{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 20.0,
  "eval_steps": 500,
  "global_step": 2440,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 78.255126953125,
      "learning_rate": 4.75e-05,
      "loss": 0.3942,
      "step": 122
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8796992481203008,
      "eval_f1": 0.8419489007724301,
      "eval_loss": 0.3128369450569153,
      "eval_precision": 0.8857758620689655,
      "eval_recall": 0.8173758865248226,
      "eval_runtime": 1.6299,
      "eval_samples_per_second": 244.801,
      "eval_steps_per_second": 30.677,
      "step": 122
    },
    {
      "epoch": 2.0,
      "grad_norm": 42.82415771484375,
      "learning_rate": 4.5e-05,
      "loss": 0.2168,
      "step": 244
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8897243107769424,
      "eval_f1": 0.8676337535436396,
      "eval_loss": 0.3043781518936157,
      "eval_precision": 0.8658613445378152,
      "eval_recall": 0.8694762684124386,
      "eval_runtime": 1.6375,
      "eval_samples_per_second": 243.661,
      "eval_steps_per_second": 30.534,
      "step": 244
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.2970781624317169,
      "learning_rate": 4.25e-05,
      "loss": 0.1372,
      "step": 366
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8897243107769424,
      "eval_f1": 0.8595250288055307,
      "eval_loss": 0.5317866802215576,
      "eval_precision": 0.885164197446576,
      "eval_recall": 0.8419712675031824,
      "eval_runtime": 1.6412,
      "eval_samples_per_second": 243.114,
      "eval_steps_per_second": 30.465,
      "step": 366
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.16418644785881042,
      "learning_rate": 4e-05,
      "loss": 0.0957,
      "step": 488
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8947368421052632,
      "eval_f1": 0.8765906680805938,
      "eval_loss": 0.47654101252555847,
      "eval_precision": 0.8675710594315245,
      "eval_recall": 0.888025095471904,
      "eval_runtime": 1.6551,
      "eval_samples_per_second": 241.073,
      "eval_steps_per_second": 30.21,
      "step": 488
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.4955180287361145,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.0674,
      "step": 610
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8872180451127819,
      "eval_f1": 0.8728804559453431,
      "eval_loss": 0.552257239818573,
      "eval_precision": 0.8576773985140519,
      "eval_recall": 0.9027095835606473,
      "eval_runtime": 1.6807,
      "eval_samples_per_second": 237.402,
      "eval_steps_per_second": 29.75,
      "step": 610
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.03946012258529663,
      "learning_rate": 3.5e-05,
      "loss": 0.0535,
      "step": 732
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9072681704260651,
      "eval_f1": 0.8878574955372402,
      "eval_loss": 0.5158531069755554,
      "eval_precision": 0.8888448885098087,
      "eval_recall": 0.8868885251863976,
      "eval_runtime": 1.6465,
      "eval_samples_per_second": 242.338,
      "eval_steps_per_second": 30.368,
      "step": 732
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.008608223870396614,
      "learning_rate": 3.2500000000000004e-05,
      "loss": 0.027,
      "step": 854
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8872180451127819,
      "eval_f1": 0.8642908431276217,
      "eval_loss": 0.5940884351730347,
      "eval_precision": 0.8633964654080464,
      "eval_recall": 0.8652027641389344,
      "eval_runtime": 1.6485,
      "eval_samples_per_second": 242.043,
      "eval_steps_per_second": 30.331,
      "step": 854
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.010127891786396503,
      "learning_rate": 3e-05,
      "loss": 0.0223,
      "step": 976
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.8796992481203008,
      "eval_f1": 0.8548827059465357,
      "eval_loss": 0.7166243195533752,
      "eval_precision": 0.8548827059465357,
      "eval_recall": 0.8548827059465357,
      "eval_runtime": 1.6562,
      "eval_samples_per_second": 240.913,
      "eval_steps_per_second": 30.19,
      "step": 976
    },
    {
      "epoch": 9.0,
      "grad_norm": 0.005933025386184454,
      "learning_rate": 2.7500000000000004e-05,
      "loss": 0.0145,
      "step": 1098
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.9022556390977443,
      "eval_f1": 0.8829621606985718,
      "eval_loss": 0.7022837996482849,
      "eval_precision": 0.8802419354838709,
      "eval_recall": 0.8858428805237315,
      "eval_runtime": 1.6595,
      "eval_samples_per_second": 240.429,
      "eval_steps_per_second": 30.129,
      "step": 1098
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.02505210041999817,
      "learning_rate": 2.5e-05,
      "loss": 0.0106,
      "step": 1220
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.9047619047619048,
      "eval_f1": 0.8839406001224739,
      "eval_loss": 0.699307918548584,
      "eval_precision": 0.8880654743486602,
      "eval_recall": 0.880114566284779,
      "eval_runtime": 1.6551,
      "eval_samples_per_second": 241.07,
      "eval_steps_per_second": 30.209,
      "step": 1220
    },
    {
      "epoch": 11.0,
      "grad_norm": 0.002501419745385647,
      "learning_rate": 2.25e-05,
      "loss": 0.0093,
      "step": 1342
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.8947368421052632,
      "eval_f1": 0.8703663593044124,
      "eval_loss": 0.8273664712905884,
      "eval_precision": 0.8789149003479912,
      "eval_recall": 0.8630205491907619,
      "eval_runtime": 1.6583,
      "eval_samples_per_second": 240.615,
      "eval_steps_per_second": 30.152,
      "step": 1342
    },
    {
      "epoch": 12.0,
      "grad_norm": 0.012166227214038372,
      "learning_rate": 2e-05,
      "loss": 0.0086,
      "step": 1464
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.8972431077694235,
      "eval_f1": 0.8744522298370696,
      "eval_loss": 0.7971612215042114,
      "eval_precision": 0.8795731707317074,
      "eval_recall": 0.8697945080923805,
      "eval_runtime": 1.6712,
      "eval_samples_per_second": 238.744,
      "eval_steps_per_second": 29.918,
      "step": 1464
    },
    {
      "epoch": 13.0,
      "grad_norm": 0.00197013420984149,
      "learning_rate": 1.75e-05,
      "loss": 0.0106,
      "step": 1586
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.8972431077694235,
      "eval_f1": 0.8787009231453675,
      "eval_loss": 0.7591652870178223,
      "eval_precision": 0.8714896214896215,
      "eval_recall": 0.8872976904891798,
      "eval_runtime": 1.6672,
      "eval_samples_per_second": 239.329,
      "eval_steps_per_second": 29.991,
      "step": 1586
    },
    {
      "epoch": 14.0,
      "grad_norm": 0.0050615849904716015,
      "learning_rate": 1.5e-05,
      "loss": 0.0072,
      "step": 1708
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.899749373433584,
      "eval_f1": 0.8813841488792438,
      "eval_loss": 0.7834069728851318,
      "eval_precision": 0.8748029197080291,
      "eval_recall": 0.8890707401345699,
      "eval_runtime": 1.6555,
      "eval_samples_per_second": 241.019,
      "eval_steps_per_second": 30.203,
      "step": 1708
    },
    {
      "epoch": 15.0,
      "grad_norm": 0.002086851978674531,
      "learning_rate": 1.25e-05,
      "loss": 0.0098,
      "step": 1830
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.899749373433584,
      "eval_f1": 0.8802521008403361,
      "eval_loss": 0.8048883676528931,
      "eval_precision": 0.8767168083714847,
      "eval_recall": 0.8840698308783415,
      "eval_runtime": 1.6591,
      "eval_samples_per_second": 240.488,
      "eval_steps_per_second": 30.136,
      "step": 1830
    },
    {
      "epoch": 16.0,
      "grad_norm": 0.0012473827227950096,
      "learning_rate": 1e-05,
      "loss": 0.0058,
      "step": 1952
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.899749373433584,
      "eval_f1": 0.8802521008403361,
      "eval_loss": 0.7670984268188477,
      "eval_precision": 0.8767168083714847,
      "eval_recall": 0.8840698308783415,
      "eval_runtime": 1.659,
      "eval_samples_per_second": 240.503,
      "eval_steps_per_second": 30.138,
      "step": 1952
    },
    {
      "epoch": 17.0,
      "grad_norm": 0.00188881263602525,
      "learning_rate": 7.5e-06,
      "loss": 0.0035,
      "step": 2074
    },
    {
      "epoch": 17.0,
      "eval_accuracy": 0.9022556390977443,
      "eval_f1": 0.8856624319419237,
      "eval_loss": 0.8084732294082642,
      "eval_precision": 0.8758364312267658,
      "eval_recall": 0.8983451536643026,
      "eval_runtime": 1.6569,
      "eval_samples_per_second": 240.816,
      "eval_steps_per_second": 30.177,
      "step": 2074
    },
    {
      "epoch": 18.0,
      "grad_norm": 0.0014366944087669253,
      "learning_rate": 5e-06,
      "loss": 0.0052,
      "step": 2196
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.899749373433584,
      "eval_f1": 0.8808243727598566,
      "eval_loss": 0.7721081972122192,
      "eval_precision": 0.875706963591375,
      "eval_recall": 0.8865702855064557,
      "eval_runtime": 1.6546,
      "eval_samples_per_second": 241.143,
      "eval_steps_per_second": 30.218,
      "step": 2196
    },
    {
      "epoch": 19.0,
      "grad_norm": 0.0011094665387645364,
      "learning_rate": 2.5e-06,
      "loss": 0.0028,
      "step": 2318
    },
    {
      "epoch": 19.0,
      "eval_accuracy": 0.8972431077694235,
      "eval_f1": 0.8792560061999484,
      "eval_loss": 0.8358559608459473,
      "eval_precision": 0.8707622232472325,
      "eval_recall": 0.889798145117294,
      "eval_runtime": 1.6584,
      "eval_samples_per_second": 240.592,
      "eval_steps_per_second": 30.149,
      "step": 2318
    },
    {
      "epoch": 20.0,
      "grad_norm": 0.0015741140814498067,
      "learning_rate": 0.0,
      "loss": 0.0033,
      "step": 2440
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.8972431077694235,
      "eval_f1": 0.8792560061999484,
      "eval_loss": 0.8335620164871216,
      "eval_precision": 0.8707622232472325,
      "eval_recall": 0.889798145117294,
      "eval_runtime": 1.6776,
      "eval_samples_per_second": 237.834,
      "eval_steps_per_second": 29.804,
      "step": 2440
    },
    {
      "epoch": 20.0,
      "step": 2440,
      "total_flos": 7584162436176000.0,
      "train_loss": 0.05526667458356404,
      "train_runtime": 862.9394,
      "train_samples_per_second": 84.316,
      "train_steps_per_second": 2.828
    }
  ],
  "logging_steps": 500,
  "max_steps": 2440,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 20,
  "save_steps": 500,
  "total_flos": 7584162436176000.0,
  "train_batch_size": 30,
  "trial_name": null,
  "trial_params": null
}