{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.984,
  "eval_steps": 1,
  "global_step": 124,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.016,
      "grad_norm": 52.25,
      "learning_rate": 2.5e-05,
      "loss": 1.2424,
      "step": 1
    },
    {
      "epoch": 0.016,
      "eval_exact_match": 0.02040816326530612,
      "eval_f1_a": 0.5754985754985755,
      "eval_f1_m": 0.30537274516866364,
      "eval_loss": 1.8040205240249634,
      "eval_runtime": 12.5246,
      "eval_samples_per_second": 19.961,
      "eval_steps_per_second": 2.555,
      "step": 1
    },
    {
      "epoch": 0.032,
      "grad_norm": 74.5,
      "learning_rate": 5e-05,
      "loss": 1.5616,
      "step": 2
    },
    {
      "epoch": 0.032,
      "eval_exact_match": 0.02040816326530612,
      "eval_f1_a": 0.5754985754985755,
      "eval_f1_m": 0.30537274516866364,
      "eval_loss": 1.6737734079360962,
      "eval_runtime": 12.5202,
      "eval_samples_per_second": 19.968,
      "eval_steps_per_second": 2.556,
      "step": 2
    },
    {
      "epoch": 0.048,
      "grad_norm": 81.5,
      "learning_rate": 4.959016393442623e-05,
      "loss": 1.9327,
      "step": 3
    },
    {
      "epoch": 0.048,
      "eval_exact_match": 0.02040816326530612,
      "eval_f1_a": 0.5798816568047337,
      "eval_f1_m": 0.3293744237621789,
      "eval_loss": 1.0940678119659424,
      "eval_runtime": 12.4172,
      "eval_samples_per_second": 20.133,
      "eval_steps_per_second": 2.577,
      "step": 3
    },
    {
      "epoch": 0.064,
      "grad_norm": 24.5,
      "learning_rate": 4.918032786885246e-05,
      "loss": 0.7502,
      "step": 4
    },
    {
      "epoch": 0.064,
      "eval_exact_match": 0.04081632653061224,
      "eval_f1_a": 0.326797385620915,
      "eval_f1_m": 0.40487871139131637,
      "eval_loss": 0.7598310708999634,
      "eval_runtime": 12.5191,
      "eval_samples_per_second": 19.969,
      "eval_steps_per_second": 2.556,
      "step": 4
    },
    {
      "epoch": 0.08,
      "grad_norm": 43.5,
      "learning_rate": 4.8770491803278687e-05,
      "loss": 1.2439,
      "step": 5
    },
    {
      "epoch": 0.08,
      "eval_exact_match": 0.0,
      "eval_f1_a": 0.11475409836065575,
      "eval_f1_m": 0.3452577201376722,
      "eval_loss": 0.8241621255874634,
      "eval_runtime": 12.412,
      "eval_samples_per_second": 20.142,
      "eval_steps_per_second": 2.578,
      "step": 5
    },
    {
      "epoch": 0.096,
      "grad_norm": 21.625,
      "learning_rate": 4.836065573770492e-05,
      "loss": 0.9487,
      "step": 6
    },
    {
      "epoch": 0.096,
      "eval_exact_match": 0.02040816326530612,
      "eval_f1_a": 0.05555555555555556,
      "eval_f1_m": 0.35745697213284255,
      "eval_loss": 0.8299877643585205,
      "eval_runtime": 12.4102,
      "eval_samples_per_second": 20.145,
      "eval_steps_per_second": 2.579,
      "step": 6
    },
    {
      "epoch": 0.112,
      "grad_norm": 114.0,
      "learning_rate": 4.795081967213115e-05,
      "loss": 1.9374,
      "step": 7
    },
    {
      "epoch": 0.112,
      "eval_exact_match": 0.0,
      "eval_f1_a": 0.08620689655172413,
      "eval_f1_m": 0.3505331636484098,
      "eval_loss": 0.7635751962661743,
      "eval_runtime": 12.5057,
      "eval_samples_per_second": 19.991,
      "eval_steps_per_second": 2.559,
      "step": 7
    },
    {
      "epoch": 0.128,
      "grad_norm": 49.25,
      "learning_rate": 4.754098360655738e-05,
      "loss": 0.9383,
      "step": 8
    },
    {
      "epoch": 0.128,
      "eval_exact_match": 0.12244897959183673,
      "eval_f1_a": 0.37499999999999994,
      "eval_f1_m": 0.4638421102706817,
      "eval_loss": 0.7212109565734863,
      "eval_runtime": 12.5257,
      "eval_samples_per_second": 19.959,
      "eval_steps_per_second": 2.555,
      "step": 8
    },
    {
      "epoch": 0.144,
      "grad_norm": 10.375,
      "learning_rate": 4.713114754098361e-05,
      "loss": 0.6205,
      "step": 9
    },
    {
      "epoch": 0.144,
      "eval_exact_match": 0.04081632653061224,
      "eval_f1_a": 0.46413502109704646,
      "eval_f1_m": 0.39599618069005815,
      "eval_loss": 0.7557714581489563,
      "eval_runtime": 12.5194,
      "eval_samples_per_second": 19.969,
      "eval_steps_per_second": 2.556,
      "step": 9
    },
    {
      "epoch": 0.16,
      "grad_norm": 11.8125,
      "learning_rate": 4.672131147540984e-05,
      "loss": 0.9441,
      "step": 10
    },
    {
      "epoch": 0.16,
      "eval_exact_match": 0.02040816326530612,
      "eval_f1_a": 0.5454545454545455,
      "eval_f1_m": 0.3639190741231557,
      "eval_loss": 0.8026367425918579,
      "eval_runtime": 12.4166,
      "eval_samples_per_second": 20.134,
      "eval_steps_per_second": 2.577,
      "step": 10
    },
    {
      "epoch": 0.176,
      "grad_norm": 32.25,
      "learning_rate": 4.631147540983607e-05,
      "loss": 0.9161,
      "step": 11
    },
    {
      "epoch": 0.176,
      "eval_exact_match": 0.02040816326530612,
      "eval_f1_a": 0.5608108108108107,
      "eval_f1_m": 0.36614026563006147,
      "eval_loss": 0.8101338148117065,
      "eval_runtime": 12.5325,
      "eval_samples_per_second": 19.948,
      "eval_steps_per_second": 2.553,
      "step": 11
    },
    {
      "epoch": 0.192,
      "grad_norm": 36.25,
      "learning_rate": 4.59016393442623e-05,
      "loss": 0.8093,
      "step": 12
    },
    {
      "epoch": 0.192,
      "eval_exact_match": 0.02040816326530612,
      "eval_f1_a": 0.5429553264604812,
      "eval_f1_m": 0.35214683275907754,
      "eval_loss": 0.7952831983566284,
      "eval_runtime": 12.4346,
      "eval_samples_per_second": 20.105,
      "eval_steps_per_second": 2.573,
      "step": 12
    },
    {
      "epoch": 0.208,
      "grad_norm": 11.0,
      "learning_rate": 4.549180327868853e-05,
      "loss": 0.6146,
      "step": 13
    },
    {
      "epoch": 0.208,
      "eval_exact_match": 0.02040816326530612,
      "eval_f1_a": 0.5441696113074205,
      "eval_f1_m": 0.3687700621374089,
      "eval_loss": 0.7707070112228394,
      "eval_runtime": 12.5246,
      "eval_samples_per_second": 19.961,
      "eval_steps_per_second": 2.555,
      "step": 13
    },
    {
      "epoch": 0.224,
      "grad_norm": 14.625,
      "learning_rate": 4.508196721311476e-05,
      "loss": 0.7804,
      "step": 14
    },
    {
      "epoch": 0.224,
      "eval_exact_match": 0.02040816326530612,
      "eval_f1_a": 0.5376344086021505,
      "eval_f1_m": 0.36071490867409234,
      "eval_loss": 0.7593163847923279,
      "eval_runtime": 12.5202,
      "eval_samples_per_second": 19.968,
      "eval_steps_per_second": 2.556,
      "step": 14
    },
    {
      "epoch": 0.24,
      "grad_norm": 12.6875,
      "learning_rate": 4.467213114754098e-05,
      "loss": 0.75,
      "step": 15
    },
    {
      "epoch": 0.24,
      "eval_exact_match": 0.0,
      "eval_f1_a": 0.5300353356890458,
      "eval_f1_m": 0.3416269105044615,
      "eval_loss": 0.7546816468238831,
      "eval_runtime": 12.53,
      "eval_samples_per_second": 19.952,
      "eval_steps_per_second": 2.554,
      "step": 15
    },
    {
      "epoch": 0.256,
      "grad_norm": 10.4375,
      "learning_rate": 4.426229508196721e-05,
      "loss": 0.9339,
      "step": 16
    },
    {
      "epoch": 0.256,
      "eval_exact_match": 0.0,
      "eval_f1_a": 0.5454545454545454,
      "eval_f1_m": 0.3630654492899391,
      "eval_loss": 0.7366230487823486,
      "eval_runtime": 12.5191,
      "eval_samples_per_second": 19.97,
      "eval_steps_per_second": 2.556,
      "step": 16
    },
    {
      "epoch": 0.272,
      "grad_norm": 36.75,
      "learning_rate": 4.3852459016393444e-05,
      "loss": 0.6286,
      "step": 17
    },
    {
      "epoch": 0.272,
      "eval_exact_match": 0.02040816326530612,
      "eval_f1_a": 0.528,
      "eval_f1_m": 0.38775957602488215,
      "eval_loss": 0.712777316570282,
      "eval_runtime": 12.5294,
      "eval_samples_per_second": 19.953,
      "eval_steps_per_second": 2.554,
      "step": 17
    },
    {
      "epoch": 0.288,
      "grad_norm": 31.75,
      "learning_rate": 4.3442622950819674e-05,
      "loss": 0.7253,
      "step": 18
    },
    {
      "epoch": 0.288,
      "eval_exact_match": 0.02040816326530612,
      "eval_f1_a": 0.5301204819277109,
      "eval_f1_m": 0.39053531502511096,
      "eval_loss": 0.7025195360183716,
      "eval_runtime": 12.5183,
      "eval_samples_per_second": 19.971,
      "eval_steps_per_second": 2.556,
      "step": 18
    },
    {
      "epoch": 0.304,
      "grad_norm": 22.0,
      "learning_rate": 4.3032786885245904e-05,
      "loss": 0.5984,
      "step": 19
    },
    {
      "epoch": 0.304,
      "eval_exact_match": 0.04081632653061224,
      "eval_f1_a": 0.5045045045045046,
      "eval_f1_m": 0.4079482875401243,
      "eval_loss": 0.6883164048194885,
      "eval_runtime": 12.5119,
      "eval_samples_per_second": 19.981,
      "eval_steps_per_second": 2.558,
      "step": 19
    },
    {
      "epoch": 0.32,
      "grad_norm": 24.75,
      "learning_rate": 4.262295081967213e-05,
      "loss": 0.6453,
      "step": 20
    },
    {
      "epoch": 0.32,
      "eval_exact_match": 0.04081632653061224,
      "eval_f1_a": 0.4931506849315069,
      "eval_f1_m": 0.4012221905079048,
      "eval_loss": 0.6845566630363464,
      "eval_runtime": 12.5289,
      "eval_samples_per_second": 19.954,
      "eval_steps_per_second": 2.554,
      "step": 20
    },
    {
      "epoch": 0.336,
      "grad_norm": 8.3125,
      "learning_rate": 4.2213114754098365e-05,
      "loss": 0.58,
      "step": 21
    },
    {
      "epoch": 0.336,
      "eval_exact_match": 0.02040816326530612,
      "eval_f1_a": 0.4859813084112149,
      "eval_f1_m": 0.4041376424029486,
      "eval_loss": 0.6770976781845093,
      "eval_runtime": 12.4164,
      "eval_samples_per_second": 20.135,
      "eval_steps_per_second": 2.577,
      "step": 21
    },
    {
      "epoch": 0.352,
      "grad_norm": 28.625,
      "learning_rate": 4.1803278688524595e-05,
      "loss": 0.6812,
      "step": 22
    },
    {
      "epoch": 0.352,
      "eval_exact_match": 0.0,
      "eval_f1_a": 0.509090909090909,
      "eval_f1_m": 0.40605698156718567,
      "eval_loss": 0.6734863519668579,
      "eval_runtime": 12.5159,
      "eval_samples_per_second": 19.975,
      "eval_steps_per_second": 2.557,
      "step": 22
    },
    {
      "epoch": 0.368,
      "grad_norm": 18.25,
      "learning_rate": 4.1393442622950826e-05,
      "loss": 0.4762,
      "step": 23
    },
    {
      "epoch": 0.368,
      "eval_exact_match": 0.02040816326530612,
      "eval_f1_a": 0.4976076555023924,
      "eval_f1_m": 0.4171717738044269,
      "eval_loss": 0.6683046817779541,
      "eval_runtime": 12.5214,
      "eval_samples_per_second": 19.966,
      "eval_steps_per_second": 2.556,
      "step": 23
    },
    {
      "epoch": 0.384,
      "grad_norm": 9.125,
      "learning_rate": 4.098360655737705e-05,
      "loss": 0.6075,
      "step": 24
    },
    {
      "epoch": 0.384,
      "eval_exact_match": 0.02040816326530612,
      "eval_f1_a": 0.48756218905472637,
      "eval_f1_m": 0.41513432878979106,
      "eval_loss": 0.6622753739356995,
      "eval_runtime": 12.4157,
      "eval_samples_per_second": 20.136,
      "eval_steps_per_second": 2.577,
      "step": 24
    },
    {
      "epoch": 0.4,
      "grad_norm": 10.1875,
      "learning_rate": 4.057377049180328e-05,
      "loss": 0.6313,
      "step": 25
    },
    {
      "epoch": 0.4,
      "eval_exact_match": 0.04081632653061224,
      "eval_f1_a": 0.48087431693989074,
      "eval_f1_m": 0.4350705716852176,
      "eval_loss": 0.6575761437416077,
      "eval_runtime": 12.5232,
      "eval_samples_per_second": 19.963,
      "eval_steps_per_second": 2.555,
      "step": 25
    },
    {
      "epoch": 0.416,
      "grad_norm": 14.0,
      "learning_rate": 4.016393442622951e-05,
      "loss": 0.6486,
      "step": 26
    },
    {
      "epoch": 0.416,
      "eval_exact_match": 0.04081632653061224,
      "eval_f1_a": 0.4606741573033708,
      "eval_f1_m": 0.44082196315089484,
      "eval_loss": 0.6529648303985596,
      "eval_runtime": 12.5257,
      "eval_samples_per_second": 19.959,
      "eval_steps_per_second": 2.555,
      "step": 26
    },
    {
      "epoch": 0.432,
      "grad_norm": 14.9375,
      "learning_rate": 3.975409836065574e-05,
      "loss": 0.8356,
      "step": 27
    },
    {
      "epoch": 0.432,
      "eval_exact_match": 0.04081632653061224,
      "eval_f1_a": 0.44311377245508976,
      "eval_f1_m": 0.4505972219057453,
      "eval_loss": 0.6500390768051147,
      "eval_runtime": 12.5271,
      "eval_samples_per_second": 19.957,
      "eval_steps_per_second": 2.554,
      "step": 27
    },
    {
      "epoch": 0.448,
      "grad_norm": 9.125,
      "learning_rate": 3.934426229508197e-05,
      "loss": 0.6429,
      "step": 28
    },
    {
      "epoch": 0.448,
      "eval_exact_match": 0.061224489795918366,
      "eval_f1_a": 0.43636363636363634,
      "eval_f1_m": 0.4468071344421886,
      "eval_loss": 0.6493672132492065,
      "eval_runtime": 12.5212,
      "eval_samples_per_second": 19.966,
      "eval_steps_per_second": 2.556,
      "step": 28
    },
    {
      "epoch": 0.464,
      "grad_norm": 15.8125,
      "learning_rate": 3.89344262295082e-05,
      "loss": 0.6248,
      "step": 29
    },
    {
      "epoch": 0.464,
      "eval_exact_match": 0.10204081632653061,
      "eval_f1_a": 0.38926174496644295,
      "eval_f1_m": 0.4801647632079804,
      "eval_loss": 0.6509453058242798,
      "eval_runtime": 12.535,
      "eval_samples_per_second": 19.944,
      "eval_steps_per_second": 2.553,
      "step": 29
    },
    {
      "epoch": 0.48,
      "grad_norm": 24.5,
      "learning_rate": 3.8524590163934424e-05,
      "loss": 0.4855,
      "step": 30
    },
    {
      "epoch": 0.48,
      "eval_exact_match": 0.061224489795918366,
      "eval_f1_a": 0.3283582089552239,
      "eval_f1_m": 0.46286459638800564,
      "eval_loss": 0.6590214967727661,
      "eval_runtime": 12.5171,
      "eval_samples_per_second": 19.973,
      "eval_steps_per_second": 2.557,
      "step": 30
    },
    {
      "epoch": 0.496,
      "grad_norm": 31.375,
      "learning_rate": 3.8114754098360655e-05,
      "loss": 0.7308,
      "step": 31
    },
    {
      "epoch": 0.496,
      "eval_exact_match": 0.02040816326530612,
      "eval_f1_a": 0.30534351145038163,
      "eval_f1_m": 0.4485788821022914,
      "eval_loss": 0.6624687314033508,
      "eval_runtime": 12.4214,
      "eval_samples_per_second": 20.127,
      "eval_steps_per_second": 2.576,
      "step": 31
    },
    {
      "epoch": 0.512,
      "grad_norm": 28.125,
      "learning_rate": 3.7704918032786885e-05,
      "loss": 0.7749,
      "step": 32
    },
    {
      "epoch": 0.512,
      "eval_exact_match": 0.04081632653061224,
      "eval_f1_a": 0.31007751937984496,
      "eval_f1_m": 0.4561152299747737,
      "eval_loss": 0.6629375219345093,
      "eval_runtime": 12.5333,
      "eval_samples_per_second": 19.947,
      "eval_steps_per_second": 2.553,
      "step": 32
    },
    {
      "epoch": 0.528,
      "grad_norm": 25.125,
      "learning_rate": 3.729508196721312e-05,
      "loss": 0.7449,
      "step": 33
    },
    {
      "epoch": 0.528,
      "eval_exact_match": 0.04081632653061224,
      "eval_f1_a": 0.34782608695652173,
      "eval_f1_m": 0.45786965468838214,
      "eval_loss": 0.6554140448570251,
      "eval_runtime": 12.5362,
      "eval_samples_per_second": 19.942,
      "eval_steps_per_second": 2.553,
      "step": 33
    },
    {
      "epoch": 0.544,
      "grad_norm": 12.1875,
      "learning_rate": 3.6885245901639346e-05,
      "loss": 0.7871,
      "step": 34
    },
    {
      "epoch": 0.544,
      "eval_exact_match": 0.02040816326530612,
      "eval_f1_a": 0.3382352941176471,
      "eval_f1_m": 0.4546140667389167,
      "eval_loss": 0.6557744145393372,
      "eval_runtime": 12.5339,
      "eval_samples_per_second": 19.946,
      "eval_steps_per_second": 2.553,
      "step": 34
    },
    {
      "epoch": 0.56,
      "grad_norm": 10.5625,
      "learning_rate": 3.6475409836065576e-05,
      "loss": 0.4428,
      "step": 35
    },
    {
      "epoch": 0.56,
      "eval_exact_match": 0.02040816326530612,
      "eval_f1_a": 0.2900763358778626,
      "eval_f1_m": 0.4422134787981126,
      "eval_loss": 0.6567177772521973,
      "eval_runtime": 12.5258,
      "eval_samples_per_second": 19.959,
      "eval_steps_per_second": 2.555,
      "step": 35
    },
    {
      "epoch": 0.576,
      "grad_norm": 37.75,
      "learning_rate": 3.6065573770491806e-05,
      "loss": 0.8155,
      "step": 36
    },
    {
      "epoch": 0.576,
      "eval_exact_match": 0.02040816326530612,
      "eval_f1_a": 0.3333333333333333,
      "eval_f1_m": 0.4582583816077213,
      "eval_loss": 0.6531835794448853,
      "eval_runtime": 12.5226,
      "eval_samples_per_second": 19.964,
      "eval_steps_per_second": 2.555,
      "step": 36
    },
    {
      "epoch": 0.592,
      "grad_norm": 6.53125,
      "learning_rate": 3.5655737704918037e-05,
      "loss": 0.6122,
      "step": 37
    },
    {
      "epoch": 0.592,
      "eval_exact_match": 0.02040816326530612,
      "eval_f1_a": 0.3230769230769231,
      "eval_f1_m": 0.45850133593230824,
      "eval_loss": 0.6510556936264038,
      "eval_runtime": 12.5321,
      "eval_samples_per_second": 19.949,
      "eval_steps_per_second": 2.553,
      "step": 37
    },
    {
      "epoch": 0.608,
      "grad_norm": 20.0,
      "learning_rate": 3.524590163934427e-05,
      "loss": 0.6843,
      "step": 38
    },
    {
      "epoch": 0.608,
      "eval_exact_match": 0.02040816326530612,
      "eval_f1_a": 0.3382352941176471,
      "eval_f1_m": 0.4538476562786286,
      "eval_loss": 0.6473007798194885,
      "eval_runtime": 12.4172,
      "eval_samples_per_second": 20.133,
      "eval_steps_per_second": 2.577,
      "step": 38
    },
    {
      "epoch": 0.624,
      "grad_norm": 22.25,
      "learning_rate": 3.483606557377049e-05,
      "loss": 0.6606,
      "step": 39
    },
    {
      "epoch": 0.624,
      "eval_exact_match": 0.061224489795918366,
      "eval_f1_a": 0.42580645161290326,
      "eval_f1_m": 0.48005761412123965,
      "eval_loss": 0.6421992182731628,
      "eval_runtime": 12.5309,
      "eval_samples_per_second": 19.951,
      "eval_steps_per_second": 2.554,
      "step": 39
    },
    {
      "epoch": 0.64,
      "grad_norm": 19.5,
      "learning_rate": 3.442622950819672e-05,
      "loss": 0.6284,
      "step": 40
    },
    {
      "epoch": 0.64,
      "eval_exact_match": 0.08163265306122448,
      "eval_f1_a": 0.4331210191082803,
      "eval_f1_m": 0.4849652914778965,
      "eval_loss": 0.6393261551856995,
      "eval_runtime": 12.4078,
      "eval_samples_per_second": 20.149,
      "eval_steps_per_second": 2.579,
      "step": 40
    },
    {
      "epoch": 0.656,
      "grad_norm": 24.0,
      "learning_rate": 3.401639344262295e-05,
      "loss": 0.6525,
      "step": 41
    },
    {
      "epoch": 0.656,
      "eval_exact_match": 0.10204081632653061,
      "eval_f1_a": 0.45977011494252873,
      "eval_f1_m": 0.4644901883497321,
      "eval_loss": 0.6380175948143005,
      "eval_runtime": 12.5392,
      "eval_samples_per_second": 19.938,
      "eval_steps_per_second": 2.552,
      "step": 41
    },
    {
      "epoch": 0.672,
      "grad_norm": 26.875,
      "learning_rate": 3.360655737704918e-05,
      "loss": 0.7537,
      "step": 42
    },
    {
      "epoch": 0.672,
      "eval_exact_match": 0.08163265306122448,
      "eval_f1_a": 0.4816753926701571,
      "eval_f1_m": 0.4603846000604704,
      "eval_loss": 0.6373457312583923,
      "eval_runtime": 12.5124,
      "eval_samples_per_second": 19.98,
      "eval_steps_per_second": 2.557,
      "step": 42
    },
    {
      "epoch": 0.688,
      "grad_norm": 8.5625,
      "learning_rate": 3.319672131147541e-05,
      "loss": 0.6196,
      "step": 43
    },
    {
      "epoch": 0.688,
      "eval_exact_match": 0.08163265306122448,
      "eval_f1_a": 0.5118483412322276,
      "eval_f1_m": 0.4378736642842286,
      "eval_loss": 0.641525387763977,
      "eval_runtime": 12.5135,
      "eval_samples_per_second": 19.978,
      "eval_steps_per_second": 2.557,
      "step": 43
    },
    {
      "epoch": 0.704,
      "grad_norm": 11.5625,
      "learning_rate": 3.2786885245901635e-05,
      "loss": 0.7105,
      "step": 44
    },
    {
      "epoch": 0.704,
      "eval_exact_match": 0.08163265306122448,
      "eval_f1_a": 0.5407725321888414,
      "eval_f1_m": 0.42079032079032075,
      "eval_loss": 0.6511328220367432,
      "eval_runtime": 12.5196,
      "eval_samples_per_second": 19.969,
      "eval_steps_per_second": 2.556,
      "step": 44
    },
    {
      "epoch": 0.72,
      "grad_norm": 41.75,
      "learning_rate": 3.237704918032787e-05,
      "loss": 0.7902,
      "step": 45
    },
    {
      "epoch": 0.72,
      "eval_exact_match": 0.061224489795918366,
      "eval_f1_a": 0.5338983050847457,
      "eval_f1_m": 0.4120925559701069,
      "eval_loss": 0.6530937552452087,
      "eval_runtime": 12.5207,
      "eval_samples_per_second": 19.967,
      "eval_steps_per_second": 2.556,
      "step": 45
    },
    {
      "epoch": 0.736,
      "grad_norm": 30.25,
      "learning_rate": 3.19672131147541e-05,
      "loss": 0.621,
      "step": 46
    },
    {
      "epoch": 0.736,
      "eval_exact_match": 0.061224489795918366,
      "eval_f1_a": 0.5423728813559323,
      "eval_f1_m": 0.4205959573306511,
      "eval_loss": 0.6510214805603027,
      "eval_runtime": 12.5251,
      "eval_samples_per_second": 19.96,
      "eval_steps_per_second": 2.555,
      "step": 46
    },
    {
      "epoch": 0.752,
      "grad_norm": 13.0,
      "learning_rate": 3.155737704918033e-05,
      "loss": 0.8848,
      "step": 47
    },
    {
      "epoch": 0.752,
      "eval_exact_match": 0.061224489795918366,
      "eval_f1_a": 0.5485232067510549,
      "eval_f1_m": 0.42849197287972796,
      "eval_loss": 0.6499238014221191,
      "eval_runtime": 12.5214,
      "eval_samples_per_second": 19.966,
      "eval_steps_per_second": 2.556,
      "step": 47
    },
    {
      "epoch": 0.768,
      "grad_norm": 51.25,
      "learning_rate": 3.114754098360656e-05,
      "loss": 0.876,
      "step": 48
    },
    {
      "epoch": 0.768,
      "eval_exact_match": 0.08163265306122448,
      "eval_f1_a": 0.5391304347826087,
      "eval_f1_m": 0.4377896479937296,
      "eval_loss": 0.6470175981521606,
      "eval_runtime": 12.5243,
      "eval_samples_per_second": 19.961,
      "eval_steps_per_second": 2.555,
      "step": 48
    },
    {
      "epoch": 0.784,
      "grad_norm": 41.25,
      "learning_rate": 3.073770491803279e-05,
      "loss": 0.7048,
      "step": 49
    },
    {
      "epoch": 0.784,
      "eval_exact_match": 0.08163265306122448,
      "eval_f1_a": 0.5302325581395348,
      "eval_f1_m": 0.4584981911512524,
      "eval_loss": 0.6386250257492065,
      "eval_runtime": 12.5286,
      "eval_samples_per_second": 19.954,
      "eval_steps_per_second": 2.554,
      "step": 49
    },
    {
      "epoch": 0.8,
      "grad_norm": 17.625,
      "learning_rate": 3.0327868852459017e-05,
      "loss": 0.8217,
      "step": 50
    },
    {
      "epoch": 0.8,
      "eval_exact_match": 0.061224489795918366,
      "eval_f1_a": 0.5258215962441315,
      "eval_f1_m": 0.45543696666145644,
      "eval_loss": 0.6384491920471191,
      "eval_runtime": 12.5207,
      "eval_samples_per_second": 19.967,
      "eval_steps_per_second": 2.556,
      "step": 50
    },
    {
      "epoch": 0.816,
      "grad_norm": 24.0,
      "learning_rate": 2.9918032786885248e-05,
      "loss": 0.6093,
      "step": 51
    },
    {
      "epoch": 0.816,
      "eval_exact_match": 0.08163265306122448,
      "eval_f1_a": 0.5125628140703518,
      "eval_f1_m": 0.48412188718311155,
      "eval_loss": 0.6337217092514038,
      "eval_runtime": 12.519,
      "eval_samples_per_second": 19.97,
      "eval_steps_per_second": 2.556,
      "step": 51
    },
    {
      "epoch": 0.832,
      "grad_norm": 28.5,
      "learning_rate": 2.9508196721311478e-05,
      "loss": 0.5875,
      "step": 52
    },
    {
      "epoch": 0.832,
      "eval_exact_match": 0.08163265306122448,
      "eval_f1_a": 0.4444444444444445,
      "eval_f1_m": 0.4678045310698371,
      "eval_loss": 0.6313378810882568,
      "eval_runtime": 12.5146,
      "eval_samples_per_second": 19.977,
      "eval_steps_per_second": 2.557,
      "step": 52
    },
    {
      "epoch": 0.848,
      "grad_norm": 46.25,
      "learning_rate": 2.9098360655737705e-05,
      "loss": 0.5525,
      "step": 53
    },
    {
      "epoch": 0.848,
      "eval_exact_match": 0.061224489795918366,
      "eval_f1_a": 0.380952380952381,
      "eval_f1_m": 0.48287293938554443,
      "eval_loss": 0.6318368911743164,
      "eval_runtime": 12.5161,
      "eval_samples_per_second": 19.974,
      "eval_steps_per_second": 2.557,
      "step": 53
    },
    {
      "epoch": 0.864,
      "grad_norm": 17.625,
      "learning_rate": 2.8688524590163935e-05,
      "loss": 0.632,
      "step": 54
    },
    {
      "epoch": 0.864,
      "eval_exact_match": 0.02040816326530612,
      "eval_f1_a": 0.2992125984251968,
      "eval_f1_m": 0.4569305710962373,
      "eval_loss": 0.6356464624404907,
      "eval_runtime": 12.5236,
      "eval_samples_per_second": 19.962,
      "eval_steps_per_second": 2.555,
      "step": 54
    },
    {
      "epoch": 0.88,
      "grad_norm": 8.875,
      "learning_rate": 2.8278688524590162e-05,
      "loss": 0.5346,
      "step": 55
    },
    {
      "epoch": 0.88,
      "eval_exact_match": 0.0,
      "eval_f1_a": 0.24999999999999997,
      "eval_f1_m": 0.4310802309601829,
      "eval_loss": 0.6434394717216492,
      "eval_runtime": 12.5233,
      "eval_samples_per_second": 19.963,
      "eval_steps_per_second": 2.555,
      "step": 55
    },
    {
      "epoch": 0.896,
      "grad_norm": 11.25,
      "learning_rate": 2.7868852459016392e-05,
      "loss": 0.4656,
      "step": 56
    },
    {
      "epoch": 0.896,
      "eval_exact_match": 0.0,
      "eval_f1_a": 0.11009174311926605,
      "eval_f1_m": 0.38233843987145305,
      "eval_loss": 0.655026376247406,
      "eval_runtime": 12.5245,
      "eval_samples_per_second": 19.961,
      "eval_steps_per_second": 2.555,
      "step": 56
    },
    {
      "epoch": 0.912,
      "grad_norm": 5.96875,
      "learning_rate": 2.7459016393442626e-05,
      "loss": 0.6506,
      "step": 57
    },
    {
      "epoch": 0.912,
      "eval_exact_match": 0.0,
      "eval_f1_a": 0.05714285714285714,
      "eval_f1_m": 0.3669373443683168,
      "eval_loss": 0.6712407469749451,
      "eval_runtime": 12.5217,
      "eval_samples_per_second": 19.965,
      "eval_steps_per_second": 2.556,
      "step": 57
    },
    {
      "epoch": 0.928,
      "grad_norm": 47.75,
      "learning_rate": 2.7049180327868856e-05,
      "loss": 0.9128,
      "step": 58
    },
    {
      "epoch": 0.928,
      "eval_exact_match": 0.0,
      "eval_f1_a": 0.038461538461538464,
      "eval_f1_m": 0.35922759380142344,
      "eval_loss": 0.6825835108757019,
      "eval_runtime": 12.5246,
      "eval_samples_per_second": 19.961,
      "eval_steps_per_second": 2.555,
      "step": 58
    },
    {
      "epoch": 0.944,
      "grad_norm": 47.25,
      "learning_rate": 2.6639344262295087e-05,
      "loss": 0.8041,
      "step": 59
    },
    {
      "epoch": 0.944,
      "eval_exact_match": 0.0,
      "eval_f1_a": 0.038461538461538464,
      "eval_f1_m": 0.35922759380142344,
      "eval_loss": 0.6874199509620667,
      "eval_runtime": 12.5257,
      "eval_samples_per_second": 19.959,
      "eval_steps_per_second": 2.555,
      "step": 59
    },
    {
      "epoch": 0.96,
      "grad_norm": 34.0,
      "learning_rate": 2.6229508196721314e-05,
      "loss": 0.7979,
      "step": 60
    },
    {
      "epoch": 0.96,
      "eval_exact_match": 0.0,
      "eval_f1_a": 0.038461538461538464,
      "eval_f1_m": 0.35922759380142344,
      "eval_loss": 0.6864975690841675,
      "eval_runtime": 12.5253,
      "eval_samples_per_second": 19.96,
      "eval_steps_per_second": 2.555,
      "step": 60
    },
    {
      "epoch": 0.976,
      "grad_norm": 27.0,
      "learning_rate": 2.5819672131147544e-05,
      "loss": 0.626,
      "step": 61
    },
    {
      "epoch": 0.976,
      "eval_exact_match": 0.0,
      "eval_f1_a": 0.038461538461538464,
      "eval_f1_m": 0.35922759380142344,
      "eval_loss": 0.6805967092514038,
      "eval_runtime": 12.5272,
      "eval_samples_per_second": 19.957,
      "eval_steps_per_second": 2.554,
      "step": 61
    },
    {
      "epoch": 0.992,
      "grad_norm": 13.1875,
      "learning_rate": 2.540983606557377e-05,
      "loss": 0.5989,
      "step": 62
    },
    {
      "epoch": 0.992,
      "eval_exact_match": 0.0,
      "eval_f1_a": 0.05714285714285714,
      "eval_f1_m": 0.3673908591075459,
      "eval_loss": 0.6689673066139221,
      "eval_runtime": 12.5192,
      "eval_samples_per_second": 19.969,
      "eval_steps_per_second": 2.556,
      "step": 62
    },
    {
      "epoch": 1.008,
      "grad_norm": 38.25,
      "learning_rate": 2.5e-05,
      "loss": 0.6907,
      "step": 63
    },
    {
      "epoch": 1.008,
      "eval_exact_match": 0.0,
      "eval_f1_a": 0.1272727272727273,
      "eval_f1_m": 0.3914323804779987,
      "eval_loss": 0.6558759808540344,
      "eval_runtime": 12.5213,
      "eval_samples_per_second": 19.966,
      "eval_steps_per_second": 2.556,
      "step": 63
    },
    {
      "epoch": 1.024,
      "grad_norm": 20.375,
      "learning_rate": 2.459016393442623e-05,
      "loss": 0.5757,
      "step": 64
    },
    {
      "epoch": 1.024,
      "eval_exact_match": 0.0,
      "eval_f1_a": 0.23333333333333334,
      "eval_f1_m": 0.426488394225489,
      "eval_loss": 0.6442788243293762,
      "eval_runtime": 12.5203,
      "eval_samples_per_second": 19.968,
      "eval_steps_per_second": 2.556,
      "step": 64
    },
    {
      "epoch": 1.04,
      "grad_norm": 36.5,
      "learning_rate": 2.418032786885246e-05,
      "loss": 0.5626,
      "step": 65
    },
    {
      "epoch": 1.04,
      "eval_exact_match": 0.02040816326530612,
      "eval_f1_a": 0.2677165354330709,
      "eval_f1_m": 0.4416973349446338,
      "eval_loss": 0.6316845417022705,
      "eval_runtime": 12.5217,
      "eval_samples_per_second": 19.965,
      "eval_steps_per_second": 2.556,
      "step": 65
    },
    {
      "epoch": 1.056,
      "grad_norm": 33.0,
      "learning_rate": 2.377049180327869e-05,
      "loss": 0.6962,
      "step": 66
    },
    {
      "epoch": 1.056,
      "eval_exact_match": 0.061224489795918366,
      "eval_f1_a": 0.3973509933774835,
      "eval_f1_m": 0.46887548559217224,
      "eval_loss": 0.6250205039978027,
      "eval_runtime": 12.5299,
      "eval_samples_per_second": 19.952,
      "eval_steps_per_second": 2.554,
      "step": 66
    },
    {
      "epoch": 1.072,
      "grad_norm": 10.25,
      "learning_rate": 2.336065573770492e-05,
      "loss": 0.5543,
      "step": 67
    },
    {
      "epoch": 1.072,
      "eval_exact_match": 0.10204081632653061,
      "eval_f1_a": 0.5082872928176796,
      "eval_f1_m": 0.5018395736583011,
      "eval_loss": 0.6262060403823853,
      "eval_runtime": 12.5279,
      "eval_samples_per_second": 19.955,
      "eval_steps_per_second": 2.554,
      "step": 67
    },
    {
      "epoch": 1.088,
      "grad_norm": 32.25,
      "learning_rate": 2.295081967213115e-05,
      "loss": 0.5034,
      "step": 68
    },
    {
      "epoch": 1.088,
      "eval_exact_match": 0.061224489795918366,
      "eval_f1_a": 0.5471698113207547,
      "eval_f1_m": 0.4798306568714732,
      "eval_loss": 0.6319238543510437,
      "eval_runtime": 12.4294,
      "eval_samples_per_second": 20.114,
      "eval_steps_per_second": 2.575,
      "step": 68
    },
    {
      "epoch": 1.104,
      "grad_norm": 44.0,
      "learning_rate": 2.254098360655738e-05,
      "loss": 0.6685,
      "step": 69
    },
    {
      "epoch": 1.104,
      "eval_exact_match": 0.08163265306122448,
      "eval_f1_a": 0.5650224215246638,
      "eval_f1_m": 0.47513240727526446,
      "eval_loss": 0.6369560360908508,
      "eval_runtime": 12.5288,
      "eval_samples_per_second": 19.954,
      "eval_steps_per_second": 2.554,
      "step": 69
    },
    {
      "epoch": 1.12,
      "grad_norm": 20.5,
      "learning_rate": 2.2131147540983607e-05,
      "loss": 0.7279,
      "step": 70
    },
    {
      "epoch": 1.12,
      "eval_exact_match": 0.08163265306122448,
      "eval_f1_a": 0.5663716814159292,
      "eval_f1_m": 0.46904937466161956,
      "eval_loss": 0.6388349533081055,
      "eval_runtime": 12.5241,
      "eval_samples_per_second": 19.961,
      "eval_steps_per_second": 2.555,
      "step": 70
    },
    {
      "epoch": 1.1360000000000001,
      "grad_norm": 26.75,
      "learning_rate": 2.1721311475409837e-05,
      "loss": 0.6486,
      "step": 71
    },
    {
      "epoch": 1.1360000000000001,
      "eval_exact_match": 0.10204081632653061,
      "eval_f1_a": 0.5826086956521739,
      "eval_f1_m": 0.4729702836845694,
      "eval_loss": 0.6392431855201721,
      "eval_runtime": 12.5238,
      "eval_samples_per_second": 19.962,
      "eval_steps_per_second": 2.555,
      "step": 71
    },
    {
      "epoch": 1.152,
      "grad_norm": 4.875,
      "learning_rate": 2.1311475409836064e-05,
      "loss": 0.7069,
      "step": 72
    },
    {
      "epoch": 1.152,
      "eval_exact_match": 0.08163265306122448,
      "eval_f1_a": 0.5836909871244634,
      "eval_f1_m": 0.47516055373198224,
      "eval_loss": 0.6390937566757202,
      "eval_runtime": 12.4268,
      "eval_samples_per_second": 20.118,
      "eval_steps_per_second": 2.575,
      "step": 72
    },
    {
      "epoch": 1.168,
      "grad_norm": 7.375,
      "learning_rate": 2.0901639344262298e-05,
      "loss": 0.597,
      "step": 73
    },
    {
      "epoch": 1.168,
      "eval_exact_match": 0.08163265306122448,
      "eval_f1_a": 0.5872340425531916,
      "eval_f1_m": 0.47219651097202114,
      "eval_loss": 0.6382343769073486,
      "eval_runtime": 12.5204,
      "eval_samples_per_second": 19.967,
      "eval_steps_per_second": 2.556,
      "step": 73
    },
    {
      "epoch": 1.184,
      "grad_norm": 6.03125,
      "learning_rate": 2.0491803278688525e-05,
      "loss": 0.5629,
      "step": 74
    },
    {
      "epoch": 1.184,
      "eval_exact_match": 0.08163265306122448,
      "eval_f1_a": 0.6050420168067226,
      "eval_f1_m": 0.4974637607290668,
      "eval_loss": 0.6368105411529541,
      "eval_runtime": 12.4204,
      "eval_samples_per_second": 20.128,
      "eval_steps_per_second": 2.576,
      "step": 74
    },
    {
      "epoch": 1.2,
      "grad_norm": 12.5625,
      "learning_rate": 2.0081967213114755e-05,
      "loss": 0.4962,
      "step": 75
    },
    {
      "epoch": 1.2,
      "eval_exact_match": 0.08163265306122448,
      "eval_f1_a": 0.5983606557377048,
      "eval_f1_m": 0.48530426489610157,
      "eval_loss": 0.6430624723434448,
      "eval_runtime": 12.5269,
      "eval_samples_per_second": 19.957,
      "eval_steps_per_second": 2.555,
      "step": 75
    },
    {
      "epoch": 1.216,
      "grad_norm": 33.5,
      "learning_rate": 1.9672131147540985e-05,
      "loss": 0.6281,
      "step": 76
    },
    {
      "epoch": 1.216,
      "eval_exact_match": 0.08163265306122448,
      "eval_f1_a": 0.6040816326530611,
      "eval_f1_m": 0.4916509227733717,
      "eval_loss": 0.6421269774436951,
      "eval_runtime": 12.5178,
      "eval_samples_per_second": 19.972,
      "eval_steps_per_second": 2.556,
      "step": 76
    },
    {
      "epoch": 1.232,
      "grad_norm": 8.375,
      "learning_rate": 1.9262295081967212e-05,
      "loss": 0.5854,
      "step": 77
    },
    {
      "epoch": 1.232,
      "eval_exact_match": 0.08163265306122448,
      "eval_f1_a": 0.5901639344262296,
      "eval_f1_m": 0.4785052249337963,
      "eval_loss": 0.640457034111023,
      "eval_runtime": 12.5109,
      "eval_samples_per_second": 19.983,
      "eval_steps_per_second": 2.558,
      "step": 77
    },
    {
      "epoch": 1.248,
      "grad_norm": 14.25,
      "learning_rate": 1.8852459016393442e-05,
      "loss": 0.6745,
      "step": 78
    },
    {
      "epoch": 1.248,
      "eval_exact_match": 0.08163265306122448,
      "eval_f1_a": 0.5934959349593495,
      "eval_f1_m": 0.4793091828806114,
      "eval_loss": 0.6414687633514404,
      "eval_runtime": 12.5277,
      "eval_samples_per_second": 19.956,
      "eval_steps_per_second": 2.554,
      "step": 78
    },
    {
      "epoch": 1.264,
      "grad_norm": 15.1875,
      "learning_rate": 1.8442622950819673e-05,
      "loss": 0.5662,
      "step": 79
    },
    {
      "epoch": 1.264,
      "eval_exact_match": 0.08163265306122448,
      "eval_f1_a": 0.5967741935483871,
      "eval_f1_m": 0.48058251272536984,
      "eval_loss": 0.6460214853286743,
      "eval_runtime": 12.5159,
      "eval_samples_per_second": 19.975,
      "eval_steps_per_second": 2.557,
      "step": 79
    },
    {
      "epoch": 1.28,
      "grad_norm": 55.25,
      "learning_rate": 1.8032786885245903e-05,
      "loss": 0.679,
      "step": 80
    },
    {
      "epoch": 1.28,
      "eval_exact_match": 0.08163265306122448,
      "eval_f1_a": 0.6007905138339921,
      "eval_f1_m": 0.4773567135812034,
      "eval_loss": 0.6468359231948853,
      "eval_runtime": 12.5291,
      "eval_samples_per_second": 19.953,
      "eval_steps_per_second": 2.554,
      "step": 80
    },
    {
      "epoch": 1.296,
      "grad_norm": 7.21875,
      "learning_rate": 1.7622950819672133e-05,
      "loss": 0.5867,
      "step": 81
    },
    {
      "epoch": 1.296,
      "eval_exact_match": 0.08163265306122448,
      "eval_f1_a": 0.6007905138339921,
      "eval_f1_m": 0.4778103982185614,
      "eval_loss": 0.6476621031761169,
      "eval_runtime": 12.5295,
      "eval_samples_per_second": 19.953,
      "eval_steps_per_second": 2.554,
      "step": 81
    },
    {
      "epoch": 1.312,
      "grad_norm": 8.125,
      "learning_rate": 1.721311475409836e-05,
      "loss": 0.5129,
      "step": 82
    },
    {
      "epoch": 1.312,
      "eval_exact_match": 0.08163265306122448,
      "eval_f1_a": 0.6007905138339921,
      "eval_f1_m": 0.47502746686420155,
      "eval_loss": 0.6478710770606995,
      "eval_runtime": 12.5269,
      "eval_samples_per_second": 19.957,
      "eval_steps_per_second": 2.554,
      "step": 82
    },
    {
      "epoch": 1.328,
      "grad_norm": 41.25,
      "learning_rate": 1.680327868852459e-05,
      "loss": 0.6594,
      "step": 83
    },
    {
      "epoch": 1.328,
      "eval_exact_match": 0.08163265306122448,
      "eval_f1_a": 0.5952380952380952,
      "eval_f1_m": 0.4711932058870834,
      "eval_loss": 0.6437314748764038,
      "eval_runtime": 12.4098,
      "eval_samples_per_second": 20.145,
      "eval_steps_per_second": 2.579,
      "step": 83
    },
    {
      "epoch": 1.3439999999999999,
      "grad_norm": 12.3125,
      "learning_rate": 1.6393442622950818e-05,
      "loss": 0.5754,
      "step": 84
    },
    {
      "epoch": 1.3439999999999999,
      "eval_exact_match": 0.08163265306122448,
      "eval_f1_a": 0.5910931174089069,
      "eval_f1_m": 0.4749946765252888,
      "eval_loss": 0.6412422060966492,
      "eval_runtime": 12.4226,
      "eval_samples_per_second": 20.125,
      "eval_steps_per_second": 2.576,
      "step": 84
    },
    {
      "epoch": 1.3599999999999999,
      "grad_norm": 12.0625,
      "learning_rate": 1.598360655737705e-05,
      "loss": 0.5544,
      "step": 85
    },
    {
      "epoch": 1.3599999999999999,
      "eval_exact_match": 0.08163265306122448,
      "eval_f1_a": 0.5887096774193548,
      "eval_f1_m": 0.47564113890644494,
      "eval_loss": 0.639517605304718,
      "eval_runtime": 12.5196,
      "eval_samples_per_second": 19.969,
      "eval_steps_per_second": 2.556,
      "step": 85
    },
    {
      "epoch": 1.376,
      "grad_norm": 13.1875,
      "learning_rate": 1.557377049180328e-05,
      "loss": 0.6295,
      "step": 86
    },
    {
      "epoch": 1.376,
      "eval_exact_match": 0.061224489795918366,
      "eval_f1_a": 0.5726141078838174,
      "eval_f1_m": 0.4691413461821626,
      "eval_loss": 0.6366972923278809,
      "eval_runtime": 12.5166,
      "eval_samples_per_second": 19.973,
      "eval_steps_per_second": 2.557,
      "step": 86
    },
    {
      "epoch": 1.392,
      "grad_norm": 5.875,
      "learning_rate": 1.5163934426229509e-05,
      "loss": 0.6123,
      "step": 87
    },
    {
      "epoch": 1.392,
      "eval_exact_match": 0.061224489795918366,
      "eval_f1_a": 0.5690376569037656,
      "eval_f1_m": 0.46367544020605245,
      "eval_loss": 0.6347695589065552,
      "eval_runtime": 12.5135,
      "eval_samples_per_second": 19.978,
      "eval_steps_per_second": 2.557,
      "step": 87
    },
    {
      "epoch": 1.408,
      "grad_norm": 41.75,
      "learning_rate": 1.4754098360655739e-05,
      "loss": 0.6,
      "step": 88
    },
    {
      "epoch": 1.408,
      "eval_exact_match": 0.061224489795918366,
      "eval_f1_a": 0.5702127659574467,
      "eval_f1_m": 0.4770136326258776,
      "eval_loss": 0.6304101347923279,
      "eval_runtime": 12.4096,
      "eval_samples_per_second": 20.146,
      "eval_steps_per_second": 2.579,
      "step": 88
    },
    {
      "epoch": 1.424,
      "grad_norm": 52.5,
      "learning_rate": 1.4344262295081968e-05,
      "loss": 0.6655,
      "step": 89
    },
    {
      "epoch": 1.424,
      "eval_exact_match": 0.061224489795918366,
      "eval_f1_a": 0.5663716814159292,
      "eval_f1_m": 0.4777696226675817,
      "eval_loss": 0.6252734661102295,
      "eval_runtime": 12.52,
      "eval_samples_per_second": 19.968,
      "eval_steps_per_second": 2.556,
      "step": 89
    },
    {
      "epoch": 1.44,
      "grad_norm": 7.5625,
      "learning_rate": 1.3934426229508196e-05,
      "loss": 0.5784,
      "step": 90
    },
    {
      "epoch": 1.44,
      "eval_exact_match": 0.08163265306122448,
      "eval_f1_a": 0.5610859728506787,
      "eval_f1_m": 0.4947867325418346,
      "eval_loss": 0.6227597594261169,
      "eval_runtime": 12.5093,
      "eval_samples_per_second": 19.985,
      "eval_steps_per_second": 2.558,
      "step": 90
    },
    {
      "epoch": 1.456,
      "grad_norm": 9.375,
      "learning_rate": 1.3524590163934428e-05,
      "loss": 0.5715,
      "step": 91
    },
    {
      "epoch": 1.456,
      "eval_exact_match": 0.08163265306122448,
      "eval_f1_a": 0.5781990521327014,
      "eval_f1_m": 0.520756964124311,
      "eval_loss": 0.6186211109161377,
      "eval_runtime": 12.7601,
      "eval_samples_per_second": 19.592,
      "eval_steps_per_second": 2.508,
      "step": 91
    },
    {
      "epoch": 1.472,
      "grad_norm": 7.21875,
      "learning_rate": 1.3114754098360657e-05,
      "loss": 0.6274,
      "step": 92
    },
    {
      "epoch": 1.472,
      "eval_exact_match": 0.08163265306122448,
      "eval_f1_a": 0.5603864734299516,
      "eval_f1_m": 0.51300422933076,
      "eval_loss": 0.6171249747276306,
      "eval_runtime": 12.9738,
      "eval_samples_per_second": 19.27,
      "eval_steps_per_second": 2.467,
      "step": 92
    },
    {
      "epoch": 1.488,
      "grad_norm": 30.0,
      "learning_rate": 1.2704918032786885e-05,
      "loss": 0.7038,
      "step": 93
    },
    {
      "epoch": 1.488,
      "eval_exact_match": 0.08163265306122448,
      "eval_f1_a": 0.5784313725490196,
      "eval_f1_m": 0.5305293346109673,
      "eval_loss": 0.6147861480712891,
      "eval_runtime": 13.3609,
      "eval_samples_per_second": 18.711,
      "eval_steps_per_second": 2.395,
      "step": 93
    },
    {
      "epoch": 1.504,
      "grad_norm": 6.6875,
      "learning_rate": 1.2295081967213116e-05,
      "loss": 0.5364,
      "step": 94
    },
    {
      "epoch": 1.504,
      "eval_exact_match": 0.08163265306122448,
      "eval_f1_a": 0.5771144278606964,
      "eval_f1_m": 0.5379805568581079,
      "eval_loss": 0.6129326224327087,
      "eval_runtime": 13.0415,
      "eval_samples_per_second": 19.17,
      "eval_steps_per_second": 2.454,
      "step": 94
    },
    {
      "epoch": 1.52,
      "grad_norm": 27.75,
      "learning_rate": 1.1885245901639344e-05,
      "loss": 0.4883,
      "step": 95
    },
    {
      "epoch": 1.52,
      "eval_exact_match": 0.08163265306122448,
      "eval_f1_a": 0.5612244897959184,
      "eval_f1_m": 0.53111536989088,
      "eval_loss": 0.6093867421150208,
      "eval_runtime": 12.5993,
      "eval_samples_per_second": 19.842,
      "eval_steps_per_second": 2.54,
      "step": 95
    },
    {
      "epoch": 1.536,
      "grad_norm": 23.0,
      "learning_rate": 1.1475409836065575e-05,
      "loss": 0.4918,
      "step": 96
    },
    {
      "epoch": 1.536,
      "eval_exact_match": 0.08163265306122448,
      "eval_f1_a": 0.5567010309278351,
      "eval_f1_m": 0.526410154471379,
      "eval_loss": 0.6084589958190918,
      "eval_runtime": 12.6824,
      "eval_samples_per_second": 19.712,
      "eval_steps_per_second": 2.523,
      "step": 96
    },
    {
      "epoch": 1.552,
      "grad_norm": 5.71875,
      "learning_rate": 1.1065573770491803e-05,
      "loss": 0.5881,
      "step": 97
    },
    {
      "epoch": 1.552,
      "eval_exact_match": 0.10204081632653061,
      "eval_f1_a": 0.5502645502645503,
      "eval_f1_m": 0.5260332071556562,
      "eval_loss": 0.6075829863548279,
      "eval_runtime": 12.5867,
      "eval_samples_per_second": 19.862,
      "eval_steps_per_second": 2.542,
      "step": 97
    },
    {
      "epoch": 1.568,
      "grad_norm": 18.0,
      "learning_rate": 1.0655737704918032e-05,
      "loss": 0.4614,
      "step": 98
    },
    {
      "epoch": 1.568,
      "eval_exact_match": 0.10204081632653061,
      "eval_f1_a": 0.5608465608465608,
      "eval_f1_m": 0.5355139475037434,
      "eval_loss": 0.6062441468238831,
      "eval_runtime": 12.567,
      "eval_samples_per_second": 19.893,
      "eval_steps_per_second": 2.546,
      "step": 98
    },
    {
      "epoch": 1.584,
      "grad_norm": 22.625,
      "learning_rate": 1.0245901639344262e-05,
      "loss": 0.5693,
      "step": 99
    },
    {
      "epoch": 1.584,
      "eval_exact_match": 0.10204081632653061,
      "eval_f1_a": 0.5608465608465608,
      "eval_f1_m": 0.5325499047437824,
      "eval_loss": 0.6053320169448853,
      "eval_runtime": 12.573,
      "eval_samples_per_second": 19.884,
      "eval_steps_per_second": 2.545,
      "step": 99
    },
    {
      "epoch": 1.6,
      "grad_norm": 22.5,
      "learning_rate": 9.836065573770493e-06,
      "loss": 0.5239,
      "step": 100
    },
    {
      "epoch": 1.6,
      "eval_exact_match": 0.08163265306122448,
      "eval_f1_a": 0.5531914893617021,
      "eval_f1_m": 0.5152986752476547,
      "eval_loss": 0.6034590005874634,
      "eval_runtime": 12.5763,
      "eval_samples_per_second": 19.879,
      "eval_steps_per_second": 2.544,
      "step": 100
    },
    {
      "epoch": 1.616,
      "grad_norm": 17.25,
      "learning_rate": 9.426229508196721e-06,
      "loss": 0.4434,
      "step": 101
    },
    {
      "epoch": 1.616,
      "eval_exact_match": 0.10204081632653061,
      "eval_f1_a": 0.5405405405405405,
      "eval_f1_m": 0.5147288369227145,
      "eval_loss": 0.602425754070282,
      "eval_runtime": 12.5712,
      "eval_samples_per_second": 19.887,
      "eval_steps_per_second": 2.546,
      "step": 101
    },
    {
      "epoch": 1.6320000000000001,
      "grad_norm": 31.875,
      "learning_rate": 9.016393442622952e-06,
      "loss": 0.6336,
      "step": 102
    },
    {
      "epoch": 1.6320000000000001,
      "eval_exact_match": 0.10204081632653061,
      "eval_f1_a": 0.53551912568306,
      "eval_f1_m": 0.5188104695757757,
      "eval_loss": 0.6012207269668579,
      "eval_runtime": 12.5726,
      "eval_samples_per_second": 19.885,
      "eval_steps_per_second": 2.545,
      "step": 102
    },
    {
      "epoch": 1.6480000000000001,
      "grad_norm": 10.875,
      "learning_rate": 8.60655737704918e-06,
      "loss": 0.5441,
      "step": 103
    },
    {
      "epoch": 1.6480000000000001,
      "eval_exact_match": 0.10204081632653061,
      "eval_f1_a": 0.5363128491620112,
      "eval_f1_m": 0.5246980627082668,
      "eval_loss": 0.6005741953849792,
      "eval_runtime": 12.5894,
      "eval_samples_per_second": 19.858,
      "eval_steps_per_second": 2.542,
      "step": 103
    },
    {
      "epoch": 1.6640000000000001,
      "grad_norm": 16.25,
      "learning_rate": 8.196721311475409e-06,
      "loss": 0.4561,
      "step": 104
    },
    {
      "epoch": 1.6640000000000001,
      "eval_exact_match": 0.10204081632653061,
      "eval_f1_a": 0.5202312138728324,
      "eval_f1_m": 0.5224427233430835,
      "eval_loss": 0.5999111533164978,
      "eval_runtime": 12.4649,
      "eval_samples_per_second": 20.056,
      "eval_steps_per_second": 2.567,
      "step": 104
    },
    {
      "epoch": 1.6800000000000002,
      "grad_norm": 12.4375,
      "learning_rate": 7.78688524590164e-06,
      "loss": 0.6999,
      "step": 105
    },
    {
      "epoch": 1.6800000000000002,
      "eval_exact_match": 0.10204081632653061,
      "eval_f1_a": 0.5207100591715976,
      "eval_f1_m": 0.5320394191642691,
      "eval_loss": 0.5994404554367065,
      "eval_runtime": 12.5816,
      "eval_samples_per_second": 19.87,
      "eval_steps_per_second": 2.543,
      "step": 105
    },
    {
      "epoch": 1.696,
      "grad_norm": 4.53125,
      "learning_rate": 7.3770491803278695e-06,
      "loss": 0.4781,
      "step": 106
    },
    {
      "epoch": 1.696,
      "eval_exact_match": 0.12244897959183673,
      "eval_f1_a": 0.524390243902439,
      "eval_f1_m": 0.5555706498383567,
      "eval_loss": 0.5982089638710022,
      "eval_runtime": 12.4629,
      "eval_samples_per_second": 20.06,
      "eval_steps_per_second": 2.568,
      "step": 106
    },
    {
      "epoch": 1.712,
      "grad_norm": 9.3125,
      "learning_rate": 6.967213114754098e-06,
      "loss": 0.4955,
      "step": 107
    },
    {
      "epoch": 1.712,
      "eval_exact_match": 0.10204081632653061,
      "eval_f1_a": 0.5000000000000001,
      "eval_f1_m": 0.5406607298263958,
      "eval_loss": 0.5973047018051147,
      "eval_runtime": 12.4809,
      "eval_samples_per_second": 20.031,
      "eval_steps_per_second": 2.564,
      "step": 107
    },
    {
      "epoch": 1.728,
      "grad_norm": 15.9375,
      "learning_rate": 6.557377049180328e-06,
      "loss": 0.7169,
      "step": 108
    },
    {
      "epoch": 1.728,
      "eval_exact_match": 0.10204081632653061,
      "eval_f1_a": 0.5061728395061729,
      "eval_f1_m": 0.5354615072802346,
      "eval_loss": 0.5978144407272339,
      "eval_runtime": 12.4727,
      "eval_samples_per_second": 20.044,
      "eval_steps_per_second": 2.566,
      "step": 108
    },
    {
      "epoch": 1.744,
      "grad_norm": 8.3125,
      "learning_rate": 6.147540983606558e-06,
      "loss": 0.5559,
      "step": 109
    },
    {
      "epoch": 1.744,
      "eval_exact_match": 0.10204081632653061,
      "eval_f1_a": 0.5185185185185185,
      "eval_f1_m": 0.538548896081909,
      "eval_loss": 0.5964921712875366,
      "eval_runtime": 12.4816,
      "eval_samples_per_second": 20.029,
      "eval_steps_per_second": 2.564,
      "step": 109
    },
    {
      "epoch": 1.76,
      "grad_norm": 14.25,
      "learning_rate": 5.737704918032787e-06,
      "loss": 0.5697,
      "step": 110
    },
    {
      "epoch": 1.76,
      "eval_exact_match": 0.10204081632653061,
      "eval_f1_a": 0.49382716049382713,
      "eval_f1_m": 0.5282943547049189,
      "eval_loss": 0.5966972708702087,
      "eval_runtime": 12.4667,
      "eval_samples_per_second": 20.053,
      "eval_steps_per_second": 2.567,
      "step": 110
    },
    {
      "epoch": 1.776,
      "grad_norm": 8.375,
      "learning_rate": 5.327868852459016e-06,
      "loss": 0.4283,
      "step": 111
    },
    {
      "epoch": 1.776,
      "eval_exact_match": 0.10204081632653061,
      "eval_f1_a": 0.5031446540880504,
      "eval_f1_m": 0.5433442707752428,
      "eval_loss": 0.5960127115249634,
      "eval_runtime": 12.4773,
      "eval_samples_per_second": 20.036,
      "eval_steps_per_second": 2.565,
      "step": 111
    },
    {
      "epoch": 1.792,
      "grad_norm": 16.75,
      "learning_rate": 4.918032786885246e-06,
      "loss": 0.6258,
      "step": 112
    },
    {
      "epoch": 1.792,
      "eval_exact_match": 0.10204081632653061,
      "eval_f1_a": 0.4774193548387097,
      "eval_f1_m": 0.5284177007466322,
      "eval_loss": 0.5976787209510803,
      "eval_runtime": 12.4731,
      "eval_samples_per_second": 20.043,
      "eval_steps_per_second": 2.566,
      "step": 112
    },
    {
      "epoch": 1.808,
      "grad_norm": 7.96875,
      "learning_rate": 4.508196721311476e-06,
      "loss": 0.5256,
      "step": 113
    },
    {
      "epoch": 1.808,
      "eval_exact_match": 0.10204081632653061,
      "eval_f1_a": 0.4615384615384615,
      "eval_f1_m": 0.5200843674132989,
      "eval_loss": 0.5962343811988831,
      "eval_runtime": 12.5684,
      "eval_samples_per_second": 19.891,
      "eval_steps_per_second": 2.546,
      "step": 113
    },
    {
      "epoch": 1.8239999999999998,
      "grad_norm": 15.125,
      "learning_rate": 4.098360655737704e-06,
      "loss": 0.5736,
      "step": 114
    },
    {
      "epoch": 1.8239999999999998,
      "eval_exact_match": 0.10204081632653061,
      "eval_f1_a": 0.4675324675324675,
      "eval_f1_m": 0.5306153330463053,
      "eval_loss": 0.5951288938522339,
      "eval_runtime": 12.5788,
      "eval_samples_per_second": 19.875,
      "eval_steps_per_second": 2.544,
      "step": 114
    },
    {
      "epoch": 1.8399999999999999,
      "grad_norm": 20.25,
      "learning_rate": 3.6885245901639347e-06,
      "loss": 0.366,
      "step": 115
    },
    {
      "epoch": 1.8399999999999999,
      "eval_exact_match": 0.10204081632653061,
      "eval_f1_a": 0.4645161290322581,
      "eval_f1_m": 0.5248462721752035,
      "eval_loss": 0.594918966293335,
      "eval_runtime": 12.4723,
      "eval_samples_per_second": 20.044,
      "eval_steps_per_second": 2.566,
      "step": 115
    },
    {
      "epoch": 1.8559999999999999,
      "grad_norm": 21.125,
      "learning_rate": 3.278688524590164e-06,
      "loss": 0.5082,
      "step": 116
    },
    {
      "epoch": 1.8559999999999999,
      "eval_exact_match": 0.10204081632653061,
      "eval_f1_a": 0.48051948051948046,
      "eval_f1_m": 0.5353904898622783,
      "eval_loss": 0.5939072370529175,
      "eval_runtime": 12.5855,
      "eval_samples_per_second": 19.864,
      "eval_steps_per_second": 2.543,
      "step": 116
    },
    {
      "epoch": 1.8719999999999999,
      "grad_norm": 11.75,
      "learning_rate": 2.8688524590163937e-06,
      "loss": 0.5421,
      "step": 117
    },
    {
      "epoch": 1.8719999999999999,
      "eval_exact_match": 0.10204081632653061,
      "eval_f1_a": 0.47435897435897434,
      "eval_f1_m": 0.5267037684204547,
      "eval_loss": 0.5957451462745667,
      "eval_runtime": 12.479,
      "eval_samples_per_second": 20.034,
      "eval_steps_per_second": 2.564,
      "step": 117
    },
    {
      "epoch": 1.888,
      "grad_norm": 8.0,
      "learning_rate": 2.459016393442623e-06,
      "loss": 0.62,
      "step": 118
    },
    {
      "epoch": 1.888,
      "eval_exact_match": 0.10204081632653061,
      "eval_f1_a": 0.4705882352941176,
      "eval_f1_m": 0.5307986531275845,
      "eval_loss": 0.5949638485908508,
      "eval_runtime": 12.4749,
      "eval_samples_per_second": 20.04,
      "eval_steps_per_second": 2.565,
      "step": 118
    },
    {
      "epoch": 1.904,
      "grad_norm": 11.375,
      "learning_rate": 2.049180327868852e-06,
      "loss": 0.6126,
      "step": 119
    },
    {
      "epoch": 1.904,
      "eval_exact_match": 0.10204081632653061,
      "eval_f1_a": 0.4774193548387097,
      "eval_f1_m": 0.5340785365095088,
      "eval_loss": 0.5944355726242065,
      "eval_runtime": 12.5766,
      "eval_samples_per_second": 19.878,
      "eval_steps_per_second": 2.544,
      "step": 119
    },
    {
      "epoch": 1.92,
      "grad_norm": 13.3125,
      "learning_rate": 1.639344262295082e-06,
      "loss": 0.527,
      "step": 120
    },
    {
      "epoch": 1.92,
      "eval_exact_match": 0.10204081632653061,
      "eval_f1_a": 0.484076433121019,
      "eval_f1_m": 0.5245061361207819,
      "eval_loss": 0.5951045155525208,
      "eval_runtime": 12.4733,
      "eval_samples_per_second": 20.043,
      "eval_steps_per_second": 2.565,
      "step": 120
    },
    {
      "epoch": 1.936,
      "grad_norm": 14.9375,
      "learning_rate": 1.2295081967213116e-06,
      "loss": 0.5308,
      "step": 121
    },
    {
      "epoch": 1.936,
      "eval_exact_match": 0.10204081632653061,
      "eval_f1_a": 0.484076433121019,
      "eval_f1_m": 0.5301669718836584,
      "eval_loss": 0.5943300724029541,
      "eval_runtime": 12.5706,
      "eval_samples_per_second": 19.888,
      "eval_steps_per_second": 2.546,
      "step": 121
    },
    {
      "epoch": 1.952,
      "grad_norm": 5.1875,
      "learning_rate": 8.19672131147541e-07,
      "loss": 0.5002,
      "step": 122
    },
    {
      "epoch": 1.952,
      "eval_exact_match": 0.10204081632653061,
      "eval_f1_a": 0.4774193548387097,
      "eval_f1_m": 0.5352071697809991,
      "eval_loss": 0.5943408012390137,
      "eval_runtime": 12.4681,
      "eval_samples_per_second": 20.051,
      "eval_steps_per_second": 2.567,
      "step": 122
    },
    {
      "epoch": 1.968,
      "grad_norm": 10.5625,
      "learning_rate": 4.098360655737705e-07,
      "loss": 0.4733,
      "step": 123
    },
    {
      "epoch": 1.968,
      "eval_exact_match": 0.10204081632653061,
      "eval_f1_a": 0.4903225806451614,
      "eval_f1_m": 0.5415129388418702,
      "eval_loss": 0.5952519774436951,
      "eval_runtime": 12.4602,
      "eval_samples_per_second": 20.064,
      "eval_steps_per_second": 2.568,
      "step": 123
    },
    {
      "epoch": 1.984,
      "grad_norm": 17.625,
      "learning_rate": 0.0,
      "loss": 0.4085,
      "step": 124
    },
    {
      "epoch": 1.984,
      "eval_exact_match": 0.10204081632653061,
      "eval_f1_a": 0.484076433121019,
      "eval_f1_m": 0.5301669718836584,
      "eval_loss": 0.5949286818504333,
      "eval_runtime": 12.5825,
      "eval_samples_per_second": 19.869,
      "eval_steps_per_second": 2.543,
      "step": 124
    },
    {
      "epoch": 1.984,
      "step": 124,
      "total_flos": 4.246810864386048e+16,
      "train_loss": 0.6719285288164693,
      "train_runtime": 1910.309,
      "train_samples_per_second": 1.047,
      "train_steps_per_second": 0.065
    }
  ],
  "logging_steps": 1,
  "max_steps": 124,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": false,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 4.246810864386048e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}