{
  "best_metric": 0.029545707628130913,
  "best_model_checkpoint": "saves/psy-course/MentaLLaMA-chat-7B/train/fold5/checkpoint-1250",
  "epoch": 4.9971563981042655,
  "eval_steps": 50,
  "global_step": 3295,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.015165876777251185,
      "grad_norm": 2.013413429260254,
      "learning_rate": 3.0303030303030305e-06,
      "loss": 1.6489,
      "step": 10
    },
    {
      "epoch": 0.03033175355450237,
      "grad_norm": 1.7110595703125,
      "learning_rate": 6.060606060606061e-06,
      "loss": 1.6363,
      "step": 20
    },
    {
      "epoch": 0.04549763033175355,
      "grad_norm": 2.262253761291504,
      "learning_rate": 9.090909090909091e-06,
      "loss": 1.6613,
      "step": 30
    },
    {
      "epoch": 0.06066350710900474,
      "grad_norm": 3.292367696762085,
      "learning_rate": 1.2121212121212122e-05,
      "loss": 1.2702,
      "step": 40
    },
    {
      "epoch": 0.07582938388625593,
      "grad_norm": 0.9315681457519531,
      "learning_rate": 1.5151515151515153e-05,
      "loss": 0.8836,
      "step": 50
    },
    {
      "epoch": 0.07582938388625593,
      "eval_loss": 0.6509825587272644,
      "eval_runtime": 177.0,
      "eval_samples_per_second": 6.627,
      "eval_steps_per_second": 6.627,
      "step": 50
    },
    {
      "epoch": 0.0909952606635071,
      "grad_norm": 1.0166341066360474,
      "learning_rate": 1.8181818181818182e-05,
      "loss": 0.7411,
      "step": 60
    },
    {
      "epoch": 0.1061611374407583,
      "grad_norm": 1.0003480911254883,
      "learning_rate": 2.1212121212121215e-05,
      "loss": 0.4577,
      "step": 70
    },
    {
      "epoch": 0.12132701421800948,
      "grad_norm": 0.6172419190406799,
      "learning_rate": 2.4242424242424244e-05,
      "loss": 0.338,
      "step": 80
    },
    {
      "epoch": 0.13649289099526066,
      "grad_norm": 0.6985596418380737,
      "learning_rate": 2.7272727272727273e-05,
      "loss": 0.1643,
      "step": 90
    },
    {
      "epoch": 0.15165876777251186,
      "grad_norm": 0.6572972536087036,
      "learning_rate": 3.0303030303030306e-05,
      "loss": 0.1276,
      "step": 100
    },
    {
      "epoch": 0.15165876777251186,
      "eval_loss": 0.11501473188400269,
      "eval_runtime": 176.988,
      "eval_samples_per_second": 6.628,
      "eval_steps_per_second": 6.628,
      "step": 100
    },
    {
      "epoch": 0.16682464454976303,
      "grad_norm": 0.4842168986797333,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.1344,
      "step": 110
    },
    {
      "epoch": 0.1819905213270142,
      "grad_norm": 0.6832391023635864,
      "learning_rate": 3.6363636363636364e-05,
      "loss": 0.0992,
      "step": 120
    },
    {
      "epoch": 0.1971563981042654,
      "grad_norm": 0.7095351815223694,
      "learning_rate": 3.939393939393939e-05,
      "loss": 0.0818,
      "step": 130
    },
    {
      "epoch": 0.2123222748815166,
      "grad_norm": 0.7936357855796814,
      "learning_rate": 4.242424242424243e-05,
      "loss": 0.0765,
      "step": 140
    },
    {
      "epoch": 0.22748815165876776,
      "grad_norm": 0.8350841999053955,
      "learning_rate": 4.545454545454546e-05,
      "loss": 0.0848,
      "step": 150
    },
    {
      "epoch": 0.22748815165876776,
      "eval_loss": 0.07310865074396133,
      "eval_runtime": 177.0391,
      "eval_samples_per_second": 6.626,
      "eval_steps_per_second": 6.626,
      "step": 150
    },
    {
      "epoch": 0.24265402843601896,
      "grad_norm": 0.9234074950218201,
      "learning_rate": 4.848484848484849e-05,
      "loss": 0.084,
      "step": 160
    },
    {
      "epoch": 0.25781990521327014,
      "grad_norm": 0.9987066984176636,
      "learning_rate": 5.151515151515152e-05,
      "loss": 0.0789,
      "step": 170
    },
    {
      "epoch": 0.2729857819905213,
      "grad_norm": 0.5734285712242126,
      "learning_rate": 5.4545454545454546e-05,
      "loss": 0.0655,
      "step": 180
    },
    {
      "epoch": 0.2881516587677725,
      "grad_norm": 0.5533661246299744,
      "learning_rate": 5.757575757575758e-05,
      "loss": 0.0537,
      "step": 190
    },
    {
      "epoch": 0.3033175355450237,
      "grad_norm": 0.674788236618042,
      "learning_rate": 6.060606060606061e-05,
      "loss": 0.0545,
      "step": 200
    },
    {
      "epoch": 0.3033175355450237,
      "eval_loss": 0.0569460466504097,
      "eval_runtime": 177.0841,
      "eval_samples_per_second": 6.624,
      "eval_steps_per_second": 6.624,
      "step": 200
    },
    {
      "epoch": 0.3184834123222749,
      "grad_norm": 0.6172575354576111,
      "learning_rate": 6.363636363636364e-05,
      "loss": 0.0637,
      "step": 210
    },
    {
      "epoch": 0.33364928909952607,
      "grad_norm": 0.2991991341114044,
      "learning_rate": 6.666666666666667e-05,
      "loss": 0.0772,
      "step": 220
    },
    {
      "epoch": 0.34881516587677724,
      "grad_norm": 0.7083033919334412,
      "learning_rate": 6.96969696969697e-05,
      "loss": 0.0482,
      "step": 230
    },
    {
      "epoch": 0.3639810426540284,
      "grad_norm": 0.3638122081756592,
      "learning_rate": 7.272727272727273e-05,
      "loss": 0.0454,
      "step": 240
    },
    {
      "epoch": 0.3791469194312796,
      "grad_norm": 0.5038028955459595,
      "learning_rate": 7.575757575757576e-05,
      "loss": 0.0542,
      "step": 250
    },
    {
      "epoch": 0.3791469194312796,
      "eval_loss": 0.04985937476158142,
      "eval_runtime": 177.0515,
      "eval_samples_per_second": 6.625,
      "eval_steps_per_second": 6.625,
      "step": 250
    },
    {
      "epoch": 0.3943127962085308,
      "grad_norm": 0.8618497252464294,
      "learning_rate": 7.878787878787879e-05,
      "loss": 0.0507,
      "step": 260
    },
    {
      "epoch": 0.409478672985782,
      "grad_norm": 0.7346657514572144,
      "learning_rate": 8.181818181818183e-05,
      "loss": 0.0433,
      "step": 270
    },
    {
      "epoch": 0.4246445497630332,
      "grad_norm": 0.42429229617118835,
      "learning_rate": 8.484848484848486e-05,
      "loss": 0.0452,
      "step": 280
    },
    {
      "epoch": 0.43981042654028435,
      "grad_norm": 0.8676588535308838,
      "learning_rate": 8.787878787878789e-05,
      "loss": 0.0555,
      "step": 290
    },
    {
      "epoch": 0.4549763033175355,
      "grad_norm": 0.42583391070365906,
      "learning_rate": 9.090909090909092e-05,
      "loss": 0.0466,
      "step": 300
    },
    {
      "epoch": 0.4549763033175355,
      "eval_loss": 0.05100926011800766,
      "eval_runtime": 176.973,
      "eval_samples_per_second": 6.628,
      "eval_steps_per_second": 6.628,
      "step": 300
    },
    {
      "epoch": 0.47014218009478675,
      "grad_norm": 0.2199127972126007,
      "learning_rate": 9.393939393939395e-05,
      "loss": 0.0506,
      "step": 310
    },
    {
      "epoch": 0.48530805687203793,
      "grad_norm": 0.45029717683792114,
      "learning_rate": 9.696969696969698e-05,
      "loss": 0.0408,
      "step": 320
    },
    {
      "epoch": 0.500473933649289,
      "grad_norm": 0.32426029443740845,
      "learning_rate": 0.0001,
      "loss": 0.0497,
      "step": 330
    },
    {
      "epoch": 0.5156398104265403,
      "grad_norm": 0.28149813413619995,
      "learning_rate": 9.999719336268101e-05,
      "loss": 0.0544,
      "step": 340
    },
    {
      "epoch": 0.5308056872037915,
      "grad_norm": 0.6782774329185486,
      "learning_rate": 9.998877376581251e-05,
      "loss": 0.0517,
      "step": 350
    },
    {
      "epoch": 0.5308056872037915,
      "eval_loss": 0.04681693762540817,
      "eval_runtime": 176.9687,
      "eval_samples_per_second": 6.628,
      "eval_steps_per_second": 6.628,
      "step": 350
    },
    {
      "epoch": 0.5459715639810426,
      "grad_norm": 0.5462453961372375,
      "learning_rate": 9.997474215462472e-05,
      "loss": 0.0539,
      "step": 360
    },
    {
      "epoch": 0.5611374407582939,
      "grad_norm": 0.3318583071231842,
      "learning_rate": 9.995510010438337e-05,
      "loss": 0.0322,
      "step": 370
    },
    {
      "epoch": 0.576303317535545,
      "grad_norm": 0.41357553005218506,
      "learning_rate": 9.992984982021295e-05,
      "loss": 0.0644,
      "step": 380
    },
    {
      "epoch": 0.5914691943127962,
      "grad_norm": 0.3177831470966339,
      "learning_rate": 9.9898994136849e-05,
      "loss": 0.0582,
      "step": 390
    },
    {
      "epoch": 0.6066350710900474,
      "grad_norm": 0.44203853607177734,
      "learning_rate": 9.986253651832005e-05,
      "loss": 0.058,
      "step": 400
    },
    {
      "epoch": 0.6066350710900474,
      "eval_loss": 0.04563398286700249,
      "eval_runtime": 176.9795,
      "eval_samples_per_second": 6.628,
      "eval_steps_per_second": 6.628,
      "step": 400
    },
    {
      "epoch": 0.6218009478672986,
      "grad_norm": 0.4002712368965149,
      "learning_rate": 9.982048105755859e-05,
      "loss": 0.0374,
      "step": 410
    },
    {
      "epoch": 0.6369668246445498,
      "grad_norm": 0.3213963806629181,
      "learning_rate": 9.977283247594166e-05,
      "loss": 0.0351,
      "step": 420
    },
    {
      "epoch": 0.6521327014218009,
      "grad_norm": 0.3570813834667206,
      "learning_rate": 9.971959612276076e-05,
      "loss": 0.0564,
      "step": 430
    },
    {
      "epoch": 0.6672985781990521,
      "grad_norm": 0.37261441349983215,
      "learning_rate": 9.966077797462129e-05,
      "loss": 0.0455,
      "step": 440
    },
    {
      "epoch": 0.6824644549763034,
      "grad_norm": 0.647449791431427,
      "learning_rate": 9.959638463477165e-05,
      "loss": 0.0521,
      "step": 450
    },
    {
      "epoch": 0.6824644549763034,
      "eval_loss": 0.04050007835030556,
      "eval_runtime": 177.0646,
      "eval_samples_per_second": 6.625,
      "eval_steps_per_second": 6.625,
      "step": 450
    },
    {
      "epoch": 0.6976303317535545,
      "grad_norm": 0.3081929385662079,
      "learning_rate": 9.952642333236186e-05,
      "loss": 0.045,
      "step": 460
    },
    {
      "epoch": 0.7127962085308057,
      "grad_norm": 0.356259822845459,
      "learning_rate": 9.945090192163202e-05,
      "loss": 0.0407,
      "step": 470
    },
    {
      "epoch": 0.7279620853080568,
      "grad_norm": 0.2743074297904968,
      "learning_rate": 9.936982888103051e-05,
      "loss": 0.0328,
      "step": 480
    },
    {
      "epoch": 0.7431279620853081,
      "grad_norm": 0.2799587547779083,
      "learning_rate": 9.928321331226219e-05,
      "loss": 0.0515,
      "step": 490
    },
    {
      "epoch": 0.7582938388625592,
      "grad_norm": 0.7967923283576965,
      "learning_rate": 9.919106493926655e-05,
      "loss": 0.0317,
      "step": 500
    },
    {
      "epoch": 0.7582938388625592,
      "eval_loss": 0.038231153041124344,
      "eval_runtime": 177.1303,
      "eval_samples_per_second": 6.622,
      "eval_steps_per_second": 6.622,
      "step": 500
    },
    {
      "epoch": 0.7734597156398104,
      "grad_norm": 0.4232497811317444,
      "learning_rate": 9.909339410712612e-05,
      "loss": 0.0453,
      "step": 510
    },
    {
      "epoch": 0.7886255924170616,
      "grad_norm": 0.17073428630828857,
      "learning_rate": 9.8990211780905e-05,
      "loss": 0.0286,
      "step": 520
    },
    {
      "epoch": 0.8037914691943128,
      "grad_norm": 0.5338394045829773,
      "learning_rate": 9.888152954441785e-05,
      "loss": 0.0489,
      "step": 530
    },
    {
      "epoch": 0.818957345971564,
      "grad_norm": 0.2233201563358307,
      "learning_rate": 9.876735959892953e-05,
      "loss": 0.0411,
      "step": 540
    },
    {
      "epoch": 0.8341232227488151,
      "grad_norm": 0.25685688853263855,
      "learning_rate": 9.864771476178522e-05,
      "loss": 0.0281,
      "step": 550
    },
    {
      "epoch": 0.8341232227488151,
      "eval_loss": 0.038973063230514526,
      "eval_runtime": 177.0687,
      "eval_samples_per_second": 6.625,
      "eval_steps_per_second": 6.625,
      "step": 550
    },
    {
      "epoch": 0.8492890995260663,
      "grad_norm": 0.2585074007511139,
      "learning_rate": 9.852260846497153e-05,
      "loss": 0.0357,
      "step": 560
    },
    {
      "epoch": 0.8644549763033176,
      "grad_norm": 0.3039783537387848,
      "learning_rate": 9.839205475360851e-05,
      "loss": 0.0425,
      "step": 570
    },
    {
      "epoch": 0.8796208530805687,
      "grad_norm": 0.34413889050483704,
      "learning_rate": 9.825606828437291e-05,
      "loss": 0.0428,
      "step": 580
    },
    {
      "epoch": 0.8947867298578199,
      "grad_norm": 0.23396067321300507,
      "learning_rate": 9.811466432385267e-05,
      "loss": 0.0472,
      "step": 590
    },
    {
      "epoch": 0.909952606635071,
      "grad_norm": 0.15519942343235016,
      "learning_rate": 9.796785874683314e-05,
      "loss": 0.0388,
      "step": 600
    },
    {
      "epoch": 0.909952606635071,
      "eval_loss": 0.03878552094101906,
      "eval_runtime": 177.1078,
      "eval_samples_per_second": 6.623,
      "eval_steps_per_second": 6.623,
      "step": 600
    },
    {
      "epoch": 0.9251184834123223,
      "grad_norm": 0.16990838944911957,
      "learning_rate": 9.781566803451475e-05,
      "loss": 0.0319,
      "step": 610
    },
    {
      "epoch": 0.9402843601895735,
      "grad_norm": 0.21462449431419373,
      "learning_rate": 9.765810927266281e-05,
      "loss": 0.0288,
      "step": 620
    },
    {
      "epoch": 0.9554502369668246,
      "grad_norm": 0.37621375918388367,
      "learning_rate": 9.749520014968934e-05,
      "loss": 0.0382,
      "step": 630
    },
    {
      "epoch": 0.9706161137440759,
      "grad_norm": 0.15480560064315796,
      "learning_rate": 9.732695895466735e-05,
      "loss": 0.0362,
      "step": 640
    },
    {
      "epoch": 0.985781990521327,
      "grad_norm": 0.2805449068546295,
      "learning_rate": 9.715340457527746e-05,
      "loss": 0.0459,
      "step": 650
    },
    {
      "epoch": 0.985781990521327,
      "eval_loss": 0.03554915264248848,
      "eval_runtime": 177.0579,
      "eval_samples_per_second": 6.625,
      "eval_steps_per_second": 6.625,
      "step": 650
    },
    {
      "epoch": 1.000947867298578,
      "grad_norm": 0.20941686630249023,
      "learning_rate": 9.697455649568761e-05,
      "loss": 0.0373,
      "step": 660
    },
    {
      "epoch": 1.0161137440758294,
      "grad_norm": 0.2722059190273285,
      "learning_rate": 9.679043479436556e-05,
      "loss": 0.0344,
      "step": 670
    },
    {
      "epoch": 1.0312796208530806,
      "grad_norm": 0.2699989080429077,
      "learning_rate": 9.660106014182489e-05,
      "loss": 0.0325,
      "step": 680
    },
    {
      "epoch": 1.0464454976303317,
      "grad_norm": 0.27695226669311523,
      "learning_rate": 9.640645379830424e-05,
      "loss": 0.0328,
      "step": 690
    },
    {
      "epoch": 1.061611374407583,
      "grad_norm": 0.16520464420318604,
      "learning_rate": 9.620663761138067e-05,
      "loss": 0.0277,
      "step": 700
    },
    {
      "epoch": 1.061611374407583,
      "eval_loss": 0.036762021481990814,
      "eval_runtime": 177.0707,
      "eval_samples_per_second": 6.624,
      "eval_steps_per_second": 6.624,
      "step": 700
    },
    {
      "epoch": 1.0767772511848341,
      "grad_norm": 0.13298264145851135,
      "learning_rate": 9.600163401351688e-05,
      "loss": 0.0281,
      "step": 710
    },
    {
      "epoch": 1.0919431279620853,
      "grad_norm": 0.26888322830200195,
      "learning_rate": 9.579146601954276e-05,
      "loss": 0.0395,
      "step": 720
    },
    {
      "epoch": 1.1071090047393364,
      "grad_norm": 0.23041464388370514,
      "learning_rate": 9.557615722407177e-05,
      "loss": 0.0356,
      "step": 730
    },
    {
      "epoch": 1.1222748815165877,
      "grad_norm": 0.2705337703227997,
      "learning_rate": 9.535573179885191e-05,
      "loss": 0.0362,
      "step": 740
    },
    {
      "epoch": 1.1374407582938388,
      "grad_norm": 0.1421772986650467,
      "learning_rate": 9.513021449005214e-05,
      "loss": 0.0342,
      "step": 750
    },
    {
      "epoch": 1.1374407582938388,
      "eval_loss": 0.036910828202962875,
      "eval_runtime": 177.1757,
      "eval_samples_per_second": 6.621,
      "eval_steps_per_second": 6.621,
      "step": 750
    },
    {
      "epoch": 1.15260663507109,
      "grad_norm": 0.10349958389997482,
      "learning_rate": 9.489963061548428e-05,
      "loss": 0.0262,
      "step": 760
    },
    {
      "epoch": 1.1677725118483413,
      "grad_norm": 0.3369249999523163,
      "learning_rate": 9.466400606176062e-05,
      "loss": 0.0335,
      "step": 770
    },
    {
      "epoch": 1.1829383886255924,
      "grad_norm": 0.28793764114379883,
      "learning_rate": 9.442336728138779e-05,
      "loss": 0.0424,
      "step": 780
    },
    {
      "epoch": 1.1981042654028435,
      "grad_norm": 0.30771031975746155,
      "learning_rate": 9.417774128979706e-05,
      "loss": 0.0344,
      "step": 790
    },
    {
      "epoch": 1.2132701421800949,
      "grad_norm": 0.20611123740673065,
      "learning_rate": 9.39271556623114e-05,
      "loss": 0.0323,
      "step": 800
    },
    {
      "epoch": 1.2132701421800949,
      "eval_loss": 0.033734481781721115,
      "eval_runtime": 177.1941,
      "eval_samples_per_second": 6.62,
      "eval_steps_per_second": 6.62,
      "step": 800
    },
    {
      "epoch": 1.228436018957346,
      "grad_norm": 0.20823083817958832,
      "learning_rate": 9.367163853104975e-05,
      "loss": 0.0301,
      "step": 810
    },
    {
      "epoch": 1.2436018957345971,
      "grad_norm": 0.18224073946475983,
      "learning_rate": 9.341121858176876e-05,
      "loss": 0.0327,
      "step": 820
    },
    {
      "epoch": 1.2587677725118485,
      "grad_norm": 0.2314557433128357,
      "learning_rate": 9.314592505064239e-05,
      "loss": 0.0292,
      "step": 830
    },
    {
      "epoch": 1.2739336492890996,
      "grad_norm": 0.20037536323070526,
      "learning_rate": 9.28757877209796e-05,
      "loss": 0.0249,
      "step": 840
    },
    {
      "epoch": 1.2890995260663507,
      "grad_norm": 0.26872518658638,
      "learning_rate": 9.260083691988084e-05,
      "loss": 0.0257,
      "step": 850
    },
    {
      "epoch": 1.2890995260663507,
      "eval_loss": 0.03507881239056587,
      "eval_runtime": 177.211,
      "eval_samples_per_second": 6.619,
      "eval_steps_per_second": 6.619,
      "step": 850
    },
    {
      "epoch": 1.3042654028436018,
      "grad_norm": 0.2439725548028946,
      "learning_rate": 9.232110351483327e-05,
      "loss": 0.0407,
      "step": 860
    },
    {
      "epoch": 1.3194312796208532,
      "grad_norm": 0.5988721251487732,
      "learning_rate": 9.203661891024547e-05,
      "loss": 0.0349,
      "step": 870
    },
    {
      "epoch": 1.3345971563981043,
      "grad_norm": 0.1617327779531479,
      "learning_rate": 9.174741504392173e-05,
      "loss": 0.029,
      "step": 880
    },
    {
      "epoch": 1.3497630331753554,
      "grad_norm": 0.11666741967201233,
      "learning_rate": 9.145352438347662e-05,
      "loss": 0.0251,
      "step": 890
    },
    {
      "epoch": 1.3649289099526065,
      "grad_norm": 0.12099427729845047,
      "learning_rate": 9.115497992268995e-05,
      "loss": 0.0218,
      "step": 900
    },
    {
      "epoch": 1.3649289099526065,
      "eval_loss": 0.034569546580314636,
      "eval_runtime": 177.2771,
      "eval_samples_per_second": 6.617,
      "eval_steps_per_second": 6.617,
      "step": 900
    },
    {
      "epoch": 1.3800947867298579,
      "grad_norm": 0.24614952504634857,
      "learning_rate": 9.085181517780273e-05,
      "loss": 0.0316,
      "step": 910
    },
    {
      "epoch": 1.395260663507109,
      "grad_norm": 0.2466830164194107,
      "learning_rate": 9.054406418375443e-05,
      "loss": 0.0406,
      "step": 920
    },
    {
      "epoch": 1.41042654028436,
      "grad_norm": 0.21008533239364624,
      "learning_rate": 9.023176149036203e-05,
      "loss": 0.0282,
      "step": 930
    },
    {
      "epoch": 1.4255924170616114,
      "grad_norm": 0.32955440878868103,
      "learning_rate": 8.991494215844132e-05,
      "loss": 0.0307,
      "step": 940
    },
    {
      "epoch": 1.4407582938388626,
      "grad_norm": 0.2300134152173996,
      "learning_rate": 8.959364175587069e-05,
      "loss": 0.0266,
      "step": 950
    },
    {
      "epoch": 1.4407582938388626,
      "eval_loss": 0.03766226768493652,
      "eval_runtime": 177.2737,
      "eval_samples_per_second": 6.617,
      "eval_steps_per_second": 6.617,
      "step": 950
    },
    {
      "epoch": 1.4559241706161137,
      "grad_norm": 0.2171100676059723,
      "learning_rate": 8.926789635359817e-05,
      "loss": 0.0327,
      "step": 960
    },
    {
      "epoch": 1.471090047393365,
      "grad_norm": 0.2776648998260498,
      "learning_rate": 8.893774252159187e-05,
      "loss": 0.0272,
      "step": 970
    },
    {
      "epoch": 1.4862559241706161,
      "grad_norm": 0.3400368392467499,
      "learning_rate": 8.860321732473439e-05,
      "loss": 0.0263,
      "step": 980
    },
    {
      "epoch": 1.5014218009478673,
      "grad_norm": 0.22292691469192505,
      "learning_rate": 8.826435831866184e-05,
      "loss": 0.0293,
      "step": 990
    },
    {
      "epoch": 1.5165876777251186,
      "grad_norm": 0.10112936049699783,
      "learning_rate": 8.79212035455475e-05,
      "loss": 0.0344,
      "step": 1000
    },
    {
      "epoch": 1.5165876777251186,
      "eval_loss": 0.03224053233861923,
      "eval_runtime": 177.3774,
      "eval_samples_per_second": 6.613,
      "eval_steps_per_second": 6.613,
      "step": 1000
    },
    {
      "epoch": 1.5317535545023697,
      "grad_norm": 0.18599554896354675,
      "learning_rate": 8.757379152983103e-05,
      "loss": 0.0254,
      "step": 1010
    },
    {
      "epoch": 1.5469194312796208,
      "grad_norm": 0.32951050996780396,
      "learning_rate": 8.722216127389363e-05,
      "loss": 0.0225,
      "step": 1020
    },
    {
      "epoch": 1.5620853080568722,
      "grad_norm": 0.22721748054027557,
      "learning_rate": 8.686635225367919e-05,
      "loss": 0.0191,
      "step": 1030
    },
    {
      "epoch": 1.577251184834123,
      "grad_norm": 0.30656304955482483,
      "learning_rate": 8.650640441426274e-05,
      "loss": 0.0393,
      "step": 1040
    },
    {
      "epoch": 1.5924170616113744,
      "grad_norm": 0.07620204240083694,
      "learning_rate": 8.614235816536582e-05,
      "loss": 0.0244,
      "step": 1050
    },
    {
      "epoch": 1.5924170616113744,
      "eval_loss": 0.03148987516760826,
      "eval_runtime": 177.3647,
      "eval_samples_per_second": 6.613,
      "eval_steps_per_second": 6.613,
      "step": 1050
    },
    {
      "epoch": 1.6075829383886255,
      "grad_norm": 0.1325739324092865,
      "learning_rate": 8.577425437681994e-05,
      "loss": 0.0298,
      "step": 1060
    },
    {
      "epoch": 1.6227488151658767,
      "grad_norm": 0.2656315863132477,
      "learning_rate": 8.540213437397833e-05,
      "loss": 0.0319,
      "step": 1070
    },
    {
      "epoch": 1.637914691943128,
      "grad_norm": 0.1368637979030609,
      "learning_rate": 8.502603993307647e-05,
      "loss": 0.0302,
      "step": 1080
    },
    {
      "epoch": 1.6530805687203791,
      "grad_norm": 0.267149418592453,
      "learning_rate": 8.464601327654207e-05,
      "loss": 0.0308,
      "step": 1090
    },
    {
      "epoch": 1.6682464454976302,
      "grad_norm": 0.09801731258630753,
      "learning_rate": 8.4262097068255e-05,
      "loss": 0.0227,
      "step": 1100
    },
    {
      "epoch": 1.6682464454976302,
      "eval_loss": 0.03317200392484665,
      "eval_runtime": 177.3004,
      "eval_samples_per_second": 6.616,
      "eval_steps_per_second": 6.616,
      "step": 1100
    },
    {
      "epoch": 1.6834123222748816,
      "grad_norm": 0.202189639210701,
      "learning_rate": 8.387433440875758e-05,
      "loss": 0.0312,
      "step": 1110
    },
    {
      "epoch": 1.6985781990521327,
      "grad_norm": 0.12905538082122803,
      "learning_rate": 8.348276883041583e-05,
      "loss": 0.0271,
      "step": 1120
    },
    {
      "epoch": 1.7137440758293838,
      "grad_norm": 0.18460094928741455,
      "learning_rate": 8.308744429253238e-05,
      "loss": 0.0272,
      "step": 1130
    },
    {
      "epoch": 1.7289099526066352,
      "grad_norm": 0.1906091421842575,
      "learning_rate": 8.268840517641123e-05,
      "loss": 0.0233,
      "step": 1140
    },
    {
      "epoch": 1.7440758293838863,
      "grad_norm": 0.21823492646217346,
      "learning_rate": 8.228569628037543e-05,
      "loss": 0.0243,
      "step": 1150
    },
    {
      "epoch": 1.7440758293838863,
      "eval_loss": 0.03176503628492355,
      "eval_runtime": 177.2561,
      "eval_samples_per_second": 6.618,
      "eval_steps_per_second": 6.618,
      "step": 1150
    },
    {
      "epoch": 1.7592417061611374,
      "grad_norm": 0.22984248399734497,
      "learning_rate": 8.187936281473758e-05,
      "loss": 0.0318,
      "step": 1160
    },
    {
      "epoch": 1.7744075829383887,
      "grad_norm": 0.1078776940703392,
      "learning_rate": 8.146945039672446e-05,
      "loss": 0.0264,
      "step": 1170
    },
    {
      "epoch": 1.7895734597156399,
      "grad_norm": 0.14255861937999725,
      "learning_rate": 8.105600504535567e-05,
      "loss": 0.02,
      "step": 1180
    },
    {
      "epoch": 1.804739336492891,
      "grad_norm": 0.19624319672584534,
      "learning_rate": 8.06390731762773e-05,
      "loss": 0.0362,
      "step": 1190
    },
    {
      "epoch": 1.8199052132701423,
      "grad_norm": 0.2134467512369156,
      "learning_rate": 8.021870159655109e-05,
      "loss": 0.03,
      "step": 1200
    },
    {
      "epoch": 1.8199052132701423,
      "eval_loss": 0.031138954684138298,
      "eval_runtime": 177.2477,
      "eval_samples_per_second": 6.618,
      "eval_steps_per_second": 6.618,
      "step": 1200
    },
    {
      "epoch": 1.8350710900473932,
      "grad_norm": 0.2627793252468109,
      "learning_rate": 7.979493749939955e-05,
      "loss": 0.0242,
      "step": 1210
    },
    {
      "epoch": 1.8502369668246446,
      "grad_norm": 0.16796381771564484,
      "learning_rate": 7.93678284589079e-05,
      "loss": 0.0242,
      "step": 1220
    },
    {
      "epoch": 1.865402843601896,
      "grad_norm": 0.23527124524116516,
      "learning_rate": 7.893742242468301e-05,
      "loss": 0.0262,
      "step": 1230
    },
    {
      "epoch": 1.8805687203791468,
      "grad_norm": 0.27677977085113525,
      "learning_rate": 7.850376771647038e-05,
      "loss": 0.0366,
      "step": 1240
    },
    {
      "epoch": 1.8957345971563981,
      "grad_norm": 0.2223452478647232,
      "learning_rate": 7.806691301872958e-05,
      "loss": 0.0307,
      "step": 1250
    },
    {
      "epoch": 1.8957345971563981,
      "eval_loss": 0.029545707628130913,
      "eval_runtime": 177.2472,
      "eval_samples_per_second": 6.618,
      "eval_steps_per_second": 6.618,
      "step": 1250
    },
    {
      "epoch": 1.9109004739336493,
      "grad_norm": 0.12168317288160324,
      "learning_rate": 7.762690737516846e-05,
      "loss": 0.0328,
      "step": 1260
    },
    {
      "epoch": 1.9260663507109004,
      "grad_norm": 0.44124069809913635,
      "learning_rate": 7.718380018323742e-05,
      "loss": 0.0373,
      "step": 1270
    },
    {
      "epoch": 1.9412322274881517,
      "grad_norm": 0.1568044126033783,
      "learning_rate": 7.673764118858371e-05,
      "loss": 0.0218,
      "step": 1280
    },
    {
      "epoch": 1.9563981042654028,
      "grad_norm": 0.3564736545085907,
      "learning_rate": 7.628848047946675e-05,
      "loss": 0.0376,
      "step": 1290
    },
    {
      "epoch": 1.971563981042654,
      "grad_norm": 0.3387766480445862,
      "learning_rate": 7.583636848113483e-05,
      "loss": 0.0344,
      "step": 1300
    },
    {
      "epoch": 1.971563981042654,
      "eval_loss": 0.030506150797009468,
      "eval_runtime": 177.2118,
      "eval_samples_per_second": 6.619,
      "eval_steps_per_second": 6.619,
      "step": 1300
    },
    {
      "epoch": 1.9867298578199053,
      "grad_norm": 0.24238063395023346,
      "learning_rate": 7.538135595016423e-05,
      "loss": 0.03,
      "step": 1310
    },
    {
      "epoch": 2.001895734597156,
      "grad_norm": 0.11398116499185562,
      "learning_rate": 7.492349396876096e-05,
      "loss": 0.0356,
      "step": 1320
    },
    {
      "epoch": 2.0170616113744075,
      "grad_norm": 0.08749754726886749,
      "learning_rate": 7.4462833939026e-05,
      "loss": 0.0198,
      "step": 1330
    },
    {
      "epoch": 2.032227488151659,
      "grad_norm": 0.06967737525701523,
      "learning_rate": 7.399942757718455e-05,
      "loss": 0.0196,
      "step": 1340
    },
    {
      "epoch": 2.0473933649289098,
      "grad_norm": 0.08232049643993378,
      "learning_rate": 7.35333269077802e-05,
      "loss": 0.0214,
      "step": 1350
    },
    {
      "epoch": 2.0473933649289098,
      "eval_loss": 0.030749354511499405,
      "eval_runtime": 176.5322,
      "eval_samples_per_second": 6.645,
      "eval_steps_per_second": 6.645,
      "step": 1350
    },
    {
      "epoch": 2.062559241706161,
      "grad_norm": 0.18698251247406006,
      "learning_rate": 7.306458425783426e-05,
      "loss": 0.0174,
      "step": 1360
    },
    {
      "epoch": 2.0777251184834125,
      "grad_norm": 0.16593368351459503,
      "learning_rate": 7.25932522509713e-05,
      "loss": 0.0177,
      "step": 1370
    },
    {
      "epoch": 2.0928909952606634,
      "grad_norm": 0.23132416605949402,
      "learning_rate": 7.211938380151133e-05,
      "loss": 0.0144,
      "step": 1380
    },
    {
      "epoch": 2.1080568720379147,
      "grad_norm": 0.1467525064945221,
      "learning_rate": 7.164303210852934e-05,
      "loss": 0.0227,
      "step": 1390
    },
    {
      "epoch": 2.123222748815166,
      "grad_norm": 0.19721190631389618,
      "learning_rate": 7.116425064988286e-05,
      "loss": 0.0178,
      "step": 1400
    },
    {
      "epoch": 2.123222748815166,
      "eval_loss": 0.031973257660865784,
      "eval_runtime": 176.3775,
      "eval_samples_per_second": 6.651,
      "eval_steps_per_second": 6.651,
      "step": 1400
    },
    {
      "epoch": 2.138388625592417,
      "grad_norm": 0.21327702701091766,
      "learning_rate": 7.068309317620827e-05,
      "loss": 0.0186,
      "step": 1410
    },
    {
      "epoch": 2.1535545023696683,
      "grad_norm": 0.19219304621219635,
      "learning_rate": 7.019961370488645e-05,
      "loss": 0.0174,
      "step": 1420
    },
    {
      "epoch": 2.1687203791469196,
      "grad_norm": 0.218804731965065,
      "learning_rate": 6.971386651397849e-05,
      "loss": 0.0189,
      "step": 1430
    },
    {
      "epoch": 2.1838862559241705,
      "grad_norm": 0.4009787142276764,
      "learning_rate": 6.922590613613211e-05,
      "loss": 0.0178,
      "step": 1440
    },
    {
      "epoch": 2.199052132701422,
      "grad_norm": 0.060959890484809875,
      "learning_rate": 6.873578735245961e-05,
      "loss": 0.0167,
      "step": 1450
    },
    {
      "epoch": 2.199052132701422,
      "eval_loss": 0.03211337700486183,
      "eval_runtime": 176.3627,
      "eval_samples_per_second": 6.651,
      "eval_steps_per_second": 6.651,
      "step": 1450
    },
    {
      "epoch": 2.2142180094786728,
      "grad_norm": 0.1430092453956604,
      "learning_rate": 6.824356518638775e-05,
      "loss": 0.017,
      "step": 1460
    },
    {
      "epoch": 2.229383886255924,
      "grad_norm": 0.18788497149944305,
      "learning_rate": 6.774929489748052e-05,
      "loss": 0.0229,
      "step": 1470
    },
    {
      "epoch": 2.2445497630331754,
      "grad_norm": 0.1279352605342865,
      "learning_rate": 6.725303197523548e-05,
      "loss": 0.0195,
      "step": 1480
    },
    {
      "epoch": 2.259715639810427,
      "grad_norm": 0.22916962206363678,
      "learning_rate": 6.675483213285412e-05,
      "loss": 0.0194,
      "step": 1490
    },
    {
      "epoch": 2.2748815165876777,
      "grad_norm": 0.11649148911237717,
      "learning_rate": 6.625475130098728e-05,
      "loss": 0.0115,
      "step": 1500
    },
    {
      "epoch": 2.2748815165876777,
      "eval_loss": 0.032478559762239456,
      "eval_runtime": 176.381,
      "eval_samples_per_second": 6.65,
      "eval_steps_per_second": 6.65,
      "step": 1500
    },
    {
      "epoch": 2.290047393364929,
      "grad_norm": 0.12490870803594589,
      "learning_rate": 6.575284562145593e-05,
      "loss": 0.0246,
      "step": 1510
    },
    {
      "epoch": 2.30521327014218,
      "grad_norm": 0.3091743290424347,
      "learning_rate": 6.524917144094851e-05,
      "loss": 0.0168,
      "step": 1520
    },
    {
      "epoch": 2.3203791469194313,
      "grad_norm": 0.18049876391887665,
      "learning_rate": 6.474378530469509e-05,
      "loss": 0.0238,
      "step": 1530
    },
    {
      "epoch": 2.3355450236966826,
      "grad_norm": 0.08443621546030045,
      "learning_rate": 6.42367439501193e-05,
      "loss": 0.0169,
      "step": 1540
    },
    {
      "epoch": 2.3507109004739335,
      "grad_norm": 0.2367069274187088,
      "learning_rate": 6.372810430046862e-05,
      "loss": 0.0192,
      "step": 1550
    },
    {
      "epoch": 2.3507109004739335,
      "eval_loss": 0.03184814006090164,
      "eval_runtime": 176.4492,
      "eval_samples_per_second": 6.648,
      "eval_steps_per_second": 6.648,
      "step": 1550
    },
    {
      "epoch": 2.365876777251185,
      "grad_norm": 0.1728668063879013,
      "learning_rate": 6.321792345842402e-05,
      "loss": 0.0151,
      "step": 1560
    },
    {
      "epoch": 2.381042654028436,
      "grad_norm": 0.3039005994796753,
      "learning_rate": 6.270625869968906e-05,
      "loss": 0.0154,
      "step": 1570
    },
    {
      "epoch": 2.396208530805687,
      "grad_norm": 0.16306930780410767,
      "learning_rate": 6.219316746656007e-05,
      "loss": 0.0175,
      "step": 1580
    },
    {
      "epoch": 2.4113744075829384,
      "grad_norm": 0.09183406829833984,
      "learning_rate": 6.167870736147713e-05,
      "loss": 0.0218,
      "step": 1590
    },
    {
      "epoch": 2.4265402843601898,
      "grad_norm": 0.3515900671482086,
      "learning_rate": 6.116293614055744e-05,
      "loss": 0.0233,
      "step": 1600
    },
    {
      "epoch": 2.4265402843601898,
      "eval_loss": 0.032726142555475235,
      "eval_runtime": 176.5984,
      "eval_samples_per_second": 6.642,
      "eval_steps_per_second": 6.642,
      "step": 1600
    },
    {
      "epoch": 2.4417061611374407,
      "grad_norm": 0.1969446986913681,
      "learning_rate": 6.06459117071113e-05,
      "loss": 0.0163,
      "step": 1610
    },
    {
      "epoch": 2.456872037914692,
      "grad_norm": 0.10922811925411224,
      "learning_rate": 6.012769210514146e-05,
      "loss": 0.0209,
      "step": 1620
    },
    {
      "epoch": 2.472037914691943,
      "grad_norm": 0.13068976998329163,
      "learning_rate": 5.9608335512826915e-05,
      "loss": 0.0153,
      "step": 1630
    },
    {
      "epoch": 2.4872037914691942,
      "grad_norm": 0.1942068338394165,
      "learning_rate": 5.908790023599144e-05,
      "loss": 0.0186,
      "step": 1640
    },
    {
      "epoch": 2.5023696682464456,
      "grad_norm": 0.20405402779579163,
      "learning_rate": 5.856644470155781e-05,
      "loss": 0.0108,
      "step": 1650
    },
    {
      "epoch": 2.5023696682464456,
      "eval_loss": 0.03395625576376915,
      "eval_runtime": 176.6011,
      "eval_samples_per_second": 6.642,
      "eval_steps_per_second": 6.642,
      "step": 1650
    },
    {
      "epoch": 2.517535545023697,
      "grad_norm": 0.1378117799758911,
      "learning_rate": 5.8044027450988546e-05,
      "loss": 0.0224,
      "step": 1660
    },
    {
      "epoch": 2.532701421800948,
      "grad_norm": 0.4658917486667633,
      "learning_rate": 5.752070713371371e-05,
      "loss": 0.0156,
      "step": 1670
    },
    {
      "epoch": 2.547867298578199,
      "grad_norm": 0.21184688806533813,
      "learning_rate": 5.699654250054662e-05,
      "loss": 0.0276,
      "step": 1680
    },
    {
      "epoch": 2.56303317535545,
      "grad_norm": 0.3075418174266815,
      "learning_rate": 5.647159239708809e-05,
      "loss": 0.0264,
      "step": 1690
    },
    {
      "epoch": 2.5781990521327014,
      "grad_norm": 0.24572160840034485,
      "learning_rate": 5.5945915757120146e-05,
      "loss": 0.0256,
      "step": 1700
    },
    {
      "epoch": 2.5781990521327014,
      "eval_loss": 0.0314929336309433,
      "eval_runtime": 176.6685,
      "eval_samples_per_second": 6.64,
      "eval_steps_per_second": 6.64,
      "step": 1700
    },
    {
      "epoch": 2.5933649289099527,
      "grad_norm": 0.30421316623687744,
      "learning_rate": 5.5419571595989825e-05,
      "loss": 0.0266,
      "step": 1710
    },
    {
      "epoch": 2.6085308056872036,
      "grad_norm": 0.1892070770263672,
      "learning_rate": 5.4892619003983734e-05,
      "loss": 0.0177,
      "step": 1720
    },
    {
      "epoch": 2.623696682464455,
      "grad_norm": 0.23157517611980438,
      "learning_rate": 5.436511713969428e-05,
      "loss": 0.0199,
      "step": 1730
    },
    {
      "epoch": 2.6388625592417063,
      "grad_norm": 0.18666303157806396,
      "learning_rate": 5.383712522337817e-05,
      "loss": 0.0221,
      "step": 1740
    },
    {
      "epoch": 2.654028436018957,
      "grad_norm": 0.0699646845459938,
      "learning_rate": 5.3308702530308076e-05,
      "loss": 0.019,
      "step": 1750
    },
    {
      "epoch": 2.654028436018957,
      "eval_loss": 0.030023746192455292,
      "eval_runtime": 176.6593,
      "eval_samples_per_second": 6.64,
      "eval_steps_per_second": 6.64,
      "step": 1750
    },
    {
      "epoch": 2.6691943127962086,
      "grad_norm": 0.16548389196395874,
      "learning_rate": 5.2779908384118025e-05,
      "loss": 0.0166,
      "step": 1760
    },
    {
      "epoch": 2.68436018957346,
      "grad_norm": 0.20772500336170197,
      "learning_rate": 5.22508021501434e-05,
      "loss": 0.0155,
      "step": 1770
    },
    {
      "epoch": 2.699526066350711,
      "grad_norm": 0.22350798547267914,
      "learning_rate": 5.1721443228756284e-05,
      "loss": 0.0152,
      "step": 1780
    },
    {
      "epoch": 2.714691943127962,
      "grad_norm": 0.10424303263425827,
      "learning_rate": 5.119189104869683e-05,
      "loss": 0.0248,
      "step": 1790
    },
    {
      "epoch": 2.729857819905213,
      "grad_norm": 0.4774502217769623,
      "learning_rate": 5.066220506040148e-05,
      "loss": 0.0205,
      "step": 1800
    },
    {
      "epoch": 2.729857819905213,
      "eval_loss": 0.030218223109841347,
      "eval_runtime": 176.6117,
      "eval_samples_per_second": 6.642,
      "eval_steps_per_second": 6.642,
      "step": 1800
    },
    {
      "epoch": 2.7450236966824644,
      "grad_norm": 0.13609637320041656,
      "learning_rate": 5.013244472932872e-05,
      "loss": 0.0199,
      "step": 1810
    },
    {
      "epoch": 2.7601895734597157,
      "grad_norm": 0.19494231045246124,
      "learning_rate": 4.960266952928316e-05,
      "loss": 0.0203,
      "step": 1820
    },
    {
      "epoch": 2.775355450236967,
      "grad_norm": 0.27867355942726135,
      "learning_rate": 4.907293893573867e-05,
      "loss": 0.0179,
      "step": 1830
    },
    {
      "epoch": 2.790521327014218,
      "grad_norm": 0.18963003158569336,
      "learning_rate": 4.8543312419161396e-05,
      "loss": 0.0104,
      "step": 1840
    },
    {
      "epoch": 2.8056872037914693,
      "grad_norm": 0.33311042189598083,
      "learning_rate": 4.8013849438333165e-05,
      "loss": 0.0197,
      "step": 1850
    },
    {
      "epoch": 2.8056872037914693,
      "eval_loss": 0.030651915818452835,
      "eval_runtime": 176.8032,
      "eval_samples_per_second": 6.634,
      "eval_steps_per_second": 6.634,
      "step": 1850
    },
    {
      "epoch": 2.82085308056872,
      "grad_norm": 0.2615697383880615,
      "learning_rate": 4.748460943367643e-05,
      "loss": 0.0203,
      "step": 1860
    },
    {
      "epoch": 2.8360189573459715,
      "grad_norm": 0.19395503401756287,
      "learning_rate": 4.695565182058113e-05,
      "loss": 0.0146,
      "step": 1870
    },
    {
      "epoch": 2.851184834123223,
      "grad_norm": 0.11207451671361923,
      "learning_rate": 4.642703598273431e-05,
      "loss": 0.0137,
      "step": 1880
    },
    {
      "epoch": 2.8663507109004738,
      "grad_norm": 0.2404172271490097,
      "learning_rate": 4.589882126545352e-05,
      "loss": 0.0158,
      "step": 1890
    },
    {
      "epoch": 2.881516587677725,
      "grad_norm": 0.11668932437896729,
      "learning_rate": 4.537106696902425e-05,
      "loss": 0.0161,
      "step": 1900
    },
    {
      "epoch": 2.881516587677725,
      "eval_loss": 0.030294641852378845,
      "eval_runtime": 176.8559,
      "eval_samples_per_second": 6.633,
      "eval_steps_per_second": 6.633,
      "step": 1900
    },
    {
      "epoch": 2.8966824644549765,
      "grad_norm": 0.12421189993619919,
      "learning_rate": 4.484383234204266e-05,
      "loss": 0.0166,
      "step": 1910
    },
    {
      "epoch": 2.9118483412322274,
      "grad_norm": 0.1858258843421936,
      "learning_rate": 4.4317176574763935e-05,
      "loss": 0.0205,
      "step": 1920
    },
    {
      "epoch": 2.9270142180094787,
      "grad_norm": 0.07033693790435791,
      "learning_rate": 4.3791158792457334e-05,
      "loss": 0.0164,
      "step": 1930
    },
    {
      "epoch": 2.94218009478673,
      "grad_norm": 0.5009151697158813,
      "learning_rate": 4.3265838048768334e-05,
      "loss": 0.0278,
      "step": 1940
    },
    {
      "epoch": 2.957345971563981,
      "grad_norm": 0.2717853784561157,
      "learning_rate": 4.274127331908915e-05,
      "loss": 0.0235,
      "step": 1950
    },
    {
      "epoch": 2.957345971563981,
      "eval_loss": 0.030197137966752052,
      "eval_runtime": 176.9858,
      "eval_samples_per_second": 6.628,
      "eval_steps_per_second": 6.628,
      "step": 1950
    },
    {
      "epoch": 2.9725118483412323,
      "grad_norm": 0.16889254748821259,
      "learning_rate": 4.221752349393758e-05,
      "loss": 0.0221,
      "step": 1960
    },
    {
      "epoch": 2.987677725118483,
      "grad_norm": 0.39976805448532104,
      "learning_rate": 4.1694647372345886e-05,
      "loss": 0.026,
      "step": 1970
    },
    {
      "epoch": 3.0028436018957345,
      "grad_norm": 0.08540892601013184,
      "learning_rate": 4.117270365525946e-05,
      "loss": 0.0212,
      "step": 1980
    },
    {
      "epoch": 3.018009478672986,
      "grad_norm": 0.22224882245063782,
      "learning_rate": 4.065175093894694e-05,
      "loss": 0.0144,
      "step": 1990
    },
    {
      "epoch": 3.0331753554502368,
      "grad_norm": 0.14129629731178284,
      "learning_rate": 4.013184770842167e-05,
      "loss": 0.01,
      "step": 2000
    },
    {
      "epoch": 3.0331753554502368,
      "eval_loss": 0.03006371296942234,
      "eval_runtime": 176.7015,
      "eval_samples_per_second": 6.638,
      "eval_steps_per_second": 6.638,
      "step": 2000
    },
    {
      "epoch": 3.048341232227488,
      "grad_norm": 0.14608627557754517,
      "learning_rate": 3.9613052330876064e-05,
      "loss": 0.0149,
      "step": 2010
    },
    {
      "epoch": 3.0635071090047394,
      "grad_norm": 0.04342254251241684,
      "learning_rate": 3.909542304912881e-05,
      "loss": 0.0124,
      "step": 2020
    },
    {
      "epoch": 3.0786729857819903,
      "grad_norm": 0.14242447912693024,
      "learning_rate": 3.857901797508628e-05,
      "loss": 0.0102,
      "step": 2030
    },
    {
      "epoch": 3.0938388625592417,
      "grad_norm": 0.22291754186153412,
      "learning_rate": 3.80638950832186e-05,
      "loss": 0.0128,
      "step": 2040
    },
    {
      "epoch": 3.109004739336493,
      "grad_norm": 0.15756551921367645,
      "learning_rate": 3.7550112204051014e-05,
      "loss": 0.0073,
      "step": 2050
    },
    {
      "epoch": 3.109004739336493,
      "eval_loss": 0.03245367109775543,
      "eval_runtime": 176.6878,
      "eval_samples_per_second": 6.639,
      "eval_steps_per_second": 6.639,
      "step": 2050
    },
    {
      "epoch": 3.124170616113744,
      "grad_norm": 0.13003289699554443,
      "learning_rate": 3.703772701767167e-05,
      "loss": 0.0132,
      "step": 2060
    },
    {
      "epoch": 3.1393364928909953,
      "grad_norm": 0.1293676346540451,
      "learning_rate": 3.652679704725596e-05,
      "loss": 0.0066,
      "step": 2070
    },
    {
      "epoch": 3.1545023696682466,
      "grad_norm": 0.1557125747203827,
      "learning_rate": 3.601737965260882e-05,
      "loss": 0.0144,
      "step": 2080
    },
    {
      "epoch": 3.1696682464454975,
      "grad_norm": 0.3026883900165558,
      "learning_rate": 3.550953202372503e-05,
      "loss": 0.011,
      "step": 2090
    },
    {
      "epoch": 3.184834123222749,
      "grad_norm": 0.15523476898670197,
      "learning_rate": 3.500331117436895e-05,
      "loss": 0.0099,
      "step": 2100
    },
    {
      "epoch": 3.184834123222749,
      "eval_loss": 0.03371278569102287,
      "eval_runtime": 176.9028,
      "eval_samples_per_second": 6.631,
      "eval_steps_per_second": 6.631,
      "step": 2100
    },
    {
      "epoch": 3.2,
      "grad_norm": 0.21841643750667572,
      "learning_rate": 3.4498773935673626e-05,
      "loss": 0.0118,
      "step": 2110
    },
    {
      "epoch": 3.215165876777251,
      "grad_norm": 0.8701363205909729,
      "learning_rate": 3.399597694976081e-05,
      "loss": 0.0115,
      "step": 2120
    },
    {
      "epoch": 3.2303317535545024,
      "grad_norm": 0.16382525861263275,
      "learning_rate": 3.349497666338187e-05,
      "loss": 0.0102,
      "step": 2130
    },
    {
      "epoch": 3.2454976303317538,
      "grad_norm": 0.17109428346157074,
      "learning_rate": 3.299582932158085e-05,
      "loss": 0.0092,
      "step": 2140
    },
    {
      "epoch": 3.2606635071090047,
      "grad_norm": 0.07784170657396317,
      "learning_rate": 3.2498590961379996e-05,
      "loss": 0.0085,
      "step": 2150
    },
    {
      "epoch": 3.2606635071090047,
      "eval_loss": 0.033658258616924286,
      "eval_runtime": 176.8981,
      "eval_samples_per_second": 6.631,
      "eval_steps_per_second": 6.631,
      "step": 2150
    },
    {
      "epoch": 3.275829383886256,
      "grad_norm": 0.23731419444084167,
      "learning_rate": 3.200331740548887e-05,
      "loss": 0.0094,
      "step": 2160
    },
    {
      "epoch": 3.2909952606635073,
      "grad_norm": 0.17054717242717743,
      "learning_rate": 3.1510064256037274e-05,
      "loss": 0.009,
      "step": 2170
    },
    {
      "epoch": 3.3061611374407582,
      "grad_norm": 0.4204874634742737,
      "learning_rate": 3.1018886888333065e-05,
      "loss": 0.0082,
      "step": 2180
    },
    {
      "epoch": 3.3213270142180096,
      "grad_norm": 0.3489740490913391,
      "learning_rate": 3.052984044464548e-05,
      "loss": 0.0139,
      "step": 2190
    },
    {
      "epoch": 3.3364928909952605,
      "grad_norm": 0.22664600610733032,
      "learning_rate": 3.0042979828014496e-05,
      "loss": 0.0076,
      "step": 2200
    },
    {
      "epoch": 3.3364928909952605,
      "eval_loss": 0.035356372594833374,
      "eval_runtime": 176.9834,
      "eval_samples_per_second": 6.628,
      "eval_steps_per_second": 6.628,
      "step": 2200
    },
    {
      "epoch": 3.351658767772512,
      "grad_norm": 0.23908552527427673,
      "learning_rate": 2.9558359696087152e-05,
      "loss": 0.0096,
      "step": 2210
    },
    {
      "epoch": 3.366824644549763,
      "grad_norm": 0.3297005295753479,
      "learning_rate": 2.9076034454981332e-05,
      "loss": 0.0097,
      "step": 2220
    },
    {
      "epoch": 3.381990521327014,
      "grad_norm": 0.1065937951207161,
      "learning_rate": 2.8596058253177932e-05,
      "loss": 0.0127,
      "step": 2230
    },
    {
      "epoch": 3.3971563981042654,
      "grad_norm": 0.4293796122074127,
      "learning_rate": 2.811848497544175e-05,
      "loss": 0.0139,
      "step": 2240
    },
    {
      "epoch": 3.4123222748815167,
      "grad_norm": 0.05868324264883995,
      "learning_rate": 2.764336823677216e-05,
      "loss": 0.0077,
      "step": 2250
    },
    {
      "epoch": 3.4123222748815167,
      "eval_loss": 0.03412007540464401,
      "eval_runtime": 177.0155,
      "eval_samples_per_second": 6.627,
      "eval_steps_per_second": 6.627,
      "step": 2250
    },
    {
      "epoch": 3.4274881516587676,
      "grad_norm": 0.3067063093185425,
      "learning_rate": 2.717076137638388e-05,
      "loss": 0.0095,
      "step": 2260
    },
    {
      "epoch": 3.442654028436019,
      "grad_norm": 0.25318676233291626,
      "learning_rate": 2.6700717451719066e-05,
      "loss": 0.0133,
      "step": 2270
    },
    {
      "epoch": 3.4578199052132703,
      "grad_norm": 0.1500188261270523,
      "learning_rate": 2.6233289232490447e-05,
      "loss": 0.0076,
      "step": 2280
    },
    {
      "epoch": 3.472985781990521,
      "grad_norm": 0.0972483903169632,
      "learning_rate": 2.5768529194757474e-05,
      "loss": 0.0118,
      "step": 2290
    },
    {
      "epoch": 3.4881516587677726,
      "grad_norm": 0.11835496872663498,
      "learning_rate": 2.5306489515034713e-05,
      "loss": 0.0107,
      "step": 2300
    },
    {
      "epoch": 3.4881516587677726,
      "eval_loss": 0.03378909453749657,
      "eval_runtime": 177.022,
      "eval_samples_per_second": 6.626,
      "eval_steps_per_second": 6.626,
      "step": 2300
    },
    {
      "epoch": 3.5033175355450235,
      "grad_norm": 0.4174239933490753,
      "learning_rate": 2.484722206443455e-05,
      "loss": 0.0146,
      "step": 2310
    },
    {
      "epoch": 3.518483412322275,
      "grad_norm": 0.14443957805633545,
      "learning_rate": 2.4390778402843605e-05,
      "loss": 0.0096,
      "step": 2320
    },
    {
      "epoch": 3.533649289099526,
      "grad_norm": 0.17293648421764374,
      "learning_rate": 2.3937209773134472e-05,
      "loss": 0.0093,
      "step": 2330
    },
    {
      "epoch": 3.5488151658767775,
      "grad_norm": 0.24881836771965027,
      "learning_rate": 2.3486567095412864e-05,
      "loss": 0.0114,
      "step": 2340
    },
    {
      "epoch": 3.5639810426540284,
      "grad_norm": 0.13191837072372437,
      "learning_rate": 2.3038900961301053e-05,
      "loss": 0.006,
      "step": 2350
    },
    {
      "epoch": 3.5639810426540284,
      "eval_loss": 0.03380955010652542,
      "eval_runtime": 176.8892,
      "eval_samples_per_second": 6.631,
      "eval_steps_per_second": 6.631,
      "step": 2350
    },
    {
      "epoch": 3.5791469194312797,
      "grad_norm": 0.3207627236843109,
      "learning_rate": 2.259426162825819e-05,
      "loss": 0.0116,
      "step": 2360
    },
    {
      "epoch": 3.5943127962085306,
      "grad_norm": 0.13752657175064087,
      "learning_rate": 2.215269901393805e-05,
      "loss": 0.0074,
      "step": 2370
    },
    {
      "epoch": 3.609478672985782,
      "grad_norm": 0.22952240705490112,
      "learning_rate": 2.1714262690585192e-05,
      "loss": 0.011,
      "step": 2380
    },
    {
      "epoch": 3.6246445497630333,
      "grad_norm": 0.20379796624183655,
      "learning_rate": 2.1279001879469424e-05,
      "loss": 0.0134,
      "step": 2390
    },
    {
      "epoch": 3.639810426540284,
      "grad_norm": 0.4424147605895996,
      "learning_rate": 2.084696544536025e-05,
      "loss": 0.0127,
      "step": 2400
    },
    {
      "epoch": 3.639810426540284,
      "eval_loss": 0.03356926888227463,
      "eval_runtime": 176.7652,
      "eval_samples_per_second": 6.636,
      "eval_steps_per_second": 6.636,
      "step": 2400
    },
    {
      "epoch": 3.6549763033175355,
      "grad_norm": 0.14149199426174164,
      "learning_rate": 2.0418201891040778e-05,
      "loss": 0.0117,
      "step": 2410
    },
    {
      "epoch": 3.670142180094787,
      "grad_norm": 0.15457390248775482,
      "learning_rate": 1.9992759351862772e-05,
      "loss": 0.0085,
      "step": 2420
    },
    {
      "epoch": 3.6853080568720378,
      "grad_norm": 0.1752661019563675,
      "learning_rate": 1.9570685590342463e-05,
      "loss": 0.0066,
      "step": 2430
    },
    {
      "epoch": 3.700473933649289,
      "grad_norm": 0.17715801298618317,
      "learning_rate": 1.9152027990798748e-05,
      "loss": 0.0053,
      "step": 2440
    },
    {
      "epoch": 3.7156398104265405,
      "grad_norm": 0.06656470149755478,
      "learning_rate": 1.8736833554033263e-05,
      "loss": 0.0099,
      "step": 2450
    },
    {
      "epoch": 3.7156398104265405,
      "eval_loss": 0.033763591200113297,
      "eval_runtime": 176.8005,
      "eval_samples_per_second": 6.635,
      "eval_steps_per_second": 6.635,
      "step": 2450
    },
    {
      "epoch": 3.7308056872037914,
      "grad_norm": 0.33968186378479004,
      "learning_rate": 1.8325148892054093e-05,
      "loss": 0.0107,
      "step": 2460
    },
    {
      "epoch": 3.7459715639810427,
      "grad_norm": 0.1185387372970581,
      "learning_rate": 1.7917020222842662e-05,
      "loss": 0.0078,
      "step": 2470
    },
    {
      "epoch": 3.7611374407582936,
      "grad_norm": 0.7232157588005066,
      "learning_rate": 1.751249336516513e-05,
      "loss": 0.0076,
      "step": 2480
    },
    {
      "epoch": 3.776303317535545,
      "grad_norm": 0.3267844021320343,
      "learning_rate": 1.7111613733428522e-05,
      "loss": 0.0144,
      "step": 2490
    },
    {
      "epoch": 3.7914691943127963,
      "grad_norm": 0.3364531397819519,
      "learning_rate": 1.671442633258218e-05,
      "loss": 0.014,
      "step": 2500
    },
    {
      "epoch": 3.7914691943127963,
      "eval_loss": 0.03367255628108978,
      "eval_runtime": 176.7942,
      "eval_samples_per_second": 6.635,
      "eval_steps_per_second": 6.635,
      "step": 2500
    },
    {
      "epoch": 3.8066350710900476,
      "grad_norm": 0.29441037774086,
      "learning_rate": 1.6320975753065453e-05,
      "loss": 0.0121,
      "step": 2510
    },
    {
      "epoch": 3.8218009478672985,
      "grad_norm": 0.1159694567322731,
      "learning_rate": 1.5931306165801452e-05,
      "loss": 0.007,
      "step": 2520
    },
    {
      "epoch": 3.83696682464455,
      "grad_norm": 0.06502162665128708,
      "learning_rate": 1.554546131723848e-05,
      "loss": 0.0102,
      "step": 2530
    },
    {
      "epoch": 3.8521327014218008,
      "grad_norm": 0.2042492926120758,
      "learning_rate": 1.5163484524438516e-05,
      "loss": 0.0091,
      "step": 2540
    },
    {
      "epoch": 3.867298578199052,
      "grad_norm": 0.16165339946746826,
      "learning_rate": 1.4785418670214496e-05,
      "loss": 0.0129,
      "step": 2550
    },
    {
      "epoch": 3.867298578199052,
      "eval_loss": 0.03390151634812355,
      "eval_runtime": 176.7582,
      "eval_samples_per_second": 6.636,
      "eval_steps_per_second": 6.636,
      "step": 2550
    },
    {
      "epoch": 3.8824644549763034,
      "grad_norm": 0.1416504681110382,
      "learning_rate": 1.4411306198315777e-05,
      "loss": 0.0078,
      "step": 2560
    },
    {
      "epoch": 3.8976303317535548,
      "grad_norm": 0.20074276626110077,
      "learning_rate": 1.4041189108663421e-05,
      "loss": 0.0076,
      "step": 2570
    },
    {
      "epoch": 3.9127962085308057,
      "grad_norm": 0.44658246636390686,
      "learning_rate": 1.3675108952634824e-05,
      "loss": 0.0121,
      "step": 2580
    },
    {
      "epoch": 3.927962085308057,
      "grad_norm": 0.2680942714214325,
      "learning_rate": 1.3313106828399147e-05,
      "loss": 0.0132,
      "step": 2590
    },
    {
      "epoch": 3.943127962085308,
      "grad_norm": 0.2255931943655014,
      "learning_rate": 1.2955223376303205e-05,
      "loss": 0.0118,
      "step": 2600
    },
    {
      "epoch": 3.943127962085308,
      "eval_loss": 0.03499338775873184,
      "eval_runtime": 176.6361,
      "eval_samples_per_second": 6.641,
      "eval_steps_per_second": 6.641,
      "step": 2600
    },
    {
      "epoch": 3.9582938388625593,
      "grad_norm": 0.21901127696037292,
      "learning_rate": 1.2601498774309112e-05,
      "loss": 0.0085,
      "step": 2610
    },
    {
      "epoch": 3.9734597156398106,
      "grad_norm": 0.17668171226978302,
      "learning_rate": 1.2251972733483612e-05,
      "loss": 0.011,
      "step": 2620
    },
    {
      "epoch": 3.9886255924170615,
      "grad_norm": 0.17130517959594727,
      "learning_rate": 1.19066844935399e-05,
      "loss": 0.0095,
      "step": 2630
    },
    {
      "epoch": 4.003791469194312,
      "grad_norm": 0.10923975706100464,
      "learning_rate": 1.156567281843241e-05,
      "loss": 0.0124,
      "step": 2640
    },
    {
      "epoch": 4.018957345971564,
      "grad_norm": 0.19639810919761658,
      "learning_rate": 1.1228975992004842e-05,
      "loss": 0.0073,
      "step": 2650
    },
    {
      "epoch": 4.018957345971564,
      "eval_loss": 0.03456135094165802,
      "eval_runtime": 176.4872,
      "eval_samples_per_second": 6.646,
      "eval_steps_per_second": 6.646,
      "step": 2650
    },
    {
      "epoch": 4.034123222748815,
      "grad_norm": 0.030663151293992996,
      "learning_rate": 1.0896631813692377e-05,
      "loss": 0.0041,
      "step": 2660
    },
    {
      "epoch": 4.049289099526066,
      "grad_norm": 0.07115893065929413,
      "learning_rate": 1.0568677594277899e-05,
      "loss": 0.0051,
      "step": 2670
    },
    {
      "epoch": 4.064454976303318,
      "grad_norm": 0.04100283235311508,
      "learning_rate": 1.0245150151703509e-05,
      "loss": 0.0043,
      "step": 2680
    },
    {
      "epoch": 4.079620853080569,
      "grad_norm": 0.08858209103345871,
      "learning_rate": 9.926085806936918e-06,
      "loss": 0.006,
      "step": 2690
    },
    {
      "epoch": 4.0947867298578196,
      "grad_norm": 0.01757189631462097,
      "learning_rate": 9.611520379894068e-06,
      "loss": 0.0048,
      "step": 2700
    },
    {
      "epoch": 4.0947867298578196,
      "eval_loss": 0.0357203334569931,
      "eval_runtime": 176.4817,
      "eval_samples_per_second": 6.647,
      "eval_steps_per_second": 6.647,
      "step": 2700
    },
    {
      "epoch": 4.109952606635071,
      "grad_norm": 0.11255928874015808,
      "learning_rate": 9.301489185417578e-06,
      "loss": 0.0052,
      "step": 2710
    },
    {
      "epoch": 4.125118483412322,
      "grad_norm": 0.12291298806667328,
      "learning_rate": 8.996027029312314e-06,
      "loss": 0.0048,
      "step": 2720
    },
    {
      "epoch": 4.140284360189574,
      "grad_norm": 0.2394372820854187,
      "learning_rate": 8.695168204437721e-06,
      "loss": 0.0042,
      "step": 2730
    },
    {
      "epoch": 4.155450236966825,
      "grad_norm": 0.09991464763879776,
      "learning_rate": 8.398946486858029e-06,
      "loss": 0.0036,
      "step": 2740
    },
    {
      "epoch": 4.170616113744076,
      "grad_norm": 0.21715304255485535,
      "learning_rate": 8.107395132050333e-06,
      "loss": 0.0059,
      "step": 2750
    },
    {
      "epoch": 4.170616113744076,
      "eval_loss": 0.037330999970436096,
      "eval_runtime": 176.5966,
      "eval_samples_per_second": 6.642,
      "eval_steps_per_second": 6.642,
      "step": 2750
    },
    {
      "epoch": 4.185781990521327,
      "grad_norm": 0.06366084516048431,
      "learning_rate": 7.820546871171142e-06,
      "loss": 0.0045,
      "step": 2760
    },
    {
      "epoch": 4.200947867298578,
      "grad_norm": 0.12914450466632843,
      "learning_rate": 7.5384339073818145e-06,
      "loss": 0.0046,
      "step": 2770
    },
    {
      "epoch": 4.216113744075829,
      "grad_norm": 0.032806962728500366,
      "learning_rate": 7.261087912233228e-06,
      "loss": 0.0055,
      "step": 2780
    },
    {
      "epoch": 4.231279620853081,
      "grad_norm": 0.1332745999097824,
      "learning_rate": 6.988540022110235e-06,
      "loss": 0.0064,
      "step": 2790
    },
    {
      "epoch": 4.246445497630332,
      "grad_norm": 0.07780851423740387,
      "learning_rate": 6.720820834735969e-06,
      "loss": 0.0053,
      "step": 2800
    },
    {
      "epoch": 4.246445497630332,
      "eval_loss": 0.037292271852493286,
      "eval_runtime": 176.612,
      "eval_samples_per_second": 6.642,
      "eval_steps_per_second": 6.642,
      "step": 2800
    },
    {
      "epoch": 4.2616113744075825,
      "grad_norm": 0.13824769854545593,
      "learning_rate": 6.457960405736968e-06,
      "loss": 0.0037,
      "step": 2810
    },
    {
      "epoch": 4.276777251184834,
      "grad_norm": 0.15241014957427979,
      "learning_rate": 6.199988245268778e-06,
      "loss": 0.0043,
      "step": 2820
    },
    {
      "epoch": 4.291943127962085,
      "grad_norm": 0.03748151287436485,
      "learning_rate": 5.946933314703157e-06,
      "loss": 0.0048,
      "step": 2830
    },
    {
      "epoch": 4.3071090047393366,
      "grad_norm": 0.011155886575579643,
      "learning_rate": 5.698824023376531e-06,
      "loss": 0.003,
      "step": 2840
    },
    {
      "epoch": 4.322274881516588,
      "grad_norm": 0.15347355604171753,
      "learning_rate": 5.455688225400802e-06,
      "loss": 0.0045,
      "step": 2850
    },
    {
      "epoch": 4.322274881516588,
      "eval_loss": 0.03810197114944458,
      "eval_runtime": 176.7279,
      "eval_samples_per_second": 6.637,
      "eval_steps_per_second": 6.637,
      "step": 2850
    },
    {
      "epoch": 4.337440758293839,
      "grad_norm": 0.18955792486667633,
      "learning_rate": 5.217553216536098e-06,
      "loss": 0.0075,
      "step": 2860
    },
    {
      "epoch": 4.35260663507109,
      "grad_norm": 0.11830997467041016,
      "learning_rate": 4.984445731126542e-06,
      "loss": 0.0063,
      "step": 2870
    },
    {
      "epoch": 4.367772511848341,
      "grad_norm": 0.09092006832361221,
      "learning_rate": 4.756391939098853e-06,
      "loss": 0.003,
      "step": 2880
    },
    {
      "epoch": 4.382938388625592,
      "grad_norm": 0.049758732318878174,
      "learning_rate": 4.533417443024374e-06,
      "loss": 0.008,
      "step": 2890
    },
    {
      "epoch": 4.398104265402844,
      "grad_norm": 0.39580151438713074,
      "learning_rate": 4.315547275244769e-06,
      "loss": 0.0054,
      "step": 2900
    },
    {
      "epoch": 4.398104265402844,
      "eval_loss": 0.03884812071919441,
      "eval_runtime": 176.624,
      "eval_samples_per_second": 6.641,
      "eval_steps_per_second": 6.641,
      "step": 2900
    },
    {
      "epoch": 4.413270142180095,
      "grad_norm": 0.11998700350522995,
      "learning_rate": 4.10280589506179e-06,
      "loss": 0.0033,
      "step": 2910
    },
    {
      "epoch": 4.4284360189573455,
      "grad_norm": 0.5768892765045166,
      "learning_rate": 3.895217185991312e-06,
      "loss": 0.0077,
      "step": 2920
    },
    {
      "epoch": 4.443601895734597,
      "grad_norm": 0.13106589019298553,
      "learning_rate": 3.692804453082038e-06,
      "loss": 0.0087,
      "step": 2930
    },
    {
      "epoch": 4.458767772511848,
      "grad_norm": 0.23035043478012085,
      "learning_rate": 3.495590420299194e-06,
      "loss": 0.0049,
      "step": 2940
    },
    {
      "epoch": 4.4739336492890995,
      "grad_norm": 0.1375955194234848,
      "learning_rate": 3.3035972279733207e-06,
      "loss": 0.0085,
      "step": 2950
    },
    {
      "epoch": 4.4739336492890995,
      "eval_loss": 0.03845488652586937,
      "eval_runtime": 176.6446,
      "eval_samples_per_second": 6.64,
      "eval_steps_per_second": 6.64,
      "step": 2950
    },
    {
      "epoch": 4.489099526066351,
      "grad_norm": 0.10981747508049011,
      "learning_rate": 3.116846430314796e-06,
      "loss": 0.0035,
      "step": 2960
    },
    {
      "epoch": 4.504265402843602,
      "grad_norm": 0.0285560991615057,
      "learning_rate": 2.9353589929939087e-06,
      "loss": 0.0042,
      "step": 2970
    },
    {
      "epoch": 4.519431279620854,
      "grad_norm": 0.22480367124080658,
      "learning_rate": 2.7591552907872574e-06,
      "loss": 0.0059,
      "step": 2980
    },
    {
      "epoch": 4.534597156398104,
      "grad_norm": 0.26764968037605286,
      "learning_rate": 2.5882551052902883e-06,
      "loss": 0.006,
      "step": 2990
    },
    {
      "epoch": 4.549763033175355,
      "grad_norm": 0.21272222697734833,
      "learning_rate": 2.4226776226965453e-06,
      "loss": 0.0066,
      "step": 3000
    },
    {
      "epoch": 4.549763033175355,
      "eval_loss": 0.03843921050429344,
      "eval_runtime": 176.6472,
      "eval_samples_per_second": 6.64,
      "eval_steps_per_second": 6.64,
      "step": 3000
    },
    {
      "epoch": 4.564928909952607,
      "grad_norm": 0.10441111028194427,
      "learning_rate": 2.262441431643697e-06,
      "loss": 0.0053,
      "step": 3010
    },
    {
      "epoch": 4.580094786729858,
      "grad_norm": 0.1366783082485199,
      "learning_rate": 2.1075645211266927e-06,
      "loss": 0.0044,
      "step": 3020
    },
    {
      "epoch": 4.595260663507109,
      "grad_norm": 0.07694575935602188,
      "learning_rate": 1.9580642784782053e-06,
      "loss": 0.0062,
      "step": 3030
    },
    {
      "epoch": 4.61042654028436,
      "grad_norm": 0.08338585495948792,
      "learning_rate": 1.813957487416651e-06,
      "loss": 0.0053,
      "step": 3040
    },
    {
      "epoch": 4.625592417061611,
      "grad_norm": 0.28469717502593994,
      "learning_rate": 1.6752603261619315e-06,
      "loss": 0.0051,
      "step": 3050
    },
    {
      "epoch": 4.625592417061611,
      "eval_loss": 0.03863969072699547,
      "eval_runtime": 176.6381,
      "eval_samples_per_second": 6.641,
      "eval_steps_per_second": 6.641,
      "step": 3050
    },
    {
      "epoch": 4.6407582938388625,
      "grad_norm": 0.21466806530952454,
      "learning_rate": 1.541988365619207e-06,
      "loss": 0.0053,
      "step": 3060
    },
    {
      "epoch": 4.655924170616114,
      "grad_norm": 0.1829688400030136,
      "learning_rate": 1.4141565676307865e-06,
      "loss": 0.005,
      "step": 3070
    },
    {
      "epoch": 4.671090047393365,
      "grad_norm": 0.08133982867002487,
      "learning_rate": 1.291779283296468e-06,
      "loss": 0.0029,
      "step": 3080
    },
    {
      "epoch": 4.6862559241706165,
      "grad_norm": 0.03812313824892044,
      "learning_rate": 1.1748702513623922e-06,
      "loss": 0.0033,
      "step": 3090
    },
    {
      "epoch": 4.701421800947867,
      "grad_norm": 0.02549022249877453,
      "learning_rate": 1.0634425966786155e-06,
      "loss": 0.0052,
      "step": 3100
    },
    {
      "epoch": 4.701421800947867,
      "eval_loss": 0.038835834711790085,
      "eval_runtime": 176.5888,
      "eval_samples_per_second": 6.643,
      "eval_steps_per_second": 6.643,
      "step": 3100
    },
    {
      "epoch": 4.716587677725118,
      "grad_norm": 0.2480955272912979,
      "learning_rate": 9.575088287257118e-07,
      "loss": 0.0038,
      "step": 3110
    },
    {
      "epoch": 4.73175355450237,
      "grad_norm": 0.22870886325836182,
      "learning_rate": 8.570808402103392e-07,
      "loss": 0.0067,
      "step": 3120
    },
    {
      "epoch": 4.746919431279621,
      "grad_norm": 0.08617883920669556,
      "learning_rate": 7.62169905730109e-07,
      "loss": 0.006,
      "step": 3130
    },
    {
      "epoch": 4.762085308056872,
      "grad_norm": 0.19280649721622467,
      "learning_rate": 6.727866805078531e-07,
      "loss": 0.0045,
      "step": 3140
    },
    {
      "epoch": 4.777251184834123,
      "grad_norm": 0.1964663714170456,
      "learning_rate": 5.889411991953975e-07,
      "loss": 0.0065,
      "step": 3150
    },
    {
      "epoch": 4.777251184834123,
      "eval_loss": 0.03892850503325462,
      "eval_runtime": 176.5708,
      "eval_samples_per_second": 6.643,
      "eval_steps_per_second": 6.643,
      "step": 3150
    },
    {
      "epoch": 4.792417061611374,
      "grad_norm": 0.022485673427581787,
      "learning_rate": 5.106428747470137e-07,
      "loss": 0.0054,
      "step": 3160
    },
    {
      "epoch": 4.8075829383886255,
      "grad_norm": 0.11611665785312653,
      "learning_rate": 4.37900497362681e-07,
      "loss": 0.0046,
      "step": 3170
    },
    {
      "epoch": 4.822748815165877,
      "grad_norm": 0.08721260726451874,
      "learning_rate": 3.7072223350124167e-07,
      "loss": 0.0054,
      "step": 3180
    },
    {
      "epoch": 4.837914691943128,
      "grad_norm": 0.1968434751033783,
      "learning_rate": 3.0911562496358517e-07,
      "loss": 0.0035,
      "step": 3190
    },
    {
      "epoch": 4.8530805687203795,
      "grad_norm": 0.10239926725625992,
      "learning_rate": 2.53087588045986e-07,
      "loss": 0.0036,
      "step": 3200
    },
    {
      "epoch": 4.8530805687203795,
      "eval_loss": 0.03907202184200287,
      "eval_runtime": 176.5342,
      "eval_samples_per_second": 6.645,
      "eval_steps_per_second": 6.645,
      "step": 3200
    },
    {
      "epoch": 4.86824644549763,
      "grad_norm": 0.19318081438541412,
      "learning_rate": 2.026444127636029e-07,
      "loss": 0.0025,
      "step": 3210
    },
    {
      "epoch": 4.883412322274881,
      "grad_norm": 0.9325800538063049,
      "learning_rate": 1.577917621443825e-07,
      "loss": 0.0066,
      "step": 3220
    },
    {
      "epoch": 4.898578199052133,
      "grad_norm": 0.1284843534231186,
      "learning_rate": 1.185346715932345e-07,
      "loss": 0.0061,
      "step": 3230
    },
    {
      "epoch": 4.913744075829384,
      "grad_norm": 0.41085031628608704,
      "learning_rate": 8.48775483267783e-08,
      "loss": 0.0072,
      "step": 3240
    },
    {
      "epoch": 4.928909952606635,
      "grad_norm": 0.053091879934072495,
      "learning_rate": 5.6824170878544414e-08,
      "loss": 0.0039,
      "step": 3250
    },
    {
      "epoch": 4.928909952606635,
      "eval_loss": 0.03911284729838371,
      "eval_runtime": 176.5645,
      "eval_samples_per_second": 6.643,
      "eval_steps_per_second": 6.643,
      "step": 3250
    },
    {
      "epoch": 4.944075829383886,
      "grad_norm": 0.18437696993350983,
      "learning_rate": 3.4377688674774866e-08,
      "loss": 0.0051,
      "step": 3260
    },
    {
      "epoch": 4.959241706161137,
      "grad_norm": 0.04348951205611229,
      "learning_rate": 1.7540621680850466e-08,
      "loss": 0.0026,
      "step": 3270
    },
    {
      "epoch": 4.9744075829383885,
      "grad_norm": 0.18638713657855988,
      "learning_rate": 6.314860118400434e-09,
      "loss": 0.0042,
      "step": 3280
    },
    {
      "epoch": 4.98957345971564,
      "grad_norm": 0.1402566283941269,
      "learning_rate": 7.016642530777162e-10,
      "loss": 0.0045,
      "step": 3290
    },
    {
      "epoch": 4.9971563981042655,
      "step": 3295,
      "total_flos": 8.907449267845202e+17,
      "train_loss": 0.04891899318543117,
      "train_runtime": 37195.0472,
      "train_samples_per_second": 1.418,
      "train_steps_per_second": 0.089
    }
  ],
  "logging_steps": 10,
  "max_steps": 3295,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 50,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 8.907449267845202e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}