checkpoint-3000/trainer_state.json

{
  "best_metric": 1.2202869653701782,
  "best_model_checkpoint": "./output/clip-finetuned-csu-p14-336-e4l58-l/checkpoint-3000",
  "epoch": 0.3199658703071672,
  "eval_steps": 500,
  "global_step": 3000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.05332764505119454,
      "grad_norm": 217.55360412597656,
      "learning_rate": 4.933340443686007e-08,
      "loss": 0.4667,
      "step": 500
    },
    {
      "epoch": 0.05332764505119454,
      "eval_loss": 1.4426143169403076,
      "eval_runtime": 61.9901,
      "eval_samples_per_second": 15.922,
      "eval_steps_per_second": 2.0,
      "step": 500
    },
    {
      "epoch": 0.10665529010238908,
      "grad_norm": 73.73340606689453,
      "learning_rate": 4.8666808873720136e-08,
      "loss": 0.4532,
      "step": 1000
    },
    {
      "epoch": 0.10665529010238908,
      "eval_loss": 1.3815597295761108,
      "eval_runtime": 62.4624,
      "eval_samples_per_second": 15.801,
      "eval_steps_per_second": 1.985,
      "step": 1000
    },
    {
      "epoch": 0.1599829351535836,
      "grad_norm": 471.4011535644531,
      "learning_rate": 4.80002133105802e-08,
      "loss": 0.3749,
      "step": 1500
    },
    {
      "epoch": 0.1599829351535836,
      "eval_loss": 1.3310539722442627,
      "eval_runtime": 63.3918,
      "eval_samples_per_second": 15.57,
      "eval_steps_per_second": 1.956,
      "step": 1500
    },
    {
      "epoch": 0.21331058020477817,
      "grad_norm": 11.962693214416504,
      "learning_rate": 4.733361774744027e-08,
      "loss": 0.336,
      "step": 2000
    },
    {
      "epoch": 0.21331058020477817,
      "eval_loss": 1.2890639305114746,
      "eval_runtime": 63.177,
      "eval_samples_per_second": 15.623,
      "eval_steps_per_second": 1.963,
      "step": 2000
    },
    {
      "epoch": 0.2666382252559727,
      "grad_norm": 375.1609191894531,
      "learning_rate": 4.666702218430034e-08,
      "loss": 0.3585,
      "step": 2500
    },
    {
      "epoch": 0.2666382252559727,
      "eval_loss": 1.2536433935165405,
      "eval_runtime": 63.1512,
      "eval_samples_per_second": 15.629,
      "eval_steps_per_second": 1.964,
      "step": 2500
    },
    {
      "epoch": 0.3199658703071672,
      "grad_norm": 0.0022936267778277397,
      "learning_rate": 4.600042662116041e-08,
      "loss": 0.303,
      "step": 3000
    },
    {
      "epoch": 0.3199658703071672,
      "eval_loss": 1.2202869653701782,
      "eval_runtime": 63.5857,
      "eval_samples_per_second": 15.522,
      "eval_steps_per_second": 1.95,
      "step": 3000
    }
  ],
  "logging_steps": 500,
  "max_steps": 37504,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1079994570482520.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}