{
  "best_metric": 0.6133951445650848,
  "best_model_checkpoint": "./step_test_microsoft_dit/checkpoint-2500",
  "epoch": 0.6942691239585963,
  "eval_steps": 50,
  "global_step": 2750,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0025246149962130774,
      "grad_norm": 1.0554239749908447,
      "learning_rate": 2.997e-05,
      "loss": 0.3197,
      "step": 10
    },
    {
      "epoch": 0.005049229992426155,
      "grad_norm": 1.5600422620773315,
      "learning_rate": 2.994e-05,
      "loss": 0.2047,
      "step": 20
    },
    {
      "epoch": 0.007573844988639233,
      "grad_norm": 2.1541621685028076,
      "learning_rate": 2.991e-05,
      "loss": 0.1528,
      "step": 30
    },
    {
      "epoch": 0.01009845998485231,
      "grad_norm": 1.805535078048706,
      "learning_rate": 2.9880000000000002e-05,
      "loss": 0.1252,
      "step": 40
    },
    {
      "epoch": 0.012623074981065387,
      "grad_norm": 1.1236392259597778,
      "learning_rate": 2.985e-05,
      "loss": 0.1165,
      "step": 50
    },
    {
      "epoch": 0.012623074981065387,
      "eval_f1": 0.4177168854259269,
      "eval_loss": 0.06423712521791458,
      "eval_runtime": 1142.2038,
      "eval_samples_per_second": 180.582,
      "eval_steps_per_second": 2.822,
      "step": 50
    },
    {
      "epoch": 0.015147689977278465,
      "grad_norm": 1.1924934387207031,
      "learning_rate": 2.982e-05,
      "loss": 0.1029,
      "step": 60
    },
    {
      "epoch": 0.017672304973491544,
      "grad_norm": 1.225701928138733,
      "learning_rate": 2.979e-05,
      "loss": 0.117,
      "step": 70
    },
    {
      "epoch": 0.02019691996970462,
      "grad_norm": 2.702486515045166,
      "learning_rate": 2.976e-05,
      "loss": 0.1103,
      "step": 80
    },
    {
      "epoch": 0.022721534965917698,
      "grad_norm": 2.0278918743133545,
      "learning_rate": 2.973e-05,
      "loss": 0.1054,
      "step": 90
    },
    {
      "epoch": 0.025246149962130773,
      "grad_norm": 1.9288796186447144,
      "learning_rate": 2.97e-05,
      "loss": 0.0942,
      "step": 100
    },
    {
      "epoch": 0.025246149962130773,
      "eval_f1": 0.4771817453963171,
      "eval_loss": 0.048530641943216324,
      "eval_runtime": 1007.4651,
      "eval_samples_per_second": 204.734,
      "eval_steps_per_second": 3.199,
      "step": 100
    },
    {
      "epoch": 0.027770764958343852,
      "grad_norm": 1.4688999652862549,
      "learning_rate": 2.967e-05,
      "loss": 0.0992,
      "step": 110
    },
    {
      "epoch": 0.03029537995455693,
      "grad_norm": 1.1097605228424072,
      "learning_rate": 2.964e-05,
      "loss": 0.1126,
      "step": 120
    },
    {
      "epoch": 0.03281999495077001,
      "grad_norm": 1.0353784561157227,
      "learning_rate": 2.961e-05,
      "loss": 0.0949,
      "step": 130
    },
    {
      "epoch": 0.03534460994698309,
      "grad_norm": 1.7303999662399292,
      "learning_rate": 2.958e-05,
      "loss": 0.1025,
      "step": 140
    },
    {
      "epoch": 0.03786922494319616,
      "grad_norm": 1.1177138090133667,
      "learning_rate": 2.955e-05,
      "loss": 0.1076,
      "step": 150
    },
    {
      "epoch": 0.03786922494319616,
      "eval_f1": 0.46432628333423487,
      "eval_loss": 0.05836363136768341,
      "eval_runtime": 980.1012,
      "eval_samples_per_second": 210.45,
      "eval_steps_per_second": 3.288,
      "step": 150
    },
    {
      "epoch": 0.04039383993940924,
      "grad_norm": 1.1965147256851196,
      "learning_rate": 2.9520000000000002e-05,
      "loss": 0.0961,
      "step": 160
    },
    {
      "epoch": 0.04291845493562232,
      "grad_norm": 1.0545780658721924,
      "learning_rate": 2.949e-05,
      "loss": 0.104,
      "step": 170
    },
    {
      "epoch": 0.045443069931835396,
      "grad_norm": 1.8348199129104614,
      "learning_rate": 2.946e-05,
      "loss": 0.0932,
      "step": 180
    },
    {
      "epoch": 0.047967684928048475,
      "grad_norm": 1.8478541374206543,
      "learning_rate": 2.943e-05,
      "loss": 0.1069,
      "step": 190
    },
    {
      "epoch": 0.05049229992426155,
      "grad_norm": 0.9377999305725098,
      "learning_rate": 2.94e-05,
      "loss": 0.1103,
      "step": 200
    },
    {
      "epoch": 0.05049229992426155,
      "eval_f1": NaN,
      "eval_loss": 0.044557176530361176,
      "eval_runtime": 978.7525,
      "eval_samples_per_second": 210.74,
      "eval_steps_per_second": 3.293,
      "step": 200
    },
    {
      "epoch": 0.053016914920474625,
      "grad_norm": 1.6204830408096313,
      "learning_rate": 2.9370000000000002e-05,
      "loss": 0.1019,
      "step": 210
    },
    {
      "epoch": 0.055541529916687704,
      "grad_norm": 1.1411000490188599,
      "learning_rate": 2.934e-05,
      "loss": 0.0969,
      "step": 220
    },
    {
      "epoch": 0.05806614491290078,
      "grad_norm": 1.1179866790771484,
      "learning_rate": 2.931e-05,
      "loss": 0.1031,
      "step": 230
    },
    {
      "epoch": 0.06059075990911386,
      "grad_norm": 1.2155176401138306,
      "learning_rate": 2.928e-05,
      "loss": 0.0851,
      "step": 240
    },
    {
      "epoch": 0.06311537490532694,
      "grad_norm": 1.4578701257705688,
      "learning_rate": 2.925e-05,
      "loss": 0.0873,
      "step": 250
    },
    {
      "epoch": 0.06311537490532694,
      "eval_f1": 0.5313367950730588,
      "eval_loss": 0.05184657499194145,
      "eval_runtime": 982.8389,
      "eval_samples_per_second": 209.863,
      "eval_steps_per_second": 3.279,
      "step": 250
    },
    {
      "epoch": 0.06563998990154002,
      "grad_norm": 1.2894303798675537,
      "learning_rate": 2.922e-05,
      "loss": 0.0876,
      "step": 260
    },
    {
      "epoch": 0.0681646048977531,
      "grad_norm": 0.8404099941253662,
      "learning_rate": 2.919e-05,
      "loss": 0.0904,
      "step": 270
    },
    {
      "epoch": 0.07068921989396618,
      "grad_norm": 2.0062506198883057,
      "learning_rate": 2.916e-05,
      "loss": 0.1009,
      "step": 280
    },
    {
      "epoch": 0.07321383489017924,
      "grad_norm": 0.8900242447853088,
      "learning_rate": 2.913e-05,
      "loss": 0.0925,
      "step": 290
    },
    {
      "epoch": 0.07573844988639232,
      "grad_norm": 1.051013708114624,
      "learning_rate": 2.91e-05,
      "loss": 0.1053,
      "step": 300
    },
    {
      "epoch": 0.07573844988639232,
      "eval_f1": 0.532925682031985,
      "eval_loss": 0.07359323650598526,
      "eval_runtime": 980.9407,
      "eval_samples_per_second": 210.27,
      "eval_steps_per_second": 3.286,
      "step": 300
    },
    {
      "epoch": 0.0782630648826054,
      "grad_norm": 0.7765111327171326,
      "learning_rate": 2.907e-05,
      "loss": 0.0848,
      "step": 310
    },
    {
      "epoch": 0.08078767987881848,
      "grad_norm": 0.9605777859687805,
      "learning_rate": 2.904e-05,
      "loss": 0.0746,
      "step": 320
    },
    {
      "epoch": 0.08331229487503156,
      "grad_norm": 1.9086962938308716,
      "learning_rate": 2.901e-05,
      "loss": 0.1023,
      "step": 330
    },
    {
      "epoch": 0.08583690987124463,
      "grad_norm": 1.5782345533370972,
      "learning_rate": 2.898e-05,
      "loss": 0.0751,
      "step": 340
    },
    {
      "epoch": 0.08836152486745771,
      "grad_norm": 1.2298818826675415,
      "learning_rate": 2.895e-05,
      "loss": 0.0797,
      "step": 350
    },
    {
      "epoch": 0.08836152486745771,
      "eval_f1": 0.5325518588749066,
      "eval_loss": 0.07257544994354248,
      "eval_runtime": 979.2135,
      "eval_samples_per_second": 210.64,
      "eval_steps_per_second": 3.291,
      "step": 350
    },
    {
      "epoch": 0.09088613986367079,
      "grad_norm": 1.1932893991470337,
      "learning_rate": 2.892e-05,
      "loss": 0.0803,
      "step": 360
    },
    {
      "epoch": 0.09341075485988387,
      "grad_norm": 0.896007776260376,
      "learning_rate": 2.889e-05,
      "loss": 0.088,
      "step": 370
    },
    {
      "epoch": 0.09593536985609695,
      "grad_norm": 2.385890483856201,
      "learning_rate": 2.8859999999999998e-05,
      "loss": 0.0886,
      "step": 380
    },
    {
      "epoch": 0.09845998485231003,
      "grad_norm": 0.966077446937561,
      "learning_rate": 2.883e-05,
      "loss": 0.1038,
      "step": 390
    },
    {
      "epoch": 0.1009845998485231,
      "grad_norm": 0.969159722328186,
      "learning_rate": 2.88e-05,
      "loss": 0.0857,
      "step": 400
    },
    {
      "epoch": 0.1009845998485231,
      "eval_f1": 0.5497736226259776,
      "eval_loss": 0.06929118931293488,
      "eval_runtime": 978.0405,
      "eval_samples_per_second": 210.893,
      "eval_steps_per_second": 3.295,
      "step": 400
    },
    {
      "epoch": 0.10350921484473617,
      "grad_norm": 0.8633397817611694,
      "learning_rate": 2.877e-05,
      "loss": 0.0895,
      "step": 410
    },
    {
      "epoch": 0.10603382984094925,
      "grad_norm": 1.163271188735962,
      "learning_rate": 2.874e-05,
      "loss": 0.0861,
      "step": 420
    },
    {
      "epoch": 0.10855844483716233,
      "grad_norm": 1.102964997291565,
      "learning_rate": 2.871e-05,
      "loss": 0.0962,
      "step": 430
    },
    {
      "epoch": 0.11108305983337541,
      "grad_norm": 1.520044207572937,
      "learning_rate": 2.868e-05,
      "loss": 0.0981,
      "step": 440
    },
    {
      "epoch": 0.11360767482958849,
      "grad_norm": 1.8637338876724243,
      "learning_rate": 2.865e-05,
      "loss": 0.0885,
      "step": 450
    },
    {
      "epoch": 0.11360767482958849,
      "eval_f1": NaN,
      "eval_loss": 0.09174469113349915,
      "eval_runtime": 1032.8967,
      "eval_samples_per_second": 199.693,
      "eval_steps_per_second": 3.12,
      "step": 450
    },
    {
      "epoch": 0.11613228982580157,
      "grad_norm": 1.1974824666976929,
      "learning_rate": 2.862e-05,
      "loss": 0.0784,
      "step": 460
    },
    {
      "epoch": 0.11865690482201464,
      "grad_norm": 1.6933320760726929,
      "learning_rate": 2.859e-05,
      "loss": 0.078,
      "step": 470
    },
    {
      "epoch": 0.12118151981822772,
      "grad_norm": 1.7774609327316284,
      "learning_rate": 2.856e-05,
      "loss": 0.0715,
      "step": 480
    },
    {
      "epoch": 0.1237061348144408,
      "grad_norm": 0.7675666213035583,
      "learning_rate": 2.853e-05,
      "loss": 0.0817,
      "step": 490
    },
    {
      "epoch": 0.12623074981065388,
      "grad_norm": 1.169325590133667,
      "learning_rate": 2.8499999999999998e-05,
      "loss": 0.102,
      "step": 500
    },
    {
      "epoch": 0.12623074981065388,
      "eval_f1": 0.5648781658864481,
      "eval_loss": 0.057994671165943146,
      "eval_runtime": 967.2924,
      "eval_samples_per_second": 213.236,
      "eval_steps_per_second": 3.332,
      "step": 500
    },
    {
      "epoch": 0.12875536480686695,
      "grad_norm": 0.9567933678627014,
      "learning_rate": 2.847e-05,
      "loss": 0.0762,
      "step": 510
    },
    {
      "epoch": 0.13127997980308004,
      "grad_norm": 0.7539889216423035,
      "learning_rate": 2.844e-05,
      "loss": 0.0655,
      "step": 520
    },
    {
      "epoch": 0.1338045947992931,
      "grad_norm": 1.873833179473877,
      "learning_rate": 2.841e-05,
      "loss": 0.0747,
      "step": 530
    },
    {
      "epoch": 0.1363292097955062,
      "grad_norm": 0.7834559082984924,
      "learning_rate": 2.838e-05,
      "loss": 0.0923,
      "step": 540
    },
    {
      "epoch": 0.13885382479171926,
      "grad_norm": 0.6193771362304688,
      "learning_rate": 2.8349999999999998e-05,
      "loss": 0.0716,
      "step": 550
    },
    {
      "epoch": 0.13885382479171926,
      "eval_f1": 0.538135593220339,
      "eval_loss": 0.07973095029592514,
      "eval_runtime": 974.1593,
      "eval_samples_per_second": 211.733,
      "eval_steps_per_second": 3.308,
      "step": 550
    },
    {
      "epoch": 0.14137843978793235,
      "grad_norm": 1.1256766319274902,
      "learning_rate": 2.832e-05,
      "loss": 0.0798,
      "step": 560
    },
    {
      "epoch": 0.14390305478414542,
      "grad_norm": 1.0669515132904053,
      "learning_rate": 2.829e-05,
      "loss": 0.0795,
      "step": 570
    },
    {
      "epoch": 0.14642766978035848,
      "grad_norm": 1.018234133720398,
      "learning_rate": 2.826e-05,
      "loss": 0.073,
      "step": 580
    },
    {
      "epoch": 0.14895228477657158,
      "grad_norm": 1.2367616891860962,
      "learning_rate": 2.823e-05,
      "loss": 0.0879,
      "step": 590
    },
    {
      "epoch": 0.15147689977278464,
      "grad_norm": 1.5840317010879517,
      "learning_rate": 2.8199999999999998e-05,
      "loss": 0.0854,
      "step": 600
    },
    {
      "epoch": 0.15147689977278464,
      "eval_f1": 0.571752762018513,
      "eval_loss": 0.07439474016427994,
      "eval_runtime": 970.5653,
      "eval_samples_per_second": 212.517,
      "eval_steps_per_second": 3.321,
      "step": 600
    },
    {
      "epoch": 0.15400151476899773,
      "grad_norm": 0.5361483097076416,
      "learning_rate": 2.817e-05,
      "loss": 0.0854,
      "step": 610
    },
    {
      "epoch": 0.1565261297652108,
      "grad_norm": 0.9658698439598083,
      "learning_rate": 2.8139999999999998e-05,
      "loss": 0.095,
      "step": 620
    },
    {
      "epoch": 0.1590507447614239,
      "grad_norm": 0.820649266242981,
      "learning_rate": 2.8110000000000004e-05,
      "loss": 0.0921,
      "step": 630
    },
    {
      "epoch": 0.16157535975763695,
      "grad_norm": 1.1583890914916992,
      "learning_rate": 2.8080000000000002e-05,
      "loss": 0.077,
      "step": 640
    },
    {
      "epoch": 0.16409997475385005,
      "grad_norm": 0.8755506277084351,
      "learning_rate": 2.805e-05,
      "loss": 0.089,
      "step": 650
    },
    {
      "epoch": 0.16409997475385005,
      "eval_f1": 0.5789600675594161,
      "eval_loss": 0.0503680482506752,
      "eval_runtime": 976.7796,
      "eval_samples_per_second": 211.165,
      "eval_steps_per_second": 3.3,
      "step": 650
    },
    {
      "epoch": 0.1666245897500631,
      "grad_norm": 0.5073147416114807,
      "learning_rate": 2.8020000000000003e-05,
      "loss": 0.0784,
      "step": 660
    },
    {
      "epoch": 0.1691492047462762,
      "grad_norm": 1.0332393646240234,
      "learning_rate": 2.799e-05,
      "loss": 0.0906,
      "step": 670
    },
    {
      "epoch": 0.17167381974248927,
      "grad_norm": 1.1538151502609253,
      "learning_rate": 2.7960000000000003e-05,
      "loss": 0.0799,
      "step": 680
    },
    {
      "epoch": 0.17419843473870233,
      "grad_norm": 1.2075843811035156,
      "learning_rate": 2.7930000000000002e-05,
      "loss": 0.0795,
      "step": 690
    },
    {
      "epoch": 0.17672304973491543,
      "grad_norm": 2.1169683933258057,
      "learning_rate": 2.79e-05,
      "loss": 0.0721,
      "step": 700
    },
    {
      "epoch": 0.17672304973491543,
      "eval_f1": 0.5727175590644663,
      "eval_loss": 0.0618172287940979,
      "eval_runtime": 975.6558,
      "eval_samples_per_second": 211.409,
      "eval_steps_per_second": 3.303,
      "step": 700
    },
    {
      "epoch": 0.1792476647311285,
      "grad_norm": 1.3094089031219482,
      "learning_rate": 2.7870000000000003e-05,
      "loss": 0.0723,
      "step": 710
    },
    {
      "epoch": 0.18177227972734158,
      "grad_norm": 0.9937088489532471,
      "learning_rate": 2.784e-05,
      "loss": 0.0704,
      "step": 720
    },
    {
      "epoch": 0.18429689472355465,
      "grad_norm": 0.6464220881462097,
      "learning_rate": 2.7810000000000003e-05,
      "loss": 0.0731,
      "step": 730
    },
    {
      "epoch": 0.18682150971976774,
      "grad_norm": 0.5544419288635254,
      "learning_rate": 2.778e-05,
      "loss": 0.0894,
      "step": 740
    },
    {
      "epoch": 0.1893461247159808,
      "grad_norm": 0.6369556188583374,
      "learning_rate": 2.7750000000000004e-05,
      "loss": 0.0721,
      "step": 750
    },
    {
      "epoch": 0.1893461247159808,
      "eval_f1": 0.5904197411394702,
      "eval_loss": 0.07033708691596985,
      "eval_runtime": 967.5811,
      "eval_samples_per_second": 213.173,
      "eval_steps_per_second": 3.331,
      "step": 750
    },
    {
      "epoch": 0.1918707397121939,
      "grad_norm": 2.0700013637542725,
      "learning_rate": 2.7720000000000002e-05,
      "loss": 0.0831,
      "step": 760
    },
    {
      "epoch": 0.19439535470840696,
      "grad_norm": 0.765533983707428,
      "learning_rate": 2.769e-05,
      "loss": 0.0707,
      "step": 770
    },
    {
      "epoch": 0.19691996970462006,
      "grad_norm": 1.6104159355163574,
      "learning_rate": 2.7660000000000003e-05,
      "loss": 0.073,
      "step": 780
    },
    {
      "epoch": 0.19944458470083312,
      "grad_norm": 1.1069729328155518,
      "learning_rate": 2.763e-05,
      "loss": 0.0702,
      "step": 790
    },
    {
      "epoch": 0.2019691996970462,
      "grad_norm": 1.6577630043029785,
      "learning_rate": 2.7600000000000003e-05,
      "loss": 0.0865,
      "step": 800
    },
    {
      "epoch": 0.2019691996970462,
      "eval_f1": 0.5952780441035476,
      "eval_loss": 0.058820515871047974,
      "eval_runtime": 917.4267,
      "eval_samples_per_second": 224.827,
      "eval_steps_per_second": 3.513,
      "step": 800
    },
    {
      "epoch": 0.20449381469325928,
      "grad_norm": 1.5197840929031372,
      "learning_rate": 2.7570000000000002e-05,
      "loss": 0.0846,
      "step": 810
    },
    {
      "epoch": 0.20701842968947234,
      "grad_norm": 1.1758556365966797,
      "learning_rate": 2.754e-05,
      "loss": 0.0813,
      "step": 820
    },
    {
      "epoch": 0.20954304468568544,
      "grad_norm": 0.5016022324562073,
      "learning_rate": 2.7510000000000003e-05,
      "loss": 0.0718,
      "step": 830
    },
    {
      "epoch": 0.2120676596818985,
      "grad_norm": 1.3600627183914185,
      "learning_rate": 2.748e-05,
      "loss": 0.0942,
      "step": 840
    },
    {
      "epoch": 0.2145922746781116,
      "grad_norm": 0.6990534067153931,
      "learning_rate": 2.7450000000000003e-05,
      "loss": 0.0767,
      "step": 850
    },
    {
      "epoch": 0.2145922746781116,
      "eval_f1": 0.5918155918155918,
      "eval_loss": 0.04372716695070267,
      "eval_runtime": 913.4291,
      "eval_samples_per_second": 225.811,
      "eval_steps_per_second": 3.528,
      "step": 850
    },
    {
      "epoch": 0.21711688967432466,
      "grad_norm": 1.0468288660049438,
      "learning_rate": 2.7420000000000002e-05,
      "loss": 0.0805,
      "step": 860
    },
    {
      "epoch": 0.21964150467053775,
      "grad_norm": 1.2046771049499512,
      "learning_rate": 2.739e-05,
      "loss": 0.0879,
      "step": 870
    },
    {
      "epoch": 0.22216611966675082,
      "grad_norm": 0.9044977426528931,
      "learning_rate": 2.7360000000000002e-05,
      "loss": 0.0597,
      "step": 880
    },
    {
      "epoch": 0.2246907346629639,
      "grad_norm": 1.145572304725647,
      "learning_rate": 2.733e-05,
      "loss": 0.1007,
      "step": 890
    },
    {
      "epoch": 0.22721534965917697,
      "grad_norm": 1.058166742324829,
      "learning_rate": 2.7300000000000003e-05,
      "loss": 0.0773,
      "step": 900
    },
    {
      "epoch": 0.22721534965917697,
      "eval_f1": 0.5956852791878172,
      "eval_loss": 0.05675825849175453,
      "eval_runtime": 923.1927,
      "eval_samples_per_second": 223.422,
      "eval_steps_per_second": 3.491,
      "step": 900
    },
    {
      "epoch": 0.22973996465539007,
      "grad_norm": 0.7665570974349976,
      "learning_rate": 2.727e-05,
      "loss": 0.084,
      "step": 910
    },
    {
      "epoch": 0.23226457965160313,
      "grad_norm": 0.8884145021438599,
      "learning_rate": 2.724e-05,
      "loss": 0.0748,
      "step": 920
    },
    {
      "epoch": 0.2347891946478162,
      "grad_norm": 0.7132917046546936,
      "learning_rate": 2.7210000000000002e-05,
      "loss": 0.0861,
      "step": 930
    },
    {
      "epoch": 0.2373138096440293,
      "grad_norm": 1.3353750705718994,
      "learning_rate": 2.718e-05,
      "loss": 0.091,
      "step": 940
    },
    {
      "epoch": 0.23983842464024235,
      "grad_norm": 1.216691255569458,
      "learning_rate": 2.7150000000000003e-05,
      "loss": 0.0748,
      "step": 950
    },
    {
      "epoch": 0.23983842464024235,
      "eval_f1": 0.5942299042601041,
      "eval_loss": 0.04645048826932907,
      "eval_runtime": 919.478,
      "eval_samples_per_second": 224.325,
      "eval_steps_per_second": 3.505,
      "step": 950
    },
    {
      "epoch": 0.24236303963645545,
      "grad_norm": 1.0420501232147217,
      "learning_rate": 2.712e-05,
      "loss": 0.0953,
      "step": 960
    },
    {
      "epoch": 0.2448876546326685,
      "grad_norm": 1.1488158702850342,
      "learning_rate": 2.709e-05,
      "loss": 0.0796,
      "step": 970
    },
    {
      "epoch": 0.2474122696288816,
      "grad_norm": 0.7872379422187805,
      "learning_rate": 2.7060000000000002e-05,
      "loss": 0.0844,
      "step": 980
    },
    {
      "epoch": 0.24993688462509467,
      "grad_norm": 0.9102885127067566,
      "learning_rate": 2.703e-05,
      "loss": 0.0792,
      "step": 990
    },
    {
      "epoch": 0.25246149962130776,
      "grad_norm": 1.040650486946106,
      "learning_rate": 2.7000000000000002e-05,
      "loss": 0.0761,
      "step": 1000
    },
    {
      "epoch": 0.25246149962130776,
      "eval_f1": NaN,
      "eval_loss": 0.06595388799905777,
      "eval_runtime": 948.4123,
      "eval_samples_per_second": 217.481,
      "eval_steps_per_second": 3.398,
      "step": 1000
    },
    {
      "epoch": 0.25498611461752085,
      "grad_norm": 1.0717836618423462,
      "learning_rate": 2.697e-05,
      "loss": 0.0569,
      "step": 1010
    },
    {
      "epoch": 0.2575107296137339,
      "grad_norm": 0.7504699230194092,
      "learning_rate": 2.6940000000000003e-05,
      "loss": 0.072,
      "step": 1020
    },
    {
      "epoch": 0.260035344609947,
      "grad_norm": 0.9767778515815735,
      "learning_rate": 2.691e-05,
      "loss": 0.0658,
      "step": 1030
    },
    {
      "epoch": 0.2625599596061601,
      "grad_norm": 0.5905674695968628,
      "learning_rate": 2.688e-05,
      "loss": 0.0775,
      "step": 1040
    },
    {
      "epoch": 0.2650845746023731,
      "grad_norm": 1.6352293491363525,
      "learning_rate": 2.6850000000000002e-05,
      "loss": 0.0855,
      "step": 1050
    },
    {
      "epoch": 0.2650845746023731,
      "eval_f1": 0.5963938973647711,
      "eval_loss": 0.04910014942288399,
      "eval_runtime": 986.8376,
      "eval_samples_per_second": 209.013,
      "eval_steps_per_second": 3.266,
      "step": 1050
    },
    {
      "epoch": 0.2676091895985862,
      "grad_norm": 0.6634190082550049,
      "learning_rate": 2.682e-05,
      "loss": 0.0741,
      "step": 1060
    },
    {
      "epoch": 0.2701338045947993,
      "grad_norm": 0.5896914601325989,
      "learning_rate": 2.6790000000000003e-05,
      "loss": 0.0713,
      "step": 1070
    },
    {
      "epoch": 0.2726584195910124,
      "grad_norm": 1.3768564462661743,
      "learning_rate": 2.676e-05,
      "loss": 0.0684,
      "step": 1080
    },
    {
      "epoch": 0.27518303458722543,
      "grad_norm": 0.7323074340820312,
      "learning_rate": 2.673e-05,
      "loss": 0.084,
      "step": 1090
    },
    {
      "epoch": 0.2777076495834385,
      "grad_norm": 0.6660707592964172,
      "learning_rate": 2.6700000000000002e-05,
      "loss": 0.0832,
      "step": 1100
    },
    {
      "epoch": 0.2777076495834385,
      "eval_f1": 0.6048397002825205,
      "eval_loss": 0.049847185611724854,
      "eval_runtime": 967.6797,
      "eval_samples_per_second": 213.151,
      "eval_steps_per_second": 3.331,
      "step": 1100
    },
    {
      "epoch": 0.2802322645796516,
      "grad_norm": 1.425309419631958,
      "learning_rate": 2.667e-05,
      "loss": 0.0793,
      "step": 1110
    },
    {
      "epoch": 0.2827568795758647,
      "grad_norm": 1.3583918809890747,
      "learning_rate": 2.6640000000000002e-05,
      "loss": 0.0808,
      "step": 1120
    },
    {
      "epoch": 0.28528149457207774,
      "grad_norm": 1.1851533651351929,
      "learning_rate": 2.661e-05,
      "loss": 0.0738,
      "step": 1130
    },
    {
      "epoch": 0.28780610956829084,
      "grad_norm": 1.4497005939483643,
      "learning_rate": 2.658e-05,
      "loss": 0.078,
      "step": 1140
    },
    {
      "epoch": 0.29033072456450393,
      "grad_norm": 1.4407027959823608,
      "learning_rate": 2.655e-05,
      "loss": 0.0821,
      "step": 1150
    },
    {
      "epoch": 0.29033072456450393,
      "eval_f1": 0.6031633616619453,
      "eval_loss": 0.059650588780641556,
      "eval_runtime": 962.0892,
      "eval_samples_per_second": 214.39,
      "eval_steps_per_second": 3.35,
      "step": 1150
    },
    {
      "epoch": 0.29285533956071697,
      "grad_norm": 1.0721668004989624,
      "learning_rate": 2.652e-05,
      "loss": 0.0706,
      "step": 1160
    },
    {
      "epoch": 0.29537995455693006,
      "grad_norm": 1.1033729314804077,
      "learning_rate": 2.6490000000000002e-05,
      "loss": 0.0737,
      "step": 1170
    },
    {
      "epoch": 0.29790456955314315,
      "grad_norm": 0.9764577746391296,
      "learning_rate": 2.646e-05,
      "loss": 0.0743,
      "step": 1180
    },
    {
      "epoch": 0.30042918454935624,
      "grad_norm": 1.2160297632217407,
      "learning_rate": 2.643e-05,
      "loss": 0.0768,
      "step": 1190
    },
    {
      "epoch": 0.3029537995455693,
      "grad_norm": 0.8387085795402527,
      "learning_rate": 2.64e-05,
      "loss": 0.0715,
      "step": 1200
    },
    {
      "epoch": 0.3029537995455693,
      "eval_f1": NaN,
      "eval_loss": 0.06428094953298569,
      "eval_runtime": 961.1037,
      "eval_samples_per_second": 214.61,
      "eval_steps_per_second": 3.353,
      "step": 1200
    },
    {
      "epoch": 0.3054784145417824,
      "grad_norm": 1.061087727546692,
      "learning_rate": 2.637e-05,
      "loss": 0.0672,
      "step": 1210
    },
    {
      "epoch": 0.30800302953799547,
      "grad_norm": 0.6768150925636292,
      "learning_rate": 2.6340000000000002e-05,
      "loss": 0.0762,
      "step": 1220
    },
    {
      "epoch": 0.31052764453420856,
      "grad_norm": 0.7020296454429626,
      "learning_rate": 2.631e-05,
      "loss": 0.0838,
      "step": 1230
    },
    {
      "epoch": 0.3130522595304216,
      "grad_norm": 0.9264736175537109,
      "learning_rate": 2.628e-05,
      "loss": 0.0769,
      "step": 1240
    },
    {
      "epoch": 0.3155768745266347,
      "grad_norm": 0.657778799533844,
      "learning_rate": 2.625e-05,
      "loss": 0.085,
      "step": 1250
    },
    {
      "epoch": 0.3155768745266347,
      "eval_f1": 0.6054250016184373,
      "eval_loss": 0.06593530625104904,
      "eval_runtime": 970.0262,
      "eval_samples_per_second": 212.635,
      "eval_steps_per_second": 3.323,
      "step": 1250
    },
    {
      "epoch": 0.3181014895228478,
      "grad_norm": 0.6904731392860413,
      "learning_rate": 2.622e-05,
      "loss": 0.0736,
      "step": 1260
    },
    {
      "epoch": 0.3206261045190608,
      "grad_norm": 1.4745820760726929,
      "learning_rate": 2.619e-05,
      "loss": 0.0832,
      "step": 1270
    },
    {
      "epoch": 0.3231507195152739,
      "grad_norm": 1.0614553689956665,
      "learning_rate": 2.616e-05,
      "loss": 0.0781,
      "step": 1280
    },
    {
      "epoch": 0.325675334511487,
      "grad_norm": 1.2228913307189941,
      "learning_rate": 2.6130000000000002e-05,
      "loss": 0.0872,
      "step": 1290
    },
    {
      "epoch": 0.3281999495077001,
      "grad_norm": 0.9905760288238525,
      "learning_rate": 2.61e-05,
      "loss": 0.0826,
      "step": 1300
    },
    {
      "epoch": 0.3281999495077001,
      "eval_f1": 0.6011740745177908,
      "eval_loss": 0.05560224503278732,
      "eval_runtime": 964.1962,
      "eval_samples_per_second": 213.921,
      "eval_steps_per_second": 3.343,
      "step": 1300
    },
    {
      "epoch": 0.33072456450391313,
      "grad_norm": 1.1195616722106934,
      "learning_rate": 2.607e-05,
      "loss": 0.0751,
      "step": 1310
    },
    {
      "epoch": 0.3332491795001262,
      "grad_norm": 0.9830445647239685,
      "learning_rate": 2.604e-05,
      "loss": 0.0694,
      "step": 1320
    },
    {
      "epoch": 0.3357737944963393,
      "grad_norm": 1.7140698432922363,
      "learning_rate": 2.601e-05,
      "loss": 0.0694,
      "step": 1330
    },
    {
      "epoch": 0.3382984094925524,
      "grad_norm": 0.9545607566833496,
      "learning_rate": 2.5980000000000002e-05,
      "loss": 0.0626,
      "step": 1340
    },
    {
      "epoch": 0.34082302448876545,
      "grad_norm": 0.8236456513404846,
      "learning_rate": 2.595e-05,
      "loss": 0.064,
      "step": 1350
    },
    {
      "epoch": 0.34082302448876545,
      "eval_f1": NaN,
      "eval_loss": 0.0564185306429863,
      "eval_runtime": 1030.8018,
      "eval_samples_per_second": 200.099,
      "eval_steps_per_second": 3.127,
      "step": 1350
    },
    {
      "epoch": 0.34334763948497854,
      "grad_norm": 1.0344712734222412,
      "learning_rate": 2.592e-05,
      "loss": 0.074,
      "step": 1360
    },
    {
      "epoch": 0.34587225448119163,
      "grad_norm": 1.647894024848938,
      "learning_rate": 2.589e-05,
      "loss": 0.0756,
      "step": 1370
    },
    {
      "epoch": 0.34839686947740467,
      "grad_norm": 1.0268642902374268,
      "learning_rate": 2.586e-05,
      "loss": 0.064,
      "step": 1380
    },
    {
      "epoch": 0.35092148447361776,
      "grad_norm": 0.6588199734687805,
      "learning_rate": 2.5830000000000002e-05,
      "loss": 0.0685,
      "step": 1390
    },
    {
      "epoch": 0.35344609946983085,
      "grad_norm": 0.8278918862342834,
      "learning_rate": 2.58e-05,
      "loss": 0.0854,
      "step": 1400
    },
    {
      "epoch": 0.35344609946983085,
      "eval_f1": NaN,
      "eval_loss": 0.05516933649778366,
      "eval_runtime": 1032.3177,
      "eval_samples_per_second": 199.805,
      "eval_steps_per_second": 3.122,
      "step": 1400
    },
    {
      "epoch": 0.35597071446604395,
      "grad_norm": 0.4216013550758362,
      "learning_rate": 2.577e-05,
      "loss": 0.0785,
      "step": 1410
    },
    {
      "epoch": 0.358495329462257,
      "grad_norm": 0.9567118287086487,
      "learning_rate": 2.574e-05,
      "loss": 0.089,
      "step": 1420
    },
    {
      "epoch": 0.3610199444584701,
      "grad_norm": 1.3202637434005737,
      "learning_rate": 2.571e-05,
      "loss": 0.0884,
      "step": 1430
    },
    {
      "epoch": 0.36354455945468317,
      "grad_norm": 1.3245704174041748,
      "learning_rate": 2.568e-05,
      "loss": 0.0739,
      "step": 1440
    },
    {
      "epoch": 0.36606917445089626,
      "grad_norm": 0.6416196823120117,
      "learning_rate": 2.565e-05,
      "loss": 0.0702,
      "step": 1450
    },
    {
      "epoch": 0.36606917445089626,
      "eval_f1": 0.6061020319393525,
      "eval_loss": 0.06748606264591217,
      "eval_runtime": 999.3826,
      "eval_samples_per_second": 206.389,
      "eval_steps_per_second": 3.225,
      "step": 1450
    },
    {
      "epoch": 0.3685937894471093,
      "grad_norm": 0.9312785267829895,
      "learning_rate": 2.562e-05,
      "loss": 0.0674,
      "step": 1460
    },
    {
      "epoch": 0.3711184044433224,
      "grad_norm": 0.9092572927474976,
      "learning_rate": 2.559e-05,
      "loss": 0.0676,
      "step": 1470
    },
    {
      "epoch": 0.3736430194395355,
      "grad_norm": 1.4935100078582764,
      "learning_rate": 2.556e-05,
      "loss": 0.0712,
      "step": 1480
    },
    {
      "epoch": 0.3761676344357485,
      "grad_norm": 0.9569060802459717,
      "learning_rate": 2.553e-05,
      "loss": 0.0747,
      "step": 1490
    },
    {
      "epoch": 0.3786922494319616,
      "grad_norm": 0.947384774684906,
      "learning_rate": 2.55e-05,
      "loss": 0.0771,
      "step": 1500
    },
    {
      "epoch": 0.3786922494319616,
      "eval_f1": NaN,
      "eval_loss": 0.057753585278987885,
      "eval_runtime": 1000.0105,
      "eval_samples_per_second": 206.26,
      "eval_steps_per_second": 3.223,
      "step": 1500
    },
    {
      "epoch": 0.3812168644281747,
      "grad_norm": 0.6996080875396729,
      "learning_rate": 2.547e-05,
      "loss": 0.0696,
      "step": 1510
    },
    {
      "epoch": 0.3837414794243878,
      "grad_norm": 0.5415595173835754,
      "learning_rate": 2.544e-05,
      "loss": 0.0757,
      "step": 1520
    },
    {
      "epoch": 0.38626609442060084,
      "grad_norm": 0.5137012600898743,
      "learning_rate": 2.541e-05,
      "loss": 0.0621,
      "step": 1530
    },
    {
      "epoch": 0.38879070941681393,
      "grad_norm": 0.9606865048408508,
      "learning_rate": 2.538e-05,
      "loss": 0.073,
      "step": 1540
    },
    {
      "epoch": 0.391315324413027,
      "grad_norm": 1.1751604080200195,
      "learning_rate": 2.535e-05,
      "loss": 0.08,
      "step": 1550
    },
    {
      "epoch": 0.391315324413027,
      "eval_f1": NaN,
      "eval_loss": 0.0491572804749012,
      "eval_runtime": 963.3777,
      "eval_samples_per_second": 214.103,
      "eval_steps_per_second": 3.346,
      "step": 1550
    },
    {
      "epoch": 0.3938399394092401,
      "grad_norm": 0.935338020324707,
      "learning_rate": 2.5319999999999998e-05,
      "loss": 0.0729,
      "step": 1560
    },
    {
      "epoch": 0.39636455440545315,
      "grad_norm": 0.7157814502716064,
      "learning_rate": 2.529e-05,
      "loss": 0.0719,
      "step": 1570
    },
    {
      "epoch": 0.39888916940166624,
      "grad_norm": 0.6739543676376343,
      "learning_rate": 2.526e-05,
      "loss": 0.0631,
      "step": 1580
    },
    {
      "epoch": 0.40141378439787934,
      "grad_norm": 0.4896785318851471,
      "learning_rate": 2.523e-05,
      "loss": 0.0746,
      "step": 1590
    },
    {
      "epoch": 0.4039383993940924,
      "grad_norm": 0.7619987726211548,
      "learning_rate": 2.52e-05,
      "loss": 0.0804,
      "step": 1600
    },
    {
      "epoch": 0.4039383993940924,
      "eval_f1": 0.6111605289687482,
      "eval_loss": 0.05378127843141556,
      "eval_runtime": 958.466,
      "eval_samples_per_second": 215.2,
      "eval_steps_per_second": 3.363,
      "step": 1600
    },
    {
      "epoch": 0.40646301439030547,
      "grad_norm": 0.7464210987091064,
      "learning_rate": 2.517e-05,
      "loss": 0.0707,
      "step": 1610
    },
    {
      "epoch": 0.40898762938651856,
      "grad_norm": 0.6707102656364441,
      "learning_rate": 2.514e-05,
      "loss": 0.0671,
      "step": 1620
    },
    {
      "epoch": 0.41151224438273165,
      "grad_norm": 1.246846079826355,
      "learning_rate": 2.511e-05,
      "loss": 0.0627,
      "step": 1630
    },
    {
      "epoch": 0.4140368593789447,
      "grad_norm": 0.9796457886695862,
      "learning_rate": 2.508e-05,
      "loss": 0.0677,
      "step": 1640
    },
    {
      "epoch": 0.4165614743751578,
      "grad_norm": 0.9717236161231995,
      "learning_rate": 2.505e-05,
      "loss": 0.083,
      "step": 1650
    },
    {
      "epoch": 0.4165614743751578,
      "eval_f1": 0.6047686163965234,
      "eval_loss": 0.057900335639715195,
      "eval_runtime": 962.7843,
      "eval_samples_per_second": 214.235,
      "eval_steps_per_second": 3.348,
      "step": 1650
    },
    {
      "epoch": 0.4190860893713709,
      "grad_norm": 1.1706446409225464,
      "learning_rate": 2.502e-05,
      "loss": 0.0764,
      "step": 1660
    },
    {
      "epoch": 0.42161070436758397,
      "grad_norm": 0.45280393958091736,
      "learning_rate": 2.499e-05,
      "loss": 0.0682,
      "step": 1670
    },
    {
      "epoch": 0.424135319363797,
      "grad_norm": 1.0100760459899902,
      "learning_rate": 2.4959999999999998e-05,
      "loss": 0.0892,
      "step": 1680
    },
    {
      "epoch": 0.4266599343600101,
      "grad_norm": 1.0506736040115356,
      "learning_rate": 2.493e-05,
      "loss": 0.0666,
      "step": 1690
    },
    {
      "epoch": 0.4291845493562232,
      "grad_norm": 0.7978639006614685,
      "learning_rate": 2.49e-05,
      "loss": 0.0701,
      "step": 1700
    },
    {
      "epoch": 0.4291845493562232,
      "eval_f1": 0.6044656147662996,
      "eval_loss": 0.06738731265068054,
      "eval_runtime": 1102.7864,
      "eval_samples_per_second": 187.037,
      "eval_steps_per_second": 2.923,
      "step": 1700
    },
    {
      "epoch": 0.4317091643524363,
      "grad_norm": 1.121317982673645,
      "learning_rate": 2.487e-05,
      "loss": 0.0771,
      "step": 1710
    },
    {
      "epoch": 0.4342337793486493,
      "grad_norm": 1.0836131572723389,
      "learning_rate": 2.484e-05,
      "loss": 0.0719,
      "step": 1720
    },
    {
      "epoch": 0.4367583943448624,
      "grad_norm": 0.61658775806427,
      "learning_rate": 2.4809999999999998e-05,
      "loss": 0.0681,
      "step": 1730
    },
    {
      "epoch": 0.4392830093410755,
      "grad_norm": 0.647393524646759,
      "learning_rate": 2.478e-05,
      "loss": 0.0668,
      "step": 1740
    },
    {
      "epoch": 0.44180762433728854,
      "grad_norm": 0.782483696937561,
      "learning_rate": 2.475e-05,
      "loss": 0.0721,
      "step": 1750
    },
    {
      "epoch": 0.44180762433728854,
      "eval_f1": 0.5979155238617663,
      "eval_loss": 0.04912808537483215,
      "eval_runtime": 1171.9033,
      "eval_samples_per_second": 176.006,
      "eval_steps_per_second": 2.75,
      "step": 1750
    },
    {
      "epoch": 0.44433223933350163,
      "grad_norm": 0.4538789987564087,
      "learning_rate": 2.472e-05,
      "loss": 0.0641,
      "step": 1760
    },
    {
      "epoch": 0.4468568543297147,
      "grad_norm": 0.7954159379005432,
      "learning_rate": 2.469e-05,
      "loss": 0.079,
      "step": 1770
    },
    {
      "epoch": 0.4493814693259278,
      "grad_norm": 0.4370203912258148,
      "learning_rate": 2.4659999999999998e-05,
      "loss": 0.0769,
      "step": 1780
    },
    {
      "epoch": 0.45190608432214086,
      "grad_norm": 1.2641068696975708,
      "learning_rate": 2.463e-05,
      "loss": 0.0649,
      "step": 1790
    },
    {
      "epoch": 0.45443069931835395,
      "grad_norm": 1.262468695640564,
      "learning_rate": 2.4599999999999998e-05,
      "loss": 0.0765,
      "step": 1800
    },
    {
      "epoch": 0.45443069931835395,
      "eval_f1": NaN,
      "eval_loss": 0.04386861249804497,
      "eval_runtime": 1166.6982,
      "eval_samples_per_second": 176.791,
      "eval_steps_per_second": 2.762,
      "step": 1800
    },
    {
      "epoch": 0.45695531431456704,
      "grad_norm": 1.0922938585281372,
      "learning_rate": 2.457e-05,
      "loss": 0.074,
      "step": 1810
    },
    {
      "epoch": 0.45947992931078013,
      "grad_norm": 0.8298421502113342,
      "learning_rate": 2.454e-05,
      "loss": 0.0778,
      "step": 1820
    },
    {
      "epoch": 0.46200454430699317,
      "grad_norm": 1.182712435722351,
      "learning_rate": 2.4509999999999997e-05,
      "loss": 0.0793,
      "step": 1830
    },
    {
      "epoch": 0.46452915930320626,
      "grad_norm": 0.7366443276405334,
      "learning_rate": 2.448e-05,
      "loss": 0.0655,
      "step": 1840
    },
    {
      "epoch": 0.46705377429941936,
      "grad_norm": 0.9185643792152405,
      "learning_rate": 2.4449999999999998e-05,
      "loss": 0.0692,
      "step": 1850
    },
    {
      "epoch": 0.46705377429941936,
      "eval_f1": 0.6057632592224568,
      "eval_loss": 0.04681675508618355,
      "eval_runtime": 1170.8075,
      "eval_samples_per_second": 176.171,
      "eval_steps_per_second": 2.753,
      "step": 1850
    },
    {
      "epoch": 0.4695783892956324,
      "grad_norm": 0.99897301197052,
      "learning_rate": 2.442e-05,
      "loss": 0.0685,
      "step": 1860
    },
    {
      "epoch": 0.4721030042918455,
      "grad_norm": 1.0028034448623657,
      "learning_rate": 2.439e-05,
      "loss": 0.0748,
      "step": 1870
    },
    {
      "epoch": 0.4746276192880586,
      "grad_norm": 2.5226945877075195,
      "learning_rate": 2.4360000000000004e-05,
      "loss": 0.0715,
      "step": 1880
    },
    {
      "epoch": 0.47715223428427167,
      "grad_norm": 0.903256893157959,
      "learning_rate": 2.4330000000000003e-05,
      "loss": 0.0709,
      "step": 1890
    },
    {
      "epoch": 0.4796768492804847,
      "grad_norm": 0.9269793629646301,
      "learning_rate": 2.43e-05,
      "loss": 0.0761,
      "step": 1900
    },
    {
      "epoch": 0.4796768492804847,
      "eval_f1": 0.6124984470120511,
      "eval_loss": 0.05741230770945549,
      "eval_runtime": 1159.3714,
      "eval_samples_per_second": 177.908,
      "eval_steps_per_second": 2.78,
      "step": 1900
    },
    {
      "epoch": 0.4822014642766978,
      "grad_norm": 1.0651170015335083,
      "learning_rate": 2.4270000000000003e-05,
      "loss": 0.0751,
      "step": 1910
    },
    {
      "epoch": 0.4847260792729109,
      "grad_norm": 1.2628437280654907,
      "learning_rate": 2.4240000000000002e-05,
      "loss": 0.0852,
      "step": 1920
    },
    {
      "epoch": 0.487250694269124,
      "grad_norm": 1.3889621496200562,
      "learning_rate": 2.4210000000000004e-05,
      "loss": 0.073,
      "step": 1930
    },
    {
      "epoch": 0.489775309265337,
      "grad_norm": 1.028456687927246,
      "learning_rate": 2.4180000000000002e-05,
      "loss": 0.0644,
      "step": 1940
    },
    {
      "epoch": 0.4922999242615501,
      "grad_norm": 0.6997565627098083,
      "learning_rate": 2.415e-05,
      "loss": 0.0757,
      "step": 1950
    },
    {
      "epoch": 0.4922999242615501,
      "eval_f1": 0.6126181795711549,
      "eval_loss": 0.05692484602332115,
      "eval_runtime": 1161.9825,
      "eval_samples_per_second": 177.509,
      "eval_steps_per_second": 2.774,
      "step": 1950
    },
    {
      "epoch": 0.4948245392577632,
      "grad_norm": 1.384186863899231,
      "learning_rate": 2.4120000000000003e-05,
      "loss": 0.0697,
      "step": 1960
    },
    {
      "epoch": 0.49734915425397624,
      "grad_norm": 0.8674394488334656,
      "learning_rate": 2.409e-05,
      "loss": 0.0739,
      "step": 1970
    },
    {
      "epoch": 0.49987376925018934,
      "grad_norm": 1.826121211051941,
      "learning_rate": 2.4060000000000003e-05,
      "loss": 0.0739,
      "step": 1980
    },
    {
      "epoch": 0.5023983842464024,
      "grad_norm": 0.6903666257858276,
      "learning_rate": 2.4030000000000002e-05,
      "loss": 0.0661,
      "step": 1990
    },
    {
      "epoch": 0.5049229992426155,
      "grad_norm": 0.7339742183685303,
      "learning_rate": 2.4e-05,
      "loss": 0.0654,
      "step": 2000
    },
    {
      "epoch": 0.5049229992426155,
      "eval_f1": 0.6095153739086423,
      "eval_loss": 0.05489746853709221,
      "eval_runtime": 1168.0449,
      "eval_samples_per_second": 176.587,
      "eval_steps_per_second": 2.759,
      "step": 2000
    },
    {
      "epoch": 0.5074476142388286,
      "grad_norm": 0.7863900065422058,
      "learning_rate": 2.3970000000000003e-05,
      "loss": 0.061,
      "step": 2010
    },
    {
      "epoch": 0.5099722292350417,
      "grad_norm": 1.0800750255584717,
      "learning_rate": 2.394e-05,
      "loss": 0.0781,
      "step": 2020
    },
    {
      "epoch": 0.5124968442312547,
      "grad_norm": 1.0992929935455322,
      "learning_rate": 2.3910000000000003e-05,
      "loss": 0.0694,
      "step": 2030
    },
    {
      "epoch": 0.5150214592274678,
      "grad_norm": 0.703554093837738,
      "learning_rate": 2.3880000000000002e-05,
      "loss": 0.0881,
      "step": 2040
    },
    {
      "epoch": 0.5175460742236809,
      "grad_norm": 1.214089274406433,
      "learning_rate": 2.385e-05,
      "loss": 0.0736,
      "step": 2050
    },
    {
      "epoch": 0.5175460742236809,
      "eval_f1": 0.612187690432663,
      "eval_loss": 0.05384594947099686,
      "eval_runtime": 1155.7771,
      "eval_samples_per_second": 178.462,
      "eval_steps_per_second": 2.789,
      "step": 2050
    },
    {
      "epoch": 0.520070689219894,
      "grad_norm": 0.8359307050704956,
      "learning_rate": 2.3820000000000002e-05,
      "loss": 0.0759,
      "step": 2060
    },
    {
      "epoch": 0.522595304216107,
      "grad_norm": 1.6299511194229126,
      "learning_rate": 2.379e-05,
      "loss": 0.076,
      "step": 2070
    },
    {
      "epoch": 0.5251199192123202,
      "grad_norm": 0.6880617737770081,
      "learning_rate": 2.3760000000000003e-05,
      "loss": 0.0745,
      "step": 2080
    },
    {
      "epoch": 0.5276445342085332,
      "grad_norm": 0.7822777032852173,
      "learning_rate": 2.373e-05,
      "loss": 0.0697,
      "step": 2090
    },
    {
      "epoch": 0.5301691492047462,
      "grad_norm": 0.7941886782646179,
      "learning_rate": 2.37e-05,
      "loss": 0.0685,
      "step": 2100
    },
    {
      "epoch": 0.5301691492047462,
      "eval_f1": 0.6104315862855695,
      "eval_loss": 0.04854836314916611,
      "eval_runtime": 1154.0649,
      "eval_samples_per_second": 178.727,
      "eval_steps_per_second": 2.793,
      "step": 2100
    },
    {
      "epoch": 0.5326937642009594,
      "grad_norm": 0.948130190372467,
      "learning_rate": 2.3670000000000002e-05,
      "loss": 0.0706,
      "step": 2110
    },
    {
      "epoch": 0.5352183791971724,
      "grad_norm": 0.959032416343689,
      "learning_rate": 2.364e-05,
      "loss": 0.0684,
      "step": 2120
    },
    {
      "epoch": 0.5377429941933856,
      "grad_norm": 1.1859666109085083,
      "learning_rate": 2.3610000000000003e-05,
      "loss": 0.0757,
      "step": 2130
    },
    {
      "epoch": 0.5402676091895986,
      "grad_norm": 0.9001142978668213,
      "learning_rate": 2.358e-05,
      "loss": 0.079,
      "step": 2140
    },
    {
      "epoch": 0.5427922241858116,
      "grad_norm": 0.47399717569351196,
      "learning_rate": 2.3550000000000003e-05,
      "loss": 0.0726,
      "step": 2150
    },
    {
      "epoch": 0.5427922241858116,
      "eval_f1": 0.611992731677771,
      "eval_loss": 0.05662121623754501,
      "eval_runtime": 1151.3771,
      "eval_samples_per_second": 179.144,
      "eval_steps_per_second": 2.799,
      "step": 2150
    },
    {
      "epoch": 0.5453168391820248,
      "grad_norm": 0.6292353272438049,
      "learning_rate": 2.3520000000000002e-05,
      "loss": 0.0677,
      "step": 2160
    },
    {
      "epoch": 0.5478414541782378,
      "grad_norm": 0.7090362906455994,
      "learning_rate": 2.349e-05,
      "loss": 0.0703,
      "step": 2170
    },
    {
      "epoch": 0.5503660691744509,
      "grad_norm": 0.6082953810691833,
      "learning_rate": 2.3460000000000002e-05,
      "loss": 0.0672,
      "step": 2180
    },
    {
      "epoch": 0.552890684170664,
      "grad_norm": 0.5937643051147461,
      "learning_rate": 2.343e-05,
      "loss": 0.0686,
      "step": 2190
    },
    {
      "epoch": 0.555415299166877,
      "grad_norm": 0.7394770979881287,
      "learning_rate": 2.3400000000000003e-05,
      "loss": 0.0731,
      "step": 2200
    },
    {
      "epoch": 0.555415299166877,
      "eval_f1": 0.6111780293905084,
      "eval_loss": 0.05852247402071953,
      "eval_runtime": 1153.1003,
      "eval_samples_per_second": 178.876,
      "eval_steps_per_second": 2.795,
      "step": 2200
    },
    {
      "epoch": 0.5579399141630901,
      "grad_norm": 0.7641323804855347,
      "learning_rate": 2.337e-05,
      "loss": 0.0732,
      "step": 2210
    },
    {
      "epoch": 0.5604645291593032,
      "grad_norm": 0.8567935824394226,
      "learning_rate": 2.334e-05,
      "loss": 0.0599,
      "step": 2220
    },
    {
      "epoch": 0.5629891441555163,
      "grad_norm": 0.9106941819190979,
      "learning_rate": 2.3310000000000002e-05,
      "loss": 0.0593,
      "step": 2230
    },
    {
      "epoch": 0.5655137591517294,
      "grad_norm": 1.5944632291793823,
      "learning_rate": 2.328e-05,
      "loss": 0.0669,
      "step": 2240
    },
    {
      "epoch": 0.5680383741479424,
      "grad_norm": 0.9120457768440247,
      "learning_rate": 2.3250000000000003e-05,
      "loss": 0.0722,
      "step": 2250
    },
    {
      "epoch": 0.5680383741479424,
      "eval_f1": 0.6139676730710583,
      "eval_loss": 0.05887339636683464,
      "eval_runtime": 1155.9087,
      "eval_samples_per_second": 178.441,
      "eval_steps_per_second": 2.788,
      "step": 2250
    },
    {
      "epoch": 0.5705629891441555,
      "grad_norm": 0.8505953550338745,
      "learning_rate": 2.322e-05,
      "loss": 0.0863,
      "step": 2260
    },
    {
      "epoch": 0.5730876041403686,
      "grad_norm": 0.9573137164115906,
      "learning_rate": 2.319e-05,
      "loss": 0.0712,
      "step": 2270
    },
    {
      "epoch": 0.5756122191365817,
      "grad_norm": 1.230735182762146,
      "learning_rate": 2.3160000000000002e-05,
      "loss": 0.0677,
      "step": 2280
    },
    {
      "epoch": 0.5781368341327947,
      "grad_norm": 1.203621745109558,
      "learning_rate": 2.313e-05,
      "loss": 0.0634,
      "step": 2290
    },
    {
      "epoch": 0.5806614491290079,
      "grad_norm": 1.3590195178985596,
      "learning_rate": 2.3100000000000002e-05,
      "loss": 0.0819,
      "step": 2300
    },
    {
      "epoch": 0.5806614491290079,
      "eval_f1": 0.6121980676328502,
      "eval_loss": 0.050494007766246796,
      "eval_runtime": 1153.6589,
      "eval_samples_per_second": 178.789,
      "eval_steps_per_second": 2.794,
      "step": 2300
    },
    {
      "epoch": 0.5831860641252209,
      "grad_norm": 0.8538402318954468,
      "learning_rate": 2.307e-05,
      "loss": 0.0674,
      "step": 2310
    },
    {
      "epoch": 0.5857106791214339,
      "grad_norm": 1.1863012313842773,
      "learning_rate": 2.304e-05,
      "loss": 0.0665,
      "step": 2320
    },
    {
      "epoch": 0.5882352941176471,
      "grad_norm": 1.0120714902877808,
      "learning_rate": 2.301e-05,
      "loss": 0.0675,
      "step": 2330
    },
    {
      "epoch": 0.5907599091138601,
      "grad_norm": 0.8394482135772705,
      "learning_rate": 2.298e-05,
      "loss": 0.0812,
      "step": 2340
    },
    {
      "epoch": 0.5932845241100733,
      "grad_norm": 0.8855767250061035,
      "learning_rate": 2.2950000000000002e-05,
      "loss": 0.0694,
      "step": 2350
    },
    {
      "epoch": 0.5932845241100733,
      "eval_f1": 0.6101251634597422,
      "eval_loss": 0.053731031715869904,
      "eval_runtime": 1147.8424,
      "eval_samples_per_second": 179.695,
      "eval_steps_per_second": 2.808,
      "step": 2350
    },
    {
      "epoch": 0.5958091391062863,
      "grad_norm": 1.241045594215393,
      "learning_rate": 2.292e-05,
      "loss": 0.0646,
      "step": 2360
    },
    {
      "epoch": 0.5983337541024993,
      "grad_norm": 2.065401315689087,
      "learning_rate": 2.289e-05,
      "loss": 0.0792,
      "step": 2370
    },
    {
      "epoch": 0.6008583690987125,
      "grad_norm": 1.0024877786636353,
      "learning_rate": 2.286e-05,
      "loss": 0.0751,
      "step": 2380
    },
    {
      "epoch": 0.6033829840949255,
      "grad_norm": 0.4943256080150604,
      "learning_rate": 2.283e-05,
      "loss": 0.076,
      "step": 2390
    },
    {
      "epoch": 0.6059075990911386,
      "grad_norm": 1.0907814502716064,
      "learning_rate": 2.2800000000000002e-05,
      "loss": 0.0705,
      "step": 2400
    },
    {
      "epoch": 0.6059075990911386,
      "eval_f1": 0.6130196664177247,
      "eval_loss": 0.06461644172668457,
      "eval_runtime": 1149.8253,
      "eval_samples_per_second": 179.386,
      "eval_steps_per_second": 2.803,
      "step": 2400
    },
    {
      "epoch": 0.6084322140873517,
      "grad_norm": 1.1304162740707397,
      "learning_rate": 2.277e-05,
      "loss": 0.0548,
      "step": 2410
    },
    {
      "epoch": 0.6109568290835647,
      "grad_norm": 1.3394097089767456,
      "learning_rate": 2.274e-05,
      "loss": 0.0607,
      "step": 2420
    },
    {
      "epoch": 0.6134814440797778,
      "grad_norm": 0.5467960834503174,
      "learning_rate": 2.271e-05,
      "loss": 0.0701,
      "step": 2430
    },
    {
      "epoch": 0.6160060590759909,
      "grad_norm": 0.5510517954826355,
      "learning_rate": 2.268e-05,
      "loss": 0.0725,
      "step": 2440
    },
    {
      "epoch": 0.618530674072204,
      "grad_norm": 0.7682734131813049,
      "learning_rate": 2.265e-05,
      "loss": 0.0702,
      "step": 2450
    },
    {
      "epoch": 0.618530674072204,
      "eval_f1": 0.6124447065762312,
      "eval_loss": 0.046234920620918274,
      "eval_runtime": 1146.4615,
      "eval_samples_per_second": 179.912,
      "eval_steps_per_second": 2.811,
      "step": 2450
    },
    {
      "epoch": 0.6210552890684171,
      "grad_norm": 0.7578818798065186,
      "learning_rate": 2.262e-05,
      "loss": 0.0703,
      "step": 2460
    },
    {
      "epoch": 0.6235799040646302,
      "grad_norm": 0.7244108319282532,
      "learning_rate": 2.2590000000000002e-05,
      "loss": 0.0635,
      "step": 2470
    },
    {
      "epoch": 0.6261045190608432,
      "grad_norm": 1.1047908067703247,
      "learning_rate": 2.256e-05,
      "loss": 0.0614,
      "step": 2480
    },
    {
      "epoch": 0.6286291340570563,
      "grad_norm": 1.0824987888336182,
      "learning_rate": 2.253e-05,
      "loss": 0.081,
      "step": 2490
    },
    {
      "epoch": 0.6311537490532694,
      "grad_norm": 1.9344598054885864,
      "learning_rate": 2.25e-05,
      "loss": 0.0709,
      "step": 2500
    },
    {
      "epoch": 0.6311537490532694,
      "eval_f1": 0.6133951445650848,
      "eval_loss": 0.04044894501566887,
      "eval_runtime": 1148.0724,
      "eval_samples_per_second": 179.659,
      "eval_steps_per_second": 2.807,
      "step": 2500
    },
    {
      "epoch": 0.6336783640494824,
      "grad_norm": 1.2797091007232666,
      "learning_rate": 2.247e-05,
      "loss": 0.072,
      "step": 2510
    },
    {
      "epoch": 0.6362029790456956,
      "grad_norm": 0.7228933572769165,
      "learning_rate": 2.2440000000000002e-05,
      "loss": 0.071,
      "step": 2520
    },
    {
      "epoch": 0.6387275940419086,
      "grad_norm": 0.9655591249465942,
      "learning_rate": 2.241e-05,
      "loss": 0.0611,
      "step": 2530
    },
    {
      "epoch": 0.6412522090381216,
      "grad_norm": 0.9924450516700745,
      "learning_rate": 2.238e-05,
      "loss": 0.0676,
      "step": 2540
    },
    {
      "epoch": 0.6437768240343348,
      "grad_norm": 1.12591552734375,
      "learning_rate": 2.235e-05,
      "loss": 0.0804,
      "step": 2550
    },
    {
      "epoch": 0.6437768240343348,
      "eval_f1": 0.612305676335696,
      "eval_loss": 0.04778852313756943,
      "eval_runtime": 1160.4576,
      "eval_samples_per_second": 177.742,
      "eval_steps_per_second": 2.777,
      "step": 2550
    },
    {
      "epoch": 0.6463014390305478,
      "grad_norm": 0.7478006482124329,
      "learning_rate": 2.232e-05,
      "loss": 0.0638,
      "step": 2560
    },
    {
      "epoch": 0.648826054026761,
      "grad_norm": 0.7661213874816895,
      "learning_rate": 2.2290000000000002e-05,
      "loss": 0.0632,
      "step": 2570
    },
    {
      "epoch": 0.651350669022974,
      "grad_norm": 0.9824168086051941,
      "learning_rate": 2.226e-05,
      "loss": 0.0602,
      "step": 2580
    },
    {
      "epoch": 0.653875284019187,
      "grad_norm": 1.1700901985168457,
      "learning_rate": 2.223e-05,
      "loss": 0.0714,
      "step": 2590
    },
    {
      "epoch": 0.6563998990154002,
      "grad_norm": 0.8846214413642883,
      "learning_rate": 2.22e-05,
      "loss": 0.0666,
      "step": 2600
    },
    {
      "epoch": 0.6563998990154002,
      "eval_f1": 0.6104417670682731,
      "eval_loss": 0.04546576738357544,
      "eval_runtime": 1160.1326,
      "eval_samples_per_second": 177.792,
      "eval_steps_per_second": 2.778,
      "step": 2600
    },
    {
      "epoch": 0.6589245140116132,
      "grad_norm": 0.7641239166259766,
      "learning_rate": 2.217e-05,
      "loss": 0.058,
      "step": 2610
    },
    {
      "epoch": 0.6614491290078263,
      "grad_norm": 0.5828648209571838,
      "learning_rate": 2.214e-05,
      "loss": 0.0686,
      "step": 2620
    },
    {
      "epoch": 0.6639737440040394,
      "grad_norm": 0.6906914710998535,
      "learning_rate": 2.211e-05,
      "loss": 0.0764,
      "step": 2630
    },
    {
      "epoch": 0.6664983590002524,
      "grad_norm": 1.3137489557266235,
      "learning_rate": 2.208e-05,
      "loss": 0.0768,
      "step": 2640
    },
    {
      "epoch": 0.6690229739964655,
      "grad_norm": 0.863865077495575,
      "learning_rate": 2.205e-05,
      "loss": 0.0749,
      "step": 2650
    },
    {
      "epoch": 0.6690229739964655,
      "eval_f1": 0.6131900703964431,
      "eval_loss": 0.04790908098220825,
      "eval_runtime": 1162.4462,
      "eval_samples_per_second": 177.438,
      "eval_steps_per_second": 2.773,
      "step": 2650
    },
    {
      "epoch": 0.6715475889926786,
      "grad_norm": 0.9182652235031128,
      "learning_rate": 2.202e-05,
      "loss": 0.0625,
      "step": 2660
    },
    {
      "epoch": 0.6740722039888917,
      "grad_norm": 1.4961283206939697,
      "learning_rate": 2.199e-05,
      "loss": 0.0726,
      "step": 2670
    },
    {
      "epoch": 0.6765968189851048,
      "grad_norm": 0.7803681492805481,
      "learning_rate": 2.196e-05,
      "loss": 0.0669,
      "step": 2680
    },
    {
      "epoch": 0.6791214339813179,
      "grad_norm": 1.0371824502944946,
      "learning_rate": 2.193e-05,
      "loss": 0.0566,
      "step": 2690
    },
    {
      "epoch": 0.6816460489775309,
      "grad_norm": 1.1832714080810547,
      "learning_rate": 2.19e-05,
      "loss": 0.067,
      "step": 2700
    },
    {
      "epoch": 0.6816460489775309,
      "eval_f1": 0.6132461161079312,
      "eval_loss": 0.055793602019548416,
      "eval_runtime": 1161.8914,
      "eval_samples_per_second": 177.523,
      "eval_steps_per_second": 2.774,
      "step": 2700
    },
    {
      "epoch": 0.684170663973744,
      "grad_norm": 0.7899573445320129,
      "learning_rate": 2.187e-05,
      "loss": 0.0763,
      "step": 2710
    },
    {
      "epoch": 0.6866952789699571,
      "grad_norm": 1.4638808965682983,
      "learning_rate": 2.184e-05,
      "loss": 0.0768,
      "step": 2720
    },
    {
      "epoch": 0.6892198939661701,
      "grad_norm": 0.7547538876533508,
      "learning_rate": 2.181e-05,
      "loss": 0.0761,
      "step": 2730
    },
    {
      "epoch": 0.6917445089623833,
      "grad_norm": 0.5143932700157166,
      "learning_rate": 2.178e-05,
      "loss": 0.0808,
      "step": 2740
    },
    {
      "epoch": 0.6942691239585963,
      "grad_norm": 1.011730432510376,
      "learning_rate": 2.175e-05,
      "loss": 0.068,
      "step": 2750
    },
    {
      "epoch": 0.6942691239585963,
      "eval_f1": 0.6108202443280978,
      "eval_loss": 0.053855251520872116,
      "eval_runtime": 1160.3338,
      "eval_samples_per_second": 177.761,
      "eval_steps_per_second": 2.778,
      "step": 2750
    },
    {
      "epoch": 0.6942691239585963,
      "step": 2750,
      "total_flos": 1.3639932886745088e+19,
      "train_loss": 0.019194319985129618,
      "train_runtime": 18605.0451,
      "train_samples_per_second": 34.399,
      "train_steps_per_second": 0.537
    }
  ],
  "logging_steps": 10,
  "max_steps": 10000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 3,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.3639932886745088e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}