{
  "best_metric": 0.8421845881825929,
  "best_model_checkpoint": "test-klue/ynat/run-3/checkpoint-2500",
  "epoch": 4.378283712784588,
  "eval_steps": 50,
  "global_step": 2500,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.09,
      "learning_rate": 3.851823303957499e-06,
      "loss": 1.934,
      "step": 50
    },
    {
      "epoch": 0.09,
      "eval_f1": 0.01705140497086806,
      "eval_loss": 2.002474546432495,
      "eval_runtime": 12.8723,
      "eval_samples_per_second": 707.486,
      "eval_steps_per_second": 1.398,
      "step": 50
    },
    {
      "epoch": 0.18,
      "learning_rate": 3.783163352015298e-06,
      "loss": 1.9174,
      "step": 100
    },
    {
      "epoch": 0.18,
      "eval_f1": 0.16773144497786774,
      "eval_loss": 1.9946870803833008,
      "eval_runtime": 13.4529,
      "eval_samples_per_second": 676.953,
      "eval_steps_per_second": 1.338,
      "step": 100
    },
    {
      "epoch": 0.26,
      "learning_rate": 3.7145034000730966e-06,
      "loss": 1.6984,
      "step": 150
    },
    {
      "epoch": 0.26,
      "eval_f1": 0.579749057690188,
      "eval_loss": 1.6373496055603027,
      "eval_runtime": 13.0114,
      "eval_samples_per_second": 699.924,
      "eval_steps_per_second": 1.383,
      "step": 150
    },
    {
      "epoch": 0.35,
      "learning_rate": 3.645843448130895e-06,
      "loss": 1.2585,
      "step": 200
    },
    {
      "epoch": 0.35,
      "eval_f1": 0.6734776928120875,
      "eval_loss": 1.2733124494552612,
      "eval_runtime": 12.6224,
      "eval_samples_per_second": 721.493,
      "eval_steps_per_second": 1.426,
      "step": 200
    },
    {
      "epoch": 0.44,
      "learning_rate": 3.5771834961886936e-06,
      "loss": 0.9487,
      "step": 250
    },
    {
      "epoch": 0.44,
      "eval_f1": 0.6814953813845488,
      "eval_loss": 1.1104496717453003,
      "eval_runtime": 12.492,
      "eval_samples_per_second": 729.025,
      "eval_steps_per_second": 1.441,
      "step": 250
    },
    {
      "epoch": 0.53,
      "learning_rate": 3.5085235442464923e-06,
      "loss": 0.8072,
      "step": 300
    },
    {
      "epoch": 0.53,
      "eval_f1": 0.6717305786202503,
      "eval_loss": 1.0609983205795288,
      "eval_runtime": 12.6873,
      "eval_samples_per_second": 717.806,
      "eval_steps_per_second": 1.419,
      "step": 300
    },
    {
      "epoch": 0.61,
      "learning_rate": 3.439863592304291e-06,
      "loss": 0.7549,
      "step": 350
    },
    {
      "epoch": 0.61,
      "eval_f1": 0.7228793656517608,
      "eval_loss": 0.9494264721870422,
      "eval_runtime": 12.8373,
      "eval_samples_per_second": 709.416,
      "eval_steps_per_second": 1.402,
      "step": 350
    },
    {
      "epoch": 0.7,
      "learning_rate": 3.3712036403620893e-06,
      "loss": 0.6644,
      "step": 400
    },
    {
      "epoch": 0.7,
      "eval_f1": 0.7719549711033966,
      "eval_loss": 0.8139678239822388,
      "eval_runtime": 12.9124,
      "eval_samples_per_second": 705.29,
      "eval_steps_per_second": 1.394,
      "step": 400
    },
    {
      "epoch": 0.79,
      "learning_rate": 3.302543688419888e-06,
      "loss": 0.6306,
      "step": 450
    },
    {
      "epoch": 0.79,
      "eval_f1": 0.7738311952589042,
      "eval_loss": 0.7844408750534058,
      "eval_runtime": 12.8236,
      "eval_samples_per_second": 710.174,
      "eval_steps_per_second": 1.404,
      "step": 450
    },
    {
      "epoch": 0.88,
      "learning_rate": 3.2338837364776867e-06,
      "loss": 0.6093,
      "step": 500
    },
    {
      "epoch": 0.88,
      "eval_f1": 0.7850534531405503,
      "eval_loss": 0.7450836300849915,
      "eval_runtime": 12.7168,
      "eval_samples_per_second": 716.14,
      "eval_steps_per_second": 1.415,
      "step": 500
    },
    {
      "epoch": 0.96,
      "learning_rate": 3.1652237845354854e-06,
      "loss": 0.599,
      "step": 550
    },
    {
      "epoch": 0.96,
      "eval_f1": 0.7932118091019561,
      "eval_loss": 0.7213732004165649,
      "eval_runtime": 12.9294,
      "eval_samples_per_second": 704.366,
      "eval_steps_per_second": 1.392,
      "step": 550
    },
    {
      "epoch": 1.05,
      "learning_rate": 3.0965638325932837e-06,
      "loss": 0.5234,
      "step": 600
    },
    {
      "epoch": 1.05,
      "eval_f1": 0.8216123467232329,
      "eval_loss": 0.6262369751930237,
      "eval_runtime": 13.578,
      "eval_samples_per_second": 670.72,
      "eval_steps_per_second": 1.326,
      "step": 600
    },
    {
      "epoch": 1.14,
      "learning_rate": 3.0279038806510824e-06,
      "loss": 0.4465,
      "step": 650
    },
    {
      "epoch": 1.14,
      "eval_f1": 0.7867065499958213,
      "eval_loss": 0.7198110818862915,
      "eval_runtime": 13.2261,
      "eval_samples_per_second": 688.561,
      "eval_steps_per_second": 1.361,
      "step": 650
    },
    {
      "epoch": 1.23,
      "learning_rate": 2.959243928708881e-06,
      "loss": 0.4521,
      "step": 700
    },
    {
      "epoch": 1.23,
      "eval_f1": 0.8063113996725187,
      "eval_loss": 0.6468276381492615,
      "eval_runtime": 12.8551,
      "eval_samples_per_second": 708.435,
      "eval_steps_per_second": 1.4,
      "step": 700
    },
    {
      "epoch": 1.31,
      "learning_rate": 2.89058397676668e-06,
      "loss": 0.4339,
      "step": 750
    },
    {
      "epoch": 1.31,
      "eval_f1": 0.8166775515876931,
      "eval_loss": 0.6203234195709229,
      "eval_runtime": 12.669,
      "eval_samples_per_second": 718.842,
      "eval_steps_per_second": 1.421,
      "step": 750
    },
    {
      "epoch": 1.4,
      "learning_rate": 2.821924024824478e-06,
      "loss": 0.4757,
      "step": 800
    },
    {
      "epoch": 1.4,
      "eval_f1": 0.8161608316711845,
      "eval_loss": 0.6249090433120728,
      "eval_runtime": 12.9976,
      "eval_samples_per_second": 700.668,
      "eval_steps_per_second": 1.385,
      "step": 800
    },
    {
      "epoch": 1.49,
      "learning_rate": 2.753264072882277e-06,
      "loss": 0.4594,
      "step": 850
    },
    {
      "epoch": 1.49,
      "eval_f1": 0.8290833266193817,
      "eval_loss": 0.5703582167625427,
      "eval_runtime": 12.8799,
      "eval_samples_per_second": 707.072,
      "eval_steps_per_second": 1.398,
      "step": 850
    },
    {
      "epoch": 1.58,
      "learning_rate": 2.6846041209400755e-06,
      "loss": 0.4032,
      "step": 900
    },
    {
      "epoch": 1.58,
      "eval_f1": 0.8215012797906703,
      "eval_loss": 0.5891036987304688,
      "eval_runtime": 12.8777,
      "eval_samples_per_second": 707.193,
      "eval_steps_per_second": 1.398,
      "step": 900
    },
    {
      "epoch": 1.66,
      "learning_rate": 2.615944168997874e-06,
      "loss": 0.4204,
      "step": 950
    },
    {
      "epoch": 1.66,
      "eval_f1": 0.8327085312817779,
      "eval_loss": 0.5532290935516357,
      "eval_runtime": 12.7986,
      "eval_samples_per_second": 711.563,
      "eval_steps_per_second": 1.406,
      "step": 950
    },
    {
      "epoch": 1.75,
      "learning_rate": 2.5472842170556725e-06,
      "loss": 0.4386,
      "step": 1000
    },
    {
      "epoch": 1.75,
      "eval_f1": 0.8286389311967884,
      "eval_loss": 0.5586702227592468,
      "eval_runtime": 12.7254,
      "eval_samples_per_second": 715.653,
      "eval_steps_per_second": 1.414,
      "step": 1000
    },
    {
      "epoch": 1.84,
      "learning_rate": 2.478624265113471e-06,
      "loss": 0.4258,
      "step": 1050
    },
    {
      "epoch": 1.84,
      "eval_f1": 0.8194650098039734,
      "eval_loss": 0.5945160388946533,
      "eval_runtime": 12.8508,
      "eval_samples_per_second": 708.674,
      "eval_steps_per_second": 1.401,
      "step": 1050
    },
    {
      "epoch": 1.93,
      "learning_rate": 2.4099643131712695e-06,
      "loss": 0.4544,
      "step": 1100
    },
    {
      "epoch": 1.93,
      "eval_f1": 0.8274559023611527,
      "eval_loss": 0.5760587453842163,
      "eval_runtime": 13.4335,
      "eval_samples_per_second": 677.933,
      "eval_steps_per_second": 1.34,
      "step": 1100
    },
    {
      "epoch": 2.01,
      "learning_rate": 2.341304361229068e-06,
      "loss": 0.418,
      "step": 1150
    },
    {
      "epoch": 2.01,
      "eval_f1": 0.8304623515332191,
      "eval_loss": 0.5600340962409973,
      "eval_runtime": 13.2414,
      "eval_samples_per_second": 687.766,
      "eval_steps_per_second": 1.359,
      "step": 1150
    },
    {
      "epoch": 2.1,
      "learning_rate": 2.2726444092868665e-06,
      "loss": 0.3918,
      "step": 1200
    },
    {
      "epoch": 2.1,
      "eval_f1": 0.8230072950822092,
      "eval_loss": 0.5960156321525574,
      "eval_runtime": 12.9264,
      "eval_samples_per_second": 704.529,
      "eval_steps_per_second": 1.393,
      "step": 1200
    },
    {
      "epoch": 2.19,
      "learning_rate": 2.203984457344665e-06,
      "loss": 0.4001,
      "step": 1250
    },
    {
      "epoch": 2.19,
      "eval_f1": 0.8370136201480073,
      "eval_loss": 0.5412853360176086,
      "eval_runtime": 12.8464,
      "eval_samples_per_second": 708.917,
      "eval_steps_per_second": 1.401,
      "step": 1250
    },
    {
      "epoch": 2.28,
      "learning_rate": 2.135324505402464e-06,
      "loss": 0.3369,
      "step": 1300
    },
    {
      "epoch": 2.28,
      "eval_f1": 0.8144549401231569,
      "eval_loss": 0.6154327392578125,
      "eval_runtime": 12.8518,
      "eval_samples_per_second": 708.616,
      "eval_steps_per_second": 1.401,
      "step": 1300
    },
    {
      "epoch": 2.36,
      "learning_rate": 2.0666645534602626e-06,
      "loss": 0.3639,
      "step": 1350
    },
    {
      "epoch": 2.36,
      "eval_f1": 0.8218463781690494,
      "eval_loss": 0.6043773889541626,
      "eval_runtime": 12.8354,
      "eval_samples_per_second": 709.522,
      "eval_steps_per_second": 1.402,
      "step": 1350
    },
    {
      "epoch": 2.45,
      "learning_rate": 1.998004601518061e-06,
      "loss": 0.305,
      "step": 1400
    },
    {
      "epoch": 2.45,
      "eval_f1": 0.8330006702059005,
      "eval_loss": 0.5544472336769104,
      "eval_runtime": 12.869,
      "eval_samples_per_second": 707.672,
      "eval_steps_per_second": 1.399,
      "step": 1400
    },
    {
      "epoch": 2.54,
      "learning_rate": 1.9293446495758596e-06,
      "loss": 0.4444,
      "step": 1450
    },
    {
      "epoch": 2.54,
      "eval_f1": 0.8346019595913722,
      "eval_loss": 0.559039294719696,
      "eval_runtime": 12.8371,
      "eval_samples_per_second": 709.431,
      "eval_steps_per_second": 1.402,
      "step": 1450
    },
    {
      "epoch": 2.63,
      "learning_rate": 1.8606846976336585e-06,
      "loss": 0.3428,
      "step": 1500
    },
    {
      "epoch": 2.63,
      "eval_f1": 0.8346071667451987,
      "eval_loss": 0.5472803115844727,
      "eval_runtime": 12.6738,
      "eval_samples_per_second": 718.567,
      "eval_steps_per_second": 1.42,
      "step": 1500
    },
    {
      "epoch": 2.71,
      "learning_rate": 1.792024745691457e-06,
      "loss": 0.473,
      "step": 1550
    },
    {
      "epoch": 2.71,
      "eval_f1": 0.8339861521101763,
      "eval_loss": 0.5484662055969238,
      "eval_runtime": 12.9188,
      "eval_samples_per_second": 704.942,
      "eval_steps_per_second": 1.393,
      "step": 1550
    },
    {
      "epoch": 2.8,
      "learning_rate": 1.7233647937492555e-06,
      "loss": 0.3423,
      "step": 1600
    },
    {
      "epoch": 2.8,
      "eval_f1": 0.834766650650793,
      "eval_loss": 0.5558911561965942,
      "eval_runtime": 13.6223,
      "eval_samples_per_second": 668.535,
      "eval_steps_per_second": 1.321,
      "step": 1600
    },
    {
      "epoch": 2.89,
      "learning_rate": 1.654704841807054e-06,
      "loss": 0.3275,
      "step": 1650
    },
    {
      "epoch": 2.89,
      "eval_f1": 0.8462797776414777,
      "eval_loss": 0.5118501782417297,
      "eval_runtime": 13.6479,
      "eval_samples_per_second": 667.283,
      "eval_steps_per_second": 1.319,
      "step": 1650
    },
    {
      "epoch": 2.98,
      "learning_rate": 1.5860448898648526e-06,
      "loss": 0.3701,
      "step": 1700
    },
    {
      "epoch": 2.98,
      "eval_f1": 0.8425787867333626,
      "eval_loss": 0.5238479375839233,
      "eval_runtime": 12.8261,
      "eval_samples_per_second": 710.034,
      "eval_steps_per_second": 1.403,
      "step": 1700
    },
    {
      "epoch": 3.06,
      "learning_rate": 1.5173849379226511e-06,
      "loss": 0.3139,
      "step": 1750
    },
    {
      "epoch": 3.06,
      "eval_f1": 0.8416221241169144,
      "eval_loss": 0.5329692959785461,
      "eval_runtime": 12.7867,
      "eval_samples_per_second": 712.223,
      "eval_steps_per_second": 1.408,
      "step": 1750
    },
    {
      "epoch": 3.15,
      "learning_rate": 1.4487249859804498e-06,
      "loss": 0.2975,
      "step": 1800
    },
    {
      "epoch": 3.15,
      "eval_f1": 0.8343340529328881,
      "eval_loss": 0.5499953627586365,
      "eval_runtime": 12.7893,
      "eval_samples_per_second": 712.078,
      "eval_steps_per_second": 1.407,
      "step": 1800
    },
    {
      "epoch": 3.24,
      "learning_rate": 1.3800650340382483e-06,
      "loss": 0.2981,
      "step": 1850
    },
    {
      "epoch": 3.24,
      "eval_f1": 0.8350192040740547,
      "eval_loss": 0.5687865018844604,
      "eval_runtime": 12.92,
      "eval_samples_per_second": 704.877,
      "eval_steps_per_second": 1.393,
      "step": 1850
    },
    {
      "epoch": 3.33,
      "learning_rate": 1.311405082096047e-06,
      "loss": 0.2378,
      "step": 1900
    },
    {
      "epoch": 3.33,
      "eval_f1": 0.8413969129271989,
      "eval_loss": 0.5414060354232788,
      "eval_runtime": 12.8936,
      "eval_samples_per_second": 706.32,
      "eval_steps_per_second": 1.396,
      "step": 1900
    },
    {
      "epoch": 3.42,
      "learning_rate": 1.2427451301538455e-06,
      "loss": 0.3733,
      "step": 1950
    },
    {
      "epoch": 3.42,
      "eval_f1": 0.8405393522238971,
      "eval_loss": 0.5475384593009949,
      "eval_runtime": 12.7243,
      "eval_samples_per_second": 715.714,
      "eval_steps_per_second": 1.415,
      "step": 1950
    },
    {
      "epoch": 3.5,
      "learning_rate": 1.1740851782116442e-06,
      "loss": 0.3825,
      "step": 2000
    },
    {
      "epoch": 3.5,
      "eval_f1": 0.8412270942316189,
      "eval_loss": 0.5363861322402954,
      "eval_runtime": 12.7385,
      "eval_samples_per_second": 714.922,
      "eval_steps_per_second": 1.413,
      "step": 2000
    },
    {
      "epoch": 3.59,
      "learning_rate": 1.1054252262694427e-06,
      "loss": 0.354,
      "step": 2050
    },
    {
      "epoch": 3.59,
      "eval_f1": 0.8435130537584258,
      "eval_loss": 0.5469411015510559,
      "eval_runtime": 13.1572,
      "eval_samples_per_second": 692.17,
      "eval_steps_per_second": 1.368,
      "step": 2050
    },
    {
      "epoch": 3.68,
      "learning_rate": 1.0367652743272414e-06,
      "loss": 0.2521,
      "step": 2100
    },
    {
      "epoch": 3.68,
      "eval_f1": 0.8351566925627497,
      "eval_loss": 0.5897064208984375,
      "eval_runtime": 13.5969,
      "eval_samples_per_second": 669.785,
      "eval_steps_per_second": 1.324,
      "step": 2100
    },
    {
      "epoch": 3.77,
      "learning_rate": 9.6810532238504e-07,
      "loss": 0.3446,
      "step": 2150
    },
    {
      "epoch": 3.77,
      "eval_f1": 0.8367972150621695,
      "eval_loss": 0.588862955570221,
      "eval_runtime": 13.2396,
      "eval_samples_per_second": 687.863,
      "eval_steps_per_second": 1.36,
      "step": 2150
    },
    {
      "epoch": 3.85,
      "learning_rate": 8.994453704428385e-07,
      "loss": 0.3163,
      "step": 2200
    },
    {
      "epoch": 3.85,
      "eval_f1": 0.8391334175384315,
      "eval_loss": 0.5625515580177307,
      "eval_runtime": 12.8463,
      "eval_samples_per_second": 708.921,
      "eval_steps_per_second": 1.401,
      "step": 2200
    },
    {
      "epoch": 3.94,
      "learning_rate": 8.307854185006371e-07,
      "loss": 0.2567,
      "step": 2250
    },
    {
      "epoch": 3.94,
      "eval_f1": 0.8368910454697663,
      "eval_loss": 0.5832029581069946,
      "eval_runtime": 12.768,
      "eval_samples_per_second": 713.27,
      "eval_steps_per_second": 1.41,
      "step": 2250
    },
    {
      "epoch": 4.03,
      "learning_rate": 7.621254665584357e-07,
      "loss": 0.3499,
      "step": 2300
    },
    {
      "epoch": 4.03,
      "eval_f1": 0.8444657615396232,
      "eval_loss": 0.5441297292709351,
      "eval_runtime": 12.9731,
      "eval_samples_per_second": 701.993,
      "eval_steps_per_second": 1.387,
      "step": 2300
    },
    {
      "epoch": 4.12,
      "learning_rate": 6.934655146162342e-07,
      "loss": 0.2326,
      "step": 2350
    },
    {
      "epoch": 4.12,
      "eval_f1": 0.8450447921852559,
      "eval_loss": 0.55018550157547,
      "eval_runtime": 13.7914,
      "eval_samples_per_second": 660.338,
      "eval_steps_per_second": 1.305,
      "step": 2350
    },
    {
      "epoch": 4.2,
      "learning_rate": 6.248055626740328e-07,
      "loss": 0.2876,
      "step": 2400
    },
    {
      "epoch": 4.2,
      "eval_f1": 0.8431017293241522,
      "eval_loss": 0.5501366257667542,
      "eval_runtime": 13.0119,
      "eval_samples_per_second": 699.898,
      "eval_steps_per_second": 1.383,
      "step": 2400
    },
    {
      "epoch": 4.29,
      "learning_rate": 5.561456107318314e-07,
      "loss": 0.2774,
      "step": 2450
    },
    {
      "epoch": 4.29,
      "eval_f1": 0.837914574132576,
      "eval_loss": 0.5795889496803284,
      "eval_runtime": 13.1344,
      "eval_samples_per_second": 693.37,
      "eval_steps_per_second": 1.37,
      "step": 2450
    },
    {
      "epoch": 4.38,
      "learning_rate": 4.8748565878963e-07,
      "loss": 0.3329,
      "step": 2500
    },
    {
      "epoch": 4.38,
      "eval_f1": 0.8421845881825929,
      "eval_loss": 0.5610998272895813,
      "eval_runtime": 12.7868,
      "eval_samples_per_second": 712.221,
      "eval_steps_per_second": 1.408,
      "step": 2500
    }
  ],
  "logging_steps": 50,
  "max_steps": 2855,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 203702323455600.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": {
    "learning_rate": 3.851823303957499e-06,
    "num_train_epochs": 5,
    "per_device_train_batch_size": 8,
    "seed": 30
  }
}