{
  "best_metric": 0.5088639855384827,
  "best_model_checkpoint": "./output/clip-finetuned-csu-p14-336-e4l57-l/checkpoint-17000",
  "epoch": 3.2240235814296243,
  "eval_steps": 500,
  "global_step": 17500,
  "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
    },
    {
      "epoch": 0.8290346352247605,
      "grad_norm": 506.12237548828125,
      "learning_rate": 3.9637067059690496e-07,
      "loss": 0.1154,
      "step": 4500
    },
    {
      "epoch": 0.8290346352247605,
      "eval_loss": 0.6744509935379028,
      "eval_runtime": 76.9493,
      "eval_samples_per_second": 15.686,
      "eval_steps_per_second": 1.962,
      "step": 4500
    },
    {
      "epoch": 0.9211495946941783,
      "grad_norm": 8.823301322991028e-05,
      "learning_rate": 3.8485630066322774e-07,
      "loss": 0.108,
      "step": 5000
    },
    {
      "epoch": 0.9211495946941783,
      "eval_loss": 0.6559990048408508,
      "eval_runtime": 75.2391,
      "eval_samples_per_second": 16.042,
      "eval_steps_per_second": 2.007,
      "step": 5000
    },
    {
      "epoch": 1.013264554163596,
      "grad_norm": 0.0008560372516512871,
      "learning_rate": 3.733419307295504e-07,
      "loss": 0.1106,
      "step": 5500
    },
    {
      "epoch": 1.013264554163596,
      "eval_loss": 0.6367496848106384,
      "eval_runtime": 77.0031,
      "eval_samples_per_second": 15.675,
      "eval_steps_per_second": 1.961,
      "step": 5500
    },
    {
      "epoch": 1.105379513633014,
      "grad_norm": 0.04510360211133957,
      "learning_rate": 3.618275607958732e-07,
      "loss": 0.0591,
      "step": 6000
    },
    {
      "epoch": 1.105379513633014,
      "eval_loss": 0.6259325742721558,
      "eval_runtime": 75.2787,
      "eval_samples_per_second": 16.034,
      "eval_steps_per_second": 2.006,
      "step": 6000
    },
    {
      "epoch": 1.1974944731024317,
      "grad_norm": 0.0005158882704563439,
      "learning_rate": 3.50313190862196e-07,
      "loss": 0.0745,
      "step": 6500
    },
    {
      "epoch": 1.1974944731024317,
      "eval_loss": 0.6209793090820312,
      "eval_runtime": 77.1989,
      "eval_samples_per_second": 15.635,
      "eval_steps_per_second": 1.956,
      "step": 6500
    },
    {
      "epoch": 1.2896094325718497,
      "grad_norm": 3.5883346072296263e-07,
      "learning_rate": 3.387988209285188e-07,
      "loss": 0.0502,
      "step": 7000
    },
    {
      "epoch": 1.2896094325718497,
      "eval_loss": 0.6133300065994263,
      "eval_runtime": 75.1116,
      "eval_samples_per_second": 16.069,
      "eval_steps_per_second": 2.01,
      "step": 7000
    },
    {
      "epoch": 1.3817243920412676,
      "grad_norm": 21.66597557067871,
      "learning_rate": 3.272844509948415e-07,
      "loss": 0.079,
      "step": 7500
    },
    {
      "epoch": 1.3817243920412676,
      "eval_loss": 0.6006803512573242,
      "eval_runtime": 75.0086,
      "eval_samples_per_second": 16.091,
      "eval_steps_per_second": 2.013,
      "step": 7500
    },
    {
      "epoch": 1.4738393515106853,
      "grad_norm": 7.992870814632624e-05,
      "learning_rate": 3.157700810611643e-07,
      "loss": 0.0776,
      "step": 8000
    },
    {
      "epoch": 1.4738393515106853,
      "eval_loss": 0.5866427421569824,
      "eval_runtime": 75.6702,
      "eval_samples_per_second": 15.951,
      "eval_steps_per_second": 1.996,
      "step": 8000
    },
    {
      "epoch": 1.565954310980103,
      "grad_norm": 0.0010381735628470778,
      "learning_rate": 3.042557111274871e-07,
      "loss": 0.0492,
      "step": 8500
    },
    {
      "epoch": 1.565954310980103,
      "eval_loss": 0.5679212212562561,
      "eval_runtime": 75.0643,
      "eval_samples_per_second": 16.08,
      "eval_steps_per_second": 2.012,
      "step": 8500
    },
    {
      "epoch": 1.658069270449521,
      "grad_norm": 1.733377956725235e-07,
      "learning_rate": 2.927413411938099e-07,
      "loss": 0.0794,
      "step": 9000
    },
    {
      "epoch": 1.658069270449521,
      "eval_loss": 0.5761673450469971,
      "eval_runtime": 76.2738,
      "eval_samples_per_second": 15.825,
      "eval_steps_per_second": 1.98,
      "step": 9000
    },
    {
      "epoch": 1.750184229918939,
      "grad_norm": 3.1063287359023306e-11,
      "learning_rate": 2.812269712601326e-07,
      "loss": 0.0677,
      "step": 9500
    },
    {
      "epoch": 1.750184229918939,
      "eval_loss": 0.5565799474716187,
      "eval_runtime": 75.6897,
      "eval_samples_per_second": 15.947,
      "eval_steps_per_second": 1.995,
      "step": 9500
    },
    {
      "epoch": 1.8422991893883567,
      "grad_norm": 0.00010032750287791714,
      "learning_rate": 2.697126013264554e-07,
      "loss": 0.0566,
      "step": 10000
    },
    {
      "epoch": 1.8422991893883567,
      "eval_loss": 0.5482199192047119,
      "eval_runtime": 75.0029,
      "eval_samples_per_second": 16.093,
      "eval_steps_per_second": 2.013,
      "step": 10000
    },
    {
      "epoch": 1.9344141488577744,
      "grad_norm": 5.553330595375883e-08,
      "learning_rate": 2.581982313927782e-07,
      "loss": 0.0828,
      "step": 10500
    },
    {
      "epoch": 1.9344141488577744,
      "eval_loss": 0.5500391125679016,
      "eval_runtime": 76.2612,
      "eval_samples_per_second": 15.827,
      "eval_steps_per_second": 1.98,
      "step": 10500
    },
    {
      "epoch": 2.026529108327192,
      "grad_norm": 1.5537947319899104e-06,
      "learning_rate": 2.466838614591009e-07,
      "loss": 0.0573,
      "step": 11000
    },
    {
      "epoch": 2.026529108327192,
      "eval_loss": 0.5341849327087402,
      "eval_runtime": 75.6131,
      "eval_samples_per_second": 15.963,
      "eval_steps_per_second": 1.997,
      "step": 11000
    },
    {
      "epoch": 2.1186440677966103,
      "grad_norm": 36.10622024536133,
      "learning_rate": 2.3516949152542374e-07,
      "loss": 0.0401,
      "step": 11500
    },
    {
      "epoch": 2.1186440677966103,
      "eval_loss": 0.5351412296295166,
      "eval_runtime": 74.8873,
      "eval_samples_per_second": 16.118,
      "eval_steps_per_second": 2.016,
      "step": 11500
    },
    {
      "epoch": 2.210759027266028,
      "grad_norm": 0.49580878019332886,
      "learning_rate": 2.236551215917465e-07,
      "loss": 0.0152,
      "step": 12000
    },
    {
      "epoch": 2.210759027266028,
      "eval_loss": 0.5349487662315369,
      "eval_runtime": 76.3109,
      "eval_samples_per_second": 15.817,
      "eval_steps_per_second": 1.979,
      "step": 12000
    },
    {
      "epoch": 2.3028739867354457,
      "grad_norm": 1.7469781637191772,
      "learning_rate": 2.1214075165806928e-07,
      "loss": 0.0638,
      "step": 12500
    },
    {
      "epoch": 2.3028739867354457,
      "eval_loss": 0.531768798828125,
      "eval_runtime": 75.5712,
      "eval_samples_per_second": 15.972,
      "eval_steps_per_second": 1.998,
      "step": 12500
    },
    {
      "epoch": 2.3949889462048635,
      "grad_norm": 0.0006845183088444173,
      "learning_rate": 2.0062638172439202e-07,
      "loss": 0.0488,
      "step": 13000
    },
    {
      "epoch": 2.3949889462048635,
      "eval_loss": 0.530552089214325,
      "eval_runtime": 75.6479,
      "eval_samples_per_second": 15.955,
      "eval_steps_per_second": 1.996,
      "step": 13000
    },
    {
      "epoch": 2.4871039056742816,
      "grad_norm": 0.0002451306500006467,
      "learning_rate": 1.891120117907148e-07,
      "loss": 0.0456,
      "step": 13500
    },
    {
      "epoch": 2.4871039056742816,
      "eval_loss": 0.5211473107337952,
      "eval_runtime": 76.4401,
      "eval_samples_per_second": 15.79,
      "eval_steps_per_second": 1.975,
      "step": 13500
    },
    {
      "epoch": 2.5792188651436994,
      "grad_norm": 0.003609905717894435,
      "learning_rate": 1.7759764185703757e-07,
      "loss": 0.0264,
      "step": 14000
    },
    {
      "epoch": 2.5792188651436994,
      "eval_loss": 0.5193639993667603,
      "eval_runtime": 74.595,
      "eval_samples_per_second": 16.181,
      "eval_steps_per_second": 2.024,
      "step": 14000
    },
    {
      "epoch": 2.671333824613117,
      "grad_norm": 6.751325276610487e-09,
      "learning_rate": 1.6608327192336035e-07,
      "loss": 0.0381,
      "step": 14500
    },
    {
      "epoch": 2.671333824613117,
      "eval_loss": 0.520633339881897,
      "eval_runtime": 75.2937,
      "eval_samples_per_second": 16.031,
      "eval_steps_per_second": 2.005,
      "step": 14500
    },
    {
      "epoch": 2.7634487840825352,
      "grad_norm": 8.452875022157968e-07,
      "learning_rate": 1.5456890198968311e-07,
      "loss": 0.0413,
      "step": 15000
    },
    {
      "epoch": 2.7634487840825352,
      "eval_loss": 0.5167532563209534,
      "eval_runtime": 75.615,
      "eval_samples_per_second": 15.962,
      "eval_steps_per_second": 1.997,
      "step": 15000
    },
    {
      "epoch": 2.855563743551953,
      "grad_norm": 2.365756202049596e-11,
      "learning_rate": 1.430545320560059e-07,
      "loss": 0.0392,
      "step": 15500
    },
    {
      "epoch": 2.855563743551953,
      "eval_loss": 0.5148531198501587,
      "eval_runtime": 74.6217,
      "eval_samples_per_second": 16.175,
      "eval_steps_per_second": 2.024,
      "step": 15500
    },
    {
      "epoch": 2.9476787030213707,
      "grad_norm": 0.32707780599594116,
      "learning_rate": 1.3154016212232866e-07,
      "loss": 0.0352,
      "step": 16000
    },
    {
      "epoch": 2.9476787030213707,
      "eval_loss": 0.5112494826316833,
      "eval_runtime": 75.142,
      "eval_samples_per_second": 16.063,
      "eval_steps_per_second": 2.01,
      "step": 16000
    },
    {
      "epoch": 3.0397936624907884,
      "grad_norm": 4.2568670810538833e-11,
      "learning_rate": 1.2002579218865142e-07,
      "loss": 0.0467,
      "step": 16500
    },
    {
      "epoch": 3.0397936624907884,
      "eval_loss": 0.5097618699073792,
      "eval_runtime": 75.705,
      "eval_samples_per_second": 15.943,
      "eval_steps_per_second": 1.995,
      "step": 16500
    },
    {
      "epoch": 3.131908621960206,
      "grad_norm": 7.157668679980134e-09,
      "learning_rate": 1.085114222549742e-07,
      "loss": 0.0366,
      "step": 17000
    },
    {
      "epoch": 3.131908621960206,
      "eval_loss": 0.5088639855384827,
      "eval_runtime": 74.5863,
      "eval_samples_per_second": 16.183,
      "eval_steps_per_second": 2.025,
      "step": 17000
    },
    {
      "epoch": 3.2240235814296243,
      "grad_norm": 0.5768880248069763,
      "learning_rate": 9.699705232129697e-08,
      "loss": 0.0454,
      "step": 17500
    },
    {
      "epoch": 3.2240235814296243,
      "eval_loss": 0.5103679299354553,
      "eval_runtime": 74.9953,
      "eval_samples_per_second": 16.094,
      "eval_steps_per_second": 2.013,
      "step": 17500
    }
  ],
  "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": 6296490663752520.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}