{
  "best_metric": 0.9392343163490295,
  "best_model_checkpoint": "./output/clip-finetuned-csu-p14-336-e4l58-l/checkpoint-12000",
  "epoch": 1.2798634812286689,
  "eval_steps": 500,
  "global_step": 12000,
  "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
    },
    {
      "epoch": 0.37329351535836175,
      "grad_norm": 15.805087089538574,
      "learning_rate": 4.5333831058020476e-08,
      "loss": 0.3242,
      "step": 3500
    },
    {
      "epoch": 0.37329351535836175,
      "eval_loss": 1.195627212524414,
      "eval_runtime": 63.4024,
      "eval_samples_per_second": 15.567,
      "eval_steps_per_second": 1.956,
      "step": 3500
    },
    {
      "epoch": 0.42662116040955633,
      "grad_norm": 5.377908229827881,
      "learning_rate": 4.4667235494880546e-08,
      "loss": 0.2427,
      "step": 4000
    },
    {
      "epoch": 0.42662116040955633,
      "eval_loss": 1.169384241104126,
      "eval_runtime": 63.5082,
      "eval_samples_per_second": 15.541,
      "eval_steps_per_second": 1.953,
      "step": 4000
    },
    {
      "epoch": 0.47994880546075086,
      "grad_norm": 150.5334930419922,
      "learning_rate": 4.4000639931740615e-08,
      "loss": 0.2993,
      "step": 4500
    },
    {
      "epoch": 0.47994880546075086,
      "eval_loss": 1.145558476448059,
      "eval_runtime": 62.3888,
      "eval_samples_per_second": 15.82,
      "eval_steps_per_second": 1.988,
      "step": 4500
    },
    {
      "epoch": 0.5332764505119454,
      "grad_norm": 0.12246542423963547,
      "learning_rate": 4.333404436860068e-08,
      "loss": 0.3183,
      "step": 5000
    },
    {
      "epoch": 0.5332764505119454,
      "eval_loss": 1.1201218366622925,
      "eval_runtime": 63.7154,
      "eval_samples_per_second": 15.491,
      "eval_steps_per_second": 1.946,
      "step": 5000
    },
    {
      "epoch": 0.58660409556314,
      "grad_norm": 0.3306196630001068,
      "learning_rate": 4.266744880546075e-08,
      "loss": 0.307,
      "step": 5500
    },
    {
      "epoch": 0.58660409556314,
      "eval_loss": 1.0982334613800049,
      "eval_runtime": 62.4882,
      "eval_samples_per_second": 15.795,
      "eval_steps_per_second": 1.984,
      "step": 5500
    },
    {
      "epoch": 0.6399317406143344,
      "grad_norm": 149.38189697265625,
      "learning_rate": 4.2000853242320816e-08,
      "loss": 0.2638,
      "step": 6000
    },
    {
      "epoch": 0.6399317406143344,
      "eval_loss": 1.0779541730880737,
      "eval_runtime": 63.4101,
      "eval_samples_per_second": 15.565,
      "eval_steps_per_second": 1.956,
      "step": 6000
    },
    {
      "epoch": 0.693259385665529,
      "grad_norm": 353.0036315917969,
      "learning_rate": 4.1334257679180886e-08,
      "loss": 0.2226,
      "step": 6500
    },
    {
      "epoch": 0.693259385665529,
      "eval_loss": 1.0613036155700684,
      "eval_runtime": 62.8292,
      "eval_samples_per_second": 15.709,
      "eval_steps_per_second": 1.974,
      "step": 6500
    },
    {
      "epoch": 0.7465870307167235,
      "grad_norm": 0.13188917934894562,
      "learning_rate": 4.0667662116040955e-08,
      "loss": 0.2453,
      "step": 7000
    },
    {
      "epoch": 0.7465870307167235,
      "eval_loss": 1.0443507432937622,
      "eval_runtime": 62.7457,
      "eval_samples_per_second": 15.73,
      "eval_steps_per_second": 1.976,
      "step": 7000
    },
    {
      "epoch": 0.7999146757679181,
      "grad_norm": 59.391639709472656,
      "learning_rate": 4.000106655290102e-08,
      "loss": 0.272,
      "step": 7500
    },
    {
      "epoch": 0.7999146757679181,
      "eval_loss": 1.0300519466400146,
      "eval_runtime": 62.3719,
      "eval_samples_per_second": 15.824,
      "eval_steps_per_second": 1.988,
      "step": 7500
    },
    {
      "epoch": 0.8532423208191127,
      "grad_norm": 495.4797668457031,
      "learning_rate": 3.9334470989761093e-08,
      "loss": 0.283,
      "step": 8000
    },
    {
      "epoch": 0.8532423208191127,
      "eval_loss": 1.0166898965835571,
      "eval_runtime": 62.0338,
      "eval_samples_per_second": 15.911,
      "eval_steps_per_second": 1.999,
      "step": 8000
    },
    {
      "epoch": 0.9065699658703071,
      "grad_norm": 159.0310821533203,
      "learning_rate": 3.866787542662116e-08,
      "loss": 0.2331,
      "step": 8500
    },
    {
      "epoch": 0.9065699658703071,
      "eval_loss": 1.003544807434082,
      "eval_runtime": 62.9704,
      "eval_samples_per_second": 15.674,
      "eval_steps_per_second": 1.969,
      "step": 8500
    },
    {
      "epoch": 0.9598976109215017,
      "grad_norm": 45.40327835083008,
      "learning_rate": 3.8001279863481225e-08,
      "loss": 0.2362,
      "step": 9000
    },
    {
      "epoch": 0.9598976109215017,
      "eval_loss": 0.992453396320343,
      "eval_runtime": 63.1405,
      "eval_samples_per_second": 15.632,
      "eval_steps_per_second": 1.964,
      "step": 9000
    },
    {
      "epoch": 1.0132252559726962,
      "grad_norm": 621.8538818359375,
      "learning_rate": 3.7334684300341295e-08,
      "loss": 0.2396,
      "step": 9500
    },
    {
      "epoch": 1.0132252559726962,
      "eval_loss": 0.9829725027084351,
      "eval_runtime": 62.3244,
      "eval_samples_per_second": 15.836,
      "eval_steps_per_second": 1.99,
      "step": 9500
    },
    {
      "epoch": 1.0665529010238908,
      "grad_norm": 461.60736083984375,
      "learning_rate": 3.6668088737201364e-08,
      "loss": 0.2013,
      "step": 10000
    },
    {
      "epoch": 1.0665529010238908,
      "eval_loss": 0.9735796451568604,
      "eval_runtime": 63.0763,
      "eval_samples_per_second": 15.648,
      "eval_steps_per_second": 1.966,
      "step": 10000
    },
    {
      "epoch": 1.1198805460750854,
      "grad_norm": 7.0309906005859375,
      "learning_rate": 3.6001493174061433e-08,
      "loss": 0.2082,
      "step": 10500
    },
    {
      "epoch": 1.1198805460750854,
      "eval_loss": 0.9639086127281189,
      "eval_runtime": 63.0238,
      "eval_samples_per_second": 15.661,
      "eval_steps_per_second": 1.968,
      "step": 10500
    },
    {
      "epoch": 1.17320819112628,
      "grad_norm": 304.01995849609375,
      "learning_rate": 3.5334897610921496e-08,
      "loss": 0.2023,
      "step": 11000
    },
    {
      "epoch": 1.17320819112628,
      "eval_loss": 0.9558469653129578,
      "eval_runtime": 62.1866,
      "eval_samples_per_second": 15.872,
      "eval_steps_per_second": 1.994,
      "step": 11000
    },
    {
      "epoch": 1.2265358361774745,
      "grad_norm": 0.0017981340643018484,
      "learning_rate": 3.4668302047781565e-08,
      "loss": 0.2331,
      "step": 11500
    },
    {
      "epoch": 1.2265358361774745,
      "eval_loss": 0.9465070962905884,
      "eval_runtime": 63.1547,
      "eval_samples_per_second": 15.628,
      "eval_steps_per_second": 1.963,
      "step": 11500
    },
    {
      "epoch": 1.2798634812286689,
      "grad_norm": 0.8147886395454407,
      "learning_rate": 3.4001706484641635e-08,
      "loss": 0.1784,
      "step": 12000
    },
    {
      "epoch": 1.2798634812286689,
      "eval_loss": 0.9392343163490295,
      "eval_runtime": 62.9221,
      "eval_samples_per_second": 15.686,
      "eval_steps_per_second": 1.971,
      "step": 12000
    }
  ],
  "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": 4317819731822520.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}