{
  "best_metric": 1.3460415601730347,
  "best_model_checkpoint": "./output/clip-finetuned-csu-p14-336-e4l59-l/checkpoint-21500",
  "epoch": 3.960943257184967,
  "eval_steps": 500,
  "global_step": 21500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.09211495946941783,
      "grad_norm": 20.952665328979492,
      "learning_rate": 4.884856300663228e-09,
      "loss": 0.3952,
      "step": 500
    },
    {
      "epoch": 0.09211495946941783,
      "eval_loss": 1.4940427541732788,
      "eval_runtime": 73.9282,
      "eval_samples_per_second": 16.327,
      "eval_steps_per_second": 2.043,
      "step": 500
    },
    {
      "epoch": 0.18422991893883567,
      "grad_norm": 59.28666687011719,
      "learning_rate": 4.769712601326456e-09,
      "loss": 0.4562,
      "step": 1000
    },
    {
      "epoch": 0.18422991893883567,
      "eval_loss": 1.4853259325027466,
      "eval_runtime": 74.6533,
      "eval_samples_per_second": 16.168,
      "eval_steps_per_second": 2.023,
      "step": 1000
    },
    {
      "epoch": 0.2763448784082535,
      "grad_norm": 1.179814338684082,
      "learning_rate": 4.654568901989683e-09,
      "loss": 0.5131,
      "step": 1500
    },
    {
      "epoch": 0.2763448784082535,
      "eval_loss": 1.4757832288742065,
      "eval_runtime": 76.2866,
      "eval_samples_per_second": 15.822,
      "eval_steps_per_second": 1.979,
      "step": 1500
    },
    {
      "epoch": 0.36845983787767134,
      "grad_norm": 0.900991678237915,
      "learning_rate": 4.5394252026529105e-09,
      "loss": 0.4481,
      "step": 2000
    },
    {
      "epoch": 0.36845983787767134,
      "eval_loss": 1.4675536155700684,
      "eval_runtime": 75.9348,
      "eval_samples_per_second": 15.895,
      "eval_steps_per_second": 1.989,
      "step": 2000
    },
    {
      "epoch": 0.46057479734708917,
      "grad_norm": 346.93853759765625,
      "learning_rate": 4.424281503316139e-09,
      "loss": 0.4839,
      "step": 2500
    },
    {
      "epoch": 0.46057479734708917,
      "eval_loss": 1.4585391283035278,
      "eval_runtime": 76.7708,
      "eval_samples_per_second": 15.722,
      "eval_steps_per_second": 1.967,
      "step": 2500
    },
    {
      "epoch": 0.552689756816507,
      "grad_norm": 318.47515869140625,
      "learning_rate": 4.3091378039793666e-09,
      "loss": 0.4377,
      "step": 3000
    },
    {
      "epoch": 0.552689756816507,
      "eval_loss": 1.4507631063461304,
      "eval_runtime": 76.5424,
      "eval_samples_per_second": 15.769,
      "eval_steps_per_second": 1.973,
      "step": 3000
    },
    {
      "epoch": 0.6448047162859248,
      "grad_norm": 112.41814422607422,
      "learning_rate": 4.193994104642594e-09,
      "loss": 0.4231,
      "step": 3500
    },
    {
      "epoch": 0.6448047162859248,
      "eval_loss": 1.4432125091552734,
      "eval_runtime": 76.4812,
      "eval_samples_per_second": 15.782,
      "eval_steps_per_second": 1.974,
      "step": 3500
    },
    {
      "epoch": 0.7369196757553427,
      "grad_norm": 83.51781463623047,
      "learning_rate": 4.078850405305822e-09,
      "loss": 0.4369,
      "step": 4000
    },
    {
      "epoch": 0.7369196757553427,
      "eval_loss": 1.4365661144256592,
      "eval_runtime": 76.4405,
      "eval_samples_per_second": 15.79,
      "eval_steps_per_second": 1.975,
      "step": 4000
    },
    {
      "epoch": 0.8290346352247605,
      "grad_norm": 460.5648498535156,
      "learning_rate": 3.963706705969049e-09,
      "loss": 0.4082,
      "step": 4500
    },
    {
      "epoch": 0.8290346352247605,
      "eval_loss": 1.4302303791046143,
      "eval_runtime": 75.042,
      "eval_samples_per_second": 16.084,
      "eval_steps_per_second": 2.012,
      "step": 4500
    },
    {
      "epoch": 0.9211495946941783,
      "grad_norm": 5.289137363433838,
      "learning_rate": 3.848563006632278e-09,
      "loss": 0.4234,
      "step": 5000
    },
    {
      "epoch": 0.9211495946941783,
      "eval_loss": 1.4243344068527222,
      "eval_runtime": 76.6198,
      "eval_samples_per_second": 15.753,
      "eval_steps_per_second": 1.971,
      "step": 5000
    },
    {
      "epoch": 1.013264554163596,
      "grad_norm": 0.6147650480270386,
      "learning_rate": 3.733419307295505e-09,
      "loss": 0.4266,
      "step": 5500
    },
    {
      "epoch": 1.013264554163596,
      "eval_loss": 1.4190788269042969,
      "eval_runtime": 75.4536,
      "eval_samples_per_second": 15.997,
      "eval_steps_per_second": 2.001,
      "step": 5500
    },
    {
      "epoch": 1.105379513633014,
      "grad_norm": 231.71815490722656,
      "learning_rate": 3.6182756079587326e-09,
      "loss": 0.4438,
      "step": 6000
    },
    {
      "epoch": 1.105379513633014,
      "eval_loss": 1.4136912822723389,
      "eval_runtime": 76.5978,
      "eval_samples_per_second": 15.758,
      "eval_steps_per_second": 1.971,
      "step": 6000
    },
    {
      "epoch": 1.1974944731024317,
      "grad_norm": 464.0763854980469,
      "learning_rate": 3.5031319086219602e-09,
      "loss": 0.3814,
      "step": 6500
    },
    {
      "epoch": 1.1974944731024317,
      "eval_loss": 1.4085203409194946,
      "eval_runtime": 75.4916,
      "eval_samples_per_second": 15.989,
      "eval_steps_per_second": 2.0,
      "step": 6500
    },
    {
      "epoch": 1.2896094325718497,
      "grad_norm": 2.5053436756134033,
      "learning_rate": 3.3879882092851882e-09,
      "loss": 0.3327,
      "step": 7000
    },
    {
      "epoch": 1.2896094325718497,
      "eval_loss": 1.4041872024536133,
      "eval_runtime": 75.2076,
      "eval_samples_per_second": 16.049,
      "eval_steps_per_second": 2.008,
      "step": 7000
    },
    {
      "epoch": 1.3817243920412676,
      "grad_norm": 126.44642639160156,
      "learning_rate": 3.2728445099484154e-09,
      "loss": 0.4045,
      "step": 7500
    },
    {
      "epoch": 1.3817243920412676,
      "eval_loss": 1.398864984512329,
      "eval_runtime": 75.3652,
      "eval_samples_per_second": 16.015,
      "eval_steps_per_second": 2.004,
      "step": 7500
    },
    {
      "epoch": 1.4738393515106853,
      "grad_norm": 1.046550989151001,
      "learning_rate": 3.1577008106116435e-09,
      "loss": 0.4038,
      "step": 8000
    },
    {
      "epoch": 1.4738393515106853,
      "eval_loss": 1.3937218189239502,
      "eval_runtime": 75.1378,
      "eval_samples_per_second": 16.064,
      "eval_steps_per_second": 2.01,
      "step": 8000
    },
    {
      "epoch": 1.565954310980103,
      "grad_norm": 212.23594665527344,
      "learning_rate": 3.042557111274871e-09,
      "loss": 0.3659,
      "step": 8500
    },
    {
      "epoch": 1.565954310980103,
      "eval_loss": 1.3894429206848145,
      "eval_runtime": 75.2403,
      "eval_samples_per_second": 16.042,
      "eval_steps_per_second": 2.007,
      "step": 8500
    },
    {
      "epoch": 1.658069270449521,
      "grad_norm": 2.719074010848999,
      "learning_rate": 2.927413411938099e-09,
      "loss": 0.4282,
      "step": 9000
    },
    {
      "epoch": 1.658069270449521,
      "eval_loss": 1.3854893445968628,
      "eval_runtime": 74.8392,
      "eval_samples_per_second": 16.128,
      "eval_steps_per_second": 2.018,
      "step": 9000
    },
    {
      "epoch": 1.750184229918939,
      "grad_norm": 4.810174465179443,
      "learning_rate": 2.8122697126013263e-09,
      "loss": 0.4173,
      "step": 9500
    },
    {
      "epoch": 1.750184229918939,
      "eval_loss": 1.3816269636154175,
      "eval_runtime": 75.6136,
      "eval_samples_per_second": 15.963,
      "eval_steps_per_second": 1.997,
      "step": 9500
    },
    {
      "epoch": 1.8422991893883567,
      "grad_norm": 630.7271118164062,
      "learning_rate": 2.6971260132645543e-09,
      "loss": 0.3758,
      "step": 10000
    },
    {
      "epoch": 1.8422991893883567,
      "eval_loss": 1.3779022693634033,
      "eval_runtime": 76.6948,
      "eval_samples_per_second": 15.738,
      "eval_steps_per_second": 1.969,
      "step": 10000
    },
    {
      "epoch": 1.9344141488577744,
      "grad_norm": 3.372347831726074,
      "learning_rate": 2.581982313927782e-09,
      "loss": 0.4105,
      "step": 10500
    },
    {
      "epoch": 1.9344141488577744,
      "eval_loss": 1.3744733333587646,
      "eval_runtime": 76.3833,
      "eval_samples_per_second": 15.802,
      "eval_steps_per_second": 1.977,
      "step": 10500
    },
    {
      "epoch": 2.026529108327192,
      "grad_norm": 0.4494275450706482,
      "learning_rate": 2.4668386145910095e-09,
      "loss": 0.3765,
      "step": 11000
    },
    {
      "epoch": 2.026529108327192,
      "eval_loss": 1.3716294765472412,
      "eval_runtime": 76.2721,
      "eval_samples_per_second": 15.825,
      "eval_steps_per_second": 1.98,
      "step": 11000
    },
    {
      "epoch": 2.1186440677966103,
      "grad_norm": 397.7524719238281,
      "learning_rate": 2.3516949152542375e-09,
      "loss": 0.3746,
      "step": 11500
    },
    {
      "epoch": 2.1186440677966103,
      "eval_loss": 1.3690176010131836,
      "eval_runtime": 74.7759,
      "eval_samples_per_second": 16.142,
      "eval_steps_per_second": 2.019,
      "step": 11500
    },
    {
      "epoch": 2.210759027266028,
      "grad_norm": 1.4716993570327759,
      "learning_rate": 2.236551215917465e-09,
      "loss": 0.3783,
      "step": 12000
    },
    {
      "epoch": 2.210759027266028,
      "eval_loss": 1.3662136793136597,
      "eval_runtime": 74.6166,
      "eval_samples_per_second": 16.176,
      "eval_steps_per_second": 2.024,
      "step": 12000
    },
    {
      "epoch": 2.3028739867354457,
      "grad_norm": 612.47119140625,
      "learning_rate": 2.1214075165806927e-09,
      "loss": 0.3832,
      "step": 12500
    },
    {
      "epoch": 2.3028739867354457,
      "eval_loss": 1.3639976978302002,
      "eval_runtime": 76.0876,
      "eval_samples_per_second": 15.863,
      "eval_steps_per_second": 1.985,
      "step": 12500
    },
    {
      "epoch": 2.3949889462048635,
      "grad_norm": 5.805767059326172,
      "learning_rate": 2.0062638172439203e-09,
      "loss": 0.3984,
      "step": 13000
    },
    {
      "epoch": 2.3949889462048635,
      "eval_loss": 1.3616890907287598,
      "eval_runtime": 75.696,
      "eval_samples_per_second": 15.945,
      "eval_steps_per_second": 1.995,
      "step": 13000
    },
    {
      "epoch": 2.4871039056742816,
      "grad_norm": 12.754598617553711,
      "learning_rate": 1.8911201179071484e-09,
      "loss": 0.4124,
      "step": 13500
    },
    {
      "epoch": 2.4871039056742816,
      "eval_loss": 1.359315276145935,
      "eval_runtime": 74.7355,
      "eval_samples_per_second": 16.15,
      "eval_steps_per_second": 2.02,
      "step": 13500
    },
    {
      "epoch": 2.5792188651436994,
      "grad_norm": 196.18942260742188,
      "learning_rate": 1.7759764185703758e-09,
      "loss": 0.3363,
      "step": 14000
    },
    {
      "epoch": 2.5792188651436994,
      "eval_loss": 1.3572332859039307,
      "eval_runtime": 76.2343,
      "eval_samples_per_second": 15.833,
      "eval_steps_per_second": 1.981,
      "step": 14000
    },
    {
      "epoch": 2.671333824613117,
      "grad_norm": 2.0238358974456787,
      "learning_rate": 1.6608327192336038e-09,
      "loss": 0.3274,
      "step": 14500
    },
    {
      "epoch": 2.671333824613117,
      "eval_loss": 1.3555117845535278,
      "eval_runtime": 75.3544,
      "eval_samples_per_second": 16.018,
      "eval_steps_per_second": 2.004,
      "step": 14500
    },
    {
      "epoch": 2.7634487840825352,
      "grad_norm": 124.35880279541016,
      "learning_rate": 1.5456890198968312e-09,
      "loss": 0.4039,
      "step": 15000
    },
    {
      "epoch": 2.7634487840825352,
      "eval_loss": 1.3538331985473633,
      "eval_runtime": 74.7014,
      "eval_samples_per_second": 16.158,
      "eval_steps_per_second": 2.021,
      "step": 15000
    },
    {
      "epoch": 2.855563743551953,
      "grad_norm": 0.13434405624866486,
      "learning_rate": 1.430545320560059e-09,
      "loss": 0.378,
      "step": 15500
    },
    {
      "epoch": 2.855563743551953,
      "eval_loss": 1.3523625135421753,
      "eval_runtime": 76.2678,
      "eval_samples_per_second": 15.826,
      "eval_steps_per_second": 1.98,
      "step": 15500
    },
    {
      "epoch": 2.9476787030213707,
      "grad_norm": 200.6512908935547,
      "learning_rate": 1.3154016212232866e-09,
      "loss": 0.3543,
      "step": 16000
    },
    {
      "epoch": 2.9476787030213707,
      "eval_loss": 1.3511295318603516,
      "eval_runtime": 75.4117,
      "eval_samples_per_second": 16.005,
      "eval_steps_per_second": 2.002,
      "step": 16000
    },
    {
      "epoch": 3.0397936624907884,
      "grad_norm": 1.6490756273269653,
      "learning_rate": 1.2002579218865144e-09,
      "loss": 0.3606,
      "step": 16500
    },
    {
      "epoch": 3.0397936624907884,
      "eval_loss": 1.3500866889953613,
      "eval_runtime": 74.798,
      "eval_samples_per_second": 16.137,
      "eval_steps_per_second": 2.019,
      "step": 16500
    },
    {
      "epoch": 3.131908621960206,
      "grad_norm": 6.0325026512146,
      "learning_rate": 1.085114222549742e-09,
      "loss": 0.4024,
      "step": 17000
    },
    {
      "epoch": 3.131908621960206,
      "eval_loss": 1.3490514755249023,
      "eval_runtime": 76.1949,
      "eval_samples_per_second": 15.841,
      "eval_steps_per_second": 1.982,
      "step": 17000
    },
    {
      "epoch": 3.2240235814296243,
      "grad_norm": 106.60160827636719,
      "learning_rate": 9.699705232129698e-10,
      "loss": 0.3182,
      "step": 17500
    },
    {
      "epoch": 3.2240235814296243,
      "eval_loss": 1.3482465744018555,
      "eval_runtime": 75.3545,
      "eval_samples_per_second": 16.018,
      "eval_steps_per_second": 2.004,
      "step": 17500
    },
    {
      "epoch": 3.316138540899042,
      "grad_norm": 35.852169036865234,
      "learning_rate": 8.548268238761974e-10,
      "loss": 0.3564,
      "step": 18000
    },
    {
      "epoch": 3.316138540899042,
      "eval_loss": 1.3475124835968018,
      "eval_runtime": 74.6762,
      "eval_samples_per_second": 16.163,
      "eval_steps_per_second": 2.022,
      "step": 18000
    },
    {
      "epoch": 3.4082535003684598,
      "grad_norm": 15.63128662109375,
      "learning_rate": 7.396831245394251e-10,
      "loss": 0.3842,
      "step": 18500
    },
    {
      "epoch": 3.4082535003684598,
      "eval_loss": 1.3470234870910645,
      "eval_runtime": 76.0753,
      "eval_samples_per_second": 15.866,
      "eval_steps_per_second": 1.985,
      "step": 18500
    },
    {
      "epoch": 3.500368459837878,
      "grad_norm": 18.28308868408203,
      "learning_rate": 6.24539425202653e-10,
      "loss": 0.352,
      "step": 19000
    },
    {
      "epoch": 3.500368459837878,
      "eval_loss": 1.3466503620147705,
      "eval_runtime": 75.4355,
      "eval_samples_per_second": 16.0,
      "eval_steps_per_second": 2.002,
      "step": 19000
    },
    {
      "epoch": 3.5924834193072956,
      "grad_norm": 3.9090354442596436,
      "learning_rate": 5.093957258658807e-10,
      "loss": 0.3828,
      "step": 19500
    },
    {
      "epoch": 3.5924834193072956,
      "eval_loss": 1.346364974975586,
      "eval_runtime": 74.044,
      "eval_samples_per_second": 16.301,
      "eval_steps_per_second": 2.039,
      "step": 19500
    },
    {
      "epoch": 3.6845983787767134,
      "grad_norm": 0.0035269868094474077,
      "learning_rate": 3.942520265291084e-10,
      "loss": 0.39,
      "step": 20000
    },
    {
      "epoch": 3.6845983787767134,
      "eval_loss": 1.3461922407150269,
      "eval_runtime": 75.6473,
      "eval_samples_per_second": 15.956,
      "eval_steps_per_second": 1.996,
      "step": 20000
    },
    {
      "epoch": 3.776713338246131,
      "grad_norm": 430.9030456542969,
      "learning_rate": 2.7910832719233604e-10,
      "loss": 0.3618,
      "step": 20500
    },
    {
      "epoch": 3.776713338246131,
      "eval_loss": 1.3460967540740967,
      "eval_runtime": 74.8382,
      "eval_samples_per_second": 16.128,
      "eval_steps_per_second": 2.018,
      "step": 20500
    },
    {
      "epoch": 3.868828297715549,
      "grad_norm": 1.444280743598938,
      "learning_rate": 1.6396462785556375e-10,
      "loss": 0.3856,
      "step": 21000
    },
    {
      "epoch": 3.868828297715549,
      "eval_loss": 1.3460530042648315,
      "eval_runtime": 74.2187,
      "eval_samples_per_second": 16.263,
      "eval_steps_per_second": 2.035,
      "step": 21000
    },
    {
      "epoch": 3.960943257184967,
      "grad_norm": 10.209774017333984,
      "learning_rate": 4.882092851879146e-11,
      "loss": 0.3586,
      "step": 21500
    },
    {
      "epoch": 3.960943257184967,
      "eval_loss": 1.3460415601730347,
      "eval_runtime": 75.5366,
      "eval_samples_per_second": 15.979,
      "eval_steps_per_second": 1.999,
      "step": 21500
    }
  ],
  "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": 7735524068792520.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}