{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.998719590268886,
  "eval_steps": 1000,
  "global_step": 9366,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00032030749519538755,
      "grad_norm": 6.058427333831787,
      "learning_rate": 0.0,
      "loss": 2.3526,
      "step": 1
    },
    {
      "epoch": 0.3203074951953876,
      "grad_norm": 3.9029946327209473,
      "learning_rate": 2.977933325424131e-05,
      "loss": 1.6965,
      "step": 1000
    },
    {
      "epoch": 0.3203074951953876,
      "eval_cosine_accuracy": 0.9661373496055603,
      "eval_loss": 0.7820664644241333,
      "eval_runtime": 28.1656,
      "eval_samples_per_second": 337.61,
      "eval_steps_per_second": 1.349,
      "step": 1000
    },
    {
      "epoch": 0.6406149903907752,
      "grad_norm": 3.5058798789978027,
      "learning_rate": 2.6220192193617276e-05,
      "loss": 1.1157,
      "step": 2000
    },
    {
      "epoch": 0.6406149903907752,
      "eval_cosine_accuracy": 0.9707645177841187,
      "eval_loss": 0.7116907835006714,
      "eval_runtime": 27.4286,
      "eval_samples_per_second": 346.681,
      "eval_steps_per_second": 1.385,
      "step": 2000
    },
    {
      "epoch": 0.9609224855861627,
      "grad_norm": 3.5930891036987305,
      "learning_rate": 2.2661051132993238e-05,
      "loss": 1.5779,
      "step": 3000
    },
    {
      "epoch": 0.9609224855861627,
      "eval_cosine_accuracy": 0.9699232578277588,
      "eval_loss": 0.7110825777053833,
      "eval_runtime": 27.7293,
      "eval_samples_per_second": 342.922,
      "eval_steps_per_second": 1.37,
      "step": 3000
    },
    {
      "epoch": 1.2810499359795133,
      "grad_norm": 4.003444671630859,
      "learning_rate": 1.9101910072369202e-05,
      "loss": 1.2526,
      "step": 4000
    },
    {
      "epoch": 1.2810499359795133,
      "eval_cosine_accuracy": 0.9773898124694824,
      "eval_loss": 0.6422452330589294,
      "eval_runtime": 27.4782,
      "eval_samples_per_second": 346.057,
      "eval_steps_per_second": 1.383,
      "step": 4000
    },
    {
      "epoch": 1.6011523687580027,
      "grad_norm": 3.0391690731048584,
      "learning_rate": 1.5549887293866415e-05,
      "loss": 1.1009,
      "step": 5000
    },
    {
      "epoch": 1.6011523687580027,
      "eval_cosine_accuracy": 0.9776001572608948,
      "eval_loss": 0.6362274885177612,
      "eval_runtime": 29.0916,
      "eval_samples_per_second": 326.864,
      "eval_steps_per_second": 1.306,
      "step": 5000
    },
    {
      "epoch": 1.9212548015364916,
      "grad_norm": 3.6326820850372314,
      "learning_rate": 1.1990746233242378e-05,
      "loss": 1.0505,
      "step": 6000
    },
    {
      "epoch": 1.9212548015364916,
      "eval_cosine_accuracy": 0.9786518216133118,
      "eval_loss": 0.6349849104881287,
      "eval_runtime": 28.408,
      "eval_samples_per_second": 334.73,
      "eval_steps_per_second": 1.338,
      "step": 6000
    },
    {
      "epoch": 2.2413572343149806,
      "grad_norm": 2.998448133468628,
      "learning_rate": 8.431605172618341e-06,
      "loss": 0.9656,
      "step": 7000
    },
    {
      "epoch": 2.2413572343149806,
      "eval_cosine_accuracy": 0.9782311320304871,
      "eval_loss": 0.6253951191902161,
      "eval_runtime": 29.1172,
      "eval_samples_per_second": 326.577,
      "eval_steps_per_second": 1.305,
      "step": 7000
    },
    {
      "epoch": 2.56145966709347,
      "grad_norm": 3.0235984325408936,
      "learning_rate": 4.876023253054929e-06,
      "loss": 0.9874,
      "step": 8000
    },
    {
      "epoch": 2.56145966709347,
      "eval_cosine_accuracy": 0.9779156446456909,
      "eval_loss": 0.6308476328849792,
      "eval_runtime": 29.1512,
      "eval_samples_per_second": 326.195,
      "eval_steps_per_second": 1.304,
      "step": 8000
    },
    {
      "epoch": 2.881562099871959,
      "grad_norm": 3.0960497856140137,
      "learning_rate": 1.3168821924308935e-06,
      "loss": 0.9667,
      "step": 9000
    },
    {
      "epoch": 2.881562099871959,
      "eval_cosine_accuracy": 0.9782311320304871,
      "eval_loss": 0.627689778804779,
      "eval_runtime": 28.5859,
      "eval_samples_per_second": 332.647,
      "eval_steps_per_second": 1.329,
      "step": 9000
    }
  ],
  "logging_steps": 1000,
  "max_steps": 9366,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}