{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 12.0,
  "eval_steps": 500,
  "global_step": 6348,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.1890359168241966,
      "grad_norm": 7.269421577453613,
      "learning_rate": 0.0005,
      "loss": 1.9195,
      "step": 100
    },
    {
      "epoch": 0.3780718336483932,
      "grad_norm": 1.071650505065918,
      "learning_rate": 0.0005,
      "loss": 1.9396,
      "step": 200
    },
    {
      "epoch": 0.5671077504725898,
      "grad_norm": 0.5662499070167542,
      "learning_rate": 0.0005,
      "loss": 2.3641,
      "step": 300
    },
    {
      "epoch": 0.7561436672967864,
      "grad_norm": 0.47088223695755005,
      "learning_rate": 0.0005,
      "loss": 1.8969,
      "step": 400
    },
    {
      "epoch": 0.945179584120983,
      "grad_norm": 0.3653932511806488,
      "learning_rate": 0.0005,
      "loss": 1.8531,
      "step": 500
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6214529147982063,
      "eval_loss": 1.5580899715423584,
      "eval_runtime": 5.56,
      "eval_samples_per_second": 89.928,
      "eval_steps_per_second": 11.331,
      "step": 529
    },
    {
      "epoch": 1.0,
      "eval_exact_match": 14.4,
      "eval_f1": 20.983333333333345,
      "eval_qa_bleu": 12.514681390986086,
      "eval_qa_exact_match": 0.102,
      "eval_recite_bleu": 12.61307158352263,
      "eval_recite_exact_match": 0.0,
      "step": 529
    },
    {
      "epoch": 1.1342155009451795,
      "grad_norm": 0.5318541526794434,
      "learning_rate": 0.0005,
      "loss": 1.6622,
      "step": 600
    },
    {
      "epoch": 1.3232514177693762,
      "grad_norm": 235.6529998779297,
      "learning_rate": 0.0005,
      "loss": 1.5596,
      "step": 700
    },
    {
      "epoch": 1.5122873345935728,
      "grad_norm": 0.5057249069213867,
      "learning_rate": 0.0005,
      "loss": 1.5296,
      "step": 800
    },
    {
      "epoch": 1.7013232514177694,
      "grad_norm": 0.5339187383651733,
      "learning_rate": 0.0005,
      "loss": 1.5036,
      "step": 900
    },
    {
      "epoch": 1.8903591682419658,
      "grad_norm": 0.5649825930595398,
      "learning_rate": 0.0005,
      "loss": 1.4884,
      "step": 1000
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6643587443946188,
      "eval_loss": 1.2189537286758423,
      "eval_runtime": 6.192,
      "eval_samples_per_second": 80.749,
      "eval_steps_per_second": 10.174,
      "step": 1058
    },
    {
      "epoch": 2.0,
      "eval_exact_match": 11.2,
      "eval_f1": 18.31238095238096,
      "eval_qa_bleu": 10.766593815425745,
      "eval_qa_exact_match": 0.096,
      "eval_recite_bleu": 13.257993783037875,
      "eval_recite_exact_match": 0.002,
      "step": 1058
    },
    {
      "epoch": 2.0793950850661624,
      "grad_norm": 0.6042512655258179,
      "learning_rate": 0.0005,
      "loss": 1.2966,
      "step": 1100
    },
    {
      "epoch": 2.268431001890359,
      "grad_norm": 965.2459106445312,
      "learning_rate": 0.0005,
      "loss": 8.8153,
      "step": 1200
    },
    {
      "epoch": 2.4574669187145557,
      "grad_norm": 3.0128896236419678,
      "learning_rate": 0.0005,
      "loss": 9.1241,
      "step": 1300
    },
    {
      "epoch": 2.6465028355387523,
      "grad_norm": 108062179328.0,
      "learning_rate": 0.0005,
      "loss": 10.5947,
      "step": 1400
    },
    {
      "epoch": 2.835538752362949,
      "grad_norm": 5002999889920.0,
      "learning_rate": 0.0005,
      "loss": 12.5972,
      "step": 1500
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.21054708520179372,
      "eval_loss": 19.724363327026367,
      "eval_runtime": 6.5285,
      "eval_samples_per_second": 76.587,
      "eval_steps_per_second": 9.65,
      "step": 1587
    },
    {
      "epoch": 3.0,
      "eval_exact_match": 0.0,
      "eval_f1": 0.0,
      "eval_qa_bleu": 0.0,
      "eval_qa_exact_match": 0.0,
      "eval_recite_bleu": 0.0,
      "eval_recite_exact_match": 0.0,
      "step": 1587
    },
    {
      "epoch": 3.0245746691871456,
      "grad_norm": 62.19309616088867,
      "learning_rate": 0.0005,
      "loss": 13.3192,
      "step": 1600
    },
    {
      "epoch": 3.213610586011342,
      "grad_norm": 4.109750747680664,
      "learning_rate": 0.0005,
      "loss": 8.4804,
      "step": 1700
    },
    {
      "epoch": 3.402646502835539,
      "grad_norm": 7.620689392089844,
      "learning_rate": 0.0005,
      "loss": 8.347,
      "step": 1800
    },
    {
      "epoch": 3.5916824196597354,
      "grad_norm": 33.092830657958984,
      "learning_rate": 0.0005,
      "loss": 7.4667,
      "step": 1900
    },
    {
      "epoch": 3.780718336483932,
      "grad_norm": 204828.203125,
      "learning_rate": 0.0005,
      "loss": 7.885,
      "step": 2000
    },
    {
      "epoch": 3.9697542533081287,
      "grad_norm": 229958729728.0,
      "learning_rate": 0.0005,
      "loss": 8.4174,
      "step": 2100
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.22234977578475337,
      "eval_loss": 9.6956148147583,
      "eval_runtime": 5.4245,
      "eval_samples_per_second": 92.174,
      "eval_steps_per_second": 11.614,
      "step": 2116
    },
    {
      "epoch": 4.0,
      "eval_exact_match": 0.0,
      "eval_f1": 0.0,
      "eval_qa_bleu": 0.0,
      "eval_qa_exact_match": 0.0,
      "eval_recite_bleu": 0.0,
      "eval_recite_exact_match": 0.0,
      "step": 2116
    },
    {
      "epoch": 4.158790170132325,
      "grad_norm": 17.098289489746094,
      "learning_rate": 0.0005,
      "loss": 10.4359,
      "step": 2200
    },
    {
      "epoch": 4.3478260869565215,
      "grad_norm": 1069.9039306640625,
      "learning_rate": 0.0005,
      "loss": 9.5369,
      "step": 2300
    },
    {
      "epoch": 4.536862003780718,
      "grad_norm": 7.694222927093506,
      "learning_rate": 0.0005,
      "loss": 7.7028,
      "step": 2400
    },
    {
      "epoch": 4.725897920604915,
      "grad_norm": Infinity,
      "learning_rate": 0.0005,
      "loss": 8.4852,
      "step": 2500
    },
    {
      "epoch": 4.914933837429111,
      "grad_norm": 396939770200064.0,
      "learning_rate": 0.0005,
      "loss": 9.5996,
      "step": 2600
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.2166188340807175,
      "eval_loss": 10.412571907043457,
      "eval_runtime": 5.8473,
      "eval_samples_per_second": 85.51,
      "eval_steps_per_second": 10.774,
      "step": 2645
    },
    {
      "epoch": 5.0,
      "eval_exact_match": 0.0,
      "eval_f1": 0.0,
      "eval_qa_bleu": 0.0,
      "eval_qa_exact_match": 0.0,
      "eval_recite_bleu": 0.0,
      "eval_recite_exact_match": 0.0,
      "step": 2645
    },
    {
      "epoch": 5.103969754253308,
      "grad_norm": 3.166937635134374e+18,
      "learning_rate": 0.0005,
      "loss": 10.3832,
      "step": 2700
    },
    {
      "epoch": 5.293005671077505,
      "grad_norm": 1.8363503072027607e+18,
      "learning_rate": 0.0005,
      "loss": 9.7717,
      "step": 2800
    },
    {
      "epoch": 5.482041587901701,
      "grad_norm": 113262112800768.0,
      "learning_rate": 0.0005,
      "loss": 8.3066,
      "step": 2900
    },
    {
      "epoch": 5.671077504725898,
      "grad_norm": 1283434880.0,
      "learning_rate": 0.0005,
      "loss": 8.6003,
      "step": 3000
    },
    {
      "epoch": 5.8601134215500945,
      "grad_norm": 1221250.5,
      "learning_rate": 0.0005,
      "loss": 9.1302,
      "step": 3100
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.21948878923766815,
      "eval_loss": 9.434032440185547,
      "eval_runtime": 6.1208,
      "eval_samples_per_second": 81.689,
      "eval_steps_per_second": 10.293,
      "step": 3174
    },
    {
      "epoch": 6.0,
      "eval_exact_match": 0.0,
      "eval_f1": 0.0,
      "eval_qa_bleu": 0.0,
      "eval_qa_exact_match": 0.0,
      "eval_recite_bleu": 0.0,
      "eval_recite_exact_match": 0.0,
      "step": 3174
    },
    {
      "epoch": 6.049149338374291,
      "grad_norm": 2019.499267578125,
      "learning_rate": 0.0005,
      "loss": 9.3999,
      "step": 3200
    },
    {
      "epoch": 6.238185255198488,
      "grad_norm": 164.9034881591797,
      "learning_rate": 0.0005,
      "loss": 8.1768,
      "step": 3300
    },
    {
      "epoch": 6.427221172022684,
      "grad_norm": 434.93157958984375,
      "learning_rate": 0.0005,
      "loss": 7.6329,
      "step": 3400
    },
    {
      "epoch": 6.616257088846881,
      "grad_norm": 1324.7332763671875,
      "learning_rate": 0.0005,
      "loss": 7.7765,
      "step": 3500
    },
    {
      "epoch": 6.805293005671078,
      "grad_norm": 52531002605568.0,
      "learning_rate": 0.0005,
      "loss": 8.006,
      "step": 3600
    },
    {
      "epoch": 6.994328922495274,
      "grad_norm": Infinity,
      "learning_rate": 0.0005,
      "loss": 15.3926,
      "step": 3700
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.23054708520179373,
      "eval_loss": 15.052428245544434,
      "eval_runtime": 5.4671,
      "eval_samples_per_second": 91.457,
      "eval_steps_per_second": 11.524,
      "step": 3703
    },
    {
      "epoch": 7.0,
      "eval_exact_match": 0.0,
      "eval_f1": 0.0,
      "eval_qa_bleu": 0.0,
      "eval_qa_exact_match": 0.0,
      "eval_recite_bleu": 0.0,
      "eval_recite_exact_match": 0.0,
      "step": 3703
    },
    {
      "epoch": 7.183364839319471,
      "grad_norm": Infinity,
      "learning_rate": 0.0005,
      "loss": 17.6263,
      "step": 3800
    },
    {
      "epoch": 7.3724007561436675,
      "grad_norm": Infinity,
      "learning_rate": 0.0005,
      "loss": 17.7669,
      "step": 3900
    },
    {
      "epoch": 7.561436672967864,
      "grad_norm": Infinity,
      "learning_rate": 0.0005,
      "loss": 17.7462,
      "step": 4000
    },
    {
      "epoch": 7.750472589792061,
      "grad_norm": Infinity,
      "learning_rate": 0.0005,
      "loss": 17.6186,
      "step": 4100
    },
    {
      "epoch": 7.939508506616257,
      "grad_norm": NaN,
      "learning_rate": 0.0005,
      "loss": 1574.3041,
      "step": 4200
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.21060089686098654,
      "eval_loss": NaN,
      "eval_runtime": 5.4376,
      "eval_samples_per_second": 91.952,
      "eval_steps_per_second": 11.586,
      "step": 4232
    },
    {
      "epoch": 8.0,
      "eval_exact_match": 0.0,
      "eval_f1": 0.0,
      "eval_qa_bleu": 0.0,
      "eval_qa_exact_match": 0.0,
      "eval_recite_bleu": 0.0,
      "eval_recite_exact_match": 0.0,
      "step": 4232
    },
    {
      "epoch": 8.128544423440454,
      "grad_norm": NaN,
      "learning_rate": 0.0005,
      "loss": 0.0,
      "step": 4300
    },
    {
      "epoch": 8.31758034026465,
      "grad_norm": NaN,
      "learning_rate": 0.0005,
      "loss": 0.0,
      "step": 4400
    },
    {
      "epoch": 8.506616257088847,
      "grad_norm": NaN,
      "learning_rate": 0.0005,
      "loss": 0.0,
      "step": 4500
    },
    {
      "epoch": 8.695652173913043,
      "grad_norm": NaN,
      "learning_rate": 0.0005,
      "loss": 0.0,
      "step": 4600
    },
    {
      "epoch": 8.88468809073724,
      "grad_norm": NaN,
      "learning_rate": 0.0005,
      "loss": 0.0,
      "step": 4700
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.21060089686098654,
      "eval_loss": NaN,
      "eval_runtime": 6.1031,
      "eval_samples_per_second": 81.925,
      "eval_steps_per_second": 10.323,
      "step": 4761
    },
    {
      "epoch": 9.0,
      "eval_exact_match": 0.0,
      "eval_f1": 0.0,
      "eval_qa_bleu": 0.0,
      "eval_qa_exact_match": 0.0,
      "eval_recite_bleu": 0.0,
      "eval_recite_exact_match": 0.0,
      "step": 4761
    },
    {
      "epoch": 9.073724007561436,
      "grad_norm": NaN,
      "learning_rate": 0.0005,
      "loss": 0.0,
      "step": 4800
    },
    {
      "epoch": 9.262759924385634,
      "grad_norm": NaN,
      "learning_rate": 0.0005,
      "loss": 0.0,
      "step": 4900
    },
    {
      "epoch": 9.45179584120983,
      "grad_norm": NaN,
      "learning_rate": 0.0005,
      "loss": 0.0,
      "step": 5000
    },
    {
      "epoch": 9.640831758034027,
      "grad_norm": NaN,
      "learning_rate": 0.0005,
      "loss": 0.0,
      "step": 5100
    },
    {
      "epoch": 9.829867674858223,
      "grad_norm": NaN,
      "learning_rate": 0.0005,
      "loss": 0.0,
      "step": 5200
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.21060089686098654,
      "eval_loss": NaN,
      "eval_runtime": 5.5143,
      "eval_samples_per_second": 90.673,
      "eval_steps_per_second": 11.425,
      "step": 5290
    },
    {
      "epoch": 10.0,
      "eval_exact_match": 0.0,
      "eval_f1": 0.0,
      "eval_qa_bleu": 0.0,
      "eval_qa_exact_match": 0.0,
      "eval_recite_bleu": 0.0,
      "eval_recite_exact_match": 0.0,
      "step": 5290
    },
    {
      "epoch": 10.01890359168242,
      "grad_norm": NaN,
      "learning_rate": 0.0005,
      "loss": 0.0,
      "step": 5300
    },
    {
      "epoch": 10.207939508506616,
      "grad_norm": NaN,
      "learning_rate": 0.0005,
      "loss": 0.0,
      "step": 5400
    },
    {
      "epoch": 10.396975425330814,
      "grad_norm": NaN,
      "learning_rate": 0.0005,
      "loss": 0.0,
      "step": 5500
    },
    {
      "epoch": 10.58601134215501,
      "grad_norm": NaN,
      "learning_rate": 0.0005,
      "loss": 0.0,
      "step": 5600
    },
    {
      "epoch": 10.775047258979207,
      "grad_norm": NaN,
      "learning_rate": 0.0005,
      "loss": 0.0,
      "step": 5700
    },
    {
      "epoch": 10.964083175803403,
      "grad_norm": NaN,
      "learning_rate": 0.0005,
      "loss": 0.0,
      "step": 5800
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.21060089686098654,
      "eval_loss": NaN,
      "eval_runtime": 6.5846,
      "eval_samples_per_second": 75.934,
      "eval_steps_per_second": 9.568,
      "step": 5819
    },
    {
      "epoch": 11.0,
      "eval_exact_match": 0.0,
      "eval_f1": 0.0,
      "eval_qa_bleu": 0.0,
      "eval_qa_exact_match": 0.0,
      "eval_recite_bleu": 0.0,
      "eval_recite_exact_match": 0.0,
      "step": 5819
    },
    {
      "epoch": 11.1531190926276,
      "grad_norm": NaN,
      "learning_rate": 0.0005,
      "loss": 0.0,
      "step": 5900
    },
    {
      "epoch": 11.342155009451796,
      "grad_norm": NaN,
      "learning_rate": 0.0005,
      "loss": 0.0,
      "step": 6000
    },
    {
      "epoch": 11.531190926275993,
      "grad_norm": NaN,
      "learning_rate": 0.0005,
      "loss": 0.0,
      "step": 6100
    },
    {
      "epoch": 11.720226843100189,
      "grad_norm": NaN,
      "learning_rate": 0.0005,
      "loss": 0.0,
      "step": 6200
    },
    {
      "epoch": 11.909262759924385,
      "grad_norm": NaN,
      "learning_rate": 0.0005,
      "loss": 0.0,
      "step": 6300
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.21060089686098654,
      "eval_loss": NaN,
      "eval_runtime": 5.4929,
      "eval_samples_per_second": 91.027,
      "eval_steps_per_second": 11.469,
      "step": 6348
    },
    {
      "epoch": 12.0,
      "eval_exact_match": 0.0,
      "eval_f1": 0.0,
      "eval_qa_bleu": 0.0,
      "eval_qa_exact_match": 0.0,
      "eval_recite_bleu": 0.0,
      "eval_recite_exact_match": 0.0,
      "step": 6348
    }
  ],
  "logging_steps": 100,
  "max_steps": 26450,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 50,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 5.785483100511273e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}