checkpoint-200/trainer_state.json

{
  "best_metric": 0.7773334980010986,
  "best_model_checkpoint": "../../models/lora/codellama-7b-multitask_cot_reflection/checkpoint-200",
  "epoch": 12.5,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.62,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.4326,
      "step": 10
    },
    {
      "epoch": 1.25,
      "learning_rate": 5.9999999999999995e-05,
      "loss": 2.3071,
      "step": 20
    },
    {
      "epoch": 1.88,
      "learning_rate": 8.699999999999999e-05,
      "loss": 2.0883,
      "step": 30
    },
    {
      "epoch": 2.5,
      "learning_rate": 0.000117,
      "loss": 1.8582,
      "step": 40
    },
    {
      "epoch": 3.12,
      "learning_rate": 0.00014399999999999998,
      "loss": 1.4434,
      "step": 50
    },
    {
      "epoch": 3.75,
      "learning_rate": 0.00017399999999999997,
      "loss": 1.1487,
      "step": 60
    },
    {
      "epoch": 4.38,
      "learning_rate": 0.000204,
      "loss": 1.042,
      "step": 70
    },
    {
      "epoch": 5.0,
      "learning_rate": 0.000234,
      "loss": 1.0089,
      "step": 80
    },
    {
      "epoch": 5.62,
      "learning_rate": 0.00026399999999999997,
      "loss": 0.9878,
      "step": 90
    },
    {
      "epoch": 6.25,
      "learning_rate": 0.000294,
      "loss": 0.9677,
      "step": 100
    },
    {
      "epoch": 6.88,
      "learning_rate": 0.0002936842105263158,
      "loss": 0.9404,
      "step": 110
    },
    {
      "epoch": 7.5,
      "learning_rate": 0.0002857894736842105,
      "loss": 0.9229,
      "step": 120
    },
    {
      "epoch": 8.12,
      "learning_rate": 0.00027789473684210523,
      "loss": 0.9233,
      "step": 130
    },
    {
      "epoch": 8.75,
      "learning_rate": 0.00027,
      "loss": 0.8922,
      "step": 140
    },
    {
      "epoch": 9.38,
      "learning_rate": 0.0002621052631578947,
      "loss": 0.8933,
      "step": 150
    },
    {
      "epoch": 10.0,
      "learning_rate": 0.00025421052631578945,
      "loss": 0.8536,
      "step": 160
    },
    {
      "epoch": 10.62,
      "learning_rate": 0.00024631578947368417,
      "loss": 0.8694,
      "step": 170
    },
    {
      "epoch": 11.25,
      "learning_rate": 0.00023842105263157895,
      "loss": 0.8344,
      "step": 180
    },
    {
      "epoch": 11.88,
      "learning_rate": 0.00023052631578947364,
      "loss": 0.8342,
      "step": 190
    },
    {
      "epoch": 12.5,
      "learning_rate": 0.0002226315789473684,
      "loss": 0.8227,
      "step": 200
    },
    {
      "epoch": 12.5,
      "eval_loss": 0.7773334980010986,
      "eval_runtime": 0.9943,
      "eval_samples_per_second": 21.12,
      "eval_steps_per_second": 3.017,
      "step": 200
    }
  ],
  "max_steps": 480,
  "num_train_epochs": 30,
  "total_flos": 2.669105373043753e+17,
  "trial_name": null,
  "trial_params": null
}