{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.8905852417302799,
  "eval_steps": 500,
  "global_step": 3500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.002544529262086514,
      "grad_norm": 8.495231628417969,
      "learning_rate": 9.160305343511451e-07,
      "loss": 0.1152,
      "step": 10
    },
    {
      "epoch": 0.005089058524173028,
      "grad_norm": 9.448420524597168,
      "learning_rate": 1.933842239185751e-06,
      "loss": 0.1589,
      "step": 20
    },
    {
      "epoch": 0.007633587786259542,
      "grad_norm": 12.052129745483398,
      "learning_rate": 2.951653944020356e-06,
      "loss": 0.1132,
      "step": 30
    },
    {
      "epoch": 0.010178117048346057,
      "grad_norm": 13.677345275878906,
      "learning_rate": 3.969465648854962e-06,
      "loss": 0.1104,
      "step": 40
    },
    {
      "epoch": 0.01272264631043257,
      "grad_norm": 9.209020614624023,
      "learning_rate": 4.987277353689568e-06,
      "loss": 0.0461,
      "step": 50
    },
    {
      "epoch": 0.015267175572519083,
      "grad_norm": 14.143465995788574,
      "learning_rate": 6.005089058524174e-06,
      "loss": 0.0482,
      "step": 60
    },
    {
      "epoch": 0.017811704834605598,
      "grad_norm": 1.9570101499557495,
      "learning_rate": 7.022900763358779e-06,
      "loss": 0.0331,
      "step": 70
    },
    {
      "epoch": 0.020356234096692113,
      "grad_norm": 6.45846700668335,
      "learning_rate": 8.040712468193384e-06,
      "loss": 0.0358,
      "step": 80
    },
    {
      "epoch": 0.022900763358778626,
      "grad_norm": 6.0744309425354,
      "learning_rate": 9.058524173027991e-06,
      "loss": 0.026,
      "step": 90
    },
    {
      "epoch": 0.02544529262086514,
      "grad_norm": 8.33021068572998,
      "learning_rate": 1.0076335877862595e-05,
      "loss": 0.0373,
      "step": 100
    },
    {
      "epoch": 0.027989821882951654,
      "grad_norm": 6.4917778968811035,
      "learning_rate": 1.1094147582697202e-05,
      "loss": 0.0186,
      "step": 110
    },
    {
      "epoch": 0.030534351145038167,
      "grad_norm": 4.850451469421387,
      "learning_rate": 1.2111959287531807e-05,
      "loss": 0.0191,
      "step": 120
    },
    {
      "epoch": 0.03307888040712468,
      "grad_norm": 3.528228521347046,
      "learning_rate": 1.3129770992366414e-05,
      "loss": 0.0244,
      "step": 130
    },
    {
      "epoch": 0.035623409669211195,
      "grad_norm": 0.09570707380771637,
      "learning_rate": 1.4147582697201019e-05,
      "loss": 0.0154,
      "step": 140
    },
    {
      "epoch": 0.03816793893129771,
      "grad_norm": 8.72260856628418,
      "learning_rate": 1.5165394402035624e-05,
      "loss": 0.0251,
      "step": 150
    },
    {
      "epoch": 0.04071246819338423,
      "grad_norm": 5.22550106048584,
      "learning_rate": 1.618320610687023e-05,
      "loss": 0.0181,
      "step": 160
    },
    {
      "epoch": 0.043256997455470736,
      "grad_norm": 7.037172317504883,
      "learning_rate": 1.7201017811704836e-05,
      "loss": 0.007,
      "step": 170
    },
    {
      "epoch": 0.04580152671755725,
      "grad_norm": 0.9109981060028076,
      "learning_rate": 1.8218829516539443e-05,
      "loss": 0.0177,
      "step": 180
    },
    {
      "epoch": 0.04834605597964377,
      "grad_norm": 6.822402477264404,
      "learning_rate": 1.923664122137405e-05,
      "loss": 0.0079,
      "step": 190
    },
    {
      "epoch": 0.05089058524173028,
      "grad_norm": 2.628634452819824,
      "learning_rate": 2.0254452926208653e-05,
      "loss": 0.0142,
      "step": 200
    },
    {
      "epoch": 0.05343511450381679,
      "grad_norm": 0.7327078580856323,
      "learning_rate": 2.127226463104326e-05,
      "loss": 0.0207,
      "step": 210
    },
    {
      "epoch": 0.05597964376590331,
      "grad_norm": 7.423736095428467,
      "learning_rate": 2.2290076335877867e-05,
      "loss": 0.0127,
      "step": 220
    },
    {
      "epoch": 0.058524173027989825,
      "grad_norm": 0.46404168009757996,
      "learning_rate": 2.330788804071247e-05,
      "loss": 0.0158,
      "step": 230
    },
    {
      "epoch": 0.061068702290076333,
      "grad_norm": 0.023200908675789833,
      "learning_rate": 2.4325699745547078e-05,
      "loss": 0.0178,
      "step": 240
    },
    {
      "epoch": 0.06361323155216285,
      "grad_norm": 3.8594086170196533,
      "learning_rate": 2.5343511450381678e-05,
      "loss": 0.0069,
      "step": 250
    },
    {
      "epoch": 0.06615776081424936,
      "grad_norm": 0.4054301679134369,
      "learning_rate": 2.6361323155216285e-05,
      "loss": 0.0223,
      "step": 260
    },
    {
      "epoch": 0.06870229007633588,
      "grad_norm": 2.2085139751434326,
      "learning_rate": 2.737913486005089e-05,
      "loss": 0.0362,
      "step": 270
    },
    {
      "epoch": 0.07124681933842239,
      "grad_norm": 0.5911164283752441,
      "learning_rate": 2.8396946564885498e-05,
      "loss": 0.0187,
      "step": 280
    },
    {
      "epoch": 0.0737913486005089,
      "grad_norm": 4.2948102951049805,
      "learning_rate": 2.9414758269720102e-05,
      "loss": 0.0107,
      "step": 290
    },
    {
      "epoch": 0.07633587786259542,
      "grad_norm": 0.2595175504684448,
      "learning_rate": 3.043256997455471e-05,
      "loss": 0.0085,
      "step": 300
    },
    {
      "epoch": 0.07888040712468193,
      "grad_norm": 5.613988876342773,
      "learning_rate": 3.145038167938931e-05,
      "loss": 0.0281,
      "step": 310
    },
    {
      "epoch": 0.08142493638676845,
      "grad_norm": 2.6039955615997314,
      "learning_rate": 3.246819338422392e-05,
      "loss": 0.0213,
      "step": 320
    },
    {
      "epoch": 0.08396946564885496,
      "grad_norm": 0.2858663499355316,
      "learning_rate": 3.3486005089058526e-05,
      "loss": 0.0141,
      "step": 330
    },
    {
      "epoch": 0.08651399491094147,
      "grad_norm": 0.19223034381866455,
      "learning_rate": 3.450381679389313e-05,
      "loss": 0.0224,
      "step": 340
    },
    {
      "epoch": 0.089058524173028,
      "grad_norm": 5.314587116241455,
      "learning_rate": 3.552162849872774e-05,
      "loss": 0.0129,
      "step": 350
    },
    {
      "epoch": 0.0916030534351145,
      "grad_norm": 4.426294326782227,
      "learning_rate": 3.653944020356235e-05,
      "loss": 0.0224,
      "step": 360
    },
    {
      "epoch": 0.09414758269720101,
      "grad_norm": 5.395482063293457,
      "learning_rate": 3.755725190839695e-05,
      "loss": 0.01,
      "step": 370
    },
    {
      "epoch": 0.09669211195928754,
      "grad_norm": 0.7296554446220398,
      "learning_rate": 3.8575063613231554e-05,
      "loss": 0.0192,
      "step": 380
    },
    {
      "epoch": 0.09923664122137404,
      "grad_norm": 0.2524012625217438,
      "learning_rate": 3.959287531806616e-05,
      "loss": 0.0178,
      "step": 390
    },
    {
      "epoch": 0.10178117048346055,
      "grad_norm": 5.55770206451416,
      "learning_rate": 3.993214588634436e-05,
      "loss": 0.0229,
      "step": 400
    },
    {
      "epoch": 0.10432569974554708,
      "grad_norm": 4.481260299682617,
      "learning_rate": 3.9819055696918295e-05,
      "loss": 0.0138,
      "step": 410
    },
    {
      "epoch": 0.10687022900763359,
      "grad_norm": 2.183645725250244,
      "learning_rate": 3.970596550749223e-05,
      "loss": 0.0201,
      "step": 420
    },
    {
      "epoch": 0.10941475826972011,
      "grad_norm": 3.418135404586792,
      "learning_rate": 3.959287531806616e-05,
      "loss": 0.0225,
      "step": 430
    },
    {
      "epoch": 0.11195928753180662,
      "grad_norm": 0.1539747565984726,
      "learning_rate": 3.947978512864009e-05,
      "loss": 0.0043,
      "step": 440
    },
    {
      "epoch": 0.11450381679389313,
      "grad_norm": 3.93865966796875,
      "learning_rate": 3.9366694939214026e-05,
      "loss": 0.0088,
      "step": 450
    },
    {
      "epoch": 0.11704834605597965,
      "grad_norm": 2.0105717182159424,
      "learning_rate": 3.925360474978796e-05,
      "loss": 0.0101,
      "step": 460
    },
    {
      "epoch": 0.11959287531806616,
      "grad_norm": 1.3878592252731323,
      "learning_rate": 3.914051456036189e-05,
      "loss": 0.0086,
      "step": 470
    },
    {
      "epoch": 0.12213740458015267,
      "grad_norm": 0.09708156436681747,
      "learning_rate": 3.902742437093582e-05,
      "loss": 0.0119,
      "step": 480
    },
    {
      "epoch": 0.12468193384223919,
      "grad_norm": 0.4155742824077606,
      "learning_rate": 3.8914334181509756e-05,
      "loss": 0.0139,
      "step": 490
    },
    {
      "epoch": 0.1272264631043257,
      "grad_norm": 0.6312187910079956,
      "learning_rate": 3.880124399208369e-05,
      "loss": 0.0185,
      "step": 500
    },
    {
      "epoch": 0.1272264631043257,
      "eval_loss": 0.0426296703517437,
      "eval_runtime": 100.8006,
      "eval_samples_per_second": 79.186,
      "eval_steps_per_second": 0.625,
      "step": 500
    },
    {
      "epoch": 0.1297709923664122,
      "grad_norm": 8.71293830871582,
      "learning_rate": 3.868815380265762e-05,
      "loss": 0.0236,
      "step": 510
    },
    {
      "epoch": 0.13231552162849872,
      "grad_norm": 0.6096652150154114,
      "learning_rate": 3.8575063613231554e-05,
      "loss": 0.01,
      "step": 520
    },
    {
      "epoch": 0.13486005089058525,
      "grad_norm": 0.9275781512260437,
      "learning_rate": 3.8461973423805486e-05,
      "loss": 0.0091,
      "step": 530
    },
    {
      "epoch": 0.13740458015267176,
      "grad_norm": 0.3549460768699646,
      "learning_rate": 3.834888323437942e-05,
      "loss": 0.0092,
      "step": 540
    },
    {
      "epoch": 0.13994910941475827,
      "grad_norm": 3.617748498916626,
      "learning_rate": 3.823579304495336e-05,
      "loss": 0.0117,
      "step": 550
    },
    {
      "epoch": 0.14249363867684478,
      "grad_norm": 3.950561761856079,
      "learning_rate": 3.8122702855527284e-05,
      "loss": 0.0134,
      "step": 560
    },
    {
      "epoch": 0.1450381679389313,
      "grad_norm": 1.9877593517303467,
      "learning_rate": 3.8009612666101216e-05,
      "loss": 0.0143,
      "step": 570
    },
    {
      "epoch": 0.1475826972010178,
      "grad_norm": 1.1295750141143799,
      "learning_rate": 3.789652247667515e-05,
      "loss": 0.0035,
      "step": 580
    },
    {
      "epoch": 0.15012722646310434,
      "grad_norm": 0.07594585418701172,
      "learning_rate": 3.778343228724908e-05,
      "loss": 0.011,
      "step": 590
    },
    {
      "epoch": 0.15267175572519084,
      "grad_norm": 0.28959858417510986,
      "learning_rate": 3.767034209782302e-05,
      "loss": 0.0034,
      "step": 600
    },
    {
      "epoch": 0.15521628498727735,
      "grad_norm": 0.31216809153556824,
      "learning_rate": 3.755725190839695e-05,
      "loss": 0.0145,
      "step": 610
    },
    {
      "epoch": 0.15776081424936386,
      "grad_norm": 0.048563506454229355,
      "learning_rate": 3.744416171897088e-05,
      "loss": 0.0136,
      "step": 620
    },
    {
      "epoch": 0.16030534351145037,
      "grad_norm": 0.18490168452262878,
      "learning_rate": 3.733107152954481e-05,
      "loss": 0.0157,
      "step": 630
    },
    {
      "epoch": 0.1628498727735369,
      "grad_norm": 4.179055690765381,
      "learning_rate": 3.721798134011875e-05,
      "loss": 0.0113,
      "step": 640
    },
    {
      "epoch": 0.16539440203562342,
      "grad_norm": 0.7711288332939148,
      "learning_rate": 3.7104891150692684e-05,
      "loss": 0.004,
      "step": 650
    },
    {
      "epoch": 0.16793893129770993,
      "grad_norm": 2.4299933910369873,
      "learning_rate": 3.699180096126661e-05,
      "loss": 0.0093,
      "step": 660
    },
    {
      "epoch": 0.17048346055979643,
      "grad_norm": 1.621607780456543,
      "learning_rate": 3.687871077184054e-05,
      "loss": 0.0071,
      "step": 670
    },
    {
      "epoch": 0.17302798982188294,
      "grad_norm": 0.06137412041425705,
      "learning_rate": 3.676562058241448e-05,
      "loss": 0.0216,
      "step": 680
    },
    {
      "epoch": 0.17557251908396945,
      "grad_norm": 1.6078226566314697,
      "learning_rate": 3.6652530392988414e-05,
      "loss": 0.0242,
      "step": 690
    },
    {
      "epoch": 0.178117048346056,
      "grad_norm": 2.2799558639526367,
      "learning_rate": 3.653944020356235e-05,
      "loss": 0.0113,
      "step": 700
    },
    {
      "epoch": 0.1806615776081425,
      "grad_norm": 2.2463083267211914,
      "learning_rate": 3.642635001413627e-05,
      "loss": 0.0054,
      "step": 710
    },
    {
      "epoch": 0.183206106870229,
      "grad_norm": 0.4119090437889099,
      "learning_rate": 3.6313259824710205e-05,
      "loss": 0.0107,
      "step": 720
    },
    {
      "epoch": 0.18575063613231552,
      "grad_norm": 3.735358476638794,
      "learning_rate": 3.6200169635284144e-05,
      "loss": 0.0096,
      "step": 730
    },
    {
      "epoch": 0.18829516539440203,
      "grad_norm": 2.4470431804656982,
      "learning_rate": 3.608707944585808e-05,
      "loss": 0.0089,
      "step": 740
    },
    {
      "epoch": 0.19083969465648856,
      "grad_norm": 0.0336996465921402,
      "learning_rate": 3.597398925643201e-05,
      "loss": 0.0111,
      "step": 750
    },
    {
      "epoch": 0.19338422391857507,
      "grad_norm": 0.6565234661102295,
      "learning_rate": 3.586089906700594e-05,
      "loss": 0.0083,
      "step": 760
    },
    {
      "epoch": 0.19592875318066158,
      "grad_norm": 9.512853622436523,
      "learning_rate": 3.5747808877579875e-05,
      "loss": 0.0051,
      "step": 770
    },
    {
      "epoch": 0.1984732824427481,
      "grad_norm": 0.5155350565910339,
      "learning_rate": 3.563471868815381e-05,
      "loss": 0.008,
      "step": 780
    },
    {
      "epoch": 0.2010178117048346,
      "grad_norm": 4.279949188232422,
      "learning_rate": 3.552162849872774e-05,
      "loss": 0.0081,
      "step": 790
    },
    {
      "epoch": 0.2035623409669211,
      "grad_norm": 2.585608959197998,
      "learning_rate": 3.540853830930167e-05,
      "loss": 0.018,
      "step": 800
    },
    {
      "epoch": 0.20610687022900764,
      "grad_norm": 1.0371241569519043,
      "learning_rate": 3.5295448119875605e-05,
      "loss": 0.0134,
      "step": 810
    },
    {
      "epoch": 0.20865139949109415,
      "grad_norm": 0.7093348503112793,
      "learning_rate": 3.518235793044954e-05,
      "loss": 0.0035,
      "step": 820
    },
    {
      "epoch": 0.21119592875318066,
      "grad_norm": 0.48215287923812866,
      "learning_rate": 3.506926774102347e-05,
      "loss": 0.0099,
      "step": 830
    },
    {
      "epoch": 0.21374045801526717,
      "grad_norm": 3.264940023422241,
      "learning_rate": 3.49561775515974e-05,
      "loss": 0.0152,
      "step": 840
    },
    {
      "epoch": 0.21628498727735368,
      "grad_norm": 0.03547951951622963,
      "learning_rate": 3.4843087362171335e-05,
      "loss": 0.0078,
      "step": 850
    },
    {
      "epoch": 0.21882951653944022,
      "grad_norm": 2.6558687686920166,
      "learning_rate": 3.472999717274527e-05,
      "loss": 0.0134,
      "step": 860
    },
    {
      "epoch": 0.22137404580152673,
      "grad_norm": 0.41831302642822266,
      "learning_rate": 3.46169069833192e-05,
      "loss": 0.0045,
      "step": 870
    },
    {
      "epoch": 0.22391857506361323,
      "grad_norm": 0.04895065724849701,
      "learning_rate": 3.450381679389313e-05,
      "loss": 0.0093,
      "step": 880
    },
    {
      "epoch": 0.22646310432569974,
      "grad_norm": 4.139157772064209,
      "learning_rate": 3.4390726604467065e-05,
      "loss": 0.0122,
      "step": 890
    },
    {
      "epoch": 0.22900763358778625,
      "grad_norm": 0.1418493539094925,
      "learning_rate": 3.4277636415041e-05,
      "loss": 0.0147,
      "step": 900
    },
    {
      "epoch": 0.23155216284987276,
      "grad_norm": 0.19460724294185638,
      "learning_rate": 3.416454622561493e-05,
      "loss": 0.0128,
      "step": 910
    },
    {
      "epoch": 0.2340966921119593,
      "grad_norm": 0.18406054377555847,
      "learning_rate": 3.405145603618886e-05,
      "loss": 0.0115,
      "step": 920
    },
    {
      "epoch": 0.2366412213740458,
      "grad_norm": 0.16981343924999237,
      "learning_rate": 3.3938365846762796e-05,
      "loss": 0.0082,
      "step": 930
    },
    {
      "epoch": 0.23918575063613232,
      "grad_norm": 0.009775158949196339,
      "learning_rate": 3.382527565733673e-05,
      "loss": 0.0027,
      "step": 940
    },
    {
      "epoch": 0.24173027989821882,
      "grad_norm": 0.1106642484664917,
      "learning_rate": 3.371218546791066e-05,
      "loss": 0.0139,
      "step": 950
    },
    {
      "epoch": 0.24427480916030533,
      "grad_norm": 0.10489284992218018,
      "learning_rate": 3.3599095278484593e-05,
      "loss": 0.002,
      "step": 960
    },
    {
      "epoch": 0.24681933842239187,
      "grad_norm": 0.26976943016052246,
      "learning_rate": 3.3486005089058526e-05,
      "loss": 0.0075,
      "step": 970
    },
    {
      "epoch": 0.24936386768447838,
      "grad_norm": 0.06751968711614609,
      "learning_rate": 3.337291489963246e-05,
      "loss": 0.015,
      "step": 980
    },
    {
      "epoch": 0.25190839694656486,
      "grad_norm": 0.07933365553617477,
      "learning_rate": 3.325982471020639e-05,
      "loss": 0.0049,
      "step": 990
    },
    {
      "epoch": 0.2544529262086514,
      "grad_norm": 1.002841830253601,
      "learning_rate": 3.3146734520780324e-05,
      "loss": 0.0082,
      "step": 1000
    },
    {
      "epoch": 0.2544529262086514,
      "eval_loss": 0.04029834270477295,
      "eval_runtime": 100.7792,
      "eval_samples_per_second": 79.203,
      "eval_steps_per_second": 0.625,
      "step": 1000
    },
    {
      "epoch": 0.25699745547073793,
      "grad_norm": 0.11880303919315338,
      "learning_rate": 3.3033644331354256e-05,
      "loss": 0.0071,
      "step": 1010
    },
    {
      "epoch": 0.2595419847328244,
      "grad_norm": 2.1375386714935303,
      "learning_rate": 3.292055414192819e-05,
      "loss": 0.0043,
      "step": 1020
    },
    {
      "epoch": 0.26208651399491095,
      "grad_norm": 1.9007747173309326,
      "learning_rate": 3.280746395250212e-05,
      "loss": 0.0056,
      "step": 1030
    },
    {
      "epoch": 0.26463104325699743,
      "grad_norm": 1.3233751058578491,
      "learning_rate": 3.2694373763076054e-05,
      "loss": 0.0122,
      "step": 1040
    },
    {
      "epoch": 0.26717557251908397,
      "grad_norm": 1.3246195316314697,
      "learning_rate": 3.2581283573649987e-05,
      "loss": 0.0082,
      "step": 1050
    },
    {
      "epoch": 0.2697201017811705,
      "grad_norm": 3.8455071449279785,
      "learning_rate": 3.246819338422392e-05,
      "loss": 0.0048,
      "step": 1060
    },
    {
      "epoch": 0.272264631043257,
      "grad_norm": 1.4055747985839844,
      "learning_rate": 3.235510319479785e-05,
      "loss": 0.008,
      "step": 1070
    },
    {
      "epoch": 0.2748091603053435,
      "grad_norm": 1.1376632452011108,
      "learning_rate": 3.224201300537179e-05,
      "loss": 0.0088,
      "step": 1080
    },
    {
      "epoch": 0.27735368956743,
      "grad_norm": 1.465824007987976,
      "learning_rate": 3.212892281594572e-05,
      "loss": 0.0082,
      "step": 1090
    },
    {
      "epoch": 0.27989821882951654,
      "grad_norm": 0.5838302373886108,
      "learning_rate": 3.201583262651965e-05,
      "loss": 0.0128,
      "step": 1100
    },
    {
      "epoch": 0.2824427480916031,
      "grad_norm": 0.008899681270122528,
      "learning_rate": 3.190274243709358e-05,
      "loss": 0.0061,
      "step": 1110
    },
    {
      "epoch": 0.28498727735368956,
      "grad_norm": 2.9608395099639893,
      "learning_rate": 3.178965224766752e-05,
      "loss": 0.0217,
      "step": 1120
    },
    {
      "epoch": 0.2875318066157761,
      "grad_norm": 0.12024850398302078,
      "learning_rate": 3.1676562058241454e-05,
      "loss": 0.0072,
      "step": 1130
    },
    {
      "epoch": 0.2900763358778626,
      "grad_norm": 0.09325462579727173,
      "learning_rate": 3.156347186881538e-05,
      "loss": 0.0079,
      "step": 1140
    },
    {
      "epoch": 0.2926208651399491,
      "grad_norm": 0.047860078513622284,
      "learning_rate": 3.145038167938931e-05,
      "loss": 0.0111,
      "step": 1150
    },
    {
      "epoch": 0.2951653944020356,
      "grad_norm": 0.14491482079029083,
      "learning_rate": 3.1337291489963245e-05,
      "loss": 0.0079,
      "step": 1160
    },
    {
      "epoch": 0.29770992366412213,
      "grad_norm": 5.476013660430908,
      "learning_rate": 3.1224201300537184e-05,
      "loss": 0.0161,
      "step": 1170
    },
    {
      "epoch": 0.30025445292620867,
      "grad_norm": 0.5506300926208496,
      "learning_rate": 3.111111111111112e-05,
      "loss": 0.0032,
      "step": 1180
    },
    {
      "epoch": 0.30279898218829515,
      "grad_norm": 0.5853263735771179,
      "learning_rate": 3.099802092168504e-05,
      "loss": 0.0107,
      "step": 1190
    },
    {
      "epoch": 0.3053435114503817,
      "grad_norm": 0.47357645630836487,
      "learning_rate": 3.0884930732258975e-05,
      "loss": 0.0046,
      "step": 1200
    },
    {
      "epoch": 0.30788804071246817,
      "grad_norm": 0.004211845807731152,
      "learning_rate": 3.0771840542832914e-05,
      "loss": 0.0099,
      "step": 1210
    },
    {
      "epoch": 0.3104325699745547,
      "grad_norm": 0.029812565073370934,
      "learning_rate": 3.065875035340685e-05,
      "loss": 0.0094,
      "step": 1220
    },
    {
      "epoch": 0.31297709923664124,
      "grad_norm": 2.437438488006592,
      "learning_rate": 3.054566016398078e-05,
      "loss": 0.014,
      "step": 1230
    },
    {
      "epoch": 0.3155216284987277,
      "grad_norm": 0.3200075626373291,
      "learning_rate": 3.043256997455471e-05,
      "loss": 0.0082,
      "step": 1240
    },
    {
      "epoch": 0.31806615776081426,
      "grad_norm": 0.023235173895955086,
      "learning_rate": 3.031947978512864e-05,
      "loss": 0.013,
      "step": 1250
    },
    {
      "epoch": 0.32061068702290074,
      "grad_norm": 0.042968690395355225,
      "learning_rate": 3.0206389595702577e-05,
      "loss": 0.0222,
      "step": 1260
    },
    {
      "epoch": 0.3231552162849873,
      "grad_norm": 0.07905975729227066,
      "learning_rate": 3.009329940627651e-05,
      "loss": 0.004,
      "step": 1270
    },
    {
      "epoch": 0.3256997455470738,
      "grad_norm": 0.47882354259490967,
      "learning_rate": 2.998020921685044e-05,
      "loss": 0.0042,
      "step": 1280
    },
    {
      "epoch": 0.3282442748091603,
      "grad_norm": 1.036494255065918,
      "learning_rate": 2.986711902742437e-05,
      "loss": 0.0032,
      "step": 1290
    },
    {
      "epoch": 0.33078880407124683,
      "grad_norm": 2.536672830581665,
      "learning_rate": 2.9754028837998307e-05,
      "loss": 0.0109,
      "step": 1300
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 2.7068772315979004,
      "learning_rate": 2.964093864857224e-05,
      "loss": 0.012,
      "step": 1310
    },
    {
      "epoch": 0.33587786259541985,
      "grad_norm": 0.03697170689702034,
      "learning_rate": 2.9527848459146173e-05,
      "loss": 0.01,
      "step": 1320
    },
    {
      "epoch": 0.3384223918575064,
      "grad_norm": 0.17284952104091644,
      "learning_rate": 2.9414758269720102e-05,
      "loss": 0.003,
      "step": 1330
    },
    {
      "epoch": 0.34096692111959287,
      "grad_norm": 3.8745925426483154,
      "learning_rate": 2.9301668080294038e-05,
      "loss": 0.0156,
      "step": 1340
    },
    {
      "epoch": 0.3435114503816794,
      "grad_norm": 0.44607672095298767,
      "learning_rate": 2.918857789086797e-05,
      "loss": 0.0115,
      "step": 1350
    },
    {
      "epoch": 0.3460559796437659,
      "grad_norm": 6.223470687866211,
      "learning_rate": 2.9075487701441903e-05,
      "loss": 0.0149,
      "step": 1360
    },
    {
      "epoch": 0.3486005089058524,
      "grad_norm": 0.10045173764228821,
      "learning_rate": 2.8962397512015835e-05,
      "loss": 0.0063,
      "step": 1370
    },
    {
      "epoch": 0.3511450381679389,
      "grad_norm": 2.6154091358184814,
      "learning_rate": 2.8849307322589765e-05,
      "loss": 0.0099,
      "step": 1380
    },
    {
      "epoch": 0.35368956743002544,
      "grad_norm": 1.6079035997390747,
      "learning_rate": 2.87362171331637e-05,
      "loss": 0.0079,
      "step": 1390
    },
    {
      "epoch": 0.356234096692112,
      "grad_norm": 0.3847286105155945,
      "learning_rate": 2.8623126943737633e-05,
      "loss": 0.0094,
      "step": 1400
    },
    {
      "epoch": 0.35877862595419846,
      "grad_norm": 0.762495219707489,
      "learning_rate": 2.8510036754311566e-05,
      "loss": 0.0138,
      "step": 1410
    },
    {
      "epoch": 0.361323155216285,
      "grad_norm": 2.905803680419922,
      "learning_rate": 2.8396946564885498e-05,
      "loss": 0.0126,
      "step": 1420
    },
    {
      "epoch": 0.3638676844783715,
      "grad_norm": 0.20091155171394348,
      "learning_rate": 2.8283856375459434e-05,
      "loss": 0.005,
      "step": 1430
    },
    {
      "epoch": 0.366412213740458,
      "grad_norm": 0.2232949584722519,
      "learning_rate": 2.8170766186033363e-05,
      "loss": 0.0157,
      "step": 1440
    },
    {
      "epoch": 0.36895674300254455,
      "grad_norm": 0.11827141046524048,
      "learning_rate": 2.8057675996607296e-05,
      "loss": 0.0076,
      "step": 1450
    },
    {
      "epoch": 0.37150127226463103,
      "grad_norm": 0.0741715356707573,
      "learning_rate": 2.794458580718123e-05,
      "loss": 0.0117,
      "step": 1460
    },
    {
      "epoch": 0.37404580152671757,
      "grad_norm": 1.2536979913711548,
      "learning_rate": 2.783149561775516e-05,
      "loss": 0.0017,
      "step": 1470
    },
    {
      "epoch": 0.37659033078880405,
      "grad_norm": 1.9891741275787354,
      "learning_rate": 2.7718405428329097e-05,
      "loss": 0.0109,
      "step": 1480
    },
    {
      "epoch": 0.3791348600508906,
      "grad_norm": 1.0660837888717651,
      "learning_rate": 2.7605315238903026e-05,
      "loss": 0.007,
      "step": 1490
    },
    {
      "epoch": 0.3816793893129771,
      "grad_norm": 0.0798741951584816,
      "learning_rate": 2.749222504947696e-05,
      "loss": 0.0139,
      "step": 1500
    },
    {
      "epoch": 0.3816793893129771,
      "eval_loss": 0.0349714532494545,
      "eval_runtime": 100.7459,
      "eval_samples_per_second": 79.229,
      "eval_steps_per_second": 0.625,
      "step": 1500
    },
    {
      "epoch": 0.3842239185750636,
      "grad_norm": 3.056962013244629,
      "learning_rate": 2.737913486005089e-05,
      "loss": 0.019,
      "step": 1510
    },
    {
      "epoch": 0.38676844783715014,
      "grad_norm": 0.41486862301826477,
      "learning_rate": 2.7266044670624827e-05,
      "loss": 0.0027,
      "step": 1520
    },
    {
      "epoch": 0.3893129770992366,
      "grad_norm": 1.0251950025558472,
      "learning_rate": 2.715295448119876e-05,
      "loss": 0.0078,
      "step": 1530
    },
    {
      "epoch": 0.39185750636132316,
      "grad_norm": 0.019924340769648552,
      "learning_rate": 2.703986429177269e-05,
      "loss": 0.0031,
      "step": 1540
    },
    {
      "epoch": 0.3944020356234097,
      "grad_norm": 3.1143898963928223,
      "learning_rate": 2.692677410234662e-05,
      "loss": 0.0054,
      "step": 1550
    },
    {
      "epoch": 0.3969465648854962,
      "grad_norm": 3.4814693927764893,
      "learning_rate": 2.6813683912920558e-05,
      "loss": 0.0169,
      "step": 1560
    },
    {
      "epoch": 0.3994910941475827,
      "grad_norm": 0.34978416562080383,
      "learning_rate": 2.670059372349449e-05,
      "loss": 0.0017,
      "step": 1570
    },
    {
      "epoch": 0.4020356234096692,
      "grad_norm": 0.17419882118701935,
      "learning_rate": 2.6587503534068423e-05,
      "loss": 0.0017,
      "step": 1580
    },
    {
      "epoch": 0.40458015267175573,
      "grad_norm": 6.330338001251221,
      "learning_rate": 2.6474413344642352e-05,
      "loss": 0.0081,
      "step": 1590
    },
    {
      "epoch": 0.4071246819338422,
      "grad_norm": 0.014109604060649872,
      "learning_rate": 2.6361323155216285e-05,
      "loss": 0.0059,
      "step": 1600
    },
    {
      "epoch": 0.40966921119592875,
      "grad_norm": 0.06557715684175491,
      "learning_rate": 2.624823296579022e-05,
      "loss": 0.0025,
      "step": 1610
    },
    {
      "epoch": 0.4122137404580153,
      "grad_norm": 0.06417052447795868,
      "learning_rate": 2.6135142776364153e-05,
      "loss": 0.001,
      "step": 1620
    },
    {
      "epoch": 0.41475826972010177,
      "grad_norm": 2.6139402389526367,
      "learning_rate": 2.6022052586938086e-05,
      "loss": 0.0049,
      "step": 1630
    },
    {
      "epoch": 0.4173027989821883,
      "grad_norm": 2.0611274242401123,
      "learning_rate": 2.5908962397512015e-05,
      "loss": 0.0062,
      "step": 1640
    },
    {
      "epoch": 0.4198473282442748,
      "grad_norm": 0.0023517871741205454,
      "learning_rate": 2.5795872208085954e-05,
      "loss": 0.0013,
      "step": 1650
    },
    {
      "epoch": 0.4223918575063613,
      "grad_norm": 0.15409255027770996,
      "learning_rate": 2.5682782018659883e-05,
      "loss": 0.0131,
      "step": 1660
    },
    {
      "epoch": 0.42493638676844786,
      "grad_norm": 0.07905858010053635,
      "learning_rate": 2.5569691829233816e-05,
      "loss": 0.0037,
      "step": 1670
    },
    {
      "epoch": 0.42748091603053434,
      "grad_norm": 0.3997408151626587,
      "learning_rate": 2.545660163980775e-05,
      "loss": 0.0064,
      "step": 1680
    },
    {
      "epoch": 0.4300254452926209,
      "grad_norm": 0.14963960647583008,
      "learning_rate": 2.5343511450381678e-05,
      "loss": 0.0107,
      "step": 1690
    },
    {
      "epoch": 0.43256997455470736,
      "grad_norm": 0.06408097594976425,
      "learning_rate": 2.5230421260955617e-05,
      "loss": 0.0072,
      "step": 1700
    },
    {
      "epoch": 0.4351145038167939,
      "grad_norm": 3.4469339847564697,
      "learning_rate": 2.5117331071529546e-05,
      "loss": 0.008,
      "step": 1710
    },
    {
      "epoch": 0.43765903307888043,
      "grad_norm": 0.004595658276230097,
      "learning_rate": 2.500424088210348e-05,
      "loss": 0.0081,
      "step": 1720
    },
    {
      "epoch": 0.4402035623409669,
      "grad_norm": 4.713712215423584,
      "learning_rate": 2.489115069267741e-05,
      "loss": 0.0107,
      "step": 1730
    },
    {
      "epoch": 0.44274809160305345,
      "grad_norm": 0.08359724283218384,
      "learning_rate": 2.4778060503251347e-05,
      "loss": 0.0018,
      "step": 1740
    },
    {
      "epoch": 0.44529262086513993,
      "grad_norm": 2.5329370498657227,
      "learning_rate": 2.466497031382528e-05,
      "loss": 0.0086,
      "step": 1750
    },
    {
      "epoch": 0.44783715012722647,
      "grad_norm": 0.017571361735463142,
      "learning_rate": 2.455188012439921e-05,
      "loss": 0.0065,
      "step": 1760
    },
    {
      "epoch": 0.45038167938931295,
      "grad_norm": 0.40600821375846863,
      "learning_rate": 2.443878993497314e-05,
      "loss": 0.0062,
      "step": 1770
    },
    {
      "epoch": 0.4529262086513995,
      "grad_norm": 3.7990636825561523,
      "learning_rate": 2.4325699745547078e-05,
      "loss": 0.0134,
      "step": 1780
    },
    {
      "epoch": 0.455470737913486,
      "grad_norm": 0.09965948760509491,
      "learning_rate": 2.421260955612101e-05,
      "loss": 0.0068,
      "step": 1790
    },
    {
      "epoch": 0.4580152671755725,
      "grad_norm": 2.918208360671997,
      "learning_rate": 2.4099519366694943e-05,
      "loss": 0.0093,
      "step": 1800
    },
    {
      "epoch": 0.46055979643765904,
      "grad_norm": 0.011430525220930576,
      "learning_rate": 2.3986429177268872e-05,
      "loss": 0.0082,
      "step": 1810
    },
    {
      "epoch": 0.4631043256997455,
      "grad_norm": 0.15496379137039185,
      "learning_rate": 2.3873338987842804e-05,
      "loss": 0.0096,
      "step": 1820
    },
    {
      "epoch": 0.46564885496183206,
      "grad_norm": 0.02492368035018444,
      "learning_rate": 2.376024879841674e-05,
      "loss": 0.0078,
      "step": 1830
    },
    {
      "epoch": 0.4681933842239186,
      "grad_norm": 0.05971188098192215,
      "learning_rate": 2.3647158608990673e-05,
      "loss": 0.0089,
      "step": 1840
    },
    {
      "epoch": 0.4707379134860051,
      "grad_norm": 0.4808364808559418,
      "learning_rate": 2.3534068419564605e-05,
      "loss": 0.0047,
      "step": 1850
    },
    {
      "epoch": 0.4732824427480916,
      "grad_norm": 2.379042863845825,
      "learning_rate": 2.3420978230138535e-05,
      "loss": 0.0118,
      "step": 1860
    },
    {
      "epoch": 0.4758269720101781,
      "grad_norm": 0.024242276325821877,
      "learning_rate": 2.330788804071247e-05,
      "loss": 0.009,
      "step": 1870
    },
    {
      "epoch": 0.47837150127226463,
      "grad_norm": 1.4494438171386719,
      "learning_rate": 2.3194797851286403e-05,
      "loss": 0.0035,
      "step": 1880
    },
    {
      "epoch": 0.48091603053435117,
      "grad_norm": 4.915983200073242,
      "learning_rate": 2.3081707661860336e-05,
      "loss": 0.0145,
      "step": 1890
    },
    {
      "epoch": 0.48346055979643765,
      "grad_norm": 2.4147846698760986,
      "learning_rate": 2.296861747243427e-05,
      "loss": 0.0078,
      "step": 1900
    },
    {
      "epoch": 0.4860050890585242,
      "grad_norm": 0.41452011466026306,
      "learning_rate": 2.2855527283008204e-05,
      "loss": 0.0036,
      "step": 1910
    },
    {
      "epoch": 0.48854961832061067,
      "grad_norm": 0.6090364456176758,
      "learning_rate": 2.2742437093582133e-05,
      "loss": 0.0054,
      "step": 1920
    },
    {
      "epoch": 0.4910941475826972,
      "grad_norm": 2.908536434173584,
      "learning_rate": 2.2629346904156066e-05,
      "loss": 0.0055,
      "step": 1930
    },
    {
      "epoch": 0.49363867684478374,
      "grad_norm": 3.4067792892456055,
      "learning_rate": 2.251625671473e-05,
      "loss": 0.0063,
      "step": 1940
    },
    {
      "epoch": 0.4961832061068702,
      "grad_norm": 0.12075275182723999,
      "learning_rate": 2.240316652530393e-05,
      "loss": 0.0075,
      "step": 1950
    },
    {
      "epoch": 0.49872773536895676,
      "grad_norm": 0.10985302925109863,
      "learning_rate": 2.2290076335877867e-05,
      "loss": 0.0032,
      "step": 1960
    },
    {
      "epoch": 0.5012722646310432,
      "grad_norm": 1.9272608757019043,
      "learning_rate": 2.2176986146451796e-05,
      "loss": 0.0069,
      "step": 1970
    },
    {
      "epoch": 0.5038167938931297,
      "grad_norm": 3.202061891555786,
      "learning_rate": 2.206389595702573e-05,
      "loss": 0.0073,
      "step": 1980
    },
    {
      "epoch": 0.5063613231552163,
      "grad_norm": 1.0483746528625488,
      "learning_rate": 2.195080576759966e-05,
      "loss": 0.0031,
      "step": 1990
    },
    {
      "epoch": 0.5089058524173028,
      "grad_norm": 1.31486976146698,
      "learning_rate": 2.1837715578173597e-05,
      "loss": 0.0025,
      "step": 2000
    },
    {
      "epoch": 0.5089058524173028,
      "eval_loss": 0.034609079360961914,
      "eval_runtime": 100.7933,
      "eval_samples_per_second": 79.192,
      "eval_steps_per_second": 0.625,
      "step": 2000
    },
    {
      "epoch": 0.5114503816793893,
      "grad_norm": 0.05615166202187538,
      "learning_rate": 2.172462538874753e-05,
      "loss": 0.0114,
      "step": 2010
    },
    {
      "epoch": 0.5139949109414759,
      "grad_norm": 1.326418161392212,
      "learning_rate": 2.161153519932146e-05,
      "loss": 0.0053,
      "step": 2020
    },
    {
      "epoch": 0.5165394402035624,
      "grad_norm": 1.4809695482254028,
      "learning_rate": 2.1498445009895392e-05,
      "loss": 0.0085,
      "step": 2030
    },
    {
      "epoch": 0.5190839694656488,
      "grad_norm": 2.975813388824463,
      "learning_rate": 2.1385354820469324e-05,
      "loss": 0.0101,
      "step": 2040
    },
    {
      "epoch": 0.5216284987277354,
      "grad_norm": 0.8962512612342834,
      "learning_rate": 2.127226463104326e-05,
      "loss": 0.0104,
      "step": 2050
    },
    {
      "epoch": 0.5241730279898219,
      "grad_norm": 3.6851797103881836,
      "learning_rate": 2.1159174441617193e-05,
      "loss": 0.0074,
      "step": 2060
    },
    {
      "epoch": 0.5267175572519084,
      "grad_norm": 2.4362735748291016,
      "learning_rate": 2.1046084252191122e-05,
      "loss": 0.0119,
      "step": 2070
    },
    {
      "epoch": 0.5292620865139949,
      "grad_norm": 3.4005730152130127,
      "learning_rate": 2.0932994062765055e-05,
      "loss": 0.0179,
      "step": 2080
    },
    {
      "epoch": 0.5318066157760815,
      "grad_norm": 0.020965535193681717,
      "learning_rate": 2.081990387333899e-05,
      "loss": 0.0108,
      "step": 2090
    },
    {
      "epoch": 0.5343511450381679,
      "grad_norm": 3.1642138957977295,
      "learning_rate": 2.0706813683912923e-05,
      "loss": 0.0089,
      "step": 2100
    },
    {
      "epoch": 0.5368956743002544,
      "grad_norm": 1.4003890752792358,
      "learning_rate": 2.0593723494486856e-05,
      "loss": 0.0086,
      "step": 2110
    },
    {
      "epoch": 0.539440203562341,
      "grad_norm": 3.2447030544281006,
      "learning_rate": 2.0480633305060785e-05,
      "loss": 0.0092,
      "step": 2120
    },
    {
      "epoch": 0.5419847328244275,
      "grad_norm": 0.20196814835071564,
      "learning_rate": 2.0367543115634724e-05,
      "loss": 0.0091,
      "step": 2130
    },
    {
      "epoch": 0.544529262086514,
      "grad_norm": 0.09086956083774567,
      "learning_rate": 2.0254452926208653e-05,
      "loss": 0.0115,
      "step": 2140
    },
    {
      "epoch": 0.5470737913486005,
      "grad_norm": 0.8166348934173584,
      "learning_rate": 2.0141362736782586e-05,
      "loss": 0.0138,
      "step": 2150
    },
    {
      "epoch": 0.549618320610687,
      "grad_norm": 1.8422399759292603,
      "learning_rate": 2.002827254735652e-05,
      "loss": 0.0041,
      "step": 2160
    },
    {
      "epoch": 0.5521628498727735,
      "grad_norm": 1.3676835298538208,
      "learning_rate": 1.991518235793045e-05,
      "loss": 0.0051,
      "step": 2170
    },
    {
      "epoch": 0.55470737913486,
      "grad_norm": 3.4715235233306885,
      "learning_rate": 1.9802092168504384e-05,
      "loss": 0.0101,
      "step": 2180
    },
    {
      "epoch": 0.5572519083969466,
      "grad_norm": 0.4825442433357239,
      "learning_rate": 1.9689001979078316e-05,
      "loss": 0.0065,
      "step": 2190
    },
    {
      "epoch": 0.5597964376590331,
      "grad_norm": 0.03331977128982544,
      "learning_rate": 1.957591178965225e-05,
      "loss": 0.0136,
      "step": 2200
    },
    {
      "epoch": 0.5623409669211196,
      "grad_norm": 0.6727198362350464,
      "learning_rate": 1.946282160022618e-05,
      "loss": 0.0023,
      "step": 2210
    },
    {
      "epoch": 0.5648854961832062,
      "grad_norm": 2.1888349056243896,
      "learning_rate": 1.9349731410800114e-05,
      "loss": 0.0095,
      "step": 2220
    },
    {
      "epoch": 0.5674300254452926,
      "grad_norm": 4.135168075561523,
      "learning_rate": 1.923664122137405e-05,
      "loss": 0.0179,
      "step": 2230
    },
    {
      "epoch": 0.5699745547073791,
      "grad_norm": 1.3800098896026611,
      "learning_rate": 1.912355103194798e-05,
      "loss": 0.0045,
      "step": 2240
    },
    {
      "epoch": 0.5725190839694656,
      "grad_norm": 0.2383573353290558,
      "learning_rate": 1.901046084252191e-05,
      "loss": 0.0154,
      "step": 2250
    },
    {
      "epoch": 0.5750636132315522,
      "grad_norm": 0.2643699645996094,
      "learning_rate": 1.8897370653095844e-05,
      "loss": 0.0065,
      "step": 2260
    },
    {
      "epoch": 0.5776081424936387,
      "grad_norm": 0.7938392162322998,
      "learning_rate": 1.8784280463669777e-05,
      "loss": 0.0074,
      "step": 2270
    },
    {
      "epoch": 0.5801526717557252,
      "grad_norm": 1.8504903316497803,
      "learning_rate": 1.8671190274243713e-05,
      "loss": 0.0093,
      "step": 2280
    },
    {
      "epoch": 0.5826972010178118,
      "grad_norm": 3.9438815116882324,
      "learning_rate": 1.8558100084817642e-05,
      "loss": 0.0065,
      "step": 2290
    },
    {
      "epoch": 0.5852417302798982,
      "grad_norm": 0.024695103988051414,
      "learning_rate": 1.8445009895391578e-05,
      "loss": 0.0007,
      "step": 2300
    },
    {
      "epoch": 0.5877862595419847,
      "grad_norm": 0.5728102326393127,
      "learning_rate": 1.8331919705965507e-05,
      "loss": 0.0014,
      "step": 2310
    },
    {
      "epoch": 0.5903307888040712,
      "grad_norm": 0.0501931831240654,
      "learning_rate": 1.8218829516539443e-05,
      "loss": 0.006,
      "step": 2320
    },
    {
      "epoch": 0.5928753180661578,
      "grad_norm": 1.13953697681427,
      "learning_rate": 1.8105739327113376e-05,
      "loss": 0.006,
      "step": 2330
    },
    {
      "epoch": 0.5954198473282443,
      "grad_norm": 2.1232895851135254,
      "learning_rate": 1.7992649137687308e-05,
      "loss": 0.0073,
      "step": 2340
    },
    {
      "epoch": 0.5979643765903307,
      "grad_norm": 0.37371209263801575,
      "learning_rate": 1.787955894826124e-05,
      "loss": 0.008,
      "step": 2350
    },
    {
      "epoch": 0.6005089058524173,
      "grad_norm": 2.0648906230926514,
      "learning_rate": 1.776646875883517e-05,
      "loss": 0.0063,
      "step": 2360
    },
    {
      "epoch": 0.6030534351145038,
      "grad_norm": 2.0410525798797607,
      "learning_rate": 1.7653378569409106e-05,
      "loss": 0.0035,
      "step": 2370
    },
    {
      "epoch": 0.6055979643765903,
      "grad_norm": 0.1869073212146759,
      "learning_rate": 1.754028837998304e-05,
      "loss": 0.0088,
      "step": 2380
    },
    {
      "epoch": 0.6081424936386769,
      "grad_norm": 2.5563478469848633,
      "learning_rate": 1.742719819055697e-05,
      "loss": 0.0092,
      "step": 2390
    },
    {
      "epoch": 0.6106870229007634,
      "grad_norm": 2.4307727813720703,
      "learning_rate": 1.7314108001130904e-05,
      "loss": 0.0101,
      "step": 2400
    },
    {
      "epoch": 0.6132315521628499,
      "grad_norm": 3.2351479530334473,
      "learning_rate": 1.7201017811704836e-05,
      "loss": 0.0104,
      "step": 2410
    },
    {
      "epoch": 0.6157760814249363,
      "grad_norm": 3.1805145740509033,
      "learning_rate": 1.708792762227877e-05,
      "loss": 0.0129,
      "step": 2420
    },
    {
      "epoch": 0.6183206106870229,
      "grad_norm": 1.25412118434906,
      "learning_rate": 1.69748374328527e-05,
      "loss": 0.0024,
      "step": 2430
    },
    {
      "epoch": 0.6208651399491094,
      "grad_norm": 0.02675499953329563,
      "learning_rate": 1.6861747243426634e-05,
      "loss": 0.0066,
      "step": 2440
    },
    {
      "epoch": 0.6234096692111959,
      "grad_norm": 5.5407819747924805,
      "learning_rate": 1.6748657054000566e-05,
      "loss": 0.007,
      "step": 2450
    },
    {
      "epoch": 0.6259541984732825,
      "grad_norm": 3.5257675647735596,
      "learning_rate": 1.66355668645745e-05,
      "loss": 0.0139,
      "step": 2460
    },
    {
      "epoch": 0.628498727735369,
      "grad_norm": 1.163690209388733,
      "learning_rate": 1.652247667514843e-05,
      "loss": 0.0041,
      "step": 2470
    },
    {
      "epoch": 0.6310432569974554,
      "grad_norm": 1.5052272081375122,
      "learning_rate": 1.6409386485722364e-05,
      "loss": 0.0162,
      "step": 2480
    },
    {
      "epoch": 0.6335877862595419,
      "grad_norm": 0.1098732277750969,
      "learning_rate": 1.6296296296296297e-05,
      "loss": 0.0077,
      "step": 2490
    },
    {
      "epoch": 0.6361323155216285,
      "grad_norm": 2.6149792671203613,
      "learning_rate": 1.618320610687023e-05,
      "loss": 0.0042,
      "step": 2500
    },
    {
      "epoch": 0.6361323155216285,
      "eval_loss": 0.030483221635222435,
      "eval_runtime": 100.851,
      "eval_samples_per_second": 79.146,
      "eval_steps_per_second": 0.625,
      "step": 2500
    },
    {
      "epoch": 0.638676844783715,
      "grad_norm": 0.5812219977378845,
      "learning_rate": 1.6070115917444162e-05,
      "loss": 0.0107,
      "step": 2510
    },
    {
      "epoch": 0.6412213740458015,
      "grad_norm": 2.341726303100586,
      "learning_rate": 1.5957025728018098e-05,
      "loss": 0.0097,
      "step": 2520
    },
    {
      "epoch": 0.6437659033078881,
      "grad_norm": 1.1993039846420288,
      "learning_rate": 1.5843935538592027e-05,
      "loss": 0.0088,
      "step": 2530
    },
    {
      "epoch": 0.6463104325699746,
      "grad_norm": 0.019058074802160263,
      "learning_rate": 1.5730845349165963e-05,
      "loss": 0.0049,
      "step": 2540
    },
    {
      "epoch": 0.648854961832061,
      "grad_norm": 1.7715319395065308,
      "learning_rate": 1.5617755159739892e-05,
      "loss": 0.0077,
      "step": 2550
    },
    {
      "epoch": 0.6513994910941476,
      "grad_norm": 2.300363302230835,
      "learning_rate": 1.5504664970313828e-05,
      "loss": 0.0289,
      "step": 2560
    },
    {
      "epoch": 0.6539440203562341,
      "grad_norm": 3.3478307723999023,
      "learning_rate": 1.539157478088776e-05,
      "loss": 0.0101,
      "step": 2570
    },
    {
      "epoch": 0.6564885496183206,
      "grad_norm": 0.5511828660964966,
      "learning_rate": 1.5278484591461693e-05,
      "loss": 0.013,
      "step": 2580
    },
    {
      "epoch": 0.6590330788804071,
      "grad_norm": 3.244690418243408,
      "learning_rate": 1.5165394402035624e-05,
      "loss": 0.0129,
      "step": 2590
    },
    {
      "epoch": 0.6615776081424937,
      "grad_norm": 0.08320210874080658,
      "learning_rate": 1.5052304212609557e-05,
      "loss": 0.004,
      "step": 2600
    },
    {
      "epoch": 0.6641221374045801,
      "grad_norm": 0.06972193717956543,
      "learning_rate": 1.493921402318349e-05,
      "loss": 0.0077,
      "step": 2610
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.010952036827802658,
      "learning_rate": 1.4826123833757422e-05,
      "loss": 0.0045,
      "step": 2620
    },
    {
      "epoch": 0.6692111959287532,
      "grad_norm": 0.1785265952348709,
      "learning_rate": 1.4713033644331356e-05,
      "loss": 0.0011,
      "step": 2630
    },
    {
      "epoch": 0.6717557251908397,
      "grad_norm": 3.3081789016723633,
      "learning_rate": 1.4599943454905287e-05,
      "loss": 0.0092,
      "step": 2640
    },
    {
      "epoch": 0.6743002544529262,
      "grad_norm": 0.011730634607374668,
      "learning_rate": 1.4486853265479221e-05,
      "loss": 0.0019,
      "step": 2650
    },
    {
      "epoch": 0.6768447837150128,
      "grad_norm": 1.933732271194458,
      "learning_rate": 1.4373763076053154e-05,
      "loss": 0.0026,
      "step": 2660
    },
    {
      "epoch": 0.6793893129770993,
      "grad_norm": 2.0810701847076416,
      "learning_rate": 1.4260672886627088e-05,
      "loss": 0.0104,
      "step": 2670
    },
    {
      "epoch": 0.6819338422391857,
      "grad_norm": 4.2147369384765625,
      "learning_rate": 1.4147582697201019e-05,
      "loss": 0.0257,
      "step": 2680
    },
    {
      "epoch": 0.6844783715012722,
      "grad_norm": 6.2209014892578125,
      "learning_rate": 1.4034492507774953e-05,
      "loss": 0.0112,
      "step": 2690
    },
    {
      "epoch": 0.6870229007633588,
      "grad_norm": 4.587314128875732,
      "learning_rate": 1.3921402318348884e-05,
      "loss": 0.007,
      "step": 2700
    },
    {
      "epoch": 0.6895674300254453,
      "grad_norm": 5.7180256843566895,
      "learning_rate": 1.3808312128922817e-05,
      "loss": 0.0107,
      "step": 2710
    },
    {
      "epoch": 0.6921119592875318,
      "grad_norm": 1.0686004161834717,
      "learning_rate": 1.369522193949675e-05,
      "loss": 0.0136,
      "step": 2720
    },
    {
      "epoch": 0.6946564885496184,
      "grad_norm": 3.1578478813171387,
      "learning_rate": 1.3582131750070682e-05,
      "loss": 0.0101,
      "step": 2730
    },
    {
      "epoch": 0.6972010178117048,
      "grad_norm": 1.4626826047897339,
      "learning_rate": 1.3469041560644616e-05,
      "loss": 0.0051,
      "step": 2740
    },
    {
      "epoch": 0.6997455470737913,
      "grad_norm": 1.409559726715088,
      "learning_rate": 1.3355951371218547e-05,
      "loss": 0.0048,
      "step": 2750
    },
    {
      "epoch": 0.7022900763358778,
      "grad_norm": 1.5923247337341309,
      "learning_rate": 1.3242861181792481e-05,
      "loss": 0.0052,
      "step": 2760
    },
    {
      "epoch": 0.7048346055979644,
      "grad_norm": 0.01753048412501812,
      "learning_rate": 1.3129770992366414e-05,
      "loss": 0.0051,
      "step": 2770
    },
    {
      "epoch": 0.7073791348600509,
      "grad_norm": 0.022448958829045296,
      "learning_rate": 1.3016680802940346e-05,
      "loss": 0.0008,
      "step": 2780
    },
    {
      "epoch": 0.7099236641221374,
      "grad_norm": 0.010822678916156292,
      "learning_rate": 1.2903590613514279e-05,
      "loss": 0.003,
      "step": 2790
    },
    {
      "epoch": 0.712468193384224,
      "grad_norm": 1.4215025901794434,
      "learning_rate": 1.2790500424088213e-05,
      "loss": 0.0061,
      "step": 2800
    },
    {
      "epoch": 0.7150127226463104,
      "grad_norm": 0.11338794231414795,
      "learning_rate": 1.2677410234662144e-05,
      "loss": 0.0045,
      "step": 2810
    },
    {
      "epoch": 0.7175572519083969,
      "grad_norm": 2.464766263961792,
      "learning_rate": 1.2564320045236076e-05,
      "loss": 0.0045,
      "step": 2820
    },
    {
      "epoch": 0.7201017811704835,
      "grad_norm": 0.09698779135942459,
      "learning_rate": 1.2451229855810009e-05,
      "loss": 0.0037,
      "step": 2830
    },
    {
      "epoch": 0.72264631043257,
      "grad_norm": 0.04683073237538338,
      "learning_rate": 1.2338139666383942e-05,
      "loss": 0.0174,
      "step": 2840
    },
    {
      "epoch": 0.7251908396946565,
      "grad_norm": 0.013533544726669788,
      "learning_rate": 1.2225049476957876e-05,
      "loss": 0.0051,
      "step": 2850
    },
    {
      "epoch": 0.727735368956743,
      "grad_norm": 1.7438645362854004,
      "learning_rate": 1.2111959287531807e-05,
      "loss": 0.0161,
      "step": 2860
    },
    {
      "epoch": 0.7302798982188295,
      "grad_norm": 1.247225046157837,
      "learning_rate": 1.1998869098105741e-05,
      "loss": 0.0006,
      "step": 2870
    },
    {
      "epoch": 0.732824427480916,
      "grad_norm": 5.181498050689697,
      "learning_rate": 1.1885778908679672e-05,
      "loss": 0.0059,
      "step": 2880
    },
    {
      "epoch": 0.7353689567430025,
      "grad_norm": 2.543138027191162,
      "learning_rate": 1.1772688719253606e-05,
      "loss": 0.015,
      "step": 2890
    },
    {
      "epoch": 0.7379134860050891,
      "grad_norm": 0.5215864777565002,
      "learning_rate": 1.1659598529827539e-05,
      "loss": 0.0144,
      "step": 2900
    },
    {
      "epoch": 0.7404580152671756,
      "grad_norm": 0.017565440386533737,
      "learning_rate": 1.1546508340401473e-05,
      "loss": 0.0213,
      "step": 2910
    },
    {
      "epoch": 0.7430025445292621,
      "grad_norm": 0.2629789113998413,
      "learning_rate": 1.1433418150975404e-05,
      "loss": 0.0051,
      "step": 2920
    },
    {
      "epoch": 0.7455470737913485,
      "grad_norm": 0.07063543051481247,
      "learning_rate": 1.1320327961549336e-05,
      "loss": 0.0024,
      "step": 2930
    },
    {
      "epoch": 0.7480916030534351,
      "grad_norm": 4.683466911315918,
      "learning_rate": 1.1207237772123269e-05,
      "loss": 0.0079,
      "step": 2940
    },
    {
      "epoch": 0.7506361323155216,
      "grad_norm": 1.2442190647125244,
      "learning_rate": 1.1094147582697202e-05,
      "loss": 0.0129,
      "step": 2950
    },
    {
      "epoch": 0.7531806615776081,
      "grad_norm": 1.3911083936691284,
      "learning_rate": 1.0981057393271136e-05,
      "loss": 0.0057,
      "step": 2960
    },
    {
      "epoch": 0.7557251908396947,
      "grad_norm": 2.5614712238311768,
      "learning_rate": 1.0867967203845067e-05,
      "loss": 0.0069,
      "step": 2970
    },
    {
      "epoch": 0.7582697201017812,
      "grad_norm": 0.22300787270069122,
      "learning_rate": 1.0754877014419001e-05,
      "loss": 0.0075,
      "step": 2980
    },
    {
      "epoch": 0.7608142493638677,
      "grad_norm": 0.11393741518259048,
      "learning_rate": 1.0641786824992932e-05,
      "loss": 0.0042,
      "step": 2990
    },
    {
      "epoch": 0.7633587786259542,
      "grad_norm": 3.6102774143218994,
      "learning_rate": 1.0528696635566866e-05,
      "loss": 0.0065,
      "step": 3000
    },
    {
      "epoch": 0.7633587786259542,
      "eval_loss": 0.028713775798678398,
      "eval_runtime": 100.7408,
      "eval_samples_per_second": 79.233,
      "eval_steps_per_second": 0.625,
      "step": 3000
    },
    {
      "epoch": 0.7659033078880407,
      "grad_norm": 0.09737461805343628,
      "learning_rate": 1.0415606446140799e-05,
      "loss": 0.0064,
      "step": 3010
    },
    {
      "epoch": 0.7684478371501272,
      "grad_norm": 0.17835134267807007,
      "learning_rate": 1.0302516256714731e-05,
      "loss": 0.0043,
      "step": 3020
    },
    {
      "epoch": 0.7709923664122137,
      "grad_norm": 7.2621588706970215,
      "learning_rate": 1.0189426067288664e-05,
      "loss": 0.0051,
      "step": 3030
    },
    {
      "epoch": 0.7735368956743003,
      "grad_norm": 1.3994181156158447,
      "learning_rate": 1.0076335877862595e-05,
      "loss": 0.0052,
      "step": 3040
    },
    {
      "epoch": 0.7760814249363868,
      "grad_norm": 0.3551107943058014,
      "learning_rate": 9.963245688436529e-06,
      "loss": 0.0028,
      "step": 3050
    },
    {
      "epoch": 0.7786259541984732,
      "grad_norm": 3.8419830799102783,
      "learning_rate": 9.850155499010461e-06,
      "loss": 0.0092,
      "step": 3060
    },
    {
      "epoch": 0.7811704834605598,
      "grad_norm": 0.03152012452483177,
      "learning_rate": 9.737065309584394e-06,
      "loss": 0.0166,
      "step": 3070
    },
    {
      "epoch": 0.7837150127226463,
      "grad_norm": 1.2947672605514526,
      "learning_rate": 9.623975120158328e-06,
      "loss": 0.0044,
      "step": 3080
    },
    {
      "epoch": 0.7862595419847328,
      "grad_norm": 0.0874599888920784,
      "learning_rate": 9.51088493073226e-06,
      "loss": 0.0054,
      "step": 3090
    },
    {
      "epoch": 0.7888040712468194,
      "grad_norm": 0.06683657318353653,
      "learning_rate": 9.397794741306192e-06,
      "loss": 0.0064,
      "step": 3100
    },
    {
      "epoch": 0.7913486005089059,
      "grad_norm": 0.09484909474849701,
      "learning_rate": 9.284704551880124e-06,
      "loss": 0.0036,
      "step": 3110
    },
    {
      "epoch": 0.7938931297709924,
      "grad_norm": 0.599434494972229,
      "learning_rate": 9.171614362454057e-06,
      "loss": 0.004,
      "step": 3120
    },
    {
      "epoch": 0.7964376590330788,
      "grad_norm": 0.015714434906840324,
      "learning_rate": 9.058524173027991e-06,
      "loss": 0.0114,
      "step": 3130
    },
    {
      "epoch": 0.7989821882951654,
      "grad_norm": 0.23739086091518402,
      "learning_rate": 8.945433983601924e-06,
      "loss": 0.0048,
      "step": 3140
    },
    {
      "epoch": 0.8015267175572519,
      "grad_norm": 0.10904980450868607,
      "learning_rate": 8.832343794175856e-06,
      "loss": 0.0052,
      "step": 3150
    },
    {
      "epoch": 0.8040712468193384,
      "grad_norm": 3.4536166191101074,
      "learning_rate": 8.719253604749789e-06,
      "loss": 0.0165,
      "step": 3160
    },
    {
      "epoch": 0.806615776081425,
      "grad_norm": 0.13291485607624054,
      "learning_rate": 8.606163415323721e-06,
      "loss": 0.0119,
      "step": 3170
    },
    {
      "epoch": 0.8091603053435115,
      "grad_norm": 1.6434555053710938,
      "learning_rate": 8.493073225897654e-06,
      "loss": 0.009,
      "step": 3180
    },
    {
      "epoch": 0.811704834605598,
      "grad_norm": 0.05683699622750282,
      "learning_rate": 8.379983036471587e-06,
      "loss": 0.0131,
      "step": 3190
    },
    {
      "epoch": 0.8142493638676844,
      "grad_norm": 3.6685760021209717,
      "learning_rate": 8.266892847045519e-06,
      "loss": 0.0277,
      "step": 3200
    },
    {
      "epoch": 0.816793893129771,
      "grad_norm": 0.07328160107135773,
      "learning_rate": 8.153802657619452e-06,
      "loss": 0.002,
      "step": 3210
    },
    {
      "epoch": 0.8193384223918575,
      "grad_norm": 4.1098480224609375,
      "learning_rate": 8.040712468193384e-06,
      "loss": 0.0046,
      "step": 3220
    },
    {
      "epoch": 0.821882951653944,
      "grad_norm": 0.4604232907295227,
      "learning_rate": 7.927622278767317e-06,
      "loss": 0.0146,
      "step": 3230
    },
    {
      "epoch": 0.8244274809160306,
      "grad_norm": 0.6996018886566162,
      "learning_rate": 7.81453208934125e-06,
      "loss": 0.0025,
      "step": 3240
    },
    {
      "epoch": 0.8269720101781171,
      "grad_norm": 0.20250293612480164,
      "learning_rate": 7.701441899915184e-06,
      "loss": 0.0046,
      "step": 3250
    },
    {
      "epoch": 0.8295165394402035,
      "grad_norm": 1.6295372247695923,
      "learning_rate": 7.588351710489115e-06,
      "loss": 0.0018,
      "step": 3260
    },
    {
      "epoch": 0.8320610687022901,
      "grad_norm": 2.616621732711792,
      "learning_rate": 7.475261521063049e-06,
      "loss": 0.0074,
      "step": 3270
    },
    {
      "epoch": 0.8346055979643766,
      "grad_norm": 3.2906081676483154,
      "learning_rate": 7.362171331636981e-06,
      "loss": 0.0143,
      "step": 3280
    },
    {
      "epoch": 0.8371501272264631,
      "grad_norm": 1.3064147233963013,
      "learning_rate": 7.249081142210914e-06,
      "loss": 0.0056,
      "step": 3290
    },
    {
      "epoch": 0.8396946564885496,
      "grad_norm": 3.039973020553589,
      "learning_rate": 7.135990952784847e-06,
      "loss": 0.0095,
      "step": 3300
    },
    {
      "epoch": 0.8422391857506362,
      "grad_norm": 0.09989529848098755,
      "learning_rate": 7.022900763358779e-06,
      "loss": 0.0014,
      "step": 3310
    },
    {
      "epoch": 0.8447837150127226,
      "grad_norm": 0.2573828399181366,
      "learning_rate": 6.909810573932712e-06,
      "loss": 0.0061,
      "step": 3320
    },
    {
      "epoch": 0.8473282442748091,
      "grad_norm": 0.2958613932132721,
      "learning_rate": 6.796720384506644e-06,
      "loss": 0.004,
      "step": 3330
    },
    {
      "epoch": 0.8498727735368957,
      "grad_norm": 0.24500970542430878,
      "learning_rate": 6.683630195080577e-06,
      "loss": 0.0071,
      "step": 3340
    },
    {
      "epoch": 0.8524173027989822,
      "grad_norm": 1.2140284776687622,
      "learning_rate": 6.57054000565451e-06,
      "loss": 0.0036,
      "step": 3350
    },
    {
      "epoch": 0.8549618320610687,
      "grad_norm": 0.9599927067756653,
      "learning_rate": 6.457449816228443e-06,
      "loss": 0.0149,
      "step": 3360
    },
    {
      "epoch": 0.8575063613231552,
      "grad_norm": 0.27447447180747986,
      "learning_rate": 6.344359626802375e-06,
      "loss": 0.0074,
      "step": 3370
    },
    {
      "epoch": 0.8600508905852418,
      "grad_norm": 0.7029807567596436,
      "learning_rate": 6.231269437376308e-06,
      "loss": 0.007,
      "step": 3380
    },
    {
      "epoch": 0.8625954198473282,
      "grad_norm": 0.054117076098918915,
      "learning_rate": 6.118179247950241e-06,
      "loss": 0.0059,
      "step": 3390
    },
    {
      "epoch": 0.8651399491094147,
      "grad_norm": 0.4443982243537903,
      "learning_rate": 6.005089058524174e-06,
      "loss": 0.006,
      "step": 3400
    },
    {
      "epoch": 0.8676844783715013,
      "grad_norm": 3.201941967010498,
      "learning_rate": 5.8919988690981064e-06,
      "loss": 0.0118,
      "step": 3410
    },
    {
      "epoch": 0.8702290076335878,
      "grad_norm": 0.044088855385780334,
      "learning_rate": 5.778908679672038e-06,
      "loss": 0.0032,
      "step": 3420
    },
    {
      "epoch": 0.8727735368956743,
      "grad_norm": 1.8481858968734741,
      "learning_rate": 5.6658184902459716e-06,
      "loss": 0.0029,
      "step": 3430
    },
    {
      "epoch": 0.8753180661577609,
      "grad_norm": 0.014196817763149738,
      "learning_rate": 5.552728300819904e-06,
      "loss": 0.0062,
      "step": 3440
    },
    {
      "epoch": 0.8778625954198473,
      "grad_norm": 0.021374225616455078,
      "learning_rate": 5.439638111393837e-06,
      "loss": 0.0087,
      "step": 3450
    },
    {
      "epoch": 0.8804071246819338,
      "grad_norm": 0.0040234895423054695,
      "learning_rate": 5.326547921967769e-06,
      "loss": 0.0059,
      "step": 3460
    },
    {
      "epoch": 0.8829516539440203,
      "grad_norm": 0.07329881191253662,
      "learning_rate": 5.213457732541703e-06,
      "loss": 0.0023,
      "step": 3470
    },
    {
      "epoch": 0.8854961832061069,
      "grad_norm": 1.4617124795913696,
      "learning_rate": 5.100367543115635e-06,
      "loss": 0.0055,
      "step": 3480
    },
    {
      "epoch": 0.8880407124681934,
      "grad_norm": 2.9887943267822266,
      "learning_rate": 4.987277353689568e-06,
      "loss": 0.0106,
      "step": 3490
    },
    {
      "epoch": 0.8905852417302799,
      "grad_norm": 0.0617634542286396,
      "learning_rate": 4.8741871642635e-06,
      "loss": 0.0216,
      "step": 3500
    },
    {
      "epoch": 0.8905852417302799,
      "eval_loss": 0.02892346866428852,
      "eval_runtime": 101.0102,
      "eval_samples_per_second": 79.022,
      "eval_steps_per_second": 0.624,
      "step": 3500
    }
  ],
  "logging_steps": 10,
  "max_steps": 3930,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.9923556583082885e+18,
  "train_batch_size": 128,
  "trial_name": null,
  "trial_params": null
}