{
  "best_metric": 0.7364675967036229,
  "best_model_checkpoint": "./CARES/checkpoints/bio-ber-stratified/run-3/checkpoint-4402",
  "epoch": 62.0,
  "global_step": 4402,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 2.0,
      "eval_loss": 0.3057152032852173,
      "eval_macro_f1": 0.0,
      "eval_macro_precision": 0.0,
      "eval_macro_recall": 0.0,
      "eval_micro_f1": 0.0,
      "eval_micro_precision": 0.0,
      "eval_micro_recall": 0.0,
      "eval_runtime": 2.8103,
      "eval_samples_per_second": 343.733,
      "eval_steps_per_second": 11.031,
      "step": 142
    },
    {
      "epoch": 4.0,
      "eval_loss": 0.2555946707725525,
      "eval_macro_f1": 0.030103995621237,
      "eval_macro_precision": 0.11450892857142857,
      "eval_macro_recall": 0.018604569384720493,
      "eval_micro_f1": 0.06492679821769573,
      "eval_micro_precision": 0.9444444444444444,
      "eval_micro_recall": 0.03361898483849703,
      "eval_runtime": 2.814,
      "eval_samples_per_second": 343.287,
      "eval_steps_per_second": 11.016,
      "step": 284
    },
    {
      "epoch": 6.0,
      "eval_loss": 0.20202794671058655,
      "eval_macro_f1": 0.20731282317116273,
      "eval_macro_precision": 0.31073371688318574,
      "eval_macro_recall": 0.17971428447401858,
      "eval_micro_f1": 0.5260370697263901,
      "eval_micro_precision": 0.7957276368491322,
      "eval_micro_recall": 0.3928806855636124,
      "eval_runtime": 2.8153,
      "eval_samples_per_second": 343.125,
      "eval_steps_per_second": 11.011,
      "step": 426
    },
    {
      "epoch": 7.04,
      "learning_rate": 3.3448027530200134e-05,
      "loss": 0.3166,
      "step": 500
    },
    {
      "epoch": 8.0,
      "eval_loss": 0.16987857222557068,
      "eval_macro_f1": 0.33747696352729556,
      "eval_macro_precision": 0.5216496556534496,
      "eval_macro_recall": 0.3003629712640149,
      "eval_micro_f1": 0.6596736596736597,
      "eval_micro_precision": 0.8032166508987701,
      "eval_micro_recall": 0.5596572181938035,
      "eval_runtime": 2.8142,
      "eval_samples_per_second": 343.259,
      "eval_steps_per_second": 11.016,
      "step": 568
    },
    {
      "epoch": 10.0,
      "eval_loss": 0.15612231194972992,
      "eval_macro_f1": 0.3962525677685097,
      "eval_macro_precision": 0.5709879791429361,
      "eval_macro_recall": 0.34789915051543907,
      "eval_micro_f1": 0.6866002214839424,
      "eval_micro_precision": 0.7802013422818792,
      "eval_micro_recall": 0.6130520764667106,
      "eval_runtime": 2.8176,
      "eval_samples_per_second": 342.846,
      "eval_steps_per_second": 11.002,
      "step": 710
    },
    {
      "epoch": 12.0,
      "eval_loss": 0.15700845420360565,
      "eval_macro_f1": 0.5033118074728004,
      "eval_macro_precision": 0.6788170359469108,
      "eval_macro_recall": 0.4359827145940205,
      "eval_micro_f1": 0.7075812274368232,
      "eval_micro_precision": 0.7821229050279329,
      "eval_micro_recall": 0.6460118655240606,
      "eval_runtime": 2.9435,
      "eval_samples_per_second": 328.182,
      "eval_steps_per_second": 10.532,
      "step": 852
    },
    {
      "epoch": 14.0,
      "eval_loss": 0.14791876077651978,
      "eval_macro_f1": 0.5201462693304842,
      "eval_macro_precision": 0.6808348875622964,
      "eval_macro_recall": 0.45464762510361434,
      "eval_micro_f1": 0.7195077813970322,
      "eval_micro_precision": 0.797752808988764,
      "eval_micro_recall": 0.6552406064601186,
      "eval_runtime": 2.814,
      "eval_samples_per_second": 343.289,
      "eval_steps_per_second": 11.017,
      "step": 994
    },
    {
      "epoch": 14.08,
      "learning_rate": 3.091408605063951e-05,
      "loss": 0.1146,
      "step": 1000
    },
    {
      "epoch": 16.0,
      "eval_loss": 0.15235331654548645,
      "eval_macro_f1": 0.560037439235141,
      "eval_macro_precision": 0.6537080668718289,
      "eval_macro_recall": 0.5093665280810307,
      "eval_micro_f1": 0.7277091906721537,
      "eval_micro_precision": 0.7583988563259471,
      "eval_micro_recall": 0.6994067237969677,
      "eval_runtime": 2.8165,
      "eval_samples_per_second": 342.975,
      "eval_steps_per_second": 11.006,
      "step": 1136
    },
    {
      "epoch": 18.0,
      "eval_loss": 0.1468934863805771,
      "eval_macro_f1": 0.5752898959236763,
      "eval_macro_precision": 0.6785878888299093,
      "eval_macro_recall": 0.5191620319354175,
      "eval_micro_f1": 0.7432950191570881,
      "eval_micro_precision": 0.7880354505169868,
      "eval_micro_recall": 0.7033618984838497,
      "eval_runtime": 2.8153,
      "eval_samples_per_second": 343.13,
      "eval_steps_per_second": 11.011,
      "step": 1278
    },
    {
      "epoch": 20.0,
      "eval_loss": 0.15318024158477783,
      "eval_macro_f1": 0.5614807929032789,
      "eval_macro_precision": 0.6891794820810673,
      "eval_macro_recall": 0.5011545072485702,
      "eval_micro_f1": 0.7351351351351351,
      "eval_micro_precision": 0.8108108108108109,
      "eval_micro_recall": 0.6723796967699407,
      "eval_runtime": 2.8153,
      "eval_samples_per_second": 343.128,
      "eval_steps_per_second": 11.011,
      "step": 1420
    },
    {
      "epoch": 21.13,
      "learning_rate": 2.83801445710789e-05,
      "loss": 0.0497,
      "step": 1500
    },
    {
      "epoch": 22.0,
      "eval_loss": 0.15478584170341492,
      "eval_macro_f1": 0.5700524565361512,
      "eval_macro_precision": 0.6592231713192768,
      "eval_macro_recall": 0.5224052472361357,
      "eval_micro_f1": 0.7377900935226879,
      "eval_micro_precision": 0.7773722627737226,
      "eval_micro_recall": 0.7020435069215557,
      "eval_runtime": 2.8156,
      "eval_samples_per_second": 343.089,
      "eval_steps_per_second": 11.01,
      "step": 1562
    },
    {
      "epoch": 24.0,
      "eval_loss": 0.16469129920005798,
      "eval_macro_f1": 0.5893809890315513,
      "eval_macro_precision": 0.7250493831689879,
      "eval_macro_recall": 0.5372381591153548,
      "eval_micro_f1": 0.7337398373983739,
      "eval_micro_precision": 0.7547038327526132,
      "eval_micro_recall": 0.7139090309822017,
      "eval_runtime": 2.8148,
      "eval_samples_per_second": 343.185,
      "eval_steps_per_second": 11.013,
      "step": 1704
    },
    {
      "epoch": 26.0,
      "eval_loss": 0.1553143858909607,
      "eval_macro_f1": 0.5787478878813199,
      "eval_macro_precision": 0.6623158627962805,
      "eval_macro_recall": 0.5281203519575204,
      "eval_micro_f1": 0.7432293452176895,
      "eval_micro_precision": 0.7742857142857142,
      "eval_micro_recall": 0.7145682267633487,
      "eval_runtime": 2.8124,
      "eval_samples_per_second": 343.477,
      "eval_steps_per_second": 11.023,
      "step": 1846
    },
    {
      "epoch": 28.0,
      "eval_loss": 0.1586785614490509,
      "eval_macro_f1": 0.5883554607456316,
      "eval_macro_precision": 0.7177006367498535,
      "eval_macro_recall": 0.5324948845496986,
      "eval_micro_f1": 0.7415575465196417,
      "eval_micro_precision": 0.7768953068592058,
      "eval_micro_recall": 0.7092946605141727,
      "eval_runtime": 2.8164,
      "eval_samples_per_second": 342.99,
      "eval_steps_per_second": 11.007,
      "step": 1988
    },
    {
      "epoch": 28.17,
      "learning_rate": 2.5846203091518283e-05,
      "loss": 0.0242,
      "step": 2000
    },
    {
      "epoch": 30.0,
      "eval_loss": 0.16228000819683075,
      "eval_macro_f1": 0.6422229361530924,
      "eval_macro_precision": 0.790092149978405,
      "eval_macro_recall": 0.5761555188160953,
      "eval_micro_f1": 0.75272599366866,
      "eval_micro_precision": 0.8069381598793364,
      "eval_micro_recall": 0.7053394858272907,
      "eval_runtime": 2.8172,
      "eval_samples_per_second": 342.897,
      "eval_steps_per_second": 11.004,
      "step": 2130
    },
    {
      "epoch": 32.0,
      "eval_loss": 0.15960222482681274,
      "eval_macro_f1": 0.6566298525236257,
      "eval_macro_precision": 0.74358025279367,
      "eval_macro_recall": 0.6057522843683747,
      "eval_micro_f1": 0.7514529914529915,
      "eval_micro_precision": 0.7805397727272727,
      "eval_micro_recall": 0.7244561634805537,
      "eval_runtime": 2.8157,
      "eval_samples_per_second": 343.071,
      "eval_steps_per_second": 11.01,
      "step": 2272
    },
    {
      "epoch": 34.0,
      "eval_loss": 0.1677185595035553,
      "eval_macro_f1": 0.6556730308579793,
      "eval_macro_precision": 0.7959295044168222,
      "eval_macro_recall": 0.5949893771939523,
      "eval_micro_f1": 0.7518796992481203,
      "eval_micro_precision": 0.7806955287437899,
      "eval_micro_recall": 0.7251153592617007,
      "eval_runtime": 2.8129,
      "eval_samples_per_second": 343.412,
      "eval_steps_per_second": 11.02,
      "step": 2414
    },
    {
      "epoch": 35.21,
      "learning_rate": 2.331226161195767e-05,
      "loss": 0.0135,
      "step": 2500
    },
    {
      "epoch": 36.0,
      "eval_loss": 0.16759739816188812,
      "eval_macro_f1": 0.6573674747389432,
      "eval_macro_precision": 0.771138294992533,
      "eval_macro_recall": 0.595599674775744,
      "eval_micro_f1": 0.7523187907935418,
      "eval_micro_precision": 0.7855093256814921,
      "eval_micro_recall": 0.7218193803559657,
      "eval_runtime": 2.8149,
      "eval_samples_per_second": 343.173,
      "eval_steps_per_second": 11.013,
      "step": 2556
    },
    {
      "epoch": 38.0,
      "eval_loss": 0.1679902970790863,
      "eval_macro_f1": 0.6639964013713953,
      "eval_macro_precision": 0.758530396471428,
      "eval_macro_recall": 0.6080264740396573,
      "eval_micro_f1": 0.7524888431170615,
      "eval_micro_precision": 0.7851002865329513,
      "eval_micro_recall": 0.7224785761371127,
      "eval_runtime": 2.8145,
      "eval_samples_per_second": 343.228,
      "eval_steps_per_second": 11.015,
      "step": 2698
    },
    {
      "epoch": 40.0,
      "eval_loss": 0.17791299521923065,
      "eval_macro_f1": 0.706771883149178,
      "eval_macro_precision": 0.8182896833237783,
      "eval_macro_recall": 0.6422691238382328,
      "eval_micro_f1": 0.7502562350529552,
      "eval_micro_precision": 0.7787234042553192,
      "eval_micro_recall": 0.7237969676994067,
      "eval_runtime": 2.8149,
      "eval_samples_per_second": 343.168,
      "eval_steps_per_second": 11.013,
      "step": 2840
    },
    {
      "epoch": 42.0,
      "eval_loss": 0.17416273057460785,
      "eval_macro_f1": 0.7042921705265328,
      "eval_macro_precision": 0.8549297777881784,
      "eval_macro_recall": 0.6327733363764919,
      "eval_micro_f1": 0.7576791808873721,
      "eval_micro_precision": 0.7855626326963907,
      "eval_micro_recall": 0.7317073170731707,
      "eval_runtime": 2.8128,
      "eval_samples_per_second": 343.433,
      "eval_steps_per_second": 11.021,
      "step": 2982
    },
    {
      "epoch": 42.25,
      "learning_rate": 2.077832013239705e-05,
      "loss": 0.0089,
      "step": 3000
    },
    {
      "epoch": 44.0,
      "eval_loss": 0.17721112072467804,
      "eval_macro_f1": 0.7133417842190977,
      "eval_macro_precision": 0.8405823975317646,
      "eval_macro_recall": 0.6551419828069049,
      "eval_micro_f1": 0.7551782682512733,
      "eval_micro_precision": 0.7787114845938375,
      "eval_micro_recall": 0.7330257086354647,
      "eval_runtime": 2.8136,
      "eval_samples_per_second": 343.328,
      "eval_steps_per_second": 11.018,
      "step": 3124
    },
    {
      "epoch": 46.0,
      "eval_loss": 0.17867934703826904,
      "eval_macro_f1": 0.7067733545432416,
      "eval_macro_precision": 0.8456299689876698,
      "eval_macro_recall": 0.6424972968645987,
      "eval_micro_f1": 0.7581743869209809,
      "eval_micro_precision": 0.7843551797040169,
      "eval_micro_recall": 0.7336849044166117,
      "eval_runtime": 2.8132,
      "eval_samples_per_second": 343.385,
      "eval_steps_per_second": 11.02,
      "step": 3266
    },
    {
      "epoch": 48.0,
      "eval_loss": 0.18271668255329132,
      "eval_macro_f1": 0.7082847532013592,
      "eval_macro_precision": 0.8433782766054858,
      "eval_macro_recall": 0.641905005936743,
      "eval_micro_f1": 0.7553444180522565,
      "eval_micro_precision": 0.7783216783216783,
      "eval_micro_recall": 0.7336849044166117,
      "eval_runtime": 2.8144,
      "eval_samples_per_second": 343.24,
      "eval_steps_per_second": 11.015,
      "step": 3408
    },
    {
      "epoch": 49.3,
      "learning_rate": 1.8244378652836435e-05,
      "loss": 0.0062,
      "step": 3500
    },
    {
      "epoch": 50.0,
      "eval_loss": 0.18576982617378235,
      "eval_macro_f1": 0.7214487691145703,
      "eval_macro_precision": 0.8418509428350153,
      "eval_macro_recall": 0.6584015327988482,
      "eval_micro_f1": 0.7551299589603283,
      "eval_micro_precision": 0.7846481876332623,
      "eval_micro_recall": 0.7277521423862887,
      "eval_runtime": 2.8151,
      "eval_samples_per_second": 343.147,
      "eval_steps_per_second": 11.012,
      "step": 3550
    },
    {
      "epoch": 52.0,
      "eval_loss": 0.18477760255336761,
      "eval_macro_f1": 0.7358242812719007,
      "eval_macro_precision": 0.8363784298641037,
      "eval_macro_recall": 0.682320650811777,
      "eval_micro_f1": 0.7623529411764705,
      "eval_micro_precision": 0.7777777777777778,
      "eval_micro_recall": 0.7475280158206987,
      "eval_runtime": 2.8152,
      "eval_samples_per_second": 343.143,
      "eval_steps_per_second": 11.012,
      "step": 3692
    },
    {
      "epoch": 54.0,
      "eval_loss": 0.18875756859779358,
      "eval_macro_f1": 0.7155614476825691,
      "eval_macro_precision": 0.8418373811449409,
      "eval_macro_recall": 0.6438779359984136,
      "eval_micro_f1": 0.7593035165585524,
      "eval_micro_precision": 0.7875354107648725,
      "eval_micro_recall": 0.7330257086354647,
      "eval_runtime": 2.8151,
      "eval_samples_per_second": 343.149,
      "eval_steps_per_second": 11.012,
      "step": 3834
    },
    {
      "epoch": 56.0,
      "eval_loss": 0.18950717151165009,
      "eval_macro_f1": 0.7224219338762576,
      "eval_macro_precision": 0.8143469166127995,
      "eval_macro_recall": 0.6672755163012613,
      "eval_micro_f1": 0.7565011820330969,
      "eval_micro_precision": 0.775623268698061,
      "eval_micro_recall": 0.7382992748846408,
      "eval_runtime": 2.8155,
      "eval_samples_per_second": 343.104,
      "eval_steps_per_second": 11.011,
      "step": 3976
    },
    {
      "epoch": 56.34,
      "learning_rate": 1.5710437173275818e-05,
      "loss": 0.0047,
      "step": 4000
    },
    {
      "epoch": 58.0,
      "eval_loss": 0.19801756739616394,
      "eval_macro_f1": 0.7241508401060731,
      "eval_macro_precision": 0.8273979735617242,
      "eval_macro_recall": 0.6621892130462943,
      "eval_micro_f1": 0.7576676777890125,
      "eval_micro_precision": 0.7751724137931034,
      "eval_micro_recall": 0.7409360580092288,
      "eval_runtime": 2.8137,
      "eval_samples_per_second": 343.316,
      "eval_steps_per_second": 11.017,
      "step": 4118
    },
    {
      "epoch": 60.0,
      "eval_loss": 0.194396510720253,
      "eval_macro_f1": 0.7287972344248086,
      "eval_macro_precision": 0.8203998323639319,
      "eval_macro_recall": 0.6775029786499779,
      "eval_micro_f1": 0.7583222370173102,
      "eval_micro_precision": 0.7659717552118359,
      "eval_micro_recall": 0.7508239947264338,
      "eval_runtime": 2.8142,
      "eval_samples_per_second": 343.264,
      "eval_steps_per_second": 11.016,
      "step": 4260
    },
    {
      "epoch": 62.0,
      "eval_loss": 0.19370318949222565,
      "eval_macro_f1": 0.7364675967036229,
      "eval_macro_precision": 0.8295903604136947,
      "eval_macro_recall": 0.6817263379239634,
      "eval_micro_f1": 0.7599051811716898,
      "eval_micro_precision": 0.7813370473537604,
      "eval_micro_recall": 0.7396176664469347,
      "eval_runtime": 2.8163,
      "eval_samples_per_second": 343.005,
      "eval_steps_per_second": 11.007,
      "step": 4402
    }
  ],
  "max_steps": 7100,
  "num_train_epochs": 100,
  "total_flos": 3.3093548336434176e+16,
  "trial_name": null,
  "trial_params": {
    "adam_epsilon": 3.039596615397574e-08,
    "learning_rate": 3.349363847683222e-05,
    "per_device_eval_batch_size": 32,
    "per_device_train_batch_size": 32,
    "seed": 326,
    "warmup_steps": 491,
    "weight_decay": 0.01214452830676255
  }
}