last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0344827586206897,
  "eval_steps": 3,
  "global_step": 15,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.07,
      "learning_rate": 0.0,
      "loss": 11.0816,
      "step": 1
    },
    {
      "epoch": 0.07,
      "eval_loss": 11.230381965637207,
      "eval_runtime": 114.2254,
      "eval_samples_per_second": 0.219,
      "eval_steps_per_second": 0.219,
      "step": 1
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0,
      "loss": 10.9422,
      "step": 2
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0,
      "loss": 11.0902,
      "step": 3
    },
    {
      "epoch": 0.21,
      "eval_loss": 11.230381965637207,
      "eval_runtime": 114.2325,
      "eval_samples_per_second": 0.219,
      "eval_steps_per_second": 0.219,
      "step": 3
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0,
      "loss": 11.0403,
      "step": 4
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0,
      "loss": 11.0741,
      "step": 5
    },
    {
      "epoch": 0.41,
      "learning_rate": 2e-05,
      "loss": 10.9716,
      "step": 6
    },
    {
      "epoch": 0.41,
      "eval_loss": 11.230381965637207,
      "eval_runtime": 114.2347,
      "eval_samples_per_second": 0.219,
      "eval_steps_per_second": 0.219,
      "step": 6
    },
    {
      "epoch": 0.48,
      "learning_rate": 4e-05,
      "loss": 11.0487,
      "step": 7
    },
    {
      "epoch": 0.55,
      "learning_rate": 6e-05,
      "loss": 10.9796,
      "step": 8
    },
    {
      "epoch": 0.62,
      "learning_rate": 8e-05,
      "loss": 10.963,
      "step": 9
    },
    {
      "epoch": 0.62,
      "eval_loss": 11.029632568359375,
      "eval_runtime": 114.2296,
      "eval_samples_per_second": 0.219,
      "eval_steps_per_second": 0.219,
      "step": 9
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0001,
      "loss": 10.8003,
      "step": 10
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.00012,
      "loss": 10.5483,
      "step": 11
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.00014,
      "loss": 10.2673,
      "step": 12
    },
    {
      "epoch": 0.83,
      "eval_loss": 10.199254035949707,
      "eval_runtime": 114.2358,
      "eval_samples_per_second": 0.219,
      "eval_steps_per_second": 0.219,
      "step": 12
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.00014,
      "loss": 9.78,
      "step": 13
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.00016,
      "loss": 9.8429,
      "step": 14
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.00018,
      "loss": 9.5715,
      "step": 15
    },
    {
      "epoch": 1.03,
      "eval_loss": 10.097070693969727,
      "eval_runtime": 114.2343,
      "eval_samples_per_second": 0.219,
      "eval_steps_per_second": 0.219,
      "step": 15
    }
  ],
  "logging_steps": 1,
  "max_steps": 42,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 2744968795914240.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}