last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.0304,
  "eval_steps": 500,
  "global_step": 380,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 62.10089111328125,
      "learning_rate": 1e-06,
      "loss": 4.5777,
      "step": 20
    },
    {
      "epoch": 0.0,
      "grad_norm": 39.39016342163086,
      "learning_rate": 2e-06,
      "loss": 4.4077,
      "step": 40
    },
    {
      "epoch": 0.0,
      "grad_norm": 54.24020767211914,
      "learning_rate": 3e-06,
      "loss": 4.4807,
      "step": 60
    },
    {
      "epoch": 0.01,
      "grad_norm": 30.161609649658203,
      "learning_rate": 4e-06,
      "loss": 4.5756,
      "step": 80
    },
    {
      "epoch": 0.01,
      "grad_norm": 40.131675720214844,
      "learning_rate": 4.9999999999999996e-06,
      "loss": 4.4352,
      "step": 100
    },
    {
      "epoch": 0.01,
      "grad_norm": 52.3621940612793,
      "learning_rate": 6e-06,
      "loss": 4.5096,
      "step": 120
    },
    {
      "epoch": 0.01,
      "grad_norm": 49.86561584472656,
      "learning_rate": 7e-06,
      "loss": 4.493,
      "step": 140
    },
    {
      "epoch": 0.01,
      "grad_norm": 20.034923553466797,
      "learning_rate": 8e-06,
      "loss": 4.4088,
      "step": 160
    },
    {
      "epoch": 0.01,
      "grad_norm": 50.790679931640625,
      "learning_rate": 9e-06,
      "loss": 4.4901,
      "step": 180
    },
    {
      "epoch": 0.02,
      "grad_norm": 48.5693473815918,
      "learning_rate": 9.999999999999999e-06,
      "loss": 4.3628,
      "step": 200
    },
    {
      "epoch": 0.02,
      "grad_norm": 37.95353698730469,
      "learning_rate": 1.1e-05,
      "loss": 4.3298,
      "step": 220
    },
    {
      "epoch": 0.02,
      "grad_norm": 35.7153434753418,
      "learning_rate": 1.2e-05,
      "loss": 4.2839,
      "step": 240
    },
    {
      "epoch": 0.02,
      "grad_norm": 91.47773742675781,
      "learning_rate": 1.3000000000000001e-05,
      "loss": 4.1238,
      "step": 260
    },
    {
      "epoch": 0.02,
      "grad_norm": 23.16193389892578,
      "learning_rate": 1.4e-05,
      "loss": 4.1245,
      "step": 280
    },
    {
      "epoch": 0.02,
      "grad_norm": 28.304485321044922,
      "learning_rate": 1.5e-05,
      "loss": 4.2198,
      "step": 300
    },
    {
      "epoch": 0.03,
      "grad_norm": 34.03230285644531,
      "learning_rate": 1.6e-05,
      "loss": 4.0958,
      "step": 320
    },
    {
      "epoch": 0.03,
      "grad_norm": 29.786975860595703,
      "learning_rate": 1.7e-05,
      "loss": 4.024,
      "step": 340
    },
    {
      "epoch": 0.03,
      "grad_norm": 33.04754638671875,
      "learning_rate": 1.8e-05,
      "loss": 4.0832,
      "step": 360
    },
    {
      "epoch": 0.03,
      "grad_norm": 28.68460464477539,
      "learning_rate": 1.9e-05,
      "loss": 3.9827,
      "step": 380
    }
  ],
  "logging_steps": 20,
  "max_steps": 20000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 20,
  "total_flos": 911840340049920.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}