checkpoint-180/trainer_state.json

{
  "best_metric": 1.1614402532577515,
  "best_model_checkpoint": "./mistral/22-04-24-Weni-WeniGPT-Agents-Mistral-1.0.12-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": 0.7089748978614807,
      "learning_rate": 0.00019999521087449523,
      "loss": 1.366,
      "step": 10
    },
    {
      "epoch": 0.35714285714285715,
      "grad_norm": 0.47184231877326965,
      "learning_rate": 0.00019942107065112286,
      "loss": 1.0042,
      "step": 20
    },
    {
      "epoch": 0.5357142857142857,
      "grad_norm": 0.4505631625652313,
      "learning_rate": 0.0001978954027238763,
      "loss": 0.8828,
      "step": 30
    },
    {
      "epoch": 0.5357142857142857,
      "eval_loss": 1.1399036645889282,
      "eval_runtime": 38.6638,
      "eval_samples_per_second": 1.19,
      "eval_steps_per_second": 0.31,
      "step": 30
    },
    {
      "epoch": 0.7142857142857143,
      "grad_norm": 0.4148388206958771,
      "learning_rate": 0.00019543280877920072,
      "loss": 0.806,
      "step": 40
    },
    {
      "epoch": 0.8928571428571429,
      "grad_norm": 0.3701043725013733,
      "learning_rate": 0.00019205685752689177,
      "loss": 0.771,
      "step": 50
    },
    {
      "epoch": 1.0714285714285714,
      "grad_norm": 0.39483487606048584,
      "learning_rate": 0.00018779985913140924,
      "loss": 0.6609,
      "step": 60
    },
    {
      "epoch": 1.0714285714285714,
      "eval_loss": 1.1470385789871216,
      "eval_runtime": 38.654,
      "eval_samples_per_second": 1.19,
      "eval_steps_per_second": 0.31,
      "step": 60
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.40163591504096985,
      "learning_rate": 0.0001827025559814854,
      "loss": 0.4089,
      "step": 70
    },
    {
      "epoch": 1.4285714285714286,
      "grad_norm": 0.3985687792301178,
      "learning_rate": 0.0001768137327575751,
      "loss": 0.3764,
      "step": 80
    },
    {
      "epoch": 1.6071428571428572,
      "grad_norm": 0.3154538869857788,
      "learning_rate": 0.00017018974952906884,
      "loss": 0.3918,
      "step": 90
    },
    {
      "epoch": 1.6071428571428572,
      "eval_loss": 1.1614402532577515,
      "eval_runtime": 38.6493,
      "eval_samples_per_second": 1.19,
      "eval_steps_per_second": 0.31,
      "step": 90
    },
    {
      "epoch": 1.7857142857142856,
      "grad_norm": 0.5162068009376526,
      "learning_rate": 0.0001628940023498477,
      "loss": 0.4058,
      "step": 100
    },
    {
      "epoch": 1.9642857142857144,
      "grad_norm": 0.2627800703048706,
      "learning_rate": 0.00015499631651465085,
      "loss": 0.2883,
      "step": 110
    },
    {
      "epoch": 2.142857142857143,
      "grad_norm": 0.42833977937698364,
      "learning_rate": 0.00014657227828320635,
      "loss": 0.1679,
      "step": 120
    },
    {
      "epoch": 2.142857142857143,
      "eval_loss": 1.2334527969360352,
      "eval_runtime": 38.6496,
      "eval_samples_per_second": 1.19,
      "eval_steps_per_second": 0.31,
      "step": 120
    },
    {
      "epoch": 2.3214285714285716,
      "grad_norm": 0.39060840010643005,
      "learning_rate": 0.000137702511467984,
      "loss": 0.1706,
      "step": 130
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.273832231760025,
      "learning_rate": 0.0001284719058091194,
      "loss": 0.1654,
      "step": 140
    },
    {
      "epoch": 2.678571428571429,
      "grad_norm": 0.3711472451686859,
      "learning_rate": 0.00011896880452149077,
      "loss": 0.1361,
      "step": 150
    },
    {
      "epoch": 2.678571428571429,
      "eval_loss": 1.2611552476882935,
      "eval_runtime": 38.655,
      "eval_samples_per_second": 1.19,
      "eval_steps_per_second": 0.31,
      "step": 150
    },
    {
      "epoch": 2.857142857142857,
      "grad_norm": 0.3359326720237732,
      "learning_rate": 0.00010928415878967781,
      "loss": 0.1492,
      "step": 160
    },
    {
      "epoch": 3.0357142857142856,
      "grad_norm": 0.3263035714626312,
      "learning_rate": 9.951065730286553e-05,
      "loss": 0.1239,
      "step": 170
    },
    {
      "epoch": 3.2142857142857144,
      "grad_norm": 0.2739813029766083,
      "learning_rate": 8.974183916063968e-05,
      "loss": 0.0571,
      "step": 180
    },
    {
      "epoch": 3.2142857142857144,
      "eval_loss": 1.3478918075561523,
      "eval_runtime": 38.6443,
      "eval_samples_per_second": 1.19,
      "eval_steps_per_second": 0.311,
      "step": 180
    }
  ],
  "logging_steps": 10,
  "max_steps": 330,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 90,
  "total_flos": 1.8702894437184307e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}