checkpoint-58/trainer_state.json

{
  "best_metric": 1.8895907402038574,
  "best_model_checkpoint": "$../logs/03-08-2024T09-56-31-detailclip-b32/checkpoint-58",
  "epoch": 0.5043478260869565,
  "eval_steps": 58,
  "global_step": 58,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.05217391304347826,
      "grad_norm": 3.0711264610290527,
      "learning_rate": 1.2e-05,
      "loss": 0.8067,
      "step": 6
    },
    {
      "epoch": 0.10434782608695652,
      "grad_norm": 3.019757032394409,
      "learning_rate": 2.4e-05,
      "loss": 0.8037,
      "step": 12
    },
    {
      "epoch": 0.1565217391304348,
      "grad_norm": 3.080796003341675,
      "learning_rate": 3.6e-05,
      "loss": 0.835,
      "step": 18
    },
    {
      "epoch": 0.20869565217391303,
      "grad_norm": 2.9880104064941406,
      "learning_rate": 4.8e-05,
      "loss": 0.6869,
      "step": 24
    },
    {
      "epoch": 0.2608695652173913,
      "grad_norm": 2.5763087272644043,
      "learning_rate": 6e-05,
      "loss": 0.5712,
      "step": 30
    },
    {
      "epoch": 0.3130434782608696,
      "grad_norm": 2.295846462249756,
      "learning_rate": 7.2e-05,
      "loss": 0.5537,
      "step": 36
    },
    {
      "epoch": 0.3652173913043478,
      "grad_norm": 2.2098076343536377,
      "learning_rate": 8.4e-05,
      "loss": 0.481,
      "step": 42
    },
    {
      "epoch": 0.41739130434782606,
      "grad_norm": 1.7029411792755127,
      "learning_rate": 9.6e-05,
      "loss": 0.3933,
      "step": 48
    },
    {
      "epoch": 0.46956521739130436,
      "grad_norm": 2.0247466564178467,
      "learning_rate": 9.999673735634258e-05,
      "loss": 0.3398,
      "step": 54
    },
    {
      "epoch": 0.5043478260869565,
      "eval_loss": 1.8895907402038574,
      "eval_runtime": 362.9542,
      "eval_samples_per_second": 6.155,
      "eval_steps_per_second": 0.05,
      "step": 58
    },
    {
      "epoch": 0.5043478260869565,
      "neg_increments": 0.3482297461360652,
      "positive_increments": 0.5847584809616501,
      "rate_increments": 0.6160316723977289,
      "step": 58
    }
  ],
  "logging_steps": 6,
  "max_steps": 1150,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 58,
  "total_flos": 1318444128207360.0,
  "train_batch_size": 128,
  "trial_name": null,
  "trial_params": null
}