727357176034,
+      "grad_norm": 0.40365155511717965,
+      "learning_rate": 3.080495356037152e-05,
+      "loss": 0.4616,
+      "step": 200
+    },
+    {
+      "epoch": 0.18671620993961913,
+      "grad_norm": 0.3462177469197346,
+      "learning_rate": 3.0959752321981425e-05,
+      "loss": 0.5022,
+      "step": 201
+    },
+    {
+      "epoch": 0.18764514630747794,
+      "grad_norm": 0.4146922057088302,
+      "learning_rate": 3.111455108359133e-05,
+      "loss": 0.463,
+      "step": 202
+    },
+    {
+      "epoch": 0.18857408267533673,
+      "grad_norm": 0.34021579524785434,
+      "learning_rate": 3.126934984520124e-05,
+      "loss": 0.4588,
+      "step": 203
+    },
+    {
+      "epoch": 0.18950301904319555,
+      "grad_norm": 0.37786067260822703,
+      "learning_rate": 3.142414860681115e-05,
+      "loss": 0.4759,
+      "step": 204
+    },
+    {
+      "epoch": 0.19043195541105434,
+      "grad_norm": 0.3871068334106478,
+      "learning_rate": 3.157894736842105e-05,
+      "loss": 0.4702,
+      "step": 205
+    },
+    {
+      "epoch": 0.19136089177891313,
+      "grad_norm": 0.30844377564643344,
+      "learning_rate": 3.173374613003096e-05,
+      "loss": 0.4506,
+      "step": 206
+    },
+    {
+      "epoch": 0.19228982814677195,
+      "grad_norm": 0.387790224919225,
+      "learning_rate": 3.188854489164087e-05,
+      "loss": 0.4554,
+      "step": 207
+    },
+    {
+      "epoch": 0.19321876451463074,
+      "grad_norm": 0.3486750036808893,
+      "learning_rate": 3.204334365325077e-05,
+      "loss": 0.4735,
+      "step": 208
+    },
+    {
+      "epoch": 0.19414770088248956,
+      "grad_norm": 0.36108244308720244,
+      "learning_rate": 3.2198142414860685e-05,
+      "loss": 0.4516,
+      "step": 209
+    },
+    {
+      "epoch": 0.19507663725034835,
+      "grad_norm": 0.34168388969435387,
+      "learning_rate": 3.235294117647059e-05,
+      "loss": 0.4716,
+      "step": 210
+    },
+    {
+      "epoch": 0.19600557361820714,
+      "grad_norm": 0.3336301263066555,
+      "learning_rate": 3.250773993808049e-05,
+      "loss": 0.462,
+      "step": 211
+    },
+    {
+      "epoch": 0.19693450998606596,
+      "grad_norm": 0.30575256557626257,
+      "learning_rate": 3.2662538699690406e-05,
+      "loss": 0.433,
+      "step": 212
+    },
+    {
+      "epoch": 0.19786344635392475,
+      "grad_norm": 0.31175854650330226,
+      "learning_rate": 3.281733746130031e-05,
+      "loss": 0.4831,
+      "step": 213
+    },
+    {
+      "epoch": 0.19879238272178357,
+      "grad_norm": 0.3195577431584941,
+      "learning_rate": 3.297213622291022e-05,
+      "loss": 0.4939,
+      "step": 214
+    },
+    {
+      "epoch": 0.19972131908964236,
+      "grad_norm": 0.324334575666152,
+      "learning_rate": 3.3126934984520126e-05,
+      "loss": 0.4674,
+      "step": 215
+    },
+    {
+      "epoch": 0.20065025545750115,
+      "grad_norm": 0.30930275701374427,
+      "learning_rate": 3.328173374613003e-05,
+      "loss": 0.4861,
+      "step": 216
+    },
+    {
+      "epoch": 0.20157919182535997,
+      "grad_norm": 0.3196175745480417,
+      "learning_rate": 3.343653250773994e-05,
+      "loss": 0.4785,
+      "step": 217
+    },
+    {
+      "epoch": 0.20250812819321876,
+      "grad_norm": 0.3460990192307274,
+      "learning_rate": 3.3591331269349846e-05,
+      "loss": 0.4843,
+      "step": 218
+    },
+    {
+      "epoch": 0.20343706456107757,
+      "grad_norm": 0.3189777288334882,
+      "learning_rate": 3.374613003095975e-05,
+      "loss": 0.5066,
+      "step": 219
+    },
+    {
+      "epoch": 0.20436600092893636,
+      "grad_norm": 0.36563447762228074,
+      "learning_rate": 3.390092879256966e-05,
+      "loss": 0.4544,
+      "step": 220
+    },
+    {
+      "epoch": 0.20529493729679518,
+      "grad_norm": 0.3437365831635904,
+      "learning_rate": 3.4055727554179566e-05,
+      "loss": 0.4644,
+      "step": 221
+    },
+    {
+      "epoch": 0.20622387366465397,
+      "grad_norm": 0.3568648135029391,
+      "learning_rate": 3.421052631578947e-05,
+      "loss": 0.4934,
+      "step": 222
+    },
+    {
+      "epoch": 0.20715281003251276,
+      "grad_norm": 0.40598865357747305,
+      "learning_rate": 3.436532507739939e-05,
+      "loss": 0.4913,
+      "step": 223
+    },
+    {
+      "epoch": 0.20808174640037158,
+      "grad_norm": 0.4241132879182254,
+      "learning_rate": 3.452012383900929e-05,
+      "loss": 0.4535,
+      "step": 224
+    },
+    {
+      "epoch": 0.20901068276823037,
+      "grad_norm": 0.35305194926225575,
+      "learning_rate": 3.46749226006192e-05,
+      "loss": 0.4682,
+      "step": 225
+    },
+    {
+      "epoch": 0.2099396191360892,
+      "grad_norm": 0.618623679289102,
+      "learning_rate": 3.482972136222911e-05,
+      "loss": 0.4818,
+      "step": 226
+    },
+    {
+      "epoch": 0.21086855550394798,
+      "grad_norm": 0.5920146651422737,
+      "learning_rate": 3.498452012383901e-05,
+      "loss": 0.4841,
+      "step": 227
+    },
+    {
+      "epoch": 0.21179749187180677,
+      "grad_norm": 0.527405882028656,
+      "learning_rate": 3.513931888544892e-05,
+      "loss": 0.4834,
+      "step": 228
+    },
+    {
+      "epoch": 0.2127264282396656,
+      "grad_norm": 0.8374512884655507,
+      "learning_rate": 3.529411764705883e-05,
+      "loss": 0.4914,
+      "step": 229
+    },
+    {
+      "epoch": 0.21365536460752438,
+      "grad_norm": 0.7597460219900155,
+      "learning_rate": 3.5448916408668734e-05,
+      "loss": 0.4609,
+      "step": 230
+    },
+    {
+      "epoch": 0.2145843009753832,
+      "grad_norm": 0.5139866484545614,
+      "learning_rate": 3.560371517027864e-05,
+      "loss": 0.457,
+      "step": 231
+    },
+    {
+      "epoch": 0.215513237343242,
+      "grad_norm": 1.235134692869283,
+      "learning_rate": 3.575851393188855e-05,
+      "loss": 0.5001,
+      "step": 232
+    },
+    {
+      "epoch": 0.21644217371110078,
+      "grad_norm": 0.41145239593782507,
+      "learning_rate": 3.5913312693498454e-05,
+      "loss": 0.4375,
+      "step": 233
+    },
+    {
+      "epoch": 0.2173711100789596,
+      "grad_norm": 0.6816112033947701,
+      "learning_rate": 3.606811145510836e-05,
+      "loss": 0.4771,
+      "step": 234
+    },
+    {
+      "epoch": 0.21830004644681839,
+      "grad_norm": 0.4343968108259159,
+      "learning_rate": 3.622291021671827e-05,
+      "loss": 0.4565,
+      "step": 235
+    },
+    {
+      "epoch": 0.2192289828146772,
+      "grad_norm": 0.6469045830088757,
+      "learning_rate": 3.6377708978328174e-05,
+      "loss": 0.4705,
+      "step": 236
+    },
+    {
+      "epoch": 0.220157919182536,
+      "grad_norm": 0.43542141035499127,
+      "learning_rate": 3.653250773993808e-05,
+      "loss": 0.4497,
+      "step": 237
+    },
+    {
+      "epoch": 0.22108685555039478,
+      "grad_norm": 0.5546407234766495,
+      "learning_rate": 3.668730650154799e-05,
+      "loss": 0.4678,
+      "step": 238
+    },
+    {
+      "epoch": 0.2220157919182536,
+      "grad_norm": 0.4764192191948784,
+      "learning_rate": 3.6842105263157895e-05,
+      "loss": 0.475,
+      "step": 239
+    },
+    {
+      "epoch": 0.2229447282861124,
+      "grad_norm": 0.43770272881976247,
+      "learning_rate": 3.69969040247678e-05,
+      "loss": 0.4427,
+      "step": 240
+    },
+    {
+      "epoch": 0.2238736646539712,
+      "grad_norm": 0.4550187245086204,
+      "learning_rate": 3.7151702786377715e-05,
+      "loss": 0.4414,
+      "step": 241
+    },
+    {
+      "epoch": 0.22480260102183,
+      "grad_norm": 0.5330169483676697,
+      "learning_rate": 3.7306501547987615e-05,
+      "loss": 0.4819,
+      "step": 242
+    },
+    {
+      "epoch": 0.22573153738968882,
+      "grad_norm": 0.4355552273324838,
+      "learning_rate": 3.746130030959752e-05,
+      "loss": 0.473,
+      "step": 243
+    },
+    {
+      "epoch": 0.2266604737575476,
+      "grad_norm": 0.529146441551152,
+      "learning_rate": 3.7616099071207435e-05,
+      "loss": 0.5233,
+      "step": 244
+    },
+    {
+      "epoch": 0.2275894101254064,
+      "grad_norm": 0.42768230047362027,
+      "learning_rate": 3.7770897832817335e-05,
+      "loss": 0.4652,
+      "step": 245
+    },
+    {
+      "epoch": 0.22851834649326522,
+      "grad_norm": 0.498811282728108,
+      "learning_rate": 3.792569659442725e-05,
+      "loss": 0.4983,
+      "step": 246
+    },
+    {
+      "epoch": 0.229447282861124,
+      "grad_norm": 0.4438510879340567,
+      "learning_rate": 3.8080495356037155e-05,
+      "loss": 0.4366,
+      "step": 247
+    },
+    {
+      "epoch": 0.23037621922898283,
+      "grad_norm": 0.455953819278577,
+      "learning_rate": 3.8235294117647055e-05,
+      "loss": 0.4506,
+      "step": 248
+    },
+    {
+      "epoch": 0.23130515559684162,
+      "grad_norm": 0.40125931844226015,
+      "learning_rate": 3.839009287925697e-05,
+      "loss": 0.4626,
+      "step": 249
+    },
+    {
+      "epoch": 0.2322340919647004,
+      "grad_norm": 0.3539141975061322,
+      "learning_rate": 3.8544891640866876e-05,
+      "loss": 0.4499,
+      "step": 250
+    },
+    {
+      "epoch": 0.23316302833255922,
+      "grad_norm": 0.5042429216492247,
+      "learning_rate": 3.869969040247678e-05,
+      "loss": 0.4647,
+      "step": 251
+    },
+    {
+      "epoch": 0.23409196470041801,
+      "grad_norm": 0.3535166606707465,
+      "learning_rate": 3.885448916408669e-05,
+      "loss": 0.4746,
+      "step": 252
+    },
+    {
+      "epoch": 0.23502090106827683,
+      "grad_norm": 0.5056187976580123,
+      "learning_rate": 3.9009287925696596e-05,
+      "loss": 0.4405,
+      "step": 253
+    },
+    {
+      "epoch": 0.23594983743613562,
+      "grad_norm": 0.33101182681558045,
+      "learning_rate": 3.91640866873065e-05,
+      "loss": 0.4589,
+      "step": 254
+    },
+    {
+      "epoch": 0.2368787738039944,
+      "grad_norm": 0.4759625175194635,
+      "learning_rate": 3.931888544891641e-05,
+      "loss": 0.4532,
+      "step": 255
+    },
+    {
+      "epoch": 0.23780771017185323,
+      "grad_norm": 0.3870086650346993,
+      "learning_rate": 3.9473684210526316e-05,
+      "loss": 0.4468,
+      "step": 256
+    },
+    {
+      "epoch": 0.23873664653971202,
+      "grad_norm": 0.408444973925042,
+      "learning_rate": 3.962848297213623e-05,
+      "loss": 0.4646,
+      "step": 257
+    },
+    {
+      "epoch": 0.23966558290757084,
+      "grad_norm": 0.44744012598105387,
+      "learning_rate": 3.978328173374613e-05,
+      "loss": 0.455,
+      "step": 258
+    },
+    {
+      "epoch": 0.24059451927542963,
+      "grad_norm": 0.37785914884066873,
+      "learning_rate": 3.9938080495356037e-05,
+      "loss": 0.4874,
+      "step": 259
+    },
+    {
+      "epoch": 0.24152345564328845,
+      "grad_norm": 0.476511379545392,
+      "learning_rate": 4.009287925696595e-05,
+      "loss": 0.4546,
+      "step": 260
+    },
+    {
+      "epoch": 0.24245239201114724,
+      "grad_norm": 0.31290275808884926,
+      "learning_rate": 4.024767801857585e-05,
+      "loss": 0.4567,
+      "step": 261
+    },
+    {
+      "epoch": 0.24338132837900603,
+      "grad_norm": 0.46969828490127746,
+      "learning_rate": 4.0402476780185764e-05,
+      "loss": 0.4921,
+      "step": 262
+    },
+    {
+      "epoch": 0.24431026474686485,
+      "grad_norm": 0.29314739835009784,
+      "learning_rate": 4.055727554179567e-05,
+      "loss": 0.4263,
+      "step": 263
+    },
+    {
+      "epoch": 0.24523920111472364,
+      "grad_norm": 0.4260667269541755,
+      "learning_rate": 4.071207430340557e-05,
+      "loss": 0.4473,
+      "step": 264
+    },
+    {
+      "epoch": 0.24616813748258246,
+      "grad_norm": 0.29731013199401535,
+      "learning_rate": 4.0866873065015484e-05,
+      "loss": 0.4657,
+      "step": 265
+    },
+    {
+      "epoch": 0.24709707385044125,
+      "grad_norm": 0.38120279832509496,
+      "learning_rate": 4.102167182662539e-05,
+      "loss": 0.4368,
+      "step": 266
+    },
+    {
+      "epoch": 0.24802601021830004,
+      "grad_norm": 0.3167100904137455,
+      "learning_rate": 4.11764705882353e-05,
+      "loss": 0.4562,
+      "step": 267
+    },
+    {
+      "epoch": 0.24895494658615885,
+      "grad_norm": 0.4814669900513275,
+      "learning_rate": 4.1331269349845204e-05,
+      "loss": 0.4461,
+      "step": 268
+    },
+    {
+      "epoch": 0.24988388295401764,
+      "grad_norm": 0.32074498184201283,
+      "learning_rate": 4.148606811145511e-05,
+      "loss": 0.4613,
+      "step": 269
+    },
+    {
+      "epoch": 0.25081281932187643,
+      "grad_norm": 0.4084599440557474,
+      "learning_rate": 4.164086687306502e-05,
+      "loss": 0.47,
+      "step": 270
+    },
+    {
+      "epoch": 0.2517417556897353,
+      "grad_norm": 0.32165189918387554,
+      "learning_rate": 4.1795665634674924e-05,
+      "loss": 0.4814,
+      "step": 271
+    },
+    {
+      "epoch": 0.25267069205759407,
+      "grad_norm": 0.32816460839997996,
+      "learning_rate": 4.195046439628483e-05,
+      "loss": 0.4256,
+      "step": 272
+    },
+    {
+      "epoch": 0.25359962842545286,
+      "grad_norm": 0.3008939493300022,
+      "learning_rate": 4.210526315789474e-05,
+      "loss": 0.473,
+      "step": 273
+    },
+    {
+      "epoch": 0.25452856479331165,
+      "grad_norm": 0.3529450610360616,
+      "learning_rate": 4.2260061919504645e-05,
+      "loss": 0.4547,
+      "step": 274
+    },
+    {
+      "epoch": 0.25545750116117044,
+      "grad_norm": 0.2972895680132811,
+      "learning_rate": 4.241486068111455e-05,
+      "loss": 0.4539,
+      "step": 275
+    },
+    {
+      "epoch": 0.2563864375290293,
+      "grad_norm": 0.30425905871603104,
+      "learning_rate": 4.256965944272446e-05,
+      "loss": 0.4468,
+      "step": 276
+    },
+    {
+      "epoch": 0.2573153738968881,
+      "grad_norm": 0.314584382041487,
+      "learning_rate": 4.2724458204334365e-05,
+      "loss": 0.4952,
+      "step": 277
+    },
+    {
+      "epoch": 0.25824431026474687,
+      "grad_norm": 0.3160101237954399,
+      "learning_rate": 4.287925696594428e-05,
+      "loss": 0.4654,
+      "step": 278
+    },
+    {
+      "epoch": 0.25917324663260566,
+      "grad_norm": 0.31709523022756986,
+      "learning_rate": 4.303405572755418e-05,
+      "loss": 0.4673,
+      "step": 279
+    },
+    {
+      "epoch": 0.26010218300046445,
+      "grad_norm": 0.2980363449021666,
+      "learning_rate": 4.3188854489164085e-05,
+      "loss": 0.4187,
+      "step": 280
+    },
+    {
+      "epoch": 0.2610311193683233,
+      "grad_norm": 0.33880341972476546,
+      "learning_rate": 4.3343653250774e-05,
+      "loss": 0.477,
+      "step": 281
+    },
+    {
+      "epoch": 0.2619600557361821,
+      "grad_norm": 0.35521193406469326,
+      "learning_rate": 4.3498452012383905e-05,
+      "loss": 0.4762,
+      "step": 282
+    },
+    {
+      "epoch": 0.2628889921040409,
+      "grad_norm": 0.39850142060231275,
+      "learning_rate": 4.365325077399381e-05,
+      "loss": 0.4634,
+      "step": 283
+    },
+    {
+      "epoch": 0.26381792847189967,
+      "grad_norm": 0.36023726265847905,
+      "learning_rate": 4.380804953560372e-05,
+      "loss": 0.4757,
+      "step": 284
+    },
+    {
+      "epoch": 0.26474686483975846,
+      "grad_norm": 0.32295320874219785,
+      "learning_rate": 4.3962848297213626e-05,
+      "loss": 0.4523,
+      "step": 285
+    },
+    {
+      "epoch": 0.2656758012076173,
+      "grad_norm": 0.42060056160022957,
+      "learning_rate": 4.411764705882353e-05,
+      "loss": 0.4928,
+      "step": 286
+    },
+    {
+      "epoch": 0.2666047375754761,
+      "grad_norm": 0.38366257834055983,
+      "learning_rate": 4.427244582043344e-05,
+      "loss": 0.4773,
+      "step": 287
+    },
+    {
+      "epoch": 0.2675336739433349,
+      "grad_norm": 0.34821572318085275,
+      "learning_rate": 4.4427244582043346e-05,
+      "loss": 0.4495,
+      "step": 288
+    },
+    {
+      "epoch": 0.2684626103111937,
+      "grad_norm": 0.35080988887460157,
+      "learning_rate": 4.458204334365325e-05,
+      "loss": 0.4619,
+      "step": 289
+    },
+    {
+      "epoch": 0.26939154667905246,
+      "grad_norm": 0.3120875797937547,
+      "learning_rate": 4.473684210526316e-05,
+      "loss": 0.4354,
+      "step": 290
+    },
+    {
+      "epoch": 0.2703204830469113,
+      "grad_norm": 0.36916544806752755,
+      "learning_rate": 4.4891640866873066e-05,
+      "loss": 0.4428,
+      "step": 291
+    },
+    {
+      "epoch": 0.2712494194147701,
+      "grad_norm": 0.3120938357971595,
+      "learning_rate": 4.504643962848297e-05,
+      "loss": 0.479,
+      "step": 292
+    },
+    {
+      "epoch": 0.2721783557826289,
+      "grad_norm": 0.33009772967549367,
+      "learning_rate": 4.520123839009288e-05,
+      "loss": 0.44,
+      "step": 293
+    },
+    {
+      "epoch": 0.2731072921504877,
+      "grad_norm": 0.3147604795938479,
+      "learning_rate": 4.535603715170279e-05,
+      "loss": 0.4873,
+      "step": 294
+    },
+    {
+      "epoch": 0.27403622851834647,
+      "grad_norm": 0.3024028343733452,
+      "learning_rate": 4.551083591331269e-05,
+      "loss": 0.4684,
+      "step": 295
+    },
+    {
+      "epoch": 0.2749651648862053,
+      "grad_norm": 0.31188391724375253,
+      "learning_rate": 4.56656346749226e-05,
+      "loss": 0.447,
+      "step": 296
+    },
+    {
+      "epoch": 0.2758941012540641,
+      "grad_norm": 0.28892910625279733,
+      "learning_rate": 4.582043343653251e-05,
+      "loss": 0.4665,
+      "step": 297
+    },
+    {
+      "epoch": 0.2768230376219229,
+      "grad_norm": 0.3736874839548374,
+      "learning_rate": 4.597523219814241e-05,
+      "loss": 0.4526,
+      "step": 298
+    },
+    {
+      "epoch": 0.2777519739897817,
+      "grad_norm": 0.28993119120770083,
+      "learning_rate": 4.613003095975233e-05,
+      "loss": 0.4526,
+      "step": 299
+    },
+    {
+      "epoch": 0.2786809103576405,
+      "grad_norm": 0.3489816006908061,
+      "learning_rate": 4.6284829721362234e-05,
+      "loss": 0.4389,
+      "step": 300
+    },
+    {
+      "epoch": 0.2796098467254993,
+      "grad_norm": 0.32821341978908847,
+      "learning_rate": 4.6439628482972134e-05,
+      "loss": 0.47,
+      "step": 301
+    },
+    {
+      "epoch": 0.2805387830933581,
+      "grad_norm": 0.3101425751832829,
+      "learning_rate": 4.659442724458205e-05,
+      "loss": 0.4602,
+      "step": 302
+    },
+    {
+      "epoch": 0.2814677194612169,
+      "grad_norm": 0.3009900967986436,
+      "learning_rate": 4.6749226006191954e-05,
+      "loss": 0.4604,
+      "step": 303
+    },
+    {
+      "epoch": 0.2823966558290757,
+      "grad_norm": 0.3103354270645873,
+      "learning_rate": 4.690402476780186e-05,
+      "loss": 0.4737,
+      "step": 304
+    },
+    {
+      "epoch": 0.2833255921969345,
+      "grad_norm": 0.30614215441269343,
+      "learning_rate": 4.705882352941177e-05,
+      "loss": 0.4707,
+      "step": 305
+    },
+    {
+      "epoch": 0.28425452856479333,
+      "grad_norm": 0.31908370093108424,
+      "learning_rate": 4.7213622291021674e-05,
+      "loss": 0.4473,
+      "step": 306
+    },
+    {
+      "epoch": 0.2851834649326521,
+      "grad_norm": 0.2959072198444788,
+      "learning_rate": 4.736842105263158e-05,
+      "loss": 0.4579,
+      "step": 307
+    },
+    {
+      "epoch": 0.2861124013005109,
+      "grad_norm": 0.3057457489708315,
+      "learning_rate": 4.752321981424149e-05,
+      "loss": 0.4199,
+      "step": 308
+    },
+    {
+      "epoch": 0.2870413376683697,
+      "grad_norm": 0.30818993057548966,
+      "learning_rate": 4.7678018575851394e-05,
+      "loss": 0.4627,
+      "step": 309
+    },
+    {
+      "epoch": 0.2879702740362285,
+      "grad_norm": 0.33780222748958855,
+      "learning_rate": 4.783281733746131e-05,
+      "loss": 0.4955,
+      "step": 310
+    },
+    {
+      "epoch": 0.28889921040408734,
+      "grad_norm": 0.3538296738454953,
+      "learning_rate": 4.798761609907121e-05,
+      "loss": 0.4528,
+      "step": 311
+    },
+    {
+      "epoch": 0.2898281467719461,
+      "grad_norm": 0.35013935319870626,
+      "learning_rate": 4.8142414860681115e-05,
+      "loss": 0.4896,
+      "step": 312
+    },
+    {
+      "epoch": 0.2907570831398049,
+      "grad_norm": 0.3919444872765707,
+      "learning_rate": 4.829721362229103e-05,
+      "loss": 0.434,
+      "step": 313
+    },
+    {
+      "epoch": 0.2916860195076637,
+      "grad_norm": 0.2873775450384677,
+      "learning_rate": 4.845201238390093e-05,
+      "loss": 0.43,
+      "step": 314
+    },
+    {
+      "epoch": 0.29261495587552255,
+      "grad_norm": 0.43278899104997853,
+      "learning_rate": 4.860681114551084e-05,
+      "loss": 0.459,
+      "step": 315
+    },
+    {
+      "epoch": 0.29354389224338134,
+      "grad_norm": 0.29762624151628303,
+      "learning_rate": 4.876160990712075e-05,
+      "loss": 0.4381,
+      "step": 316
+    },
+    {
+      "epoch": 0.29447282861124013,
+      "grad_norm": 0.38552936527851245,
+      "learning_rate": 4.891640866873065e-05,
+      "loss": 0.4488,
+      "step": 317
+    },
+    {
+      "epoch": 0.2954017649790989,
+      "grad_norm": 0.3262364914608491,
+      "learning_rate": 4.907120743034056e-05,
+      "loss": 0.4745,
+      "step": 318
+    },
+    {
+      "epoch": 0.2963307013469577,
+      "grad_norm": 0.3368476064425339,
+      "learning_rate": 4.922600619195047e-05,
+      "loss": 0.4588,
+      "step": 319
+    },
+    {
+      "epoch": 0.29725963771481656,
+      "grad_norm": 0.3312500891089113,
+      "learning_rate": 4.9380804953560375e-05,
+      "loss": 0.4794,
+      "step": 320
+    },
+    {
+      "epoch": 0.29818857408267535,
+      "grad_norm": 0.3390444459850631,
+      "learning_rate": 4.953560371517028e-05,
+      "loss": 0.4604,
+      "step": 321
+    },
+    {
+      "epoch": 0.29911751045053414,
+      "grad_norm": 0.3387119384494132,
+      "learning_rate": 4.969040247678019e-05,
+      "loss": 0.4602,
+      "step": 322
+    },
+    {
+      "epoch": 0.30004644681839293,
+      "grad_norm": 0.37541797983485836,
+      "learning_rate": 4.9845201238390096e-05,
+      "loss": 0.4857,
+      "step": 323
+    },
+    {
+      "epoch": 0.3009753831862517,
+      "grad_norm": 0.36204268391898264,
+      "learning_rate": 5e-05,
+      "loss": 0.4487,
+      "step": 324
+    },
+    {
+      "epoch": 0.30190431955411057,
+      "grad_norm": 0.5069674260137537,
+      "learning_rate": 4.998278829604131e-05,
+      "loss": 0.4795,
+      "step": 325
+    },
+    {
+      "epoch": 0.30283325592196936,
+      "grad_norm": 0.31607540585629734,
+      "learning_rate": 4.996557659208262e-05,
+      "loss": 0.459,
+      "step": 326
+    },
+    {
+      "epoch": 0.30376219228982815,
+      "grad_norm": 0.5375774496764807,
+      "learning_rate": 4.9948364888123924e-05,
+      "loss": 0.465,
+      "step": 327
+    },
+    {
+      "epoch": 0.30469112865768694,
+      "grad_norm": 0.34299271466845627,
+      "learning_rate": 4.9931153184165236e-05,
+      "loss": 0.4693,
+      "step": 328
+    },
+    {
+      "epoch": 0.30562006502554573,
+      "grad_norm": 0.4143539281353358,
+      "learning_rate": 4.991394148020654e-05,
+      "loss": 0.4606,
+      "step": 329
+    },
+    {
+      "epoch": 0.3065490013934046,
+      "grad_norm": 0.4855004584252198,
+      "learning_rate": 4.989672977624785e-05,
+      "loss": 0.4783,
+      "step": 330
+    },
+    {
+      "epoch": 0.30747793776126336,
+      "grad_norm": 0.3423755850607668,
+      "learning_rate": 4.987951807228916e-05,
+      "loss": 0.4444,
+      "step": 331
+    },
+    {
+      "epoch": 0.30840687412912215,
+      "grad_norm": 0.3583846261043207,
+      "learning_rate": 4.986230636833047e-05,
+      "loss": 0.473,
+      "step": 332
+    },
+    {
+      "epoch": 0.30933581049698095,
+      "grad_norm": 0.3138663780309369,
+      "learning_rate": 4.9845094664371775e-05,
+      "loss": 0.4353,
+      "step": 333
+    },
+    {
+      "epoch": 0.31026474686483974,
+      "grad_norm": 0.3232503486528393,
+      "learning_rate": 4.9827882960413086e-05,
+      "loss": 0.4662,
+      "step": 334
+    },
+    {
+      "epoch": 0.3111936832326986,
+      "grad_norm": 0.33077321915284624,
+      "learning_rate": 4.981067125645439e-05,
+      "loss": 0.4389,
+      "step": 335
+    },
+    {
+      "epoch": 0.31212261960055737,
+      "grad_norm": 0.32940453207532444,
+      "learning_rate": 4.9793459552495696e-05,
+      "loss": 0.4549,
+      "step": 336
+    },
+    {
+      "epoch": 0.31305155596841616,
+      "grad_norm": 0.3489785623343884,
+      "learning_rate": 4.977624784853701e-05,
+      "loss": 0.4519,
+      "step": 337
+    },
+    {
+      "epoch": 0.31398049233627495,
+      "grad_norm": 0.3380056249650716,
+      "learning_rate": 4.975903614457831e-05,
+      "loss": 0.4501,
+      "step": 338
+    },
+    {
+      "epoch": 0.31490942870413374,
+      "grad_norm": 0.39621724081434734,
+      "learning_rate": 4.9741824440619625e-05,
+      "loss": 0.4886,
+      "step": 339
+    },
+    {
+      "epoch": 0.3158383650719926,
+      "grad_norm": 0.34090797197414546,
+      "learning_rate": 4.972461273666093e-05,
+      "loss": 0.4738,
+      "step": 340
+    },
+    {
+      "epoch": 0.3167673014398514,
+      "grad_norm": 0.35008163528496755,
+      "learning_rate": 4.970740103270224e-05,
+      "loss": 0.4533,
+      "step": 341
+    },
+    {
+      "epoch": 0.31769623780771017,
+      "grad_norm": 0.35189206816774354,
+      "learning_rate": 4.969018932874355e-05,
+      "loss": 0.4286,
+      "step": 342
+    },
+    {
+      "epoch": 0.31862517417556896,
+      "grad_norm": 0.3640074292494907,
+      "learning_rate": 4.967297762478486e-05,
+      "loss": 0.4412,
+      "step": 343
+    },
+    {
+      "epoch": 0.31955411054342775,
+      "grad_norm": 0.3532873296836122,
+      "learning_rate": 4.9655765920826164e-05,
+      "loss": 0.4686,
+      "step": 344
+    },
+    {
+      "epoch": 0.3204830469112866,
+      "grad_norm": 0.37025094985949847,
+      "learning_rate": 4.9638554216867475e-05,
+      "loss": 0.4338,
+      "step": 345
+    },
+    {
+      "epoch": 0.3214119832791454,
+      "grad_norm": 0.33462244732386653,
+      "learning_rate": 4.962134251290878e-05,
+      "loss": 0.4392,
+      "step": 346
+    },
+    {
+      "epoch": 0.3223409196470042,
+      "grad_norm": 0.3820555101933481,
+      "learning_rate": 4.960413080895009e-05,
+      "loss": 0.4898,
+      "step": 347
+    },
+    {
+      "epoch": 0.32326985601486297,
+      "grad_norm": 0.35860399216197664,
+      "learning_rate": 4.95869191049914e-05,
+      "loss": 0.4876,
+      "step": 348
+    },
+    {
+      "epoch": 0.32419879238272176,
+      "grad_norm": 0.3419630528737662,
+      "learning_rate": 4.95697074010327e-05,
+      "loss": 0.4698,
+      "step": 349
+    },
+    {
+      "epoch": 0.3251277287505806,
+      "grad_norm": 0.35528988609754913,
+      "learning_rate": 4.9552495697074014e-05,
+      "loss": 0.4275,
+      "step": 350
+    },
+    {
+      "epoch": 0.3260566651184394,
+      "grad_norm": 0.38379201396008805,
+      "learning_rate": 4.953528399311532e-05,
+      "loss": 0.4445,
+      "step": 351
+    },
+    {
+      "epoch": 0.3269856014862982,
+      "grad_norm": 0.3398302347977464,
+      "learning_rate": 4.951807228915663e-05,
+      "loss": 0.4554,
+      "step": 352
+    },
+    {
+      "epoch": 0.327914537854157,
+      "grad_norm": 0.3897934147080159,
+      "learning_rate": 4.9500860585197936e-05,
+      "loss": 0.4396,
+      "step": 353
+    },
+    {
+      "epoch": 0.3288434742220158,
+      "grad_norm": 0.3704282311872543,
+      "learning_rate": 4.948364888123925e-05,
+      "loss": 0.4476,
+      "step": 354
+    },
+    {
+      "epoch": 0.3297724105898746,
+      "grad_norm": 0.3792694514662743,
+      "learning_rate": 4.946643717728055e-05,
+      "loss": 0.4153,
+      "step": 355
+    },
+    {
+      "epoch": 0.3307013469577334,
+      "grad_norm": 0.3900734100570611,
+      "learning_rate": 4.944922547332186e-05,
+      "loss": 0.4347,
+      "step": 356
+    },
+    {
+      "epoch": 0.3316302833255922,
+      "grad_norm": 0.3409495630328528,
+      "learning_rate": 4.943201376936317e-05,
+      "loss": 0.4242,
+      "step": 357
+    },
+    {
+      "epoch": 0.332559219693451,
+      "grad_norm": 0.3661404637247587,
+      "learning_rate": 4.9414802065404474e-05,
+      "loss": 0.4426,
+      "step": 358
+    },
+    {
+      "epoch": 0.3334881560613098,
+      "grad_norm": 0.4272666657937548,
+      "learning_rate": 4.9397590361445786e-05,
+      "loss": 0.4468,
+      "step": 359
+    },
+    {
+      "epoch": 0.3344170924291686,
+      "grad_norm": 0.2754985787418611,
+      "learning_rate": 4.938037865748709e-05,
+      "loss": 0.4187,
+      "step": 360
+    },
+    {
+      "epoch": 0.3353460287970274,
+      "grad_norm": 0.38303057883878117,
+      "learning_rate": 4.93631669535284e-05,
+      "loss": 0.4594,
+      "step": 361
+    },
+    {
+      "epoch": 0.3362749651648862,
+      "grad_norm": 0.3015558469101507,
+      "learning_rate": 4.934595524956971e-05,
+      "loss": 0.4337,
+      "step": 362
+    },
+    {
+      "epoch": 0.337203901532745,
+      "grad_norm": 0.36850129972653306,
+      "learning_rate": 4.932874354561102e-05,
+      "loss": 0.4666,
+      "step": 363
+    },
+    {
+      "epoch": 0.33813283790060383,
+      "grad_norm": 0.2884585481636936,
+      "learning_rate": 4.9311531841652325e-05,
+      "loss": 0.4437,
+      "step": 364
+    },
+    {
+      "epoch": 0.3390617742684626,
+      "grad_norm": 0.3163751089768913,
+      "learning_rate": 4.929432013769364e-05,
+      "loss": 0.4591,
+      "step": 365
+    },
+    {
+      "epoch": 0.3399907106363214,
+      "grad_norm": 0.28834325339704836,
+      "learning_rate": 4.927710843373494e-05,
+      "loss": 0.4339,
+      "step": 366
+    },
+    {
+      "epoch": 0.3409196470041802,
+      "grad_norm": 0.3089186351649892,
+      "learning_rate": 4.9259896729776253e-05,
+      "loss": 0.4413,
+      "step": 367
+    },
+    {
+      "epoch": 0.341848583372039,
+      "grad_norm": 0.35241634143696166,
+      "learning_rate": 4.924268502581756e-05,
+      "loss": 0.4571,
+      "step": 368
+    },
+    {
+      "epoch": 0.34277751973989784,
+      "grad_norm": 0.3101976948447532,
+      "learning_rate": 4.922547332185887e-05,
+      "loss": 0.4558,
+      "step": 369
+    },
+    {
+      "epoch": 0.34370645610775663,
+      "grad_norm": 0.35319592477754186,
+      "learning_rate": 4.9208261617900175e-05,
+      "loss": 0.4429,
+      "step": 370
+    },
+    {
+      "epoch": 0.3446353924756154,
+      "grad_norm": 0.3322284047796149,
+      "learning_rate": 4.919104991394149e-05,
+      "loss": 0.4524,
+      "step": 371
+    },
+    {
+      "epoch": 0.3455643288434742,
+      "grad_norm": 0.37262088448309033,
+      "learning_rate": 4.917383820998279e-05,
+      "loss": 0.4325,
+      "step": 372
+    },
+    {
+      "epoch": 0.346493265211333,
+      "grad_norm": 0.336028619967502,
+      "learning_rate": 4.9156626506024104e-05,
+      "loss": 0.4488,
+      "step": 373
+    },
+    {
+      "epoch": 0.34742220157919185,
+      "grad_norm": 0.37466782782498514,
+      "learning_rate": 4.913941480206541e-05,
+      "loss": 0.458,
+      "step": 374
+    },
+    {
+      "epoch": 0.34835113794705064,
+      "grad_norm": 0.3083157732259951,
+      "learning_rate": 4.9122203098106714e-05,
+      "loss": 0.4522,
+      "step": 375
+    },
+    {
+      "epoch": 0.34928007431490943,
+      "grad_norm": 0.3696409621615052,
+      "learning_rate": 4.9104991394148026e-05,
+      "loss": 0.4769,
+      "step": 376
+    },
+    {
+      "epoch": 0.3502090106827682,
+      "grad_norm": 0.37589325200592977,
+      "learning_rate": 4.908777969018933e-05,
+      "loss": 0.4535,
+      "step": 377
+    },
+    {
+      "epoch": 0.351137947050627,
+      "grad_norm": 0.290205612168557,
+      "learning_rate": 4.9070567986230636e-05,
+      "loss": 0.4669,
+      "step": 378
+    },
+    {
+      "epoch": 0.35206688341848585,
+      "grad_norm": 0.37740647897617646,
+      "learning_rate": 4.905335628227195e-05,
+      "loss": 0.4384,
+      "step": 379
+    },
+    {
+      "epoch": 0.35299581978634464,
+      "grad_norm": 0.28374091712444127,
+      "learning_rate": 4.903614457831325e-05,
+      "loss": 0.4437,
+      "step": 380
+    },
+    {
+      "epoch": 0.35392475615420343,
+      "grad_norm": 0.3555291712519551,
+      "learning_rate": 4.9018932874354564e-05,
+      "loss": 0.4437,
+      "step": 381
+    },
+    {
+      "epoch": 0.3548536925220622,
+      "grad_norm": 0.3021908819889071,
+      "learning_rate": 4.900172117039587e-05,
+      "loss": 0.4529,
+      "step": 382
+    },
+    {
+      "epoch": 0.355782628889921,
+      "grad_norm": 0.28070022204204287,
+      "learning_rate": 4.898450946643718e-05,
+      "loss": 0.4445,
+      "step": 383
+    },
+    {
+      "epoch": 0.35671156525777986,
+      "grad_norm": 0.34406702126509014,
+      "learning_rate": 4.8967297762478486e-05,
+      "loss": 0.4532,
+      "step": 384
+    },
+    {
+      "epoch": 0.35764050162563865,
+      "grad_norm": 0.2888201625795591,
+      "learning_rate": 4.89500860585198e-05,
+      "loss": 0.4636,
+      "step": 385
+    },
+    {
+      "epoch": 0.35856943799349744,
+      "grad_norm": 0.3598719298462701,
+      "learning_rate": 4.89328743545611e-05,
+      "loss": 0.4402,
+      "step": 386
+    },
+    {
+      "epoch": 0.35949837436135623,
+      "grad_norm": 0.2895073326465502,
+      "learning_rate": 4.8915662650602415e-05,
+      "loss": 0.4625,
+      "step": 387
+    },
+    {
+      "epoch": 0.360427310729215,
+      "grad_norm": 0.323679285493869,
+      "learning_rate": 4.889845094664372e-05,
+      "loss": 0.4253,
+      "step": 388
+    },
+    {
+      "epoch": 0.36135624709707387,
+      "grad_norm": 0.32852184199735984,
+      "learning_rate": 4.888123924268503e-05,
+      "loss": 0.4404,
+      "step": 389
+    },
+    {
+      "epoch": 0.36228518346493266,
+      "grad_norm": 0.3825310829770502,
+      "learning_rate": 4.8864027538726336e-05,
+      "loss": 0.4324,
+      "step": 390
+    },
+    {
+      "epoch": 0.36321411983279145,
+      "grad_norm": 0.3147280199593259,
+      "learning_rate": 4.884681583476765e-05,
+      "loss": 0.4576,
+      "step": 391
+    },
+    {
+      "epoch": 0.36414305620065024,
+      "grad_norm": 0.33189632998575647,
+      "learning_rate": 4.882960413080895e-05,
+      "loss": 0.4508,
+      "step": 392
+    },
+    {
+      "epoch": 0.36507199256850903,
+      "grad_norm": 0.31161211897471663,
+      "learning_rate": 4.8812392426850265e-05,
+      "loss": 0.433,
+      "step": 393
+    },
+    {
+      "epoch": 0.3660009289363679,
+      "grad_norm": 0.29323440533256345,
+      "learning_rate": 4.879518072289157e-05,
+      "loss": 0.4426,
+      "step": 394
+    },
+    {
+      "epoch": 0.36692986530422667,
+      "grad_norm": 0.3044769147297849,
+      "learning_rate": 4.877796901893288e-05,
+      "loss": 0.4383,
+      "step": 395
+    },
+    {
+      "epoch": 0.36785880167208546,
+      "grad_norm": 0.33105623962975367,
+      "learning_rate": 4.876075731497419e-05,
+      "loss": 0.4511,
+      "step": 396
+    },
+    {
+      "epoch": 0.36878773803994425,
+      "grad_norm": 0.2926145124782726,
+      "learning_rate": 4.87435456110155e-05,
+      "loss": 0.4494,
+      "step": 397
+    },
+    {
+      "epoch": 0.3697166744078031,
+      "grad_norm": 0.2785219519909291,
+      "learning_rate": 4.8726333907056804e-05,
+      "loss": 0.4193,
+      "step": 398
+    },
+    {
+      "epoch": 0.3706456107756619,
+      "grad_norm": 0.3190152822347888,
+      "learning_rate": 4.870912220309811e-05,
+      "loss": 0.4348,
+      "step": 399
+    },
+    {
+      "epoch": 0.3715745471435207,
+      "grad_norm": 0.2944642128721377,
+      "learning_rate": 4.8691910499139414e-05,
+      "loss": 0.4413,
+      "step": 400
+    },
+    {
+      "epoch": 0.37250348351137946,
+      "grad_norm": 0.2874818565004639,
+      "learning_rate": 4.8674698795180725e-05,
+      "loss": 0.4379,
+      "step": 401
+    },
+    {
+      "epoch": 0.37343241987923825,
+      "grad_norm": 0.27777508008099533,
+      "learning_rate": 4.865748709122203e-05,
+      "loss": 0.4241,
+      "step": 402
+    },
+    {
+      "epoch": 0.3743613562470971,
+      "grad_norm": 0.32978399945466425,
+      "learning_rate": 4.864027538726334e-05,
+      "loss": 0.447,
+      "step": 403
+    },
+    {
+      "epoch": 0.3752902926149559,
+      "grad_norm": 0.3094967262792066,
+      "learning_rate": 4.862306368330465e-05,
+      "loss": 0.4466,
+      "step": 404
+    },
+    {
+      "epoch": 0.3762192289828147,
+      "grad_norm": 0.29574208650459116,
+      "learning_rate": 4.860585197934596e-05,
+      "loss": 0.436,
+      "step": 405
+    },
+    {
+      "epoch": 0.37714816535067347,
+      "grad_norm": 0.288270087489997,
+      "learning_rate": 4.8588640275387264e-05,
+      "loss": 0.4441,
+      "step": 406
+    },
+    {
+      "epoch": 0.37807710171853226,
+      "grad_norm": 0.30162172795880676,
+      "learning_rate": 4.8571428571428576e-05,
+      "loss": 0.4145,
+      "step": 407
+    },
+    {
+      "epoch": 0.3790060380863911,
+      "grad_norm": 0.2618604647255097,
+      "learning_rate": 4.855421686746988e-05,
+      "loss": 0.4162,
+      "step": 408
+    },
+    {
+      "epoch": 0.3799349744542499,
+      "grad_norm": 0.30851370446236615,
+      "learning_rate": 4.853700516351119e-05,
+      "loss": 0.4224,
+      "step": 409
+    },
+    {
+      "epoch": 0.3808639108221087,
+      "grad_norm": 0.3006905343379606,
+      "learning_rate": 4.85197934595525e-05,
+      "loss": 0.4879,
+      "step": 410
+    },
+    {
+      "epoch": 0.3817928471899675,
+      "grad_norm": 0.28307585696243226,
+      "learning_rate": 4.850258175559381e-05,
+      "loss": 0.4068,
+      "step": 411
+    },
+    {
+      "epoch": 0.38272178355782627,
+      "grad_norm": 0.26102123895109014,
+      "learning_rate": 4.8485370051635114e-05,
+      "loss": 0.428,
+      "step": 412
+    },
+    {
+      "epoch": 0.3836507199256851,
+      "grad_norm": 0.26532852818511954,
+      "learning_rate": 4.8468158347676426e-05,
+      "loss": 0.4356,
+      "step": 413
+    },
+    {
+      "epoch": 0.3845796562935439,
+      "grad_norm": 0.27713412833380047,
+      "learning_rate": 4.845094664371773e-05,
+      "loss": 0.4166,
+      "step": 414
+    },
+    {
+      "epoch": 0.3855085926614027,
+      "grad_norm": 0.2721011127848934,
+      "learning_rate": 4.843373493975904e-05,
+      "loss": 0.4588,
+      "step": 415
+    },
+    {
+      "epoch": 0.3864375290292615,
+      "grad_norm": 0.270987247231762,
+      "learning_rate": 4.841652323580035e-05,
+      "loss": 0.4418,
+      "step": 416
+    },
+    {
+      "epoch": 0.3873664653971203,
+      "grad_norm": 0.2815915571880338,
+      "learning_rate": 4.839931153184166e-05,
+      "loss": 0.438,
+      "step": 417
+    },
+    {
+      "epoch": 0.3882954017649791,
+      "grad_norm": 0.2900554851874775,
+      "learning_rate": 4.8382099827882965e-05,
+      "loss": 0.4495,
+      "step": 418
+    },
+    {
+      "epoch": 0.3892243381328379,
+      "grad_norm": 0.3334842549517787,
+      "learning_rate": 4.836488812392428e-05,
+      "loss": 0.4646,
+      "step": 419
+    },
+    {
+      "epoch": 0.3901532745006967,
+      "grad_norm": 0.24007212579976095,
+      "learning_rate": 4.8347676419965575e-05,
+      "loss": 0.4238,
+      "step": 420
+    },
+    {
+      "epoch": 0.3910822108685555,
+      "grad_norm": 0.3077325041506196,
+      "learning_rate": 4.833046471600689e-05,
+      "loss": 0.4447,
+      "step": 421
+    },
+    {
+      "epoch": 0.3920111472364143,
+      "grad_norm": 0.2863122225447303,
+      "learning_rate": 4.831325301204819e-05,
+      "loss": 0.4365,
+      "step": 422
+    },
+    {
+      "epoch": 0.3929400836042731,
+      "grad_norm": 0.30580235165952646,
+      "learning_rate": 4.8296041308089504e-05,
+      "loss": 0.4453,
+      "step": 423
+    },
+    {
+      "epoch": 0.3938690199721319,
+      "grad_norm": 0.26472608629055344,
+      "learning_rate": 4.827882960413081e-05,
+      "loss": 0.445,
+      "step": 424
+    },
+    {
+      "epoch": 0.3947979563399907,
+      "grad_norm": 0.30850387581287186,
+      "learning_rate": 4.826161790017212e-05,
+      "loss": 0.4459,
+      "step": 425
+    },
+    {
+      "epoch": 0.3957268927078495,
+      "grad_norm": 0.32129564773476693,
+      "learning_rate": 4.8244406196213425e-05,
+      "loss": 0.4302,
+      "step": 426
+    },
+    {
+      "epoch": 0.3966558290757083,
+      "grad_norm": 0.2856821353622809,
+      "learning_rate": 4.822719449225474e-05,
+      "loss": 0.4366,
+      "step": 427
+    },
+    {
+      "epoch": 0.39758476544356713,
+      "grad_norm": 0.27875816280973154,
+      "learning_rate": 4.820998278829604e-05,
+      "loss": 0.4177,
+      "step": 428
+    },
+    {
+      "epoch": 0.3985137018114259,
+      "grad_norm": 0.375391127760771,
+      "learning_rate": 4.8192771084337354e-05,
+      "loss": 0.4549,
+      "step": 429
+    },
+    {
+      "epoch": 0.3994426381792847,
+      "grad_norm": 0.2919547832552509,
+      "learning_rate": 4.817555938037866e-05,
+      "loss": 0.4377,
+      "step": 430
+    },
+    {
+      "epoch": 0.4003715745471435,
+      "grad_norm": 0.35769067629140655,
+      "learning_rate": 4.815834767641997e-05,
+      "loss": 0.4106,
+      "step": 431
+    },
+    {
+      "epoch": 0.4013005109150023,
+      "grad_norm": 0.3169078733636773,
+      "learning_rate": 4.8141135972461276e-05,
+      "loss": 0.4344,
+      "step": 432
+    },
+    {
+      "epoch": 0.40222944728286114,
+      "grad_norm": 0.38647441373213326,
+      "learning_rate": 4.812392426850259e-05,
+      "loss": 0.4662,
+      "step": 433
+    },
+    {
+      "epoch": 0.40315838365071993,
+      "grad_norm": 0.3093808502182461,
+      "learning_rate": 4.810671256454389e-05,
+      "loss": 0.4268,
+      "step": 434
+    },
+    {
+      "epoch": 0.4040873200185787,
+      "grad_norm": 0.3428360410015756,
+      "learning_rate": 4.8089500860585204e-05,
+      "loss": 0.4248,
+      "step": 435
+    },
+    {
+      "epoch": 0.4050162563864375,
+      "grad_norm": 0.39226349195208254,
+      "learning_rate": 4.807228915662651e-05,
+      "loss": 0.4828,
+      "step": 436
+    },
+    {
+      "epoch": 0.40594519275429636,
+      "grad_norm": 0.3124911712167451,
+      "learning_rate": 4.805507745266782e-05,
+      "loss": 0.42,
+      "step": 437
+    },
+    {
+      "epoch": 0.40687412912215515,
+      "grad_norm": 0.3325357418649828,
+      "learning_rate": 4.8037865748709126e-05,
+      "loss": 0.4183,
+      "step": 438
+    },
+    {
+      "epoch": 0.40780306549001394,
+      "grad_norm": 0.2967793995799997,
+      "learning_rate": 4.802065404475044e-05,
+      "loss": 0.4522,
+      "step": 439
+    },
+    {
+      "epoch": 0.40873200185787273,
+      "grad_norm": 0.3542440897290239,
+      "learning_rate": 4.800344234079174e-05,
+      "loss": 0.4463,
+      "step": 440
+    },
+    {
+      "epoch": 0.4096609382257315,
+      "grad_norm": 0.307305266388483,
+      "learning_rate": 4.798623063683305e-05,
+      "loss": 0.4444,
+      "step": 441
+    },
+    {
+      "epoch": 0.41058987459359036,
+      "grad_norm": 0.3009613722666484,
+      "learning_rate": 4.796901893287435e-05,
+      "loss": 0.42,
+      "step": 442
+    },
+    {
+      "epoch": 0.41151881096144916,
+      "grad_norm": 0.2942089986687925,
+      "learning_rate": 4.7951807228915665e-05,
+      "loss": 0.4408,
+      "step": 443
+    },
+    {
+      "epoch": 0.41244774732930795,
+      "grad_norm": 0.25994860698808103,
+      "learning_rate": 4.793459552495697e-05,
+      "loss": 0.422,
+      "step": 444
+    },
+    {
+      "epoch": 0.41337668369716674,
+      "grad_norm": 0.30165826806753704,
+      "learning_rate": 4.791738382099828e-05,
+      "loss": 0.4433,
+      "step": 445
+    },
+    {
+      "epoch": 0.4143056200650255,
+      "grad_norm": 0.2591476595221208,
+      "learning_rate": 4.7900172117039587e-05,
+      "loss": 0.4035,
+      "step": 446
+    },
+    {
+      "epoch": 0.41523455643288437,
+      "grad_norm": 0.31443538659749376,
+      "learning_rate": 4.78829604130809e-05,
+      "loss": 0.4451,
+      "step": 447
+    },
+    {
+      "epoch": 0.41616349280074316,
+      "grad_norm": 0.2845675404201316,
+      "learning_rate": 4.78657487091222e-05,
+      "loss": 0.4499,
+      "step": 448
+    },
+    {
+      "epoch": 0.41709242916860195,
+      "grad_norm": 0.3082724543796442,
+      "learning_rate": 4.7848537005163515e-05,
+      "loss": 0.4567,
+      "step": 449
+    },
+    {
+      "epoch": 0.41802136553646074,
+      "grad_norm": 0.2886233884056253,
+      "learning_rate": 4.783132530120482e-05,
+      "loss": 0.4351,
+      "step": 450
+    },
+    {
+      "epoch": 0.41895030190431953,
+      "grad_norm": 0.32824185907530123,
+      "learning_rate": 4.781411359724613e-05,
+      "loss": 0.4343,
+      "step": 451
+    },
+    {
+      "epoch": 0.4198792382721784,
+      "grad_norm": 0.3045951917878721,
+      "learning_rate": 4.779690189328744e-05,
+      "loss": 0.4233,
+      "step": 452
+    },
+    {
+      "epoch": 0.42080817464003717,
+      "grad_norm": 0.30065030364940504,
+      "learning_rate": 4.777969018932875e-05,
+      "loss": 0.4449,
+      "step": 453
+    },
+    {
+      "epoch": 0.42173711100789596,
+      "grad_norm": 0.28194914295465456,
+      "learning_rate": 4.7762478485370054e-05,
+      "loss": 0.4216,
+      "step": 454
+    },
+    {
+      "epoch": 0.42266604737575475,
+      "grad_norm": 0.358326292324705,
+      "learning_rate": 4.7745266781411366e-05,
+      "loss": 0.4545,
+      "step": 455
+    },
+    {
+      "epoch": 0.42359498374361354,
+      "grad_norm": 0.32509662830836533,
+      "learning_rate": 4.772805507745267e-05,
+      "loss": 0.4173,
+      "step": 456
+    },
+    {
+      "epoch": 0.4245239201114724,
+      "grad_norm": 0.33738335206220016,
+      "learning_rate": 4.771084337349398e-05,
+      "loss": 0.4433,
+      "step": 457
+    },
+    {
+      "epoch": 0.4254528564793312,
+      "grad_norm": 0.3607433985711699,
+      "learning_rate": 4.769363166953529e-05,
+      "loss": 0.4317,
+      "step": 458
+    },
+    {
+      "epoch": 0.42638179284718997,
+      "grad_norm": 0.32470986956497677,
+      "learning_rate": 4.76764199655766e-05,
+      "loss": 0.4278,
+      "step": 459
+    },
+    {
+      "epoch": 0.42731072921504876,
+      "grad_norm": 0.31596561843145204,
+      "learning_rate": 4.7659208261617904e-05,
+      "loss": 0.4507,
+      "step": 460
+    },
+    {
+      "epoch": 0.42823966558290755,
+      "grad_norm": 0.3145646642534761,
+      "learning_rate": 4.7641996557659216e-05,
+      "loss": 0.4353,
+      "step": 461
+    },
+    {
+      "epoch": 0.4291686019507664,
+      "grad_norm": 0.28777196903769586,
+      "learning_rate": 4.7624784853700514e-05,
+      "loss": 0.4606,
+      "step": 462
+    },
+    {
+      "epoch": 0.4300975383186252,
+      "grad_norm": 0.34643745463301656,
+      "learning_rate": 4.7607573149741826e-05,
+      "loss": 0.4264,
+      "step": 463
+    },
+    {
+      "epoch": 0.431026474686484,
+      "grad_norm": 0.3022418092258536,
+      "learning_rate": 4.759036144578313e-05,
+      "loss": 0.4046,
+      "step": 464
+    },
+    {
+      "epoch": 0.43195541105434276,
+      "grad_norm": 0.27983510155632235,
+      "learning_rate": 4.757314974182444e-05,
+      "loss": 0.4122,
+      "step": 465
+    },
+    {
+      "epoch": 0.43288434742220155,
+      "grad_norm": 0.28454933014088335,
+      "learning_rate": 4.755593803786575e-05,
+      "loss": 0.4558,
+      "step": 466
+    },
+    {
+      "epoch": 0.4338132837900604,
+      "grad_norm": 0.32254214458961294,
+      "learning_rate": 4.753872633390706e-05,
+      "loss": 0.4272,
+      "step": 467
+    },
+    {
+      "epoch": 0.4347422201579192,
+      "grad_norm": 0.2613178418721715,
+      "learning_rate": 4.7521514629948365e-05,
+      "loss": 0.4418,
+      "step": 468
+    },
+    {
+      "epoch": 0.435671156525778,
+      "grad_norm": 0.3056014331601247,
+      "learning_rate": 4.7504302925989676e-05,
+      "loss": 0.421,
+      "step": 469
+    },
+    {
+      "epoch": 0.43660009289363677,
+      "grad_norm": 0.2695626668562775,
+      "learning_rate": 4.748709122203098e-05,
+      "loss": 0.4321,
+      "step": 470
+    },
+    {
+      "epoch": 0.43752902926149556,
+      "grad_norm": 0.27446080453869376,
+      "learning_rate": 4.746987951807229e-05,
+      "loss": 0.4375,
+      "step": 471
+    },
+    {
+      "epoch": 0.4384579656293544,
+      "grad_norm": 0.29987194356039903,
+      "learning_rate": 4.74526678141136e-05,
+      "loss": 0.4436,
+      "step": 472
+    },
+    {
+      "epoch": 0.4393869019972132,
+      "grad_norm": 0.2973778889764395,
+      "learning_rate": 4.743545611015491e-05,
+      "loss": 0.4508,
+      "step": 473
+    },
+    {
+      "epoch": 0.440315838365072,
+      "grad_norm": 0.28858718688747587,
+      "learning_rate": 4.7418244406196215e-05,
+      "loss": 0.4088,
+      "step": 474
+    },
+    {
+      "epoch": 0.4412447747329308,
+      "grad_norm": 0.2776236267106209,
+      "learning_rate": 4.740103270223753e-05,
+      "loss": 0.4472,
+      "step": 475
+    },
+    {
+      "epoch": 0.44217371110078957,
+      "grad_norm": 0.2673174324156048,
+      "learning_rate": 4.738382099827883e-05,
+      "loss": 0.4318,
+      "step": 476
+    },
+    {
+      "epoch": 0.4431026474686484,
+      "grad_norm": 0.2789675691022897,
+      "learning_rate": 4.7366609294320144e-05,
+      "loss": 0.4256,
+      "step": 477
+    },
+    {
+      "epoch": 0.4440315838365072,
+      "grad_norm": 0.3242561622863543,
+      "learning_rate": 4.734939759036145e-05,
+      "loss": 0.4314,
+      "step": 478
+    },
+    {
+      "epoch": 0.444960520204366,
+      "grad_norm": 0.26245305270546254,
+      "learning_rate": 4.733218588640276e-05,
+      "loss": 0.432,
+      "step": 479
+    },
+    {
+      "epoch": 0.4458894565722248,
+      "grad_norm": 0.2908962068774102,
+      "learning_rate": 4.7314974182444065e-05,
+      "loss": 0.4262,
+      "step": 480
+    },
+    {
+      "epoch": 0.44681839294008363,
+      "grad_norm": 0.27171256702754587,
+      "learning_rate": 4.729776247848538e-05,
+      "loss": 0.4352,
+      "step": 481
+    },
+    {
+      "epoch": 0.4477473293079424,
+      "grad_norm": 0.3640009107002527,
+      "learning_rate": 4.728055077452668e-05,
+      "loss": 0.4337,
+      "step": 482
+    },
+    {
+      "epoch": 0.4486762656758012,
+      "grad_norm": 0.3035811404578588,
+      "learning_rate": 4.7263339070567994e-05,
+      "loss": 0.453,
+      "step": 483
+    },
+    {
+      "epoch": 0.44960520204366,
+      "grad_norm": 0.3635508538266957,
+      "learning_rate": 4.724612736660929e-05,
+      "loss": 0.4296,
+      "step": 484
+    },
+    {
+      "epoch": 0.4505341384115188,
+      "grad_norm": 0.3170039262392314,
+      "learning_rate": 4.7228915662650604e-05,
+      "loss": 0.4207,
+      "step": 485
+    },
+    {
+      "epoch": 0.45146307477937764,
+      "grad_norm": 0.31774438891510837,
+      "learning_rate": 4.721170395869191e-05,
+      "loss": 0.4406,
+      "step": 486
+    },
+    {
+      "epoch": 0.45239201114723643,
+      "grad_norm": 0.34019517987590486,
+      "learning_rate": 4.719449225473322e-05,
+      "loss": 0.4902,
+      "step": 487
+    },
+    {
+      "epoch": 0.4533209475150952,
+      "grad_norm": 0.2793264697380186,
+      "learning_rate": 4.7177280550774526e-05,
+      "loss": 0.4454,
+      "step": 488
+    },
+    {
+      "epoch": 0.454249883882954,
+      "grad_norm": 0.28144894598721126,
+      "learning_rate": 4.716006884681584e-05,
+      "loss": 0.4187,
+      "step": 489
+    },
+    {
+      "epoch": 0.4551788202508128,
+      "grad_norm": 0.30680476206969465,
+      "learning_rate": 4.714285714285714e-05,
+      "loss": 0.4458,
+      "step": 490
+    },
+    {
+      "epoch": 0.45610775661867164,
+      "grad_norm": 0.2761255587228729,
+      "learning_rate": 4.7125645438898454e-05,
+      "loss": 0.4141,
+      "step": 491
+    },
+    {
+      "epoch": 0.45703669298653044,
+      "grad_norm": 0.2747933627980402,
+      "learning_rate": 4.710843373493976e-05,
+      "loss": 0.4558,
+      "step": 492
+    },
+    {
+      "epoch": 0.4579656293543892,
+      "grad_norm": 0.2914395069490274,
+      "learning_rate": 4.709122203098107e-05,
+      "loss": 0.4353,
+      "step": 493
+    },
+    {
+      "epoch": 0.458894565722248,
+      "grad_norm": 0.26839160122030176,
+      "learning_rate": 4.7074010327022376e-05,
+      "loss": 0.4131,
+      "step": 494
+    },
+    {
+      "epoch": 0.4598235020901068,
+      "grad_norm": 0.2822993839047011,
+      "learning_rate": 4.705679862306369e-05,
+      "loss": 0.4265,
+      "step": 495
+    },
+    {
+      "epoch": 0.46075243845796565,
+      "grad_norm": 0.2657294003991585,
+      "learning_rate": 4.703958691910499e-05,
+      "loss": 0.4408,
+      "step": 496
+    },
+    {
+      "epoch": 0.46168137482582444,
+      "grad_norm": 0.29482740291064663,
+      "learning_rate": 4.7022375215146305e-05,
+      "loss": 0.4368,
+      "step": 497
+    },
+    {
+      "epoch": 0.46261031119368323,
+      "grad_norm": 0.26269631652127934,
+      "learning_rate": 4.700516351118761e-05,
+      "loss": 0.4313,
+      "step": 498
+    },
+    {
+      "epoch": 0.463539247561542,
+      "grad_norm": 0.33705615306936615,
+      "learning_rate": 4.698795180722892e-05,
+      "loss": 0.4529,
+      "step": 499
+    },
+    {
+      "epoch": 0.4644681839294008,
+      "grad_norm": 0.24499796930149098,
+      "learning_rate": 4.6970740103270227e-05,
+      "loss": 0.4385,
+      "step": 500
+    },
+    {
+      "epoch": 0.46539712029725966,
+      "grad_norm": 0.30948537181680397,
+      "learning_rate": 4.695352839931154e-05,
+      "loss": 0.4449,
+      "step": 501
+    },
+    {
+      "epoch": 0.46632605666511845,
+      "grad_norm": 0.29180609911144206,
+      "learning_rate": 4.6936316695352843e-05,
+      "loss": 0.4352,
+      "step": 502
+    },
+    {
+      "epoch": 0.46725499303297724,
+      "grad_norm": 0.2891708931616031,
+      "learning_rate": 4.6919104991394155e-05,
+      "loss": 0.439,
+      "step": 503
+    },
+    {
+      "epoch": 0.46818392940083603,
+      "grad_norm": 0.2686671320742356,
+      "learning_rate": 4.690189328743546e-05,
+      "loss": 0.4306,
+      "step": 504
+    },
+    {
+      "epoch": 0.4691128657686948,
+      "grad_norm": 0.30780838224951595,
+      "learning_rate": 4.6884681583476765e-05,
+      "loss": 0.431,
+      "step": 505
+    },
+    {
+      "epoch": 0.47004180213655367,
+      "grad_norm": 0.25568604631287883,
+      "learning_rate": 4.686746987951807e-05,
+      "loss": 0.458,
+      "step": 506
+    },
+    {
+      "epoch": 0.47097073850441246,
+      "grad_norm": 0.37480261887391536,
+      "learning_rate": 4.685025817555938e-05,
+      "loss": 0.4343,
+      "step": 507
+    },
+    {
+      "epoch": 0.47189967487227125,
+      "grad_norm": 0.2552649038864197,
+      "learning_rate": 4.683304647160069e-05,
+      "loss": 0.3908,
+      "step": 508
+    },
+    {
+      "epoch": 0.47282861124013004,
+      "grad_norm": 0.35376195689446877,
+      "learning_rate": 4.6815834767642e-05,
+      "loss": 0.4491,
+      "step": 509
+    },
+    {
+      "epoch": 0.4737575476079888,
+      "grad_norm": 0.3016203905265542,
+      "learning_rate": 4.6798623063683304e-05,
+      "loss": 0.4291,
+      "step": 510
+    },
+    {
+      "epoch": 0.4746864839758477,
+      "grad_norm": 0.29055972391783796,
+      "learning_rate": 4.6781411359724616e-05,
+      "loss": 0.4195,
+      "step": 511
+    },
+    {
+      "epoch": 0.47561542034370646,
+      "grad_norm": 0.3190347010255864,
+      "learning_rate": 4.676419965576592e-05,
+      "loss": 0.4108,
+      "step": 512
+    },
+    {
+      "epoch": 0.47654435671156525,
+      "grad_norm": 0.3294349955740729,
+      "learning_rate": 4.674698795180723e-05,
+      "loss": 0.4693,
+      "step": 513
+    },
+    {
+      "epoch": 0.47747329307942404,
+      "grad_norm": 0.31781079968735304,
+      "learning_rate": 4.672977624784854e-05,
+      "loss": 0.4435,
+      "step": 514
+    },
+    {
+      "epoch": 0.47840222944728283,
+      "grad_norm": 0.3226175038410287,
+      "learning_rate": 4.671256454388985e-05,
+      "loss": 0.4335,
+      "step": 515
+    },
+    {
+      "epoch": 0.4793311658151417,
+      "grad_norm": 0.2637433153066965,
+      "learning_rate": 4.6695352839931154e-05,
+      "loss": 0.4415,
+      "step": 516
+    },
+    {
+      "epoch": 0.48026010218300047,
+      "grad_norm": 0.3003332421321039,
+      "learning_rate": 4.6678141135972466e-05,
+      "loss": 0.4202,
+      "step": 517
+    },
+    {
+      "epoch": 0.48118903855085926,
+      "grad_norm": 0.37384718084964086,
+      "learning_rate": 4.666092943201377e-05,
+      "loss": 0.4603,
+      "step": 518
+    },
+    {
+      "epoch": 0.48211797491871805,
+      "grad_norm": 0.2715471956417477,
+      "learning_rate": 4.664371772805508e-05,
+      "loss": 0.4507,
+      "step": 519
+    },
+    {
+      "epoch": 0.4830469112865769,
+      "grad_norm": 0.262436801625057,
+      "learning_rate": 4.662650602409639e-05,
+      "loss": 0.4209,
+      "step": 520
+    },
+    {
+      "epoch": 0.4839758476544357,
+      "grad_norm": 0.3007149616543295,
+      "learning_rate": 4.66092943201377e-05,
+      "loss": 0.4493,
+      "step": 521
+    },
+    {
+      "epoch": 0.4849047840222945,
+      "grad_norm": 0.2804336311294239,
+      "learning_rate": 4.6592082616179005e-05,
+      "loss": 0.4286,
+      "step": 522
+    },
+    {
+      "epoch": 0.48583372039015327,
+      "grad_norm": 0.26330683566772206,
+      "learning_rate": 4.6574870912220316e-05,
+      "loss": 0.4202,
+      "step": 523
+    },
+    {
+      "epoch": 0.48676265675801206,
+      "grad_norm": 0.286309727815638,
+      "learning_rate": 4.655765920826162e-05,
+      "loss": 0.4301,
+      "step": 524
+    },
+    {
+      "epoch": 0.4876915931258709,
+      "grad_norm": 0.2529965819385294,
+      "learning_rate": 4.654044750430293e-05,
+      "loss": 0.4325,
+      "step": 525
+    },
+    {
+      "epoch": 0.4886205294937297,
+      "grad_norm": 0.3253005644243001,
+      "learning_rate": 4.652323580034423e-05,
+      "loss": 0.42,
+      "step": 526
+    },
+    {
+      "epoch": 0.4895494658615885,
+      "grad_norm": 0.2658924013835058,
+      "learning_rate": 4.650602409638554e-05,
+      "loss": 0.4193,
+      "step": 527
+    },
+    {
+      "epoch": 0.4904784022294473,
+      "grad_norm": 0.27995881407155443,
+      "learning_rate": 4.648881239242685e-05,
+      "loss": 0.4216,
+      "step": 528
+    },
+    {
+      "epoch": 0.49140733859730606,
+      "grad_norm": 0.3025444675822039,
+      "learning_rate": 4.647160068846816e-05,
+      "loss": 0.4621,
+      "step": 529
+    },
+    {
+      "epoch": 0.4923362749651649,
+      "grad_norm": 0.2737570425581585,
+      "learning_rate": 4.6454388984509465e-05,
+      "loss": 0.4324,
+      "step": 530
+    },
+    {
+      "epoch": 0.4932652113330237,
+      "grad_norm": 0.2582930487746826,
+      "learning_rate": 4.643717728055078e-05,
+      "loss": 0.4423,
+      "step": 531
+    },
+    {
+      "epoch": 0.4941941477008825,
+      "grad_norm": 0.28559852630894345,
+      "learning_rate": 4.641996557659208e-05,
+      "loss": 0.4477,
+      "step": 532
+    },
+    {
+      "epoch": 0.4951230840687413,
+      "grad_norm": 0.29483695154838563,
+      "learning_rate": 4.6402753872633394e-05,
+      "loss": 0.4156,
+      "step": 533
+    },
+    {
+      "epoch": 0.49605202043660007,
+      "grad_norm": 0.2613379393803091,
+      "learning_rate": 4.63855421686747e-05,
+      "loss": 0.432,
+      "step": 534
+    },
+    {
+      "epoch": 0.4969809568044589,
+      "grad_norm": 0.22622325752066913,
+      "learning_rate": 4.636833046471601e-05,
+      "loss": 0.4297,
+      "step": 535
+    },
+    {
+      "epoch": 0.4979098931723177,
+      "grad_norm": 0.28353494710194255,
+      "learning_rate": 4.6351118760757315e-05,
+      "loss": 0.4533,
+      "step": 536
+    },
+    {
+      "epoch": 0.4988388295401765,
+      "grad_norm": 0.31712551450474613,
+      "learning_rate": 4.633390705679863e-05,
+      "loss": 0.4667,
+      "step": 537
+    },
+    {
+      "epoch": 0.4997677659080353,
+      "grad_norm": 0.25690414290083796,
+      "learning_rate": 4.631669535283993e-05,
+      "loss": 0.434,
+      "step": 538
+    },
+    {
+      "epoch": 0.5006967022758941,
+      "grad_norm": 0.26005133208971615,
+      "learning_rate": 4.6299483648881244e-05,
+      "loss": 0.4385,
+      "step": 539
+    },
+    {
+      "epoch": 0.5016256386437529,
+      "grad_norm": 0.34410172574131104,
+      "learning_rate": 4.628227194492255e-05,
+      "loss": 0.4508,
+      "step": 540
+    },
+    {
+      "epoch": 0.5025545750116117,
+      "grad_norm": 0.25043115978998853,
+      "learning_rate": 4.626506024096386e-05,
+      "loss": 0.4098,
+      "step": 541
+    },
+    {
+      "epoch": 0.5034835113794706,
+      "grad_norm": 0.3104609566884535,
+      "learning_rate": 4.6247848537005166e-05,
+      "loss": 0.4305,
+      "step": 542
+    },
+    {
+      "epoch": 0.5044124477473293,
+      "grad_norm": 0.293951951581098,
+      "learning_rate": 4.623063683304648e-05,
+      "loss": 0.4461,
+      "step": 543
+    },
+    {
+      "epoch": 0.5053413841151881,
+      "grad_norm": 0.2560108568907451,
+      "learning_rate": 4.621342512908778e-05,
+      "loss": 0.4355,
+      "step": 544
+    },
+    {
+      "epoch": 0.5062703204830469,
+      "grad_norm": 0.358412433789302,
+      "learning_rate": 4.6196213425129094e-05,
+      "loss": 0.4425,
+      "step": 545
+    },
+    {
+      "epoch": 0.5071992568509057,
+      "grad_norm": 0.3056935351314926,
+      "learning_rate": 4.61790017211704e-05,
+      "loss": 0.4428,
+      "step": 546
+    },
+    {
+      "epoch": 0.5081281932187646,
+      "grad_norm": 0.24667559189967134,
+      "learning_rate": 4.6161790017211704e-05,
+      "loss": 0.3957,
+      "step": 547
+    },
+    {
+      "epoch": 0.5090571295866233,
+      "grad_norm": 0.33367259063635524,
+      "learning_rate": 4.614457831325301e-05,
+      "loss": 0.4355,
+      "step": 548
+    },
+    {
+      "epoch": 0.5099860659544821,
+      "grad_norm": 0.27853496333383976,
+      "learning_rate": 4.612736660929432e-05,
+      "loss": 0.4126,
+      "step": 549
+    },
+    {
+      "epoch": 0.5109150023223409,
+      "grad_norm": 0.268125448502846,
+      "learning_rate": 4.6110154905335626e-05,
+      "loss": 0.4303,
+      "step": 550
+    },
+    {
+      "epoch": 0.5118439386901997,
+      "grad_norm": 0.3441204665083962,
+      "learning_rate": 4.609294320137694e-05,
+      "loss": 0.449,
+      "step": 551
+    },
+    {
+      "epoch": 0.5127728750580586,
+      "grad_norm": 0.24236240929630035,
+      "learning_rate": 4.607573149741824e-05,
+      "loss": 0.414,
+      "step": 552
+    },
+    {
+      "epoch": 0.5137018114259173,
+      "grad_norm": 0.3015664010797303,
+      "learning_rate": 4.6058519793459555e-05,
+      "loss": 0.4204,
+      "step": 553
+    },
+    {
+      "epoch": 0.5146307477937762,
+      "grad_norm": 0.25242151768392607,
+      "learning_rate": 4.604130808950086e-05,
+      "loss": 0.4273,
+      "step": 554
+    },
+    {
+      "epoch": 0.5155596841616349,
+      "grad_norm": 0.30744068298559074,
+      "learning_rate": 4.602409638554217e-05,
+      "loss": 0.4508,
+      "step": 555
+    },
+    {
+      "epoch": 0.5164886205294937,
+      "grad_norm": 0.30718731865640203,
+      "learning_rate": 4.600688468158348e-05,
+      "loss": 0.4474,
+      "step": 556
+    },
+    {
+      "epoch": 0.5174175568973526,
+      "grad_norm": 0.29846939639927766,
+      "learning_rate": 4.598967297762479e-05,
+      "loss": 0.4302,
+      "step": 557
+    },
+    {
+      "epoch": 0.5183464932652113,
+      "grad_norm": 0.29336388381608647,
+      "learning_rate": 4.5972461273666093e-05,
+      "loss": 0.4227,
+      "step": 558
+    },
+    {
+      "epoch": 0.5192754296330702,
+      "grad_norm": 0.28098300576052854,
+      "learning_rate": 4.5955249569707405e-05,
+      "loss": 0.4225,
+      "step": 559
+    },
+    {
+      "epoch": 0.5202043660009289,
+      "grad_norm": 0.27826523322374,
+      "learning_rate": 4.593803786574871e-05,
+      "loss": 0.4395,
+      "step": 560
+    },
+    {
+      "epoch": 0.5211333023687877,
+      "grad_norm": 0.2857326733803591,
+      "learning_rate": 4.592082616179002e-05,
+      "loss": 0.4251,
+      "step": 561
+    },
+    {
+      "epoch": 0.5220622387366466,
+      "grad_norm": 0.24460900653251463,
+      "learning_rate": 4.590361445783133e-05,
+      "loss": 0.3983,
+      "step": 562
+    },
+    {
+      "epoch": 0.5229911751045053,
+      "grad_norm": 0.31606867032450164,
+      "learning_rate": 4.588640275387264e-05,
+      "loss": 0.3987,
+      "step": 563
+    },
+    {
+      "epoch": 0.5239201114723642,
+      "grad_norm": 0.2720837508546668,
+      "learning_rate": 4.5869191049913944e-05,
+      "loss": 0.4187,
+      "step": 564
+    },
+    {
+      "epoch": 0.5248490478402229,
+      "grad_norm": 0.3692339108368577,
+      "learning_rate": 4.5851979345955256e-05,
+      "loss": 0.4027,
+      "step": 565
+    },
+    {
+      "epoch": 0.5257779842080818,
+      "grad_norm": 0.2621237616477022,
+      "learning_rate": 4.583476764199656e-05,
+      "loss": 0.4069,
+      "step": 566
+    },
+    {
+      "epoch": 0.5267069205759406,
+      "grad_norm": 0.28576357238482947,
+      "learning_rate": 4.581755593803787e-05,
+      "loss": 0.4083,
+      "step": 567
+    },
+    {
+      "epoch": 0.5276358569437993,
+      "grad_norm": 0.2876324843394776,
+      "learning_rate": 4.580034423407917e-05,
+      "loss": 0.4053,
+      "step": 568
+    },
+    {
+      "epoch": 0.5285647933116582,
+      "grad_norm": 0.2952514491419477,
+      "learning_rate": 4.578313253012048e-05,
+      "loss": 0.4372,
+      "step": 569
+    },
+    {
+      "epoch": 0.5294937296795169,
+      "grad_norm": 0.2413668924475377,
+      "learning_rate": 4.576592082616179e-05,
+      "loss": 0.3872,
+      "step": 570
+    },
+    {
+      "epoch": 0.5304226660473758,
+      "grad_norm": 0.31322703800100615,
+      "learning_rate": 4.57487091222031e-05,
+      "loss": 0.415,
+      "step": 571
+    },
+    {
+      "epoch": 0.5313516024152346,
+      "grad_norm": 0.29838477602444063,
+      "learning_rate": 4.5731497418244404e-05,
+      "loss": 0.4227,
+      "step": 572
+    },
+    {
+      "epoch": 0.5322805387830933,
+      "grad_norm": 0.2743433087149783,
+      "learning_rate": 4.5714285714285716e-05,
+      "loss": 0.4069,
+      "step": 573
+    },
+    {
+      "epoch": 0.5332094751509522,
+      "grad_norm": 0.2681736163650413,
+      "learning_rate": 4.569707401032702e-05,
+      "loss": 0.4124,
+      "step": 574
+    },
+    {
+      "epoch": 0.5341384115188109,
+      "grad_norm": 0.27958710415501165,
+      "learning_rate": 4.567986230636833e-05,
+      "loss": 0.4479,
+      "step": 575
+    },
+    {
+      "epoch": 0.5350673478866698,
+      "grad_norm": 0.3107133186979922,
+      "learning_rate": 4.566265060240964e-05,
+      "loss": 0.4309,
+      "step": 576
+    },
+    {
+      "epoch": 0.5359962842545286,
+      "grad_norm": 0.3008304998300067,
+      "learning_rate": 4.564543889845095e-05,
+      "loss": 0.4215,
+      "step": 577
+    },
+    {
+      "epoch": 0.5369252206223873,
+      "grad_norm": 0.3077480260868722,
+      "learning_rate": 4.5628227194492255e-05,
+      "loss": 0.4177,
+      "step": 578
+    },
+    {
+      "epoch": 0.5378541569902462,
+      "grad_norm": 0.2856706754554965,
+      "learning_rate": 4.5611015490533566e-05,
+      "loss": 0.4298,
+      "step": 579
+    },
+    {
+      "epoch": 0.5387830933581049,
+      "grad_norm": 0.3062434202958225,
+      "learning_rate": 4.559380378657487e-05,
+      "loss": 0.4489,
+      "step": 580
+    },
+    {
+      "epoch": 0.5397120297259638,
+      "grad_norm": 0.26880576068838663,
+      "learning_rate": 4.557659208261618e-05,
+      "loss": 0.4378,
+      "step": 581
+    },
+    {
+      "epoch": 0.5406409660938226,
+      "grad_norm": 0.2504523646966336,
+      "learning_rate": 4.555938037865749e-05,
+      "loss": 0.4308,
+      "step": 582
+    },
+    {
+      "epoch": 0.5415699024616814,
+      "grad_norm": 0.26076967760757586,
+      "learning_rate": 4.55421686746988e-05,
+      "loss": 0.422,
+      "step": 583
+    },
+    {
+      "epoch": 0.5424988388295402,
+      "grad_norm": 0.235898250335184,
+      "learning_rate": 4.5524956970740105e-05,
+      "loss": 0.4317,
+      "step": 584
+    },
+    {
+      "epoch": 0.5434277751973989,
+      "grad_norm": 0.24591979617927925,
+      "learning_rate": 4.550774526678142e-05,
+      "loss": 0.4576,
+      "step": 585
+    },
+    {
+      "epoch": 0.5443567115652578,
+      "grad_norm": 0.2336427626086619,
+      "learning_rate": 4.549053356282272e-05,
+      "loss": 0.4229,
+      "step": 586
+    },
+    {
+      "epoch": 0.5452856479331166,
+      "grad_norm": 0.24258696081299802,
+      "learning_rate": 4.5473321858864034e-05,
+      "loss": 0.4314,
+      "step": 587
+    },
+    {
+      "epoch": 0.5462145843009754,
+      "grad_norm": 0.2448812822768551,
+      "learning_rate": 4.545611015490534e-05,
+      "loss": 0.4148,
+      "step": 588
+    },
+    {
+      "epoch": 0.5471435206688342,
+      "grad_norm": 0.23590842458952918,
+      "learning_rate": 4.543889845094665e-05,
+      "loss": 0.42,
+      "step": 589
+    },
+    {
+      "epoch": 0.5480724570366929,
+      "grad_norm": 0.2522688595104704,
+      "learning_rate": 4.542168674698795e-05,
+      "loss": 0.4314,
+      "step": 590
+    },
+    {
+      "epoch": 0.5490013934045518,
+      "grad_norm": 0.26737150464975795,
+      "learning_rate": 4.540447504302926e-05,
+      "loss": 0.4086,
+      "step": 591
+    },
+    {
+      "epoch": 0.5499303297724106,
+      "grad_norm": 0.2559167030387441,
+      "learning_rate": 4.5387263339070566e-05,
+      "loss": 0.4365,
+      "step": 592
+    },
+    {
+      "epoch": 0.5508592661402694,
+      "grad_norm": 0.2767220404229008,
+      "learning_rate": 4.537005163511188e-05,
+      "loss": 0.4148,
+      "step": 593
+    },
+    {
+      "epoch": 0.5517882025081282,
+      "grad_norm": 0.27937423564051256,
+      "learning_rate": 4.535283993115318e-05,
+      "loss": 0.4253,
+      "step": 594
+    },
+    {
+      "epoch": 0.552717138875987,
+      "grad_norm": 0.2900121711086677,
+      "learning_rate": 4.5335628227194494e-05,
+      "loss": 0.4325,
+      "step": 595
+    },
+    {
+      "epoch": 0.5536460752438458,
+      "grad_norm": 0.27657569986368546,
+      "learning_rate": 4.53184165232358e-05,
+      "loss": 0.4505,
+      "step": 596
+    },
+    {
+      "epoch": 0.5545750116117046,
+      "grad_norm": 0.3126165251767002,
+      "learning_rate": 4.530120481927711e-05,
+      "loss": 0.4358,
+      "step": 597
+    },
+    {
+      "epoch": 0.5555039479795634,
+      "grad_norm": 0.27088886464887985,
+      "learning_rate": 4.5283993115318416e-05,
+      "loss": 0.4393,
+      "step": 598
+    },
+    {
+      "epoch": 0.5564328843474222,
+      "grad_norm": 0.2504180754270926,
+      "learning_rate": 4.526678141135973e-05,
+      "loss": 0.4188,
+      "step": 599
+    },
+    {
+      "epoch": 0.557361820715281,
+      "grad_norm": 0.30044985875880187,
+      "learning_rate": 4.524956970740103e-05,
+      "loss": 0.4272,
+      "step": 600
+    },
+    {
+      "epoch": 0.5582907570831398,
+      "grad_norm": 0.26657267822610964,
+      "learning_rate": 4.5232358003442345e-05,
+      "loss": 0.4351,
+      "step": 601
+    },
+    {
+      "epoch": 0.5592196934509986,
+      "grad_norm": 0.2534161771476428,
+      "learning_rate": 4.521514629948365e-05,
+      "loss": 0.4171,
+      "step": 602
+    },
+    {
+      "epoch": 0.5601486298188574,
+      "grad_norm": 0.2856683409630207,
+      "learning_rate": 4.519793459552496e-05,
+      "loss": 0.4229,
+      "step": 603
+    },
+    {
+      "epoch": 0.5610775661867162,
+      "grad_norm": 0.2232216928555761,
+      "learning_rate": 4.5180722891566266e-05,
+      "loss": 0.3916,
+      "step": 604
+    },
+    {
+      "epoch": 0.562006502554575,
+      "grad_norm": 0.3018832675679175,
+      "learning_rate": 4.516351118760758e-05,
+      "loss": 0.4179,
+      "step": 605
+    },
+    {
+      "epoch": 0.5629354389224338,
+      "grad_norm": 0.2569380913397442,
+      "learning_rate": 4.514629948364888e-05,
+      "loss": 0.4311,
+      "step": 606
+    },
+    {
+      "epoch": 0.5638643752902927,
+      "grad_norm": 0.2722108299381111,
+      "learning_rate": 4.5129087779690195e-05,
+      "loss": 0.4227,
+      "step": 607
+    },
+    {
+      "epoch": 0.5647933116581514,
+      "grad_norm": 0.24245246643994783,
+      "learning_rate": 4.51118760757315e-05,
+      "loss": 0.4058,
+      "step": 608
+    },
+    {
+      "epoch": 0.5657222480260102,
+      "grad_norm": 0.2895449101358798,
+      "learning_rate": 4.509466437177281e-05,
+      "loss": 0.4239,
+      "step": 609
+    },
+    {
+      "epoch": 0.566651184393869,
+      "grad_norm": 0.22872993936692906,
+      "learning_rate": 4.507745266781412e-05,
+      "loss": 0.4194,
+      "step": 610
+    },
+    {
+      "epoch": 0.5675801207617278,
+      "grad_norm": 0.2951044959391205,
+      "learning_rate": 4.506024096385542e-05,
+      "loss": 0.4159,
+      "step": 611
+    },
+    {
+      "epoch": 0.5685090571295867,
+      "grad_norm": 0.2509311292631651,
+      "learning_rate": 4.504302925989673e-05,
+      "loss": 0.4237,
+      "step": 612
+    },
+    {
+      "epoch": 0.5694379934974454,
+      "grad_norm": 0.32512599427391997,
+      "learning_rate": 4.502581755593804e-05,
+      "loss": 0.416,
+      "step": 613
+    },
+    {
+      "epoch": 0.5703669298653042,
+      "grad_norm": 0.2819827511805377,
+      "learning_rate": 4.5008605851979344e-05,
+      "loss": 0.3945,
+      "step": 614
+    },
+    {
+      "epoch": 0.571295866233163,
+      "grad_norm": 0.26137765186363743,
+      "learning_rate": 4.4991394148020655e-05,
+      "loss": 0.4364,
+      "step": 615
+    },
+    {
+      "epoch": 0.5722248026010218,
+      "grad_norm": 0.24986429511060695,
+      "learning_rate": 4.497418244406196e-05,
+      "loss": 0.4192,
+      "step": 616
+    },
+    {
+      "epoch": 0.5731537389688807,
+      "grad_norm": 0.32510562396239795,
+      "learning_rate": 4.495697074010327e-05,
+      "loss": 0.4564,
+      "step": 617
+    },
+    {
+      "epoch": 0.5740826753367394,
+      "grad_norm": 0.2437246846388565,
+      "learning_rate": 4.493975903614458e-05,
+      "loss": 0.4491,
+      "step": 618
+    },
+    {
+      "epoch": 0.5750116117045982,
+      "grad_norm": 0.28270839310005524,
+      "learning_rate": 4.492254733218589e-05,
+      "loss": 0.4165,
+      "step": 619
+    },
+    {
+      "epoch": 0.575940548072457,
+      "grad_norm": 0.2978639151525471,
+      "learning_rate": 4.4905335628227194e-05,
+      "loss": 0.4181,
+      "step": 620
+    },
+    {
+      "epoch": 0.5768694844403158,
+      "grad_norm": 0.283277054225362,
+      "learning_rate": 4.4888123924268506e-05,
+      "loss": 0.4506,
+      "step": 621
+    },
+    {
+      "epoch": 0.5777984208081747,
+      "grad_norm": 0.24187266585240402,
+      "learning_rate": 4.487091222030981e-05,
+      "loss": 0.401,
+      "step": 622
+    },
+    {
+      "epoch": 0.5787273571760334,
+      "grad_norm": 0.2863505365930887,
+      "learning_rate": 4.485370051635112e-05,
+      "loss": 0.4117,
+      "step": 623
+    },
+    {
+      "epoch": 0.5796562935438923,
+      "grad_norm": 0.2773113061348817,
+      "learning_rate": 4.483648881239243e-05,
+      "loss": 0.4244,
+      "step": 624
+    },
+    {
+      "epoch": 0.5805852299117511,
+      "grad_norm": 0.23781370664986884,
+      "learning_rate": 4.481927710843374e-05,
+      "loss": 0.4172,
+      "step": 625
+    },
+    {
+      "epoch": 0.5815141662796098,
+      "grad_norm": 0.2899327516074531,
+      "learning_rate": 4.4802065404475044e-05,
+      "loss": 0.414,
+      "step": 626
+    },
+    {
+      "epoch": 0.5824431026474687,
+      "grad_norm": 0.21917991809985152,
+      "learning_rate": 4.4784853700516356e-05,
+      "loss": 0.3858,
+      "step": 627
+    },
+    {
+      "epoch": 0.5833720390153274,
+      "grad_norm": 0.29585581964621416,
+      "learning_rate": 4.476764199655766e-05,
+      "loss": 0.4262,
+      "step": 628
+    },
+    {
+      "epoch": 0.5843009753831863,
+      "grad_norm": 0.24103700840664655,
+      "learning_rate": 4.475043029259897e-05,
+      "loss": 0.4061,
+      "step": 629
+    },
+    {
+      "epoch": 0.5852299117510451,
+      "grad_norm": 0.30538358771552415,
+      "learning_rate": 4.473321858864028e-05,
+      "loss": 0.4068,
+      "step": 630
+    },
+    {
+      "epoch": 0.5861588481189038,
+      "grad_norm": 0.28929577954162256,
+      "learning_rate": 4.471600688468159e-05,
+      "loss": 0.4307,
+      "step": 631
+    },
+    {
+      "epoch": 0.5870877844867627,
+      "grad_norm": 0.28414615427121404,
+      "learning_rate": 4.469879518072289e-05,
+      "loss": 0.4201,
+      "step": 632
+    },
+    {
+      "epoch": 0.5880167208546214,
+      "grad_norm": 0.26330534375194914,
+      "learning_rate": 4.46815834767642e-05,
+      "loss": 0.4346,
+      "step": 633
+    },
+    {
+      "epoch": 0.5889456572224803,
+      "grad_norm": 0.24452133530493197,
+      "learning_rate": 4.4664371772805505e-05,
+      "loss": 0.4283,
+      "step": 634
+    },
+    {
+      "epoch": 0.5898745935903391,
+      "grad_norm": 0.275064839540402,
+      "learning_rate": 4.4647160068846817e-05,
+      "loss": 0.4103,
+      "step": 635
+    },
+    {
+      "epoch": 0.5908035299581978,
+      "grad_norm": 0.23295221964562113,
+      "learning_rate": 4.462994836488812e-05,
+      "loss": 0.4265,
+      "step": 636
+    },
+    {
+      "epoch": 0.5917324663260567,
+      "grad_norm": 0.29348555430997814,
+      "learning_rate": 4.461273666092943e-05,
+      "loss": 0.4498,
+      "step": 637
+    },
+    {
+      "epoch": 0.5926614026939154,
+      "grad_norm": 0.2387454374843579,
+      "learning_rate": 4.459552495697074e-05,
+      "loss": 0.4267,
+      "step": 638
+    },
+    {
+      "epoch": 0.5935903390617743,
+      "grad_norm": 0.24669696423956647,
+      "learning_rate": 4.457831325301205e-05,
+      "loss": 0.4109,
+      "step": 639
+    },
+    {
+      "epoch": 0.5945192754296331,
+      "grad_norm": 0.2569187304772452,
+      "learning_rate": 4.4561101549053355e-05,
+      "loss": 0.4142,
+      "step": 640
+    },
+    {
+      "epoch": 0.5954482117974919,
+      "grad_norm": 0.23582837954343758,
+      "learning_rate": 4.454388984509467e-05,
+      "loss": 0.4361,
+      "step": 641
+    },
+    {
+      "epoch": 0.5963771481653507,
+      "grad_norm": 0.2577453345455404,
+      "learning_rate": 4.452667814113597e-05,
+      "loss": 0.4208,
+      "step": 642
+    },
+    {
+      "epoch": 0.5973060845332094,
+      "grad_norm": 0.2575075232080959,
+      "learning_rate": 4.4509466437177284e-05,
+      "loss": 0.4175,
+      "step": 643
+    },
+    {
+      "epoch": 0.5982350209010683,
+      "grad_norm": 0.2592496028490341,
+      "learning_rate": 4.449225473321859e-05,
+      "loss": 0.4243,
+      "step": 644
+    },
+    {
+      "epoch": 0.5991639572689271,
+      "grad_norm": 0.2518897193108137,
+      "learning_rate": 4.44750430292599e-05,
+      "loss": 0.4238,
+      "step": 645
+    },
+    {
+      "epoch": 0.6000928936367859,
+      "grad_norm": 0.27689384611411444,
+      "learning_rate": 4.4457831325301206e-05,
+      "loss": 0.4145,
+      "step": 646
+    },
+    {
+      "epoch": 0.6010218300046447,
+      "grad_norm": 0.25485645505304355,
+      "learning_rate": 4.444061962134252e-05,
+      "loss": 0.4201,
+      "step": 647
+    },
+    {
+      "epoch": 0.6019507663725034,
+      "grad_norm": 0.27128573378890863,
+      "learning_rate": 4.442340791738382e-05,
+      "loss": 0.4316,
+      "step": 648
+    },
+    {
+      "epoch": 0.6028797027403623,
+      "grad_norm": 0.2578065387198813,
+      "learning_rate": 4.4406196213425134e-05,
+      "loss": 0.4193,
+      "step": 649
+    },
+    {
+      "epoch": 0.6038086391082211,
+      "grad_norm": 0.2667247647777713,
+      "learning_rate": 4.438898450946644e-05,
+      "loss": 0.3986,
+      "step": 650
+    },
+    {
+      "epoch": 0.6047375754760799,
+      "grad_norm": 0.26585574043938304,
+      "learning_rate": 4.437177280550775e-05,
+      "loss": 0.4433,
+      "step": 651
+    },
+    {
+      "epoch": 0.6056665118439387,
+      "grad_norm": 0.29551030914401943,
+      "learning_rate": 4.4354561101549056e-05,
+      "loss": 0.4461,
+      "step": 652
+    },
+    {
+      "epoch": 0.6065954482117974,
+      "grad_norm": 0.2841049021245725,
+      "learning_rate": 4.433734939759036e-05,
+      "loss": 0.4071,
+      "step": 653
+    },
+    {
+      "epoch": 0.6075243845796563,
+      "grad_norm": 0.30886770292110666,
+      "learning_rate": 4.4320137693631666e-05,
+      "loss": 0.4544,
+      "step": 654
+    },
+    {
+      "epoch": 0.6084533209475151,
+      "grad_norm": 0.2686093425730839,
+      "learning_rate": 4.430292598967298e-05,
+      "loss": 0.407,
+      "step": 655
+    },
+    {
+      "epoch": 0.6093822573153739,
+      "grad_norm": 0.30412706468771006,
+      "learning_rate": 4.428571428571428e-05,
+      "loss": 0.3915,
+      "step": 656
+    },
+    {
+      "epoch": 0.6103111936832327,
+      "grad_norm": 0.35364500741842214,
+      "learning_rate": 4.4268502581755595e-05,
+      "loss": 0.453,
+      "step": 657
+    },
+    {
+      "epoch": 0.6112401300510915,
+      "grad_norm": 0.26460657385536474,
+      "learning_rate": 4.42512908777969e-05,
+      "loss": 0.3909,
+      "step": 658
+    },
+    {
+      "epoch": 0.6121690664189503,
+      "grad_norm": 0.4428486110797581,
+      "learning_rate": 4.423407917383821e-05,
+      "loss": 0.4269,
+      "step": 659
+    },
+    {
+      "epoch": 0.6130980027868091,
+      "grad_norm": 0.2866328626559702,
+      "learning_rate": 4.4216867469879516e-05,
+      "loss": 0.4204,
+      "step": 660
+    },
+    {
+      "epoch": 0.6140269391546679,
+      "grad_norm": 0.36462567528091633,
+      "learning_rate": 4.419965576592083e-05,
+      "loss": 0.4353,
+      "step": 661
+    },
+    {
+      "epoch": 0.6149558755225267,
+      "grad_norm": 0.3087527803770854,
+      "learning_rate": 4.418244406196213e-05,
+      "loss": 0.4204,
+      "step": 662
+    },
+    {
+      "epoch": 0.6158848118903855,
+      "grad_norm": 0.3103535653539977,
+      "learning_rate": 4.4165232358003445e-05,
+      "loss": 0.4344,
+      "step": 663
+    },
+    {
+      "epoch": 0.6168137482582443,
+      "grad_norm": 0.28523746975783004,
+      "learning_rate": 4.414802065404475e-05,
+      "loss": 0.4163,
+      "step": 664
+    },
+    {
+      "epoch": 0.6177426846261032,
+      "grad_norm": 0.2842390326622571,
+      "learning_rate": 4.413080895008606e-05,
+      "loss": 0.408,
+      "step": 665
+    },
+    {
+      "epoch": 0.6186716209939619,
+      "grad_norm": 0.23234413274784502,
+      "learning_rate": 4.411359724612737e-05,
+      "loss": 0.4194,
+      "step": 666
+    },
+    {
+      "epoch": 0.6196005573618207,
+      "grad_norm": 0.275004251162571,
+      "learning_rate": 4.409638554216868e-05,
+      "loss": 0.4323,
+      "step": 667
+    },
+    {
+      "epoch": 0.6205294937296795,
+      "grad_norm": 0.31847768258839976,
+      "learning_rate": 4.4079173838209984e-05,
+      "loss": 0.4268,
+      "step": 668
+    },
+    {
+      "epoch": 0.6214584300975383,
+      "grad_norm": 0.23406432898908414,
+      "learning_rate": 4.4061962134251295e-05,
+      "loss": 0.4137,
+      "step": 669
+    },
+    {
+      "epoch": 0.6223873664653972,
+      "grad_norm": 0.3093054020461385,
+      "learning_rate": 4.40447504302926e-05,
+      "loss": 0.3857,
+      "step": 670
+    },
+    {
+      "epoch": 0.6233163028332559,
+      "grad_norm": 0.25338402995679804,
+      "learning_rate": 4.402753872633391e-05,
+      "loss": 0.4007,
+      "step": 671
+    },
+    {
+      "epoch": 0.6242452392011147,
+      "grad_norm": 0.2935153845414147,
+      "learning_rate": 4.401032702237522e-05,
+      "loss": 0.4253,
+      "step": 672
+    },
+    {
+      "epoch": 0.6251741755689735,
+      "grad_norm": 0.3257007027536486,
+      "learning_rate": 4.399311531841653e-05,
+      "loss": 0.4573,
+      "step": 673
+    },
+    {
+      "epoch": 0.6261031119368323,
+      "grad_norm": 0.29349741747875213,
+      "learning_rate": 4.3975903614457834e-05,
+      "loss": 0.4016,
+      "step": 674
+    },
+    {
+      "epoch": 0.6270320483046912,
+      "grad_norm": 0.3229849235540013,
+      "learning_rate": 4.395869191049914e-05,
+      "loss": 0.4272,
+      "step": 675
+    },
+    {
+      "epoch": 0.6279609846725499,
+      "grad_norm": 0.36116533623931474,
+      "learning_rate": 4.3941480206540444e-05,
+      "loss": 0.4144,
+      "step": 676
+    },
+    {
+      "epoch": 0.6288899210404088,
+      "grad_norm": 0.2563978464326882,
+      "learning_rate": 4.3924268502581756e-05,
+      "loss": 0.4194,
+      "step": 677
+    },
+    {
+      "epoch": 0.6298188574082675,
+      "grad_norm": 0.3196063755723246,
+      "learning_rate": 4.390705679862306e-05,
+      "loss": 0.393,
+      "step": 678
+    },
+    {
+      "epoch": 0.6307477937761263,
+      "grad_norm": 0.35436425630405965,
+      "learning_rate": 4.388984509466437e-05,
+      "loss": 0.4171,
+      "step": 679
+    },
+    {
+      "epoch": 0.6316767301439852,
+      "grad_norm": 0.30135253821167374,
+      "learning_rate": 4.387263339070568e-05,
+      "loss": 0.4256,
+      "step": 680
+    },
+    {
+      "epoch": 0.6326056665118439,
+      "grad_norm": 0.30674680549750183,
+      "learning_rate": 4.385542168674699e-05,
+      "loss": 0.3959,
+      "step": 681
+    },
+    {
+      "epoch": 0.6335346028797028,
+      "grad_norm": 0.32187516875469857,
+      "learning_rate": 4.3838209982788294e-05,
+      "loss": 0.419,
+      "step": 682
+    },
+    {
+      "epoch": 0.6344635392475615,
+      "grad_norm": 0.2896684222699531,
+      "learning_rate": 4.3820998278829606e-05,
+      "loss": 0.4324,
+      "step": 683
+    },
+    {
+      "epoch": 0.6353924756154203,
+      "grad_norm": 0.2799876389868912,
+      "learning_rate": 4.380378657487091e-05,
+      "loss": 0.4238,
+      "step": 684
+    },
+    {
+      "epoch": 0.6363214119832792,
+      "grad_norm": 0.26251094611425313,
+      "learning_rate": 4.378657487091222e-05,
+      "loss": 0.4537,
+      "step": 685
+    },
+    {
+      "epoch": 0.6372503483511379,
+      "grad_norm": 0.24365180702477157,
+      "learning_rate": 4.376936316695353e-05,
+      "loss": 0.4201,
+      "step": 686
+    },
+    {
+      "epoch": 0.6381792847189968,
+      "grad_norm": 0.34911025101281234,
+      "learning_rate": 4.375215146299484e-05,
+      "loss": 0.4591,
+      "step": 687
+    },
+    {
+      "epoch": 0.6391082210868555,
+      "grad_norm": 0.25889134972689726,
+      "learning_rate": 4.3734939759036145e-05,
+      "loss": 0.4272,
+      "step": 688
+    },
+    {
+      "epoch": 0.6400371574547143,
+      "grad_norm": 0.28475140108094354,
+      "learning_rate": 4.371772805507746e-05,
+      "loss": 0.4267,
+      "step": 689
+    },
+    {
+      "epoch": 0.6409660938225732,
+      "grad_norm": 0.3161029094501537,
+      "learning_rate": 4.370051635111876e-05,
+      "loss": 0.4418,
+      "step": 690
+    },
+    {
+      "epoch": 0.6418950301904319,
+      "grad_norm": 0.2396023845575197,
+      "learning_rate": 4.3683304647160073e-05,
+      "loss": 0.4421,
+      "step": 691
+    },
+    {
+      "epoch": 0.6428239665582908,
+      "grad_norm": 0.2832054932462056,
+      "learning_rate": 4.366609294320138e-05,
+      "loss": 0.4091,
+      "step": 692
+    },
+    {
+      "epoch": 0.6437529029261495,
+      "grad_norm": 0.28376936431369904,
+      "learning_rate": 4.364888123924269e-05,
+      "loss": 0.3974,
+      "step": 693
+    },
+    {
+      "epoch": 0.6446818392940084,
+      "grad_norm": 0.24743406449667463,
+      "learning_rate": 4.3631669535283995e-05,
+      "loss": 0.411,
+      "step": 694
+    },
+    {
+      "epoch": 0.6456107756618672,
+      "grad_norm": 0.2730246000748479,
+      "learning_rate": 4.361445783132531e-05,
+      "loss": 0.4277,
+      "step": 695
+    },
+    {
+      "epoch": 0.6465397120297259,
+      "grad_norm": 0.2650695341327053,
+      "learning_rate": 4.359724612736661e-05,
+      "loss": 0.3944,
+      "step": 696
+    },
+    {
+      "epoch": 0.6474686483975848,
+      "grad_norm": 0.321320466947324,
+      "learning_rate": 4.358003442340792e-05,
+      "loss": 0.4464,
+      "step": 697
+    },
+    {
+      "epoch": 0.6483975847654435,
+      "grad_norm": 0.3178561485682785,
+      "learning_rate": 4.356282271944923e-05,
+      "loss": 0.4503,
+      "step": 698
+    },
+    {
+      "epoch": 0.6493265211333024,
+      "grad_norm": 0.25967628951798444,
+      "learning_rate": 4.3545611015490534e-05,
+      "loss": 0.4225,
+      "step": 699
+    },
+    {
+      "epoch": 0.6502554575011612,
+      "grad_norm": 0.2895967588454303,
+      "learning_rate": 4.3528399311531846e-05,
+      "loss": 0.4181,
+      "step": 700
+    },
+    {
+      "epoch": 0.6511843938690199,
+      "grad_norm": 0.278239598528544,
+      "learning_rate": 4.351118760757315e-05,
+      "loss": 0.4196,
+      "step": 701
+    },
+    {
+      "epoch": 0.6521133302368788,
+      "grad_norm": 0.24741676508319585,
+      "learning_rate": 4.3493975903614456e-05,
+      "loss": 0.4183,
+      "step": 702
+    },
+    {
+      "epoch": 0.6530422666047375,
+      "grad_norm": 0.23960910227402535,
+      "learning_rate": 4.347676419965577e-05,
+      "loss": 0.4154,
+      "step": 703
+    },
+    {
+      "epoch": 0.6539712029725964,
+      "grad_norm": 0.30614533128558064,
+      "learning_rate": 4.345955249569707e-05,
+      "loss": 0.4132,
+      "step": 704
+    },
+    {
+      "epoch": 0.6549001393404552,
+      "grad_norm": 0.2690126725576619,
+      "learning_rate": 4.3442340791738384e-05,
+      "loss": 0.4252,
+      "step": 705
+    },
+    {
+      "epoch": 0.655829075708314,
+      "grad_norm": 0.2849912129415421,
+      "learning_rate": 4.342512908777969e-05,
+      "loss": 0.4179,
+      "step": 706
+    },
+    {
+      "epoch": 0.6567580120761728,
+      "grad_norm": 0.2480212482763167,
+      "learning_rate": 4.3407917383821e-05,
+      "loss": 0.429,
+      "step": 707
+    },
+    {
+      "epoch": 0.6576869484440316,
+      "grad_norm": 0.2804131416188235,
+      "learning_rate": 4.3390705679862306e-05,
+      "loss": 0.4068,
+      "step": 708
+    },
+    {
+      "epoch": 0.6586158848118904,
+      "grad_norm": 0.2768683400232619,
+      "learning_rate": 4.337349397590362e-05,
+      "loss": 0.423,
+      "step": 709
+    },
+    {
+      "epoch": 0.6595448211797492,
+      "grad_norm": 0.24148977148375103,
+      "learning_rate": 4.335628227194492e-05,
+      "loss": 0.4322,
+      "step": 710
+    },
+    {
+      "epoch": 0.660473757547608,
+      "grad_norm": 0.2405167807911801,
+      "learning_rate": 4.3339070567986235e-05,
+      "loss": 0.4046,
+      "step": 711
+    },
+    {
+      "epoch": 0.6614026939154668,
+      "grad_norm": 0.22245312753439328,
+      "learning_rate": 4.332185886402754e-05,
+      "loss": 0.4126,
+      "step": 712
+    },
+    {
+      "epoch": 0.6623316302833256,
+      "grad_norm": 0.2669023790429907,
+      "learning_rate": 4.330464716006885e-05,
+      "loss": 0.4338,
+      "step": 713
+    },
+    {
+      "epoch": 0.6632605666511844,
+      "grad_norm": 0.2521812276262092,
+      "learning_rate": 4.3287435456110156e-05,
+      "loss": 0.4448,
+      "step": 714
+    },
+    {
+      "epoch": 0.6641895030190432,
+      "grad_norm": 0.2625757212246767,
+      "learning_rate": 4.327022375215147e-05,
+      "loss": 0.414,
+      "step": 715
+    },
+    {
+      "epoch": 0.665118439386902,
+      "grad_norm": 0.25033293810292023,
+      "learning_rate": 4.325301204819277e-05,
+      "loss": 0.423,
+      "step": 716
+    },
+    {
+      "epoch": 0.6660473757547608,
+      "grad_norm": 0.22334251988732523,
+      "learning_rate": 4.323580034423408e-05,
+      "loss": 0.4237,
+      "step": 717
+    },
+    {
+      "epoch": 0.6669763121226197,
+      "grad_norm": 0.2397245440218477,
+      "learning_rate": 4.321858864027539e-05,
+      "loss": 0.4137,
+      "step": 718
+    },
+    {
+      "epoch": 0.6679052484904784,
+      "grad_norm": 0.30143292571229535,
+      "learning_rate": 4.3201376936316695e-05,
+      "loss": 0.3724,
+      "step": 719
+    },
+    {
+      "epoch": 0.6688341848583372,
+      "grad_norm": 0.23079068591492524,
+      "learning_rate": 4.318416523235801e-05,
+      "loss": 0.4007,
+      "step": 720
+    },
+    {
+      "epoch": 0.669763121226196,
+      "grad_norm": 0.30620216556806684,
+      "learning_rate": 4.316695352839931e-05,
+      "loss": 0.4384,
+      "step": 721
+    },
+    {
+      "epoch": 0.6706920575940548,
+      "grad_norm": 0.29486394422042084,
+      "learning_rate": 4.3149741824440624e-05,
+      "loss": 0.4046,
+      "step": 722
+    },
+    {
+      "epoch": 0.6716209939619137,
+      "grad_norm": 0.3305581175194426,
+      "learning_rate": 4.313253012048193e-05,
+      "loss": 0.4242,
+      "step": 723
+    },
+    {
+      "epoch": 0.6725499303297724,
+      "grad_norm": 0.2679330466442912,
+      "learning_rate": 4.311531841652324e-05,
+      "loss": 0.4257,
+      "step": 724
+    },
+    {
+      "epoch": 0.6734788666976312,
+      "grad_norm": 0.35811562362234445,
+      "learning_rate": 4.3098106712564545e-05,
+      "loss": 0.4384,
+      "step": 725
+    },
+    {
+      "epoch": 0.67440780306549,
+      "grad_norm": 0.29334272663787064,
+      "learning_rate": 4.308089500860585e-05,
+      "loss": 0.4444,
+      "step": 726
+    },
+    {
+      "epoch": 0.6753367394333488,
+      "grad_norm": 0.24401338629880093,
+      "learning_rate": 4.306368330464716e-05,
+      "loss": 0.4193,
+      "step": 727
+    },
+    {
+      "epoch": 0.6762656758012077,
+      "grad_norm": 0.28052429245661187,
+      "learning_rate": 4.304647160068847e-05,
+      "loss": 0.4088,
+      "step": 728
+    },
+    {
+      "epoch": 0.6771946121690664,
+      "grad_norm": 0.25925872113831155,
+      "learning_rate": 4.302925989672978e-05,
+      "loss": 0.4073,
+      "step": 729
+    },
+    {
+      "epoch": 0.6781235485369252,
+      "grad_norm": 0.2736489940435994,
+      "learning_rate": 4.3012048192771084e-05,
+      "loss": 0.4237,
+      "step": 730
+    },
+    {
+      "epoch": 0.679052484904784,
+      "grad_norm": 0.25905656215957673,
+      "learning_rate": 4.2994836488812396e-05,
+      "loss": 0.4162,
+      "step": 731
+    },
+    {
+      "epoch": 0.6799814212726428,
+      "grad_norm": 0.24390493228872448,
+      "learning_rate": 4.29776247848537e-05,
+      "loss": 0.4307,
+      "step": 732
+    },
+    {
+      "epoch": 0.6809103576405017,
+      "grad_norm": 0.24108332455669904,
+      "learning_rate": 4.296041308089501e-05,
+      "loss": 0.4408,
+      "step": 733
+    },
+    {
+      "epoch": 0.6818392940083604,
+      "grad_norm": 0.23861930125640354,
+      "learning_rate": 4.294320137693632e-05,
+      "loss": 0.4097,
+      "step": 734
+    },
+    {
+      "epoch": 0.6827682303762193,
+      "grad_norm": 0.2342264218537004,
+      "learning_rate": 4.292598967297763e-05,
+      "loss": 0.4006,
+      "step": 735
+    },
+    {
+      "epoch": 0.683697166744078,
+      "grad_norm": 0.2534426133410622,
+      "learning_rate": 4.2908777969018934e-05,
+      "loss": 0.4127,
+      "step": 736
+    },
+    {
+      "epoch": 0.6846261031119368,
+      "grad_norm": 0.2711574107583699,
+      "learning_rate": 4.2891566265060246e-05,
+      "loss": 0.4334,
+      "step": 737
+    },
+    {
+      "epoch": 0.6855550394797957,
+      "grad_norm": 0.31772185693408067,
+      "learning_rate": 4.287435456110155e-05,
+      "loss": 0.4187,
+      "step": 738
+    },
+    {
+      "epoch": 0.6864839758476544,
+      "grad_norm": 0.30536491271045085,
+      "learning_rate": 4.2857142857142856e-05,
+      "loss": 0.4215,
+      "step": 739
+    },
+    {
+      "epoch": 0.6874129122155133,
+      "grad_norm": 0.33006635156656183,
+      "learning_rate": 4.283993115318417e-05,
+      "loss": 0.405,
+      "step": 740
+    },
+    {
+      "epoch": 0.688341848583372,
+      "grad_norm": 0.2964099136976024,
+      "learning_rate": 4.282271944922547e-05,
+      "loss": 0.419,
+      "step": 741
+    },
+    {
+      "epoch": 0.6892707849512308,
+      "grad_norm": 0.2931220669174559,
+      "learning_rate": 4.2805507745266785e-05,
+      "loss": 0.4422,
+      "step": 742
+    },
+    {
+      "epoch": 0.6901997213190897,
+      "grad_norm": 0.2983381792229504,
+      "learning_rate": 4.278829604130809e-05,
+      "loss": 0.4411,
+      "step": 743
+    },
+    {
+      "epoch": 0.6911286576869484,
+      "grad_norm": 0.2648675526907866,
+      "learning_rate": 4.27710843373494e-05,
+      "loss": 0.4429,
+      "step": 744
+    },
+    {
+      "epoch": 0.6920575940548073,
+      "grad_norm": 0.2602456162921885,
+      "learning_rate": 4.275387263339071e-05,
+      "loss": 0.4198,
+      "step": 745
+    },
+    {
+      "epoch": 0.692986530422666,
+      "grad_norm": 0.23353204937477282,
+      "learning_rate": 4.273666092943202e-05,
+      "loss": 0.4028,
+      "step": 746
+    },
+    {
+      "epoch": 0.6939154667905248,
+      "grad_norm": 0.2991360217695013,
+      "learning_rate": 4.2719449225473323e-05,
+      "loss": 0.4143,
+      "step": 747
+    },
+    {
+      "epoch": 0.6948444031583837,
+      "grad_norm": 0.25702159816645936,
+      "learning_rate": 4.2702237521514635e-05,
+      "loss": 0.4213,
+      "step": 748
+    },
+    {
+      "epoch": 0.6957733395262424,
+      "grad_norm": 0.2838721908179337,
+      "learning_rate": 4.268502581755594e-05,
+      "loss": 0.4387,
+      "step": 749
+    },
+    {
+      "epoch": 0.6967022758941013,
+      "grad_norm": 0.24149559875694418,
+      "learning_rate": 4.2667814113597245e-05,
+      "loss": 0.4184,
+      "step": 750
+    },
+    {
+      "epoch": 0.69763121226196,
+      "grad_norm": 0.2952086958075859,
+      "learning_rate": 4.265060240963856e-05,
+      "loss": 0.4167,
+      "step": 751
+    },
+    {
+      "epoch": 0.6985601486298189,
+      "grad_norm": 0.23972843797863336,
+      "learning_rate": 4.263339070567986e-05,
+      "loss": 0.4141,
+      "step": 752
+    },
+    {
+      "epoch": 0.6994890849976777,
+      "grad_norm": 0.2746023756298842,
+      "learning_rate": 4.2616179001721174e-05,
+      "loss": 0.4364,
+      "step": 753
+    },
+    {
+      "epoch": 0.7004180213655364,
+      "grad_norm": 0.29765586671383076,
+      "learning_rate": 4.259896729776248e-05,
+      "loss": 0.4089,
+      "step": 754
+    },
+    {
+      "epoch": 0.7013469577333953,
+      "grad_norm": 0.21800684863628034,
+      "learning_rate": 4.258175559380379e-05,
+      "loss": 0.4143,
+      "step": 755
+    },
+    {
+      "epoch": 0.702275894101254,
+      "grad_norm": 0.28610650003068566,
+      "learning_rate": 4.2564543889845096e-05,
+      "loss": 0.4138,
+      "step": 756
+    },
+    {
+      "epoch": 0.7032048304691129,
+      "grad_norm": 0.2640665130452579,
+      "learning_rate": 4.254733218588641e-05,
+      "loss": 0.4549,
+      "step": 757
+    },
+    {
+      "epoch": 0.7041337668369717,
+      "grad_norm": 0.25026626980377065,
+      "learning_rate": 4.253012048192771e-05,
+      "loss": 0.4038,
+      "step": 758
+    },
+    {
+      "epoch": 0.7050627032048304,
+      "grad_norm": 0.275452109464895,
+      "learning_rate": 4.251290877796902e-05,
+      "loss": 0.4222,
+      "step": 759
+    },
+    {
+      "epoch": 0.7059916395726893,
+      "grad_norm": 0.25082813720469543,
+      "learning_rate": 4.249569707401033e-05,
+      "loss": 0.414,
+      "step": 760
+    },
+    {
+      "epoch": 0.706920575940548,
+      "grad_norm": 0.2607187023425719,
+      "learning_rate": 4.2478485370051634e-05,
+      "loss": 0.4186,
+      "step": 761
+    },
+    {
+      "epoch": 0.7078495123084069,
+      "grad_norm": 0.2628577880876619,
+      "learning_rate": 4.2461273666092946e-05,
+      "loss": 0.4313,
+      "step": 762
+    },
+    {
+      "epoch": 0.7087784486762657,
+      "grad_norm": 0.256368283425763,
+      "learning_rate": 4.244406196213425e-05,
+      "loss": 0.4371,
+      "step": 763
+    },
+    {
+      "epoch": 0.7097073850441245,
+      "grad_norm": 0.27295950833164523,
+      "learning_rate": 4.242685025817556e-05,
+      "loss": 0.4224,
+      "step": 764
+    },
+    {
+      "epoch": 0.7106363214119833,
+      "grad_norm": 0.2788051820721469,
+      "learning_rate": 4.240963855421687e-05,
+      "loss": 0.4261,
+      "step": 765
+    },
+    {
+      "epoch": 0.711565257779842,
+      "grad_norm": 0.2966812172764707,
+      "learning_rate": 4.239242685025818e-05,
+      "loss": 0.4622,
+      "step": 766
+    },
+    {
+      "epoch": 0.7124941941477009,
+      "grad_norm": 0.26841840271189216,
+      "learning_rate": 4.2375215146299485e-05,
+      "loss": 0.4065,
+      "step": 767
+    },
+    {
+      "epoch": 0.7134231305155597,
+      "grad_norm": 0.262959590522867,
+      "learning_rate": 4.2358003442340797e-05,
+      "loss": 0.3935,
+      "step": 768
+    },
+    {
+      "epoch": 0.7143520668834185,
+      "grad_norm": 0.24938352992988017,
+      "learning_rate": 4.23407917383821e-05,
+      "loss": 0.4231,
+      "step": 769
+    },
+    {
+      "epoch": 0.7152810032512773,
+      "grad_norm": 0.29239937087989554,
+      "learning_rate": 4.232358003442341e-05,
+      "loss": 0.4064,
+      "step": 770
+    },
+    {
+      "epoch": 0.716209939619136,
+      "grad_norm": 0.23536282870494213,
+      "learning_rate": 4.230636833046472e-05,
+      "loss": 0.402,
+      "step": 771
+    },
+    {
+      "epoch": 0.7171388759869949,
+      "grad_norm": 0.2714220152937965,
+      "learning_rate": 4.228915662650603e-05,
+      "loss": 0.3936,
+      "step": 772
+    },
+    {
+      "epoch": 0.7180678123548537,
+      "grad_norm": 0.2747257504638425,
+      "learning_rate": 4.2271944922547335e-05,
+      "loss": 0.4163,
+      "step": 773
+    },
+    {
+      "epoch": 0.7189967487227125,
+      "grad_norm": 0.2723713156748802,
+      "learning_rate": 4.225473321858865e-05,
+      "loss": 0.4066,
+      "step": 774
+    },
+    {
+      "epoch": 0.7199256850905713,
+      "grad_norm": 0.2557791781141218,
+      "learning_rate": 4.223752151462995e-05,
+      "loss": 0.4087,
+      "step": 775
+    },
+    {
+      "epoch": 0.72085462145843,
+      "grad_norm": 0.23079182269430176,
+      "learning_rate": 4.222030981067126e-05,
+      "loss": 0.3846,
+      "step": 776
+    },
+    {
+      "epoch": 0.7217835578262889,
+      "grad_norm": 0.28776961647603927,
+      "learning_rate": 4.220309810671257e-05,
+      "loss": 0.4035,
+      "step": 777
+    },
+    {
+      "epoch": 0.7227124941941477,
+      "grad_norm": 0.25625135388254744,
+      "learning_rate": 4.2185886402753874e-05,
+      "loss": 0.4196,
+      "step": 778
+    },
+    {
+      "epoch": 0.7236414305620065,
+      "grad_norm": 0.29595393487275606,
+      "learning_rate": 4.2168674698795186e-05,
+      "loss": 0.4448,
+      "step": 779
+    },
+    {
+      "epoch": 0.7245703669298653,
+      "grad_norm": 0.24841926756950633,
+      "learning_rate": 4.215146299483649e-05,
+      "loss": 0.4153,
+      "step": 780
+    },
+    {
+      "epoch": 0.725499303297724,
+      "grad_norm": 0.28459325293580745,
+      "learning_rate": 4.2134251290877796e-05,
+      "loss": 0.444,
+      "step": 781
+    },
+    {
+      "epoch": 0.7264282396655829,
+      "grad_norm": 0.26193828352352555,
+      "learning_rate": 4.211703958691911e-05,
+      "loss": 0.4302,
+      "step": 782
+    },
+    {
+      "epoch": 0.7273571760334417,
+      "grad_norm": 0.2479500695456553,
+      "learning_rate": 4.209982788296041e-05,
+      "loss": 0.4073,
+      "step": 783
+    },
+    {
+      "epoch": 0.7282861124013005,
+      "grad_norm": 0.28792928032141457,
+      "learning_rate": 4.2082616179001724e-05,
+      "loss": 0.4006,
+      "step": 784
+    },
+    {
+      "epoch": 0.7292150487691593,
+      "grad_norm": 0.22643696354988968,
+      "learning_rate": 4.206540447504303e-05,
+      "loss": 0.4186,
+      "step": 785
+    },
+    {
+      "epoch": 0.7301439851370181,
+      "grad_norm": 0.30992055940652685,
+      "learning_rate": 4.204819277108434e-05,
+      "loss": 0.4013,
+      "step": 786
+    },
+    {
+      "epoch": 0.7310729215048769,
+      "grad_norm": 0.22983459619899754,
+      "learning_rate": 4.2030981067125646e-05,
+      "loss": 0.4181,
+      "step": 787
+    },
+    {
+      "epoch": 0.7320018578727358,
+      "grad_norm": 0.25635099005397793,
+      "learning_rate": 4.201376936316696e-05,
+      "loss": 0.4099,
+      "step": 788
+    },
+    {
+      "epoch": 0.7329307942405945,
+      "grad_norm": 0.2562041424665725,
+      "learning_rate": 4.199655765920826e-05,
+      "loss": 0.4108,
+      "step": 789
+    },
+    {
+      "epoch": 0.7338597306084533,
+      "grad_norm": 0.25150699905041096,
+      "learning_rate": 4.1979345955249575e-05,
+      "loss": 0.4085,
+      "step": 790
+    },
+    {
+      "epoch": 0.7347886669763122,
+      "grad_norm": 0.2714063520253769,
+      "learning_rate": 4.196213425129088e-05,
+      "loss": 0.4214,
+      "step": 791
+    },
+    {
+      "epoch": 0.7357176033441709,
+      "grad_norm": 0.2665920594786692,
+      "learning_rate": 4.194492254733219e-05,
+      "loss": 0.4199,
+      "step": 792
+    },
+    {
+      "epoch": 0.7366465397120298,
+      "grad_norm": 0.2671300207460533,
+      "learning_rate": 4.1927710843373496e-05,
+      "loss": 0.4161,
+      "step": 793
+    },
+    {
+      "epoch": 0.7375754760798885,
+      "grad_norm": 0.27034859657353083,
+      "learning_rate": 4.191049913941481e-05,
+      "loss": 0.4271,
+      "step": 794
+    },
+    {
+      "epoch": 0.7385044124477473,
+      "grad_norm": 0.26690699071818325,
+      "learning_rate": 4.189328743545611e-05,
+      "loss": 0.4192,
+      "step": 795
+    },
+    {
+      "epoch": 0.7394333488156062,
+      "grad_norm": 0.3051440328850412,
+      "learning_rate": 4.1876075731497425e-05,
+      "loss": 0.4132,
+      "step": 796
+    },
+    {
+      "epoch": 0.7403622851834649,
+      "grad_norm": 0.21974143916170535,
+      "learning_rate": 4.185886402753873e-05,
+      "loss": 0.3979,
+      "step": 797
+    },
+    {
+      "epoch": 0.7412912215513238,
+      "grad_norm": 0.24983461807683474,
+      "learning_rate": 4.184165232358004e-05,
+      "loss": 0.4079,
+      "step": 798
+    },
+    {
+      "epoch": 0.7422201579191825,
+      "grad_norm": 0.24224059591231967,
+      "learning_rate": 4.182444061962135e-05,
+      "loss": 0.4032,
+      "step": 799
+    },
+    {
+      "epoch": 0.7431490942870413,
+      "grad_norm": 0.2502956003038763,
+      "learning_rate": 4.180722891566265e-05,
+      "loss": 0.409,
+      "step": 800
+    },
+    {
+      "epoch": 0.7440780306549002,
+      "grad_norm": 0.24589196452215084,
+      "learning_rate": 4.1790017211703964e-05,
+      "loss": 0.4111,
+      "step": 801
+    },
+    {
+      "epoch": 0.7450069670227589,
+      "grad_norm": 0.22442063047390312,
+      "learning_rate": 4.177280550774527e-05,
+      "loss": 0.4212,
+      "step": 802
+    },
+    {
+      "epoch": 0.7459359033906178,
+      "grad_norm": 0.2360026228130957,
+      "learning_rate": 4.1755593803786574e-05,
+      "loss": 0.3978,
+      "step": 803
+    },
+    {
+      "epoch": 0.7468648397584765,
+      "grad_norm": 0.26711637554464907,
+      "learning_rate": 4.1738382099827885e-05,
+      "loss": 0.4028,
+      "step": 804
+    },
+    {
+      "epoch": 0.7477937761263354,
+      "grad_norm": 0.2956104037100123,
+      "learning_rate": 4.172117039586919e-05,
+      "loss": 0.417,
+      "step": 805
+    },
+    {
+      "epoch": 0.7487227124941942,
+      "grad_norm": 0.2764502588713264,
+      "learning_rate": 4.17039586919105e-05,
+      "loss": 0.4075,
+      "step": 806
+    },
+    {
+      "epoch": 0.7496516488620529,
+      "grad_norm": 0.26416939396871475,
+      "learning_rate": 4.168674698795181e-05,
+      "loss": 0.4306,
+      "step": 807
+    },
+    {
+      "epoch": 0.7505805852299118,
+      "grad_norm": 0.29657741806318,
+      "learning_rate": 4.166953528399312e-05,
+      "loss": 0.4213,
+      "step": 808
+    },
+    {
+      "epoch": 0.7515095215977705,
+      "grad_norm": 0.262465514495409,
+      "learning_rate": 4.1652323580034424e-05,
+      "loss": 0.4211,
+      "step": 809
+    },
+    {
+      "epoch": 0.7524384579656294,
+      "grad_norm": 0.29476951737672735,
+      "learning_rate": 4.1635111876075736e-05,
+      "loss": 0.4163,
+      "step": 810
+    },
+    {
+      "epoch": 0.7533673943334882,
+      "grad_norm": 0.25271095485048567,
+      "learning_rate": 4.161790017211704e-05,
+      "loss": 0.4064,
+      "step": 811
+    },
+    {
+      "epoch": 0.7542963307013469,
+      "grad_norm": 0.30865411128822723,
+      "learning_rate": 4.160068846815835e-05,
+      "loss": 0.4151,
+      "step": 812
+    },
+    {
+      "epoch": 0.7552252670692058,
+      "grad_norm": 0.24802250876527565,
+      "learning_rate": 4.158347676419966e-05,
+      "loss": 0.4445,
+      "step": 813
+    },
+    {
+      "epoch": 0.7561542034370645,
+      "grad_norm": 0.3701424649307489,
+      "learning_rate": 4.156626506024097e-05,
+      "loss": 0.4183,
+      "step": 814
+    },
+    {
+      "epoch": 0.7570831398049234,
+      "grad_norm": 0.2337360342099123,
+      "learning_rate": 4.1549053356282274e-05,
+      "loss": 0.3858,
+      "step": 815
+    },
+    {
+      "epoch": 0.7580120761727822,
+      "grad_norm": 0.2910522135380579,
+      "learning_rate": 4.1531841652323586e-05,
+      "loss": 0.4122,
+      "step": 816
+    },
+    {
+      "epoch": 0.758941012540641,
+      "grad_norm": 0.236795985808718,
+      "learning_rate": 4.151462994836489e-05,
+      "loss": 0.4322,
+      "step": 817
+    },
+    {
+      "epoch": 0.7598699489084998,
+      "grad_norm": 0.28675347251155037,
+      "learning_rate": 4.14974182444062e-05,
+      "loss": 0.411,
+      "step": 818
+    },
+    {
+      "epoch": 0.7607988852763585,
+      "grad_norm": 0.24135506233109558,
+      "learning_rate": 4.148020654044751e-05,
+      "loss": 0.4033,
+      "step": 819
+    },
+    {
+      "epoch": 0.7617278216442174,
+      "grad_norm": 0.29903357931708696,
+      "learning_rate": 4.146299483648882e-05,
+      "loss": 0.4028,
+      "step": 820
+    },
+    {
+      "epoch": 0.7626567580120762,
+      "grad_norm": 0.2666535342900728,
+      "learning_rate": 4.1445783132530125e-05,
+      "loss": 0.4064,
+      "step": 821
+    },
+    {
+      "epoch": 0.763585694379935,
+      "grad_norm": 0.33097109067206043,
+      "learning_rate": 4.1428571428571437e-05,
+      "loss": 0.4167,
+      "step": 822
+    },
+    {
+      "epoch": 0.7645146307477938,
+      "grad_norm": 0.27705874830384963,
+      "learning_rate": 4.1411359724612735e-05,
+      "loss": 0.4317,
+      "step": 823
+    },
+    {
+      "epoch": 0.7654435671156525,
+      "grad_norm": 0.27006115913006945,
+      "learning_rate": 4.1394148020654047e-05,
+      "loss": 0.4011,
+      "step": 824
+    },
+    {
+      "epoch": 0.7663725034835114,
+      "grad_norm": 0.2191843080256796,
+      "learning_rate": 4.137693631669535e-05,
+      "loss": 0.3877,
+      "step": 825
+    },
+    {
+      "epoch": 0.7673014398513702,
+      "grad_norm": 0.26765533335243286,
+      "learning_rate": 4.135972461273666e-05,
+      "loss": 0.416,
+      "step": 826
+    },
+    {
+      "epoch": 0.768230376219229,
+      "grad_norm": 0.28444860627784857,
+      "learning_rate": 4.134251290877797e-05,
+      "loss": 0.4142,
+      "step": 827
+    },
+    {
+      "epoch": 0.7691593125870878,
+      "grad_norm": 0.2590581112218012,
+      "learning_rate": 4.132530120481928e-05,
+      "loss": 0.455,
+      "step": 828
+    },
+    {
+      "epoch": 0.7700882489549465,
+      "grad_norm": 0.28511612916261636,
+      "learning_rate": 4.1308089500860585e-05,
+      "loss": 0.4439,
+      "step": 829
+    },
+    {
+      "epoch": 0.7710171853228054,
+      "grad_norm": 0.2724311438434457,
+      "learning_rate": 4.12908777969019e-05,
+      "loss": 0.4105,
+      "step": 830
+    },
+    {
+      "epoch": 0.7719461216906642,
+      "grad_norm": 0.23702064984429358,
+      "learning_rate": 4.12736660929432e-05,
+      "loss": 0.4136,
+      "step": 831
+    },
+    {
+      "epoch": 0.772875058058523,
+      "grad_norm": 0.2892255007964407,
+      "learning_rate": 4.1256454388984514e-05,
+      "loss": 0.4089,
+      "step": 832
+    },
+    {
+      "epoch": 0.7738039944263818,
+      "grad_norm": 0.24990575367096712,
+      "learning_rate": 4.123924268502582e-05,
+      "loss": 0.443,
+      "step": 833
+    },
+    {
+      "epoch": 0.7747329307942405,
+      "grad_norm": 0.3184046525164543,
+      "learning_rate": 4.122203098106713e-05,
+      "loss": 0.4318,
+      "step": 834
+    },
+    {
+      "epoch": 0.7756618671620994,
+      "grad_norm": 0.2673099414060095,
+      "learning_rate": 4.1204819277108436e-05,
+      "loss": 0.4013,
+      "step": 835
+    },
+    {
+      "epoch": 0.7765908035299582,
+      "grad_norm": 0.2925594820844854,
+      "learning_rate": 4.118760757314975e-05,
+      "loss": 0.4271,
+      "step": 836
+    },
+    {
+      "epoch": 0.777519739897817,
+      "grad_norm": 0.22953985012166775,
+      "learning_rate": 4.117039586919105e-05,
+      "loss": 0.3745,
+      "step": 837
+    },
+    {
+      "epoch": 0.7784486762656758,
+      "grad_norm": 0.33609789574689314,
+      "learning_rate": 4.1153184165232364e-05,
+      "loss": 0.4037,
+      "step": 838
+    },
+    {
+      "epoch": 0.7793776126335346,
+      "grad_norm": 0.24182783755842846,
+      "learning_rate": 4.113597246127367e-05,
+      "loss": 0.3996,
+      "step": 839
+    },
+    {
+      "epoch": 0.7803065490013934,
+      "grad_norm": 0.24661825820680278,
+      "learning_rate": 4.111876075731498e-05,
+      "loss": 0.4,
+      "step": 840
+    },
+    {
+      "epoch": 0.7812354853692522,
+      "grad_norm": 0.2676051683765745,
+      "learning_rate": 4.1101549053356286e-05,
+      "loss": 0.4341,
+      "step": 841
+    },
+    {
+      "epoch": 0.782164421737111,
+      "grad_norm": 0.2302592197123923,
+      "learning_rate": 4.10843373493976e-05,
+      "loss": 0.3956,
+      "step": 842
+    },
+    {
+      "epoch": 0.7830933581049698,
+      "grad_norm": 0.26234103881814524,
+      "learning_rate": 4.10671256454389e-05,
+      "loss": 0.4011,
+      "step": 843
+    },
+    {
+      "epoch": 0.7840222944728286,
+      "grad_norm": 0.2801794293079212,
+      "learning_rate": 4.104991394148021e-05,
+      "loss": 0.4329,
+      "step": 844
+    },
+    {
+      "epoch": 0.7849512308406874,
+      "grad_norm": 0.23397262641748529,
+      "learning_rate": 4.103270223752151e-05,
+      "loss": 0.3709,
+      "step": 845
+    },
+    {
+      "epoch": 0.7858801672085463,
+      "grad_norm": 0.3544703249382208,
+      "learning_rate": 4.1015490533562825e-05,
+      "loss": 0.4171,
+      "step": 846
+    },
+    {
+      "epoch": 0.786809103576405,
+      "grad_norm": 0.2722275854852537,
+      "learning_rate": 4.099827882960413e-05,
+      "loss": 0.4337,
+      "step": 847
+    },
+    {
+      "epoch": 0.7877380399442638,
+      "grad_norm": 0.26553648559179016,
+      "learning_rate": 4.098106712564544e-05,
+      "loss": 0.4173,
+      "step": 848
+    },
+    {
+      "epoch": 0.7886669763121226,
+      "grad_norm": 0.2678696587952733,
+      "learning_rate": 4.0963855421686746e-05,
+      "loss": 0.4111,
+      "step": 849
+    },
+    {
+      "epoch": 0.7895959126799814,
+      "grad_norm": 0.2657158376139536,
+      "learning_rate": 4.094664371772806e-05,
+      "loss": 0.4126,
+      "step": 850
+    },
+    {
+      "epoch": 0.7905248490478403,
+      "grad_norm": 0.287284238795583,
+      "learning_rate": 4.092943201376936e-05,
+      "loss": 0.4207,
+      "step": 851
+    },
+    {
+      "epoch": 0.791453785415699,
+      "grad_norm": 0.2754176953959211,
+      "learning_rate": 4.0912220309810675e-05,
+      "loss": 0.4259,
+      "step": 852
+    },
+    {
+      "epoch": 0.7923827217835578,
+      "grad_norm": 0.262107512607669,
+      "learning_rate": 4.089500860585198e-05,
+      "loss": 0.4127,
+      "step": 853
+    },
+    {
+      "epoch": 0.7933116581514166,
+      "grad_norm": 0.24162465041652548,
+      "learning_rate": 4.087779690189329e-05,
+      "loss": 0.4009,
+      "step": 854
+    },
+    {
+      "epoch": 0.7942405945192754,
+      "grad_norm": 0.266448063011791,
+      "learning_rate": 4.08605851979346e-05,
+      "loss": 0.4271,
+      "step": 855
+    },
+    {
+      "epoch": 0.7951695308871343,
+      "grad_norm": 0.24051279422477578,
+      "learning_rate": 4.084337349397591e-05,
+      "loss": 0.4089,
+      "step": 856
+    },
+    {
+      "epoch": 0.796098467254993,
+      "grad_norm": 0.2729628116438206,
+      "learning_rate": 4.0826161790017214e-05,
+      "loss": 0.4167,
+      "step": 857
+    },
+    {
+      "epoch": 0.7970274036228518,
+      "grad_norm": 0.23410719385943532,
+      "learning_rate": 4.0808950086058525e-05,
+      "loss": 0.4179,
+      "step": 858
+    },
+    {
+      "epoch": 0.7979563399907106,
+      "grad_norm": 0.30219591202715035,
+      "learning_rate": 4.079173838209983e-05,
+      "loss": 0.4138,
+      "step": 859
+    },
+    {
+      "epoch": 0.7988852763585694,
+      "grad_norm": 0.23985576065241604,
+      "learning_rate": 4.077452667814114e-05,
+      "loss": 0.4075,
+      "step": 860
+    },
+    {
+      "epoch": 0.7998142127264283,
+      "grad_norm": 0.25029357115071,
+      "learning_rate": 4.075731497418245e-05,
+      "loss": 0.4118,
+      "step": 861
+    },
+    {
+      "epoch": 0.800743149094287,
+      "grad_norm": 0.29345741652304597,
+      "learning_rate": 4.074010327022376e-05,
+      "loss": 0.397,
+      "step": 862
+    },
+    {
+      "epoch": 0.8016720854621459,
+      "grad_norm": 0.24385532620239805,
+      "learning_rate": 4.0722891566265064e-05,
+      "loss": 0.4,
+      "step": 863
+    },
+    {
+      "epoch": 0.8026010218300046,
+      "grad_norm": 0.2637648833416592,
+      "learning_rate": 4.0705679862306376e-05,
+      "loss": 0.4219,
+      "step": 864
+    },
+    {
+      "epoch": 0.8035299581978634,
+      "grad_norm": 0.25527572185866787,
+      "learning_rate": 4.0688468158347674e-05,
+      "loss": 0.4007,
+      "step": 865
+    },
+    {
+      "epoch": 0.8044588945657223,
+      "grad_norm": 0.2513201236957386,
+      "learning_rate": 4.0671256454388986e-05,
+      "loss": 0.3955,
+      "step": 866
+    },
+    {
+      "epoch": 0.805387830933581,
+      "grad_norm": 0.25660774924232826,
+      "learning_rate": 4.065404475043029e-05,
+      "loss": 0.3787,
+      "step": 867
+    },
+    {
+      "epoch": 0.8063167673014399,
+      "grad_norm": 0.26206437729581916,
+      "learning_rate": 4.06368330464716e-05,
+      "loss": 0.4196,
+      "step": 868
+    },
+    {
+      "epoch": 0.8072457036692986,
+      "grad_norm": 0.24835895191817717,
+      "learning_rate": 4.061962134251291e-05,
+      "loss": 0.3994,
+      "step": 869
+    },
+    {
+      "epoch": 0.8081746400371574,
+      "grad_norm": 0.26426141442192896,
+      "learning_rate": 4.060240963855422e-05,
+      "loss": 0.4254,
+      "step": 870
+    },
+    {
+      "epoch": 0.8091035764050163,
+      "grad_norm": 0.24819792871771582,
+      "learning_rate": 4.0585197934595524e-05,
+      "loss": 0.4124,
+      "step": 871
+    },
+    {
+      "epoch": 0.810032512772875,
+      "grad_norm": 0.2444630039170481,
+      "learning_rate": 4.0567986230636836e-05,
+      "loss": 0.4374,
+      "step": 872
+    },
+    {
+      "epoch": 0.8109614491407339,
+      "grad_norm": 0.25882143197020263,
+      "learning_rate": 4.055077452667814e-05,
+      "loss": 0.4224,
+      "step": 873
+    },
+    {
+      "epoch": 0.8118903855085927,
+      "grad_norm": 0.22956628354423742,
+      "learning_rate": 4.053356282271945e-05,
+      "loss": 0.4014,
+      "step": 874
+    },
+    {
+      "epoch": 0.8128193218764515,
+      "grad_norm": 0.264956195080769,
+      "learning_rate": 4.051635111876076e-05,
+      "loss": 0.3938,
+      "step": 875
+    },
+    {
+      "epoch": 0.8137482582443103,
+      "grad_norm": 0.24985343436278015,
+      "learning_rate": 4.049913941480207e-05,
+      "loss": 0.4223,
+      "step": 876
+    },
+    {
+      "epoch": 0.814677194612169,
+      "grad_norm": 0.2864398731906139,
+      "learning_rate": 4.0481927710843375e-05,
+      "loss": 0.3934,
+      "step": 877
+    },
+    {
+      "epoch": 0.8156061309800279,
+      "grad_norm": 0.24292187852748645,
+      "learning_rate": 4.046471600688469e-05,
+      "loss": 0.3944,
+      "step": 878
+    },
+    {
+      "epoch": 0.8165350673478867,
+      "grad_norm": 0.2860359553745768,
+      "learning_rate": 4.044750430292599e-05,
+      "loss": 0.4127,
+      "step": 879
+    },
+    {
+      "epoch": 0.8174640037157455,
+      "grad_norm": 0.29668643281017715,
+      "learning_rate": 4.0430292598967303e-05,
+      "loss": 0.3998,
+      "step": 880
+    },
+    {
+      "epoch": 0.8183929400836043,
+      "grad_norm": 0.278439860888243,
+      "learning_rate": 4.041308089500861e-05,
+      "loss": 0.3744,
+      "step": 881
+    },
+    {
+      "epoch": 0.819321876451463,
+      "grad_norm": 0.2763340216330821,
+      "learning_rate": 4.039586919104992e-05,
+      "loss": 0.4046,
+      "step": 882
+    },
+    {
+      "epoch": 0.8202508128193219,
+      "grad_norm": 0.2578084168559668,
+      "learning_rate": 4.0378657487091225e-05,
+      "loss": 0.4144,
+      "step": 883
+    },
+    {
+      "epoch": 0.8211797491871807,
+      "grad_norm": 0.2704411317528433,
+      "learning_rate": 4.036144578313254e-05,
+      "loss": 0.3938,
+      "step": 884
+    },
+    {
+      "epoch": 0.8221086855550395,
+      "grad_norm": 0.2216930759924396,
+      "learning_rate": 4.034423407917384e-05,
+      "loss": 0.3976,
+      "step": 885
+    },
+    {
+      "epoch": 0.8230376219228983,
+      "grad_norm": 0.23736907014407435,
+      "learning_rate": 4.032702237521515e-05,
+      "loss": 0.4032,
+      "step": 886
+    },
+    {
+      "epoch": 0.823966558290757,
+      "grad_norm": 0.2751484069936401,
+      "learning_rate": 4.030981067125645e-05,
+      "loss": 0.4142,
+      "step": 887
+    },
+    {
+      "epoch": 0.8248954946586159,
+      "grad_norm": 0.22573606403051907,
+      "learning_rate": 4.0292598967297764e-05,
+      "loss": 0.3922,
+      "step": 888
+    },
+    {
+      "epoch": 0.8258244310264747,
+      "grad_norm": 0.2602795860148468,
+      "learning_rate": 4.027538726333907e-05,
+      "loss": 0.4139,
+      "step": 889
+    },
+    {
+      "epoch": 0.8267533673943335,
+      "grad_norm": 0.24581668806400517,
+      "learning_rate": 4.025817555938038e-05,
+      "loss": 0.4173,
+      "step": 890
+    },
+    {
+      "epoch": 0.8276823037621923,
+      "grad_norm": 0.23586598943899656,
+      "learning_rate": 4.0240963855421686e-05,
+      "loss": 0.397,
+      "step": 891
+    },
+    {
+      "epoch": 0.828611240130051,
+      "grad_norm": 0.22698815324965912,
+      "learning_rate": 4.0223752151463e-05,
+      "loss": 0.397,
+      "step": 892
+    },
+    {
+      "epoch": 0.8295401764979099,
+      "grad_norm": 0.251756454856816,
+      "learning_rate": 4.02065404475043e-05,
+      "loss": 0.43,
+      "step": 893
+    },
+    {
+      "epoch": 0.8304691128657687,
+      "grad_norm": 0.23474799972973084,
+      "learning_rate": 4.0189328743545614e-05,
+      "loss": 0.4232,
+      "step": 894
+    },
+    {
+      "epoch": 0.8313980492336275,
+      "grad_norm": 0.2284304060585405,
+      "learning_rate": 4.017211703958692e-05,
+      "loss": 0.4363,
+      "step": 895
+    },
+    {
+      "epoch": 0.8323269856014863,
+      "grad_norm": 0.22189967592933565,
+      "learning_rate": 4.015490533562823e-05,
+      "loss": 0.3996,
+      "step": 896
+    },
+    {
+      "epoch": 0.8332559219693451,
+      "grad_norm": 0.24727551944573103,
+      "learning_rate": 4.0137693631669536e-05,
+      "loss": 0.3881,
+      "step": 897
+    },
+    {
+      "epoch": 0.8341848583372039,
+      "grad_norm": 0.26502357788364,
+      "learning_rate": 4.012048192771085e-05,
+      "loss": 0.4345,
+      "step": 898
+    },
+    {
+      "epoch": 0.8351137947050628,
+      "grad_norm": 0.24506267621395414,
+      "learning_rate": 4.010327022375215e-05,
+      "loss": 0.408,
+      "step": 899
+    },
+    {
+      "epoch": 0.8360427310729215,
+      "grad_norm": 0.2686607976921608,
+      "learning_rate": 4.0086058519793465e-05,
+      "loss": 0.4284,
+      "step": 900
+    },
+    {
+      "epoch": 0.8369716674407803,
+      "grad_norm": 0.22658254041478237,
+      "learning_rate": 4.006884681583477e-05,
+      "loss": 0.4036,
+      "step": 901
+    },
+    {
+      "epoch": 0.8379006038086391,
+      "grad_norm": 0.24269121548237235,
+      "learning_rate": 4.005163511187608e-05,
+      "loss": 0.4158,
+      "step": 902
+    },
+    {
+      "epoch": 0.8388295401764979,
+      "grad_norm": 0.27624389649170866,
+      "learning_rate": 4.0034423407917386e-05,
+      "loss": 0.4067,
+      "step": 903
+    },
+    {
+      "epoch": 0.8397584765443568,
+      "grad_norm": 0.24834115738955598,
+      "learning_rate": 4.00172117039587e-05,
+      "loss": 0.4068,
+      "step": 904
+    },
+    {
+      "epoch": 0.8406874129122155,
+      "grad_norm": 0.25359081776709796,
+      "learning_rate": 4e-05,
+      "loss": 0.4013,
+      "step": 905
+    },
+    {
+      "epoch": 0.8416163492800743,
+      "grad_norm": 0.2621593127454798,
+      "learning_rate": 3.9982788296041315e-05,
+      "loss": 0.4019,
+      "step": 906
+    },
+    {
+      "epoch": 0.8425452856479331,
+      "grad_norm": 0.22773835872678272,
+      "learning_rate": 3.996557659208262e-05,
+      "loss": 0.3898,
+      "step": 907
+    },
+    {
+      "epoch": 0.8434742220157919,
+      "grad_norm": 0.25454356735572764,
+      "learning_rate": 3.9948364888123925e-05,
+      "loss": 0.4041,
+      "step": 908
+    },
+    {
+      "epoch": 0.8444031583836508,
+      "grad_norm": 0.24027047666859136,
+      "learning_rate": 3.993115318416523e-05,
+      "loss": 0.4158,
+      "step": 909
+    },
+    {
+      "epoch": 0.8453320947515095,
+      "grad_norm": 0.2485921595211011,
+      "learning_rate": 3.991394148020654e-05,
+      "loss": 0.4101,
+      "step": 910
+    },
+    {
+      "epoch": 0.8462610311193683,
+      "grad_norm": 0.22285216317499973,
+      "learning_rate": 3.989672977624785e-05,
+      "loss": 0.3969,
+      "step": 911
+    },
+    {
+      "epoch": 0.8471899674872271,
+      "grad_norm": 0.2596638733050711,
+      "learning_rate": 3.987951807228916e-05,
+      "loss": 0.42,
+      "step": 912
+    },
+    {
+      "epoch": 0.8481189038550859,
+      "grad_norm": 0.23550833352781333,
+      "learning_rate": 3.9862306368330464e-05,
+      "loss": 0.4136,
+      "step": 913
+    },
+    {
+      "epoch": 0.8490478402229448,
+      "grad_norm": 0.24273175946950082,
+      "learning_rate": 3.9845094664371775e-05,
+      "loss": 0.4347,
+      "step": 914
+    },
+    {
+      "epoch": 0.8499767765908035,
+      "grad_norm": 0.2571365495317302,
+      "learning_rate": 3.982788296041308e-05,
+      "loss": 0.4322,
+      "step": 915
+    },
+    {
+      "epoch": 0.8509057129586624,
+      "grad_norm": 0.22272682206410346,
+      "learning_rate": 3.981067125645439e-05,
+      "loss": 0.382,
+      "step": 916
+    },
+    {
+      "epoch": 0.8518346493265211,
+      "grad_norm": 0.22974977779029027,
+      "learning_rate": 3.97934595524957e-05,
+      "loss": 0.3874,
+      "step": 917
+    },
+    {
+      "epoch": 0.8527635856943799,
+      "grad_norm": 0.23626085060684773,
+      "learning_rate": 3.977624784853701e-05,
+      "loss": 0.4205,
+      "step": 918
+    },
+    {
+      "epoch": 0.8536925220622388,
+      "grad_norm": 0.2711925943505794,
+      "learning_rate": 3.9759036144578314e-05,
+      "loss": 0.4048,
+      "step": 919
+    },
+    {
+      "epoch": 0.8546214584300975,
+      "grad_norm": 0.23373429510160257,
+      "learning_rate": 3.9741824440619626e-05,
+      "loss": 0.4183,
+      "step": 920
+    },
+    {
+      "epoch": 0.8555503947979564,
+      "grad_norm": 0.269906941097679,
+      "learning_rate": 3.972461273666093e-05,
+      "loss": 0.4187,
+      "step": 921
+    },
+    {
+      "epoch": 0.8564793311658151,
+      "grad_norm": 0.21569823380266548,
+      "learning_rate": 3.970740103270224e-05,
+      "loss": 0.3984,
+      "step": 922
+    },
+    {
+      "epoch": 0.8574082675336739,
+      "grad_norm": 0.23805027630937817,
+      "learning_rate": 3.969018932874355e-05,
+      "loss": 0.4166,
+      "step": 923
+    },
+    {
+      "epoch": 0.8583372039015328,
+      "grad_norm": 0.23069245503461422,
+      "learning_rate": 3.967297762478486e-05,
+      "loss": 0.4262,
+      "step": 924
+    },
+    {
+      "epoch": 0.8592661402693915,
+      "grad_norm": 0.2428482973970677,
+      "learning_rate": 3.9655765920826164e-05,
+      "loss": 0.4168,
+      "step": 925
+    },
+    {
+      "epoch": 0.8601950766372504,
+      "grad_norm": 0.2183870165831691,
+      "learning_rate": 3.9638554216867476e-05,
+      "loss": 0.4051,
+      "step": 926
+    },
+    {
+      "epoch": 0.8611240130051091,
+      "grad_norm": 0.25786863909064817,
+      "learning_rate": 3.962134251290878e-05,
+      "loss": 0.4024,
+      "step": 927
+    },
+    {
+      "epoch": 0.862052949372968,
+      "grad_norm": 0.21274681020345573,
+      "learning_rate": 3.960413080895009e-05,
+      "loss": 0.4056,
+      "step": 928
+    },
+    {
+      "epoch": 0.8629818857408268,
+      "grad_norm": 0.2805951596096335,
+      "learning_rate": 3.958691910499139e-05,
+      "loss": 0.4251,
+      "step": 929
+    },
+    {
+      "epoch": 0.8639108221086855,
+      "grad_norm": 0.22683405969287646,
+      "learning_rate": 3.95697074010327e-05,
+      "loss": 0.4217,
+      "step": 930
+    },
+    {
+      "epoch": 0.8648397584765444,
+      "grad_norm": 0.26021009912177695,
+      "learning_rate": 3.955249569707401e-05,
+      "loss": 0.4219,
+      "step": 931
+    },
+    {
+      "epoch": 0.8657686948444031,
+      "grad_norm": 0.24224641967846924,
+      "learning_rate": 3.953528399311532e-05,
+      "loss": 0.424,
+      "step": 932
+    },
+    {
+      "epoch": 0.866697631212262,
+      "grad_norm": 0.22247989096036871,
+      "learning_rate": 3.9518072289156625e-05,
+      "loss": 0.3968,
+      "step": 933
+    },
+    {
+      "epoch": 0.8676265675801208,
+      "grad_norm": 0.22730388207221008,
+      "learning_rate": 3.950086058519794e-05,
+      "loss": 0.4027,
+      "step": 934
+    },
+    {
+      "epoch": 0.8685555039479795,
+      "grad_norm": 0.22984491491807468,
+      "learning_rate": 3.948364888123924e-05,
+      "loss": 0.4015,
+      "step": 935
+    },
+    {
+      "epoch": 0.8694844403158384,
+      "grad_norm": 0.2706739171118266,
+      "learning_rate": 3.9466437177280554e-05,
+      "loss": 0.4057,
+      "step": 936
+    },
+    {
+      "epoch": 0.8704133766836971,
+      "grad_norm": 0.2869836202377347,
+      "learning_rate": 3.944922547332186e-05,
+      "loss": 0.4169,
+      "step": 937
+    },
+    {
+      "epoch": 0.871342313051556,
+      "grad_norm": 0.24144907861227421,
+      "learning_rate": 3.943201376936317e-05,
+      "loss": 0.3895,
+      "step": 938
+    },
+    {
+      "epoch": 0.8722712494194148,
+      "grad_norm": 0.2318365608955253,
+      "learning_rate": 3.9414802065404475e-05,
+      "loss": 0.3887,
+      "step": 939
+    },
+    {
+      "epoch": 0.8732001857872735,
+      "grad_norm": 0.24391204560861726,
+      "learning_rate": 3.939759036144579e-05,
+      "loss": 0.4144,
+      "step": 940
+    },
+    {
+      "epoch": 0.8741291221551324,
+      "grad_norm": 0.24279276074096975,
+      "learning_rate": 3.938037865748709e-05,
+      "loss": 0.4105,
+      "step": 941
+    },
+    {
+      "epoch": 0.8750580585229911,
+      "grad_norm": 0.2206737590128525,
+      "learning_rate": 3.9363166953528404e-05,
+      "loss": 0.4055,
+      "step": 942
+    },
+    {
+      "epoch": 0.87598699489085,
+      "grad_norm": 0.2501829527519811,
+      "learning_rate": 3.934595524956971e-05,
+      "loss": 0.419,
+      "step": 943
+    },
+    {
+      "epoch": 0.8769159312587088,
+      "grad_norm": 0.20113559459810643,
+      "learning_rate": 3.932874354561102e-05,
+      "loss": 0.3995,
+      "step": 944
+    },
+    {
+      "epoch": 0.8778448676265675,
+      "grad_norm": 0.2590015788880731,
+      "learning_rate": 3.9311531841652326e-05,
+      "loss": 0.4106,
+      "step": 945
+    },
+    {
+      "epoch": 0.8787738039944264,
+      "grad_norm": 0.21875502199045596,
+      "learning_rate": 3.929432013769364e-05,
+      "loss": 0.4137,
+      "step": 946
+    },
+    {
+      "epoch": 0.8797027403622851,
+      "grad_norm": 0.22619403593853593,
+      "learning_rate": 3.927710843373494e-05,
+      "loss": 0.3897,
+      "step": 947
+    },
+    {
+      "epoch": 0.880631676730144,
+      "grad_norm": 0.28274876745197,
+      "learning_rate": 3.9259896729776254e-05,
+      "loss": 0.4243,
+      "step": 948
+    },
+    {
+      "epoch": 0.8815606130980028,
+      "grad_norm": 0.2550591340833801,
+      "learning_rate": 3.924268502581756e-05,
+      "loss": 0.4262,
+      "step": 949
+    },
+    {
+      "epoch": 0.8824895494658616,
+      "grad_norm": 0.28300258572852405,
+      "learning_rate": 3.9225473321858864e-05,
+      "loss": 0.3981,
+      "step": 950
+    },
+    {
+      "epoch": 0.8834184858337204,
+      "grad_norm": 0.2651505216641192,
+      "learning_rate": 3.920826161790017e-05,
+      "loss": 0.3987,
+      "step": 951
+    },
+    {
+      "epoch": 0.8843474222015791,
+      "grad_norm": 0.2699525095302782,
+      "learning_rate": 3.919104991394148e-05,
+      "loss": 0.3963,
+      "step": 952
+    },
+    {
+      "epoch": 0.885276358569438,
+      "grad_norm": 0.24632775459420023,
+      "learning_rate": 3.9173838209982786e-05,
+      "loss": 0.3793,
+      "step": 953
+    },
+    {
+      "epoch": 0.8862052949372968,
+      "grad_norm": 0.2729975548782007,
+      "learning_rate": 3.91566265060241e-05,
+      "loss": 0.4259,
+      "step": 954
+    },
+    {
+      "epoch": 0.8871342313051556,
+      "grad_norm": 0.23113626254645106,
+      "learning_rate": 3.91394148020654e-05,
+      "loss": 0.4118,
+      "step": 955
+    },
+    {
+      "epoch": 0.8880631676730144,
+      "grad_norm": 0.2493442523902469,
+      "learning_rate": 3.9122203098106715e-05,
+      "loss": 0.4277,
+      "step": 956
+    },
+    {
+      "epoch": 0.8889921040408733,
+      "grad_norm": 0.24785873510336145,
+      "learning_rate": 3.910499139414802e-05,
+      "loss": 0.4065,
+      "step": 957
+    },
+    {
+      "epoch": 0.889921040408732,
+      "grad_norm": 0.24341153477797958,
+      "learning_rate": 3.908777969018933e-05,
+      "loss": 0.4037,
+      "step": 958
+    },
+    {
+      "epoch": 0.8908499767765908,
+      "grad_norm": 0.23299336958401565,
+      "learning_rate": 3.9070567986230637e-05,
+      "loss": 0.4018,
+      "step": 959
+    },
+    {
+      "epoch": 0.8917789131444496,
+      "grad_norm": 0.27717768327110737,
+      "learning_rate": 3.905335628227195e-05,
+      "loss": 0.4136,
+      "step": 960
+    },
+    {
+      "epoch": 0.8927078495123084,
+      "grad_norm": 0.2634600076134501,
+      "learning_rate": 3.903614457831325e-05,
+      "loss": 0.4151,
+      "step": 961
+    },
+    {
+      "epoch": 0.8936367858801673,
+      "grad_norm": 0.2431673757388936,
+      "learning_rate": 3.9018932874354565e-05,
+      "loss": 0.4046,
+      "step": 962
+    },
+    {
+      "epoch": 0.894565722248026,
+      "grad_norm": 0.2873899288522361,
+      "learning_rate": 3.900172117039587e-05,
+      "loss": 0.4029,
+      "step": 963
+    },
+    {
+      "epoch": 0.8954946586158848,
+      "grad_norm": 0.2068486702586044,
+      "learning_rate": 3.898450946643718e-05,
+      "loss": 0.3862,
+      "step": 964
+    },
+    {
+      "epoch": 0.8964235949837436,
+      "grad_norm": 0.21678196583034384,
+      "learning_rate": 3.896729776247849e-05,
+      "loss": 0.3926,
+      "step": 965
+    },
+    {
+      "epoch": 0.8973525313516024,
+      "grad_norm": 0.23384577673120308,
+      "learning_rate": 3.89500860585198e-05,
+      "loss": 0.3986,
+      "step": 966
+    },
+    {
+      "epoch": 0.8982814677194613,
+      "grad_norm": 0.2168835780899871,
+      "learning_rate": 3.8932874354561104e-05,
+      "loss": 0.4117,
+      "step": 967
+    },
+    {
+      "epoch": 0.89921040408732,
+      "grad_norm": 0.2182530040951482,
+      "learning_rate": 3.8915662650602416e-05,
+      "loss": 0.3993,
+      "step": 968
+    },
+    {
+      "epoch": 0.9001393404551788,
+      "grad_norm": 0.2279085259514714,
+      "learning_rate": 3.889845094664372e-05,
+      "loss": 0.4179,
+      "step": 969
+    },
+    {
+      "epoch": 0.9010682768230376,
+      "grad_norm": 0.2161021312161298,
+      "learning_rate": 3.888123924268503e-05,
+      "loss": 0.4196,
+      "step": 970
+    },
+    {
+      "epoch": 0.9019972131908964,
+      "grad_norm": 0.23665046914396706,
+      "learning_rate": 3.886402753872633e-05,
+      "loss": 0.4151,
+      "step": 971
+    },
+    {
+      "epoch": 0.9029261495587553,
+      "grad_norm": 0.2153174695085264,
+      "learning_rate": 3.884681583476764e-05,
+      "loss": 0.392,
+      "step": 972
+    },
+    {
+      "epoch": 0.903855085926614,
+      "grad_norm": 0.22865056532137162,
+      "learning_rate": 3.882960413080895e-05,
+      "loss": 0.4285,
+      "step": 973
+    },
+    {
+      "epoch": 0.9047840222944729,
+      "grad_norm": 0.24540292705433664,
+      "learning_rate": 3.881239242685026e-05,
+      "loss": 0.4085,
+      "step": 974
+    },
+    {
+      "epoch": 0.9057129586623316,
+      "grad_norm": 0.21512013945160077,
+      "learning_rate": 3.8795180722891564e-05,
+      "loss": 0.4001,
+      "step": 975
+    },
+    {
+      "epoch": 0.9066418950301904,
+      "grad_norm": 0.2543665399130075,
+      "learning_rate": 3.8777969018932876e-05,
+      "loss": 0.4233,
+      "step": 976
+    },
+    {
+      "epoch": 0.9075708313980493,
+      "grad_norm": 0.2604590648729307,
+      "learning_rate": 3.876075731497418e-05,
+      "loss": 0.4315,
+      "step": 977
+    },
+    {
+      "epoch": 0.908499767765908,
+      "grad_norm": 0.2426498446435362,
+      "learning_rate": 3.874354561101549e-05,
+      "loss": 0.4194,
+      "step": 978
+    },
+    {
+      "epoch": 0.9094287041337669,
+      "grad_norm": 0.21692216609161416,
+      "learning_rate": 3.87263339070568e-05,
+      "loss": 0.392,
+      "step": 979
+    },
+    {
+      "epoch": 0.9103576405016256,
+      "grad_norm": 0.24908430824930783,
+      "learning_rate": 3.870912220309811e-05,
+      "loss": 0.3719,
+      "step": 980
+    },
+    {
+      "epoch": 0.9112865768694844,
+      "grad_norm": 0.2159682412651961,
+      "learning_rate": 3.8691910499139415e-05,
+      "loss": 0.3898,
+      "step": 981
+    },
+    {
+      "epoch": 0.9122155132373433,
+      "grad_norm": 0.21127012379911123,
+      "learning_rate": 3.8674698795180726e-05,
+      "loss": 0.4248,
+      "step": 982
+    },
+    {
+      "epoch": 0.913144449605202,
+      "grad_norm": 0.22813635269217872,
+      "learning_rate": 3.865748709122203e-05,
+      "loss": 0.4244,
+      "step": 983
+    },
+    {
+      "epoch": 0.9140733859730609,
+      "grad_norm": 0.23743928654284238,
+      "learning_rate": 3.864027538726334e-05,
+      "loss": 0.4214,
+      "step": 984
+    },
+    {
+      "epoch": 0.9150023223409196,
+      "grad_norm": 0.2186026650640482,
+      "learning_rate": 3.862306368330465e-05,
+      "loss": 0.405,
+      "step": 985
+    },
+    {
+      "epoch": 0.9159312587087785,
+      "grad_norm": 0.257356309199639,
+      "learning_rate": 3.860585197934596e-05,
+      "loss": 0.3917,
+      "step": 986
+    },
+    {
+      "epoch": 0.9168601950766373,
+      "grad_norm": 0.19464718995992264,
+      "learning_rate": 3.8588640275387265e-05,
+      "loss": 0.4049,
+      "step": 987
+    },
+    {
+      "epoch": 0.917789131444496,
+      "grad_norm": 0.25657951632861403,
+      "learning_rate": 3.857142857142858e-05,
+      "loss": 0.4186,
+      "step": 988
+    },
+    {
+      "epoch": 0.9187180678123549,
+      "grad_norm": 0.22465676748227606,
+      "learning_rate": 3.855421686746988e-05,
+      "loss": 0.4345,
+      "step": 989
+    },
+    {
+      "epoch": 0.9196470041802136,
+      "grad_norm": 0.2506754334740293,
+      "learning_rate": 3.8537005163511194e-05,
+      "loss": 0.4025,
+      "step": 990
+    },
+    {
+      "epoch": 0.9205759405480725,
+      "grad_norm": 0.2340373514090616,
+      "learning_rate": 3.85197934595525e-05,
+      "loss": 0.4256,
+      "step": 991
+    },
+    {
+      "epoch": 0.9215048769159313,
+      "grad_norm": 0.23036227594793113,
+      "learning_rate": 3.8502581755593804e-05,
+      "loss": 0.4176,
+      "step": 992
+    },
+    {
+      "epoch": 0.92243381328379,
+      "grad_norm": 0.23093971462681168,
+      "learning_rate": 3.848537005163511e-05,
+      "loss": 0.4007,
+      "step": 993
+    },
+    {
+      "epoch": 0.9233627496516489,
+      "grad_norm": 0.2317551394646806,
+      "learning_rate": 3.846815834767642e-05,
+      "loss": 0.405,
+      "step": 994
+    },
+    {
+      "epoch": 0.9242916860195076,
+      "grad_norm": 0.21158502861181783,
+      "learning_rate": 3.8450946643717725e-05,
+      "loss": 0.3942,
+      "step": 995
+    },
+    {
+      "epoch": 0.9252206223873665,
+      "grad_norm": 0.2537032563068634,
+      "learning_rate": 3.843373493975904e-05,
+      "loss": 0.3791,
+      "step": 996
+    },
+    {
+      "epoch": 0.9261495587552253,
+      "grad_norm": 0.2239536974059551,
+      "learning_rate": 3.841652323580034e-05,
+      "loss": 0.4064,
+      "step": 997
+    },
+    {
+      "epoch": 0.927078495123084,
+      "grad_norm": 0.25068704163807093,
+      "learning_rate": 3.8399311531841654e-05,
+      "loss": 0.4273,
+      "step": 998
+    },
+    {
+      "epoch": 0.9280074314909429,
+      "grad_norm": 0.2457256835693048,
+      "learning_rate": 3.838209982788296e-05,
+      "loss": 0.4096,
+      "step": 999
+    },
+    {
+      "epoch": 0.9289363678588016,
+      "grad_norm": 0.23667198447703158,
+      "learning_rate": 3.836488812392427e-05,
+      "loss": 0.3934,
+      "step": 1000
+    },
+    {
+      "epoch": 0.9298653042266605,
+      "grad_norm": 0.24449737642939945,
+      "learning_rate": 3.8347676419965576e-05,
+      "loss": 0.409,
+      "step": 1001
+    },
+    {
+      "epoch": 0.9307942405945193,
+      "grad_norm": 0.24650595873011708,
+      "learning_rate": 3.833046471600689e-05,
+      "loss": 0.4053,
+      "step": 1002
+    },
+    {
+      "epoch": 0.931723176962378,
+      "grad_norm": 0.26151738161916,
+      "learning_rate": 3.831325301204819e-05,
+      "loss": 0.4104,
+      "step": 1003
+    },
+    {
+      "epoch": 0.9326521133302369,
+      "grad_norm": 0.2310178657482112,
+      "learning_rate": 3.8296041308089504e-05,
+      "loss": 0.3937,
+      "step": 1004
+    },
+    {
+      "epoch": 0.9335810496980956,
+      "grad_norm": 0.23546622207591916,
+      "learning_rate": 3.827882960413081e-05,
+      "loss": 0.4041,
+      "step": 1005
+    },
+    {
+      "epoch": 0.9345099860659545,
+      "grad_norm": 0.2607486009527593,
+      "learning_rate": 3.826161790017212e-05,
+      "loss": 0.4353,
+      "step": 1006
+    },
+    {
+      "epoch": 0.9354389224338133,
+      "grad_norm": 0.2392967181255042,
+      "learning_rate": 3.8244406196213426e-05,
+      "loss": 0.3939,
+      "step": 1007
+    },
+    {
+      "epoch": 0.9363678588016721,
+      "grad_norm": 0.21674134867792574,
+      "learning_rate": 3.822719449225474e-05,
+      "loss": 0.388,
+      "step": 1008
+    },
+    {
+      "epoch": 0.9372967951695309,
+      "grad_norm": 0.24940419360087127,
+      "learning_rate": 3.820998278829604e-05,
+      "loss": 0.4239,
+      "step": 1009
+    },
+    {
+      "epoch": 0.9382257315373896,
+      "grad_norm": 0.2596709039538196,
+      "learning_rate": 3.8192771084337355e-05,
+      "loss": 0.4005,
+      "step": 1010
+    },
+    {
+      "epoch": 0.9391546679052485,
+      "grad_norm": 0.23606867754298808,
+      "learning_rate": 3.817555938037866e-05,
+      "loss": 0.4374,
+      "step": 1011
+    },
+    {
+      "epoch": 0.9400836042731073,
+      "grad_norm": 0.27125016523755546,
+      "learning_rate": 3.815834767641997e-05,
+      "loss": 0.4426,
+      "step": 1012
+    },
+    {
+      "epoch": 0.9410125406409661,
+      "grad_norm": 0.24501601889271793,
+      "learning_rate": 3.8141135972461277e-05,
+      "loss": 0.3874,
+      "step": 1013
+    },
+    {
+      "epoch": 0.9419414770088249,
+      "grad_norm": 0.2442683405582566,
+      "learning_rate": 3.812392426850258e-05,
+      "loss": 0.4047,
+      "step": 1014
+    },
+    {
+      "epoch": 0.9428704133766836,
+      "grad_norm": 0.26685613518926404,
+      "learning_rate": 3.810671256454389e-05,
+      "loss": 0.436,
+      "step": 1015
+    },
+    {
+      "epoch": 0.9437993497445425,
+      "grad_norm": 0.23983650732436032,
+      "learning_rate": 3.80895008605852e-05,
+      "loss": 0.4119,
+      "step": 1016
+    },
+    {
+      "epoch": 0.9447282861124013,
+      "grad_norm": 0.2570741433796242,
+      "learning_rate": 3.8072289156626503e-05,
+      "loss": 0.4236,
+      "step": 1017
+    },
+    {
+      "epoch": 0.9456572224802601,
+      "grad_norm": 0.2544931404350317,
+      "learning_rate": 3.8055077452667815e-05,
+      "loss": 0.3768,
+      "step": 1018
+    },
+    {
+      "epoch": 0.9465861588481189,
+      "grad_norm": 0.2345446843587865,
+      "learning_rate": 3.803786574870912e-05,
+      "loss": 0.4143,
+      "step": 1019
+    },
+    {
+      "epoch": 0.9475150952159777,
+      "grad_norm": 0.24384024290773332,
+      "learning_rate": 3.802065404475043e-05,
+      "loss": 0.4044,
+      "step": 1020
+    },
+    {
+      "epoch": 0.9484440315838365,
+      "grad_norm": 0.21208048637232893,
+      "learning_rate": 3.800344234079174e-05,
+      "loss": 0.4217,
+      "step": 1021
+    },
+    {
+      "epoch": 0.9493729679516953,
+      "grad_norm": 0.2599635780830963,
+      "learning_rate": 3.798623063683305e-05,
+      "loss": 0.4122,
+      "step": 1022
+    },
+    {
+      "epoch": 0.9503019043195541,
+      "grad_norm": 0.2391907225652313,
+      "learning_rate": 3.7969018932874354e-05,
+      "loss": 0.3861,
+      "step": 1023
+    },
+    {
+      "epoch": 0.9512308406874129,
+      "grad_norm": 0.21784120153158443,
+      "learning_rate": 3.7951807228915666e-05,
+      "loss": 0.4017,
+      "step": 1024
+    },
+    {
+      "epoch": 0.9521597770552717,
+      "grad_norm": 0.2868793994136997,
+      "learning_rate": 3.793459552495697e-05,
+      "loss": 0.4078,
+      "step": 1025
+    },
+    {
+      "epoch": 0.9530887134231305,
+      "grad_norm": 0.2303322060460401,
+      "learning_rate": 3.791738382099828e-05,
+      "loss": 0.3959,
+      "step": 1026
+    },
+    {
+      "epoch": 0.9540176497909894,
+      "grad_norm": 0.21929777651344276,
+      "learning_rate": 3.790017211703959e-05,
+      "loss": 0.4378,
+      "step": 1027
+    },
+    {
+      "epoch": 0.9549465861588481,
+      "grad_norm": 0.29921819323800025,
+      "learning_rate": 3.78829604130809e-05,
+      "loss": 0.4114,
+      "step": 1028
+    },
+    {
+      "epoch": 0.9558755225267069,
+      "grad_norm": 0.24792104039539412,
+      "learning_rate": 3.7865748709122204e-05,
+      "loss": 0.3957,
+      "step": 1029
+    },
+    {
+      "epoch": 0.9568044588945657,
+      "grad_norm": 0.2547675598430285,
+      "learning_rate": 3.7848537005163516e-05,
+      "loss": 0.4139,
+      "step": 1030
+    },
+    {
+      "epoch": 0.9577333952624245,
+      "grad_norm": 0.23635621173173713,
+      "learning_rate": 3.783132530120482e-05,
+      "loss": 0.4302,
+      "step": 1031
+    },
+    {
+      "epoch": 0.9586623316302834,
+      "grad_norm": 0.23539058473689725,
+      "learning_rate": 3.781411359724613e-05,
+      "loss": 0.4106,
+      "step": 1032
+    },
+    {
+      "epoch": 0.9595912679981421,
+      "grad_norm": 0.2252626765187598,
+      "learning_rate": 3.779690189328744e-05,
+      "loss": 0.4201,
+      "step": 1033
+    },
+    {
+      "epoch": 0.9605202043660009,
+      "grad_norm": 0.23605984311517597,
+      "learning_rate": 3.777969018932875e-05,
+      "loss": 0.4014,
+      "step": 1034
+    },
+    {
+      "epoch": 0.9614491407338597,
+      "grad_norm": 0.22411023295920124,
+      "learning_rate": 3.776247848537005e-05,
+      "loss": 0.3995,
+      "step": 1035
+    },
+    {
+      "epoch": 0.9623780771017185,
+      "grad_norm": 0.22444324386702214,
+      "learning_rate": 3.774526678141136e-05,
+      "loss": 0.396,
+      "step": 1036
+    },
+    {
+      "epoch": 0.9633070134695774,
+      "grad_norm": 0.2392457838225579,
+      "learning_rate": 3.7728055077452665e-05,
+      "loss": 0.3868,
+      "step": 1037
+    },
+    {
+      "epoch": 0.9642359498374361,
+      "grad_norm": 0.24259642738031353,
+      "learning_rate": 3.7710843373493976e-05,
+      "loss": 0.4059,
+      "step": 1038
+    },
+    {
+      "epoch": 0.965164886205295,
+      "grad_norm": 0.22122843588430882,
+      "learning_rate": 3.769363166953528e-05,
+      "loss": 0.3932,
+      "step": 1039
+    },
+    {
+      "epoch": 0.9660938225731538,
+      "grad_norm": 0.39575069207466423,
+      "learning_rate": 3.767641996557659e-05,
+      "loss": 0.4106,
+      "step": 1040
+    },
+    {
+      "epoch": 0.9670227589410125,
+      "grad_norm": 0.20970275754321213,
+      "learning_rate": 3.76592082616179e-05,
+      "loss": 0.3826,
+      "step": 1041
+    },
+    {
+      "epoch": 0.9679516953088714,
+      "grad_norm": 0.2689513010745634,
+      "learning_rate": 3.764199655765921e-05,
+      "loss": 0.4246,
+      "step": 1042
+    },
+    {
+      "epoch": 0.9688806316767301,
+      "grad_norm": 0.21716008885897425,
+      "learning_rate": 3.7624784853700515e-05,
+      "loss": 0.372,
+      "step": 1043
+    },
+    {
+      "epoch": 0.969809568044589,
+      "grad_norm": 0.19032663763219618,
+      "learning_rate": 3.760757314974183e-05,
+      "loss": 0.3871,
+      "step": 1044
+    },
+    {
+      "epoch": 0.9707385044124478,
+      "grad_norm": 0.24689207170253144,
+      "learning_rate": 3.759036144578313e-05,
+      "loss": 0.3887,
+      "step": 1045
+    },
+    {
+      "epoch": 0.9716674407803065,
+      "grad_norm": 0.23664100739316893,
+      "learning_rate": 3.7573149741824444e-05,
+      "loss": 0.403,
+      "step": 1046
+    },
+    {
+      "epoch": 0.9725963771481654,
+      "grad_norm": 0.22874968861174907,
+      "learning_rate": 3.755593803786575e-05,
+      "loss": 0.429,
+      "step": 1047
+    },
+    {
+      "epoch": 0.9735253135160241,
+      "grad_norm": 0.213785249209633,
+      "learning_rate": 3.753872633390706e-05,
+      "loss": 0.3864,
+      "step": 1048
+    },
+    {
+      "epoch": 0.974454249883883,
+      "grad_norm": 0.24065036290868722,
+      "learning_rate": 3.7521514629948365e-05,
+      "loss": 0.433,
+      "step": 1049
+    },
+    {
+      "epoch": 0.9753831862517418,
+      "grad_norm": 0.22801923121906967,
+      "learning_rate": 3.750430292598968e-05,
+      "loss": 0.3929,
+      "step": 1050
+    },
+    {
+      "epoch": 0.9763121226196005,
+      "grad_norm": 0.24919400482036166,
+      "learning_rate": 3.748709122203098e-05,
+      "loss": 0.4002,
+      "step": 1051
+    },
+    {
+      "epoch": 0.9772410589874594,
+      "grad_norm": 0.24145964608043208,
+      "learning_rate": 3.7469879518072294e-05,
+      "loss": 0.4318,
+      "step": 1052
+    },
+    {
+      "epoch": 0.9781699953553181,
+      "grad_norm": 0.20174232619370144,
+      "learning_rate": 3.74526678141136e-05,
+      "loss": 0.3912,
+      "step": 1053
+    },
+    {
+      "epoch": 0.979098931723177,
+      "grad_norm": 0.21629692614591498,
+      "learning_rate": 3.743545611015491e-05,
+      "loss": 0.3904,
+      "step": 1054
+    },
+    {
+      "epoch": 0.9800278680910358,
+      "grad_norm": 0.22440046201731037,
+      "learning_rate": 3.7418244406196216e-05,
+      "loss": 0.4137,
+      "step": 1055
+    },
+    {
+      "epoch": 0.9809568044588945,
+      "grad_norm": 0.22672281532519148,
+      "learning_rate": 3.740103270223752e-05,
+      "loss": 0.3976,
+      "step": 1056
+    },
+    {
+      "epoch": 0.9818857408267534,
+      "grad_norm": 0.20668391208886464,
+      "learning_rate": 3.7383820998278826e-05,
+      "loss": 0.3951,
+      "step": 1057
+    },
+    {
+      "epoch": 0.9828146771946121,
+      "grad_norm": 0.2142818046995827,
+      "learning_rate": 3.736660929432014e-05,
+      "loss": 0.393,
+      "step": 1058
+    },
+    {
+      "epoch": 0.983743613562471,
+      "grad_norm": 0.21718324928035726,
+      "learning_rate": 3.734939759036144e-05,
+      "loss": 0.398,
+      "step": 1059
+    },
+    {
+      "epoch": 0.9846725499303298,
+      "grad_norm": 0.23550777849566212,
+      "learning_rate": 3.7332185886402754e-05,
+      "loss": 0.4065,
+      "step": 1060
+    },
+    {
+      "epoch": 0.9856014862981886,
+      "grad_norm": 0.2167543874679464,
+      "learning_rate": 3.731497418244406e-05,
+      "loss": 0.3927,
+      "step": 1061
+    },
+    {
+      "epoch": 0.9865304226660474,
+      "grad_norm": 0.2530841425618813,
+      "learning_rate": 3.729776247848537e-05,
+      "loss": 0.4099,
+      "step": 1062
+    },
+    {
+      "epoch": 0.9874593590339061,
+      "grad_norm": 0.2376035990853862,
+      "learning_rate": 3.7280550774526676e-05,
+      "loss": 0.4056,
+      "step": 1063
+    },
+    {
+      "epoch": 0.988388295401765,
+      "grad_norm": 0.22485376239690827,
+      "learning_rate": 3.726333907056799e-05,
+      "loss": 0.411,
+      "step": 1064
+    },
+    {
+      "epoch": 0.9893172317696238,
+      "grad_norm": 0.2364195030509185,
+      "learning_rate": 3.724612736660929e-05,
+      "loss": 0.4024,
+      "step": 1065
+    },
+    {
+      "epoch": 0.9902461681374826,
+      "grad_norm": 0.2741768987248631,
+      "learning_rate": 3.7228915662650605e-05,
+      "loss": 0.4172,
+      "step": 1066
+    },
+    {
+      "epoch": 0.9911751045053414,
+      "grad_norm": 0.2224572039465709,
+      "learning_rate": 3.721170395869191e-05,
+      "loss": 0.4063,
+      "step": 1067
+    },
+    {
+      "epoch": 0.9921040408732001,
+      "grad_norm": 0.25384861733232483,
+      "learning_rate": 3.719449225473322e-05,
+      "loss": 0.3874,
+      "step": 1068
+    },
+    {
+      "epoch": 0.993032977241059,
+      "grad_norm": 0.21782354887827174,
+      "learning_rate": 3.717728055077453e-05,
+      "loss": 0.38,
+      "step": 1069
+    },
+    {
+      "epoch": 0.9939619136089178,
+      "grad_norm": 0.23692026091455595,
+      "learning_rate": 3.716006884681584e-05,
+      "loss": 0.4025,
+      "step": 1070
+    },
+    {
+      "epoch": 0.9948908499767766,
+      "grad_norm": 0.2628360369682366,
+      "learning_rate": 3.7142857142857143e-05,
+      "loss": 0.4011,
+      "step": 1071
+    },
+    {
+      "epoch": 0.9958197863446354,
+      "grad_norm": 0.23586526214391343,
+      "learning_rate": 3.7125645438898455e-05,
+      "loss": 0.4004,
+      "step": 1072
+    },
+    {
+      "epoch": 0.9967487227124942,
+      "grad_norm": 0.2699525810140143,
+      "learning_rate": 3.710843373493976e-05,
+      "loss": 0.4211,
+      "step": 1073
+    },
+    {
+      "epoch": 0.997677659080353,
+      "grad_norm": 0.24858789264268025,
+      "learning_rate": 3.709122203098107e-05,
+      "loss": 0.4055,
+      "step": 1074
+    },
+    {
+      "epoch": 0.9986065954482118,
+      "grad_norm": 0.21173668049329192,
+      "learning_rate": 3.707401032702238e-05,
+      "loss": 0.3769,
+      "step": 1075
+    },
+    {
+      "epoch": 0.9995355318160706,
+      "grad_norm": 0.24262016662999178,
+      "learning_rate": 3.705679862306369e-05,
+      "loss": 0.4221,
+      "step": 1076
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.24262016662999178,
+      "learning_rate": 3.703958691910499e-05,
+      "loss": 0.4056,
+      "step": 1077
+    },
+    {
+      "epoch": 1.0009289363678588,
+      "grad_norm": 0.42400479184527856,
+      "learning_rate": 3.70223752151463e-05,
+      "loss": 0.359,
+      "step": 1078
+    },
+    {
+      "epoch": 1.0018578727357177,
+      "grad_norm": 0.25465365326859013,
+      "learning_rate": 3.7005163511187604e-05,
+      "loss": 0.3744,
+      "step": 1079
+    },
+    {
+      "epoch": 1.0027868091035763,
+      "grad_norm": 0.2749579483536481,
+      "learning_rate": 3.6987951807228916e-05,
+      "loss": 0.3215,
+      "step": 1080
+    },
+    {
+      "epoch": 1.0037157454714352,
+      "grad_norm": 0.25882510922444807,
+      "learning_rate": 3.697074010327022e-05,
+      "loss": 0.3664,
+      "step": 1081
+    },
+    {
+      "epoch": 1.004644681839294,
+      "grad_norm": 0.25257387474158305,
+      "learning_rate": 3.695352839931153e-05,
+      "loss": 0.3439,
+      "step": 1082
+    },
+    {
+      "epoch": 1.0055736182071529,
+      "grad_norm": 0.2818136562808609,
+      "learning_rate": 3.693631669535284e-05,
+      "loss": 0.3495,
+      "step": 1083
+    },
+    {
+      "epoch": 1.0065025545750117,
+      "grad_norm": 0.2796090387992452,
+      "learning_rate": 3.691910499139415e-05,
+      "loss": 0.3577,
+      "step": 1084
+    },
+    {
+      "epoch": 1.0074314909428703,
+      "grad_norm": 0.25921344914520555,
+      "learning_rate": 3.6901893287435454e-05,
+      "loss": 0.3474,
+      "step": 1085
+    },
+    {
+      "epoch": 1.0083604273107292,
+      "grad_norm": 0.26684815525436895,
+      "learning_rate": 3.6884681583476766e-05,
+      "loss": 0.3476,
+      "step": 1086
+    },
+    {
+      "epoch": 1.009289363678588,
+      "grad_norm": 0.27357439621922414,
+      "learning_rate": 3.686746987951807e-05,
+      "loss": 0.3437,
+      "step": 1087
+    },
+    {
+      "epoch": 1.0102183000464469,
+      "grad_norm": 0.24447102717487307,
+      "learning_rate": 3.685025817555938e-05,
+      "loss": 0.3221,
+      "step": 1088
+    },
+    {
+      "epoch": 1.0111472364143057,
+      "grad_norm": 0.258464222591644,
+      "learning_rate": 3.683304647160069e-05,
+      "loss": 0.356,
+      "step": 1089
+    },
+    {
+      "epoch": 1.0120761727821643,
+      "grad_norm": 0.27925086544400457,
+      "learning_rate": 3.6815834767642e-05,
+      "loss": 0.3701,
+      "step": 1090
+    },
+    {
+      "epoch": 1.0130051091500232,
+      "grad_norm": 0.2457044337838858,
+      "learning_rate": 3.6798623063683305e-05,
+      "loss": 0.3458,
+      "step": 1091
+    },
+    {
+      "epoch": 1.013934045517882,
+      "grad_norm": 0.2973360259118452,
+      "learning_rate": 3.6781411359724616e-05,
+      "loss": 0.3503,
+      "step": 1092
+    },
+    {
+      "epoch": 1.0148629818857409,
+      "grad_norm": 0.24684519541371244,
+      "learning_rate": 3.676419965576592e-05,
+      "loss": 0.3555,
+      "step": 1093
+    },
+    {
+      "epoch": 1.0157919182535997,
+      "grad_norm": 0.2553488145462288,
+      "learning_rate": 3.674698795180723e-05,
+      "loss": 0.339,
+      "step": 1094
+    },
+    {
+      "epoch": 1.0167208546214583,
+      "grad_norm": 0.27698275528497907,
+      "learning_rate": 3.672977624784854e-05,
+      "loss": 0.3555,
+      "step": 1095
+    },
+    {
+      "epoch": 1.0176497909893172,
+      "grad_norm": 0.25512853436869026,
+      "learning_rate": 3.671256454388985e-05,
+      "loss": 0.3508,
+      "step": 1096
+    },
+    {
+      "epoch": 1.018578727357176,
+      "grad_norm": 0.23687230499469353,
+      "learning_rate": 3.6695352839931155e-05,
+      "loss": 0.3541,
+      "step": 1097
+    },
+    {
+      "epoch": 1.0195076637250349,
+      "grad_norm": 0.2363717396988948,
+      "learning_rate": 3.667814113597246e-05,
+      "loss": 0.3533,
+      "step": 1098
+    },
+    {
+      "epoch": 1.0204366000928937,
+      "grad_norm": 0.2303201035801806,
+      "learning_rate": 3.666092943201377e-05,
+      "loss": 0.3632,
+      "step": 1099
+    },
+    {
+      "epoch": 1.0213655364607523,
+      "grad_norm": 0.2459782206416615,
+      "learning_rate": 3.664371772805508e-05,
+      "loss": 0.3618,
+      "step": 1100
+    },
+    {
+      "epoch": 1.0222944728286112,
+      "grad_norm": 0.296616176626298,
+      "learning_rate": 3.662650602409639e-05,
+      "loss": 0.3922,
+      "step": 1101
+    },
+    {
+      "epoch": 1.02322340919647,
+      "grad_norm": 0.2147341891639701,
+      "learning_rate": 3.6609294320137694e-05,
+      "loss": 0.3519,
+      "step": 1102
+    },
+    {
+      "epoch": 1.0241523455643289,
+      "grad_norm": 0.266276617171416,
+      "learning_rate": 3.6592082616179e-05,
+      "loss": 0.3377,
+      "step": 1103
+    },
+    {
+      "epoch": 1.0250812819321877,
+      "grad_norm": 0.23338087166757537,
+      "learning_rate": 3.657487091222031e-05,
+      "loss": 0.3436,
+      "step": 1104
+    },
+    {
+      "epoch": 1.0260102183000464,
+      "grad_norm": 0.26374694027225837,
+      "learning_rate": 3.6557659208261616e-05,
+      "loss": 0.3478,
+      "step": 1105
+    },
+    {
+      "epoch": 1.0269391546679052,
+      "grad_norm": 0.2337037433490987,
+      "learning_rate": 3.654044750430293e-05,
+      "loss": 0.3517,
+      "step": 1106
+    },
+    {
+      "epoch": 1.027868091035764,
+      "grad_norm": 0.2596231255028893,
+      "learning_rate": 3.652323580034423e-05,
+      "loss": 0.365,
+      "step": 1107
+    },
+    {
+      "epoch": 1.0287970274036229,
+      "grad_norm": 0.240620940997802,
+      "learning_rate": 3.6506024096385544e-05,
+      "loss": 0.3628,
+      "step": 1108
+    },
+    {
+      "epoch": 1.0297259637714817,
+      "grad_norm": 0.21426737173957083,
+      "learning_rate": 3.648881239242685e-05,
+      "loss": 0.3335,
+      "step": 1109
+    },
+    {
+      "epoch": 1.0306549001393404,
+      "grad_norm": 0.2229960824044303,
+      "learning_rate": 3.647160068846816e-05,
+      "loss": 0.3549,
+      "step": 1110
+    },
+    {
+      "epoch": 1.0315838365071992,
+      "grad_norm": 0.27079181149762227,
+      "learning_rate": 3.6454388984509466e-05,
+      "loss": 0.3767,
+      "step": 1111
+    },
+    {
+      "epoch": 1.032512772875058,
+      "grad_norm": 0.22945807604481355,
+      "learning_rate": 3.643717728055078e-05,
+      "loss": 0.3442,
+      "step": 1112
+    },
+    {
+      "epoch": 1.033441709242917,
+      "grad_norm": 0.20955895984599512,
+      "learning_rate": 3.641996557659208e-05,
+      "loss": 0.3248,
+      "step": 1113
+    },
+    {
+      "epoch": 1.0343706456107757,
+      "grad_norm": 0.23194107498456232,
+      "learning_rate": 3.6402753872633395e-05,
+      "loss": 0.3502,
+      "step": 1114
+    },
+    {
+      "epoch": 1.0352995819786344,
+      "grad_norm": 0.2564535374989993,
+      "learning_rate": 3.63855421686747e-05,
+      "loss": 0.3694,
+      "step": 1115
+    },
+    {
+      "epoch": 1.0362285183464932,
+      "grad_norm": 0.2280355951773234,
+      "learning_rate": 3.636833046471601e-05,
+      "loss": 0.3373,
+      "step": 1116
+    },
+    {
+      "epoch": 1.037157454714352,
+      "grad_norm": 0.23682133916062154,
+      "learning_rate": 3.6351118760757316e-05,
+      "loss": 0.3355,
+      "step": 1117
+    },
+    {
+      "epoch": 1.038086391082211,
+      "grad_norm": 0.216976014046176,
+      "learning_rate": 3.633390705679863e-05,
+      "loss": 0.3476,
+      "step": 1118
+    },
+    {
+      "epoch": 1.0390153274500697,
+      "grad_norm": 0.2067060037353485,
+      "learning_rate": 3.631669535283993e-05,
+      "loss": 0.3515,
+      "step": 1119
+    },
+    {
+      "epoch": 1.0399442638179284,
+      "grad_norm": 0.26852478563809756,
+      "learning_rate": 3.629948364888124e-05,
+      "loss": 0.3687,
+      "step": 1120
+    },
+    {
+      "epoch": 1.0408732001857872,
+      "grad_norm": 0.23440938063759156,
+      "learning_rate": 3.628227194492255e-05,
+      "loss": 0.3676,
+      "step": 1121
+    },
+    {
+      "epoch": 1.041802136553646,
+      "grad_norm": 0.22839083712400235,
+      "learning_rate": 3.6265060240963855e-05,
+      "loss": 0.3483,
+      "step": 1122
+    },
+    {
+      "epoch": 1.042731072921505,
+      "grad_norm": 0.2868994905545559,
+      "learning_rate": 3.624784853700517e-05,
+      "loss": 0.3495,
+      "step": 1123
+    },
+    {
+      "epoch": 1.0436600092893638,
+      "grad_norm": 0.21384049452616608,
+      "learning_rate": 3.623063683304647e-05,
+      "loss": 0.3419,
+      "step": 1124
+    },
+    {
+      "epoch": 1.0445889456572224,
+      "grad_norm": 0.22545699928123575,
+      "learning_rate": 3.6213425129087784e-05,
+      "loss": 0.3507,
+      "step": 1125
+    },
+    {
+      "epoch": 1.0455178820250812,
+      "grad_norm": 0.2616098875142772,
+      "learning_rate": 3.619621342512909e-05,
+      "loss": 0.3684,
+      "step": 1126
+    },
+    {
+      "epoch": 1.04644681839294,
+      "grad_norm": 0.20808430540664596,
+      "learning_rate": 3.6179001721170394e-05,
+      "loss": 0.331,
+      "step": 1127
+    },
+    {
+      "epoch": 1.047375754760799,
+      "grad_norm": 0.2447603473682486,
+      "learning_rate": 3.6161790017211705e-05,
+      "loss": 0.3455,
+      "step": 1128
+    },
+    {
+      "epoch": 1.0483046911286578,
+      "grad_norm": 0.2500131614712927,
+      "learning_rate": 3.614457831325301e-05,
+      "loss": 0.3361,
+      "step": 1129
+    },
+    {
+      "epoch": 1.0492336274965164,
+      "grad_norm": 0.2325076708728905,
+      "learning_rate": 3.612736660929432e-05,
+      "loss": 0.3345,
+      "step": 1130
+    },
+    {
+      "epoch": 1.0501625638643752,
+      "grad_norm": 0.3372982804928654,
+      "learning_rate": 3.611015490533563e-05,
+      "loss": 0.3442,
+      "step": 1131
+    },
+    {
+      "epoch": 1.051091500232234,
+      "grad_norm": 0.26364292805811546,
+      "learning_rate": 3.609294320137694e-05,
+      "loss": 0.355,
+      "step": 1132
+    },
+    {
+      "epoch": 1.052020436600093,
+      "grad_norm": 0.2527413633654713,
+      "learning_rate": 3.6075731497418244e-05,
+      "loss": 0.3629,
+      "step": 1133
+    },
+    {
+      "epoch": 1.0529493729679518,
+      "grad_norm": 0.2554529746289192,
+      "learning_rate": 3.6058519793459556e-05,
+      "loss": 0.3383,
+      "step": 1134
+    },
+    {
+      "epoch": 1.0538783093358104,
+      "grad_norm": 0.27047127903067114,
+      "learning_rate": 3.604130808950086e-05,
+      "loss": 0.3282,
+      "step": 1135
+    },
+    {
+      "epoch": 1.0548072457036692,
+      "grad_norm": 0.19736989479380834,
+      "learning_rate": 3.602409638554217e-05,
+      "loss": 0.3713,
+      "step": 1136
+    },
+    {
+      "epoch": 1.055736182071528,
+      "grad_norm": 0.2945278332205401,
+      "learning_rate": 3.600688468158348e-05,
+      "loss": 0.3537,
+      "step": 1137
+    },
+    {
+      "epoch": 1.056665118439387,
+      "grad_norm": 0.2583946369570546,
+      "learning_rate": 3.598967297762479e-05,
+      "loss": 0.3963,
+      "step": 1138
+    },
+    {
+      "epoch": 1.0575940548072458,
+      "grad_norm": 0.50245090499772,
+      "learning_rate": 3.5972461273666094e-05,
+      "loss": 0.3514,
+      "step": 1139
+    },
+    {
+      "epoch": 1.0585229911751044,
+      "grad_norm": 0.21531395236728787,
+      "learning_rate": 3.5955249569707406e-05,
+      "loss": 0.3562,
+      "step": 1140
+    },
+    {
+      "epoch": 1.0594519275429632,
+      "grad_norm": 0.24061017735313153,
+      "learning_rate": 3.593803786574871e-05,
+      "loss": 0.3694,
+      "step": 1141
+    },
+    {
+      "epoch": 1.060380863910822,
+      "grad_norm": 0.2268314841653852,
+      "learning_rate": 3.5920826161790016e-05,
+      "loss": 0.3422,
+      "step": 1142
+    },
+    {
+      "epoch": 1.061309800278681,
+      "grad_norm": 0.22078553445878826,
+      "learning_rate": 3.590361445783133e-05,
+      "loss": 0.3606,
+      "step": 1143
+    },
+    {
+      "epoch": 1.0622387366465398,
+      "grad_norm": 0.22090392465288902,
+      "learning_rate": 3.588640275387263e-05,
+      "loss": 0.3448,
+      "step": 1144
+    },
+    {
+      "epoch": 1.0631676730143984,
+      "grad_norm": 0.23293244274551053,
+      "learning_rate": 3.5869191049913945e-05,
+      "loss": 0.3415,
+      "step": 1145
+    },
+    {
+      "epoch": 1.0640966093822573,
+      "grad_norm": 0.21096983770192937,
+      "learning_rate": 3.585197934595525e-05,
+      "loss": 0.3563,
+      "step": 1146
+    },
+    {
+      "epoch": 1.065025545750116,
+      "grad_norm": 0.22212524809608433,
+      "learning_rate": 3.583476764199656e-05,
+      "loss": 0.3316,
+      "step": 1147
+    },
+    {
+      "epoch": 1.065954482117975,
+      "grad_norm": 0.20899680745058694,
+      "learning_rate": 3.5817555938037867e-05,
+      "loss": 0.3649,
+      "step": 1148
+    },
+    {
+      "epoch": 1.0668834184858338,
+      "grad_norm": 0.2224898644473964,
+      "learning_rate": 3.580034423407918e-05,
+      "loss": 0.3485,
+      "step": 1149
+    },
+    {
+      "epoch": 1.0678123548536926,
+      "grad_norm": 0.2487544062552923,
+      "learning_rate": 3.578313253012048e-05,
+      "loss": 0.3522,
+      "step": 1150
+    },
+    {
+      "epoch": 1.0687412912215513,
+      "grad_norm": 0.23055813456250082,
+      "learning_rate": 3.576592082616179e-05,
+      "loss": 0.3618,
+      "step": 1151
+    },
+    {
+      "epoch": 1.06967022758941,
+      "grad_norm": 0.20028327166610577,
+      "learning_rate": 3.57487091222031e-05,
+      "loss": 0.3358,
+      "step": 1152
+    },
+    {
+      "epoch": 1.070599163957269,
+      "grad_norm": 0.22311914856614992,
+      "learning_rate": 3.5731497418244405e-05,
+      "loss": 0.3549,
+      "step": 1153
+    },
+    {
+      "epoch": 1.0715281003251278,
+      "grad_norm": 0.23560718370813885,
+      "learning_rate": 3.571428571428572e-05,
+      "loss": 0.374,
+      "step": 1154
+    },
+    {
+      "epoch": 1.0724570366929864,
+      "grad_norm": 0.4148928972512194,
+      "learning_rate": 3.569707401032702e-05,
+      "loss": 0.3528,
+      "step": 1155
+    },
+    {
+      "epoch": 1.0733859730608453,
+      "grad_norm": 0.25885861158094975,
+      "learning_rate": 3.5679862306368334e-05,
+      "loss": 0.3412,
+      "step": 1156
+    },
+    {
+      "epoch": 1.0743149094287041,
+      "grad_norm": 0.21300196539669833,
+      "learning_rate": 3.566265060240964e-05,
+      "loss": 0.3355,
+      "step": 1157
+    },
+    {
+      "epoch": 1.075243845796563,
+      "grad_norm": 0.235345413141929,
+      "learning_rate": 3.564543889845095e-05,
+      "loss": 0.3629,
+      "step": 1158
+    },
+    {
+      "epoch": 1.0761727821644218,
+      "grad_norm": 0.23367347817298498,
+      "learning_rate": 3.5628227194492256e-05,
+      "loss": 0.3613,
+      "step": 1159
+    },
+    {
+      "epoch": 1.0771017185322806,
+      "grad_norm": 0.2326105412483254,
+      "learning_rate": 3.561101549053357e-05,
+      "loss": 0.3461,
+      "step": 1160
+    },
+    {
+      "epoch": 1.0780306549001393,
+      "grad_norm": 0.23402340603110036,
+      "learning_rate": 3.559380378657487e-05,
+      "loss": 0.3383,
+      "step": 1161
+    },
+    {
+      "epoch": 1.0789595912679981,
+      "grad_norm": 0.2392078922693959,
+      "learning_rate": 3.557659208261618e-05,
+      "loss": 0.3481,
+      "step": 1162
+    },
+    {
+      "epoch": 1.079888527635857,
+      "grad_norm": 0.21367450129437765,
+      "learning_rate": 3.555938037865749e-05,
+      "loss": 0.3552,
+      "step": 1163
+    },
+    {
+      "epoch": 1.0808174640037158,
+      "grad_norm": 0.2715676998763513,
+      "learning_rate": 3.5542168674698794e-05,
+      "loss": 0.3379,
+      "step": 1164
+    },
+    {
+      "epoch": 1.0817464003715744,
+      "grad_norm": 0.2386244987904004,
+      "learning_rate": 3.5524956970740106e-05,
+      "loss": 0.3681,
+      "step": 1165
+    },
+    {
+      "epoch": 1.0826753367394333,
+      "grad_norm": 0.22559148972950677,
+      "learning_rate": 3.550774526678141e-05,
+      "loss": 0.3784,
+      "step": 1166
+    },
+    {
+      "epoch": 1.0836042731072921,
+      "grad_norm": 0.2525819984114936,
+      "learning_rate": 3.549053356282272e-05,
+      "loss": 0.3547,
+      "step": 1167
+    },
+    {
+      "epoch": 1.084533209475151,
+      "grad_norm": 0.21367915648446467,
+      "learning_rate": 3.547332185886403e-05,
+      "loss": 0.3396,
+      "step": 1168
+    },
+    {
+      "epoch": 1.0854621458430098,
+      "grad_norm": 0.2521020982590234,
+      "learning_rate": 3.545611015490534e-05,
+      "loss": 0.3592,
+      "step": 1169
+    },
+    {
+      "epoch": 1.0863910822108687,
+      "grad_norm": 0.21484751786346046,
+      "learning_rate": 3.5438898450946645e-05,
+      "loss": 0.3401,
+      "step": 1170
+    },
+    {
+      "epoch": 1.0873200185787273,
+      "grad_norm": 0.22930787394886787,
+      "learning_rate": 3.5421686746987956e-05,
+      "loss": 0.3503,
+      "step": 1171
+    },
+    {
+      "epoch": 1.0882489549465861,
+      "grad_norm": 0.21234758307608245,
+      "learning_rate": 3.540447504302926e-05,
+      "loss": 0.342,
+      "step": 1172
+    },
+    {
+      "epoch": 1.089177891314445,
+      "grad_norm": 0.1870818694435957,
+      "learning_rate": 3.538726333907057e-05,
+      "loss": 0.3388,
+      "step": 1173
+    },
+    {
+      "epoch": 1.0901068276823038,
+      "grad_norm": 0.2075435312021659,
+      "learning_rate": 3.537005163511188e-05,
+      "loss": 0.3534,
+      "step": 1174
+    },
+    {
+      "epoch": 1.0910357640501624,
+      "grad_norm": 0.2080276884082451,
+      "learning_rate": 3.535283993115319e-05,
+      "loss": 0.3308,
+      "step": 1175
+    },
+    {
+      "epoch": 1.0919647004180213,
+      "grad_norm": 0.21516330313139653,
+      "learning_rate": 3.5335628227194495e-05,
+      "loss": 0.3632,
+      "step": 1176
+    },
+    {
+      "epoch": 1.0928936367858801,
+      "grad_norm": 0.24747236372423687,
+      "learning_rate": 3.53184165232358e-05,
+      "loss": 0.3819,
+      "step": 1177
+    },
+    {
+      "epoch": 1.093822573153739,
+      "grad_norm": 0.2084513425036716,
+      "learning_rate": 3.530120481927711e-05,
+      "loss": 0.3284,
+      "step": 1178
+    },
+    {
+      "epoch": 1.0947515095215978,
+      "grad_norm": 0.2206018613707994,
+      "learning_rate": 3.528399311531842e-05,
+      "loss": 0.3736,
+      "step": 1179
+    },
+    {
+      "epoch": 1.0956804458894567,
+      "grad_norm": 0.26865785499633366,
+      "learning_rate": 3.526678141135973e-05,
+      "loss": 0.3811,
+      "step": 1180
+    },
+    {
+      "epoch": 1.0966093822573153,
+      "grad_norm": 0.2208182545530636,
+      "learning_rate": 3.5249569707401034e-05,
+      "loss": 0.3613,
+      "step": 1181
+    },
+    {
+      "epoch": 1.0975383186251741,
+      "grad_norm": 0.2654033751873286,
+      "learning_rate": 3.5232358003442345e-05,
+      "loss": 0.3426,
+      "step": 1182
+    },
+    {
+      "epoch": 1.098467254993033,
+      "grad_norm": 0.22186733925795926,
+      "learning_rate": 3.521514629948365e-05,
+      "loss": 0.3542,
+      "step": 1183
+    },
+    {
+      "epoch": 1.0993961913608918,
+      "grad_norm": 0.24207554296990225,
+      "learning_rate": 3.5197934595524955e-05,
+      "loss": 0.3839,
+      "step": 1184
+    },
+    {
+      "epoch": 1.1003251277287505,
+      "grad_norm": 0.2720135985816513,
+      "learning_rate": 3.518072289156627e-05,
+      "loss": 0.3818,
+      "step": 1185
+    },
+    {
+      "epoch": 1.1012540640966093,
+      "grad_norm": 0.23520239556784836,
+      "learning_rate": 3.516351118760757e-05,
+      "loss": 0.3573,
+      "step": 1186
+    },
+    {
+      "epoch": 1.1021830004644682,
+      "grad_norm": 0.20611704267650519,
+      "learning_rate": 3.5146299483648884e-05,
+      "loss": 0.3636,
+      "step": 1187
+    },
+    {
+      "epoch": 1.103111936832327,
+      "grad_norm": 0.23271916548723978,
+      "learning_rate": 3.512908777969019e-05,
+      "loss": 0.3563,
+      "step": 1188
+    },
+    {
+      "epoch": 1.1040408732001858,
+      "grad_norm": 0.22320870615937669,
+      "learning_rate": 3.51118760757315e-05,
+      "loss": 0.353,
+      "step": 1189
+    },
+    {
+      "epoch": 1.1049698095680447,
+      "grad_norm": 0.24142802201840471,
+      "learning_rate": 3.5094664371772806e-05,
+      "loss": 0.3674,
+      "step": 1190
+    },
+    {
+      "epoch": 1.1058987459359033,
+      "grad_norm": 0.24032397770102173,
+      "learning_rate": 3.507745266781412e-05,
+      "loss": 0.3396,
+      "step": 1191
+    },
+    {
+      "epoch": 1.1068276823037622,
+      "grad_norm": 0.23859301700052904,
+      "learning_rate": 3.506024096385542e-05,
+      "loss": 0.3709,
+      "step": 1192
+    },
+    {
+      "epoch": 1.107756618671621,
+      "grad_norm": 0.20993929975570846,
+      "learning_rate": 3.5043029259896734e-05,
+      "loss": 0.3532,
+      "step": 1193
+    },
+    {
+      "epoch": 1.1086855550394799,
+      "grad_norm": 0.21525862778600577,
+      "learning_rate": 3.502581755593804e-05,
+      "loss": 0.3414,
+      "step": 1194
+    },
+    {
+      "epoch": 1.1096144914073387,
+      "grad_norm": 0.2153151742361438,
+      "learning_rate": 3.500860585197935e-05,
+      "loss": 0.3479,
+      "step": 1195
+    },
+    {
+      "epoch": 1.1105434277751973,
+      "grad_norm": 0.2341045687466886,
+      "learning_rate": 3.4991394148020656e-05,
+      "loss": 0.3514,
+      "step": 1196
+    },
+    {
+      "epoch": 1.1114723641430562,
+      "grad_norm": 0.22041252024232932,
+      "learning_rate": 3.497418244406197e-05,
+      "loss": 0.3517,
+      "step": 1197
+    },
+    {
+      "epoch": 1.112401300510915,
+      "grad_norm": 0.2421034130602699,
+      "learning_rate": 3.495697074010327e-05,
+      "loss": 0.355,
+      "step": 1198
+    },
+    {
+      "epoch": 1.1133302368787739,
+      "grad_norm": 0.2148087803076203,
+      "learning_rate": 3.4939759036144585e-05,
+      "loss": 0.3433,
+      "step": 1199
+    },
+    {
+      "epoch": 1.1142591732466327,
+      "grad_norm": 0.23749094706036994,
+      "learning_rate": 3.492254733218589e-05,
+      "loss": 0.3566,
+      "step": 1200
+    },
+    {
+      "epoch": 1.1151881096144913,
+      "grad_norm": 0.24636426776494177,
+      "learning_rate": 3.4905335628227195e-05,
+      "loss": 0.3415,
+      "step": 1201
+    },
+    {
+      "epoch": 1.1161170459823502,
+      "grad_norm": 0.21408882618052283,
+      "learning_rate": 3.488812392426851e-05,
+      "loss": 0.348,
+      "step": 1202
+    },
+    {
+      "epoch": 1.117045982350209,
+      "grad_norm": 0.20922550693268657,
+      "learning_rate": 3.487091222030981e-05,
+      "loss": 0.3385,
+      "step": 1203
+    },
+    {
+      "epoch": 1.1179749187180679,
+      "grad_norm": 0.20482695943117954,
+      "learning_rate": 3.485370051635112e-05,
+      "loss": 0.3714,
+      "step": 1204
+    },
+    {
+      "epoch": 1.1189038550859267,
+      "grad_norm": 0.21259189935478204,
+      "learning_rate": 3.483648881239243e-05,
+      "loss": 0.3583,
+      "step": 1205
+    },
+    {
+      "epoch": 1.1198327914537853,
+      "grad_norm": 0.23993781373108236,
+      "learning_rate": 3.4819277108433733e-05,
+      "loss": 0.3549,
+      "step": 1206
+    },
+    {
+      "epoch": 1.1207617278216442,
+      "grad_norm": 0.21359898911362096,
+      "learning_rate": 3.4802065404475045e-05,
+      "loss": 0.3915,
+      "step": 1207
+    },
+    {
+      "epoch": 1.121690664189503,
+      "grad_norm": 0.24339391904887306,
+      "learning_rate": 3.478485370051635e-05,
+      "loss": 0.3481,
+      "step": 1208
+    },
+    {
+      "epoch": 1.1226196005573619,
+      "grad_norm": 0.18696533553937125,
+      "learning_rate": 3.476764199655766e-05,
+      "loss": 0.3489,
+      "step": 1209
+    },
+    {
+      "epoch": 1.1235485369252207,
+      "grad_norm": 0.22213428030739601,
+      "learning_rate": 3.475043029259897e-05,
+      "loss": 0.3442,
+      "step": 1210
+    },
+    {
+      "epoch": 1.1244774732930793,
+      "grad_norm": 0.22004231403214247,
+      "learning_rate": 3.473321858864028e-05,
+      "loss": 0.3567,
+      "step": 1211
+    },
+    {
+      "epoch": 1.1254064096609382,
+      "grad_norm": 0.24987319354407675,
+      "learning_rate": 3.4716006884681584e-05,
+      "loss": 0.3608,
+      "step": 1212
+    },
+    {
+      "epoch": 1.126335346028797,
+      "grad_norm": 0.22593282852581212,
+      "learning_rate": 3.4698795180722896e-05,
+      "loss": 0.3604,
+      "step": 1213
+    },
+    {
+      "epoch": 1.1272642823966559,
+      "grad_norm": 0.22913929079537423,
+      "learning_rate": 3.46815834767642e-05,
+      "loss": 0.3308,
+      "step": 1214
+    },
+    {
+      "epoch": 1.1281932187645147,
+      "grad_norm": 0.22512936563970043,
+      "learning_rate": 3.466437177280551e-05,
+      "loss": 0.3539,
+      "step": 1215
+    },
+    {
+      "epoch": 1.1291221551323734,
+      "grad_norm": 0.2051404174738686,
+      "learning_rate": 3.464716006884682e-05,
+      "loss": 0.363,
+      "step": 1216
+    },
+    {
+      "epoch": 1.1300510915002322,
+      "grad_norm": 0.21490274319372904,
+      "learning_rate": 3.462994836488813e-05,
+      "loss": 0.3354,
+      "step": 1217
+    },
+    {
+      "epoch": 1.130980027868091,
+      "grad_norm": 0.2174571977823224,
+      "learning_rate": 3.4612736660929434e-05,
+      "loss": 0.3513,
+      "step": 1218
+    },
+    {
+      "epoch": 1.1319089642359499,
+      "grad_norm": 0.19767325158530422,
+      "learning_rate": 3.4595524956970746e-05,
+      "loss": 0.3435,
+      "step": 1219
+    },
+    {
+      "epoch": 1.1328379006038087,
+      "grad_norm": 0.21719464735692653,
+      "learning_rate": 3.457831325301205e-05,
+      "loss": 0.3574,
+      "step": 1220
+    },
+    {
+      "epoch": 1.1337668369716674,
+      "grad_norm": 0.20581929044171807,
+      "learning_rate": 3.456110154905336e-05,
+      "loss": 0.3341,
+      "step": 1221
+    },
+    {
+      "epoch": 1.1346957733395262,
+      "grad_norm": 0.19844253892744268,
+      "learning_rate": 3.454388984509467e-05,
+      "loss": 0.3362,
+      "step": 1222
+    },
+    {
+      "epoch": 1.135624709707385,
+      "grad_norm": 0.2120233032755566,
+      "learning_rate": 3.452667814113598e-05,
+      "loss": 0.3521,
+      "step": 1223
+    },
+    {
+      "epoch": 1.136553646075244,
+      "grad_norm": 0.2038168421319054,
+      "learning_rate": 3.4509466437177285e-05,
+      "loss": 0.3831,
+      "step": 1224
+    },
+    {
+      "epoch": 1.1374825824431027,
+      "grad_norm": 0.25162161041366465,
+      "learning_rate": 3.449225473321859e-05,
+      "loss": 0.3629,
+      "step": 1225
+    },
+    {
+      "epoch": 1.1384115188109614,
+      "grad_norm": 0.23246017940495503,
+      "learning_rate": 3.4475043029259895e-05,
+      "loss": 0.3361,
+      "step": 1226
+    },
+    {
+      "epoch": 1.1393404551788202,
+      "grad_norm": 0.20778768722697064,
+      "learning_rate": 3.4457831325301206e-05,
+      "loss": 0.3555,
+      "step": 1227
+    },
+    {
+      "epoch": 1.140269391546679,
+      "grad_norm": 0.2153164058775418,
+      "learning_rate": 3.444061962134251e-05,
+      "loss": 0.3351,
+      "step": 1228
+    },
+    {
+      "epoch": 1.141198327914538,
+      "grad_norm": 0.20123624716901134,
+      "learning_rate": 3.442340791738382e-05,
+      "loss": 0.3443,
+      "step": 1229
+    },
+    {
+      "epoch": 1.1421272642823967,
+      "grad_norm": 0.21058058376175867,
+      "learning_rate": 3.440619621342513e-05,
+      "loss": 0.3587,
+      "step": 1230
+    },
+    {
+      "epoch": 1.1430562006502554,
+      "grad_norm": 0.2263383068441551,
+      "learning_rate": 3.438898450946644e-05,
+      "loss": 0.3305,
+      "step": 1231
+    },
+    {
+      "epoch": 1.1439851370181142,
+      "grad_norm": 0.21243006539833154,
+      "learning_rate": 3.4371772805507745e-05,
+      "loss": 0.3554,
+      "step": 1232
+    },
+    {
+      "epoch": 1.144914073385973,
+      "grad_norm": 0.2102541546923477,
+      "learning_rate": 3.435456110154906e-05,
+      "loss": 0.3567,
+      "step": 1233
+    },
+    {
+      "epoch": 1.145843009753832,
+      "grad_norm": 0.23046254004079486,
+      "learning_rate": 3.433734939759036e-05,
+      "loss": 0.3617,
+      "step": 1234
+    },
+    {
+      "epoch": 1.1467719461216908,
+      "grad_norm": 0.2930500467524406,
+      "learning_rate": 3.4320137693631674e-05,
+      "loss": 0.3555,
+      "step": 1235
+    },
+    {
+      "epoch": 1.1477008824895494,
+      "grad_norm": 0.23152478794927303,
+      "learning_rate": 3.430292598967298e-05,
+      "loss": 0.3734,
+      "step": 1236
+    },
+    {
+      "epoch": 1.1486298188574082,
+      "grad_norm": 0.2549389816652615,
+      "learning_rate": 3.428571428571429e-05,
+      "loss": 0.3293,
+      "step": 1237
+    },
+    {
+      "epoch": 1.149558755225267,
+      "grad_norm": 0.2167026826205843,
+      "learning_rate": 3.4268502581755595e-05,
+      "loss": 0.3411,
+      "step": 1238
+    },
+    {
+      "epoch": 1.150487691593126,
+      "grad_norm": 0.2485030988375794,
+      "learning_rate": 3.425129087779691e-05,
+      "loss": 0.3559,
+      "step": 1239
+    },
+    {
+      "epoch": 1.1514166279609848,
+      "grad_norm": 0.24163510390409695,
+      "learning_rate": 3.423407917383821e-05,
+      "loss": 0.3448,
+      "step": 1240
+    },
+    {
+      "epoch": 1.1523455643288434,
+      "grad_norm": 0.23020837321077686,
+      "learning_rate": 3.4216867469879524e-05,
+      "loss": 0.3346,
+      "step": 1241
+    },
+    {
+      "epoch": 1.1532745006967022,
+      "grad_norm": 0.21827258140007752,
+      "learning_rate": 3.419965576592083e-05,
+      "loss": 0.3594,
+      "step": 1242
+    },
+    {
+      "epoch": 1.154203437064561,
+      "grad_norm": 0.2259959955019267,
+      "learning_rate": 3.418244406196214e-05,
+      "loss": 0.3263,
+      "step": 1243
+    },
+    {
+      "epoch": 1.15513237343242,
+      "grad_norm": 0.2235516711875459,
+      "learning_rate": 3.4165232358003446e-05,
+      "loss": 0.3315,
+      "step": 1244
+    },
+    {
+      "epoch": 1.1560613098002788,
+      "grad_norm": 0.20721762489645518,
+      "learning_rate": 3.414802065404476e-05,
+      "loss": 0.3604,
+      "step": 1245
+    },
+    {
+      "epoch": 1.1569902461681374,
+      "grad_norm": 0.2231484555937936,
+      "learning_rate": 3.413080895008606e-05,
+      "loss": 0.3476,
+      "step": 1246
+    },
+    {
+      "epoch": 1.1579191825359962,
+      "grad_norm": 0.22679340464244963,
+      "learning_rate": 3.411359724612737e-05,
+      "loss": 0.3788,
+      "step": 1247
+    },
+    {
+      "epoch": 1.158848118903855,
+      "grad_norm": 0.2183757050580824,
+      "learning_rate": 3.409638554216867e-05,
+      "loss": 0.3308,
+      "step": 1248
+    },
+    {
+      "epoch": 1.159777055271714,
+      "grad_norm": 0.23012791532153595,
+      "learning_rate": 3.4079173838209984e-05,
+      "loss": 0.3475,
+      "step": 1249
+    },
+    {
+      "epoch": 1.1607059916395728,
+      "grad_norm": 0.2151078725178325,
+      "learning_rate": 3.406196213425129e-05,
+      "loss": 0.3542,
+      "step": 1250
+    },
+    {
+      "epoch": 1.1616349280074314,
+      "grad_norm": 0.20530486932614164,
+      "learning_rate": 3.40447504302926e-05,
+      "loss": 0.3362,
+      "step": 1251
+    },
+    {
+      "epoch": 1.1625638643752902,
+      "grad_norm": 0.21039678917229546,
+      "learning_rate": 3.4027538726333906e-05,
+      "loss": 0.3602,
+      "step": 1252
+    },
+    {
+      "epoch": 1.163492800743149,
+      "grad_norm": 0.2292399032989059,
+      "learning_rate": 3.401032702237522e-05,
+      "loss": 0.3678,
+      "step": 1253
+    },
+    {
+      "epoch": 1.164421737111008,
+      "grad_norm": 0.23493270267966537,
+      "learning_rate": 3.399311531841652e-05,
+      "loss": 0.3531,
+      "step": 1254
+    },
+    {
+      "epoch": 1.1653506734788668,
+      "grad_norm": 0.20806480245707265,
+      "learning_rate": 3.3975903614457835e-05,
+      "loss": 0.3474,
+      "step": 1255
+    },
+    {
+      "epoch": 1.1662796098467254,
+      "grad_norm": 0.23970461641766203,
+      "learning_rate": 3.395869191049914e-05,
+      "loss": 0.3555,
+      "step": 1256
+    },
+    {
+      "epoch": 1.1672085462145843,
+      "grad_norm": 0.23969672692118035,
+      "learning_rate": 3.394148020654045e-05,
+      "loss": 0.3634,
+      "step": 1257
+    },
+    {
+      "epoch": 1.168137482582443,
+      "grad_norm": 0.2285160457892427,
+      "learning_rate": 3.392426850258176e-05,
+      "loss": 0.3442,
+      "step": 1258
+    },
+    {
+      "epoch": 1.169066418950302,
+      "grad_norm": 0.22369375804224814,
+      "learning_rate": 3.390705679862307e-05,
+      "loss": 0.3311,
+      "step": 1259
+    },
+    {
+      "epoch": 1.1699953553181608,
+      "grad_norm": 0.21415743915884217,
+      "learning_rate": 3.3889845094664373e-05,
+      "loss": 0.3469,
+      "step": 1260
+    },
+    {
+      "epoch": 1.1709242916860194,
+      "grad_norm": 0.19956161096816913,
+      "learning_rate": 3.3872633390705685e-05,
+      "loss": 0.3665,
+      "step": 1261
+    },
+    {
+      "epoch": 1.1718532280538783,
+      "grad_norm": 0.22850970980578056,
+      "learning_rate": 3.385542168674699e-05,
+      "loss": 0.3311,
+      "step": 1262
+    },
+    {
+      "epoch": 1.172782164421737,
+      "grad_norm": 0.22798171419609636,
+      "learning_rate": 3.38382099827883e-05,
+      "loss": 0.3487,
+      "step": 1263
+    },
+    {
+      "epoch": 1.173711100789596,
+      "grad_norm": 0.23282279392195235,
+      "learning_rate": 3.382099827882961e-05,
+      "loss": 0.3698,
+      "step": 1264
+    },
+    {
+      "epoch": 1.1746400371574548,
+      "grad_norm": 0.2703092707739479,
+      "learning_rate": 3.380378657487092e-05,
+      "loss": 0.3394,
+      "step": 1265
+    },
+    {
+      "epoch": 1.1755689735253134,
+      "grad_norm": 0.22016376237672852,
+      "learning_rate": 3.3786574870912224e-05,
+      "loss": 0.3546,
+      "step": 1266
+    },
+    {
+      "epoch": 1.1764979098931723,
+      "grad_norm": 0.20478813537848256,
+      "learning_rate": 3.3769363166953536e-05,
+      "loss": 0.3528,
+      "step": 1267
+    },
+    {
+      "epoch": 1.1774268462610311,
+      "grad_norm": 0.2561543579105953,
+      "learning_rate": 3.3752151462994834e-05,
+      "loss": 0.383,
+      "step": 1268
+    },
+    {
+      "epoch": 1.17835578262889,
+      "grad_norm": 0.2192358895821466,
+      "learning_rate": 3.3734939759036146e-05,
+      "loss": 0.3548,
+      "step": 1269
+    },
+    {
+      "epoch": 1.1792847189967488,
+      "grad_norm": 0.21639896112955978,
+      "learning_rate": 3.371772805507745e-05,
+      "loss": 0.3438,
+      "step": 1270
+    },
+    {
+      "epoch": 1.1802136553646074,
+      "grad_norm": 0.23897229606319104,
+      "learning_rate": 3.370051635111876e-05,
+      "loss": 0.3408,
+      "step": 1271
+    },
+    {
+      "epoch": 1.1811425917324663,
+      "grad_norm": 0.21230871449016278,
+      "learning_rate": 3.368330464716007e-05,
+      "loss": 0.3529,
+      "step": 1272
+    },
+    {
+      "epoch": 1.1820715281003251,
+      "grad_norm": 0.2225131278174084,
+      "learning_rate": 3.366609294320138e-05,
+      "loss": 0.3443,
+      "step": 1273
+    },
+    {
+      "epoch": 1.183000464468184,
+      "grad_norm": 0.22179982347649585,
+      "learning_rate": 3.3648881239242684e-05,
+      "loss": 0.3667,
+      "step": 1274
+    },
+    {
+      "epoch": 1.1839294008360428,
+      "grad_norm": 0.2363874317205004,
+      "learning_rate": 3.3631669535283996e-05,
+      "loss": 0.3786,
+      "step": 1275
+    },
+    {
+      "epoch": 1.1848583372039014,
+      "grad_norm": 0.2193585078333849,
+      "learning_rate": 3.36144578313253e-05,
+      "loss": 0.3297,
+      "step": 1276
+    },
+    {
+      "epoch": 1.1857872735717603,
+      "grad_norm": 0.2284696248734332,
+      "learning_rate": 3.359724612736661e-05,
+      "loss": 0.3437,
+      "step": 1277
+    },
+    {
+      "epoch": 1.1867162099396191,
+      "grad_norm": 0.20793086153649565,
+      "learning_rate": 3.358003442340792e-05,
+      "loss": 0.3531,
+      "step": 1278
+    },
+    {
+      "epoch": 1.187645146307478,
+      "grad_norm": 0.21567184317250074,
+      "learning_rate": 3.356282271944923e-05,
+      "loss": 0.3406,
+      "step": 1279
+    },
+    {
+      "epoch": 1.1885740826753368,
+      "grad_norm": 0.24433871916587774,
+      "learning_rate": 3.3545611015490535e-05,
+      "loss": 0.3716,
+      "step": 1280
+    },
+    {
+      "epoch": 1.1895030190431957,
+      "grad_norm": 0.22152344071069677,
+      "learning_rate": 3.3528399311531847e-05,
+      "loss": 0.3356,
+      "step": 1281
+    },
+    {
+      "epoch": 1.1904319554110543,
+      "grad_norm": 0.21020957516671623,
+      "learning_rate": 3.351118760757315e-05,
+      "loss": 0.374,
+      "step": 1282
+    },
+    {
+      "epoch": 1.1913608917789131,
+      "grad_norm": 0.5970058115054317,
+      "learning_rate": 3.349397590361446e-05,
+      "loss": 0.3366,
+      "step": 1283
+    },
+    {
+      "epoch": 1.192289828146772,
+      "grad_norm": 0.22658123670184568,
+      "learning_rate": 3.347676419965577e-05,
+      "loss": 0.345,
+      "step": 1284
+    },
+    {
+      "epoch": 1.1932187645146308,
+      "grad_norm": 0.20134140896250374,
+      "learning_rate": 3.345955249569708e-05,
+      "loss": 0.3369,
+      "step": 1285
+    },
+    {
+      "epoch": 1.1941477008824894,
+      "grad_norm": 0.2203877272353749,
+      "learning_rate": 3.3442340791738385e-05,
+      "loss": 0.343,
+      "step": 1286
+    },
+    {
+      "epoch": 1.1950766372503483,
+      "grad_norm": 0.20330351316025363,
+      "learning_rate": 3.34251290877797e-05,
+      "loss": 0.3546,
+      "step": 1287
+    },
+    {
+      "epoch": 1.1960055736182071,
+      "grad_norm": 0.20833906150996628,
+      "learning_rate": 3.3407917383821e-05,
+      "loss": 0.3653,
+      "step": 1288
+    },
+    {
+      "epoch": 1.196934509986066,
+      "grad_norm": 0.21852984142409648,
+      "learning_rate": 3.339070567986231e-05,
+      "loss": 0.3576,
+      "step": 1289
+    },
+    {
+      "epoch": 1.1978634463539248,
+      "grad_norm": 0.23067032362394563,
+      "learning_rate": 3.337349397590361e-05,
+      "loss": 0.3237,
+      "step": 1290
+    },
+    {
+      "epoch": 1.1987923827217837,
+      "grad_norm": 0.21203185556352686,
+      "learning_rate": 3.3356282271944924e-05,
+      "loss": 0.3413,
+      "step": 1291
+    },
+    {
+      "epoch": 1.1997213190896423,
+      "grad_norm": 0.19814905094151894,
+      "learning_rate": 3.333907056798623e-05,
+      "loss": 0.3187,
+      "step": 1292
+    },
+    {
+      "epoch": 1.2006502554575011,
+      "grad_norm": 0.23683735633528624,
+      "learning_rate": 3.332185886402754e-05,
+      "loss": 0.3692,
+      "step": 1293
+    },
+    {
+      "epoch": 1.20157919182536,
+      "grad_norm": 0.23935032392727087,
+      "learning_rate": 3.3304647160068846e-05,
+      "loss": 0.341,
+      "step": 1294
+    },
+    {
+      "epoch": 1.2025081281932188,
+      "grad_norm": 0.21718405839661278,
+      "learning_rate": 3.328743545611016e-05,
+      "loss": 0.339,
+      "step": 1295
+    },
+    {
+      "epoch": 1.2034370645610775,
+      "grad_norm": 0.2333890322775293,
+      "learning_rate": 3.327022375215146e-05,
+      "loss": 0.3522,
+      "step": 1296
+    },
+    {
+      "epoch": 1.2043660009289363,
+      "grad_norm": 0.23203001135682452,
+      "learning_rate": 3.3253012048192774e-05,
+      "loss": 0.3328,
+      "step": 1297
+    },
+    {
+      "epoch": 1.2052949372967952,
+      "grad_norm": 0.20690983108544067,
+      "learning_rate": 3.323580034423408e-05,
+      "loss": 0.3527,
+      "step": 1298
+    },
+    {
+      "epoch": 1.206223873664654,
+      "grad_norm": 0.2466034408066261,
+      "learning_rate": 3.321858864027539e-05,
+      "loss": 0.3629,
+      "step": 1299
+    },
+    {
+      "epoch": 1.2071528100325128,
+      "grad_norm": 0.21156982100561744,
+      "learning_rate": 3.3201376936316696e-05,
+      "loss": 0.3669,
+      "step": 1300
+    },
+    {
+      "epoch": 1.2080817464003717,
+      "grad_norm": 0.2160203305108374,
+      "learning_rate": 3.318416523235801e-05,
+      "loss": 0.3528,
+      "step": 1301
+    },
+    {
+      "epoch": 1.2090106827682303,
+      "grad_norm": 0.24371827812631483,
+      "learning_rate": 3.316695352839931e-05,
+      "loss": 0.3665,
+      "step": 1302
+    },
+    {
+      "epoch": 1.2099396191360892,
+      "grad_norm": 0.22457004262344515,
+      "learning_rate": 3.3149741824440625e-05,
+      "loss": 0.3458,
+      "step": 1303
+    },
+    {
+      "epoch": 1.210868555503948,
+      "grad_norm": 0.22055226828004412,
+      "learning_rate": 3.313253012048193e-05,
+      "loss": 0.3738,
+      "step": 1304
+    },
+    {
+      "epoch": 1.2117974918718069,
+      "grad_norm": 0.22285967677243848,
+      "learning_rate": 3.311531841652324e-05,
+      "loss": 0.3616,
+      "step": 1305
+    },
+    {
+      "epoch": 1.2127264282396655,
+      "grad_norm": 0.2161533855354827,
+      "learning_rate": 3.3098106712564546e-05,
+      "loss": 0.3699,
+      "step": 1306
+    },
+    {
+      "epoch": 1.2136553646075243,
+      "grad_norm": 0.22228225680627903,
+      "learning_rate": 3.308089500860586e-05,
+      "loss": 0.3363,
+      "step": 1307
+    },
+    {
+      "epoch": 1.2145843009753832,
+      "grad_norm": 0.20607924665127353,
+      "learning_rate": 3.306368330464716e-05,
+      "loss": 0.3303,
+      "step": 1308
+    },
+    {
+      "epoch": 1.215513237343242,
+      "grad_norm": 0.18066272382062054,
+      "learning_rate": 3.3046471600688475e-05,
+      "loss": 0.3569,
+      "step": 1309
+    },
+    {
+      "epoch": 1.2164421737111009,
+      "grad_norm": 0.21118001457268404,
+      "learning_rate": 3.302925989672978e-05,
+      "loss": 0.347,
+      "step": 1310
+    },
+    {
+      "epoch": 1.2173711100789597,
+      "grad_norm": 0.20561719917149285,
+      "learning_rate": 3.3012048192771085e-05,
+      "loss": 0.3546,
+      "step": 1311
+    },
+    {
+      "epoch": 1.2183000464468183,
+      "grad_norm": 0.22731459097307283,
+      "learning_rate": 3.299483648881239e-05,
+      "loss": 0.3745,
+      "step": 1312
+    },
+    {
+      "epoch": 1.2192289828146772,
+      "grad_norm": 0.2424041406528847,
+      "learning_rate": 3.29776247848537e-05,
+      "loss": 0.3486,
+      "step": 1313
+    },
+    {
+      "epoch": 1.220157919182536,
+      "grad_norm": 0.20069413390660085,
+      "learning_rate": 3.296041308089501e-05,
+      "loss": 0.3383,
+      "step": 1314
+    },
+    {
+      "epoch": 1.2210868555503949,
+      "grad_norm": 0.2281772093622241,
+      "learning_rate": 3.294320137693632e-05,
+      "loss": 0.3434,
+      "step": 1315
+    },
+    {
+      "epoch": 1.2220157919182535,
+      "grad_norm": 0.19211768424635667,
+      "learning_rate": 3.2925989672977624e-05,
+      "loss": 0.3352,
+      "step": 1316
+    },
+    {
+      "epoch": 1.2229447282861123,
+      "grad_norm": 0.2513611309761323,
+      "learning_rate": 3.2908777969018935e-05,
+      "loss": 0.345,
+      "step": 1317
+    },
+    {
+      "epoch": 1.2238736646539712,
+      "grad_norm": 0.24961481880569977,
+      "learning_rate": 3.289156626506024e-05,
+      "loss": 0.3231,
+      "step": 1318
+    },
+    {
+      "epoch": 1.22480260102183,
+      "grad_norm": 0.20092458352098808,
+      "learning_rate": 3.287435456110155e-05,
+      "loss": 0.3563,
+      "step": 1319
+    },
+    {
+      "epoch": 1.2257315373896889,
+      "grad_norm": 0.2436154621559124,
+      "learning_rate": 3.285714285714286e-05,
+      "loss": 0.3831,
+      "step": 1320
+    },
+    {
+      "epoch": 1.2266604737575477,
+      "grad_norm": 0.24743764188275466,
+      "learning_rate": 3.283993115318417e-05,
+      "loss": 0.3574,
+      "step": 1321
+    },
+    {
+      "epoch": 1.2275894101254063,
+      "grad_norm": 0.19895641348642534,
+      "learning_rate": 3.2822719449225474e-05,
+      "loss": 0.3482,
+      "step": 1322
+    },
+    {
+      "epoch": 1.2285183464932652,
+      "grad_norm": 0.22955056964532125,
+      "learning_rate": 3.2805507745266786e-05,
+      "loss": 0.3532,
+      "step": 1323
+    },
+    {
+      "epoch": 1.229447282861124,
+      "grad_norm": 0.2100380867776916,
+      "learning_rate": 3.278829604130809e-05,
+      "loss": 0.3348,
+      "step": 1324
+    },
+    {
+      "epoch": 1.2303762192289829,
+      "grad_norm": 0.18910582670755707,
+      "learning_rate": 3.27710843373494e-05,
+      "loss": 0.343,
+      "step": 1325
+    },
+    {
+      "epoch": 1.2313051555968415,
+      "grad_norm": 0.23613291094866282,
+      "learning_rate": 3.275387263339071e-05,
+      "loss": 0.3567,
+      "step": 1326
+    },
+    {
+      "epoch": 1.2322340919647004,
+      "grad_norm": 0.19368690581896822,
+      "learning_rate": 3.273666092943202e-05,
+      "loss": 0.3561,
+      "step": 1327
+    },
+    {
+      "epoch": 1.2331630283325592,
+      "grad_norm": 0.20487266897523065,
+      "learning_rate": 3.2719449225473324e-05,
+      "loss": 0.3477,
+      "step": 1328
+    },
+    {
+      "epoch": 1.234091964700418,
+      "grad_norm": 0.19205460972365668,
+      "learning_rate": 3.2702237521514636e-05,
+      "loss": 0.353,
+      "step": 1329
+    },
+    {
+      "epoch": 1.2350209010682769,
+      "grad_norm": 0.22085132737761864,
+      "learning_rate": 3.268502581755594e-05,
+      "loss": 0.3499,
+      "step": 1330
+    },
+    {
+      "epoch": 1.2359498374361357,
+      "grad_norm": 0.19737628387896256,
+      "learning_rate": 3.266781411359725e-05,
+      "loss": 0.3396,
+      "step": 1331
+    },
+    {
+      "epoch": 1.2368787738039944,
+      "grad_norm": 0.19772707162756117,
+      "learning_rate": 3.265060240963855e-05,
+      "loss": 0.3364,
+      "step": 1332
+    },
+    {
+      "epoch": 1.2378077101718532,
+      "grad_norm": 0.1990220518002121,
+      "learning_rate": 3.263339070567986e-05,
+      "loss": 0.3543,
+      "step": 1333
+    },
+    {
+      "epoch": 1.238736646539712,
+      "grad_norm": 0.2088922466620167,
+      "learning_rate": 3.261617900172117e-05,
+      "loss": 0.3423,
+      "step": 1334
+    },
+    {
+      "epoch": 1.239665582907571,
+      "grad_norm": 0.23503440075942228,
+      "learning_rate": 3.259896729776248e-05,
+      "loss": 0.378,
+      "step": 1335
+    },
+    {
+      "epoch": 1.2405945192754295,
+      "grad_norm": 0.20957487854697568,
+      "learning_rate": 3.2581755593803785e-05,
+      "loss": 0.3497,
+      "step": 1336
+    },
+    {
+      "epoch": 1.2415234556432884,
+      "grad_norm": 0.20536395635402577,
+      "learning_rate": 3.2564543889845097e-05,
+      "loss": 0.3621,
+      "step": 1337
+    },
+    {
+      "epoch": 1.2424523920111472,
+      "grad_norm": 0.19566124028745266,
+      "learning_rate": 3.25473321858864e-05,
+      "loss": 0.3392,
+      "step": 1338
+    },
+    {
+      "epoch": 1.243381328379006,
+      "grad_norm": 0.19859624787750643,
+      "learning_rate": 3.253012048192771e-05,
+      "loss": 0.3305,
+      "step": 1339
+    },
+    {
+      "epoch": 1.244310264746865,
+      "grad_norm": 0.20220312062723214,
+      "learning_rate": 3.251290877796902e-05,
+      "loss": 0.3429,
+      "step": 1340
+    },
+    {
+      "epoch": 1.2452392011147237,
+      "grad_norm": 0.20306062927254576,
+      "learning_rate": 3.249569707401033e-05,
+      "loss": 0.3515,
+      "step": 1341
+    },
+    {
+      "epoch": 1.2461681374825824,
+      "grad_norm": 0.22899674919736335,
+      "learning_rate": 3.2478485370051635e-05,
+      "loss": 0.3512,
+      "step": 1342
+    },
+    {
+      "epoch": 1.2470970738504412,
+      "grad_norm": 0.20719067431783017,
+      "learning_rate": 3.246127366609295e-05,
+      "loss": 0.3687,
+      "step": 1343
+    },
+    {
+      "epoch": 1.2480260102183,
+      "grad_norm": 0.19102206124283935,
+      "learning_rate": 3.244406196213425e-05,
+      "loss": 0.3562,
+      "step": 1344
+    },
+    {
+      "epoch": 1.248954946586159,
+      "grad_norm": 0.1986375316054701,
+      "learning_rate": 3.2426850258175564e-05,
+      "loss": 0.3549,
+      "step": 1345
+    },
+    {
+      "epoch": 1.2498838829540175,
+      "grad_norm": 0.20320312735004203,
+      "learning_rate": 3.240963855421687e-05,
+      "loss": 0.344,
+      "step": 1346
+    },
+    {
+      "epoch": 1.2508128193218764,
+      "grad_norm": 0.20833632211864914,
+      "learning_rate": 3.239242685025818e-05,
+      "loss": 0.3419,
+      "step": 1347
+    },
+    {
+      "epoch": 1.2517417556897352,
+      "grad_norm": 0.20209089611428194,
+      "learning_rate": 3.2375215146299486e-05,
+      "loss": 0.378,
+      "step": 1348
+    },
+    {
+      "epoch": 1.252670692057594,
+      "grad_norm": 0.23317589217939863,
+      "learning_rate": 3.23580034423408e-05,
+      "loss": 0.3472,
+      "step": 1349
+    },
+    {
+      "epoch": 1.253599628425453,
+      "grad_norm": 0.23889292842454002,
+      "learning_rate": 3.23407917383821e-05,
+      "loss": 0.3546,
+      "step": 1350
+    },
+    {
+      "epoch": 1.2545285647933118,
+      "grad_norm": 0.23185784512316487,
+      "learning_rate": 3.2323580034423414e-05,
+      "loss": 0.353,
+      "step": 1351
+    },
+    {
+      "epoch": 1.2554575011611704,
+      "grad_norm": 0.25338080904879534,
+      "learning_rate": 3.230636833046472e-05,
+      "loss": 0.3369,
+      "step": 1352
+    },
+    {
+      "epoch": 1.2563864375290292,
+      "grad_norm": 0.2099328571771037,
+      "learning_rate": 3.2289156626506024e-05,
+      "loss": 0.3393,
+      "step": 1353
+    },
+    {
+      "epoch": 1.257315373896888,
+      "grad_norm": 0.24756023680513764,
+      "learning_rate": 3.227194492254733e-05,
+      "loss": 0.3448,
+      "step": 1354
+    },
+    {
+      "epoch": 1.258244310264747,
+      "grad_norm": 0.23036929610075849,
+      "learning_rate": 3.225473321858864e-05,
+      "loss": 0.3428,
+      "step": 1355
+    },
+    {
+      "epoch": 1.2591732466326055,
+      "grad_norm": 0.26373689904363057,
+      "learning_rate": 3.2237521514629946e-05,
+      "loss": 0.3525,
+      "step": 1356
+    },
+    {
+      "epoch": 1.2601021830004644,
+      "grad_norm": 0.20482417525178498,
+      "learning_rate": 3.222030981067126e-05,
+      "loss": 0.3463,
+      "step": 1357
+    },
+    {
+      "epoch": 1.2610311193683232,
+      "grad_norm": 0.24978474205561277,
+      "learning_rate": 3.220309810671256e-05,
+      "loss": 0.3616,
+      "step": 1358
+    },
+    {
+      "epoch": 1.261960055736182,
+      "grad_norm": 0.2200819349873659,
+      "learning_rate": 3.2185886402753875e-05,
+      "loss": 0.3438,
+      "step": 1359
+    },
+    {
+      "epoch": 1.262888992104041,
+      "grad_norm": 0.23973808094498264,
+      "learning_rate": 3.216867469879518e-05,
+      "loss": 0.3552,
+      "step": 1360
+    },
+    {
+      "epoch": 1.2638179284718998,
+      "grad_norm": 0.22114441784933195,
+      "learning_rate": 3.215146299483649e-05,
+      "loss": 0.3752,
+      "step": 1361
+    },
+    {
+      "epoch": 1.2647468648397584,
+      "grad_norm": 0.27751924776177905,
+      "learning_rate": 3.2134251290877796e-05,
+      "loss": 0.3515,
+      "step": 1362
+    },
+    {
+      "epoch": 1.2656758012076172,
+      "grad_norm": 0.21042269563943283,
+      "learning_rate": 3.211703958691911e-05,
+      "loss": 0.3465,
+      "step": 1363
+    },
+    {
+      "epoch": 1.266604737575476,
+      "grad_norm": 0.28456491769206277,
+      "learning_rate": 3.209982788296041e-05,
+      "loss": 0.3694,
+      "step": 1364
+    },
+    {
+      "epoch": 1.267533673943335,
+      "grad_norm": 0.2307337278461796,
+      "learning_rate": 3.2082616179001725e-05,
+      "loss": 0.3686,
+      "step": 1365
+    },
+    {
+      "epoch": 1.2684626103111936,
+      "grad_norm": 0.2295732638687118,
+      "learning_rate": 3.206540447504303e-05,
+      "loss": 0.3595,
+      "step": 1366
+    },
+    {
+      "epoch": 1.2693915466790524,
+      "grad_norm": 0.25011196671438235,
+      "learning_rate": 3.204819277108434e-05,
+      "loss": 0.3676,
+      "step": 1367
+    },
+    {
+      "epoch": 1.2703204830469113,
+      "grad_norm": 0.21793396400635984,
+      "learning_rate": 3.203098106712565e-05,
+      "loss": 0.3505,
+      "step": 1368
+    },
+    {
+      "epoch": 1.27124941941477,
+      "grad_norm": 0.19814582335576053,
+      "learning_rate": 3.201376936316696e-05,
+      "loss": 0.333,
+      "step": 1369
+    },
+    {
+      "epoch": 1.272178355782629,
+      "grad_norm": 0.22394942492694084,
+      "learning_rate": 3.1996557659208264e-05,
+      "loss": 0.358,
+      "step": 1370
+    },
+    {
+      "epoch": 1.2731072921504878,
+      "grad_norm": 0.2021914187434177,
+      "learning_rate": 3.1979345955249575e-05,
+      "loss": 0.3385,
+      "step": 1371
+    },
+    {
+      "epoch": 1.2740362285183464,
+      "grad_norm": 0.18777768433600583,
+      "learning_rate": 3.196213425129088e-05,
+      "loss": 0.3599,
+      "step": 1372
+    },
+    {
+      "epoch": 1.2749651648862053,
+      "grad_norm": 0.20249518397349203,
+      "learning_rate": 3.194492254733219e-05,
+      "loss": 0.3548,
+      "step": 1373
+    },
+    {
+      "epoch": 1.275894101254064,
+      "grad_norm": 0.2269731964103628,
+      "learning_rate": 3.192771084337349e-05,
+      "loss": 0.3537,
+      "step": 1374
+    },
+    {
+      "epoch": 1.276823037621923,
+      "grad_norm": 0.19146870047473466,
+      "learning_rate": 3.19104991394148e-05,
+      "loss": 0.3555,
+      "step": 1375
+    },
+    {
+      "epoch": 1.2777519739897816,
+      "grad_norm": 0.2073759588674991,
+      "learning_rate": 3.189328743545611e-05,
+      "loss": 0.351,
+      "step": 1376
+    },
+    {
+      "epoch": 1.2786809103576404,
+      "grad_norm": 0.2157939027150171,
+      "learning_rate": 3.187607573149742e-05,
+      "loss": 0.368,
+      "step": 1377
+    },
+    {
+      "epoch": 1.2796098467254993,
+      "grad_norm": 0.21802601041303335,
+      "learning_rate": 3.1858864027538724e-05,
+      "loss": 0.3697,
+      "step": 1378
+    },
+    {
+      "epoch": 1.2805387830933581,
+      "grad_norm": 0.20338664825260214,
+      "learning_rate": 3.1841652323580036e-05,
+      "loss": 0.3553,
+      "step": 1379
+    },
+    {
+      "epoch": 1.281467719461217,
+      "grad_norm": 0.21747341979747276,
+      "learning_rate": 3.182444061962134e-05,
+      "loss": 0.3434,
+      "step": 1380
+    },
+    {
+      "epoch": 1.2823966558290758,
+      "grad_norm": 0.19657412183584266,
+      "learning_rate": 3.180722891566265e-05,
+      "loss": 0.354,
+      "step": 1381
+    },
+    {
+      "epoch": 1.2833255921969344,
+      "grad_norm": 0.203905731577841,
+      "learning_rate": 3.179001721170396e-05,
+      "loss": 0.3639,
+      "step": 1382
+    },
+    {
+      "epoch": 1.2842545285647933,
+      "grad_norm": 0.19642280259503783,
+      "learning_rate": 3.177280550774527e-05,
+      "loss": 0.3429,
+      "step": 1383
+    },
+    {
+      "epoch": 1.2851834649326521,
+      "grad_norm": 0.20991675969465276,
+      "learning_rate": 3.1755593803786574e-05,
+      "loss": 0.3521,
+      "step": 1384
+    },
+    {
+      "epoch": 1.286112401300511,
+      "grad_norm": 0.20253615601371597,
+      "learning_rate": 3.1738382099827886e-05,
+      "loss": 0.3473,
+      "step": 1385
+    },
+    {
+      "epoch": 1.2870413376683696,
+      "grad_norm": 0.198688892854731,
+      "learning_rate": 3.172117039586919e-05,
+      "loss": 0.3436,
+      "step": 1386
+    },
+    {
+      "epoch": 1.2879702740362284,
+      "grad_norm": 2.4661548054153672,
+      "learning_rate": 3.17039586919105e-05,
+      "loss": 0.3541,
+      "step": 1387
+    },
+    {
+      "epoch": 1.2888992104040873,
+      "grad_norm": 0.20742528996459592,
+      "learning_rate": 3.168674698795181e-05,
+      "loss": 0.3517,
+      "step": 1388
+    },
+    {
+      "epoch": 1.2898281467719461,
+      "grad_norm": 0.20329525205803542,
+      "learning_rate": 3.166953528399312e-05,
+      "loss": 0.3424,
+      "step": 1389
+    },
+    {
+      "epoch": 1.290757083139805,
+      "grad_norm": 0.2112927190742448,
+      "learning_rate": 3.1652323580034425e-05,
+      "loss": 0.3556,
+      "step": 1390
+    },
+    {
+      "epoch": 1.2916860195076638,
+      "grad_norm": 0.23636619370443224,
+      "learning_rate": 3.163511187607574e-05,
+      "loss": 0.3735,
+      "step": 1391
+    },
+    {
+      "epoch": 1.2926149558755227,
+      "grad_norm": 0.22276923652505054,
+      "learning_rate": 3.161790017211704e-05,
+      "loss": 0.3579,
+      "step": 1392
+    },
+    {
+      "epoch": 1.2935438922433813,
+      "grad_norm": 0.2005541094030619,
+      "learning_rate": 3.1600688468158353e-05,
+      "loss": 0.3554,
+      "step": 1393
+    },
+    {
+      "epoch": 1.2944728286112401,
+      "grad_norm": 0.22155762592559106,
+      "learning_rate": 3.158347676419966e-05,
+      "loss": 0.3546,
+      "step": 1394
+    },
+    {
+      "epoch": 1.295401764979099,
+      "grad_norm": 0.20946083857237227,
+      "learning_rate": 3.1566265060240963e-05,
+      "loss": 0.349,
+      "step": 1395
+    },
+    {
+      "epoch": 1.2963307013469576,
+      "grad_norm": 0.20702763325289733,
+      "learning_rate": 3.154905335628227e-05,
+      "loss": 0.3649,
+      "step": 1396
+    },
+    {
+      "epoch": 1.2972596377148164,
+      "grad_norm": 0.2229383054425248,
+      "learning_rate": 3.153184165232358e-05,
+      "loss": 0.349,
+      "step": 1397
+    },
+    {
+      "epoch": 1.2981885740826753,
+      "grad_norm": 0.23124550374061495,
+      "learning_rate": 3.1514629948364885e-05,
+      "loss": 0.336,
+      "step": 1398
+    },
+    {
+      "epoch": 1.2991175104505341,
+      "grad_norm": 0.19990916764328137,
+      "learning_rate": 3.14974182444062e-05,
+      "loss": 0.3619,
+      "step": 1399
+    },
+    {
+      "epoch": 1.300046446818393,
+      "grad_norm": 0.21799120394800936,
+      "learning_rate": 3.14802065404475e-05,
+      "loss": 0.3492,
+      "step": 1400
+    },
+    {
+      "epoch": 1.3009753831862518,
+      "grad_norm": 0.3118082548093683,
+      "learning_rate": 3.1462994836488814e-05,
+      "loss": 0.3486,
+      "step": 1401
+    },
+    {
+      "epoch": 1.3019043195541107,
+      "grad_norm": 0.21875180339324876,
+      "learning_rate": 3.144578313253012e-05,
+      "loss": 0.345,
+      "step": 1402
+    },
+    {
+      "epoch": 1.3028332559219693,
+      "grad_norm": 0.20622030938134586,
+      "learning_rate": 3.142857142857143e-05,
+      "loss": 0.3395,
+      "step": 1403
+    },
+    {
+      "epoch": 1.3037621922898281,
+      "grad_norm": 0.22507581862454673,
+      "learning_rate": 3.1411359724612736e-05,
+      "loss": 0.3462,
+      "step": 1404
+    },
+    {
+      "epoch": 1.304691128657687,
+      "grad_norm": 0.20997485299055227,
+      "learning_rate": 3.139414802065405e-05,
+      "loss": 0.3248,
+      "step": 1405
+    },
+    {
+      "epoch": 1.3056200650255456,
+      "grad_norm": 0.20777883901322605,
+      "learning_rate": 3.137693631669535e-05,
+      "loss": 0.3632,
+      "step": 1406
+    },
+    {
+      "epoch": 1.3065490013934045,
+      "grad_norm": 0.22237908122375719,
+      "learning_rate": 3.1359724612736664e-05,
+      "loss": 0.3547,
+      "step": 1407
+    },
+    {
+      "epoch": 1.3074779377612633,
+      "grad_norm": 0.22989418042498605,
+      "learning_rate": 3.134251290877797e-05,
+      "loss": 0.3386,
+      "step": 1408
+    },
+    {
+      "epoch": 1.3084068741291222,
+      "grad_norm": 0.18232306340600674,
+      "learning_rate": 3.132530120481928e-05,
+      "loss": 0.3387,
+      "step": 1409
+    },
+    {
+      "epoch": 1.309335810496981,
+      "grad_norm": 0.22379285894618078,
+      "learning_rate": 3.1308089500860586e-05,
+      "loss": 0.351,
+      "step": 1410
+    },
+    {
+      "epoch": 1.3102647468648398,
+      "grad_norm": 0.22957367785503446,
+      "learning_rate": 3.12908777969019e-05,
+      "loss": 0.3538,
+      "step": 1411
+    },
+    {
+      "epoch": 1.3111936832326987,
+      "grad_norm": 0.2213926772964687,
+      "learning_rate": 3.12736660929432e-05,
+      "loss": 0.3501,
+      "step": 1412
+    },
+    {
+      "epoch": 1.3121226196005573,
+      "grad_norm": 0.2199668333695165,
+      "learning_rate": 3.1256454388984515e-05,
+      "loss": 0.3357,
+      "step": 1413
+    },
+    {
+      "epoch": 1.3130515559684162,
+      "grad_norm": 0.22045412078084195,
+      "learning_rate": 3.123924268502582e-05,
+      "loss": 0.3665,
+      "step": 1414
+    },
+    {
+      "epoch": 1.313980492336275,
+      "grad_norm": 0.23937083341287288,
+      "learning_rate": 3.122203098106713e-05,
+      "loss": 0.3541,
+      "step": 1415
+    },
+    {
+      "epoch": 1.3149094287041336,
+      "grad_norm": 0.22142560842558442,
+      "learning_rate": 3.1204819277108436e-05,
+      "loss": 0.3589,
+      "step": 1416
+    },
+    {
+      "epoch": 1.3158383650719925,
+      "grad_norm": 0.25259554318810806,
+      "learning_rate": 3.118760757314974e-05,
+      "loss": 0.3558,
+      "step": 1417
+    },
+    {
+      "epoch": 1.3167673014398513,
+      "grad_norm": 0.23469823002390106,
+      "learning_rate": 3.1170395869191046e-05,
+      "loss": 0.3604,
+      "step": 1418
+    },
+    {
+      "epoch": 1.3176962378077102,
+      "grad_norm": 0.2318292616166613,
+      "learning_rate": 3.115318416523236e-05,
+      "loss": 0.3231,
+      "step": 1419
+    },
+    {
+      "epoch": 1.318625174175569,
+      "grad_norm": 0.23736119043390674,
+      "learning_rate": 3.113597246127366e-05,
+      "loss": 0.3644,
+      "step": 1420
+    },
+    {
+      "epoch": 1.3195541105434279,
+      "grad_norm": 0.2164485499905474,
+      "learning_rate": 3.1118760757314975e-05,
+      "loss": 0.3662,
+      "step": 1421
+    },
+    {
+      "epoch": 1.3204830469112867,
+      "grad_norm": 0.21673871274658302,
+      "learning_rate": 3.110154905335628e-05,
+      "loss": 0.3451,
+      "step": 1422
+    },
+    {
+      "epoch": 1.3214119832791453,
+      "grad_norm": 0.2472895979807864,
+      "learning_rate": 3.108433734939759e-05,
+      "loss": 0.3543,
+      "step": 1423
+    },
+    {
+      "epoch": 1.3223409196470042,
+      "grad_norm": 0.2245120667209746,
+      "learning_rate": 3.10671256454389e-05,
+      "loss": 0.3624,
+      "step": 1424
+    },
+    {
+      "epoch": 1.323269856014863,
+      "grad_norm": 0.24985123445714202,
+      "learning_rate": 3.104991394148021e-05,
+      "loss": 0.3387,
+      "step": 1425
+    },
+    {
+      "epoch": 1.3241987923827216,
+      "grad_norm": 0.2142027736341417,
+      "learning_rate": 3.1032702237521514e-05,
+      "loss": 0.3576,
+      "step": 1426
+    },
+    {
+      "epoch": 1.3251277287505805,
+      "grad_norm": 0.21501285847890808,
+      "learning_rate": 3.1015490533562825e-05,
+      "loss": 0.3259,
+      "step": 1427
+    },
+    {
+      "epoch": 1.3260566651184393,
+      "grad_norm": 0.22544501880772352,
+      "learning_rate": 3.099827882960413e-05,
+      "loss": 0.3363,
+      "step": 1428
+    },
+    {
+      "epoch": 1.3269856014862982,
+      "grad_norm": 0.23996880906387041,
+      "learning_rate": 3.098106712564544e-05,
+      "loss": 0.3681,
+      "step": 1429
+    },
+    {
+      "epoch": 1.327914537854157,
+      "grad_norm": 0.22942067334804397,
+      "learning_rate": 3.096385542168675e-05,
+      "loss": 0.3505,
+      "step": 1430
+    },
+    {
+      "epoch": 1.3288434742220159,
+      "grad_norm": 0.23432507600874153,
+      "learning_rate": 3.094664371772806e-05,
+      "loss": 0.3703,
+      "step": 1431
+    },
+    {
+      "epoch": 1.3297724105898747,
+      "grad_norm": 0.19874580942818373,
+      "learning_rate": 3.0929432013769364e-05,
+      "loss": 0.3629,
+      "step": 1432
+    },
+    {
+      "epoch": 1.3307013469577333,
+      "grad_norm": 0.2312124529538691,
+      "learning_rate": 3.0912220309810676e-05,
+      "loss": 0.3382,
+      "step": 1433
+    },
+    {
+      "epoch": 1.3316302833255922,
+      "grad_norm": 0.20766547973250402,
+      "learning_rate": 3.089500860585198e-05,
+      "loss": 0.3572,
+      "step": 1434
+    },
+    {
+      "epoch": 1.332559219693451,
+      "grad_norm": 0.20786382857619148,
+      "learning_rate": 3.087779690189329e-05,
+      "loss": 0.358,
+      "step": 1435
+    },
+    {
+      "epoch": 1.3334881560613099,
+      "grad_norm": 0.20447235064175212,
+      "learning_rate": 3.08605851979346e-05,
+      "loss": 0.3364,
+      "step": 1436
+    },
+    {
+      "epoch": 1.3344170924291685,
+      "grad_norm": 0.1996680493576843,
+      "learning_rate": 3.084337349397591e-05,
+      "loss": 0.3557,
+      "step": 1437
+    },
+    {
+      "epoch": 1.3353460287970274,
+      "grad_norm": 0.19501800099094735,
+      "learning_rate": 3.082616179001721e-05,
+      "loss": 0.3515,
+      "step": 1438
+    },
+    {
+      "epoch": 1.3362749651648862,
+      "grad_norm": 0.23158157472952676,
+      "learning_rate": 3.080895008605852e-05,
+      "loss": 0.3582,
+      "step": 1439
+    },
+    {
+      "epoch": 1.337203901532745,
+      "grad_norm": 0.6667951671559779,
+      "learning_rate": 3.0791738382099825e-05,
+      "loss": 0.3487,
+      "step": 1440
+    },
+    {
+      "epoch": 1.3381328379006039,
+      "grad_norm": 0.20918867974295047,
+      "learning_rate": 3.0774526678141136e-05,
+      "loss": 0.3528,
+      "step": 1441
+    },
+    {
+      "epoch": 1.3390617742684627,
+      "grad_norm": 0.22293293277353693,
+      "learning_rate": 3.075731497418244e-05,
+      "loss": 0.3409,
+      "step": 1442
+    },
+    {
+      "epoch": 1.3399907106363214,
+      "grad_norm": 0.1829660308906723,
+      "learning_rate": 3.074010327022375e-05,
+      "loss": 0.3455,
+      "step": 1443
+    },
+    {
+      "epoch": 1.3409196470041802,
+      "grad_norm": 0.2000567256104809,
+      "learning_rate": 3.072289156626506e-05,
+      "loss": 0.3643,
+      "step": 1444
+    },
+    {
+      "epoch": 1.341848583372039,
+      "grad_norm": 0.2115519953166791,
+      "learning_rate": 3.070567986230637e-05,
+      "loss": 0.3661,
+      "step": 1445
+    },
+    {
+      "epoch": 1.342777519739898,
+      "grad_norm": 0.21732387886026813,
+      "learning_rate": 3.0688468158347675e-05,
+      "loss": 0.3316,
+      "step": 1446
+    },
+    {
+      "epoch": 1.3437064561077565,
+      "grad_norm": 0.20075318769521555,
+      "learning_rate": 3.067125645438899e-05,
+      "loss": 0.3554,
+      "step": 1447
+    },
+    {
+      "epoch": 1.3446353924756154,
+      "grad_norm": 0.20861534949300697,
+      "learning_rate": 3.065404475043029e-05,
+      "loss": 0.3505,
+      "step": 1448
+    },
+    {
+      "epoch": 1.3455643288434742,
+      "grad_norm": 0.254657487659329,
+      "learning_rate": 3.0636833046471604e-05,
+      "loss": 0.3625,
+      "step": 1449
+    },
+    {
+      "epoch": 1.346493265211333,
+      "grad_norm": 0.2070643343223196,
+      "learning_rate": 3.061962134251291e-05,
+      "loss": 0.3239,
+      "step": 1450
+    },
+    {
+      "epoch": 1.347422201579192,
+      "grad_norm": 0.1928014077888377,
+      "learning_rate": 3.060240963855422e-05,
+      "loss": 0.3558,
+      "step": 1451
+    },
+    {
+      "epoch": 1.3483511379470507,
+      "grad_norm": 0.2294446621025945,
+      "learning_rate": 3.0585197934595525e-05,
+      "loss": 0.3674,
+      "step": 1452
+    },
+    {
+      "epoch": 1.3492800743149094,
+      "grad_norm": 0.18874531273124526,
+      "learning_rate": 3.056798623063684e-05,
+      "loss": 0.3561,
+      "step": 1453
+    },
+    {
+      "epoch": 1.3502090106827682,
+      "grad_norm": 0.22251661860503666,
+      "learning_rate": 3.055077452667814e-05,
+      "loss": 0.3603,
+      "step": 1454
+    },
+    {
+      "epoch": 1.351137947050627,
+      "grad_norm": 0.2235668510889753,
+      "learning_rate": 3.0533562822719454e-05,
+      "loss": 0.3517,
+      "step": 1455
+    },
+    {
+      "epoch": 1.352066883418486,
+      "grad_norm": 0.21432492708979067,
+      "learning_rate": 3.0516351118760762e-05,
+      "loss": 0.3496,
+      "step": 1456
+    },
+    {
+      "epoch": 1.3529958197863445,
+      "grad_norm": 0.20494031548874958,
+      "learning_rate": 3.049913941480207e-05,
+      "loss": 0.3342,
+      "step": 1457
+    },
+    {
+      "epoch": 1.3539247561542034,
+      "grad_norm": 0.20672906781500622,
+      "learning_rate": 3.048192771084338e-05,
+      "loss": 0.366,
+      "step": 1458
+    },
+    {
+      "epoch": 1.3548536925220622,
+      "grad_norm": 0.2042325486051598,
+      "learning_rate": 3.046471600688468e-05,
+      "loss": 0.3293,
+      "step": 1459
+    },
+    {
+      "epoch": 1.355782628889921,
+      "grad_norm": 0.20799195423693348,
+      "learning_rate": 3.044750430292599e-05,
+      "loss": 0.3273,
+      "step": 1460
+    },
+    {
+      "epoch": 1.35671156525778,
+      "grad_norm": 0.1915937763700306,
+      "learning_rate": 3.0430292598967298e-05,
+      "loss": 0.3195,
+      "step": 1461
+    },
+    {
+      "epoch": 1.3576405016256388,
+      "grad_norm": 0.18941206561090246,
+      "learning_rate": 3.0413080895008606e-05,
+      "loss": 0.3603,
+      "step": 1462
+    },
+    {
+      "epoch": 1.3585694379934974,
+      "grad_norm": 0.22011411157308233,
+      "learning_rate": 3.0395869191049914e-05,
+      "loss": 0.3389,
+      "step": 1463
+    },
+    {
+      "epoch": 1.3594983743613562,
+      "grad_norm": 0.20454140375565522,
+      "learning_rate": 3.0378657487091223e-05,
+      "loss": 0.3412,
+      "step": 1464
+    },
+    {
+      "epoch": 1.360427310729215,
+      "grad_norm": 0.204807974919683,
+      "learning_rate": 3.036144578313253e-05,
+      "loss": 0.3688,
+      "step": 1465
+    },
+    {
+      "epoch": 1.361356247097074,
+      "grad_norm": 0.2500844338812536,
+      "learning_rate": 3.034423407917384e-05,
+      "loss": 0.3528,
+      "step": 1466
+    },
+    {
+      "epoch": 1.3622851834649325,
+      "grad_norm": 0.19177649858240708,
+      "learning_rate": 3.0327022375215148e-05,
+      "loss": 0.3371,
+      "step": 1467
+    },
+    {
+      "epoch": 1.3632141198327914,
+      "grad_norm": 0.23935175027569758,
+      "learning_rate": 3.0309810671256456e-05,
+      "loss": 0.3659,
+      "step": 1468
+    },
+    {
+      "epoch": 1.3641430562006502,
+      "grad_norm": 0.2132668493989873,
+      "learning_rate": 3.0292598967297765e-05,
+      "loss": 0.3447,
+      "step": 1469
+    },
+    {
+      "epoch": 1.365071992568509,
+      "grad_norm": 0.2060605613886273,
+      "learning_rate": 3.0275387263339073e-05,
+      "loss": 0.3522,
+      "step": 1470
+    },
+    {
+      "epoch": 1.366000928936368,
+      "grad_norm": 0.19961249600782785,
+      "learning_rate": 3.025817555938038e-05,
+      "loss": 0.3369,
+      "step": 1471
+    },
+    {
+      "epoch": 1.3669298653042268,
+      "grad_norm": 0.20880345374087075,
+      "learning_rate": 3.024096385542169e-05,
+      "loss": 0.3371,
+      "step": 1472
+    },
+    {
+      "epoch": 1.3678588016720854,
+      "grad_norm": 0.18622587429732448,
+      "learning_rate": 3.0223752151463e-05,
+      "loss": 0.3546,
+      "step": 1473
+    },
+    {
+      "epoch": 1.3687877380399442,
+      "grad_norm": 0.21866694803984443,
+      "learning_rate": 3.0206540447504307e-05,
+      "loss": 0.3582,
+      "step": 1474
+    },
+    {
+      "epoch": 1.369716674407803,
+      "grad_norm": 0.1952217882301225,
+      "learning_rate": 3.0189328743545615e-05,
+      "loss": 0.338,
+      "step": 1475
+    },
+    {
+      "epoch": 1.370645610775662,
+      "grad_norm": 0.1935932378498297,
+      "learning_rate": 3.0172117039586924e-05,
+      "loss": 0.3311,
+      "step": 1476
+    },
+    {
+      "epoch": 1.3715745471435206,
+      "grad_norm": 0.21516220551975052,
+      "learning_rate": 3.0154905335628232e-05,
+      "loss": 0.3438,
+      "step": 1477
+    },
+    {
+      "epoch": 1.3725034835113794,
+      "grad_norm": 0.18251550631463553,
+      "learning_rate": 3.013769363166954e-05,
+      "loss": 0.355,
+      "step": 1478
+    },
+    {
+      "epoch": 1.3734324198792383,
+      "grad_norm": 0.21335773806259328,
+      "learning_rate": 3.012048192771085e-05,
+      "loss": 0.3555,
+      "step": 1479
+    },
+    {
+      "epoch": 1.374361356247097,
+      "grad_norm": 0.19458561617263528,
+      "learning_rate": 3.010327022375215e-05,
+      "loss": 0.3334,
+      "step": 1480
+    },
+    {
+      "epoch": 1.375290292614956,
+      "grad_norm": 0.2006684368240207,
+      "learning_rate": 3.008605851979346e-05,
+      "loss": 0.3541,
+      "step": 1481
+    },
+    {
+      "epoch": 1.3762192289828148,
+      "grad_norm": 0.20772428446640925,
+      "learning_rate": 3.0068846815834767e-05,
+      "loss": 0.3289,
+      "step": 1482
+    },
+    {
+      "epoch": 1.3771481653506734,
+      "grad_norm": 0.1940082452881049,
+      "learning_rate": 3.0051635111876076e-05,
+      "loss": 0.3464,
+      "step": 1483
+    },
+    {
+      "epoch": 1.3780771017185323,
+      "grad_norm": 0.21437646933932283,
+      "learning_rate": 3.0034423407917384e-05,
+      "loss": 0.3589,
+      "step": 1484
+    },
+    {
+      "epoch": 1.379006038086391,
+      "grad_norm": 0.20429035849628274,
+      "learning_rate": 3.0017211703958692e-05,
+      "loss": 0.3367,
+      "step": 1485
+    },
+    {
+      "epoch": 1.37993497445425,
+      "grad_norm": 0.19153881560052258,
+      "learning_rate": 3e-05,
+      "loss": 0.3595,
+      "step": 1486
+    },
+    {
+      "epoch": 1.3808639108221086,
+      "grad_norm": 0.21368509231169802,
+      "learning_rate": 2.998278829604131e-05,
+      "loss": 0.338,
+      "step": 1487
+    },
+    {
+      "epoch": 1.3817928471899674,
+      "grad_norm": 0.19137787166733314,
+      "learning_rate": 2.9965576592082618e-05,
+      "loss": 0.3409,
+      "step": 1488
+    },
+    {
+      "epoch": 1.3827217835578263,
+      "grad_norm": 0.20518740175365516,
+      "learning_rate": 2.9948364888123926e-05,
+      "loss": 0.3458,
+      "step": 1489
+    },
+    {
+      "epoch": 1.3836507199256851,
+      "grad_norm": 0.21123625494807322,
+      "learning_rate": 2.9931153184165234e-05,
+      "loss": 0.3577,
+      "step": 1490
+    },
+    {
+      "epoch": 1.384579656293544,
+      "grad_norm": 0.20634934940948052,
+      "learning_rate": 2.9913941480206543e-05,
+      "loss": 0.3472,
+      "step": 1491
+    },
+    {
+      "epoch": 1.3855085926614028,
+      "grad_norm": 0.20463549838354728,
+      "learning_rate": 2.989672977624785e-05,
+      "loss": 0.3457,
+      "step": 1492
+    },
+    {
+      "epoch": 1.3864375290292614,
+      "grad_norm": 0.24704845466896294,
+      "learning_rate": 2.987951807228916e-05,
+      "loss": 0.3598,
+      "step": 1493
+    },
+    {
+      "epoch": 1.3873664653971203,
+      "grad_norm": 0.1988515336691101,
+      "learning_rate": 2.9862306368330468e-05,
+      "loss": 0.3372,
+      "step": 1494
+    },
+    {
+      "epoch": 1.3882954017649791,
+      "grad_norm": 0.20847564536416982,
+      "learning_rate": 2.9845094664371776e-05,
+      "loss": 0.3428,
+      "step": 1495
+    },
+    {
+      "epoch": 1.389224338132838,
+      "grad_norm": 0.21413822989945702,
+      "learning_rate": 2.9827882960413085e-05,
+      "loss": 0.3787,
+      "step": 1496
+    },
+    {
+      "epoch": 1.3901532745006966,
+      "grad_norm": 0.22368254215474134,
+      "learning_rate": 2.9810671256454393e-05,
+      "loss": 0.3139,
+      "step": 1497
+    },
+    {
+      "epoch": 1.3910822108685554,
+      "grad_norm": 0.19657862058957046,
+      "learning_rate": 2.97934595524957e-05,
+      "loss": 0.3386,
+      "step": 1498
+    },
+    {
+      "epoch": 1.3920111472364143,
+      "grad_norm": 0.20285755404055592,
+      "learning_rate": 2.977624784853701e-05,
+      "loss": 0.3551,
+      "step": 1499
+    },
+    {
+      "epoch": 1.3929400836042731,
+      "grad_norm": 0.24237808422101634,
+      "learning_rate": 2.975903614457832e-05,
+      "loss": 0.3325,
+      "step": 1500
+    },
+    {
+      "epoch": 1.393869019972132,
+      "grad_norm": 0.1939612180735908,
+      "learning_rate": 2.974182444061962e-05,
+      "loss": 0.3643,
+      "step": 1501
+    },
+    {
+      "epoch": 1.3947979563399908,
+      "grad_norm": 0.25538025094523903,
+      "learning_rate": 2.972461273666093e-05,
+      "loss": 0.3472,
+      "step": 1502
+    },
+    {
+      "epoch": 1.3957268927078494,
+      "grad_norm": 0.2411165671533172,
+      "learning_rate": 2.9707401032702237e-05,
+      "loss": 0.3618,
+      "step": 1503
+    },
+    {
+      "epoch": 1.3966558290757083,
+      "grad_norm": 0.18225449502294644,
+      "learning_rate": 2.9690189328743545e-05,
+      "loss": 0.3378,
+      "step": 1504
+    },
+    {
+      "epoch": 1.3975847654435671,
+      "grad_norm": 0.2222434673585961,
+      "learning_rate": 2.9672977624784854e-05,
+      "loss": 0.3501,
+      "step": 1505
+    },
+    {
+      "epoch": 1.398513701811426,
+      "grad_norm": 0.22332973721087676,
+      "learning_rate": 2.9655765920826162e-05,
+      "loss": 0.3361,
+      "step": 1506
+    },
+    {
+      "epoch": 1.3994426381792846,
+      "grad_norm": 0.204654731810479,
+      "learning_rate": 2.963855421686747e-05,
+      "loss": 0.3594,
+      "step": 1507
+    },
+    {
+      "epoch": 1.4003715745471434,
+      "grad_norm": 0.18130868120983315,
+      "learning_rate": 2.962134251290878e-05,
+      "loss": 0.3504,
+      "step": 1508
+    },
+    {
+      "epoch": 1.4013005109150023,
+      "grad_norm": 0.22439741167196856,
+      "learning_rate": 2.9604130808950087e-05,
+      "loss": 0.3291,
+      "step": 1509
+    },
+    {
+      "epoch": 1.4022294472828611,
+      "grad_norm": 0.2513401337007612,
+      "learning_rate": 2.9586919104991396e-05,
+      "loss": 0.3703,
+      "step": 1510
+    },
+    {
+      "epoch": 1.40315838365072,
+      "grad_norm": 0.20610567460377152,
+      "learning_rate": 2.9569707401032704e-05,
+      "loss": 0.3347,
+      "step": 1511
+    },
+    {
+      "epoch": 1.4040873200185788,
+      "grad_norm": 0.2134355620033519,
+      "learning_rate": 2.9552495697074012e-05,
+      "loss": 0.3328,
+      "step": 1512
+    },
+    {
+      "epoch": 1.4050162563864375,
+      "grad_norm": 0.26437614395333625,
+      "learning_rate": 2.953528399311532e-05,
+      "loss": 0.3456,
+      "step": 1513
+    },
+    {
+      "epoch": 1.4059451927542963,
+      "grad_norm": 0.20273923027295387,
+      "learning_rate": 2.951807228915663e-05,
+      "loss": 0.3522,
+      "step": 1514
+    },
+    {
+      "epoch": 1.4068741291221551,
+      "grad_norm": 0.23385915941470425,
+      "learning_rate": 2.9500860585197938e-05,
+      "loss": 0.3654,
+      "step": 1515
+    },
+    {
+      "epoch": 1.407803065490014,
+      "grad_norm": 0.24126554734693004,
+      "learning_rate": 2.9483648881239246e-05,
+      "loss": 0.3486,
+      "step": 1516
+    },
+    {
+      "epoch": 1.4087320018578726,
+      "grad_norm": 0.23539838223695975,
+      "learning_rate": 2.9466437177280554e-05,
+      "loss": 0.3526,
+      "step": 1517
+    },
+    {
+      "epoch": 1.4096609382257315,
+      "grad_norm": 0.202514576386571,
+      "learning_rate": 2.9449225473321863e-05,
+      "loss": 0.347,
+      "step": 1518
+    },
+    {
+      "epoch": 1.4105898745935903,
+      "grad_norm": 0.21679264136440565,
+      "learning_rate": 2.943201376936317e-05,
+      "loss": 0.3446,
+      "step": 1519
+    },
+    {
+      "epoch": 1.4115188109614492,
+      "grad_norm": 0.20985791433208453,
+      "learning_rate": 2.941480206540448e-05,
+      "loss": 0.3336,
+      "step": 1520
+    },
+    {
+      "epoch": 1.412447747329308,
+      "grad_norm": 0.20082124256632042,
+      "learning_rate": 2.9397590361445788e-05,
+      "loss": 0.3627,
+      "step": 1521
+    },
+    {
+      "epoch": 1.4133766836971668,
+      "grad_norm": 0.21260803696246838,
+      "learning_rate": 2.9380378657487096e-05,
+      "loss": 0.3279,
+      "step": 1522
+    },
+    {
+      "epoch": 1.4143056200650255,
+      "grad_norm": 0.19228212989651475,
+      "learning_rate": 2.9363166953528398e-05,
+      "loss": 0.3536,
+      "step": 1523
+    },
+    {
+      "epoch": 1.4152345564328843,
+      "grad_norm": 0.1964531621998223,
+      "learning_rate": 2.9345955249569706e-05,
+      "loss": 0.3244,
+      "step": 1524
+    },
+    {
+      "epoch": 1.4161634928007432,
+      "grad_norm": 0.20227705650558617,
+      "learning_rate": 2.9328743545611015e-05,
+      "loss": 0.3637,
+      "step": 1525
+    },
+    {
+      "epoch": 1.417092429168602,
+      "grad_norm": 0.19644430453193695,
+      "learning_rate": 2.9311531841652323e-05,
+      "loss": 0.3379,
+      "step": 1526
+    },
+    {
+      "epoch": 1.4180213655364606,
+      "grad_norm": 0.19015970604275795,
+      "learning_rate": 2.929432013769363e-05,
+      "loss": 0.3505,
+      "step": 1527
+    },
+    {
+      "epoch": 1.4189503019043195,
+      "grad_norm": 0.19835877082909076,
+      "learning_rate": 2.927710843373494e-05,
+      "loss": 0.3479,
+      "step": 1528
+    },
+    {
+      "epoch": 1.4198792382721783,
+      "grad_norm": 0.20490877773794255,
+      "learning_rate": 2.925989672977625e-05,
+      "loss": 0.3656,
+      "step": 1529
+    },
+    {
+      "epoch": 1.4208081746400372,
+      "grad_norm": 0.2700572256039223,
+      "learning_rate": 2.9242685025817557e-05,
+      "loss": 0.35,
+      "step": 1530
+    },
+    {
+      "epoch": 1.421737111007896,
+      "grad_norm": 0.19596937388189786,
+      "learning_rate": 2.9225473321858865e-05,
+      "loss": 0.3539,
+      "step": 1531
+    },
+    {
+      "epoch": 1.4226660473757549,
+      "grad_norm": 0.21738021003203584,
+      "learning_rate": 2.9208261617900174e-05,
+      "loss": 0.3581,
+      "step": 1532
+    },
+    {
+      "epoch": 1.4235949837436135,
+      "grad_norm": 0.20702370533181025,
+      "learning_rate": 2.9191049913941482e-05,
+      "loss": 0.3408,
+      "step": 1533
+    },
+    {
+      "epoch": 1.4245239201114723,
+      "grad_norm": 0.18957085537630466,
+      "learning_rate": 2.917383820998279e-05,
+      "loss": 0.3734,
+      "step": 1534
+    },
+    {
+      "epoch": 1.4254528564793312,
+      "grad_norm": 0.22813119163304144,
+      "learning_rate": 2.91566265060241e-05,
+      "loss": 0.3694,
+      "step": 1535
+    },
+    {
+      "epoch": 1.42638179284719,
+      "grad_norm": 0.20400390978454386,
+      "learning_rate": 2.9139414802065407e-05,
+      "loss": 0.3435,
+      "step": 1536
+    },
+    {
+      "epoch": 1.4273107292150486,
+      "grad_norm": 0.18107256509762557,
+      "learning_rate": 2.9122203098106716e-05,
+      "loss": 0.3558,
+      "step": 1537
+    },
+    {
+      "epoch": 1.4282396655829075,
+      "grad_norm": 0.23378269273098054,
+      "learning_rate": 2.9104991394148024e-05,
+      "loss": 0.3488,
+      "step": 1538
+    },
+    {
+      "epoch": 1.4291686019507663,
+      "grad_norm": 0.1881671167013014,
+      "learning_rate": 2.9087779690189332e-05,
+      "loss": 0.3399,
+      "step": 1539
+    },
+    {
+      "epoch": 1.4300975383186252,
+      "grad_norm": 0.19700102593332322,
+      "learning_rate": 2.907056798623064e-05,
+      "loss": 0.3562,
+      "step": 1540
+    },
+    {
+      "epoch": 1.431026474686484,
+      "grad_norm": 0.2013437891388278,
+      "learning_rate": 2.905335628227195e-05,
+      "loss": 0.342,
+      "step": 1541
+    },
+    {
+      "epoch": 1.4319554110543429,
+      "grad_norm": 0.19061395252656796,
+      "learning_rate": 2.9036144578313258e-05,
+      "loss": 0.3286,
+      "step": 1542
+    },
+    {
+      "epoch": 1.4328843474222015,
+      "grad_norm": 0.2133955681118407,
+      "learning_rate": 2.9018932874354566e-05,
+      "loss": 0.355,
+      "step": 1543
+    },
+    {
+      "epoch": 1.4338132837900603,
+      "grad_norm": 0.21653616877161203,
+      "learning_rate": 2.9001721170395868e-05,
+      "loss": 0.3637,
+      "step": 1544
+    },
+    {
+      "epoch": 1.4347422201579192,
+      "grad_norm": 0.2241202614347583,
+      "learning_rate": 2.8984509466437176e-05,
+      "loss": 0.3392,
+      "step": 1545
+    },
+    {
+      "epoch": 1.435671156525778,
+      "grad_norm": 0.2024615955265462,
+      "learning_rate": 2.8967297762478484e-05,
+      "loss": 0.3543,
+      "step": 1546
+    },
+    {
+      "epoch": 1.4366000928936367,
+      "grad_norm": 0.1955822427239985,
+      "learning_rate": 2.8950086058519793e-05,
+      "loss": 0.3364,
+      "step": 1547
+    },
+    {
+      "epoch": 1.4375290292614955,
+      "grad_norm": 0.20873100814874504,
+      "learning_rate": 2.89328743545611e-05,
+      "loss": 0.3517,
+      "step": 1548
+    },
+    {
+      "epoch": 1.4384579656293544,
+      "grad_norm": 0.2112000677453288,
+      "learning_rate": 2.891566265060241e-05,
+      "loss": 0.3469,
+      "step": 1549
+    },
+    {
+      "epoch": 1.4393869019972132,
+      "grad_norm": 0.20592979867590985,
+      "learning_rate": 2.8898450946643718e-05,
+      "loss": 0.3601,
+      "step": 1550
+    },
+    {
+      "epoch": 1.440315838365072,
+      "grad_norm": 0.21192866699319832,
+      "learning_rate": 2.8881239242685026e-05,
+      "loss": 0.3469,
+      "step": 1551
+    },
+    {
+      "epoch": 1.4412447747329309,
+      "grad_norm": 0.18398546591558737,
+      "learning_rate": 2.8864027538726335e-05,
+      "loss": 0.3441,
+      "step": 1552
+    },
+    {
+      "epoch": 1.4421737111007895,
+      "grad_norm": 0.2017578725992054,
+      "learning_rate": 2.8846815834767643e-05,
+      "loss": 0.3617,
+      "step": 1553
+    },
+    {
+      "epoch": 1.4431026474686484,
+      "grad_norm": 0.20148355147165561,
+      "learning_rate": 2.882960413080895e-05,
+      "loss": 0.3574,
+      "step": 1554
+    },
+    {
+      "epoch": 1.4440315838365072,
+      "grad_norm": 0.18503169901772,
+      "learning_rate": 2.881239242685026e-05,
+      "loss": 0.3602,
+      "step": 1555
+    },
+    {
+      "epoch": 1.444960520204366,
+      "grad_norm": 0.19444893478794364,
+      "learning_rate": 2.879518072289157e-05,
+      "loss": 0.3423,
+      "step": 1556
+    },
+    {
+      "epoch": 1.4458894565722247,
+      "grad_norm": 0.17839868562302205,
+      "learning_rate": 2.8777969018932877e-05,
+      "loss": 0.3228,
+      "step": 1557
+    },
+    {
+      "epoch": 1.4468183929400835,
+      "grad_norm": 0.17929509233834304,
+      "learning_rate": 2.8760757314974185e-05,
+      "loss": 0.3096,
+      "step": 1558
+    },
+    {
+      "epoch": 1.4477473293079424,
+      "grad_norm": 0.18214616689218266,
+      "learning_rate": 2.8743545611015494e-05,
+      "loss": 0.3522,
+      "step": 1559
+    },
+    {
+      "epoch": 1.4486762656758012,
+      "grad_norm": 0.2012672967397137,
+      "learning_rate": 2.8726333907056802e-05,
+      "loss": 0.358,
+      "step": 1560
+    },
+    {
+      "epoch": 1.44960520204366,
+      "grad_norm": 0.19930879626694126,
+      "learning_rate": 2.870912220309811e-05,
+      "loss": 0.3431,
+      "step": 1561
+    },
+    {
+      "epoch": 1.450534138411519,
+      "grad_norm": 0.17903764713191608,
+      "learning_rate": 2.869191049913942e-05,
+      "loss": 0.3396,
+      "step": 1562
+    },
+    {
+      "epoch": 1.4514630747793777,
+      "grad_norm": 0.19529230647348247,
+      "learning_rate": 2.8674698795180727e-05,
+      "loss": 0.3592,
+      "step": 1563
+    },
+    {
+      "epoch": 1.4523920111472364,
+      "grad_norm": 0.19975868842982136,
+      "learning_rate": 2.8657487091222036e-05,
+      "loss": 0.355,
+      "step": 1564
+    },
+    {
+      "epoch": 1.4533209475150952,
+      "grad_norm": 0.20358611197286738,
+      "learning_rate": 2.8640275387263337e-05,
+      "loss": 0.3527,
+      "step": 1565
+    },
+    {
+      "epoch": 1.454249883882954,
+      "grad_norm": 0.19856276668969844,
+      "learning_rate": 2.8623063683304646e-05,
+      "loss": 0.3266,
+      "step": 1566
+    },
+    {
+      "epoch": 1.4551788202508127,
+      "grad_norm": 0.19124828137152422,
+      "learning_rate": 2.8605851979345954e-05,
+      "loss": 0.3324,
+      "step": 1567
+    },
+    {
+      "epoch": 1.4561077566186715,
+      "grad_norm": 0.20725907417756995,
+      "learning_rate": 2.8588640275387262e-05,
+      "loss": 0.3357,
+      "step": 1568
+    },
+    {
+      "epoch": 1.4570366929865304,
+      "grad_norm": 0.2176938171580828,
+      "learning_rate": 2.857142857142857e-05,
+      "loss": 0.3435,
+      "step": 1569
+    },
+    {
+      "epoch": 1.4579656293543892,
+      "grad_norm": 0.21108437489228465,
+      "learning_rate": 2.855421686746988e-05,
+      "loss": 0.3466,
+      "step": 1570
+    },
+    {
+      "epoch": 1.458894565722248,
+      "grad_norm": 0.22480057073204013,
+      "learning_rate": 2.8537005163511188e-05,
+      "loss": 0.3606,
+      "step": 1571
+    },
+    {
+      "epoch": 1.459823502090107,
+      "grad_norm": 0.20864206196181312,
+      "learning_rate": 2.8519793459552496e-05,
+      "loss": 0.339,
+      "step": 1572
+    },
+    {
+      "epoch": 1.4607524384579658,
+      "grad_norm": 0.21922099540081025,
+      "learning_rate": 2.8502581755593804e-05,
+      "loss": 0.3444,
+      "step": 1573
+    },
+    {
+      "epoch": 1.4616813748258244,
+      "grad_norm": 0.2276494800159974,
+      "learning_rate": 2.8485370051635113e-05,
+      "loss": 0.3599,
+      "step": 1574
+    },
+    {
+      "epoch": 1.4626103111936832,
+      "grad_norm": 0.2288239429440535,
+      "learning_rate": 2.846815834767642e-05,
+      "loss": 0.3491,
+      "step": 1575
+    },
+    {
+      "epoch": 1.463539247561542,
+      "grad_norm": 0.17841881990882816,
+      "learning_rate": 2.845094664371773e-05,
+      "loss": 0.343,
+      "step": 1576
+    },
+    {
+      "epoch": 1.4644681839294007,
+      "grad_norm": 0.21286928712739886,
+      "learning_rate": 2.8433734939759038e-05,
+      "loss": 0.3521,
+      "step": 1577
+    },
+    {
+      "epoch": 1.4653971202972595,
+      "grad_norm": 0.21833463235565898,
+      "learning_rate": 2.8416523235800346e-05,
+      "loss": 0.3438,
+      "step": 1578
+    },
+    {
+      "epoch": 1.4663260566651184,
+      "grad_norm": 0.18733866679034886,
+      "learning_rate": 2.8399311531841655e-05,
+      "loss": 0.3382,
+      "step": 1579
+    },
+    {
+      "epoch": 1.4672549930329772,
+      "grad_norm": 0.20885255362383456,
+      "learning_rate": 2.8382099827882963e-05,
+      "loss": 0.3352,
+      "step": 1580
+    },
+    {
+      "epoch": 1.468183929400836,
+      "grad_norm": 0.2110857036563971,
+      "learning_rate": 2.836488812392427e-05,
+      "loss": 0.3451,
+      "step": 1581
+    },
+    {
+      "epoch": 1.469112865768695,
+      "grad_norm": 0.21499377814488893,
+      "learning_rate": 2.834767641996558e-05,
+      "loss": 0.344,
+      "step": 1582
+    },
+    {
+      "epoch": 1.4700418021365538,
+      "grad_norm": 0.2241849573695223,
+      "learning_rate": 2.833046471600689e-05,
+      "loss": 0.3292,
+      "step": 1583
+    },
+    {
+      "epoch": 1.4709707385044124,
+      "grad_norm": 0.21699607258286505,
+      "learning_rate": 2.8313253012048197e-05,
+      "loss": 0.334,
+      "step": 1584
+    },
+    {
+      "epoch": 1.4718996748722712,
+      "grad_norm": 0.22542594584645712,
+      "learning_rate": 2.8296041308089505e-05,
+      "loss": 0.3477,
+      "step": 1585
+    },
+    {
+      "epoch": 1.47282861124013,
+      "grad_norm": 0.1970265391341249,
+      "learning_rate": 2.8278829604130807e-05,
+      "loss": 0.3338,
+      "step": 1586
+    },
+    {
+      "epoch": 1.4737575476079887,
+      "grad_norm": 0.21144181981047877,
+      "learning_rate": 2.8261617900172115e-05,
+      "loss": 0.3657,
+      "step": 1587
+    },
+    {
+      "epoch": 1.4746864839758476,
+      "grad_norm": 0.22030587436035384,
+      "learning_rate": 2.8244406196213424e-05,
+      "loss": 0.3263,
+      "step": 1588
+    },
+    {
+      "epoch": 1.4756154203437064,
+      "grad_norm": 0.21592987334276875,
+      "learning_rate": 2.8227194492254732e-05,
+      "loss": 0.3691,
+      "step": 1589
+    },
+    {
+      "epoch": 1.4765443567115653,
+      "grad_norm": 0.20142452673876848,
+      "learning_rate": 2.820998278829604e-05,
+      "loss": 0.3485,
+      "step": 1590
+    },
+    {
+      "epoch": 1.477473293079424,
+      "grad_norm": 0.21246785539377971,
+      "learning_rate": 2.819277108433735e-05,
+      "loss": 0.3379,
+      "step": 1591
+    },
+    {
+      "epoch": 1.478402229447283,
+      "grad_norm": 0.20419932770516422,
+      "learning_rate": 2.8175559380378657e-05,
+      "loss": 0.3664,
+      "step": 1592
+    },
+    {
+      "epoch": 1.4793311658151418,
+      "grad_norm": 0.23868328892673205,
+      "learning_rate": 2.8158347676419966e-05,
+      "loss": 0.3579,
+      "step": 1593
+    },
+    {
+      "epoch": 1.4802601021830004,
+      "grad_norm": 0.18142148540515865,
+      "learning_rate": 2.8141135972461274e-05,
+      "loss": 0.3293,
+      "step": 1594
+    },
+    {
+      "epoch": 1.4811890385508593,
+      "grad_norm": 0.190821964829027,
+      "learning_rate": 2.8123924268502582e-05,
+      "loss": 0.3336,
+      "step": 1595
+    },
+    {
+      "epoch": 1.482117974918718,
+      "grad_norm": 0.1990392036313888,
+      "learning_rate": 2.810671256454389e-05,
+      "loss": 0.3639,
+      "step": 1596
+    },
+    {
+      "epoch": 1.483046911286577,
+      "grad_norm": 0.20621738698691292,
+      "learning_rate": 2.80895008605852e-05,
+      "loss": 0.3524,
+      "step": 1597
+    },
+    {
+      "epoch": 1.4839758476544356,
+      "grad_norm": 0.21378160126323356,
+      "learning_rate": 2.8072289156626508e-05,
+      "loss": 0.3507,
+      "step": 1598
+    },
+    {
+      "epoch": 1.4849047840222944,
+      "grad_norm": 0.2159950509848571,
+      "learning_rate": 2.8055077452667816e-05,
+      "loss": 0.3549,
+      "step": 1599
+    },
+    {
+      "epoch": 1.4858337203901533,
+      "grad_norm": 0.18702549069355595,
+      "learning_rate": 2.8037865748709124e-05,
+      "loss": 0.3366,
+      "step": 1600
+    },
+    {
+      "epoch": 1.4867626567580121,
+      "grad_norm": 0.1994967299413536,
+      "learning_rate": 2.8020654044750433e-05,
+      "loss": 0.3256,
+      "step": 1601
+    },
+    {
+      "epoch": 1.487691593125871,
+      "grad_norm": 0.21942189823000416,
+      "learning_rate": 2.800344234079174e-05,
+      "loss": 0.3581,
+      "step": 1602
+    },
+    {
+      "epoch": 1.4886205294937298,
+      "grad_norm": 0.1766879531892212,
+      "learning_rate": 2.798623063683305e-05,
+      "loss": 0.3457,
+      "step": 1603
+    },
+    {
+      "epoch": 1.4895494658615884,
+      "grad_norm": 0.19258517625663005,
+      "learning_rate": 2.7969018932874358e-05,
+      "loss": 0.3425,
+      "step": 1604
+    },
+    {
+      "epoch": 1.4904784022294473,
+      "grad_norm": 0.2101457920763198,
+      "learning_rate": 2.7951807228915666e-05,
+      "loss": 0.3442,
+      "step": 1605
+    },
+    {
+      "epoch": 1.4914073385973061,
+      "grad_norm": 0.17432791499600817,
+      "learning_rate": 2.7934595524956975e-05,
+      "loss": 0.3537,
+      "step": 1606
+    },
+    {
+      "epoch": 1.492336274965165,
+      "grad_norm": 0.21787927461103915,
+      "learning_rate": 2.7917383820998277e-05,
+      "loss": 0.3375,
+      "step": 1607
+    },
+    {
+      "epoch": 1.4932652113330236,
+      "grad_norm": 0.2135791752018399,
+      "learning_rate": 2.7900172117039585e-05,
+      "loss": 0.3642,
+      "step": 1608
+    },
+    {
+      "epoch": 1.4941941477008824,
+      "grad_norm": 0.19691607764966712,
+      "learning_rate": 2.7882960413080893e-05,
+      "loss": 0.3492,
+      "step": 1609
+    },
+    {
+      "epoch": 1.4951230840687413,
+      "grad_norm": 0.19492765300345521,
+      "learning_rate": 2.7865748709122202e-05,
+      "loss": 0.3657,
+      "step": 1610
+    },
+    {
+      "epoch": 1.4960520204366001,
+      "grad_norm": 0.20105927888888867,
+      "learning_rate": 2.784853700516351e-05,
+      "loss": 0.3522,
+      "step": 1611
+    },
+    {
+      "epoch": 1.496980956804459,
+      "grad_norm": 0.2022257310559514,
+      "learning_rate": 2.783132530120482e-05,
+      "loss": 0.3352,
+      "step": 1612
+    },
+    {
+      "epoch": 1.4979098931723178,
+      "grad_norm": 0.18412786976848391,
+      "learning_rate": 2.7814113597246127e-05,
+      "loss": 0.3569,
+      "step": 1613
+    },
+    {
+      "epoch": 1.4988388295401764,
+      "grad_norm": 0.18521152850671202,
+      "learning_rate": 2.7796901893287435e-05,
+      "loss": 0.3291,
+      "step": 1614
+    },
+    {
+      "epoch": 1.4997677659080353,
+      "grad_norm": 0.18484257371830262,
+      "learning_rate": 2.7779690189328744e-05,
+      "loss": 0.3471,
+      "step": 1615
+    },
+    {
+      "epoch": 1.5006967022758941,
+      "grad_norm": 0.20613823228990855,
+      "learning_rate": 2.7762478485370052e-05,
+      "loss": 0.3523,
+      "step": 1616
+    },
+    {
+      "epoch": 1.5016256386437528,
+      "grad_norm": 0.21700406423639299,
+      "learning_rate": 2.774526678141136e-05,
+      "loss": 0.3402,
+      "step": 1617
+    },
+    {
+      "epoch": 1.5025545750116116,
+      "grad_norm": 0.1825892710960065,
+      "learning_rate": 2.772805507745267e-05,
+      "loss": 0.3463,
+      "step": 1618
+    },
+    {
+      "epoch": 1.5034835113794704,
+      "grad_norm": 0.21767334282347994,
+      "learning_rate": 2.7710843373493977e-05,
+      "loss": 0.3591,
+      "step": 1619
+    },
+    {
+      "epoch": 1.5044124477473293,
+      "grad_norm": 0.2240701912543056,
+      "learning_rate": 2.7693631669535286e-05,
+      "loss": 0.3491,
+      "step": 1620
+    },
+    {
+      "epoch": 1.5053413841151881,
+      "grad_norm": 0.20794265732288209,
+      "learning_rate": 2.7676419965576594e-05,
+      "loss": 0.3597,
+      "step": 1621
+    },
+    {
+      "epoch": 1.506270320483047,
+      "grad_norm": 0.1942386308673407,
+      "learning_rate": 2.7659208261617903e-05,
+      "loss": 0.3553,
+      "step": 1622
+    },
+    {
+      "epoch": 1.5071992568509058,
+      "grad_norm": 0.23953241896174912,
+      "learning_rate": 2.764199655765921e-05,
+      "loss": 0.3353,
+      "step": 1623
+    },
+    {
+      "epoch": 1.5081281932187647,
+      "grad_norm": 0.19064283645632496,
+      "learning_rate": 2.762478485370052e-05,
+      "loss": 0.3309,
+      "step": 1624
+    },
+    {
+      "epoch": 1.5090571295866233,
+      "grad_norm": 0.19455048912942696,
+      "learning_rate": 2.7607573149741828e-05,
+      "loss": 0.3512,
+      "step": 1625
+    },
+    {
+      "epoch": 1.5099860659544821,
+      "grad_norm": 0.24734707148619361,
+      "learning_rate": 2.7590361445783136e-05,
+      "loss": 0.3369,
+      "step": 1626
+    },
+    {
+      "epoch": 1.5109150023223408,
+      "grad_norm": 0.19596045941652765,
+      "learning_rate": 2.7573149741824445e-05,
+      "loss": 0.3447,
+      "step": 1627
+    },
+    {
+      "epoch": 1.5118439386901996,
+      "grad_norm": 0.20593155094983895,
+      "learning_rate": 2.7555938037865753e-05,
+      "loss": 0.3281,
+      "step": 1628
+    },
+    {
+      "epoch": 1.5127728750580585,
+      "grad_norm": 0.18985986735657961,
+      "learning_rate": 2.7538726333907055e-05,
+      "loss": 0.3444,
+      "step": 1629
+    },
+    {
+      "epoch": 1.5137018114259173,
+      "grad_norm": 0.20048398993083572,
+      "learning_rate": 2.7521514629948363e-05,
+      "loss": 0.3352,
+      "step": 1630
+    },
+    {
+      "epoch": 1.5146307477937762,
+      "grad_norm": 0.2340324824637472,
+      "learning_rate": 2.750430292598967e-05,
+      "loss": 0.3547,
+      "step": 1631
+    },
+    {
+      "epoch": 1.515559684161635,
+      "grad_norm": 0.1979649285811828,
+      "learning_rate": 2.748709122203098e-05,
+      "loss": 0.3282,
+      "step": 1632
+    },
+    {
+      "epoch": 1.5164886205294938,
+      "grad_norm": 0.22728561652019158,
+      "learning_rate": 2.7469879518072288e-05,
+      "loss": 0.3433,
+      "step": 1633
+    },
+    {
+      "epoch": 1.5174175568973527,
+      "grad_norm": 0.1909605625979193,
+      "learning_rate": 2.7452667814113597e-05,
+      "loss": 0.3481,
+      "step": 1634
+    },
+    {
+      "epoch": 1.5183464932652113,
+      "grad_norm": 0.21650749973633887,
+      "learning_rate": 2.7435456110154905e-05,
+      "loss": 0.3316,
+      "step": 1635
+    },
+    {
+      "epoch": 1.5192754296330702,
+      "grad_norm": 0.20766956553484378,
+      "learning_rate": 2.7418244406196213e-05,
+      "loss": 0.3553,
+      "step": 1636
+    },
+    {
+      "epoch": 1.5202043660009288,
+      "grad_norm": 0.19184440178799098,
+      "learning_rate": 2.7401032702237522e-05,
+      "loss": 0.3479,
+      "step": 1637
+    },
+    {
+      "epoch": 1.5211333023687876,
+      "grad_norm": 0.19041440103788876,
+      "learning_rate": 2.738382099827883e-05,
+      "loss": 0.3229,
+      "step": 1638
+    },
+    {
+      "epoch": 1.5220622387366465,
+      "grad_norm": 0.19777552122771166,
+      "learning_rate": 2.736660929432014e-05,
+      "loss": 0.3535,
+      "step": 1639
+    },
+    {
+      "epoch": 1.5229911751045053,
+      "grad_norm": 0.2004208302269382,
+      "learning_rate": 2.7349397590361447e-05,
+      "loss": 0.34,
+      "step": 1640
+    },
+    {
+      "epoch": 1.5239201114723642,
+      "grad_norm": 0.20336163065057344,
+      "learning_rate": 2.7332185886402755e-05,
+      "loss": 0.3521,
+      "step": 1641
+    },
+    {
+      "epoch": 1.524849047840223,
+      "grad_norm": 0.19703431204687524,
+      "learning_rate": 2.7314974182444064e-05,
+      "loss": 0.3414,
+      "step": 1642
+    },
+    {
+      "epoch": 1.5257779842080819,
+      "grad_norm": 0.19845951237979154,
+      "learning_rate": 2.7297762478485372e-05,
+      "loss": 0.3677,
+      "step": 1643
+    },
+    {
+      "epoch": 1.5267069205759407,
+      "grad_norm": 0.19034132547856192,
+      "learning_rate": 2.728055077452668e-05,
+      "loss": 0.3406,
+      "step": 1644
+    },
+    {
+      "epoch": 1.5276358569437993,
+      "grad_norm": 0.18091025641587147,
+      "learning_rate": 2.726333907056799e-05,
+      "loss": 0.3313,
+      "step": 1645
+    },
+    {
+      "epoch": 1.5285647933116582,
+      "grad_norm": 0.18588920730588274,
+      "learning_rate": 2.7246127366609297e-05,
+      "loss": 0.3379,
+      "step": 1646
+    },
+    {
+      "epoch": 1.5294937296795168,
+      "grad_norm": 0.183194983638107,
+      "learning_rate": 2.7228915662650606e-05,
+      "loss": 0.3514,
+      "step": 1647
+    },
+    {
+      "epoch": 1.5304226660473756,
+      "grad_norm": 0.20313603959857246,
+      "learning_rate": 2.7211703958691914e-05,
+      "loss": 0.3148,
+      "step": 1648
+    },
+    {
+      "epoch": 1.5313516024152345,
+      "grad_norm": 0.18413859078792266,
+      "learning_rate": 2.7194492254733223e-05,
+      "loss": 0.3406,
+      "step": 1649
+    },
+    {
+      "epoch": 1.5322805387830933,
+      "grad_norm": 0.2218660778677224,
+      "learning_rate": 2.7177280550774524e-05,
+      "loss": 0.3473,
+      "step": 1650
+    },
+    {
+      "epoch": 1.5332094751509522,
+      "grad_norm": 0.20256666218542452,
+      "learning_rate": 2.7160068846815833e-05,
+      "loss": 0.327,
+      "step": 1651
+    },
+    {
+      "epoch": 1.534138411518811,
+      "grad_norm": 0.20973362475741678,
+      "learning_rate": 2.714285714285714e-05,
+      "loss": 0.3382,
+      "step": 1652
+    },
+    {
+      "epoch": 1.5350673478866699,
+      "grad_norm": 0.21420987345417866,
+      "learning_rate": 2.712564543889845e-05,
+      "loss": 0.3549,
+      "step": 1653
+    },
+    {
+      "epoch": 1.5359962842545287,
+      "grad_norm": 0.20547490967457238,
+      "learning_rate": 2.7108433734939758e-05,
+      "loss": 0.3692,
+      "step": 1654
+    },
+    {
+      "epoch": 1.5369252206223873,
+      "grad_norm": 0.20579644597140412,
+      "learning_rate": 2.7091222030981066e-05,
+      "loss": 0.3603,
+      "step": 1655
+    },
+    {
+      "epoch": 1.5378541569902462,
+      "grad_norm": 0.17767519418921987,
+      "learning_rate": 2.7074010327022375e-05,
+      "loss": 0.3207,
+      "step": 1656
+    },
+    {
+      "epoch": 1.5387830933581048,
+      "grad_norm": 0.21067954281906232,
+      "learning_rate": 2.7056798623063683e-05,
+      "loss": 0.3668,
+      "step": 1657
+    },
+    {
+      "epoch": 1.5397120297259637,
+      "grad_norm": 0.1962228244666741,
+      "learning_rate": 2.703958691910499e-05,
+      "loss": 0.3492,
+      "step": 1658
+    },
+    {
+      "epoch": 1.5406409660938225,
+      "grad_norm": 0.19557765329976629,
+      "learning_rate": 2.70223752151463e-05,
+      "loss": 0.3474,
+      "step": 1659
+    },
+    {
+      "epoch": 1.5415699024616814,
+      "grad_norm": 0.18500343001861966,
+      "learning_rate": 2.7005163511187608e-05,
+      "loss": 0.3416,
+      "step": 1660
+    },
+    {
+      "epoch": 1.5424988388295402,
+      "grad_norm": 0.2023975455255552,
+      "learning_rate": 2.6987951807228917e-05,
+      "loss": 0.3406,
+      "step": 1661
+    },
+    {
+      "epoch": 1.543427775197399,
+      "grad_norm": 0.19197373382978503,
+      "learning_rate": 2.6970740103270225e-05,
+      "loss": 0.3494,
+      "step": 1662
+    },
+    {
+      "epoch": 1.544356711565258,
+      "grad_norm": 0.200189895799683,
+      "learning_rate": 2.6953528399311533e-05,
+      "loss": 0.3437,
+      "step": 1663
+    },
+    {
+      "epoch": 1.5452856479331167,
+      "grad_norm": 0.2224630313261313,
+      "learning_rate": 2.6936316695352842e-05,
+      "loss": 0.362,
+      "step": 1664
+    },
+    {
+      "epoch": 1.5462145843009754,
+      "grad_norm": 0.22940474152039902,
+      "learning_rate": 2.691910499139415e-05,
+      "loss": 0.3469,
+      "step": 1665
+    },
+    {
+      "epoch": 1.5471435206688342,
+      "grad_norm": 0.2206713442498346,
+      "learning_rate": 2.690189328743546e-05,
+      "loss": 0.3565,
+      "step": 1666
+    },
+    {
+      "epoch": 1.5480724570366928,
+      "grad_norm": 0.19182785647681846,
+      "learning_rate": 2.6884681583476767e-05,
+      "loss": 0.3279,
+      "step": 1667
+    },
+    {
+      "epoch": 1.5490013934045517,
+      "grad_norm": 0.23133211363227132,
+      "learning_rate": 2.6867469879518075e-05,
+      "loss": 0.3571,
+      "step": 1668
+    },
+    {
+      "epoch": 1.5499303297724105,
+      "grad_norm": 0.19494744045107942,
+      "learning_rate": 2.6850258175559384e-05,
+      "loss": 0.325,
+      "step": 1669
+    },
+    {
+      "epoch": 1.5508592661402694,
+      "grad_norm": 0.19903204750297201,
+      "learning_rate": 2.6833046471600692e-05,
+      "loss": 0.3385,
+      "step": 1670
+    },
+    {
+      "epoch": 1.5517882025081282,
+      "grad_norm": 0.1941975771860589,
+      "learning_rate": 2.6815834767641994e-05,
+      "loss": 0.3472,
+      "step": 1671
+    },
+    {
+      "epoch": 1.552717138875987,
+      "grad_norm": 0.20247277746107317,
+      "learning_rate": 2.6798623063683302e-05,
+      "loss": 0.3473,
+      "step": 1672
+    },
+    {
+      "epoch": 1.553646075243846,
+      "grad_norm": 0.21609956845102268,
+      "learning_rate": 2.678141135972461e-05,
+      "loss": 0.3259,
+      "step": 1673
+    },
+    {
+      "epoch": 1.5545750116117047,
+      "grad_norm": 0.19593104013837537,
+      "learning_rate": 2.676419965576592e-05,
+      "loss": 0.354,
+      "step": 1674
+    },
+    {
+      "epoch": 1.5555039479795634,
+      "grad_norm": 0.18796258655354342,
+      "learning_rate": 2.6746987951807227e-05,
+      "loss": 0.3506,
+      "step": 1675
+    },
+    {
+      "epoch": 1.5564328843474222,
+      "grad_norm": 0.2059985043490576,
+      "learning_rate": 2.6729776247848536e-05,
+      "loss": 0.3533,
+      "step": 1676
+    },
+    {
+      "epoch": 1.5573618207152808,
+      "grad_norm": 0.2268856053908675,
+      "learning_rate": 2.6712564543889844e-05,
+      "loss": 0.3617,
+      "step": 1677
+    },
+    {
+      "epoch": 1.5582907570831397,
+      "grad_norm": 0.19081957209925454,
+      "learning_rate": 2.6695352839931153e-05,
+      "loss": 0.3425,
+      "step": 1678
+    },
+    {
+      "epoch": 1.5592196934509985,
+      "grad_norm": 0.21370626444738328,
+      "learning_rate": 2.667814113597246e-05,
+      "loss": 0.3435,
+      "step": 1679
+    },
+    {
+      "epoch": 1.5601486298188574,
+      "grad_norm": 0.20241828803392062,
+      "learning_rate": 2.666092943201377e-05,
+      "loss": 0.3471,
+      "step": 1680
+    },
+    {
+      "epoch": 1.5610775661867162,
+      "grad_norm": 0.2126563580465179,
+      "learning_rate": 2.6643717728055078e-05,
+      "loss": 0.3345,
+      "step": 1681
+    },
+    {
+      "epoch": 1.562006502554575,
+      "grad_norm": 0.19313249003359623,
+      "learning_rate": 2.6626506024096386e-05,
+      "loss": 0.3367,
+      "step": 1682
+    },
+    {
+      "epoch": 1.562935438922434,
+      "grad_norm": 0.20080971884495982,
+      "learning_rate": 2.6609294320137695e-05,
+      "loss": 0.3512,
+      "step": 1683
+    },
+    {
+      "epoch": 1.5638643752902928,
+      "grad_norm": 0.2326488096574912,
+      "learning_rate": 2.6592082616179003e-05,
+      "loss": 0.3507,
+      "step": 1684
+    },
+    {
+      "epoch": 1.5647933116581514,
+      "grad_norm": 0.20822523814151891,
+      "learning_rate": 2.657487091222031e-05,
+      "loss": 0.3483,
+      "step": 1685
+    },
+    {
+      "epoch": 1.5657222480260102,
+      "grad_norm": 0.21602169294812015,
+      "learning_rate": 2.655765920826162e-05,
+      "loss": 0.3451,
+      "step": 1686
+    },
+    {
+      "epoch": 1.5666511843938689,
+      "grad_norm": 0.2026495811587196,
+      "learning_rate": 2.6540447504302928e-05,
+      "loss": 0.3401,
+      "step": 1687
+    },
+    {
+      "epoch": 1.5675801207617277,
+      "grad_norm": 0.20373496856303852,
+      "learning_rate": 2.6523235800344237e-05,
+      "loss": 0.3472,
+      "step": 1688
+    },
+    {
+      "epoch": 1.5685090571295865,
+      "grad_norm": 0.202815214563734,
+      "learning_rate": 2.6506024096385545e-05,
+      "loss": 0.3443,
+      "step": 1689
+    },
+    {
+      "epoch": 1.5694379934974454,
+      "grad_norm": 0.20807192948042533,
+      "learning_rate": 2.6488812392426853e-05,
+      "loss": 0.3244,
+      "step": 1690
+    },
+    {
+      "epoch": 1.5703669298653042,
+      "grad_norm": 0.19319517687258816,
+      "learning_rate": 2.6471600688468162e-05,
+      "loss": 0.3382,
+      "step": 1691
+    },
+    {
+      "epoch": 1.571295866233163,
+      "grad_norm": 0.22053885061161396,
+      "learning_rate": 2.6454388984509463e-05,
+      "loss": 0.3666,
+      "step": 1692
+    },
+    {
+      "epoch": 1.572224802601022,
+      "grad_norm": 0.207890820090718,
+      "learning_rate": 2.6437177280550772e-05,
+      "loss": 0.3436,
+      "step": 1693
+    },
+    {
+      "epoch": 1.5731537389688808,
+      "grad_norm": 0.19245242691550665,
+      "learning_rate": 2.641996557659208e-05,
+      "loss": 0.3539,
+      "step": 1694
+    },
+    {
+      "epoch": 1.5740826753367394,
+      "grad_norm": 0.21843726413189252,
+      "learning_rate": 2.640275387263339e-05,
+      "loss": 0.3421,
+      "step": 1695
+    },
+    {
+      "epoch": 1.5750116117045982,
+      "grad_norm": 0.220646507516627,
+      "learning_rate": 2.6385542168674697e-05,
+      "loss": 0.3541,
+      "step": 1696
+    },
+    {
+      "epoch": 1.5759405480724569,
+      "grad_norm": 0.22537203693214305,
+      "learning_rate": 2.6368330464716005e-05,
+      "loss": 0.3277,
+      "step": 1697
+    },
+    {
+      "epoch": 1.5768694844403157,
+      "grad_norm": 0.20461650445631827,
+      "learning_rate": 2.6351118760757314e-05,
+      "loss": 0.3387,
+      "step": 1698
+    },
+    {
+      "epoch": 1.5777984208081746,
+      "grad_norm": 0.2136483563667655,
+      "learning_rate": 2.6333907056798622e-05,
+      "loss": 0.3426,
+      "step": 1699
+    },
+    {
+      "epoch": 1.5787273571760334,
+      "grad_norm": 0.19981232379812758,
+      "learning_rate": 2.631669535283993e-05,
+      "loss": 0.3454,
+      "step": 1700
+    },
+    {
+      "epoch": 1.5796562935438923,
+      "grad_norm": 0.23165196856655018,
+      "learning_rate": 2.629948364888124e-05,
+      "loss": 0.3775,
+      "step": 1701
+    },
+    {
+      "epoch": 1.580585229911751,
+      "grad_norm": 0.2055708819872716,
+      "learning_rate": 2.6282271944922547e-05,
+      "loss": 0.3398,
+      "step": 1702
+    },
+    {
+      "epoch": 1.58151416627961,
+      "grad_norm": 0.20549199744598382,
+      "learning_rate": 2.6265060240963856e-05,
+      "loss": 0.3303,
+      "step": 1703
+    },
+    {
+      "epoch": 1.5824431026474688,
+      "grad_norm": 0.19735611925338659,
+      "learning_rate": 2.6247848537005164e-05,
+      "loss": 0.328,
+      "step": 1704
+    },
+    {
+      "epoch": 1.5833720390153274,
+      "grad_norm": 0.18310842492165016,
+      "learning_rate": 2.6230636833046473e-05,
+      "loss": 0.343,
+      "step": 1705
+    },
+    {
+      "epoch": 1.5843009753831863,
+      "grad_norm": 0.1873807736634079,
+      "learning_rate": 2.621342512908778e-05,
+      "loss": 0.3481,
+      "step": 1706
+    },
+    {
+      "epoch": 1.585229911751045,
+      "grad_norm": 0.20000318958063903,
+      "learning_rate": 2.619621342512909e-05,
+      "loss": 0.3356,
+      "step": 1707
+    },
+    {
+      "epoch": 1.5861588481189037,
+      "grad_norm": 0.2015136507544033,
+      "learning_rate": 2.6179001721170398e-05,
+      "loss": 0.3324,
+      "step": 1708
+    },
+    {
+      "epoch": 1.5870877844867626,
+      "grad_norm": 0.18262820920953574,
+      "learning_rate": 2.6161790017211706e-05,
+      "loss": 0.3532,
+      "step": 1709
+    },
+    {
+      "epoch": 1.5880167208546214,
+      "grad_norm": 0.23344245143312825,
+      "learning_rate": 2.6144578313253015e-05,
+      "loss": 0.3775,
+      "step": 1710
+    },
+    {
+      "epoch": 1.5889456572224803,
+      "grad_norm": 0.2199511607804858,
+      "learning_rate": 2.6127366609294323e-05,
+      "loss": 0.3168,
+      "step": 1711
+    },
+    {
+      "epoch": 1.5898745935903391,
+      "grad_norm": 0.212983265606066,
+      "learning_rate": 2.611015490533563e-05,
+      "loss": 0.3437,
+      "step": 1712
+    },
+    {
+      "epoch": 1.590803529958198,
+      "grad_norm": 0.213128136916506,
+      "learning_rate": 2.6092943201376936e-05,
+      "loss": 0.3754,
+      "step": 1713
+    },
+    {
+      "epoch": 1.5917324663260568,
+      "grad_norm": 0.2448986605848652,
+      "learning_rate": 2.6075731497418245e-05,
+      "loss": 0.352,
+      "step": 1714
+    },
+    {
+      "epoch": 1.5926614026939154,
+      "grad_norm": 0.20674274942405582,
+      "learning_rate": 2.605851979345955e-05,
+      "loss": 0.3496,
+      "step": 1715
+    },
+    {
+      "epoch": 1.5935903390617743,
+      "grad_norm": 0.18776586917192614,
+      "learning_rate": 2.6041308089500858e-05,
+      "loss": 0.3322,
+      "step": 1716
+    },
+    {
+      "epoch": 1.5945192754296331,
+      "grad_norm": 0.20953790690241902,
+      "learning_rate": 2.6024096385542167e-05,
+      "loss": 0.331,
+      "step": 1717
+    },
+    {
+      "epoch": 1.5954482117974917,
+      "grad_norm": 0.20314003333327893,
+      "learning_rate": 2.6006884681583475e-05,
+      "loss": 0.3357,
+      "step": 1718
+    },
+    {
+      "epoch": 1.5963771481653506,
+      "grad_norm": 0.20922503716379712,
+      "learning_rate": 2.5989672977624783e-05,
+      "loss": 0.3572,
+      "step": 1719
+    },
+    {
+      "epoch": 1.5973060845332094,
+      "grad_norm": 0.20988177401915958,
+      "learning_rate": 2.5972461273666092e-05,
+      "loss": 0.352,
+      "step": 1720
+    },
+    {
+      "epoch": 1.5982350209010683,
+      "grad_norm": 0.21315542074548124,
+      "learning_rate": 2.59552495697074e-05,
+      "loss": 0.3626,
+      "step": 1721
+    },
+    {
+      "epoch": 1.5991639572689271,
+      "grad_norm": 0.20813871660567065,
+      "learning_rate": 2.593803786574871e-05,
+      "loss": 0.3474,
+      "step": 1722
+    },
+    {
+      "epoch": 1.600092893636786,
+      "grad_norm": 0.1872599355028666,
+      "learning_rate": 2.5920826161790017e-05,
+      "loss": 0.3445,
+      "step": 1723
+    },
+    {
+      "epoch": 1.6010218300046448,
+      "grad_norm": 0.1911940128780578,
+      "learning_rate": 2.5903614457831325e-05,
+      "loss": 0.3391,
+      "step": 1724
+    },
+    {
+      "epoch": 1.6019507663725034,
+      "grad_norm": 0.2092461298477511,
+      "learning_rate": 2.5886402753872634e-05,
+      "loss": 0.3539,
+      "step": 1725
+    },
+    {
+      "epoch": 1.6028797027403623,
+      "grad_norm": 0.20596276428098503,
+      "learning_rate": 2.5869191049913942e-05,
+      "loss": 0.3539,
+      "step": 1726
+    },
+    {
+      "epoch": 1.6038086391082211,
+      "grad_norm": 0.1932818475108648,
+      "learning_rate": 2.585197934595525e-05,
+      "loss": 0.3526,
+      "step": 1727
+    },
+    {
+      "epoch": 1.6047375754760798,
+      "grad_norm": 0.21283715898106823,
+      "learning_rate": 2.583476764199656e-05,
+      "loss": 0.3468,
+      "step": 1728
+    },
+    {
+      "epoch": 1.6056665118439386,
+      "grad_norm": 0.20794295159851667,
+      "learning_rate": 2.5817555938037867e-05,
+      "loss": 0.3496,
+      "step": 1729
+    },
+    {
+      "epoch": 1.6065954482117974,
+      "grad_norm": 0.20156095064197782,
+      "learning_rate": 2.5800344234079176e-05,
+      "loss": 0.3396,
+      "step": 1730
+    },
+    {
+      "epoch": 1.6075243845796563,
+      "grad_norm": 0.18701079703686946,
+      "learning_rate": 2.5783132530120484e-05,
+      "loss": 0.3491,
+      "step": 1731
+    },
+    {
+      "epoch": 1.6084533209475151,
+      "grad_norm": 0.19536223716914838,
+      "learning_rate": 2.5765920826161793e-05,
+      "loss": 0.3377,
+      "step": 1732
+    },
+    {
+      "epoch": 1.609382257315374,
+      "grad_norm": 0.17791846865411473,
+      "learning_rate": 2.57487091222031e-05,
+      "loss": 0.3472,
+      "step": 1733
+    },
+    {
+      "epoch": 1.6103111936832328,
+      "grad_norm": 0.2129004966083023,
+      "learning_rate": 2.573149741824441e-05,
+      "loss": 0.329,
+      "step": 1734
+    },
+    {
+      "epoch": 1.6112401300510915,
+      "grad_norm": 0.18814499058025813,
+      "learning_rate": 2.5714285714285714e-05,
+      "loss": 0.3429,
+      "step": 1735
+    },
+    {
+      "epoch": 1.6121690664189503,
+      "grad_norm": 0.1809854014670112,
+      "learning_rate": 2.5697074010327023e-05,
+      "loss": 0.3237,
+      "step": 1736
+    },
+    {
+      "epoch": 1.6130980027868091,
+      "grad_norm": 0.18538402717591057,
+      "learning_rate": 2.567986230636833e-05,
+      "loss": 0.3497,
+      "step": 1737
+    },
+    {
+      "epoch": 1.6140269391546678,
+      "grad_norm": 0.18559136167726656,
+      "learning_rate": 2.566265060240964e-05,
+      "loss": 0.354,
+      "step": 1738
+    },
+    {
+      "epoch": 1.6149558755225266,
+      "grad_norm": 0.20263301316144824,
+      "learning_rate": 2.5645438898450948e-05,
+      "loss": 0.3672,
+      "step": 1739
+    },
+    {
+      "epoch": 1.6158848118903855,
+      "grad_norm": 0.196423315030862,
+      "learning_rate": 2.5628227194492253e-05,
+      "loss": 0.3408,
+      "step": 1740
+    },
+    {
+      "epoch": 1.6168137482582443,
+      "grad_norm": 0.19368812454813836,
+      "learning_rate": 2.561101549053356e-05,
+      "loss": 0.348,
+      "step": 1741
+    },
+    {
+      "epoch": 1.6177426846261032,
+      "grad_norm": 0.19885581508283579,
+      "learning_rate": 2.559380378657487e-05,
+      "loss": 0.3371,
+      "step": 1742
+    },
+    {
+      "epoch": 1.618671620993962,
+      "grad_norm": 0.21769073543185363,
+      "learning_rate": 2.5576592082616178e-05,
+      "loss": 0.3652,
+      "step": 1743
+    },
+    {
+      "epoch": 1.6196005573618208,
+      "grad_norm": 0.215179619394168,
+      "learning_rate": 2.5559380378657487e-05,
+      "loss": 0.3412,
+      "step": 1744
+    },
+    {
+      "epoch": 1.6205294937296795,
+      "grad_norm": 0.1898709834286627,
+      "learning_rate": 2.5542168674698795e-05,
+      "loss": 0.3352,
+      "step": 1745
+    },
+    {
+      "epoch": 1.6214584300975383,
+      "grad_norm": 0.24343732302612947,
+      "learning_rate": 2.5524956970740103e-05,
+      "loss": 0.3271,
+      "step": 1746
+    },
+    {
+      "epoch": 1.6223873664653972,
+      "grad_norm": 0.19406338350847918,
+      "learning_rate": 2.5507745266781412e-05,
+      "loss": 0.3207,
+      "step": 1747
+    },
+    {
+      "epoch": 1.6233163028332558,
+      "grad_norm": 0.18606967526005735,
+      "learning_rate": 2.549053356282272e-05,
+      "loss": 0.3402,
+      "step": 1748
+    },
+    {
+      "epoch": 1.6242452392011146,
+      "grad_norm": 0.24944407718719433,
+      "learning_rate": 2.547332185886403e-05,
+      "loss": 0.3327,
+      "step": 1749
+    },
+    {
+      "epoch": 1.6251741755689735,
+      "grad_norm": 0.20574480952491447,
+      "learning_rate": 2.5456110154905337e-05,
+      "loss": 0.3353,
+      "step": 1750
+    },
+    {
+      "epoch": 1.6261031119368323,
+      "grad_norm": 0.1899167111957658,
+      "learning_rate": 2.5438898450946645e-05,
+      "loss": 0.3533,
+      "step": 1751
+    },
+    {
+      "epoch": 1.6270320483046912,
+      "grad_norm": 0.2355371100922719,
+      "learning_rate": 2.5421686746987954e-05,
+      "loss": 0.3442,
+      "step": 1752
+    },
+    {
+      "epoch": 1.62796098467255,
+      "grad_norm": 0.22707474270252473,
+      "learning_rate": 2.5404475043029262e-05,
+      "loss": 0.3485,
+      "step": 1753
+    },
+    {
+      "epoch": 1.6288899210404089,
+      "grad_norm": 0.23297359604796236,
+      "learning_rate": 2.538726333907057e-05,
+      "loss": 0.3492,
+      "step": 1754
+    },
+    {
+      "epoch": 1.6298188574082675,
+      "grad_norm": 0.18949532826143803,
+      "learning_rate": 2.537005163511188e-05,
+      "loss": 0.3479,
+      "step": 1755
+    },
+    {
+      "epoch": 1.6307477937761263,
+      "grad_norm": 0.18362249976581027,
+      "learning_rate": 2.5352839931153184e-05,
+      "loss": 0.3581,
+      "step": 1756
+    },
+    {
+      "epoch": 1.6316767301439852,
+      "grad_norm": 0.2554058616714679,
+      "learning_rate": 2.5335628227194492e-05,
+      "loss": 0.3546,
+      "step": 1757
+    },
+    {
+      "epoch": 1.6326056665118438,
+      "grad_norm": 0.22540032891893913,
+      "learning_rate": 2.53184165232358e-05,
+      "loss": 0.3619,
+      "step": 1758
+    },
+    {
+      "epoch": 1.6335346028797026,
+      "grad_norm": 0.20418915234656554,
+      "learning_rate": 2.530120481927711e-05,
+      "loss": 0.361,
+      "step": 1759
+    },
+    {
+      "epoch": 1.6344635392475615,
+      "grad_norm": 0.19977889622190362,
+      "learning_rate": 2.5283993115318418e-05,
+      "loss": 0.3377,
+      "step": 1760
+    },
+    {
+      "epoch": 1.6353924756154203,
+      "grad_norm": 0.22689586847036763,
+      "learning_rate": 2.5266781411359726e-05,
+      "loss": 0.3282,
+      "step": 1761
+    },
+    {
+      "epoch": 1.6363214119832792,
+      "grad_norm": 0.20137363650236248,
+      "learning_rate": 2.5249569707401034e-05,
+      "loss": 0.3444,
+      "step": 1762
+    },
+    {
+      "epoch": 1.637250348351138,
+      "grad_norm": 0.20785537438471768,
+      "learning_rate": 2.5232358003442343e-05,
+      "loss": 0.3476,
+      "step": 1763
+    },
+    {
+      "epoch": 1.6381792847189969,
+      "grad_norm": 0.21247019504395587,
+      "learning_rate": 2.521514629948365e-05,
+      "loss": 0.3579,
+      "step": 1764
+    },
+    {
+      "epoch": 1.6391082210868555,
+      "grad_norm": 0.20923349346429376,
+      "learning_rate": 2.5197934595524956e-05,
+      "loss": 0.3547,
+      "step": 1765
+    },
+    {
+      "epoch": 1.6400371574547143,
+      "grad_norm": 0.19648308507644077,
+      "learning_rate": 2.5180722891566265e-05,
+      "loss": 0.3606,
+      "step": 1766
+    },
+    {
+      "epoch": 1.6409660938225732,
+      "grad_norm": 0.2302898022257594,
+      "learning_rate": 2.5163511187607573e-05,
+      "loss": 0.3521,
+      "step": 1767
+    },
+    {
+      "epoch": 1.6418950301904318,
+      "grad_norm": 0.2324588872518842,
+      "learning_rate": 2.514629948364888e-05,
+      "loss": 0.3551,
+      "step": 1768
+    },
+    {
+      "epoch": 1.6428239665582907,
+      "grad_norm": 0.22587569032785934,
+      "learning_rate": 2.512908777969019e-05,
+      "loss": 0.3874,
+      "step": 1769
+    },
+    {
+      "epoch": 1.6437529029261495,
+      "grad_norm": 0.2281758443165584,
+      "learning_rate": 2.5111876075731498e-05,
+      "loss": 0.3501,
+      "step": 1770
+    },
+    {
+      "epoch": 1.6446818392940084,
+      "grad_norm": 0.20674500113194846,
+      "learning_rate": 2.5094664371772807e-05,
+      "loss": 0.3321,
+      "step": 1771
+    },
+    {
+      "epoch": 1.6456107756618672,
+      "grad_norm": 0.18216341746652498,
+      "learning_rate": 2.5077452667814115e-05,
+      "loss": 0.3451,
+      "step": 1772
+    },
+    {
+      "epoch": 1.646539712029726,
+      "grad_norm": 0.24083907570929008,
+      "learning_rate": 2.5060240963855423e-05,
+      "loss": 0.3571,
+      "step": 1773
+    },
+    {
+      "epoch": 1.647468648397585,
+      "grad_norm": 0.21053850263699328,
+      "learning_rate": 2.5043029259896732e-05,
+      "loss": 0.3393,
+      "step": 1774
+    },
+    {
+      "epoch": 1.6483975847654435,
+      "grad_norm": 0.21190369434405645,
+      "learning_rate": 2.502581755593804e-05,
+      "loss": 0.3562,
+      "step": 1775
+    },
+    {
+      "epoch": 1.6493265211333024,
+      "grad_norm": 0.22135502539564378,
+      "learning_rate": 2.500860585197935e-05,
+      "loss": 0.3525,
+      "step": 1776
+    },
+    {
+      "epoch": 1.6502554575011612,
+      "grad_norm": 0.22367773820582154,
+      "learning_rate": 2.4991394148020654e-05,
+      "loss": 0.3466,
+      "step": 1777
+    },
+    {
+      "epoch": 1.6511843938690198,
+      "grad_norm": 0.21451556990315485,
+      "learning_rate": 2.4974182444061962e-05,
+      "loss": 0.3338,
+      "step": 1778
+    },
+    {
+      "epoch": 1.6521133302368787,
+      "grad_norm": 0.2465298762993094,
+      "learning_rate": 2.495697074010327e-05,
+      "loss": 0.333,
+      "step": 1779
+    },
+    {
+      "epoch": 1.6530422666047375,
+      "grad_norm": 0.2547044770751165,
+      "learning_rate": 2.493975903614458e-05,
+      "loss": 0.4001,
+      "step": 1780
+    },
+    {
+      "epoch": 1.6539712029725964,
+      "grad_norm": 0.22774781080068754,
+      "learning_rate": 2.4922547332185887e-05,
+      "loss": 0.3461,
+      "step": 1781
+    },
+    {
+      "epoch": 1.6549001393404552,
+      "grad_norm": 0.21204449913798434,
+      "learning_rate": 2.4905335628227196e-05,
+      "loss": 0.3274,
+      "step": 1782
+    },
+    {
+      "epoch": 1.655829075708314,
+      "grad_norm": 0.21580956854064373,
+      "learning_rate": 2.4888123924268504e-05,
+      "loss": 0.3359,
+      "step": 1783
+    },
+    {
+      "epoch": 1.656758012076173,
+      "grad_norm": 0.24076908609466494,
+      "learning_rate": 2.4870912220309813e-05,
+      "loss": 0.3588,
+      "step": 1784
+    },
+    {
+      "epoch": 1.6576869484440317,
+      "grad_norm": 0.2125733983681432,
+      "learning_rate": 2.485370051635112e-05,
+      "loss": 0.3295,
+      "step": 1785
+    },
+    {
+      "epoch": 1.6586158848118904,
+      "grad_norm": 0.1890708531709907,
+      "learning_rate": 2.483648881239243e-05,
+      "loss": 0.3418,
+      "step": 1786
+    },
+    {
+      "epoch": 1.6595448211797492,
+      "grad_norm": 0.19675942962384732,
+      "learning_rate": 2.4819277108433738e-05,
+      "loss": 0.3472,
+      "step": 1787
+    },
+    {
+      "epoch": 1.6604737575476078,
+      "grad_norm": 0.2170796522409692,
+      "learning_rate": 2.4802065404475046e-05,
+      "loss": 0.3457,
+      "step": 1788
+    },
+    {
+      "epoch": 1.6614026939154667,
+      "grad_norm": 0.21149308045148213,
+      "learning_rate": 2.478485370051635e-05,
+      "loss": 0.3759,
+      "step": 1789
+    },
+    {
+      "epoch": 1.6623316302833255,
+      "grad_norm": 0.20813548096362552,
+      "learning_rate": 2.476764199655766e-05,
+      "loss": 0.3582,
+      "step": 1790
+    },
+    {
+      "epoch": 1.6632605666511844,
+      "grad_norm": 0.19034774277541827,
+      "learning_rate": 2.4750430292598968e-05,
+      "loss": 0.3489,
+      "step": 1791
+    },
+    {
+      "epoch": 1.6641895030190432,
+      "grad_norm": 0.19380501073483952,
+      "learning_rate": 2.4733218588640276e-05,
+      "loss": 0.3339,
+      "step": 1792
+    },
+    {
+      "epoch": 1.665118439386902,
+      "grad_norm": 0.1834320423794678,
+      "learning_rate": 2.4716006884681585e-05,
+      "loss": 0.3617,
+      "step": 1793
+    },
+    {
+      "epoch": 1.666047375754761,
+      "grad_norm": 0.24296412127242056,
+      "learning_rate": 2.4698795180722893e-05,
+      "loss": 0.3596,
+      "step": 1794
+    },
+    {
+      "epoch": 1.6669763121226198,
+      "grad_norm": 0.2280578339681279,
+      "learning_rate": 2.46815834767642e-05,
+      "loss": 0.3584,
+      "step": 1795
+    },
+    {
+      "epoch": 1.6679052484904784,
+      "grad_norm": 0.19120766695593186,
+      "learning_rate": 2.466437177280551e-05,
+      "loss": 0.3479,
+      "step": 1796
+    },
+    {
+      "epoch": 1.6688341848583372,
+      "grad_norm": 0.2310771090381013,
+      "learning_rate": 2.464716006884682e-05,
+      "loss": 0.3555,
+      "step": 1797
+    },
+    {
+      "epoch": 1.6697631212261959,
+      "grad_norm": 0.18985210461541444,
+      "learning_rate": 2.4629948364888127e-05,
+      "loss": 0.3356,
+      "step": 1798
+    },
+    {
+      "epoch": 1.6706920575940547,
+      "grad_norm": 0.19786178896000547,
+      "learning_rate": 2.4612736660929435e-05,
+      "loss": 0.3374,
+      "step": 1799
+    },
+    {
+      "epoch": 1.6716209939619135,
+      "grad_norm": 0.23092029223763727,
+      "learning_rate": 2.4595524956970744e-05,
+      "loss": 0.3447,
+      "step": 1800
+    },
+    {
+      "epoch": 1.6725499303297724,
+      "grad_norm": 0.1922509738346118,
+      "learning_rate": 2.4578313253012052e-05,
+      "loss": 0.3292,
+      "step": 1801
+    },
+    {
+      "epoch": 1.6734788666976312,
+      "grad_norm": 0.21074032243268462,
+      "learning_rate": 2.4561101549053357e-05,
+      "loss": 0.3591,
+      "step": 1802
+    },
+    {
+      "epoch": 1.67440780306549,
+      "grad_norm": 0.2210145887813637,
+      "learning_rate": 2.4543889845094665e-05,
+      "loss": 0.3408,
+      "step": 1803
+    },
+    {
+      "epoch": 1.675336739433349,
+      "grad_norm": 0.1955049882758258,
+      "learning_rate": 2.4526678141135974e-05,
+      "loss": 0.3252,
+      "step": 1804
+    },
+    {
+      "epoch": 1.6762656758012078,
+      "grad_norm": 0.20630880882991626,
+      "learning_rate": 2.4509466437177282e-05,
+      "loss": 0.3343,
+      "step": 1805
+    },
+    {
+      "epoch": 1.6771946121690664,
+      "grad_norm": 0.2068467693733224,
+      "learning_rate": 2.449225473321859e-05,
+      "loss": 0.3718,
+      "step": 1806
+    },
+    {
+      "epoch": 1.6781235485369252,
+      "grad_norm": 0.21746922099285282,
+      "learning_rate": 2.44750430292599e-05,
+      "loss": 0.3473,
+      "step": 1807
+    },
+    {
+      "epoch": 1.6790524849047839,
+      "grad_norm": 0.22531439643757564,
+      "learning_rate": 2.4457831325301207e-05,
+      "loss": 0.3378,
+      "step": 1808
+    },
+    {
+      "epoch": 1.6799814212726427,
+      "grad_norm": 0.16961930850382043,
+      "learning_rate": 2.4440619621342516e-05,
+      "loss": 0.3307,
+      "step": 1809
+    },
+    {
+      "epoch": 1.6809103576405016,
+      "grad_norm": 0.2184425585470951,
+      "learning_rate": 2.4423407917383824e-05,
+      "loss": 0.3444,
+      "step": 1810
+    },
+    {
+      "epoch": 1.6818392940083604,
+      "grad_norm": 0.20341565894023106,
+      "learning_rate": 2.4406196213425133e-05,
+      "loss": 0.341,
+      "step": 1811
+    },
+    {
+      "epoch": 1.6827682303762193,
+      "grad_norm": 0.20083870829005598,
+      "learning_rate": 2.438898450946644e-05,
+      "loss": 0.3561,
+      "step": 1812
+    },
+    {
+      "epoch": 1.683697166744078,
+      "grad_norm": 0.22620645826766164,
+      "learning_rate": 2.437177280550775e-05,
+      "loss": 0.3585,
+      "step": 1813
+    },
+    {
+      "epoch": 1.684626103111937,
+      "grad_norm": 0.1941948141031554,
+      "learning_rate": 2.4354561101549054e-05,
+      "loss": 0.3443,
+      "step": 1814
+    },
+    {
+      "epoch": 1.6855550394797958,
+      "grad_norm": 0.2277987840149738,
+      "learning_rate": 2.4337349397590363e-05,
+      "loss": 0.3595,
+      "step": 1815
+    },
+    {
+      "epoch": 1.6864839758476544,
+      "grad_norm": 0.19485898183092634,
+      "learning_rate": 2.432013769363167e-05,
+      "loss": 0.322,
+      "step": 1816
+    },
+    {
+      "epoch": 1.6874129122155133,
+      "grad_norm": 0.18603940637738384,
+      "learning_rate": 2.430292598967298e-05,
+      "loss": 0.3582,
+      "step": 1817
+    },
+    {
+      "epoch": 1.6883418485833719,
+      "grad_norm": 0.23071978533124235,
+      "learning_rate": 2.4285714285714288e-05,
+      "loss": 0.3644,
+      "step": 1818
+    },
+    {
+      "epoch": 1.6892707849512307,
+      "grad_norm": 0.20747311082414033,
+      "learning_rate": 2.4268502581755596e-05,
+      "loss": 0.3373,
+      "step": 1819
+    },
+    {
+      "epoch": 1.6901997213190896,
+      "grad_norm": 0.19539815282798248,
+      "learning_rate": 2.4251290877796905e-05,
+      "loss": 0.3348,
+      "step": 1820
+    },
+    {
+      "epoch": 1.6911286576869484,
+      "grad_norm": 0.21365984840479363,
+      "learning_rate": 2.4234079173838213e-05,
+      "loss": 0.344,
+      "step": 1821
+    },
+    {
+      "epoch": 1.6920575940548073,
+      "grad_norm": 0.21504209447331304,
+      "learning_rate": 2.421686746987952e-05,
+      "loss": 0.3439,
+      "step": 1822
+    },
+    {
+      "epoch": 1.6929865304226661,
+      "grad_norm": 0.1982570782906239,
+      "learning_rate": 2.419965576592083e-05,
+      "loss": 0.3507,
+      "step": 1823
+    },
+    {
+      "epoch": 1.693915466790525,
+      "grad_norm": 0.19337549611458893,
+      "learning_rate": 2.418244406196214e-05,
+      "loss": 0.3591,
+      "step": 1824
+    },
+    {
+      "epoch": 1.6948444031583838,
+      "grad_norm": 0.20348014807541867,
+      "learning_rate": 2.4165232358003443e-05,
+      "loss": 0.325,
+      "step": 1825
+    },
+    {
+      "epoch": 1.6957733395262424,
+      "grad_norm": 0.20540644281014037,
+      "learning_rate": 2.4148020654044752e-05,
+      "loss": 0.3411,
+      "step": 1826
+    },
+    {
+      "epoch": 1.6967022758941013,
+      "grad_norm": 0.20236522736446938,
+      "learning_rate": 2.413080895008606e-05,
+      "loss": 0.3551,
+      "step": 1827
+    },
+    {
+      "epoch": 1.69763121226196,
+      "grad_norm": 0.20198047155898852,
+      "learning_rate": 2.411359724612737e-05,
+      "loss": 0.3477,
+      "step": 1828
+    },
+    {
+      "epoch": 1.6985601486298187,
+      "grad_norm": 0.21306903457439758,
+      "learning_rate": 2.4096385542168677e-05,
+      "loss": 0.3423,
+      "step": 1829
+    },
+    {
+      "epoch": 1.6994890849976776,
+      "grad_norm": 0.19666890014745034,
+      "learning_rate": 2.4079173838209985e-05,
+      "loss": 0.3444,
+      "step": 1830
+    },
+    {
+      "epoch": 1.7004180213655364,
+      "grad_norm": 0.19584912929425907,
+      "learning_rate": 2.4061962134251294e-05,
+      "loss": 0.3541,
+      "step": 1831
+    },
+    {
+      "epoch": 1.7013469577333953,
+      "grad_norm": 0.20466324373230668,
+      "learning_rate": 2.4044750430292602e-05,
+      "loss": 0.3537,
+      "step": 1832
+    },
+    {
+      "epoch": 1.7022758941012541,
+      "grad_norm": 0.20414327533805723,
+      "learning_rate": 2.402753872633391e-05,
+      "loss": 0.3491,
+      "step": 1833
+    },
+    {
+      "epoch": 1.703204830469113,
+      "grad_norm": 0.17877641915198147,
+      "learning_rate": 2.401032702237522e-05,
+      "loss": 0.3306,
+      "step": 1834
+    },
+    {
+      "epoch": 1.7041337668369718,
+      "grad_norm": 0.19756753438210706,
+      "learning_rate": 2.3993115318416524e-05,
+      "loss": 0.3522,
+      "step": 1835
+    },
+    {
+      "epoch": 1.7050627032048304,
+      "grad_norm": 0.20624963074350838,
+      "learning_rate": 2.3975903614457832e-05,
+      "loss": 0.3271,
+      "step": 1836
+    },
+    {
+      "epoch": 1.7059916395726893,
+      "grad_norm": 0.18278177540349813,
+      "learning_rate": 2.395869191049914e-05,
+      "loss": 0.3285,
+      "step": 1837
+    },
+    {
+      "epoch": 1.706920575940548,
+      "grad_norm": 0.18016020662628543,
+      "learning_rate": 2.394148020654045e-05,
+      "loss": 0.3512,
+      "step": 1838
+    },
+    {
+      "epoch": 1.7078495123084068,
+      "grad_norm": 0.19957513588627965,
+      "learning_rate": 2.3924268502581758e-05,
+      "loss": 0.342,
+      "step": 1839
+    },
+    {
+      "epoch": 1.7087784486762656,
+      "grad_norm": 0.19966706741908602,
+      "learning_rate": 2.3907056798623066e-05,
+      "loss": 0.346,
+      "step": 1840
+    },
+    {
+      "epoch": 1.7097073850441245,
+      "grad_norm": 0.18844002311194907,
+      "learning_rate": 2.3889845094664374e-05,
+      "loss": 0.3448,
+      "step": 1841
+    },
+    {
+      "epoch": 1.7106363214119833,
+      "grad_norm": 0.19040969107600308,
+      "learning_rate": 2.3872633390705683e-05,
+      "loss": 0.3606,
+      "step": 1842
+    },
+    {
+      "epoch": 1.7115652577798421,
+      "grad_norm": 0.19757417243583347,
+      "learning_rate": 2.385542168674699e-05,
+      "loss": 0.3299,
+      "step": 1843
+    },
+    {
+      "epoch": 1.712494194147701,
+      "grad_norm": 0.19669847871271992,
+      "learning_rate": 2.38382099827883e-05,
+      "loss": 0.3628,
+      "step": 1844
+    },
+    {
+      "epoch": 1.7134231305155598,
+      "grad_norm": 0.1817638016455849,
+      "learning_rate": 2.3820998278829608e-05,
+      "loss": 0.3723,
+      "step": 1845
+    },
+    {
+      "epoch": 1.7143520668834185,
+      "grad_norm": 0.19044959623520683,
+      "learning_rate": 2.3803786574870913e-05,
+      "loss": 0.336,
+      "step": 1846
+    },
+    {
+      "epoch": 1.7152810032512773,
+      "grad_norm": 0.20454387824331371,
+      "learning_rate": 2.378657487091222e-05,
+      "loss": 0.3561,
+      "step": 1847
+    },
+    {
+      "epoch": 1.716209939619136,
+      "grad_norm": 0.2101073375770777,
+      "learning_rate": 2.376936316695353e-05,
+      "loss": 0.3373,
+      "step": 1848
+    },
+    {
+      "epoch": 1.7171388759869948,
+      "grad_norm": 0.200320939170345,
+      "learning_rate": 2.3752151462994838e-05,
+      "loss": 0.3404,
+      "step": 1849
+    },
+    {
+      "epoch": 1.7180678123548536,
+      "grad_norm": 0.1952985352033432,
+      "learning_rate": 2.3734939759036147e-05,
+      "loss": 0.3518,
+      "step": 1850
+    },
+    {
+      "epoch": 1.7189967487227125,
+      "grad_norm": 0.19447811004540178,
+      "learning_rate": 2.3717728055077455e-05,
+      "loss": 0.3428,
+      "step": 1851
+    },
+    {
+      "epoch": 1.7199256850905713,
+      "grad_norm": 0.23780369028264597,
+      "learning_rate": 2.3700516351118763e-05,
+      "loss": 0.3454,
+      "step": 1852
+    },
+    {
+      "epoch": 1.7208546214584302,
+      "grad_norm": 0.21663178518511156,
+      "learning_rate": 2.3683304647160072e-05,
+      "loss": 0.3585,
+      "step": 1853
+    },
+    {
+      "epoch": 1.721783557826289,
+      "grad_norm": 0.1920341932268259,
+      "learning_rate": 2.366609294320138e-05,
+      "loss": 0.3373,
+      "step": 1854
+    },
+    {
+      "epoch": 1.7227124941941478,
+      "grad_norm": 0.20088003096202445,
+      "learning_rate": 2.364888123924269e-05,
+      "loss": 0.3408,
+      "step": 1855
+    },
+    {
+      "epoch": 1.7236414305620065,
+      "grad_norm": 0.19365614048712398,
+      "learning_rate": 2.3631669535283997e-05,
+      "loss": 0.3297,
+      "step": 1856
+    },
+    {
+      "epoch": 1.7245703669298653,
+      "grad_norm": 0.1906200226199188,
+      "learning_rate": 2.3614457831325302e-05,
+      "loss": 0.352,
+      "step": 1857
+    },
+    {
+      "epoch": 1.725499303297724,
+      "grad_norm": 0.20155735822067727,
+      "learning_rate": 2.359724612736661e-05,
+      "loss": 0.34,
+      "step": 1858
+    },
+    {
+      "epoch": 1.7264282396655828,
+      "grad_norm": 0.19158534386586176,
+      "learning_rate": 2.358003442340792e-05,
+      "loss": 0.3243,
+      "step": 1859
+    },
+    {
+      "epoch": 1.7273571760334416,
+      "grad_norm": 0.20500769628661117,
+      "learning_rate": 2.3562822719449227e-05,
+      "loss": 0.3318,
+      "step": 1860
+    },
+    {
+      "epoch": 1.7282861124013005,
+      "grad_norm": 0.211428858561031,
+      "learning_rate": 2.3545611015490536e-05,
+      "loss": 0.3332,
+      "step": 1861
+    },
+    {
+      "epoch": 1.7292150487691593,
+      "grad_norm": 0.18958780180075827,
+      "learning_rate": 2.3528399311531844e-05,
+      "loss": 0.3391,
+      "step": 1862
+    },
+    {
+      "epoch": 1.7301439851370182,
+      "grad_norm": 0.19932551881534102,
+      "learning_rate": 2.3511187607573152e-05,
+      "loss": 0.3354,
+      "step": 1863
+    },
+    {
+      "epoch": 1.731072921504877,
+      "grad_norm": 0.19686069428098435,
+      "learning_rate": 2.349397590361446e-05,
+      "loss": 0.3343,
+      "step": 1864
+    },
+    {
+      "epoch": 1.7320018578727359,
+      "grad_norm": 0.1874048143051562,
+      "learning_rate": 2.347676419965577e-05,
+      "loss": 0.3331,
+      "step": 1865
+    },
+    {
+      "epoch": 1.7329307942405945,
+      "grad_norm": 0.22051015667350352,
+      "learning_rate": 2.3459552495697078e-05,
+      "loss": 0.3406,
+      "step": 1866
+    },
+    {
+      "epoch": 1.7338597306084533,
+      "grad_norm": 0.19652476043071881,
+      "learning_rate": 2.3442340791738383e-05,
+      "loss": 0.3413,
+      "step": 1867
+    },
+    {
+      "epoch": 1.7347886669763122,
+      "grad_norm": 0.18429305065219817,
+      "learning_rate": 2.342512908777969e-05,
+      "loss": 0.3392,
+      "step": 1868
+    },
+    {
+      "epoch": 1.7357176033441708,
+      "grad_norm": 0.21153170060670934,
+      "learning_rate": 2.3407917383821e-05,
+      "loss": 0.3537,
+      "step": 1869
+    },
+    {
+      "epoch": 1.7366465397120296,
+      "grad_norm": 0.19456879357822598,
+      "learning_rate": 2.3390705679862308e-05,
+      "loss": 0.3405,
+      "step": 1870
+    },
+    {
+      "epoch": 1.7375754760798885,
+      "grad_norm": 0.19705924651045284,
+      "learning_rate": 2.3373493975903616e-05,
+      "loss": 0.3518,
+      "step": 1871
+    },
+    {
+      "epoch": 1.7385044124477473,
+      "grad_norm": 0.23371156033466298,
+      "learning_rate": 2.3356282271944925e-05,
+      "loss": 0.348,
+      "step": 1872
+    },
+    {
+      "epoch": 1.7394333488156062,
+      "grad_norm": 0.17949864900216111,
+      "learning_rate": 2.3339070567986233e-05,
+      "loss": 0.3411,
+      "step": 1873
+    },
+    {
+      "epoch": 1.740362285183465,
+      "grad_norm": 0.19886700512246908,
+      "learning_rate": 2.332185886402754e-05,
+      "loss": 0.3447,
+      "step": 1874
+    },
+    {
+      "epoch": 1.7412912215513239,
+      "grad_norm": 0.2084202011104834,
+      "learning_rate": 2.330464716006885e-05,
+      "loss": 0.3571,
+      "step": 1875
+    },
+    {
+      "epoch": 1.7422201579191825,
+      "grad_norm": 0.18145128888963527,
+      "learning_rate": 2.3287435456110158e-05,
+      "loss": 0.339,
+      "step": 1876
+    },
+    {
+      "epoch": 1.7431490942870413,
+      "grad_norm": 0.19420189258092047,
+      "learning_rate": 2.3270223752151467e-05,
+      "loss": 0.3496,
+      "step": 1877
+    },
+    {
+      "epoch": 1.7440780306549002,
+      "grad_norm": 0.19829094715505624,
+      "learning_rate": 2.325301204819277e-05,
+      "loss": 0.3137,
+      "step": 1878
+    },
+    {
+      "epoch": 1.7450069670227588,
+      "grad_norm": 0.20101000694540003,
+      "learning_rate": 2.323580034423408e-05,
+      "loss": 0.3544,
+      "step": 1879
+    },
+    {
+      "epoch": 1.7459359033906177,
+      "grad_norm": 0.1905563892060915,
+      "learning_rate": 2.321858864027539e-05,
+      "loss": 0.3345,
+      "step": 1880
+    },
+    {
+      "epoch": 1.7468648397584765,
+      "grad_norm": 0.19097935275198497,
+      "learning_rate": 2.3201376936316697e-05,
+      "loss": 0.3483,
+      "step": 1881
+    },
+    {
+      "epoch": 1.7477937761263354,
+      "grad_norm": 0.1991523378048746,
+      "learning_rate": 2.3184165232358005e-05,
+      "loss": 0.3345,
+      "step": 1882
+    },
+    {
+      "epoch": 1.7487227124941942,
+      "grad_norm": 0.1960091282122507,
+      "learning_rate": 2.3166953528399314e-05,
+      "loss": 0.3309,
+      "step": 1883
+    },
+    {
+      "epoch": 1.749651648862053,
+      "grad_norm": 0.18402497230694953,
+      "learning_rate": 2.3149741824440622e-05,
+      "loss": 0.3414,
+      "step": 1884
+    },
+    {
+      "epoch": 1.750580585229912,
+      "grad_norm": 0.18159596100294748,
+      "learning_rate": 2.313253012048193e-05,
+      "loss": 0.3571,
+      "step": 1885
+    },
+    {
+      "epoch": 1.7515095215977705,
+      "grad_norm": 0.20590477472666366,
+      "learning_rate": 2.311531841652324e-05,
+      "loss": 0.3403,
+      "step": 1886
+    },
+    {
+      "epoch": 1.7524384579656294,
+      "grad_norm": 0.18739986818056353,
+      "learning_rate": 2.3098106712564547e-05,
+      "loss": 0.368,
+      "step": 1887
+    },
+    {
+      "epoch": 1.7533673943334882,
+      "grad_norm": 0.1922357485644034,
+      "learning_rate": 2.3080895008605852e-05,
+      "loss": 0.3622,
+      "step": 1888
+    },
+    {
+      "epoch": 1.7542963307013468,
+      "grad_norm": 0.1907104113911057,
+      "learning_rate": 2.306368330464716e-05,
+      "loss": 0.337,
+      "step": 1889
+    },
+    {
+      "epoch": 1.7552252670692057,
+      "grad_norm": 0.18926167592617857,
+      "learning_rate": 2.304647160068847e-05,
+      "loss": 0.3637,
+      "step": 1890
+    },
+    {
+      "epoch": 1.7561542034370645,
+      "grad_norm": 0.19061055691996645,
+      "learning_rate": 2.3029259896729777e-05,
+      "loss": 0.3605,
+      "step": 1891
+    },
+    {
+      "epoch": 1.7570831398049234,
+      "grad_norm": 0.301161005754076,
+      "learning_rate": 2.3012048192771086e-05,
+      "loss": 0.3274,
+      "step": 1892
+    },
+    {
+      "epoch": 1.7580120761727822,
+      "grad_norm": 0.17079160860066211,
+      "learning_rate": 2.2994836488812394e-05,
+      "loss": 0.3499,
+      "step": 1893
+    },
+    {
+      "epoch": 1.758941012540641,
+      "grad_norm": 0.20510016168695175,
+      "learning_rate": 2.2977624784853703e-05,
+      "loss": 0.3718,
+      "step": 1894
+    },
+    {
+      "epoch": 1.7598699489085,
+      "grad_norm": 0.1860961861877371,
+      "learning_rate": 2.296041308089501e-05,
+      "loss": 0.3446,
+      "step": 1895
+    },
+    {
+      "epoch": 1.7607988852763585,
+      "grad_norm": 0.1823482266004319,
+      "learning_rate": 2.294320137693632e-05,
+      "loss": 0.3444,
+      "step": 1896
+    },
+    {
+      "epoch": 1.7617278216442174,
+      "grad_norm": 0.181834177614641,
+      "learning_rate": 2.2925989672977628e-05,
+      "loss": 0.352,
+      "step": 1897
+    },
+    {
+      "epoch": 1.7626567580120762,
+      "grad_norm": 0.189848159943081,
+      "learning_rate": 2.2908777969018936e-05,
+      "loss": 0.3502,
+      "step": 1898
+    },
+    {
+      "epoch": 1.7635856943799348,
+      "grad_norm": 0.19580044122071402,
+      "learning_rate": 2.289156626506024e-05,
+      "loss": 0.343,
+      "step": 1899
+    },
+    {
+      "epoch": 1.7645146307477937,
+      "grad_norm": 0.1958254403858763,
+      "learning_rate": 2.287435456110155e-05,
+      "loss": 0.3301,
+      "step": 1900
+    },
+    {
+      "epoch": 1.7654435671156525,
+      "grad_norm": 0.1714082079009884,
+      "learning_rate": 2.2857142857142858e-05,
+      "loss": 0.3136,
+      "step": 1901
+    },
+    {
+      "epoch": 1.7663725034835114,
+      "grad_norm": 0.188207144023739,
+      "learning_rate": 2.2839931153184166e-05,
+      "loss": 0.3496,
+      "step": 1902
+    },
+    {
+      "epoch": 1.7673014398513702,
+      "grad_norm": 0.19149325753512772,
+      "learning_rate": 2.2822719449225475e-05,
+      "loss": 0.3471,
+      "step": 1903
+    },
+    {
+      "epoch": 1.768230376219229,
+      "grad_norm": 0.20844243247605607,
+      "learning_rate": 2.2805507745266783e-05,
+      "loss": 0.3635,
+      "step": 1904
+    },
+    {
+      "epoch": 1.769159312587088,
+      "grad_norm": 0.21302605597258154,
+      "learning_rate": 2.278829604130809e-05,
+      "loss": 0.3586,
+      "step": 1905
+    },
+    {
+      "epoch": 1.7700882489549465,
+      "grad_norm": 0.18803392186519952,
+      "learning_rate": 2.27710843373494e-05,
+      "loss": 0.3236,
+      "step": 1906
+    },
+    {
+      "epoch": 1.7710171853228054,
+      "grad_norm": 0.19461739487306387,
+      "learning_rate": 2.275387263339071e-05,
+      "loss": 0.3336,
+      "step": 1907
+    },
+    {
+      "epoch": 1.7719461216906642,
+      "grad_norm": 0.19855123181901438,
+      "learning_rate": 2.2736660929432017e-05,
+      "loss": 0.328,
+      "step": 1908
+    },
+    {
+      "epoch": 1.7728750580585229,
+      "grad_norm": 0.193296807754737,
+      "learning_rate": 2.2719449225473325e-05,
+      "loss": 0.3429,
+      "step": 1909
+    },
+    {
+      "epoch": 1.7738039944263817,
+      "grad_norm": 0.18250714695301445,
+      "learning_rate": 2.270223752151463e-05,
+      "loss": 0.3402,
+      "step": 1910
+    },
+    {
+      "epoch": 1.7747329307942405,
+      "grad_norm": 0.2207728048782478,
+      "learning_rate": 2.268502581755594e-05,
+      "loss": 0.3648,
+      "step": 1911
+    },
+    {
+      "epoch": 1.7756618671620994,
+      "grad_norm": 0.20763889726318494,
+      "learning_rate": 2.2667814113597247e-05,
+      "loss": 0.335,
+      "step": 1912
+    },
+    {
+      "epoch": 1.7765908035299582,
+      "grad_norm": 0.1802847803327561,
+      "learning_rate": 2.2650602409638555e-05,
+      "loss": 0.3659,
+      "step": 1913
+    },
+    {
+      "epoch": 1.777519739897817,
+      "grad_norm": 0.2017058059062457,
+      "learning_rate": 2.2633390705679864e-05,
+      "loss": 0.3229,
+      "step": 1914
+    },
+    {
+      "epoch": 1.778448676265676,
+      "grad_norm": 0.18380005279157755,
+      "learning_rate": 2.2616179001721172e-05,
+      "loss": 0.3297,
+      "step": 1915
+    },
+    {
+      "epoch": 1.7793776126335346,
+      "grad_norm": 0.20980299060848648,
+      "learning_rate": 2.259896729776248e-05,
+      "loss": 0.3603,
+      "step": 1916
+    },
+    {
+      "epoch": 1.7803065490013934,
+      "grad_norm": 0.21845300686706248,
+      "learning_rate": 2.258175559380379e-05,
+      "loss": 0.3526,
+      "step": 1917
+    },
+    {
+      "epoch": 1.7812354853692522,
+      "grad_norm": 0.18306805762943487,
+      "learning_rate": 2.2564543889845097e-05,
+      "loss": 0.3427,
+      "step": 1918
+    },
+    {
+      "epoch": 1.7821644217371109,
+      "grad_norm": 0.22267478907178537,
+      "learning_rate": 2.2547332185886406e-05,
+      "loss": 0.3502,
+      "step": 1919
+    },
+    {
+      "epoch": 1.7830933581049697,
+      "grad_norm": 0.2114169092592862,
+      "learning_rate": 2.253012048192771e-05,
+      "loss": 0.337,
+      "step": 1920
+    },
+    {
+      "epoch": 1.7840222944728286,
+      "grad_norm": 0.2114698629778862,
+      "learning_rate": 2.251290877796902e-05,
+      "loss": 0.3331,
+      "step": 1921
+    },
+    {
+      "epoch": 1.7849512308406874,
+      "grad_norm": 0.1737401942597242,
+      "learning_rate": 2.2495697074010328e-05,
+      "loss": 0.3402,
+      "step": 1922
+    },
+    {
+      "epoch": 1.7858801672085463,
+      "grad_norm": 0.21860276556477812,
+      "learning_rate": 2.2478485370051636e-05,
+      "loss": 0.3785,
+      "step": 1923
+    },
+    {
+      "epoch": 1.786809103576405,
+      "grad_norm": 0.21990594737894525,
+      "learning_rate": 2.2461273666092944e-05,
+      "loss": 0.333,
+      "step": 1924
+    },
+    {
+      "epoch": 1.787738039944264,
+      "grad_norm": 0.17965595351731087,
+      "learning_rate": 2.2444061962134253e-05,
+      "loss": 0.3161,
+      "step": 1925
+    },
+    {
+      "epoch": 1.7886669763121226,
+      "grad_norm": 0.18037444567864072,
+      "learning_rate": 2.242685025817556e-05,
+      "loss": 0.3218,
+      "step": 1926
+    },
+    {
+      "epoch": 1.7895959126799814,
+      "grad_norm": 0.19458012919376982,
+      "learning_rate": 2.240963855421687e-05,
+      "loss": 0.3445,
+      "step": 1927
+    },
+    {
+      "epoch": 1.7905248490478403,
+      "grad_norm": 0.1870071334265635,
+      "learning_rate": 2.2392426850258178e-05,
+      "loss": 0.3467,
+      "step": 1928
+    },
+    {
+      "epoch": 1.7914537854156989,
+      "grad_norm": 0.20143024738216037,
+      "learning_rate": 2.2375215146299486e-05,
+      "loss": 0.3667,
+      "step": 1929
+    },
+    {
+      "epoch": 1.7923827217835577,
+      "grad_norm": 0.18517571958598458,
+      "learning_rate": 2.2358003442340795e-05,
+      "loss": 0.347,
+      "step": 1930
+    },
+    {
+      "epoch": 1.7933116581514166,
+      "grad_norm": 0.18972740321401171,
+      "learning_rate": 2.23407917383821e-05,
+      "loss": 0.3561,
+      "step": 1931
+    },
+    {
+      "epoch": 1.7942405945192754,
+      "grad_norm": 0.18681340130076599,
+      "learning_rate": 2.2323580034423408e-05,
+      "loss": 0.3526,
+      "step": 1932
+    },
+    {
+      "epoch": 1.7951695308871343,
+      "grad_norm": 0.17839062435491831,
+      "learning_rate": 2.2306368330464717e-05,
+      "loss": 0.3492,
+      "step": 1933
+    },
+    {
+      "epoch": 1.7960984672549931,
+      "grad_norm": 0.19307814521598352,
+      "learning_rate": 2.2289156626506025e-05,
+      "loss": 0.3543,
+      "step": 1934
+    },
+    {
+      "epoch": 1.797027403622852,
+      "grad_norm": 0.18708013233718124,
+      "learning_rate": 2.2271944922547333e-05,
+      "loss": 0.3459,
+      "step": 1935
+    },
+    {
+      "epoch": 1.7979563399907106,
+      "grad_norm": 0.17553651305256152,
+      "learning_rate": 2.2254733218588642e-05,
+      "loss": 0.3598,
+      "step": 1936
+    },
+    {
+      "epoch": 1.7988852763585694,
+      "grad_norm": 0.18618042918495897,
+      "learning_rate": 2.223752151462995e-05,
+      "loss": 0.3248,
+      "step": 1937
+    },
+    {
+      "epoch": 1.7998142127264283,
+      "grad_norm": 0.17742356291112643,
+      "learning_rate": 2.222030981067126e-05,
+      "loss": 0.3367,
+      "step": 1938
+    },
+    {
+      "epoch": 1.800743149094287,
+      "grad_norm": 0.1783779824099535,
+      "learning_rate": 2.2203098106712567e-05,
+      "loss": 0.3459,
+      "step": 1939
+    },
+    {
+      "epoch": 1.8016720854621457,
+      "grad_norm": 0.1873864409342349,
+      "learning_rate": 2.2185886402753875e-05,
+      "loss": 0.3411,
+      "step": 1940
+    },
+    {
+      "epoch": 1.8026010218300046,
+      "grad_norm": 0.19820459741704952,
+      "learning_rate": 2.216867469879518e-05,
+      "loss": 0.3622,
+      "step": 1941
+    },
+    {
+      "epoch": 1.8035299581978634,
+      "grad_norm": 0.19579829256800121,
+      "learning_rate": 2.215146299483649e-05,
+      "loss": 0.3488,
+      "step": 1942
+    },
+    {
+      "epoch": 1.8044588945657223,
+      "grad_norm": 0.18994419202784169,
+      "learning_rate": 2.2134251290877797e-05,
+      "loss": 0.3274,
+      "step": 1943
+    },
+    {
+      "epoch": 1.8053878309335811,
+      "grad_norm": 0.17560731190213005,
+      "learning_rate": 2.2117039586919106e-05,
+      "loss": 0.3249,
+      "step": 1944
+    },
+    {
+      "epoch": 1.80631676730144,
+      "grad_norm": 0.18332672388432894,
+      "learning_rate": 2.2099827882960414e-05,
+      "loss": 0.3287,
+      "step": 1945
+    },
+    {
+      "epoch": 1.8072457036692986,
+      "grad_norm": 0.17542344527664755,
+      "learning_rate": 2.2082616179001722e-05,
+      "loss": 0.3438,
+      "step": 1946
+    },
+    {
+      "epoch": 1.8081746400371574,
+      "grad_norm": 0.19087927519310427,
+      "learning_rate": 2.206540447504303e-05,
+      "loss": 0.3422,
+      "step": 1947
+    },
+    {
+      "epoch": 1.8091035764050163,
+      "grad_norm": 0.17849449852961385,
+      "learning_rate": 2.204819277108434e-05,
+      "loss": 0.3461,
+      "step": 1948
+    },
+    {
+      "epoch": 1.810032512772875,
+      "grad_norm": 0.1994999151338678,
+      "learning_rate": 2.2030981067125648e-05,
+      "loss": 0.3394,
+      "step": 1949
+    },
+    {
+      "epoch": 1.8109614491407338,
+      "grad_norm": 0.1907734719321139,
+      "learning_rate": 2.2013769363166956e-05,
+      "loss": 0.3503,
+      "step": 1950
+    },
+    {
+      "epoch": 1.8118903855085926,
+      "grad_norm": 0.19648653560232324,
+      "learning_rate": 2.1996557659208264e-05,
+      "loss": 0.3316,
+      "step": 1951
+    },
+    {
+      "epoch": 1.8128193218764515,
+      "grad_norm": 0.1896427151890389,
+      "learning_rate": 2.197934595524957e-05,
+      "loss": 0.3392,
+      "step": 1952
+    },
+    {
+      "epoch": 1.8137482582443103,
+      "grad_norm": 0.19027376064820928,
+      "learning_rate": 2.1962134251290878e-05,
+      "loss": 0.3537,
+      "step": 1953
+    },
+    {
+      "epoch": 1.8146771946121691,
+      "grad_norm": 0.1855808710590735,
+      "learning_rate": 2.1944922547332186e-05,
+      "loss": 0.3187,
+      "step": 1954
+    },
+    {
+      "epoch": 1.815606130980028,
+      "grad_norm": 0.1984498093903945,
+      "learning_rate": 2.1927710843373495e-05,
+      "loss": 0.3613,
+      "step": 1955
+    },
+    {
+      "epoch": 1.8165350673478868,
+      "grad_norm": 0.2242505101068666,
+      "learning_rate": 2.1910499139414803e-05,
+      "loss": 0.3266,
+      "step": 1956
+    },
+    {
+      "epoch": 1.8174640037157455,
+      "grad_norm": 0.18137823080536702,
+      "learning_rate": 2.189328743545611e-05,
+      "loss": 0.3399,
+      "step": 1957
+    },
+    {
+      "epoch": 1.8183929400836043,
+      "grad_norm": 0.22546513335231178,
+      "learning_rate": 2.187607573149742e-05,
+      "loss": 0.3301,
+      "step": 1958
+    },
+    {
+      "epoch": 1.819321876451463,
+      "grad_norm": 0.20216140374779656,
+      "learning_rate": 2.185886402753873e-05,
+      "loss": 0.3401,
+      "step": 1959
+    },
+    {
+      "epoch": 1.8202508128193218,
+      "grad_norm": 0.19247489072112667,
+      "learning_rate": 2.1841652323580037e-05,
+      "loss": 0.3451,
+      "step": 1960
+    },
+    {
+      "epoch": 1.8211797491871806,
+      "grad_norm": 0.217041904392123,
+      "learning_rate": 2.1824440619621345e-05,
+      "loss": 0.3609,
+      "step": 1961
+    },
+    {
+      "epoch": 1.8221086855550395,
+      "grad_norm": 0.20709511532226713,
+      "learning_rate": 2.1807228915662654e-05,
+      "loss": 0.3389,
+      "step": 1962
+    },
+    {
+      "epoch": 1.8230376219228983,
+      "grad_norm": 0.20249280813186266,
+      "learning_rate": 2.179001721170396e-05,
+      "loss": 0.3367,
+      "step": 1963
+    },
+    {
+      "epoch": 1.8239665582907572,
+      "grad_norm": 0.21759029061763033,
+      "learning_rate": 2.1772805507745267e-05,
+      "loss": 0.3532,
+      "step": 1964
+    },
+    {
+      "epoch": 1.824895494658616,
+      "grad_norm": 0.19698131495510754,
+      "learning_rate": 2.1755593803786575e-05,
+      "loss": 0.3453,
+      "step": 1965
+    },
+    {
+      "epoch": 1.8258244310264748,
+      "grad_norm": 0.222032209900618,
+      "learning_rate": 2.1738382099827884e-05,
+      "loss": 0.3358,
+      "step": 1966
+    },
+    {
+      "epoch": 1.8267533673943335,
+      "grad_norm": 0.2098001700888135,
+      "learning_rate": 2.1721170395869192e-05,
+      "loss": 0.342,
+      "step": 1967
+    },
+    {
+      "epoch": 1.8276823037621923,
+      "grad_norm": 0.20422523735242673,
+      "learning_rate": 2.17039586919105e-05,
+      "loss": 0.321,
+      "step": 1968
+    },
+    {
+      "epoch": 1.828611240130051,
+      "grad_norm": 0.2012686527729587,
+      "learning_rate": 2.168674698795181e-05,
+      "loss": 0.3442,
+      "step": 1969
+    },
+    {
+      "epoch": 1.8295401764979098,
+      "grad_norm": 0.21283216710918032,
+      "learning_rate": 2.1669535283993117e-05,
+      "loss": 0.3588,
+      "step": 1970
+    },
+    {
+      "epoch": 1.8304691128657686,
+      "grad_norm": 0.19417470172069212,
+      "learning_rate": 2.1652323580034426e-05,
+      "loss": 0.3283,
+      "step": 1971
+    },
+    {
+      "epoch": 1.8313980492336275,
+      "grad_norm": 0.20490789352440664,
+      "learning_rate": 2.1635111876075734e-05,
+      "loss": 0.3637,
+      "step": 1972
+    },
+    {
+      "epoch": 1.8323269856014863,
+      "grad_norm": 0.2069872797582916,
+      "learning_rate": 2.161790017211704e-05,
+      "loss": 0.3346,
+      "step": 1973
+    },
+    {
+      "epoch": 1.8332559219693452,
+      "grad_norm": 0.18465378730661033,
+      "learning_rate": 2.1600688468158348e-05,
+      "loss": 0.3514,
+      "step": 1974
+    },
+    {
+      "epoch": 1.834184858337204,
+      "grad_norm": 0.1772459699016406,
+      "learning_rate": 2.1583476764199656e-05,
+      "loss": 0.3446,
+      "step": 1975
+    },
+    {
+      "epoch": 1.8351137947050629,
+      "grad_norm": 0.2246086295913219,
+      "learning_rate": 2.1566265060240964e-05,
+      "loss": 0.3426,
+      "step": 1976
+    },
+    {
+      "epoch": 1.8360427310729215,
+      "grad_norm": 0.19999809607634841,
+      "learning_rate": 2.1549053356282273e-05,
+      "loss": 0.3266,
+      "step": 1977
+    },
+    {
+      "epoch": 1.8369716674407803,
+      "grad_norm": 0.18937275360959321,
+      "learning_rate": 2.153184165232358e-05,
+      "loss": 0.3362,
+      "step": 1978
+    },
+    {
+      "epoch": 1.837900603808639,
+      "grad_norm": 0.1894381875679942,
+      "learning_rate": 2.151462994836489e-05,
+      "loss": 0.3555,
+      "step": 1979
+    },
+    {
+      "epoch": 1.8388295401764978,
+      "grad_norm": 0.2163255652366772,
+      "learning_rate": 2.1497418244406198e-05,
+      "loss": 0.3509,
+      "step": 1980
+    },
+    {
+      "epoch": 1.8397584765443566,
+      "grad_norm": 0.21158701402876343,
+      "learning_rate": 2.1480206540447506e-05,
+      "loss": 0.3419,
+      "step": 1981
+    },
+    {
+      "epoch": 1.8406874129122155,
+      "grad_norm": 0.1832944886654045,
+      "learning_rate": 2.1462994836488815e-05,
+      "loss": 0.3519,
+      "step": 1982
+    },
+    {
+      "epoch": 1.8416163492800743,
+      "grad_norm": 0.20724791335919457,
+      "learning_rate": 2.1445783132530123e-05,
+      "loss": 0.3332,
+      "step": 1983
+    },
+    {
+      "epoch": 1.8425452856479332,
+      "grad_norm": 0.2026632632941415,
+      "learning_rate": 2.1428571428571428e-05,
+      "loss": 0.3535,
+      "step": 1984
+    },
+    {
+      "epoch": 1.843474222015792,
+      "grad_norm": 0.18716762184495547,
+      "learning_rate": 2.1411359724612737e-05,
+      "loss": 0.3388,
+      "step": 1985
+    },
+    {
+      "epoch": 1.8444031583836509,
+      "grad_norm": 0.19811432373555643,
+      "learning_rate": 2.1394148020654045e-05,
+      "loss": 0.338,
+      "step": 1986
+    },
+    {
+      "epoch": 1.8453320947515095,
+      "grad_norm": 0.2078115035058908,
+      "learning_rate": 2.1376936316695353e-05,
+      "loss": 0.3175,
+      "step": 1987
+    },
+    {
+      "epoch": 1.8462610311193683,
+      "grad_norm": 0.19475185295752664,
+      "learning_rate": 2.1359724612736662e-05,
+      "loss": 0.3284,
+      "step": 1988
+    },
+    {
+      "epoch": 1.847189967487227,
+      "grad_norm": 0.19643704859931127,
+      "learning_rate": 2.134251290877797e-05,
+      "loss": 0.3579,
+      "step": 1989
+    },
+    {
+      "epoch": 1.8481189038550858,
+      "grad_norm": 0.19642321023564513,
+      "learning_rate": 2.132530120481928e-05,
+      "loss": 0.3215,
+      "step": 1990
+    },
+    {
+      "epoch": 1.8490478402229447,
+      "grad_norm": 0.18968037438997556,
+      "learning_rate": 2.1308089500860587e-05,
+      "loss": 0.3206,
+      "step": 1991
+    },
+    {
+      "epoch": 1.8499767765908035,
+      "grad_norm": 0.19787881644954247,
+      "learning_rate": 2.1290877796901895e-05,
+      "loss": 0.3362,
+      "step": 1992
+    },
+    {
+      "epoch": 1.8509057129586624,
+      "grad_norm": 0.19024488827058927,
+      "learning_rate": 2.1273666092943204e-05,
+      "loss": 0.3398,
+      "step": 1993
+    },
+    {
+      "epoch": 1.8518346493265212,
+      "grad_norm": 0.19182808398964474,
+      "learning_rate": 2.125645438898451e-05,
+      "loss": 0.3428,
+      "step": 1994
+    },
+    {
+      "epoch": 1.85276358569438,
+      "grad_norm": 0.21009107426495674,
+      "learning_rate": 2.1239242685025817e-05,
+      "loss": 0.3434,
+      "step": 1995
+    },
+    {
+      "epoch": 1.853692522062239,
+      "grad_norm": 0.1914880152985732,
+      "learning_rate": 2.1222030981067126e-05,
+      "loss": 0.327,
+      "step": 1996
+    },
+    {
+      "epoch": 1.8546214584300975,
+      "grad_norm": 0.1913970478509932,
+      "learning_rate": 2.1204819277108434e-05,
+      "loss": 0.34,
+      "step": 1997
+    },
+    {
+      "epoch": 1.8555503947979564,
+      "grad_norm": 0.2155756840317901,
+      "learning_rate": 2.1187607573149742e-05,
+      "loss": 0.3245,
+      "step": 1998
+    },
+    {
+      "epoch": 1.856479331165815,
+      "grad_norm": 0.18445074398921807,
+      "learning_rate": 2.117039586919105e-05,
+      "loss": 0.3429,
+      "step": 1999
+    },
+    {
+      "epoch": 1.8574082675336738,
+      "grad_norm": 0.18560311928741094,
+      "learning_rate": 2.115318416523236e-05,
+      "loss": 0.328,
+      "step": 2000
+    },
+    {
+      "epoch": 1.8583372039015327,
+      "grad_norm": 0.17532760828844168,
+      "learning_rate": 2.1135972461273668e-05,
+      "loss": 0.3499,
+      "step": 2001
+    },
+    {
+      "epoch": 1.8592661402693915,
+      "grad_norm": 0.18320848398237594,
+      "learning_rate": 2.1118760757314976e-05,
+      "loss": 0.3371,
+      "step": 2002
+    },
+    {
+      "epoch": 1.8601950766372504,
+      "grad_norm": 0.18645021050922914,
+      "learning_rate": 2.1101549053356284e-05,
+      "loss": 0.3712,
+      "step": 2003
+    },
+    {
+      "epoch": 1.8611240130051092,
+      "grad_norm": 0.1930981725638318,
+      "learning_rate": 2.1084337349397593e-05,
+      "loss": 0.3631,
+      "step": 2004
+    },
+    {
+      "epoch": 1.862052949372968,
+      "grad_norm": 0.23294279233206924,
+      "learning_rate": 2.1067125645438898e-05,
+      "loss": 0.3627,
+      "step": 2005
+    },
+    {
+      "epoch": 1.862981885740827,
+      "grad_norm": 0.22252132694320956,
+      "learning_rate": 2.1049913941480206e-05,
+      "loss": 0.3251,
+      "step": 2006
+    },
+    {
+      "epoch": 1.8639108221086855,
+      "grad_norm": 0.18325558308090584,
+      "learning_rate": 2.1032702237521515e-05,
+      "loss": 0.3516,
+      "step": 2007
+    },
+    {
+      "epoch": 1.8648397584765444,
+      "grad_norm": 0.18803938364991366,
+      "learning_rate": 2.1015490533562823e-05,
+      "loss": 0.335,
+      "step": 2008
+    },
+    {
+      "epoch": 1.865768694844403,
+      "grad_norm": 0.21560136159819376,
+      "learning_rate": 2.099827882960413e-05,
+      "loss": 0.3502,
+      "step": 2009
+    },
+    {
+      "epoch": 1.8666976312122618,
+      "grad_norm": 0.2075112616048167,
+      "learning_rate": 2.098106712564544e-05,
+      "loss": 0.3448,
+      "step": 2010
+    },
+    {
+      "epoch": 1.8676265675801207,
+      "grad_norm": 0.1794063031438016,
+      "learning_rate": 2.0963855421686748e-05,
+      "loss": 0.3589,
+      "step": 2011
+    },
+    {
+      "epoch": 1.8685555039479795,
+      "grad_norm": 0.2088734523300138,
+      "learning_rate": 2.0946643717728057e-05,
+      "loss": 0.3401,
+      "step": 2012
+    },
+    {
+      "epoch": 1.8694844403158384,
+      "grad_norm": 0.20514199980300488,
+      "learning_rate": 2.0929432013769365e-05,
+      "loss": 0.3562,
+      "step": 2013
+    },
+    {
+      "epoch": 1.8704133766836972,
+      "grad_norm": 0.1888327050014192,
+      "learning_rate": 2.0912220309810673e-05,
+      "loss": 0.3104,
+      "step": 2014
+    },
+    {
+      "epoch": 1.871342313051556,
+      "grad_norm": 0.18729822949477454,
+      "learning_rate": 2.0895008605851982e-05,
+      "loss": 0.3547,
+      "step": 2015
+    },
+    {
+      "epoch": 1.872271249419415,
+      "grad_norm": 0.22612417730319762,
+      "learning_rate": 2.0877796901893287e-05,
+      "loss": 0.3499,
+      "step": 2016
+    },
+    {
+      "epoch": 1.8732001857872735,
+      "grad_norm": 0.19668595579111908,
+      "learning_rate": 2.0860585197934595e-05,
+      "loss": 0.3296,
+      "step": 2017
+    },
+    {
+      "epoch": 1.8741291221551324,
+      "grad_norm": 0.18571506463595566,
+      "learning_rate": 2.0843373493975904e-05,
+      "loss": 0.3312,
+      "step": 2018
+    },
+    {
+      "epoch": 1.875058058522991,
+      "grad_norm": 0.18777769470240502,
+      "learning_rate": 2.0826161790017212e-05,
+      "loss": 0.3234,
+      "step": 2019
+    },
+    {
+      "epoch": 1.8759869948908499,
+      "grad_norm": 0.18583762914371052,
+      "learning_rate": 2.080895008605852e-05,
+      "loss": 0.3271,
+      "step": 2020
+    },
+    {
+      "epoch": 1.8769159312587087,
+      "grad_norm": 0.1960771159495725,
+      "learning_rate": 2.079173838209983e-05,
+      "loss": 0.3581,
+      "step": 2021
+    },
+    {
+      "epoch": 1.8778448676265675,
+      "grad_norm": 0.19599306930659696,
+      "learning_rate": 2.0774526678141137e-05,
+      "loss": 0.357,
+      "step": 2022
+    },
+    {
+      "epoch": 1.8787738039944264,
+      "grad_norm": 0.1951564097933878,
+      "learning_rate": 2.0757314974182446e-05,
+      "loss": 0.3591,
+      "step": 2023
+    },
+    {
+      "epoch": 1.8797027403622852,
+      "grad_norm": 0.19235765747681716,
+      "learning_rate": 2.0740103270223754e-05,
+      "loss": 0.3609,
+      "step": 2024
+    },
+    {
+      "epoch": 1.880631676730144,
+      "grad_norm": 0.19639988332184477,
+      "learning_rate": 2.0722891566265062e-05,
+      "loss": 0.3397,
+      "step": 2025
+    },
+    {
+      "epoch": 1.881560613098003,
+      "grad_norm": 0.19618499182075172,
+      "learning_rate": 2.0705679862306367e-05,
+      "loss": 0.3319,
+      "step": 2026
+    },
+    {
+      "epoch": 1.8824895494658616,
+      "grad_norm": 0.18391264325568676,
+      "learning_rate": 2.0688468158347676e-05,
+      "loss": 0.3668,
+      "step": 2027
+    },
+    {
+      "epoch": 1.8834184858337204,
+      "grad_norm": 0.1865315427514266,
+      "learning_rate": 2.0671256454388984e-05,
+      "loss": 0.357,
+      "step": 2028
+    },
+    {
+      "epoch": 1.884347422201579,
+      "grad_norm": 0.1965123947388705,
+      "learning_rate": 2.0654044750430293e-05,
+      "loss": 0.3317,
+      "step": 2029
+    },
+    {
+      "epoch": 1.8852763585694379,
+      "grad_norm": 0.19090795755847184,
+      "learning_rate": 2.06368330464716e-05,
+      "loss": 0.3409,
+      "step": 2030
+    },
+    {
+      "epoch": 1.8862052949372967,
+      "grad_norm": 0.2011381053988122,
+      "learning_rate": 2.061962134251291e-05,
+      "loss": 0.3579,
+      "step": 2031
+    },
+    {
+      "epoch": 1.8871342313051556,
+      "grad_norm": 0.17714274475948777,
+      "learning_rate": 2.0602409638554218e-05,
+      "loss": 0.3486,
+      "step": 2032
+    },
+    {
+      "epoch": 1.8880631676730144,
+      "grad_norm": 0.20684528947484848,
+      "learning_rate": 2.0585197934595526e-05,
+      "loss": 0.3315,
+      "step": 2033
+    },
+    {
+      "epoch": 1.8889921040408733,
+      "grad_norm": 0.20594746537045297,
+      "learning_rate": 2.0567986230636835e-05,
+      "loss": 0.3518,
+      "step": 2034
+    },
+    {
+      "epoch": 1.889921040408732,
+      "grad_norm": 0.20843330387434464,
+      "learning_rate": 2.0550774526678143e-05,
+      "loss": 0.3377,
+      "step": 2035
+    },
+    {
+      "epoch": 1.890849976776591,
+      "grad_norm": 0.19119948120789587,
+      "learning_rate": 2.053356282271945e-05,
+      "loss": 0.3508,
+      "step": 2036
+    },
+    {
+      "epoch": 1.8917789131444496,
+      "grad_norm": 0.20499362964321996,
+      "learning_rate": 2.0516351118760756e-05,
+      "loss": 0.3386,
+      "step": 2037
+    },
+    {
+      "epoch": 1.8927078495123084,
+      "grad_norm": 0.20951246709031993,
+      "learning_rate": 2.0499139414802065e-05,
+      "loss": 0.3454,
+      "step": 2038
+    },
+    {
+      "epoch": 1.8936367858801673,
+      "grad_norm": 0.1933469877666157,
+      "learning_rate": 2.0481927710843373e-05,
+      "loss": 0.3498,
+      "step": 2039
+    },
+    {
+      "epoch": 1.8945657222480259,
+      "grad_norm": 0.2033030466421948,
+      "learning_rate": 2.046471600688468e-05,
+      "loss": 0.3496,
+      "step": 2040
+    },
+    {
+      "epoch": 1.8954946586158847,
+      "grad_norm": 0.21435793489097013,
+      "learning_rate": 2.044750430292599e-05,
+      "loss": 0.35,
+      "step": 2041
+    },
+    {
+      "epoch": 1.8964235949837436,
+      "grad_norm": 0.21062698065040608,
+      "learning_rate": 2.04302925989673e-05,
+      "loss": 0.3598,
+      "step": 2042
+    },
+    {
+      "epoch": 1.8973525313516024,
+      "grad_norm": 0.19925766321798463,
+      "learning_rate": 2.0413080895008607e-05,
+      "loss": 0.3301,
+      "step": 2043
+    },
+    {
+      "epoch": 1.8982814677194613,
+      "grad_norm": 0.19489181490424323,
+      "learning_rate": 2.0395869191049915e-05,
+      "loss": 0.325,
+      "step": 2044
+    },
+    {
+      "epoch": 1.8992104040873201,
+      "grad_norm": 0.20174771996042817,
+      "learning_rate": 2.0378657487091224e-05,
+      "loss": 0.3308,
+      "step": 2045
+    },
+    {
+      "epoch": 1.900139340455179,
+      "grad_norm": 0.20408788253809196,
+      "learning_rate": 2.0361445783132532e-05,
+      "loss": 0.3211,
+      "step": 2046
+    },
+    {
+      "epoch": 1.9010682768230376,
+      "grad_norm": 0.1878516530540329,
+      "learning_rate": 2.0344234079173837e-05,
+      "loss": 0.3414,
+      "step": 2047
+    },
+    {
+      "epoch": 1.9019972131908964,
+      "grad_norm": 0.21257975400270182,
+      "learning_rate": 2.0327022375215145e-05,
+      "loss": 0.3781,
+      "step": 2048
+    },
+    {
+      "epoch": 1.9029261495587553,
+      "grad_norm": 0.20595091660025142,
+      "learning_rate": 2.0309810671256454e-05,
+      "loss": 0.3373,
+      "step": 2049
+    },
+    {
+      "epoch": 1.903855085926614,
+      "grad_norm": 0.19877200066200787,
+      "learning_rate": 2.0292598967297762e-05,
+      "loss": 0.3417,
+      "step": 2050
+    },
+    {
+      "epoch": 1.9047840222944727,
+      "grad_norm": 0.20104969915961177,
+      "learning_rate": 2.027538726333907e-05,
+      "loss": 0.3467,
+      "step": 2051
+    },
+    {
+      "epoch": 1.9057129586623316,
+      "grad_norm": 0.2327660603720691,
+      "learning_rate": 2.025817555938038e-05,
+      "loss": 0.3568,
+      "step": 2052
+    },
+    {
+      "epoch": 1.9066418950301904,
+      "grad_norm": 0.18750059187569587,
+      "learning_rate": 2.0240963855421687e-05,
+      "loss": 0.3427,
+      "step": 2053
+    },
+    {
+      "epoch": 1.9075708313980493,
+      "grad_norm": 0.18529718656829258,
+      "learning_rate": 2.0223752151462996e-05,
+      "loss": 0.3657,
+      "step": 2054
+    },
+    {
+      "epoch": 1.9084997677659081,
+      "grad_norm": 0.21378367120455413,
+      "learning_rate": 2.0206540447504304e-05,
+      "loss": 0.3448,
+      "step": 2055
+    },
+    {
+      "epoch": 1.909428704133767,
+      "grad_norm": 0.19816996642108167,
+      "learning_rate": 2.0189328743545613e-05,
+      "loss": 0.3379,
+      "step": 2056
+    },
+    {
+      "epoch": 1.9103576405016256,
+      "grad_norm": 0.17952569303828486,
+      "learning_rate": 2.017211703958692e-05,
+      "loss": 0.3495,
+      "step": 2057
+    },
+    {
+      "epoch": 1.9112865768694844,
+      "grad_norm": 0.1887115752212257,
+      "learning_rate": 2.0154905335628226e-05,
+      "loss": 0.3528,
+      "step": 2058
+    },
+    {
+      "epoch": 1.9122155132373433,
+      "grad_norm": 0.19795487285379446,
+      "learning_rate": 2.0137693631669534e-05,
+      "loss": 0.312,
+      "step": 2059
+    },
+    {
+      "epoch": 1.913144449605202,
+      "grad_norm": 0.1913042487899833,
+      "learning_rate": 2.0120481927710843e-05,
+      "loss": 0.3474,
+      "step": 2060
+    },
+    {
+      "epoch": 1.9140733859730608,
+      "grad_norm": 0.18702333471919602,
+      "learning_rate": 2.010327022375215e-05,
+      "loss": 0.3313,
+      "step": 2061
+    },
+    {
+      "epoch": 1.9150023223409196,
+      "grad_norm": 0.18216724495400727,
+      "learning_rate": 2.008605851979346e-05,
+      "loss": 0.3273,
+      "step": 2062
+    },
+    {
+      "epoch": 1.9159312587087785,
+      "grad_norm": 0.1870733559834162,
+      "learning_rate": 2.0068846815834768e-05,
+      "loss": 0.328,
+      "step": 2063
+    },
+    {
+      "epoch": 1.9168601950766373,
+      "grad_norm": 0.20069109494783774,
+      "learning_rate": 2.0051635111876076e-05,
+      "loss": 0.3468,
+      "step": 2064
+    },
+    {
+      "epoch": 1.9177891314444961,
+      "grad_norm": 0.1788738115806139,
+      "learning_rate": 2.0034423407917385e-05,
+      "loss": 0.3186,
+      "step": 2065
+    },
+    {
+      "epoch": 1.918718067812355,
+      "grad_norm": 0.18264576767730586,
+      "learning_rate": 2.0017211703958693e-05,
+      "loss": 0.3399,
+      "step": 2066
+    },
+    {
+      "epoch": 1.9196470041802136,
+      "grad_norm": 0.1798033926304608,
+      "learning_rate": 2e-05,
+      "loss": 0.3475,
+      "step": 2067
+    },
+    {
+      "epoch": 1.9205759405480725,
+      "grad_norm": 0.20622934405855825,
+      "learning_rate": 1.998278829604131e-05,
+      "loss": 0.3448,
+      "step": 2068
+    },
+    {
+      "epoch": 1.9215048769159313,
+      "grad_norm": 0.20482612442847245,
+      "learning_rate": 1.9965576592082615e-05,
+      "loss": 0.3479,
+      "step": 2069
+    },
+    {
+      "epoch": 1.92243381328379,
+      "grad_norm": 0.19506902393707024,
+      "learning_rate": 1.9948364888123923e-05,
+      "loss": 0.3428,
+      "step": 2070
+    },
+    {
+      "epoch": 1.9233627496516488,
+      "grad_norm": 0.20966049670336326,
+      "learning_rate": 1.9931153184165232e-05,
+      "loss": 0.344,
+      "step": 2071
+    },
+    {
+      "epoch": 1.9242916860195076,
+      "grad_norm": 0.1865501634784815,
+      "learning_rate": 1.991394148020654e-05,
+      "loss": 0.3616,
+      "step": 2072
+    },
+    {
+      "epoch": 1.9252206223873665,
+      "grad_norm": 0.196751064921281,
+      "learning_rate": 1.989672977624785e-05,
+      "loss": 0.3318,
+      "step": 2073
+    },
+    {
+      "epoch": 1.9261495587552253,
+      "grad_norm": 0.1804349938052275,
+      "learning_rate": 1.9879518072289157e-05,
+      "loss": 0.3416,
+      "step": 2074
+    },
+    {
+      "epoch": 1.9270784951230842,
+      "grad_norm": 0.18106072316653846,
+      "learning_rate": 1.9862306368330465e-05,
+      "loss": 0.3336,
+      "step": 2075
+    },
+    {
+      "epoch": 1.928007431490943,
+      "grad_norm": 0.18984032415869853,
+      "learning_rate": 1.9845094664371774e-05,
+      "loss": 0.35,
+      "step": 2076
+    },
+    {
+      "epoch": 1.9289363678588016,
+      "grad_norm": 0.21925027722818818,
+      "learning_rate": 1.9827882960413082e-05,
+      "loss": 0.3428,
+      "step": 2077
+    },
+    {
+      "epoch": 1.9298653042266605,
+      "grad_norm": 0.17538934019640326,
+      "learning_rate": 1.981067125645439e-05,
+      "loss": 0.3489,
+      "step": 2078
+    },
+    {
+      "epoch": 1.9307942405945193,
+      "grad_norm": 0.21247914753683844,
+      "learning_rate": 1.9793459552495696e-05,
+      "loss": 0.3436,
+      "step": 2079
+    },
+    {
+      "epoch": 1.931723176962378,
+      "grad_norm": 0.20170277926352845,
+      "learning_rate": 1.9776247848537004e-05,
+      "loss": 0.3665,
+      "step": 2080
+    },
+    {
+      "epoch": 1.9326521133302368,
+      "grad_norm": 0.21489372804535314,
+      "learning_rate": 1.9759036144578312e-05,
+      "loss": 0.3271,
+      "step": 2081
+    },
+    {
+      "epoch": 1.9335810496980956,
+      "grad_norm": 0.19404730364157935,
+      "learning_rate": 1.974182444061962e-05,
+      "loss": 0.3395,
+      "step": 2082
+    },
+    {
+      "epoch": 1.9345099860659545,
+      "grad_norm": 0.1768311883365791,
+      "learning_rate": 1.972461273666093e-05,
+      "loss": 0.3304,
+      "step": 2083
+    },
+    {
+      "epoch": 1.9354389224338133,
+      "grad_norm": 0.19513776903048077,
+      "learning_rate": 1.9707401032702238e-05,
+      "loss": 0.3355,
+      "step": 2084
+    },
+    {
+      "epoch": 1.9363678588016722,
+      "grad_norm": 0.20552603770095754,
+      "learning_rate": 1.9690189328743546e-05,
+      "loss": 0.3586,
+      "step": 2085
+    },
+    {
+      "epoch": 1.937296795169531,
+      "grad_norm": 0.20606350494140052,
+      "learning_rate": 1.9672977624784854e-05,
+      "loss": 0.341,
+      "step": 2086
+    },
+    {
+      "epoch": 1.9382257315373896,
+      "grad_norm": 0.21117517653181553,
+      "learning_rate": 1.9655765920826163e-05,
+      "loss": 0.3556,
+      "step": 2087
+    },
+    {
+      "epoch": 1.9391546679052485,
+      "grad_norm": 0.19889606660583597,
+      "learning_rate": 1.963855421686747e-05,
+      "loss": 0.3754,
+      "step": 2088
+    },
+    {
+      "epoch": 1.9400836042731073,
+      "grad_norm": 0.2315272588969166,
+      "learning_rate": 1.962134251290878e-05,
+      "loss": 0.3625,
+      "step": 2089
+    },
+    {
+      "epoch": 1.941012540640966,
+      "grad_norm": 0.1801392600297905,
+      "learning_rate": 1.9604130808950085e-05,
+      "loss": 0.3678,
+      "step": 2090
+    },
+    {
+      "epoch": 1.9419414770088248,
+      "grad_norm": 0.20819332360704426,
+      "learning_rate": 1.9586919104991393e-05,
+      "loss": 0.3413,
+      "step": 2091
+    },
+    {
+      "epoch": 1.9428704133766836,
+      "grad_norm": 0.20855084760144948,
+      "learning_rate": 1.95697074010327e-05,
+      "loss": 0.3479,
+      "step": 2092
+    },
+    {
+      "epoch": 1.9437993497445425,
+      "grad_norm": 0.1934591269154208,
+      "learning_rate": 1.955249569707401e-05,
+      "loss": 0.3292,
+      "step": 2093
+    },
+    {
+      "epoch": 1.9447282861124013,
+      "grad_norm": 0.19888631185262237,
+      "learning_rate": 1.9535283993115318e-05,
+      "loss": 0.3517,
+      "step": 2094
+    },
+    {
+      "epoch": 1.9456572224802602,
+      "grad_norm": 0.23696777928124235,
+      "learning_rate": 1.9518072289156627e-05,
+      "loss": 0.3603,
+      "step": 2095
+    },
+    {
+      "epoch": 1.946586158848119,
+      "grad_norm": 0.1930501835027152,
+      "learning_rate": 1.9500860585197935e-05,
+      "loss": 0.3511,
+      "step": 2096
+    },
+    {
+      "epoch": 1.9475150952159777,
+      "grad_norm": 0.1847048438548015,
+      "learning_rate": 1.9483648881239243e-05,
+      "loss": 0.3256,
+      "step": 2097
+    },
+    {
+      "epoch": 1.9484440315838365,
+      "grad_norm": 0.17904191429627417,
+      "learning_rate": 1.9466437177280552e-05,
+      "loss": 0.3526,
+      "step": 2098
+    },
+    {
+      "epoch": 1.9493729679516953,
+      "grad_norm": 0.21564625082297997,
+      "learning_rate": 1.944922547332186e-05,
+      "loss": 0.3336,
+      "step": 2099
+    },
+    {
+      "epoch": 1.950301904319554,
+      "grad_norm": 0.18093745627216662,
+      "learning_rate": 1.9432013769363165e-05,
+      "loss": 0.337,
+      "step": 2100
+    },
+    {
+      "epoch": 1.9512308406874128,
+      "grad_norm": 0.20861506817234635,
+      "learning_rate": 1.9414802065404474e-05,
+      "loss": 0.3545,
+      "step": 2101
+    },
+    {
+      "epoch": 1.9521597770552717,
+      "grad_norm": 0.18651110392988504,
+      "learning_rate": 1.9397590361445782e-05,
+      "loss": 0.3382,
+      "step": 2102
+    },
+    {
+      "epoch": 1.9530887134231305,
+      "grad_norm": 0.19498741795933872,
+      "learning_rate": 1.938037865748709e-05,
+      "loss": 0.3196,
+      "step": 2103
+    },
+    {
+      "epoch": 1.9540176497909894,
+      "grad_norm": 0.19938426190865807,
+      "learning_rate": 1.93631669535284e-05,
+      "loss": 0.3439,
+      "step": 2104
+    },
+    {
+      "epoch": 1.9549465861588482,
+      "grad_norm": 0.20167566769499962,
+      "learning_rate": 1.9345955249569707e-05,
+      "loss": 0.3422,
+      "step": 2105
+    },
+    {
+      "epoch": 1.955875522526707,
+      "grad_norm": 0.1858484130472836,
+      "learning_rate": 1.9328743545611016e-05,
+      "loss": 0.3334,
+      "step": 2106
+    },
+    {
+      "epoch": 1.9568044588945657,
+      "grad_norm": 0.19150012147084414,
+      "learning_rate": 1.9311531841652324e-05,
+      "loss": 0.3199,
+      "step": 2107
+    },
+    {
+      "epoch": 1.9577333952624245,
+      "grad_norm": 0.20387383733567602,
+      "learning_rate": 1.9294320137693632e-05,
+      "loss": 0.3531,
+      "step": 2108
+    },
+    {
+      "epoch": 1.9586623316302834,
+      "grad_norm": 0.18764227954582366,
+      "learning_rate": 1.927710843373494e-05,
+      "loss": 0.3485,
+      "step": 2109
+    },
+    {
+      "epoch": 1.959591267998142,
+      "grad_norm": 0.19184067561980478,
+      "learning_rate": 1.925989672977625e-05,
+      "loss": 0.3375,
+      "step": 2110
+    },
+    {
+      "epoch": 1.9605202043660008,
+      "grad_norm": 0.18361672644385202,
+      "learning_rate": 1.9242685025817554e-05,
+      "loss": 0.3411,
+      "step": 2111
+    },
+    {
+      "epoch": 1.9614491407338597,
+      "grad_norm": 0.19465647573588937,
+      "learning_rate": 1.9225473321858863e-05,
+      "loss": 0.3273,
+      "step": 2112
+    },
+    {
+      "epoch": 1.9623780771017185,
+      "grad_norm": 0.18345387692416829,
+      "learning_rate": 1.920826161790017e-05,
+      "loss": 0.3534,
+      "step": 2113
+    },
+    {
+      "epoch": 1.9633070134695774,
+      "grad_norm": 0.18440352667913198,
+      "learning_rate": 1.919104991394148e-05,
+      "loss": 0.3338,
+      "step": 2114
+    },
+    {
+      "epoch": 1.9642359498374362,
+      "grad_norm": 0.1905629308788521,
+      "learning_rate": 1.9173838209982788e-05,
+      "loss": 0.348,
+      "step": 2115
+    },
+    {
+      "epoch": 1.965164886205295,
+      "grad_norm": 0.17738331253996667,
+      "learning_rate": 1.9156626506024096e-05,
+      "loss": 0.338,
+      "step": 2116
+    },
+    {
+      "epoch": 1.966093822573154,
+      "grad_norm": 0.19317990178767253,
+      "learning_rate": 1.9139414802065405e-05,
+      "loss": 0.3463,
+      "step": 2117
+    },
+    {
+      "epoch": 1.9670227589410125,
+      "grad_norm": 0.19118447547256318,
+      "learning_rate": 1.9122203098106713e-05,
+      "loss": 0.3205,
+      "step": 2118
+    },
+    {
+      "epoch": 1.9679516953088714,
+      "grad_norm": 0.1941495864820665,
+      "learning_rate": 1.910499139414802e-05,
+      "loss": 0.3539,
+      "step": 2119
+    },
+    {
+      "epoch": 1.96888063167673,
+      "grad_norm": 0.19760267671496076,
+      "learning_rate": 1.908777969018933e-05,
+      "loss": 0.3464,
+      "step": 2120
+    },
+    {
+      "epoch": 1.9698095680445888,
+      "grad_norm": 0.20254146860119207,
+      "learning_rate": 1.9070567986230638e-05,
+      "loss": 0.3499,
+      "step": 2121
+    },
+    {
+      "epoch": 1.9707385044124477,
+      "grad_norm": 0.1843345665414209,
+      "learning_rate": 1.9053356282271943e-05,
+      "loss": 0.3208,
+      "step": 2122
+    },
+    {
+      "epoch": 1.9716674407803065,
+      "grad_norm": 0.18931583414329914,
+      "learning_rate": 1.9036144578313252e-05,
+      "loss": 0.3314,
+      "step": 2123
+    },
+    {
+      "epoch": 1.9725963771481654,
+      "grad_norm": 0.17544270577499563,
+      "learning_rate": 1.901893287435456e-05,
+      "loss": 0.3287,
+      "step": 2124
+    },
+    {
+      "epoch": 1.9735253135160242,
+      "grad_norm": 0.19688051277717727,
+      "learning_rate": 1.900172117039587e-05,
+      "loss": 0.3522,
+      "step": 2125
+    },
+    {
+      "epoch": 1.974454249883883,
+      "grad_norm": 0.19873447267182606,
+      "learning_rate": 1.8984509466437177e-05,
+      "loss": 0.3312,
+      "step": 2126
+    },
+    {
+      "epoch": 1.975383186251742,
+      "grad_norm": 0.19087379498613669,
+      "learning_rate": 1.8967297762478485e-05,
+      "loss": 0.3676,
+      "step": 2127
+    },
+    {
+      "epoch": 1.9763121226196005,
+      "grad_norm": 0.19614887292658442,
+      "learning_rate": 1.8950086058519794e-05,
+      "loss": 0.3592,
+      "step": 2128
+    },
+    {
+      "epoch": 1.9772410589874594,
+      "grad_norm": 0.20900078157895577,
+      "learning_rate": 1.8932874354561102e-05,
+      "loss": 0.3431,
+      "step": 2129
+    },
+    {
+      "epoch": 1.978169995355318,
+      "grad_norm": 0.19572251469376092,
+      "learning_rate": 1.891566265060241e-05,
+      "loss": 0.3266,
+      "step": 2130
+    },
+    {
+      "epoch": 1.9790989317231769,
+      "grad_norm": 0.17443311056477603,
+      "learning_rate": 1.889845094664372e-05,
+      "loss": 0.3338,
+      "step": 2131
+    },
+    {
+      "epoch": 1.9800278680910357,
+      "grad_norm": 0.20884670381056047,
+      "learning_rate": 1.8881239242685024e-05,
+      "loss": 0.3434,
+      "step": 2132
+    },
+    {
+      "epoch": 1.9809568044588945,
+      "grad_norm": 0.19322917648542368,
+      "learning_rate": 1.8864027538726332e-05,
+      "loss": 0.3201,
+      "step": 2133
+    },
+    {
+      "epoch": 1.9818857408267534,
+      "grad_norm": 0.18161054306051905,
+      "learning_rate": 1.884681583476764e-05,
+      "loss": 0.3369,
+      "step": 2134
+    },
+    {
+      "epoch": 1.9828146771946122,
+      "grad_norm": 0.19995407241341367,
+      "learning_rate": 1.882960413080895e-05,
+      "loss": 0.3286,
+      "step": 2135
+    },
+    {
+      "epoch": 1.983743613562471,
+      "grad_norm": 0.1989383783308988,
+      "learning_rate": 1.8812392426850258e-05,
+      "loss": 0.3627,
+      "step": 2136
+    },
+    {
+      "epoch": 1.98467254993033,
+      "grad_norm": 0.20119453420571734,
+      "learning_rate": 1.8795180722891566e-05,
+      "loss": 0.3256,
+      "step": 2137
+    },
+    {
+      "epoch": 1.9856014862981886,
+      "grad_norm": 0.2182021666836405,
+      "learning_rate": 1.8777969018932874e-05,
+      "loss": 0.3461,
+      "step": 2138
+    },
+    {
+      "epoch": 1.9865304226660474,
+      "grad_norm": 0.2102127576334096,
+      "learning_rate": 1.8760757314974183e-05,
+      "loss": 0.3416,
+      "step": 2139
+    },
+    {
+      "epoch": 1.987459359033906,
+      "grad_norm": 0.18281161210727,
+      "learning_rate": 1.874354561101549e-05,
+      "loss": 0.3288,
+      "step": 2140
+    },
+    {
+      "epoch": 1.9883882954017649,
+      "grad_norm": 0.21447362464818037,
+      "learning_rate": 1.87263339070568e-05,
+      "loss": 0.3259,
+      "step": 2141
+    },
+    {
+      "epoch": 1.9893172317696237,
+      "grad_norm": 0.2115018838882039,
+      "learning_rate": 1.8709122203098108e-05,
+      "loss": 0.3453,
+      "step": 2142
+    },
+    {
+      "epoch": 1.9902461681374826,
+      "grad_norm": 0.1960898048792348,
+      "learning_rate": 1.8691910499139413e-05,
+      "loss": 0.3445,
+      "step": 2143
+    },
+    {
+      "epoch": 1.9911751045053414,
+      "grad_norm": 0.1930005268145748,
+      "learning_rate": 1.867469879518072e-05,
+      "loss": 0.3257,
+      "step": 2144
+    },
+    {
+      "epoch": 1.9921040408732003,
+      "grad_norm": 0.22568106040346528,
+      "learning_rate": 1.865748709122203e-05,
+      "loss": 0.3308,
+      "step": 2145
+    },
+    {
+      "epoch": 1.993032977241059,
+      "grad_norm": 0.1984125411749985,
+      "learning_rate": 1.8640275387263338e-05,
+      "loss": 0.3465,
+      "step": 2146
+    },
+    {
+      "epoch": 1.993961913608918,
+      "grad_norm": 0.19270028218557986,
+      "learning_rate": 1.8623063683304647e-05,
+      "loss": 0.3766,
+      "step": 2147
+    },
+    {
+      "epoch": 1.9948908499767766,
+      "grad_norm": 0.21961399982931906,
+      "learning_rate": 1.8605851979345955e-05,
+      "loss": 0.364,
+      "step": 2148
+    },
+    {
+      "epoch": 1.9958197863446354,
+      "grad_norm": 0.19683946052220597,
+      "learning_rate": 1.8588640275387263e-05,
+      "loss": 0.3494,
+      "step": 2149
+    },
+    {
+      "epoch": 1.996748722712494,
+      "grad_norm": 0.2216547899837589,
+      "learning_rate": 1.8571428571428572e-05,
+      "loss": 0.3343,
+      "step": 2150
+    },
+    {
+      "epoch": 1.9976776590803529,
+      "grad_norm": 0.18513919900783316,
+      "learning_rate": 1.855421686746988e-05,
+      "loss": 0.3333,
+      "step": 2151
+    },
+    {
+      "epoch": 1.9986065954482117,
+      "grad_norm": 0.19337964070214114,
+      "learning_rate": 1.853700516351119e-05,
+      "loss": 0.349,
+      "step": 2152
+    },
+    {
+      "epoch": 1.9995355318160706,
+      "grad_norm": 0.22054539434718806,
+      "learning_rate": 1.8519793459552494e-05,
+      "loss": 0.3425,
+      "step": 2153
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.3469287314164476,
+      "learning_rate": 1.8502581755593802e-05,
+      "loss": 0.3178,
+      "step": 2154
+    },
+    {
+      "epoch": 2.000928936367859,
+      "grad_norm": 0.2157058376159686,
+      "learning_rate": 1.848537005163511e-05,
+      "loss": 0.2774,
+      "step": 2155
+    },
+    {
+      "epoch": 2.0018578727357177,
+      "grad_norm": 0.23148826903740533,
+      "learning_rate": 1.846815834767642e-05,
+      "loss": 0.2857,
+      "step": 2156
+    },
+    {
+      "epoch": 2.0027868091035765,
+      "grad_norm": 0.2528270886829438,
+      "learning_rate": 1.8450946643717727e-05,
+      "loss": 0.2855,
+      "step": 2157
+    },
+    {
+      "epoch": 2.0037157454714354,
+      "grad_norm": 0.24255204959272156,
+      "learning_rate": 1.8433734939759036e-05,
+      "loss": 0.2678,
+      "step": 2158
+    },
+    {
+      "epoch": 2.0046446818392942,
+      "grad_norm": 0.21988723244876518,
+      "learning_rate": 1.8416523235800344e-05,
+      "loss": 0.2666,
+      "step": 2159
+    },
+    {
+      "epoch": 2.0055736182071526,
+      "grad_norm": 0.230258825927274,
+      "learning_rate": 1.8399311531841652e-05,
+      "loss": 0.2855,
+      "step": 2160
+    },
+    {
+      "epoch": 2.0065025545750115,
+      "grad_norm": 0.24750969996912872,
+      "learning_rate": 1.838209982788296e-05,
+      "loss": 0.274,
+      "step": 2161
+    },
+    {
+      "epoch": 2.0074314909428703,
+      "grad_norm": 0.21415269672959325,
+      "learning_rate": 1.836488812392427e-05,
+      "loss": 0.2559,
+      "step": 2162
+    },
+    {
+      "epoch": 2.008360427310729,
+      "grad_norm": 0.24277353543089242,
+      "learning_rate": 1.8347676419965578e-05,
+      "loss": 0.2881,
+      "step": 2163
+    },
+    {
+      "epoch": 2.009289363678588,
+      "grad_norm": 0.22462647600480273,
+      "learning_rate": 1.8330464716006886e-05,
+      "loss": 0.273,
+      "step": 2164
+    },
+    {
+      "epoch": 2.010218300046447,
+      "grad_norm": 0.23734616553826293,
+      "learning_rate": 1.8313253012048194e-05,
+      "loss": 0.2876,
+      "step": 2165
+    },
+    {
+      "epoch": 2.0111472364143057,
+      "grad_norm": 0.24635100347524322,
+      "learning_rate": 1.82960413080895e-05,
+      "loss": 0.2943,
+      "step": 2166
+    },
+    {
+      "epoch": 2.0120761727821646,
+      "grad_norm": 0.23031332092273404,
+      "learning_rate": 1.8278829604130808e-05,
+      "loss": 0.2724,
+      "step": 2167
+    },
+    {
+      "epoch": 2.0130051091500234,
+      "grad_norm": 0.20356233261897416,
+      "learning_rate": 1.8261617900172116e-05,
+      "loss": 0.2768,
+      "step": 2168
+    },
+    {
+      "epoch": 2.0139340455178822,
+      "grad_norm": 0.24232305321170966,
+      "learning_rate": 1.8244406196213425e-05,
+      "loss": 0.278,
+      "step": 2169
+    },
+    {
+      "epoch": 2.0148629818857406,
+      "grad_norm": 0.20514711352879814,
+      "learning_rate": 1.8227194492254733e-05,
+      "loss": 0.2889,
+      "step": 2170
+    },
+    {
+      "epoch": 2.0157919182535995,
+      "grad_norm": 0.21172830872100393,
+      "learning_rate": 1.820998278829604e-05,
+      "loss": 0.2795,
+      "step": 2171
+    },
+    {
+      "epoch": 2.0167208546214583,
+      "grad_norm": 0.19343812296165566,
+      "learning_rate": 1.819277108433735e-05,
+      "loss": 0.27,
+      "step": 2172
+    },
+    {
+      "epoch": 2.017649790989317,
+      "grad_norm": 0.22135939544753655,
+      "learning_rate": 1.8175559380378658e-05,
+      "loss": 0.2653,
+      "step": 2173
+    },
+    {
+      "epoch": 2.018578727357176,
+      "grad_norm": 0.21690056177642822,
+      "learning_rate": 1.8158347676419967e-05,
+      "loss": 0.2739,
+      "step": 2174
+    },
+    {
+      "epoch": 2.019507663725035,
+      "grad_norm": 0.21806288455919925,
+      "learning_rate": 1.8141135972461275e-05,
+      "loss": 0.2973,
+      "step": 2175
+    },
+    {
+      "epoch": 2.0204366000928937,
+      "grad_norm": 0.22552713294809715,
+      "learning_rate": 1.8123924268502583e-05,
+      "loss": 0.2854,
+      "step": 2176
+    },
+    {
+      "epoch": 2.0213655364607526,
+      "grad_norm": 0.2055813988009532,
+      "learning_rate": 1.8106712564543892e-05,
+      "loss": 0.2806,
+      "step": 2177
+    },
+    {
+      "epoch": 2.0222944728286114,
+      "grad_norm": 0.20183039247919923,
+      "learning_rate": 1.8089500860585197e-05,
+      "loss": 0.2812,
+      "step": 2178
+    },
+    {
+      "epoch": 2.0232234091964703,
+      "grad_norm": 0.19722763146647482,
+      "learning_rate": 1.8072289156626505e-05,
+      "loss": 0.2677,
+      "step": 2179
+    },
+    {
+      "epoch": 2.0241523455643287,
+      "grad_norm": 0.21974854118511616,
+      "learning_rate": 1.8055077452667814e-05,
+      "loss": 0.2869,
+      "step": 2180
+    },
+    {
+      "epoch": 2.0250812819321875,
+      "grad_norm": 0.20162885066764683,
+      "learning_rate": 1.8037865748709122e-05,
+      "loss": 0.2793,
+      "step": 2181
+    },
+    {
+      "epoch": 2.0260102183000464,
+      "grad_norm": 0.18375708438721025,
+      "learning_rate": 1.802065404475043e-05,
+      "loss": 0.2693,
+      "step": 2182
+    },
+    {
+      "epoch": 2.026939154667905,
+      "grad_norm": 0.2180582338151327,
+      "learning_rate": 1.800344234079174e-05,
+      "loss": 0.2778,
+      "step": 2183
+    },
+    {
+      "epoch": 2.027868091035764,
+      "grad_norm": 0.1904035088310135,
+      "learning_rate": 1.7986230636833047e-05,
+      "loss": 0.267,
+      "step": 2184
+    },
+    {
+      "epoch": 2.028797027403623,
+      "grad_norm": 0.21014267773827153,
+      "learning_rate": 1.7969018932874356e-05,
+      "loss": 0.2888,
+      "step": 2185
+    },
+    {
+      "epoch": 2.0297259637714817,
+      "grad_norm": 0.19891687553756965,
+      "learning_rate": 1.7951807228915664e-05,
+      "loss": 0.2874,
+      "step": 2186
+    },
+    {
+      "epoch": 2.0306549001393406,
+      "grad_norm": 0.20789519384380667,
+      "learning_rate": 1.7934595524956972e-05,
+      "loss": 0.2623,
+      "step": 2187
+    },
+    {
+      "epoch": 2.0315838365071994,
+      "grad_norm": 0.20429364046527473,
+      "learning_rate": 1.791738382099828e-05,
+      "loss": 0.2784,
+      "step": 2188
+    },
+    {
+      "epoch": 2.0325127728750583,
+      "grad_norm": 0.19034811630615903,
+      "learning_rate": 1.790017211703959e-05,
+      "loss": 0.292,
+      "step": 2189
+    },
+    {
+      "epoch": 2.0334417092429167,
+      "grad_norm": 0.20841543986466363,
+      "learning_rate": 1.7882960413080894e-05,
+      "loss": 0.2819,
+      "step": 2190
+    },
+    {
+      "epoch": 2.0343706456107755,
+      "grad_norm": 0.1955048400721127,
+      "learning_rate": 1.7865748709122203e-05,
+      "loss": 0.2641,
+      "step": 2191
+    },
+    {
+      "epoch": 2.0352995819786344,
+      "grad_norm": 0.1830392111751671,
+      "learning_rate": 1.784853700516351e-05,
+      "loss": 0.2671,
+      "step": 2192
+    },
+    {
+      "epoch": 2.036228518346493,
+      "grad_norm": 0.1902652316296126,
+      "learning_rate": 1.783132530120482e-05,
+      "loss": 0.2767,
+      "step": 2193
+    },
+    {
+      "epoch": 2.037157454714352,
+      "grad_norm": 0.19416968904801624,
+      "learning_rate": 1.7814113597246128e-05,
+      "loss": 0.2768,
+      "step": 2194
+    },
+    {
+      "epoch": 2.038086391082211,
+      "grad_norm": 0.22500831983539799,
+      "learning_rate": 1.7796901893287436e-05,
+      "loss": 0.2738,
+      "step": 2195
+    },
+    {
+      "epoch": 2.0390153274500697,
+      "grad_norm": 0.1928120950604119,
+      "learning_rate": 1.7779690189328745e-05,
+      "loss": 0.2811,
+      "step": 2196
+    },
+    {
+      "epoch": 2.0399442638179286,
+      "grad_norm": 0.17208435269872213,
+      "learning_rate": 1.7762478485370053e-05,
+      "loss": 0.2755,
+      "step": 2197
+    },
+    {
+      "epoch": 2.0408732001857874,
+      "grad_norm": 0.18090731260063056,
+      "learning_rate": 1.774526678141136e-05,
+      "loss": 0.2544,
+      "step": 2198
+    },
+    {
+      "epoch": 2.0418021365536463,
+      "grad_norm": 0.183776818234785,
+      "learning_rate": 1.772805507745267e-05,
+      "loss": 0.2866,
+      "step": 2199
+    },
+    {
+      "epoch": 2.0427310729215047,
+      "grad_norm": 0.18371201328625772,
+      "learning_rate": 1.7710843373493978e-05,
+      "loss": 0.2704,
+      "step": 2200
+    },
+    {
+      "epoch": 2.0436600092893635,
+      "grad_norm": 0.18123129067098215,
+      "learning_rate": 1.7693631669535287e-05,
+      "loss": 0.3001,
+      "step": 2201
+    },
+    {
+      "epoch": 2.0445889456572224,
+      "grad_norm": 0.1811746843377487,
+      "learning_rate": 1.7676419965576595e-05,
+      "loss": 0.2676,
+      "step": 2202
+    },
+    {
+      "epoch": 2.0455178820250812,
+      "grad_norm": 0.18554403639433137,
+      "learning_rate": 1.76592082616179e-05,
+      "loss": 0.2738,
+      "step": 2203
+    },
+    {
+      "epoch": 2.04644681839294,
+      "grad_norm": 0.19747699767180618,
+      "learning_rate": 1.764199655765921e-05,
+      "loss": 0.2763,
+      "step": 2204
+    },
+    {
+      "epoch": 2.047375754760799,
+      "grad_norm": 0.1935650970013363,
+      "learning_rate": 1.7624784853700517e-05,
+      "loss": 0.2715,
+      "step": 2205
+    },
+    {
+      "epoch": 2.0483046911286578,
+      "grad_norm": 0.1854401984756031,
+      "learning_rate": 1.7607573149741825e-05,
+      "loss": 0.2789,
+      "step": 2206
+    },
+    {
+      "epoch": 2.0492336274965166,
+      "grad_norm": 0.1878981380085049,
+      "learning_rate": 1.7590361445783134e-05,
+      "loss": 0.2747,
+      "step": 2207
+    },
+    {
+      "epoch": 2.0501625638643755,
+      "grad_norm": 0.18824729243879953,
+      "learning_rate": 1.7573149741824442e-05,
+      "loss": 0.2766,
+      "step": 2208
+    },
+    {
+      "epoch": 2.0510915002322343,
+      "grad_norm": 0.18511176681612157,
+      "learning_rate": 1.755593803786575e-05,
+      "loss": 0.2808,
+      "step": 2209
+    },
+    {
+      "epoch": 2.0520204366000927,
+      "grad_norm": 0.17356200032688698,
+      "learning_rate": 1.753872633390706e-05,
+      "loss": 0.2739,
+      "step": 2210
+    },
+    {
+      "epoch": 2.0529493729679515,
+      "grad_norm": 0.1751736860021714,
+      "learning_rate": 1.7521514629948367e-05,
+      "loss": 0.2685,
+      "step": 2211
+    },
+    {
+      "epoch": 2.0538783093358104,
+      "grad_norm": 0.18256208438125998,
+      "learning_rate": 1.7504302925989676e-05,
+      "loss": 0.2604,
+      "step": 2212
+    },
+    {
+      "epoch": 2.0548072457036692,
+      "grad_norm": 0.1767328307937035,
+      "learning_rate": 1.7487091222030984e-05,
+      "loss": 0.2673,
+      "step": 2213
+    },
+    {
+      "epoch": 2.055736182071528,
+      "grad_norm": 0.18036991264224542,
+      "learning_rate": 1.7469879518072292e-05,
+      "loss": 0.2735,
+      "step": 2214
+    },
+    {
+      "epoch": 2.056665118439387,
+      "grad_norm": 0.17618684914959876,
+      "learning_rate": 1.7452667814113597e-05,
+      "loss": 0.2671,
+      "step": 2215
+    },
+    {
+      "epoch": 2.0575940548072458,
+      "grad_norm": 0.22278489819230726,
+      "learning_rate": 1.7435456110154906e-05,
+      "loss": 0.2953,
+      "step": 2216
+    },
+    {
+      "epoch": 2.0585229911751046,
+      "grad_norm": 0.1957544879322214,
+      "learning_rate": 1.7418244406196214e-05,
+      "loss": 0.29,
+      "step": 2217
+    },
+    {
+      "epoch": 2.0594519275429635,
+      "grad_norm": 0.20251022082815323,
+      "learning_rate": 1.7401032702237523e-05,
+      "loss": 0.2655,
+      "step": 2218
+    },
+    {
+      "epoch": 2.0603808639108223,
+      "grad_norm": 0.18210373540439084,
+      "learning_rate": 1.738382099827883e-05,
+      "loss": 0.2719,
+      "step": 2219
+    },
+    {
+      "epoch": 2.0613098002786807,
+      "grad_norm": 0.19081176507689804,
+      "learning_rate": 1.736660929432014e-05,
+      "loss": 0.289,
+      "step": 2220
+    },
+    {
+      "epoch": 2.0622387366465396,
+      "grad_norm": 0.17312113263605755,
+      "learning_rate": 1.7349397590361448e-05,
+      "loss": 0.2707,
+      "step": 2221
+    },
+    {
+      "epoch": 2.0631676730143984,
+      "grad_norm": 0.2042563838051068,
+      "learning_rate": 1.7332185886402756e-05,
+      "loss": 0.2734,
+      "step": 2222
+    },
+    {
+      "epoch": 2.0640966093822573,
+      "grad_norm": 0.19562693428663167,
+      "learning_rate": 1.7314974182444065e-05,
+      "loss": 0.2735,
+      "step": 2223
+    },
+    {
+      "epoch": 2.065025545750116,
+      "grad_norm": 0.207464339658486,
+      "learning_rate": 1.7297762478485373e-05,
+      "loss": 0.2771,
+      "step": 2224
+    },
+    {
+      "epoch": 2.065954482117975,
+      "grad_norm": 0.20297657867438057,
+      "learning_rate": 1.728055077452668e-05,
+      "loss": 0.2945,
+      "step": 2225
+    },
+    {
+      "epoch": 2.066883418485834,
+      "grad_norm": 0.19308466101824381,
+      "learning_rate": 1.726333907056799e-05,
+      "loss": 0.2761,
+      "step": 2226
+    },
+    {
+      "epoch": 2.0678123548536926,
+      "grad_norm": 0.1940231498846366,
+      "learning_rate": 1.7246127366609295e-05,
+      "loss": 0.2781,
+      "step": 2227
+    },
+    {
+      "epoch": 2.0687412912215515,
+      "grad_norm": 0.19959094783677078,
+      "learning_rate": 1.7228915662650603e-05,
+      "loss": 0.283,
+      "step": 2228
+    },
+    {
+      "epoch": 2.0696702275894103,
+      "grad_norm": 0.17718713745827044,
+      "learning_rate": 1.721170395869191e-05,
+      "loss": 0.2731,
+      "step": 2229
+    },
+    {
+      "epoch": 2.0705991639572687,
+      "grad_norm": 0.20114808299866702,
+      "learning_rate": 1.719449225473322e-05,
+      "loss": 0.2682,
+      "step": 2230
+    },
+    {
+      "epoch": 2.0715281003251276,
+      "grad_norm": 0.18891743595084226,
+      "learning_rate": 1.717728055077453e-05,
+      "loss": 0.268,
+      "step": 2231
+    },
+    {
+      "epoch": 2.0724570366929864,
+      "grad_norm": 0.19305694440730528,
+      "learning_rate": 1.7160068846815837e-05,
+      "loss": 0.2864,
+      "step": 2232
+    },
+    {
+      "epoch": 2.0733859730608453,
+      "grad_norm": 0.18612754148970131,
+      "learning_rate": 1.7142857142857145e-05,
+      "loss": 0.2804,
+      "step": 2233
+    },
+    {
+      "epoch": 2.074314909428704,
+      "grad_norm": 0.21870942336091884,
+      "learning_rate": 1.7125645438898454e-05,
+      "loss": 0.2803,
+      "step": 2234
+    },
+    {
+      "epoch": 2.075243845796563,
+      "grad_norm": 0.1905749129757181,
+      "learning_rate": 1.7108433734939762e-05,
+      "loss": 0.2769,
+      "step": 2235
+    },
+    {
+      "epoch": 2.076172782164422,
+      "grad_norm": 0.17958345546132004,
+      "learning_rate": 1.709122203098107e-05,
+      "loss": 0.2676,
+      "step": 2236
+    },
+    {
+      "epoch": 2.0771017185322806,
+      "grad_norm": 0.17978922180228651,
+      "learning_rate": 1.707401032702238e-05,
+      "loss": 0.2566,
+      "step": 2237
+    },
+    {
+      "epoch": 2.0780306549001395,
+      "grad_norm": 0.19619208069729835,
+      "learning_rate": 1.7056798623063684e-05,
+      "loss": 0.2847,
+      "step": 2238
+    },
+    {
+      "epoch": 2.0789595912679983,
+      "grad_norm": 0.2028901050066484,
+      "learning_rate": 1.7039586919104992e-05,
+      "loss": 0.2839,
+      "step": 2239
+    },
+    {
+      "epoch": 2.0798885276358567,
+      "grad_norm": 0.20553977912065274,
+      "learning_rate": 1.70223752151463e-05,
+      "loss": 0.27,
+      "step": 2240
+    },
+    {
+      "epoch": 2.0808174640037156,
+      "grad_norm": 0.2175363885394602,
+      "learning_rate": 1.700516351118761e-05,
+      "loss": 0.2827,
+      "step": 2241
+    },
+    {
+      "epoch": 2.0817464003715744,
+      "grad_norm": 0.17374801708666573,
+      "learning_rate": 1.6987951807228917e-05,
+      "loss": 0.2514,
+      "step": 2242
+    },
+    {
+      "epoch": 2.0826753367394333,
+      "grad_norm": 0.2009093330228981,
+      "learning_rate": 1.6970740103270226e-05,
+      "loss": 0.2695,
+      "step": 2243
+    },
+    {
+      "epoch": 2.083604273107292,
+      "grad_norm": 0.18927752447294688,
+      "learning_rate": 1.6953528399311534e-05,
+      "loss": 0.2576,
+      "step": 2244
+    },
+    {
+      "epoch": 2.084533209475151,
+      "grad_norm": 0.23624065141040287,
+      "learning_rate": 1.6936316695352843e-05,
+      "loss": 0.2998,
+      "step": 2245
+    },
+    {
+      "epoch": 2.08546214584301,
+      "grad_norm": 0.18922983252395634,
+      "learning_rate": 1.691910499139415e-05,
+      "loss": 0.2805,
+      "step": 2246
+    },
+    {
+      "epoch": 2.0863910822108687,
+      "grad_norm": 0.19589618442669648,
+      "learning_rate": 1.690189328743546e-05,
+      "loss": 0.2683,
+      "step": 2247
+    },
+    {
+      "epoch": 2.0873200185787275,
+      "grad_norm": 0.19640479008012807,
+      "learning_rate": 1.6884681583476768e-05,
+      "loss": 0.2744,
+      "step": 2248
+    },
+    {
+      "epoch": 2.0882489549465864,
+      "grad_norm": 0.19901030382264118,
+      "learning_rate": 1.6867469879518073e-05,
+      "loss": 0.264,
+      "step": 2249
+    },
+    {
+      "epoch": 2.0891778913144448,
+      "grad_norm": 0.1952682814125198,
+      "learning_rate": 1.685025817555938e-05,
+      "loss": 0.2958,
+      "step": 2250
+    },
+    {
+      "epoch": 2.0901068276823036,
+      "grad_norm": 0.19766258576899515,
+      "learning_rate": 1.683304647160069e-05,
+      "loss": 0.2786,
+      "step": 2251
+    },
+    {
+      "epoch": 2.0910357640501624,
+      "grad_norm": 0.19455969769783835,
+      "learning_rate": 1.6815834767641998e-05,
+      "loss": 0.2798,
+      "step": 2252
+    },
+    {
+      "epoch": 2.0919647004180213,
+      "grad_norm": 0.18818819171389964,
+      "learning_rate": 1.6798623063683306e-05,
+      "loss": 0.2781,
+      "step": 2253
+    },
+    {
+      "epoch": 2.09289363678588,
+      "grad_norm": 0.1915501131624007,
+      "learning_rate": 1.6781411359724615e-05,
+      "loss": 0.2744,
+      "step": 2254
+    },
+    {
+      "epoch": 2.093822573153739,
+      "grad_norm": 0.18831604411563369,
+      "learning_rate": 1.6764199655765923e-05,
+      "loss": 0.2619,
+      "step": 2255
+    },
+    {
+      "epoch": 2.094751509521598,
+      "grad_norm": 0.19348282994201058,
+      "learning_rate": 1.674698795180723e-05,
+      "loss": 0.2751,
+      "step": 2256
+    },
+    {
+      "epoch": 2.0956804458894567,
+      "grad_norm": 0.19615082676185192,
+      "learning_rate": 1.672977624784854e-05,
+      "loss": 0.2888,
+      "step": 2257
+    },
+    {
+      "epoch": 2.0966093822573155,
+      "grad_norm": 0.1847380252200025,
+      "learning_rate": 1.671256454388985e-05,
+      "loss": 0.2854,
+      "step": 2258
+    },
+    {
+      "epoch": 2.0975383186251744,
+      "grad_norm": 0.20348389567918942,
+      "learning_rate": 1.6695352839931153e-05,
+      "loss": 0.2873,
+      "step": 2259
+    },
+    {
+      "epoch": 2.0984672549930328,
+      "grad_norm": 0.19400279600413042,
+      "learning_rate": 1.6678141135972462e-05,
+      "loss": 0.2813,
+      "step": 2260
+    },
+    {
+      "epoch": 2.0993961913608916,
+      "grad_norm": 0.18111282515696545,
+      "learning_rate": 1.666092943201377e-05,
+      "loss": 0.2983,
+      "step": 2261
+    },
+    {
+      "epoch": 2.1003251277287505,
+      "grad_norm": 0.18813347341361206,
+      "learning_rate": 1.664371772805508e-05,
+      "loss": 0.269,
+      "step": 2262
+    },
+    {
+      "epoch": 2.1012540640966093,
+      "grad_norm": 0.17424998589403623,
+      "learning_rate": 1.6626506024096387e-05,
+      "loss": 0.2607,
+      "step": 2263
+    },
+    {
+      "epoch": 2.102183000464468,
+      "grad_norm": 0.17051368168715583,
+      "learning_rate": 1.6609294320137695e-05,
+      "loss": 0.2632,
+      "step": 2264
+    },
+    {
+      "epoch": 2.103111936832327,
+      "grad_norm": 0.19333768098843307,
+      "learning_rate": 1.6592082616179004e-05,
+      "loss": 0.2739,
+      "step": 2265
+    },
+    {
+      "epoch": 2.104040873200186,
+      "grad_norm": 0.17969196315731728,
+      "learning_rate": 1.6574870912220312e-05,
+      "loss": 0.2642,
+      "step": 2266
+    },
+    {
+      "epoch": 2.1049698095680447,
+      "grad_norm": 0.17290109863089814,
+      "learning_rate": 1.655765920826162e-05,
+      "loss": 0.274,
+      "step": 2267
+    },
+    {
+      "epoch": 2.1058987459359035,
+      "grad_norm": 0.17856097137064214,
+      "learning_rate": 1.654044750430293e-05,
+      "loss": 0.2735,
+      "step": 2268
+    },
+    {
+      "epoch": 2.1068276823037624,
+      "grad_norm": 0.18795624301701083,
+      "learning_rate": 1.6523235800344237e-05,
+      "loss": 0.2811,
+      "step": 2269
+    },
+    {
+      "epoch": 2.107756618671621,
+      "grad_norm": 0.17878323498135645,
+      "learning_rate": 1.6506024096385542e-05,
+      "loss": 0.2757,
+      "step": 2270
+    },
+    {
+      "epoch": 2.1086855550394796,
+      "grad_norm": 0.1754157334553124,
+      "learning_rate": 1.648881239242685e-05,
+      "loss": 0.2719,
+      "step": 2271
+    },
+    {
+      "epoch": 2.1096144914073385,
+      "grad_norm": 0.18969010978993706,
+      "learning_rate": 1.647160068846816e-05,
+      "loss": 0.2805,
+      "step": 2272
+    },
+    {
+      "epoch": 2.1105434277751973,
+      "grad_norm": 0.2042293911687927,
+      "learning_rate": 1.6454388984509468e-05,
+      "loss": 0.2869,
+      "step": 2273
+    },
+    {
+      "epoch": 2.111472364143056,
+      "grad_norm": 0.1733527665317661,
+      "learning_rate": 1.6437177280550776e-05,
+      "loss": 0.2684,
+      "step": 2274
+    },
+    {
+      "epoch": 2.112401300510915,
+      "grad_norm": 0.18133085027013407,
+      "learning_rate": 1.6419965576592084e-05,
+      "loss": 0.2783,
+      "step": 2275
+    },
+    {
+      "epoch": 2.113330236878774,
+      "grad_norm": 0.19680333317378604,
+      "learning_rate": 1.6402753872633393e-05,
+      "loss": 0.2804,
+      "step": 2276
+    },
+    {
+      "epoch": 2.1142591732466327,
+      "grad_norm": 0.18081135463697318,
+      "learning_rate": 1.63855421686747e-05,
+      "loss": 0.2605,
+      "step": 2277
+    },
+    {
+      "epoch": 2.1151881096144916,
+      "grad_norm": 0.19740057774259165,
+      "learning_rate": 1.636833046471601e-05,
+      "loss": 0.2789,
+      "step": 2278
+    },
+    {
+      "epoch": 2.1161170459823504,
+      "grad_norm": 0.18568459190301742,
+      "learning_rate": 1.6351118760757318e-05,
+      "loss": 0.2828,
+      "step": 2279
+    },
+    {
+      "epoch": 2.117045982350209,
+      "grad_norm": 0.19365683468329514,
+      "learning_rate": 1.6333907056798626e-05,
+      "loss": 0.2688,
+      "step": 2280
+    },
+    {
+      "epoch": 2.1179749187180676,
+      "grad_norm": 0.20252419703650162,
+      "learning_rate": 1.631669535283993e-05,
+      "loss": 0.2713,
+      "step": 2281
+    },
+    {
+      "epoch": 2.1189038550859265,
+      "grad_norm": 0.1792890569752184,
+      "learning_rate": 1.629948364888124e-05,
+      "loss": 0.262,
+      "step": 2282
+    },
+    {
+      "epoch": 2.1198327914537853,
+      "grad_norm": 0.20524389028390555,
+      "learning_rate": 1.6282271944922548e-05,
+      "loss": 0.2843,
+      "step": 2283
+    },
+    {
+      "epoch": 2.120761727821644,
+      "grad_norm": 0.28341598447437294,
+      "learning_rate": 1.6265060240963857e-05,
+      "loss": 0.3053,
+      "step": 2284
+    },
+    {
+      "epoch": 2.121690664189503,
+      "grad_norm": 0.20203034410437481,
+      "learning_rate": 1.6247848537005165e-05,
+      "loss": 0.2854,
+      "step": 2285
+    },
+    {
+      "epoch": 2.122619600557362,
+      "grad_norm": 0.1953939465453228,
+      "learning_rate": 1.6230636833046473e-05,
+      "loss": 0.2899,
+      "step": 2286
+    },
+    {
+      "epoch": 2.1235485369252207,
+      "grad_norm": 0.1793459966677846,
+      "learning_rate": 1.6213425129087782e-05,
+      "loss": 0.2796,
+      "step": 2287
+    },
+    {
+      "epoch": 2.1244774732930796,
+      "grad_norm": 0.17783901924314915,
+      "learning_rate": 1.619621342512909e-05,
+      "loss": 0.2626,
+      "step": 2288
+    },
+    {
+      "epoch": 2.1254064096609384,
+      "grad_norm": 0.22343298968404166,
+      "learning_rate": 1.61790017211704e-05,
+      "loss": 0.2811,
+      "step": 2289
+    },
+    {
+      "epoch": 2.126335346028797,
+      "grad_norm": 0.1814389246560736,
+      "learning_rate": 1.6161790017211707e-05,
+      "loss": 0.2598,
+      "step": 2290
+    },
+    {
+      "epoch": 2.1272642823966557,
+      "grad_norm": 0.18055709417523266,
+      "learning_rate": 1.6144578313253012e-05,
+      "loss": 0.2798,
+      "step": 2291
+    },
+    {
+      "epoch": 2.1281932187645145,
+      "grad_norm": 0.17888962889126264,
+      "learning_rate": 1.612736660929432e-05,
+      "loss": 0.2682,
+      "step": 2292
+    },
+    {
+      "epoch": 2.1291221551323734,
+      "grad_norm": 0.19721301430366406,
+      "learning_rate": 1.611015490533563e-05,
+      "loss": 0.2719,
+      "step": 2293
+    },
+    {
+      "epoch": 2.130051091500232,
+      "grad_norm": 0.19444402714079692,
+      "learning_rate": 1.6092943201376937e-05,
+      "loss": 0.2783,
+      "step": 2294
+    },
+    {
+      "epoch": 2.130980027868091,
+      "grad_norm": 0.1838632683011167,
+      "learning_rate": 1.6075731497418246e-05,
+      "loss": 0.2768,
+      "step": 2295
+    },
+    {
+      "epoch": 2.13190896423595,
+      "grad_norm": 0.1884248262044888,
+      "learning_rate": 1.6058519793459554e-05,
+      "loss": 0.2707,
+      "step": 2296
+    },
+    {
+      "epoch": 2.1328379006038087,
+      "grad_norm": 0.19204342794555168,
+      "learning_rate": 1.6041308089500863e-05,
+      "loss": 0.2775,
+      "step": 2297
+    },
+    {
+      "epoch": 2.1337668369716676,
+      "grad_norm": 0.18608945592808687,
+      "learning_rate": 1.602409638554217e-05,
+      "loss": 0.2685,
+      "step": 2298
+    },
+    {
+      "epoch": 2.1346957733395264,
+      "grad_norm": 0.18587485562812903,
+      "learning_rate": 1.600688468158348e-05,
+      "loss": 0.2786,
+      "step": 2299
+    },
+    {
+      "epoch": 2.1356247097073853,
+      "grad_norm": 0.19587986603415075,
+      "learning_rate": 1.5989672977624788e-05,
+      "loss": 0.2856,
+      "step": 2300
+    },
+    {
+      "epoch": 2.1365536460752437,
+      "grad_norm": 0.1942871240310058,
+      "learning_rate": 1.5972461273666096e-05,
+      "loss": 0.2803,
+      "step": 2301
+    },
+    {
+      "epoch": 2.1374825824431025,
+      "grad_norm": 0.18749941680014653,
+      "learning_rate": 1.59552495697074e-05,
+      "loss": 0.2917,
+      "step": 2302
+    },
+    {
+      "epoch": 2.1384115188109614,
+      "grad_norm": 0.1887106230362805,
+      "learning_rate": 1.593803786574871e-05,
+      "loss": 0.2872,
+      "step": 2303
+    },
+    {
+      "epoch": 2.13934045517882,
+      "grad_norm": 0.18424120413217845,
+      "learning_rate": 1.5920826161790018e-05,
+      "loss": 0.2744,
+      "step": 2304
+    },
+    {
+      "epoch": 2.140269391546679,
+      "grad_norm": 0.2091266962240203,
+      "learning_rate": 1.5903614457831326e-05,
+      "loss": 0.2872,
+      "step": 2305
+    },
+    {
+      "epoch": 2.141198327914538,
+      "grad_norm": 0.167622736533787,
+      "learning_rate": 1.5886402753872635e-05,
+      "loss": 0.2579,
+      "step": 2306
+    },
+    {
+      "epoch": 2.1421272642823967,
+      "grad_norm": 0.18300097585692288,
+      "learning_rate": 1.5869191049913943e-05,
+      "loss": 0.2979,
+      "step": 2307
+    },
+    {
+      "epoch": 2.1430562006502556,
+      "grad_norm": 0.18348416995061526,
+      "learning_rate": 1.585197934595525e-05,
+      "loss": 0.261,
+      "step": 2308
+    },
+    {
+      "epoch": 2.1439851370181144,
+      "grad_norm": 0.1890149322297578,
+      "learning_rate": 1.583476764199656e-05,
+      "loss": 0.2918,
+      "step": 2309
+    },
+    {
+      "epoch": 2.144914073385973,
+      "grad_norm": 0.1925329558848419,
+      "learning_rate": 1.581755593803787e-05,
+      "loss": 0.2823,
+      "step": 2310
+    },
+    {
+      "epoch": 2.1458430097538317,
+      "grad_norm": 0.1805093088723906,
+      "learning_rate": 1.5800344234079177e-05,
+      "loss": 0.279,
+      "step": 2311
+    },
+    {
+      "epoch": 2.1467719461216905,
+      "grad_norm": 0.1896794503755551,
+      "learning_rate": 1.5783132530120482e-05,
+      "loss": 0.2791,
+      "step": 2312
+    },
+    {
+      "epoch": 2.1477008824895494,
+      "grad_norm": 0.19321251388893304,
+      "learning_rate": 1.576592082616179e-05,
+      "loss": 0.2947,
+      "step": 2313
+    },
+    {
+      "epoch": 2.1486298188574082,
+      "grad_norm": 0.17234766420937184,
+      "learning_rate": 1.57487091222031e-05,
+      "loss": 0.2516,
+      "step": 2314
+    },
+    {
+      "epoch": 2.149558755225267,
+      "grad_norm": 0.1822533993953394,
+      "learning_rate": 1.5731497418244407e-05,
+      "loss": 0.2823,
+      "step": 2315
+    },
+    {
+      "epoch": 2.150487691593126,
+      "grad_norm": 0.17706007655705194,
+      "learning_rate": 1.5714285714285715e-05,
+      "loss": 0.2874,
+      "step": 2316
+    },
+    {
+      "epoch": 2.1514166279609848,
+      "grad_norm": 0.19007988270275075,
+      "learning_rate": 1.5697074010327024e-05,
+      "loss": 0.2964,
+      "step": 2317
+    },
+    {
+      "epoch": 2.1523455643288436,
+      "grad_norm": 0.18529312530513586,
+      "learning_rate": 1.5679862306368332e-05,
+      "loss": 0.2622,
+      "step": 2318
+    },
+    {
+      "epoch": 2.1532745006967025,
+      "grad_norm": 0.17979646607166183,
+      "learning_rate": 1.566265060240964e-05,
+      "loss": 0.2873,
+      "step": 2319
+    },
+    {
+      "epoch": 2.1542034370645613,
+      "grad_norm": 0.17790060931246865,
+      "learning_rate": 1.564543889845095e-05,
+      "loss": 0.2684,
+      "step": 2320
+    },
+    {
+      "epoch": 2.1551323734324197,
+      "grad_norm": 0.18215704599038443,
+      "learning_rate": 1.5628227194492257e-05,
+      "loss": 0.2541,
+      "step": 2321
+    },
+    {
+      "epoch": 2.1560613098002785,
+      "grad_norm": 0.1892490048553774,
+      "learning_rate": 1.5611015490533566e-05,
+      "loss": 0.2669,
+      "step": 2322
+    },
+    {
+      "epoch": 2.1569902461681374,
+      "grad_norm": 0.19249753393355534,
+      "learning_rate": 1.559380378657487e-05,
+      "loss": 0.2761,
+      "step": 2323
+    },
+    {
+      "epoch": 2.1579191825359962,
+      "grad_norm": 0.1839816309966501,
+      "learning_rate": 1.557659208261618e-05,
+      "loss": 0.2677,
+      "step": 2324
+    },
+    {
+      "epoch": 2.158848118903855,
+      "grad_norm": 0.1810375798407296,
+      "learning_rate": 1.5559380378657488e-05,
+      "loss": 0.2773,
+      "step": 2325
+    },
+    {
+      "epoch": 2.159777055271714,
+      "grad_norm": 0.18831165733890587,
+      "learning_rate": 1.5542168674698796e-05,
+      "loss": 0.2848,
+      "step": 2326
+    },
+    {
+      "epoch": 2.1607059916395728,
+      "grad_norm": 0.19377240510871852,
+      "learning_rate": 1.5524956970740104e-05,
+      "loss": 0.2941,
+      "step": 2327
+    },
+    {
+      "epoch": 2.1616349280074316,
+      "grad_norm": 0.1761716521658807,
+      "learning_rate": 1.5507745266781413e-05,
+      "loss": 0.2829,
+      "step": 2328
+    },
+    {
+      "epoch": 2.1625638643752905,
+      "grad_norm": 0.19274957074290722,
+      "learning_rate": 1.549053356282272e-05,
+      "loss": 0.2732,
+      "step": 2329
+    },
+    {
+      "epoch": 2.163492800743149,
+      "grad_norm": 0.18470851789842668,
+      "learning_rate": 1.547332185886403e-05,
+      "loss": 0.2829,
+      "step": 2330
+    },
+    {
+      "epoch": 2.1644217371110077,
+      "grad_norm": 0.17644911709437527,
+      "learning_rate": 1.5456110154905338e-05,
+      "loss": 0.2639,
+      "step": 2331
+    },
+    {
+      "epoch": 2.1653506734788666,
+      "grad_norm": 0.1860663550583413,
+      "learning_rate": 1.5438898450946646e-05,
+      "loss": 0.265,
+      "step": 2332
+    },
+    {
+      "epoch": 2.1662796098467254,
+      "grad_norm": 0.1732712738965505,
+      "learning_rate": 1.5421686746987955e-05,
+      "loss": 0.275,
+      "step": 2333
+    },
+    {
+      "epoch": 2.1672085462145843,
+      "grad_norm": 0.20738176414089443,
+      "learning_rate": 1.540447504302926e-05,
+      "loss": 0.2922,
+      "step": 2334
+    },
+    {
+      "epoch": 2.168137482582443,
+      "grad_norm": 0.18615130026642834,
+      "learning_rate": 1.5387263339070568e-05,
+      "loss": 0.2749,
+      "step": 2335
+    },
+    {
+      "epoch": 2.169066418950302,
+      "grad_norm": 0.18330688472974185,
+      "learning_rate": 1.5370051635111877e-05,
+      "loss": 0.2829,
+      "step": 2336
+    },
+    {
+      "epoch": 2.169995355318161,
+      "grad_norm": 0.19173214451709955,
+      "learning_rate": 1.5352839931153185e-05,
+      "loss": 0.2915,
+      "step": 2337
+    },
+    {
+      "epoch": 2.1709242916860196,
+      "grad_norm": 0.186766371199974,
+      "learning_rate": 1.5335628227194493e-05,
+      "loss": 0.2658,
+      "step": 2338
+    },
+    {
+      "epoch": 2.1718532280538785,
+      "grad_norm": 0.18396142682701574,
+      "learning_rate": 1.5318416523235802e-05,
+      "loss": 0.2899,
+      "step": 2339
+    },
+    {
+      "epoch": 2.1727821644217373,
+      "grad_norm": 0.1780638489393759,
+      "learning_rate": 1.530120481927711e-05,
+      "loss": 0.2787,
+      "step": 2340
+    },
+    {
+      "epoch": 2.1737111007895957,
+      "grad_norm": 0.1752883172604511,
+      "learning_rate": 1.528399311531842e-05,
+      "loss": 0.2748,
+      "step": 2341
+    },
+    {
+      "epoch": 2.1746400371574546,
+      "grad_norm": 0.177303228639095,
+      "learning_rate": 1.5266781411359727e-05,
+      "loss": 0.2783,
+      "step": 2342
+    },
+    {
+      "epoch": 2.1755689735253134,
+      "grad_norm": 0.18826552255300416,
+      "learning_rate": 1.5249569707401035e-05,
+      "loss": 0.2684,
+      "step": 2343
+    },
+    {
+      "epoch": 2.1764979098931723,
+      "grad_norm": 0.17442496731786208,
+      "learning_rate": 1.523235800344234e-05,
+      "loss": 0.2749,
+      "step": 2344
+    },
+    {
+      "epoch": 2.177426846261031,
+      "grad_norm": 0.1737699610108496,
+      "learning_rate": 1.5215146299483649e-05,
+      "loss": 0.2703,
+      "step": 2345
+    },
+    {
+      "epoch": 2.17835578262889,
+      "grad_norm": 0.17176047456146679,
+      "learning_rate": 1.5197934595524957e-05,
+      "loss": 0.2683,
+      "step": 2346
+    },
+    {
+      "epoch": 2.179284718996749,
+      "grad_norm": 0.17869107629479228,
+      "learning_rate": 1.5180722891566266e-05,
+      "loss": 0.2689,
+      "step": 2347
+    },
+    {
+      "epoch": 2.1802136553646077,
+      "grad_norm": 0.188008817213917,
+      "learning_rate": 1.5163511187607574e-05,
+      "loss": 0.2735,
+      "step": 2348
+    },
+    {
+      "epoch": 2.1811425917324665,
+      "grad_norm": 0.17523593662500742,
+      "learning_rate": 1.5146299483648882e-05,
+      "loss": 0.2763,
+      "step": 2349
+    },
+    {
+      "epoch": 2.182071528100325,
+      "grad_norm": 0.18284465023530216,
+      "learning_rate": 1.512908777969019e-05,
+      "loss": 0.2792,
+      "step": 2350
+    },
+    {
+      "epoch": 2.1830004644681837,
+      "grad_norm": 0.1868107881910502,
+      "learning_rate": 1.51118760757315e-05,
+      "loss": 0.2841,
+      "step": 2351
+    },
+    {
+      "epoch": 2.1839294008360426,
+      "grad_norm": 0.1814772554278054,
+      "learning_rate": 1.5094664371772808e-05,
+      "loss": 0.2818,
+      "step": 2352
+    },
+    {
+      "epoch": 2.1848583372039014,
+      "grad_norm": 0.1982916282455446,
+      "learning_rate": 1.5077452667814116e-05,
+      "loss": 0.2909,
+      "step": 2353
+    },
+    {
+      "epoch": 2.1857872735717603,
+      "grad_norm": 0.18547018214356015,
+      "learning_rate": 1.5060240963855424e-05,
+      "loss": 0.2633,
+      "step": 2354
+    },
+    {
+      "epoch": 2.186716209939619,
+      "grad_norm": 0.18629229106963163,
+      "learning_rate": 1.504302925989673e-05,
+      "loss": 0.2714,
+      "step": 2355
+    },
+    {
+      "epoch": 2.187645146307478,
+      "grad_norm": 0.17792953090605365,
+      "learning_rate": 1.5025817555938038e-05,
+      "loss": 0.2804,
+      "step": 2356
+    },
+    {
+      "epoch": 2.188574082675337,
+      "grad_norm": 0.19411643483321742,
+      "learning_rate": 1.5008605851979346e-05,
+      "loss": 0.2834,
+      "step": 2357
+    },
+    {
+      "epoch": 2.1895030190431957,
+      "grad_norm": 0.17838088091084456,
+      "learning_rate": 1.4991394148020655e-05,
+      "loss": 0.2894,
+      "step": 2358
+    },
+    {
+      "epoch": 2.1904319554110545,
+      "grad_norm": 0.18798746896736152,
+      "learning_rate": 1.4974182444061963e-05,
+      "loss": 0.2825,
+      "step": 2359
+    },
+    {
+      "epoch": 2.1913608917789134,
+      "grad_norm": 0.18713834475410582,
+      "learning_rate": 1.4956970740103271e-05,
+      "loss": 0.2753,
+      "step": 2360
+    },
+    {
+      "epoch": 2.1922898281467718,
+      "grad_norm": 0.21091756592860836,
+      "learning_rate": 1.493975903614458e-05,
+      "loss": 0.2841,
+      "step": 2361
+    },
+    {
+      "epoch": 2.1932187645146306,
+      "grad_norm": 0.18950367949638064,
+      "learning_rate": 1.4922547332185888e-05,
+      "loss": 0.2786,
+      "step": 2362
+    },
+    {
+      "epoch": 2.1941477008824894,
+      "grad_norm": 0.18906056192602488,
+      "learning_rate": 1.4905335628227197e-05,
+      "loss": 0.2792,
+      "step": 2363
+    },
+    {
+      "epoch": 2.1950766372503483,
+      "grad_norm": 0.18613054799021314,
+      "learning_rate": 1.4888123924268505e-05,
+      "loss": 0.2681,
+      "step": 2364
+    },
+    {
+      "epoch": 2.196005573618207,
+      "grad_norm": 0.18070972999393983,
+      "learning_rate": 1.487091222030981e-05,
+      "loss": 0.2689,
+      "step": 2365
+    },
+    {
+      "epoch": 2.196934509986066,
+      "grad_norm": 0.17512758804049097,
+      "learning_rate": 1.4853700516351118e-05,
+      "loss": 0.2702,
+      "step": 2366
+    },
+    {
+      "epoch": 2.197863446353925,
+      "grad_norm": 0.1977826508419132,
+      "learning_rate": 1.4836488812392427e-05,
+      "loss": 0.2714,
+      "step": 2367
+    },
+    {
+      "epoch": 2.1987923827217837,
+      "grad_norm": 0.19045916194894494,
+      "learning_rate": 1.4819277108433735e-05,
+      "loss": 0.2898,
+      "step": 2368
+    },
+    {
+      "epoch": 2.1997213190896425,
+      "grad_norm": 0.1894952240845188,
+      "learning_rate": 1.4802065404475044e-05,
+      "loss": 0.2793,
+      "step": 2369
+    },
+    {
+      "epoch": 2.200650255457501,
+      "grad_norm": 0.18712603681731005,
+      "learning_rate": 1.4784853700516352e-05,
+      "loss": 0.2875,
+      "step": 2370
+    },
+    {
+      "epoch": 2.2015791918253598,
+      "grad_norm": 0.18017707880699574,
+      "learning_rate": 1.476764199655766e-05,
+      "loss": 0.2676,
+      "step": 2371
+    },
+    {
+      "epoch": 2.2025081281932186,
+      "grad_norm": 0.1872086868065948,
+      "learning_rate": 1.4750430292598969e-05,
+      "loss": 0.2812,
+      "step": 2372
+    },
+    {
+      "epoch": 2.2034370645610775,
+      "grad_norm": 0.18279582654005014,
+      "learning_rate": 1.4733218588640277e-05,
+      "loss": 0.2775,
+      "step": 2373
+    },
+    {
+      "epoch": 2.2043660009289363,
+      "grad_norm": 0.1742546303606481,
+      "learning_rate": 1.4716006884681586e-05,
+      "loss": 0.2672,
+      "step": 2374
+    },
+    {
+      "epoch": 2.205294937296795,
+      "grad_norm": 0.19829823980108882,
+      "learning_rate": 1.4698795180722894e-05,
+      "loss": 0.2932,
+      "step": 2375
+    },
+    {
+      "epoch": 2.206223873664654,
+      "grad_norm": 0.1907823494693345,
+      "learning_rate": 1.4681583476764199e-05,
+      "loss": 0.2759,
+      "step": 2376
+    },
+    {
+      "epoch": 2.207152810032513,
+      "grad_norm": 0.1760244498190223,
+      "learning_rate": 1.4664371772805507e-05,
+      "loss": 0.2675,
+      "step": 2377
+    },
+    {
+      "epoch": 2.2080817464003717,
+      "grad_norm": 0.17827743105039992,
+      "learning_rate": 1.4647160068846816e-05,
+      "loss": 0.2737,
+      "step": 2378
+    },
+    {
+      "epoch": 2.2090106827682305,
+      "grad_norm": 0.19304042505289626,
+      "learning_rate": 1.4629948364888124e-05,
+      "loss": 0.2798,
+      "step": 2379
+    },
+    {
+      "epoch": 2.2099396191360894,
+      "grad_norm": 0.16940952191421077,
+      "learning_rate": 1.4612736660929433e-05,
+      "loss": 0.2736,
+      "step": 2380
+    },
+    {
+      "epoch": 2.210868555503948,
+      "grad_norm": 2.077281485156273,
+      "learning_rate": 1.4595524956970741e-05,
+      "loss": 0.3002,
+      "step": 2381
+    },
+    {
+      "epoch": 2.2117974918718066,
+      "grad_norm": 0.19341272255759445,
+      "learning_rate": 1.457831325301205e-05,
+      "loss": 0.2777,
+      "step": 2382
+    },
+    {
+      "epoch": 2.2127264282396655,
+      "grad_norm": 0.20112600841398107,
+      "learning_rate": 1.4561101549053358e-05,
+      "loss": 0.2887,
+      "step": 2383
+    },
+    {
+      "epoch": 2.2136553646075243,
+      "grad_norm": 0.19413557575902854,
+      "learning_rate": 1.4543889845094666e-05,
+      "loss": 0.2947,
+      "step": 2384
+    },
+    {
+      "epoch": 2.214584300975383,
+      "grad_norm": 0.19780119460533946,
+      "learning_rate": 1.4526678141135975e-05,
+      "loss": 0.2839,
+      "step": 2385
+    },
+    {
+      "epoch": 2.215513237343242,
+      "grad_norm": 0.17913396155866435,
+      "learning_rate": 1.4509466437177283e-05,
+      "loss": 0.2733,
+      "step": 2386
+    },
+    {
+      "epoch": 2.216442173711101,
+      "grad_norm": 0.18886643483257676,
+      "learning_rate": 1.4492254733218588e-05,
+      "loss": 0.2864,
+      "step": 2387
+    },
+    {
+      "epoch": 2.2173711100789597,
+      "grad_norm": 0.1995227456484706,
+      "learning_rate": 1.4475043029259896e-05,
+      "loss": 0.2838,
+      "step": 2388
+    },
+    {
+      "epoch": 2.2183000464468186,
+      "grad_norm": 0.18895265517020549,
+      "learning_rate": 1.4457831325301205e-05,
+      "loss": 0.2706,
+      "step": 2389
+    },
+    {
+      "epoch": 2.2192289828146774,
+      "grad_norm": 0.19334993843744272,
+      "learning_rate": 1.4440619621342513e-05,
+      "loss": 0.2787,
+      "step": 2390
+    },
+    {
+      "epoch": 2.220157919182536,
+      "grad_norm": 0.20236064176192772,
+      "learning_rate": 1.4423407917383822e-05,
+      "loss": 0.2788,
+      "step": 2391
+    },
+    {
+      "epoch": 2.2210868555503946,
+      "grad_norm": 0.3655805333065711,
+      "learning_rate": 1.440619621342513e-05,
+      "loss": 0.2708,
+      "step": 2392
+    },
+    {
+      "epoch": 2.2220157919182535,
+      "grad_norm": 0.18758244632384213,
+      "learning_rate": 1.4388984509466438e-05,
+      "loss": 0.2825,
+      "step": 2393
+    },
+    {
+      "epoch": 2.2229447282861123,
+      "grad_norm": 0.17757060628418905,
+      "learning_rate": 1.4371772805507747e-05,
+      "loss": 0.2779,
+      "step": 2394
+    },
+    {
+      "epoch": 2.223873664653971,
+      "grad_norm": 0.19104858972751676,
+      "learning_rate": 1.4354561101549055e-05,
+      "loss": 0.272,
+      "step": 2395
+    },
+    {
+      "epoch": 2.22480260102183,
+      "grad_norm": 0.19793684191220914,
+      "learning_rate": 1.4337349397590364e-05,
+      "loss": 0.2728,
+      "step": 2396
+    },
+    {
+      "epoch": 2.225731537389689,
+      "grad_norm": 0.18934791385839767,
+      "learning_rate": 1.4320137693631669e-05,
+      "loss": 0.2796,
+      "step": 2397
+    },
+    {
+      "epoch": 2.2266604737575477,
+      "grad_norm": 0.17614906901226796,
+      "learning_rate": 1.4302925989672977e-05,
+      "loss": 0.2812,
+      "step": 2398
+    },
+    {
+      "epoch": 2.2275894101254066,
+      "grad_norm": 0.17586759106445218,
+      "learning_rate": 1.4285714285714285e-05,
+      "loss": 0.2677,
+      "step": 2399
+    },
+    {
+      "epoch": 2.2285183464932654,
+      "grad_norm": 0.1872402948258275,
+      "learning_rate": 1.4268502581755594e-05,
+      "loss": 0.2555,
+      "step": 2400
+    },
+    {
+      "epoch": 2.229447282861124,
+      "grad_norm": 0.19771821410158671,
+      "learning_rate": 1.4251290877796902e-05,
+      "loss": 0.2677,
+      "step": 2401
+    },
+    {
+      "epoch": 2.2303762192289827,
+      "grad_norm": 0.17818595832024386,
+      "learning_rate": 1.423407917383821e-05,
+      "loss": 0.2607,
+      "step": 2402
+    },
+    {
+      "epoch": 2.2313051555968415,
+      "grad_norm": 0.1776547560819404,
+      "learning_rate": 1.4216867469879519e-05,
+      "loss": 0.2749,
+      "step": 2403
+    },
+    {
+      "epoch": 2.2322340919647004,
+      "grad_norm": 0.170957875139368,
+      "learning_rate": 1.4199655765920827e-05,
+      "loss": 0.2591,
+      "step": 2404
+    },
+    {
+      "epoch": 2.233163028332559,
+      "grad_norm": 0.1742647844913809,
+      "learning_rate": 1.4182444061962136e-05,
+      "loss": 0.2738,
+      "step": 2405
+    },
+    {
+      "epoch": 2.234091964700418,
+      "grad_norm": 0.1925283348131071,
+      "learning_rate": 1.4165232358003444e-05,
+      "loss": 0.2812,
+      "step": 2406
+    },
+    {
+      "epoch": 2.235020901068277,
+      "grad_norm": 0.19555384576301638,
+      "learning_rate": 1.4148020654044753e-05,
+      "loss": 0.2945,
+      "step": 2407
+    },
+    {
+      "epoch": 2.2359498374361357,
+      "grad_norm": 0.17682561155125856,
+      "learning_rate": 1.4130808950086058e-05,
+      "loss": 0.2614,
+      "step": 2408
+    },
+    {
+      "epoch": 2.2368787738039946,
+      "grad_norm": 0.21751007119184168,
+      "learning_rate": 1.4113597246127366e-05,
+      "loss": 0.2985,
+      "step": 2409
+    },
+    {
+      "epoch": 2.2378077101718534,
+      "grad_norm": 0.18932523578727964,
+      "learning_rate": 1.4096385542168674e-05,
+      "loss": 0.2706,
+      "step": 2410
+    },
+    {
+      "epoch": 2.238736646539712,
+      "grad_norm": 0.1741531414035046,
+      "learning_rate": 1.4079173838209983e-05,
+      "loss": 0.2775,
+      "step": 2411
+    },
+    {
+      "epoch": 2.2396655829075707,
+      "grad_norm": 0.1790079893951182,
+      "learning_rate": 1.4061962134251291e-05,
+      "loss": 0.2716,
+      "step": 2412
+    },
+    {
+      "epoch": 2.2405945192754295,
+      "grad_norm": 0.1897657068346121,
+      "learning_rate": 1.40447504302926e-05,
+      "loss": 0.2887,
+      "step": 2413
+    },
+    {
+      "epoch": 2.2415234556432884,
+      "grad_norm": 0.17563878014722972,
+      "learning_rate": 1.4027538726333908e-05,
+      "loss": 0.2678,
+      "step": 2414
+    },
+    {
+      "epoch": 2.242452392011147,
+      "grad_norm": 0.1734405490085664,
+      "learning_rate": 1.4010327022375216e-05,
+      "loss": 0.2863,
+      "step": 2415
+    },
+    {
+      "epoch": 2.243381328379006,
+      "grad_norm": 0.18332196872544093,
+      "learning_rate": 1.3993115318416525e-05,
+      "loss": 0.2734,
+      "step": 2416
+    },
+    {
+      "epoch": 2.244310264746865,
+      "grad_norm": 0.17947567690760682,
+      "learning_rate": 1.3975903614457833e-05,
+      "loss": 0.27,
+      "step": 2417
+    },
+    {
+      "epoch": 2.2452392011147237,
+      "grad_norm": 0.18735527285985934,
+      "learning_rate": 1.3958691910499138e-05,
+      "loss": 0.2851,
+      "step": 2418
+    },
+    {
+      "epoch": 2.2461681374825826,
+      "grad_norm": 0.19049022365781365,
+      "learning_rate": 1.3941480206540447e-05,
+      "loss": 0.2795,
+      "step": 2419
+    },
+    {
+      "epoch": 2.2470970738504414,
+      "grad_norm": 0.17554046191220224,
+      "learning_rate": 1.3924268502581755e-05,
+      "loss": 0.2703,
+      "step": 2420
+    },
+    {
+      "epoch": 2.2480260102183,
+      "grad_norm": 0.18762826003233593,
+      "learning_rate": 1.3907056798623063e-05,
+      "loss": 0.2928,
+      "step": 2421
+    },
+    {
+      "epoch": 2.2489549465861587,
+      "grad_norm": 0.18234373655440445,
+      "learning_rate": 1.3889845094664372e-05,
+      "loss": 0.2817,
+      "step": 2422
+    },
+    {
+      "epoch": 2.2498838829540175,
+      "grad_norm": 0.18049105683824865,
+      "learning_rate": 1.387263339070568e-05,
+      "loss": 0.2744,
+      "step": 2423
+    },
+    {
+      "epoch": 2.2508128193218764,
+      "grad_norm": 0.1770550838421612,
+      "learning_rate": 1.3855421686746989e-05,
+      "loss": 0.2694,
+      "step": 2424
+    },
+    {
+      "epoch": 2.2517417556897352,
+      "grad_norm": 0.1756612487036032,
+      "learning_rate": 1.3838209982788297e-05,
+      "loss": 0.2647,
+      "step": 2425
+    },
+    {
+      "epoch": 2.252670692057594,
+      "grad_norm": 0.1811227177376827,
+      "learning_rate": 1.3820998278829605e-05,
+      "loss": 0.2732,
+      "step": 2426
+    },
+    {
+      "epoch": 2.253599628425453,
+      "grad_norm": 0.1895660967425297,
+      "learning_rate": 1.3803786574870914e-05,
+      "loss": 0.2883,
+      "step": 2427
+    },
+    {
+      "epoch": 2.2545285647933118,
+      "grad_norm": 0.1759719292934546,
+      "learning_rate": 1.3786574870912222e-05,
+      "loss": 0.2704,
+      "step": 2428
+    },
+    {
+      "epoch": 2.2554575011611706,
+      "grad_norm": 0.1852248917014425,
+      "learning_rate": 1.3769363166953527e-05,
+      "loss": 0.2714,
+      "step": 2429
+    },
+    {
+      "epoch": 2.2563864375290295,
+      "grad_norm": 0.1872463960739118,
+      "learning_rate": 1.3752151462994836e-05,
+      "loss": 0.2884,
+      "step": 2430
+    },
+    {
+      "epoch": 2.257315373896888,
+      "grad_norm": 0.19160635539373605,
+      "learning_rate": 1.3734939759036144e-05,
+      "loss": 0.2653,
+      "step": 2431
+    },
+    {
+      "epoch": 2.2582443102647467,
+      "grad_norm": 0.17691605824841325,
+      "learning_rate": 1.3717728055077452e-05,
+      "loss": 0.2587,
+      "step": 2432
+    },
+    {
+      "epoch": 2.2591732466326055,
+      "grad_norm": 0.1840601443984862,
+      "learning_rate": 1.3700516351118761e-05,
+      "loss": 0.2863,
+      "step": 2433
+    },
+    {
+      "epoch": 2.2601021830004644,
+      "grad_norm": 0.17288469873242204,
+      "learning_rate": 1.368330464716007e-05,
+      "loss": 0.2547,
+      "step": 2434
+    },
+    {
+      "epoch": 2.2610311193683232,
+      "grad_norm": 0.20393520871661627,
+      "learning_rate": 1.3666092943201378e-05,
+      "loss": 0.2918,
+      "step": 2435
+    },
+    {
+      "epoch": 2.261960055736182,
+      "grad_norm": 0.185024404089295,
+      "learning_rate": 1.3648881239242686e-05,
+      "loss": 0.2779,
+      "step": 2436
+    },
+    {
+      "epoch": 2.262888992104041,
+      "grad_norm": 0.1901769547372566,
+      "learning_rate": 1.3631669535283994e-05,
+      "loss": 0.2931,
+      "step": 2437
+    },
+    {
+      "epoch": 2.2638179284718998,
+      "grad_norm": 0.2214837074293375,
+      "learning_rate": 1.3614457831325303e-05,
+      "loss": 0.3066,
+      "step": 2438
+    },
+    {
+      "epoch": 2.2647468648397586,
+      "grad_norm": 0.18890674277222674,
+      "learning_rate": 1.3597246127366611e-05,
+      "loss": 0.273,
+      "step": 2439
+    },
+    {
+      "epoch": 2.2656758012076175,
+      "grad_norm": 0.17542645415927108,
+      "learning_rate": 1.3580034423407916e-05,
+      "loss": 0.2746,
+      "step": 2440
+    },
+    {
+      "epoch": 2.2666047375754763,
+      "grad_norm": 0.19723610546547288,
+      "learning_rate": 1.3562822719449225e-05,
+      "loss": 0.2862,
+      "step": 2441
+    },
+    {
+      "epoch": 2.2675336739433347,
+      "grad_norm": 0.18409020452903432,
+      "learning_rate": 1.3545611015490533e-05,
+      "loss": 0.2756,
+      "step": 2442
+    },
+    {
+      "epoch": 2.2684626103111936,
+      "grad_norm": 0.17662475275167705,
+      "learning_rate": 1.3528399311531841e-05,
+      "loss": 0.2678,
+      "step": 2443
+    },
+    {
+      "epoch": 2.2693915466790524,
+      "grad_norm": 0.1791475766551058,
+      "learning_rate": 1.351118760757315e-05,
+      "loss": 0.275,
+      "step": 2444
+    },
+    {
+      "epoch": 2.2703204830469113,
+      "grad_norm": 0.1854099470575039,
+      "learning_rate": 1.3493975903614458e-05,
+      "loss": 0.2728,
+      "step": 2445
+    },
+    {
+      "epoch": 2.27124941941477,
+      "grad_norm": 0.1852589646237944,
+      "learning_rate": 1.3476764199655767e-05,
+      "loss": 0.277,
+      "step": 2446
+    },
+    {
+      "epoch": 2.272178355782629,
+      "grad_norm": 0.18383715705433082,
+      "learning_rate": 1.3459552495697075e-05,
+      "loss": 0.2889,
+      "step": 2447
+    },
+    {
+      "epoch": 2.273107292150488,
+      "grad_norm": 0.1951503760520118,
+      "learning_rate": 1.3442340791738383e-05,
+      "loss": 0.2805,
+      "step": 2448
+    },
+    {
+      "epoch": 2.2740362285183466,
+      "grad_norm": 0.20073372805491246,
+      "learning_rate": 1.3425129087779692e-05,
+      "loss": 0.2731,
+      "step": 2449
+    },
+    {
+      "epoch": 2.2749651648862055,
+      "grad_norm": 0.16310800213885338,
+      "learning_rate": 1.3407917383820997e-05,
+      "loss": 0.2639,
+      "step": 2450
+    },
+    {
+      "epoch": 2.275894101254064,
+      "grad_norm": 0.17606453823569151,
+      "learning_rate": 1.3390705679862305e-05,
+      "loss": 0.2694,
+      "step": 2451
+    },
+    {
+      "epoch": 2.2768230376219227,
+      "grad_norm": 0.21268495266034193,
+      "learning_rate": 1.3373493975903614e-05,
+      "loss": 0.273,
+      "step": 2452
+    },
+    {
+      "epoch": 2.2777519739897816,
+      "grad_norm": 0.17975510622715113,
+      "learning_rate": 1.3356282271944922e-05,
+      "loss": 0.2721,
+      "step": 2453
+    },
+    {
+      "epoch": 2.2786809103576404,
+      "grad_norm": 0.17495866372601926,
+      "learning_rate": 1.333907056798623e-05,
+      "loss": 0.2608,
+      "step": 2454
+    },
+    {
+      "epoch": 2.2796098467254993,
+      "grad_norm": 0.1799744906204507,
+      "learning_rate": 1.3321858864027539e-05,
+      "loss": 0.2773,
+      "step": 2455
+    },
+    {
+      "epoch": 2.280538783093358,
+      "grad_norm": 0.20386977048097904,
+      "learning_rate": 1.3304647160068847e-05,
+      "loss": 0.298,
+      "step": 2456
+    },
+    {
+      "epoch": 2.281467719461217,
+      "grad_norm": 0.17543050916264705,
+      "learning_rate": 1.3287435456110156e-05,
+      "loss": 0.2682,
+      "step": 2457
+    },
+    {
+      "epoch": 2.282396655829076,
+      "grad_norm": 0.18115557846011818,
+      "learning_rate": 1.3270223752151464e-05,
+      "loss": 0.2963,
+      "step": 2458
+    },
+    {
+      "epoch": 2.2833255921969347,
+      "grad_norm": 0.1825233749262127,
+      "learning_rate": 1.3253012048192772e-05,
+      "loss": 0.2759,
+      "step": 2459
+    },
+    {
+      "epoch": 2.2842545285647935,
+      "grad_norm": 0.18810580131761556,
+      "learning_rate": 1.3235800344234081e-05,
+      "loss": 0.2736,
+      "step": 2460
+    },
+    {
+      "epoch": 2.2851834649326523,
+      "grad_norm": 0.20406387820144806,
+      "learning_rate": 1.3218588640275386e-05,
+      "loss": 0.2885,
+      "step": 2461
+    },
+    {
+      "epoch": 2.2861124013005107,
+      "grad_norm": 0.17866786419081768,
+      "learning_rate": 1.3201376936316694e-05,
+      "loss": 0.27,
+      "step": 2462
+    },
+    {
+      "epoch": 2.2870413376683696,
+      "grad_norm": 0.19558916886417338,
+      "learning_rate": 1.3184165232358003e-05,
+      "loss": 0.2745,
+      "step": 2463
+    },
+    {
+      "epoch": 2.2879702740362284,
+      "grad_norm": 0.17178727146349201,
+      "learning_rate": 1.3166953528399311e-05,
+      "loss": 0.2695,
+      "step": 2464
+    },
+    {
+      "epoch": 2.2888992104040873,
+      "grad_norm": 0.18186780650265671,
+      "learning_rate": 1.314974182444062e-05,
+      "loss": 0.2768,
+      "step": 2465
+    },
+    {
+      "epoch": 2.289828146771946,
+      "grad_norm": 0.1851530367223113,
+      "learning_rate": 1.3132530120481928e-05,
+      "loss": 0.2867,
+      "step": 2466
+    },
+    {
+      "epoch": 2.290757083139805,
+      "grad_norm": 0.16670097051368127,
+      "learning_rate": 1.3115318416523236e-05,
+      "loss": 0.2793,
+      "step": 2467
+    },
+    {
+      "epoch": 2.291686019507664,
+      "grad_norm": 0.1856617082893609,
+      "learning_rate": 1.3098106712564545e-05,
+      "loss": 0.2852,
+      "step": 2468
+    },
+    {
+      "epoch": 2.2926149558755227,
+      "grad_norm": 0.16657158904737376,
+      "learning_rate": 1.3080895008605853e-05,
+      "loss": 0.2648,
+      "step": 2469
+    },
+    {
+      "epoch": 2.2935438922433815,
+      "grad_norm": 0.1817338062264846,
+      "learning_rate": 1.3063683304647162e-05,
+      "loss": 0.2888,
+      "step": 2470
+    },
+    {
+      "epoch": 2.29447282861124,
+      "grad_norm": 0.16841896976761195,
+      "learning_rate": 1.3046471600688468e-05,
+      "loss": 0.2765,
+      "step": 2471
+    },
+    {
+      "epoch": 2.2954017649790988,
+      "grad_norm": 0.16883158219004724,
+      "learning_rate": 1.3029259896729775e-05,
+      "loss": 0.2684,
+      "step": 2472
+    },
+    {
+      "epoch": 2.2963307013469576,
+      "grad_norm": 0.17717975998990523,
+      "learning_rate": 1.3012048192771083e-05,
+      "loss": 0.268,
+      "step": 2473
+    },
+    {
+      "epoch": 2.2972596377148164,
+      "grad_norm": 0.18560225793278562,
+      "learning_rate": 1.2994836488812392e-05,
+      "loss": 0.3009,
+      "step": 2474
+    },
+    {
+      "epoch": 2.2981885740826753,
+      "grad_norm": 0.17167104424094876,
+      "learning_rate": 1.29776247848537e-05,
+      "loss": 0.2759,
+      "step": 2475
+    },
+    {
+      "epoch": 2.299117510450534,
+      "grad_norm": 0.18419703054460113,
+      "learning_rate": 1.2960413080895009e-05,
+      "loss": 0.2842,
+      "step": 2476
+    },
+    {
+      "epoch": 2.300046446818393,
+      "grad_norm": 0.18156266735689566,
+      "learning_rate": 1.2943201376936317e-05,
+      "loss": 0.2676,
+      "step": 2477
+    },
+    {
+      "epoch": 2.300975383186252,
+      "grad_norm": 0.18273226393990982,
+      "learning_rate": 1.2925989672977625e-05,
+      "loss": 0.2852,
+      "step": 2478
+    },
+    {
+      "epoch": 2.3019043195541107,
+      "grad_norm": 0.18012479146698604,
+      "learning_rate": 1.2908777969018934e-05,
+      "loss": 0.2817,
+      "step": 2479
+    },
+    {
+      "epoch": 2.3028332559219695,
+      "grad_norm": 0.18611898476583089,
+      "learning_rate": 1.2891566265060242e-05,
+      "loss": 0.2734,
+      "step": 2480
+    },
+    {
+      "epoch": 2.3037621922898284,
+      "grad_norm": 0.2045432229041082,
+      "learning_rate": 1.287435456110155e-05,
+      "loss": 0.2868,
+      "step": 2481
+    },
+    {
+      "epoch": 2.3046911286576868,
+      "grad_norm": 0.18166021704998359,
+      "learning_rate": 1.2857142857142857e-05,
+      "loss": 0.2742,
+      "step": 2482
+    },
+    {
+      "epoch": 2.3056200650255456,
+      "grad_norm": 0.18583776084377682,
+      "learning_rate": 1.2839931153184166e-05,
+      "loss": 0.2878,
+      "step": 2483
+    },
+    {
+      "epoch": 2.3065490013934045,
+      "grad_norm": 0.18056875244820153,
+      "learning_rate": 1.2822719449225474e-05,
+      "loss": 0.2722,
+      "step": 2484
+    },
+    {
+      "epoch": 2.3074779377612633,
+      "grad_norm": 0.1775475098336737,
+      "learning_rate": 1.280550774526678e-05,
+      "loss": 0.2621,
+      "step": 2485
+    },
+    {
+      "epoch": 2.308406874129122,
+      "grad_norm": 0.19416763735743434,
+      "learning_rate": 1.2788296041308089e-05,
+      "loss": 0.2777,
+      "step": 2486
+    },
+    {
+      "epoch": 2.309335810496981,
+      "grad_norm": 0.17930364275371594,
+      "learning_rate": 1.2771084337349398e-05,
+      "loss": 0.2686,
+      "step": 2487
+    },
+    {
+      "epoch": 2.31026474686484,
+      "grad_norm": 0.19195008866900118,
+      "learning_rate": 1.2753872633390706e-05,
+      "loss": 0.2767,
+      "step": 2488
+    },
+    {
+      "epoch": 2.3111936832326987,
+      "grad_norm": 0.2102981226089247,
+      "learning_rate": 1.2736660929432014e-05,
+      "loss": 0.2975,
+      "step": 2489
+    },
+    {
+      "epoch": 2.3121226196005575,
+      "grad_norm": 0.1838111249569394,
+      "learning_rate": 1.2719449225473323e-05,
+      "loss": 0.2701,
+      "step": 2490
+    },
+    {
+      "epoch": 2.313051555968416,
+      "grad_norm": 0.1834891735582498,
+      "learning_rate": 1.2702237521514631e-05,
+      "loss": 0.2668,
+      "step": 2491
+    },
+    {
+      "epoch": 2.313980492336275,
+      "grad_norm": 0.1748837067912934,
+      "learning_rate": 1.268502581755594e-05,
+      "loss": 0.2536,
+      "step": 2492
+    },
+    {
+      "epoch": 2.3149094287041336,
+      "grad_norm": 0.18170866744646788,
+      "learning_rate": 1.2667814113597246e-05,
+      "loss": 0.2791,
+      "step": 2493
+    },
+    {
+      "epoch": 2.3158383650719925,
+      "grad_norm": 0.17517417710104063,
+      "learning_rate": 1.2650602409638555e-05,
+      "loss": 0.2697,
+      "step": 2494
+    },
+    {
+      "epoch": 2.3167673014398513,
+      "grad_norm": 0.1696007724114442,
+      "learning_rate": 1.2633390705679863e-05,
+      "loss": 0.2701,
+      "step": 2495
+    },
+    {
+      "epoch": 2.31769623780771,
+      "grad_norm": 0.19983708306954123,
+      "learning_rate": 1.2616179001721171e-05,
+      "loss": 0.2759,
+      "step": 2496
+    },
+    {
+      "epoch": 2.318625174175569,
+      "grad_norm": 0.18007852898360852,
+      "learning_rate": 1.2598967297762478e-05,
+      "loss": 0.2738,
+      "step": 2497
+    },
+    {
+      "epoch": 2.319554110543428,
+      "grad_norm": 0.17640021787489446,
+      "learning_rate": 1.2581755593803787e-05,
+      "loss": 0.2568,
+      "step": 2498
+    },
+    {
+      "epoch": 2.3204830469112867,
+      "grad_norm": 0.17426432406739611,
+      "learning_rate": 1.2564543889845095e-05,
+      "loss": 0.2813,
+      "step": 2499
+    },
+    {
+      "epoch": 2.3214119832791456,
+      "grad_norm": 0.1819597764792421,
+      "learning_rate": 1.2547332185886403e-05,
+      "loss": 0.2843,
+      "step": 2500
+    },
+    {
+      "epoch": 2.3223409196470044,
+      "grad_norm": 0.6807628441338488,
+      "learning_rate": 1.2530120481927712e-05,
+      "loss": 0.2916,
+      "step": 2501
+    },
+    {
+      "epoch": 2.323269856014863,
+      "grad_norm": 0.18961215732510447,
+      "learning_rate": 1.251290877796902e-05,
+      "loss": 0.2885,
+      "step": 2502
+    },
+    {
+      "epoch": 2.3241987923827216,
+      "grad_norm": 0.185281322046881,
+      "learning_rate": 1.2495697074010327e-05,
+      "loss": 0.261,
+      "step": 2503
+    },
+    {
+      "epoch": 2.3251277287505805,
+      "grad_norm": 0.18212697020003069,
+      "learning_rate": 1.2478485370051635e-05,
+      "loss": 0.2824,
+      "step": 2504
+    },
+    {
+      "epoch": 2.3260566651184393,
+      "grad_norm": 0.19341133551329,
+      "learning_rate": 1.2461273666092944e-05,
+      "loss": 0.2861,
+      "step": 2505
+    },
+    {
+      "epoch": 2.326985601486298,
+      "grad_norm": 0.17022345986335427,
+      "learning_rate": 1.2444061962134252e-05,
+      "loss": 0.2693,
+      "step": 2506
+    },
+    {
+      "epoch": 2.327914537854157,
+      "grad_norm": 0.18536850713483183,
+      "learning_rate": 1.242685025817556e-05,
+      "loss": 0.2746,
+      "step": 2507
+    },
+    {
+      "epoch": 2.328843474222016,
+      "grad_norm": 0.16506173986908923,
+      "learning_rate": 1.2409638554216869e-05,
+      "loss": 0.265,
+      "step": 2508
+    },
+    {
+      "epoch": 2.3297724105898747,
+      "grad_norm": 0.17182803727776838,
+      "learning_rate": 1.2392426850258176e-05,
+      "loss": 0.2736,
+      "step": 2509
+    },
+    {
+      "epoch": 2.3307013469577336,
+      "grad_norm": 0.1742820002887796,
+      "learning_rate": 1.2375215146299484e-05,
+      "loss": 0.2618,
+      "step": 2510
+    },
+    {
+      "epoch": 2.331630283325592,
+      "grad_norm": 0.19642680627052156,
+      "learning_rate": 1.2358003442340792e-05,
+      "loss": 0.274,
+      "step": 2511
+    },
+    {
+      "epoch": 2.332559219693451,
+      "grad_norm": 0.18642669549275,
+      "learning_rate": 1.23407917383821e-05,
+      "loss": 0.284,
+      "step": 2512
+    },
+    {
+      "epoch": 2.3334881560613097,
+      "grad_norm": 0.17054316401763836,
+      "learning_rate": 1.232358003442341e-05,
+      "loss": 0.2869,
+      "step": 2513
+    },
+    {
+      "epoch": 2.3344170924291685,
+      "grad_norm": 0.18369081567830525,
+      "learning_rate": 1.2306368330464718e-05,
+      "loss": 0.2767,
+      "step": 2514
+    },
+    {
+      "epoch": 2.3353460287970274,
+      "grad_norm": 0.20114946960190172,
+      "learning_rate": 1.2289156626506026e-05,
+      "loss": 0.2814,
+      "step": 2515
+    },
+    {
+      "epoch": 2.336274965164886,
+      "grad_norm": 0.18437735158767538,
+      "learning_rate": 1.2271944922547333e-05,
+      "loss": 0.2919,
+      "step": 2516
+    },
+    {
+      "epoch": 2.337203901532745,
+      "grad_norm": 0.18161588561862224,
+      "learning_rate": 1.2254733218588641e-05,
+      "loss": 0.287,
+      "step": 2517
+    },
+    {
+      "epoch": 2.338132837900604,
+      "grad_norm": 0.18870854980187435,
+      "learning_rate": 1.223752151462995e-05,
+      "loss": 0.2805,
+      "step": 2518
+    },
+    {
+      "epoch": 2.3390617742684627,
+      "grad_norm": 0.1768654024578103,
+      "learning_rate": 1.2220309810671258e-05,
+      "loss": 0.269,
+      "step": 2519
+    },
+    {
+      "epoch": 2.3399907106363216,
+      "grad_norm": 0.17490366879789768,
+      "learning_rate": 1.2203098106712566e-05,
+      "loss": 0.2783,
+      "step": 2520
+    },
+    {
+      "epoch": 2.3409196470041804,
+      "grad_norm": 0.18896883370592954,
+      "learning_rate": 1.2185886402753875e-05,
+      "loss": 0.268,
+      "step": 2521
+    },
+    {
+      "epoch": 2.341848583372039,
+      "grad_norm": 0.1801247883571675,
+      "learning_rate": 1.2168674698795181e-05,
+      "loss": 0.2719,
+      "step": 2522
+    },
+    {
+      "epoch": 2.3427775197398977,
+      "grad_norm": 0.18694993552175548,
+      "learning_rate": 1.215146299483649e-05,
+      "loss": 0.2757,
+      "step": 2523
+    },
+    {
+      "epoch": 2.3437064561077565,
+      "grad_norm": 0.18275519642880572,
+      "learning_rate": 1.2134251290877798e-05,
+      "loss": 0.263,
+      "step": 2524
+    },
+    {
+      "epoch": 2.3446353924756154,
+      "grad_norm": 0.17351016085246834,
+      "learning_rate": 1.2117039586919107e-05,
+      "loss": 0.2565,
+      "step": 2525
+    },
+    {
+      "epoch": 2.345564328843474,
+      "grad_norm": 0.18379515482313682,
+      "learning_rate": 1.2099827882960415e-05,
+      "loss": 0.2696,
+      "step": 2526
+    },
+    {
+      "epoch": 2.346493265211333,
+      "grad_norm": 0.1720315048361915,
+      "learning_rate": 1.2082616179001722e-05,
+      "loss": 0.2641,
+      "step": 2527
+    },
+    {
+      "epoch": 2.347422201579192,
+      "grad_norm": 0.18590275177538543,
+      "learning_rate": 1.206540447504303e-05,
+      "loss": 0.2733,
+      "step": 2528
+    },
+    {
+      "epoch": 2.3483511379470507,
+      "grad_norm": 0.169610311172933,
+      "learning_rate": 1.2048192771084338e-05,
+      "loss": 0.2702,
+      "step": 2529
+    },
+    {
+      "epoch": 2.3492800743149096,
+      "grad_norm": 0.1963508763656982,
+      "learning_rate": 1.2030981067125647e-05,
+      "loss": 0.2652,
+      "step": 2530
+    },
+    {
+      "epoch": 2.350209010682768,
+      "grad_norm": 0.18587076519396542,
+      "learning_rate": 1.2013769363166955e-05,
+      "loss": 0.2789,
+      "step": 2531
+    },
+    {
+      "epoch": 2.351137947050627,
+      "grad_norm": 0.1854453938930749,
+      "learning_rate": 1.1996557659208262e-05,
+      "loss": 0.2732,
+      "step": 2532
+    },
+    {
+      "epoch": 2.3520668834184857,
+      "grad_norm": 0.1942839700749339,
+      "learning_rate": 1.197934595524957e-05,
+      "loss": 0.2803,
+      "step": 2533
+    },
+    {
+      "epoch": 2.3529958197863445,
+      "grad_norm": 0.19659561231939945,
+      "learning_rate": 1.1962134251290879e-05,
+      "loss": 0.2913,
+      "step": 2534
+    },
+    {
+      "epoch": 2.3539247561542034,
+      "grad_norm": 0.17480801654404732,
+      "learning_rate": 1.1944922547332187e-05,
+      "loss": 0.271,
+      "step": 2535
+    },
+    {
+      "epoch": 2.3548536925220622,
+      "grad_norm": 0.1944663657157406,
+      "learning_rate": 1.1927710843373496e-05,
+      "loss": 0.2641,
+      "step": 2536
+    },
+    {
+      "epoch": 2.355782628889921,
+      "grad_norm": 0.18816619868751366,
+      "learning_rate": 1.1910499139414804e-05,
+      "loss": 0.2833,
+      "step": 2537
+    },
+    {
+      "epoch": 2.35671156525778,
+      "grad_norm": 0.20018181591126397,
+      "learning_rate": 1.189328743545611e-05,
+      "loss": 0.2852,
+      "step": 2538
+    },
+    {
+      "epoch": 2.3576405016256388,
+      "grad_norm": 0.1756427462956202,
+      "learning_rate": 1.1876075731497419e-05,
+      "loss": 0.2778,
+      "step": 2539
+    },
+    {
+      "epoch": 2.3585694379934976,
+      "grad_norm": 0.18983605510608587,
+      "learning_rate": 1.1858864027538727e-05,
+      "loss": 0.2769,
+      "step": 2540
+    },
+    {
+      "epoch": 2.3594983743613565,
+      "grad_norm": 0.20928473702526423,
+      "learning_rate": 1.1841652323580036e-05,
+      "loss": 0.2856,
+      "step": 2541
+    },
+    {
+      "epoch": 2.360427310729215,
+      "grad_norm": 0.18099417160933323,
+      "learning_rate": 1.1824440619621344e-05,
+      "loss": 0.2758,
+      "step": 2542
+    },
+    {
+      "epoch": 2.3613562470970737,
+      "grad_norm": 0.1788276756913518,
+      "learning_rate": 1.1807228915662651e-05,
+      "loss": 0.2809,
+      "step": 2543
+    },
+    {
+      "epoch": 2.3622851834649325,
+      "grad_norm": 0.18196101771025752,
+      "learning_rate": 1.179001721170396e-05,
+      "loss": 0.2714,
+      "step": 2544
+    },
+    {
+      "epoch": 2.3632141198327914,
+      "grad_norm": 0.16952857307776525,
+      "learning_rate": 1.1772805507745268e-05,
+      "loss": 0.2562,
+      "step": 2545
+    },
+    {
+      "epoch": 2.3641430562006502,
+      "grad_norm": 0.18206933762294697,
+      "learning_rate": 1.1755593803786576e-05,
+      "loss": 0.2761,
+      "step": 2546
+    },
+    {
+      "epoch": 2.365071992568509,
+      "grad_norm": 0.19380265765568364,
+      "learning_rate": 1.1738382099827885e-05,
+      "loss": 0.2869,
+      "step": 2547
+    },
+    {
+      "epoch": 2.366000928936368,
+      "grad_norm": 0.16983950092360575,
+      "learning_rate": 1.1721170395869191e-05,
+      "loss": 0.2745,
+      "step": 2548
+    },
+    {
+      "epoch": 2.3669298653042268,
+      "grad_norm": 0.19191163491346017,
+      "learning_rate": 1.17039586919105e-05,
+      "loss": 0.2946,
+      "step": 2549
+    },
+    {
+      "epoch": 2.3678588016720856,
+      "grad_norm": 0.17125497389397198,
+      "learning_rate": 1.1686746987951808e-05,
+      "loss": 0.2735,
+      "step": 2550
+    },
+    {
+      "epoch": 2.368787738039944,
+      "grad_norm": 0.19949567009374117,
+      "learning_rate": 1.1669535283993117e-05,
+      "loss": 0.2946,
+      "step": 2551
+    },
+    {
+      "epoch": 2.369716674407803,
+      "grad_norm": 0.18799349450204403,
+      "learning_rate": 1.1652323580034425e-05,
+      "loss": 0.2838,
+      "step": 2552
+    },
+    {
+      "epoch": 2.3706456107756617,
+      "grad_norm": 0.17431131561492344,
+      "learning_rate": 1.1635111876075733e-05,
+      "loss": 0.2662,
+      "step": 2553
+    },
+    {
+      "epoch": 2.3715745471435206,
+      "grad_norm": 0.17729505501913617,
+      "learning_rate": 1.161790017211704e-05,
+      "loss": 0.2845,
+      "step": 2554
+    },
+    {
+      "epoch": 2.3725034835113794,
+      "grad_norm": 0.17567637231257308,
+      "learning_rate": 1.1600688468158348e-05,
+      "loss": 0.2828,
+      "step": 2555
+    },
+    {
+      "epoch": 2.3734324198792383,
+      "grad_norm": 0.18325562355641456,
+      "learning_rate": 1.1583476764199657e-05,
+      "loss": 0.2881,
+      "step": 2556
+    },
+    {
+      "epoch": 2.374361356247097,
+      "grad_norm": 0.20307829278142425,
+      "learning_rate": 1.1566265060240965e-05,
+      "loss": 0.2797,
+      "step": 2557
+    },
+    {
+      "epoch": 2.375290292614956,
+      "grad_norm": 0.1760215816447019,
+      "learning_rate": 1.1549053356282274e-05,
+      "loss": 0.2866,
+      "step": 2558
+    },
+    {
+      "epoch": 2.376219228982815,
+      "grad_norm": 0.16869374727160663,
+      "learning_rate": 1.153184165232358e-05,
+      "loss": 0.2677,
+      "step": 2559
+    },
+    {
+      "epoch": 2.3771481653506736,
+      "grad_norm": 0.19634176148081195,
+      "learning_rate": 1.1514629948364889e-05,
+      "loss": 0.2826,
+      "step": 2560
+    },
+    {
+      "epoch": 2.3780771017185325,
+      "grad_norm": 0.1901365638963679,
+      "learning_rate": 1.1497418244406197e-05,
+      "loss": 0.2717,
+      "step": 2561
+    },
+    {
+      "epoch": 2.3790060380863913,
+      "grad_norm": 0.17289531409457187,
+      "learning_rate": 1.1480206540447506e-05,
+      "loss": 0.2745,
+      "step": 2562
+    },
+    {
+      "epoch": 2.3799349744542497,
+      "grad_norm": 0.1863744077538756,
+      "learning_rate": 1.1462994836488814e-05,
+      "loss": 0.2861,
+      "step": 2563
+    },
+    {
+      "epoch": 2.3808639108221086,
+      "grad_norm": 0.19647354430195266,
+      "learning_rate": 1.144578313253012e-05,
+      "loss": 0.2953,
+      "step": 2564
+    },
+    {
+      "epoch": 2.3817928471899674,
+      "grad_norm": 0.16997392062698125,
+      "learning_rate": 1.1428571428571429e-05,
+      "loss": 0.2662,
+      "step": 2565
+    },
+    {
+      "epoch": 2.3827217835578263,
+      "grad_norm": 0.17741241720060805,
+      "learning_rate": 1.1411359724612737e-05,
+      "loss": 0.2801,
+      "step": 2566
+    },
+    {
+      "epoch": 2.383650719925685,
+      "grad_norm": 0.1734889437651768,
+      "learning_rate": 1.1394148020654046e-05,
+      "loss": 0.2715,
+      "step": 2567
+    },
+    {
+      "epoch": 2.384579656293544,
+      "grad_norm": 0.18052716242882721,
+      "learning_rate": 1.1376936316695354e-05,
+      "loss": 0.2861,
+      "step": 2568
+    },
+    {
+      "epoch": 2.385508592661403,
+      "grad_norm": 0.17565238572476038,
+      "learning_rate": 1.1359724612736663e-05,
+      "loss": 0.2714,
+      "step": 2569
+    },
+    {
+      "epoch": 2.3864375290292617,
+      "grad_norm": 0.1706278981440593,
+      "learning_rate": 1.134251290877797e-05,
+      "loss": 0.2762,
+      "step": 2570
+    },
+    {
+      "epoch": 2.38736646539712,
+      "grad_norm": 0.17926154917329037,
+      "learning_rate": 1.1325301204819278e-05,
+      "loss": 0.2854,
+      "step": 2571
+    },
+    {
+      "epoch": 2.388295401764979,
+      "grad_norm": 0.1802656856974113,
+      "learning_rate": 1.1308089500860586e-05,
+      "loss": 0.2807,
+      "step": 2572
+    },
+    {
+      "epoch": 2.3892243381328377,
+      "grad_norm": 0.17506633593361373,
+      "learning_rate": 1.1290877796901895e-05,
+      "loss": 0.2779,
+      "step": 2573
+    },
+    {
+      "epoch": 2.3901532745006966,
+      "grad_norm": 0.1728471092867977,
+      "learning_rate": 1.1273666092943203e-05,
+      "loss": 0.2765,
+      "step": 2574
+    },
+    {
+      "epoch": 2.3910822108685554,
+      "grad_norm": 0.19565702280776281,
+      "learning_rate": 1.125645438898451e-05,
+      "loss": 0.2635,
+      "step": 2575
+    },
+    {
+      "epoch": 2.3920111472364143,
+      "grad_norm": 0.18314447255310315,
+      "learning_rate": 1.1239242685025818e-05,
+      "loss": 0.2737,
+      "step": 2576
+    },
+    {
+      "epoch": 2.392940083604273,
+      "grad_norm": 0.1750548586458759,
+      "learning_rate": 1.1222030981067126e-05,
+      "loss": 0.2642,
+      "step": 2577
+    },
+    {
+      "epoch": 2.393869019972132,
+      "grad_norm": 0.1832451160171968,
+      "learning_rate": 1.1204819277108435e-05,
+      "loss": 0.2745,
+      "step": 2578
+    },
+    {
+      "epoch": 2.394797956339991,
+      "grad_norm": 0.18031801543289583,
+      "learning_rate": 1.1187607573149743e-05,
+      "loss": 0.2684,
+      "step": 2579
+    },
+    {
+      "epoch": 2.3957268927078497,
+      "grad_norm": 0.1799011721479443,
+      "learning_rate": 1.117039586919105e-05,
+      "loss": 0.2807,
+      "step": 2580
+    },
+    {
+      "epoch": 2.3966558290757085,
+      "grad_norm": 0.17467422843455963,
+      "learning_rate": 1.1153184165232358e-05,
+      "loss": 0.2661,
+      "step": 2581
+    },
+    {
+      "epoch": 2.3975847654435674,
+      "grad_norm": 0.182510619647452,
+      "learning_rate": 1.1135972461273667e-05,
+      "loss": 0.2661,
+      "step": 2582
+    },
+    {
+      "epoch": 2.3985137018114258,
+      "grad_norm": 0.19000712669188646,
+      "learning_rate": 1.1118760757314975e-05,
+      "loss": 0.2785,
+      "step": 2583
+    },
+    {
+      "epoch": 2.3994426381792846,
+      "grad_norm": 0.16903359729376602,
+      "learning_rate": 1.1101549053356284e-05,
+      "loss": 0.2704,
+      "step": 2584
+    },
+    {
+      "epoch": 2.4003715745471434,
+      "grad_norm": 0.18217095684501244,
+      "learning_rate": 1.108433734939759e-05,
+      "loss": 0.2634,
+      "step": 2585
+    },
+    {
+      "epoch": 2.4013005109150023,
+      "grad_norm": 0.17717495586427512,
+      "learning_rate": 1.1067125645438899e-05,
+      "loss": 0.2551,
+      "step": 2586
+    },
+    {
+      "epoch": 2.402229447282861,
+      "grad_norm": 0.1838135913410062,
+      "learning_rate": 1.1049913941480207e-05,
+      "loss": 0.2876,
+      "step": 2587
+    },
+    {
+      "epoch": 2.40315838365072,
+      "grad_norm": 0.17031644253992032,
+      "learning_rate": 1.1032702237521515e-05,
+      "loss": 0.2624,
+      "step": 2588
+    },
+    {
+      "epoch": 2.404087320018579,
+      "grad_norm": 0.17556071203202087,
+      "learning_rate": 1.1015490533562824e-05,
+      "loss": 0.2753,
+      "step": 2589
+    },
+    {
+      "epoch": 2.4050162563864377,
+      "grad_norm": 0.1834407267830562,
+      "learning_rate": 1.0998278829604132e-05,
+      "loss": 0.2746,
+      "step": 2590
+    },
+    {
+      "epoch": 2.4059451927542965,
+      "grad_norm": 0.19926877874180607,
+      "learning_rate": 1.0981067125645439e-05,
+      "loss": 0.29,
+      "step": 2591
+    },
+    {
+      "epoch": 2.406874129122155,
+      "grad_norm": 0.17906776663032192,
+      "learning_rate": 1.0963855421686747e-05,
+      "loss": 0.2844,
+      "step": 2592
+    },
+    {
+      "epoch": 2.4078030654900138,
+      "grad_norm": 0.16699051055445974,
+      "learning_rate": 1.0946643717728056e-05,
+      "loss": 0.2588,
+      "step": 2593
+    },
+    {
+      "epoch": 2.4087320018578726,
+      "grad_norm": 0.1751714906923023,
+      "learning_rate": 1.0929432013769364e-05,
+      "loss": 0.2717,
+      "step": 2594
+    },
+    {
+      "epoch": 2.4096609382257315,
+      "grad_norm": 0.17331285320106854,
+      "learning_rate": 1.0912220309810673e-05,
+      "loss": 0.251,
+      "step": 2595
+    },
+    {
+      "epoch": 2.4105898745935903,
+      "grad_norm": 0.1739226946282106,
+      "learning_rate": 1.089500860585198e-05,
+      "loss": 0.2883,
+      "step": 2596
+    },
+    {
+      "epoch": 2.411518810961449,
+      "grad_norm": 0.19012644739523568,
+      "learning_rate": 1.0877796901893288e-05,
+      "loss": 0.2806,
+      "step": 2597
+    },
+    {
+      "epoch": 2.412447747329308,
+      "grad_norm": 0.18129586130288267,
+      "learning_rate": 1.0860585197934596e-05,
+      "loss": 0.2769,
+      "step": 2598
+    },
+    {
+      "epoch": 2.413376683697167,
+      "grad_norm": 0.17128766688292568,
+      "learning_rate": 1.0843373493975904e-05,
+      "loss": 0.2725,
+      "step": 2599
+    },
+    {
+      "epoch": 2.4143056200650257,
+      "grad_norm": 0.1723024173039281,
+      "learning_rate": 1.0826161790017213e-05,
+      "loss": 0.2712,
+      "step": 2600
+    },
+    {
+      "epoch": 2.4152345564328845,
+      "grad_norm": 0.18914686577326056,
+      "learning_rate": 1.080895008605852e-05,
+      "loss": 0.2823,
+      "step": 2601
+    },
+    {
+      "epoch": 2.4161634928007434,
+      "grad_norm": 0.17962076856198506,
+      "learning_rate": 1.0791738382099828e-05,
+      "loss": 0.2738,
+      "step": 2602
+    },
+    {
+      "epoch": 2.417092429168602,
+      "grad_norm": 0.18223450453434567,
+      "learning_rate": 1.0774526678141136e-05,
+      "loss": 0.2817,
+      "step": 2603
+    },
+    {
+      "epoch": 2.4180213655364606,
+      "grad_norm": 0.17204461955432634,
+      "learning_rate": 1.0757314974182445e-05,
+      "loss": 0.2784,
+      "step": 2604
+    },
+    {
+      "epoch": 2.4189503019043195,
+      "grad_norm": 0.18734269004984067,
+      "learning_rate": 1.0740103270223753e-05,
+      "loss": 0.2798,
+      "step": 2605
+    },
+    {
+      "epoch": 2.4198792382721783,
+      "grad_norm": 0.18627841159310524,
+      "learning_rate": 1.0722891566265062e-05,
+      "loss": 0.2736,
+      "step": 2606
+    },
+    {
+      "epoch": 2.420808174640037,
+      "grad_norm": 0.18288440859506122,
+      "learning_rate": 1.0705679862306368e-05,
+      "loss": 0.2762,
+      "step": 2607
+    },
+    {
+      "epoch": 2.421737111007896,
+      "grad_norm": 0.16671344490112985,
+      "learning_rate": 1.0688468158347677e-05,
+      "loss": 0.2583,
+      "step": 2608
+    },
+    {
+      "epoch": 2.422666047375755,
+      "grad_norm": 0.16774244865276433,
+      "learning_rate": 1.0671256454388985e-05,
+      "loss": 0.2667,
+      "step": 2609
+    },
+    {
+      "epoch": 2.4235949837436137,
+      "grad_norm": 0.1906502737894114,
+      "learning_rate": 1.0654044750430293e-05,
+      "loss": 0.2729,
+      "step": 2610
+    },
+    {
+      "epoch": 2.4245239201114726,
+      "grad_norm": 0.19139371036181707,
+      "learning_rate": 1.0636833046471602e-05,
+      "loss": 0.2917,
+      "step": 2611
+    },
+    {
+      "epoch": 2.425452856479331,
+      "grad_norm": 0.1830525571294439,
+      "learning_rate": 1.0619621342512909e-05,
+      "loss": 0.273,
+      "step": 2612
+    },
+    {
+      "epoch": 2.42638179284719,
+      "grad_norm": 0.17846541802522353,
+      "learning_rate": 1.0602409638554217e-05,
+      "loss": 0.2697,
+      "step": 2613
+    },
+    {
+      "epoch": 2.4273107292150486,
+      "grad_norm": 0.1912546863200559,
+      "learning_rate": 1.0585197934595525e-05,
+      "loss": 0.2614,
+      "step": 2614
+    },
+    {
+      "epoch": 2.4282396655829075,
+      "grad_norm": 0.18087505849060975,
+      "learning_rate": 1.0567986230636834e-05,
+      "loss": 0.2615,
+      "step": 2615
+    },
+    {
+      "epoch": 2.4291686019507663,
+      "grad_norm": 0.18672321455235916,
+      "learning_rate": 1.0550774526678142e-05,
+      "loss": 0.2829,
+      "step": 2616
+    },
+    {
+      "epoch": 2.430097538318625,
+      "grad_norm": 0.17819321647845873,
+      "learning_rate": 1.0533562822719449e-05,
+      "loss": 0.2781,
+      "step": 2617
+    },
+    {
+      "epoch": 2.431026474686484,
+      "grad_norm": 0.1802526099580208,
+      "learning_rate": 1.0516351118760757e-05,
+      "loss": 0.2799,
+      "step": 2618
+    },
+    {
+      "epoch": 2.431955411054343,
+      "grad_norm": 0.1732325664289292,
+      "learning_rate": 1.0499139414802066e-05,
+      "loss": 0.267,
+      "step": 2619
+    },
+    {
+      "epoch": 2.4328843474222017,
+      "grad_norm": 0.1862818608754717,
+      "learning_rate": 1.0481927710843374e-05,
+      "loss": 0.257,
+      "step": 2620
+    },
+    {
+      "epoch": 2.4338132837900606,
+      "grad_norm": 0.19978393037967157,
+      "learning_rate": 1.0464716006884682e-05,
+      "loss": 0.2982,
+      "step": 2621
+    },
+    {
+      "epoch": 2.4347422201579194,
+      "grad_norm": 0.16756039896886504,
+      "learning_rate": 1.0447504302925991e-05,
+      "loss": 0.2713,
+      "step": 2622
+    },
+    {
+      "epoch": 2.435671156525778,
+      "grad_norm": 0.1812223551466504,
+      "learning_rate": 1.0430292598967298e-05,
+      "loss": 0.271,
+      "step": 2623
+    },
+    {
+      "epoch": 2.4366000928936367,
+      "grad_norm": 0.18945960502714276,
+      "learning_rate": 1.0413080895008606e-05,
+      "loss": 0.2799,
+      "step": 2624
+    },
+    {
+      "epoch": 2.4375290292614955,
+      "grad_norm": 0.17519311685029668,
+      "learning_rate": 1.0395869191049914e-05,
+      "loss": 0.2765,
+      "step": 2625
+    },
+    {
+      "epoch": 2.4384579656293544,
+      "grad_norm": 0.18324447068348415,
+      "learning_rate": 1.0378657487091223e-05,
+      "loss": 0.2762,
+      "step": 2626
+    },
+    {
+      "epoch": 2.439386901997213,
+      "grad_norm": 0.17760314979609604,
+      "learning_rate": 1.0361445783132531e-05,
+      "loss": 0.2686,
+      "step": 2627
+    },
+    {
+      "epoch": 2.440315838365072,
+      "grad_norm": 0.17054771582356815,
+      "learning_rate": 1.0344234079173838e-05,
+      "loss": 0.2615,
+      "step": 2628
+    },
+    {
+      "epoch": 2.441244774732931,
+      "grad_norm": 0.1826470589868783,
+      "learning_rate": 1.0327022375215146e-05,
+      "loss": 0.2942,
+      "step": 2629
+    },
+    {
+      "epoch": 2.4421737111007897,
+      "grad_norm": 0.1949199675365578,
+      "learning_rate": 1.0309810671256455e-05,
+      "loss": 0.2861,
+      "step": 2630
+    },
+    {
+      "epoch": 2.4431026474686486,
+      "grad_norm": 0.1960583056723856,
+      "learning_rate": 1.0292598967297763e-05,
+      "loss": 0.272,
+      "step": 2631
+    },
+    {
+      "epoch": 2.444031583836507,
+      "grad_norm": 0.17233454553942215,
+      "learning_rate": 1.0275387263339071e-05,
+      "loss": 0.2778,
+      "step": 2632
+    },
+    {
+      "epoch": 2.444960520204366,
+      "grad_norm": 0.1954500467677132,
+      "learning_rate": 1.0258175559380378e-05,
+      "loss": 0.2769,
+      "step": 2633
+    },
+    {
+      "epoch": 2.4458894565722247,
+      "grad_norm": 0.1833072097746944,
+      "learning_rate": 1.0240963855421687e-05,
+      "loss": 0.2931,
+      "step": 2634
+    },
+    {
+      "epoch": 2.4468183929400835,
+      "grad_norm": 0.18965157114105274,
+      "learning_rate": 1.0223752151462995e-05,
+      "loss": 0.2877,
+      "step": 2635
+    },
+    {
+      "epoch": 2.4477473293079424,
+      "grad_norm": 0.20978733790727866,
+      "learning_rate": 1.0206540447504303e-05,
+      "loss": 0.2854,
+      "step": 2636
+    },
+    {
+      "epoch": 2.448676265675801,
+      "grad_norm": 0.17423882357451637,
+      "learning_rate": 1.0189328743545612e-05,
+      "loss": 0.2801,
+      "step": 2637
+    },
+    {
+      "epoch": 2.44960520204366,
+      "grad_norm": 0.17370390492230156,
+      "learning_rate": 1.0172117039586919e-05,
+      "loss": 0.2756,
+      "step": 2638
+    },
+    {
+      "epoch": 2.450534138411519,
+      "grad_norm": 0.16986015376541683,
+      "learning_rate": 1.0154905335628227e-05,
+      "loss": 0.2684,
+      "step": 2639
+    },
+    {
+      "epoch": 2.4514630747793777,
+      "grad_norm": 0.18055842985320075,
+      "learning_rate": 1.0137693631669535e-05,
+      "loss": 0.2691,
+      "step": 2640
+    },
+    {
+      "epoch": 2.4523920111472366,
+      "grad_norm": 0.1879144310139025,
+      "learning_rate": 1.0120481927710844e-05,
+      "loss": 0.2829,
+      "step": 2641
+    },
+    {
+      "epoch": 2.4533209475150954,
+      "grad_norm": 0.16956291193537515,
+      "learning_rate": 1.0103270223752152e-05,
+      "loss": 0.2836,
+      "step": 2642
+    },
+    {
+      "epoch": 2.454249883882954,
+      "grad_norm": 0.17981281771166743,
+      "learning_rate": 1.008605851979346e-05,
+      "loss": 0.2885,
+      "step": 2643
+    },
+    {
+      "epoch": 2.4551788202508127,
+      "grad_norm": 0.17947003101922124,
+      "learning_rate": 1.0068846815834767e-05,
+      "loss": 0.2693,
+      "step": 2644
+    },
+    {
+      "epoch": 2.4561077566186715,
+      "grad_norm": 0.1940400400563871,
+      "learning_rate": 1.0051635111876076e-05,
+      "loss": 0.2846,
+      "step": 2645
+    },
+    {
+      "epoch": 2.4570366929865304,
+      "grad_norm": 0.18487882046134174,
+      "learning_rate": 1.0034423407917384e-05,
+      "loss": 0.2577,
+      "step": 2646
+    },
+    {
+      "epoch": 2.4579656293543892,
+      "grad_norm": 0.17290264600688812,
+      "learning_rate": 1.0017211703958692e-05,
+      "loss": 0.2787,
+      "step": 2647
+    },
+    {
+      "epoch": 2.458894565722248,
+      "grad_norm": 0.17023515664022007,
+      "learning_rate": 1e-05,
+      "loss": 0.2673,
+      "step": 2648
+    },
+    {
+      "epoch": 2.459823502090107,
+      "grad_norm": 0.1776155129313746,
+      "learning_rate": 9.982788296041308e-06,
+      "loss": 0.2759,
+      "step": 2649
+    },
+    {
+      "epoch": 2.4607524384579658,
+      "grad_norm": 0.1922956087310196,
+      "learning_rate": 9.965576592082616e-06,
+      "loss": 0.2784,
+      "step": 2650
+    },
+    {
+      "epoch": 2.4616813748258246,
+      "grad_norm": 0.19175819536969194,
+      "learning_rate": 9.948364888123924e-06,
+      "loss": 0.271,
+      "step": 2651
+    },
+    {
+      "epoch": 2.462610311193683,
+      "grad_norm": 0.17557011388898683,
+      "learning_rate": 9.931153184165233e-06,
+      "loss": 0.2803,
+      "step": 2652
+    },
+    {
+      "epoch": 2.463539247561542,
+      "grad_norm": 0.17863208746544765,
+      "learning_rate": 9.913941480206541e-06,
+      "loss": 0.2717,
+      "step": 2653
+    },
+    {
+      "epoch": 2.4644681839294007,
+      "grad_norm": 0.18402079757460876,
+      "learning_rate": 9.896729776247848e-06,
+      "loss": 0.2705,
+      "step": 2654
+    },
+    {
+      "epoch": 2.4653971202972595,
+      "grad_norm": 0.18480632945427394,
+      "learning_rate": 9.879518072289156e-06,
+      "loss": 0.2939,
+      "step": 2655
+    },
+    {
+      "epoch": 2.4663260566651184,
+      "grad_norm": 0.1699533541431954,
+      "learning_rate": 9.862306368330465e-06,
+      "loss": 0.2773,
+      "step": 2656
+    },
+    {
+      "epoch": 2.4672549930329772,
+      "grad_norm": 0.17129023030549725,
+      "learning_rate": 9.845094664371773e-06,
+      "loss": 0.2748,
+      "step": 2657
+    },
+    {
+      "epoch": 2.468183929400836,
+      "grad_norm": 0.17801465299435518,
+      "learning_rate": 9.827882960413081e-06,
+      "loss": 0.2862,
+      "step": 2658
+    },
+    {
+      "epoch": 2.469112865768695,
+      "grad_norm": 0.16935169755624885,
+      "learning_rate": 9.81067125645439e-06,
+      "loss": 0.2682,
+      "step": 2659
+    },
+    {
+      "epoch": 2.4700418021365538,
+      "grad_norm": 0.2000237770209295,
+      "learning_rate": 9.793459552495697e-06,
+      "loss": 0.2813,
+      "step": 2660
+    },
+    {
+      "epoch": 2.4709707385044126,
+      "grad_norm": 0.18784791548400018,
+      "learning_rate": 9.776247848537005e-06,
+      "loss": 0.2829,
+      "step": 2661
+    },
+    {
+      "epoch": 2.4718996748722715,
+      "grad_norm": 0.17733443093063525,
+      "learning_rate": 9.759036144578313e-06,
+      "loss": 0.278,
+      "step": 2662
+    },
+    {
+      "epoch": 2.47282861124013,
+      "grad_norm": 0.17325580348112374,
+      "learning_rate": 9.741824440619622e-06,
+      "loss": 0.2773,
+      "step": 2663
+    },
+    {
+      "epoch": 2.4737575476079887,
+      "grad_norm": 0.17758599669012587,
+      "learning_rate": 9.72461273666093e-06,
+      "loss": 0.2691,
+      "step": 2664
+    },
+    {
+      "epoch": 2.4746864839758476,
+      "grad_norm": 0.2084107357635697,
+      "learning_rate": 9.707401032702237e-06,
+      "loss": 0.2772,
+      "step": 2665
+    },
+    {
+      "epoch": 2.4756154203437064,
+      "grad_norm": 0.18325821857011707,
+      "learning_rate": 9.690189328743545e-06,
+      "loss": 0.2731,
+      "step": 2666
+    },
+    {
+      "epoch": 2.4765443567115653,
+      "grad_norm": 0.18181226539042997,
+      "learning_rate": 9.672977624784854e-06,
+      "loss": 0.2745,
+      "step": 2667
+    },
+    {
+      "epoch": 2.477473293079424,
+      "grad_norm": 0.18724076787159039,
+      "learning_rate": 9.655765920826162e-06,
+      "loss": 0.2882,
+      "step": 2668
+    },
+    {
+      "epoch": 2.478402229447283,
+      "grad_norm": 0.19008040074552002,
+      "learning_rate": 9.63855421686747e-06,
+      "loss": 0.296,
+      "step": 2669
+    },
+    {
+      "epoch": 2.479331165815142,
+      "grad_norm": 0.1683701319992699,
+      "learning_rate": 9.621342512908777e-06,
+      "loss": 0.2805,
+      "step": 2670
+    },
+    {
+      "epoch": 2.4802601021830006,
+      "grad_norm": 0.17949764929549564,
+      "learning_rate": 9.604130808950086e-06,
+      "loss": 0.2715,
+      "step": 2671
+    },
+    {
+      "epoch": 2.481189038550859,
+      "grad_norm": 0.18184976004626205,
+      "learning_rate": 9.586919104991394e-06,
+      "loss": 0.2773,
+      "step": 2672
+    },
+    {
+      "epoch": 2.482117974918718,
+      "grad_norm": 0.17775807892598655,
+      "learning_rate": 9.569707401032702e-06,
+      "loss": 0.283,
+      "step": 2673
+    },
+    {
+      "epoch": 2.4830469112865767,
+      "grad_norm": 0.18923652551120143,
+      "learning_rate": 9.55249569707401e-06,
+      "loss": 0.2845,
+      "step": 2674
+    },
+    {
+      "epoch": 2.4839758476544356,
+      "grad_norm": 0.17906213434493118,
+      "learning_rate": 9.535283993115319e-06,
+      "loss": 0.2729,
+      "step": 2675
+    },
+    {
+      "epoch": 2.4849047840222944,
+      "grad_norm": 0.18853045947261501,
+      "learning_rate": 9.518072289156626e-06,
+      "loss": 0.2749,
+      "step": 2676
+    },
+    {
+      "epoch": 2.4858337203901533,
+      "grad_norm": 0.1725220515656567,
+      "learning_rate": 9.500860585197934e-06,
+      "loss": 0.2556,
+      "step": 2677
+    },
+    {
+      "epoch": 2.486762656758012,
+      "grad_norm": 0.18222882747239028,
+      "learning_rate": 9.483648881239243e-06,
+      "loss": 0.2761,
+      "step": 2678
+    },
+    {
+      "epoch": 2.487691593125871,
+      "grad_norm": 0.18770105080124452,
+      "learning_rate": 9.466437177280551e-06,
+      "loss": 0.2759,
+      "step": 2679
+    },
+    {
+      "epoch": 2.48862052949373,
+      "grad_norm": 0.1797766621607397,
+      "learning_rate": 9.44922547332186e-06,
+      "loss": 0.2931,
+      "step": 2680
+    },
+    {
+      "epoch": 2.4895494658615887,
+      "grad_norm": 0.1842354032637381,
+      "learning_rate": 9.432013769363166e-06,
+      "loss": 0.2826,
+      "step": 2681
+    },
+    {
+      "epoch": 2.4904784022294475,
+      "grad_norm": 0.164114716412252,
+      "learning_rate": 9.414802065404475e-06,
+      "loss": 0.2659,
+      "step": 2682
+    },
+    {
+      "epoch": 2.491407338597306,
+      "grad_norm": 0.16405362051068478,
+      "learning_rate": 9.397590361445783e-06,
+      "loss": 0.2658,
+      "step": 2683
+    },
+    {
+      "epoch": 2.4923362749651647,
+      "grad_norm": 0.17956863474404955,
+      "learning_rate": 9.380378657487091e-06,
+      "loss": 0.2781,
+      "step": 2684
+    },
+    {
+      "epoch": 2.4932652113330236,
+      "grad_norm": 0.17737289818146046,
+      "learning_rate": 9.3631669535284e-06,
+      "loss": 0.296,
+      "step": 2685
+    },
+    {
+      "epoch": 2.4941941477008824,
+      "grad_norm": 0.18585541574048775,
+      "learning_rate": 9.345955249569706e-06,
+      "loss": 0.3013,
+      "step": 2686
+    },
+    {
+      "epoch": 2.4951230840687413,
+      "grad_norm": 0.1643243554438021,
+      "learning_rate": 9.328743545611015e-06,
+      "loss": 0.2623,
+      "step": 2687
+    },
+    {
+      "epoch": 2.4960520204366,
+      "grad_norm": 0.17270213602505902,
+      "learning_rate": 9.311531841652323e-06,
+      "loss": 0.2766,
+      "step": 2688
+    },
+    {
+      "epoch": 2.496980956804459,
+      "grad_norm": 0.16782827955554608,
+      "learning_rate": 9.294320137693632e-06,
+      "loss": 0.2665,
+      "step": 2689
+    },
+    {
+      "epoch": 2.497909893172318,
+      "grad_norm": 0.17122748158137155,
+      "learning_rate": 9.27710843373494e-06,
+      "loss": 0.2705,
+      "step": 2690
+    },
+    {
+      "epoch": 2.4988388295401767,
+      "grad_norm": 0.17869194306118438,
+      "learning_rate": 9.259896729776247e-06,
+      "loss": 0.281,
+      "step": 2691
+    },
+    {
+      "epoch": 2.499767765908035,
+      "grad_norm": 0.1695879134783124,
+      "learning_rate": 9.242685025817555e-06,
+      "loss": 0.2537,
+      "step": 2692
+    },
+    {
+      "epoch": 2.500696702275894,
+      "grad_norm": 0.1662584266248051,
+      "learning_rate": 9.225473321858864e-06,
+      "loss": 0.2632,
+      "step": 2693
+    },
+    {
+      "epoch": 2.5016256386437528,
+      "grad_norm": 0.17093570405435482,
+      "learning_rate": 9.208261617900172e-06,
+      "loss": 0.2792,
+      "step": 2694
+    },
+    {
+      "epoch": 2.5025545750116116,
+      "grad_norm": 0.17307510808065985,
+      "learning_rate": 9.19104991394148e-06,
+      "loss": 0.2769,
+      "step": 2695
+    },
+    {
+      "epoch": 2.5034835113794704,
+      "grad_norm": 0.17692551001812182,
+      "learning_rate": 9.173838209982789e-06,
+      "loss": 0.2648,
+      "step": 2696
+    },
+    {
+      "epoch": 2.5044124477473293,
+      "grad_norm": 0.17879082159343984,
+      "learning_rate": 9.156626506024097e-06,
+      "loss": 0.2763,
+      "step": 2697
+    },
+    {
+      "epoch": 2.505341384115188,
+      "grad_norm": 0.19597822806679713,
+      "learning_rate": 9.139414802065404e-06,
+      "loss": 0.2869,
+      "step": 2698
+    },
+    {
+      "epoch": 2.506270320483047,
+      "grad_norm": 0.17477995496206758,
+      "learning_rate": 9.122203098106712e-06,
+      "loss": 0.2787,
+      "step": 2699
+    },
+    {
+      "epoch": 2.507199256850906,
+      "grad_norm": 0.19752698933537774,
+      "learning_rate": 9.10499139414802e-06,
+      "loss": 0.2796,
+      "step": 2700
+    },
+    {
+      "epoch": 2.5081281932187647,
+      "grad_norm": 0.19121547602167308,
+      "learning_rate": 9.087779690189329e-06,
+      "loss": 0.2763,
+      "step": 2701
+    },
+    {
+      "epoch": 2.5090571295866235,
+      "grad_norm": 0.19101906802675103,
+      "learning_rate": 9.070567986230637e-06,
+      "loss": 0.3006,
+      "step": 2702
+    },
+    {
+      "epoch": 2.5099860659544824,
+      "grad_norm": 0.17615835392415316,
+      "learning_rate": 9.053356282271946e-06,
+      "loss": 0.2846,
+      "step": 2703
+    },
+    {
+      "epoch": 2.5109150023223408,
+      "grad_norm": 0.1762519374159477,
+      "learning_rate": 9.036144578313253e-06,
+      "loss": 0.2693,
+      "step": 2704
+    },
+    {
+      "epoch": 2.5118439386901996,
+      "grad_norm": 0.20903449277468536,
+      "learning_rate": 9.018932874354561e-06,
+      "loss": 0.2829,
+      "step": 2705
+    },
+    {
+      "epoch": 2.5127728750580585,
+      "grad_norm": 0.17834560785229894,
+      "learning_rate": 9.00172117039587e-06,
+      "loss": 0.2674,
+      "step": 2706
+    },
+    {
+      "epoch": 2.5137018114259173,
+      "grad_norm": 0.1793333142413588,
+      "learning_rate": 8.984509466437178e-06,
+      "loss": 0.2592,
+      "step": 2707
+    },
+    {
+      "epoch": 2.514630747793776,
+      "grad_norm": 0.18220038218018145,
+      "learning_rate": 8.967297762478486e-06,
+      "loss": 0.2738,
+      "step": 2708
+    },
+    {
+      "epoch": 2.515559684161635,
+      "grad_norm": 0.1767274685035216,
+      "learning_rate": 8.950086058519795e-06,
+      "loss": 0.2777,
+      "step": 2709
+    },
+    {
+      "epoch": 2.516488620529494,
+      "grad_norm": 0.1838314403491137,
+      "learning_rate": 8.932874354561101e-06,
+      "loss": 0.2811,
+      "step": 2710
+    },
+    {
+      "epoch": 2.5174175568973527,
+      "grad_norm": 0.2000434374752791,
+      "learning_rate": 8.91566265060241e-06,
+      "loss": 0.2744,
+      "step": 2711
+    },
+    {
+      "epoch": 2.518346493265211,
+      "grad_norm": 0.1821868758548907,
+      "learning_rate": 8.898450946643718e-06,
+      "loss": 0.2768,
+      "step": 2712
+    },
+    {
+      "epoch": 2.51927542963307,
+      "grad_norm": 0.29230691057904484,
+      "learning_rate": 8.881239242685026e-06,
+      "loss": 0.2734,
+      "step": 2713
+    },
+    {
+      "epoch": 2.520204366000929,
+      "grad_norm": 0.167813313251146,
+      "learning_rate": 8.864027538726335e-06,
+      "loss": 0.2601,
+      "step": 2714
+    },
+    {
+      "epoch": 2.5211333023687876,
+      "grad_norm": 0.1769085034080416,
+      "learning_rate": 8.846815834767643e-06,
+      "loss": 0.2726,
+      "step": 2715
+    },
+    {
+      "epoch": 2.5220622387366465,
+      "grad_norm": 0.20220842083808388,
+      "learning_rate": 8.82960413080895e-06,
+      "loss": 0.2817,
+      "step": 2716
+    },
+    {
+      "epoch": 2.5229911751045053,
+      "grad_norm": 0.18444616770204614,
+      "learning_rate": 8.812392426850258e-06,
+      "loss": 0.2773,
+      "step": 2717
+    },
+    {
+      "epoch": 2.523920111472364,
+      "grad_norm": 0.18296071897423907,
+      "learning_rate": 8.795180722891567e-06,
+      "loss": 0.2797,
+      "step": 2718
+    },
+    {
+      "epoch": 2.524849047840223,
+      "grad_norm": 0.1817067050377542,
+      "learning_rate": 8.777969018932875e-06,
+      "loss": 0.2641,
+      "step": 2719
+    },
+    {
+      "epoch": 2.525777984208082,
+      "grad_norm": 0.1794572761384637,
+      "learning_rate": 8.760757314974184e-06,
+      "loss": 0.2813,
+      "step": 2720
+    },
+    {
+      "epoch": 2.5267069205759407,
+      "grad_norm": 0.1676740839883334,
+      "learning_rate": 8.743545611015492e-06,
+      "loss": 0.2808,
+      "step": 2721
+    },
+    {
+      "epoch": 2.5276358569437996,
+      "grad_norm": 0.17116798658772298,
+      "learning_rate": 8.726333907056799e-06,
+      "loss": 0.2778,
+      "step": 2722
+    },
+    {
+      "epoch": 2.5285647933116584,
+      "grad_norm": 0.17943680035895654,
+      "learning_rate": 8.709122203098107e-06,
+      "loss": 0.2713,
+      "step": 2723
+    },
+    {
+      "epoch": 2.529493729679517,
+      "grad_norm": 0.18500653140503157,
+      "learning_rate": 8.691910499139416e-06,
+      "loss": 0.2748,
+      "step": 2724
+    },
+    {
+      "epoch": 2.5304226660473756,
+      "grad_norm": 0.18656247536015494,
+      "learning_rate": 8.674698795180724e-06,
+      "loss": 0.2931,
+      "step": 2725
+    },
+    {
+      "epoch": 2.5313516024152345,
+      "grad_norm": 0.17519545206291068,
+      "learning_rate": 8.657487091222032e-06,
+      "loss": 0.2636,
+      "step": 2726
+    },
+    {
+      "epoch": 2.5322805387830933,
+      "grad_norm": 0.18579678715189885,
+      "learning_rate": 8.64027538726334e-06,
+      "loss": 0.2609,
+      "step": 2727
+    },
+    {
+      "epoch": 2.533209475150952,
+      "grad_norm": 0.17685544278047252,
+      "learning_rate": 8.623063683304647e-06,
+      "loss": 0.269,
+      "step": 2728
+    },
+    {
+      "epoch": 2.534138411518811,
+      "grad_norm": 0.1716684760379682,
+      "learning_rate": 8.605851979345956e-06,
+      "loss": 0.273,
+      "step": 2729
+    },
+    {
+      "epoch": 2.53506734788667,
+      "grad_norm": 0.18997722560178743,
+      "learning_rate": 8.588640275387264e-06,
+      "loss": 0.2735,
+      "step": 2730
+    },
+    {
+      "epoch": 2.5359962842545287,
+      "grad_norm": 0.1725056612650081,
+      "learning_rate": 8.571428571428573e-06,
+      "loss": 0.2782,
+      "step": 2731
+    },
+    {
+      "epoch": 2.536925220622387,
+      "grad_norm": 0.1846867810643451,
+      "learning_rate": 8.554216867469881e-06,
+      "loss": 0.281,
+      "step": 2732
+    },
+    {
+      "epoch": 2.537854156990246,
+      "grad_norm": 0.1746259889225677,
+      "learning_rate": 8.53700516351119e-06,
+      "loss": 0.2802,
+      "step": 2733
+    },
+    {
+      "epoch": 2.538783093358105,
+      "grad_norm": 0.18034001099175637,
+      "learning_rate": 8.519793459552496e-06,
+      "loss": 0.2983,
+      "step": 2734
+    },
+    {
+      "epoch": 2.5397120297259637,
+      "grad_norm": 0.17285283661416923,
+      "learning_rate": 8.502581755593805e-06,
+      "loss": 0.267,
+      "step": 2735
+    },
+    {
+      "epoch": 2.5406409660938225,
+      "grad_norm": 0.16623526260482924,
+      "learning_rate": 8.485370051635113e-06,
+      "loss": 0.2714,
+      "step": 2736
+    },
+    {
+      "epoch": 2.5415699024616814,
+      "grad_norm": 0.17140669097240416,
+      "learning_rate": 8.468158347676421e-06,
+      "loss": 0.2764,
+      "step": 2737
+    },
+    {
+      "epoch": 2.54249883882954,
+      "grad_norm": 0.17006832454927703,
+      "learning_rate": 8.45094664371773e-06,
+      "loss": 0.2911,
+      "step": 2738
+    },
+    {
+      "epoch": 2.543427775197399,
+      "grad_norm": 0.16385080260835502,
+      "learning_rate": 8.433734939759036e-06,
+      "loss": 0.2677,
+      "step": 2739
+    },
+    {
+      "epoch": 2.544356711565258,
+      "grad_norm": 0.1798411378111758,
+      "learning_rate": 8.416523235800345e-06,
+      "loss": 0.2876,
+      "step": 2740
+    },
+    {
+      "epoch": 2.5452856479331167,
+      "grad_norm": 0.17749376274284356,
+      "learning_rate": 8.399311531841653e-06,
+      "loss": 0.2809,
+      "step": 2741
+    },
+    {
+      "epoch": 2.5462145843009756,
+      "grad_norm": 0.17811015681571632,
+      "learning_rate": 8.382099827882962e-06,
+      "loss": 0.2758,
+      "step": 2742
+    },
+    {
+      "epoch": 2.5471435206688344,
+      "grad_norm": 0.1791864260525822,
+      "learning_rate": 8.36488812392427e-06,
+      "loss": 0.2868,
+      "step": 2743
+    },
+    {
+      "epoch": 2.548072457036693,
+      "grad_norm": 0.18193816987523823,
+      "learning_rate": 8.347676419965577e-06,
+      "loss": 0.2759,
+      "step": 2744
+    },
+    {
+      "epoch": 2.5490013934045517,
+      "grad_norm": 0.18066533942488144,
+      "learning_rate": 8.330464716006885e-06,
+      "loss": 0.2671,
+      "step": 2745
+    },
+    {
+      "epoch": 2.5499303297724105,
+      "grad_norm": 0.17448722669503738,
+      "learning_rate": 8.313253012048194e-06,
+      "loss": 0.2747,
+      "step": 2746
+    },
+    {
+      "epoch": 2.5508592661402694,
+      "grad_norm": 0.1818779797983094,
+      "learning_rate": 8.296041308089502e-06,
+      "loss": 0.2836,
+      "step": 2747
+    },
+    {
+      "epoch": 2.551788202508128,
+      "grad_norm": 0.1802872411024581,
+      "learning_rate": 8.27882960413081e-06,
+      "loss": 0.2762,
+      "step": 2748
+    },
+    {
+      "epoch": 2.552717138875987,
+      "grad_norm": 0.17919426431481447,
+      "learning_rate": 8.261617900172119e-06,
+      "loss": 0.2844,
+      "step": 2749
+    },
+    {
+      "epoch": 2.553646075243846,
+      "grad_norm": 0.17435336893217318,
+      "learning_rate": 8.244406196213425e-06,
+      "loss": 0.2803,
+      "step": 2750
+    },
+    {
+      "epoch": 2.5545750116117047,
+      "grad_norm": 0.19650206355264208,
+      "learning_rate": 8.227194492254734e-06,
+      "loss": 0.2995,
+      "step": 2751
+    },
+    {
+      "epoch": 2.555503947979563,
+      "grad_norm": 0.1751970358659085,
+      "learning_rate": 8.209982788296042e-06,
+      "loss": 0.272,
+      "step": 2752
+    },
+    {
+      "epoch": 2.556432884347422,
+      "grad_norm": 0.1690503222258136,
+      "learning_rate": 8.19277108433735e-06,
+      "loss": 0.2739,
+      "step": 2753
+    },
+    {
+      "epoch": 2.557361820715281,
+      "grad_norm": 0.17620039647154925,
+      "learning_rate": 8.175559380378659e-06,
+      "loss": 0.2805,
+      "step": 2754
+    },
+    {
+      "epoch": 2.5582907570831397,
+      "grad_norm": 0.17388899053880985,
+      "learning_rate": 8.158347676419966e-06,
+      "loss": 0.2704,
+      "step": 2755
+    },
+    {
+      "epoch": 2.5592196934509985,
+      "grad_norm": 0.17685736263136762,
+      "learning_rate": 8.141135972461274e-06,
+      "loss": 0.2649,
+      "step": 2756
+    },
+    {
+      "epoch": 2.5601486298188574,
+      "grad_norm": 0.17315801191584784,
+      "learning_rate": 8.123924268502583e-06,
+      "loss": 0.2783,
+      "step": 2757
+    },
+    {
+      "epoch": 2.5610775661867162,
+      "grad_norm": 0.17710989290259285,
+      "learning_rate": 8.106712564543891e-06,
+      "loss": 0.2833,
+      "step": 2758
+    },
+    {
+      "epoch": 2.562006502554575,
+      "grad_norm": 0.17693674550109328,
+      "learning_rate": 8.0895008605852e-06,
+      "loss": 0.28,
+      "step": 2759
+    },
+    {
+      "epoch": 2.562935438922434,
+      "grad_norm": 0.18006819440338806,
+      "learning_rate": 8.072289156626506e-06,
+      "loss": 0.2793,
+      "step": 2760
+    },
+    {
+      "epoch": 2.5638643752902928,
+      "grad_norm": 0.17402028938476224,
+      "learning_rate": 8.055077452667814e-06,
+      "loss": 0.2806,
+      "step": 2761
+    },
+    {
+      "epoch": 2.5647933116581516,
+      "grad_norm": 0.1756389111979832,
+      "learning_rate": 8.037865748709123e-06,
+      "loss": 0.2797,
+      "step": 2762
+    },
+    {
+      "epoch": 2.5657222480260105,
+      "grad_norm": 0.1798104235898651,
+      "learning_rate": 8.020654044750431e-06,
+      "loss": 0.2801,
+      "step": 2763
+    },
+    {
+      "epoch": 2.566651184393869,
+      "grad_norm": 0.17595147246541074,
+      "learning_rate": 8.00344234079174e-06,
+      "loss": 0.2796,
+      "step": 2764
+    },
+    {
+      "epoch": 2.5675801207617277,
+      "grad_norm": 0.1866139051974165,
+      "learning_rate": 7.986230636833048e-06,
+      "loss": 0.2792,
+      "step": 2765
+    },
+    {
+      "epoch": 2.5685090571295865,
+      "grad_norm": 0.17242021849415873,
+      "learning_rate": 7.969018932874355e-06,
+      "loss": 0.2799,
+      "step": 2766
+    },
+    {
+      "epoch": 2.5694379934974454,
+      "grad_norm": 0.1722563276561945,
+      "learning_rate": 7.951807228915663e-06,
+      "loss": 0.2658,
+      "step": 2767
+    },
+    {
+      "epoch": 2.5703669298653042,
+      "grad_norm": 0.18138158515978361,
+      "learning_rate": 7.934595524956972e-06,
+      "loss": 0.2937,
+      "step": 2768
+    },
+    {
+      "epoch": 2.571295866233163,
+      "grad_norm": 0.17876412039688724,
+      "learning_rate": 7.91738382099828e-06,
+      "loss": 0.2749,
+      "step": 2769
+    },
+    {
+      "epoch": 2.572224802601022,
+      "grad_norm": 0.17551798053405512,
+      "learning_rate": 7.900172117039588e-06,
+      "loss": 0.2755,
+      "step": 2770
+    },
+    {
+      "epoch": 2.5731537389688808,
+      "grad_norm": 0.17206297732377063,
+      "learning_rate": 7.882960413080895e-06,
+      "loss": 0.2729,
+      "step": 2771
+    },
+    {
+      "epoch": 2.574082675336739,
+      "grad_norm": 0.17868585578978832,
+      "learning_rate": 7.865748709122203e-06,
+      "loss": 0.2958,
+      "step": 2772
+    },
+    {
+      "epoch": 2.575011611704598,
+      "grad_norm": 0.17935135793930987,
+      "learning_rate": 7.848537005163512e-06,
+      "loss": 0.2679,
+      "step": 2773
+    },
+    {
+      "epoch": 2.575940548072457,
+      "grad_norm": 0.16765676035780458,
+      "learning_rate": 7.83132530120482e-06,
+      "loss": 0.2648,
+      "step": 2774
+    },
+    {
+      "epoch": 2.5768694844403157,
+      "grad_norm": 0.18548752107627353,
+      "learning_rate": 7.814113597246129e-06,
+      "loss": 0.2953,
+      "step": 2775
+    },
+    {
+      "epoch": 2.5777984208081746,
+      "grad_norm": 0.18625401386712584,
+      "learning_rate": 7.796901893287435e-06,
+      "loss": 0.2903,
+      "step": 2776
+    },
+    {
+      "epoch": 2.5787273571760334,
+      "grad_norm": 0.17797790691377569,
+      "learning_rate": 7.779690189328744e-06,
+      "loss": 0.2587,
+      "step": 2777
+    },
+    {
+      "epoch": 2.5796562935438923,
+      "grad_norm": 0.17169551543009043,
+      "learning_rate": 7.762478485370052e-06,
+      "loss": 0.2591,
+      "step": 2778
+    },
+    {
+      "epoch": 2.580585229911751,
+      "grad_norm": 0.17885288708688196,
+      "learning_rate": 7.74526678141136e-06,
+      "loss": 0.2644,
+      "step": 2779
+    },
+    {
+      "epoch": 2.58151416627961,
+      "grad_norm": 0.17569046883577338,
+      "learning_rate": 7.728055077452669e-06,
+      "loss": 0.2802,
+      "step": 2780
+    },
+    {
+      "epoch": 2.582443102647469,
+      "grad_norm": 0.18426157444373517,
+      "learning_rate": 7.710843373493977e-06,
+      "loss": 0.2734,
+      "step": 2781
+    },
+    {
+      "epoch": 2.5833720390153276,
+      "grad_norm": 0.18868415686038092,
+      "learning_rate": 7.693631669535284e-06,
+      "loss": 0.2862,
+      "step": 2782
+    },
+    {
+      "epoch": 2.5843009753831865,
+      "grad_norm": 0.1826628963868718,
+      "learning_rate": 7.676419965576592e-06,
+      "loss": 0.275,
+      "step": 2783
+    },
+    {
+      "epoch": 2.5852299117510453,
+      "grad_norm": 0.18305810292468144,
+      "learning_rate": 7.659208261617901e-06,
+      "loss": 0.2783,
+      "step": 2784
+    },
+    {
+      "epoch": 2.5861588481189037,
+      "grad_norm": 0.19042265177468026,
+      "learning_rate": 7.64199655765921e-06,
+      "loss": 0.2836,
+      "step": 2785
+    },
+    {
+      "epoch": 2.5870877844867626,
+      "grad_norm": 0.17182388584361694,
+      "learning_rate": 7.624784853700518e-06,
+      "loss": 0.2758,
+      "step": 2786
+    },
+    {
+      "epoch": 2.5880167208546214,
+      "grad_norm": 0.16171186896342973,
+      "learning_rate": 7.607573149741824e-06,
+      "loss": 0.27,
+      "step": 2787
+    },
+    {
+      "epoch": 2.5889456572224803,
+      "grad_norm": 0.16522516692804265,
+      "learning_rate": 7.590361445783133e-06,
+      "loss": 0.2518,
+      "step": 2788
+    },
+    {
+      "epoch": 2.589874593590339,
+      "grad_norm": 0.18331019909753513,
+      "learning_rate": 7.573149741824441e-06,
+      "loss": 0.2866,
+      "step": 2789
+    },
+    {
+      "epoch": 2.590803529958198,
+      "grad_norm": 0.18035348067205553,
+      "learning_rate": 7.55593803786575e-06,
+      "loss": 0.2613,
+      "step": 2790
+    },
+    {
+      "epoch": 2.591732466326057,
+      "grad_norm": 0.18295229566848326,
+      "learning_rate": 7.538726333907058e-06,
+      "loss": 0.272,
+      "step": 2791
+    },
+    {
+      "epoch": 2.592661402693915,
+      "grad_norm": 0.16902236952248076,
+      "learning_rate": 7.521514629948365e-06,
+      "loss": 0.2676,
+      "step": 2792
+    },
+    {
+      "epoch": 2.593590339061774,
+      "grad_norm": 0.18044289901142607,
+      "learning_rate": 7.504302925989673e-06,
+      "loss": 0.2907,
+      "step": 2793
+    },
+    {
+      "epoch": 2.594519275429633,
+      "grad_norm": 0.18658630020919675,
+      "learning_rate": 7.4870912220309815e-06,
+      "loss": 0.2898,
+      "step": 2794
+    },
+    {
+      "epoch": 2.5954482117974917,
+      "grad_norm": 0.1659203986129596,
+      "learning_rate": 7.46987951807229e-06,
+      "loss": 0.2767,
+      "step": 2795
+    },
+    {
+      "epoch": 2.5963771481653506,
+      "grad_norm": 0.18224464598969245,
+      "learning_rate": 7.452667814113598e-06,
+      "loss": 0.2744,
+      "step": 2796
+    },
+    {
+      "epoch": 2.5973060845332094,
+      "grad_norm": 0.18798345715683007,
+      "learning_rate": 7.435456110154905e-06,
+      "loss": 0.2803,
+      "step": 2797
+    },
+    {
+      "epoch": 2.5982350209010683,
+      "grad_norm": 0.17498384353384416,
+      "learning_rate": 7.418244406196213e-06,
+      "loss": 0.2752,
+      "step": 2798
+    },
+    {
+      "epoch": 2.599163957268927,
+      "grad_norm": 0.17390827410315354,
+      "learning_rate": 7.401032702237522e-06,
+      "loss": 0.2713,
+      "step": 2799
+    },
+    {
+      "epoch": 2.600092893636786,
+      "grad_norm": 0.17905924555277583,
+      "learning_rate": 7.38382099827883e-06,
+      "loss": 0.272,
+      "step": 2800
+    },
+    {
+      "epoch": 2.601021830004645,
+      "grad_norm": 0.19237045567028638,
+      "learning_rate": 7.366609294320139e-06,
+      "loss": 0.2754,
+      "step": 2801
+    },
+    {
+      "epoch": 2.6019507663725037,
+      "grad_norm": 0.19081730075922945,
+      "learning_rate": 7.349397590361447e-06,
+      "loss": 0.2784,
+      "step": 2802
+    },
+    {
+      "epoch": 2.6028797027403625,
+      "grad_norm": 0.1773815781690385,
+      "learning_rate": 7.332185886402754e-06,
+      "loss": 0.2812,
+      "step": 2803
+    },
+    {
+      "epoch": 2.6038086391082214,
+      "grad_norm": 0.18905163088336832,
+      "learning_rate": 7.314974182444062e-06,
+      "loss": 0.2798,
+      "step": 2804
+    },
+    {
+      "epoch": 2.6047375754760798,
+      "grad_norm": 0.17873348755643859,
+      "learning_rate": 7.2977624784853705e-06,
+      "loss": 0.2716,
+      "step": 2805
+    },
+    {
+      "epoch": 2.6056665118439386,
+      "grad_norm": 0.17030522579544236,
+      "learning_rate": 7.280550774526679e-06,
+      "loss": 0.2685,
+      "step": 2806
+    },
+    {
+      "epoch": 2.6065954482117974,
+      "grad_norm": 0.1862035284607427,
+      "learning_rate": 7.263339070567987e-06,
+      "loss": 0.2788,
+      "step": 2807
+    },
+    {
+      "epoch": 2.6075243845796563,
+      "grad_norm": 0.17337568922147115,
+      "learning_rate": 7.246127366609294e-06,
+      "loss": 0.2642,
+      "step": 2808
+    },
+    {
+      "epoch": 2.608453320947515,
+      "grad_norm": 0.19450967596793517,
+      "learning_rate": 7.228915662650602e-06,
+      "loss": 0.2727,
+      "step": 2809
+    },
+    {
+      "epoch": 2.609382257315374,
+      "grad_norm": 0.1893548613084817,
+      "learning_rate": 7.211703958691911e-06,
+      "loss": 0.2791,
+      "step": 2810
+    },
+    {
+      "epoch": 2.610311193683233,
+      "grad_norm": 0.18755670041525413,
+      "learning_rate": 7.194492254733219e-06,
+      "loss": 0.2893,
+      "step": 2811
+    },
+    {
+      "epoch": 2.6112401300510912,
+      "grad_norm": 0.1737293113823649,
+      "learning_rate": 7.177280550774528e-06,
+      "loss": 0.2654,
+      "step": 2812
+    },
+    {
+      "epoch": 2.61216906641895,
+      "grad_norm": 0.17980094681636927,
+      "learning_rate": 7.160068846815834e-06,
+      "loss": 0.2801,
+      "step": 2813
+    },
+    {
+      "epoch": 2.613098002786809,
+      "grad_norm": 0.17904359523084443,
+      "learning_rate": 7.142857142857143e-06,
+      "loss": 0.2767,
+      "step": 2814
+    },
+    {
+      "epoch": 2.6140269391546678,
+      "grad_norm": 0.17364837747203463,
+      "learning_rate": 7.125645438898451e-06,
+      "loss": 0.2702,
+      "step": 2815
+    },
+    {
+      "epoch": 2.6149558755225266,
+      "grad_norm": 0.17617804961029349,
+      "learning_rate": 7.1084337349397595e-06,
+      "loss": 0.2642,
+      "step": 2816
+    },
+    {
+      "epoch": 2.6158848118903855,
+      "grad_norm": 0.1730303401260583,
+      "learning_rate": 7.091222030981068e-06,
+      "loss": 0.2778,
+      "step": 2817
+    },
+    {
+      "epoch": 2.6168137482582443,
+      "grad_norm": 0.1831938804451245,
+      "learning_rate": 7.074010327022376e-06,
+      "loss": 0.2811,
+      "step": 2818
+    },
+    {
+      "epoch": 2.617742684626103,
+      "grad_norm": 0.1753374582449086,
+      "learning_rate": 7.056798623063683e-06,
+      "loss": 0.2717,
+      "step": 2819
+    },
+    {
+      "epoch": 2.618671620993962,
+      "grad_norm": 0.18343820977770026,
+      "learning_rate": 7.039586919104991e-06,
+      "loss": 0.3012,
+      "step": 2820
+    },
+    {
+      "epoch": 2.619600557361821,
+      "grad_norm": 0.18184849598170855,
+      "learning_rate": 7.0223752151463e-06,
+      "loss": 0.2697,
+      "step": 2821
+    },
+    {
+      "epoch": 2.6205294937296797,
+      "grad_norm": 0.167860177827926,
+      "learning_rate": 7.005163511187608e-06,
+      "loss": 0.2784,
+      "step": 2822
+    },
+    {
+      "epoch": 2.6214584300975385,
+      "grad_norm": 0.17080098742603222,
+      "learning_rate": 6.987951807228917e-06,
+      "loss": 0.2577,
+      "step": 2823
+    },
+    {
+      "epoch": 2.6223873664653974,
+      "grad_norm": 0.1883622438215358,
+      "learning_rate": 6.970740103270223e-06,
+      "loss": 0.2879,
+      "step": 2824
+    },
+    {
+      "epoch": 2.623316302833256,
+      "grad_norm": 0.1744115106332342,
+      "learning_rate": 6.953528399311532e-06,
+      "loss": 0.2739,
+      "step": 2825
+    },
+    {
+      "epoch": 2.6242452392011146,
+      "grad_norm": 0.16819532461761638,
+      "learning_rate": 6.93631669535284e-06,
+      "loss": 0.262,
+      "step": 2826
+    },
+    {
+      "epoch": 2.6251741755689735,
+      "grad_norm": 0.1760702927558675,
+      "learning_rate": 6.9191049913941485e-06,
+      "loss": 0.2698,
+      "step": 2827
+    },
+    {
+      "epoch": 2.6261031119368323,
+      "grad_norm": 0.17497265930471442,
+      "learning_rate": 6.901893287435457e-06,
+      "loss": 0.268,
+      "step": 2828
+    },
+    {
+      "epoch": 2.627032048304691,
+      "grad_norm": 0.23004773700222614,
+      "learning_rate": 6.884681583476764e-06,
+      "loss": 0.2662,
+      "step": 2829
+    },
+    {
+      "epoch": 2.62796098467255,
+      "grad_norm": 0.1757075749648394,
+      "learning_rate": 6.867469879518072e-06,
+      "loss": 0.2788,
+      "step": 2830
+    },
+    {
+      "epoch": 2.628889921040409,
+      "grad_norm": 0.1817582870520524,
+      "learning_rate": 6.8502581755593804e-06,
+      "loss": 0.2864,
+      "step": 2831
+    },
+    {
+      "epoch": 2.6298188574082673,
+      "grad_norm": 0.17943345206651734,
+      "learning_rate": 6.833046471600689e-06,
+      "loss": 0.2709,
+      "step": 2832
+    },
+    {
+      "epoch": 2.630747793776126,
+      "grad_norm": 0.18193170849342033,
+      "learning_rate": 6.815834767641997e-06,
+      "loss": 0.2776,
+      "step": 2833
+    },
+    {
+      "epoch": 2.631676730143985,
+      "grad_norm": 0.17203280602668908,
+      "learning_rate": 6.798623063683306e-06,
+      "loss": 0.2619,
+      "step": 2834
+    },
+    {
+      "epoch": 2.632605666511844,
+      "grad_norm": 0.17778486134836455,
+      "learning_rate": 6.781411359724612e-06,
+      "loss": 0.2825,
+      "step": 2835
+    },
+    {
+      "epoch": 2.6335346028797026,
+      "grad_norm": 0.17844868569560363,
+      "learning_rate": 6.764199655765921e-06,
+      "loss": 0.2858,
+      "step": 2836
+    },
+    {
+      "epoch": 2.6344635392475615,
+      "grad_norm": 0.17726568802995427,
+      "learning_rate": 6.746987951807229e-06,
+      "loss": 0.2811,
+      "step": 2837
+    },
+    {
+      "epoch": 2.6353924756154203,
+      "grad_norm": 0.16952025898385292,
+      "learning_rate": 6.7297762478485375e-06,
+      "loss": 0.271,
+      "step": 2838
+    },
+    {
+      "epoch": 2.636321411983279,
+      "grad_norm": 0.16547292535807162,
+      "learning_rate": 6.712564543889846e-06,
+      "loss": 0.2685,
+      "step": 2839
+    },
+    {
+      "epoch": 2.637250348351138,
+      "grad_norm": 0.17302245425655283,
+      "learning_rate": 6.695352839931153e-06,
+      "loss": 0.271,
+      "step": 2840
+    },
+    {
+      "epoch": 2.638179284718997,
+      "grad_norm": 0.16762567096195974,
+      "learning_rate": 6.678141135972461e-06,
+      "loss": 0.266,
+      "step": 2841
+    },
+    {
+      "epoch": 2.6391082210868557,
+      "grad_norm": 0.1699246368225437,
+      "learning_rate": 6.6609294320137694e-06,
+      "loss": 0.2696,
+      "step": 2842
+    },
+    {
+      "epoch": 2.6400371574547146,
+      "grad_norm": 0.16340294143468365,
+      "learning_rate": 6.643717728055078e-06,
+      "loss": 0.2732,
+      "step": 2843
+    },
+    {
+      "epoch": 2.6409660938225734,
+      "grad_norm": 0.17885057916542138,
+      "learning_rate": 6.626506024096386e-06,
+      "loss": 0.2896,
+      "step": 2844
+    },
+    {
+      "epoch": 2.641895030190432,
+      "grad_norm": 0.17028443868648646,
+      "learning_rate": 6.609294320137693e-06,
+      "loss": 0.2771,
+      "step": 2845
+    },
+    {
+      "epoch": 2.6428239665582907,
+      "grad_norm": 0.17146793678084205,
+      "learning_rate": 6.592082616179001e-06,
+      "loss": 0.277,
+      "step": 2846
+    },
+    {
+      "epoch": 2.6437529029261495,
+      "grad_norm": 0.18046897103380846,
+      "learning_rate": 6.57487091222031e-06,
+      "loss": 0.2712,
+      "step": 2847
+    },
+    {
+      "epoch": 2.6446818392940084,
+      "grad_norm": 0.17765626048887875,
+      "learning_rate": 6.557659208261618e-06,
+      "loss": 0.2755,
+      "step": 2848
+    },
+    {
+      "epoch": 2.645610775661867,
+      "grad_norm": 0.16752452222262293,
+      "learning_rate": 6.5404475043029266e-06,
+      "loss": 0.2676,
+      "step": 2849
+    },
+    {
+      "epoch": 2.646539712029726,
+      "grad_norm": 0.17019441380452832,
+      "learning_rate": 6.523235800344234e-06,
+      "loss": 0.27,
+      "step": 2850
+    },
+    {
+      "epoch": 2.647468648397585,
+      "grad_norm": 0.20022193692643978,
+      "learning_rate": 6.506024096385542e-06,
+      "loss": 0.2719,
+      "step": 2851
+    },
+    {
+      "epoch": 2.6483975847654433,
+      "grad_norm": 0.18034574748663765,
+      "learning_rate": 6.48881239242685e-06,
+      "loss": 0.2815,
+      "step": 2852
+    },
+    {
+      "epoch": 2.649326521133302,
+      "grad_norm": 0.17256922959908166,
+      "learning_rate": 6.4716006884681585e-06,
+      "loss": 0.2822,
+      "step": 2853
+    },
+    {
+      "epoch": 2.650255457501161,
+      "grad_norm": 0.17557669549784571,
+      "learning_rate": 6.454388984509467e-06,
+      "loss": 0.2819,
+      "step": 2854
+    },
+    {
+      "epoch": 2.65118439386902,
+      "grad_norm": 0.18357507050027355,
+      "learning_rate": 6.437177280550775e-06,
+      "loss": 0.2856,
+      "step": 2855
+    },
+    {
+      "epoch": 2.6521133302368787,
+      "grad_norm": 0.17883677795624237,
+      "learning_rate": 6.419965576592083e-06,
+      "loss": 0.278,
+      "step": 2856
+    },
+    {
+      "epoch": 2.6530422666047375,
+      "grad_norm": 0.1626862574701004,
+      "learning_rate": 6.40275387263339e-06,
+      "loss": 0.2574,
+      "step": 2857
+    },
+    {
+      "epoch": 2.6539712029725964,
+      "grad_norm": 0.16383951586471682,
+      "learning_rate": 6.385542168674699e-06,
+      "loss": 0.2752,
+      "step": 2858
+    },
+    {
+      "epoch": 2.654900139340455,
+      "grad_norm": 0.16774157905432333,
+      "learning_rate": 6.368330464716007e-06,
+      "loss": 0.2788,
+      "step": 2859
+    },
+    {
+      "epoch": 2.655829075708314,
+      "grad_norm": 0.17064891524490505,
+      "learning_rate": 6.3511187607573156e-06,
+      "loss": 0.2744,
+      "step": 2860
+    },
+    {
+      "epoch": 2.656758012076173,
+      "grad_norm": 0.18019181299443282,
+      "learning_rate": 6.333907056798623e-06,
+      "loss": 0.2749,
+      "step": 2861
+    },
+    {
+      "epoch": 2.6576869484440317,
+      "grad_norm": 0.18065522881499407,
+      "learning_rate": 6.3166953528399315e-06,
+      "loss": 0.2781,
+      "step": 2862
+    },
+    {
+      "epoch": 2.6586158848118906,
+      "grad_norm": 0.16699071545783015,
+      "learning_rate": 6.299483648881239e-06,
+      "loss": 0.2699,
+      "step": 2863
+    },
+    {
+      "epoch": 2.6595448211797494,
+      "grad_norm": 0.17960486746690257,
+      "learning_rate": 6.2822719449225475e-06,
+      "loss": 0.2701,
+      "step": 2864
+    },
+    {
+      "epoch": 2.660473757547608,
+      "grad_norm": 0.16751383073112,
+      "learning_rate": 6.265060240963856e-06,
+      "loss": 0.2706,
+      "step": 2865
+    },
+    {
+      "epoch": 2.6614026939154667,
+      "grad_norm": 0.17012354333185065,
+      "learning_rate": 6.2478485370051634e-06,
+      "loss": 0.2603,
+      "step": 2866
+    },
+    {
+      "epoch": 2.6623316302833255,
+      "grad_norm": 0.16323188266871058,
+      "learning_rate": 6.230636833046472e-06,
+      "loss": 0.2587,
+      "step": 2867
+    },
+    {
+      "epoch": 2.6632605666511844,
+      "grad_norm": 0.16800338995254763,
+      "learning_rate": 6.21342512908778e-06,
+      "loss": 0.2844,
+      "step": 2868
+    },
+    {
+      "epoch": 2.6641895030190432,
+      "grad_norm": 0.17523825857623324,
+      "learning_rate": 6.196213425129088e-06,
+      "loss": 0.2664,
+      "step": 2869
+    },
+    {
+      "epoch": 2.665118439386902,
+      "grad_norm": 0.17259113929735792,
+      "learning_rate": 6.179001721170396e-06,
+      "loss": 0.2682,
+      "step": 2870
+    },
+    {
+      "epoch": 2.666047375754761,
+      "grad_norm": 0.15843773315912996,
+      "learning_rate": 6.161790017211705e-06,
+      "loss": 0.2526,
+      "step": 2871
+    },
+    {
+      "epoch": 2.6669763121226198,
+      "grad_norm": 0.1667368341714137,
+      "learning_rate": 6.144578313253013e-06,
+      "loss": 0.2836,
+      "step": 2872
+    },
+    {
+      "epoch": 2.667905248490478,
+      "grad_norm": 0.18356977684527812,
+      "learning_rate": 6.1273666092943205e-06,
+      "loss": 0.2879,
+      "step": 2873
+    },
+    {
+      "epoch": 2.668834184858337,
+      "grad_norm": 0.1718844280644391,
+      "learning_rate": 6.110154905335629e-06,
+      "loss": 0.2691,
+      "step": 2874
+    },
+    {
+      "epoch": 2.669763121226196,
+      "grad_norm": 0.17542563761145286,
+      "learning_rate": 6.092943201376937e-06,
+      "loss": 0.2835,
+      "step": 2875
+    },
+    {
+      "epoch": 2.6706920575940547,
+      "grad_norm": 0.18112601734776523,
+      "learning_rate": 6.075731497418245e-06,
+      "loss": 0.2813,
+      "step": 2876
+    },
+    {
+      "epoch": 2.6716209939619135,
+      "grad_norm": 0.17517423987429506,
+      "learning_rate": 6.058519793459553e-06,
+      "loss": 0.2769,
+      "step": 2877
+    },
+    {
+      "epoch": 2.6725499303297724,
+      "grad_norm": 0.17260026662133704,
+      "learning_rate": 6.041308089500861e-06,
+      "loss": 0.2764,
+      "step": 2878
+    },
+    {
+      "epoch": 2.6734788666976312,
+      "grad_norm": 0.17274079216676472,
+      "learning_rate": 6.024096385542169e-06,
+      "loss": 0.2633,
+      "step": 2879
+    },
+    {
+      "epoch": 2.67440780306549,
+      "grad_norm": 0.1749655577705151,
+      "learning_rate": 6.006884681583478e-06,
+      "loss": 0.2801,
+      "step": 2880
+    },
+    {
+      "epoch": 2.675336739433349,
+      "grad_norm": 0.18151574461910705,
+      "learning_rate": 5.989672977624785e-06,
+      "loss": 0.2866,
+      "step": 2881
+    },
+    {
+      "epoch": 2.6762656758012078,
+      "grad_norm": 0.1603999496686991,
+      "learning_rate": 5.972461273666094e-06,
+      "loss": 0.2582,
+      "step": 2882
+    },
+    {
+      "epoch": 2.6771946121690666,
+      "grad_norm": 0.1757361166235524,
+      "learning_rate": 5.955249569707402e-06,
+      "loss": 0.2844,
+      "step": 2883
+    },
+    {
+      "epoch": 2.6781235485369255,
+      "grad_norm": 0.17402452532430263,
+      "learning_rate": 5.9380378657487095e-06,
+      "loss": 0.2765,
+      "step": 2884
+    },
+    {
+      "epoch": 2.679052484904784,
+      "grad_norm": 0.1709408589895161,
+      "learning_rate": 5.920826161790018e-06,
+      "loss": 0.2838,
+      "step": 2885
+    },
+    {
+      "epoch": 2.6799814212726427,
+      "grad_norm": 0.17361142927299925,
+      "learning_rate": 5.9036144578313255e-06,
+      "loss": 0.2868,
+      "step": 2886
+    },
+    {
+      "epoch": 2.6809103576405016,
+      "grad_norm": 0.1597414246640021,
+      "learning_rate": 5.886402753872634e-06,
+      "loss": 0.2611,
+      "step": 2887
+    },
+    {
+      "epoch": 2.6818392940083604,
+      "grad_norm": 0.16013784375486953,
+      "learning_rate": 5.869191049913942e-06,
+      "loss": 0.2557,
+      "step": 2888
+    },
+    {
+      "epoch": 2.6827682303762193,
+      "grad_norm": 0.15920257619823422,
+      "learning_rate": 5.85197934595525e-06,
+      "loss": 0.2537,
+      "step": 2889
+    },
+    {
+      "epoch": 2.683697166744078,
+      "grad_norm": 0.19602228120008208,
+      "learning_rate": 5.834767641996558e-06,
+      "loss": 0.2767,
+      "step": 2890
+    },
+    {
+      "epoch": 2.684626103111937,
+      "grad_norm": 0.16026967401774392,
+      "learning_rate": 5.817555938037867e-06,
+      "loss": 0.2652,
+      "step": 2891
+    },
+    {
+      "epoch": 2.685555039479796,
+      "grad_norm": 0.16763626760535624,
+      "learning_rate": 5.800344234079174e-06,
+      "loss": 0.2797,
+      "step": 2892
+    },
+    {
+      "epoch": 2.686483975847654,
+      "grad_norm": 0.1706032665023045,
+      "learning_rate": 5.783132530120483e-06,
+      "loss": 0.2744,
+      "step": 2893
+    },
+    {
+      "epoch": 2.687412912215513,
+      "grad_norm": 0.18732977118836014,
+      "learning_rate": 5.76592082616179e-06,
+      "loss": 0.2924,
+      "step": 2894
+    },
+    {
+      "epoch": 2.688341848583372,
+      "grad_norm": 0.17185758153715494,
+      "learning_rate": 5.7487091222030986e-06,
+      "loss": 0.2655,
+      "step": 2895
+    },
+    {
+      "epoch": 2.6892707849512307,
+      "grad_norm": 0.17530485016367162,
+      "learning_rate": 5.731497418244407e-06,
+      "loss": 0.2847,
+      "step": 2896
+    },
+    {
+      "epoch": 2.6901997213190896,
+      "grad_norm": 0.15863803214886732,
+      "learning_rate": 5.7142857142857145e-06,
+      "loss": 0.2704,
+      "step": 2897
+    },
+    {
+      "epoch": 2.6911286576869484,
+      "grad_norm": 0.1691861266842899,
+      "learning_rate": 5.697074010327023e-06,
+      "loss": 0.29,
+      "step": 2898
+    },
+    {
+      "epoch": 2.6920575940548073,
+      "grad_norm": 0.16970518454179528,
+      "learning_rate": 5.679862306368331e-06,
+      "loss": 0.2614,
+      "step": 2899
+    },
+    {
+      "epoch": 2.692986530422666,
+      "grad_norm": 0.1741414021917801,
+      "learning_rate": 5.662650602409639e-06,
+      "loss": 0.2822,
+      "step": 2900
+    },
+    {
+      "epoch": 2.693915466790525,
+      "grad_norm": 0.1681316865512381,
+      "learning_rate": 5.645438898450947e-06,
+      "loss": 0.2577,
+      "step": 2901
+    },
+    {
+      "epoch": 2.694844403158384,
+      "grad_norm": 0.16184757987994441,
+      "learning_rate": 5.628227194492255e-06,
+      "loss": 0.2583,
+      "step": 2902
+    },
+    {
+      "epoch": 2.6957733395262427,
+      "grad_norm": 0.16281128048095134,
+      "learning_rate": 5.611015490533563e-06,
+      "loss": 0.2641,
+      "step": 2903
+    },
+    {
+      "epoch": 2.6967022758941015,
+      "grad_norm": 0.16822832216234318,
+      "learning_rate": 5.593803786574872e-06,
+      "loss": 0.2674,
+      "step": 2904
+    },
+    {
+      "epoch": 2.69763121226196,
+      "grad_norm": 0.1660704396611372,
+      "learning_rate": 5.576592082616179e-06,
+      "loss": 0.2693,
+      "step": 2905
+    },
+    {
+      "epoch": 2.6985601486298187,
+      "grad_norm": 0.17603534664952974,
+      "learning_rate": 5.5593803786574876e-06,
+      "loss": 0.2913,
+      "step": 2906
+    },
+    {
+      "epoch": 2.6994890849976776,
+      "grad_norm": 0.16173314477849363,
+      "learning_rate": 5.542168674698795e-06,
+      "loss": 0.2565,
+      "step": 2907
+    },
+    {
+      "epoch": 2.7004180213655364,
+      "grad_norm": 0.1663575560714465,
+      "learning_rate": 5.5249569707401035e-06,
+      "loss": 0.2748,
+      "step": 2908
+    },
+    {
+      "epoch": 2.7013469577333953,
+      "grad_norm": 0.1721550238413437,
+      "learning_rate": 5.507745266781412e-06,
+      "loss": 0.2912,
+      "step": 2909
+    },
+    {
+      "epoch": 2.702275894101254,
+      "grad_norm": 0.16768308926561104,
+      "learning_rate": 5.4905335628227195e-06,
+      "loss": 0.2666,
+      "step": 2910
+    },
+    {
+      "epoch": 2.703204830469113,
+      "grad_norm": 0.18026118938082336,
+      "learning_rate": 5.473321858864028e-06,
+      "loss": 0.2742,
+      "step": 2911
+    },
+    {
+      "epoch": 2.704133766836972,
+      "grad_norm": 0.17114811293445578,
+      "learning_rate": 5.456110154905336e-06,
+      "loss": 0.2869,
+      "step": 2912
+    },
+    {
+      "epoch": 2.70506270320483,
+      "grad_norm": 0.1897353218307368,
+      "learning_rate": 5.438898450946644e-06,
+      "loss": 0.2891,
+      "step": 2913
+    },
+    {
+      "epoch": 2.705991639572689,
+      "grad_norm": 0.17270248612365974,
+      "learning_rate": 5.421686746987952e-06,
+      "loss": 0.262,
+      "step": 2914
+    },
+    {
+      "epoch": 2.706920575940548,
+      "grad_norm": 0.16787775314760653,
+      "learning_rate": 5.40447504302926e-06,
+      "loss": 0.2698,
+      "step": 2915
+    },
+    {
+      "epoch": 2.7078495123084068,
+      "grad_norm": 0.1969016770528086,
+      "learning_rate": 5.387263339070568e-06,
+      "loss": 0.2822,
+      "step": 2916
+    },
+    {
+      "epoch": 2.7087784486762656,
+      "grad_norm": 0.16996615370604629,
+      "learning_rate": 5.370051635111877e-06,
+      "loss": 0.2695,
+      "step": 2917
+    },
+    {
+      "epoch": 2.7097073850441245,
+      "grad_norm": 0.16671749967465915,
+      "learning_rate": 5.352839931153184e-06,
+      "loss": 0.2673,
+      "step": 2918
+    },
+    {
+      "epoch": 2.7106363214119833,
+      "grad_norm": 0.17148870032680638,
+      "learning_rate": 5.3356282271944925e-06,
+      "loss": 0.2736,
+      "step": 2919
+    },
+    {
+      "epoch": 2.711565257779842,
+      "grad_norm": 0.16182282721435845,
+      "learning_rate": 5.318416523235801e-06,
+      "loss": 0.255,
+      "step": 2920
+    },
+    {
+      "epoch": 2.712494194147701,
+      "grad_norm": 0.18004793360917773,
+      "learning_rate": 5.3012048192771085e-06,
+      "loss": 0.3005,
+      "step": 2921
+    },
+    {
+      "epoch": 2.71342313051556,
+      "grad_norm": 0.16893534756516349,
+      "learning_rate": 5.283993115318417e-06,
+      "loss": 0.2614,
+      "step": 2922
+    },
+    {
+      "epoch": 2.7143520668834187,
+      "grad_norm": 0.17993627923927366,
+      "learning_rate": 5.2667814113597244e-06,
+      "loss": 0.2829,
+      "step": 2923
+    },
+    {
+      "epoch": 2.7152810032512775,
+      "grad_norm": 0.16985404866566564,
+      "learning_rate": 5.249569707401033e-06,
+      "loss": 0.2714,
+      "step": 2924
+    },
+    {
+      "epoch": 2.716209939619136,
+      "grad_norm": 0.16997003285554874,
+      "learning_rate": 5.232358003442341e-06,
+      "loss": 0.2723,
+      "step": 2925
+    },
+    {
+      "epoch": 2.7171388759869948,
+      "grad_norm": 0.17280395561719686,
+      "learning_rate": 5.215146299483649e-06,
+      "loss": 0.2806,
+      "step": 2926
+    },
+    {
+      "epoch": 2.7180678123548536,
+      "grad_norm": 0.18591179771276334,
+      "learning_rate": 5.197934595524957e-06,
+      "loss": 0.2839,
+      "step": 2927
+    },
+    {
+      "epoch": 2.7189967487227125,
+      "grad_norm": 0.16848714989466346,
+      "learning_rate": 5.180722891566266e-06,
+      "loss": 0.27,
+      "step": 2928
+    },
+    {
+      "epoch": 2.7199256850905713,
+      "grad_norm": 0.17396159174476428,
+      "learning_rate": 5.163511187607573e-06,
+      "loss": 0.2887,
+      "step": 2929
+    },
+    {
+      "epoch": 2.72085462145843,
+      "grad_norm": 0.16362910998344926,
+      "learning_rate": 5.1462994836488815e-06,
+      "loss": 0.2677,
+      "step": 2930
+    },
+    {
+      "epoch": 2.721783557826289,
+      "grad_norm": 0.17616701253337705,
+      "learning_rate": 5.129087779690189e-06,
+      "loss": 0.2805,
+      "step": 2931
+    },
+    {
+      "epoch": 2.722712494194148,
+      "grad_norm": 0.16554867783146326,
+      "learning_rate": 5.1118760757314975e-06,
+      "loss": 0.2604,
+      "step": 2932
+    },
+    {
+      "epoch": 2.7236414305620062,
+      "grad_norm": 0.1830057396941038,
+      "learning_rate": 5.094664371772806e-06,
+      "loss": 0.2661,
+      "step": 2933
+    },
+    {
+      "epoch": 2.724570366929865,
+      "grad_norm": 0.16804873136561682,
+      "learning_rate": 5.0774526678141135e-06,
+      "loss": 0.2761,
+      "step": 2934
+    },
+    {
+      "epoch": 2.725499303297724,
+      "grad_norm": 0.17881514628477865,
+      "learning_rate": 5.060240963855422e-06,
+      "loss": 0.2833,
+      "step": 2935
+    },
+    {
+      "epoch": 2.726428239665583,
+      "grad_norm": 0.18028085375147532,
+      "learning_rate": 5.04302925989673e-06,
+      "loss": 0.28,
+      "step": 2936
+    },
+    {
+      "epoch": 2.7273571760334416,
+      "grad_norm": 0.17128865792041942,
+      "learning_rate": 5.025817555938038e-06,
+      "loss": 0.2827,
+      "step": 2937
+    },
+    {
+      "epoch": 2.7282861124013005,
+      "grad_norm": 0.17508069891904077,
+      "learning_rate": 5.008605851979346e-06,
+      "loss": 0.2846,
+      "step": 2938
+    },
+    {
+      "epoch": 2.7292150487691593,
+      "grad_norm": 0.1624442446983855,
+      "learning_rate": 4.991394148020654e-06,
+      "loss": 0.2605,
+      "step": 2939
+    },
+    {
+      "epoch": 2.730143985137018,
+      "grad_norm": 0.16717663468832003,
+      "learning_rate": 4.974182444061962e-06,
+      "loss": 0.2562,
+      "step": 2940
+    },
+    {
+      "epoch": 2.731072921504877,
+      "grad_norm": 0.17435792481065424,
+      "learning_rate": 4.9569707401032706e-06,
+      "loss": 0.2641,
+      "step": 2941
+    },
+    {
+      "epoch": 2.732001857872736,
+      "grad_norm": 0.18206523972349564,
+      "learning_rate": 4.939759036144578e-06,
+      "loss": 0.2906,
+      "step": 2942
+    },
+    {
+      "epoch": 2.7329307942405947,
+      "grad_norm": 0.1660518040509173,
+      "learning_rate": 4.9225473321858865e-06,
+      "loss": 0.2616,
+      "step": 2943
+    },
+    {
+      "epoch": 2.7338597306084536,
+      "grad_norm": 0.17641521165468432,
+      "learning_rate": 4.905335628227195e-06,
+      "loss": 0.279,
+      "step": 2944
+    },
+    {
+      "epoch": 2.7347886669763124,
+      "grad_norm": 0.17021194005960247,
+      "learning_rate": 4.8881239242685025e-06,
+      "loss": 0.2662,
+      "step": 2945
+    },
+    {
+      "epoch": 2.735717603344171,
+      "grad_norm": 0.18680530424480912,
+      "learning_rate": 4.870912220309811e-06,
+      "loss": 0.279,
+      "step": 2946
+    },
+    {
+      "epoch": 2.7366465397120296,
+      "grad_norm": 0.16624112372533945,
+      "learning_rate": 4.853700516351118e-06,
+      "loss": 0.2603,
+      "step": 2947
+    },
+    {
+      "epoch": 2.7375754760798885,
+      "grad_norm": 0.1658532363601099,
+      "learning_rate": 4.836488812392427e-06,
+      "loss": 0.2573,
+      "step": 2948
+    },
+    {
+      "epoch": 2.7385044124477473,
+      "grad_norm": 0.17508939946629776,
+      "learning_rate": 4.819277108433735e-06,
+      "loss": 0.2748,
+      "step": 2949
+    },
+    {
+      "epoch": 2.739433348815606,
+      "grad_norm": 0.1697082547434485,
+      "learning_rate": 4.802065404475043e-06,
+      "loss": 0.277,
+      "step": 2950
+    },
+    {
+      "epoch": 2.740362285183465,
+      "grad_norm": 0.17265474187461444,
+      "learning_rate": 4.784853700516351e-06,
+      "loss": 0.2766,
+      "step": 2951
+    },
+    {
+      "epoch": 2.741291221551324,
+      "grad_norm": 0.1801581954287595,
+      "learning_rate": 4.7676419965576596e-06,
+      "loss": 0.2865,
+      "step": 2952
+    },
+    {
+      "epoch": 2.7422201579191823,
+      "grad_norm": 0.1671113061115391,
+      "learning_rate": 4.750430292598967e-06,
+      "loss": 0.2789,
+      "step": 2953
+    },
+    {
+      "epoch": 2.743149094287041,
+      "grad_norm": 0.1810191012298334,
+      "learning_rate": 4.7332185886402755e-06,
+      "loss": 0.2829,
+      "step": 2954
+    },
+    {
+      "epoch": 2.7440780306549,
+      "grad_norm": 0.1873382578762812,
+      "learning_rate": 4.716006884681583e-06,
+      "loss": 0.3014,
+      "step": 2955
+    },
+    {
+      "epoch": 2.745006967022759,
+      "grad_norm": 0.16836258371111407,
+      "learning_rate": 4.6987951807228915e-06,
+      "loss": 0.2777,
+      "step": 2956
+    },
+    {
+      "epoch": 2.7459359033906177,
+      "grad_norm": 0.17397396442719273,
+      "learning_rate": 4.6815834767642e-06,
+      "loss": 0.2794,
+      "step": 2957
+    },
+    {
+      "epoch": 2.7468648397584765,
+      "grad_norm": 0.17012803273497898,
+      "learning_rate": 4.6643717728055074e-06,
+      "loss": 0.2735,
+      "step": 2958
+    },
+    {
+      "epoch": 2.7477937761263354,
+      "grad_norm": 0.1726453541385025,
+      "learning_rate": 4.647160068846816e-06,
+      "loss": 0.2728,
+      "step": 2959
+    },
+    {
+      "epoch": 2.748722712494194,
+      "grad_norm": 0.1607579980714856,
+      "learning_rate": 4.629948364888123e-06,
+      "loss": 0.2652,
+      "step": 2960
+    },
+    {
+      "epoch": 2.749651648862053,
+      "grad_norm": 0.19173078813688318,
+      "learning_rate": 4.612736660929432e-06,
+      "loss": 0.2889,
+      "step": 2961
+    },
+    {
+      "epoch": 2.750580585229912,
+      "grad_norm": 0.16695194838007188,
+      "learning_rate": 4.59552495697074e-06,
+      "loss": 0.2711,
+      "step": 2962
+    },
+    {
+      "epoch": 2.7515095215977707,
+      "grad_norm": 0.1644552191700174,
+      "learning_rate": 4.578313253012049e-06,
+      "loss": 0.2612,
+      "step": 2963
+    },
+    {
+      "epoch": 2.7524384579656296,
+      "grad_norm": 0.16861408449465942,
+      "learning_rate": 4.561101549053356e-06,
+      "loss": 0.2672,
+      "step": 2964
+    },
+    {
+      "epoch": 2.7533673943334884,
+      "grad_norm": 0.16556411207676844,
+      "learning_rate": 4.5438898450946645e-06,
+      "loss": 0.2843,
+      "step": 2965
+    },
+    {
+      "epoch": 2.754296330701347,
+      "grad_norm": 0.1653424279361049,
+      "learning_rate": 4.526678141135973e-06,
+      "loss": 0.2698,
+      "step": 2966
+    },
+    {
+      "epoch": 2.7552252670692057,
+      "grad_norm": 0.17268687936725427,
+      "learning_rate": 4.5094664371772805e-06,
+      "loss": 0.2753,
+      "step": 2967
+    },
+    {
+      "epoch": 2.7561542034370645,
+      "grad_norm": 0.16679010719063728,
+      "learning_rate": 4.492254733218589e-06,
+      "loss": 0.2658,
+      "step": 2968
+    },
+    {
+      "epoch": 2.7570831398049234,
+      "grad_norm": 0.1731384049435442,
+      "learning_rate": 4.475043029259897e-06,
+      "loss": 0.2776,
+      "step": 2969
+    },
+    {
+      "epoch": 2.758012076172782,
+      "grad_norm": 0.17817981617485937,
+      "learning_rate": 4.457831325301205e-06,
+      "loss": 0.2723,
+      "step": 2970
+    },
+    {
+      "epoch": 2.758941012540641,
+      "grad_norm": 0.17631736307329407,
+      "learning_rate": 4.440619621342513e-06,
+      "loss": 0.29,
+      "step": 2971
+    },
+    {
+      "epoch": 2.7598699489085,
+      "grad_norm": 0.16557688303929388,
+      "learning_rate": 4.423407917383822e-06,
+      "loss": 0.2715,
+      "step": 2972
+    },
+    {
+      "epoch": 2.7607988852763583,
+      "grad_norm": 0.16830413475646644,
+      "learning_rate": 4.406196213425129e-06,
+      "loss": 0.2755,
+      "step": 2973
+    },
+    {
+      "epoch": 2.761727821644217,
+      "grad_norm": 0.16634439441042223,
+      "learning_rate": 4.388984509466438e-06,
+      "loss": 0.2821,
+      "step": 2974
+    },
+    {
+      "epoch": 2.762656758012076,
+      "grad_norm": 0.17593720968564058,
+      "learning_rate": 4.371772805507746e-06,
+      "loss": 0.2661,
+      "step": 2975
+    },
+    {
+      "epoch": 2.763585694379935,
+      "grad_norm": 0.16971061649067173,
+      "learning_rate": 4.3545611015490536e-06,
+      "loss": 0.2611,
+      "step": 2976
+    },
+    {
+      "epoch": 2.7645146307477937,
+      "grad_norm": 0.18290133340663603,
+      "learning_rate": 4.337349397590362e-06,
+      "loss": 0.296,
+      "step": 2977
+    },
+    {
+      "epoch": 2.7654435671156525,
+      "grad_norm": 0.16094186761731016,
+      "learning_rate": 4.32013769363167e-06,
+      "loss": 0.2599,
+      "step": 2978
+    },
+    {
+      "epoch": 2.7663725034835114,
+      "grad_norm": 0.1700225427982041,
+      "learning_rate": 4.302925989672978e-06,
+      "loss": 0.2851,
+      "step": 2979
+    },
+    {
+      "epoch": 2.7673014398513702,
+      "grad_norm": 0.169245573986097,
+      "learning_rate": 4.285714285714286e-06,
+      "loss": 0.2543,
+      "step": 2980
+    },
+    {
+      "epoch": 2.768230376219229,
+      "grad_norm": 0.16461588181504275,
+      "learning_rate": 4.268502581755595e-06,
+      "loss": 0.2671,
+      "step": 2981
+    },
+    {
+      "epoch": 2.769159312587088,
+      "grad_norm": 0.16325022819287804,
+      "learning_rate": 4.251290877796902e-06,
+      "loss": 0.2917,
+      "step": 2982
+    },
+    {
+      "epoch": 2.7700882489549468,
+      "grad_norm": 0.17634053929655194,
+      "learning_rate": 4.234079173838211e-06,
+      "loss": 0.289,
+      "step": 2983
+    },
+    {
+      "epoch": 2.7710171853228056,
+      "grad_norm": 0.17119702526297131,
+      "learning_rate": 4.216867469879518e-06,
+      "loss": 0.283,
+      "step": 2984
+    },
+    {
+      "epoch": 2.7719461216906645,
+      "grad_norm": 0.16474730190343942,
+      "learning_rate": 4.199655765920827e-06,
+      "loss": 0.2639,
+      "step": 2985
+    },
+    {
+      "epoch": 2.772875058058523,
+      "grad_norm": 0.33285250375326086,
+      "learning_rate": 4.182444061962135e-06,
+      "loss": 0.267,
+      "step": 2986
+    },
+    {
+      "epoch": 2.7738039944263817,
+      "grad_norm": 0.16364044631808572,
+      "learning_rate": 4.1652323580034426e-06,
+      "loss": 0.2717,
+      "step": 2987
+    },
+    {
+      "epoch": 2.7747329307942405,
+      "grad_norm": 0.16681061485901677,
+      "learning_rate": 4.148020654044751e-06,
+      "loss": 0.2683,
+      "step": 2988
+    },
+    {
+      "epoch": 2.7756618671620994,
+      "grad_norm": 0.1651724380142913,
+      "learning_rate": 4.130808950086059e-06,
+      "loss": 0.2704,
+      "step": 2989
+    },
+    {
+      "epoch": 2.7765908035299582,
+      "grad_norm": 0.1709084288247249,
+      "learning_rate": 4.113597246127367e-06,
+      "loss": 0.2819,
+      "step": 2990
+    },
+    {
+      "epoch": 2.777519739897817,
+      "grad_norm": 0.16357948699205382,
+      "learning_rate": 4.096385542168675e-06,
+      "loss": 0.2875,
+      "step": 2991
+    },
+    {
+      "epoch": 2.778448676265676,
+      "grad_norm": 0.17167993006882395,
+      "learning_rate": 4.079173838209983e-06,
+      "loss": 0.2745,
+      "step": 2992
+    },
+    {
+      "epoch": 2.7793776126335343,
+      "grad_norm": 0.1675771078403747,
+      "learning_rate": 4.061962134251291e-06,
+      "loss": 0.2823,
+      "step": 2993
+    },
+    {
+      "epoch": 2.780306549001393,
+      "grad_norm": 0.16584264394434609,
+      "learning_rate": 4.0447504302926e-06,
+      "loss": 0.2744,
+      "step": 2994
+    },
+    {
+      "epoch": 2.781235485369252,
+      "grad_norm": 0.16662698036780293,
+      "learning_rate": 4.027538726333907e-06,
+      "loss": 0.2688,
+      "step": 2995
+    },
+    {
+      "epoch": 2.782164421737111,
+      "grad_norm": 0.16794738348660668,
+      "learning_rate": 4.010327022375216e-06,
+      "loss": 0.2888,
+      "step": 2996
+    },
+    {
+      "epoch": 2.7830933581049697,
+      "grad_norm": 0.17789218429995182,
+      "learning_rate": 3.993115318416524e-06,
+      "loss": 0.2859,
+      "step": 2997
+    },
+    {
+      "epoch": 2.7840222944728286,
+      "grad_norm": 0.16091287332048385,
+      "learning_rate": 3.975903614457832e-06,
+      "loss": 0.2746,
+      "step": 2998
+    },
+    {
+      "epoch": 2.7849512308406874,
+      "grad_norm": 0.18944455859801637,
+      "learning_rate": 3.95869191049914e-06,
+      "loss": 0.2716,
+      "step": 2999
+    },
+    {
+      "epoch": 2.7858801672085463,
+      "grad_norm": 0.2954117399236272,
+      "learning_rate": 3.9414802065404475e-06,
+      "loss": 0.2902,
+      "step": 3000
+    },
+    {
+      "epoch": 2.786809103576405,
+      "grad_norm": 0.16573385628794401,
+      "learning_rate": 3.924268502581756e-06,
+      "loss": 0.2726,
+      "step": 3001
+    },
+    {
+      "epoch": 2.787738039944264,
+      "grad_norm": 0.17004238681555403,
+      "learning_rate": 3.907056798623064e-06,
+      "loss": 0.2831,
+      "step": 3002
+    },
+    {
+      "epoch": 2.788666976312123,
+      "grad_norm": 0.17663477423381163,
+      "learning_rate": 3.889845094664372e-06,
+      "loss": 0.2651,
+      "step": 3003
+    },
+    {
+      "epoch": 2.7895959126799816,
+      "grad_norm": 0.17378992478360958,
+      "learning_rate": 3.87263339070568e-06,
+      "loss": 0.2743,
+      "step": 3004
+    },
+    {
+      "epoch": 2.7905248490478405,
+      "grad_norm": 0.16494900421379727,
+      "learning_rate": 3.855421686746989e-06,
+      "loss": 0.2689,
+      "step": 3005
+    },
+    {
+      "epoch": 2.791453785415699,
+      "grad_norm": 0.1603907236344023,
+      "learning_rate": 3.838209982788296e-06,
+      "loss": 0.2657,
+      "step": 3006
+    },
+    {
+      "epoch": 2.7923827217835577,
+      "grad_norm": 0.1567572624521738,
+      "learning_rate": 3.820998278829605e-06,
+      "loss": 0.2637,
+      "step": 3007
+    },
+    {
+      "epoch": 2.7933116581514166,
+      "grad_norm": 0.18404937236372684,
+      "learning_rate": 3.803786574870912e-06,
+      "loss": 0.2755,
+      "step": 3008
+    },
+    {
+      "epoch": 2.7942405945192754,
+      "grad_norm": 0.16868307932572335,
+      "learning_rate": 3.7865748709122206e-06,
+      "loss": 0.2778,
+      "step": 3009
+    },
+    {
+      "epoch": 2.7951695308871343,
+      "grad_norm": 0.17073649004440103,
+      "learning_rate": 3.769363166953529e-06,
+      "loss": 0.264,
+      "step": 3010
+    },
+    {
+      "epoch": 2.796098467254993,
+      "grad_norm": 0.16505500534365,
+      "learning_rate": 3.7521514629948365e-06,
+      "loss": 0.2672,
+      "step": 3011
+    },
+    {
+      "epoch": 2.797027403622852,
+      "grad_norm": 0.16381885442693822,
+      "learning_rate": 3.734939759036145e-06,
+      "loss": 0.272,
+      "step": 3012
+    },
+    {
+      "epoch": 2.7979563399907104,
+      "grad_norm": 0.16292544827439895,
+      "learning_rate": 3.7177280550774525e-06,
+      "loss": 0.2737,
+      "step": 3013
+    },
+    {
+      "epoch": 2.798885276358569,
+      "grad_norm": 0.17228455569690837,
+      "learning_rate": 3.700516351118761e-06,
+      "loss": 0.2811,
+      "step": 3014
+    },
+    {
+      "epoch": 2.799814212726428,
+      "grad_norm": 0.19524306180865297,
+      "learning_rate": 3.6833046471600693e-06,
+      "loss": 0.2713,
+      "step": 3015
+    },
+    {
+      "epoch": 2.800743149094287,
+      "grad_norm": 0.16070401680293828,
+      "learning_rate": 3.666092943201377e-06,
+      "loss": 0.2692,
+      "step": 3016
+    },
+    {
+      "epoch": 2.8016720854621457,
+      "grad_norm": 0.16705220705216192,
+      "learning_rate": 3.6488812392426853e-06,
+      "loss": 0.2744,
+      "step": 3017
+    },
+    {
+      "epoch": 2.8026010218300046,
+      "grad_norm": 0.16539813777059287,
+      "learning_rate": 3.6316695352839937e-06,
+      "loss": 0.2747,
+      "step": 3018
+    },
+    {
+      "epoch": 2.8035299581978634,
+      "grad_norm": 0.16555633870974384,
+      "learning_rate": 3.614457831325301e-06,
+      "loss": 0.2899,
+      "step": 3019
+    },
+    {
+      "epoch": 2.8044588945657223,
+      "grad_norm": 0.16101648042383737,
+      "learning_rate": 3.5972461273666096e-06,
+      "loss": 0.268,
+      "step": 3020
+    },
+    {
+      "epoch": 2.805387830933581,
+      "grad_norm": 0.16506138181326488,
+      "learning_rate": 3.580034423407917e-06,
+      "loss": 0.2653,
+      "step": 3021
+    },
+    {
+      "epoch": 2.80631676730144,
+      "grad_norm": 0.1741809446855172,
+      "learning_rate": 3.5628227194492256e-06,
+      "loss": 0.2626,
+      "step": 3022
+    },
+    {
+      "epoch": 2.807245703669299,
+      "grad_norm": 0.16483290521015864,
+      "learning_rate": 3.545611015490534e-06,
+      "loss": 0.286,
+      "step": 3023
+    },
+    {
+      "epoch": 2.8081746400371577,
+      "grad_norm": 0.1571851560823419,
+      "learning_rate": 3.5283993115318415e-06,
+      "loss": 0.2495,
+      "step": 3024
+    },
+    {
+      "epoch": 2.8091035764050165,
+      "grad_norm": 0.16235471506982116,
+      "learning_rate": 3.51118760757315e-06,
+      "loss": 0.2778,
+      "step": 3025
+    },
+    {
+      "epoch": 2.810032512772875,
+      "grad_norm": 0.15756406208488002,
+      "learning_rate": 3.4939759036144583e-06,
+      "loss": 0.2643,
+      "step": 3026
+    },
+    {
+      "epoch": 2.8109614491407338,
+      "grad_norm": 0.17055705625316472,
+      "learning_rate": 3.476764199655766e-06,
+      "loss": 0.2996,
+      "step": 3027
+    },
+    {
+      "epoch": 2.8118903855085926,
+      "grad_norm": 0.1796767292475761,
+      "learning_rate": 3.4595524956970743e-06,
+      "loss": 0.276,
+      "step": 3028
+    },
+    {
+      "epoch": 2.8128193218764515,
+      "grad_norm": 0.1717885163183135,
+      "learning_rate": 3.442340791738382e-06,
+      "loss": 0.2933,
+      "step": 3029
+    },
+    {
+      "epoch": 2.8137482582443103,
+      "grad_norm": 0.17849394515997952,
+      "learning_rate": 3.4251290877796902e-06,
+      "loss": 0.2835,
+      "step": 3030
+    },
+    {
+      "epoch": 2.814677194612169,
+      "grad_norm": 0.1724053731347566,
+      "learning_rate": 3.4079173838209986e-06,
+      "loss": 0.2829,
+      "step": 3031
+    },
+    {
+      "epoch": 2.815606130980028,
+      "grad_norm": 0.17700506216444933,
+      "learning_rate": 3.390705679862306e-06,
+      "loss": 0.2836,
+      "step": 3032
+    },
+    {
+      "epoch": 2.816535067347887,
+      "grad_norm": 0.1711437698846605,
+      "learning_rate": 3.3734939759036146e-06,
+      "loss": 0.2848,
+      "step": 3033
+    },
+    {
+      "epoch": 2.8174640037157452,
+      "grad_norm": 0.1681253963815512,
+      "learning_rate": 3.356282271944923e-06,
+      "loss": 0.2792,
+      "step": 3034
+    },
+    {
+      "epoch": 2.818392940083604,
+      "grad_norm": 0.15981392784911405,
+      "learning_rate": 3.3390705679862305e-06,
+      "loss": 0.2699,
+      "step": 3035
+    },
+    {
+      "epoch": 2.819321876451463,
+      "grad_norm": 0.1740927276763424,
+      "learning_rate": 3.321858864027539e-06,
+      "loss": 0.2841,
+      "step": 3036
+    },
+    {
+      "epoch": 2.8202508128193218,
+      "grad_norm": 0.17717015170291478,
+      "learning_rate": 3.3046471600688465e-06,
+      "loss": 0.2819,
+      "step": 3037
+    },
+    {
+      "epoch": 2.8211797491871806,
+      "grad_norm": 0.1677501092435031,
+      "learning_rate": 3.287435456110155e-06,
+      "loss": 0.273,
+      "step": 3038
+    },
+    {
+      "epoch": 2.8221086855550395,
+      "grad_norm": 0.16673210395588484,
+      "learning_rate": 3.2702237521514633e-06,
+      "loss": 0.2693,
+      "step": 3039
+    },
+    {
+      "epoch": 2.8230376219228983,
+      "grad_norm": 0.16028997391776514,
+      "learning_rate": 3.253012048192771e-06,
+      "loss": 0.2624,
+      "step": 3040
+    },
+    {
+      "epoch": 2.823966558290757,
+      "grad_norm": 0.15950671745325568,
+      "learning_rate": 3.2358003442340792e-06,
+      "loss": 0.266,
+      "step": 3041
+    },
+    {
+      "epoch": 2.824895494658616,
+      "grad_norm": 0.18954997486322914,
+      "learning_rate": 3.2185886402753876e-06,
+      "loss": 0.2705,
+      "step": 3042
+    },
+    {
+      "epoch": 2.825824431026475,
+      "grad_norm": 0.18313245510676568,
+      "learning_rate": 3.201376936316695e-06,
+      "loss": 0.2903,
+      "step": 3043
+    },
+    {
+      "epoch": 2.8267533673943337,
+      "grad_norm": 0.17647582446439805,
+      "learning_rate": 3.1841652323580036e-06,
+      "loss": 0.299,
+      "step": 3044
+    },
+    {
+      "epoch": 2.8276823037621925,
+      "grad_norm": 0.18666596433252766,
+      "learning_rate": 3.1669535283993116e-06,
+      "loss": 0.2691,
+      "step": 3045
+    },
+    {
+      "epoch": 2.828611240130051,
+      "grad_norm": 0.1621297164024695,
+      "learning_rate": 3.1497418244406195e-06,
+      "loss": 0.2581,
+      "step": 3046
+    },
+    {
+      "epoch": 2.82954017649791,
+      "grad_norm": 0.16497540494920462,
+      "learning_rate": 3.132530120481928e-06,
+      "loss": 0.2674,
+      "step": 3047
+    },
+    {
+      "epoch": 2.8304691128657686,
+      "grad_norm": 0.18187531575636953,
+      "learning_rate": 3.115318416523236e-06,
+      "loss": 0.3102,
+      "step": 3048
+    },
+    {
+      "epoch": 2.8313980492336275,
+      "grad_norm": 0.19013570305586183,
+      "learning_rate": 3.098106712564544e-06,
+      "loss": 0.2816,
+      "step": 3049
+    },
+    {
+      "epoch": 2.8323269856014863,
+      "grad_norm": 0.16717134338818446,
+      "learning_rate": 3.0808950086058523e-06,
+      "loss": 0.2771,
+      "step": 3050
+    },
+    {
+      "epoch": 2.833255921969345,
+      "grad_norm": 0.16409077988979776,
+      "learning_rate": 3.0636833046471603e-06,
+      "loss": 0.2907,
+      "step": 3051
+    },
+    {
+      "epoch": 2.834184858337204,
+      "grad_norm": 0.16156022525948335,
+      "learning_rate": 3.0464716006884687e-06,
+      "loss": 0.2707,
+      "step": 3052
+    },
+    {
+      "epoch": 2.835113794705063,
+      "grad_norm": 0.1628946794644337,
+      "learning_rate": 3.0292598967297766e-06,
+      "loss": 0.2525,
+      "step": 3053
+    },
+    {
+      "epoch": 2.8360427310729213,
+      "grad_norm": 0.17817137797577384,
+      "learning_rate": 3.0120481927710846e-06,
+      "loss": 0.287,
+      "step": 3054
+    },
+    {
+      "epoch": 2.83697166744078,
+      "grad_norm": 0.1574612277457272,
+      "learning_rate": 2.9948364888123926e-06,
+      "loss": 0.2691,
+      "step": 3055
+    },
+    {
+      "epoch": 2.837900603808639,
+      "grad_norm": 0.1596146649010269,
+      "learning_rate": 2.977624784853701e-06,
+      "loss": 0.2618,
+      "step": 3056
+    },
+    {
+      "epoch": 2.838829540176498,
+      "grad_norm": 0.1666069390299578,
+      "learning_rate": 2.960413080895009e-06,
+      "loss": 0.2788,
+      "step": 3057
+    },
+    {
+      "epoch": 2.8397584765443566,
+      "grad_norm": 0.16942939086250466,
+      "learning_rate": 2.943201376936317e-06,
+      "loss": 0.278,
+      "step": 3058
+    },
+    {
+      "epoch": 2.8406874129122155,
+      "grad_norm": 0.16780692378544415,
+      "learning_rate": 2.925989672977625e-06,
+      "loss": 0.2864,
+      "step": 3059
+    },
+    {
+      "epoch": 2.8416163492800743,
+      "grad_norm": 0.16519686217403176,
+      "learning_rate": 2.9087779690189333e-06,
+      "loss": 0.2743,
+      "step": 3060
+    },
+    {
+      "epoch": 2.842545285647933,
+      "grad_norm": 0.18887259058754785,
+      "learning_rate": 2.8915662650602413e-06,
+      "loss": 0.28,
+      "step": 3061
+    },
+    {
+      "epoch": 2.843474222015792,
+      "grad_norm": 0.16104392704343692,
+      "learning_rate": 2.8743545611015493e-06,
+      "loss": 0.2651,
+      "step": 3062
+    },
+    {
+      "epoch": 2.844403158383651,
+      "grad_norm": 0.16267194525224002,
+      "learning_rate": 2.8571428571428573e-06,
+      "loss": 0.2743,
+      "step": 3063
+    },
+    {
+      "epoch": 2.8453320947515097,
+      "grad_norm": 0.16408592399590138,
+      "learning_rate": 2.8399311531841657e-06,
+      "loss": 0.2703,
+      "step": 3064
+    },
+    {
+      "epoch": 2.8462610311193686,
+      "grad_norm": 0.17434242637401323,
+      "learning_rate": 2.8227194492254736e-06,
+      "loss": 0.2797,
+      "step": 3065
+    },
+    {
+      "epoch": 2.847189967487227,
+      "grad_norm": 0.16647472284974027,
+      "learning_rate": 2.8055077452667816e-06,
+      "loss": 0.2623,
+      "step": 3066
+    },
+    {
+      "epoch": 2.848118903855086,
+      "grad_norm": 0.16968808669769664,
+      "learning_rate": 2.7882960413080896e-06,
+      "loss": 0.278,
+      "step": 3067
+    },
+    {
+      "epoch": 2.8490478402229447,
+      "grad_norm": 0.17519157938858387,
+      "learning_rate": 2.7710843373493976e-06,
+      "loss": 0.2864,
+      "step": 3068
+    },
+    {
+      "epoch": 2.8499767765908035,
+      "grad_norm": 0.15968806347904377,
+      "learning_rate": 2.753872633390706e-06,
+      "loss": 0.2781,
+      "step": 3069
+    },
+    {
+      "epoch": 2.8509057129586624,
+      "grad_norm": 0.1690915914984628,
+      "learning_rate": 2.736660929432014e-06,
+      "loss": 0.2761,
+      "step": 3070
+    },
+    {
+      "epoch": 2.851834649326521,
+      "grad_norm": 0.16230516310874854,
+      "learning_rate": 2.719449225473322e-06,
+      "loss": 0.268,
+      "step": 3071
+    },
+    {
+      "epoch": 2.85276358569438,
+      "grad_norm": 0.17552261229791508,
+      "learning_rate": 2.70223752151463e-06,
+      "loss": 0.2862,
+      "step": 3072
+    },
+    {
+      "epoch": 2.853692522062239,
+      "grad_norm": 0.16542085819719532,
+      "learning_rate": 2.6850258175559383e-06,
+      "loss": 0.2707,
+      "step": 3073
+    },
+    {
+      "epoch": 2.8546214584300973,
+      "grad_norm": 0.17209167653272706,
+      "learning_rate": 2.6678141135972463e-06,
+      "loss": 0.2745,
+      "step": 3074
+    },
+    {
+      "epoch": 2.855550394797956,
+      "grad_norm": 0.1801208337324367,
+      "learning_rate": 2.6506024096385542e-06,
+      "loss": 0.275,
+      "step": 3075
+    },
+    {
+      "epoch": 2.856479331165815,
+      "grad_norm": 0.1658175112453202,
+      "learning_rate": 2.6333907056798622e-06,
+      "loss": 0.28,
+      "step": 3076
+    },
+    {
+      "epoch": 2.857408267533674,
+      "grad_norm": 0.1682675080594553,
+      "learning_rate": 2.6161790017211706e-06,
+      "loss": 0.272,
+      "step": 3077
+    },
+    {
+      "epoch": 2.8583372039015327,
+      "grad_norm": 0.16424665159704155,
+      "learning_rate": 2.5989672977624786e-06,
+      "loss": 0.2722,
+      "step": 3078
+    },
+    {
+      "epoch": 2.8592661402693915,
+      "grad_norm": 0.17646757447622446,
+      "learning_rate": 2.5817555938037866e-06,
+      "loss": 0.2743,
+      "step": 3079
+    },
+    {
+      "epoch": 2.8601950766372504,
+      "grad_norm": 0.16214982779968065,
+      "learning_rate": 2.5645438898450946e-06,
+      "loss": 0.2636,
+      "step": 3080
+    },
+    {
+      "epoch": 2.861124013005109,
+      "grad_norm": 0.16427321944645545,
+      "learning_rate": 2.547332185886403e-06,
+      "loss": 0.2687,
+      "step": 3081
+    },
+    {
+      "epoch": 2.862052949372968,
+      "grad_norm": 0.1651934825129748,
+      "learning_rate": 2.530120481927711e-06,
+      "loss": 0.2716,
+      "step": 3082
+    },
+    {
+      "epoch": 2.862981885740827,
+      "grad_norm": 0.1801065723272109,
+      "learning_rate": 2.512908777969019e-06,
+      "loss": 0.2907,
+      "step": 3083
+    },
+    {
+      "epoch": 2.8639108221086857,
+      "grad_norm": 0.1625252701117974,
+      "learning_rate": 2.495697074010327e-06,
+      "loss": 0.2694,
+      "step": 3084
+    },
+    {
+      "epoch": 2.8648397584765446,
+      "grad_norm": 0.1693264127024217,
+      "learning_rate": 2.4784853700516353e-06,
+      "loss": 0.2719,
+      "step": 3085
+    },
+    {
+      "epoch": 2.865768694844403,
+      "grad_norm": 0.18597914567062376,
+      "learning_rate": 2.4612736660929433e-06,
+      "loss": 0.2759,
+      "step": 3086
+    },
+    {
+      "epoch": 2.866697631212262,
+      "grad_norm": 0.1680321028316844,
+      "learning_rate": 2.4440619621342512e-06,
+      "loss": 0.2582,
+      "step": 3087
+    },
+    {
+      "epoch": 2.8676265675801207,
+      "grad_norm": 0.17018851398423457,
+      "learning_rate": 2.426850258175559e-06,
+      "loss": 0.2741,
+      "step": 3088
+    },
+    {
+      "epoch": 2.8685555039479795,
+      "grad_norm": 0.17493183425794315,
+      "learning_rate": 2.4096385542168676e-06,
+      "loss": 0.2824,
+      "step": 3089
+    },
+    {
+      "epoch": 2.8694844403158384,
+      "grad_norm": 0.17392176050130073,
+      "learning_rate": 2.3924268502581756e-06,
+      "loss": 0.2861,
+      "step": 3090
+    },
+    {
+      "epoch": 2.8704133766836972,
+      "grad_norm": 0.16101372989259485,
+      "learning_rate": 2.3752151462994836e-06,
+      "loss": 0.2729,
+      "step": 3091
+    },
+    {
+      "epoch": 2.871342313051556,
+      "grad_norm": 0.1811947560081589,
+      "learning_rate": 2.3580034423407915e-06,
+      "loss": 0.29,
+      "step": 3092
+    },
+    {
+      "epoch": 2.872271249419415,
+      "grad_norm": 0.16656066225329413,
+      "learning_rate": 2.3407917383821e-06,
+      "loss": 0.2582,
+      "step": 3093
+    },
+    {
+      "epoch": 2.8732001857872733,
+      "grad_norm": 0.15996052741989927,
+      "learning_rate": 2.323580034423408e-06,
+      "loss": 0.2597,
+      "step": 3094
+    },
+    {
+      "epoch": 2.874129122155132,
+      "grad_norm": 0.16380652320388664,
+      "learning_rate": 2.306368330464716e-06,
+      "loss": 0.2705,
+      "step": 3095
+    },
+    {
+      "epoch": 2.875058058522991,
+      "grad_norm": 0.1628946736235343,
+      "learning_rate": 2.2891566265060243e-06,
+      "loss": 0.2723,
+      "step": 3096
+    },
+    {
+      "epoch": 2.87598699489085,
+      "grad_norm": 0.17560963055242468,
+      "learning_rate": 2.2719449225473323e-06,
+      "loss": 0.2798,
+      "step": 3097
+    },
+    {
+      "epoch": 2.8769159312587087,
+      "grad_norm": 0.16780655023233185,
+      "learning_rate": 2.2547332185886402e-06,
+      "loss": 0.2657,
+      "step": 3098
+    },
+    {
+      "epoch": 2.8778448676265675,
+      "grad_norm": 0.17403931947708165,
+      "learning_rate": 2.2375215146299486e-06,
+      "loss": 0.2785,
+      "step": 3099
+    },
+    {
+      "epoch": 2.8787738039944264,
+      "grad_norm": 0.17010520777366847,
+      "learning_rate": 2.2203098106712566e-06,
+      "loss": 0.2614,
+      "step": 3100
+    },
+    {
+      "epoch": 2.8797027403622852,
+      "grad_norm": 0.16812351916262241,
+      "learning_rate": 2.2030981067125646e-06,
+      "loss": 0.281,
+      "step": 3101
+    },
+    {
+      "epoch": 2.880631676730144,
+      "grad_norm": 0.16691194577486648,
+      "learning_rate": 2.185886402753873e-06,
+      "loss": 0.2662,
+      "step": 3102
+    },
+    {
+      "epoch": 2.881560613098003,
+      "grad_norm": 0.1759493066311429,
+      "learning_rate": 2.168674698795181e-06,
+      "loss": 0.2747,
+      "step": 3103
+    },
+    {
+      "epoch": 2.8824895494658618,
+      "grad_norm": 0.1707537478590751,
+      "learning_rate": 2.151462994836489e-06,
+      "loss": 0.2707,
+      "step": 3104
+    },
+    {
+      "epoch": 2.8834184858337206,
+      "grad_norm": 0.18135021613947166,
+      "learning_rate": 2.1342512908777974e-06,
+      "loss": 0.2905,
+      "step": 3105
+    },
+    {
+      "epoch": 2.884347422201579,
+      "grad_norm": 0.16500668342734362,
+      "learning_rate": 2.1170395869191053e-06,
+      "loss": 0.287,
+      "step": 3106
+    },
+    {
+      "epoch": 2.885276358569438,
+      "grad_norm": 0.1719432107365479,
+      "learning_rate": 2.0998278829604133e-06,
+      "loss": 0.2813,
+      "step": 3107
+    },
+    {
+      "epoch": 2.8862052949372967,
+      "grad_norm": 0.15985021847497927,
+      "learning_rate": 2.0826161790017213e-06,
+      "loss": 0.263,
+      "step": 3108
+    },
+    {
+      "epoch": 2.8871342313051556,
+      "grad_norm": 0.16354165474456403,
+      "learning_rate": 2.0654044750430297e-06,
+      "loss": 0.2836,
+      "step": 3109
+    },
+    {
+      "epoch": 2.8880631676730144,
+      "grad_norm": 0.16878198938643335,
+      "learning_rate": 2.0481927710843377e-06,
+      "loss": 0.2785,
+      "step": 3110
+    },
+    {
+      "epoch": 2.8889921040408733,
+      "grad_norm": 0.16589172961440266,
+      "learning_rate": 2.0309810671256456e-06,
+      "loss": 0.2784,
+      "step": 3111
+    },
+    {
+      "epoch": 2.889921040408732,
+      "grad_norm": 0.17438950890163252,
+      "learning_rate": 2.0137693631669536e-06,
+      "loss": 0.2714,
+      "step": 3112
+    },
+    {
+      "epoch": 2.890849976776591,
+      "grad_norm": 0.16242581403999212,
+      "learning_rate": 1.996557659208262e-06,
+      "loss": 0.2629,
+      "step": 3113
+    },
+    {
+      "epoch": 2.8917789131444493,
+      "grad_norm": 0.1744527543075875,
+      "learning_rate": 1.97934595524957e-06,
+      "loss": 0.2707,
+      "step": 3114
+    },
+    {
+      "epoch": 2.892707849512308,
+      "grad_norm": 0.16426414428304387,
+      "learning_rate": 1.962134251290878e-06,
+      "loss": 0.2816,
+      "step": 3115
+    },
+    {
+      "epoch": 2.893636785880167,
+      "grad_norm": 0.1621871796090516,
+      "learning_rate": 1.944922547332186e-06,
+      "loss": 0.2751,
+      "step": 3116
+    },
+    {
+      "epoch": 2.894565722248026,
+      "grad_norm": 0.1775919652596439,
+      "learning_rate": 1.9277108433734943e-06,
+      "loss": 0.2895,
+      "step": 3117
+    },
+    {
+      "epoch": 2.8954946586158847,
+      "grad_norm": 0.17006770205203484,
+      "learning_rate": 1.9104991394148023e-06,
+      "loss": 0.2688,
+      "step": 3118
+    },
+    {
+      "epoch": 2.8964235949837436,
+      "grad_norm": 0.16894126178473537,
+      "learning_rate": 1.8932874354561103e-06,
+      "loss": 0.2609,
+      "step": 3119
+    },
+    {
+      "epoch": 2.8973525313516024,
+      "grad_norm": 0.16582301619354106,
+      "learning_rate": 1.8760757314974183e-06,
+      "loss": 0.2732,
+      "step": 3120
+    },
+    {
+      "epoch": 2.8982814677194613,
+      "grad_norm": 0.17231531336949352,
+      "learning_rate": 1.8588640275387262e-06,
+      "loss": 0.2734,
+      "step": 3121
+    },
+    {
+      "epoch": 2.89921040408732,
+      "grad_norm": 0.16228298439571703,
+      "learning_rate": 1.8416523235800346e-06,
+      "loss": 0.2662,
+      "step": 3122
+    },
+    {
+      "epoch": 2.900139340455179,
+      "grad_norm": 0.1655198446070943,
+      "learning_rate": 1.8244406196213426e-06,
+      "loss": 0.2678,
+      "step": 3123
+    },
+    {
+      "epoch": 2.901068276823038,
+      "grad_norm": 0.17406281157074208,
+      "learning_rate": 1.8072289156626506e-06,
+      "loss": 0.2802,
+      "step": 3124
+    },
+    {
+      "epoch": 2.9019972131908967,
+      "grad_norm": 0.16674227317723664,
+      "learning_rate": 1.7900172117039586e-06,
+      "loss": 0.266,
+      "step": 3125
+    },
+    {
+      "epoch": 2.9029261495587555,
+      "grad_norm": 0.1706960267209148,
+      "learning_rate": 1.772805507745267e-06,
+      "loss": 0.2974,
+      "step": 3126
+    },
+    {
+      "epoch": 2.903855085926614,
+      "grad_norm": 0.1658872040790518,
+      "learning_rate": 1.755593803786575e-06,
+      "loss": 0.2694,
+      "step": 3127
+    },
+    {
+      "epoch": 2.9047840222944727,
+      "grad_norm": 0.16658354730048636,
+      "learning_rate": 1.738382099827883e-06,
+      "loss": 0.2587,
+      "step": 3128
+    },
+    {
+      "epoch": 2.9057129586623316,
+      "grad_norm": 0.16444858436297213,
+      "learning_rate": 1.721170395869191e-06,
+      "loss": 0.2813,
+      "step": 3129
+    },
+    {
+      "epoch": 2.9066418950301904,
+      "grad_norm": 0.16336194076854949,
+      "learning_rate": 1.7039586919104993e-06,
+      "loss": 0.2732,
+      "step": 3130
+    },
+    {
+      "epoch": 2.9075708313980493,
+      "grad_norm": 0.171145553329007,
+      "learning_rate": 1.6867469879518073e-06,
+      "loss": 0.2723,
+      "step": 3131
+    },
+    {
+      "epoch": 2.908499767765908,
+      "grad_norm": 0.16347568718682626,
+      "learning_rate": 1.6695352839931153e-06,
+      "loss": 0.2762,
+      "step": 3132
+    },
+    {
+      "epoch": 2.909428704133767,
+      "grad_norm": 0.16509176449938368,
+      "learning_rate": 1.6523235800344232e-06,
+      "loss": 0.2684,
+      "step": 3133
+    },
+    {
+      "epoch": 2.9103576405016254,
+      "grad_norm": 0.16259817795344064,
+      "learning_rate": 1.6351118760757316e-06,
+      "loss": 0.2672,
+      "step": 3134
+    },
+    {
+      "epoch": 2.911286576869484,
+      "grad_norm": 0.16536212949320228,
+      "learning_rate": 1.6179001721170396e-06,
+      "loss": 0.2712,
+      "step": 3135
+    },
+    {
+      "epoch": 2.912215513237343,
+      "grad_norm": 0.17142664756129727,
+      "learning_rate": 1.6006884681583476e-06,
+      "loss": 0.2781,
+      "step": 3136
+    },
+    {
+      "epoch": 2.913144449605202,
+      "grad_norm": 0.16836163730587553,
+      "learning_rate": 1.5834767641996558e-06,
+      "loss": 0.2808,
+      "step": 3137
+    },
+    {
+      "epoch": 2.9140733859730608,
+      "grad_norm": 0.16607118988603203,
+      "learning_rate": 1.566265060240964e-06,
+      "loss": 0.2766,
+      "step": 3138
+    },
+    {
+      "epoch": 2.9150023223409196,
+      "grad_norm": 0.1672718943537726,
+      "learning_rate": 1.549053356282272e-06,
+      "loss": 0.2704,
+      "step": 3139
+    },
+    {
+      "epoch": 2.9159312587087785,
+      "grad_norm": 0.16835726638255913,
+      "learning_rate": 1.5318416523235801e-06,
+      "loss": 0.2769,
+      "step": 3140
+    },
+    {
+      "epoch": 2.9168601950766373,
+      "grad_norm": 0.161048165231964,
+      "learning_rate": 1.5146299483648883e-06,
+      "loss": 0.2837,
+      "step": 3141
+    },
+    {
+      "epoch": 2.917789131444496,
+      "grad_norm": 0.16282977150245906,
+      "learning_rate": 1.4974182444061963e-06,
+      "loss": 0.2638,
+      "step": 3142
+    },
+    {
+      "epoch": 2.918718067812355,
+      "grad_norm": 0.17323001324895482,
+      "learning_rate": 1.4802065404475045e-06,
+      "loss": 0.2815,
+      "step": 3143
+    },
+    {
+      "epoch": 2.919647004180214,
+      "grad_norm": 0.16388686409084324,
+      "learning_rate": 1.4629948364888125e-06,
+      "loss": 0.2827,
+      "step": 3144
+    },
+    {
+      "epoch": 2.9205759405480727,
+      "grad_norm": 0.16384234361312092,
+      "learning_rate": 1.4457831325301207e-06,
+      "loss": 0.282,
+      "step": 3145
+    },
+    {
+      "epoch": 2.9215048769159315,
+      "grad_norm": 0.16568266746026178,
+      "learning_rate": 1.4285714285714286e-06,
+      "loss": 0.2704,
+      "step": 3146
+    },
+    {
+      "epoch": 2.92243381328379,
+      "grad_norm": 0.16915537105295747,
+      "learning_rate": 1.4113597246127368e-06,
+      "loss": 0.2796,
+      "step": 3147
+    },
+    {
+      "epoch": 2.9233627496516488,
+      "grad_norm": 0.15803350350704365,
+      "learning_rate": 1.3941480206540448e-06,
+      "loss": 0.2632,
+      "step": 3148
+    },
+    {
+      "epoch": 2.9242916860195076,
+      "grad_norm": 0.1724406036070289,
+      "learning_rate": 1.376936316695353e-06,
+      "loss": 0.2811,
+      "step": 3149
+    },
+    {
+      "epoch": 2.9252206223873665,
+      "grad_norm": 0.16469469642126497,
+      "learning_rate": 1.359724612736661e-06,
+      "loss": 0.255,
+      "step": 3150
+    },
+    {
+      "epoch": 2.9261495587552253,
+      "grad_norm": 0.18760969517301657,
+      "learning_rate": 1.3425129087779691e-06,
+      "loss": 0.2682,
+      "step": 3151
+    },
+    {
+      "epoch": 2.927078495123084,
+      "grad_norm": 0.16665139220788266,
+      "learning_rate": 1.3253012048192771e-06,
+      "loss": 0.268,
+      "step": 3152
+    },
+    {
+      "epoch": 2.928007431490943,
+      "grad_norm": 0.17530978095431596,
+      "learning_rate": 1.3080895008605853e-06,
+      "loss": 0.2624,
+      "step": 3153
+    },
+    {
+      "epoch": 2.9289363678588014,
+      "grad_norm": 0.1631271421954177,
+      "learning_rate": 1.2908777969018933e-06,
+      "loss": 0.2727,
+      "step": 3154
+    },
+    {
+      "epoch": 2.9298653042266602,
+      "grad_norm": 0.15966257157616312,
+      "learning_rate": 1.2736660929432015e-06,
+      "loss": 0.2688,
+      "step": 3155
+    },
+    {
+      "epoch": 2.930794240594519,
+      "grad_norm": 0.15325566348484762,
+      "learning_rate": 1.2564543889845095e-06,
+      "loss": 0.2559,
+      "step": 3156
+    },
+    {
+      "epoch": 2.931723176962378,
+      "grad_norm": 0.16790370946361619,
+      "learning_rate": 1.2392426850258176e-06,
+      "loss": 0.286,
+      "step": 3157
+    },
+    {
+      "epoch": 2.932652113330237,
+      "grad_norm": 0.16679108887381452,
+      "learning_rate": 1.2220309810671256e-06,
+      "loss": 0.2863,
+      "step": 3158
+    },
+    {
+      "epoch": 2.9335810496980956,
+      "grad_norm": 0.17093563722555055,
+      "learning_rate": 1.2048192771084338e-06,
+      "loss": 0.283,
+      "step": 3159
+    },
+    {
+      "epoch": 2.9345099860659545,
+      "grad_norm": 0.1671496062211597,
+      "learning_rate": 1.1876075731497418e-06,
+      "loss": 0.2792,
+      "step": 3160
+    },
+    {
+      "epoch": 2.9354389224338133,
+      "grad_norm": 0.1599449296648143,
+      "learning_rate": 1.17039586919105e-06,
+      "loss": 0.2448,
+      "step": 3161
+    },
+    {
+      "epoch": 2.936367858801672,
+      "grad_norm": 0.15847396108522563,
+      "learning_rate": 1.153184165232358e-06,
+      "loss": 0.2704,
+      "step": 3162
+    },
+    {
+      "epoch": 2.937296795169531,
+      "grad_norm": 0.16199214200992662,
+      "learning_rate": 1.1359724612736661e-06,
+      "loss": 0.2516,
+      "step": 3163
+    },
+    {
+      "epoch": 2.93822573153739,
+      "grad_norm": 0.1626770624143151,
+      "learning_rate": 1.1187607573149743e-06,
+      "loss": 0.2682,
+      "step": 3164
+    },
+    {
+      "epoch": 2.9391546679052487,
+      "grad_norm": 0.16395321212716582,
+      "learning_rate": 1.1015490533562823e-06,
+      "loss": 0.2722,
+      "step": 3165
+    },
+    {
+      "epoch": 2.9400836042731076,
+      "grad_norm": 0.15821372199935677,
+      "learning_rate": 1.0843373493975905e-06,
+      "loss": 0.2687,
+      "step": 3166
+    },
+    {
+      "epoch": 2.941012540640966,
+      "grad_norm": 0.15843662175473597,
+      "learning_rate": 1.0671256454388987e-06,
+      "loss": 0.2681,
+      "step": 3167
+    },
+    {
+      "epoch": 2.941941477008825,
+      "grad_norm": 0.16333393084594955,
+      "learning_rate": 1.0499139414802067e-06,
+      "loss": 0.2856,
+      "step": 3168
+    },
+    {
+      "epoch": 2.9428704133766836,
+      "grad_norm": 0.17542548208560677,
+      "learning_rate": 1.0327022375215148e-06,
+      "loss": 0.2798,
+      "step": 3169
+    },
+    {
+      "epoch": 2.9437993497445425,
+      "grad_norm": 0.17024423281462964,
+      "learning_rate": 1.0154905335628228e-06,
+      "loss": 0.2742,
+      "step": 3170
+    },
+    {
+      "epoch": 2.9447282861124013,
+      "grad_norm": 0.1576061324066094,
+      "learning_rate": 9.98278829604131e-07,
+      "loss": 0.2781,
+      "step": 3171
+    },
+    {
+      "epoch": 2.94565722248026,
+      "grad_norm": 0.1601221693884007,
+      "learning_rate": 9.81067125645439e-07,
+      "loss": 0.2693,
+      "step": 3172
+    },
+    {
+      "epoch": 2.946586158848119,
+      "grad_norm": 0.16786872565481428,
+      "learning_rate": 9.638554216867472e-07,
+      "loss": 0.2753,
+      "step": 3173
+    },
+    {
+      "epoch": 2.9475150952159774,
+      "grad_norm": 0.16660337771357872,
+      "learning_rate": 9.466437177280551e-07,
+      "loss": 0.2758,
+      "step": 3174
+    },
+    {
+      "epoch": 2.9484440315838363,
+      "grad_norm": 0.1691333863317382,
+      "learning_rate": 9.294320137693631e-07,
+      "loss": 0.2795,
+      "step": 3175
+    },
+    {
+      "epoch": 2.949372967951695,
+      "grad_norm": 0.1722484989104376,
+      "learning_rate": 9.122203098106713e-07,
+      "loss": 0.2802,
+      "step": 3176
+    },
+    {
+      "epoch": 2.950301904319554,
+      "grad_norm": 0.16958831531505558,
+      "learning_rate": 8.950086058519793e-07,
+      "loss": 0.2901,
+      "step": 3177
+    },
+    {
+      "epoch": 2.951230840687413,
+      "grad_norm": 0.17023309066648035,
+      "learning_rate": 8.777969018932875e-07,
+      "loss": 0.2717,
+      "step": 3178
+    },
+    {
+      "epoch": 2.9521597770552717,
+      "grad_norm": 0.15809523592675392,
+      "learning_rate": 8.605851979345955e-07,
+      "loss": 0.265,
+      "step": 3179
+    },
+    {
+      "epoch": 2.9530887134231305,
+      "grad_norm": 0.1675472901902543,
+      "learning_rate": 8.433734939759036e-07,
+      "loss": 0.2849,
+      "step": 3180
+    },
+    {
+      "epoch": 2.9540176497909894,
+      "grad_norm": 0.1654082507246669,
+      "learning_rate": 8.261617900172116e-07,
+      "loss": 0.2791,
+      "step": 3181
+    },
+    {
+      "epoch": 2.954946586158848,
+      "grad_norm": 0.162707952994883,
+      "learning_rate": 8.089500860585198e-07,
+      "loss": 0.2813,
+      "step": 3182
+    },
+    {
+      "epoch": 2.955875522526707,
+      "grad_norm": 0.16446726694121308,
+      "learning_rate": 7.917383820998279e-07,
+      "loss": 0.2872,
+      "step": 3183
+    },
+    {
+      "epoch": 2.956804458894566,
+      "grad_norm": 0.1637824540494551,
+      "learning_rate": 7.74526678141136e-07,
+      "loss": 0.2795,
+      "step": 3184
+    },
+    {
+      "epoch": 2.9577333952624247,
+      "grad_norm": 0.18638420051088697,
+      "learning_rate": 7.573149741824442e-07,
+      "loss": 0.2709,
+      "step": 3185
+    },
+    {
+      "epoch": 2.9586623316302836,
+      "grad_norm": 0.15707600467510602,
+      "learning_rate": 7.401032702237522e-07,
+      "loss": 0.2628,
+      "step": 3186
+    },
+    {
+      "epoch": 2.959591267998142,
+      "grad_norm": 0.16140242751307413,
+      "learning_rate": 7.228915662650603e-07,
+      "loss": 0.2735,
+      "step": 3187
+    },
+    {
+      "epoch": 2.960520204366001,
+      "grad_norm": 0.16654689577483472,
+      "learning_rate": 7.056798623063684e-07,
+      "loss": 0.2831,
+      "step": 3188
+    },
+    {
+      "epoch": 2.9614491407338597,
+      "grad_norm": 0.16057309604016964,
+      "learning_rate": 6.884681583476765e-07,
+      "loss": 0.2635,
+      "step": 3189
+    },
+    {
+      "epoch": 2.9623780771017185,
+      "grad_norm": 0.16490464189295742,
+      "learning_rate": 6.712564543889846e-07,
+      "loss": 0.2752,
+      "step": 3190
+    },
+    {
+      "epoch": 2.9633070134695774,
+      "grad_norm": 0.1671405024647208,
+      "learning_rate": 6.540447504302927e-07,
+      "loss": 0.269,
+      "step": 3191
+    },
+    {
+      "epoch": 2.964235949837436,
+      "grad_norm": 0.1710576565622639,
+      "learning_rate": 6.368330464716007e-07,
+      "loss": 0.2843,
+      "step": 3192
+    },
+    {
+      "epoch": 2.965164886205295,
+      "grad_norm": 0.16276635918619017,
+      "learning_rate": 6.196213425129088e-07,
+      "loss": 0.2813,
+      "step": 3193
+    },
+    {
+      "epoch": 2.966093822573154,
+      "grad_norm": 0.15281133331696603,
+      "learning_rate": 6.024096385542169e-07,
+      "loss": 0.2639,
+      "step": 3194
+    },
+    {
+      "epoch": 2.9670227589410123,
+      "grad_norm": 0.16323794995058483,
+      "learning_rate": 5.85197934595525e-07,
+      "loss": 0.2663,
+      "step": 3195
+    },
+    {
+      "epoch": 2.967951695308871,
+      "grad_norm": 0.1624282394962894,
+      "learning_rate": 5.679862306368331e-07,
+      "loss": 0.2752,
+      "step": 3196
+    },
+    {
+      "epoch": 2.96888063167673,
+      "grad_norm": 0.16134159864744876,
+      "learning_rate": 5.507745266781412e-07,
+      "loss": 0.2702,
+      "step": 3197
+    },
+    {
+      "epoch": 2.969809568044589,
+      "grad_norm": 0.16589417651717428,
+      "learning_rate": 5.335628227194493e-07,
+      "loss": 0.2828,
+      "step": 3198
+    },
+    {
+      "epoch": 2.9707385044124477,
+      "grad_norm": 0.16183155315903366,
+      "learning_rate": 5.163511187607574e-07,
+      "loss": 0.2852,
+      "step": 3199
+    },
+    {
+      "epoch": 2.9716674407803065,
+      "grad_norm": 0.16678141742547176,
+      "learning_rate": 4.991394148020655e-07,
+      "loss": 0.2813,
+      "step": 3200
+    },
+    {
+      "epoch": 2.9725963771481654,
+      "grad_norm": 0.1619658189257669,
+      "learning_rate": 4.819277108433736e-07,
+      "loss": 0.2649,
+      "step": 3201
+    },
+    {
+      "epoch": 2.9735253135160242,
+      "grad_norm": 0.15993238852155908,
+      "learning_rate": 4.6471600688468156e-07,
+      "loss": 0.2764,
+      "step": 3202
+    },
+    {
+      "epoch": 2.974454249883883,
+      "grad_norm": 0.16228626601674104,
+      "learning_rate": 4.4750430292598964e-07,
+      "loss": 0.2738,
+      "step": 3203
+    },
+    {
+      "epoch": 2.975383186251742,
+      "grad_norm": 0.16945807739892618,
+      "learning_rate": 4.3029259896729773e-07,
+      "loss": 0.2875,
+      "step": 3204
+    },
+    {
+      "epoch": 2.9763121226196008,
+      "grad_norm": 0.16673802928277734,
+      "learning_rate": 4.130808950086058e-07,
+      "loss": 0.2751,
+      "step": 3205
+    },
+    {
+      "epoch": 2.9772410589874596,
+      "grad_norm": 0.16066420702808745,
+      "learning_rate": 3.9586919104991394e-07,
+      "loss": 0.2691,
+      "step": 3206
+    },
+    {
+      "epoch": 2.978169995355318,
+      "grad_norm": 0.167644174523662,
+      "learning_rate": 3.786574870912221e-07,
+      "loss": 0.276,
+      "step": 3207
+    },
+    {
+      "epoch": 2.979098931723177,
+      "grad_norm": 0.16405491071346914,
+      "learning_rate": 3.6144578313253016e-07,
+      "loss": 0.27,
+      "step": 3208
+    },
+    {
+      "epoch": 2.9800278680910357,
+      "grad_norm": 0.15776111203718893,
+      "learning_rate": 3.4423407917383825e-07,
+      "loss": 0.2754,
+      "step": 3209
+    },
+    {
+      "epoch": 2.9809568044588945,
+      "grad_norm": 0.16291385160189412,
+      "learning_rate": 3.2702237521514633e-07,
+      "loss": 0.2801,
+      "step": 3210
+    },
+    {
+      "epoch": 2.9818857408267534,
+      "grad_norm": 0.15798932483422015,
+      "learning_rate": 3.098106712564544e-07,
+      "loss": 0.2655,
+      "step": 3211
+    },
+    {
+      "epoch": 2.9828146771946122,
+      "grad_norm": 0.16945630512529875,
+      "learning_rate": 2.925989672977625e-07,
+      "loss": 0.274,
+      "step": 3212
+    },
+    {
+      "epoch": 2.983743613562471,
+      "grad_norm": 0.1708554921170849,
+      "learning_rate": 2.753872633390706e-07,
+      "loss": 0.272,
+      "step": 3213
+    },
+    {
+      "epoch": 2.98467254993033,
+      "grad_norm": 0.15872652564709347,
+      "learning_rate": 2.581755593803787e-07,
+      "loss": 0.258,
+      "step": 3214
+    },
+    {
+      "epoch": 2.9856014862981883,
+      "grad_norm": 0.1830290046487256,
+      "learning_rate": 2.409638554216868e-07,
+      "loss": 0.2794,
+      "step": 3215
+    },
+    {
+      "epoch": 2.986530422666047,
+      "grad_norm": 0.15745598014130835,
+      "learning_rate": 2.2375215146299482e-07,
+      "loss": 0.2645,
+      "step": 3216
+    },
+    {
+      "epoch": 2.987459359033906,
+      "grad_norm": 0.16220353029392465,
+      "learning_rate": 2.065404475043029e-07,
+      "loss": 0.2789,
+      "step": 3217
+    },
+    {
+      "epoch": 2.988388295401765,
+      "grad_norm": 0.16537567907705764,
+      "learning_rate": 1.8932874354561104e-07,
+      "loss": 0.2724,
+      "step": 3218
+    },
+    {
+      "epoch": 2.9893172317696237,
+      "grad_norm": 0.15966135311906623,
+      "learning_rate": 1.7211703958691912e-07,
+      "loss": 0.2603,
+      "step": 3219
+    },
+    {
+      "epoch": 2.9902461681374826,
+      "grad_norm": 0.16670677322460914,
+      "learning_rate": 1.549053356282272e-07,
+      "loss": 0.277,
+      "step": 3220
+    },
+    {
+      "epoch": 2.9911751045053414,
+      "grad_norm": 0.16236390105210605,
+      "learning_rate": 1.376936316695353e-07,
+      "loss": 0.2778,
+      "step": 3221
+    },
+    {
+      "epoch": 2.9921040408732003,
+      "grad_norm": 0.15740425470314215,
+      "learning_rate": 1.204819277108434e-07,
+      "loss": 0.263,
+      "step": 3222
+    },
+    {
+      "epoch": 2.993032977241059,
+      "grad_norm": 0.16389828021365552,
+      "learning_rate": 1.0327022375215145e-07,
+      "loss": 0.276,
+      "step": 3223
+    },
+    {
+      "epoch": 2.993961913608918,
+      "grad_norm": 0.1578402137230674,
+      "learning_rate": 8.605851979345956e-08,
+      "loss": 0.26,
+      "step": 3224
+    },
+    {
+      "epoch": 2.994890849976777,
+      "grad_norm": 0.1705942119258943,
+      "learning_rate": 6.884681583476764e-08,
+      "loss": 0.2776,
+      "step": 3225
+    },
+    {
+      "epoch": 2.9958197863446356,
+      "grad_norm": 0.16036831167171559,
+      "learning_rate": 5.1635111876075726e-08,
+      "loss": 0.2761,
+      "step": 3226
+    },
+    {
+      "epoch": 2.996748722712494,
+      "grad_norm": 0.17533866907265988,
+      "learning_rate": 3.442340791738382e-08,
+      "loss": 0.2913,
+      "step": 3227
+    },
+    {
+      "epoch": 2.997677659080353,
+      "grad_norm": 0.1660516196575767,
+      "learning_rate": 1.721170395869191e-08,
+      "loss": 0.2709,
+      "step": 3228
+    },
+    {
+      "epoch": 2.997677659080353,
+      "step": 3228,
+      "total_flos": 3.5889551683046343e+19,
+      "train_loss": 0.40990160300847617,
+      "train_runtime": 91578.7681,
+      "train_samples_per_second": 0.564,
+      "train_steps_per_second": 0.035
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 3228,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.5889551683046343e+19,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}