checkpoint-180/trainer_state.json

{
  "best_metric": 1.1286808252334595,
  "best_model_checkpoint": "./mistral/22-04-24-Weni-WeniGPT-Agents-Mistral-1.0.11-SFT_Experiment with SFT and a new tokenizer configuration for chat template of mistral-2_max_steps-330_batch_8_2024-04-22_ppid_9/checkpoint-90",
  "epoch": 3.2142857142857144,
  "eval_steps": 30,
  "global_step": 180,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.17857142857142858,
      "grad_norm": 0.45132189989089966,
      "learning_rate": 0.00019999521087449523,
      "loss": 1.5157,
      "step": 10
    },
    {
      "epoch": 0.35714285714285715,
      "grad_norm": 0.25432512164115906,
      "learning_rate": 0.00019942107065112286,
      "loss": 1.1691,
      "step": 20
    },
    {
      "epoch": 0.5357142857142857,
      "grad_norm": 0.30843719840049744,
      "learning_rate": 0.0001978954027238763,
      "loss": 1.0466,
      "step": 30
    },
    {
      "epoch": 0.5357142857142857,
      "eval_loss": 1.1547143459320068,
      "eval_runtime": 36.25,
      "eval_samples_per_second": 1.269,
      "eval_steps_per_second": 0.331,
      "step": 30
    },
    {
      "epoch": 0.7142857142857143,
      "grad_norm": 0.35677382349967957,
      "learning_rate": 0.00019543280877920072,
      "loss": 1.0075,
      "step": 40
    },
    {
      "epoch": 0.8928571428571429,
      "grad_norm": 0.3124666213989258,
      "learning_rate": 0.00019205685752689177,
      "loss": 0.9662,
      "step": 50
    },
    {
      "epoch": 1.0714285714285714,
      "grad_norm": 0.3985981047153473,
      "learning_rate": 0.00018779985913140924,
      "loss": 0.8783,
      "step": 60
    },
    {
      "epoch": 1.0714285714285714,
      "eval_loss": 1.12884521484375,
      "eval_runtime": 36.2483,
      "eval_samples_per_second": 1.269,
      "eval_steps_per_second": 0.331,
      "step": 60
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.40197721123695374,
      "learning_rate": 0.0001827025559814854,
      "loss": 0.7525,
      "step": 70
    },
    {
      "epoch": 1.4285714285714286,
      "grad_norm": 0.4119006395339966,
      "learning_rate": 0.0001768137327575751,
      "loss": 0.6832,
      "step": 80
    },
    {
      "epoch": 1.6071428571428572,
      "grad_norm": 0.3348449766635895,
      "learning_rate": 0.00017018974952906884,
      "loss": 0.6778,
      "step": 90
    },
    {
      "epoch": 1.6071428571428572,
      "eval_loss": 1.1286808252334595,
      "eval_runtime": 36.2515,
      "eval_samples_per_second": 1.269,
      "eval_steps_per_second": 0.331,
      "step": 90
    },
    {
      "epoch": 1.7857142857142856,
      "grad_norm": 0.49415308237075806,
      "learning_rate": 0.0001628940023498477,
      "loss": 0.682,
      "step": 100
    },
    {
      "epoch": 1.9642857142857144,
      "grad_norm": 0.28949180245399475,
      "learning_rate": 0.00015499631651465085,
      "loss": 0.544,
      "step": 110
    },
    {
      "epoch": 2.142857142857143,
      "grad_norm": 0.4506300687789917,
      "learning_rate": 0.00014657227828320635,
      "loss": 0.4609,
      "step": 120
    },
    {
      "epoch": 2.142857142857143,
      "eval_loss": 1.1452628374099731,
      "eval_runtime": 36.2339,
      "eval_samples_per_second": 1.27,
      "eval_steps_per_second": 0.331,
      "step": 120
    },
    {
      "epoch": 2.3214285714285716,
      "grad_norm": 0.49903905391693115,
      "learning_rate": 0.000137702511467984,
      "loss": 0.5106,
      "step": 130
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.422453910112381,
      "learning_rate": 0.0001284719058091194,
      "loss": 0.4149,
      "step": 140
    },
    {
      "epoch": 2.678571428571429,
      "grad_norm": 0.536363959312439,
      "learning_rate": 0.00011896880452149077,
      "loss": 0.3946,
      "step": 150
    },
    {
      "epoch": 2.678571428571429,
      "eval_loss": 1.1766327619552612,
      "eval_runtime": 36.2209,
      "eval_samples_per_second": 1.27,
      "eval_steps_per_second": 0.331,
      "step": 150
    },
    {
      "epoch": 2.857142857142857,
      "grad_norm": 0.5005641579627991,
      "learning_rate": 0.00010928415878967781,
      "loss": 0.4512,
      "step": 160
    },
    {
      "epoch": 3.0357142857142856,
      "grad_norm": 0.4409847557544708,
      "learning_rate": 9.951065730286553e-05,
      "loss": 0.3759,
      "step": 170
    },
    {
      "epoch": 3.2142857142857144,
      "grad_norm": 0.33086544275283813,
      "learning_rate": 8.974183916063968e-05,
      "loss": 0.2669,
      "step": 180
    },
    {
      "epoch": 3.2142857142857144,
      "eval_loss": 1.212720274925232,
      "eval_runtime": 36.2554,
      "eval_samples_per_second": 1.269,
      "eval_steps_per_second": 0.331,
      "step": 180
    }
  ],
  "logging_steps": 10,
  "max_steps": 330,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 90,
  "total_flos": 1.7291593631399936e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}