checkpoint-4000/trainer_state.json

{
  "best_metric": 0.6952972412109375,
  "best_model_checkpoint": "./output/clip-finetuned-csu-p14-336-e4l57-l/checkpoint-4000",
  "epoch": 0.7369196757553427,
  "eval_steps": 500,
  "global_step": 4000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.09211495946941783,
      "grad_norm": 3.0293657779693604,
      "learning_rate": 4.884856300663227e-07,
      "loss": 0.271,
      "step": 500
    },
    {
      "epoch": 0.09211495946941783,
      "eval_loss": 1.069305419921875,
      "eval_runtime": 75.1409,
      "eval_samples_per_second": 16.063,
      "eval_steps_per_second": 2.01,
      "step": 500
    },
    {
      "epoch": 0.18422991893883567,
      "grad_norm": 7.3040852546691895,
      "learning_rate": 4.769712601326456e-07,
      "loss": 0.2493,
      "step": 1000
    },
    {
      "epoch": 0.18422991893883567,
      "eval_loss": 0.9427077770233154,
      "eval_runtime": 74.8926,
      "eval_samples_per_second": 16.116,
      "eval_steps_per_second": 2.016,
      "step": 1000
    },
    {
      "epoch": 0.2763448784082535,
      "grad_norm": 0.020736657083034515,
      "learning_rate": 4.654568901989683e-07,
      "loss": 0.2348,
      "step": 1500
    },
    {
      "epoch": 0.2763448784082535,
      "eval_loss": 0.8726764917373657,
      "eval_runtime": 76.6621,
      "eval_samples_per_second": 15.744,
      "eval_steps_per_second": 1.97,
      "step": 1500
    },
    {
      "epoch": 0.36845983787767134,
      "grad_norm": 5.3738909628009424e-05,
      "learning_rate": 4.5394252026529107e-07,
      "loss": 0.1552,
      "step": 2000
    },
    {
      "epoch": 0.36845983787767134,
      "eval_loss": 0.8325753808021545,
      "eval_runtime": 76.6127,
      "eval_samples_per_second": 15.755,
      "eval_steps_per_second": 1.971,
      "step": 2000
    },
    {
      "epoch": 0.46057479734708917,
      "grad_norm": 404.510009765625,
      "learning_rate": 4.4242815033161386e-07,
      "loss": 0.1753,
      "step": 2500
    },
    {
      "epoch": 0.46057479734708917,
      "eval_loss": 0.7550467848777771,
      "eval_runtime": 76.8024,
      "eval_samples_per_second": 15.716,
      "eval_steps_per_second": 1.966,
      "step": 2500
    },
    {
      "epoch": 0.552689756816507,
      "grad_norm": 370.7579650878906,
      "learning_rate": 4.3091378039793665e-07,
      "loss": 0.1659,
      "step": 3000
    },
    {
      "epoch": 0.552689756816507,
      "eval_loss": 0.719167172908783,
      "eval_runtime": 76.6327,
      "eval_samples_per_second": 15.75,
      "eval_steps_per_second": 1.97,
      "step": 3000
    },
    {
      "epoch": 0.6448047162859248,
      "grad_norm": 24.97766876220703,
      "learning_rate": 4.193994104642594e-07,
      "loss": 0.105,
      "step": 3500
    },
    {
      "epoch": 0.6448047162859248,
      "eval_loss": 0.7117515802383423,
      "eval_runtime": 77.1286,
      "eval_samples_per_second": 15.649,
      "eval_steps_per_second": 1.958,
      "step": 3500
    },
    {
      "epoch": 0.7369196757553427,
      "grad_norm": 305.62060546875,
      "learning_rate": 4.0788504053058217e-07,
      "loss": 0.1336,
      "step": 4000
    },
    {
      "epoch": 0.7369196757553427,
      "eval_loss": 0.6952972412109375,
      "eval_runtime": 76.7253,
      "eval_samples_per_second": 15.731,
      "eval_steps_per_second": 1.968,
      "step": 4000
    }
  ],
  "logging_steps": 500,
  "max_steps": 21712,
  "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": 1439752921742520.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}