low-shot-task-specific/social_i_qa/best_model/trainer_state.json

{
  "best_metric": 0.21922020614147186,
  "best_model_checkpoint": "checkpoints/instrucode/low-shot-task-specific/social_i_qa/checkpoint-100",
  "epoch": 4.0,
  "eval_steps": 500,
  "global_step": 100,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.4,
      "learning_rate": 5.9999999999999995e-05,
      "loss": 6.145,
      "step": 10
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.00011999999999999999,
      "loss": 3.2951,
      "step": 20
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.47254127264022827,
      "eval_runtime": 8.3699,
      "eval_samples_per_second": 23.895,
      "eval_steps_per_second": 2.987,
      "step": 25
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.00017999999999999998,
      "loss": 0.5553,
      "step": 30
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.00023999999999999998,
      "loss": 0.3415,
      "step": 40
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0003,
      "loss": 0.3055,
      "step": 50
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.29911044239997864,
      "eval_runtime": 8.3444,
      "eval_samples_per_second": 23.968,
      "eval_steps_per_second": 2.996,
      "step": 50
    },
    {
      "epoch": 2.4,
      "learning_rate": 0.000285,
      "loss": 0.2157,
      "step": 60
    },
    {
      "epoch": 2.8,
      "learning_rate": 0.00027,
      "loss": 0.1871,
      "step": 70
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.2219252735376358,
      "eval_runtime": 8.3281,
      "eval_samples_per_second": 24.015,
      "eval_steps_per_second": 3.002,
      "step": 75
    },
    {
      "epoch": 3.2,
      "learning_rate": 0.00025499999999999996,
      "loss": 0.1417,
      "step": 80
    },
    {
      "epoch": 3.6,
      "learning_rate": 0.00023999999999999998,
      "loss": 0.12,
      "step": 90
    },
    {
      "epoch": 4.0,
      "learning_rate": 0.000225,
      "loss": 0.1079,
      "step": 100
    },
    {
      "epoch": 4.0,
      "eval_loss": 0.21922020614147186,
      "eval_runtime": 8.385,
      "eval_samples_per_second": 23.852,
      "eval_steps_per_second": 2.981,
      "step": 100
    }
  ],
  "logging_steps": 10,
  "max_steps": 250,
  "num_train_epochs": 10,
  "save_steps": 500,
  "total_flos": 1.656903891124224e+16,
  "trial_name": null,
  "trial_params": null
}