{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.27692307692307694,
  "eval_steps": 10,
  "global_step": 90,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.006153846153846154,
      "eval_loss": 1.4845428466796875,
      "eval_runtime": 1.2728,
      "eval_samples_per_second": 89.569,
      "eval_steps_per_second": 4.714,
      "step": 2
    },
    {
      "epoch": 0.03076923076923077,
      "grad_norm": 15.541728973388672,
      "learning_rate": 6.060606060606061e-06,
      "loss": 1.4404,
      "step": 10
    },
    {
      "epoch": 0.03076923076923077,
      "eval_loss": 0.9616736769676208,
      "eval_runtime": 1.1678,
      "eval_samples_per_second": 97.621,
      "eval_steps_per_second": 5.138,
      "step": 10
    },
    {
      "epoch": 0.06153846153846154,
      "grad_norm": 12.678512573242188,
      "learning_rate": 1.2121212121212122e-05,
      "loss": 0.8308,
      "step": 20
    },
    {
      "epoch": 0.06153846153846154,
      "eval_loss": 0.7126240134239197,
      "eval_runtime": 1.1678,
      "eval_samples_per_second": 97.621,
      "eval_steps_per_second": 5.138,
      "step": 20
    },
    {
      "epoch": 0.09230769230769231,
      "grad_norm": 10.91329574584961,
      "learning_rate": 1.8181818181818182e-05,
      "loss": 0.8042,
      "step": 30
    },
    {
      "epoch": 0.09230769230769231,
      "eval_loss": 0.7227696180343628,
      "eval_runtime": 1.17,
      "eval_samples_per_second": 97.438,
      "eval_steps_per_second": 5.128,
      "step": 30
    },
    {
      "epoch": 0.12307692307692308,
      "grad_norm": 3.0046517848968506,
      "learning_rate": 1.9971653800228782e-05,
      "loss": 0.7589,
      "step": 40
    },
    {
      "epoch": 0.12307692307692308,
      "eval_loss": 0.6917378902435303,
      "eval_runtime": 1.169,
      "eval_samples_per_second": 97.523,
      "eval_steps_per_second": 5.133,
      "step": 40
    },
    {
      "epoch": 0.15384615384615385,
      "grad_norm": 4.805281162261963,
      "learning_rate": 1.9833201993305452e-05,
      "loss": 0.7804,
      "step": 50
    },
    {
      "epoch": 0.15384615384615385,
      "eval_loss": 0.7115028500556946,
      "eval_runtime": 1.1687,
      "eval_samples_per_second": 97.543,
      "eval_steps_per_second": 5.134,
      "step": 50
    },
    {
      "epoch": 0.18461538461538463,
      "grad_norm": 4.5656962394714355,
      "learning_rate": 1.9581037207470382e-05,
      "loss": 0.6548,
      "step": 60
    },
    {
      "epoch": 0.18461538461538463,
      "eval_loss": 0.7130362391471863,
      "eval_runtime": 1.1701,
      "eval_samples_per_second": 97.425,
      "eval_steps_per_second": 5.128,
      "step": 60
    },
    {
      "epoch": 0.2153846153846154,
      "grad_norm": 2.2376620769500732,
      "learning_rate": 1.9218075523263104e-05,
      "loss": 0.8105,
      "step": 70
    },
    {
      "epoch": 0.2153846153846154,
      "eval_loss": 0.7137090563774109,
      "eval_runtime": 1.1699,
      "eval_samples_per_second": 97.446,
      "eval_steps_per_second": 5.129,
      "step": 70
    },
    {
      "epoch": 0.24615384615384617,
      "grad_norm": 2.6304614543914795,
      "learning_rate": 1.874851429718765e-05,
      "loss": 0.6829,
      "step": 80
    },
    {
      "epoch": 0.24615384615384617,
      "eval_loss": 0.7175354361534119,
      "eval_runtime": 1.1691,
      "eval_samples_per_second": 97.512,
      "eval_steps_per_second": 5.132,
      "step": 80
    },
    {
      "epoch": 0.27692307692307694,
      "grad_norm": 3.8461191654205322,
      "learning_rate": 1.8177783622700328e-05,
      "loss": 0.6773,
      "step": 90
    },
    {
      "epoch": 0.27692307692307694,
      "eval_loss": 0.7199242115020752,
      "eval_runtime": 1.1693,
      "eval_samples_per_second": 97.494,
      "eval_steps_per_second": 5.131,
      "step": 90
    }
  ],
  "logging_steps": 10,
  "max_steps": 325,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 5,
  "total_flos": 5627427715809280.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}