{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 6.0159010600706715,
  "eval_steps": 500,
  "global_step": 3405,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0176678445229682,
      "grad_norm": 8.093123435974121,
      "learning_rate": 1e-05,
      "loss": 3.4829,
      "step": 10
    },
    {
      "epoch": 0.0353356890459364,
      "grad_norm": 0.5401085615158081,
      "learning_rate": 2e-05,
      "loss": 3.4785,
      "step": 20
    },
    {
      "epoch": 0.053003533568904596,
      "grad_norm": 0.6342999339103699,
      "learning_rate": 1.9999569325372924e-05,
      "loss": 2.9312,
      "step": 30
    },
    {
      "epoch": 0.0706713780918728,
      "grad_norm": 0.7516262531280518,
      "learning_rate": 1.9998277338587826e-05,
      "loss": 2.9621,
      "step": 40
    },
    {
      "epoch": 0.08833922261484099,
      "grad_norm": 0.4993094205856323,
      "learning_rate": 1.9996124150929886e-05,
      "loss": 2.5796,
      "step": 50
    },
    {
      "epoch": 0.10600706713780919,
      "grad_norm": 1.0893628597259521,
      "learning_rate": 1.9993109947863768e-05,
      "loss": 2.5345,
      "step": 60
    },
    {
      "epoch": 0.12367491166077739,
      "grad_norm": 0.7348760962486267,
      "learning_rate": 1.9989234989017622e-05,
      "loss": 2.1804,
      "step": 70
    },
    {
      "epoch": 0.1413427561837456,
      "grad_norm": 0.7738515734672546,
      "learning_rate": 1.9984499608160744e-05,
      "loss": 1.9985,
      "step": 80
    },
    {
      "epoch": 0.15901060070671377,
      "grad_norm": 5.662350654602051,
      "learning_rate": 1.9978904213174812e-05,
      "loss": 2.0175,
      "step": 90
    },
    {
      "epoch": 0.17667844522968199,
      "grad_norm": 0.9150586724281311,
      "learning_rate": 1.997244928601875e-05,
      "loss": 1.8988,
      "step": 100
    },
    {
      "epoch": 0.19434628975265017,
      "grad_norm": 8.785883903503418,
      "learning_rate": 1.9965135382687235e-05,
      "loss": 1.7734,
      "step": 110
    },
    {
      "epoch": 0.21201413427561838,
      "grad_norm": 0.5746702551841736,
      "learning_rate": 1.9956963133162776e-05,
      "loss": 1.683,
      "step": 120
    },
    {
      "epoch": 0.22968197879858657,
      "grad_norm": 0.7510441541671753,
      "learning_rate": 1.9947933241361482e-05,
      "loss": 1.6329,
      "step": 130
    },
    {
      "epoch": 0.24734982332155478,
      "grad_norm": 0.7327967882156372,
      "learning_rate": 1.993804648507241e-05,
      "loss": 1.5906,
      "step": 140
    },
    {
      "epoch": 0.26501766784452296,
      "grad_norm": 0.7026442885398865,
      "learning_rate": 1.9927303715890573e-05,
      "loss": 1.5329,
      "step": 150
    },
    {
      "epoch": 0.2826855123674912,
      "grad_norm": 0.689364492893219,
      "learning_rate": 1.9915705859143597e-05,
      "loss": 1.3812,
      "step": 160
    },
    {
      "epoch": 0.3003533568904594,
      "grad_norm": 0.577924370765686,
      "learning_rate": 1.9903253913812003e-05,
      "loss": 1.3581,
      "step": 170
    },
    {
      "epoch": 0.31802120141342755,
      "grad_norm": 0.5152636766433716,
      "learning_rate": 1.9889948952443174e-05,
      "loss": 1.3432,
      "step": 180
    },
    {
      "epoch": 0.33568904593639576,
      "grad_norm": 0.5361068844795227,
      "learning_rate": 1.987579212105897e-05,
      "loss": 1.2808,
      "step": 190
    },
    {
      "epoch": 0.35335689045936397,
      "grad_norm": 0.562074601650238,
      "learning_rate": 1.9860784639057e-05,
      "loss": 1.2738,
      "step": 200
    },
    {
      "epoch": 0.3710247349823322,
      "grad_norm": 0.6112928986549377,
      "learning_rate": 1.9844927799105615e-05,
      "loss": 1.258,
      "step": 210
    },
    {
      "epoch": 0.38869257950530034,
      "grad_norm": 0.6768578290939331,
      "learning_rate": 1.9828222967032533e-05,
      "loss": 1.2198,
      "step": 220
    },
    {
      "epoch": 0.40636042402826855,
      "grad_norm": 0.5859869122505188,
      "learning_rate": 1.9810671581707223e-05,
      "loss": 1.2073,
      "step": 230
    },
    {
      "epoch": 0.42402826855123676,
      "grad_norm": 0.5085058808326721,
      "learning_rate": 1.979227515491695e-05,
      "loss": 1.2316,
      "step": 240
    },
    {
      "epoch": 0.4416961130742049,
      "grad_norm": 0.5284629464149475,
      "learning_rate": 1.9773035271236566e-05,
      "loss": 1.1577,
      "step": 250
    },
    {
      "epoch": 0.45936395759717313,
      "grad_norm": 0.5618858933448792,
      "learning_rate": 1.9752953587892013e-05,
      "loss": 1.1296,
      "step": 260
    },
    {
      "epoch": 0.47703180212014135,
      "grad_norm": 0.5541247725486755,
      "learning_rate": 1.973203183461759e-05,
      "loss": 1.1492,
      "step": 270
    },
    {
      "epoch": 0.49469964664310956,
      "grad_norm": 0.5731683969497681,
      "learning_rate": 1.9710271813506954e-05,
      "loss": 1.2129,
      "step": 280
    },
    {
      "epoch": 0.5123674911660777,
      "grad_norm": 1.798783302307129,
      "learning_rate": 1.96876753988579e-05,
      "loss": 1.1975,
      "step": 290
    },
    {
      "epoch": 0.5300353356890459,
      "grad_norm": 0.6141424179077148,
      "learning_rate": 1.9664244537010924e-05,
      "loss": 1.1168,
      "step": 300
    },
    {
      "epoch": 0.5477031802120141,
      "grad_norm": 0.49130862951278687,
      "learning_rate": 1.9639981246181557e-05,
      "loss": 1.1087,
      "step": 310
    },
    {
      "epoch": 0.5653710247349824,
      "grad_norm": 0.518574059009552,
      "learning_rate": 1.9614887616286544e-05,
      "loss": 1.0381,
      "step": 320
    },
    {
      "epoch": 0.5830388692579506,
      "grad_norm": 0.5865027904510498,
      "learning_rate": 1.958896580876383e-05,
      "loss": 1.0836,
      "step": 330
    },
    {
      "epoch": 0.6007067137809188,
      "grad_norm": 0.7070803046226501,
      "learning_rate": 1.9562218056386366e-05,
      "loss": 1.1234,
      "step": 340
    },
    {
      "epoch": 0.6183745583038869,
      "grad_norm": 0.5792056322097778,
      "learning_rate": 1.9534646663069816e-05,
      "loss": 1.0202,
      "step": 350
    },
    {
      "epoch": 0.6360424028268551,
      "grad_norm": 0.6844910383224487,
      "learning_rate": 1.9506254003674084e-05,
      "loss": 1.1047,
      "step": 360
    },
    {
      "epoch": 0.6537102473498233,
      "grad_norm": 0.5521841049194336,
      "learning_rate": 1.9477042523798762e-05,
      "loss": 1.0834,
      "step": 370
    },
    {
      "epoch": 0.6713780918727915,
      "grad_norm": 0.578198254108429,
      "learning_rate": 1.9447014739572503e-05,
      "loss": 0.9927,
      "step": 380
    },
    {
      "epoch": 0.6890459363957597,
      "grad_norm": 0.47503796219825745,
      "learning_rate": 1.9416173237436252e-05,
      "loss": 1.051,
      "step": 390
    },
    {
      "epoch": 0.7067137809187279,
      "grad_norm": 0.5625858902931213,
      "learning_rate": 1.9384520673920502e-05,
      "loss": 1.0239,
      "step": 400
    },
    {
      "epoch": 0.7243816254416962,
      "grad_norm": 0.535770833492279,
      "learning_rate": 1.9352059775416442e-05,
      "loss": 1.0245,
      "step": 410
    },
    {
      "epoch": 0.7420494699646644,
      "grad_norm": 0.5579429864883423,
      "learning_rate": 1.931879333794115e-05,
      "loss": 0.9748,
      "step": 420
    },
    {
      "epoch": 0.7597173144876325,
      "grad_norm": 0.5888713598251343,
      "learning_rate": 1.928472422689674e-05,
      "loss": 1.0526,
      "step": 430
    },
    {
      "epoch": 0.7773851590106007,
      "grad_norm": 0.562355637550354,
      "learning_rate": 1.9249855376823542e-05,
      "loss": 0.9808,
      "step": 440
    },
    {
      "epoch": 0.7950530035335689,
      "grad_norm": 0.45564863085746765,
      "learning_rate": 1.9214189791147363e-05,
      "loss": 0.9943,
      "step": 450
    },
    {
      "epoch": 0.8127208480565371,
      "grad_norm": 0.5389313101768494,
      "learning_rate": 1.917773054192076e-05,
      "loss": 0.9906,
      "step": 460
    },
    {
      "epoch": 0.8303886925795053,
      "grad_norm": 0.5588345527648926,
      "learning_rate": 1.9140480769558448e-05,
      "loss": 1.0077,
      "step": 470
    },
    {
      "epoch": 0.8480565371024735,
      "grad_norm": 0.6023375988006592,
      "learning_rate": 1.9102443682566792e-05,
      "loss": 0.9995,
      "step": 480
    },
    {
      "epoch": 0.8657243816254417,
      "grad_norm": 0.5838338732719421,
      "learning_rate": 1.9063622557267443e-05,
      "loss": 0.9542,
      "step": 490
    },
    {
      "epoch": 0.8833922261484098,
      "grad_norm": 0.6777709722518921,
      "learning_rate": 1.9024020737515135e-05,
      "loss": 1.0053,
      "step": 500
    },
    {
      "epoch": 0.901060070671378,
      "grad_norm": 0.6219585537910461,
      "learning_rate": 1.8983641634409657e-05,
      "loss": 0.9962,
      "step": 510
    },
    {
      "epoch": 0.9187279151943463,
      "grad_norm": 0.6651351451873779,
      "learning_rate": 1.894248872600204e-05,
      "loss": 0.9596,
      "step": 520
    },
    {
      "epoch": 0.9363957597173145,
      "grad_norm": 0.5591141581535339,
      "learning_rate": 1.8900565556994986e-05,
      "loss": 0.9638,
      "step": 530
    },
    {
      "epoch": 0.9540636042402827,
      "grad_norm": 0.6066205501556396,
      "learning_rate": 1.8857875738437526e-05,
      "loss": 0.9468,
      "step": 540
    },
    {
      "epoch": 0.9717314487632509,
      "grad_norm": 0.584261417388916,
      "learning_rate": 1.8814422947414e-05,
      "loss": 0.9694,
      "step": 550
    },
    {
      "epoch": 0.9893992932862191,
      "grad_norm": 0.553644061088562,
      "learning_rate": 1.877021092672732e-05,
      "loss": 0.9886,
      "step": 560
    },
    {
      "epoch": 1.0070671378091873,
      "grad_norm": 0.877346396446228,
      "learning_rate": 1.872524348457659e-05,
      "loss": 1.0424,
      "step": 570
    },
    {
      "epoch": 1.0247349823321554,
      "grad_norm": 0.5153814554214478,
      "learning_rate": 1.867952449422909e-05,
      "loss": 0.9541,
      "step": 580
    },
    {
      "epoch": 1.0424028268551238,
      "grad_norm": 0.6072525382041931,
      "learning_rate": 1.863305789368664e-05,
      "loss": 0.9835,
      "step": 590
    },
    {
      "epoch": 1.0600706713780919,
      "grad_norm": 2.0284619331359863,
      "learning_rate": 1.8585847685346415e-05,
      "loss": 0.9128,
      "step": 600
    },
    {
      "epoch": 1.0777385159010602,
      "grad_norm": 0.5690125226974487,
      "learning_rate": 1.853789793565618e-05,
      "loss": 0.9317,
      "step": 610
    },
    {
      "epoch": 1.0954063604240283,
      "grad_norm": 0.5882884860038757,
      "learning_rate": 1.8489212774764064e-05,
      "loss": 0.9644,
      "step": 620
    },
    {
      "epoch": 1.1130742049469964,
      "grad_norm": 0.7221271991729736,
      "learning_rate": 1.8439796396162756e-05,
      "loss": 1.024,
      "step": 630
    },
    {
      "epoch": 1.1307420494699647,
      "grad_norm": 0.6504462361335754,
      "learning_rate": 1.8389653056328344e-05,
      "loss": 0.9529,
      "step": 640
    },
    {
      "epoch": 1.1484098939929328,
      "grad_norm": 0.5796203017234802,
      "learning_rate": 1.833878707435367e-05,
      "loss": 0.9081,
      "step": 650
    },
    {
      "epoch": 1.1660777385159011,
      "grad_norm": 0.5094664096832275,
      "learning_rate": 1.8287202831576292e-05,
      "loss": 0.9413,
      "step": 660
    },
    {
      "epoch": 1.1837455830388692,
      "grad_norm": 0.5529056191444397,
      "learning_rate": 1.8234904771201115e-05,
      "loss": 0.9494,
      "step": 670
    },
    {
      "epoch": 1.2014134275618376,
      "grad_norm": 0.5121185779571533,
      "learning_rate": 1.8181897397917672e-05,
      "loss": 0.8958,
      "step": 680
    },
    {
      "epoch": 1.2190812720848057,
      "grad_norm": 0.6205973029136658,
      "learning_rate": 1.8128185277512106e-05,
      "loss": 0.9541,
      "step": 690
    },
    {
      "epoch": 1.2367491166077738,
      "grad_norm": 0.4824579954147339,
      "learning_rate": 1.80737730364739e-05,
      "loss": 0.9269,
      "step": 700
    },
    {
      "epoch": 1.254416961130742,
      "grad_norm": 0.6322911381721497,
      "learning_rate": 1.8018665361597378e-05,
      "loss": 0.9151,
      "step": 710
    },
    {
      "epoch": 1.2720848056537102,
      "grad_norm": 0.5373772382736206,
      "learning_rate": 1.7962866999578005e-05,
      "loss": 0.8693,
      "step": 720
    },
    {
      "epoch": 1.2897526501766785,
      "grad_norm": 0.640831470489502,
      "learning_rate": 1.7906382756603536e-05,
      "loss": 0.9646,
      "step": 730
    },
    {
      "epoch": 1.3074204946996466,
      "grad_norm": 0.6338053941726685,
      "learning_rate": 1.784921749794002e-05,
      "loss": 0.9903,
      "step": 740
    },
    {
      "epoch": 1.325088339222615,
      "grad_norm": 0.692629337310791,
      "learning_rate": 1.7791376147512754e-05,
      "loss": 0.9846,
      "step": 750
    },
    {
      "epoch": 1.342756183745583,
      "grad_norm": 0.7451064586639404,
      "learning_rate": 1.773286368748214e-05,
      "loss": 0.9787,
      "step": 760
    },
    {
      "epoch": 1.3604240282685511,
      "grad_norm": 0.7188582420349121,
      "learning_rate": 1.7673685157814556e-05,
      "loss": 0.9304,
      "step": 770
    },
    {
      "epoch": 1.3780918727915195,
      "grad_norm": 0.5633394718170166,
      "learning_rate": 1.761384565584825e-05,
      "loss": 0.9507,
      "step": 780
    },
    {
      "epoch": 1.3957597173144876,
      "grad_norm": 0.5163728594779968,
      "learning_rate": 1.7553350335854253e-05,
      "loss": 0.9334,
      "step": 790
    },
    {
      "epoch": 1.4134275618374559,
      "grad_norm": 0.5537968277931213,
      "learning_rate": 1.7492204408592447e-05,
      "loss": 0.8753,
      "step": 800
    },
    {
      "epoch": 1.431095406360424,
      "grad_norm": 0.5419259071350098,
      "learning_rate": 1.7430413140862705e-05,
      "loss": 0.8902,
      "step": 810
    },
    {
      "epoch": 1.4487632508833923,
      "grad_norm": 0.5655599236488342,
      "learning_rate": 1.7367981855051275e-05,
      "loss": 0.8664,
      "step": 820
    },
    {
      "epoch": 1.4664310954063604,
      "grad_norm": 0.6520758867263794,
      "learning_rate": 1.73049159286723e-05,
      "loss": 0.8684,
      "step": 830
    },
    {
      "epoch": 1.4840989399293285,
      "grad_norm": 0.5785622596740723,
      "learning_rate": 1.7241220793904644e-05,
      "loss": 0.9331,
      "step": 840
    },
    {
      "epoch": 1.5017667844522968,
      "grad_norm": 0.5621282458305359,
      "learning_rate": 1.717690193712399e-05,
      "loss": 0.9055,
      "step": 850
    },
    {
      "epoch": 1.5194346289752652,
      "grad_norm": 0.5962981581687927,
      "learning_rate": 1.711196489843027e-05,
      "loss": 0.8962,
      "step": 860
    },
    {
      "epoch": 1.5371024734982333,
      "grad_norm": 1.2902908325195312,
      "learning_rate": 1.704641527117047e-05,
      "loss": 0.9052,
      "step": 870
    },
    {
      "epoch": 1.5547703180212014,
      "grad_norm": 0.5082823038101196,
      "learning_rate": 1.6980258701456843e-05,
      "loss": 0.8873,
      "step": 880
    },
    {
      "epoch": 1.5724381625441697,
      "grad_norm": 0.5958746075630188,
      "learning_rate": 1.6913500887680588e-05,
      "loss": 0.9228,
      "step": 890
    },
    {
      "epoch": 1.5901060070671378,
      "grad_norm": 0.7719606161117554,
      "learning_rate": 1.6846147580021016e-05,
      "loss": 0.921,
      "step": 900
    },
    {
      "epoch": 1.6077738515901059,
      "grad_norm": 2.6116905212402344,
      "learning_rate": 1.6778204579950258e-05,
      "loss": 0.8997,
      "step": 910
    },
    {
      "epoch": 1.6254416961130742,
      "grad_norm": 0.5398052930831909,
      "learning_rate": 1.6709677739733555e-05,
      "loss": 0.8912,
      "step": 920
    },
    {
      "epoch": 1.6431095406360425,
      "grad_norm": 0.6152315735816956,
      "learning_rate": 1.6640572961925182e-05,
      "loss": 0.893,
      "step": 930
    },
    {
      "epoch": 1.6607773851590106,
      "grad_norm": 0.5041419863700867,
      "learning_rate": 1.657089619886002e-05,
      "loss": 0.8951,
      "step": 940
    },
    {
      "epoch": 1.6784452296819787,
      "grad_norm": 0.5210781693458557,
      "learning_rate": 1.650065345214086e-05,
      "loss": 0.9144,
      "step": 950
    },
    {
      "epoch": 1.696113074204947,
      "grad_norm": 0.6072030067443848,
      "learning_rate": 1.6429850772121448e-05,
      "loss": 0.9113,
      "step": 960
    },
    {
      "epoch": 1.7137809187279152,
      "grad_norm": 0.6967266201972961,
      "learning_rate": 1.635849425738535e-05,
      "loss": 0.9556,
      "step": 970
    },
    {
      "epoch": 1.7314487632508833,
      "grad_norm": 0.5619155168533325,
      "learning_rate": 1.6286590054220643e-05,
      "loss": 0.8637,
      "step": 980
    },
    {
      "epoch": 1.7491166077738516,
      "grad_norm": 0.6261915564537048,
      "learning_rate": 1.6214144356090494e-05,
      "loss": 0.877,
      "step": 990
    },
    {
      "epoch": 1.76678445229682,
      "grad_norm": 0.5935543775558472,
      "learning_rate": 1.6141163403099716e-05,
      "loss": 0.9611,
      "step": 1000
    },
    {
      "epoch": 1.784452296819788,
      "grad_norm": 0.6743911504745483,
      "learning_rate": 1.6067653481457254e-05,
      "loss": 0.9007,
      "step": 1010
    },
    {
      "epoch": 1.802120141342756,
      "grad_norm": 0.6876866221427917,
      "learning_rate": 1.5993620922934716e-05,
      "loss": 0.8605,
      "step": 1020
    },
    {
      "epoch": 1.8197879858657244,
      "grad_norm": 2.510390043258667,
      "learning_rate": 1.591907210432102e-05,
      "loss": 0.864,
      "step": 1030
    },
    {
      "epoch": 1.8374558303886925,
      "grad_norm": 1.431142807006836,
      "learning_rate": 1.5844013446873087e-05,
      "loss": 0.9,
      "step": 1040
    },
    {
      "epoch": 1.8551236749116606,
      "grad_norm": 0.6549361348152161,
      "learning_rate": 1.5768451415762784e-05,
      "loss": 0.8726,
      "step": 1050
    },
    {
      "epoch": 1.872791519434629,
      "grad_norm": 0.5458822846412659,
      "learning_rate": 1.5692392519520022e-05,
      "loss": 0.9468,
      "step": 1060
    },
    {
      "epoch": 1.8904593639575973,
      "grad_norm": 1.6610078811645508,
      "learning_rate": 1.5615843309472162e-05,
      "loss": 0.8806,
      "step": 1070
    },
    {
      "epoch": 1.9081272084805654,
      "grad_norm": 0.5835666656494141,
      "learning_rate": 1.5538810379179694e-05,
      "loss": 0.8934,
      "step": 1080
    },
    {
      "epoch": 1.9257950530035335,
      "grad_norm": 0.5027960538864136,
      "learning_rate": 1.5461300363868326e-05,
      "loss": 0.8641,
      "step": 1090
    },
    {
      "epoch": 1.9434628975265018,
      "grad_norm": 0.5919517278671265,
      "learning_rate": 1.538331993985745e-05,
      "loss": 0.8959,
      "step": 1100
    },
    {
      "epoch": 1.96113074204947,
      "grad_norm": 0.7145887017250061,
      "learning_rate": 1.5304875823985067e-05,
      "loss": 0.9284,
      "step": 1110
    },
    {
      "epoch": 1.978798586572438,
      "grad_norm": 0.6099452376365662,
      "learning_rate": 1.5225974773029246e-05,
      "loss": 0.8896,
      "step": 1120
    },
    {
      "epoch": 1.9964664310954063,
      "grad_norm": 0.5641235709190369,
      "learning_rate": 1.5146623583126134e-05,
      "loss": 0.8888,
      "step": 1130
    },
    {
      "epoch": 2.0141342756183747,
      "grad_norm": 3.2261722087860107,
      "learning_rate": 1.5066829089184545e-05,
      "loss": 0.8495,
      "step": 1140
    },
    {
      "epoch": 2.0318021201413425,
      "grad_norm": 0.5612754821777344,
      "learning_rate": 1.498659816429727e-05,
      "loss": 0.8446,
      "step": 1150
    },
    {
      "epoch": 2.049469964664311,
      "grad_norm": 0.5648794174194336,
      "learning_rate": 1.4905937719149038e-05,
      "loss": 0.9185,
      "step": 1160
    },
    {
      "epoch": 2.067137809187279,
      "grad_norm": 0.5706191658973694,
      "learning_rate": 1.4824854701421277e-05,
      "loss": 0.8588,
      "step": 1170
    },
    {
      "epoch": 2.0848056537102475,
      "grad_norm": 0.5422844290733337,
      "learning_rate": 1.4743356095193665e-05,
      "loss": 0.8906,
      "step": 1180
    },
    {
      "epoch": 2.1024734982332154,
      "grad_norm": 0.6885594725608826,
      "learning_rate": 1.4661448920342585e-05,
      "loss": 0.871,
      "step": 1190
    },
    {
      "epoch": 2.1201413427561837,
      "grad_norm": 3.2255871295928955,
      "learning_rate": 1.4579140231936415e-05,
      "loss": 0.8621,
      "step": 1200
    },
    {
      "epoch": 2.137809187279152,
      "grad_norm": 0.5476865172386169,
      "learning_rate": 1.4496437119627907e-05,
      "loss": 0.8193,
      "step": 1210
    },
    {
      "epoch": 2.1554770318021204,
      "grad_norm": 0.6046047210693359,
      "learning_rate": 1.4413346707043467e-05,
      "loss": 0.8533,
      "step": 1220
    },
    {
      "epoch": 2.1731448763250882,
      "grad_norm": 0.5473910570144653,
      "learning_rate": 1.4329876151169581e-05,
      "loss": 0.8563,
      "step": 1230
    },
    {
      "epoch": 2.1908127208480566,
      "grad_norm": 0.5437634587287903,
      "learning_rate": 1.4246032641736362e-05,
      "loss": 0.8801,
      "step": 1240
    },
    {
      "epoch": 2.208480565371025,
      "grad_norm": 0.5767475366592407,
      "learning_rate": 1.4161823400598234e-05,
      "loss": 0.8802,
      "step": 1250
    },
    {
      "epoch": 2.2261484098939928,
      "grad_norm": 0.5995182991027832,
      "learning_rate": 1.4077255681111905e-05,
      "loss": 0.921,
      "step": 1260
    },
    {
      "epoch": 2.243816254416961,
      "grad_norm": 1.4571260213851929,
      "learning_rate": 1.3992336767511585e-05,
      "loss": 0.8508,
      "step": 1270
    },
    {
      "epoch": 2.2614840989399294,
      "grad_norm": 0.6092621088027954,
      "learning_rate": 1.3907073974281562e-05,
      "loss": 0.8886,
      "step": 1280
    },
    {
      "epoch": 2.2791519434628977,
      "grad_norm": 0.6601437926292419,
      "learning_rate": 1.3821474645526174e-05,
      "loss": 0.9181,
      "step": 1290
    },
    {
      "epoch": 2.2968197879858656,
      "grad_norm": 0.5523033738136292,
      "learning_rate": 1.3735546154337218e-05,
      "loss": 0.87,
      "step": 1300
    },
    {
      "epoch": 2.314487632508834,
      "grad_norm": 0.5680422782897949,
      "learning_rate": 1.3649295902158874e-05,
      "loss": 0.9032,
      "step": 1310
    },
    {
      "epoch": 2.3321554770318023,
      "grad_norm": 0.5610331892967224,
      "learning_rate": 1.3562731318150177e-05,
      "loss": 0.8743,
      "step": 1320
    },
    {
      "epoch": 2.34982332155477,
      "grad_norm": 0.6986725926399231,
      "learning_rate": 1.3475859858545121e-05,
      "loss": 0.8976,
      "step": 1330
    },
    {
      "epoch": 2.3674911660777385,
      "grad_norm": 0.6468251943588257,
      "learning_rate": 1.3388689006010394e-05,
      "loss": 0.8701,
      "step": 1340
    },
    {
      "epoch": 2.385159010600707,
      "grad_norm": 0.6808676719665527,
      "learning_rate": 1.330122626900088e-05,
      "loss": 0.8447,
      "step": 1350
    },
    {
      "epoch": 2.402826855123675,
      "grad_norm": 0.9036700129508972,
      "learning_rate": 1.3213479181112906e-05,
      "loss": 0.8917,
      "step": 1360
    },
    {
      "epoch": 2.420494699646643,
      "grad_norm": 0.5925520062446594,
      "learning_rate": 1.3125455300435343e-05,
      "loss": 0.8911,
      "step": 1370
    },
    {
      "epoch": 2.4381625441696113,
      "grad_norm": 0.610488772392273,
      "learning_rate": 1.303716220889859e-05,
      "loss": 0.8397,
      "step": 1380
    },
    {
      "epoch": 2.4558303886925796,
      "grad_norm": 0.7890755534172058,
      "learning_rate": 1.2948607511621498e-05,
      "loss": 0.9072,
      "step": 1390
    },
    {
      "epoch": 2.4734982332155475,
      "grad_norm": 0.5136253237724304,
      "learning_rate": 1.2859798836256316e-05,
      "loss": 0.847,
      "step": 1400
    },
    {
      "epoch": 2.491166077738516,
      "grad_norm": 0.5742533206939697,
      "learning_rate": 1.277074383233167e-05,
      "loss": 0.8632,
      "step": 1410
    },
    {
      "epoch": 2.508833922261484,
      "grad_norm": 0.6180789470672607,
      "learning_rate": 1.2681450170593683e-05,
      "loss": 0.8637,
      "step": 1420
    },
    {
      "epoch": 2.5265017667844525,
      "grad_norm": 0.604252815246582,
      "learning_rate": 1.2591925542345244e-05,
      "loss": 0.865,
      "step": 1430
    },
    {
      "epoch": 2.5441696113074204,
      "grad_norm": 0.565585732460022,
      "learning_rate": 1.2502177658783538e-05,
      "loss": 0.8873,
      "step": 1440
    },
    {
      "epoch": 2.5618374558303887,
      "grad_norm": 0.5197896957397461,
      "learning_rate": 1.2412214250335815e-05,
      "loss": 0.8654,
      "step": 1450
    },
    {
      "epoch": 2.579505300353357,
      "grad_norm": 0.5932325720787048,
      "learning_rate": 1.2322043065993556e-05,
      "loss": 0.8402,
      "step": 1460
    },
    {
      "epoch": 2.597173144876325,
      "grad_norm": 0.4967307150363922,
      "learning_rate": 1.2231671872644995e-05,
      "loss": 0.8662,
      "step": 1470
    },
    {
      "epoch": 2.614840989399293,
      "grad_norm": 0.5910283327102661,
      "learning_rate": 1.214110845440613e-05,
      "loss": 0.8501,
      "step": 1480
    },
    {
      "epoch": 2.6325088339222615,
      "grad_norm": 0.5208580493927002,
      "learning_rate": 1.2050360611950245e-05,
      "loss": 0.8566,
      "step": 1490
    },
    {
      "epoch": 2.65017667844523,
      "grad_norm": 0.6784213781356812,
      "learning_rate": 1.1959436161835971e-05,
      "loss": 0.8649,
      "step": 1500
    },
    {
      "epoch": 2.6678445229681977,
      "grad_norm": 0.6994915008544922,
      "learning_rate": 1.1868342935834043e-05,
      "loss": 0.8613,
      "step": 1510
    },
    {
      "epoch": 2.685512367491166,
      "grad_norm": 0.5287041068077087,
      "learning_rate": 1.1777088780252688e-05,
      "loss": 0.8911,
      "step": 1520
    },
    {
      "epoch": 2.7031802120141344,
      "grad_norm": 0.5094373822212219,
      "learning_rate": 1.1685681555261788e-05,
      "loss": 0.8625,
      "step": 1530
    },
    {
      "epoch": 2.7208480565371023,
      "grad_norm": 0.6024549603462219,
      "learning_rate": 1.1594129134215852e-05,
      "loss": 0.8821,
      "step": 1540
    },
    {
      "epoch": 2.7385159010600706,
      "grad_norm": 0.5689032673835754,
      "learning_rate": 1.1502439402975842e-05,
      "loss": 0.8474,
      "step": 1550
    },
    {
      "epoch": 2.756183745583039,
      "grad_norm": 0.5587700009346008,
      "learning_rate": 1.141062025922991e-05,
      "loss": 0.8609,
      "step": 1560
    },
    {
      "epoch": 2.7738515901060072,
      "grad_norm": 0.6656576991081238,
      "learning_rate": 1.1318679611813166e-05,
      "loss": 0.8577,
      "step": 1570
    },
    {
      "epoch": 2.791519434628975,
      "grad_norm": 0.6215320825576782,
      "learning_rate": 1.1226625380026407e-05,
      "loss": 0.8226,
      "step": 1580
    },
    {
      "epoch": 2.8091872791519434,
      "grad_norm": 0.5720105171203613,
      "learning_rate": 1.1134465492954028e-05,
      "loss": 0.8981,
      "step": 1590
    },
    {
      "epoch": 2.8268551236749118,
      "grad_norm": 0.5954132080078125,
      "learning_rate": 1.1042207888781031e-05,
      "loss": 0.8478,
      "step": 1600
    },
    {
      "epoch": 2.8445229681978796,
      "grad_norm": 0.6065002083778381,
      "learning_rate": 1.0949860514109265e-05,
      "loss": 0.8569,
      "step": 1610
    },
    {
      "epoch": 2.862190812720848,
      "grad_norm": 0.5527032017707825,
      "learning_rate": 1.085743132327296e-05,
      "loss": 0.868,
      "step": 1620
    },
    {
      "epoch": 2.8798586572438163,
      "grad_norm": 0.5579487681388855,
      "learning_rate": 1.0764928277653577e-05,
      "loss": 0.8946,
      "step": 1630
    },
    {
      "epoch": 2.8975265017667846,
      "grad_norm": 0.5760249495506287,
      "learning_rate": 1.067235934499405e-05,
      "loss": 0.8359,
      "step": 1640
    },
    {
      "epoch": 2.9151943462897525,
      "grad_norm": 0.6175970435142517,
      "learning_rate": 1.057973249871249e-05,
      "loss": 0.8355,
      "step": 1650
    },
    {
      "epoch": 2.932862190812721,
      "grad_norm": 0.5847891569137573,
      "learning_rate": 1.0487055717215394e-05,
      "loss": 0.8035,
      "step": 1660
    },
    {
      "epoch": 2.950530035335689,
      "grad_norm": 1.2150473594665527,
      "learning_rate": 1.039433698321042e-05,
      "loss": 0.8211,
      "step": 1670
    },
    {
      "epoch": 2.968197879858657,
      "grad_norm": 0.663726270198822,
      "learning_rate": 1.0301584283018813e-05,
      "loss": 0.8213,
      "step": 1680
    },
    {
      "epoch": 2.9858657243816253,
      "grad_norm": 0.5659819841384888,
      "learning_rate": 1.020880560588748e-05,
      "loss": 0.8943,
      "step": 1690
    },
    {
      "epoch": 3.0035335689045937,
      "grad_norm": 0.5880382657051086,
      "learning_rate": 1.0116008943300852e-05,
      "loss": 0.8894,
      "step": 1700
    },
    {
      "epoch": 3.021201413427562,
      "grad_norm": 0.5816975831985474,
      "learning_rate": 1.0023202288292552e-05,
      "loss": 0.8256,
      "step": 1710
    },
    {
      "epoch": 3.03886925795053,
      "grad_norm": 0.6161662936210632,
      "learning_rate": 9.930393634756877e-06,
      "loss": 0.8375,
      "step": 1720
    },
    {
      "epoch": 3.056537102473498,
      "grad_norm": 0.5282244682312012,
      "learning_rate": 9.837590976760283e-06,
      "loss": 0.8738,
      "step": 1730
    },
    {
      "epoch": 3.0742049469964665,
      "grad_norm": 0.5946238040924072,
      "learning_rate": 9.744802307852794e-06,
      "loss": 0.8679,
      "step": 1740
    },
    {
      "epoch": 3.091872791519435,
      "grad_norm": 0.5638343095779419,
      "learning_rate": 9.652035620379481e-06,
      "loss": 0.8447,
      "step": 1750
    },
    {
      "epoch": 3.1095406360424027,
      "grad_norm": 1.077945351600647,
      "learning_rate": 9.559298904792054e-06,
      "loss": 0.8454,
      "step": 1760
    },
    {
      "epoch": 3.127208480565371,
      "grad_norm": 0.6322164535522461,
      "learning_rate": 9.466600148960597e-06,
      "loss": 0.8682,
      "step": 1770
    },
    {
      "epoch": 3.1448763250883394,
      "grad_norm": 0.5520142912864685,
      "learning_rate": 9.373947337485521e-06,
      "loss": 0.7946,
      "step": 1780
    },
    {
      "epoch": 3.1625441696113072,
      "grad_norm": 0.5950481295585632,
      "learning_rate": 9.281348451009837e-06,
      "loss": 0.8241,
      "step": 1790
    },
    {
      "epoch": 3.1802120141342756,
      "grad_norm": 0.614825427532196,
      "learning_rate": 9.188811465531725e-06,
      "loss": 0.8697,
      "step": 1800
    },
    {
      "epoch": 3.197879858657244,
      "grad_norm": 0.5201759934425354,
      "learning_rate": 9.096344351717528e-06,
      "loss": 0.9154,
      "step": 1810
    },
    {
      "epoch": 3.215547703180212,
      "grad_norm": 0.6227285861968994,
      "learning_rate": 9.003955074215198e-06,
      "loss": 0.8995,
      "step": 1820
    },
    {
      "epoch": 3.23321554770318,
      "grad_norm": 0.5124209523200989,
      "learning_rate": 8.911651590968259e-06,
      "loss": 0.8432,
      "step": 1830
    },
    {
      "epoch": 3.2508833922261484,
      "grad_norm": 0.5678322911262512,
      "learning_rate": 8.819441852530358e-06,
      "loss": 0.837,
      "step": 1840
    },
    {
      "epoch": 3.2685512367491167,
      "grad_norm": 0.7585355639457703,
      "learning_rate": 8.72733380138044e-06,
      "loss": 0.8434,
      "step": 1850
    },
    {
      "epoch": 3.2862190812720846,
      "grad_norm": 0.6334419250488281,
      "learning_rate": 8.63533537123861e-06,
      "loss": 0.8203,
      "step": 1860
    },
    {
      "epoch": 3.303886925795053,
      "grad_norm": 0.7814059853553772,
      "learning_rate": 8.543454486382803e-06,
      "loss": 0.8214,
      "step": 1870
    },
    {
      "epoch": 3.3215547703180213,
      "grad_norm": 0.5241668820381165,
      "learning_rate": 8.451699060966174e-06,
      "loss": 0.8379,
      "step": 1880
    },
    {
      "epoch": 3.3392226148409896,
      "grad_norm": 0.5536594390869141,
      "learning_rate": 8.360076998335447e-06,
      "loss": 0.8338,
      "step": 1890
    },
    {
      "epoch": 3.3568904593639575,
      "grad_norm": 0.5642170310020447,
      "learning_rate": 8.268596190350158e-06,
      "loss": 0.8628,
      "step": 1900
    },
    {
      "epoch": 3.374558303886926,
      "grad_norm": 0.5817282199859619,
      "learning_rate": 8.177264516702875e-06,
      "loss": 0.8496,
      "step": 1910
    },
    {
      "epoch": 3.392226148409894,
      "grad_norm": 0.6194543838500977,
      "learning_rate": 8.086089844240495e-06,
      "loss": 0.8059,
      "step": 1920
    },
    {
      "epoch": 3.4098939929328624,
      "grad_norm": 0.737491250038147,
      "learning_rate": 7.995080026286632e-06,
      "loss": 0.8449,
      "step": 1930
    },
    {
      "epoch": 3.4275618374558303,
      "grad_norm": 0.5627506971359253,
      "learning_rate": 7.904242901965171e-06,
      "loss": 0.8297,
      "step": 1940
    },
    {
      "epoch": 3.4452296819787986,
      "grad_norm": 0.8182918429374695,
      "learning_rate": 7.81358629552504e-06,
      "loss": 0.8627,
      "step": 1950
    },
    {
      "epoch": 3.462897526501767,
      "grad_norm": 0.5409490466117859,
      "learning_rate": 7.723118015666266e-06,
      "loss": 0.8837,
      "step": 1960
    },
    {
      "epoch": 3.480565371024735,
      "grad_norm": 0.5023618340492249,
      "learning_rate": 7.632845854867393e-06,
      "loss": 0.8592,
      "step": 1970
    },
    {
      "epoch": 3.498233215547703,
      "grad_norm": 0.5892724394798279,
      "learning_rate": 7.542777588714256e-06,
      "loss": 0.8318,
      "step": 1980
    },
    {
      "epoch": 3.5159010600706715,
      "grad_norm": 0.5022056698799133,
      "learning_rate": 7.452920975230247e-06,
      "loss": 0.8231,
      "step": 1990
    },
    {
      "epoch": 3.53356890459364,
      "grad_norm": 0.6234381198883057,
      "learning_rate": 7.363283754208061e-06,
      "loss": 0.7936,
      "step": 2000
    },
    {
      "epoch": 3.5512367491166077,
      "grad_norm": 0.5600544214248657,
      "learning_rate": 7.273873646543044e-06,
      "loss": 0.9023,
      "step": 2010
    },
    {
      "epoch": 3.568904593639576,
      "grad_norm": 0.5818344354629517,
      "learning_rate": 7.184698353568157e-06,
      "loss": 0.8629,
      "step": 2020
    },
    {
      "epoch": 3.586572438162544,
      "grad_norm": 0.5921304225921631,
      "learning_rate": 7.095765556390606e-06,
      "loss": 0.8372,
      "step": 2030
    },
    {
      "epoch": 3.604240282685512,
      "grad_norm": 0.6511384844779968,
      "learning_rate": 7.007082915230247e-06,
      "loss": 0.8143,
      "step": 2040
    },
    {
      "epoch": 3.6219081272084805,
      "grad_norm": 0.5591942071914673,
      "learning_rate": 6.918658068759754e-06,
      "loss": 0.8483,
      "step": 2050
    },
    {
      "epoch": 3.639575971731449,
      "grad_norm": 0.5008967518806458,
      "learning_rate": 6.8304986334466884e-06,
      "loss": 0.8219,
      "step": 2060
    },
    {
      "epoch": 3.657243816254417,
      "grad_norm": 0.5557330250740051,
      "learning_rate": 6.742612202897436e-06,
      "loss": 0.8225,
      "step": 2070
    },
    {
      "epoch": 3.674911660777385,
      "grad_norm": 0.524795413017273,
      "learning_rate": 6.655006347203128e-06,
      "loss": 0.8677,
      "step": 2080
    },
    {
      "epoch": 3.6925795053003534,
      "grad_norm": 0.5734356641769409,
      "learning_rate": 6.567688612287625e-06,
      "loss": 0.835,
      "step": 2090
    },
    {
      "epoch": 3.7102473498233217,
      "grad_norm": 0.4921402037143707,
      "learning_rate": 6.480666519257501e-06,
      "loss": 0.8554,
      "step": 2100
    },
    {
      "epoch": 3.7279151943462896,
      "grad_norm": 0.5890780687332153,
      "learning_rate": 6.393947563754253e-06,
      "loss": 0.7895,
      "step": 2110
    },
    {
      "epoch": 3.745583038869258,
      "grad_norm": 0.7719208598136902,
      "learning_rate": 6.307539215308644e-06,
      "loss": 0.8313,
      "step": 2120
    },
    {
      "epoch": 3.7632508833922262,
      "grad_norm": 0.5820373296737671,
      "learning_rate": 6.2214489166973235e-06,
      "loss": 0.8796,
      "step": 2130
    },
    {
      "epoch": 3.7809187279151946,
      "grad_norm": 0.5791803598403931,
      "learning_rate": 6.135684083301738e-06,
      "loss": 0.8228,
      "step": 2140
    },
    {
      "epoch": 3.7985865724381624,
      "grad_norm": 0.4887305498123169,
      "learning_rate": 6.050252102469417e-06,
      "loss": 0.8304,
      "step": 2150
    },
    {
      "epoch": 3.8162544169611308,
      "grad_norm": 0.7655726671218872,
      "learning_rate": 5.9651603328776606e-06,
      "loss": 0.8734,
      "step": 2160
    },
    {
      "epoch": 3.833922261484099,
      "grad_norm": 0.5560609698295593,
      "learning_rate": 5.880416103899696e-06,
      "loss": 0.8274,
      "step": 2170
    },
    {
      "epoch": 3.851590106007067,
      "grad_norm": 0.5723090171813965,
      "learning_rate": 5.796026714973359e-06,
      "loss": 0.8734,
      "step": 2180
    },
    {
      "epoch": 3.8692579505300353,
      "grad_norm": 0.5722220540046692,
      "learning_rate": 5.711999434972378e-06,
      "loss": 0.8711,
      "step": 2190
    },
    {
      "epoch": 3.8869257950530036,
      "grad_norm": 0.48828595876693726,
      "learning_rate": 5.628341501580246e-06,
      "loss": 0.7963,
      "step": 2200
    },
    {
      "epoch": 3.904593639575972,
      "grad_norm": 0.6057942509651184,
      "learning_rate": 5.545060120666812e-06,
      "loss": 0.8188,
      "step": 2210
    },
    {
      "epoch": 3.92226148409894,
      "grad_norm": 0.5612859129905701,
      "learning_rate": 5.462162465667614e-06,
      "loss": 0.8261,
      "step": 2220
    },
    {
      "epoch": 3.939929328621908,
      "grad_norm": 0.5768514275550842,
      "learning_rate": 5.379655676965984e-06,
      "loss": 0.8937,
      "step": 2230
    },
    {
      "epoch": 3.9575971731448765,
      "grad_norm": 0.8001552820205688,
      "learning_rate": 5.297546861278013e-06,
      "loss": 0.8398,
      "step": 2240
    },
    {
      "epoch": 3.9752650176678443,
      "grad_norm": 0.5258967876434326,
      "learning_rate": 5.215843091040409e-06,
      "loss": 0.84,
      "step": 2250
    },
    {
      "epoch": 3.9929328621908127,
      "grad_norm": 0.5343855023384094,
      "learning_rate": 5.134551403801336e-06,
      "loss": 0.8507,
      "step": 2260
    },
    {
      "epoch": 4.010600706713781,
      "grad_norm": 0.5460066199302673,
      "learning_rate": 5.053678801614205e-06,
      "loss": 0.8737,
      "step": 2270
    },
    {
      "epoch": 4.028268551236749,
      "grad_norm": 0.6175852417945862,
      "learning_rate": 4.973232250434579e-06,
      "loss": 0.8511,
      "step": 2280
    },
    {
      "epoch": 4.045936395759718,
      "grad_norm": 0.6113592982292175,
      "learning_rate": 4.893218679520137e-06,
      "loss": 0.8204,
      "step": 2290
    },
    {
      "epoch": 4.063604240282685,
      "grad_norm": 0.9578297138214111,
      "learning_rate": 4.813644980833851e-06,
      "loss": 0.8181,
      "step": 2300
    },
    {
      "epoch": 4.081272084805653,
      "grad_norm": 0.5546936988830566,
      "learning_rate": 4.734518008450312e-06,
      "loss": 0.789,
      "step": 2310
    },
    {
      "epoch": 4.098939929328622,
      "grad_norm": 0.680998682975769,
      "learning_rate": 4.6558445779653946e-06,
      "loss": 0.8605,
      "step": 2320
    },
    {
      "epoch": 4.11660777385159,
      "grad_norm": 0.6179676055908203,
      "learning_rate": 4.57763146590916e-06,
      "loss": 0.8559,
      "step": 2330
    },
    {
      "epoch": 4.134275618374558,
      "grad_norm": 0.6207692623138428,
      "learning_rate": 4.49988540916218e-06,
      "loss": 0.8733,
      "step": 2340
    },
    {
      "epoch": 4.151943462897527,
      "grad_norm": 0.7949970960617065,
      "learning_rate": 4.422613104375259e-06,
      "loss": 0.8521,
      "step": 2350
    },
    {
      "epoch": 4.169611307420495,
      "grad_norm": 0.5593691468238831,
      "learning_rate": 4.345821207392605e-06,
      "loss": 0.8163,
      "step": 2360
    },
    {
      "epoch": 4.1872791519434625,
      "grad_norm": 0.7125366926193237,
      "learning_rate": 4.269516332678529e-06,
      "loss": 0.8531,
      "step": 2370
    },
    {
      "epoch": 4.204946996466431,
      "grad_norm": 0.5829963088035583,
      "learning_rate": 4.193705052747737e-06,
      "loss": 0.8855,
      "step": 2380
    },
    {
      "epoch": 4.222614840989399,
      "grad_norm": 0.6220327615737915,
      "learning_rate": 4.1183938975991644e-06,
      "loss": 0.8152,
      "step": 2390
    },
    {
      "epoch": 4.240282685512367,
      "grad_norm": 0.5507341623306274,
      "learning_rate": 4.043589354153541e-06,
      "loss": 0.805,
      "step": 2400
    },
    {
      "epoch": 4.257950530035336,
      "grad_norm": 0.5985105633735657,
      "learning_rate": 3.969297865694641e-06,
      "loss": 0.8132,
      "step": 2410
    },
    {
      "epoch": 4.275618374558304,
      "grad_norm": 0.8096944093704224,
      "learning_rate": 3.895525831314282e-06,
      "loss": 0.8731,
      "step": 2420
    },
    {
      "epoch": 4.293286219081272,
      "grad_norm": 0.6014131307601929,
      "learning_rate": 3.822279605361138e-06,
      "loss": 0.8497,
      "step": 2430
    },
    {
      "epoch": 4.310954063604241,
      "grad_norm": 0.5453154444694519,
      "learning_rate": 3.74956549689342e-06,
      "loss": 0.8213,
      "step": 2440
    },
    {
      "epoch": 4.328621908127208,
      "grad_norm": 0.4849812090396881,
      "learning_rate": 3.677389769135444e-06,
      "loss": 0.8196,
      "step": 2450
    },
    {
      "epoch": 4.3462897526501765,
      "grad_norm": 0.5745995044708252,
      "learning_rate": 3.6057586389381326e-06,
      "loss": 0.8202,
      "step": 2460
    },
    {
      "epoch": 4.363957597173145,
      "grad_norm": 0.5921375155448914,
      "learning_rate": 3.5346782762435383e-06,
      "loss": 0.8623,
      "step": 2470
    },
    {
      "epoch": 4.381625441696113,
      "grad_norm": 1.4667394161224365,
      "learning_rate": 3.464154803553408e-06,
      "loss": 0.8438,
      "step": 2480
    },
    {
      "epoch": 4.3992932862190814,
      "grad_norm": 0.6135275959968567,
      "learning_rate": 3.394194295401796e-06,
      "loss": 0.886,
      "step": 2490
    },
    {
      "epoch": 4.41696113074205,
      "grad_norm": 0.5192072987556458,
      "learning_rate": 3.3248027778318593e-06,
      "loss": 0.8115,
      "step": 2500
    },
    {
      "epoch": 4.434628975265017,
      "grad_norm": 0.4936811327934265,
      "learning_rate": 3.255986227876782e-06,
      "loss": 0.8178,
      "step": 2510
    },
    {
      "epoch": 4.4522968197879855,
      "grad_norm": 0.5864696502685547,
      "learning_rate": 3.1877505730449677e-06,
      "loss": 0.8582,
      "step": 2520
    },
    {
      "epoch": 4.469964664310954,
      "grad_norm": 0.5196579694747925,
      "learning_rate": 3.1201016908094518e-06,
      "loss": 0.8544,
      "step": 2530
    },
    {
      "epoch": 4.487632508833922,
      "grad_norm": 0.7537738084793091,
      "learning_rate": 3.0530454081016637e-06,
      "loss": 0.897,
      "step": 2540
    },
    {
      "epoch": 4.5053003533568905,
      "grad_norm": 0.5827525854110718,
      "learning_rate": 2.9865875008095114e-06,
      "loss": 0.8611,
      "step": 2550
    },
    {
      "epoch": 4.522968197879859,
      "grad_norm": 0.6028100848197937,
      "learning_rate": 2.920733693279879e-06,
      "loss": 0.7945,
      "step": 2560
    },
    {
      "epoch": 4.540636042402827,
      "grad_norm": 0.6480982899665833,
      "learning_rate": 2.855489657825573e-06,
      "loss": 0.8389,
      "step": 2570
    },
    {
      "epoch": 4.5583038869257955,
      "grad_norm": 0.6590670347213745,
      "learning_rate": 2.7908610142367144e-06,
      "loss": 0.8061,
      "step": 2580
    },
    {
      "epoch": 4.575971731448763,
      "grad_norm": 0.6050382852554321,
      "learning_rate": 2.7268533292967026e-06,
      "loss": 0.8249,
      "step": 2590
    },
    {
      "epoch": 4.593639575971731,
      "grad_norm": 0.6227623224258423,
      "learning_rate": 2.6634721163027076e-06,
      "loss": 0.8171,
      "step": 2600
    },
    {
      "epoch": 4.6113074204946995,
      "grad_norm": 0.4837830662727356,
      "learning_rate": 2.600722834590781e-06,
      "loss": 0.7877,
      "step": 2610
    },
    {
      "epoch": 4.628975265017668,
      "grad_norm": 0.5412755608558655,
      "learning_rate": 2.538610889065619e-06,
      "loss": 0.8309,
      "step": 2620
    },
    {
      "epoch": 4.646643109540636,
      "grad_norm": 0.5128341317176819,
      "learning_rate": 2.477141629735025e-06,
      "loss": 0.8903,
      "step": 2630
    },
    {
      "epoch": 4.6643109540636045,
      "grad_norm": 0.5534719228744507,
      "learning_rate": 2.416320351249062e-06,
      "loss": 0.8418,
      "step": 2640
    },
    {
      "epoch": 4.681978798586572,
      "grad_norm": 0.6710190773010254,
      "learning_rate": 2.3561522924440127e-06,
      "loss": 0.829,
      "step": 2650
    },
    {
      "epoch": 4.69964664310954,
      "grad_norm": 0.5560805797576904,
      "learning_rate": 2.2966426358911387e-06,
      "loss": 0.8572,
      "step": 2660
    },
    {
      "epoch": 4.717314487632509,
      "grad_norm": 0.5557115077972412,
      "learning_rate": 2.237796507450272e-06,
      "loss": 0.8317,
      "step": 2670
    },
    {
      "epoch": 4.734982332155477,
      "grad_norm": 0.7195934653282166,
      "learning_rate": 2.1796189758282917e-06,
      "loss": 0.832,
      "step": 2680
    },
    {
      "epoch": 4.752650176678445,
      "grad_norm": 0.5557926893234253,
      "learning_rate": 2.122115052142545e-06,
      "loss": 0.7884,
      "step": 2690
    },
    {
      "epoch": 4.770318021201414,
      "grad_norm": 0.5075626373291016,
      "learning_rate": 2.065289689489213e-06,
      "loss": 0.7874,
      "step": 2700
    },
    {
      "epoch": 4.787985865724382,
      "grad_norm": 0.7519310712814331,
      "learning_rate": 2.0091477825166637e-06,
      "loss": 0.8451,
      "step": 2710
    },
    {
      "epoch": 4.80565371024735,
      "grad_norm": 0.5555654764175415,
      "learning_rate": 1.9536941670038745e-06,
      "loss": 0.8339,
      "step": 2720
    },
    {
      "epoch": 4.823321554770318,
      "grad_norm": 0.5968272686004639,
      "learning_rate": 1.8989336194438756e-06,
      "loss": 0.8268,
      "step": 2730
    },
    {
      "epoch": 4.840989399293286,
      "grad_norm": 0.7066864967346191,
      "learning_rate": 1.8448708566323504e-06,
      "loss": 0.8263,
      "step": 2740
    },
    {
      "epoch": 4.858657243816254,
      "grad_norm": 0.46801695227622986,
      "learning_rate": 1.7915105352613382e-06,
      "loss": 0.8091,
      "step": 2750
    },
    {
      "epoch": 4.876325088339223,
      "grad_norm": 0.7215123772621155,
      "learning_rate": 1.7388572515181445e-06,
      "loss": 0.8339,
      "step": 2760
    },
    {
      "epoch": 4.893992932862191,
      "grad_norm": 0.6271361708641052,
      "learning_rate": 1.6869155406894344e-06,
      "loss": 0.8223,
      "step": 2770
    },
    {
      "epoch": 4.911660777385159,
      "grad_norm": 0.9216979742050171,
      "learning_rate": 1.6356898767705954e-06,
      "loss": 0.8052,
      "step": 2780
    },
    {
      "epoch": 4.929328621908128,
      "grad_norm": 0.5909148454666138,
      "learning_rate": 1.585184672080371e-06,
      "loss": 0.8026,
      "step": 2790
    },
    {
      "epoch": 4.946996466431095,
      "grad_norm": 0.5720986723899841,
      "learning_rate": 1.5354042768807976e-06,
      "loss": 0.831,
      "step": 2800
    },
    {
      "epoch": 4.964664310954063,
      "grad_norm": 0.5404096245765686,
      "learning_rate": 1.4863529790025033e-06,
      "loss": 0.8301,
      "step": 2810
    },
    {
      "epoch": 4.982332155477032,
      "grad_norm": 0.5945255756378174,
      "learning_rate": 1.4380350034753766e-06,
      "loss": 0.8786,
      "step": 2820
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.2441853284835815,
      "learning_rate": 1.3904545121646319e-06,
      "loss": 0.8814,
      "step": 2830
    },
    {
      "epoch": 5.017667844522968,
      "grad_norm": 0.655418872833252,
      "learning_rate": 1.3436156034123383e-06,
      "loss": 0.8598,
      "step": 2840
    },
    {
      "epoch": 5.035335689045937,
      "grad_norm": 0.5294393301010132,
      "learning_rate": 1.2975223116844115e-06,
      "loss": 0.831,
      "step": 2850
    },
    {
      "epoch": 5.053003533568905,
      "grad_norm": 0.5677767395973206,
      "learning_rate": 1.2521786072230935e-06,
      "loss": 0.838,
      "step": 2860
    },
    {
      "epoch": 5.070671378091872,
      "grad_norm": 0.6083729267120361,
      "learning_rate": 1.2075883957049862e-06,
      "loss": 0.8866,
      "step": 2870
    },
    {
      "epoch": 5.088339222614841,
      "grad_norm": 0.6290938854217529,
      "learning_rate": 1.1637555179046344e-06,
      "loss": 0.8751,
      "step": 2880
    },
    {
      "epoch": 5.106007067137809,
      "grad_norm": 0.5518789291381836,
      "learning_rate": 1.1206837493636992e-06,
      "loss": 0.8597,
      "step": 2890
    },
    {
      "epoch": 5.123674911660777,
      "grad_norm": 0.5620294809341431,
      "learning_rate": 1.078376800065749e-06,
      "loss": 0.8105,
      "step": 2900
    },
    {
      "epoch": 5.141342756183746,
      "grad_norm": 0.6572288870811462,
      "learning_rate": 1.036838314116706e-06,
      "loss": 0.8057,
      "step": 2910
    },
    {
      "epoch": 5.159010600706714,
      "grad_norm": 0.5638129115104675,
      "learning_rate": 9.960718694309623e-07,
      "loss": 0.8214,
      "step": 2920
    },
    {
      "epoch": 5.176678445229682,
      "grad_norm": 0.7003311514854431,
      "learning_rate": 9.560809774231872e-07,
      "loss": 0.8544,
      "step": 2930
    },
    {
      "epoch": 5.19434628975265,
      "grad_norm": 0.5848519802093506,
      "learning_rate": 9.168690827058813e-07,
      "loss": 0.8273,
      "step": 2940
    },
    {
      "epoch": 5.212014134275618,
      "grad_norm": 0.5222687721252441,
      "learning_rate": 8.784395627926734e-07,
      "loss": 0.8567,
      "step": 2950
    },
    {
      "epoch": 5.229681978798586,
      "grad_norm": 0.5781293511390686,
      "learning_rate": 8.407957278073952e-07,
      "loss": 0.8236,
      "step": 2960
    },
    {
      "epoch": 5.247349823321555,
      "grad_norm": 0.6143786311149597,
      "learning_rate": 8.039408201989618e-07,
      "loss": 0.8415,
      "step": 2970
    },
    {
      "epoch": 5.265017667844523,
      "grad_norm": 0.5558165311813354,
      "learning_rate": 7.678780144620956e-07,
      "loss": 0.7737,
      "step": 2980
    },
    {
      "epoch": 5.282685512367491,
      "grad_norm": 0.6024080514907837,
      "learning_rate": 7.32610416863877e-07,
      "loss": 0.8291,
      "step": 2990
    },
    {
      "epoch": 5.30035335689046,
      "grad_norm": 0.5599005818367004,
      "learning_rate": 6.981410651761933e-07,
      "loss": 0.869,
      "step": 3000
    },
    {
      "epoch": 5.318021201413427,
      "grad_norm": 0.640021562576294,
      "learning_rate": 6.644729284140828e-07,
      "loss": 0.8184,
      "step": 3010
    },
    {
      "epoch": 5.3356890459363955,
      "grad_norm": 0.5697632431983948,
      "learning_rate": 6.316089065799958e-07,
      "loss": 0.841,
      "step": 3020
    },
    {
      "epoch": 5.353356890459364,
      "grad_norm": 0.5342921018600464,
      "learning_rate": 5.995518304139991e-07,
      "loss": 0.8244,
      "step": 3030
    },
    {
      "epoch": 5.371024734982332,
      "grad_norm": 0.5991451144218445,
      "learning_rate": 5.6830446114996e-07,
      "loss": 0.8567,
      "step": 3040
    },
    {
      "epoch": 5.3886925795053005,
      "grad_norm": 0.5282171964645386,
      "learning_rate": 5.37869490277697e-07,
      "loss": 0.8427,
      "step": 3050
    },
    {
      "epoch": 5.406360424028269,
      "grad_norm": 0.6702190637588501,
      "learning_rate": 5.082495393111564e-07,
      "loss": 0.817,
      "step": 3060
    },
    {
      "epoch": 5.424028268551237,
      "grad_norm": 0.5918974280357361,
      "learning_rate": 4.794471595626071e-07,
      "loss": 0.8249,
      "step": 3070
    },
    {
      "epoch": 5.4416961130742045,
      "grad_norm": 0.6387214064598083,
      "learning_rate": 4.514648319228798e-07,
      "loss": 0.8695,
      "step": 3080
    },
    {
      "epoch": 5.459363957597173,
      "grad_norm": 0.5860491394996643,
      "learning_rate": 4.243049666476784e-07,
      "loss": 0.7571,
      "step": 3090
    },
    {
      "epoch": 5.477031802120141,
      "grad_norm": 0.64732426404953,
      "learning_rate": 3.9796990314997176e-07,
      "loss": 0.8065,
      "step": 3100
    },
    {
      "epoch": 5.4946996466431095,
      "grad_norm": 0.4854739010334015,
      "learning_rate": 3.7246190979849164e-07,
      "loss": 0.8417,
      "step": 3110
    },
    {
      "epoch": 5.512367491166078,
      "grad_norm": 0.5119847059249878,
      "learning_rate": 3.477831837223433e-07,
      "loss": 0.8222,
      "step": 3120
    },
    {
      "epoch": 5.530035335689046,
      "grad_norm": 1.499524474143982,
      "learning_rate": 3.239358506217549e-07,
      "loss": 0.8278,
      "step": 3130
    },
    {
      "epoch": 5.5477031802120145,
      "grad_norm": 0.6199241280555725,
      "learning_rate": 3.009219645849859e-07,
      "loss": 0.8005,
      "step": 3140
    },
    {
      "epoch": 5.565371024734983,
      "grad_norm": 0.6573869585990906,
      "learning_rate": 2.7874350791139203e-07,
      "loss": 0.8338,
      "step": 3150
    },
    {
      "epoch": 5.58303886925795,
      "grad_norm": 0.5395724773406982,
      "learning_rate": 2.574023909406853e-07,
      "loss": 0.8363,
      "step": 3160
    },
    {
      "epoch": 5.6007067137809186,
      "grad_norm": 0.834705114364624,
      "learning_rate": 2.369004518883855e-07,
      "loss": 0.8207,
      "step": 3170
    },
    {
      "epoch": 5.618374558303887,
      "grad_norm": 0.5336258411407471,
      "learning_rate": 2.1723945668748248e-07,
      "loss": 0.8153,
      "step": 3180
    },
    {
      "epoch": 5.636042402826855,
      "grad_norm": 0.7850385904312134,
      "learning_rate": 1.984210988363311e-07,
      "loss": 0.8114,
      "step": 3190
    },
    {
      "epoch": 5.6537102473498235,
      "grad_norm": 0.597079336643219,
      "learning_rate": 1.8044699925278242e-07,
      "loss": 0.834,
      "step": 3200
    },
    {
      "epoch": 5.671378091872792,
      "grad_norm": 0.6741916537284851,
      "learning_rate": 1.6331870613456423e-07,
      "loss": 0.822,
      "step": 3210
    },
    {
      "epoch": 5.689045936395759,
      "grad_norm": 0.7204360961914062,
      "learning_rate": 1.4703769482592335e-07,
      "loss": 0.8481,
      "step": 3220
    },
    {
      "epoch": 5.706713780918728,
      "grad_norm": 0.6026734113693237,
      "learning_rate": 1.3160536769055708e-07,
      "loss": 0.845,
      "step": 3230
    },
    {
      "epoch": 5.724381625441696,
      "grad_norm": 0.568855345249176,
      "learning_rate": 1.1702305399081349e-07,
      "loss": 0.8179,
      "step": 3240
    },
    {
      "epoch": 5.742049469964664,
      "grad_norm": 0.5504450798034668,
      "learning_rate": 1.0329200977319265e-07,
      "loss": 0.8236,
      "step": 3250
    },
    {
      "epoch": 5.759717314487633,
      "grad_norm": 0.60431307554245,
      "learning_rate": 9.041341776016565e-08,
      "loss": 0.8486,
      "step": 3260
    },
    {
      "epoch": 5.777385159010601,
      "grad_norm": 0.5151515007019043,
      "learning_rate": 7.83883872482949e-08,
      "loss": 0.8344,
      "step": 3270
    },
    {
      "epoch": 5.795053003533569,
      "grad_norm": 0.6697881817817688,
      "learning_rate": 6.721795401268493e-08,
      "loss": 0.8232,
      "step": 3280
    },
    {
      "epoch": 5.8127208480565375,
      "grad_norm": 0.5603944659233093,
      "learning_rate": 5.69030802177728e-08,
      "loss": 0.8249,
      "step": 3290
    },
    {
      "epoch": 5.830388692579505,
      "grad_norm": 0.5996559262275696,
      "learning_rate": 4.744465433443979e-08,
      "loss": 0.8186,
      "step": 3300
    },
    {
      "epoch": 5.848056537102473,
      "grad_norm": 0.5992182493209839,
      "learning_rate": 3.884349106349716e-08,
      "loss": 0.8801,
      "step": 3310
    },
    {
      "epoch": 5.865724381625442,
      "grad_norm": 0.5472813248634338,
      "learning_rate": 3.110033126549894e-08,
      "loss": 0.8299,
      "step": 3320
    },
    {
      "epoch": 5.88339222614841,
      "grad_norm": 0.6012226343154907,
      "learning_rate": 2.4215841896938486e-08,
      "loss": 0.8276,
      "step": 3330
    },
    {
      "epoch": 5.901060070671378,
      "grad_norm": 0.5417101979255676,
      "learning_rate": 1.8190615952794477e-08,
      "loss": 0.8673,
      "step": 3340
    },
    {
      "epoch": 5.918727915194347,
      "grad_norm": 0.5450870990753174,
      "learning_rate": 1.3025172415451758e-08,
      "loss": 0.8526,
      "step": 3350
    },
    {
      "epoch": 5.936395759717314,
      "grad_norm": 0.5599738359451294,
      "learning_rate": 8.719956210007096e-09,
      "loss": 0.7772,
      "step": 3360
    },
    {
      "epoch": 5.954063604240282,
      "grad_norm": 0.5253913998603821,
      "learning_rate": 5.275338165935395e-09,
      "loss": 0.8169,
      "step": 3370
    },
    {
      "epoch": 5.971731448763251,
      "grad_norm": 0.6051644682884216,
      "learning_rate": 2.6916149851563542e-09,
      "loss": 0.8312,
      "step": 3380
    },
    {
      "epoch": 5.989399293286219,
      "grad_norm": 0.6155297160148621,
      "learning_rate": 9.690092164715835e-10,
      "loss": 0.8493,
      "step": 3390
    },
    {
      "epoch": 6.007067137809187,
      "grad_norm": 0.7502905130386353,
      "learning_rate": 1.0766923640215254e-10,
      "loss": 0.7903,
      "step": 3400
    }
  ],
  "logging_steps": 10,
  "max_steps": 3405,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "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": 4.855684396713247e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}