checkpoint-13/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.04,
  "eval_steps": 4,
  "global_step": 13,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.08,
      "grad_norm": 17.890169143676758,
      "learning_rate": 2e-05,
      "loss": 6.6299,
      "step": 1
    },
    {
      "epoch": 0.08,
      "eval_loss": 6.932005405426025,
      "eval_runtime": 5.2841,
      "eval_samples_per_second": 2.271,
      "eval_steps_per_second": 1.135,
      "step": 1
    },
    {
      "epoch": 0.16,
      "grad_norm": 17.0613956451416,
      "learning_rate": 4e-05,
      "loss": 6.8062,
      "step": 2
    },
    {
      "epoch": 0.24,
      "grad_norm": 19.356693267822266,
      "learning_rate": 6e-05,
      "loss": 7.0697,
      "step": 3
    },
    {
      "epoch": 0.32,
      "grad_norm": 12.688406944274902,
      "learning_rate": 8e-05,
      "loss": 5.9686,
      "step": 4
    },
    {
      "epoch": 0.32,
      "eval_loss": 4.446321964263916,
      "eval_runtime": 5.3077,
      "eval_samples_per_second": 2.261,
      "eval_steps_per_second": 1.13,
      "step": 4
    },
    {
      "epoch": 0.4,
      "grad_norm": 11.240882873535156,
      "learning_rate": 0.0001,
      "loss": 4.5161,
      "step": 5
    },
    {
      "epoch": 0.48,
      "grad_norm": 14.287125587463379,
      "learning_rate": 0.00012,
      "loss": 2.5736,
      "step": 6
    },
    {
      "epoch": 0.56,
      "grad_norm": 4.680168628692627,
      "learning_rate": 0.00014,
      "loss": 0.9119,
      "step": 7
    },
    {
      "epoch": 0.64,
      "grad_norm": 3.75272536277771,
      "learning_rate": 0.00016,
      "loss": 0.5956,
      "step": 8
    },
    {
      "epoch": 0.64,
      "eval_loss": 0.5577284693717957,
      "eval_runtime": 5.323,
      "eval_samples_per_second": 2.254,
      "eval_steps_per_second": 1.127,
      "step": 8
    },
    {
      "epoch": 0.72,
      "grad_norm": 7.569386959075928,
      "learning_rate": 0.00018,
      "loss": 0.6499,
      "step": 9
    },
    {
      "epoch": 0.8,
      "grad_norm": 14.337709426879883,
      "learning_rate": 0.0002,
      "loss": 1.6675,
      "step": 10
    },
    {
      "epoch": 0.88,
      "grad_norm": 3.6579396724700928,
      "learning_rate": 0.00019981755542233177,
      "loss": 0.5612,
      "step": 11
    },
    {
      "epoch": 0.96,
      "grad_norm": 2.5605533123016357,
      "learning_rate": 0.0001992708874098054,
      "loss": 0.4848,
      "step": 12
    },
    {
      "epoch": 0.96,
      "eval_loss": 0.8369883894920349,
      "eval_runtime": 5.327,
      "eval_samples_per_second": 2.253,
      "eval_steps_per_second": 1.126,
      "step": 12
    },
    {
      "epoch": 1.04,
      "grad_norm": 8.139214515686035,
      "learning_rate": 0.00019836199069471437,
      "loss": 0.9289,
      "step": 13
    }
  ],
  "logging_steps": 1,
  "max_steps": 62,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 13,
  "total_flos": 8463871748603904.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}