{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 14.99819298879653,
  "global_step": 41500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "compression_loss": 0.0,
      "distillation_loss": 10.801469802856445,
      "epoch": 0.0,
      "learning_rate": 6.999996751507258e-05,
      "loss": 12.3405,
      "step": 10,
      "task_loss": 3.04644775390625
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 5.152049541473389,
      "epoch": 0.01,
      "learning_rate": 6.999976899629009e-05,
      "loss": 7.0823,
      "step": 20,
      "task_loss": 2.0041749477386475
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 3.413750171661377,
      "epoch": 0.01,
      "learning_rate": 6.999939000692937e-05,
      "loss": 3.5252,
      "step": 30,
      "task_loss": 2.1580514907836914
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 2.6377995014190674,
      "epoch": 0.01,
      "learning_rate": 6.999883054894466e-05,
      "loss": 2.9633,
      "step": 40,
      "task_loss": 1.704897165298462
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 2.8139700889587402,
      "epoch": 0.02,
      "learning_rate": 6.999809062522068e-05,
      "loss": 2.4806,
      "step": 50,
      "task_loss": 1.2986035346984863
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 2.0872642993927,
      "epoch": 0.02,
      "learning_rate": 6.999717023957274e-05,
      "loss": 2.3172,
      "step": 60,
      "task_loss": 1.510830044746399
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 2.322617530822754,
      "epoch": 0.03,
      "learning_rate": 6.999606939674666e-05,
      "loss": 2.0705,
      "step": 70,
      "task_loss": 1.6173977851867676
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.930269718170166,
      "epoch": 0.03,
      "learning_rate": 6.999478810241875e-05,
      "loss": 1.8671,
      "step": 80,
      "task_loss": 1.3637044429779053
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 2.235046863555908,
      "epoch": 0.03,
      "learning_rate": 6.999332636319578e-05,
      "loss": 1.9769,
      "step": 90,
      "task_loss": 1.294555425643921
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 2.0402920246124268,
      "epoch": 0.04,
      "learning_rate": 6.999168418661498e-05,
      "loss": 1.9396,
      "step": 100,
      "task_loss": 1.5622482299804688
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.7098126411437988,
      "epoch": 0.04,
      "learning_rate": 6.9989861581144e-05,
      "loss": 1.8838,
      "step": 110,
      "task_loss": 1.0829225778579712
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.186718225479126,
      "epoch": 0.04,
      "learning_rate": 6.998785855618076e-05,
      "loss": 1.7517,
      "step": 120,
      "task_loss": 0.6862260699272156
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 2.135342597961426,
      "epoch": 0.05,
      "learning_rate": 6.998567512205353e-05,
      "loss": 1.8173,
      "step": 130,
      "task_loss": 1.4869736433029175
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.74942946434021,
      "epoch": 0.05,
      "learning_rate": 6.998331129002086e-05,
      "loss": 1.962,
      "step": 140,
      "task_loss": 1.1613945960998535
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.7254133224487305,
      "epoch": 0.05,
      "learning_rate": 6.998076707227144e-05,
      "loss": 1.7815,
      "step": 150,
      "task_loss": 1.527439832687378
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 2.164252758026123,
      "epoch": 0.06,
      "learning_rate": 6.997804248192412e-05,
      "loss": 1.7949,
      "step": 160,
      "task_loss": 1.0921921730041504
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 2.0819597244262695,
      "epoch": 0.06,
      "learning_rate": 6.997513753302782e-05,
      "loss": 1.765,
      "step": 170,
      "task_loss": 1.229843020439148
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.4309144020080566,
      "epoch": 0.07,
      "learning_rate": 6.99720522405614e-05,
      "loss": 1.6999,
      "step": 180,
      "task_loss": 1.0473328828811646
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.8363723754882812,
      "epoch": 0.07,
      "learning_rate": 6.996878662043369e-05,
      "loss": 1.6825,
      "step": 190,
      "task_loss": 1.4055719375610352
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.8027558326721191,
      "epoch": 0.07,
      "learning_rate": 6.996534068948332e-05,
      "loss": 1.7122,
      "step": 200,
      "task_loss": 1.052909255027771
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.6521565914154053,
      "epoch": 0.08,
      "learning_rate": 6.996171446547867e-05,
      "loss": 1.7636,
      "step": 210,
      "task_loss": 0.9150123596191406
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 2.1441216468811035,
      "epoch": 0.08,
      "learning_rate": 6.995790796711777e-05,
      "loss": 1.6195,
      "step": 220,
      "task_loss": 1.9888694286346436
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.6174334287643433,
      "epoch": 0.08,
      "learning_rate": 6.99539212140282e-05,
      "loss": 1.5489,
      "step": 230,
      "task_loss": 1.0295372009277344
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.3089139461517334,
      "epoch": 0.09,
      "learning_rate": 6.994975422676701e-05,
      "loss": 1.5913,
      "step": 240,
      "task_loss": 1.1961848735809326
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.507627010345459,
      "epoch": 0.09,
      "learning_rate": 6.994540702682057e-05,
      "loss": 1.5598,
      "step": 250,
      "task_loss": 1.0574748516082764
    },
    {
      "epoch": 0.09,
      "eval_exact_match": 75.62913907284768,
      "eval_f1": 85.12845450683153,
      "step": 250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.5679240226745605,
      "epoch": 0.09,
      "learning_rate": 6.994087963660452e-05,
      "loss": 1.5927,
      "step": 260,
      "task_loss": 0.7596392631530762
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.3486156463623047,
      "epoch": 0.1,
      "learning_rate": 6.99361720794636e-05,
      "loss": 1.5363,
      "step": 270,
      "task_loss": 0.8923729062080383
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.6968021392822266,
      "epoch": 0.1,
      "learning_rate": 6.993128437967155e-05,
      "loss": 1.5517,
      "step": 280,
      "task_loss": 1.2004411220550537
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.994243860244751,
      "epoch": 0.1,
      "learning_rate": 6.992621656243103e-05,
      "loss": 1.4897,
      "step": 290,
      "task_loss": 1.4071820974349976
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.4140949249267578,
      "epoch": 0.11,
      "learning_rate": 6.992096865387336e-05,
      "loss": 1.5032,
      "step": 300,
      "task_loss": 0.9100825786590576
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.4215054512023926,
      "epoch": 0.11,
      "learning_rate": 6.991554068105856e-05,
      "loss": 1.4617,
      "step": 310,
      "task_loss": 0.948255181312561
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.661365270614624,
      "epoch": 0.12,
      "learning_rate": 6.990993267197504e-05,
      "loss": 1.4116,
      "step": 320,
      "task_loss": 1.1182901859283447
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2898075580596924,
      "epoch": 0.12,
      "learning_rate": 6.990414465553964e-05,
      "loss": 1.5381,
      "step": 330,
      "task_loss": 0.8843156695365906
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2767791748046875,
      "epoch": 0.12,
      "learning_rate": 6.989817666159727e-05,
      "loss": 1.4393,
      "step": 340,
      "task_loss": 0.9783102869987488
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.4274914264678955,
      "epoch": 0.13,
      "learning_rate": 6.989202872092093e-05,
      "loss": 1.5174,
      "step": 350,
      "task_loss": 0.9928803443908691
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2917709350585938,
      "epoch": 0.13,
      "learning_rate": 6.988570086521147e-05,
      "loss": 1.3964,
      "step": 360,
      "task_loss": 1.268720269203186
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.3270267248153687,
      "epoch": 0.13,
      "learning_rate": 6.987919312709744e-05,
      "loss": 1.4736,
      "step": 370,
      "task_loss": 1.111161708831787
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.4867007732391357,
      "epoch": 0.14,
      "learning_rate": 6.987250554013494e-05,
      "loss": 1.5575,
      "step": 380,
      "task_loss": 1.3485462665557861
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.7727158069610596,
      "epoch": 0.14,
      "learning_rate": 6.98656381388074e-05,
      "loss": 1.399,
      "step": 390,
      "task_loss": 1.4341893196105957
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.7645524740219116,
      "epoch": 0.14,
      "learning_rate": 6.985859095852546e-05,
      "loss": 1.4392,
      "step": 400,
      "task_loss": 0.8381137847900391
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.3841367959976196,
      "epoch": 0.15,
      "learning_rate": 6.985136403562674e-05,
      "loss": 1.4883,
      "step": 410,
      "task_loss": 0.6091371774673462
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.4980037212371826,
      "epoch": 0.15,
      "learning_rate": 6.984395740737568e-05,
      "loss": 1.4821,
      "step": 420,
      "task_loss": 0.7701343297958374
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.4909740686416626,
      "epoch": 0.16,
      "learning_rate": 6.983637111196338e-05,
      "loss": 1.4347,
      "step": 430,
      "task_loss": 1.2313339710235596
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2218682765960693,
      "epoch": 0.16,
      "learning_rate": 6.982860518850731e-05,
      "loss": 1.355,
      "step": 440,
      "task_loss": 0.7704342603683472
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.997610330581665,
      "epoch": 0.16,
      "learning_rate": 6.982065967705115e-05,
      "loss": 1.3387,
      "step": 450,
      "task_loss": 0.5290695428848267
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.199681282043457,
      "epoch": 0.17,
      "learning_rate": 6.981253461856468e-05,
      "loss": 1.5446,
      "step": 460,
      "task_loss": 1.1124558448791504
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.501362681388855,
      "epoch": 0.17,
      "learning_rate": 6.980423005494342e-05,
      "loss": 1.3709,
      "step": 470,
      "task_loss": 0.956885814666748
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.6759693622589111,
      "epoch": 0.17,
      "learning_rate": 6.979574602900849e-05,
      "loss": 1.409,
      "step": 480,
      "task_loss": 1.1642181873321533
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.4492310285568237,
      "epoch": 0.18,
      "learning_rate": 6.978708258450635e-05,
      "loss": 1.4329,
      "step": 490,
      "task_loss": 1.134920358657837
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.5545307397842407,
      "epoch": 0.18,
      "learning_rate": 6.977823976610867e-05,
      "loss": 1.4856,
      "step": 500,
      "task_loss": 1.1181845664978027
    },
    {
      "epoch": 0.18,
      "eval_exact_match": 77.1050141911069,
      "eval_f1": 86.09609925606219,
      "step": 500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.4672131538391113,
      "epoch": 0.18,
      "learning_rate": 6.9769217619412e-05,
      "loss": 1.3902,
      "step": 510,
      "task_loss": 0.6892313957214355
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.4067072868347168,
      "epoch": 0.19,
      "learning_rate": 6.976001619093752e-05,
      "loss": 1.3474,
      "step": 520,
      "task_loss": 0.6431139707565308
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.3734104633331299,
      "epoch": 0.19,
      "learning_rate": 6.975063552813094e-05,
      "loss": 1.3806,
      "step": 530,
      "task_loss": 0.8170199990272522
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1358160972595215,
      "epoch": 0.2,
      "learning_rate": 6.974107567936206e-05,
      "loss": 1.3843,
      "step": 540,
      "task_loss": 0.5692929029464722
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.7749069929122925,
      "epoch": 0.2,
      "learning_rate": 6.973133669392473e-05,
      "loss": 1.4445,
      "step": 550,
      "task_loss": 1.323117971420288
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.689734697341919,
      "epoch": 0.2,
      "learning_rate": 6.97214186220364e-05,
      "loss": 1.3183,
      "step": 560,
      "task_loss": 1.1061980724334717
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.3462835550308228,
      "epoch": 0.21,
      "learning_rate": 6.971132151483798e-05,
      "loss": 1.2587,
      "step": 570,
      "task_loss": 0.793758749961853
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.4587080478668213,
      "epoch": 0.21,
      "learning_rate": 6.970104542439358e-05,
      "loss": 1.4055,
      "step": 580,
      "task_loss": 0.8984804153442383
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.171175241470337,
      "epoch": 0.21,
      "learning_rate": 6.969059040369014e-05,
      "loss": 1.2979,
      "step": 590,
      "task_loss": 0.8862461447715759
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.400400161743164,
      "epoch": 0.22,
      "learning_rate": 6.967995650663729e-05,
      "loss": 1.2604,
      "step": 600,
      "task_loss": 1.041499137878418
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.278060793876648,
      "epoch": 0.22,
      "learning_rate": 6.966914378806696e-05,
      "loss": 1.3942,
      "step": 610,
      "task_loss": 1.0624408721923828
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.905518651008606,
      "epoch": 0.22,
      "learning_rate": 6.965815230373319e-05,
      "loss": 1.3035,
      "step": 620,
      "task_loss": 0.7998524904251099
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2518730163574219,
      "epoch": 0.23,
      "learning_rate": 6.964698211031174e-05,
      "loss": 1.3988,
      "step": 630,
      "task_loss": 1.4076586961746216
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0938901901245117,
      "epoch": 0.23,
      "learning_rate": 6.963563326539988e-05,
      "loss": 1.3236,
      "step": 640,
      "task_loss": 0.5017621517181396
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.3720566034317017,
      "epoch": 0.23,
      "learning_rate": 6.962410582751611e-05,
      "loss": 1.3655,
      "step": 650,
      "task_loss": 1.3073033094406128
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.45675528049469,
      "epoch": 0.24,
      "learning_rate": 6.961239985609975e-05,
      "loss": 1.3654,
      "step": 660,
      "task_loss": 1.0983452796936035
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.594604730606079,
      "epoch": 0.24,
      "learning_rate": 6.960051541151072e-05,
      "loss": 1.2799,
      "step": 670,
      "task_loss": 1.1729819774627686
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.3936469554901123,
      "epoch": 0.25,
      "learning_rate": 6.958845255502922e-05,
      "loss": 1.2039,
      "step": 680,
      "task_loss": 0.9447198510169983
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.3353527784347534,
      "epoch": 0.25,
      "learning_rate": 6.95762113488554e-05,
      "loss": 1.2488,
      "step": 690,
      "task_loss": 1.1083095073699951
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.8204264640808105,
      "epoch": 0.25,
      "learning_rate": 6.956379185610903e-05,
      "loss": 1.3971,
      "step": 700,
      "task_loss": 1.080670714378357
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.3583054542541504,
      "epoch": 0.26,
      "learning_rate": 6.955119414082917e-05,
      "loss": 1.2583,
      "step": 710,
      "task_loss": 0.8659499883651733
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.4246025085449219,
      "epoch": 0.26,
      "learning_rate": 6.953841826797391e-05,
      "loss": 1.3295,
      "step": 720,
      "task_loss": 1.0010908842086792
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0018075704574585,
      "epoch": 0.26,
      "learning_rate": 6.952546430341994e-05,
      "loss": 1.3176,
      "step": 730,
      "task_loss": 0.6832550764083862
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1276683807373047,
      "epoch": 0.27,
      "learning_rate": 6.951233231396224e-05,
      "loss": 1.2458,
      "step": 740,
      "task_loss": 0.6161213517189026
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.3984005451202393,
      "epoch": 0.27,
      "learning_rate": 6.949902236731379e-05,
      "loss": 1.3199,
      "step": 750,
      "task_loss": 1.2446082830429077
    },
    {
      "epoch": 0.27,
      "eval_exact_match": 77.99432355723746,
      "eval_f1": 86.68746599929251,
      "step": 750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1631975173950195,
      "epoch": 0.27,
      "learning_rate": 6.948553453210512e-05,
      "loss": 1.2846,
      "step": 760,
      "task_loss": 0.5416216850280762
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2278193235397339,
      "epoch": 0.28,
      "learning_rate": 6.947186887788406e-05,
      "loss": 1.1666,
      "step": 770,
      "task_loss": 0.6427184343338013
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8423769474029541,
      "epoch": 0.28,
      "learning_rate": 6.945802547511531e-05,
      "loss": 1.1551,
      "step": 780,
      "task_loss": 0.5025490522384644
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0702672004699707,
      "epoch": 0.29,
      "learning_rate": 6.944400439518014e-05,
      "loss": 1.3203,
      "step": 790,
      "task_loss": 0.5747840404510498
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1548950672149658,
      "epoch": 0.29,
      "learning_rate": 6.94298057103759e-05,
      "loss": 1.325,
      "step": 800,
      "task_loss": 0.7011659741401672
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.390937089920044,
      "epoch": 0.29,
      "learning_rate": 6.941542949391584e-05,
      "loss": 1.3387,
      "step": 810,
      "task_loss": 0.6473915576934814
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9621462821960449,
      "epoch": 0.3,
      "learning_rate": 6.940087581992852e-05,
      "loss": 1.3016,
      "step": 820,
      "task_loss": 0.5471967458724976
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.5039995908737183,
      "epoch": 0.3,
      "learning_rate": 6.93861447634576e-05,
      "loss": 1.3924,
      "step": 830,
      "task_loss": 0.7582854628562927
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2158305644989014,
      "epoch": 0.3,
      "learning_rate": 6.937123640046135e-05,
      "loss": 1.339,
      "step": 840,
      "task_loss": 0.984902560710907
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.307792067527771,
      "epoch": 0.31,
      "learning_rate": 6.935615080781233e-05,
      "loss": 1.2595,
      "step": 850,
      "task_loss": 0.8387401103973389
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1716396808624268,
      "epoch": 0.31,
      "learning_rate": 6.934088806329689e-05,
      "loss": 1.1884,
      "step": 860,
      "task_loss": 0.7160657644271851
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.267090082168579,
      "epoch": 0.31,
      "learning_rate": 6.932544824561489e-05,
      "loss": 1.2543,
      "step": 870,
      "task_loss": 0.8390096426010132
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0325896739959717,
      "epoch": 0.32,
      "learning_rate": 6.930983143437924e-05,
      "loss": 1.245,
      "step": 880,
      "task_loss": 0.729494035243988
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.7409932613372803,
      "epoch": 0.32,
      "learning_rate": 6.929403771011544e-05,
      "loss": 1.2879,
      "step": 890,
      "task_loss": 1.1845016479492188
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8987354636192322,
      "epoch": 0.33,
      "learning_rate": 6.927806715426126e-05,
      "loss": 1.1892,
      "step": 900,
      "task_loss": 0.8024774193763733
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.344858169555664,
      "epoch": 0.33,
      "learning_rate": 6.926191984916625e-05,
      "loss": 1.2256,
      "step": 910,
      "task_loss": 0.9374756217002869
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.851283609867096,
      "epoch": 0.33,
      "learning_rate": 6.924559587809135e-05,
      "loss": 1.1071,
      "step": 920,
      "task_loss": 0.7503871917724609
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0525801181793213,
      "epoch": 0.34,
      "learning_rate": 6.922909532520844e-05,
      "loss": 1.323,
      "step": 930,
      "task_loss": 0.9204051494598389
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.303235411643982,
      "epoch": 0.34,
      "learning_rate": 6.921241827559993e-05,
      "loss": 1.307,
      "step": 940,
      "task_loss": 0.7826588153839111
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.189795970916748,
      "epoch": 0.34,
      "learning_rate": 6.919556481525826e-05,
      "loss": 1.44,
      "step": 950,
      "task_loss": 0.9468139410018921
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2204010486602783,
      "epoch": 0.35,
      "learning_rate": 6.917853503108559e-05,
      "loss": 1.25,
      "step": 960,
      "task_loss": 0.7199236154556274
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0518145561218262,
      "epoch": 0.35,
      "learning_rate": 6.91613290108932e-05,
      "loss": 1.296,
      "step": 970,
      "task_loss": 0.6551309823989868
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.3444669246673584,
      "epoch": 0.35,
      "learning_rate": 6.914394684340107e-05,
      "loss": 1.0248,
      "step": 980,
      "task_loss": 0.5821710824966431
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2943898439407349,
      "epoch": 0.36,
      "learning_rate": 6.912638861823755e-05,
      "loss": 1.0796,
      "step": 990,
      "task_loss": 0.9846506714820862
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1659986972808838,
      "epoch": 0.36,
      "learning_rate": 6.910865442593876e-05,
      "loss": 1.1452,
      "step": 1000,
      "task_loss": 0.8229370713233948
    },
    {
      "epoch": 0.36,
      "eval_exact_match": 78.06054872280038,
      "eval_f1": 86.791557332193,
      "step": 1000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.6198532581329346,
      "epoch": 0.37,
      "learning_rate": 6.909074435794811e-05,
      "loss": 1.4119,
      "step": 1010,
      "task_loss": 1.3580809831619263
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.59236741065979,
      "epoch": 0.37,
      "learning_rate": 6.907265850661598e-05,
      "loss": 1.2508,
      "step": 1020,
      "task_loss": 1.45587158203125
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1220039129257202,
      "epoch": 0.37,
      "learning_rate": 6.905439696519907e-05,
      "loss": 1.1633,
      "step": 1030,
      "task_loss": 0.9460667967796326
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.548439860343933,
      "epoch": 0.38,
      "learning_rate": 6.903595982786005e-05,
      "loss": 1.3139,
      "step": 1040,
      "task_loss": 1.329020380973816
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7319619059562683,
      "epoch": 0.38,
      "learning_rate": 6.901734718966698e-05,
      "loss": 1.1286,
      "step": 1050,
      "task_loss": 0.6876011490821838
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.02553391456604,
      "epoch": 0.38,
      "learning_rate": 6.89985591465929e-05,
      "loss": 1.1894,
      "step": 1060,
      "task_loss": 1.0004324913024902
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.4067671298980713,
      "epoch": 0.39,
      "learning_rate": 6.897959579551526e-05,
      "loss": 1.2614,
      "step": 1070,
      "task_loss": 0.950558066368103
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9523043632507324,
      "epoch": 0.39,
      "learning_rate": 6.89604572342155e-05,
      "loss": 1.1757,
      "step": 1080,
      "task_loss": 0.894980251789093
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.028804063796997,
      "epoch": 0.39,
      "learning_rate": 6.894114356137845e-05,
      "loss": 1.2074,
      "step": 1090,
      "task_loss": 0.6418243050575256
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8996524810791016,
      "epoch": 0.4,
      "learning_rate": 6.89216548765919e-05,
      "loss": 1.0727,
      "step": 1100,
      "task_loss": 0.49160271883010864
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2093851566314697,
      "epoch": 0.4,
      "learning_rate": 6.890199128034611e-05,
      "loss": 1.1026,
      "step": 1110,
      "task_loss": 0.8884088397026062
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.4207491874694824,
      "epoch": 0.4,
      "learning_rate": 6.888215287403314e-05,
      "loss": 1.2668,
      "step": 1120,
      "task_loss": 1.6956202983856201
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8989244699478149,
      "epoch": 0.41,
      "learning_rate": 6.88621397599465e-05,
      "loss": 1.1698,
      "step": 1130,
      "task_loss": 0.8216118216514587
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.310133934020996,
      "epoch": 0.41,
      "learning_rate": 6.884195204128056e-05,
      "loss": 1.2178,
      "step": 1140,
      "task_loss": 0.9177054166793823
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8592339158058167,
      "epoch": 0.42,
      "learning_rate": 6.882158982212994e-05,
      "loss": 1.1676,
      "step": 1150,
      "task_loss": 0.7672922611236572
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.31768798828125,
      "epoch": 0.42,
      "learning_rate": 6.880105320748914e-05,
      "loss": 1.2468,
      "step": 1160,
      "task_loss": 1.0583456754684448
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2563865184783936,
      "epoch": 0.42,
      "learning_rate": 6.878034230325181e-05,
      "loss": 1.2145,
      "step": 1170,
      "task_loss": 0.8758540749549866
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.582754373550415,
      "epoch": 0.43,
      "learning_rate": 6.875945721621036e-05,
      "loss": 1.2769,
      "step": 1180,
      "task_loss": 1.3132786750793457
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1652511358261108,
      "epoch": 0.43,
      "learning_rate": 6.873839805405531e-05,
      "loss": 1.1476,
      "step": 1190,
      "task_loss": 0.6254457831382751
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1324548721313477,
      "epoch": 0.43,
      "learning_rate": 6.871716492537479e-05,
      "loss": 1.1278,
      "step": 1200,
      "task_loss": 0.786486029624939
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2423601150512695,
      "epoch": 0.44,
      "learning_rate": 6.869575793965394e-05,
      "loss": 1.2086,
      "step": 1210,
      "task_loss": 0.7296125292778015
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.30234956741333,
      "epoch": 0.44,
      "learning_rate": 6.86741772072744e-05,
      "loss": 1.136,
      "step": 1220,
      "task_loss": 0.875715970993042
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0005595684051514,
      "epoch": 0.44,
      "learning_rate": 6.865242283951365e-05,
      "loss": 1.2373,
      "step": 1230,
      "task_loss": 1.0233049392700195
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.4597254991531372,
      "epoch": 0.45,
      "learning_rate": 6.863049494854456e-05,
      "loss": 1.2707,
      "step": 1240,
      "task_loss": 0.9770482182502747
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8547976016998291,
      "epoch": 0.45,
      "learning_rate": 6.86083936474347e-05,
      "loss": 1.0618,
      "step": 1250,
      "task_loss": 0.6300672888755798
    },
    {
      "epoch": 0.45,
      "eval_exact_match": 78.79848628193,
      "eval_f1": 87.37934185870961,
      "step": 1250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.3759208917617798,
      "epoch": 0.46,
      "learning_rate": 6.858611905014577e-05,
      "loss": 1.2313,
      "step": 1260,
      "task_loss": 1.2745630741119385
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9745676517486572,
      "epoch": 0.46,
      "learning_rate": 6.856367127153315e-05,
      "loss": 1.2901,
      "step": 1270,
      "task_loss": 0.6641950607299805
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1643112897872925,
      "epoch": 0.46,
      "learning_rate": 6.854105042734507e-05,
      "loss": 1.1501,
      "step": 1280,
      "task_loss": 1.071144938468933
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0921354293823242,
      "epoch": 0.47,
      "learning_rate": 6.851825663422227e-05,
      "loss": 1.1568,
      "step": 1290,
      "task_loss": 0.5675968527793884
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0819282531738281,
      "epoch": 0.47,
      "learning_rate": 6.849529000969717e-05,
      "loss": 0.9992,
      "step": 1300,
      "task_loss": 1.3584985733032227
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9455113410949707,
      "epoch": 0.47,
      "learning_rate": 6.847215067219341e-05,
      "loss": 1.0654,
      "step": 1310,
      "task_loss": 0.7708945870399475
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0378615856170654,
      "epoch": 0.48,
      "learning_rate": 6.844883874102523e-05,
      "loss": 1.2306,
      "step": 1320,
      "task_loss": 0.8547934293746948
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9078717231750488,
      "epoch": 0.48,
      "learning_rate": 6.842535433639674e-05,
      "loss": 1.094,
      "step": 1330,
      "task_loss": 0.7515894174575806
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2001479864120483,
      "epoch": 0.48,
      "learning_rate": 6.840169757940148e-05,
      "loss": 1.1639,
      "step": 1340,
      "task_loss": 0.9455994367599487
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.205193042755127,
      "epoch": 0.49,
      "learning_rate": 6.837786859202159e-05,
      "loss": 1.265,
      "step": 1350,
      "task_loss": 0.8357219696044922
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1558539867401123,
      "epoch": 0.49,
      "learning_rate": 6.835386749712738e-05,
      "loss": 1.1953,
      "step": 1360,
      "task_loss": 0.9024025797843933
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8633468151092529,
      "epoch": 0.5,
      "learning_rate": 6.832969441847655e-05,
      "loss": 1.1829,
      "step": 1370,
      "task_loss": 1.0007044076919556
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1772587299346924,
      "epoch": 0.5,
      "learning_rate": 6.830534948071362e-05,
      "loss": 1.1232,
      "step": 1380,
      "task_loss": 0.6399517059326172
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.416799783706665,
      "epoch": 0.5,
      "learning_rate": 6.828083280936928e-05,
      "loss": 1.1562,
      "step": 1390,
      "task_loss": 0.8481613397598267
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0655555725097656,
      "epoch": 0.51,
      "learning_rate": 6.825614453085974e-05,
      "loss": 1.2153,
      "step": 1400,
      "task_loss": 0.9690505266189575
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7563189268112183,
      "epoch": 0.51,
      "learning_rate": 6.823128477248606e-05,
      "loss": 1.0047,
      "step": 1410,
      "task_loss": 0.8691654205322266
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.11103355884552,
      "epoch": 0.51,
      "learning_rate": 6.82062536624335e-05,
      "loss": 1.2397,
      "step": 1420,
      "task_loss": 0.7043032050132751
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2859797477722168,
      "epoch": 0.52,
      "learning_rate": 6.81810513297709e-05,
      "loss": 1.1573,
      "step": 1430,
      "task_loss": 1.0251799821853638
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.4492874145507812,
      "epoch": 0.52,
      "learning_rate": 6.815567790444994e-05,
      "loss": 1.1774,
      "step": 1440,
      "task_loss": 0.8829349875450134
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.20381760597229,
      "epoch": 0.52,
      "learning_rate": 6.813013351730457e-05,
      "loss": 1.0484,
      "step": 1450,
      "task_loss": 1.0177350044250488
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2550363540649414,
      "epoch": 0.53,
      "learning_rate": 6.810441830005021e-05,
      "loss": 1.0561,
      "step": 1460,
      "task_loss": 0.7583893537521362
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1922001838684082,
      "epoch": 0.53,
      "learning_rate": 6.807853238528316e-05,
      "loss": 1.0369,
      "step": 1470,
      "task_loss": 0.8423135876655579
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.057039737701416,
      "epoch": 0.53,
      "learning_rate": 6.805247590647992e-05,
      "loss": 1.06,
      "step": 1480,
      "task_loss": 0.8067488670349121
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1017229557037354,
      "epoch": 0.54,
      "learning_rate": 6.802624899799646e-05,
      "loss": 1.2062,
      "step": 1490,
      "task_loss": 0.9069002866744995
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9764423370361328,
      "epoch": 0.54,
      "learning_rate": 6.799985179506753e-05,
      "loss": 1.2209,
      "step": 1500,
      "task_loss": 1.0442379713058472
    },
    {
      "epoch": 0.54,
      "eval_exact_match": 79.47965941343425,
      "eval_f1": 87.69873817803351,
      "step": 1500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9164371490478516,
      "epoch": 0.55,
      "learning_rate": 6.797328443380597e-05,
      "loss": 1.2163,
      "step": 1510,
      "task_loss": 1.1710762977600098
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0816839933395386,
      "epoch": 0.55,
      "learning_rate": 6.794654705120207e-05,
      "loss": 1.0914,
      "step": 1520,
      "task_loss": 1.251457691192627
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1803267002105713,
      "epoch": 0.55,
      "learning_rate": 6.791963978512273e-05,
      "loss": 1.0023,
      "step": 1530,
      "task_loss": 0.8527973890304565
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.8129152059555054,
      "epoch": 0.56,
      "learning_rate": 6.789256277431087e-05,
      "loss": 1.1583,
      "step": 1540,
      "task_loss": 1.6529321670532227
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.4472854137420654,
      "epoch": 0.56,
      "learning_rate": 6.786531615838467e-05,
      "loss": 1.2427,
      "step": 1550,
      "task_loss": 1.1761155128479004
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.314471960067749,
      "epoch": 0.56,
      "learning_rate": 6.783790007783683e-05,
      "loss": 1.2151,
      "step": 1560,
      "task_loss": 0.894078254699707
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9429694414138794,
      "epoch": 0.57,
      "learning_rate": 6.78103146740339e-05,
      "loss": 1.077,
      "step": 1570,
      "task_loss": 0.5526290535926819
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1317658424377441,
      "epoch": 0.57,
      "learning_rate": 6.778256008921548e-05,
      "loss": 1.0742,
      "step": 1580,
      "task_loss": 0.6145908236503601
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.015599250793457,
      "epoch": 0.57,
      "learning_rate": 6.775463646649355e-05,
      "loss": 1.0751,
      "step": 1590,
      "task_loss": 1.2883775234222412
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.221526861190796,
      "epoch": 0.58,
      "learning_rate": 6.77265439498517e-05,
      "loss": 1.0519,
      "step": 1600,
      "task_loss": 1.0645751953125
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1089495420455933,
      "epoch": 0.58,
      "learning_rate": 6.769828268414439e-05,
      "loss": 1.2171,
      "step": 1610,
      "task_loss": 0.4754822254180908
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7438245415687561,
      "epoch": 0.59,
      "learning_rate": 6.76698528150962e-05,
      "loss": 1.0148,
      "step": 1620,
      "task_loss": 0.8168939352035522
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9932070374488831,
      "epoch": 0.59,
      "learning_rate": 6.764125448930112e-05,
      "loss": 1.094,
      "step": 1630,
      "task_loss": 0.9427931308746338
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2198781967163086,
      "epoch": 0.59,
      "learning_rate": 6.761248785422172e-05,
      "loss": 1.1677,
      "step": 1640,
      "task_loss": 0.8359280824661255
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0543707609176636,
      "epoch": 0.6,
      "learning_rate": 6.758355305818843e-05,
      "loss": 1.121,
      "step": 1650,
      "task_loss": 0.6798961162567139
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9245864152908325,
      "epoch": 0.6,
      "learning_rate": 6.755445025039881e-05,
      "loss": 0.9735,
      "step": 1660,
      "task_loss": 0.8748812079429626
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9669818878173828,
      "epoch": 0.6,
      "learning_rate": 6.752517958091671e-05,
      "loss": 1.071,
      "step": 1670,
      "task_loss": 0.763600766658783
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2453193664550781,
      "epoch": 0.61,
      "learning_rate": 6.749574120067155e-05,
      "loss": 1.06,
      "step": 1680,
      "task_loss": 1.083929419517517
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5505318641662598,
      "epoch": 0.61,
      "learning_rate": 6.746613526145752e-05,
      "loss": 1.1745,
      "step": 1690,
      "task_loss": 0.41013821959495544
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.723199486732483,
      "epoch": 0.61,
      "learning_rate": 6.743636191593279e-05,
      "loss": 1.1728,
      "step": 1700,
      "task_loss": 1.5259857177734375
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8400850296020508,
      "epoch": 0.62,
      "learning_rate": 6.740642131761876e-05,
      "loss": 1.1087,
      "step": 1710,
      "task_loss": 0.8797809481620789
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.503807783126831,
      "epoch": 0.62,
      "learning_rate": 6.737631362089919e-05,
      "loss": 1.1093,
      "step": 1720,
      "task_loss": 1.2856862545013428
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.252486228942871,
      "epoch": 0.63,
      "learning_rate": 6.734603898101952e-05,
      "loss": 1.1645,
      "step": 1730,
      "task_loss": 1.240208625793457
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9471865892410278,
      "epoch": 0.63,
      "learning_rate": 6.731559755408595e-05,
      "loss": 1.0735,
      "step": 1740,
      "task_loss": 0.6639693975448608
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.4272645711898804,
      "epoch": 0.63,
      "learning_rate": 6.728498949706473e-05,
      "loss": 1.0731,
      "step": 1750,
      "task_loss": 1.374558448791504
    },
    {
      "epoch": 0.63,
      "eval_exact_match": 79.47019867549669,
      "eval_f1": 87.62942304951152,
      "step": 1750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1430143117904663,
      "epoch": 0.64,
      "learning_rate": 6.72542149677813e-05,
      "loss": 1.1903,
      "step": 1760,
      "task_loss": 1.1442666053771973
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9708960652351379,
      "epoch": 0.64,
      "learning_rate": 6.722327412491946e-05,
      "loss": 1.0846,
      "step": 1770,
      "task_loss": 0.7830320596694946
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9312682151794434,
      "epoch": 0.64,
      "learning_rate": 6.71921671280206e-05,
      "loss": 1.1422,
      "step": 1780,
      "task_loss": 0.8345022201538086
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.4343688488006592,
      "epoch": 0.65,
      "learning_rate": 6.716089413748289e-05,
      "loss": 1.031,
      "step": 1790,
      "task_loss": 1.1940171718597412
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1692103147506714,
      "epoch": 0.65,
      "learning_rate": 6.712945531456035e-05,
      "loss": 1.2076,
      "step": 1800,
      "task_loss": 0.9186697006225586
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1725080013275146,
      "epoch": 0.65,
      "learning_rate": 6.709785082136213e-05,
      "loss": 1.0877,
      "step": 1810,
      "task_loss": 1.2045694589614868
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9919499158859253,
      "epoch": 0.66,
      "learning_rate": 6.706608082085164e-05,
      "loss": 1.0455,
      "step": 1820,
      "task_loss": 0.9436089396476746
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1209893226623535,
      "epoch": 0.66,
      "learning_rate": 6.703414547684568e-05,
      "loss": 1.0689,
      "step": 1830,
      "task_loss": 0.7971674203872681
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8886414766311646,
      "epoch": 0.66,
      "learning_rate": 6.700204495401361e-05,
      "loss": 1.093,
      "step": 1840,
      "task_loss": 1.0051020383834839
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.602144718170166,
      "epoch": 0.67,
      "learning_rate": 6.696977941787651e-05,
      "loss": 1.1886,
      "step": 1850,
      "task_loss": 0.9657487273216248
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0558702945709229,
      "epoch": 0.67,
      "learning_rate": 6.693734903480639e-05,
      "loss": 1.0341,
      "step": 1860,
      "task_loss": 0.9395914673805237
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9531676173210144,
      "epoch": 0.68,
      "learning_rate": 6.690475397202515e-05,
      "loss": 0.9256,
      "step": 1870,
      "task_loss": 0.9502599239349365
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1095507144927979,
      "epoch": 0.68,
      "learning_rate": 6.687199439760392e-05,
      "loss": 1.0397,
      "step": 1880,
      "task_loss": 0.970070481300354
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0300889015197754,
      "epoch": 0.68,
      "learning_rate": 6.683907048046209e-05,
      "loss": 1.0262,
      "step": 1890,
      "task_loss": 0.8158546686172485
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9780499935150146,
      "epoch": 0.69,
      "learning_rate": 6.680598239036641e-05,
      "loss": 1.1005,
      "step": 1900,
      "task_loss": 0.8497573137283325
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0767104625701904,
      "epoch": 0.69,
      "learning_rate": 6.677273029793025e-05,
      "loss": 1.001,
      "step": 1910,
      "task_loss": 0.9059280157089233
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9213865995407104,
      "epoch": 0.69,
      "learning_rate": 6.673931437461255e-05,
      "loss": 1.098,
      "step": 1920,
      "task_loss": 0.615307629108429
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2905583381652832,
      "epoch": 0.7,
      "learning_rate": 6.670573479271705e-05,
      "loss": 1.1162,
      "step": 1930,
      "task_loss": 0.8289541006088257
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1863908767700195,
      "epoch": 0.7,
      "learning_rate": 6.667199172539137e-05,
      "loss": 1.0084,
      "step": 1940,
      "task_loss": 1.479468822479248
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2633111476898193,
      "epoch": 0.7,
      "learning_rate": 6.663808534662611e-05,
      "loss": 1.1662,
      "step": 1950,
      "task_loss": 0.7809505462646484
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9272475242614746,
      "epoch": 0.71,
      "learning_rate": 6.660401583125397e-05,
      "loss": 1.0658,
      "step": 1960,
      "task_loss": 1.3027081489562988
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0278266668319702,
      "epoch": 0.71,
      "learning_rate": 6.65697833549488e-05,
      "loss": 1.0035,
      "step": 1970,
      "task_loss": 0.5536607503890991
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1361572742462158,
      "epoch": 0.72,
      "learning_rate": 6.653538809422479e-05,
      "loss": 1.1669,
      "step": 1980,
      "task_loss": 1.4308404922485352
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8886505365371704,
      "epoch": 0.72,
      "learning_rate": 6.650083022643546e-05,
      "loss": 1.0813,
      "step": 1990,
      "task_loss": 1.0401955842971802
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9133187532424927,
      "epoch": 0.72,
      "learning_rate": 6.646610992977279e-05,
      "loss": 1.0828,
      "step": 2000,
      "task_loss": 0.5507737994194031
    },
    {
      "epoch": 0.72,
      "eval_exact_match": 79.2620624408704,
      "eval_f1": 87.39585165916291,
      "step": 2000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7194988131523132,
      "epoch": 0.73,
      "learning_rate": 6.643122738326632e-05,
      "loss": 1.0682,
      "step": 2010,
      "task_loss": 0.5585941076278687
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.098526120185852,
      "epoch": 0.73,
      "learning_rate": 6.639618276678217e-05,
      "loss": 1.0775,
      "step": 2020,
      "task_loss": 0.8350554704666138
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1135523319244385,
      "epoch": 0.73,
      "learning_rate": 6.636097626102219e-05,
      "loss": 1.1848,
      "step": 2030,
      "task_loss": 1.121304988861084
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9656478762626648,
      "epoch": 0.74,
      "learning_rate": 6.632560804752294e-05,
      "loss": 1.0541,
      "step": 2040,
      "task_loss": 0.7606736421585083
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.807013750076294,
      "epoch": 0.74,
      "learning_rate": 6.629007830865483e-05,
      "loss": 0.9216,
      "step": 2050,
      "task_loss": 0.5676593780517578
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.4709911346435547,
      "epoch": 0.74,
      "learning_rate": 6.625438722762114e-05,
      "loss": 1.0567,
      "step": 2060,
      "task_loss": 1.668703317642212
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.243257999420166,
      "epoch": 0.75,
      "learning_rate": 6.62185349884571e-05,
      "loss": 1.0476,
      "step": 2070,
      "task_loss": 1.2251255512237549
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8128871321678162,
      "epoch": 0.75,
      "learning_rate": 6.618252177602889e-05,
      "loss": 0.9901,
      "step": 2080,
      "task_loss": 0.7867209911346436
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.8095636367797852,
      "epoch": 0.76,
      "learning_rate": 6.614634777603274e-05,
      "loss": 1.1872,
      "step": 2090,
      "task_loss": 1.47059965133667
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0928575992584229,
      "epoch": 0.76,
      "learning_rate": 6.611001317499399e-05,
      "loss": 1.1672,
      "step": 2100,
      "task_loss": 1.0058388710021973
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8655899167060852,
      "epoch": 0.76,
      "learning_rate": 6.607351816026604e-05,
      "loss": 0.925,
      "step": 2110,
      "task_loss": 0.7716636657714844
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.830997109413147,
      "epoch": 0.77,
      "learning_rate": 6.603686292002945e-05,
      "loss": 1.0954,
      "step": 2120,
      "task_loss": 1.1273469924926758
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0073375701904297,
      "epoch": 0.77,
      "learning_rate": 6.600004764329098e-05,
      "loss": 1.0318,
      "step": 2130,
      "task_loss": 0.7276365756988525
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.026449203491211,
      "epoch": 0.77,
      "learning_rate": 6.596307251988257e-05,
      "loss": 1.1058,
      "step": 2140,
      "task_loss": 1.2019627094268799
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1347987651824951,
      "epoch": 0.78,
      "learning_rate": 6.592593774046039e-05,
      "loss": 1.0494,
      "step": 2150,
      "task_loss": 0.947940468788147
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9246293306350708,
      "epoch": 0.78,
      "learning_rate": 6.588864349650386e-05,
      "loss": 0.9727,
      "step": 2160,
      "task_loss": 0.8415713310241699
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9354982972145081,
      "epoch": 0.78,
      "learning_rate": 6.585118998031462e-05,
      "loss": 1.015,
      "step": 2170,
      "task_loss": 0.8214912414550781
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0691757202148438,
      "epoch": 0.79,
      "learning_rate": 6.581357738501561e-05,
      "loss": 1.0531,
      "step": 2180,
      "task_loss": 0.8481631278991699
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2228312492370605,
      "epoch": 0.79,
      "learning_rate": 6.577580590455002e-05,
      "loss": 1.1091,
      "step": 2190,
      "task_loss": 0.9836097955703735
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9064592123031616,
      "epoch": 0.8,
      "learning_rate": 6.57378757336803e-05,
      "loss": 0.9704,
      "step": 2200,
      "task_loss": 0.4665876626968384
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7420713305473328,
      "epoch": 0.8,
      "learning_rate": 6.569978706798717e-05,
      "loss": 1.0118,
      "step": 2210,
      "task_loss": 0.4307766556739807
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.6637189388275146,
      "epoch": 0.8,
      "learning_rate": 6.566154010386858e-05,
      "loss": 1.1637,
      "step": 2220,
      "task_loss": 1.3755857944488525
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.168283462524414,
      "epoch": 0.81,
      "learning_rate": 6.562313503853876e-05,
      "loss": 1.1656,
      "step": 2230,
      "task_loss": 0.7027386426925659
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8026998043060303,
      "epoch": 0.81,
      "learning_rate": 6.558457207002713e-05,
      "loss": 1.0892,
      "step": 2240,
      "task_loss": 0.700858473777771
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1535958051681519,
      "epoch": 0.81,
      "learning_rate": 6.554585139717734e-05,
      "loss": 1.0732,
      "step": 2250,
      "task_loss": 0.8288143873214722
    },
    {
      "epoch": 0.81,
      "eval_exact_match": 80.06622516556291,
      "eval_f1": 87.992821077547,
      "step": 2250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2591572999954224,
      "epoch": 0.82,
      "learning_rate": 6.55069732196462e-05,
      "loss": 1.144,
      "step": 2260,
      "task_loss": 0.8866115212440491
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0764415264129639,
      "epoch": 0.82,
      "learning_rate": 6.546793773790262e-05,
      "loss": 0.9642,
      "step": 2270,
      "task_loss": 0.891737699508667
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1543341875076294,
      "epoch": 0.82,
      "learning_rate": 6.542874515322672e-05,
      "loss": 0.9274,
      "step": 2280,
      "task_loss": 1.1221840381622314
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.189033031463623,
      "epoch": 0.83,
      "learning_rate": 6.538939566770863e-05,
      "loss": 0.946,
      "step": 2290,
      "task_loss": 1.2177841663360596
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1238713264465332,
      "epoch": 0.83,
      "learning_rate": 6.534988948424753e-05,
      "loss": 0.9513,
      "step": 2300,
      "task_loss": 0.7087922096252441
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0527355670928955,
      "epoch": 0.83,
      "learning_rate": 6.531022680655057e-05,
      "loss": 1.1627,
      "step": 2310,
      "task_loss": 1.0492477416992188
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1853148937225342,
      "epoch": 0.84,
      "learning_rate": 6.527040783913188e-05,
      "loss": 0.9871,
      "step": 2320,
      "task_loss": 1.1989734172821045
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1268264055252075,
      "epoch": 0.84,
      "learning_rate": 6.523043278731144e-05,
      "loss": 1.1276,
      "step": 2330,
      "task_loss": 0.7949765920639038
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.072948694229126,
      "epoch": 0.85,
      "learning_rate": 6.519030185721406e-05,
      "loss": 1.0278,
      "step": 2340,
      "task_loss": 0.9652178287506104
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2226154804229736,
      "epoch": 0.85,
      "learning_rate": 6.515001525576829e-05,
      "loss": 0.9951,
      "step": 2350,
      "task_loss": 0.9827218055725098
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9792837500572205,
      "epoch": 0.85,
      "learning_rate": 6.510957319070542e-05,
      "loss": 0.9745,
      "step": 2360,
      "task_loss": 0.744950532913208
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.3859732151031494,
      "epoch": 0.86,
      "learning_rate": 6.50689758705583e-05,
      "loss": 1.019,
      "step": 2370,
      "task_loss": 1.170689582824707
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8908461332321167,
      "epoch": 0.86,
      "learning_rate": 6.502822350466042e-05,
      "loss": 1.1045,
      "step": 2380,
      "task_loss": 0.838524580001831
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.280137062072754,
      "epoch": 0.86,
      "learning_rate": 6.498731630314464e-05,
      "loss": 0.948,
      "step": 2390,
      "task_loss": 1.2174863815307617
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.02490234375,
      "epoch": 0.87,
      "learning_rate": 6.494625447694226e-05,
      "loss": 1.0909,
      "step": 2400,
      "task_loss": 0.8274341821670532
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8467527627944946,
      "epoch": 0.87,
      "learning_rate": 6.490503823778187e-05,
      "loss": 1.0749,
      "step": 2410,
      "task_loss": 0.5999952554702759
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.4467501640319824,
      "epoch": 0.87,
      "learning_rate": 6.486366779818823e-05,
      "loss": 1.0028,
      "step": 2420,
      "task_loss": 0.9743595123291016
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9197993874549866,
      "epoch": 0.88,
      "learning_rate": 6.482214337148128e-05,
      "loss": 1.0118,
      "step": 2430,
      "task_loss": 0.9321801662445068
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1022918224334717,
      "epoch": 0.88,
      "learning_rate": 6.478046517177489e-05,
      "loss": 1.1273,
      "step": 2440,
      "task_loss": 0.7594503164291382
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9985268115997314,
      "epoch": 0.89,
      "learning_rate": 6.47386334139759e-05,
      "loss": 0.9987,
      "step": 2450,
      "task_loss": 0.5284379720687866
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5598468780517578,
      "epoch": 0.89,
      "learning_rate": 6.469664831378291e-05,
      "loss": 1.0219,
      "step": 2460,
      "task_loss": 0.8020057678222656
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7448583841323853,
      "epoch": 0.89,
      "learning_rate": 6.46545100876852e-05,
      "loss": 1.0504,
      "step": 2470,
      "task_loss": 0.7287012338638306
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8835182189941406,
      "epoch": 0.9,
      "learning_rate": 6.461221895296166e-05,
      "loss": 1.083,
      "step": 2480,
      "task_loss": 1.5722730159759521
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7862961292266846,
      "epoch": 0.9,
      "learning_rate": 6.45697751276796e-05,
      "loss": 1.0801,
      "step": 2490,
      "task_loss": 0.4984385371208191
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.5525429248809814,
      "epoch": 0.9,
      "learning_rate": 6.452717883069362e-05,
      "loss": 0.9646,
      "step": 2500,
      "task_loss": 1.5134918689727783
    },
    {
      "epoch": 0.9,
      "eval_exact_match": 80.54872280037843,
      "eval_f1": 88.31187998291739,
      "step": 2500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9102578163146973,
      "epoch": 0.91,
      "learning_rate": 6.448443028164458e-05,
      "loss": 1.0809,
      "step": 2510,
      "task_loss": 0.8942407369613647
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9936506152153015,
      "epoch": 0.91,
      "learning_rate": 6.444152970095839e-05,
      "loss": 1.0757,
      "step": 2520,
      "task_loss": 0.9664222598075867
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.041625738143921,
      "epoch": 0.91,
      "learning_rate": 6.439847730984484e-05,
      "loss": 1.1217,
      "step": 2530,
      "task_loss": 0.7623254656791687
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.021484375,
      "epoch": 0.92,
      "learning_rate": 6.435527333029654e-05,
      "loss": 0.8477,
      "step": 2540,
      "task_loss": 0.7754542231559753
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1402299404144287,
      "epoch": 0.92,
      "learning_rate": 6.431191798508773e-05,
      "loss": 1.0529,
      "step": 2550,
      "task_loss": 0.698540449142456
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9655864238739014,
      "epoch": 0.93,
      "learning_rate": 6.426841149777318e-05,
      "loss": 1.0713,
      "step": 2560,
      "task_loss": 0.7478283643722534
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8475412726402283,
      "epoch": 0.93,
      "learning_rate": 6.422475409268694e-05,
      "loss": 1.0224,
      "step": 2570,
      "task_loss": 0.922493040561676
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2475067377090454,
      "epoch": 0.93,
      "learning_rate": 6.418094599494128e-05,
      "loss": 1.0774,
      "step": 2580,
      "task_loss": 1.2594256401062012
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0052337646484375,
      "epoch": 0.94,
      "learning_rate": 6.413698743042548e-05,
      "loss": 0.8954,
      "step": 2590,
      "task_loss": 0.7200769186019897
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8628212213516235,
      "epoch": 0.94,
      "learning_rate": 6.409287862580469e-05,
      "loss": 0.9528,
      "step": 2600,
      "task_loss": 0.936751127243042
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1267863512039185,
      "epoch": 0.94,
      "learning_rate": 6.404861980851874e-05,
      "loss": 1.0709,
      "step": 2610,
      "task_loss": 0.9654757976531982
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2044198513031006,
      "epoch": 0.95,
      "learning_rate": 6.400421120678095e-05,
      "loss": 1.0257,
      "step": 2620,
      "task_loss": 1.0389150381088257
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0975178480148315,
      "epoch": 0.95,
      "learning_rate": 6.395965304957704e-05,
      "loss": 1.0731,
      "step": 2630,
      "task_loss": 1.0233360528945923
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9345778226852417,
      "epoch": 0.95,
      "learning_rate": 6.391494556666383e-05,
      "loss": 1.0423,
      "step": 2640,
      "task_loss": 0.7960381507873535
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6801820397377014,
      "epoch": 0.96,
      "learning_rate": 6.387008898856813e-05,
      "loss": 0.9883,
      "step": 2650,
      "task_loss": 0.6798236966133118
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0460097789764404,
      "epoch": 0.96,
      "learning_rate": 6.382508354658555e-05,
      "loss": 1.0479,
      "step": 2660,
      "task_loss": 0.7182983160018921
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1210675239562988,
      "epoch": 0.96,
      "learning_rate": 6.377992947277927e-05,
      "loss": 1.037,
      "step": 2670,
      "task_loss": 0.8833307027816772
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.166975498199463,
      "epoch": 0.97,
      "learning_rate": 6.373462699997891e-05,
      "loss": 1.0224,
      "step": 2680,
      "task_loss": 0.8321333527565002
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.843514621257782,
      "epoch": 0.97,
      "learning_rate": 6.368917636177923e-05,
      "loss": 0.9786,
      "step": 2690,
      "task_loss": 0.6727123260498047
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0732271671295166,
      "epoch": 0.98,
      "learning_rate": 6.3643577792539e-05,
      "loss": 1.0344,
      "step": 2700,
      "task_loss": 1.1724574565887451
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.704067051410675,
      "epoch": 0.98,
      "learning_rate": 6.35978315273798e-05,
      "loss": 0.9464,
      "step": 2710,
      "task_loss": 0.4790229797363281
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8885566592216492,
      "epoch": 0.98,
      "learning_rate": 6.355193780218473e-05,
      "loss": 1.0417,
      "step": 2720,
      "task_loss": 1.1946167945861816
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6279390454292297,
      "epoch": 0.99,
      "learning_rate": 6.350589685359728e-05,
      "loss": 0.9042,
      "step": 2730,
      "task_loss": 0.38034892082214355
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8930590748786926,
      "epoch": 0.99,
      "learning_rate": 6.345970891902006e-05,
      "loss": 0.8633,
      "step": 2740,
      "task_loss": 0.6923383474349976
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.864410936832428,
      "epoch": 0.99,
      "learning_rate": 6.341337423661359e-05,
      "loss": 0.9801,
      "step": 2750,
      "task_loss": 0.8051791191101074
    },
    {
      "epoch": 0.99,
      "eval_exact_match": 80.6244087038789,
      "eval_f1": 88.24949743347717,
      "step": 2750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5681467652320862,
      "epoch": 1.0,
      "learning_rate": 6.33668930452951e-05,
      "loss": 1.0045,
      "step": 2760,
      "task_loss": 0.5351274013519287
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7320226430892944,
      "epoch": 1.0,
      "learning_rate": 6.33202655847372e-05,
      "loss": 0.8719,
      "step": 2770,
      "task_loss": 0.5473893880844116
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.155122995376587,
      "epoch": 1.0,
      "learning_rate": 6.327349209536678e-05,
      "loss": 0.9349,
      "step": 2780,
      "task_loss": 1.0080279111862183
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2032737731933594,
      "epoch": 1.01,
      "learning_rate": 6.322657281836369e-05,
      "loss": 0.7598,
      "step": 2790,
      "task_loss": 0.9358059167861938
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6500468850135803,
      "epoch": 1.01,
      "learning_rate": 6.317950799565947e-05,
      "loss": 0.7994,
      "step": 2800,
      "task_loss": 0.6531230807304382
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.904877781867981,
      "epoch": 1.02,
      "learning_rate": 6.31322978699362e-05,
      "loss": 0.8276,
      "step": 2810,
      "task_loss": 0.6158158779144287
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7756827473640442,
      "epoch": 1.02,
      "learning_rate": 6.308494268462514e-05,
      "loss": 0.7618,
      "step": 2820,
      "task_loss": 0.48452329635620117
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9801332950592041,
      "epoch": 1.02,
      "learning_rate": 6.303744268390556e-05,
      "loss": 0.7582,
      "step": 2830,
      "task_loss": 0.8897899389266968
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8067210912704468,
      "epoch": 1.03,
      "learning_rate": 6.298979811270345e-05,
      "loss": 0.79,
      "step": 2840,
      "task_loss": 0.7257287502288818
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7176787853240967,
      "epoch": 1.03,
      "learning_rate": 6.294200921669022e-05,
      "loss": 0.7634,
      "step": 2850,
      "task_loss": 0.6690226197242737
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6467972993850708,
      "epoch": 1.03,
      "learning_rate": 6.28940762422815e-05,
      "loss": 0.8435,
      "step": 2860,
      "task_loss": 0.4089285135269165
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6949372291564941,
      "epoch": 1.04,
      "learning_rate": 6.284599943663583e-05,
      "loss": 0.7328,
      "step": 2870,
      "task_loss": 0.5266913175582886
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6310738325119019,
      "epoch": 1.04,
      "learning_rate": 6.279777904765335e-05,
      "loss": 0.6816,
      "step": 2880,
      "task_loss": 0.659895658493042
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7482784986495972,
      "epoch": 1.04,
      "learning_rate": 6.274941532397464e-05,
      "loss": 0.7569,
      "step": 2890,
      "task_loss": 0.8065102100372314
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9670886993408203,
      "epoch": 1.05,
      "learning_rate": 6.27009085149793e-05,
      "loss": 0.7821,
      "step": 2900,
      "task_loss": 0.7383887767791748
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6161856651306152,
      "epoch": 1.05,
      "learning_rate": 6.265225887078478e-05,
      "loss": 0.7253,
      "step": 2910,
      "task_loss": 0.9445866346359253
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8717264533042908,
      "epoch": 1.06,
      "learning_rate": 6.260346664224497e-05,
      "loss": 0.7941,
      "step": 2920,
      "task_loss": 1.126537799835205
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6857512593269348,
      "epoch": 1.06,
      "learning_rate": 6.255453208094901e-05,
      "loss": 0.77,
      "step": 2930,
      "task_loss": 0.8074723482131958
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7998137474060059,
      "epoch": 1.06,
      "learning_rate": 6.250545543921997e-05,
      "loss": 0.85,
      "step": 2940,
      "task_loss": 0.6924498677253723
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1140186786651611,
      "epoch": 1.07,
      "learning_rate": 6.245623697011352e-05,
      "loss": 0.8453,
      "step": 2950,
      "task_loss": 1.2529780864715576
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.808302640914917,
      "epoch": 1.07,
      "learning_rate": 6.240687692741663e-05,
      "loss": 0.8449,
      "step": 2960,
      "task_loss": 0.704319953918457
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6823419332504272,
      "epoch": 1.07,
      "learning_rate": 6.235737556564626e-05,
      "loss": 0.698,
      "step": 2970,
      "task_loss": 0.5391907095909119
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6212811470031738,
      "epoch": 1.08,
      "learning_rate": 6.230773314004811e-05,
      "loss": 0.773,
      "step": 2980,
      "task_loss": 0.802199125289917
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6568988561630249,
      "epoch": 1.08,
      "learning_rate": 6.225794990659519e-05,
      "loss": 0.7828,
      "step": 2990,
      "task_loss": 0.8078853487968445
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8658432364463806,
      "epoch": 1.08,
      "learning_rate": 6.220802612198661e-05,
      "loss": 0.7315,
      "step": 3000,
      "task_loss": 0.9337005615234375
    },
    {
      "epoch": 1.08,
      "eval_exact_match": 80.64333017975402,
      "eval_f1": 88.53171644067697,
      "step": 3000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9142760038375854,
      "epoch": 1.09,
      "learning_rate": 6.215796204364618e-05,
      "loss": 0.7422,
      "step": 3010,
      "task_loss": 0.8946501016616821
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0131399631500244,
      "epoch": 1.09,
      "learning_rate": 6.210775792972113e-05,
      "loss": 0.829,
      "step": 3020,
      "task_loss": 0.694104790687561
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.858983039855957,
      "epoch": 1.1,
      "learning_rate": 6.205741403908076e-05,
      "loss": 0.7699,
      "step": 3030,
      "task_loss": 0.8205258846282959
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8396916389465332,
      "epoch": 1.1,
      "learning_rate": 6.20069306313151e-05,
      "loss": 0.8486,
      "step": 3040,
      "task_loss": 0.6872456073760986
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7519895434379578,
      "epoch": 1.1,
      "learning_rate": 6.195630796673355e-05,
      "loss": 0.8115,
      "step": 3050,
      "task_loss": 1.1500433683395386
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8241223096847534,
      "epoch": 1.11,
      "learning_rate": 6.19055463063636e-05,
      "loss": 0.7984,
      "step": 3060,
      "task_loss": 0.8155335187911987
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6022705435752869,
      "epoch": 1.11,
      "learning_rate": 6.185464591194946e-05,
      "loss": 0.7076,
      "step": 3070,
      "task_loss": 0.594499945640564
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.895301103591919,
      "epoch": 1.11,
      "learning_rate": 6.180360704595069e-05,
      "loss": 0.7757,
      "step": 3080,
      "task_loss": 1.0475142002105713
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9936012029647827,
      "epoch": 1.12,
      "learning_rate": 6.17524299715408e-05,
      "loss": 0.7569,
      "step": 3090,
      "task_loss": 0.5248420238494873
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6736662983894348,
      "epoch": 1.12,
      "learning_rate": 6.170111495260603e-05,
      "loss": 0.779,
      "step": 3100,
      "task_loss": 0.8802880644798279
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9580499529838562,
      "epoch": 1.12,
      "learning_rate": 6.164966225374386e-05,
      "loss": 0.7582,
      "step": 3110,
      "task_loss": 1.5431132316589355
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7854373455047607,
      "epoch": 1.13,
      "learning_rate": 6.15980721402617e-05,
      "loss": 0.819,
      "step": 3120,
      "task_loss": 0.8485944271087646
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1915910243988037,
      "epoch": 1.13,
      "learning_rate": 6.154634487817555e-05,
      "loss": 0.7818,
      "step": 3130,
      "task_loss": 1.0657076835632324
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5495825409889221,
      "epoch": 1.13,
      "learning_rate": 6.149448073420854e-05,
      "loss": 0.7273,
      "step": 3140,
      "task_loss": 0.436737596988678
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0557966232299805,
      "epoch": 1.14,
      "learning_rate": 6.144247997578965e-05,
      "loss": 0.8162,
      "step": 3150,
      "task_loss": 1.5712409019470215
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5940508246421814,
      "epoch": 1.14,
      "learning_rate": 6.139034287105228e-05,
      "loss": 0.7228,
      "step": 3160,
      "task_loss": 0.7547852993011475
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9270653128623962,
      "epoch": 1.15,
      "learning_rate": 6.133806968883286e-05,
      "loss": 0.7985,
      "step": 3170,
      "task_loss": 0.7847113609313965
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6993736624717712,
      "epoch": 1.15,
      "learning_rate": 6.12856606986695e-05,
      "loss": 0.7397,
      "step": 3180,
      "task_loss": 0.8840616941452026
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7587177157402039,
      "epoch": 1.15,
      "learning_rate": 6.123311617080057e-05,
      "loss": 0.7178,
      "step": 3190,
      "task_loss": 0.6237931251525879
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0422003269195557,
      "epoch": 1.16,
      "learning_rate": 6.118043637616336e-05,
      "loss": 0.822,
      "step": 3200,
      "task_loss": 1.0040562152862549
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7708929777145386,
      "epoch": 1.16,
      "learning_rate": 6.112762158639257e-05,
      "loss": 0.8651,
      "step": 3210,
      "task_loss": 0.8678411245346069
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7193692922592163,
      "epoch": 1.16,
      "learning_rate": 6.107467207381903e-05,
      "loss": 0.7916,
      "step": 3220,
      "task_loss": 0.6456923484802246
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6213294267654419,
      "epoch": 1.17,
      "learning_rate": 6.1021588111468214e-05,
      "loss": 0.7506,
      "step": 3230,
      "task_loss": 0.430205762386322
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.717927098274231,
      "epoch": 1.17,
      "learning_rate": 6.0968369973058906e-05,
      "loss": 0.6777,
      "step": 3240,
      "task_loss": 0.7677985429763794
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5344014167785645,
      "epoch": 1.17,
      "learning_rate": 6.0915017933001704e-05,
      "loss": 0.6799,
      "step": 3250,
      "task_loss": 0.6910272836685181
    },
    {
      "epoch": 1.17,
      "eval_exact_match": 80.70955534531693,
      "eval_f1": 88.61805230583994,
      "step": 3250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7944083213806152,
      "epoch": 1.18,
      "learning_rate": 6.086153226639769e-05,
      "loss": 0.7754,
      "step": 3260,
      "task_loss": 0.8809820413589478
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7685368657112122,
      "epoch": 1.18,
      "learning_rate": 6.080791324903693e-05,
      "loss": 0.8346,
      "step": 3270,
      "task_loss": 1.361288070678711
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8159446716308594,
      "epoch": 1.19,
      "learning_rate": 6.075416115739711e-05,
      "loss": 0.783,
      "step": 3280,
      "task_loss": 0.8244059085845947
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9105017185211182,
      "epoch": 1.19,
      "learning_rate": 6.070027626864208e-05,
      "loss": 0.7819,
      "step": 3290,
      "task_loss": 1.182206630706787
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6835358142852783,
      "epoch": 1.19,
      "learning_rate": 6.064625886062046e-05,
      "loss": 0.7313,
      "step": 3300,
      "task_loss": 0.7755618095397949
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7198119163513184,
      "epoch": 1.2,
      "learning_rate": 6.059210921186417e-05,
      "loss": 0.7644,
      "step": 3310,
      "task_loss": 0.6677089929580688
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1201332807540894,
      "epoch": 1.2,
      "learning_rate": 6.053782760158701e-05,
      "loss": 0.9287,
      "step": 3320,
      "task_loss": 1.3147826194763184
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7347714900970459,
      "epoch": 1.2,
      "learning_rate": 6.04834143096832e-05,
      "loss": 0.7157,
      "step": 3330,
      "task_loss": 1.1088364124298096
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5882917642593384,
      "epoch": 1.21,
      "learning_rate": 6.0428869616725985e-05,
      "loss": 0.7162,
      "step": 3340,
      "task_loss": 0.697338342666626
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6236642003059387,
      "epoch": 1.21,
      "learning_rate": 6.0374193803966136e-05,
      "loss": 0.7029,
      "step": 3350,
      "task_loss": 1.0311845541000366
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6423226594924927,
      "epoch": 1.21,
      "learning_rate": 6.0319387153330536e-05,
      "loss": 0.7465,
      "step": 3360,
      "task_loss": 0.8737854957580566
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9397047758102417,
      "epoch": 1.22,
      "learning_rate": 6.0264449947420696e-05,
      "loss": 0.8276,
      "step": 3370,
      "task_loss": 0.5499547719955444
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9513425230979919,
      "epoch": 1.22,
      "learning_rate": 6.0209382469511336e-05,
      "loss": 0.7756,
      "step": 3380,
      "task_loss": 1.0776315927505493
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.955211877822876,
      "epoch": 1.23,
      "learning_rate": 6.015418500354887e-05,
      "loss": 0.8155,
      "step": 3390,
      "task_loss": 0.8473688364028931
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7617285251617432,
      "epoch": 1.23,
      "learning_rate": 6.009885783415001e-05,
      "loss": 0.8277,
      "step": 3400,
      "task_loss": 0.9811973571777344
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8025141954421997,
      "epoch": 1.23,
      "learning_rate": 6.0043401246600244e-05,
      "loss": 0.8092,
      "step": 3410,
      "task_loss": 0.78231281042099
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.732769787311554,
      "epoch": 1.24,
      "learning_rate": 5.998781552685238e-05,
      "loss": 0.7231,
      "step": 3420,
      "task_loss": 0.6331707239151001
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7150098085403442,
      "epoch": 1.24,
      "learning_rate": 5.9932100961525085e-05,
      "loss": 0.7696,
      "step": 3430,
      "task_loss": 0.65977543592453
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.913851261138916,
      "epoch": 1.24,
      "learning_rate": 5.9876257837901394e-05,
      "loss": 0.8205,
      "step": 3440,
      "task_loss": 0.8993183970451355
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5872275829315186,
      "epoch": 1.25,
      "learning_rate": 5.982028644392723e-05,
      "loss": 0.7224,
      "step": 3450,
      "task_loss": 0.657208263874054
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7204796671867371,
      "epoch": 1.25,
      "learning_rate": 5.976418706820992e-05,
      "loss": 0.7553,
      "step": 3460,
      "task_loss": 0.8270089626312256
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8491474390029907,
      "epoch": 1.25,
      "learning_rate": 5.9707960000016706e-05,
      "loss": 0.8146,
      "step": 3470,
      "task_loss": 0.7125909328460693
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.630764365196228,
      "epoch": 1.26,
      "learning_rate": 5.9651605529273264e-05,
      "loss": 0.7561,
      "step": 3480,
      "task_loss": 0.798701286315918
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6272002458572388,
      "epoch": 1.26,
      "learning_rate": 5.9595123946562196e-05,
      "loss": 0.7211,
      "step": 3490,
      "task_loss": 0.7057316899299622
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7990823984146118,
      "epoch": 1.26,
      "learning_rate": 5.9538515543121524e-05,
      "loss": 0.8313,
      "step": 3500,
      "task_loss": 0.7466528415679932
    },
    {
      "epoch": 1.26,
      "eval_exact_match": 80.89877010406812,
      "eval_f1": 88.60748473078176,
      "step": 3500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5751552581787109,
      "epoch": 1.27,
      "learning_rate": 5.948178061084323e-05,
      "loss": 0.7492,
      "step": 3510,
      "task_loss": 0.5113732814788818
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8527777194976807,
      "epoch": 1.27,
      "learning_rate": 5.9424919442271693e-05,
      "loss": 0.6718,
      "step": 3520,
      "task_loss": 0.9590537548065186
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6820343732833862,
      "epoch": 1.28,
      "learning_rate": 5.9367932330602216e-05,
      "loss": 0.7317,
      "step": 3530,
      "task_loss": 0.775328516960144
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7279667854309082,
      "epoch": 1.28,
      "learning_rate": 5.931081956967951e-05,
      "loss": 0.8587,
      "step": 3540,
      "task_loss": 0.5757768750190735
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7633707523345947,
      "epoch": 1.28,
      "learning_rate": 5.925358145399619e-05,
      "loss": 0.7386,
      "step": 3550,
      "task_loss": 0.6215412616729736
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9111154079437256,
      "epoch": 1.29,
      "learning_rate": 5.9196218278691204e-05,
      "loss": 0.8404,
      "step": 3560,
      "task_loss": 0.9098265171051025
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.968406081199646,
      "epoch": 1.29,
      "learning_rate": 5.913873033954839e-05,
      "loss": 0.8175,
      "step": 3570,
      "task_loss": 0.8037402629852295
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5799151062965393,
      "epoch": 1.29,
      "learning_rate": 5.908111793299489e-05,
      "loss": 0.7005,
      "step": 3580,
      "task_loss": 0.7509677410125732
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7919946908950806,
      "epoch": 1.3,
      "learning_rate": 5.902338135609965e-05,
      "loss": 0.784,
      "step": 3590,
      "task_loss": 0.75666344165802
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5682597160339355,
      "epoch": 1.3,
      "learning_rate": 5.896552090657185e-05,
      "loss": 0.7353,
      "step": 3600,
      "task_loss": 0.6385715007781982
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8308883905410767,
      "epoch": 1.3,
      "learning_rate": 5.890753688275946e-05,
      "loss": 0.8532,
      "step": 3610,
      "task_loss": 0.8819625377655029
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8280819058418274,
      "epoch": 1.31,
      "learning_rate": 5.884942958364759e-05,
      "loss": 0.8406,
      "step": 3620,
      "task_loss": 0.8358502984046936
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8390671014785767,
      "epoch": 1.31,
      "learning_rate": 5.8791199308857e-05,
      "loss": 0.7029,
      "step": 3630,
      "task_loss": 1.1161457300186157
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6460493803024292,
      "epoch": 1.32,
      "learning_rate": 5.873284635864259e-05,
      "loss": 0.9247,
      "step": 3640,
      "task_loss": 0.6301670074462891
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6909412741661072,
      "epoch": 1.32,
      "learning_rate": 5.867437103389178e-05,
      "loss": 0.774,
      "step": 3650,
      "task_loss": 0.8153108954429626
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6475903987884521,
      "epoch": 1.32,
      "learning_rate": 5.861577363612302e-05,
      "loss": 0.7276,
      "step": 3660,
      "task_loss": 0.5706870555877686
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4457024335861206,
      "epoch": 1.33,
      "learning_rate": 5.855705446748418e-05,
      "loss": 0.7223,
      "step": 3670,
      "task_loss": 0.7826858162879944
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6906821131706238,
      "epoch": 1.33,
      "learning_rate": 5.849821383075105e-05,
      "loss": 0.831,
      "step": 3680,
      "task_loss": 0.6492674946784973
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7505589723587036,
      "epoch": 1.33,
      "learning_rate": 5.8439252029325734e-05,
      "loss": 0.8624,
      "step": 3690,
      "task_loss": 0.8882350921630859
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9934232831001282,
      "epoch": 1.34,
      "learning_rate": 5.838016936723511e-05,
      "loss": 0.8199,
      "step": 3700,
      "task_loss": 1.0231308937072754
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5626651048660278,
      "epoch": 1.34,
      "learning_rate": 5.832096614912926e-05,
      "loss": 0.7692,
      "step": 3710,
      "task_loss": 0.5882710814476013
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5242562294006348,
      "epoch": 1.34,
      "learning_rate": 5.826164268027987e-05,
      "loss": 0.7655,
      "step": 3720,
      "task_loss": 0.5058928728103638
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6647918820381165,
      "epoch": 1.35,
      "learning_rate": 5.8202199266578704e-05,
      "loss": 0.7998,
      "step": 3730,
      "task_loss": 1.0364404916763306
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5112893581390381,
      "epoch": 1.35,
      "learning_rate": 5.8142636214535976e-05,
      "loss": 0.7007,
      "step": 3740,
      "task_loss": 0.8115373253822327
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5241594314575195,
      "epoch": 1.36,
      "learning_rate": 5.808295383127883e-05,
      "loss": 0.769,
      "step": 3750,
      "task_loss": 0.4635359048843384
    },
    {
      "epoch": 1.36,
      "eval_exact_match": 80.60548722800378,
      "eval_f1": 88.48461872007243,
      "step": 3750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7614384293556213,
      "epoch": 1.36,
      "learning_rate": 5.80231524245497e-05,
      "loss": 0.7911,
      "step": 3760,
      "task_loss": 0.5848983526229858
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7333616614341736,
      "epoch": 1.36,
      "learning_rate": 5.796323230270474e-05,
      "loss": 0.8573,
      "step": 3770,
      "task_loss": 0.6257777214050293
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7030531764030457,
      "epoch": 1.37,
      "learning_rate": 5.790319377471224e-05,
      "loss": 0.7254,
      "step": 3780,
      "task_loss": 0.4509909152984619
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8695291876792908,
      "epoch": 1.37,
      "learning_rate": 5.7843037150151074e-05,
      "loss": 0.7335,
      "step": 3790,
      "task_loss": 0.8156959414482117
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6627474427223206,
      "epoch": 1.37,
      "learning_rate": 5.778276273920899e-05,
      "loss": 0.7666,
      "step": 3800,
      "task_loss": 0.5233876705169678
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0300884246826172,
      "epoch": 1.38,
      "learning_rate": 5.772237085268114e-05,
      "loss": 0.7196,
      "step": 3810,
      "task_loss": 0.7798508405685425
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5791784524917603,
      "epoch": 1.38,
      "learning_rate": 5.766186180196839e-05,
      "loss": 0.7791,
      "step": 3820,
      "task_loss": 0.6178016066551208
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.891481876373291,
      "epoch": 1.38,
      "learning_rate": 5.7601235899075745e-05,
      "loss": 0.7834,
      "step": 3830,
      "task_loss": 0.8527568578720093
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6954624652862549,
      "epoch": 1.39,
      "learning_rate": 5.7540493456610766e-05,
      "loss": 0.7348,
      "step": 3840,
      "task_loss": 0.8668036460876465
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5850803256034851,
      "epoch": 1.39,
      "learning_rate": 5.747963478778191e-05,
      "loss": 0.7299,
      "step": 3850,
      "task_loss": 0.4395766258239746
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8359133005142212,
      "epoch": 1.4,
      "learning_rate": 5.741866020639692e-05,
      "loss": 0.7206,
      "step": 3860,
      "task_loss": 1.7376631498336792
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5748704671859741,
      "epoch": 1.4,
      "learning_rate": 5.735757002686126e-05,
      "loss": 0.7436,
      "step": 3870,
      "task_loss": 0.460451602935791
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7052741050720215,
      "epoch": 1.4,
      "learning_rate": 5.7296364564176416e-05,
      "loss": 0.7916,
      "step": 3880,
      "task_loss": 0.82174152135849
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0083122253417969,
      "epoch": 1.41,
      "learning_rate": 5.7235044133938356e-05,
      "loss": 0.8267,
      "step": 3890,
      "task_loss": 0.8673921823501587
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5959757566452026,
      "epoch": 1.41,
      "learning_rate": 5.717360905233584e-05,
      "loss": 0.7126,
      "step": 3900,
      "task_loss": 0.7061013579368591
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6814622282981873,
      "epoch": 1.41,
      "learning_rate": 5.711205963614879e-05,
      "loss": 0.8038,
      "step": 3910,
      "task_loss": 0.4971301853656769
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2168264389038086,
      "epoch": 1.42,
      "learning_rate": 5.70503962027467e-05,
      "loss": 0.773,
      "step": 3920,
      "task_loss": 1.144862413406372
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.63560950756073,
      "epoch": 1.42,
      "learning_rate": 5.698861907008698e-05,
      "loss": 0.797,
      "step": 3930,
      "task_loss": 0.6747443079948425
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7305340766906738,
      "epoch": 1.42,
      "learning_rate": 5.692672855671328e-05,
      "loss": 0.7419,
      "step": 3940,
      "task_loss": 0.39194267988204956
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7829553484916687,
      "epoch": 1.43,
      "learning_rate": 5.6864724981753914e-05,
      "loss": 0.7298,
      "step": 3950,
      "task_loss": 0.7858966588973999
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5866974592208862,
      "epoch": 1.43,
      "learning_rate": 5.680260866492015e-05,
      "loss": 0.7619,
      "step": 3960,
      "task_loss": 0.8394837379455566
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8175315260887146,
      "epoch": 1.43,
      "learning_rate": 5.674037992650463e-05,
      "loss": 0.7816,
      "step": 3970,
      "task_loss": 0.7513538599014282
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9726064205169678,
      "epoch": 1.44,
      "learning_rate": 5.667803908737963e-05,
      "loss": 0.689,
      "step": 3980,
      "task_loss": 0.914472758769989
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5984624624252319,
      "epoch": 1.44,
      "learning_rate": 5.6615586468995484e-05,
      "loss": 0.7236,
      "step": 3990,
      "task_loss": 0.4769936203956604
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6598918437957764,
      "epoch": 1.45,
      "learning_rate": 5.65530223933789e-05,
      "loss": 0.716,
      "step": 4000,
      "task_loss": 1.0444022417068481
    },
    {
      "epoch": 1.45,
      "eval_exact_match": 81.05014191106906,
      "eval_f1": 88.78854316118291,
      "step": 4000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4541385769844055,
      "epoch": 1.45,
      "learning_rate": 5.6490347183131293e-05,
      "loss": 0.7595,
      "step": 4010,
      "task_loss": 0.45822444558143616
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8845678567886353,
      "epoch": 1.45,
      "learning_rate": 5.6427561161427114e-05,
      "loss": 0.7587,
      "step": 4020,
      "task_loss": 1.0624125003814697
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6516546010971069,
      "epoch": 1.46,
      "learning_rate": 5.6364664652012204e-05,
      "loss": 0.692,
      "step": 4030,
      "task_loss": 0.6902788877487183
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8230467438697815,
      "epoch": 1.46,
      "learning_rate": 5.630165797920211e-05,
      "loss": 0.7548,
      "step": 4040,
      "task_loss": 1.0239936113357544
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.565695583820343,
      "epoch": 1.46,
      "learning_rate": 5.623854146788044e-05,
      "loss": 0.7528,
      "step": 4050,
      "task_loss": 0.9347711801528931
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.48703664541244507,
      "epoch": 1.47,
      "learning_rate": 5.6175315443497146e-05,
      "loss": 0.795,
      "step": 4060,
      "task_loss": 0.5277222394943237
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.571143388748169,
      "epoch": 1.47,
      "learning_rate": 5.611198023206687e-05,
      "loss": 0.8009,
      "step": 4070,
      "task_loss": 0.8106241822242737
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5028181672096252,
      "epoch": 1.47,
      "learning_rate": 5.6048536160167244e-05,
      "loss": 0.7494,
      "step": 4080,
      "task_loss": 0.39358311891555786
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8799566030502319,
      "epoch": 1.48,
      "learning_rate": 5.598498355493726e-05,
      "loss": 0.7456,
      "step": 4090,
      "task_loss": 0.6059544682502747
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8514795303344727,
      "epoch": 1.48,
      "learning_rate": 5.592132274407551e-05,
      "loss": 0.7538,
      "step": 4100,
      "task_loss": 0.9229733943939209
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.305832862854004,
      "epoch": 1.49,
      "learning_rate": 5.5857554055838546e-05,
      "loss": 0.7833,
      "step": 4110,
      "task_loss": 1.1757216453552246
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7599394917488098,
      "epoch": 1.49,
      "learning_rate": 5.5793677819039163e-05,
      "loss": 0.7546,
      "step": 4120,
      "task_loss": 0.8550896644592285
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9693371057510376,
      "epoch": 1.49,
      "learning_rate": 5.572969436304471e-05,
      "loss": 0.8094,
      "step": 4130,
      "task_loss": 1.0503398180007935
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6661901473999023,
      "epoch": 1.5,
      "learning_rate": 5.566560401777542e-05,
      "loss": 0.7682,
      "step": 4140,
      "task_loss": 0.47326183319091797
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5503145456314087,
      "epoch": 1.5,
      "learning_rate": 5.560140711370266e-05,
      "loss": 0.6929,
      "step": 4150,
      "task_loss": 0.5086473226547241
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6122060418128967,
      "epoch": 1.5,
      "learning_rate": 5.553710398184724e-05,
      "loss": 0.7253,
      "step": 4160,
      "task_loss": 0.7502968311309814
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8065118789672852,
      "epoch": 1.51,
      "learning_rate": 5.547269495377774e-05,
      "loss": 0.7417,
      "step": 4170,
      "task_loss": 0.8612374663352966
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8094028234481812,
      "epoch": 1.51,
      "learning_rate": 5.5408180361608775e-05,
      "loss": 0.7633,
      "step": 4180,
      "task_loss": 0.8984237909317017
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7946018576622009,
      "epoch": 1.51,
      "learning_rate": 5.5343560537999265e-05,
      "loss": 0.7919,
      "step": 4190,
      "task_loss": 0.8055480718612671
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9102916121482849,
      "epoch": 1.52,
      "learning_rate": 5.527883581615077e-05,
      "loss": 0.7809,
      "step": 4200,
      "task_loss": 0.8656786680221558
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7173982858657837,
      "epoch": 1.52,
      "learning_rate": 5.521400652980568e-05,
      "loss": 0.7714,
      "step": 4210,
      "task_loss": 0.7693835496902466
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8590863943099976,
      "epoch": 1.53,
      "learning_rate": 5.514907301324564e-05,
      "loss": 0.7208,
      "step": 4220,
      "task_loss": 0.7576629519462585
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7024766206741333,
      "epoch": 1.53,
      "learning_rate": 5.5084035601289674e-05,
      "loss": 0.7241,
      "step": 4230,
      "task_loss": 0.5862898230552673
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6137065291404724,
      "epoch": 1.53,
      "learning_rate": 5.501889462929254e-05,
      "loss": 0.659,
      "step": 4240,
      "task_loss": 0.6031349897384644
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.734519362449646,
      "epoch": 1.54,
      "learning_rate": 5.4953650433143015e-05,
      "loss": 0.777,
      "step": 4250,
      "task_loss": 0.5666317343711853
    },
    {
      "epoch": 1.54,
      "eval_exact_match": 80.90823084200568,
      "eval_f1": 88.62694215371808,
      "step": 4250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8583282232284546,
      "epoch": 1.54,
      "learning_rate": 5.4888303349262105e-05,
      "loss": 0.7682,
      "step": 4260,
      "task_loss": 0.5990405082702637
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7635049223899841,
      "epoch": 1.54,
      "learning_rate": 5.4822853714601336e-05,
      "loss": 0.8497,
      "step": 4270,
      "task_loss": 0.9380366206169128
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5275091528892517,
      "epoch": 1.55,
      "learning_rate": 5.475730186664106e-05,
      "loss": 0.7873,
      "step": 4280,
      "task_loss": 0.5869154334068298
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.918228030204773,
      "epoch": 1.55,
      "learning_rate": 5.4691648143388625e-05,
      "loss": 0.7501,
      "step": 4290,
      "task_loss": 0.8605305552482605
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9519420862197876,
      "epoch": 1.55,
      "learning_rate": 5.462589288337671e-05,
      "loss": 0.8131,
      "step": 4300,
      "task_loss": 0.8920116424560547
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5668010115623474,
      "epoch": 1.56,
      "learning_rate": 5.456003642566156e-05,
      "loss": 0.7325,
      "step": 4310,
      "task_loss": 0.7306046485900879
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3651421070098877,
      "epoch": 1.56,
      "learning_rate": 5.44940791098212e-05,
      "loss": 0.7889,
      "step": 4320,
      "task_loss": 0.42743948101997375
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6651080250740051,
      "epoch": 1.56,
      "learning_rate": 5.442802127595373e-05,
      "loss": 0.7759,
      "step": 4330,
      "task_loss": 1.0754146575927734
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.752009391784668,
      "epoch": 1.57,
      "learning_rate": 5.4361863264675574e-05,
      "loss": 0.7438,
      "step": 4340,
      "task_loss": 0.8343137502670288
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5974752902984619,
      "epoch": 1.57,
      "learning_rate": 5.4295605417119656e-05,
      "loss": 0.8421,
      "step": 4350,
      "task_loss": 0.6965165138244629
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5219227075576782,
      "epoch": 1.58,
      "learning_rate": 5.422924807493375e-05,
      "loss": 0.7804,
      "step": 4360,
      "task_loss": 1.1046372652053833
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7944738864898682,
      "epoch": 1.58,
      "learning_rate": 5.41627915802786e-05,
      "loss": 0.7957,
      "step": 4370,
      "task_loss": 1.3685147762298584
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7721883058547974,
      "epoch": 1.58,
      "learning_rate": 5.409623627582625e-05,
      "loss": 0.7248,
      "step": 4380,
      "task_loss": 0.6489783525466919
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7032918930053711,
      "epoch": 1.59,
      "learning_rate": 5.402958250475823e-05,
      "loss": 0.7312,
      "step": 4390,
      "task_loss": 0.7790108919143677
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6564103364944458,
      "epoch": 1.59,
      "learning_rate": 5.396283061076379e-05,
      "loss": 0.7982,
      "step": 4400,
      "task_loss": 0.7662302255630493
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9732493162155151,
      "epoch": 1.59,
      "learning_rate": 5.389598093803816e-05,
      "loss": 0.8468,
      "step": 4410,
      "task_loss": 0.800963282585144
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5395100116729736,
      "epoch": 1.6,
      "learning_rate": 5.3829033831280716e-05,
      "loss": 0.7181,
      "step": 4420,
      "task_loss": 0.6361441612243652
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7411954998970032,
      "epoch": 1.6,
      "learning_rate": 5.376198963569325e-05,
      "loss": 0.7124,
      "step": 4430,
      "task_loss": 0.6355388164520264
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6182745695114136,
      "epoch": 1.6,
      "learning_rate": 5.36948486969782e-05,
      "loss": 0.6804,
      "step": 4440,
      "task_loss": 0.7747174501419067
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.772506058216095,
      "epoch": 1.61,
      "learning_rate": 5.36276113613368e-05,
      "loss": 0.7687,
      "step": 4450,
      "task_loss": 0.6414505839347839
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9038888216018677,
      "epoch": 1.61,
      "learning_rate": 5.356027797546736e-05,
      "loss": 0.7532,
      "step": 4460,
      "task_loss": 0.7040343880653381
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6857190728187561,
      "epoch": 1.62,
      "learning_rate": 5.349284888656348e-05,
      "loss": 0.7633,
      "step": 4470,
      "task_loss": 0.8633918166160583
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7742190957069397,
      "epoch": 1.62,
      "learning_rate": 5.342532444231222e-05,
      "loss": 0.7441,
      "step": 4480,
      "task_loss": 0.7828219532966614
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0203568935394287,
      "epoch": 1.62,
      "learning_rate": 5.33577049908923e-05,
      "loss": 0.7315,
      "step": 4490,
      "task_loss": 0.8948928713798523
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5982668399810791,
      "epoch": 1.63,
      "learning_rate": 5.328999088097236e-05,
      "loss": 0.7708,
      "step": 4500,
      "task_loss": 0.6327080726623535
    },
    {
      "epoch": 1.63,
      "eval_exact_match": 80.85146641438033,
      "eval_f1": 88.73605738995374,
      "step": 4500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8688200116157532,
      "epoch": 1.63,
      "learning_rate": 5.322218246170911e-05,
      "loss": 0.8681,
      "step": 4510,
      "task_loss": 1.0566356182098389
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9127258062362671,
      "epoch": 1.63,
      "learning_rate": 5.315428008274558e-05,
      "loss": 0.8109,
      "step": 4520,
      "task_loss": 0.9632719159126282
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6907234787940979,
      "epoch": 1.64,
      "learning_rate": 5.308628409420925e-05,
      "loss": 0.7828,
      "step": 4530,
      "task_loss": 0.7649120688438416
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6892116069793701,
      "epoch": 1.64,
      "learning_rate": 5.301819484671032e-05,
      "loss": 0.7633,
      "step": 4540,
      "task_loss": 0.5884895324707031
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8199399709701538,
      "epoch": 1.64,
      "learning_rate": 5.295001269133984e-05,
      "loss": 0.729,
      "step": 4550,
      "task_loss": 0.7061965465545654
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1325392723083496,
      "epoch": 1.65,
      "learning_rate": 5.2881737979667926e-05,
      "loss": 0.8629,
      "step": 4560,
      "task_loss": 1.1424285173416138
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5671732425689697,
      "epoch": 1.65,
      "learning_rate": 5.2813371063741945e-05,
      "loss": 0.6932,
      "step": 4570,
      "task_loss": 0.61430424451828
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9136245846748352,
      "epoch": 1.66,
      "learning_rate": 5.2744912296084714e-05,
      "loss": 0.8854,
      "step": 4580,
      "task_loss": 1.0167670249938965
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6893622875213623,
      "epoch": 1.66,
      "learning_rate": 5.2676362029692664e-05,
      "loss": 0.817,
      "step": 4590,
      "task_loss": 0.501244068145752
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6975424289703369,
      "epoch": 1.66,
      "learning_rate": 5.260772061803401e-05,
      "loss": 0.6955,
      "step": 4600,
      "task_loss": 0.8912298679351807
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6169401407241821,
      "epoch": 1.67,
      "learning_rate": 5.2538988415046944e-05,
      "loss": 0.7314,
      "step": 4610,
      "task_loss": 0.5644417405128479
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5297744274139404,
      "epoch": 1.67,
      "learning_rate": 5.247016577513784e-05,
      "loss": 0.7156,
      "step": 4620,
      "task_loss": 0.539046585559845
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6868759393692017,
      "epoch": 1.67,
      "learning_rate": 5.2401253053179365e-05,
      "loss": 0.7758,
      "step": 4630,
      "task_loss": 0.574796736240387
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7551968097686768,
      "epoch": 1.68,
      "learning_rate": 5.2332250604508675e-05,
      "loss": 0.7911,
      "step": 4640,
      "task_loss": 0.7167572379112244
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6903712749481201,
      "epoch": 1.68,
      "learning_rate": 5.226315878492561e-05,
      "loss": 0.7262,
      "step": 4650,
      "task_loss": 0.8787746429443359
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8185775876045227,
      "epoch": 1.68,
      "learning_rate": 5.219397795069082e-05,
      "loss": 0.8305,
      "step": 4660,
      "task_loss": 0.8529611825942993
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6731795072555542,
      "epoch": 1.69,
      "learning_rate": 5.212470845852394e-05,
      "loss": 0.7395,
      "step": 4670,
      "task_loss": 0.8858544230461121
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9355323910713196,
      "epoch": 1.69,
      "learning_rate": 5.205535066560179e-05,
      "loss": 0.767,
      "step": 4680,
      "task_loss": 0.7263451218605042
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7926836609840393,
      "epoch": 1.69,
      "learning_rate": 5.198590492955645e-05,
      "loss": 0.6869,
      "step": 4690,
      "task_loss": 0.5331249833106995
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9048827886581421,
      "epoch": 1.7,
      "learning_rate": 5.1916371608473494e-05,
      "loss": 0.7267,
      "step": 4700,
      "task_loss": 1.2360985279083252
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5285090208053589,
      "epoch": 1.7,
      "learning_rate": 5.1846751060890114e-05,
      "loss": 0.8697,
      "step": 4710,
      "task_loss": 0.521661639213562
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0250056982040405,
      "epoch": 1.71,
      "learning_rate": 5.1777043645793254e-05,
      "loss": 0.6634,
      "step": 4720,
      "task_loss": 0.7965474724769592
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6251044273376465,
      "epoch": 1.71,
      "learning_rate": 5.17072497226178e-05,
      "loss": 0.7716,
      "step": 4730,
      "task_loss": 0.3452892303466797
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8053218126296997,
      "epoch": 1.71,
      "learning_rate": 5.1637369651244666e-05,
      "loss": 0.6825,
      "step": 4740,
      "task_loss": 0.6003007292747498
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5973554253578186,
      "epoch": 1.72,
      "learning_rate": 5.156740379199902e-05,
      "loss": 0.7522,
      "step": 4750,
      "task_loss": 0.8139755725860596
    },
    {
      "epoch": 1.72,
      "eval_exact_match": 81.36234626300852,
      "eval_f1": 88.9705388336572,
      "step": 4750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7501943111419678,
      "epoch": 1.72,
      "learning_rate": 5.149735250564836e-05,
      "loss": 0.7183,
      "step": 4760,
      "task_loss": 0.6855126619338989
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5616030693054199,
      "epoch": 1.72,
      "learning_rate": 5.142721615340066e-05,
      "loss": 0.6749,
      "step": 4770,
      "task_loss": 0.7784498929977417
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0238778591156006,
      "epoch": 1.73,
      "learning_rate": 5.135699509690255e-05,
      "loss": 0.7818,
      "step": 4780,
      "task_loss": 0.8857766389846802
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5700663328170776,
      "epoch": 1.73,
      "learning_rate": 5.128668969823739e-05,
      "loss": 0.6645,
      "step": 4790,
      "task_loss": 0.6657969951629639
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5193660259246826,
      "epoch": 1.73,
      "learning_rate": 5.121630031992348e-05,
      "loss": 0.7029,
      "step": 4800,
      "task_loss": 0.7439495325088501
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5604043006896973,
      "epoch": 1.74,
      "learning_rate": 5.114582732491213e-05,
      "loss": 0.7065,
      "step": 4810,
      "task_loss": 0.7198078036308289
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7965430021286011,
      "epoch": 1.74,
      "learning_rate": 5.107527107658576e-05,
      "loss": 0.7062,
      "step": 4820,
      "task_loss": 0.5083843469619751
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6073727607727051,
      "epoch": 1.75,
      "learning_rate": 5.1004631938756154e-05,
      "loss": 0.7453,
      "step": 4830,
      "task_loss": 0.4514174461364746
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8077729940414429,
      "epoch": 1.75,
      "learning_rate": 5.093391027566244e-05,
      "loss": 0.6939,
      "step": 4840,
      "task_loss": 0.6148980855941772
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.566129207611084,
      "epoch": 1.75,
      "learning_rate": 5.086310645196928e-05,
      "loss": 0.6517,
      "step": 4850,
      "task_loss": 0.5865195393562317
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7375544309616089,
      "epoch": 1.76,
      "learning_rate": 5.079222083276504e-05,
      "loss": 0.7073,
      "step": 4860,
      "task_loss": 0.4971655607223511
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4711465835571289,
      "epoch": 1.76,
      "learning_rate": 5.072125378355978e-05,
      "loss": 0.7163,
      "step": 4870,
      "task_loss": 0.36320602893829346
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.90312659740448,
      "epoch": 1.76,
      "learning_rate": 5.0650205670283475e-05,
      "loss": 0.6884,
      "step": 4880,
      "task_loss": 0.7906779050827026
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7571483850479126,
      "epoch": 1.77,
      "learning_rate": 5.057907685928408e-05,
      "loss": 0.6455,
      "step": 4890,
      "task_loss": 0.610438346862793
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6640625,
      "epoch": 1.77,
      "learning_rate": 5.050786771732567e-05,
      "loss": 0.6774,
      "step": 4900,
      "task_loss": 0.9800470471382141
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9504784345626831,
      "epoch": 1.77,
      "learning_rate": 5.043657861158653e-05,
      "loss": 0.7658,
      "step": 4910,
      "task_loss": 1.0992169380187988
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6472917795181274,
      "epoch": 1.78,
      "learning_rate": 5.036520990965726e-05,
      "loss": 0.7905,
      "step": 4920,
      "task_loss": 0.5526524782180786
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9225701093673706,
      "epoch": 1.78,
      "learning_rate": 5.029376197953888e-05,
      "loss": 0.7936,
      "step": 4930,
      "task_loss": 1.0977343320846558
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8606763482093811,
      "epoch": 1.79,
      "learning_rate": 5.022223518964095e-05,
      "loss": 0.7437,
      "step": 4940,
      "task_loss": 0.95201575756073
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7826287150382996,
      "epoch": 1.79,
      "learning_rate": 5.015062990877964e-05,
      "loss": 0.7283,
      "step": 4950,
      "task_loss": 0.7915252447128296
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8127055168151855,
      "epoch": 1.79,
      "learning_rate": 5.007894650617588e-05,
      "loss": 0.7323,
      "step": 4960,
      "task_loss": 1.1264145374298096
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9009028673171997,
      "epoch": 1.8,
      "learning_rate": 5.0007185351453374e-05,
      "loss": 0.7694,
      "step": 4970,
      "task_loss": 0.7590454816818237
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.088401198387146,
      "epoch": 1.8,
      "learning_rate": 4.9935346814636785e-05,
      "loss": 0.7431,
      "step": 4980,
      "task_loss": 1.1049809455871582
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7166246175765991,
      "epoch": 1.8,
      "learning_rate": 4.9863431266149745e-05,
      "loss": 0.6911,
      "step": 4990,
      "task_loss": 0.7473670840263367
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5155638456344604,
      "epoch": 1.81,
      "learning_rate": 4.979143907681301e-05,
      "loss": 0.7239,
      "step": 5000,
      "task_loss": 0.6666215062141418
    },
    {
      "epoch": 1.81,
      "eval_exact_match": 81.53263954588458,
      "eval_f1": 89.04411412196261,
      "step": 5000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46164464950561523,
      "epoch": 1.81,
      "learning_rate": 4.9719370617842544e-05,
      "loss": 0.6567,
      "step": 5010,
      "task_loss": 0.2584804892539978
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.095167636871338,
      "epoch": 1.81,
      "learning_rate": 4.964722626084752e-05,
      "loss": 0.7566,
      "step": 5020,
      "task_loss": 1.149203896522522
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6007382869720459,
      "epoch": 1.82,
      "learning_rate": 4.9575006377828535e-05,
      "loss": 0.7119,
      "step": 5030,
      "task_loss": 0.8033612966537476
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7142719030380249,
      "epoch": 1.82,
      "learning_rate": 4.9502711341175553e-05,
      "loss": 0.7405,
      "step": 5040,
      "task_loss": 1.000267744064331
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7234902381896973,
      "epoch": 1.83,
      "learning_rate": 4.9430341523666124e-05,
      "loss": 0.7188,
      "step": 5050,
      "task_loss": 0.7795543670654297
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7521551847457886,
      "epoch": 1.83,
      "learning_rate": 4.935789729846335e-05,
      "loss": 0.6675,
      "step": 5060,
      "task_loss": 0.6602778434753418
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.48747414350509644,
      "epoch": 1.83,
      "learning_rate": 4.928537903911402e-05,
      "loss": 0.8129,
      "step": 5070,
      "task_loss": 0.5280594825744629
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6115528345108032,
      "epoch": 1.84,
      "learning_rate": 4.921278711954666e-05,
      "loss": 0.7775,
      "step": 5080,
      "task_loss": 0.9519979953765869
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4798819422721863,
      "epoch": 1.84,
      "learning_rate": 4.9140121914069626e-05,
      "loss": 0.6354,
      "step": 5090,
      "task_loss": 0.4596359133720398
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5512056350708008,
      "epoch": 1.84,
      "learning_rate": 4.9067383797369146e-05,
      "loss": 0.706,
      "step": 5100,
      "task_loss": 0.275431364774704
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6959718465805054,
      "epoch": 1.85,
      "learning_rate": 4.899457314450742e-05,
      "loss": 0.6351,
      "step": 5110,
      "task_loss": 0.7794166803359985
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6612961888313293,
      "epoch": 1.85,
      "learning_rate": 4.8921690330920637e-05,
      "loss": 0.7879,
      "step": 5120,
      "task_loss": 0.6663255095481873
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7638074159622192,
      "epoch": 1.85,
      "learning_rate": 4.884873573241711e-05,
      "loss": 0.7483,
      "step": 5130,
      "task_loss": 0.5485991835594177
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5604839324951172,
      "epoch": 1.86,
      "learning_rate": 4.8775709725175277e-05,
      "loss": 0.6781,
      "step": 5140,
      "task_loss": 0.41115280985832214
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4681781530380249,
      "epoch": 1.86,
      "learning_rate": 4.870261268574178e-05,
      "loss": 0.7632,
      "step": 5150,
      "task_loss": 0.3362809121608734
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6780769228935242,
      "epoch": 1.86,
      "learning_rate": 4.8629444991029544e-05,
      "loss": 0.6914,
      "step": 5160,
      "task_loss": 0.7485142350196838
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5212565660476685,
      "epoch": 1.87,
      "learning_rate": 4.8556207018315793e-05,
      "loss": 0.7206,
      "step": 5170,
      "task_loss": 0.4486411511898041
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5345607995986938,
      "epoch": 1.87,
      "learning_rate": 4.848289914524017e-05,
      "loss": 0.7443,
      "step": 5180,
      "task_loss": 0.8513927459716797
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.781924843788147,
      "epoch": 1.88,
      "learning_rate": 4.840952174980269e-05,
      "loss": 0.8009,
      "step": 5190,
      "task_loss": 0.5516729950904846
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7696242332458496,
      "epoch": 1.88,
      "learning_rate": 4.8336075210361884e-05,
      "loss": 0.7267,
      "step": 5200,
      "task_loss": 1.0042166709899902
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1635730266571045,
      "epoch": 1.88,
      "learning_rate": 4.826255990563282e-05,
      "loss": 0.7788,
      "step": 5210,
      "task_loss": 1.0315825939178467
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9389041662216187,
      "epoch": 1.89,
      "learning_rate": 4.8188976214685125e-05,
      "loss": 0.7449,
      "step": 5220,
      "task_loss": 0.7496635913848877
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4800257086753845,
      "epoch": 1.89,
      "learning_rate": 4.811532451694104e-05,
      "loss": 0.7274,
      "step": 5230,
      "task_loss": 0.6168285608291626
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5006886720657349,
      "epoch": 1.89,
      "learning_rate": 4.8041605192173505e-05,
      "loss": 0.6992,
      "step": 5240,
      "task_loss": 0.5821633338928223
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7733217477798462,
      "epoch": 1.9,
      "learning_rate": 4.796781862050413e-05,
      "loss": 0.7236,
      "step": 5250,
      "task_loss": 0.7929372787475586
    },
    {
      "epoch": 1.9,
      "eval_exact_match": 81.43803216650899,
      "eval_f1": 88.85914255552852,
      "step": 5250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6627964973449707,
      "epoch": 1.9,
      "learning_rate": 4.789396518240132e-05,
      "loss": 0.7888,
      "step": 5260,
      "task_loss": 0.8579301834106445
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6447975635528564,
      "epoch": 1.9,
      "learning_rate": 4.782004525867819e-05,
      "loss": 0.6355,
      "step": 5270,
      "task_loss": 0.42783862352371216
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5919076204299927,
      "epoch": 1.91,
      "learning_rate": 4.774605923049076e-05,
      "loss": 0.7468,
      "step": 5280,
      "task_loss": 0.7170625329017639
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6702316999435425,
      "epoch": 1.91,
      "learning_rate": 4.767200747933586e-05,
      "loss": 0.6474,
      "step": 5290,
      "task_loss": 0.8280285596847534
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1130871772766113,
      "epoch": 1.92,
      "learning_rate": 4.759789038704922e-05,
      "loss": 0.7438,
      "step": 5300,
      "task_loss": 1.4849894046783447
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5206938982009888,
      "epoch": 1.92,
      "learning_rate": 4.7523708335803504e-05,
      "loss": 0.7162,
      "step": 5310,
      "task_loss": 0.5600756406784058
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5643234252929688,
      "epoch": 1.92,
      "learning_rate": 4.7449461708106276e-05,
      "loss": 0.6722,
      "step": 5320,
      "task_loss": 0.6859425902366638
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7487398386001587,
      "epoch": 1.93,
      "learning_rate": 4.7375150886798146e-05,
      "loss": 0.7401,
      "step": 5330,
      "task_loss": 0.9123652577400208
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5706630945205688,
      "epoch": 1.93,
      "learning_rate": 4.73007762550507e-05,
      "loss": 0.7382,
      "step": 5340,
      "task_loss": 0.4185778498649597
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4675668776035309,
      "epoch": 1.93,
      "learning_rate": 4.722633819636453e-05,
      "loss": 0.6022,
      "step": 5350,
      "task_loss": 0.4251108169555664
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5795633792877197,
      "epoch": 1.94,
      "learning_rate": 4.715183709456731e-05,
      "loss": 0.65,
      "step": 5360,
      "task_loss": 0.5892939567565918
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6301298141479492,
      "epoch": 1.94,
      "learning_rate": 4.707727333381177e-05,
      "loss": 0.7081,
      "step": 5370,
      "task_loss": 0.43882763385772705
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6167176961898804,
      "epoch": 1.94,
      "learning_rate": 4.7002647298573725e-05,
      "loss": 0.7195,
      "step": 5380,
      "task_loss": 0.5531620383262634
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6640338897705078,
      "epoch": 1.95,
      "learning_rate": 4.692795937365013e-05,
      "loss": 0.6945,
      "step": 5390,
      "task_loss": 0.5559107661247253
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9159687161445618,
      "epoch": 1.95,
      "learning_rate": 4.685320994415701e-05,
      "loss": 0.6897,
      "step": 5400,
      "task_loss": 1.0885941982269287
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.5111383199691772,
      "epoch": 1.96,
      "learning_rate": 4.677839939552759e-05,
      "loss": 0.8984,
      "step": 5410,
      "task_loss": 1.1360901594161987
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49151280522346497,
      "epoch": 1.96,
      "learning_rate": 4.670352811351019e-05,
      "loss": 0.7072,
      "step": 5420,
      "task_loss": 0.2888219654560089
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6618350744247437,
      "epoch": 1.96,
      "learning_rate": 4.662859648416633e-05,
      "loss": 0.727,
      "step": 5430,
      "task_loss": 0.5811130404472351
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7194912433624268,
      "epoch": 1.97,
      "learning_rate": 4.655360489386869e-05,
      "loss": 0.6801,
      "step": 5440,
      "task_loss": 0.6658434867858887
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9585258960723877,
      "epoch": 1.97,
      "learning_rate": 4.647855372929912e-05,
      "loss": 0.8352,
      "step": 5450,
      "task_loss": 0.8630807399749756
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6633427739143372,
      "epoch": 1.97,
      "learning_rate": 4.640344337744667e-05,
      "loss": 0.7305,
      "step": 5460,
      "task_loss": 0.8915340900421143
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8834680318832397,
      "epoch": 1.98,
      "learning_rate": 4.6328274225605556e-05,
      "loss": 0.7421,
      "step": 5470,
      "task_loss": 0.8303129076957703
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.506862223148346,
      "epoch": 1.98,
      "learning_rate": 4.625304666137321e-05,
      "loss": 0.6902,
      "step": 5480,
      "task_loss": 0.46703067421913147
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5266671180725098,
      "epoch": 1.98,
      "learning_rate": 4.617776107264826e-05,
      "loss": 0.6921,
      "step": 5490,
      "task_loss": 0.5322936177253723
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4714794158935547,
      "epoch": 1.99,
      "learning_rate": 4.6102417847628494e-05,
      "loss": 0.7554,
      "step": 5500,
      "task_loss": 0.4454458951950073
    },
    {
      "epoch": 1.99,
      "eval_exact_match": 81.52317880794702,
      "eval_f1": 88.9781389750277,
      "step": 5500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6718301177024841,
      "epoch": 1.99,
      "learning_rate": 4.602701737480895e-05,
      "loss": 0.7268,
      "step": 5510,
      "task_loss": 0.6084522008895874
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5852054953575134,
      "epoch": 1.99,
      "learning_rate": 4.595156004297978e-05,
      "loss": 0.7398,
      "step": 5520,
      "task_loss": 0.9601466655731201
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7977070808410645,
      "epoch": 2.0,
      "learning_rate": 4.5876046241224384e-05,
      "loss": 0.7316,
      "step": 5530,
      "task_loss": 1.0793871879577637
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5240362286567688,
      "epoch": 2.0,
      "learning_rate": 4.580047635891733e-05,
      "loss": 0.5617,
      "step": 5540,
      "task_loss": 0.4422588050365448
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47274863719940186,
      "epoch": 2.01,
      "learning_rate": 4.572485078572231e-05,
      "loss": 0.5784,
      "step": 5550,
      "task_loss": 0.5353361964225769
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4247085452079773,
      "epoch": 2.01,
      "learning_rate": 4.564916991159023e-05,
      "loss": 0.6764,
      "step": 5560,
      "task_loss": 0.5559691190719604
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5100902318954468,
      "epoch": 2.01,
      "learning_rate": 4.5573434126757125e-05,
      "loss": 0.6411,
      "step": 5570,
      "task_loss": 0.6340029835700989
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6564936637878418,
      "epoch": 2.02,
      "learning_rate": 4.5497643821742164e-05,
      "loss": 0.5141,
      "step": 5580,
      "task_loss": 0.8005180358886719
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8931950330734253,
      "epoch": 2.02,
      "learning_rate": 4.5421799387345656e-05,
      "loss": 0.6289,
      "step": 5590,
      "task_loss": 0.5530489683151245
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4969043731689453,
      "epoch": 2.02,
      "learning_rate": 4.534590121464699e-05,
      "loss": 0.6045,
      "step": 5600,
      "task_loss": 0.6437075138092041
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.48622632026672363,
      "epoch": 2.03,
      "learning_rate": 4.526994969500268e-05,
      "loss": 0.5805,
      "step": 5610,
      "task_loss": 0.7267904281616211
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5548010468482971,
      "epoch": 2.03,
      "learning_rate": 4.5193945220044304e-05,
      "loss": 0.5796,
      "step": 5620,
      "task_loss": 0.8219677209854126
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7763128280639648,
      "epoch": 2.03,
      "learning_rate": 4.5117888181676476e-05,
      "loss": 0.5508,
      "step": 5630,
      "task_loss": 0.6904866099357605
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5509345531463623,
      "epoch": 2.04,
      "learning_rate": 4.5041778972074884e-05,
      "loss": 0.5718,
      "step": 5640,
      "task_loss": 0.36285942792892456
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5179515480995178,
      "epoch": 2.04,
      "learning_rate": 4.4965617983684185e-05,
      "loss": 0.5675,
      "step": 5650,
      "task_loss": 0.6883585453033447
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5797591209411621,
      "epoch": 2.05,
      "learning_rate": 4.4889405609216065e-05,
      "loss": 0.5678,
      "step": 5660,
      "task_loss": 0.6791222095489502
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.865078330039978,
      "epoch": 2.05,
      "learning_rate": 4.481314224164713e-05,
      "loss": 0.6235,
      "step": 5670,
      "task_loss": 0.8147318363189697
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5414189100265503,
      "epoch": 2.05,
      "learning_rate": 4.473682827421697e-05,
      "loss": 0.525,
      "step": 5680,
      "task_loss": 0.7834112644195557
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.529310941696167,
      "epoch": 2.06,
      "learning_rate": 4.4660464100426054e-05,
      "loss": 0.6026,
      "step": 5690,
      "task_loss": 0.5471886396408081
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43340983986854553,
      "epoch": 2.06,
      "learning_rate": 4.458405011403374e-05,
      "loss": 0.5505,
      "step": 5700,
      "task_loss": 0.3781987428665161
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6081724166870117,
      "epoch": 2.06,
      "learning_rate": 4.450758670905624e-05,
      "loss": 0.5351,
      "step": 5710,
      "task_loss": 0.5465576648712158
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4424320161342621,
      "epoch": 2.07,
      "learning_rate": 4.4431074279764585e-05,
      "loss": 0.5218,
      "step": 5720,
      "task_loss": 0.5846661329269409
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5387907028198242,
      "epoch": 2.07,
      "learning_rate": 4.4354513220682575e-05,
      "loss": 0.5726,
      "step": 5730,
      "task_loss": 0.5176105499267578
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.696904182434082,
      "epoch": 2.07,
      "learning_rate": 4.427790392658479e-05,
      "loss": 0.6044,
      "step": 5740,
      "task_loss": 0.7294101715087891
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6790659427642822,
      "epoch": 2.08,
      "learning_rate": 4.4201246792494514e-05,
      "loss": 0.5654,
      "step": 5750,
      "task_loss": 1.0561041831970215
    },
    {
      "epoch": 2.08,
      "eval_exact_match": 81.59886471144749,
      "eval_f1": 89.04378469586636,
      "step": 5750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4576905071735382,
      "epoch": 2.08,
      "learning_rate": 4.41245422136817e-05,
      "loss": 0.6122,
      "step": 5760,
      "task_loss": 0.7306819558143616
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3834877014160156,
      "epoch": 2.09,
      "learning_rate": 4.404779058566094e-05,
      "loss": 0.5291,
      "step": 5770,
      "task_loss": 0.35700392723083496
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6859825849533081,
      "epoch": 2.09,
      "learning_rate": 4.397099230418947e-05,
      "loss": 0.561,
      "step": 5780,
      "task_loss": 1.0357601642608643
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5586205720901489,
      "epoch": 2.09,
      "learning_rate": 4.3894147765265014e-05,
      "loss": 0.5312,
      "step": 5790,
      "task_loss": 0.5139092206954956
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7590255737304688,
      "epoch": 2.1,
      "learning_rate": 4.381725736512387e-05,
      "loss": 0.6398,
      "step": 5800,
      "task_loss": 0.6853212118148804
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5403786897659302,
      "epoch": 2.1,
      "learning_rate": 4.374032150023881e-05,
      "loss": 0.5527,
      "step": 5810,
      "task_loss": 0.8133782744407654
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7060180306434631,
      "epoch": 2.1,
      "learning_rate": 4.366334056731701e-05,
      "loss": 0.5431,
      "step": 5820,
      "task_loss": 0.6709779500961304
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5633669495582581,
      "epoch": 2.11,
      "learning_rate": 4.358631496329804e-05,
      "loss": 0.549,
      "step": 5830,
      "task_loss": 0.44238489866256714
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45574522018432617,
      "epoch": 2.11,
      "learning_rate": 4.350924508535184e-05,
      "loss": 0.5134,
      "step": 5840,
      "task_loss": 0.5507422089576721
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5343673229217529,
      "epoch": 2.11,
      "learning_rate": 4.3432131330876594e-05,
      "loss": 0.5119,
      "step": 5850,
      "task_loss": 0.6400316953659058
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4289189577102661,
      "epoch": 2.12,
      "learning_rate": 4.335497409749677e-05,
      "loss": 0.5306,
      "step": 5860,
      "task_loss": 0.35904762148857117
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5090552568435669,
      "epoch": 2.12,
      "learning_rate": 4.327777378306099e-05,
      "loss": 0.5495,
      "step": 5870,
      "task_loss": 0.43965592980384827
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3385082185268402,
      "epoch": 2.13,
      "learning_rate": 4.3200530785640036e-05,
      "loss": 0.5618,
      "step": 5880,
      "task_loss": 0.5122158527374268
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5285344123840332,
      "epoch": 2.13,
      "learning_rate": 4.312324550352479e-05,
      "loss": 0.5729,
      "step": 5890,
      "task_loss": 0.6291429400444031
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37837162613868713,
      "epoch": 2.13,
      "learning_rate": 4.3045918335224154e-05,
      "loss": 0.5985,
      "step": 5900,
      "task_loss": 0.3076714277267456
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4923848509788513,
      "epoch": 2.14,
      "learning_rate": 4.296854967946301e-05,
      "loss": 0.5914,
      "step": 5910,
      "task_loss": 0.39471399784088135
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9474149942398071,
      "epoch": 2.14,
      "learning_rate": 4.289113993518015e-05,
      "loss": 0.5734,
      "step": 5920,
      "task_loss": 1.258808970451355
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4284988045692444,
      "epoch": 2.14,
      "learning_rate": 4.2813689501526246e-05,
      "loss": 0.5348,
      "step": 5930,
      "task_loss": 0.6401066184043884
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6566985845565796,
      "epoch": 2.15,
      "learning_rate": 4.2736198777861805e-05,
      "loss": 0.6076,
      "step": 5940,
      "task_loss": 0.7970720529556274
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5755891799926758,
      "epoch": 2.15,
      "learning_rate": 4.265866816375503e-05,
      "loss": 0.5311,
      "step": 5950,
      "task_loss": 0.42543694376945496
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41077885031700134,
      "epoch": 2.15,
      "learning_rate": 4.258109805897985e-05,
      "loss": 0.5489,
      "step": 5960,
      "task_loss": 0.32989516854286194
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1504120826721191,
      "epoch": 2.16,
      "learning_rate": 4.2503488863513794e-05,
      "loss": 0.6178,
      "step": 5970,
      "task_loss": 1.1779499053955078
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4549116790294647,
      "epoch": 2.16,
      "learning_rate": 4.242584097753599e-05,
      "loss": 0.5549,
      "step": 5980,
      "task_loss": 0.8180230855941772
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5225155353546143,
      "epoch": 2.16,
      "learning_rate": 4.2348154801425044e-05,
      "loss": 0.5836,
      "step": 5990,
      "task_loss": 1.0198224782943726
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8069769144058228,
      "epoch": 2.17,
      "learning_rate": 4.227043073575699e-05,
      "loss": 0.5687,
      "step": 6000,
      "task_loss": 0.9743910431861877
    },
    {
      "epoch": 2.17,
      "eval_exact_match": 81.69347209082308,
      "eval_f1": 89.0618531655906,
      "step": 6000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.795231282711029,
      "epoch": 2.17,
      "learning_rate": 4.219266918130327e-05,
      "loss": 0.5749,
      "step": 6010,
      "task_loss": 0.9268602132797241
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6040865182876587,
      "epoch": 2.18,
      "learning_rate": 4.211487053902858e-05,
      "loss": 0.5849,
      "step": 6020,
      "task_loss": 0.6653480529785156
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5331346988677979,
      "epoch": 2.18,
      "learning_rate": 4.203703521008891e-05,
      "loss": 0.6451,
      "step": 6030,
      "task_loss": 0.5994354486465454
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4456799328327179,
      "epoch": 2.18,
      "learning_rate": 4.195916359582938e-05,
      "loss": 0.5041,
      "step": 6040,
      "task_loss": 0.989681601524353
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4132393002510071,
      "epoch": 2.19,
      "learning_rate": 4.188125609778219e-05,
      "loss": 0.5796,
      "step": 6050,
      "task_loss": 1.066744089126587
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.91560298204422,
      "epoch": 2.19,
      "learning_rate": 4.180331311766464e-05,
      "loss": 0.6226,
      "step": 6060,
      "task_loss": 0.8165582418441772
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.430245965719223,
      "epoch": 2.19,
      "learning_rate": 4.1725335057376915e-05,
      "loss": 0.5809,
      "step": 6070,
      "task_loss": 0.5763449668884277
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5203709006309509,
      "epoch": 2.2,
      "learning_rate": 4.164732231900013e-05,
      "loss": 0.5706,
      "step": 6080,
      "task_loss": 0.4892015755176544
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.434043824672699,
      "epoch": 2.2,
      "learning_rate": 4.15692753047942e-05,
      "loss": 0.6104,
      "step": 6090,
      "task_loss": 0.4391113817691803
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.492077112197876,
      "epoch": 2.2,
      "learning_rate": 4.149119441719576e-05,
      "loss": 0.6113,
      "step": 6100,
      "task_loss": 0.33191928267478943
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3813079595565796,
      "epoch": 2.21,
      "learning_rate": 4.141308005881614e-05,
      "loss": 0.5083,
      "step": 6110,
      "task_loss": 0.3728364109992981
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44847387075424194,
      "epoch": 2.21,
      "learning_rate": 4.133493263243922e-05,
      "loss": 0.5311,
      "step": 6120,
      "task_loss": 0.34665340185165405
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5354032516479492,
      "epoch": 2.22,
      "learning_rate": 4.1256752541019415e-05,
      "loss": 0.5935,
      "step": 6130,
      "task_loss": 0.7769221067428589
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5569686889648438,
      "epoch": 2.22,
      "learning_rate": 4.1178540187679585e-05,
      "loss": 0.6126,
      "step": 6140,
      "task_loss": 0.6996185779571533
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5548681020736694,
      "epoch": 2.22,
      "learning_rate": 4.1100295975708904e-05,
      "loss": 0.5007,
      "step": 6150,
      "task_loss": 0.4657728672027588
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.650769829750061,
      "epoch": 2.23,
      "learning_rate": 4.102202030856085e-05,
      "loss": 0.6611,
      "step": 6160,
      "task_loss": 0.7665435075759888
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4635891020298004,
      "epoch": 2.23,
      "learning_rate": 4.0943713589851066e-05,
      "loss": 0.5795,
      "step": 6170,
      "task_loss": 0.41729113459587097
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4042726457118988,
      "epoch": 2.23,
      "learning_rate": 4.086537622335534e-05,
      "loss": 0.5163,
      "step": 6180,
      "task_loss": 0.31872618198394775
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45881569385528564,
      "epoch": 2.24,
      "learning_rate": 4.0787008613007484e-05,
      "loss": 0.5228,
      "step": 6190,
      "task_loss": 0.5137763023376465
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45852431654930115,
      "epoch": 2.24,
      "learning_rate": 4.070861116289723e-05,
      "loss": 0.5515,
      "step": 6200,
      "task_loss": 0.5384382009506226
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8377902507781982,
      "epoch": 2.24,
      "learning_rate": 4.063018427726821e-05,
      "loss": 0.5644,
      "step": 6210,
      "task_loss": 0.8768316507339478
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5842976570129395,
      "epoch": 2.25,
      "learning_rate": 4.05517283605158e-05,
      "loss": 0.5803,
      "step": 6220,
      "task_loss": 0.6779041886329651
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6453624963760376,
      "epoch": 2.25,
      "learning_rate": 4.047324381718511e-05,
      "loss": 0.6245,
      "step": 6230,
      "task_loss": 0.44167518615722656
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6187850832939148,
      "epoch": 2.26,
      "learning_rate": 4.039473105196883e-05,
      "loss": 0.5975,
      "step": 6240,
      "task_loss": 0.5326038599014282
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37316951155662537,
      "epoch": 2.26,
      "learning_rate": 4.031619046970517e-05,
      "loss": 0.5677,
      "step": 6250,
      "task_loss": 0.4080279767513275
    },
    {
      "epoch": 2.26,
      "eval_exact_match": 81.9205298013245,
      "eval_f1": 89.22495923106058,
      "step": 6250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6016359329223633,
      "epoch": 2.26,
      "learning_rate": 4.02376224753758e-05,
      "loss": 0.599,
      "step": 6260,
      "task_loss": 0.4683157503604889
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8311920762062073,
      "epoch": 2.27,
      "learning_rate": 4.01590274741037e-05,
      "loss": 0.6095,
      "step": 6270,
      "task_loss": 1.272101640701294
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6363869309425354,
      "epoch": 2.27,
      "learning_rate": 4.008040587115112e-05,
      "loss": 0.5894,
      "step": 6280,
      "task_loss": 1.0684162378311157
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3956725001335144,
      "epoch": 2.27,
      "learning_rate": 4.000175807191752e-05,
      "loss": 0.5567,
      "step": 6290,
      "task_loss": 0.3566119074821472
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7650632858276367,
      "epoch": 2.28,
      "learning_rate": 3.992308448193736e-05,
      "loss": 0.674,
      "step": 6300,
      "task_loss": 0.5824695825576782
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7076523303985596,
      "epoch": 2.28,
      "learning_rate": 3.984438550687815e-05,
      "loss": 0.5626,
      "step": 6310,
      "task_loss": 1.0556983947753906
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47868841886520386,
      "epoch": 2.28,
      "learning_rate": 3.976566155253826e-05,
      "loss": 0.5895,
      "step": 6320,
      "task_loss": 0.6626901626586914
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5259031057357788,
      "epoch": 2.29,
      "learning_rate": 3.9686913024844855e-05,
      "loss": 0.5349,
      "step": 6330,
      "task_loss": 0.3104013502597809
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4891437590122223,
      "epoch": 2.29,
      "learning_rate": 3.960814032985186e-05,
      "loss": 0.5576,
      "step": 6340,
      "task_loss": 0.6433727741241455
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5479400753974915,
      "epoch": 2.29,
      "learning_rate": 3.952934387373775e-05,
      "loss": 0.5763,
      "step": 6350,
      "task_loss": 0.6837672591209412
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6273237466812134,
      "epoch": 2.3,
      "learning_rate": 3.945052406280356e-05,
      "loss": 0.5408,
      "step": 6360,
      "task_loss": 1.0349431037902832
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.48982030153274536,
      "epoch": 2.3,
      "learning_rate": 3.937168130347074e-05,
      "loss": 0.5978,
      "step": 6370,
      "task_loss": 0.8352075815200806
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6010058522224426,
      "epoch": 2.31,
      "learning_rate": 3.9292816002279055e-05,
      "loss": 0.5013,
      "step": 6380,
      "task_loss": 0.645653486251831
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6257983446121216,
      "epoch": 2.31,
      "learning_rate": 3.921392856588455e-05,
      "loss": 0.4692,
      "step": 6390,
      "task_loss": 0.7384665012359619
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5193866491317749,
      "epoch": 2.31,
      "learning_rate": 3.9135019401057337e-05,
      "loss": 0.5861,
      "step": 6400,
      "task_loss": 0.575605571269989
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6001657247543335,
      "epoch": 2.32,
      "learning_rate": 3.9056088914679635e-05,
      "loss": 0.585,
      "step": 6410,
      "task_loss": 0.48626482486724854
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.630062460899353,
      "epoch": 2.32,
      "learning_rate": 3.897713751374355e-05,
      "loss": 0.5203,
      "step": 6420,
      "task_loss": 0.657923698425293
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44871628284454346,
      "epoch": 2.32,
      "learning_rate": 3.889816560534907e-05,
      "loss": 0.5597,
      "step": 6430,
      "task_loss": 0.691073477268219
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4313535690307617,
      "epoch": 2.33,
      "learning_rate": 3.881917359670191e-05,
      "loss": 0.5059,
      "step": 6440,
      "task_loss": 0.70628821849823
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.483176052570343,
      "epoch": 2.33,
      "learning_rate": 3.874016189511141e-05,
      "loss": 0.5833,
      "step": 6450,
      "task_loss": 0.5552691221237183
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.582874059677124,
      "epoch": 2.33,
      "learning_rate": 3.8661130907988505e-05,
      "loss": 0.5192,
      "step": 6460,
      "task_loss": 0.6283879280090332
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45941537618637085,
      "epoch": 2.34,
      "learning_rate": 3.85820810428435e-05,
      "loss": 0.59,
      "step": 6470,
      "task_loss": 0.8111543655395508
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6040874719619751,
      "epoch": 2.34,
      "learning_rate": 3.850301270728412e-05,
      "loss": 0.5399,
      "step": 6480,
      "task_loss": 0.8921731114387512
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6502740979194641,
      "epoch": 2.35,
      "learning_rate": 3.8423926309013267e-05,
      "loss": 0.5538,
      "step": 6490,
      "task_loss": 0.6972618699073792
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5272783041000366,
      "epoch": 2.35,
      "learning_rate": 3.834482225582701e-05,
      "loss": 0.5857,
      "step": 6500,
      "task_loss": 0.8147091865539551
    },
    {
      "epoch": 2.35,
      "eval_exact_match": 81.54210028382214,
      "eval_f1": 89.0964633965607,
      "step": 6500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7971393465995789,
      "epoch": 2.35,
      "learning_rate": 3.826570095561245e-05,
      "loss": 0.5413,
      "step": 6510,
      "task_loss": 0.8758788108825684
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6447325944900513,
      "epoch": 2.36,
      "learning_rate": 3.8186562816345615e-05,
      "loss": 0.5911,
      "step": 6520,
      "task_loss": 0.6143375635147095
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5946791172027588,
      "epoch": 2.36,
      "learning_rate": 3.8107408246089365e-05,
      "loss": 0.5965,
      "step": 6530,
      "task_loss": 0.712870717048645
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5228013396263123,
      "epoch": 2.36,
      "learning_rate": 3.802823765299128e-05,
      "loss": 0.5602,
      "step": 6540,
      "task_loss": 0.8571695685386658
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6156158447265625,
      "epoch": 2.37,
      "learning_rate": 3.794905144528156e-05,
      "loss": 0.5644,
      "step": 6550,
      "task_loss": 0.9497164487838745
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6857894659042358,
      "epoch": 2.37,
      "learning_rate": 3.7869850031270916e-05,
      "loss": 0.5606,
      "step": 6560,
      "task_loss": 0.6282366514205933
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5698332190513611,
      "epoch": 2.37,
      "learning_rate": 3.7790633819348476e-05,
      "loss": 0.5349,
      "step": 6570,
      "task_loss": 0.48486849665641785
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.551287055015564,
      "epoch": 2.38,
      "learning_rate": 3.771140321797967e-05,
      "loss": 0.5857,
      "step": 6580,
      "task_loss": 0.6155105829238892
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4047469198703766,
      "epoch": 2.38,
      "learning_rate": 3.7632158635704116e-05,
      "loss": 0.5965,
      "step": 6590,
      "task_loss": 0.68361496925354
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5632671117782593,
      "epoch": 2.39,
      "learning_rate": 3.755290048113352e-05,
      "loss": 0.5808,
      "step": 6600,
      "task_loss": 1.0737969875335693
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5189325213432312,
      "epoch": 2.39,
      "learning_rate": 3.747362916294959e-05,
      "loss": 0.5092,
      "step": 6610,
      "task_loss": 1.006134033203125
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5590713024139404,
      "epoch": 2.39,
      "learning_rate": 3.7394345089901914e-05,
      "loss": 0.5843,
      "step": 6620,
      "task_loss": 1.0642428398132324
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6743557453155518,
      "epoch": 2.4,
      "learning_rate": 3.731504867080579e-05,
      "loss": 0.6301,
      "step": 6630,
      "task_loss": 1.1177798509597778
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6843348741531372,
      "epoch": 2.4,
      "learning_rate": 3.723574031454026e-05,
      "loss": 0.6255,
      "step": 6640,
      "task_loss": 1.0057365894317627
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5847528576850891,
      "epoch": 2.4,
      "learning_rate": 3.715642043004586e-05,
      "loss": 0.5272,
      "step": 6650,
      "task_loss": 0.7699944972991943
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.48671820759773254,
      "epoch": 2.41,
      "learning_rate": 3.7077089426322586e-05,
      "loss": 0.5762,
      "step": 6660,
      "task_loss": 0.7598029375076294
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5621806383132935,
      "epoch": 2.41,
      "learning_rate": 3.699774771242778e-05,
      "loss": 0.5534,
      "step": 6670,
      "task_loss": 0.47044456005096436
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.499066561460495,
      "epoch": 2.41,
      "learning_rate": 3.691839569747399e-05,
      "loss": 0.5246,
      "step": 6680,
      "task_loss": 0.48376011848449707
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6511626839637756,
      "epoch": 2.42,
      "learning_rate": 3.68390337906269e-05,
      "loss": 0.6135,
      "step": 6690,
      "task_loss": 0.9496736526489258
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7415065169334412,
      "epoch": 2.42,
      "learning_rate": 3.675966240110316e-05,
      "loss": 0.5912,
      "step": 6700,
      "task_loss": 0.7771897315979004
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6975048780441284,
      "epoch": 2.43,
      "learning_rate": 3.668028193816837e-05,
      "loss": 0.6043,
      "step": 6710,
      "task_loss": 0.7732299566268921
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5505435466766357,
      "epoch": 2.43,
      "learning_rate": 3.6600892811134887e-05,
      "loss": 0.5488,
      "step": 6720,
      "task_loss": 0.47356608510017395
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6242953538894653,
      "epoch": 2.43,
      "learning_rate": 3.652149542935974e-05,
      "loss": 0.5837,
      "step": 6730,
      "task_loss": 0.6268038749694824
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5890005826950073,
      "epoch": 2.44,
      "learning_rate": 3.644209020224254e-05,
      "loss": 0.5406,
      "step": 6740,
      "task_loss": 0.503425121307373
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.510681688785553,
      "epoch": 2.44,
      "learning_rate": 3.6362677539223316e-05,
      "loss": 0.4745,
      "step": 6750,
      "task_loss": 0.6263338327407837
    },
    {
      "epoch": 2.44,
      "eval_exact_match": 81.9205298013245,
      "eval_f1": 89.17913767330805,
      "step": 6750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.48122429847717285,
      "epoch": 2.44,
      "learning_rate": 3.628325784978048e-05,
      "loss": 0.6206,
      "step": 6760,
      "task_loss": 0.7160674333572388
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5861722230911255,
      "epoch": 2.45,
      "learning_rate": 3.620383154342866e-05,
      "loss": 0.5832,
      "step": 6770,
      "task_loss": 0.7612639665603638
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35261979699134827,
      "epoch": 2.45,
      "learning_rate": 3.612439902971659e-05,
      "loss": 0.6246,
      "step": 6780,
      "task_loss": 0.43168213963508606
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4552045464515686,
      "epoch": 2.45,
      "learning_rate": 3.604496071822503e-05,
      "loss": 0.5272,
      "step": 6790,
      "task_loss": 0.5549889206886292
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5096551775932312,
      "epoch": 2.46,
      "learning_rate": 3.596551701856461e-05,
      "loss": 0.4898,
      "step": 6800,
      "task_loss": 0.6168457865715027
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44083502888679504,
      "epoch": 2.46,
      "learning_rate": 3.5886068340373774e-05,
      "loss": 0.5555,
      "step": 6810,
      "task_loss": 0.525178074836731
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7820392847061157,
      "epoch": 2.46,
      "learning_rate": 3.580661509331662e-05,
      "loss": 0.5692,
      "step": 6820,
      "task_loss": 0.7302403450012207
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5217739343643188,
      "epoch": 2.47,
      "learning_rate": 3.572715768708081e-05,
      "loss": 0.5292,
      "step": 6830,
      "task_loss": 0.6324975490570068
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6764619946479797,
      "epoch": 2.47,
      "learning_rate": 3.564769653137545e-05,
      "loss": 0.5874,
      "step": 6840,
      "task_loss": 0.6163495779037476
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45639386773109436,
      "epoch": 2.48,
      "learning_rate": 3.556823203592897e-05,
      "loss": 0.5314,
      "step": 6850,
      "task_loss": 0.3424451947212219
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6600396633148193,
      "epoch": 2.48,
      "learning_rate": 3.548876461048703e-05,
      "loss": 0.5602,
      "step": 6860,
      "task_loss": 0.7038537263870239
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6123368740081787,
      "epoch": 2.48,
      "learning_rate": 3.5409294664810414e-05,
      "loss": 0.6264,
      "step": 6870,
      "task_loss": 0.5839701890945435
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.50824373960495,
      "epoch": 2.49,
      "learning_rate": 3.5329822608672863e-05,
      "loss": 0.5268,
      "step": 6880,
      "task_loss": 0.5268850922584534
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5141448974609375,
      "epoch": 2.49,
      "learning_rate": 3.5250348851859044e-05,
      "loss": 0.5426,
      "step": 6890,
      "task_loss": 0.3040235638618469
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.53406822681427,
      "epoch": 2.49,
      "learning_rate": 3.517087380416235e-05,
      "loss": 0.5786,
      "step": 6900,
      "task_loss": 0.5800484418869019
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7829024791717529,
      "epoch": 2.5,
      "learning_rate": 3.509139787538286e-05,
      "loss": 0.5593,
      "step": 6910,
      "task_loss": 0.8509390950202942
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7222518920898438,
      "epoch": 2.5,
      "learning_rate": 3.501192147532521e-05,
      "loss": 0.5797,
      "step": 6920,
      "task_loss": 1.3599720001220703
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5438934564590454,
      "epoch": 2.5,
      "learning_rate": 3.49324450137964e-05,
      "loss": 0.6199,
      "step": 6930,
      "task_loss": 0.6372125148773193
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4705803394317627,
      "epoch": 2.51,
      "learning_rate": 3.485296890060384e-05,
      "loss": 0.5347,
      "step": 6940,
      "task_loss": 0.7346433401107788
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5038604140281677,
      "epoch": 2.51,
      "learning_rate": 3.4773493545553046e-05,
      "loss": 0.5929,
      "step": 6950,
      "task_loss": 0.8522573709487915
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4030439853668213,
      "epoch": 2.52,
      "learning_rate": 3.469401935844572e-05,
      "loss": 0.5134,
      "step": 6960,
      "task_loss": 0.23669208586215973
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5648401975631714,
      "epoch": 2.52,
      "learning_rate": 3.461454674907746e-05,
      "loss": 0.5579,
      "step": 6970,
      "task_loss": 0.5249247550964355
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7584888339042664,
      "epoch": 2.52,
      "learning_rate": 3.453507612723578e-05,
      "loss": 0.5626,
      "step": 6980,
      "task_loss": 0.7458637356758118
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6533717513084412,
      "epoch": 2.53,
      "learning_rate": 3.445560790269793e-05,
      "loss": 0.5996,
      "step": 6990,
      "task_loss": 0.7032555341720581
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4984493553638458,
      "epoch": 2.53,
      "learning_rate": 3.437614248522879e-05,
      "loss": 0.6306,
      "step": 7000,
      "task_loss": 0.46757519245147705
    },
    {
      "epoch": 2.53,
      "eval_exact_match": 81.42857142857143,
      "eval_f1": 88.92804168222406,
      "step": 7000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4259212017059326,
      "epoch": 2.53,
      "learning_rate": 3.429668028457878e-05,
      "loss": 0.5477,
      "step": 7010,
      "task_loss": 0.58353590965271
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7622242569923401,
      "epoch": 2.54,
      "learning_rate": 3.421722171048173e-05,
      "loss": 0.54,
      "step": 7020,
      "task_loss": 0.8156880140304565
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6038123369216919,
      "epoch": 2.54,
      "learning_rate": 3.413776717265275e-05,
      "loss": 0.6034,
      "step": 7030,
      "task_loss": 0.7022565603256226
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33076730370521545,
      "epoch": 2.54,
      "learning_rate": 3.4058317080786186e-05,
      "loss": 0.5226,
      "step": 7040,
      "task_loss": 0.26581063866615295
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7431491613388062,
      "epoch": 2.55,
      "learning_rate": 3.397887184455339e-05,
      "loss": 0.5195,
      "step": 7050,
      "task_loss": 0.5218563079833984
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6481354236602783,
      "epoch": 2.55,
      "learning_rate": 3.389943187360075e-05,
      "loss": 0.629,
      "step": 7060,
      "task_loss": 0.6381015777587891
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5212134122848511,
      "epoch": 2.56,
      "learning_rate": 3.381999757754745e-05,
      "loss": 0.5217,
      "step": 7070,
      "task_loss": 0.6918044686317444
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42733877897262573,
      "epoch": 2.56,
      "learning_rate": 3.374056936598344e-05,
      "loss": 0.5712,
      "step": 7080,
      "task_loss": 0.6634185314178467
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36479300260543823,
      "epoch": 2.56,
      "learning_rate": 3.3661147648467286e-05,
      "loss": 0.4548,
      "step": 7090,
      "task_loss": 0.9132073521614075
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43868011236190796,
      "epoch": 2.57,
      "learning_rate": 3.3581732834524065e-05,
      "loss": 0.5306,
      "step": 7100,
      "task_loss": 0.44713014364242554
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.738867461681366,
      "epoch": 2.57,
      "learning_rate": 3.350232533364326e-05,
      "loss": 0.541,
      "step": 7110,
      "task_loss": 0.9076246023178101
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47398871183395386,
      "epoch": 2.57,
      "learning_rate": 3.342292555527666e-05,
      "loss": 0.5282,
      "step": 7120,
      "task_loss": 0.48428577184677124
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6713889837265015,
      "epoch": 2.58,
      "learning_rate": 3.3343533908836205e-05,
      "loss": 0.6313,
      "step": 7130,
      "task_loss": 1.2264108657836914
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7997803688049316,
      "epoch": 2.58,
      "learning_rate": 3.326415080369194e-05,
      "loss": 0.5618,
      "step": 7140,
      "task_loss": 1.1236743927001953
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4207914471626282,
      "epoch": 2.58,
      "learning_rate": 3.3184776649169834e-05,
      "loss": 0.5202,
      "step": 7150,
      "task_loss": 0.477580726146698
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5528711080551147,
      "epoch": 2.59,
      "learning_rate": 3.3105411854549734e-05,
      "loss": 0.5922,
      "step": 7160,
      "task_loss": 0.9368165731430054
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4040418267250061,
      "epoch": 2.59,
      "learning_rate": 3.302605682906319e-05,
      "loss": 0.553,
      "step": 7170,
      "task_loss": 0.4502519369125366
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5803967714309692,
      "epoch": 2.59,
      "learning_rate": 3.294671198189141e-05,
      "loss": 0.5853,
      "step": 7180,
      "task_loss": 0.8804277181625366
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7026090621948242,
      "epoch": 2.6,
      "learning_rate": 3.2867377722163106e-05,
      "loss": 0.6469,
      "step": 7190,
      "task_loss": 0.45715072751045227
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45515570044517517,
      "epoch": 2.6,
      "learning_rate": 3.2788054458952395e-05,
      "loss": 0.4855,
      "step": 7200,
      "task_loss": 1.0894179344177246
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6373165249824524,
      "epoch": 2.61,
      "learning_rate": 3.270874260127669e-05,
      "loss": 0.5423,
      "step": 7210,
      "task_loss": 0.43047064542770386
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5695480108261108,
      "epoch": 2.61,
      "learning_rate": 3.2637372019106546e-05,
      "loss": 0.602,
      "step": 7220,
      "task_loss": 0.5894209742546082
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35404130816459656,
      "epoch": 2.61,
      "learning_rate": 3.255808295857803e-05,
      "loss": 0.5373,
      "step": 7230,
      "task_loss": 0.6537626385688782
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5384463667869568,
      "epoch": 2.62,
      "learning_rate": 3.247880648939492e-05,
      "loss": 0.5688,
      "step": 7240,
      "task_loss": 0.7174534797668457
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5153848528862,
      "epoch": 2.62,
      "learning_rate": 3.239954302033335e-05,
      "loss": 0.5985,
      "step": 7250,
      "task_loss": 0.8951561450958252
    },
    {
      "epoch": 2.62,
      "eval_exact_match": 82.03405865657521,
      "eval_f1": 89.46367633239403,
      "step": 7250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36988773941993713,
      "epoch": 2.62,
      "learning_rate": 3.232029296010241e-05,
      "loss": 0.508,
      "step": 7260,
      "task_loss": 0.371121346950531
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5778599977493286,
      "epoch": 2.63,
      "learning_rate": 3.224105671734206e-05,
      "loss": 0.5983,
      "step": 7270,
      "task_loss": 0.6856058239936829
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5628077983856201,
      "epoch": 2.63,
      "learning_rate": 3.2161834700621026e-05,
      "loss": 0.525,
      "step": 7280,
      "task_loss": 0.6911593675613403
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7071762681007385,
      "epoch": 2.63,
      "learning_rate": 3.2082627318434634e-05,
      "loss": 0.5646,
      "step": 7290,
      "task_loss": 1.0483005046844482
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.677939236164093,
      "epoch": 2.64,
      "learning_rate": 3.2003434979202836e-05,
      "loss": 0.5433,
      "step": 7300,
      "task_loss": 0.41416674852371216
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5552082061767578,
      "epoch": 2.64,
      "learning_rate": 3.19242580912679e-05,
      "loss": 0.5089,
      "step": 7310,
      "task_loss": 0.7204807996749878
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5466291904449463,
      "epoch": 2.65,
      "learning_rate": 3.1845097062892545e-05,
      "loss": 0.6588,
      "step": 7320,
      "task_loss": 1.0191127061843872
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5075401067733765,
      "epoch": 2.65,
      "learning_rate": 3.1765952302257604e-05,
      "loss": 0.5275,
      "step": 7330,
      "task_loss": 0.4421289265155792
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.720165491104126,
      "epoch": 2.65,
      "learning_rate": 3.1686824217460105e-05,
      "loss": 0.5934,
      "step": 7340,
      "task_loss": 1.0821608304977417
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7632290124893188,
      "epoch": 2.66,
      "learning_rate": 3.160771321651105e-05,
      "loss": 0.5672,
      "step": 7350,
      "task_loss": 0.9648917317390442
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4810241758823395,
      "epoch": 2.66,
      "learning_rate": 3.152861970733336e-05,
      "loss": 0.589,
      "step": 7360,
      "task_loss": 0.7089782953262329
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4056462049484253,
      "epoch": 2.66,
      "learning_rate": 3.144954409775978e-05,
      "loss": 0.5529,
      "step": 7370,
      "task_loss": 0.30260807275772095
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5448493361473083,
      "epoch": 2.67,
      "learning_rate": 3.1370486795530724e-05,
      "loss": 0.5954,
      "step": 7380,
      "task_loss": 0.7051540613174438
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4768620729446411,
      "epoch": 2.67,
      "learning_rate": 3.129144820829223e-05,
      "loss": 0.5176,
      "step": 7390,
      "task_loss": 0.4055364727973938
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5303025841712952,
      "epoch": 2.67,
      "learning_rate": 3.1212428743593856e-05,
      "loss": 0.4851,
      "step": 7400,
      "task_loss": 0.6226521730422974
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4235970675945282,
      "epoch": 2.68,
      "learning_rate": 3.113342880888649e-05,
      "loss": 0.486,
      "step": 7410,
      "task_loss": 0.5380692481994629
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5234407186508179,
      "epoch": 2.68,
      "learning_rate": 3.10544488115204e-05,
      "loss": 0.5456,
      "step": 7420,
      "task_loss": 0.6824461221694946
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6319477558135986,
      "epoch": 2.69,
      "learning_rate": 3.097548915874299e-05,
      "loss": 0.593,
      "step": 7430,
      "task_loss": 0.5957967042922974
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5229743123054504,
      "epoch": 2.69,
      "learning_rate": 3.08965502576968e-05,
      "loss": 0.4986,
      "step": 7440,
      "task_loss": 0.44954097270965576
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5310331583023071,
      "epoch": 2.69,
      "learning_rate": 3.081763251541732e-05,
      "loss": 0.5477,
      "step": 7450,
      "task_loss": 0.8785396218299866
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6118483543395996,
      "epoch": 2.7,
      "learning_rate": 3.0738736338830997e-05,
      "loss": 0.5452,
      "step": 7460,
      "task_loss": 0.8830308318138123
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7529491186141968,
      "epoch": 2.7,
      "learning_rate": 3.0659862134753025e-05,
      "loss": 0.5697,
      "step": 7470,
      "task_loss": 1.0144604444503784
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46560198068618774,
      "epoch": 2.7,
      "learning_rate": 3.0581010309885335e-05,
      "loss": 0.5672,
      "step": 7480,
      "task_loss": 0.5504160523414612
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44873565435409546,
      "epoch": 2.71,
      "learning_rate": 3.0502181270814433e-05,
      "loss": 0.5469,
      "step": 7490,
      "task_loss": 0.730833888053894
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8784873485565186,
      "epoch": 2.71,
      "learning_rate": 3.042337542400939e-05,
      "loss": 0.5677,
      "step": 7500,
      "task_loss": 0.9325830936431885
    },
    {
      "epoch": 2.71,
      "eval_exact_match": 82.02459791863765,
      "eval_f1": 89.43948705493302,
      "step": 7500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4937174320220947,
      "epoch": 2.71,
      "learning_rate": 3.0344593175819606e-05,
      "loss": 0.5672,
      "step": 7510,
      "task_loss": 0.693834662437439
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5106741189956665,
      "epoch": 2.72,
      "learning_rate": 3.026583493247288e-05,
      "loss": 0.5034,
      "step": 7520,
      "task_loss": 0.9328588843345642
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.594123363494873,
      "epoch": 2.72,
      "learning_rate": 3.018710110007318e-05,
      "loss": 0.514,
      "step": 7530,
      "task_loss": 1.1014786958694458
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5339140892028809,
      "epoch": 2.72,
      "learning_rate": 3.010839208459863e-05,
      "loss": 0.5621,
      "step": 7540,
      "task_loss": 0.6708189249038696
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6270914673805237,
      "epoch": 2.73,
      "learning_rate": 3.0029708291899378e-05,
      "loss": 0.5361,
      "step": 7550,
      "task_loss": 0.8918893337249756
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36830925941467285,
      "epoch": 2.73,
      "learning_rate": 2.9951050127695518e-05,
      "loss": 0.5261,
      "step": 7560,
      "task_loss": 0.5323593616485596
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4848215579986572,
      "epoch": 2.74,
      "learning_rate": 2.9872417997574987e-05,
      "loss": 0.5538,
      "step": 7570,
      "task_loss": 0.719284176826477
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4425901472568512,
      "epoch": 2.74,
      "learning_rate": 2.979381230699151e-05,
      "loss": 0.5173,
      "step": 7580,
      "task_loss": 0.6015424132347107
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5119283199310303,
      "epoch": 2.74,
      "learning_rate": 2.9715233461262427e-05,
      "loss": 0.5088,
      "step": 7590,
      "task_loss": 0.641405463218689
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.48643380403518677,
      "epoch": 2.75,
      "learning_rate": 2.9636681865566735e-05,
      "loss": 0.566,
      "step": 7600,
      "task_loss": 0.47257906198501587
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45301955938339233,
      "epoch": 2.75,
      "learning_rate": 2.9558157924942824e-05,
      "loss": 0.5988,
      "step": 7610,
      "task_loss": 0.802404522895813
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41510939598083496,
      "epoch": 2.75,
      "learning_rate": 2.947966204428658e-05,
      "loss": 0.5027,
      "step": 7620,
      "task_loss": 0.6679592132568359
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5907465815544128,
      "epoch": 2.76,
      "learning_rate": 2.940119462834914e-05,
      "loss": 0.5256,
      "step": 7630,
      "task_loss": 1.0826715230941772
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6161787509918213,
      "epoch": 2.76,
      "learning_rate": 2.9322756081734894e-05,
      "loss": 0.595,
      "step": 7640,
      "task_loss": 0.7296005487442017
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4290156960487366,
      "epoch": 2.76,
      "learning_rate": 2.9244346808899347e-05,
      "loss": 0.4999,
      "step": 7650,
      "task_loss": 0.6461764574050903
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5703807473182678,
      "epoch": 2.77,
      "learning_rate": 2.91659672141471e-05,
      "loss": 0.5634,
      "step": 7660,
      "task_loss": 0.7666523456573486
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42602774500846863,
      "epoch": 2.77,
      "learning_rate": 2.9087617701629685e-05,
      "loss": 0.5413,
      "step": 7670,
      "task_loss": 0.5653133392333984
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5586769580841064,
      "epoch": 2.78,
      "learning_rate": 2.9009298675343544e-05,
      "loss": 0.5102,
      "step": 7680,
      "task_loss": 1.4154021739959717
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5477520227432251,
      "epoch": 2.78,
      "learning_rate": 2.8931010539127894e-05,
      "loss": 0.5165,
      "step": 7690,
      "task_loss": 0.6499199867248535
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46430933475494385,
      "epoch": 2.78,
      "learning_rate": 2.8852753696662726e-05,
      "loss": 0.5086,
      "step": 7700,
      "task_loss": 0.4722907543182373
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2965603768825531,
      "epoch": 2.79,
      "learning_rate": 2.877452855146659e-05,
      "loss": 0.5257,
      "step": 7710,
      "task_loss": 0.3856130540370941
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7141191363334656,
      "epoch": 2.79,
      "learning_rate": 2.8696335506894678e-05,
      "loss": 0.6047,
      "step": 7720,
      "task_loss": 1.0723342895507812
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5581351518630981,
      "epoch": 2.79,
      "learning_rate": 2.8618174966136597e-05,
      "loss": 0.5185,
      "step": 7730,
      "task_loss": 0.6514792442321777
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5451016426086426,
      "epoch": 2.8,
      "learning_rate": 2.8540047332214402e-05,
      "loss": 0.6023,
      "step": 7740,
      "task_loss": 0.6417021751403809
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4637037217617035,
      "epoch": 2.8,
      "learning_rate": 2.8461953007980413e-05,
      "loss": 0.5725,
      "step": 7750,
      "task_loss": 0.5279682278633118
    },
    {
      "epoch": 2.8,
      "eval_exact_match": 81.86376537369915,
      "eval_f1": 89.34181998595834,
      "step": 7750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.622160792350769,
      "epoch": 2.8,
      "learning_rate": 2.8383892396115255e-05,
      "loss": 0.5184,
      "step": 7760,
      "task_loss": 0.8936756253242493
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39960283041000366,
      "epoch": 2.81,
      "learning_rate": 2.830586589912568e-05,
      "loss": 0.5541,
      "step": 7770,
      "task_loss": 0.635195791721344
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6131960153579712,
      "epoch": 2.81,
      "learning_rate": 2.8227873919342545e-05,
      "loss": 0.5496,
      "step": 7780,
      "task_loss": 0.845227837562561
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.48792392015457153,
      "epoch": 2.82,
      "learning_rate": 2.814991685891871e-05,
      "loss": 0.4879,
      "step": 7790,
      "task_loss": 0.7151265144348145
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.659151554107666,
      "epoch": 2.82,
      "learning_rate": 2.807199511982703e-05,
      "loss": 0.6531,
      "step": 7800,
      "task_loss": 1.1711469888687134
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4288066029548645,
      "epoch": 2.82,
      "learning_rate": 2.7994109103858147e-05,
      "loss": 0.5295,
      "step": 7810,
      "task_loss": 0.40471869707107544
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46426552534103394,
      "epoch": 2.83,
      "learning_rate": 2.7916259212618592e-05,
      "loss": 0.5089,
      "step": 7820,
      "task_loss": 0.44435638189315796
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6056751608848572,
      "epoch": 2.83,
      "learning_rate": 2.7838445847528554e-05,
      "loss": 0.5387,
      "step": 7830,
      "task_loss": 0.74387526512146
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3924906253814697,
      "epoch": 2.83,
      "learning_rate": 2.776066940981993e-05,
      "loss": 0.5363,
      "step": 7840,
      "task_loss": 0.6373250484466553
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5626404285430908,
      "epoch": 2.84,
      "learning_rate": 2.7682930300534165e-05,
      "loss": 0.5712,
      "step": 7850,
      "task_loss": 0.7083041667938232
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6474617719650269,
      "epoch": 2.84,
      "learning_rate": 2.7605228920520273e-05,
      "loss": 0.5376,
      "step": 7860,
      "task_loss": 0.6635825634002686
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7015373706817627,
      "epoch": 2.84,
      "learning_rate": 2.7527565670432668e-05,
      "loss": 0.6284,
      "step": 7870,
      "task_loss": 1.1057612895965576
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.592483401298523,
      "epoch": 2.85,
      "learning_rate": 2.7449940950729213e-05,
      "loss": 0.5704,
      "step": 7880,
      "task_loss": 0.8539618253707886
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5118753910064697,
      "epoch": 2.85,
      "learning_rate": 2.737235516166903e-05,
      "loss": 0.5517,
      "step": 7890,
      "task_loss": 0.6258624792098999
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47284987568855286,
      "epoch": 2.86,
      "learning_rate": 2.729480870331058e-05,
      "loss": 0.4877,
      "step": 7900,
      "task_loss": 0.6712979674339294
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5549010038375854,
      "epoch": 2.86,
      "learning_rate": 2.721730197550944e-05,
      "loss": 0.5215,
      "step": 7910,
      "task_loss": 1.0644373893737793
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6102703809738159,
      "epoch": 2.86,
      "learning_rate": 2.7139835377916394e-05,
      "loss": 0.5548,
      "step": 7920,
      "task_loss": 0.5565172433853149
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49971044063568115,
      "epoch": 2.87,
      "learning_rate": 2.7062409309975242e-05,
      "loss": 0.5297,
      "step": 7930,
      "task_loss": 1.1361805200576782
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6527148485183716,
      "epoch": 2.87,
      "learning_rate": 2.6985024170920843e-05,
      "loss": 0.6014,
      "step": 7940,
      "task_loss": 0.7518099546432495
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5662878155708313,
      "epoch": 2.87,
      "learning_rate": 2.690768035977699e-05,
      "loss": 0.5916,
      "step": 7950,
      "task_loss": 0.7932103872299194
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6615817546844482,
      "epoch": 2.88,
      "learning_rate": 2.6830378275354378e-05,
      "loss": 0.54,
      "step": 7960,
      "task_loss": 0.6306872367858887
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4921113848686218,
      "epoch": 2.88,
      "learning_rate": 2.6753118316248545e-05,
      "loss": 0.4873,
      "step": 7970,
      "task_loss": 0.5558697581291199
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7511510848999023,
      "epoch": 2.88,
      "learning_rate": 2.6675900880837846e-05,
      "loss": 0.5769,
      "step": 7980,
      "task_loss": 0.8355304598808289
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5489118099212646,
      "epoch": 2.89,
      "learning_rate": 2.65987263672813e-05,
      "loss": 0.5927,
      "step": 7990,
      "task_loss": 1.078818440437317
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46048545837402344,
      "epoch": 2.89,
      "learning_rate": 2.652159517351669e-05,
      "loss": 0.6023,
      "step": 8000,
      "task_loss": 0.5179476737976074
    },
    {
      "epoch": 2.89,
      "eval_exact_match": 82.23273415326395,
      "eval_f1": 89.49494539305317,
      "step": 8000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4600314497947693,
      "epoch": 2.89,
      "learning_rate": 2.644450769725837e-05,
      "loss": 0.6,
      "step": 8010,
      "task_loss": 1.121453046798706
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5110729336738586,
      "epoch": 2.9,
      "learning_rate": 2.6367464335995296e-05,
      "loss": 0.5577,
      "step": 8020,
      "task_loss": 0.517736554145813
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6211916208267212,
      "epoch": 2.9,
      "learning_rate": 2.6290465486988934e-05,
      "loss": 0.5561,
      "step": 8030,
      "task_loss": 0.9467544555664062
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.451476514339447,
      "epoch": 2.91,
      "learning_rate": 2.6213511547271256e-05,
      "loss": 0.5386,
      "step": 8040,
      "task_loss": 0.6299540400505066
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.397861510515213,
      "epoch": 2.91,
      "learning_rate": 2.613660291364264e-05,
      "loss": 0.5725,
      "step": 8050,
      "task_loss": 0.6148253679275513
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42952167987823486,
      "epoch": 2.91,
      "learning_rate": 2.6059739982669875e-05,
      "loss": 0.5305,
      "step": 8060,
      "task_loss": 0.49444225430488586
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4712672829627991,
      "epoch": 2.92,
      "learning_rate": 2.5982923150684062e-05,
      "loss": 0.5201,
      "step": 8070,
      "task_loss": 0.3702479302883148
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5742895603179932,
      "epoch": 2.92,
      "learning_rate": 2.5906152813778642e-05,
      "loss": 0.5038,
      "step": 8080,
      "task_loss": 0.6904934644699097
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44902893900871277,
      "epoch": 2.92,
      "learning_rate": 2.582942936780726e-05,
      "loss": 0.5227,
      "step": 8090,
      "task_loss": 0.5900508165359497
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5167535543441772,
      "epoch": 2.93,
      "learning_rate": 2.575275320838183e-05,
      "loss": 0.5737,
      "step": 8100,
      "task_loss": 0.7928590178489685
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40067818760871887,
      "epoch": 2.93,
      "learning_rate": 2.5676124730870394e-05,
      "loss": 0.456,
      "step": 8110,
      "task_loss": 0.3831639885902405
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3589646816253662,
      "epoch": 2.93,
      "learning_rate": 2.5599544330395164e-05,
      "loss": 0.537,
      "step": 8120,
      "task_loss": 0.42124050855636597
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6571911573410034,
      "epoch": 2.94,
      "learning_rate": 2.5523012401830417e-05,
      "loss": 0.5946,
      "step": 8130,
      "task_loss": 0.8380374312400818
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5014092326164246,
      "epoch": 2.94,
      "learning_rate": 2.5446529339800535e-05,
      "loss": 0.5636,
      "step": 8140,
      "task_loss": 0.4395813047885895
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7437175512313843,
      "epoch": 2.95,
      "learning_rate": 2.5370095538677883e-05,
      "loss": 0.5487,
      "step": 8150,
      "task_loss": 1.2434213161468506
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4760008454322815,
      "epoch": 2.95,
      "learning_rate": 2.529371139258086e-05,
      "loss": 0.5244,
      "step": 8160,
      "task_loss": 0.5885391235351562
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6417983770370483,
      "epoch": 2.95,
      "learning_rate": 2.5217377295371787e-05,
      "loss": 0.5579,
      "step": 8170,
      "task_loss": 0.9728447198867798
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6246418952941895,
      "epoch": 2.96,
      "learning_rate": 2.5141093640654972e-05,
      "loss": 0.6308,
      "step": 8180,
      "task_loss": 0.5883153676986694
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.605870246887207,
      "epoch": 2.96,
      "learning_rate": 2.506486082177455e-05,
      "loss": 0.5425,
      "step": 8190,
      "task_loss": 0.861560583114624
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5695328712463379,
      "epoch": 2.96,
      "learning_rate": 2.4988679231812602e-05,
      "loss": 0.5019,
      "step": 8200,
      "task_loss": 0.6430089473724365
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5561754703521729,
      "epoch": 2.97,
      "learning_rate": 2.4912549263587004e-05,
      "loss": 0.522,
      "step": 8210,
      "task_loss": 0.7524732351303101
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4856584072113037,
      "epoch": 2.97,
      "learning_rate": 2.4836471309649488e-05,
      "loss": 0.5376,
      "step": 8220,
      "task_loss": 0.44399377703666687
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.417047381401062,
      "epoch": 2.97,
      "learning_rate": 2.4760445762283548e-05,
      "loss": 0.5393,
      "step": 8230,
      "task_loss": 0.7637783288955688
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.418597936630249,
      "epoch": 2.98,
      "learning_rate": 2.468447301350249e-05,
      "loss": 0.4973,
      "step": 8240,
      "task_loss": 0.5073602199554443
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4748353362083435,
      "epoch": 2.98,
      "learning_rate": 2.4608553455047338e-05,
      "loss": 0.4767,
      "step": 8250,
      "task_loss": 0.853038489818573
    },
    {
      "epoch": 2.98,
      "eval_exact_match": 82.20435193945127,
      "eval_f1": 89.58123307407982,
      "step": 8250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44185155630111694,
      "epoch": 2.99,
      "learning_rate": 2.4532687478384877e-05,
      "loss": 0.5337,
      "step": 8260,
      "task_loss": 0.7460501790046692
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4590781331062317,
      "epoch": 2.99,
      "learning_rate": 2.4456875474705573e-05,
      "loss": 0.5245,
      "step": 8270,
      "task_loss": 0.6354435086250305
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4764559864997864,
      "epoch": 2.99,
      "learning_rate": 2.4381117834921653e-05,
      "loss": 0.5644,
      "step": 8280,
      "task_loss": 0.44803768396377563
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4564550817012787,
      "epoch": 3.0,
      "learning_rate": 2.4305414949664922e-05,
      "loss": 0.6295,
      "step": 8290,
      "task_loss": 0.7989727258682251
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7786110639572144,
      "epoch": 3.0,
      "learning_rate": 2.422976720928497e-05,
      "loss": 0.5651,
      "step": 8300,
      "task_loss": 1.1269173622131348
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40761882066726685,
      "epoch": 3.0,
      "learning_rate": 2.415417500384695e-05,
      "loss": 0.4447,
      "step": 8310,
      "task_loss": 0.45310571789741516
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.48288601636886597,
      "epoch": 3.01,
      "learning_rate": 2.4078638723129704e-05,
      "loss": 0.4888,
      "step": 8320,
      "task_loss": 0.739985466003418
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2868315875530243,
      "epoch": 3.01,
      "learning_rate": 2.40031587566237e-05,
      "loss": 0.4326,
      "step": 8330,
      "task_loss": 0.24925296008586884
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4214899241924286,
      "epoch": 3.01,
      "learning_rate": 2.3927735493529002e-05,
      "loss": 0.4645,
      "step": 8340,
      "task_loss": 0.9287468194961548
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5637334585189819,
      "epoch": 3.02,
      "learning_rate": 2.385236932275336e-05,
      "loss": 0.4484,
      "step": 8350,
      "task_loss": 0.9525846838951111
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40637287497520447,
      "epoch": 3.02,
      "learning_rate": 2.377706063291005e-05,
      "loss": 0.4851,
      "step": 8360,
      "task_loss": 0.7950841784477234
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3816872835159302,
      "epoch": 3.02,
      "learning_rate": 2.3701809812316033e-05,
      "loss": 0.4679,
      "step": 8370,
      "task_loss": 0.6461036205291748
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4131890535354614,
      "epoch": 3.03,
      "learning_rate": 2.3626617248989822e-05,
      "loss": 0.4446,
      "step": 8380,
      "task_loss": 0.47412508726119995
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4663524627685547,
      "epoch": 3.03,
      "learning_rate": 2.3551483330649577e-05,
      "loss": 0.478,
      "step": 8390,
      "task_loss": 0.5707106590270996
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6307777166366577,
      "epoch": 3.04,
      "learning_rate": 2.3476408444711027e-05,
      "loss": 0.474,
      "step": 8400,
      "task_loss": 0.7121979594230652
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4564501941204071,
      "epoch": 3.04,
      "learning_rate": 2.3401392978285548e-05,
      "loss": 0.4281,
      "step": 8410,
      "task_loss": 0.8050240278244019
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35332757234573364,
      "epoch": 3.04,
      "learning_rate": 2.3326437318178093e-05,
      "loss": 0.461,
      "step": 8420,
      "task_loss": 0.6041104197502136
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5488748550415039,
      "epoch": 3.05,
      "learning_rate": 2.3251541850885265e-05,
      "loss": 0.4821,
      "step": 8430,
      "task_loss": 0.7619943618774414
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3080117106437683,
      "epoch": 3.05,
      "learning_rate": 2.3176706962593264e-05,
      "loss": 0.4015,
      "step": 8440,
      "task_loss": 0.44871240854263306
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5030428767204285,
      "epoch": 3.05,
      "learning_rate": 2.310193303917596e-05,
      "loss": 0.4905,
      "step": 8450,
      "task_loss": 0.7244405150413513
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36335813999176025,
      "epoch": 3.06,
      "learning_rate": 2.302722046619281e-05,
      "loss": 0.4854,
      "step": 8460,
      "task_loss": 0.7136834263801575
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3732087016105652,
      "epoch": 3.06,
      "learning_rate": 2.2952569628887004e-05,
      "loss": 0.445,
      "step": 8470,
      "task_loss": 0.5682865381240845
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5076338052749634,
      "epoch": 3.06,
      "learning_rate": 2.2877980912183336e-05,
      "loss": 0.4496,
      "step": 8480,
      "task_loss": 0.5227811336517334
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28474703431129456,
      "epoch": 3.07,
      "learning_rate": 2.2803454700686325e-05,
      "loss": 0.4638,
      "step": 8490,
      "task_loss": 0.3647550940513611
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5656081438064575,
      "epoch": 3.07,
      "learning_rate": 2.2728991378678165e-05,
      "loss": 0.4743,
      "step": 8500,
      "task_loss": 0.40611183643341064
    },
    {
      "epoch": 3.07,
      "eval_exact_match": 81.90160832544939,
      "eval_f1": 89.43148494540868,
      "step": 8500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4654257893562317,
      "epoch": 3.08,
      "learning_rate": 2.2654591330116794e-05,
      "loss": 0.4993,
      "step": 8510,
      "task_loss": 0.9065669775009155
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4588896632194519,
      "epoch": 3.08,
      "learning_rate": 2.258025493863388e-05,
      "loss": 0.4447,
      "step": 8520,
      "task_loss": 0.7351205348968506
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.451030969619751,
      "epoch": 3.08,
      "learning_rate": 2.250598258753285e-05,
      "loss": 0.4581,
      "step": 8530,
      "task_loss": 0.4265928268432617
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5207918882369995,
      "epoch": 3.09,
      "learning_rate": 2.243177465978691e-05,
      "loss": 0.4342,
      "step": 8540,
      "task_loss": 0.8488381505012512
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2967930734157562,
      "epoch": 3.09,
      "learning_rate": 2.2357631538037123e-05,
      "loss": 0.394,
      "step": 8550,
      "task_loss": 0.2662898600101471
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2869606018066406,
      "epoch": 3.09,
      "learning_rate": 2.2283553604590325e-05,
      "loss": 0.4646,
      "step": 8560,
      "task_loss": 0.317579984664917
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5217782855033875,
      "epoch": 3.1,
      "learning_rate": 2.220954124141727e-05,
      "loss": 0.43,
      "step": 8570,
      "task_loss": 0.6486431360244751
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5660613775253296,
      "epoch": 3.1,
      "learning_rate": 2.2135594830150573e-05,
      "loss": 0.489,
      "step": 8580,
      "task_loss": 0.7349064350128174
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4088625907897949,
      "epoch": 3.1,
      "learning_rate": 2.2061714752082815e-05,
      "loss": 0.4138,
      "step": 8590,
      "task_loss": 0.62070232629776
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4042903184890747,
      "epoch": 3.11,
      "learning_rate": 2.1987901388164503e-05,
      "loss": 0.433,
      "step": 8600,
      "task_loss": 0.39558160305023193
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4541303217411041,
      "epoch": 3.11,
      "learning_rate": 2.1914155119002177e-05,
      "loss": 0.5357,
      "step": 8610,
      "task_loss": 0.6944977045059204
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5707521438598633,
      "epoch": 3.12,
      "learning_rate": 2.1840476324856386e-05,
      "loss": 0.441,
      "step": 8620,
      "task_loss": 0.8816947937011719
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5475913286209106,
      "epoch": 3.12,
      "learning_rate": 2.1766865385639794e-05,
      "loss": 0.4894,
      "step": 8630,
      "task_loss": 0.7322568893432617
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4617810845375061,
      "epoch": 3.12,
      "learning_rate": 2.169332268091512e-05,
      "loss": 0.4456,
      "step": 8640,
      "task_loss": 0.43918541073799133
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3960902690887451,
      "epoch": 3.13,
      "learning_rate": 2.161984858989331e-05,
      "loss": 0.4294,
      "step": 8650,
      "task_loss": 0.6775773763656616
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.549976110458374,
      "epoch": 3.13,
      "learning_rate": 2.154644349143147e-05,
      "loss": 0.4091,
      "step": 8660,
      "task_loss": 0.8885008096694946
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4155174791812897,
      "epoch": 3.13,
      "learning_rate": 2.1473107764030978e-05,
      "loss": 0.4348,
      "step": 8670,
      "task_loss": 0.8198248744010925
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47810494899749756,
      "epoch": 3.14,
      "learning_rate": 2.1399841785835495e-05,
      "loss": 0.4994,
      "step": 8680,
      "task_loss": 0.44902503490448
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4945884346961975,
      "epoch": 3.14,
      "learning_rate": 2.1326645934629052e-05,
      "loss": 0.4552,
      "step": 8690,
      "task_loss": 1.0011985301971436
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4940509796142578,
      "epoch": 3.14,
      "learning_rate": 2.1253520587834062e-05,
      "loss": 0.4673,
      "step": 8700,
      "task_loss": 0.6739383935928345
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45873111486434937,
      "epoch": 3.15,
      "learning_rate": 2.1180466122509413e-05,
      "loss": 0.4844,
      "step": 8710,
      "task_loss": 0.9118214845657349
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5551230907440186,
      "epoch": 3.15,
      "learning_rate": 2.1107482915348477e-05,
      "loss": 0.4288,
      "step": 8720,
      "task_loss": 0.6939976215362549
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4679955244064331,
      "epoch": 3.16,
      "learning_rate": 2.1034571342677242e-05,
      "loss": 0.5021,
      "step": 8730,
      "task_loss": 0.3576204776763916
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34198206663131714,
      "epoch": 3.16,
      "learning_rate": 2.0961731780452256e-05,
      "loss": 0.4387,
      "step": 8740,
      "task_loss": 0.5113054513931274
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41313353180885315,
      "epoch": 3.16,
      "learning_rate": 2.0888964604258828e-05,
      "loss": 0.4645,
      "step": 8750,
      "task_loss": 0.6293415427207947
    },
    {
      "epoch": 3.16,
      "eval_exact_match": 82.30842005676443,
      "eval_f1": 89.4770482916315,
      "step": 8750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45705389976501465,
      "epoch": 3.17,
      "learning_rate": 2.0816270189308964e-05,
      "loss": 0.4733,
      "step": 8760,
      "task_loss": 0.9850953817367554
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3981572091579437,
      "epoch": 3.17,
      "learning_rate": 2.0743648910439537e-05,
      "loss": 0.4788,
      "step": 8770,
      "task_loss": 1.0468618869781494
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44604384899139404,
      "epoch": 3.17,
      "learning_rate": 2.0671101142110257e-05,
      "loss": 0.4668,
      "step": 8780,
      "task_loss": 0.6613199710845947
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4133979380130768,
      "epoch": 3.18,
      "learning_rate": 2.0598627258401834e-05,
      "loss": 0.482,
      "step": 8790,
      "task_loss": 0.5200196504592896
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4611911475658417,
      "epoch": 3.18,
      "learning_rate": 2.0526227633013956e-05,
      "loss": 0.4248,
      "step": 8800,
      "task_loss": 0.6944129467010498
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3930094838142395,
      "epoch": 3.18,
      "learning_rate": 2.045390263926347e-05,
      "loss": 0.4411,
      "step": 8810,
      "task_loss": 0.529641330242157
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4318530559539795,
      "epoch": 3.19,
      "learning_rate": 2.0381652650082324e-05,
      "loss": 0.5077,
      "step": 8820,
      "task_loss": 0.4306020140647888
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40004050731658936,
      "epoch": 3.19,
      "learning_rate": 2.0309478038015807e-05,
      "loss": 0.4431,
      "step": 8830,
      "task_loss": 0.5305246710777283
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4837936758995056,
      "epoch": 3.19,
      "learning_rate": 2.023737917522045e-05,
      "loss": 0.4706,
      "step": 8840,
      "task_loss": 0.4521811902523041
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4394914507865906,
      "epoch": 3.2,
      "learning_rate": 2.016535643346224e-05,
      "loss": 0.4401,
      "step": 8850,
      "task_loss": 0.5930649042129517
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3942365348339081,
      "epoch": 3.2,
      "learning_rate": 2.0093410184114667e-05,
      "loss": 0.459,
      "step": 8860,
      "task_loss": 0.3833394944667816
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4827319383621216,
      "epoch": 3.21,
      "learning_rate": 2.0021540798156785e-05,
      "loss": 0.5369,
      "step": 8870,
      "task_loss": 0.8428436517715454
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39570751786231995,
      "epoch": 3.21,
      "learning_rate": 1.9949748646171282e-05,
      "loss": 0.4456,
      "step": 8880,
      "task_loss": 1.0797185897827148
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4631249010562897,
      "epoch": 3.21,
      "learning_rate": 1.9878034098342688e-05,
      "loss": 0.4824,
      "step": 8890,
      "task_loss": 0.9745497107505798
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3670880198478699,
      "epoch": 3.22,
      "learning_rate": 1.980639752445529e-05,
      "loss": 0.4259,
      "step": 8900,
      "task_loss": 0.38064563274383545
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41449859738349915,
      "epoch": 3.22,
      "learning_rate": 1.973483929389136e-05,
      "loss": 0.4552,
      "step": 8910,
      "task_loss": 0.5112131237983704
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6296032071113586,
      "epoch": 3.22,
      "learning_rate": 1.9663359775629198e-05,
      "loss": 0.5241,
      "step": 8920,
      "task_loss": 0.9333739280700684
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3224326968193054,
      "epoch": 3.23,
      "learning_rate": 1.959195933824125e-05,
      "loss": 0.4352,
      "step": 8930,
      "task_loss": 0.7056103944778442
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.700102686882019,
      "epoch": 3.23,
      "learning_rate": 1.9520638349892158e-05,
      "loss": 0.4737,
      "step": 8940,
      "task_loss": 0.7871214151382446
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4205997884273529,
      "epoch": 3.23,
      "learning_rate": 1.9449397178336928e-05,
      "loss": 0.4576,
      "step": 8950,
      "task_loss": 0.888438880443573
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46629273891448975,
      "epoch": 3.24,
      "learning_rate": 1.9378236190919002e-05,
      "loss": 0.4272,
      "step": 8960,
      "task_loss": 0.6628609299659729
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47699084877967834,
      "epoch": 3.24,
      "learning_rate": 1.9307155754568368e-05,
      "loss": 0.5153,
      "step": 8970,
      "task_loss": 0.3555866479873657
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4512190520763397,
      "epoch": 3.25,
      "learning_rate": 1.9236156235799624e-05,
      "loss": 0.4604,
      "step": 8980,
      "task_loss": 0.7414048910140991
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34125715494155884,
      "epoch": 3.25,
      "learning_rate": 1.9165238000710218e-05,
      "loss": 0.4164,
      "step": 8990,
      "task_loss": 0.6094802618026733
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5650801658630371,
      "epoch": 3.25,
      "learning_rate": 1.9094401414978365e-05,
      "loss": 0.4677,
      "step": 9000,
      "task_loss": 0.7018455266952515
    },
    {
      "epoch": 3.25,
      "eval_exact_match": 82.35572374645223,
      "eval_f1": 89.5948549052119,
      "step": 9000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40010178089141846,
      "epoch": 3.26,
      "learning_rate": 1.9023646843861352e-05,
      "loss": 0.4391,
      "step": 9010,
      "task_loss": 0.5890183448791504
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4083946943283081,
      "epoch": 3.26,
      "learning_rate": 1.8952974652193525e-05,
      "loss": 0.4411,
      "step": 9020,
      "task_loss": 0.4983852207660675
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46902433037757874,
      "epoch": 3.26,
      "learning_rate": 1.888238520438449e-05,
      "loss": 0.4439,
      "step": 9030,
      "task_loss": 0.5505859851837158
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43851032853126526,
      "epoch": 3.27,
      "learning_rate": 1.8811878864417145e-05,
      "loss": 0.4677,
      "step": 9040,
      "task_loss": 0.7526010870933533
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36553966999053955,
      "epoch": 3.27,
      "learning_rate": 1.8741455995845906e-05,
      "loss": 0.4361,
      "step": 9050,
      "task_loss": 0.4793585240840912
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3230721354484558,
      "epoch": 3.27,
      "learning_rate": 1.8671116961794767e-05,
      "loss": 0.4381,
      "step": 9060,
      "task_loss": 0.685264527797699
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4076061546802521,
      "epoch": 3.28,
      "learning_rate": 1.860086212495545e-05,
      "loss": 0.4927,
      "step": 9070,
      "task_loss": 0.5918831825256348
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4831254780292511,
      "epoch": 3.28,
      "learning_rate": 1.8530691847585468e-05,
      "loss": 0.4676,
      "step": 9080,
      "task_loss": 0.8359205722808838
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.48880958557128906,
      "epoch": 3.29,
      "learning_rate": 1.8460606491506437e-05,
      "loss": 0.44,
      "step": 9090,
      "task_loss": 0.9667303562164307
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.572074294090271,
      "epoch": 3.29,
      "learning_rate": 1.8390606418101974e-05,
      "loss": 0.4763,
      "step": 9100,
      "task_loss": 0.7159460783004761
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28445014357566833,
      "epoch": 3.29,
      "learning_rate": 1.8320691988316008e-05,
      "loss": 0.4735,
      "step": 9110,
      "task_loss": 0.3898680806159973
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.403300404548645,
      "epoch": 3.3,
      "learning_rate": 1.8250863562650848e-05,
      "loss": 0.4487,
      "step": 9120,
      "task_loss": 0.6166889667510986
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.48287683725357056,
      "epoch": 3.3,
      "learning_rate": 1.8181121501165358e-05,
      "loss": 0.4543,
      "step": 9130,
      "task_loss": 0.7038086652755737
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4671033024787903,
      "epoch": 3.3,
      "learning_rate": 1.811146616347302e-05,
      "loss": 0.4809,
      "step": 9140,
      "task_loss": 0.6968979239463806
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40570127964019775,
      "epoch": 3.31,
      "learning_rate": 1.8041897908740205e-05,
      "loss": 0.4245,
      "step": 9150,
      "task_loss": 0.5374201536178589
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.565341055393219,
      "epoch": 3.31,
      "learning_rate": 1.7972417095684214e-05,
      "loss": 0.4669,
      "step": 9160,
      "task_loss": 0.586516261100769
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5152424573898315,
      "epoch": 3.31,
      "learning_rate": 1.790302408257151e-05,
      "loss": 0.4803,
      "step": 9170,
      "task_loss": 0.8241206407546997
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.487447053194046,
      "epoch": 3.32,
      "learning_rate": 1.7833719227215755e-05,
      "loss": 0.4344,
      "step": 9180,
      "task_loss": 0.8506260514259338
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5072003602981567,
      "epoch": 3.32,
      "learning_rate": 1.7764502886976142e-05,
      "loss": 0.4939,
      "step": 9190,
      "task_loss": 0.8282922506332397
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6491203904151917,
      "epoch": 3.32,
      "learning_rate": 1.769537541875536e-05,
      "loss": 0.4493,
      "step": 9200,
      "task_loss": 0.9028392434120178
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35966378450393677,
      "epoch": 3.33,
      "learning_rate": 1.7626337178997885e-05,
      "loss": 0.4309,
      "step": 9210,
      "task_loss": 0.4382772445678711
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44351670145988464,
      "epoch": 3.33,
      "learning_rate": 1.755738852368811e-05,
      "loss": 0.5132,
      "step": 9220,
      "task_loss": 0.5579593777656555
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5415867567062378,
      "epoch": 3.34,
      "learning_rate": 1.748852980834849e-05,
      "loss": 0.5029,
      "step": 9230,
      "task_loss": 0.8813443183898926
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5128371715545654,
      "epoch": 3.34,
      "learning_rate": 1.7419761388037698e-05,
      "loss": 0.4431,
      "step": 9240,
      "task_loss": 0.7216111421585083
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42158788442611694,
      "epoch": 3.34,
      "learning_rate": 1.7351083617348838e-05,
      "loss": 0.4622,
      "step": 9250,
      "task_loss": 0.6663612723350525
    },
    {
      "epoch": 3.34,
      "eval_exact_match": 82.61116367076632,
      "eval_f1": 89.75797445902968,
      "step": 9250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5766861438751221,
      "epoch": 3.35,
      "learning_rate": 1.7282496850407593e-05,
      "loss": 0.4458,
      "step": 9260,
      "task_loss": 0.5448006391525269
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5536984205245972,
      "epoch": 3.35,
      "learning_rate": 1.721400144087041e-05,
      "loss": 0.4782,
      "step": 9270,
      "task_loss": 0.9958784580230713
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.394481360912323,
      "epoch": 3.35,
      "learning_rate": 1.714559774192262e-05,
      "loss": 0.4619,
      "step": 9280,
      "task_loss": 0.7101040482521057
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.356433629989624,
      "epoch": 3.36,
      "learning_rate": 1.707728610627674e-05,
      "loss": 0.4389,
      "step": 9290,
      "task_loss": 0.619407057762146
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4516601264476776,
      "epoch": 3.36,
      "learning_rate": 1.7009066886170497e-05,
      "loss": 0.484,
      "step": 9300,
      "task_loss": 0.5669988393783569
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4672652781009674,
      "epoch": 3.36,
      "learning_rate": 1.6940940433365148e-05,
      "loss": 0.4677,
      "step": 9310,
      "task_loss": 0.8607810735702515
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33394983410835266,
      "epoch": 3.37,
      "learning_rate": 1.6872907099143585e-05,
      "loss": 0.4618,
      "step": 9320,
      "task_loss": 0.43108028173446655
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5350024700164795,
      "epoch": 3.37,
      "learning_rate": 1.6804967234308577e-05,
      "loss": 0.5019,
      "step": 9330,
      "task_loss": 0.9354665279388428
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5334118604660034,
      "epoch": 3.38,
      "learning_rate": 1.673712118918088e-05,
      "loss": 0.4638,
      "step": 9340,
      "task_loss": 0.7986032962799072
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3785496652126312,
      "epoch": 3.38,
      "learning_rate": 1.6669369313597535e-05,
      "loss": 0.4518,
      "step": 9350,
      "task_loss": 0.8192988038063049
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4895620346069336,
      "epoch": 3.38,
      "learning_rate": 1.6601711956909998e-05,
      "loss": 0.4556,
      "step": 9360,
      "task_loss": 0.7032527923583984
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41272759437561035,
      "epoch": 3.39,
      "learning_rate": 1.653414946798235e-05,
      "loss": 0.5252,
      "step": 9370,
      "task_loss": 0.4573909044265747
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4580502510070801,
      "epoch": 3.39,
      "learning_rate": 1.646668219518948e-05,
      "loss": 0.4687,
      "step": 9380,
      "task_loss": 0.6156324148178101
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34736135601997375,
      "epoch": 3.39,
      "learning_rate": 1.639931048641538e-05,
      "loss": 0.4566,
      "step": 9390,
      "task_loss": 0.42701518535614014
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36195576190948486,
      "epoch": 3.4,
      "learning_rate": 1.6332034689051194e-05,
      "loss": 0.4718,
      "step": 9400,
      "task_loss": 0.3998080790042877
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3932255804538727,
      "epoch": 3.4,
      "learning_rate": 1.6264855149993574e-05,
      "loss": 0.4301,
      "step": 9410,
      "task_loss": 0.7975203990936279
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40939027070999146,
      "epoch": 3.4,
      "learning_rate": 1.6197772215642807e-05,
      "loss": 0.4091,
      "step": 9420,
      "task_loss": 0.643854022026062
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39888107776641846,
      "epoch": 3.41,
      "learning_rate": 1.6130786231901073e-05,
      "loss": 0.3992,
      "step": 9430,
      "task_loss": 0.543474555015564
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46289771795272827,
      "epoch": 3.41,
      "learning_rate": 1.606389754417061e-05,
      "loss": 0.4787,
      "step": 9440,
      "task_loss": 0.9618111848831177
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4148654341697693,
      "epoch": 3.42,
      "learning_rate": 1.5997106497351993e-05,
      "loss": 0.4771,
      "step": 9450,
      "task_loss": 0.5072375535964966
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49600380659103394,
      "epoch": 3.42,
      "learning_rate": 1.593041343584232e-05,
      "loss": 0.4332,
      "step": 9460,
      "task_loss": 0.7816798686981201
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34944719076156616,
      "epoch": 3.42,
      "learning_rate": 1.5863818703533445e-05,
      "loss": 0.453,
      "step": 9470,
      "task_loss": 0.6673324108123779
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4825136959552765,
      "epoch": 3.43,
      "learning_rate": 1.5797322643810196e-05,
      "loss": 0.4553,
      "step": 9480,
      "task_loss": 0.6439677476882935
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5216967463493347,
      "epoch": 3.43,
      "learning_rate": 1.5730925599548637e-05,
      "loss": 0.4688,
      "step": 9490,
      "task_loss": 0.5796299576759338
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.504321277141571,
      "epoch": 3.43,
      "learning_rate": 1.5664627913114222e-05,
      "loss": 0.4935,
      "step": 9500,
      "task_loss": 0.5632234215736389
    },
    {
      "epoch": 3.43,
      "eval_exact_match": 82.52601702932829,
      "eval_f1": 89.72896108835486,
      "step": 9500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37664616107940674,
      "epoch": 3.44,
      "learning_rate": 1.5598429926360136e-05,
      "loss": 0.4524,
      "step": 9510,
      "task_loss": 0.47461533546447754
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38051480054855347,
      "epoch": 3.44,
      "learning_rate": 1.5532331980625454e-05,
      "loss": 0.4261,
      "step": 9520,
      "task_loss": 0.6248530149459839
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4424772262573242,
      "epoch": 3.44,
      "learning_rate": 1.5466334416733425e-05,
      "loss": 0.4905,
      "step": 9530,
      "task_loss": 0.7271965146064758
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3982722759246826,
      "epoch": 3.45,
      "learning_rate": 1.5400437574989648e-05,
      "loss": 0.4329,
      "step": 9540,
      "task_loss": 0.5037179589271545
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5665621757507324,
      "epoch": 3.45,
      "learning_rate": 1.5334641795180442e-05,
      "loss": 0.4662,
      "step": 9550,
      "task_loss": 0.8826358318328857
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4354783892631531,
      "epoch": 3.46,
      "learning_rate": 1.5268947416570933e-05,
      "loss": 0.4084,
      "step": 9560,
      "task_loss": 0.9508954286575317
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5282255411148071,
      "epoch": 3.46,
      "learning_rate": 1.5203354777903448e-05,
      "loss": 0.4618,
      "step": 9570,
      "task_loss": 0.6860237121582031
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3957028090953827,
      "epoch": 3.46,
      "learning_rate": 1.5137864217395681e-05,
      "loss": 0.4416,
      "step": 9580,
      "task_loss": 0.7152292728424072
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3824961483478546,
      "epoch": 3.47,
      "learning_rate": 1.5072476072739005e-05,
      "loss": 0.4534,
      "step": 9590,
      "task_loss": 0.379571795463562
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34231045842170715,
      "epoch": 3.47,
      "learning_rate": 1.5007190681096652e-05,
      "loss": 0.4064,
      "step": 9600,
      "task_loss": 0.44205284118652344
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3934401869773865,
      "epoch": 3.47,
      "learning_rate": 1.4942008379102068e-05,
      "loss": 0.5002,
      "step": 9610,
      "task_loss": 0.36241433024406433
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35561633110046387,
      "epoch": 3.48,
      "learning_rate": 1.4876929502857115e-05,
      "loss": 0.4556,
      "step": 9620,
      "task_loss": 0.6349127292633057
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30296555161476135,
      "epoch": 3.48,
      "learning_rate": 1.4811954387930384e-05,
      "loss": 0.4188,
      "step": 9630,
      "task_loss": 0.9882563352584839
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2964661717414856,
      "epoch": 3.48,
      "learning_rate": 1.474708336935538e-05,
      "loss": 0.4593,
      "step": 9640,
      "task_loss": 0.3808435797691345
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45735228061676025,
      "epoch": 3.49,
      "learning_rate": 1.4682316781628928e-05,
      "loss": 0.4316,
      "step": 9650,
      "task_loss": 0.5127953886985779
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29700976610183716,
      "epoch": 3.49,
      "learning_rate": 1.4617654958709308e-05,
      "loss": 0.443,
      "step": 9660,
      "task_loss": 0.3418108820915222
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4318172037601471,
      "epoch": 3.49,
      "learning_rate": 1.4553098234014623e-05,
      "loss": 0.4149,
      "step": 9670,
      "task_loss": 0.5029164552688599
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42593926191329956,
      "epoch": 3.5,
      "learning_rate": 1.4488646940421064e-05,
      "loss": 0.4384,
      "step": 9680,
      "task_loss": 0.6561027765274048
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4087527394294739,
      "epoch": 3.5,
      "learning_rate": 1.4424301410261182e-05,
      "loss": 0.4536,
      "step": 9690,
      "task_loss": 0.29617658257484436
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3533773124217987,
      "epoch": 3.51,
      "learning_rate": 1.4360061975322142e-05,
      "loss": 0.4615,
      "step": 9700,
      "task_loss": 0.45025789737701416
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4536002278327942,
      "epoch": 3.51,
      "learning_rate": 1.4295928966844085e-05,
      "loss": 0.4683,
      "step": 9710,
      "task_loss": 0.6415740251541138
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38968485593795776,
      "epoch": 3.51,
      "learning_rate": 1.423190271551837e-05,
      "loss": 0.4867,
      "step": 9720,
      "task_loss": 0.3250332474708557
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38814741373062134,
      "epoch": 3.52,
      "learning_rate": 1.4167983551485887e-05,
      "loss": 0.4569,
      "step": 9730,
      "task_loss": 0.923711895942688
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4382472038269043,
      "epoch": 3.52,
      "learning_rate": 1.4104171804335311e-05,
      "loss": 0.4405,
      "step": 9740,
      "task_loss": 0.7725265026092529
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4493694305419922,
      "epoch": 3.52,
      "learning_rate": 1.404046780310151e-05,
      "loss": 0.4066,
      "step": 9750,
      "task_loss": 0.39111417531967163
    },
    {
      "epoch": 3.52,
      "eval_exact_match": 82.35572374645223,
      "eval_f1": 89.55060641123724,
      "step": 9750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4694647789001465,
      "epoch": 3.53,
      "learning_rate": 1.39768718762637e-05,
      "loss": 0.5056,
      "step": 9760,
      "task_loss": 0.6846310496330261
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5013730525970459,
      "epoch": 3.53,
      "learning_rate": 1.391338435174388e-05,
      "loss": 0.4405,
      "step": 9770,
      "task_loss": 0.47695648670196533
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36243951320648193,
      "epoch": 3.53,
      "learning_rate": 1.3850005556905072e-05,
      "loss": 0.4278,
      "step": 9780,
      "task_loss": 0.41346102952957153
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.500864565372467,
      "epoch": 3.54,
      "learning_rate": 1.3786735818549667e-05,
      "loss": 0.4443,
      "step": 9790,
      "task_loss": 0.7879412770271301
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4714551866054535,
      "epoch": 3.54,
      "learning_rate": 1.372357546291769e-05,
      "loss": 0.4504,
      "step": 9800,
      "task_loss": 0.5894259214401245
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.490497887134552,
      "epoch": 3.55,
      "learning_rate": 1.3660524815685188e-05,
      "loss": 0.4955,
      "step": 9810,
      "task_loss": 0.5708312392234802
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5023337602615356,
      "epoch": 3.55,
      "learning_rate": 1.359758420196249e-05,
      "loss": 0.4816,
      "step": 9820,
      "task_loss": 0.570620059967041
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5660896301269531,
      "epoch": 3.55,
      "learning_rate": 1.3534753946292586e-05,
      "loss": 0.4025,
      "step": 9830,
      "task_loss": 0.5451761484146118
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3981624245643616,
      "epoch": 3.56,
      "learning_rate": 1.347203437264936e-05,
      "loss": 0.4706,
      "step": 9840,
      "task_loss": 0.7182565927505493
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5007762908935547,
      "epoch": 3.56,
      "learning_rate": 1.3409425804436078e-05,
      "loss": 0.4933,
      "step": 9850,
      "task_loss": 0.520271360874176
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36946970224380493,
      "epoch": 3.56,
      "learning_rate": 1.3346928564483535e-05,
      "loss": 0.477,
      "step": 9860,
      "task_loss": 0.9104056358337402
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4460700750350952,
      "epoch": 3.57,
      "learning_rate": 1.3284542975048519e-05,
      "loss": 0.4966,
      "step": 9870,
      "task_loss": 0.5107903480529785
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40421903133392334,
      "epoch": 3.57,
      "learning_rate": 1.3222269357812115e-05,
      "loss": 0.4479,
      "step": 9880,
      "task_loss": 0.3044702708721161
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37104153633117676,
      "epoch": 3.57,
      "learning_rate": 1.3160108033878046e-05,
      "loss": 0.3931,
      "step": 9890,
      "task_loss": 0.5892044305801392
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.572177529335022,
      "epoch": 3.58,
      "learning_rate": 1.3098059323770975e-05,
      "loss": 0.4358,
      "step": 9900,
      "task_loss": 0.8855876922607422
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4747053384780884,
      "epoch": 3.58,
      "learning_rate": 1.3036123547434927e-05,
      "loss": 0.4494,
      "step": 9910,
      "task_loss": 0.758881688117981
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4006056785583496,
      "epoch": 3.59,
      "learning_rate": 1.2974301024231595e-05,
      "loss": 0.4463,
      "step": 9920,
      "task_loss": 0.4315032362937927
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41052359342575073,
      "epoch": 3.59,
      "learning_rate": 1.2912592072938709e-05,
      "loss": 0.4069,
      "step": 9930,
      "task_loss": 0.3629281520843506
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37371566891670227,
      "epoch": 3.59,
      "learning_rate": 1.2850997011748333e-05,
      "loss": 0.4324,
      "step": 9940,
      "task_loss": 0.8833544254302979
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5907098650932312,
      "epoch": 3.6,
      "learning_rate": 1.2789516158265369e-05,
      "loss": 0.4559,
      "step": 9950,
      "task_loss": 0.8144785165786743
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4753618538379669,
      "epoch": 3.6,
      "learning_rate": 1.272814982950573e-05,
      "loss": 0.4728,
      "step": 9960,
      "task_loss": 0.9171135425567627
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3954315781593323,
      "epoch": 3.6,
      "learning_rate": 1.2666898341894853e-05,
      "loss": 0.4018,
      "step": 9970,
      "task_loss": 0.590205192565918
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3371639549732208,
      "epoch": 3.61,
      "learning_rate": 1.2605762011266012e-05,
      "loss": 0.4689,
      "step": 9980,
      "task_loss": 0.644167423248291
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35054704546928406,
      "epoch": 3.61,
      "learning_rate": 1.2544741152858692e-05,
      "loss": 0.4749,
      "step": 9990,
      "task_loss": 0.4521157145500183
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45311444997787476,
      "epoch": 3.61,
      "learning_rate": 1.248383608131694e-05,
      "loss": 0.4068,
      "step": 10000,
      "task_loss": 0.718034029006958
    },
    {
      "epoch": 3.61,
      "eval_exact_match": 82.56385998107852,
      "eval_f1": 89.63155210704089,
      "step": 10000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2593729794025421,
      "epoch": 3.62,
      "learning_rate": 1.2423047110687794e-05,
      "loss": 0.428,
      "step": 10010,
      "task_loss": 0.4959852695465088
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5165232419967651,
      "epoch": 3.62,
      "learning_rate": 1.2362374554419625e-05,
      "loss": 0.4786,
      "step": 10020,
      "task_loss": 0.510176420211792
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4051707088947296,
      "epoch": 3.62,
      "learning_rate": 1.2301818725360548e-05,
      "loss": 0.4223,
      "step": 10030,
      "task_loss": 0.7712464332580566
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44198185205459595,
      "epoch": 3.63,
      "learning_rate": 1.2241379935756749e-05,
      "loss": 0.4015,
      "step": 10040,
      "task_loss": 0.7498027086257935
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4870222210884094,
      "epoch": 3.63,
      "learning_rate": 1.2181058497250998e-05,
      "loss": 0.4714,
      "step": 10050,
      "task_loss": 0.29204386472702026
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3149113059043884,
      "epoch": 3.64,
      "learning_rate": 1.2120854720880877e-05,
      "loss": 0.4304,
      "step": 10060,
      "task_loss": 0.31059128046035767
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40889161825180054,
      "epoch": 3.64,
      "learning_rate": 1.206076891707731e-05,
      "loss": 0.4754,
      "step": 10070,
      "task_loss": 0.5112392902374268
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40656381845474243,
      "epoch": 3.64,
      "learning_rate": 1.2000801395662918e-05,
      "loss": 0.4317,
      "step": 10080,
      "task_loss": 0.867332398891449
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32227474451065063,
      "epoch": 3.65,
      "learning_rate": 1.1940952465850405e-05,
      "loss": 0.4197,
      "step": 10090,
      "task_loss": 0.4937788248062134
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40178176760673523,
      "epoch": 3.65,
      "learning_rate": 1.1881222436240966e-05,
      "loss": 0.4706,
      "step": 10100,
      "task_loss": 0.5314730405807495
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3422333598136902,
      "epoch": 3.65,
      "learning_rate": 1.1821611614822724e-05,
      "loss": 0.4499,
      "step": 10110,
      "task_loss": 0.6373083591461182
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4214498996734619,
      "epoch": 3.66,
      "learning_rate": 1.176212030896912e-05,
      "loss": 0.4645,
      "step": 10120,
      "task_loss": 0.6125006079673767
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4471798539161682,
      "epoch": 3.66,
      "learning_rate": 1.1702748825437348e-05,
      "loss": 0.4203,
      "step": 10130,
      "task_loss": 0.8206405639648438
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5667499303817749,
      "epoch": 3.66,
      "learning_rate": 1.16434974703667e-05,
      "loss": 0.4428,
      "step": 10140,
      "task_loss": 0.7912921905517578
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44536837935447693,
      "epoch": 3.67,
      "learning_rate": 1.1584366549277138e-05,
      "loss": 0.4994,
      "step": 10150,
      "task_loss": 0.7000430226325989
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43894514441490173,
      "epoch": 3.67,
      "learning_rate": 1.1525356367067541e-05,
      "loss": 0.4713,
      "step": 10160,
      "task_loss": 0.7236070036888123
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4418973922729492,
      "epoch": 3.68,
      "learning_rate": 1.1466467228014262e-05,
      "loss": 0.4531,
      "step": 10170,
      "task_loss": 1.1132642030715942
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47217002511024475,
      "epoch": 3.68,
      "learning_rate": 1.14076994357695e-05,
      "loss": 0.457,
      "step": 10180,
      "task_loss": 0.7230472564697266
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3841246962547302,
      "epoch": 3.68,
      "learning_rate": 1.134905329335976e-05,
      "loss": 0.5203,
      "step": 10190,
      "task_loss": 0.5335608720779419
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45422253012657166,
      "epoch": 3.69,
      "learning_rate": 1.1290529103184282e-05,
      "loss": 0.4765,
      "step": 10200,
      "task_loss": 0.7611352801322937
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.511451005935669,
      "epoch": 3.69,
      "learning_rate": 1.1232127167013436e-05,
      "loss": 0.4526,
      "step": 10210,
      "task_loss": 0.935626208782196
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4894786477088928,
      "epoch": 3.69,
      "learning_rate": 1.1173847785987288e-05,
      "loss": 0.4566,
      "step": 10220,
      "task_loss": 0.316392183303833
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.56304931640625,
      "epoch": 3.7,
      "learning_rate": 1.1115691260613889e-05,
      "loss": 0.4464,
      "step": 10230,
      "task_loss": 0.7091505527496338
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3679378628730774,
      "epoch": 3.7,
      "learning_rate": 1.1057657890767854e-05,
      "loss": 0.4543,
      "step": 10240,
      "task_loss": 0.48593321442604065
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44522807002067566,
      "epoch": 3.7,
      "learning_rate": 1.0999747975688754e-05,
      "loss": 0.4578,
      "step": 10250,
      "task_loss": 0.6225043535232544
    },
    {
      "epoch": 3.7,
      "eval_exact_match": 82.55439924314096,
      "eval_f1": 89.67409457937354,
      "step": 10250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40633970499038696,
      "epoch": 3.71,
      "learning_rate": 1.094196181397959e-05,
      "loss": 0.4626,
      "step": 10260,
      "task_loss": 0.4013146758079529
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39074042439460754,
      "epoch": 3.71,
      "learning_rate": 1.0884299703605228e-05,
      "loss": 0.4468,
      "step": 10270,
      "task_loss": 0.5073047876358032
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.48598477244377136,
      "epoch": 3.72,
      "learning_rate": 1.0826761941890906e-05,
      "loss": 0.4369,
      "step": 10280,
      "task_loss": 0.933097243309021
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5473797917366028,
      "epoch": 3.72,
      "learning_rate": 1.0769348825520669e-05,
      "loss": 0.4574,
      "step": 10290,
      "task_loss": 1.121228575706482
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34773969650268555,
      "epoch": 3.72,
      "learning_rate": 1.0712060650535867e-05,
      "loss": 0.4501,
      "step": 10300,
      "task_loss": 0.27765679359436035
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.421405553817749,
      "epoch": 3.73,
      "learning_rate": 1.0654897712333557e-05,
      "loss": 0.5029,
      "step": 10310,
      "task_loss": 0.7494446039199829
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5293338298797607,
      "epoch": 3.73,
      "learning_rate": 1.059786030566512e-05,
      "loss": 0.4723,
      "step": 10320,
      "task_loss": 1.1732032299041748
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3712977468967438,
      "epoch": 3.73,
      "learning_rate": 1.0540948724634564e-05,
      "loss": 0.4501,
      "step": 10330,
      "task_loss": 0.8228405714035034
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43470335006713867,
      "epoch": 3.74,
      "learning_rate": 1.0484163262697156e-05,
      "loss": 0.4082,
      "step": 10340,
      "task_loss": 0.4585592746734619
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34602972865104675,
      "epoch": 3.74,
      "learning_rate": 1.042750421265783e-05,
      "loss": 0.4309,
      "step": 10350,
      "task_loss": 0.63847815990448
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3850482702255249,
      "epoch": 3.74,
      "learning_rate": 1.0370971866669712e-05,
      "loss": 0.4563,
      "step": 10360,
      "task_loss": 0.3458658456802368
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4272826910018921,
      "epoch": 3.75,
      "learning_rate": 1.0314566516232566e-05,
      "loss": 0.4322,
      "step": 10370,
      "task_loss": 0.5228935480117798
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42405521869659424,
      "epoch": 3.75,
      "learning_rate": 1.0263910522432138e-05,
      "loss": 0.4398,
      "step": 10380,
      "task_loss": 0.7290164232254028
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4163869023323059,
      "epoch": 3.75,
      "learning_rate": 1.020774726427728e-05,
      "loss": 0.4492,
      "step": 10390,
      "task_loss": 0.8367542028427124
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6735716462135315,
      "epoch": 3.76,
      "learning_rate": 1.0151711843314342e-05,
      "loss": 0.4848,
      "step": 10400,
      "task_loss": 0.6577698588371277
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40833598375320435,
      "epoch": 3.76,
      "learning_rate": 1.0095804548480798e-05,
      "loss": 0.4422,
      "step": 10410,
      "task_loss": 0.502683162689209
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3372454047203064,
      "epoch": 3.77,
      "learning_rate": 1.004002566805347e-05,
      "loss": 0.4176,
      "step": 10420,
      "task_loss": 0.6605957746505737
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4468865990638733,
      "epoch": 3.77,
      "learning_rate": 9.984375489647e-06,
      "loss": 0.443,
      "step": 10430,
      "task_loss": 0.3105127215385437
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40403956174850464,
      "epoch": 3.77,
      "learning_rate": 9.92885430021243e-06,
      "loss": 0.5085,
      "step": 10440,
      "task_loss": 0.5763041973114014
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35261040925979614,
      "epoch": 3.78,
      "learning_rate": 9.87346238603569e-06,
      "loss": 0.4012,
      "step": 10450,
      "task_loss": 0.4983939826488495
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43633800745010376,
      "epoch": 3.78,
      "learning_rate": 9.81820003273612e-06,
      "loss": 0.4526,
      "step": 10460,
      "task_loss": 0.5212877988815308
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4117966294288635,
      "epoch": 3.78,
      "learning_rate": 9.763067525264964e-06,
      "loss": 0.4783,
      "step": 10470,
      "task_loss": 0.40295490622520447
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3620610237121582,
      "epoch": 3.79,
      "learning_rate": 9.708065147904013e-06,
      "loss": 0.4127,
      "step": 10480,
      "task_loss": 0.4528779983520508
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4549371302127838,
      "epoch": 3.79,
      "learning_rate": 9.653193184263991e-06,
      "loss": 0.4901,
      "step": 10490,
      "task_loss": 0.4713435769081116
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4598999619483948,
      "epoch": 3.79,
      "learning_rate": 9.598451917283206e-06,
      "loss": 0.4689,
      "step": 10500,
      "task_loss": 0.7843847274780273
    },
    {
      "epoch": 3.79,
      "eval_exact_match": 82.639545884579,
      "eval_f1": 89.74238716713347,
      "step": 10500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4294615387916565,
      "epoch": 3.8,
      "learning_rate": 9.543841629226033e-06,
      "loss": 0.4288,
      "step": 10510,
      "task_loss": 0.38632312417030334
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3702619969844818,
      "epoch": 3.8,
      "learning_rate": 9.489362601681498e-06,
      "loss": 0.4508,
      "step": 10520,
      "task_loss": 0.6389187574386597
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5605703592300415,
      "epoch": 3.81,
      "learning_rate": 9.435015115561758e-06,
      "loss": 0.4653,
      "step": 10530,
      "task_loss": 0.8252593278884888
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29313477873802185,
      "epoch": 3.81,
      "learning_rate": 9.380799451100741e-06,
      "loss": 0.5044,
      "step": 10540,
      "task_loss": 0.4142262637615204
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36696240305900574,
      "epoch": 3.81,
      "learning_rate": 9.326715887852645e-06,
      "loss": 0.4497,
      "step": 10550,
      "task_loss": 0.4621296525001526
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5388485193252563,
      "epoch": 3.82,
      "learning_rate": 9.272764704690518e-06,
      "loss": 0.4476,
      "step": 10560,
      "task_loss": 0.5078322291374207
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38809970021247864,
      "epoch": 3.82,
      "learning_rate": 9.218946179804766e-06,
      "loss": 0.4435,
      "step": 10570,
      "task_loss": 0.4280998706817627
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5328583717346191,
      "epoch": 3.82,
      "learning_rate": 9.16526059070185e-06,
      "loss": 0.4354,
      "step": 10580,
      "task_loss": 0.7839690446853638
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41611891984939575,
      "epoch": 3.83,
      "learning_rate": 9.11170821420269e-06,
      "loss": 0.5341,
      "step": 10590,
      "task_loss": 0.4555775821208954
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4526450037956238,
      "epoch": 3.83,
      "learning_rate": 9.058289326441349e-06,
      "loss": 0.442,
      "step": 10600,
      "task_loss": 0.5670619606971741
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4477660357952118,
      "epoch": 3.83,
      "learning_rate": 9.005004202863581e-06,
      "loss": 0.4255,
      "step": 10610,
      "task_loss": 1.0125763416290283
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5518783330917358,
      "epoch": 3.84,
      "learning_rate": 8.951853118225413e-06,
      "loss": 0.5069,
      "step": 10620,
      "task_loss": 0.4689984619617462
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39925044775009155,
      "epoch": 3.84,
      "learning_rate": 8.898836346591686e-06,
      "loss": 0.4361,
      "step": 10630,
      "task_loss": 0.7678056955337524
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40966880321502686,
      "epoch": 3.85,
      "learning_rate": 8.845954161334712e-06,
      "loss": 0.4182,
      "step": 10640,
      "task_loss": 0.4853663742542267
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5304992198944092,
      "epoch": 3.85,
      "learning_rate": 8.793206835132823e-06,
      "loss": 0.4547,
      "step": 10650,
      "task_loss": 0.6776725053787231
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3908195495605469,
      "epoch": 3.85,
      "learning_rate": 8.74059463996898e-06,
      "loss": 0.4477,
      "step": 10660,
      "task_loss": 0.4654383063316345
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42154234647750854,
      "epoch": 3.86,
      "learning_rate": 8.688117847129323e-06,
      "loss": 0.4834,
      "step": 10670,
      "task_loss": 0.6709318161010742
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4247440695762634,
      "epoch": 3.86,
      "learning_rate": 8.635776727201879e-06,
      "loss": 0.4722,
      "step": 10680,
      "task_loss": 0.5344759225845337
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3206929564476013,
      "epoch": 3.86,
      "learning_rate": 8.583571550075038e-06,
      "loss": 0.4014,
      "step": 10690,
      "task_loss": 0.5088073015213013
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3521346151828766,
      "epoch": 3.87,
      "learning_rate": 8.531502584936257e-06,
      "loss": 0.4669,
      "step": 10700,
      "task_loss": 0.546211838722229
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43378379940986633,
      "epoch": 3.87,
      "learning_rate": 8.479570100270628e-06,
      "loss": 0.4134,
      "step": 10710,
      "task_loss": 0.3153116703033447
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3812393844127655,
      "epoch": 3.87,
      "learning_rate": 8.427774363859511e-06,
      "loss": 0.4282,
      "step": 10720,
      "task_loss": 0.4867439568042755
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47613561153411865,
      "epoch": 3.88,
      "learning_rate": 8.37611564277913e-06,
      "loss": 0.4332,
      "step": 10730,
      "task_loss": 0.5595312714576721
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.384829044342041,
      "epoch": 3.88,
      "learning_rate": 8.324594203399223e-06,
      "loss": 0.4575,
      "step": 10740,
      "task_loss": 0.8852812051773071
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5001972913742065,
      "epoch": 3.89,
      "learning_rate": 8.273210311381659e-06,
      "loss": 0.4724,
      "step": 10750,
      "task_loss": 0.8678057193756104
    },
    {
      "epoch": 3.89,
      "eval_exact_match": 82.64900662251655,
      "eval_f1": 89.66552807530321,
      "step": 10750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5500179529190063,
      "epoch": 3.89,
      "learning_rate": 8.221964231679075e-06,
      "loss": 0.4497,
      "step": 10760,
      "task_loss": 0.6250838041305542
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3807331919670105,
      "epoch": 3.89,
      "learning_rate": 8.170856228533461e-06,
      "loss": 0.4349,
      "step": 10770,
      "task_loss": 0.48087456822395325
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4343903958797455,
      "epoch": 3.9,
      "learning_rate": 8.119886565474906e-06,
      "loss": 0.438,
      "step": 10780,
      "task_loss": 0.7211394309997559
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.291828989982605,
      "epoch": 3.9,
      "learning_rate": 8.069055505320102e-06,
      "loss": 0.4236,
      "step": 10790,
      "task_loss": 0.4594219923019409
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4952172338962555,
      "epoch": 3.9,
      "learning_rate": 8.018363310171097e-06,
      "loss": 0.4617,
      "step": 10800,
      "task_loss": 0.6142617464065552
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3711337447166443,
      "epoch": 3.91,
      "learning_rate": 7.9678102414139e-06,
      "loss": 0.4719,
      "step": 10810,
      "task_loss": 0.5707615613937378
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34765708446502686,
      "epoch": 3.91,
      "learning_rate": 7.917396559717134e-06,
      "loss": 0.4705,
      "step": 10820,
      "task_loss": 0.5658267140388489
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3544590473175049,
      "epoch": 3.91,
      "learning_rate": 7.867122525030685e-06,
      "loss": 0.4342,
      "step": 10830,
      "task_loss": 1.0905160903930664
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44853660464286804,
      "epoch": 3.92,
      "learning_rate": 7.816988396584382e-06,
      "loss": 0.4501,
      "step": 10840,
      "task_loss": 0.6044590473175049
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4930655360221863,
      "epoch": 3.92,
      "learning_rate": 7.766994432886658e-06,
      "loss": 0.4838,
      "step": 10850,
      "task_loss": 0.8701499104499817
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3618244528770447,
      "epoch": 3.92,
      "learning_rate": 7.7171408917232e-06,
      "loss": 0.4294,
      "step": 10860,
      "task_loss": 0.5858584046363831
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5228819847106934,
      "epoch": 3.93,
      "learning_rate": 7.667428030155615e-06,
      "loss": 0.4462,
      "step": 10870,
      "task_loss": 0.40236058831214905
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39159971475601196,
      "epoch": 3.93,
      "learning_rate": 7.61785610452016e-06,
      "loss": 0.468,
      "step": 10880,
      "task_loss": 0.874620795249939
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.437158465385437,
      "epoch": 3.94,
      "learning_rate": 7.568425370426332e-06,
      "loss": 0.4786,
      "step": 10890,
      "task_loss": 0.7076048254966736
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4604962170124054,
      "epoch": 3.94,
      "learning_rate": 7.5191360827556245e-06,
      "loss": 0.4425,
      "step": 10900,
      "task_loss": 0.425723135471344
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40846937894821167,
      "epoch": 3.94,
      "learning_rate": 7.4699884956601825e-06,
      "loss": 0.4462,
      "step": 10910,
      "task_loss": 0.376265287399292
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5597336888313293,
      "epoch": 3.95,
      "learning_rate": 7.420982862561493e-06,
      "loss": 0.4404,
      "step": 10920,
      "task_loss": 0.7780491709709167
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3314722776412964,
      "epoch": 3.95,
      "learning_rate": 7.372119436149068e-06,
      "loss": 0.4198,
      "step": 10930,
      "task_loss": 0.6404004096984863
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41364434361457825,
      "epoch": 3.95,
      "learning_rate": 7.323398468379176e-06,
      "loss": 0.446,
      "step": 10940,
      "task_loss": 0.7026541829109192
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.512392520904541,
      "epoch": 3.96,
      "learning_rate": 7.2748202104735e-06,
      "loss": 0.4714,
      "step": 10950,
      "task_loss": 0.7072031497955322
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2855644226074219,
      "epoch": 3.96,
      "learning_rate": 7.226384912917893e-06,
      "loss": 0.3969,
      "step": 10960,
      "task_loss": 0.496487557888031
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5489861965179443,
      "epoch": 3.96,
      "learning_rate": 7.178092825461004e-06,
      "loss": 0.4441,
      "step": 10970,
      "task_loss": 0.6558412313461304
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3471415042877197,
      "epoch": 3.97,
      "learning_rate": 7.129944197113112e-06,
      "loss": 0.4202,
      "step": 10980,
      "task_loss": 0.5004458427429199
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5401880741119385,
      "epoch": 3.97,
      "learning_rate": 7.081939276144708e-06,
      "loss": 0.4106,
      "step": 10990,
      "task_loss": 0.4734864830970764
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49857717752456665,
      "epoch": 3.98,
      "learning_rate": 7.034078310085315e-06,
      "loss": 0.4135,
      "step": 11000,
      "task_loss": 0.8906891942024231
    },
    {
      "epoch": 3.98,
      "eval_exact_match": 82.34626300851467,
      "eval_f1": 89.59125330964915,
      "step": 11000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38503700494766235,
      "epoch": 3.98,
      "learning_rate": 6.986361545722173e-06,
      "loss": 0.4515,
      "step": 11010,
      "task_loss": 0.42232775688171387
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.322620689868927,
      "epoch": 3.98,
      "learning_rate": 6.938789229098967e-06,
      "loss": 0.4919,
      "step": 11020,
      "task_loss": 0.25172334909439087
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5034948587417603,
      "epoch": 3.99,
      "learning_rate": 6.8913616055145456e-06,
      "loss": 0.4706,
      "step": 11030,
      "task_loss": 0.6424163579940796
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4091174602508545,
      "epoch": 3.99,
      "learning_rate": 6.8440789195216915e-06,
      "loss": 0.4358,
      "step": 11040,
      "task_loss": 0.3128480315208435
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43555712699890137,
      "epoch": 3.99,
      "learning_rate": 6.796941414925831e-06,
      "loss": 0.411,
      "step": 11050,
      "task_loss": 0.4940514862537384
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4541236162185669,
      "epoch": 4.0,
      "learning_rate": 6.7499493347838e-06,
      "loss": 0.395,
      "step": 11060,
      "task_loss": 0.4762210249900818
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35681986808776855,
      "epoch": 4.0,
      "learning_rate": 6.703102921402538e-06,
      "loss": 0.4215,
      "step": 11070,
      "task_loss": 0.577312171459198
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5038410425186157,
      "epoch": 4.0,
      "learning_rate": 6.656402416337944e-06,
      "loss": 0.4241,
      "step": 11080,
      "task_loss": 0.5972139835357666
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3108671307563782,
      "epoch": 4.01,
      "learning_rate": 6.609848060393494e-06,
      "loss": 0.3705,
      "step": 11090,
      "task_loss": 0.5614968538284302
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5272907614707947,
      "epoch": 4.01,
      "learning_rate": 6.563440093619115e-06,
      "loss": 0.4174,
      "step": 11100,
      "task_loss": 1.0142651796340942
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4231366515159607,
      "epoch": 4.02,
      "learning_rate": 6.5171787553098855e-06,
      "loss": 0.3767,
      "step": 11110,
      "task_loss": 0.5615295171737671
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37049567699432373,
      "epoch": 4.02,
      "learning_rate": 6.471064284004835e-06,
      "loss": 0.4511,
      "step": 11120,
      "task_loss": 0.5118829011917114
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4037175476551056,
      "epoch": 4.02,
      "learning_rate": 6.4250969174856635e-06,
      "loss": 0.4546,
      "step": 11130,
      "task_loss": 0.40569519996643066
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43485087156295776,
      "epoch": 4.03,
      "learning_rate": 6.379276892775579e-06,
      "loss": 0.4269,
      "step": 11140,
      "task_loss": 0.9028794169425964
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3825688362121582,
      "epoch": 4.03,
      "learning_rate": 6.3336044461380385e-06,
      "loss": 0.3916,
      "step": 11150,
      "task_loss": 0.5466604232788086
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3695110082626343,
      "epoch": 4.03,
      "learning_rate": 6.288079813075543e-06,
      "loss": 0.4324,
      "step": 11160,
      "task_loss": 0.3143315315246582
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3524971008300781,
      "epoch": 4.04,
      "learning_rate": 6.242703228328384e-06,
      "loss": 0.3955,
      "step": 11170,
      "task_loss": 0.49497929215431213
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5673324465751648,
      "epoch": 4.04,
      "learning_rate": 6.197474925873525e-06,
      "loss": 0.4068,
      "step": 11180,
      "task_loss": 0.5235052704811096
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39023974537849426,
      "epoch": 4.04,
      "learning_rate": 6.1523951389232715e-06,
      "loss": 0.4131,
      "step": 11190,
      "task_loss": 0.7273709177970886
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.544334352016449,
      "epoch": 4.05,
      "learning_rate": 6.107464099924179e-06,
      "loss": 0.4283,
      "step": 11200,
      "task_loss": 0.5451278686523438
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4094650149345398,
      "epoch": 4.05,
      "learning_rate": 6.062682040555794e-06,
      "loss": 0.3935,
      "step": 11210,
      "task_loss": 0.6262540817260742
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3312651813030243,
      "epoch": 4.05,
      "learning_rate": 6.018049191729475e-06,
      "loss": 0.4116,
      "step": 11220,
      "task_loss": 0.5048397183418274
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37144196033477783,
      "epoch": 4.06,
      "learning_rate": 5.9735657835871894e-06,
      "loss": 0.387,
      "step": 11230,
      "task_loss": 0.6016961336135864
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38478124141693115,
      "epoch": 4.06,
      "learning_rate": 5.929232045500351e-06,
      "loss": 0.4095,
      "step": 11240,
      "task_loss": 0.7590740919113159
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3207855820655823,
      "epoch": 4.07,
      "learning_rate": 5.885048206068627e-06,
      "loss": 0.4483,
      "step": 11250,
      "task_loss": 0.3834182024002075
    },
    {
      "epoch": 4.07,
      "eval_exact_match": 82.45033112582782,
      "eval_f1": 89.61437887010287,
      "step": 11250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33126235008239746,
      "epoch": 4.07,
      "learning_rate": 5.841014493118753e-06,
      "loss": 0.373,
      "step": 11260,
      "task_loss": 0.4069299101829529
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32440701127052307,
      "epoch": 4.07,
      "learning_rate": 5.797131133703335e-06,
      "loss": 0.4187,
      "step": 11270,
      "task_loss": 0.3777425289154053
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38386988639831543,
      "epoch": 4.08,
      "learning_rate": 5.753398354099773e-06,
      "loss": 0.4056,
      "step": 11280,
      "task_loss": 0.4748387038707733
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40675050020217896,
      "epoch": 4.08,
      "learning_rate": 5.7098163798089505e-06,
      "loss": 0.4438,
      "step": 11290,
      "task_loss": 0.5782796144485474
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41655170917510986,
      "epoch": 4.08,
      "learning_rate": 5.666385435554192e-06,
      "loss": 0.3852,
      "step": 11300,
      "task_loss": 0.415513813495636
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3474239706993103,
      "epoch": 4.09,
      "learning_rate": 5.623105745280054e-06,
      "loss": 0.4068,
      "step": 11310,
      "task_loss": 0.7074331045150757
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49055156111717224,
      "epoch": 4.09,
      "learning_rate": 5.5799775321511795e-06,
      "loss": 0.419,
      "step": 11320,
      "task_loss": 0.5741523504257202
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35831817984580994,
      "epoch": 4.09,
      "learning_rate": 5.53700101855112e-06,
      "loss": 0.3905,
      "step": 11330,
      "task_loss": 0.35589608550071716
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3777387738227844,
      "epoch": 4.1,
      "learning_rate": 5.494176426081241e-06,
      "loss": 0.3921,
      "step": 11340,
      "task_loss": 0.5492821931838989
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5529429316520691,
      "epoch": 4.1,
      "learning_rate": 5.45150397555954e-06,
      "loss": 0.4691,
      "step": 11350,
      "task_loss": 0.39025944471359253
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40030044317245483,
      "epoch": 4.11,
      "learning_rate": 5.4089838870195325e-06,
      "loss": 0.4313,
      "step": 11360,
      "task_loss": 0.5147404670715332
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4512081742286682,
      "epoch": 4.11,
      "learning_rate": 5.366616379709066e-06,
      "loss": 0.4517,
      "step": 11370,
      "task_loss": 0.6353756189346313
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40005961060523987,
      "epoch": 4.11,
      "learning_rate": 5.324401672089289e-06,
      "loss": 0.4823,
      "step": 11380,
      "task_loss": 0.8307480216026306
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3458729386329651,
      "epoch": 4.12,
      "learning_rate": 5.282339981833406e-06,
      "loss": 0.4048,
      "step": 11390,
      "task_loss": 0.6542222499847412
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6230404376983643,
      "epoch": 4.12,
      "learning_rate": 5.240431525825641e-06,
      "loss": 0.4144,
      "step": 11400,
      "task_loss": 0.8542134761810303
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4207066297531128,
      "epoch": 4.12,
      "learning_rate": 5.198676520160087e-06,
      "loss": 0.4289,
      "step": 11410,
      "task_loss": 0.49778881669044495
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4467724561691284,
      "epoch": 4.13,
      "learning_rate": 5.157075180139602e-06,
      "loss": 0.4279,
      "step": 11420,
      "task_loss": 1.0865116119384766
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43505191802978516,
      "epoch": 4.13,
      "learning_rate": 5.1156277202746694e-06,
      "loss": 0.4192,
      "step": 11430,
      "task_loss": 0.5657995939254761
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3828865885734558,
      "epoch": 4.13,
      "learning_rate": 5.074334354282335e-06,
      "loss": 0.3946,
      "step": 11440,
      "task_loss": 0.3974769115447998
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44511428475379944,
      "epoch": 4.14,
      "learning_rate": 5.033195295085081e-06,
      "loss": 0.4205,
      "step": 11450,
      "task_loss": 0.5045269727706909
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4921976327896118,
      "epoch": 4.14,
      "learning_rate": 4.992210754809733e-06,
      "loss": 0.3766,
      "step": 11460,
      "task_loss": 0.8440622091293335
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34847381711006165,
      "epoch": 4.15,
      "learning_rate": 4.951380944786361e-06,
      "loss": 0.4214,
      "step": 11470,
      "task_loss": 0.4547831416130066
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5087148547172546,
      "epoch": 4.15,
      "learning_rate": 4.910706075547198e-06,
      "loss": 0.4062,
      "step": 11480,
      "task_loss": 0.9185361862182617
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.427493155002594,
      "epoch": 4.15,
      "learning_rate": 4.8701863568255366e-06,
      "loss": 0.4216,
      "step": 11490,
      "task_loss": 0.5520419478416443
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5256533026695251,
      "epoch": 4.16,
      "learning_rate": 4.829821997554683e-06,
      "loss": 0.4155,
      "step": 11500,
      "task_loss": 0.7243201732635498
    },
    {
      "epoch": 4.16,
      "eval_exact_match": 82.75307473982971,
      "eval_f1": 89.83958785566995,
      "step": 11500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33289045095443726,
      "epoch": 4.16,
      "learning_rate": 4.789613205866839e-06,
      "loss": 0.4289,
      "step": 11510,
      "task_loss": 0.4773218631744385
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5849810838699341,
      "epoch": 4.16,
      "learning_rate": 4.749560189092066e-06,
      "loss": 0.4359,
      "step": 11520,
      "task_loss": 1.2744464874267578
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.378299355506897,
      "epoch": 4.17,
      "learning_rate": 4.709663153757165e-06,
      "loss": 0.3864,
      "step": 11530,
      "task_loss": 0.4580954313278198
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4290379285812378,
      "epoch": 4.17,
      "learning_rate": 4.669922305584701e-06,
      "loss": 0.4094,
      "step": 11540,
      "task_loss": 0.649642825126648
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31375768780708313,
      "epoch": 4.17,
      "learning_rate": 4.630337849491818e-06,
      "loss": 0.4181,
      "step": 11550,
      "task_loss": 0.4781082272529602
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29374831914901733,
      "epoch": 4.18,
      "learning_rate": 4.5909099895892995e-06,
      "loss": 0.3873,
      "step": 11560,
      "task_loss": 0.4056081771850586
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32750147581100464,
      "epoch": 4.18,
      "learning_rate": 4.551638929180444e-06,
      "loss": 0.4488,
      "step": 11570,
      "task_loss": 0.5592796802520752
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31511664390563965,
      "epoch": 4.19,
      "learning_rate": 4.512524870760054e-06,
      "loss": 0.3618,
      "step": 11580,
      "task_loss": 0.4754389226436615
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3744344711303711,
      "epoch": 4.19,
      "learning_rate": 4.473568016013349e-06,
      "loss": 0.4034,
      "step": 11590,
      "task_loss": 0.6776789426803589
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4056161344051361,
      "epoch": 4.19,
      "learning_rate": 4.4347685658149885e-06,
      "loss": 0.3945,
      "step": 11600,
      "task_loss": 0.6023022532463074
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3327157199382782,
      "epoch": 4.2,
      "learning_rate": 4.396126720227975e-06,
      "loss": 0.3806,
      "step": 11610,
      "task_loss": 0.5939061641693115
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5144981145858765,
      "epoch": 4.2,
      "learning_rate": 4.357642678502669e-06,
      "loss": 0.3975,
      "step": 11620,
      "task_loss": 0.972696840763092
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46392837166786194,
      "epoch": 4.2,
      "learning_rate": 4.319316639075711e-06,
      "loss": 0.4097,
      "step": 11630,
      "task_loss": 0.30408644676208496
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26904988288879395,
      "epoch": 4.21,
      "learning_rate": 4.281148799569073e-06,
      "loss": 0.4108,
      "step": 11640,
      "task_loss": 0.5989107489585876
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.409515380859375,
      "epoch": 4.21,
      "learning_rate": 4.243139356788951e-06,
      "loss": 0.4074,
      "step": 11650,
      "task_loss": 0.4942619800567627
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3208807706832886,
      "epoch": 4.21,
      "learning_rate": 4.205288506724823e-06,
      "loss": 0.3768,
      "step": 11660,
      "task_loss": 0.46400171518325806
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4018038213253021,
      "epoch": 4.22,
      "learning_rate": 4.167596444548396e-06,
      "loss": 0.4199,
      "step": 11670,
      "task_loss": 0.6154822111129761
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30387967824935913,
      "epoch": 4.22,
      "learning_rate": 4.130063364612621e-06,
      "loss": 0.401,
      "step": 11680,
      "task_loss": 0.5835756063461304
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5109657049179077,
      "epoch": 4.22,
      "learning_rate": 4.0926894604506725e-06,
      "loss": 0.4202,
      "step": 11690,
      "task_loss": 0.9396585822105408
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3630220890045166,
      "epoch": 4.23,
      "learning_rate": 4.055474924774975e-06,
      "loss": 0.4453,
      "step": 11700,
      "task_loss": 1.1204054355621338
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3808678388595581,
      "epoch": 4.23,
      "learning_rate": 4.01841994947618e-06,
      "loss": 0.4071,
      "step": 11710,
      "task_loss": 0.6350036859512329
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36871522665023804,
      "epoch": 4.24,
      "learning_rate": 3.981524725622215e-06,
      "loss": 0.4058,
      "step": 11720,
      "task_loss": 0.48085230588912964
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.379439115524292,
      "epoch": 4.24,
      "learning_rate": 3.944789443457237e-06,
      "loss": 0.3727,
      "step": 11730,
      "task_loss": 0.4141382873058319
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49233728647232056,
      "epoch": 4.24,
      "learning_rate": 3.908214292400743e-06,
      "loss": 0.4215,
      "step": 11740,
      "task_loss": 0.3407425284385681
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40830057859420776,
      "epoch": 4.25,
      "learning_rate": 3.871799461046491e-06,
      "loss": 0.4168,
      "step": 11750,
      "task_loss": 0.7030295133590698
    },
    {
      "epoch": 4.25,
      "eval_exact_match": 82.67738883632923,
      "eval_f1": 89.83433781229459,
      "step": 11750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5997090339660645,
      "epoch": 4.25,
      "learning_rate": 3.8355451371616e-06,
      "loss": 0.4535,
      "step": 11760,
      "task_loss": 0.7779182195663452
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3632296621799469,
      "epoch": 4.25,
      "learning_rate": 3.799451507685557e-06,
      "loss": 0.408,
      "step": 11770,
      "task_loss": 0.6163282990455627
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3645058274269104,
      "epoch": 4.26,
      "learning_rate": 3.7635187587292616e-06,
      "loss": 0.4098,
      "step": 11780,
      "task_loss": 0.4656088948249817
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3899230659008026,
      "epoch": 4.26,
      "learning_rate": 3.727747075574033e-06,
      "loss": 0.3789,
      "step": 11790,
      "task_loss": 0.37167468667030334
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3266602158546448,
      "epoch": 4.26,
      "learning_rate": 3.6921366426707064e-06,
      "loss": 0.393,
      "step": 11800,
      "task_loss": 0.42901307344436646
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.48315972089767456,
      "epoch": 4.27,
      "learning_rate": 3.656687643638646e-06,
      "loss": 0.4205,
      "step": 11810,
      "task_loss": 0.883271336555481
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3944903016090393,
      "epoch": 4.27,
      "learning_rate": 3.6214002612648187e-06,
      "loss": 0.4132,
      "step": 11820,
      "task_loss": 0.3511267900466919
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3845597505569458,
      "epoch": 4.28,
      "learning_rate": 3.5862746775028274e-06,
      "loss": 0.3884,
      "step": 11830,
      "task_loss": 0.8081369400024414
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.52008455991745,
      "epoch": 4.28,
      "learning_rate": 3.5513110734719947e-06,
      "loss": 0.388,
      "step": 11840,
      "task_loss": 0.6470053195953369
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32628077268600464,
      "epoch": 4.28,
      "learning_rate": 3.5165096294564294e-06,
      "loss": 0.412,
      "step": 11850,
      "task_loss": 0.5431479215621948
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3263109624385834,
      "epoch": 4.29,
      "learning_rate": 3.481870524904063e-06,
      "loss": 0.4327,
      "step": 11860,
      "task_loss": 0.4655928313732147
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33923470973968506,
      "epoch": 4.29,
      "learning_rate": 3.4473939384257713e-06,
      "loss": 0.4195,
      "step": 11870,
      "task_loss": 0.5043308734893799
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3827700614929199,
      "epoch": 4.29,
      "learning_rate": 3.4130800477944353e-06,
      "loss": 0.4015,
      "step": 11880,
      "task_loss": 0.5053728222846985
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.418369323015213,
      "epoch": 4.3,
      "learning_rate": 3.378929029944014e-06,
      "loss": 0.3662,
      "step": 11890,
      "task_loss": 0.5845623016357422
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3653680384159088,
      "epoch": 4.3,
      "learning_rate": 3.3449410609686222e-06,
      "loss": 0.4075,
      "step": 11900,
      "task_loss": 0.5306881070137024
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38282740116119385,
      "epoch": 4.3,
      "learning_rate": 3.311116316121681e-06,
      "loss": 0.4187,
      "step": 11910,
      "task_loss": 0.5226006507873535
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3763877749443054,
      "epoch": 4.31,
      "learning_rate": 3.27745496981493e-06,
      "loss": 0.374,
      "step": 11920,
      "task_loss": 0.5572165846824646
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4219665825366974,
      "epoch": 4.31,
      "learning_rate": 3.2439571956175955e-06,
      "loss": 0.4161,
      "step": 11930,
      "task_loss": 0.3990279734134674
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36016109585762024,
      "epoch": 4.32,
      "learning_rate": 3.2106231662554687e-06,
      "loss": 0.4118,
      "step": 11940,
      "task_loss": 0.7736810445785522
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4748755395412445,
      "epoch": 4.32,
      "learning_rate": 3.17745305361002e-06,
      "loss": 0.4131,
      "step": 11950,
      "task_loss": 0.7507423758506775
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32424837350845337,
      "epoch": 4.32,
      "learning_rate": 3.144447028717492e-06,
      "loss": 0.3758,
      "step": 11960,
      "task_loss": 0.3291553854942322
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4312622547149658,
      "epoch": 4.33,
      "learning_rate": 3.1116052617680475e-06,
      "loss": 0.3978,
      "step": 11970,
      "task_loss": 0.6131731271743774
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3169502019882202,
      "epoch": 4.33,
      "learning_rate": 3.078927922104888e-06,
      "loss": 0.3739,
      "step": 11980,
      "task_loss": 0.7729911804199219
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5403449535369873,
      "epoch": 4.33,
      "learning_rate": 3.04641517822337e-06,
      "loss": 0.429,
      "step": 11990,
      "task_loss": 0.7878259420394897
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5041810870170593,
      "epoch": 4.34,
      "learning_rate": 3.0140671977701126e-06,
      "loss": 0.4378,
      "step": 12000,
      "task_loss": 0.8172563314437866
    },
    {
      "epoch": 4.34,
      "eval_exact_match": 82.39356669820246,
      "eval_f1": 89.55316666275424,
      "step": 12000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45394831895828247,
      "epoch": 4.34,
      "learning_rate": 2.9818841475422074e-06,
      "loss": 0.4096,
      "step": 12010,
      "task_loss": 0.7580146789550781
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30688315629959106,
      "epoch": 4.34,
      "learning_rate": 2.9498661934862625e-06,
      "loss": 0.4064,
      "step": 12020,
      "task_loss": 0.5445766448974609
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.324246883392334,
      "epoch": 4.35,
      "learning_rate": 2.9180135006976253e-06,
      "loss": 0.402,
      "step": 12030,
      "task_loss": 0.7565941214561462
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37556836009025574,
      "epoch": 4.35,
      "learning_rate": 2.8863262334194892e-06,
      "loss": 0.4264,
      "step": 12040,
      "task_loss": 0.7802863121032715
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3691438138484955,
      "epoch": 4.35,
      "learning_rate": 2.854804555042066e-06,
      "loss": 0.4111,
      "step": 12050,
      "task_loss": 0.4967532455921173
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3941887617111206,
      "epoch": 4.36,
      "learning_rate": 2.823448628101714e-06,
      "loss": 0.4082,
      "step": 12060,
      "task_loss": 0.6636197566986084
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3338344693183899,
      "epoch": 4.36,
      "learning_rate": 2.792258614280147e-06,
      "loss": 0.4258,
      "step": 12070,
      "task_loss": 0.4560825824737549
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39394477009773254,
      "epoch": 4.37,
      "learning_rate": 2.761234674403565e-06,
      "loss": 0.3904,
      "step": 12080,
      "task_loss": 0.5815662145614624
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3504999876022339,
      "epoch": 4.37,
      "learning_rate": 2.730376968441837e-06,
      "loss": 0.3928,
      "step": 12090,
      "task_loss": 0.5948781371116638
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30044350028038025,
      "epoch": 4.37,
      "learning_rate": 2.699685655507652e-06,
      "loss": 0.4128,
      "step": 12100,
      "task_loss": 0.3113638758659363
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4110683500766754,
      "epoch": 4.38,
      "learning_rate": 2.6691608938557673e-06,
      "loss": 0.4011,
      "step": 12110,
      "task_loss": 0.7891151309013367
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39706042408943176,
      "epoch": 4.38,
      "learning_rate": 2.638802840882099e-06,
      "loss": 0.3894,
      "step": 12120,
      "task_loss": 0.6502903699874878
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.419207364320755,
      "epoch": 4.38,
      "learning_rate": 2.608611653122982e-06,
      "loss": 0.4157,
      "step": 12130,
      "task_loss": 0.9563859701156616
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3630548119544983,
      "epoch": 4.39,
      "learning_rate": 2.5785874862543364e-06,
      "loss": 0.395,
      "step": 12140,
      "task_loss": 0.5200846195220947
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4006337821483612,
      "epoch": 4.39,
      "learning_rate": 2.548730495090867e-06,
      "loss": 0.4044,
      "step": 12150,
      "task_loss": 0.7663620710372925
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44687414169311523,
      "epoch": 4.39,
      "learning_rate": 2.5190408335852423e-06,
      "loss": 0.3852,
      "step": 12160,
      "task_loss": 0.5466241240501404
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36376717686653137,
      "epoch": 4.4,
      "learning_rate": 2.4895186548273553e-06,
      "loss": 0.4024,
      "step": 12170,
      "task_loss": 0.4370715618133545
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47231391072273254,
      "epoch": 4.4,
      "learning_rate": 2.4601641110434842e-06,
      "loss": 0.4036,
      "step": 12180,
      "task_loss": 0.820824921131134
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5579500198364258,
      "epoch": 4.41,
      "learning_rate": 2.430977353595531e-06,
      "loss": 0.4426,
      "step": 12190,
      "task_loss": 0.6717907190322876
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43694451451301575,
      "epoch": 4.41,
      "learning_rate": 2.401958532980213e-06,
      "loss": 0.4222,
      "step": 12200,
      "task_loss": 0.6733720302581787
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27389103174209595,
      "epoch": 4.41,
      "learning_rate": 2.3731077988283558e-06,
      "loss": 0.3659,
      "step": 12210,
      "task_loss": 0.3571043610572815
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37224647402763367,
      "epoch": 4.42,
      "learning_rate": 2.34442529990403e-06,
      "loss": 0.3988,
      "step": 12220,
      "task_loss": 0.3586083948612213
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35565561056137085,
      "epoch": 4.42,
      "learning_rate": 2.315911184103842e-06,
      "loss": 0.3741,
      "step": 12230,
      "task_loss": 0.5137349367141724
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29059457778930664,
      "epoch": 4.42,
      "learning_rate": 2.287565598456169e-06,
      "loss": 0.3942,
      "step": 12240,
      "task_loss": 0.54352867603302
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2987663745880127,
      "epoch": 4.43,
      "learning_rate": 2.259388689120386e-06,
      "loss": 0.3893,
      "step": 12250,
      "task_loss": 0.3862733840942383
    },
    {
      "epoch": 4.43,
      "eval_exact_match": 82.63008514664143,
      "eval_f1": 89.73722178822344,
      "step": 12250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35016316175460815,
      "epoch": 4.43,
      "learning_rate": 2.2313806013861007e-06,
      "loss": 0.3867,
      "step": 12260,
      "task_loss": 0.46879592537879944
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3823249340057373,
      "epoch": 4.43,
      "learning_rate": 2.2035414796724365e-06,
      "loss": 0.4238,
      "step": 12270,
      "task_loss": 0.630598783493042
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4345713257789612,
      "epoch": 4.44,
      "learning_rate": 2.1758714675272663e-06,
      "loss": 0.4307,
      "step": 12280,
      "task_loss": 0.5999336242675781
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5260065793991089,
      "epoch": 4.44,
      "learning_rate": 2.1483707076264827e-06,
      "loss": 0.4394,
      "step": 12290,
      "task_loss": 0.7119778394699097
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37995320558547974,
      "epoch": 4.45,
      "learning_rate": 2.121039341773235e-06,
      "loss": 0.4377,
      "step": 12300,
      "task_loss": 0.5297107696533203
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4561520218849182,
      "epoch": 4.45,
      "learning_rate": 2.0938775108972628e-06,
      "loss": 0.4092,
      "step": 12310,
      "task_loss": 0.6727806329727173
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.326673686504364,
      "epoch": 4.45,
      "learning_rate": 2.0668853550540775e-06,
      "loss": 0.4197,
      "step": 12320,
      "task_loss": 0.27783846855163574
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43794888257980347,
      "epoch": 4.46,
      "learning_rate": 2.0400630134243162e-06,
      "loss": 0.3867,
      "step": 12330,
      "task_loss": 0.5385734438896179
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30467817187309265,
      "epoch": 4.46,
      "learning_rate": 2.013410624312997e-06,
      "loss": 0.3803,
      "step": 12340,
      "task_loss": 0.4139289855957031
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35417941212654114,
      "epoch": 4.46,
      "learning_rate": 1.9869283251488022e-06,
      "loss": 0.3369,
      "step": 12350,
      "task_loss": 0.5619122982025146
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36347198486328125,
      "epoch": 4.47,
      "learning_rate": 1.9606162524833615e-06,
      "loss": 0.4038,
      "step": 12360,
      "task_loss": 0.6940585970878601
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5780143737792969,
      "epoch": 4.47,
      "learning_rate": 1.9344745419905726e-06,
      "loss": 0.4446,
      "step": 12370,
      "task_loss": 1.0974355936050415
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.327537477016449,
      "epoch": 4.47,
      "learning_rate": 1.9085033284658917e-06,
      "loss": 0.4227,
      "step": 12380,
      "task_loss": 0.4338206648826599
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5237728357315063,
      "epoch": 4.48,
      "learning_rate": 1.8827027458256239e-06,
      "loss": 0.4358,
      "step": 12390,
      "task_loss": 0.8812257051467896
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45310911536216736,
      "epoch": 4.48,
      "learning_rate": 1.857072927106244e-06,
      "loss": 0.4478,
      "step": 12400,
      "task_loss": 0.6072947978973389
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49674561619758606,
      "epoch": 4.49,
      "learning_rate": 1.831614004463733e-06,
      "loss": 0.4305,
      "step": 12410,
      "task_loss": 0.7922726273536682
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3848080039024353,
      "epoch": 4.49,
      "learning_rate": 1.806326109172838e-06,
      "loss": 0.3823,
      "step": 12420,
      "task_loss": 0.37194693088531494
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3583149015903473,
      "epoch": 4.49,
      "learning_rate": 1.781209371626461e-06,
      "loss": 0.393,
      "step": 12430,
      "task_loss": 0.391252726316452
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5838788151741028,
      "epoch": 4.5,
      "learning_rate": 1.75626392133494e-06,
      "loss": 0.4088,
      "step": 12440,
      "task_loss": 0.6965578198432922
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5408843159675598,
      "epoch": 4.5,
      "learning_rate": 1.7314898869254157e-06,
      "loss": 0.4157,
      "step": 12450,
      "task_loss": 0.4436303377151489
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4076458513736725,
      "epoch": 4.5,
      "learning_rate": 1.7068873961411294e-06,
      "loss": 0.4012,
      "step": 12460,
      "task_loss": 0.6320613026618958
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3804321885108948,
      "epoch": 4.51,
      "learning_rate": 1.6824565758408027e-06,
      "loss": 0.3988,
      "step": 12470,
      "task_loss": 0.7089783549308777
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2783825993537903,
      "epoch": 4.51,
      "learning_rate": 1.6581975519979562e-06,
      "loss": 0.4198,
      "step": 12480,
      "task_loss": 0.660349428653717
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33828794956207275,
      "epoch": 4.51,
      "learning_rate": 1.6341104497002816e-06,
      "loss": 0.3879,
      "step": 12490,
      "task_loss": 0.23362329602241516
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4729890823364258,
      "epoch": 4.52,
      "learning_rate": 1.6101953931489643e-06,
      "loss": 0.3607,
      "step": 12500,
      "task_loss": 0.6899687051773071
    },
    {
      "epoch": 4.52,
      "eval_exact_match": 82.639545884579,
      "eval_f1": 89.7036436184772,
      "step": 12500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3680706024169922,
      "epoch": 4.52,
      "learning_rate": 1.5864525056580967e-06,
      "loss": 0.4391,
      "step": 12510,
      "task_loss": 0.5775980949401855
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4175216555595398,
      "epoch": 4.52,
      "learning_rate": 1.5628819096539764e-06,
      "loss": 0.4019,
      "step": 12520,
      "task_loss": 0.9011520743370056
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4379788041114807,
      "epoch": 4.53,
      "learning_rate": 1.5394837266745246e-06,
      "loss": 0.3794,
      "step": 12530,
      "task_loss": 0.49954891204833984
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4514831304550171,
      "epoch": 4.53,
      "learning_rate": 1.5162580773686438e-06,
      "loss": 0.4316,
      "step": 12540,
      "task_loss": 0.4557006359100342
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41194820404052734,
      "epoch": 4.54,
      "learning_rate": 1.4932050814955942e-06,
      "loss": 0.4016,
      "step": 12550,
      "task_loss": 0.3731171488761902
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32936277985572815,
      "epoch": 4.54,
      "learning_rate": 1.470324857924365e-06,
      "loss": 0.3858,
      "step": 12560,
      "task_loss": 0.384751558303833
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.610994815826416,
      "epoch": 4.54,
      "learning_rate": 1.4476175246330906e-06,
      "loss": 0.4325,
      "step": 12570,
      "task_loss": 1.0964241027832031
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41487258672714233,
      "epoch": 4.55,
      "learning_rate": 1.4250831987084155e-06,
      "loss": 0.3689,
      "step": 12580,
      "task_loss": 0.9095247387886047
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39539405703544617,
      "epoch": 4.55,
      "learning_rate": 1.4027219963449087e-06,
      "loss": 0.3798,
      "step": 12590,
      "task_loss": 0.6725915670394897
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31662312150001526,
      "epoch": 4.55,
      "learning_rate": 1.3805340328444287e-06,
      "loss": 0.3512,
      "step": 12600,
      "task_loss": 0.5368452072143555
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47574758529663086,
      "epoch": 4.56,
      "learning_rate": 1.358519422615601e-06,
      "loss": 0.4363,
      "step": 12610,
      "task_loss": 0.40942999720573425
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29873067140579224,
      "epoch": 4.56,
      "learning_rate": 1.3366782791731396e-06,
      "loss": 0.3548,
      "step": 12620,
      "task_loss": 0.4708471894264221
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37882232666015625,
      "epoch": 4.56,
      "learning_rate": 1.315010715137333e-06,
      "loss": 0.3903,
      "step": 12630,
      "task_loss": 0.5266144871711731
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44176942110061646,
      "epoch": 4.57,
      "learning_rate": 1.2935168422334307e-06,
      "loss": 0.4213,
      "step": 12640,
      "task_loss": 0.44183528423309326
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3719322085380554,
      "epoch": 4.57,
      "learning_rate": 1.2721967712910648e-06,
      "loss": 0.4138,
      "step": 12650,
      "task_loss": 0.46417778730392456
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24051040410995483,
      "epoch": 4.58,
      "learning_rate": 1.2510506122436886e-06,
      "loss": 0.3619,
      "step": 12660,
      "task_loss": 0.325649619102478
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46371278166770935,
      "epoch": 4.58,
      "learning_rate": 1.2300784741280078e-06,
      "loss": 0.4077,
      "step": 12670,
      "task_loss": 0.7890371084213257
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31556272506713867,
      "epoch": 4.58,
      "learning_rate": 1.209280465083427e-06,
      "loss": 0.4136,
      "step": 12680,
      "task_loss": 0.7762376070022583
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4381011128425598,
      "epoch": 4.59,
      "learning_rate": 1.1886566923514634e-06,
      "loss": 0.4207,
      "step": 12690,
      "task_loss": 0.9917271733283997
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3886127173900604,
      "epoch": 4.59,
      "learning_rate": 1.1682072622752342e-06,
      "loss": 0.4295,
      "step": 12700,
      "task_loss": 0.6097781658172607
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4845806956291199,
      "epoch": 4.59,
      "learning_rate": 1.147932280298877e-06,
      "loss": 0.4321,
      "step": 12710,
      "task_loss": 0.8994347453117371
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41579073667526245,
      "epoch": 4.6,
      "learning_rate": 1.127831850967007e-06,
      "loss": 0.3906,
      "step": 12720,
      "task_loss": 0.45600637793540955
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45800602436065674,
      "epoch": 4.6,
      "learning_rate": 1.1079060779242022e-06,
      "loss": 0.4268,
      "step": 12730,
      "task_loss": 0.7129201889038086
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4211260676383972,
      "epoch": 4.6,
      "learning_rate": 1.0881550639144531e-06,
      "loss": 0.4522,
      "step": 12740,
      "task_loss": 0.772409200668335
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39386406540870667,
      "epoch": 4.61,
      "learning_rate": 1.0685789107806264e-06,
      "loss": 0.3954,
      "step": 12750,
      "task_loss": 0.2882859408855438
    },
    {
      "epoch": 4.61,
      "eval_exact_match": 82.55439924314096,
      "eval_f1": 89.76019468353418,
      "step": 12750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3273109793663025,
      "epoch": 4.61,
      "learning_rate": 1.0491777194639433e-06,
      "loss": 0.4323,
      "step": 12760,
      "task_loss": 0.9542597532272339
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3854421377182007,
      "epoch": 4.62,
      "learning_rate": 1.029951590003487e-06,
      "loss": 0.4066,
      "step": 12770,
      "task_loss": 0.6295945048332214
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34793558716773987,
      "epoch": 4.62,
      "learning_rate": 1.0109006215356314e-06,
      "loss": 0.4065,
      "step": 12780,
      "task_loss": 0.3474682569503784
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4002053141593933,
      "epoch": 4.62,
      "learning_rate": 9.920249122935775e-07,
      "loss": 0.4288,
      "step": 12790,
      "task_loss": 0.3302229046821594
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31503939628601074,
      "epoch": 4.63,
      "learning_rate": 9.733245596068346e-07,
      "loss": 0.3968,
      "step": 12800,
      "task_loss": 0.30184412002563477
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4567040801048279,
      "epoch": 4.63,
      "learning_rate": 9.547996599007135e-07,
      "loss": 0.4107,
      "step": 12810,
      "task_loss": 0.48562997579574585
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30701467394828796,
      "epoch": 4.63,
      "learning_rate": 9.364503086958147e-07,
      "loss": 0.375,
      "step": 12820,
      "task_loss": 0.616033673286438
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37898629903793335,
      "epoch": 4.64,
      "learning_rate": 9.182766006075692e-07,
      "loss": 0.3696,
      "step": 12830,
      "task_loss": 0.42548850178718567
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35982203483581543,
      "epoch": 4.64,
      "learning_rate": 9.002786293457259e-07,
      "loss": 0.4362,
      "step": 12840,
      "task_loss": 0.7381753921508789
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2988097667694092,
      "epoch": 4.64,
      "learning_rate": 8.824564877138851e-07,
      "loss": 0.423,
      "step": 12850,
      "task_loss": 0.6503967642784119
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4052898585796356,
      "epoch": 4.65,
      "learning_rate": 8.648102676089857e-07,
      "loss": 0.381,
      "step": 12860,
      "task_loss": 0.38243943452835083
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42070329189300537,
      "epoch": 4.65,
      "learning_rate": 8.473400600208857e-07,
      "loss": 0.4464,
      "step": 12870,
      "task_loss": 0.5559731721878052
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45240452885627747,
      "epoch": 4.65,
      "learning_rate": 8.30045955031845e-07,
      "loss": 0.3897,
      "step": 12880,
      "task_loss": 0.5374152064323425
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41727572679519653,
      "epoch": 4.66,
      "learning_rate": 8.129280418160867e-07,
      "loss": 0.3942,
      "step": 12890,
      "task_loss": 0.5857281684875488
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3383082151412964,
      "epoch": 4.66,
      "learning_rate": 7.959864086393225e-07,
      "loss": 0.3923,
      "step": 12900,
      "task_loss": 0.2979002594947815
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2600701153278351,
      "epoch": 4.67,
      "learning_rate": 7.792211428583184e-07,
      "loss": 0.4077,
      "step": 12910,
      "task_loss": 0.3120166063308716
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39484894275665283,
      "epoch": 4.67,
      "learning_rate": 7.626323309204157e-07,
      "loss": 0.3848,
      "step": 12920,
      "task_loss": 0.8494405150413513
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3408588469028473,
      "epoch": 4.67,
      "learning_rate": 7.462200583631045e-07,
      "loss": 0.4072,
      "step": 12930,
      "task_loss": 0.6850634813308716
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4030011296272278,
      "epoch": 4.68,
      "learning_rate": 7.299844098135915e-07,
      "loss": 0.4062,
      "step": 12940,
      "task_loss": 0.4810425639152527
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3025147318840027,
      "epoch": 4.68,
      "learning_rate": 7.139254689883305e-07,
      "loss": 0.3951,
      "step": 12950,
      "task_loss": 0.593737006187439
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3802594542503357,
      "epoch": 4.68,
      "learning_rate": 6.980433186926221e-07,
      "loss": 0.4089,
      "step": 12960,
      "task_loss": 0.588676929473877
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3437938988208771,
      "epoch": 4.69,
      "learning_rate": 6.823380408201817e-07,
      "loss": 0.4062,
      "step": 12970,
      "task_loss": 0.3986343741416931
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3688074052333832,
      "epoch": 4.69,
      "learning_rate": 6.668097163526936e-07,
      "loss": 0.4375,
      "step": 12980,
      "task_loss": 0.48088422417640686
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3709104061126709,
      "epoch": 4.69,
      "learning_rate": 6.514584253594218e-07,
      "loss": 0.4188,
      "step": 12990,
      "task_loss": 0.47096890211105347
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4205417037010193,
      "epoch": 4.7,
      "learning_rate": 6.362842469967905e-07,
      "loss": 0.4339,
      "step": 13000,
      "task_loss": 0.3885349631309509
    },
    {
      "epoch": 4.7,
      "eval_exact_match": 82.63008514664143,
      "eval_f1": 89.77533433718209,
      "step": 13000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33247077465057373,
      "epoch": 4.7,
      "learning_rate": 6.212872595079643e-07,
      "loss": 0.4131,
      "step": 13010,
      "task_loss": 0.4318593144416809
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.48946917057037354,
      "epoch": 4.71,
      "learning_rate": 6.064675402224444e-07,
      "loss": 0.4139,
      "step": 13020,
      "task_loss": 0.5803698301315308
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3440093398094177,
      "epoch": 4.71,
      "learning_rate": 5.918251655556994e-07,
      "loss": 0.4133,
      "step": 13030,
      "task_loss": 0.7929366827011108
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30020883679389954,
      "epoch": 4.71,
      "learning_rate": 5.773602110087295e-07,
      "loss": 0.4317,
      "step": 13040,
      "task_loss": 0.6749934554100037
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.637604296207428,
      "epoch": 4.72,
      "learning_rate": 5.630727511677097e-07,
      "loss": 0.4341,
      "step": 13050,
      "task_loss": 0.8321160078048706
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37140244245529175,
      "epoch": 4.72,
      "learning_rate": 5.489628597035817e-07,
      "loss": 0.3955,
      "step": 13060,
      "task_loss": 0.8422946929931641
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.442553848028183,
      "epoch": 4.72,
      "learning_rate": 5.350306093716961e-07,
      "loss": 0.4234,
      "step": 13070,
      "task_loss": 0.6642134785652161
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43634718656539917,
      "epoch": 4.73,
      "learning_rate": 5.212760720114123e-07,
      "loss": 0.3571,
      "step": 13080,
      "task_loss": 0.622330904006958
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4799298644065857,
      "epoch": 4.73,
      "learning_rate": 5.076993185457568e-07,
      "loss": 0.4098,
      "step": 13090,
      "task_loss": 0.46641305088996887
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35526013374328613,
      "epoch": 4.73,
      "learning_rate": 4.943004189810379e-07,
      "loss": 0.3992,
      "step": 13100,
      "task_loss": 0.5122348070144653
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46897608041763306,
      "epoch": 4.74,
      "learning_rate": 4.810794424064813e-07,
      "loss": 0.4203,
      "step": 13110,
      "task_loss": 0.6970722675323486
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4693143367767334,
      "epoch": 4.74,
      "learning_rate": 4.6803645699389105e-07,
      "loss": 0.4392,
      "step": 13120,
      "task_loss": 0.5941569209098816
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40301787853240967,
      "epoch": 4.75,
      "learning_rate": 4.551715299972852e-07,
      "loss": 0.466,
      "step": 13130,
      "task_loss": 0.7612343430519104
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3860597014427185,
      "epoch": 4.75,
      "learning_rate": 4.424847277525534e-07,
      "loss": 0.4076,
      "step": 13140,
      "task_loss": 0.6143934726715088
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46132004261016846,
      "epoch": 4.75,
      "learning_rate": 4.299761156771148e-07,
      "loss": 0.4,
      "step": 13150,
      "task_loss": 0.5275061130523682
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3260010778903961,
      "epoch": 4.76,
      "learning_rate": 4.1764575826957634e-07,
      "loss": 0.3797,
      "step": 13160,
      "task_loss": 0.3990122973918915
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3616238832473755,
      "epoch": 4.76,
      "learning_rate": 4.054937191094143e-07,
      "loss": 0.4277,
      "step": 13170,
      "task_loss": 0.44919294118881226
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39060574769973755,
      "epoch": 4.76,
      "learning_rate": 3.935200608566241e-07,
      "loss": 0.4464,
      "step": 13180,
      "task_loss": 0.495510458946228
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47807246446609497,
      "epoch": 4.77,
      "learning_rate": 3.817248452514138e-07,
      "loss": 0.4065,
      "step": 13190,
      "task_loss": 0.6110713481903076
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43341243267059326,
      "epoch": 4.77,
      "learning_rate": 3.701081331138772e-07,
      "loss": 0.4318,
      "step": 13200,
      "task_loss": 0.6513692140579224
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40219298005104065,
      "epoch": 4.77,
      "learning_rate": 3.586699843436991e-07,
      "loss": 0.3551,
      "step": 13210,
      "task_loss": 0.33632349967956543
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4364554286003113,
      "epoch": 4.78,
      "learning_rate": 3.474104579198128e-07,
      "loss": 0.4083,
      "step": 13220,
      "task_loss": 0.4315539300441742
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3684564232826233,
      "epoch": 4.78,
      "learning_rate": 3.363296119001246e-07,
      "loss": 0.3922,
      "step": 13230,
      "task_loss": 0.7234320640563965
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35623037815093994,
      "epoch": 4.78,
      "learning_rate": 3.254275034212028e-07,
      "loss": 0.3792,
      "step": 13240,
      "task_loss": 0.6207572221755981
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5407557487487793,
      "epoch": 4.79,
      "learning_rate": 3.1470418869799007e-07,
      "loss": 0.4309,
      "step": 13250,
      "task_loss": 0.5729117393493652
    },
    {
      "epoch": 4.79,
      "eval_exact_match": 82.36518448438979,
      "eval_f1": 89.62599752726967,
      "step": 13250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3366709351539612,
      "epoch": 4.79,
      "learning_rate": 3.041597230234888e-07,
      "loss": 0.3861,
      "step": 13260,
      "task_loss": 0.7600011825561523
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3440953195095062,
      "epoch": 4.8,
      "learning_rate": 2.9379416076852405e-07,
      "loss": 0.3966,
      "step": 13270,
      "task_loss": 0.48770636320114136
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35563838481903076,
      "epoch": 4.8,
      "learning_rate": 2.836075553814171e-07,
      "loss": 0.4175,
      "step": 13280,
      "task_loss": 0.3995950520038605
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5269416570663452,
      "epoch": 4.8,
      "learning_rate": 2.735999593877253e-07,
      "loss": 0.4154,
      "step": 13290,
      "task_loss": 0.6878436803817749
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3388027548789978,
      "epoch": 4.81,
      "learning_rate": 2.6377142438998134e-07,
      "loss": 0.3958,
      "step": 13300,
      "task_loss": 0.4817107319831848
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4826379418373108,
      "epoch": 4.81,
      "learning_rate": 2.5412200106742166e-07,
      "loss": 0.3647,
      "step": 13310,
      "task_loss": 1.065226435661316
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30741727352142334,
      "epoch": 4.81,
      "learning_rate": 2.4465173917571023e-07,
      "loss": 0.3859,
      "step": 13320,
      "task_loss": 0.5023446679115295
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4362955093383789,
      "epoch": 4.82,
      "learning_rate": 2.3536068754670568e-07,
      "loss": 0.4087,
      "step": 13330,
      "task_loss": 0.7960480451583862
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2628311514854431,
      "epoch": 4.82,
      "learning_rate": 2.262488940881968e-07,
      "loss": 0.3649,
      "step": 13340,
      "task_loss": 0.4985067844390869
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3144613802433014,
      "epoch": 4.82,
      "learning_rate": 2.1731640578365016e-07,
      "loss": 0.3909,
      "step": 13350,
      "task_loss": 0.2815832495689392
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37193599343299866,
      "epoch": 4.83,
      "learning_rate": 2.0856326869198082e-07,
      "loss": 0.398,
      "step": 13360,
      "task_loss": 0.8057834506034851
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4489399194717407,
      "epoch": 4.83,
      "learning_rate": 1.999895279473074e-07,
      "loss": 0.4062,
      "step": 13370,
      "task_loss": 0.6343175768852234
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.288999080657959,
      "epoch": 4.84,
      "learning_rate": 1.9159522775871906e-07,
      "loss": 0.4113,
      "step": 13380,
      "task_loss": 0.7062314748764038
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46635884046554565,
      "epoch": 4.84,
      "learning_rate": 1.83380411410054e-07,
      "loss": 0.399,
      "step": 13390,
      "task_loss": 0.716343879699707
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33257097005844116,
      "epoch": 4.84,
      "learning_rate": 1.7534512125966237e-07,
      "loss": 0.4012,
      "step": 13400,
      "task_loss": 0.857595682144165
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4126928448677063,
      "epoch": 4.85,
      "learning_rate": 1.6748939874020818e-07,
      "loss": 0.4223,
      "step": 13410,
      "task_loss": 0.6019068956375122
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3106433153152466,
      "epoch": 4.85,
      "learning_rate": 1.598132843584321e-07,
      "loss": 0.4124,
      "step": 13420,
      "task_loss": 0.6166244149208069
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3775056004524231,
      "epoch": 4.85,
      "learning_rate": 1.5231681769496517e-07,
      "loss": 0.4144,
      "step": 13430,
      "task_loss": 0.7219608426094055
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4089081883430481,
      "epoch": 4.86,
      "learning_rate": 1.4500003740410715e-07,
      "loss": 0.39,
      "step": 13440,
      "task_loss": 0.619770348072052
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3215952217578888,
      "epoch": 4.86,
      "learning_rate": 1.3786298121364392e-07,
      "loss": 0.4686,
      "step": 13450,
      "task_loss": 0.5262433290481567
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2864310145378113,
      "epoch": 4.86,
      "learning_rate": 1.3090568592462603e-07,
      "loss": 0.4086,
      "step": 13460,
      "task_loss": 0.46462225914001465
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3616725206375122,
      "epoch": 4.87,
      "learning_rate": 1.24128187411221e-07,
      "loss": 0.4201,
      "step": 13470,
      "task_loss": 0.4158579111099243
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4216560125350952,
      "epoch": 4.87,
      "learning_rate": 1.1753052062048018e-07,
      "loss": 0.3923,
      "step": 13480,
      "task_loss": 0.676395058631897
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45745834708213806,
      "epoch": 4.88,
      "learning_rate": 1.1111271957219503e-07,
      "loss": 0.4328,
      "step": 13490,
      "task_loss": 0.7130784392356873
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46026918292045593,
      "epoch": 4.88,
      "learning_rate": 1.0487481735870662e-07,
      "loss": 0.4589,
      "step": 13500,
      "task_loss": 0.6129114627838135
    },
    {
      "epoch": 4.88,
      "eval_exact_match": 82.58278145695364,
      "eval_f1": 89.74308413444165,
      "step": 13500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36996591091156006,
      "epoch": 4.88,
      "learning_rate": 9.881684614473861e-08,
      "loss": 0.3992,
      "step": 13510,
      "task_loss": 0.39129143953323364
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.327798068523407,
      "epoch": 4.89,
      "learning_rate": 9.29388371672224e-08,
      "loss": 0.3677,
      "step": 13520,
      "task_loss": 0.5077568292617798
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38932156562805176,
      "epoch": 4.89,
      "learning_rate": 8.724082073515326e-08,
      "loss": 0.4236,
      "step": 13530,
      "task_loss": 0.7245796918869019
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4578113257884979,
      "epoch": 4.89,
      "learning_rate": 8.172282622941562e-08,
      "loss": 0.39,
      "step": 13540,
      "task_loss": 1.0679997205734253
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42175060510635376,
      "epoch": 4.9,
      "learning_rate": 7.638488210265081e-08,
      "loss": 0.4656,
      "step": 13550,
      "task_loss": 0.6959472894668579
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4087962508201599,
      "epoch": 4.9,
      "learning_rate": 7.122701587908619e-08,
      "loss": 0.3932,
      "step": 13560,
      "task_loss": 1.1434520483016968
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3045280873775482,
      "epoch": 4.9,
      "learning_rate": 6.62492541544224e-08,
      "loss": 0.412,
      "step": 13570,
      "task_loss": 0.6849862933158875
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3511340320110321,
      "epoch": 4.91,
      "learning_rate": 6.145162259566628e-08,
      "loss": 0.458,
      "step": 13580,
      "task_loss": 0.5146377086639404
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45488184690475464,
      "epoch": 4.91,
      "learning_rate": 5.683414594102209e-08,
      "loss": 0.4581,
      "step": 13590,
      "task_loss": 0.9683970808982849
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.493674635887146,
      "epoch": 4.92,
      "learning_rate": 5.2396847999751635e-08,
      "loss": 0.3936,
      "step": 13600,
      "task_loss": 1.0288453102111816
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3300265669822693,
      "epoch": 4.92,
      "learning_rate": 4.813975165205763e-08,
      "loss": 0.428,
      "step": 13610,
      "task_loss": 0.43383991718292236
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32362163066864014,
      "epoch": 4.92,
      "learning_rate": 4.406287884895943e-08,
      "loss": 0.4261,
      "step": 13620,
      "task_loss": 0.5160417556762695
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5967044234275818,
      "epoch": 4.93,
      "learning_rate": 4.016625061218026e-08,
      "loss": 0.4086,
      "step": 13630,
      "task_loss": 1.1938400268554688
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3653201758861542,
      "epoch": 4.93,
      "learning_rate": 3.644988703405016e-08,
      "loss": 0.3941,
      "step": 13640,
      "task_loss": 0.6336909532546997
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36608558893203735,
      "epoch": 4.93,
      "learning_rate": 3.2913807277385464e-08,
      "loss": 0.3878,
      "step": 13650,
      "task_loss": 0.532042920589447
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37139713764190674,
      "epoch": 4.94,
      "learning_rate": 2.9558029575407204e-08,
      "loss": 0.4069,
      "step": 13660,
      "task_loss": 0.7530549764633179
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3533708453178406,
      "epoch": 4.94,
      "learning_rate": 2.6382571231628456e-08,
      "loss": 0.397,
      "step": 13670,
      "task_loss": 0.47629207372665405
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4421239495277405,
      "epoch": 4.94,
      "learning_rate": 2.3387448619784367e-08,
      "loss": 0.4319,
      "step": 13680,
      "task_loss": 0.5352086424827576
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5323693156242371,
      "epoch": 4.95,
      "learning_rate": 2.0572677183731122e-08,
      "loss": 0.3947,
      "step": 13690,
      "task_loss": 0.48410868644714355
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5018866062164307,
      "epoch": 4.95,
      "learning_rate": 1.7938271437379913e-08,
      "loss": 0.4427,
      "step": 13700,
      "task_loss": 0.4523616433143616
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3665663003921509,
      "epoch": 4.95,
      "learning_rate": 1.5484244964611427e-08,
      "loss": 0.3782,
      "step": 13710,
      "task_loss": 0.48350703716278076
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4491804242134094,
      "epoch": 4.96,
      "learning_rate": 1.3210610419213673e-08,
      "loss": 0.4083,
      "step": 13720,
      "task_loss": 0.6975788474082947
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3747801184654236,
      "epoch": 4.96,
      "learning_rate": 1.1117379524804293e-08,
      "loss": 0.3877,
      "step": 13730,
      "task_loss": 0.46945083141326904
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3005831837654114,
      "epoch": 4.97,
      "learning_rate": 9.204563074791671e-09,
      "loss": 0.4103,
      "step": 13740,
      "task_loss": 0.437521755695343
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49657440185546875,
      "epoch": 4.97,
      "learning_rate": 7.47217093229724e-09,
      "loss": 0.4278,
      "step": 13750,
      "task_loss": 0.9666340947151184
    },
    {
      "epoch": 4.97,
      "eval_exact_match": 82.72469252601702,
      "eval_f1": 89.75211707399707,
      "step": 13750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5844108462333679,
      "epoch": 4.97,
      "learning_rate": 5.9202120301204966e-09,
      "loss": 0.4205,
      "step": 13760,
      "task_loss": 0.8376038074493408
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5542365312576294,
      "epoch": 4.98,
      "learning_rate": 4.548694370680728e-09,
      "loss": 0.4406,
      "step": 13770,
      "task_loss": 0.7455954551696777
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3821340501308441,
      "epoch": 4.98,
      "learning_rate": 3.3576250259859106e-09,
      "loss": 0.4466,
      "step": 13780,
      "task_loss": 0.43582969903945923
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38751429319381714,
      "epoch": 4.98,
      "learning_rate": 2.3470101375860917e-09,
      "loss": 0.4559,
      "step": 13790,
      "task_loss": 0.4809384346008301
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40226811170578003,
      "epoch": 4.99,
      "learning_rate": 1.5168549165578415e-09,
      "loss": 0.4583,
      "step": 13800,
      "task_loss": 0.4811709523200989
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.500964343547821,
      "epoch": 4.99,
      "learning_rate": 8.671636434576246e-10,
      "loss": 0.3806,
      "step": 13810,
      "task_loss": 0.6287086009979248
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3395386338233948,
      "epoch": 4.99,
      "learning_rate": 3.97939668314029e-10,
      "loss": 0.4134,
      "step": 13820,
      "task_loss": 0.35841506719589233
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3570495843887329,
      "epoch": 5.0,
      "learning_rate": 1.0918541060056518e-10,
      "loss": 0.4233,
      "step": 13830,
      "task_loss": 0.5482752323150635
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35532039403915405,
      "epoch": 5.0,
      "learning_rate": 9.023592317802808e-13,
      "loss": 0.3687,
      "step": 13840,
      "task_loss": 0.36665064096450806
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35440778732299805,
      "epoch": 5.01,
      "learning_rate": 6.999992690892745e-05,
      "loss": 0.3737,
      "step": 13850,
      "task_loss": 0.6965274810791016
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.48675596714019775,
      "epoch": 5.01,
      "learning_rate": 6.999967424882167e-05,
      "loss": 0.5142,
      "step": 13860,
      "task_loss": 0.669815182685852
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3265790343284607,
      "epoch": 5.01,
      "learning_rate": 6.999924111862622e-05,
      "loss": 0.3724,
      "step": 13870,
      "task_loss": 0.48695600032806396
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5362226366996765,
      "epoch": 5.02,
      "learning_rate": 6.999862752057449e-05,
      "loss": 0.419,
      "step": 13880,
      "task_loss": 0.6971112489700317
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4938550889492035,
      "epoch": 5.02,
      "learning_rate": 6.999783345783038e-05,
      "loss": 0.4973,
      "step": 13890,
      "task_loss": 0.6833978891372681
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4292599558830261,
      "epoch": 5.02,
      "learning_rate": 6.999685893448834e-05,
      "loss": 0.4466,
      "step": 13900,
      "task_loss": 0.7468749284744263
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3701336979866028,
      "epoch": 5.03,
      "learning_rate": 6.999570395557335e-05,
      "loss": 0.4276,
      "step": 13910,
      "task_loss": 0.3578958511352539
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4636634886264801,
      "epoch": 5.03,
      "learning_rate": 6.999436852704088e-05,
      "loss": 0.4046,
      "step": 13920,
      "task_loss": 0.5777134895324707
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34819674491882324,
      "epoch": 5.03,
      "learning_rate": 6.999285265577683e-05,
      "loss": 0.3958,
      "step": 13930,
      "task_loss": 0.7134641408920288
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4702969789505005,
      "epoch": 5.04,
      "learning_rate": 6.999115634959755e-05,
      "loss": 0.4482,
      "step": 13940,
      "task_loss": 0.7330292463302612
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4083574414253235,
      "epoch": 5.04,
      "learning_rate": 6.998927961724976e-05,
      "loss": 0.4519,
      "step": 13950,
      "task_loss": 0.5460534691810608
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40470677614212036,
      "epoch": 5.05,
      "learning_rate": 6.998722246841052e-05,
      "loss": 0.42,
      "step": 13960,
      "task_loss": 0.8615512847900391
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.501833438873291,
      "epoch": 5.05,
      "learning_rate": 6.99849849136872e-05,
      "loss": 0.4176,
      "step": 13970,
      "task_loss": 0.6687218546867371
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.48426300287246704,
      "epoch": 5.05,
      "learning_rate": 6.998256696461737e-05,
      "loss": 0.4576,
      "step": 13980,
      "task_loss": 0.5862292051315308
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5200551748275757,
      "epoch": 5.06,
      "learning_rate": 6.997996863366879e-05,
      "loss": 0.4555,
      "step": 13990,
      "task_loss": 0.7281656861305237
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5517895817756653,
      "epoch": 5.06,
      "learning_rate": 6.997718993423931e-05,
      "loss": 0.4606,
      "step": 14000,
      "task_loss": 0.4926382005214691
    },
    {
      "epoch": 5.06,
      "eval_exact_match": 82.39356669820246,
      "eval_f1": 89.59166032418017,
      "step": 14000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4386638402938843,
      "epoch": 5.06,
      "learning_rate": 6.997423088065688e-05,
      "loss": 0.4664,
      "step": 14010,
      "task_loss": 0.6251969337463379
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44594234228134155,
      "epoch": 5.07,
      "learning_rate": 6.997109148817935e-05,
      "loss": 0.4358,
      "step": 14020,
      "task_loss": 0.4952147305011749
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3212473690509796,
      "epoch": 5.07,
      "learning_rate": 6.996777177299449e-05,
      "loss": 0.4691,
      "step": 14030,
      "task_loss": 0.2904544472694397
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3446178734302521,
      "epoch": 5.07,
      "learning_rate": 6.996427175221984e-05,
      "loss": 0.414,
      "step": 14040,
      "task_loss": 0.5980858206748962
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3580099642276764,
      "epoch": 5.08,
      "learning_rate": 6.996059144390272e-05,
      "loss": 0.4941,
      "step": 14050,
      "task_loss": 0.28045976161956787
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4789047837257385,
      "epoch": 5.08,
      "learning_rate": 6.995673086702003e-05,
      "loss": 0.5185,
      "step": 14060,
      "task_loss": 0.751612663269043
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33268317580223083,
      "epoch": 5.08,
      "learning_rate": 6.99526900414782e-05,
      "loss": 0.4453,
      "step": 14070,
      "task_loss": 0.2582695186138153
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5172496438026428,
      "epoch": 5.09,
      "learning_rate": 6.994846898811308e-05,
      "loss": 0.4784,
      "step": 14080,
      "task_loss": 0.5228416919708252
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3266606330871582,
      "epoch": 5.09,
      "learning_rate": 6.994406772868984e-05,
      "loss": 0.5077,
      "step": 14090,
      "task_loss": 0.6353722214698792
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46428632736206055,
      "epoch": 5.1,
      "learning_rate": 6.993948628590286e-05,
      "loss": 0.4278,
      "step": 14100,
      "task_loss": 0.3377443253993988
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5082235336303711,
      "epoch": 5.1,
      "learning_rate": 6.993472468337559e-05,
      "loss": 0.447,
      "step": 14110,
      "task_loss": 0.6395173072814941
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3576199412345886,
      "epoch": 5.1,
      "learning_rate": 6.992978294566047e-05,
      "loss": 0.5164,
      "step": 14120,
      "task_loss": 0.6690906286239624
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40793901681900024,
      "epoch": 5.11,
      "learning_rate": 6.992466109823875e-05,
      "loss": 0.4775,
      "step": 14130,
      "task_loss": 0.8127427101135254
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4388144910335541,
      "epoch": 5.11,
      "learning_rate": 6.99193591675204e-05,
      "loss": 0.4678,
      "step": 14140,
      "task_loss": 0.5744211673736572
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.551649272441864,
      "epoch": 5.11,
      "learning_rate": 6.991387718084395e-05,
      "loss": 0.4929,
      "step": 14150,
      "task_loss": 0.5900485515594482
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35258394479751587,
      "epoch": 5.12,
      "learning_rate": 6.990821516647637e-05,
      "loss": 0.4451,
      "step": 14160,
      "task_loss": 0.7655850648880005
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41839784383773804,
      "epoch": 5.12,
      "learning_rate": 6.990237315361293e-05,
      "loss": 0.4519,
      "step": 14170,
      "task_loss": 0.4728570878505707
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7037980556488037,
      "epoch": 5.12,
      "learning_rate": 6.9896351172377e-05,
      "loss": 0.4855,
      "step": 14180,
      "task_loss": 0.881999135017395
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4729958474636078,
      "epoch": 5.13,
      "learning_rate": 6.98901492538199e-05,
      "loss": 0.4912,
      "step": 14190,
      "task_loss": 0.4268884062767029
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40473929047584534,
      "epoch": 5.13,
      "learning_rate": 6.988376742992088e-05,
      "loss": 0.4478,
      "step": 14200,
      "task_loss": 0.701400876045227
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3299444913864136,
      "epoch": 5.14,
      "learning_rate": 6.987720573358672e-05,
      "loss": 0.4597,
      "step": 14210,
      "task_loss": 0.4115176796913147
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39635318517684937,
      "epoch": 5.14,
      "learning_rate": 6.987046419865176e-05,
      "loss": 0.4558,
      "step": 14220,
      "task_loss": 0.5531545877456665
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3222476840019226,
      "epoch": 5.14,
      "learning_rate": 6.986354285987759e-05,
      "loss": 0.4689,
      "step": 14230,
      "task_loss": 0.35380634665489197
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39281386137008667,
      "epoch": 5.15,
      "learning_rate": 6.985644175295299e-05,
      "loss": 0.4214,
      "step": 14240,
      "task_loss": 0.5921287536621094
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3763327896595001,
      "epoch": 5.15,
      "learning_rate": 6.984916091449365e-05,
      "loss": 0.4603,
      "step": 14250,
      "task_loss": 0.30917519330978394
    },
    {
      "epoch": 5.15,
      "eval_exact_match": 82.19489120151371,
      "eval_f1": 89.36849233451807,
      "step": 14250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5615275502204895,
      "epoch": 5.15,
      "learning_rate": 6.984170038204199e-05,
      "loss": 0.4981,
      "step": 14260,
      "task_loss": 0.6949574947357178
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47236767411231995,
      "epoch": 5.16,
      "learning_rate": 6.983406019406708e-05,
      "loss": 0.5041,
      "step": 14270,
      "task_loss": 0.5212032794952393
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5434203743934631,
      "epoch": 5.16,
      "learning_rate": 6.982624038996425e-05,
      "loss": 0.4769,
      "step": 14280,
      "task_loss": 0.6844668388366699
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36416149139404297,
      "epoch": 5.16,
      "learning_rate": 6.981824101005505e-05,
      "loss": 0.5013,
      "step": 14290,
      "task_loss": 0.5063308477401733
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46571964025497437,
      "epoch": 5.17,
      "learning_rate": 6.9810062095587e-05,
      "loss": 0.4407,
      "step": 14300,
      "task_loss": 0.5314142107963562
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4478929340839386,
      "epoch": 5.17,
      "learning_rate": 6.98017036887333e-05,
      "loss": 0.4981,
      "step": 14310,
      "task_loss": 0.43525004386901855
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3345271348953247,
      "epoch": 5.18,
      "learning_rate": 6.979316583259272e-05,
      "loss": 0.4351,
      "step": 14320,
      "task_loss": 0.249060720205307
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5977888703346252,
      "epoch": 5.18,
      "learning_rate": 6.978444857118933e-05,
      "loss": 0.5105,
      "step": 14330,
      "task_loss": 0.731816828250885
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3528749942779541,
      "epoch": 5.18,
      "learning_rate": 6.977555194947223e-05,
      "loss": 0.4528,
      "step": 14340,
      "task_loss": 0.28791236877441406
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4057888984680176,
      "epoch": 5.19,
      "learning_rate": 6.976647601331542e-05,
      "loss": 0.4821,
      "step": 14350,
      "task_loss": 0.7864526510238647
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5908790826797485,
      "epoch": 5.19,
      "learning_rate": 6.975722080951746e-05,
      "loss": 0.5164,
      "step": 14360,
      "task_loss": 1.0292553901672363
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5158011317253113,
      "epoch": 5.19,
      "learning_rate": 6.97477863858013e-05,
      "loss": 0.5173,
      "step": 14370,
      "task_loss": 0.44501206278800964
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.644179105758667,
      "epoch": 5.2,
      "learning_rate": 6.973817279081402e-05,
      "loss": 0.4935,
      "step": 14380,
      "task_loss": 0.6516858339309692
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3709027171134949,
      "epoch": 5.2,
      "learning_rate": 6.972838007412653e-05,
      "loss": 0.4828,
      "step": 14390,
      "task_loss": 0.5542988181114197
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3429642617702484,
      "epoch": 5.2,
      "learning_rate": 6.971840828623337e-05,
      "loss": 0.4879,
      "step": 14400,
      "task_loss": 0.3810282349586487
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41922885179519653,
      "epoch": 5.21,
      "learning_rate": 6.970825747855243e-05,
      "loss": 0.4552,
      "step": 14410,
      "task_loss": 0.7148609757423401
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47263821959495544,
      "epoch": 5.21,
      "learning_rate": 6.96979277034247e-05,
      "loss": 0.5117,
      "step": 14420,
      "task_loss": 0.8885880708694458
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47650307416915894,
      "epoch": 5.22,
      "learning_rate": 6.968847793264618e-05,
      "loss": 0.5001,
      "step": 14430,
      "task_loss": 0.48124372959136963
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3229578137397766,
      "epoch": 5.22,
      "learning_rate": 6.967780826687376e-05,
      "loss": 0.4864,
      "step": 14440,
      "task_loss": 0.2538095712661743
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4400070309638977,
      "epoch": 5.22,
      "learning_rate": 6.966695979066094e-05,
      "loss": 0.5075,
      "step": 14450,
      "task_loss": 0.5561120510101318
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.450109601020813,
      "epoch": 5.23,
      "learning_rate": 6.965593255994606e-05,
      "loss": 0.4603,
      "step": 14460,
      "task_loss": 0.457144170999527
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7270958423614502,
      "epoch": 5.23,
      "learning_rate": 6.964472663158928e-05,
      "loss": 0.5067,
      "step": 14470,
      "task_loss": 0.876134991645813
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4349331855773926,
      "epoch": 5.23,
      "learning_rate": 6.96333420633721e-05,
      "loss": 0.4893,
      "step": 14480,
      "task_loss": 0.48338618874549866
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.386316180229187,
      "epoch": 5.24,
      "learning_rate": 6.962177891399719e-05,
      "loss": 0.4693,
      "step": 14490,
      "task_loss": 0.5884779095649719
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5028120875358582,
      "epoch": 5.24,
      "learning_rate": 6.961003724308804e-05,
      "loss": 0.4485,
      "step": 14500,
      "task_loss": 0.5481046438217163
    },
    {
      "epoch": 5.24,
      "eval_exact_match": 82.15704824976348,
      "eval_f1": 89.39737203821956,
      "step": 14500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6019538044929504,
      "epoch": 5.24,
      "learning_rate": 6.959811711118866e-05,
      "loss": 0.5045,
      "step": 14510,
      "task_loss": 0.676105260848999
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45118898153305054,
      "epoch": 5.25,
      "learning_rate": 6.958601857976325e-05,
      "loss": 0.4433,
      "step": 14520,
      "task_loss": 0.6423743963241577
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5065474510192871,
      "epoch": 5.25,
      "learning_rate": 6.957374171119591e-05,
      "loss": 0.5745,
      "step": 14530,
      "task_loss": 0.7627721428871155
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45010310411453247,
      "epoch": 5.25,
      "learning_rate": 6.956128656879031e-05,
      "loss": 0.5029,
      "step": 14540,
      "task_loss": 0.8689513206481934
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45659008622169495,
      "epoch": 5.26,
      "learning_rate": 6.954865321676934e-05,
      "loss": 0.5493,
      "step": 14550,
      "task_loss": 0.7313786745071411
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4748663902282715,
      "epoch": 5.26,
      "learning_rate": 6.953584172027481e-05,
      "loss": 0.4825,
      "step": 14560,
      "task_loss": 0.6921977996826172
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43287229537963867,
      "epoch": 5.27,
      "learning_rate": 6.95228521453671e-05,
      "loss": 0.4728,
      "step": 14570,
      "task_loss": 0.41803377866744995
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5268940329551697,
      "epoch": 5.27,
      "learning_rate": 6.950968455902485e-05,
      "loss": 0.4534,
      "step": 14580,
      "task_loss": 0.6837814450263977
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.538477897644043,
      "epoch": 5.27,
      "learning_rate": 6.949633902914455e-05,
      "loss": 0.5323,
      "step": 14590,
      "task_loss": 0.6758741736412048
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3660999834537506,
      "epoch": 5.28,
      "learning_rate": 6.948281562454023e-05,
      "loss": 0.447,
      "step": 14600,
      "task_loss": 0.39962196350097656
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.508263885974884,
      "epoch": 5.28,
      "learning_rate": 6.946911441494312e-05,
      "loss": 0.5491,
      "step": 14610,
      "task_loss": 0.38159501552581787
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.475387305021286,
      "epoch": 5.28,
      "learning_rate": 6.945523547100128e-05,
      "loss": 0.5048,
      "step": 14620,
      "task_loss": 0.46061187982559204
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6909676790237427,
      "epoch": 5.29,
      "learning_rate": 6.944117886427917e-05,
      "loss": 0.5538,
      "step": 14630,
      "task_loss": 1.0875879526138306
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49635446071624756,
      "epoch": 5.29,
      "learning_rate": 6.942694466725742e-05,
      "loss": 0.5124,
      "step": 14640,
      "task_loss": 1.1167693138122559
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47104009985923767,
      "epoch": 5.29,
      "learning_rate": 6.941253295333232e-05,
      "loss": 0.4429,
      "step": 14650,
      "task_loss": 0.5356221199035645
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4203311502933502,
      "epoch": 5.3,
      "learning_rate": 6.939794379681553e-05,
      "loss": 0.4496,
      "step": 14660,
      "task_loss": 0.6668910980224609
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39971667528152466,
      "epoch": 5.3,
      "learning_rate": 6.938317727293362e-05,
      "loss": 0.4446,
      "step": 14670,
      "task_loss": 0.8728127479553223
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46285662055015564,
      "epoch": 5.31,
      "learning_rate": 6.936823345782777e-05,
      "loss": 0.5239,
      "step": 14680,
      "task_loss": 0.6147375106811523
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44424545764923096,
      "epoch": 5.31,
      "learning_rate": 6.935311242855331e-05,
      "loss": 0.4991,
      "step": 14690,
      "task_loss": 0.25694024562835693
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5805097222328186,
      "epoch": 5.31,
      "learning_rate": 6.933781426307934e-05,
      "loss": 0.4837,
      "step": 14700,
      "task_loss": 0.7926748991012573
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5545125603675842,
      "epoch": 5.32,
      "learning_rate": 6.932233904028838e-05,
      "loss": 0.4692,
      "step": 14710,
      "task_loss": 0.8997831344604492
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3451814651489258,
      "epoch": 5.32,
      "learning_rate": 6.930668683997586e-05,
      "loss": 0.44,
      "step": 14720,
      "task_loss": 0.3470460772514343
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5201274156570435,
      "epoch": 5.32,
      "learning_rate": 6.929085774284978e-05,
      "loss": 0.4791,
      "step": 14730,
      "task_loss": 0.5729685425758362
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6551119089126587,
      "epoch": 5.33,
      "learning_rate": 6.92748518305303e-05,
      "loss": 0.4964,
      "step": 14740,
      "task_loss": 0.686562180519104
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.408242791891098,
      "epoch": 5.33,
      "learning_rate": 6.925866918554928e-05,
      "loss": 0.4589,
      "step": 14750,
      "task_loss": 0.7477397918701172
    },
    {
      "epoch": 5.33,
      "eval_exact_match": 81.8543046357616,
      "eval_f1": 88.98935594178413,
      "step": 14750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4235571026802063,
      "epoch": 5.33,
      "learning_rate": 6.924230989134988e-05,
      "loss": 0.5103,
      "step": 14760,
      "task_loss": 0.44014132022857666
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46613240242004395,
      "epoch": 5.34,
      "learning_rate": 6.922577403228613e-05,
      "loss": 0.4946,
      "step": 14770,
      "task_loss": 0.4285615086555481
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44623327255249023,
      "epoch": 5.34,
      "learning_rate": 6.920906169362248e-05,
      "loss": 0.4949,
      "step": 14780,
      "task_loss": 0.2965691089630127
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46191853284835815,
      "epoch": 5.35,
      "learning_rate": 6.919217296153334e-05,
      "loss": 0.4888,
      "step": 14790,
      "task_loss": 0.44412630796432495
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.48645418882369995,
      "epoch": 5.35,
      "learning_rate": 6.917510792310273e-05,
      "loss": 0.4859,
      "step": 14800,
      "task_loss": 0.7811020612716675
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5859341621398926,
      "epoch": 5.35,
      "learning_rate": 6.915786666632371e-05,
      "loss": 0.4877,
      "step": 14810,
      "task_loss": 0.5216963291168213
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8215360641479492,
      "epoch": 5.36,
      "learning_rate": 6.914044928009801e-05,
      "loss": 0.5259,
      "step": 14820,
      "task_loss": 0.6187966465950012
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5492468476295471,
      "epoch": 5.36,
      "learning_rate": 6.91228558542355e-05,
      "loss": 0.5066,
      "step": 14830,
      "task_loss": 0.8015393614768982
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5337282419204712,
      "epoch": 5.36,
      "learning_rate": 6.910508647945382e-05,
      "loss": 0.4744,
      "step": 14840,
      "task_loss": 0.6190358400344849
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4521459639072418,
      "epoch": 5.37,
      "learning_rate": 6.908714124737785e-05,
      "loss": 0.5216,
      "step": 14850,
      "task_loss": 0.6480435729026794
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4558612108230591,
      "epoch": 5.37,
      "learning_rate": 6.906902025053921e-05,
      "loss": 0.5544,
      "step": 14860,
      "task_loss": 0.7935390472412109
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4677425026893616,
      "epoch": 5.37,
      "learning_rate": 6.905072358237589e-05,
      "loss": 0.5016,
      "step": 14870,
      "task_loss": 0.836277961730957
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.601864218711853,
      "epoch": 5.38,
      "learning_rate": 6.903225133723164e-05,
      "loss": 0.4917,
      "step": 14880,
      "task_loss": 0.6987572908401489
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44784462451934814,
      "epoch": 5.38,
      "learning_rate": 6.901360361035558e-05,
      "loss": 0.5084,
      "step": 14890,
      "task_loss": 0.40038225054740906
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49896252155303955,
      "epoch": 5.38,
      "learning_rate": 6.899478049790166e-05,
      "loss": 0.4782,
      "step": 14900,
      "task_loss": 1.08770751953125
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5893778800964355,
      "epoch": 5.39,
      "learning_rate": 6.897578209692816e-05,
      "loss": 0.5051,
      "step": 14910,
      "task_loss": 0.8863139152526855
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.596195638179779,
      "epoch": 5.39,
      "learning_rate": 6.895660850539724e-05,
      "loss": 0.5446,
      "step": 14920,
      "task_loss": 0.41577187180519104
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49143242835998535,
      "epoch": 5.4,
      "learning_rate": 6.89372598221744e-05,
      "loss": 0.4568,
      "step": 14930,
      "task_loss": 0.6072136163711548
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4943245053291321,
      "epoch": 5.4,
      "learning_rate": 6.891773614702792e-05,
      "loss": 0.4862,
      "step": 14940,
      "task_loss": 0.7640190124511719
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.660693347454071,
      "epoch": 5.4,
      "learning_rate": 6.889803758062846e-05,
      "loss": 0.5279,
      "step": 14950,
      "task_loss": 0.9499300718307495
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46248817443847656,
      "epoch": 5.41,
      "learning_rate": 6.887816422454846e-05,
      "loss": 0.4533,
      "step": 14960,
      "task_loss": 0.48348772525787354
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6976580023765564,
      "epoch": 5.41,
      "learning_rate": 6.885811618126159e-05,
      "loss": 0.4955,
      "step": 14970,
      "task_loss": 0.6008883714675903
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5361173152923584,
      "epoch": 5.41,
      "learning_rate": 6.883789355414233e-05,
      "loss": 0.5052,
      "step": 14980,
      "task_loss": 0.601929783821106
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45808547735214233,
      "epoch": 5.42,
      "learning_rate": 6.881749644746535e-05,
      "loss": 0.4697,
      "step": 14990,
      "task_loss": 0.6224072575569153
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44869574904441833,
      "epoch": 5.42,
      "learning_rate": 6.879692496640498e-05,
      "loss": 0.5494,
      "step": 15000,
      "task_loss": 0.5406436920166016
    },
    {
      "epoch": 5.42,
      "eval_exact_match": 81.91106906338695,
      "eval_f1": 89.24159681473488,
      "step": 15000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44767728447914124,
      "epoch": 5.42,
      "learning_rate": 6.877617921703468e-05,
      "loss": 0.4827,
      "step": 15010,
      "task_loss": 0.668562650680542
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5660895109176636,
      "epoch": 5.43,
      "learning_rate": 6.875525930632653e-05,
      "loss": 0.5327,
      "step": 15020,
      "task_loss": 1.034258246421814
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35861772298812866,
      "epoch": 5.43,
      "learning_rate": 6.873416534215064e-05,
      "loss": 0.4806,
      "step": 15030,
      "task_loss": 0.42596346139907837
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7044867277145386,
      "epoch": 5.44,
      "learning_rate": 6.871289743327455e-05,
      "loss": 0.5416,
      "step": 15040,
      "task_loss": 0.7539388537406921
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43850019574165344,
      "epoch": 5.44,
      "learning_rate": 6.869145568936275e-05,
      "loss": 0.4962,
      "step": 15050,
      "task_loss": 0.5783032178878784
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4840550124645233,
      "epoch": 5.44,
      "learning_rate": 6.866984022097612e-05,
      "loss": 0.5386,
      "step": 15060,
      "task_loss": 0.5743952989578247
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.663632869720459,
      "epoch": 5.45,
      "learning_rate": 6.864805113957123e-05,
      "loss": 0.4495,
      "step": 15070,
      "task_loss": 0.8771752119064331
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41818153858184814,
      "epoch": 5.45,
      "learning_rate": 6.862608855749995e-05,
      "loss": 0.5517,
      "step": 15080,
      "task_loss": 0.8001173734664917
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4510330557823181,
      "epoch": 5.45,
      "learning_rate": 6.860395258800871e-05,
      "loss": 0.5632,
      "step": 15090,
      "task_loss": 0.5343620181083679
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5964381694793701,
      "epoch": 5.46,
      "learning_rate": 6.858164334523806e-05,
      "loss": 0.5241,
      "step": 15100,
      "task_loss": 0.8330484628677368
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5306263566017151,
      "epoch": 5.46,
      "learning_rate": 6.85591609442219e-05,
      "loss": 0.4959,
      "step": 15110,
      "task_loss": 0.6922684907913208
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46480873227119446,
      "epoch": 5.46,
      "learning_rate": 6.853650550088709e-05,
      "loss": 0.5503,
      "step": 15120,
      "task_loss": 0.5510062575340271
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4726712107658386,
      "epoch": 5.47,
      "learning_rate": 6.85136771320527e-05,
      "loss": 0.4861,
      "step": 15130,
      "task_loss": 0.7079441547393799
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.635477602481842,
      "epoch": 5.47,
      "learning_rate": 6.849067595542946e-05,
      "loss": 0.5073,
      "step": 15140,
      "task_loss": 0.8118971586227417
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5491691827774048,
      "epoch": 5.48,
      "learning_rate": 6.846750208961921e-05,
      "loss": 0.5041,
      "step": 15150,
      "task_loss": 0.645523726940155
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33065298199653625,
      "epoch": 5.48,
      "learning_rate": 6.844415565411416e-05,
      "loss": 0.4876,
      "step": 15160,
      "task_loss": 0.7148600816726685
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44501185417175293,
      "epoch": 5.48,
      "learning_rate": 6.84206367692964e-05,
      "loss": 0.45,
      "step": 15170,
      "task_loss": 0.5103074908256531
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.437028169631958,
      "epoch": 5.49,
      "learning_rate": 6.83969455564372e-05,
      "loss": 0.5025,
      "step": 15180,
      "task_loss": 0.6442121863365173
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46236348152160645,
      "epoch": 5.49,
      "learning_rate": 6.837308213769643e-05,
      "loss": 0.4766,
      "step": 15190,
      "task_loss": 0.6244036555290222
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4882710576057434,
      "epoch": 5.49,
      "learning_rate": 6.834904663612188e-05,
      "loss": 0.5092,
      "step": 15200,
      "task_loss": 0.4923517107963562
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3791254758834839,
      "epoch": 5.5,
      "learning_rate": 6.832483917564871e-05,
      "loss": 0.4735,
      "step": 15210,
      "task_loss": 0.476325660943985
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4039973020553589,
      "epoch": 5.5,
      "learning_rate": 6.830045988109869e-05,
      "loss": 0.5144,
      "step": 15220,
      "task_loss": 0.6185945868492126
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6345975995063782,
      "epoch": 5.5,
      "learning_rate": 6.827590887817969e-05,
      "loss": 0.4869,
      "step": 15230,
      "task_loss": 0.6183285713195801
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5272897481918335,
      "epoch": 5.51,
      "learning_rate": 6.825118629348493e-05,
      "loss": 0.4712,
      "step": 15240,
      "task_loss": 0.9741804599761963
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34770917892456055,
      "epoch": 5.51,
      "learning_rate": 6.822629225449237e-05,
      "loss": 0.4669,
      "step": 15250,
      "task_loss": 0.3193807899951935
    },
    {
      "epoch": 5.51,
      "eval_exact_match": 81.79754020813624,
      "eval_f1": 89.026956018613,
      "step": 15250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5802989602088928,
      "epoch": 5.51,
      "learning_rate": 6.820122688956404e-05,
      "loss": 0.5421,
      "step": 15260,
      "task_loss": 0.6691771745681763
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4500080645084381,
      "epoch": 5.52,
      "learning_rate": 6.817599032794539e-05,
      "loss": 0.478,
      "step": 15270,
      "task_loss": 0.7451514005661011
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5461714267730713,
      "epoch": 5.52,
      "learning_rate": 6.815058269976462e-05,
      "loss": 0.5201,
      "step": 15280,
      "task_loss": 0.36617377400398254
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5489475727081299,
      "epoch": 5.53,
      "learning_rate": 6.8125004136032e-05,
      "loss": 0.5483,
      "step": 15290,
      "task_loss": 0.44468170404434204
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5517098307609558,
      "epoch": 5.53,
      "learning_rate": 6.809925476863924e-05,
      "loss": 0.5487,
      "step": 15300,
      "task_loss": 0.901218831539154
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3634655177593231,
      "epoch": 5.53,
      "learning_rate": 6.807333473035868e-05,
      "loss": 0.4936,
      "step": 15310,
      "task_loss": 0.552862823009491
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6331474184989929,
      "epoch": 5.54,
      "learning_rate": 6.804724415484278e-05,
      "loss": 0.5806,
      "step": 15320,
      "task_loss": 0.6623814105987549
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5808169841766357,
      "epoch": 5.54,
      "learning_rate": 6.802098317662334e-05,
      "loss": 0.5248,
      "step": 15330,
      "task_loss": 0.5808233022689819
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.482837051153183,
      "epoch": 5.54,
      "learning_rate": 6.799455193111077e-05,
      "loss": 0.4598,
      "step": 15340,
      "task_loss": 0.7158746719360352
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6607789993286133,
      "epoch": 5.55,
      "learning_rate": 6.796795055459347e-05,
      "loss": 0.5437,
      "step": 15350,
      "task_loss": 1.0560665130615234
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4435341954231262,
      "epoch": 5.55,
      "learning_rate": 6.794117918423706e-05,
      "loss": 0.5636,
      "step": 15360,
      "task_loss": 0.648557186126709
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4403044581413269,
      "epoch": 5.55,
      "learning_rate": 6.791423795808378e-05,
      "loss": 0.4541,
      "step": 15370,
      "task_loss": 0.4011836051940918
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4444364309310913,
      "epoch": 5.56,
      "learning_rate": 6.78871270150516e-05,
      "loss": 0.5244,
      "step": 15380,
      "task_loss": 0.5661810636520386
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7234218716621399,
      "epoch": 5.56,
      "learning_rate": 6.785984649493368e-05,
      "loss": 0.5102,
      "step": 15390,
      "task_loss": 1.0226433277130127
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4961375296115875,
      "epoch": 5.57,
      "learning_rate": 6.783239653839755e-05,
      "loss": 0.4689,
      "step": 15400,
      "task_loss": 0.6303238868713379
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4570309519767761,
      "epoch": 5.57,
      "learning_rate": 6.780477728698443e-05,
      "loss": 0.4629,
      "step": 15410,
      "task_loss": 0.5990092158317566
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4736815094947815,
      "epoch": 5.57,
      "learning_rate": 6.777698888310844e-05,
      "loss": 0.5097,
      "step": 15420,
      "task_loss": 0.5480905175209045
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6659838557243347,
      "epoch": 5.58,
      "learning_rate": 6.774903147005596e-05,
      "loss": 0.4949,
      "step": 15430,
      "task_loss": 0.7123420834541321
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5285578966140747,
      "epoch": 5.58,
      "learning_rate": 6.772090519198481e-05,
      "loss": 0.4649,
      "step": 15440,
      "task_loss": 0.6163167953491211
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.564858615398407,
      "epoch": 5.58,
      "learning_rate": 6.769261019392352e-05,
      "loss": 0.4921,
      "step": 15450,
      "task_loss": 1.2447395324707031
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47404301166534424,
      "epoch": 5.59,
      "learning_rate": 6.766414662177064e-05,
      "loss": 0.5179,
      "step": 15460,
      "task_loss": 0.5387969613075256
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47900664806365967,
      "epoch": 5.59,
      "learning_rate": 6.76355146222939e-05,
      "loss": 0.5314,
      "step": 15470,
      "task_loss": 0.7243887186050415
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.48273512721061707,
      "epoch": 5.59,
      "learning_rate": 6.760671434312953e-05,
      "loss": 0.4921,
      "step": 15480,
      "task_loss": 0.7808024287223816
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.52320396900177,
      "epoch": 5.6,
      "learning_rate": 6.757774593278144e-05,
      "loss": 0.5465,
      "step": 15490,
      "task_loss": 0.8334228992462158
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.531326949596405,
      "epoch": 5.6,
      "learning_rate": 6.75486095406205e-05,
      "loss": 0.5283,
      "step": 15500,
      "task_loss": 1.0520827770233154
    },
    {
      "epoch": 5.6,
      "eval_exact_match": 82.081362346263,
      "eval_f1": 89.29942274768753,
      "step": 15500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6806000471115112,
      "epoch": 5.61,
      "learning_rate": 6.751930531688376e-05,
      "loss": 0.5496,
      "step": 15510,
      "task_loss": 0.7352679967880249
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7652064561843872,
      "epoch": 5.61,
      "learning_rate": 6.748983341267363e-05,
      "loss": 0.5732,
      "step": 15520,
      "task_loss": 0.9669309854507446
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3168491721153259,
      "epoch": 5.61,
      "learning_rate": 6.746019397995717e-05,
      "loss": 0.547,
      "step": 15530,
      "task_loss": 0.6193458437919617
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2838849723339081,
      "epoch": 5.62,
      "learning_rate": 6.743038717156527e-05,
      "loss": 0.5339,
      "step": 15540,
      "task_loss": 0.47925713658332825
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5559385418891907,
      "epoch": 5.62,
      "learning_rate": 6.740041314119183e-05,
      "loss": 0.5182,
      "step": 15550,
      "task_loss": 0.6141558885574341
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45640748739242554,
      "epoch": 5.62,
      "learning_rate": 6.737027204339306e-05,
      "loss": 0.519,
      "step": 15560,
      "task_loss": 0.5364757776260376
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4492012858390808,
      "epoch": 5.63,
      "learning_rate": 6.733996403358657e-05,
      "loss": 0.4872,
      "step": 15570,
      "task_loss": 0.5102091431617737
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4823322892189026,
      "epoch": 5.63,
      "learning_rate": 6.730948926805068e-05,
      "loss": 0.4538,
      "step": 15580,
      "task_loss": 0.6215452551841736
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6813780069351196,
      "epoch": 5.63,
      "learning_rate": 6.72788479039235e-05,
      "loss": 0.4924,
      "step": 15590,
      "task_loss": 0.8112049698829651
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4826444387435913,
      "epoch": 5.64,
      "learning_rate": 6.724804009920222e-05,
      "loss": 0.5361,
      "step": 15600,
      "task_loss": 0.40426796674728394
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3255290389060974,
      "epoch": 5.64,
      "learning_rate": 6.721706601274225e-05,
      "loss": 0.5145,
      "step": 15610,
      "task_loss": 0.38015216588974
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44477927684783936,
      "epoch": 5.65,
      "learning_rate": 6.718592580425637e-05,
      "loss": 0.4861,
      "step": 15620,
      "task_loss": 0.6369027495384216
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7055771946907043,
      "epoch": 5.65,
      "learning_rate": 6.715461963431398e-05,
      "loss": 0.5927,
      "step": 15630,
      "task_loss": 1.0351436138153076
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4551340341567993,
      "epoch": 5.65,
      "learning_rate": 6.712314766434023e-05,
      "loss": 0.5257,
      "step": 15640,
      "task_loss": 0.7271813154220581
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.621050238609314,
      "epoch": 5.66,
      "learning_rate": 6.709151005661517e-05,
      "loss": 0.4921,
      "step": 15650,
      "task_loss": 1.096197247505188
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5006800293922424,
      "epoch": 5.66,
      "learning_rate": 6.705970697427294e-05,
      "loss": 0.517,
      "step": 15660,
      "task_loss": 0.6352694630622864
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6934675574302673,
      "epoch": 5.66,
      "learning_rate": 6.702773858130093e-05,
      "loss": 0.5508,
      "step": 15670,
      "task_loss": 0.721714437007904
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3144254684448242,
      "epoch": 5.67,
      "learning_rate": 6.699560504253894e-05,
      "loss": 0.4457,
      "step": 15680,
      "task_loss": 0.34663382172584534
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.48802831768989563,
      "epoch": 5.67,
      "learning_rate": 6.696330652367827e-05,
      "loss": 0.6041,
      "step": 15690,
      "task_loss": 0.6961667537689209
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4313547611236572,
      "epoch": 5.67,
      "learning_rate": 6.693084319126098e-05,
      "loss": 0.5342,
      "step": 15700,
      "task_loss": 0.3994500935077667
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4418693780899048,
      "epoch": 5.68,
      "learning_rate": 6.68982152126789e-05,
      "loss": 0.5396,
      "step": 15710,
      "task_loss": 0.8848820924758911
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5503214597702026,
      "epoch": 5.68,
      "learning_rate": 6.686542275617286e-05,
      "loss": 0.5216,
      "step": 15720,
      "task_loss": 0.5391052961349487
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.53646320104599,
      "epoch": 5.68,
      "learning_rate": 6.683246599083182e-05,
      "loss": 0.5336,
      "step": 15730,
      "task_loss": 1.1558949947357178
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3631494641304016,
      "epoch": 5.69,
      "learning_rate": 6.679934508659192e-05,
      "loss": 0.4783,
      "step": 15740,
      "task_loss": 0.4442519545555115
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5602353811264038,
      "epoch": 5.69,
      "learning_rate": 6.676606021423567e-05,
      "loss": 0.5216,
      "step": 15750,
      "task_loss": 0.6735915541648865
    },
    {
      "epoch": 5.69,
      "eval_exact_match": 82.52601702932829,
      "eval_f1": 89.46075198123665,
      "step": 15750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5539681911468506,
      "epoch": 5.7,
      "learning_rate": 6.673261154539109e-05,
      "loss": 0.4855,
      "step": 15760,
      "task_loss": 0.7846614718437195
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4747917652130127,
      "epoch": 5.7,
      "learning_rate": 6.669899925253077e-05,
      "loss": 0.4877,
      "step": 15770,
      "task_loss": 0.5126049518585205
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4245802164077759,
      "epoch": 5.7,
      "learning_rate": 6.666522350897096e-05,
      "loss": 0.4613,
      "step": 15780,
      "task_loss": 0.37426063418388367
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5295876264572144,
      "epoch": 5.71,
      "learning_rate": 6.663128448887077e-05,
      "loss": 0.5046,
      "step": 15790,
      "task_loss": 0.9215235710144043
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3816160559654236,
      "epoch": 5.71,
      "learning_rate": 6.65971823672312e-05,
      "loss": 0.4756,
      "step": 15800,
      "task_loss": 0.7307689189910889
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38228023052215576,
      "epoch": 5.71,
      "learning_rate": 6.656291731989428e-05,
      "loss": 0.5442,
      "step": 15810,
      "task_loss": 0.6373156309127808
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5242533683776855,
      "epoch": 5.72,
      "learning_rate": 6.65319396218302e-05,
      "loss": 0.5753,
      "step": 15820,
      "task_loss": 0.5102771520614624
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5250728726387024,
      "epoch": 5.72,
      "learning_rate": 6.649736550312107e-05,
      "loss": 0.5102,
      "step": 15830,
      "task_loss": 0.2875036597251892
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6699880957603455,
      "epoch": 5.72,
      "learning_rate": 6.646262897340388e-05,
      "loss": 0.5943,
      "step": 15840,
      "task_loss": 0.606027364730835
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4970390796661377,
      "epoch": 5.73,
      "learning_rate": 6.642773021179186e-05,
      "loss": 0.5663,
      "step": 15850,
      "task_loss": 0.7026304006576538
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47369736433029175,
      "epoch": 5.73,
      "learning_rate": 6.639266939823477e-05,
      "loss": 0.491,
      "step": 15860,
      "task_loss": 0.6342217326164246
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3805849254131317,
      "epoch": 5.74,
      "learning_rate": 6.635744671351794e-05,
      "loss": 0.4797,
      "step": 15870,
      "task_loss": 0.6766672730445862
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.502840518951416,
      "epoch": 5.74,
      "learning_rate": 6.632206233926139e-05,
      "loss": 0.5094,
      "step": 15880,
      "task_loss": 0.918179452419281
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36127620935440063,
      "epoch": 5.74,
      "learning_rate": 6.628651645791885e-05,
      "loss": 0.462,
      "step": 15890,
      "task_loss": 0.4396474361419678
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6621323823928833,
      "epoch": 5.75,
      "learning_rate": 6.625080925277681e-05,
      "loss": 0.557,
      "step": 15900,
      "task_loss": 0.8044998049736023
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38126713037490845,
      "epoch": 5.75,
      "learning_rate": 6.621494090795367e-05,
      "loss": 0.54,
      "step": 15910,
      "task_loss": 0.8128173351287842
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4991961121559143,
      "epoch": 5.75,
      "learning_rate": 6.617891160839865e-05,
      "loss": 0.478,
      "step": 15920,
      "task_loss": 0.7899810671806335
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36616766452789307,
      "epoch": 5.76,
      "learning_rate": 6.614272153989095e-05,
      "loss": 0.5037,
      "step": 15930,
      "task_loss": 0.43909940123558044
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4640388488769531,
      "epoch": 5.76,
      "learning_rate": 6.61063708890387e-05,
      "loss": 0.5291,
      "step": 15940,
      "task_loss": 0.7651124000549316
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43500208854675293,
      "epoch": 5.76,
      "learning_rate": 6.606985984327813e-05,
      "loss": 0.5923,
      "step": 15950,
      "task_loss": 0.37657132744789124
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4381626844406128,
      "epoch": 5.77,
      "learning_rate": 6.603318859087243e-05,
      "loss": 0.4844,
      "step": 15960,
      "task_loss": 0.4404764771461487
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4864192306995392,
      "epoch": 5.77,
      "learning_rate": 6.599635732091092e-05,
      "loss": 0.456,
      "step": 15970,
      "task_loss": 0.7914941310882568
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49039989709854126,
      "epoch": 5.78,
      "learning_rate": 6.595936622330802e-05,
      "loss": 0.5051,
      "step": 15980,
      "task_loss": 0.4711025357246399
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36692336201667786,
      "epoch": 5.78,
      "learning_rate": 6.592221548880224e-05,
      "loss": 0.4975,
      "step": 15990,
      "task_loss": 0.5510057210922241
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6534066200256348,
      "epoch": 5.78,
      "learning_rate": 6.58849053089553e-05,
      "loss": 0.4932,
      "step": 16000,
      "task_loss": 0.8516380786895752
    },
    {
      "epoch": 5.78,
      "eval_exact_match": 82.42194891201514,
      "eval_f1": 89.36242276057637,
      "step": 16000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7806046605110168,
      "epoch": 5.79,
      "learning_rate": 6.584743587615102e-05,
      "loss": 0.5337,
      "step": 16010,
      "task_loss": 0.8906459808349609
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4783743619918823,
      "epoch": 5.79,
      "learning_rate": 6.580980738359438e-05,
      "loss": 0.4975,
      "step": 16020,
      "task_loss": 0.3754185140132904
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5132753849029541,
      "epoch": 5.79,
      "learning_rate": 6.577202002531056e-05,
      "loss": 0.4836,
      "step": 16030,
      "task_loss": 0.824865460395813
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.708114743232727,
      "epoch": 5.8,
      "learning_rate": 6.573407399614388e-05,
      "loss": 0.5799,
      "step": 16040,
      "task_loss": 0.8528615832328796
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8225008845329285,
      "epoch": 5.8,
      "learning_rate": 6.569596949175681e-05,
      "loss": 0.5178,
      "step": 16050,
      "task_loss": 1.1864490509033203
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.663002073764801,
      "epoch": 5.8,
      "learning_rate": 6.5657706708629e-05,
      "loss": 0.5239,
      "step": 16060,
      "task_loss": 0.9480224251747131
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7957673072814941,
      "epoch": 5.81,
      "learning_rate": 6.561928584405624e-05,
      "loss": 0.5899,
      "step": 16070,
      "task_loss": 0.778226912021637
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47262027859687805,
      "epoch": 5.81,
      "learning_rate": 6.558070709614942e-05,
      "loss": 0.5108,
      "step": 16080,
      "task_loss": 0.9569321870803833
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5292235612869263,
      "epoch": 5.81,
      "learning_rate": 6.55419706638335e-05,
      "loss": 0.5378,
      "step": 16090,
      "task_loss": 0.8886809349060059
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40345680713653564,
      "epoch": 5.82,
      "learning_rate": 6.550307674684662e-05,
      "loss": 0.503,
      "step": 16100,
      "task_loss": 0.5904554128646851
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41043514013290405,
      "epoch": 5.82,
      "learning_rate": 6.546402554573885e-05,
      "loss": 0.4817,
      "step": 16110,
      "task_loss": 0.40754806995391846
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5127359628677368,
      "epoch": 5.83,
      "learning_rate": 6.54248172618713e-05,
      "loss": 0.5225,
      "step": 16120,
      "task_loss": 0.7525652647018433
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7493129968643188,
      "epoch": 5.83,
      "learning_rate": 6.538545209741511e-05,
      "loss": 0.4626,
      "step": 16130,
      "task_loss": 0.5919231176376343
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5420757532119751,
      "epoch": 5.83,
      "learning_rate": 6.534593025535028e-05,
      "loss": 0.5729,
      "step": 16140,
      "task_loss": 0.7412921190261841
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6038966178894043,
      "epoch": 5.84,
      "learning_rate": 6.530625193946472e-05,
      "loss": 0.5041,
      "step": 16150,
      "task_loss": 0.8964495658874512
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46763065457344055,
      "epoch": 5.84,
      "learning_rate": 6.526641735435317e-05,
      "loss": 0.4891,
      "step": 16160,
      "task_loss": 0.6115823984146118
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5795122385025024,
      "epoch": 5.84,
      "learning_rate": 6.522642670541613e-05,
      "loss": 0.5359,
      "step": 16170,
      "task_loss": 0.57799232006073
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40503954887390137,
      "epoch": 5.85,
      "learning_rate": 6.518628019885888e-05,
      "loss": 0.4867,
      "step": 16180,
      "task_loss": 0.34715238213539124
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5611152648925781,
      "epoch": 5.85,
      "learning_rate": 6.514597804169025e-05,
      "loss": 0.5164,
      "step": 16190,
      "task_loss": 0.9298588633537292
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6309351325035095,
      "epoch": 5.85,
      "learning_rate": 6.510552044172176e-05,
      "loss": 0.4337,
      "step": 16200,
      "task_loss": 0.5857481956481934
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5015100240707397,
      "epoch": 5.86,
      "learning_rate": 6.506490760756639e-05,
      "loss": 0.4809,
      "step": 16210,
      "task_loss": 0.5657186508178711
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44960838556289673,
      "epoch": 5.86,
      "learning_rate": 6.502413974863753e-05,
      "loss": 0.4203,
      "step": 16220,
      "task_loss": 0.8054425120353699
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40969163179397583,
      "epoch": 5.87,
      "learning_rate": 6.498321707514802e-05,
      "loss": 0.448,
      "step": 16230,
      "task_loss": 0.49577945470809937
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6154443025588989,
      "epoch": 5.87,
      "learning_rate": 6.494213979810891e-05,
      "loss": 0.5039,
      "step": 16240,
      "task_loss": 0.8134857416152954
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5586275458335876,
      "epoch": 5.87,
      "learning_rate": 6.490090812932844e-05,
      "loss": 0.5517,
      "step": 16250,
      "task_loss": 0.3241579532623291
    },
    {
      "epoch": 5.87,
      "eval_exact_match": 81.8070009460738,
      "eval_f1": 89.10508584535329,
      "step": 16250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5156681537628174,
      "epoch": 5.88,
      "learning_rate": 6.485952228141097e-05,
      "loss": 0.4911,
      "step": 16260,
      "task_loss": 0.801060140132904
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5016605854034424,
      "epoch": 5.88,
      "learning_rate": 6.481798246775586e-05,
      "loss": 0.5218,
      "step": 16270,
      "task_loss": 0.8665897846221924
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36766764521598816,
      "epoch": 5.88,
      "learning_rate": 6.477628890255634e-05,
      "loss": 0.5,
      "step": 16280,
      "task_loss": 0.3850083351135254
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44555163383483887,
      "epoch": 5.89,
      "learning_rate": 6.473444180079845e-05,
      "loss": 0.4808,
      "step": 16290,
      "task_loss": 0.6322591304779053
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5872722268104553,
      "epoch": 5.89,
      "learning_rate": 6.469244137825993e-05,
      "loss": 0.4952,
      "step": 16300,
      "task_loss": 0.8481529951095581
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4090435802936554,
      "epoch": 5.89,
      "learning_rate": 6.465028785150908e-05,
      "loss": 0.6102,
      "step": 16310,
      "task_loss": 0.7143661975860596
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3229040503501892,
      "epoch": 5.9,
      "learning_rate": 6.460798143790366e-05,
      "loss": 0.5282,
      "step": 16320,
      "task_loss": 0.5130259990692139
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6251620650291443,
      "epoch": 5.9,
      "learning_rate": 6.456552235558976e-05,
      "loss": 0.5402,
      "step": 16330,
      "task_loss": 0.684100866317749
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.48610326647758484,
      "epoch": 5.91,
      "learning_rate": 6.452291082350068e-05,
      "loss": 0.4923,
      "step": 16340,
      "task_loss": 0.6871728897094727
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42472466826438904,
      "epoch": 5.91,
      "learning_rate": 6.448014706135582e-05,
      "loss": 0.4976,
      "step": 16350,
      "task_loss": 0.5411844253540039
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4997144937515259,
      "epoch": 5.91,
      "learning_rate": 6.443723128965951e-05,
      "loss": 0.5019,
      "step": 16360,
      "task_loss": 0.5415253639221191
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4492264986038208,
      "epoch": 5.92,
      "learning_rate": 6.439416372969992e-05,
      "loss": 0.5249,
      "step": 16370,
      "task_loss": 0.951138973236084
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6077766418457031,
      "epoch": 5.92,
      "learning_rate": 6.435094460354784e-05,
      "loss": 0.5132,
      "step": 16380,
      "task_loss": 0.8307206630706787
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5573058724403381,
      "epoch": 5.92,
      "learning_rate": 6.430757413405562e-05,
      "loss": 0.5879,
      "step": 16390,
      "task_loss": 0.5353271961212158
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4729651212692261,
      "epoch": 5.93,
      "learning_rate": 6.426405254485603e-05,
      "loss": 0.5174,
      "step": 16400,
      "task_loss": 0.5413171052932739
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5121825337409973,
      "epoch": 5.93,
      "learning_rate": 6.422038006036097e-05,
      "loss": 0.5453,
      "step": 16410,
      "task_loss": 0.8983778357505798
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4299796223640442,
      "epoch": 5.93,
      "learning_rate": 6.417655690576046e-05,
      "loss": 0.6098,
      "step": 16420,
      "task_loss": 0.8138757944107056
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7405078411102295,
      "epoch": 5.94,
      "learning_rate": 6.413258330702145e-05,
      "loss": 0.5537,
      "step": 16430,
      "task_loss": 0.8167030811309814
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4909617304801941,
      "epoch": 5.94,
      "learning_rate": 6.408845949088657e-05,
      "loss": 0.4987,
      "step": 16440,
      "task_loss": 0.5977470874786377
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49547016620635986,
      "epoch": 5.95,
      "learning_rate": 6.404418568487308e-05,
      "loss": 0.5069,
      "step": 16450,
      "task_loss": 0.8176012635231018
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3618711531162262,
      "epoch": 5.95,
      "learning_rate": 6.39997621172716e-05,
      "loss": 0.4346,
      "step": 16460,
      "task_loss": 0.3624705672264099
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4783799648284912,
      "epoch": 5.95,
      "learning_rate": 6.395518901714497e-05,
      "loss": 0.5274,
      "step": 16470,
      "task_loss": 0.9623761177062988
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33645719289779663,
      "epoch": 5.96,
      "learning_rate": 6.391046661432711e-05,
      "loss": 0.5243,
      "step": 16480,
      "task_loss": 0.6531622409820557
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4208140969276428,
      "epoch": 5.96,
      "learning_rate": 6.386559513942175e-05,
      "loss": 0.4774,
      "step": 16490,
      "task_loss": 0.6000040769577026
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6129087209701538,
      "epoch": 5.96,
      "learning_rate": 6.38205748238013e-05,
      "loss": 0.5298,
      "step": 16500,
      "task_loss": 0.7589020729064941
    },
    {
      "epoch": 5.96,
      "eval_exact_match": 81.72185430463576,
      "eval_f1": 89.09431894529595,
      "step": 16500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.560535192489624,
      "epoch": 5.97,
      "learning_rate": 6.377540589960567e-05,
      "loss": 0.5601,
      "step": 16510,
      "task_loss": 0.5373163819313049
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5602094531059265,
      "epoch": 5.97,
      "learning_rate": 6.373008859974099e-05,
      "loss": 0.5261,
      "step": 16520,
      "task_loss": 1.0255897045135498
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4283807873725891,
      "epoch": 5.97,
      "learning_rate": 6.36846231578785e-05,
      "loss": 0.5167,
      "step": 16530,
      "task_loss": 0.655482292175293
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4438017010688782,
      "epoch": 5.98,
      "learning_rate": 6.363900980845333e-05,
      "loss": 0.4951,
      "step": 16540,
      "task_loss": 0.4216102361679077
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5425772070884705,
      "epoch": 5.98,
      "learning_rate": 6.359324878666324e-05,
      "loss": 0.5276,
      "step": 16550,
      "task_loss": 0.9083492755889893
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.610127329826355,
      "epoch": 5.98,
      "learning_rate": 6.354734032846744e-05,
      "loss": 0.5713,
      "step": 16560,
      "task_loss": 1.105069875717163
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40713953971862793,
      "epoch": 5.99,
      "learning_rate": 6.350128467058539e-05,
      "loss": 0.4681,
      "step": 16570,
      "task_loss": 0.43503206968307495
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7117682695388794,
      "epoch": 5.99,
      "learning_rate": 6.345508205049552e-05,
      "loss": 0.5597,
      "step": 16580,
      "task_loss": 0.6516414880752563
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7078899145126343,
      "epoch": 6.0,
      "learning_rate": 6.340873270643411e-05,
      "loss": 0.5851,
      "step": 16590,
      "task_loss": 0.6792216300964355
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6714178323745728,
      "epoch": 6.0,
      "learning_rate": 6.336223687739394e-05,
      "loss": 0.5065,
      "step": 16600,
      "task_loss": 0.7536165118217468
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32467812299728394,
      "epoch": 6.0,
      "learning_rate": 6.331559480312315e-05,
      "loss": 0.4598,
      "step": 16610,
      "task_loss": 0.392375111579895
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4250606298446655,
      "epoch": 6.01,
      "learning_rate": 6.326880672412396e-05,
      "loss": 0.3817,
      "step": 16620,
      "task_loss": 0.5436099171638489
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40928852558135986,
      "epoch": 6.01,
      "learning_rate": 6.322187288165144e-05,
      "loss": 0.4347,
      "step": 16630,
      "task_loss": 0.4154428243637085
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.593374490737915,
      "epoch": 6.01,
      "learning_rate": 6.317479351771226e-05,
      "loss": 0.4448,
      "step": 16640,
      "task_loss": 1.1337848901748657
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6083024144172668,
      "epoch": 6.02,
      "learning_rate": 6.312756887506345e-05,
      "loss": 0.4901,
      "step": 16650,
      "task_loss": 1.052654504776001
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39533329010009766,
      "epoch": 6.02,
      "learning_rate": 6.308019919721113e-05,
      "loss": 0.4397,
      "step": 16660,
      "task_loss": 0.5021288394927979
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42493703961372375,
      "epoch": 6.02,
      "learning_rate": 6.303268472840934e-05,
      "loss": 0.4398,
      "step": 16670,
      "task_loss": 0.6849949359893799
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39546722173690796,
      "epoch": 6.03,
      "learning_rate": 6.29850257136586e-05,
      "loss": 0.4657,
      "step": 16680,
      "task_loss": 0.6547192931175232
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.57149338722229,
      "epoch": 6.03,
      "learning_rate": 6.293722239870485e-05,
      "loss": 0.4044,
      "step": 16690,
      "task_loss": 0.834121584892273
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3922726511955261,
      "epoch": 6.04,
      "learning_rate": 6.288927503003805e-05,
      "loss": 0.429,
      "step": 16700,
      "task_loss": 0.40333420038223267
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3624475598335266,
      "epoch": 6.04,
      "learning_rate": 6.284118385489095e-05,
      "loss": 0.4963,
      "step": 16710,
      "task_loss": 0.7349679470062256
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40686216950416565,
      "epoch": 6.04,
      "learning_rate": 6.279294912123784e-05,
      "loss": 0.409,
      "step": 16720,
      "task_loss": 0.7947590947151184
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.395913302898407,
      "epoch": 6.05,
      "learning_rate": 6.27445710777932e-05,
      "loss": 0.3941,
      "step": 16730,
      "task_loss": 0.6178956031799316
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42935115098953247,
      "epoch": 6.05,
      "learning_rate": 6.269604997401051e-05,
      "loss": 0.4276,
      "step": 16740,
      "task_loss": 0.6020127534866333
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41655611991882324,
      "epoch": 6.05,
      "learning_rate": 6.264738606008087e-05,
      "loss": 0.4512,
      "step": 16750,
      "task_loss": 0.4740205407142639
    },
    {
      "epoch": 6.05,
      "eval_exact_match": 82.19489120151371,
      "eval_f1": 89.3732801354867,
      "step": 16750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4346901774406433,
      "epoch": 6.06,
      "learning_rate": 6.259857958693182e-05,
      "loss": 0.3777,
      "step": 16760,
      "task_loss": 0.8214448690414429
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5657368302345276,
      "epoch": 6.06,
      "learning_rate": 6.254963080622591e-05,
      "loss": 0.4877,
      "step": 16770,
      "task_loss": 0.7357831001281738
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5232205986976624,
      "epoch": 6.06,
      "learning_rate": 6.250053997035956e-05,
      "loss": 0.4555,
      "step": 16780,
      "task_loss": 1.022857904434204
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43613356351852417,
      "epoch": 6.07,
      "learning_rate": 6.245130733246159e-05,
      "loss": 0.4097,
      "step": 16790,
      "task_loss": 1.0982863903045654
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4957405924797058,
      "epoch": 6.07,
      "learning_rate": 6.240193314639205e-05,
      "loss": 0.4478,
      "step": 16800,
      "task_loss": 1.0710291862487793
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3766104280948639,
      "epoch": 6.08,
      "learning_rate": 6.235241766674084e-05,
      "loss": 0.4634,
      "step": 16810,
      "task_loss": 0.608417272567749
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38640034198760986,
      "epoch": 6.08,
      "learning_rate": 6.230276114882642e-05,
      "loss": 0.3841,
      "step": 16820,
      "task_loss": 0.48826298117637634
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4803047776222229,
      "epoch": 6.08,
      "learning_rate": 6.225296384869451e-05,
      "loss": 0.4249,
      "step": 16830,
      "task_loss": 0.7844042778015137
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3516848683357239,
      "epoch": 6.09,
      "learning_rate": 6.220302602311674e-05,
      "loss": 0.4653,
      "step": 16840,
      "task_loss": 0.49346923828125
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.499299556016922,
      "epoch": 6.09,
      "learning_rate": 6.21529479295893e-05,
      "loss": 0.4503,
      "step": 16850,
      "task_loss": 0.5536627769470215
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29132914543151855,
      "epoch": 6.09,
      "learning_rate": 6.21027298263317e-05,
      "loss": 0.3977,
      "step": 16860,
      "task_loss": 0.4496367275714874
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4155663847923279,
      "epoch": 6.1,
      "learning_rate": 6.205237197228537e-05,
      "loss": 0.4114,
      "step": 16870,
      "task_loss": 0.8739005327224731
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4325769245624542,
      "epoch": 6.1,
      "learning_rate": 6.200187462711232e-05,
      "loss": 0.457,
      "step": 16880,
      "task_loss": 0.3728763461112976
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3581392168998718,
      "epoch": 6.1,
      "learning_rate": 6.195123805119386e-05,
      "loss": 0.4158,
      "step": 16890,
      "task_loss": 0.6587432622909546
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34714964032173157,
      "epoch": 6.11,
      "learning_rate": 6.19004625056292e-05,
      "loss": 0.4553,
      "step": 16900,
      "task_loss": 0.939296305179596
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46273887157440186,
      "epoch": 6.11,
      "learning_rate": 6.184954825223412e-05,
      "loss": 0.444,
      "step": 16910,
      "task_loss": 0.7927017211914062
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3514648973941803,
      "epoch": 6.11,
      "learning_rate": 6.179849555353966e-05,
      "loss": 0.4455,
      "step": 16920,
      "task_loss": 0.5540809631347656
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33696162700653076,
      "epoch": 6.12,
      "learning_rate": 6.174730467279065e-05,
      "loss": 0.438,
      "step": 16930,
      "task_loss": 0.6107865571975708
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5959919095039368,
      "epoch": 6.12,
      "learning_rate": 6.169597587394453e-05,
      "loss": 0.4469,
      "step": 16940,
      "task_loss": 0.7320927977561951
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.616447389125824,
      "epoch": 6.13,
      "learning_rate": 6.164450942166983e-05,
      "loss": 0.4048,
      "step": 16950,
      "task_loss": 0.6176183819770813
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5620734691619873,
      "epoch": 6.13,
      "learning_rate": 6.159290558134487e-05,
      "loss": 0.4278,
      "step": 16960,
      "task_loss": 0.6595306396484375
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3424335718154907,
      "epoch": 6.13,
      "learning_rate": 6.154116461905642e-05,
      "loss": 0.4275,
      "step": 16970,
      "task_loss": 0.6462994813919067
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6205884218215942,
      "epoch": 6.14,
      "learning_rate": 6.14892868015983e-05,
      "loss": 0.4362,
      "step": 16980,
      "task_loss": 0.7719211578369141
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35494518280029297,
      "epoch": 6.14,
      "learning_rate": 6.143727239646995e-05,
      "loss": 0.3998,
      "step": 16990,
      "task_loss": 0.2818894386291504
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5920026302337646,
      "epoch": 6.14,
      "learning_rate": 6.138512167187514e-05,
      "loss": 0.4274,
      "step": 17000,
      "task_loss": 0.43748384714126587
    },
    {
      "epoch": 6.14,
      "eval_exact_match": 82.16650898770104,
      "eval_f1": 89.3841314898369,
      "step": 17000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6196434497833252,
      "epoch": 6.15,
      "learning_rate": 6.133283489672054e-05,
      "loss": 0.4509,
      "step": 17010,
      "task_loss": 1.4450292587280273
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49087828397750854,
      "epoch": 6.15,
      "learning_rate": 6.128041234061437e-05,
      "loss": 0.4569,
      "step": 17020,
      "task_loss": 0.7537893056869507
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42802369594573975,
      "epoch": 6.15,
      "learning_rate": 6.122785427386493e-05,
      "loss": 0.3961,
      "step": 17030,
      "task_loss": 0.43074288964271545
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43579283356666565,
      "epoch": 6.16,
      "learning_rate": 6.117516096747929e-05,
      "loss": 0.4457,
      "step": 17040,
      "task_loss": 0.842292070388794
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5460898876190186,
      "epoch": 6.16,
      "learning_rate": 6.112233269316187e-05,
      "loss": 0.4811,
      "step": 17050,
      "task_loss": 0.6467231512069702
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4218035638332367,
      "epoch": 6.17,
      "learning_rate": 6.106936972331298e-05,
      "loss": 0.4519,
      "step": 17060,
      "task_loss": 0.5069524645805359
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5733527541160583,
      "epoch": 6.17,
      "learning_rate": 6.101627233102756e-05,
      "loss": 0.5014,
      "step": 17070,
      "task_loss": 1.1166313886642456
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29548192024230957,
      "epoch": 6.17,
      "learning_rate": 6.096304079009358e-05,
      "loss": 0.4181,
      "step": 17080,
      "task_loss": 0.5228471159934998
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39930418133735657,
      "epoch": 6.18,
      "learning_rate": 6.090967537499077e-05,
      "loss": 0.4604,
      "step": 17090,
      "task_loss": 0.322740763425827
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3291262984275818,
      "epoch": 6.18,
      "learning_rate": 6.085617636088917e-05,
      "loss": 0.4097,
      "step": 17100,
      "task_loss": 0.35267138481140137
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4520864188671112,
      "epoch": 6.18,
      "learning_rate": 6.080254402364767e-05,
      "loss": 0.4308,
      "step": 17110,
      "task_loss": 0.6222145557403564
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39761024713516235,
      "epoch": 6.19,
      "learning_rate": 6.074877863981264e-05,
      "loss": 0.3809,
      "step": 17120,
      "task_loss": 0.6330628395080566
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38448137044906616,
      "epoch": 6.19,
      "learning_rate": 6.06948804866165e-05,
      "loss": 0.3893,
      "step": 17130,
      "task_loss": 0.5772709846496582
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39714741706848145,
      "epoch": 6.19,
      "learning_rate": 6.0640849841976206e-05,
      "loss": 0.4316,
      "step": 17140,
      "task_loss": 0.3851231336593628
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34323185682296753,
      "epoch": 6.2,
      "learning_rate": 6.058668698449197e-05,
      "loss": 0.4345,
      "step": 17150,
      "task_loss": 0.42983704805374146
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5064968466758728,
      "epoch": 6.2,
      "learning_rate": 6.0532392193445684e-05,
      "loss": 0.4824,
      "step": 17160,
      "task_loss": 0.749971866607666
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30737966299057007,
      "epoch": 6.21,
      "learning_rate": 6.047796574879955e-05,
      "loss": 0.4194,
      "step": 17170,
      "task_loss": 0.5119467973709106
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.531447172164917,
      "epoch": 6.21,
      "learning_rate": 6.0423407931194616e-05,
      "loss": 0.465,
      "step": 17180,
      "task_loss": 0.43263059854507446
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33223626017570496,
      "epoch": 6.21,
      "learning_rate": 6.036871902194934e-05,
      "loss": 0.3791,
      "step": 17190,
      "task_loss": 0.6570825576782227
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5735246539115906,
      "epoch": 6.22,
      "learning_rate": 6.031389930305813e-05,
      "loss": 0.4322,
      "step": 17200,
      "task_loss": 0.7190378308296204
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47174426913261414,
      "epoch": 6.22,
      "learning_rate": 6.025894905718988e-05,
      "loss": 0.4194,
      "step": 17210,
      "task_loss": 0.8839192390441895
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5682074427604675,
      "epoch": 6.22,
      "learning_rate": 6.0203868567686544e-05,
      "loss": 0.4753,
      "step": 17220,
      "task_loss": 0.7814291715621948
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5652934908866882,
      "epoch": 6.23,
      "learning_rate": 6.014865811856164e-05,
      "loss": 0.4415,
      "step": 17230,
      "task_loss": 0.7998194694519043
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3965378999710083,
      "epoch": 6.23,
      "learning_rate": 6.009331799449884e-05,
      "loss": 0.4436,
      "step": 17240,
      "task_loss": 0.5939300656318665
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35603681206703186,
      "epoch": 6.23,
      "learning_rate": 6.003784848085037e-05,
      "loss": 0.4256,
      "step": 17250,
      "task_loss": 0.35019153356552124
    },
    {
      "epoch": 6.23,
      "eval_exact_match": 82.64900662251655,
      "eval_f1": 89.77520373181453,
      "step": 17250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49196216464042664,
      "epoch": 6.24,
      "learning_rate": 5.998224986363576e-05,
      "loss": 0.4834,
      "step": 17260,
      "task_loss": 1.102785587310791
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5928846597671509,
      "epoch": 6.24,
      "learning_rate": 5.992652242954014e-05,
      "loss": 0.4438,
      "step": 17270,
      "task_loss": 0.4663623571395874
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45873433351516724,
      "epoch": 6.25,
      "learning_rate": 5.987066646591291e-05,
      "loss": 0.4328,
      "step": 17280,
      "task_loss": 0.49007028341293335
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3176501989364624,
      "epoch": 6.25,
      "learning_rate": 5.9814682260766194e-05,
      "loss": 0.4403,
      "step": 17290,
      "task_loss": 0.5457800030708313
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3774525225162506,
      "epoch": 6.25,
      "learning_rate": 5.975857010277339e-05,
      "loss": 0.3944,
      "step": 17300,
      "task_loss": 0.7043144106864929
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34644702076911926,
      "epoch": 6.26,
      "learning_rate": 5.9702330281267646e-05,
      "loss": 0.4241,
      "step": 17310,
      "task_loss": 0.34404587745666504
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3558131754398346,
      "epoch": 6.26,
      "learning_rate": 5.9645963086240405e-05,
      "loss": 0.389,
      "step": 17320,
      "task_loss": 0.5505707263946533
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4252021908760071,
      "epoch": 6.26,
      "learning_rate": 5.958946880833986e-05,
      "loss": 0.4379,
      "step": 17330,
      "task_loss": 0.6420973539352417
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36581557989120483,
      "epoch": 6.27,
      "learning_rate": 5.953284773886952e-05,
      "loss": 0.4012,
      "step": 17340,
      "task_loss": 0.37061429023742676
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.48756957054138184,
      "epoch": 6.27,
      "learning_rate": 5.947610016978664e-05,
      "loss": 0.4317,
      "step": 17350,
      "task_loss": 0.8574424982070923
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45216867327690125,
      "epoch": 6.27,
      "learning_rate": 5.94192263937008e-05,
      "loss": 0.4094,
      "step": 17360,
      "task_loss": 0.5074952840805054
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6077759265899658,
      "epoch": 6.28,
      "learning_rate": 5.936222670387228e-05,
      "loss": 0.438,
      "step": 17370,
      "task_loss": 0.6022177338600159
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46468818187713623,
      "epoch": 6.28,
      "learning_rate": 5.930510139421068e-05,
      "loss": 0.4133,
      "step": 17380,
      "task_loss": 0.602958619594574
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2893716096878052,
      "epoch": 6.28,
      "learning_rate": 5.924785075927328e-05,
      "loss": 0.4087,
      "step": 17390,
      "task_loss": 0.3538118600845337
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36932480335235596,
      "epoch": 6.29,
      "learning_rate": 5.919047509426362e-05,
      "loss": 0.4407,
      "step": 17400,
      "task_loss": 0.4979381561279297
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6854021549224854,
      "epoch": 6.29,
      "learning_rate": 5.913297469502991e-05,
      "loss": 0.443,
      "step": 17410,
      "task_loss": 0.5967050790786743
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5585907697677612,
      "epoch": 6.3,
      "learning_rate": 5.907534985806355e-05,
      "loss": 0.4198,
      "step": 17420,
      "task_loss": 0.7241256833076477
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33727091550827026,
      "epoch": 6.3,
      "learning_rate": 5.901760088049758e-05,
      "loss": 0.4343,
      "step": 17430,
      "task_loss": 0.4912257492542267
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3864668309688568,
      "epoch": 6.3,
      "learning_rate": 5.895972806010516e-05,
      "loss": 0.4325,
      "step": 17440,
      "task_loss": 0.8334698677062988
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36405840516090393,
      "epoch": 6.31,
      "learning_rate": 5.890173169529798e-05,
      "loss": 0.4259,
      "step": 17450,
      "task_loss": 0.6634342670440674
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30496692657470703,
      "epoch": 6.31,
      "learning_rate": 5.884361208512483e-05,
      "loss": 0.443,
      "step": 17460,
      "task_loss": 0.5642157196998596
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42527127265930176,
      "epoch": 6.31,
      "learning_rate": 5.878536952926994e-05,
      "loss": 0.4407,
      "step": 17470,
      "task_loss": 0.6014546751976013
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4635869860649109,
      "epoch": 6.32,
      "learning_rate": 5.872700432805154e-05,
      "loss": 0.4449,
      "step": 17480,
      "task_loss": 0.7708749175071716
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6213729977607727,
      "epoch": 6.32,
      "learning_rate": 5.866851678242021e-05,
      "loss": 0.4247,
      "step": 17490,
      "task_loss": 0.6365109086036682
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3652482032775879,
      "epoch": 6.32,
      "learning_rate": 5.86099071939574e-05,
      "loss": 0.4398,
      "step": 17500,
      "task_loss": 0.36544913053512573
    },
    {
      "epoch": 6.32,
      "eval_exact_match": 82.37464522232735,
      "eval_f1": 89.41607657285384,
      "step": 17500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3291020393371582,
      "epoch": 6.33,
      "learning_rate": 5.855117586487388e-05,
      "loss": 0.3939,
      "step": 17510,
      "task_loss": 0.3664117455482483
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4412246346473694,
      "epoch": 6.33,
      "learning_rate": 5.849232309800812e-05,
      "loss": 0.4223,
      "step": 17520,
      "task_loss": 0.7607033252716064
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5140312910079956,
      "epoch": 6.34,
      "learning_rate": 5.8433349196824764e-05,
      "loss": 0.3941,
      "step": 17530,
      "task_loss": 1.1523401737213135
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39518868923187256,
      "epoch": 6.34,
      "learning_rate": 5.83742544654131e-05,
      "loss": 0.4747,
      "step": 17540,
      "task_loss": 0.605819582939148
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6016864776611328,
      "epoch": 6.34,
      "learning_rate": 5.831503920848542e-05,
      "loss": 0.4245,
      "step": 17550,
      "task_loss": 0.6810322403907776
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4904579222202301,
      "epoch": 6.35,
      "learning_rate": 5.825570373137551e-05,
      "loss": 0.4252,
      "step": 17560,
      "task_loss": 0.44462671875953674
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5831754207611084,
      "epoch": 6.35,
      "learning_rate": 5.819624834003702e-05,
      "loss": 0.4582,
      "step": 17570,
      "task_loss": 0.7443471550941467
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5073849558830261,
      "epoch": 6.35,
      "learning_rate": 5.8136673341041975e-05,
      "loss": 0.4428,
      "step": 17580,
      "task_loss": 0.38186776638031006
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41992390155792236,
      "epoch": 6.36,
      "learning_rate": 5.807697904157908e-05,
      "loss": 0.4408,
      "step": 17590,
      "task_loss": 0.7133923172950745
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3772296905517578,
      "epoch": 6.36,
      "learning_rate": 5.801716574945222e-05,
      "loss": 0.4687,
      "step": 17600,
      "task_loss": 0.5094341039657593
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47077786922454834,
      "epoch": 6.36,
      "learning_rate": 5.795723377307885e-05,
      "loss": 0.4438,
      "step": 17610,
      "task_loss": 0.4625624418258667
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3527754545211792,
      "epoch": 6.37,
      "learning_rate": 5.789718342148839e-05,
      "loss": 0.3928,
      "step": 17620,
      "task_loss": 0.2926297187805176
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5881116390228271,
      "epoch": 6.37,
      "learning_rate": 5.783701500432064e-05,
      "loss": 0.4462,
      "step": 17630,
      "task_loss": 0.8188558220863342
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4231293797492981,
      "epoch": 6.38,
      "learning_rate": 5.777672883182419e-05,
      "loss": 0.451,
      "step": 17640,
      "task_loss": 0.427749902009964
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4589058756828308,
      "epoch": 6.38,
      "learning_rate": 5.771632521485482e-05,
      "loss": 0.4511,
      "step": 17650,
      "task_loss": 0.6190907955169678
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4041414260864258,
      "epoch": 6.38,
      "learning_rate": 5.76558044648739e-05,
      "loss": 0.442,
      "step": 17660,
      "task_loss": 0.3975062370300293
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4810088276863098,
      "epoch": 6.39,
      "learning_rate": 5.7595166893946776e-05,
      "loss": 0.4074,
      "step": 17670,
      "task_loss": 0.8788394927978516
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3534824252128601,
      "epoch": 6.39,
      "learning_rate": 5.7534412814741126e-05,
      "loss": 0.4732,
      "step": 17680,
      "task_loss": 0.41670382022857666
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36095911264419556,
      "epoch": 6.39,
      "learning_rate": 5.747354254052542e-05,
      "loss": 0.4306,
      "step": 17690,
      "task_loss": 0.6991243958473206
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3893577754497528,
      "epoch": 6.4,
      "learning_rate": 5.741255638516727e-05,
      "loss": 0.4119,
      "step": 17700,
      "task_loss": 0.6416240930557251
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33806878328323364,
      "epoch": 6.4,
      "learning_rate": 5.7351454663131803e-05,
      "loss": 0.4067,
      "step": 17710,
      "task_loss": 0.6782964468002319
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3822578191757202,
      "epoch": 6.4,
      "learning_rate": 5.729023768948003e-05,
      "loss": 0.4106,
      "step": 17720,
      "task_loss": 0.5011782050132751
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5085767507553101,
      "epoch": 6.41,
      "learning_rate": 5.7228905779867264e-05,
      "loss": 0.3913,
      "step": 17730,
      "task_loss": 0.5804252624511719
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5053203105926514,
      "epoch": 6.41,
      "learning_rate": 5.7167459250541455e-05,
      "loss": 0.457,
      "step": 17740,
      "task_loss": 0.5962353944778442
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4938843846321106,
      "epoch": 6.41,
      "learning_rate": 5.710589841834156e-05,
      "loss": 0.449,
      "step": 17750,
      "task_loss": 0.9457273483276367
    },
    {
      "epoch": 6.41,
      "eval_exact_match": 82.39356669820246,
      "eval_f1": 89.4370835165859,
      "step": 17750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3040074110031128,
      "epoch": 6.42,
      "learning_rate": 5.704422360069595e-05,
      "loss": 0.4009,
      "step": 17760,
      "task_loss": 0.37006545066833496
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.289045125246048,
      "epoch": 6.42,
      "learning_rate": 5.69824351156207e-05,
      "loss": 0.4099,
      "step": 17770,
      "task_loss": 0.9054403305053711
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3423129916191101,
      "epoch": 6.43,
      "learning_rate": 5.692053328171803e-05,
      "loss": 0.3902,
      "step": 17780,
      "task_loss": 0.7568475008010864
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4408140480518341,
      "epoch": 6.43,
      "learning_rate": 5.685851841817462e-05,
      "loss": 0.4457,
      "step": 17790,
      "task_loss": 0.9160474538803101
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3867258131504059,
      "epoch": 6.43,
      "learning_rate": 5.679639084475993e-05,
      "loss": 0.4964,
      "step": 17800,
      "task_loss": 0.3529086112976074
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3333200216293335,
      "epoch": 6.44,
      "learning_rate": 5.6734150881824656e-05,
      "loss": 0.3877,
      "step": 17810,
      "task_loss": 0.3891972303390503
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38562119007110596,
      "epoch": 6.44,
      "learning_rate": 5.667179885029895e-05,
      "loss": 0.4839,
      "step": 17820,
      "task_loss": 0.43628987669944763
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28627991676330566,
      "epoch": 6.44,
      "learning_rate": 5.660933507169086e-05,
      "loss": 0.4551,
      "step": 17830,
      "task_loss": 0.36466485261917114
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36773768067359924,
      "epoch": 6.45,
      "learning_rate": 5.654675986808465e-05,
      "loss": 0.4461,
      "step": 17840,
      "task_loss": 0.4903709292411804
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4122188091278076,
      "epoch": 6.45,
      "learning_rate": 5.6484073562139083e-05,
      "loss": 0.4621,
      "step": 17850,
      "task_loss": 0.27144575119018555
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3822740614414215,
      "epoch": 6.45,
      "learning_rate": 5.642127647708586e-05,
      "loss": 0.4331,
      "step": 17860,
      "task_loss": 0.3758048713207245
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4019128978252411,
      "epoch": 6.46,
      "learning_rate": 5.635836893672784e-05,
      "loss": 0.464,
      "step": 17870,
      "task_loss": 0.2744655907154083
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5253816246986389,
      "epoch": 6.46,
      "learning_rate": 5.6295351265437475e-05,
      "loss": 0.4224,
      "step": 17880,
      "task_loss": 0.9804195165634155
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42899399995803833,
      "epoch": 6.47,
      "learning_rate": 5.6232223788155074e-05,
      "loss": 0.4357,
      "step": 17890,
      "task_loss": 0.5716034173965454
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.390700101852417,
      "epoch": 6.47,
      "learning_rate": 5.616898683038712e-05,
      "loss": 0.4725,
      "step": 17900,
      "task_loss": 0.6366195678710938
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.328586220741272,
      "epoch": 6.47,
      "learning_rate": 5.610564071820462e-05,
      "loss": 0.4537,
      "step": 17910,
      "task_loss": 0.6518372893333435
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38455188274383545,
      "epoch": 6.48,
      "learning_rate": 5.6042185778241474e-05,
      "loss": 0.4551,
      "step": 17920,
      "task_loss": 0.21707725524902344
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5620262622833252,
      "epoch": 6.48,
      "learning_rate": 5.5978622337692656e-05,
      "loss": 0.4955,
      "step": 17930,
      "task_loss": 1.0111944675445557
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44680696725845337,
      "epoch": 6.48,
      "learning_rate": 5.5914950724312644e-05,
      "loss": 0.4513,
      "step": 17940,
      "task_loss": 0.4735015630722046
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5101405382156372,
      "epoch": 6.49,
      "learning_rate": 5.5851171266413694e-05,
      "loss": 0.4333,
      "step": 17950,
      "task_loss": 0.7333424687385559
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46911144256591797,
      "epoch": 6.49,
      "learning_rate": 5.578728429286414e-05,
      "loss": 0.4418,
      "step": 17960,
      "task_loss": 0.9456169605255127
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40388286113739014,
      "epoch": 6.49,
      "learning_rate": 5.5723290133086686e-05,
      "loss": 0.4104,
      "step": 17970,
      "task_loss": 0.5725729465484619
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3859924077987671,
      "epoch": 6.5,
      "learning_rate": 5.5659189117056755e-05,
      "loss": 0.4469,
      "step": 17980,
      "task_loss": 0.5803688764572144
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.443365216255188,
      "epoch": 6.5,
      "learning_rate": 5.5594981575300746e-05,
      "loss": 0.4438,
      "step": 17990,
      "task_loss": 0.5173107385635376
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36923152208328247,
      "epoch": 6.51,
      "learning_rate": 5.5530667838894306e-05,
      "loss": 0.4254,
      "step": 18000,
      "task_loss": 0.6575000286102295
    },
    {
      "epoch": 6.51,
      "eval_exact_match": 82.64900662251655,
      "eval_f1": 89.58216558481634,
      "step": 18000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43531203269958496,
      "epoch": 6.51,
      "learning_rate": 5.546624823946071e-05,
      "loss": 0.4421,
      "step": 18010,
      "task_loss": 0.6014666557312012
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41215750575065613,
      "epoch": 6.51,
      "learning_rate": 5.5401723109169074e-05,
      "loss": 0.4351,
      "step": 18020,
      "task_loss": 0.6380587816238403
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.508256196975708,
      "epoch": 6.52,
      "learning_rate": 5.5337092780732664e-05,
      "loss": 0.4143,
      "step": 18030,
      "task_loss": 0.6583020687103271
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31025072932243347,
      "epoch": 6.52,
      "learning_rate": 5.5272357587407176e-05,
      "loss": 0.4536,
      "step": 18040,
      "task_loss": 0.39846885204315186
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42146748304367065,
      "epoch": 6.52,
      "learning_rate": 5.520751786298905e-05,
      "loss": 0.4195,
      "step": 18050,
      "task_loss": 0.45628345012664795
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35557764768600464,
      "epoch": 6.53,
      "learning_rate": 5.51425739418137e-05,
      "loss": 0.4741,
      "step": 18060,
      "task_loss": 0.46955522894859314
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46768325567245483,
      "epoch": 6.53,
      "learning_rate": 5.507752615875383e-05,
      "loss": 0.4104,
      "step": 18070,
      "task_loss": 0.7435535192489624
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43830177187919617,
      "epoch": 6.53,
      "learning_rate": 5.501237484921767e-05,
      "loss": 0.5065,
      "step": 18080,
      "task_loss": 0.5534304976463318
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41444259881973267,
      "epoch": 6.54,
      "learning_rate": 5.494712034914728e-05,
      "loss": 0.478,
      "step": 18090,
      "task_loss": 0.4837718605995178
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3985566198825836,
      "epoch": 6.54,
      "learning_rate": 5.488176299501683e-05,
      "loss": 0.4515,
      "step": 18100,
      "task_loss": 0.48695647716522217
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38144785165786743,
      "epoch": 6.54,
      "learning_rate": 5.4816303123830796e-05,
      "loss": 0.4537,
      "step": 18110,
      "task_loss": 0.7126217484474182
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46962636709213257,
      "epoch": 6.55,
      "learning_rate": 5.4750741073122284e-05,
      "loss": 0.4182,
      "step": 18120,
      "task_loss": 0.6314877867698669
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.532020628452301,
      "epoch": 6.55,
      "learning_rate": 5.4685077180951276e-05,
      "loss": 0.4502,
      "step": 18130,
      "task_loss": 0.4621773660182953
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3801581561565399,
      "epoch": 6.56,
      "learning_rate": 5.461931178590289e-05,
      "loss": 0.4401,
      "step": 18140,
      "task_loss": 0.508671224117279
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4196373522281647,
      "epoch": 6.56,
      "learning_rate": 5.4553445227085605e-05,
      "loss": 0.4309,
      "step": 18150,
      "task_loss": 0.39120665192604065
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34238317608833313,
      "epoch": 6.56,
      "learning_rate": 5.4487477844129556e-05,
      "loss": 0.407,
      "step": 18160,
      "task_loss": 0.4414762854576111
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37008988857269287,
      "epoch": 6.57,
      "learning_rate": 5.442802127595373e-05,
      "loss": 0.4509,
      "step": 18170,
      "task_loss": 0.2949051260948181
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42825210094451904,
      "epoch": 6.57,
      "learning_rate": 5.4361863264675574e-05,
      "loss": 0.424,
      "step": 18180,
      "task_loss": 0.5720166563987732
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44686269760131836,
      "epoch": 6.57,
      "learning_rate": 5.4295605417119656e-05,
      "loss": 0.4321,
      "step": 18190,
      "task_loss": 0.8985435366630554
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5419995784759521,
      "epoch": 6.58,
      "learning_rate": 5.422924807493375e-05,
      "loss": 0.3999,
      "step": 18200,
      "task_loss": 0.8705387115478516
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44984179735183716,
      "epoch": 6.58,
      "learning_rate": 5.41627915802786e-05,
      "loss": 0.4632,
      "step": 18210,
      "task_loss": 0.6009798645973206
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4739626944065094,
      "epoch": 6.58,
      "learning_rate": 5.409623627582625e-05,
      "loss": 0.4673,
      "step": 18220,
      "task_loss": 0.7192704677581787
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3950083255767822,
      "epoch": 6.59,
      "learning_rate": 5.402958250475823e-05,
      "loss": 0.4619,
      "step": 18230,
      "task_loss": 0.44106048345565796
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23980626463890076,
      "epoch": 6.59,
      "learning_rate": 5.396283061076379e-05,
      "loss": 0.404,
      "step": 18240,
      "task_loss": 0.3170390725135803
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4144637882709503,
      "epoch": 6.6,
      "learning_rate": 5.389598093803816e-05,
      "loss": 0.4758,
      "step": 18250,
      "task_loss": 0.45880258083343506
    },
    {
      "epoch": 6.6,
      "eval_exact_match": 82.47871333964049,
      "eval_f1": 89.40149266393212,
      "step": 18250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5653073191642761,
      "epoch": 6.6,
      "learning_rate": 5.3829033831280716e-05,
      "loss": 0.4511,
      "step": 18260,
      "task_loss": 0.917737603187561
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36813783645629883,
      "epoch": 6.6,
      "learning_rate": 5.376198963569325e-05,
      "loss": 0.4015,
      "step": 18270,
      "task_loss": 0.44203007221221924
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38101470470428467,
      "epoch": 6.61,
      "learning_rate": 5.36948486969782e-05,
      "loss": 0.4641,
      "step": 18280,
      "task_loss": 0.5209448337554932
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3684188425540924,
      "epoch": 6.61,
      "learning_rate": 5.36276113613368e-05,
      "loss": 0.3796,
      "step": 18290,
      "task_loss": 0.3748064935207367
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4009804129600525,
      "epoch": 6.61,
      "learning_rate": 5.356027797546736e-05,
      "loss": 0.4367,
      "step": 18300,
      "task_loss": 0.4126998484134674
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4069993495941162,
      "epoch": 6.62,
      "learning_rate": 5.349284888656348e-05,
      "loss": 0.4818,
      "step": 18310,
      "task_loss": 0.4426262080669403
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5031300187110901,
      "epoch": 6.62,
      "learning_rate": 5.342532444231222e-05,
      "loss": 0.4451,
      "step": 18320,
      "task_loss": 0.9891221523284912
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38253137469291687,
      "epoch": 6.62,
      "learning_rate": 5.33577049908923e-05,
      "loss": 0.4383,
      "step": 18330,
      "task_loss": 0.5382572412490845
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3650605380535126,
      "epoch": 6.63,
      "learning_rate": 5.328999088097236e-05,
      "loss": 0.4475,
      "step": 18340,
      "task_loss": 0.4127558171749115
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3485070466995239,
      "epoch": 6.63,
      "learning_rate": 5.322218246170911e-05,
      "loss": 0.4303,
      "step": 18350,
      "task_loss": 0.5886161923408508
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.442766010761261,
      "epoch": 6.64,
      "learning_rate": 5.315428008274558e-05,
      "loss": 0.4622,
      "step": 18360,
      "task_loss": 0.5299726128578186
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45492517948150635,
      "epoch": 6.64,
      "learning_rate": 5.308628409420925e-05,
      "loss": 0.4739,
      "step": 18370,
      "task_loss": 0.6444842219352722
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49286532402038574,
      "epoch": 6.64,
      "learning_rate": 5.301819484671032e-05,
      "loss": 0.4572,
      "step": 18380,
      "task_loss": 0.6773360967636108
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4356333017349243,
      "epoch": 6.65,
      "learning_rate": 5.295001269133984e-05,
      "loss": 0.3974,
      "step": 18390,
      "task_loss": 0.6576051115989685
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3914899230003357,
      "epoch": 6.65,
      "learning_rate": 5.2881737979667926e-05,
      "loss": 0.4884,
      "step": 18400,
      "task_loss": 0.7598779797554016
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4002273678779602,
      "epoch": 6.65,
      "learning_rate": 5.2813371063741945e-05,
      "loss": 0.4135,
      "step": 18410,
      "task_loss": 0.5003056526184082
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5130934715270996,
      "epoch": 6.66,
      "learning_rate": 5.2744912296084714e-05,
      "loss": 0.5041,
      "step": 18420,
      "task_loss": 0.7868928909301758
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42638254165649414,
      "epoch": 6.66,
      "learning_rate": 5.2676362029692664e-05,
      "loss": 0.4332,
      "step": 18430,
      "task_loss": 0.7337071895599365
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5077611207962036,
      "epoch": 6.66,
      "learning_rate": 5.260772061803401e-05,
      "loss": 0.434,
      "step": 18440,
      "task_loss": 0.7167735695838928
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3096952438354492,
      "epoch": 6.67,
      "learning_rate": 5.2538988415046944e-05,
      "loss": 0.3743,
      "step": 18450,
      "task_loss": 0.235990971326828
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40252023935317993,
      "epoch": 6.67,
      "learning_rate": 5.247016577513784e-05,
      "loss": 0.4,
      "step": 18460,
      "task_loss": 0.6203321814537048
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4693078398704529,
      "epoch": 6.68,
      "learning_rate": 5.2401253053179365e-05,
      "loss": 0.4146,
      "step": 18470,
      "task_loss": 0.6614182591438293
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27100706100463867,
      "epoch": 6.68,
      "learning_rate": 5.2332250604508675e-05,
      "loss": 0.4092,
      "step": 18480,
      "task_loss": 0.3299637734889984
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3161439895629883,
      "epoch": 6.68,
      "learning_rate": 5.226315878492561e-05,
      "loss": 0.4102,
      "step": 18490,
      "task_loss": 0.8195310235023499
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.48115524649620056,
      "epoch": 6.69,
      "learning_rate": 5.219397795069082e-05,
      "loss": 0.4006,
      "step": 18500,
      "task_loss": 0.6623936891555786
    },
    {
      "epoch": 6.69,
      "eval_exact_match": 82.81929990539263,
      "eval_f1": 89.71049915364028,
      "step": 18500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3616160750389099,
      "epoch": 6.69,
      "learning_rate": 5.212470845852394e-05,
      "loss": 0.4761,
      "step": 18510,
      "task_loss": 0.5478072762489319
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3384724259376526,
      "epoch": 6.69,
      "learning_rate": 5.205535066560179e-05,
      "loss": 0.4039,
      "step": 18520,
      "task_loss": 0.8222288489341736
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4261634051799774,
      "epoch": 6.7,
      "learning_rate": 5.198590492955645e-05,
      "loss": 0.4239,
      "step": 18530,
      "task_loss": 0.7383376359939575
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4749411344528198,
      "epoch": 6.7,
      "learning_rate": 5.1916371608473494e-05,
      "loss": 0.4562,
      "step": 18540,
      "task_loss": 0.7253599762916565
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4970089793205261,
      "epoch": 6.7,
      "learning_rate": 5.1846751060890114e-05,
      "loss": 0.3964,
      "step": 18550,
      "task_loss": 0.7168070077896118
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4212992787361145,
      "epoch": 6.71,
      "learning_rate": 5.1777043645793254e-05,
      "loss": 0.4283,
      "step": 18560,
      "task_loss": 0.71849524974823
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3591970205307007,
      "epoch": 6.71,
      "learning_rate": 5.17072497226178e-05,
      "loss": 0.4409,
      "step": 18570,
      "task_loss": 0.37877827882766724
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4818565249443054,
      "epoch": 6.71,
      "learning_rate": 5.1637369651244666e-05,
      "loss": 0.3911,
      "step": 18580,
      "task_loss": 0.7044258117675781
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3378729820251465,
      "epoch": 6.72,
      "learning_rate": 5.156740379199902e-05,
      "loss": 0.4253,
      "step": 18590,
      "task_loss": 0.3735387921333313
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5539082884788513,
      "epoch": 6.72,
      "learning_rate": 5.149735250564836e-05,
      "loss": 0.4326,
      "step": 18600,
      "task_loss": 0.6768410205841064
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4786291718482971,
      "epoch": 6.73,
      "learning_rate": 5.142721615340066e-05,
      "loss": 0.4563,
      "step": 18610,
      "task_loss": 0.5143426656723022
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47057053446769714,
      "epoch": 6.73,
      "learning_rate": 5.135699509690255e-05,
      "loss": 0.461,
      "step": 18620,
      "task_loss": 0.5533486604690552
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3846384584903717,
      "epoch": 6.73,
      "learning_rate": 5.128668969823739e-05,
      "loss": 0.4523,
      "step": 18630,
      "task_loss": 0.45162439346313477
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.372709721326828,
      "epoch": 6.74,
      "learning_rate": 5.121630031992348e-05,
      "loss": 0.4113,
      "step": 18640,
      "task_loss": 0.47227999567985535
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3604407012462616,
      "epoch": 6.74,
      "learning_rate": 5.114582732491213e-05,
      "loss": 0.445,
      "step": 18650,
      "task_loss": 0.5071597099304199
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45643842220306396,
      "epoch": 6.74,
      "learning_rate": 5.107527107658576e-05,
      "loss": 0.4598,
      "step": 18660,
      "task_loss": 0.6165838241577148
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5041736960411072,
      "epoch": 6.75,
      "learning_rate": 5.1004631938756154e-05,
      "loss": 0.4796,
      "step": 18670,
      "task_loss": 0.7193949222564697
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31500351428985596,
      "epoch": 6.75,
      "learning_rate": 5.093391027566244e-05,
      "loss": 0.4015,
      "step": 18680,
      "task_loss": 0.5422650575637817
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47139471769332886,
      "epoch": 6.75,
      "learning_rate": 5.086310645196928e-05,
      "loss": 0.4027,
      "step": 18690,
      "task_loss": 0.46124646067619324
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35555168986320496,
      "epoch": 6.76,
      "learning_rate": 5.079222083276504e-05,
      "loss": 0.4077,
      "step": 18700,
      "task_loss": 0.37314948439598083
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35326719284057617,
      "epoch": 6.76,
      "learning_rate": 5.072125378355978e-05,
      "loss": 0.4241,
      "step": 18710,
      "task_loss": 0.5533696413040161
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42909830808639526,
      "epoch": 6.77,
      "learning_rate": 5.0650205670283475e-05,
      "loss": 0.4068,
      "step": 18720,
      "task_loss": 0.39843881130218506
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41295886039733887,
      "epoch": 6.77,
      "learning_rate": 5.057907685928408e-05,
      "loss": 0.4629,
      "step": 18730,
      "task_loss": 0.5897053480148315
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43410900235176086,
      "epoch": 6.77,
      "learning_rate": 5.050786771732567e-05,
      "loss": 0.438,
      "step": 18740,
      "task_loss": 0.6660215854644775
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3357173800468445,
      "epoch": 6.78,
      "learning_rate": 5.043657861158653e-05,
      "loss": 0.4601,
      "step": 18750,
      "task_loss": 0.6065893173217773
    },
    {
      "epoch": 6.78,
      "eval_exact_match": 82.45033112582782,
      "eval_f1": 89.46843856351111,
      "step": 18750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4543573558330536,
      "epoch": 6.78,
      "learning_rate": 5.036520990965726e-05,
      "loss": 0.4089,
      "step": 18760,
      "task_loss": 0.6887750625610352
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39767420291900635,
      "epoch": 6.78,
      "learning_rate": 5.029376197953888e-05,
      "loss": 0.4149,
      "step": 18770,
      "task_loss": 0.6786099672317505
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38811343908309937,
      "epoch": 6.79,
      "learning_rate": 5.022223518964095e-05,
      "loss": 0.4349,
      "step": 18780,
      "task_loss": 0.4405578374862671
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47207069396972656,
      "epoch": 6.79,
      "learning_rate": 5.015062990877964e-05,
      "loss": 0.4619,
      "step": 18790,
      "task_loss": 0.5872414112091064
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3899841904640198,
      "epoch": 6.79,
      "learning_rate": 5.007894650617588e-05,
      "loss": 0.4326,
      "step": 18800,
      "task_loss": 0.42872917652130127
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46909916400909424,
      "epoch": 6.8,
      "learning_rate": 5.0007185351453374e-05,
      "loss": 0.4121,
      "step": 18810,
      "task_loss": 0.6106085181236267
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45315465331077576,
      "epoch": 6.8,
      "learning_rate": 4.9935346814636785e-05,
      "loss": 0.3855,
      "step": 18820,
      "task_loss": 0.7110905647277832
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4155175983905792,
      "epoch": 6.81,
      "learning_rate": 4.9863431266149745e-05,
      "loss": 0.441,
      "step": 18830,
      "task_loss": 0.6808761954307556
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37727808952331543,
      "epoch": 6.81,
      "learning_rate": 4.979143907681301e-05,
      "loss": 0.4181,
      "step": 18840,
      "task_loss": 0.44022470712661743
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6061004996299744,
      "epoch": 6.81,
      "learning_rate": 4.9719370617842544e-05,
      "loss": 0.51,
      "step": 18850,
      "task_loss": 0.6300066113471985
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47508275508880615,
      "epoch": 6.82,
      "learning_rate": 4.964722626084752e-05,
      "loss": 0.4558,
      "step": 18860,
      "task_loss": 0.6476293206214905
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.48660221695899963,
      "epoch": 6.82,
      "learning_rate": 4.9575006377828535e-05,
      "loss": 0.3868,
      "step": 18870,
      "task_loss": 0.5755065679550171
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.527796745300293,
      "epoch": 6.82,
      "learning_rate": 4.9502711341175553e-05,
      "loss": 0.4417,
      "step": 18880,
      "task_loss": 0.6073551774024963
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34590059518814087,
      "epoch": 6.83,
      "learning_rate": 4.9430341523666124e-05,
      "loss": 0.4138,
      "step": 18890,
      "task_loss": 0.6458975672721863
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3756251335144043,
      "epoch": 6.83,
      "learning_rate": 4.935789729846335e-05,
      "loss": 0.4016,
      "step": 18900,
      "task_loss": 0.5915954113006592
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4537370204925537,
      "epoch": 6.83,
      "learning_rate": 4.928537903911402e-05,
      "loss": 0.3831,
      "step": 18910,
      "task_loss": 0.7734125852584839
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4709266722202301,
      "epoch": 6.84,
      "learning_rate": 4.921278711954666e-05,
      "loss": 0.4286,
      "step": 18920,
      "task_loss": 0.9293255805969238
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4475264251232147,
      "epoch": 6.84,
      "learning_rate": 4.9140121914069626e-05,
      "loss": 0.4389,
      "step": 18930,
      "task_loss": 0.5489974021911621
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33601874113082886,
      "epoch": 6.84,
      "learning_rate": 4.9067383797369146e-05,
      "loss": 0.3989,
      "step": 18940,
      "task_loss": 0.38476261496543884
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3520219326019287,
      "epoch": 6.85,
      "learning_rate": 4.899457314450742e-05,
      "loss": 0.3766,
      "step": 18950,
      "task_loss": 0.6894028186798096
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38159820437431335,
      "epoch": 6.85,
      "learning_rate": 4.8921690330920637e-05,
      "loss": 0.4241,
      "step": 18960,
      "task_loss": 0.7299265265464783
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47799068689346313,
      "epoch": 6.86,
      "learning_rate": 4.884873573241711e-05,
      "loss": 0.4292,
      "step": 18970,
      "task_loss": 0.6850833892822266
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6151865720748901,
      "epoch": 6.86,
      "learning_rate": 4.8775709725175277e-05,
      "loss": 0.418,
      "step": 18980,
      "task_loss": 0.6462631225585938
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47672033309936523,
      "epoch": 6.86,
      "learning_rate": 4.870261268574178e-05,
      "loss": 0.444,
      "step": 18990,
      "task_loss": 0.8364094495773315
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3276512920856476,
      "epoch": 6.87,
      "learning_rate": 4.8629444991029544e-05,
      "loss": 0.4532,
      "step": 19000,
      "task_loss": 0.3989708423614502
    },
    {
      "epoch": 6.87,
      "eval_exact_match": 82.45979186376537,
      "eval_f1": 89.5439109204433,
      "step": 19000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44405293464660645,
      "epoch": 6.87,
      "learning_rate": 4.8556207018315793e-05,
      "loss": 0.3897,
      "step": 19010,
      "task_loss": 0.5321621894836426
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35056599974632263,
      "epoch": 6.87,
      "learning_rate": 4.848289914524017e-05,
      "loss": 0.444,
      "step": 19020,
      "task_loss": 0.3490481674671173
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4751790761947632,
      "epoch": 6.88,
      "learning_rate": 4.840952174980269e-05,
      "loss": 0.3982,
      "step": 19030,
      "task_loss": 0.663070797920227
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45868778228759766,
      "epoch": 6.88,
      "learning_rate": 4.8336075210361884e-05,
      "loss": 0.4686,
      "step": 19040,
      "task_loss": 0.5684219598770142
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5517913699150085,
      "epoch": 6.88,
      "learning_rate": 4.826255990563282e-05,
      "loss": 0.4379,
      "step": 19050,
      "task_loss": 0.8500944972038269
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36011481285095215,
      "epoch": 6.89,
      "learning_rate": 4.8188976214685125e-05,
      "loss": 0.4354,
      "step": 19060,
      "task_loss": 0.45963573455810547
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4780333638191223,
      "epoch": 6.89,
      "learning_rate": 4.81226927362005e-05,
      "loss": 0.4268,
      "step": 19070,
      "task_loss": 0.5274924039840698
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3727762699127197,
      "epoch": 6.9,
      "learning_rate": 4.804898015703557e-05,
      "loss": 0.418,
      "step": 19080,
      "task_loss": 0.5746783018112183
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24984151124954224,
      "epoch": 6.9,
      "learning_rate": 4.797520029294102e-05,
      "loss": 0.378,
      "step": 19090,
      "task_loss": 0.3299369812011719
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3473876416683197,
      "epoch": 6.9,
      "learning_rate": 4.7901353524350616e-05,
      "loss": 0.4693,
      "step": 19100,
      "task_loss": 0.7796521186828613
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4735308885574341,
      "epoch": 6.91,
      "learning_rate": 4.7827440232043156e-05,
      "loss": 0.4624,
      "step": 19110,
      "task_loss": 0.862983226776123
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4470568895339966,
      "epoch": 6.91,
      "learning_rate": 4.77534607971404e-05,
      "loss": 0.469,
      "step": 19120,
      "task_loss": 0.3181147575378418
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42912188172340393,
      "epoch": 6.91,
      "learning_rate": 4.767941560110523e-05,
      "loss": 0.4015,
      "step": 19130,
      "task_loss": 0.623262882232666
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4887217581272125,
      "epoch": 6.92,
      "learning_rate": 4.7605305025739525e-05,
      "loss": 0.4325,
      "step": 19140,
      "task_loss": 0.5894296765327454
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34331122040748596,
      "epoch": 6.92,
      "learning_rate": 4.753112945318238e-05,
      "loss": 0.3854,
      "step": 19150,
      "task_loss": 0.4348488450050354
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4002523720264435,
      "epoch": 6.92,
      "learning_rate": 4.745688926590796e-05,
      "loss": 0.4437,
      "step": 19160,
      "task_loss": 0.5534040927886963
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.612992525100708,
      "epoch": 6.93,
      "learning_rate": 4.738258484672365e-05,
      "loss": 0.4264,
      "step": 19170,
      "task_loss": 0.8351370096206665
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40844202041625977,
      "epoch": 6.93,
      "learning_rate": 4.730821657876802e-05,
      "loss": 0.4385,
      "step": 19180,
      "task_loss": 0.5529130697250366
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41695636510849,
      "epoch": 6.94,
      "learning_rate": 4.7233784845508855e-05,
      "loss": 0.456,
      "step": 19190,
      "task_loss": 0.2549833059310913
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5138141512870789,
      "epoch": 6.94,
      "learning_rate": 4.715929003074122e-05,
      "loss": 0.4156,
      "step": 19200,
      "task_loss": 0.43483638763427734
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4193929433822632,
      "epoch": 6.94,
      "learning_rate": 4.7084732518585406e-05,
      "loss": 0.4429,
      "step": 19210,
      "task_loss": 0.6967915296554565
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4237646460533142,
      "epoch": 6.95,
      "learning_rate": 4.701011269348504e-05,
      "loss": 0.4296,
      "step": 19220,
      "task_loss": 1.0524966716766357
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3706055283546448,
      "epoch": 6.95,
      "learning_rate": 4.6935430940205026e-05,
      "loss": 0.4508,
      "step": 19230,
      "task_loss": 0.848009467124939
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5015447735786438,
      "epoch": 6.95,
      "learning_rate": 4.686068764382958e-05,
      "loss": 0.4599,
      "step": 19240,
      "task_loss": 0.8581211566925049
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4016169309616089,
      "epoch": 6.96,
      "learning_rate": 4.67858831897603e-05,
      "loss": 0.4012,
      "step": 19250,
      "task_loss": 0.6866574883460999
    },
    {
      "epoch": 6.96,
      "eval_exact_match": 82.77199621570483,
      "eval_f1": 89.62177861298487,
      "step": 19250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37276118993759155,
      "epoch": 6.96,
      "learning_rate": 4.671101796371409e-05,
      "loss": 0.4105,
      "step": 19260,
      "task_loss": 0.38845476508140564
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37246400117874146,
      "epoch": 6.96,
      "learning_rate": 4.663609235172122e-05,
      "loss": 0.4296,
      "step": 19270,
      "task_loss": 0.49262505769729614
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.426718533039093,
      "epoch": 6.97,
      "learning_rate": 4.6561106740123356e-05,
      "loss": 0.4424,
      "step": 19280,
      "task_loss": 0.4604375958442688
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47683387994766235,
      "epoch": 6.97,
      "learning_rate": 4.648606151557153e-05,
      "loss": 0.367,
      "step": 19290,
      "task_loss": 0.4319005012512207
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5281187891960144,
      "epoch": 6.98,
      "learning_rate": 4.641095706502413e-05,
      "loss": 0.4282,
      "step": 19300,
      "task_loss": 0.8619099259376526
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3414451777935028,
      "epoch": 6.98,
      "learning_rate": 4.633579377574499e-05,
      "loss": 0.4111,
      "step": 19310,
      "task_loss": 0.3021755814552307
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3465237021446228,
      "epoch": 6.98,
      "learning_rate": 4.6260572035301296e-05,
      "loss": 0.3609,
      "step": 19320,
      "task_loss": 0.49031707644462585
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3119080066680908,
      "epoch": 6.99,
      "learning_rate": 4.6185292231561614e-05,
      "loss": 0.4461,
      "step": 19330,
      "task_loss": 0.6686820387840271
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40698978304862976,
      "epoch": 6.99,
      "learning_rate": 4.610995475269395e-05,
      "loss": 0.38,
      "step": 19340,
      "task_loss": 0.751046895980835
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4363614320755005,
      "epoch": 6.99,
      "learning_rate": 4.603455998716366e-05,
      "loss": 0.4459,
      "step": 19350,
      "task_loss": 0.6054350137710571
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3699437975883484,
      "epoch": 7.0,
      "learning_rate": 4.595910832373155e-05,
      "loss": 0.4081,
      "step": 19360,
      "task_loss": 0.47203072905540466
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.476010262966156,
      "epoch": 7.0,
      "learning_rate": 4.58836001514517e-05,
      "loss": 0.4631,
      "step": 19370,
      "task_loss": 0.6500152349472046
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25550246238708496,
      "epoch": 7.0,
      "learning_rate": 4.58080358596697e-05,
      "loss": 0.3536,
      "step": 19380,
      "task_loss": 0.41859322786331177
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37610524892807007,
      "epoch": 7.01,
      "learning_rate": 4.57324158380204e-05,
      "loss": 0.3568,
      "step": 19390,
      "task_loss": 0.3900749087333679
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2958787977695465,
      "epoch": 7.01,
      "learning_rate": 4.56567404764261e-05,
      "loss": 0.3335,
      "step": 19400,
      "task_loss": 0.531352698802948
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3521242141723633,
      "epoch": 7.01,
      "learning_rate": 4.558101016509439e-05,
      "loss": 0.3586,
      "step": 19410,
      "task_loss": 0.43806248903274536
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32949671149253845,
      "epoch": 7.02,
      "learning_rate": 4.550522529451622e-05,
      "loss": 0.3663,
      "step": 19420,
      "task_loss": 0.31022578477859497
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3833150565624237,
      "epoch": 7.02,
      "learning_rate": 4.542938625546389e-05,
      "loss": 0.3462,
      "step": 19430,
      "task_loss": 0.7389860153198242
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4690796136856079,
      "epoch": 7.03,
      "learning_rate": 4.5353493438988946e-05,
      "loss": 0.3397,
      "step": 19440,
      "task_loss": 0.8385846018791199
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2842397093772888,
      "epoch": 7.03,
      "learning_rate": 4.527754723642031e-05,
      "loss": 0.3836,
      "step": 19450,
      "task_loss": 0.49422913789749146
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24541881680488586,
      "epoch": 7.03,
      "learning_rate": 4.520154803936211e-05,
      "loss": 0.3828,
      "step": 19460,
      "task_loss": 0.4191051721572876
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3875623345375061,
      "epoch": 7.04,
      "learning_rate": 4.512549623969179e-05,
      "loss": 0.379,
      "step": 19470,
      "task_loss": 0.6146583557128906
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31408095359802246,
      "epoch": 7.04,
      "learning_rate": 4.5049392229557984e-05,
      "loss": 0.3425,
      "step": 19480,
      "task_loss": 0.7459331750869751
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3380616009235382,
      "epoch": 7.04,
      "learning_rate": 4.497323640137856e-05,
      "loss": 0.358,
      "step": 19490,
      "task_loss": 0.5370668172836304
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3334946036338806,
      "epoch": 7.05,
      "learning_rate": 4.489702914783859e-05,
      "loss": 0.3863,
      "step": 19500,
      "task_loss": 0.6622374057769775
    },
    {
      "epoch": 7.05,
      "eval_exact_match": 82.37464522232735,
      "eval_f1": 89.41990224068103,
      "step": 19500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.342751681804657,
      "epoch": 7.05,
      "learning_rate": 4.482077086188828e-05,
      "loss": 0.3715,
      "step": 19510,
      "task_loss": 0.6515509486198425
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3435029983520508,
      "epoch": 7.05,
      "learning_rate": 4.4744461936741006e-05,
      "loss": 0.3807,
      "step": 19520,
      "task_loss": 0.5341594219207764
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3486357033252716,
      "epoch": 7.06,
      "learning_rate": 4.466810276587125e-05,
      "loss": 0.3861,
      "step": 19530,
      "task_loss": 0.5828717350959778
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27860820293426514,
      "epoch": 7.06,
      "learning_rate": 4.459169374301257e-05,
      "loss": 0.3819,
      "step": 19540,
      "task_loss": 0.3259779214859009
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2861538529396057,
      "epoch": 7.07,
      "learning_rate": 4.4515235262155585e-05,
      "loss": 0.3759,
      "step": 19550,
      "task_loss": 0.5208934545516968
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2555454969406128,
      "epoch": 7.07,
      "learning_rate": 4.443872771754593e-05,
      "loss": 0.3568,
      "step": 19560,
      "task_loss": 0.6315025091171265
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4431646466255188,
      "epoch": 7.07,
      "learning_rate": 4.436217150368223e-05,
      "loss": 0.3419,
      "step": 19570,
      "task_loss": 0.617892861366272
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2903032600879669,
      "epoch": 7.08,
      "learning_rate": 4.4285567015314056e-05,
      "loss": 0.3874,
      "step": 19580,
      "task_loss": 0.39440637826919556
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3276922106742859,
      "epoch": 7.08,
      "learning_rate": 4.4208914647439934e-05,
      "loss": 0.3447,
      "step": 19590,
      "task_loss": 0.7188665270805359
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3929433822631836,
      "epoch": 7.08,
      "learning_rate": 4.413221479530524e-05,
      "loss": 0.3543,
      "step": 19600,
      "task_loss": 0.48741626739501953
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4890502393245697,
      "epoch": 7.09,
      "learning_rate": 4.40554678544002e-05,
      "loss": 0.3848,
      "step": 19610,
      "task_loss": 0.8010128736495972
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28317737579345703,
      "epoch": 7.09,
      "learning_rate": 4.397867422045783e-05,
      "loss": 0.3642,
      "step": 19620,
      "task_loss": 0.4237869679927826
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32048285007476807,
      "epoch": 7.09,
      "learning_rate": 4.390183428945197e-05,
      "loss": 0.3484,
      "step": 19630,
      "task_loss": 0.42599737644195557
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30844658613204956,
      "epoch": 7.1,
      "learning_rate": 4.38249484575951e-05,
      "loss": 0.3503,
      "step": 19640,
      "task_loss": 0.29508674144744873
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3091406226158142,
      "epoch": 7.1,
      "learning_rate": 4.374801712133647e-05,
      "loss": 0.353,
      "step": 19650,
      "task_loss": 0.5975150465965271
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39048248529434204,
      "epoch": 7.11,
      "learning_rate": 4.367104067735987e-05,
      "loss": 0.3897,
      "step": 19660,
      "task_loss": 0.770504355430603
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38784366846084595,
      "epoch": 7.11,
      "learning_rate": 4.359401952258176e-05,
      "loss": 0.3453,
      "step": 19670,
      "task_loss": 0.5350626707077026
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3483542203903198,
      "epoch": 7.11,
      "learning_rate": 4.351695405414913e-05,
      "loss": 0.3466,
      "step": 19680,
      "task_loss": 0.7391943335533142
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3177662491798401,
      "epoch": 7.12,
      "learning_rate": 4.3439844669437396e-05,
      "loss": 0.3558,
      "step": 19690,
      "task_loss": 0.3970339298248291
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3653816878795624,
      "epoch": 7.12,
      "learning_rate": 4.336269176604852e-05,
      "loss": 0.3538,
      "step": 19700,
      "task_loss": 0.48417383432388306
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5231342315673828,
      "epoch": 7.12,
      "learning_rate": 4.32854957418088e-05,
      "loss": 0.3773,
      "step": 19710,
      "task_loss": 0.5919955372810364
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2906436026096344,
      "epoch": 7.13,
      "learning_rate": 4.3208256994766895e-05,
      "loss": 0.3596,
      "step": 19720,
      "task_loss": 0.7230117321014404
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31748613715171814,
      "epoch": 7.13,
      "learning_rate": 4.313097592319176e-05,
      "loss": 0.3458,
      "step": 19730,
      "task_loss": 0.3181889057159424
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33267971873283386,
      "epoch": 7.13,
      "learning_rate": 4.305365292557059e-05,
      "loss": 0.3904,
      "step": 19740,
      "task_loss": 0.5057215094566345
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3165643811225891,
      "epoch": 7.14,
      "learning_rate": 4.297628840060676e-05,
      "loss": 0.3824,
      "step": 19750,
      "task_loss": 0.43284088373184204
    },
    {
      "epoch": 7.14,
      "eval_exact_match": 82.87606433301798,
      "eval_f1": 89.81716000383479,
      "step": 19750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2485005110502243,
      "epoch": 7.14,
      "learning_rate": 4.2898882747217766e-05,
      "loss": 0.3264,
      "step": 19760,
      "task_loss": 0.29056209325790405
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3244897127151489,
      "epoch": 7.14,
      "learning_rate": 4.2821436364533194e-05,
      "loss": 0.3762,
      "step": 19770,
      "task_loss": 0.5531866550445557
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3738449513912201,
      "epoch": 7.15,
      "learning_rate": 4.274394965189265e-05,
      "loss": 0.342,
      "step": 19780,
      "task_loss": 0.4225583076477051
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.302204966545105,
      "epoch": 7.15,
      "learning_rate": 4.266642300884366e-05,
      "loss": 0.3548,
      "step": 19790,
      "task_loss": 0.776357114315033
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36438536643981934,
      "epoch": 7.16,
      "learning_rate": 4.258885683513967e-05,
      "loss": 0.3736,
      "step": 19800,
      "task_loss": 0.48018282651901245
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2947632670402527,
      "epoch": 7.16,
      "learning_rate": 4.2511251530737956e-05,
      "loss": 0.4031,
      "step": 19810,
      "task_loss": 0.49369704723358154
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3188723921775818,
      "epoch": 7.16,
      "learning_rate": 4.243360749579756e-05,
      "loss": 0.377,
      "step": 19820,
      "task_loss": 0.4446810483932495
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2981301248073578,
      "epoch": 7.17,
      "learning_rate": 4.2355925130677255e-05,
      "loss": 0.3348,
      "step": 19830,
      "task_loss": 0.25408676266670227
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35322049260139465,
      "epoch": 7.17,
      "learning_rate": 4.22782048359334e-05,
      "loss": 0.3721,
      "step": 19840,
      "task_loss": 0.41666746139526367
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2933061718940735,
      "epoch": 7.17,
      "learning_rate": 4.2200447012318006e-05,
      "loss": 0.4241,
      "step": 19850,
      "task_loss": 0.7377125024795532
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2902827262878418,
      "epoch": 7.18,
      "learning_rate": 4.212265206077655e-05,
      "loss": 0.3631,
      "step": 19860,
      "task_loss": 0.4684586524963379
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3492077589035034,
      "epoch": 7.18,
      "learning_rate": 4.204482038244594e-05,
      "loss": 0.3792,
      "step": 19870,
      "task_loss": 0.5240708589553833
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3437349200248718,
      "epoch": 7.18,
      "learning_rate": 4.1966952378652515e-05,
      "loss": 0.3532,
      "step": 19880,
      "task_loss": 0.39789509773254395
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2762902081012726,
      "epoch": 7.19,
      "learning_rate": 4.1889048450909854e-05,
      "loss": 0.3529,
      "step": 19890,
      "task_loss": 0.5627302527427673
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3159010410308838,
      "epoch": 7.19,
      "learning_rate": 4.181110900091682e-05,
      "loss": 0.3713,
      "step": 19900,
      "task_loss": 0.6435883045196533
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3086598515510559,
      "epoch": 7.2,
      "learning_rate": 4.173313443055545e-05,
      "loss": 0.3744,
      "step": 19910,
      "task_loss": 0.38989460468292236
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44235122203826904,
      "epoch": 7.2,
      "learning_rate": 4.165512514188881e-05,
      "loss": 0.3728,
      "step": 19920,
      "task_loss": 0.6314876079559326
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2719871997833252,
      "epoch": 7.2,
      "learning_rate": 4.157708153715904e-05,
      "loss": 0.3254,
      "step": 19930,
      "task_loss": 0.3361320495605469
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3495417833328247,
      "epoch": 7.21,
      "learning_rate": 4.1499004018785207e-05,
      "loss": 0.3469,
      "step": 19940,
      "task_loss": 0.40324363112449646
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31385642290115356,
      "epoch": 7.21,
      "learning_rate": 4.1420892989361255e-05,
      "loss": 0.352,
      "step": 19950,
      "task_loss": 0.46830520033836365
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.386014461517334,
      "epoch": 7.21,
      "learning_rate": 4.134274885165392e-05,
      "loss": 0.3731,
      "step": 19960,
      "task_loss": 0.8087267875671387
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37941837310791016,
      "epoch": 7.22,
      "learning_rate": 4.1264572008600645e-05,
      "loss": 0.3761,
      "step": 19970,
      "task_loss": 0.23961874842643738
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29875701665878296,
      "epoch": 7.22,
      "learning_rate": 4.118636286330753e-05,
      "loss": 0.3597,
      "step": 19980,
      "task_loss": 0.43934816122055054
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3207720220088959,
      "epoch": 7.22,
      "learning_rate": 4.110812181904722e-05,
      "loss": 0.3418,
      "step": 19990,
      "task_loss": 0.5765659809112549
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2805137634277344,
      "epoch": 7.23,
      "learning_rate": 4.102984927925685e-05,
      "loss": 0.3795,
      "step": 20000,
      "task_loss": 0.48947030305862427
    },
    {
      "epoch": 7.23,
      "eval_exact_match": 82.61116367076632,
      "eval_f1": 89.78752648269042,
      "step": 20000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35584986209869385,
      "epoch": 7.23,
      "learning_rate": 4.095154564753595e-05,
      "loss": 0.3677,
      "step": 20010,
      "task_loss": 0.5476221442222595
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4366397261619568,
      "epoch": 7.24,
      "learning_rate": 4.0873211327644396e-05,
      "loss": 0.3743,
      "step": 20020,
      "task_loss": 1.0048655271530151
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36649996042251587,
      "epoch": 7.24,
      "learning_rate": 4.079484672350027e-05,
      "loss": 0.3642,
      "step": 20030,
      "task_loss": 0.49900856614112854
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26360058784484863,
      "epoch": 7.24,
      "learning_rate": 4.0716452239177816e-05,
      "loss": 0.3499,
      "step": 20040,
      "task_loss": 0.32680177688598633
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49425458908081055,
      "epoch": 7.25,
      "learning_rate": 4.0638028278905354e-05,
      "loss": 0.3466,
      "step": 20050,
      "task_loss": 0.7145361304283142
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31015101075172424,
      "epoch": 7.25,
      "learning_rate": 4.055957524706322e-05,
      "loss": 0.3908,
      "step": 20060,
      "task_loss": 0.5189313888549805
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3288698196411133,
      "epoch": 7.25,
      "learning_rate": 4.048109354818161e-05,
      "loss": 0.378,
      "step": 20070,
      "task_loss": 0.42466723918914795
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3703783452510834,
      "epoch": 7.26,
      "learning_rate": 4.040258358693856e-05,
      "loss": 0.3968,
      "step": 20080,
      "task_loss": 0.9877618551254272
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41583582758903503,
      "epoch": 7.26,
      "learning_rate": 4.032404576815781e-05,
      "loss": 0.3541,
      "step": 20090,
      "task_loss": 0.7476043701171875
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.310701847076416,
      "epoch": 7.26,
      "learning_rate": 4.024548049680678e-05,
      "loss": 0.3608,
      "step": 20100,
      "task_loss": 0.7368062138557434
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42681705951690674,
      "epoch": 7.27,
      "learning_rate": 4.016688817799444e-05,
      "loss": 0.3524,
      "step": 20110,
      "task_loss": 0.6136888861656189
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3238581120967865,
      "epoch": 7.27,
      "learning_rate": 4.0088269216969186e-05,
      "loss": 0.3522,
      "step": 20120,
      "task_loss": 0.47326260805130005
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3412288725376129,
      "epoch": 7.28,
      "learning_rate": 4.0009624019116845e-05,
      "loss": 0.3413,
      "step": 20130,
      "task_loss": 0.6561700105667114
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24474942684173584,
      "epoch": 7.28,
      "learning_rate": 3.993095298995848e-05,
      "loss": 0.3571,
      "step": 20140,
      "task_loss": 0.2822563648223877
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33786123991012573,
      "epoch": 7.28,
      "learning_rate": 3.985225653514839e-05,
      "loss": 0.3548,
      "step": 20150,
      "task_loss": 0.5039522647857666
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26680850982666016,
      "epoch": 7.29,
      "learning_rate": 3.977353506047195e-05,
      "loss": 0.3808,
      "step": 20160,
      "task_loss": 0.4330366849899292
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3075251579284668,
      "epoch": 7.29,
      "learning_rate": 3.9694788971843556e-05,
      "loss": 0.3602,
      "step": 20170,
      "task_loss": 0.4514367878437042
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26794716715812683,
      "epoch": 7.29,
      "learning_rate": 3.961601867530452e-05,
      "loss": 0.3607,
      "step": 20180,
      "task_loss": 0.46283090114593506
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2926703095436096,
      "epoch": 7.3,
      "learning_rate": 3.953722457702098e-05,
      "loss": 0.3783,
      "step": 20190,
      "task_loss": 0.5470341444015503
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3019060492515564,
      "epoch": 7.3,
      "learning_rate": 3.9458407083281805e-05,
      "loss": 0.3466,
      "step": 20200,
      "task_loss": 0.4646940529346466
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3612298369407654,
      "epoch": 7.3,
      "learning_rate": 3.93795666004965e-05,
      "loss": 0.3733,
      "step": 20210,
      "task_loss": 0.6525784730911255
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41849541664123535,
      "epoch": 7.31,
      "learning_rate": 3.930070353519308e-05,
      "loss": 0.358,
      "step": 20220,
      "task_loss": 0.6312863826751709
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33530235290527344,
      "epoch": 7.31,
      "learning_rate": 3.9221818294016066e-05,
      "loss": 0.3541,
      "step": 20230,
      "task_loss": 0.5233701467514038
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28744643926620483,
      "epoch": 7.31,
      "learning_rate": 3.914291128372426e-05,
      "loss": 0.3271,
      "step": 20240,
      "task_loss": 0.5557354092597961
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4122638404369354,
      "epoch": 7.32,
      "learning_rate": 3.9063982911188756e-05,
      "loss": 0.3687,
      "step": 20250,
      "task_loss": 0.5771534442901611
    },
    {
      "epoch": 7.32,
      "eval_exact_match": 82.98013245033113,
      "eval_f1": 89.94478162383577,
      "step": 20250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40526461601257324,
      "epoch": 7.32,
      "learning_rate": 3.8985033583390775e-05,
      "loss": 0.388,
      "step": 20260,
      "task_loss": 0.8766403198242188
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5623233318328857,
      "epoch": 7.33,
      "learning_rate": 3.8906063707419606e-05,
      "loss": 0.3682,
      "step": 20270,
      "task_loss": 0.39787179231643677
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2818659543991089,
      "epoch": 7.33,
      "learning_rate": 3.882707369047048e-05,
      "loss": 0.3711,
      "step": 20280,
      "task_loss": 0.590981125831604
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.468313992023468,
      "epoch": 7.33,
      "learning_rate": 3.874806393984249e-05,
      "loss": 0.3643,
      "step": 20290,
      "task_loss": 0.6314665079116821
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2927176356315613,
      "epoch": 7.34,
      "learning_rate": 3.866903486293648e-05,
      "loss": 0.3748,
      "step": 20300,
      "task_loss": 0.6189347505569458
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32553645968437195,
      "epoch": 7.34,
      "learning_rate": 3.8589986867252935e-05,
      "loss": 0.3557,
      "step": 20310,
      "task_loss": 0.8069854974746704
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.287088006734848,
      "epoch": 7.34,
      "learning_rate": 3.851092036038991e-05,
      "loss": 0.3733,
      "step": 20320,
      "task_loss": 0.851733922958374
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41035404801368713,
      "epoch": 7.35,
      "learning_rate": 3.843183575004091e-05,
      "loss": 0.3983,
      "step": 20330,
      "task_loss": 0.53264319896698
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39696526527404785,
      "epoch": 7.35,
      "learning_rate": 3.835273344399275e-05,
      "loss": 0.3934,
      "step": 20340,
      "task_loss": 0.460975706577301
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3248482048511505,
      "epoch": 7.35,
      "learning_rate": 3.827361385012354e-05,
      "loss": 0.3619,
      "step": 20350,
      "task_loss": 0.5883316993713379
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.488412082195282,
      "epoch": 7.36,
      "learning_rate": 3.819447737640053e-05,
      "loss": 0.4008,
      "step": 20360,
      "task_loss": 0.699428141117096
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3935892581939697,
      "epoch": 7.36,
      "learning_rate": 3.8115324430877966e-05,
      "loss": 0.3865,
      "step": 20370,
      "task_loss": 0.6628360748291016
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29902222752571106,
      "epoch": 7.37,
      "learning_rate": 3.8036155421695045e-05,
      "loss": 0.3458,
      "step": 20380,
      "task_loss": 0.6145927309989929
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30101636052131653,
      "epoch": 7.37,
      "learning_rate": 3.7956970757073814e-05,
      "loss": 0.3705,
      "step": 20390,
      "task_loss": 0.4143185019493103
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22915220260620117,
      "epoch": 7.37,
      "learning_rate": 3.787777084531704e-05,
      "loss": 0.3822,
      "step": 20400,
      "task_loss": 0.35247260332107544
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3068666458129883,
      "epoch": 7.38,
      "learning_rate": 3.779855609480608e-05,
      "loss": 0.3304,
      "step": 20410,
      "task_loss": 0.7493861317634583
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3240140974521637,
      "epoch": 7.38,
      "learning_rate": 3.771932691399883e-05,
      "loss": 0.3831,
      "step": 20420,
      "task_loss": 0.4175964593887329
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30583804845809937,
      "epoch": 7.38,
      "learning_rate": 3.76400837114276e-05,
      "loss": 0.352,
      "step": 20430,
      "task_loss": 0.42632582783699036
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3739269971847534,
      "epoch": 7.39,
      "learning_rate": 3.7560826895696975e-05,
      "loss": 0.3465,
      "step": 20440,
      "task_loss": 0.38233160972595215
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32636404037475586,
      "epoch": 7.39,
      "learning_rate": 3.748155687548175e-05,
      "loss": 0.3609,
      "step": 20450,
      "task_loss": 0.560936689376831
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2921903729438782,
      "epoch": 7.39,
      "learning_rate": 3.740227405952482e-05,
      "loss": 0.3442,
      "step": 20460,
      "task_loss": 0.7394105792045593
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28189751505851746,
      "epoch": 7.4,
      "learning_rate": 3.732297885663502e-05,
      "loss": 0.3615,
      "step": 20470,
      "task_loss": 0.3475574851036072
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32151737809181213,
      "epoch": 7.4,
      "learning_rate": 3.7243671675685105e-05,
      "loss": 0.3768,
      "step": 20480,
      "task_loss": 0.35703375935554504
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4484822750091553,
      "epoch": 7.41,
      "learning_rate": 3.716435292560956e-05,
      "loss": 0.3516,
      "step": 20490,
      "task_loss": 0.8057897090911865
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2865402400493622,
      "epoch": 7.41,
      "learning_rate": 3.708502301540252e-05,
      "loss": 0.3487,
      "step": 20500,
      "task_loss": 0.7167019248008728
    },
    {
      "epoch": 7.41,
      "eval_exact_match": 82.96121097445601,
      "eval_f1": 89.79021405051186,
      "step": 20500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29780441522598267,
      "epoch": 7.41,
      "learning_rate": 3.70056823541157e-05,
      "loss": 0.4029,
      "step": 20510,
      "task_loss": 0.36201319098472595
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47126126289367676,
      "epoch": 7.42,
      "learning_rate": 3.692633135085621e-05,
      "loss": 0.3639,
      "step": 20520,
      "task_loss": 0.5977838039398193
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2689071297645569,
      "epoch": 7.42,
      "learning_rate": 3.684697041478451e-05,
      "loss": 0.358,
      "step": 20530,
      "task_loss": 0.38315659761428833
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28154638409614563,
      "epoch": 7.42,
      "learning_rate": 3.6767599955112285e-05,
      "loss": 0.3681,
      "step": 20540,
      "task_loss": 0.4523980915546417
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33390653133392334,
      "epoch": 7.43,
      "learning_rate": 3.66882203811003e-05,
      "loss": 0.3368,
      "step": 20550,
      "task_loss": 0.3338510990142822
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36619308590888977,
      "epoch": 7.43,
      "learning_rate": 3.6608832102056345e-05,
      "loss": 0.3908,
      "step": 20560,
      "task_loss": 0.49176403880119324
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4774571657180786,
      "epoch": 7.43,
      "learning_rate": 3.6529435527333065e-05,
      "loss": 0.3801,
      "step": 20570,
      "task_loss": 0.9242376089096069
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3975795805454254,
      "epoch": 7.44,
      "learning_rate": 3.645003106632591e-05,
      "loss": 0.3505,
      "step": 20580,
      "task_loss": 0.6916407346725464
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3709920048713684,
      "epoch": 7.44,
      "learning_rate": 3.637061912847099e-05,
      "loss": 0.3534,
      "step": 20590,
      "task_loss": 0.495505690574646
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2470513880252838,
      "epoch": 7.44,
      "learning_rate": 3.629120012324294e-05,
      "loss": 0.3301,
      "step": 20600,
      "task_loss": 0.14731258153915405
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2766827344894409,
      "epoch": 7.45,
      "learning_rate": 3.621177446015287e-05,
      "loss": 0.3582,
      "step": 20610,
      "task_loss": 0.2590910494327545
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34083378314971924,
      "epoch": 7.45,
      "learning_rate": 3.61323425487462e-05,
      "loss": 0.3602,
      "step": 20620,
      "task_loss": 0.4858570396900177
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27848386764526367,
      "epoch": 7.46,
      "learning_rate": 3.605290479860057e-05,
      "loss": 0.38,
      "step": 20630,
      "task_loss": 0.314744770526886
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2510620951652527,
      "epoch": 7.46,
      "learning_rate": 3.597346161932376e-05,
      "loss": 0.3617,
      "step": 20640,
      "task_loss": 0.33013665676116943
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31609711050987244,
      "epoch": 7.46,
      "learning_rate": 3.5894013420551494e-05,
      "loss": 0.3399,
      "step": 20650,
      "task_loss": 0.42119982838630676
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40757670998573303,
      "epoch": 7.47,
      "learning_rate": 3.58145606119454e-05,
      "loss": 0.3584,
      "step": 20660,
      "task_loss": 0.5284794569015503
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25075843930244446,
      "epoch": 7.47,
      "learning_rate": 3.573510360319088e-05,
      "loss": 0.3632,
      "step": 20670,
      "task_loss": 0.2154819518327713
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3524804413318634,
      "epoch": 7.47,
      "learning_rate": 3.565564280399499e-05,
      "loss": 0.3769,
      "step": 20680,
      "task_loss": 0.5062075257301331
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2267530858516693,
      "epoch": 7.48,
      "learning_rate": 3.557617862408434e-05,
      "loss": 0.3603,
      "step": 20690,
      "task_loss": 0.3778666853904724
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2886289954185486,
      "epoch": 7.48,
      "learning_rate": 3.5496711473202944e-05,
      "loss": 0.3234,
      "step": 20700,
      "task_loss": 0.4648386240005493
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30359941720962524,
      "epoch": 7.48,
      "learning_rate": 3.541724176111018e-05,
      "loss": 0.3388,
      "step": 20710,
      "task_loss": 0.5805556178092957
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.390267550945282,
      "epoch": 7.49,
      "learning_rate": 3.533776989757858e-05,
      "loss": 0.3511,
      "step": 20720,
      "task_loss": 0.9671446084976196
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39545518159866333,
      "epoch": 7.49,
      "learning_rate": 3.52582962923918e-05,
      "loss": 0.3803,
      "step": 20730,
      "task_loss": 0.4439846873283386
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25991690158843994,
      "epoch": 7.5,
      "learning_rate": 3.517882135534249e-05,
      "loss": 0.3708,
      "step": 20740,
      "task_loss": 0.40725821256637573
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3476290702819824,
      "epoch": 7.5,
      "learning_rate": 3.509934549623014e-05,
      "loss": 0.3794,
      "step": 20750,
      "task_loss": 0.40585631132125854
    },
    {
      "epoch": 7.5,
      "eval_exact_match": 83.14096499526963,
      "eval_f1": 90.06066489311088,
      "step": 20750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26795804500579834,
      "epoch": 7.5,
      "learning_rate": 3.501986912485901e-05,
      "loss": 0.3313,
      "step": 20760,
      "task_loss": 0.318131685256958
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27528253197669983,
      "epoch": 7.51,
      "learning_rate": 3.494039265103597e-05,
      "loss": 0.3785,
      "step": 20770,
      "task_loss": 0.6479158401489258
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4581003189086914,
      "epoch": 7.51,
      "learning_rate": 3.486091648456848e-05,
      "loss": 0.3914,
      "step": 20780,
      "task_loss": 0.5169153213500977
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28237971663475037,
      "epoch": 7.51,
      "learning_rate": 3.478144103526236e-05,
      "loss": 0.3394,
      "step": 20790,
      "task_loss": 0.4905843734741211
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35639041662216187,
      "epoch": 7.52,
      "learning_rate": 3.470196671291976e-05,
      "loss": 0.3705,
      "step": 20800,
      "task_loss": 0.5656434297561646
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3621225953102112,
      "epoch": 7.52,
      "learning_rate": 3.4622493927337014e-05,
      "loss": 0.3687,
      "step": 20810,
      "task_loss": 0.799970269203186
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2872578501701355,
      "epoch": 7.52,
      "learning_rate": 3.4543023088302515e-05,
      "loss": 0.3459,
      "step": 20820,
      "task_loss": 0.2272077053785324
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37445396184921265,
      "epoch": 7.53,
      "learning_rate": 3.4463554605594636e-05,
      "loss": 0.3734,
      "step": 20830,
      "task_loss": 0.6421491503715515
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3490419387817383,
      "epoch": 7.53,
      "learning_rate": 3.438408888897961e-05,
      "loss": 0.3361,
      "step": 20840,
      "task_loss": 0.5615388751029968
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3245730400085449,
      "epoch": 7.54,
      "learning_rate": 3.4304626348209364e-05,
      "loss": 0.3551,
      "step": 20850,
      "task_loss": 0.3907738924026489
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3850998282432556,
      "epoch": 7.54,
      "learning_rate": 3.4225167393019515e-05,
      "loss": 0.397,
      "step": 20860,
      "task_loss": 0.6859105825424194
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37591278553009033,
      "epoch": 7.54,
      "learning_rate": 3.4145712433127126e-05,
      "loss": 0.3667,
      "step": 20870,
      "task_loss": 0.4931633472442627
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35777273774147034,
      "epoch": 7.55,
      "learning_rate": 3.4066261878228706e-05,
      "loss": 0.3744,
      "step": 20880,
      "task_loss": 0.4974115490913391
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36851346492767334,
      "epoch": 7.55,
      "learning_rate": 3.398681613799803e-05,
      "loss": 0.3769,
      "step": 20890,
      "task_loss": 0.4662131071090698
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33928361535072327,
      "epoch": 7.55,
      "learning_rate": 3.390737562208405e-05,
      "loss": 0.4119,
      "step": 20900,
      "task_loss": 0.5498688220977783
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4270703196525574,
      "epoch": 7.56,
      "learning_rate": 3.3827940740108764e-05,
      "loss": 0.3709,
      "step": 20910,
      "task_loss": 0.481827050447464
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43292000889778137,
      "epoch": 7.56,
      "learning_rate": 3.374851190166516e-05,
      "loss": 0.3787,
      "step": 20920,
      "task_loss": 0.9506334066390991
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2657639980316162,
      "epoch": 7.56,
      "learning_rate": 3.366908951631501e-05,
      "loss": 0.3494,
      "step": 20930,
      "task_loss": 0.4249715209007263
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.394223690032959,
      "epoch": 7.57,
      "learning_rate": 3.3589673993586866e-05,
      "loss": 0.3941,
      "step": 20940,
      "task_loss": 0.5982447266578674
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2714296579360962,
      "epoch": 7.57,
      "learning_rate": 3.351026574297384e-05,
      "loss": 0.3427,
      "step": 20950,
      "task_loss": 0.6363648176193237
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3336118459701538,
      "epoch": 7.57,
      "learning_rate": 3.343086517393162e-05,
      "loss": 0.394,
      "step": 20960,
      "task_loss": 0.7720279693603516
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30867213010787964,
      "epoch": 7.58,
      "learning_rate": 3.335147269587619e-05,
      "loss": 0.3789,
      "step": 20970,
      "task_loss": 0.40268540382385254
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4145679771900177,
      "epoch": 7.58,
      "learning_rate": 3.327208871818189e-05,
      "loss": 0.3476,
      "step": 20980,
      "task_loss": 0.7717650532722473
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31620293855667114,
      "epoch": 7.59,
      "learning_rate": 3.3192713650179204e-05,
      "loss": 0.3984,
      "step": 20990,
      "task_loss": 0.5271072387695312
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3817560076713562,
      "epoch": 7.59,
      "learning_rate": 3.311334790115267e-05,
      "loss": 0.3455,
      "step": 21000,
      "task_loss": 0.6588694453239441
    },
    {
      "epoch": 7.59,
      "eval_exact_match": 83.01797540208136,
      "eval_f1": 90.0024701105002,
      "step": 21000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.304983913898468,
      "epoch": 7.59,
      "learning_rate": 3.303399188033876e-05,
      "loss": 0.3438,
      "step": 21010,
      "task_loss": 0.49317359924316406
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3328720033168793,
      "epoch": 7.6,
      "learning_rate": 3.295464599692384e-05,
      "loss": 0.3604,
      "step": 21020,
      "task_loss": 0.6254655718803406
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31305211782455444,
      "epoch": 7.6,
      "learning_rate": 3.287531066004193e-05,
      "loss": 0.3649,
      "step": 21030,
      "task_loss": 0.5049108266830444
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38831958174705505,
      "epoch": 7.6,
      "learning_rate": 3.279598627877273e-05,
      "loss": 0.368,
      "step": 21040,
      "task_loss": 0.3947518467903137
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29545751214027405,
      "epoch": 7.61,
      "learning_rate": 3.2716673262139395e-05,
      "loss": 0.3442,
      "step": 21050,
      "task_loss": 0.9915268421173096
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33492806553840637,
      "epoch": 7.61,
      "learning_rate": 3.2637372019106546e-05,
      "loss": 0.3608,
      "step": 21060,
      "task_loss": 0.49239182472229004
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31485992670059204,
      "epoch": 7.61,
      "learning_rate": 3.255808295857803e-05,
      "loss": 0.3431,
      "step": 21070,
      "task_loss": 0.811964750289917
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32304200530052185,
      "epoch": 7.62,
      "learning_rate": 3.247880648939492e-05,
      "loss": 0.3626,
      "step": 21080,
      "task_loss": 0.3488319218158722
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4019070267677307,
      "epoch": 7.62,
      "learning_rate": 3.239954302033335e-05,
      "loss": 0.3459,
      "step": 21090,
      "task_loss": 0.6527318954467773
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35567885637283325,
      "epoch": 7.63,
      "learning_rate": 3.232029296010241e-05,
      "loss": 0.3526,
      "step": 21100,
      "task_loss": 0.4953433871269226
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.248754620552063,
      "epoch": 7.63,
      "learning_rate": 3.224105671734206e-05,
      "loss": 0.3247,
      "step": 21110,
      "task_loss": 0.5917302370071411
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3231223523616791,
      "epoch": 7.63,
      "learning_rate": 3.2161834700621026e-05,
      "loss": 0.3262,
      "step": 21120,
      "task_loss": 0.6599835157394409
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3281194567680359,
      "epoch": 7.64,
      "learning_rate": 3.2082627318434634e-05,
      "loss": 0.3392,
      "step": 21130,
      "task_loss": 0.7458435297012329
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37381210923194885,
      "epoch": 7.64,
      "learning_rate": 3.2003434979202836e-05,
      "loss": 0.3412,
      "step": 21140,
      "task_loss": 0.6678844690322876
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22763706743717194,
      "epoch": 7.64,
      "learning_rate": 3.19242580912679e-05,
      "loss": 0.3519,
      "step": 21150,
      "task_loss": 0.42863303422927856
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2760619819164276,
      "epoch": 7.65,
      "learning_rate": 3.1845097062892545e-05,
      "loss": 0.3365,
      "step": 21160,
      "task_loss": 0.6848987340927124
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38911378383636475,
      "epoch": 7.65,
      "learning_rate": 3.1765952302257604e-05,
      "loss": 0.3308,
      "step": 21170,
      "task_loss": 0.2971108555793762
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39150434732437134,
      "epoch": 7.65,
      "learning_rate": 3.1686824217460105e-05,
      "loss": 0.3895,
      "step": 21180,
      "task_loss": 0.7148730158805847
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.48236069083213806,
      "epoch": 7.66,
      "learning_rate": 3.160771321651105e-05,
      "loss": 0.3615,
      "step": 21190,
      "task_loss": 0.5340760350227356
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2843457758426666,
      "epoch": 7.66,
      "learning_rate": 3.152861970733336e-05,
      "loss": 0.3503,
      "step": 21200,
      "task_loss": 0.5471694469451904
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37405431270599365,
      "epoch": 7.67,
      "learning_rate": 3.144954409775978e-05,
      "loss": 0.353,
      "step": 21210,
      "task_loss": 0.5435091853141785
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3254290223121643,
      "epoch": 7.67,
      "learning_rate": 3.1370486795530724e-05,
      "loss": 0.3726,
      "step": 21220,
      "task_loss": 0.44784238934516907
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3736199736595154,
      "epoch": 7.67,
      "learning_rate": 3.129144820829223e-05,
      "loss": 0.3367,
      "step": 21230,
      "task_loss": 0.684989869594574
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2720557749271393,
      "epoch": 7.68,
      "learning_rate": 3.1212428743593856e-05,
      "loss": 0.3924,
      "step": 21240,
      "task_loss": 0.43417856097221375
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3352609872817993,
      "epoch": 7.68,
      "learning_rate": 3.113342880888649e-05,
      "loss": 0.3355,
      "step": 21250,
      "task_loss": 0.5457616448402405
    },
    {
      "epoch": 7.68,
      "eval_exact_match": 82.97067171239357,
      "eval_f1": 89.75270243191726,
      "step": 21250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2917085886001587,
      "epoch": 7.68,
      "learning_rate": 3.10544488115204e-05,
      "loss": 0.3251,
      "step": 21260,
      "task_loss": 0.4778927266597748
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3367916941642761,
      "epoch": 7.69,
      "learning_rate": 3.097548915874299e-05,
      "loss": 0.3507,
      "step": 21270,
      "task_loss": 0.5044569373130798
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3709861636161804,
      "epoch": 7.69,
      "learning_rate": 3.08965502576968e-05,
      "loss": 0.4027,
      "step": 21280,
      "task_loss": 0.5584628582000732
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35419756174087524,
      "epoch": 7.69,
      "learning_rate": 3.081763251541732e-05,
      "loss": 0.3693,
      "step": 21290,
      "task_loss": 0.6167153120040894
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35388609766960144,
      "epoch": 7.7,
      "learning_rate": 3.0738736338830997e-05,
      "loss": 0.3349,
      "step": 21300,
      "task_loss": 0.420735239982605
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28986817598342896,
      "epoch": 7.7,
      "learning_rate": 3.0659862134753025e-05,
      "loss": 0.3477,
      "step": 21310,
      "task_loss": 0.3834241032600403
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4581122100353241,
      "epoch": 7.71,
      "learning_rate": 3.0581010309885335e-05,
      "loss": 0.3659,
      "step": 21320,
      "task_loss": 0.6090573072433472
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2977628707885742,
      "epoch": 7.71,
      "learning_rate": 3.0502181270814433e-05,
      "loss": 0.3684,
      "step": 21330,
      "task_loss": 0.28831470012664795
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32104063034057617,
      "epoch": 7.71,
      "learning_rate": 3.042337542400939e-05,
      "loss": 0.3712,
      "step": 21340,
      "task_loss": 0.4596444368362427
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39884674549102783,
      "epoch": 7.72,
      "learning_rate": 3.0344593175819606e-05,
      "loss": 0.3462,
      "step": 21350,
      "task_loss": 0.8089415431022644
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33032548427581787,
      "epoch": 7.72,
      "learning_rate": 3.026583493247288e-05,
      "loss": 0.3524,
      "step": 21360,
      "task_loss": 0.6769004464149475
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37094488739967346,
      "epoch": 7.72,
      "learning_rate": 3.018710110007318e-05,
      "loss": 0.3836,
      "step": 21370,
      "task_loss": 0.5415042638778687
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2806081175804138,
      "epoch": 7.73,
      "learning_rate": 3.010839208459863e-05,
      "loss": 0.3942,
      "step": 21380,
      "task_loss": 0.598015308380127
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.363537073135376,
      "epoch": 7.73,
      "learning_rate": 3.0029708291899378e-05,
      "loss": 0.3359,
      "step": 21390,
      "task_loss": 0.7415146231651306
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32732537388801575,
      "epoch": 7.73,
      "learning_rate": 2.9951050127695518e-05,
      "loss": 0.3601,
      "step": 21400,
      "task_loss": 0.7806991338729858
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44270604848861694,
      "epoch": 7.74,
      "learning_rate": 2.9872417997574987e-05,
      "loss": 0.3318,
      "step": 21410,
      "task_loss": 0.5888664722442627
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2828146815299988,
      "epoch": 7.74,
      "learning_rate": 2.979381230699151e-05,
      "loss": 0.3797,
      "step": 21420,
      "task_loss": 0.6815419793128967
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3201327919960022,
      "epoch": 7.74,
      "learning_rate": 2.9715233461262427e-05,
      "loss": 0.3256,
      "step": 21430,
      "task_loss": 0.7906991243362427
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2204771190881729,
      "epoch": 7.75,
      "learning_rate": 2.9636681865566735e-05,
      "loss": 0.378,
      "step": 21440,
      "task_loss": 0.3519682288169861
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3597320020198822,
      "epoch": 7.75,
      "learning_rate": 2.9558157924942824e-05,
      "loss": 0.3507,
      "step": 21450,
      "task_loss": 0.8138501048088074
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21947142481803894,
      "epoch": 7.76,
      "learning_rate": 2.947966204428658e-05,
      "loss": 0.3732,
      "step": 21460,
      "task_loss": 0.6017472743988037
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3029401898384094,
      "epoch": 7.76,
      "learning_rate": 2.940119462834914e-05,
      "loss": 0.3162,
      "step": 21470,
      "task_loss": 0.4510578513145447
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.48782584071159363,
      "epoch": 7.76,
      "learning_rate": 2.9322756081734894e-05,
      "loss": 0.3629,
      "step": 21480,
      "task_loss": 0.8380551338195801
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34943652153015137,
      "epoch": 7.77,
      "learning_rate": 2.9244346808899347e-05,
      "loss": 0.3598,
      "step": 21490,
      "task_loss": 0.8652399182319641
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.343707799911499,
      "epoch": 7.77,
      "learning_rate": 2.91659672141471e-05,
      "loss": 0.3269,
      "step": 21500,
      "task_loss": 0.6488476991653442
    },
    {
      "epoch": 7.77,
      "eval_exact_match": 83.22611163670766,
      "eval_f1": 90.06873548509186,
      "step": 21500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28079891204833984,
      "epoch": 7.77,
      "learning_rate": 2.9087617701629685e-05,
      "loss": 0.3729,
      "step": 21510,
      "task_loss": 0.5381488800048828
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30641067028045654,
      "epoch": 7.78,
      "learning_rate": 2.9009298675343544e-05,
      "loss": 0.3535,
      "step": 21520,
      "task_loss": 0.3132174015045166
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4152531921863556,
      "epoch": 7.78,
      "learning_rate": 2.8931010539127894e-05,
      "loss": 0.3741,
      "step": 21530,
      "task_loss": 0.7610316276550293
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36251795291900635,
      "epoch": 7.78,
      "learning_rate": 2.8852753696662726e-05,
      "loss": 0.3404,
      "step": 21540,
      "task_loss": 0.6292003393173218
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40879419445991516,
      "epoch": 7.79,
      "learning_rate": 2.877452855146659e-05,
      "loss": 0.3618,
      "step": 21550,
      "task_loss": 0.7074788808822632
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33598774671554565,
      "epoch": 7.79,
      "learning_rate": 2.8696335506894678e-05,
      "loss": 0.3758,
      "step": 21560,
      "task_loss": 0.9449920654296875
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3502233624458313,
      "epoch": 7.8,
      "learning_rate": 2.8618174966136597e-05,
      "loss": 0.3865,
      "step": 21570,
      "task_loss": 0.4362756311893463
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44666188955307007,
      "epoch": 7.8,
      "learning_rate": 2.8540047332214402e-05,
      "loss": 0.3597,
      "step": 21580,
      "task_loss": 0.5449387431144714
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3455656170845032,
      "epoch": 7.8,
      "learning_rate": 2.8461953007980413e-05,
      "loss": 0.3728,
      "step": 21590,
      "task_loss": 0.37384119629859924
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37466901540756226,
      "epoch": 7.81,
      "learning_rate": 2.8383892396115255e-05,
      "loss": 0.3396,
      "step": 21600,
      "task_loss": 0.4856455326080322
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30749914050102234,
      "epoch": 7.81,
      "learning_rate": 2.830586589912568e-05,
      "loss": 0.3426,
      "step": 21610,
      "task_loss": 0.5648226737976074
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35594284534454346,
      "epoch": 7.81,
      "learning_rate": 2.8227873919342545e-05,
      "loss": 0.3668,
      "step": 21620,
      "task_loss": 0.42687565088272095
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3522205650806427,
      "epoch": 7.82,
      "learning_rate": 2.814991685891871e-05,
      "loss": 0.3696,
      "step": 21630,
      "task_loss": 0.823002278804779
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3458482027053833,
      "epoch": 7.82,
      "learning_rate": 2.807199511982703e-05,
      "loss": 0.3495,
      "step": 21640,
      "task_loss": 0.43529629707336426
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2593701481819153,
      "epoch": 7.82,
      "learning_rate": 2.7994109103858147e-05,
      "loss": 0.3302,
      "step": 21650,
      "task_loss": 0.564799964427948
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3371639549732208,
      "epoch": 7.83,
      "learning_rate": 2.7916259212618592e-05,
      "loss": 0.3596,
      "step": 21660,
      "task_loss": 0.8256035447120667
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4070214033126831,
      "epoch": 7.83,
      "learning_rate": 2.7838445847528554e-05,
      "loss": 0.3238,
      "step": 21670,
      "task_loss": 0.38024699687957764
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3013300895690918,
      "epoch": 7.84,
      "learning_rate": 2.776066940981993e-05,
      "loss": 0.3573,
      "step": 21680,
      "task_loss": 0.42121320962905884
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33513176441192627,
      "epoch": 7.84,
      "learning_rate": 2.7682930300534165e-05,
      "loss": 0.3414,
      "step": 21690,
      "task_loss": 0.5897566080093384
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3103760778903961,
      "epoch": 7.84,
      "learning_rate": 2.7605228920520273e-05,
      "loss": 0.3171,
      "step": 21700,
      "task_loss": 0.6985391974449158
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32781437039375305,
      "epoch": 7.85,
      "learning_rate": 2.7527565670432668e-05,
      "loss": 0.3543,
      "step": 21710,
      "task_loss": 0.6251723766326904
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2737557291984558,
      "epoch": 7.85,
      "learning_rate": 2.7449940950729213e-05,
      "loss": 0.3419,
      "step": 21720,
      "task_loss": 0.6766575574874878
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27434462308883667,
      "epoch": 7.85,
      "learning_rate": 2.737235516166903e-05,
      "loss": 0.3292,
      "step": 21730,
      "task_loss": 0.4648294150829315
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29078248143196106,
      "epoch": 7.86,
      "learning_rate": 2.729480870331058e-05,
      "loss": 0.3265,
      "step": 21740,
      "task_loss": 0.6191574335098267
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26681390404701233,
      "epoch": 7.86,
      "learning_rate": 2.721730197550944e-05,
      "loss": 0.3432,
      "step": 21750,
      "task_loss": 0.4409167766571045
    },
    {
      "epoch": 7.86,
      "eval_exact_match": 82.74361400189214,
      "eval_f1": 89.81621091651192,
      "step": 21750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45901256799697876,
      "epoch": 7.86,
      "learning_rate": 2.7139835377916394e-05,
      "loss": 0.3586,
      "step": 21760,
      "task_loss": 0.7206564545631409
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3710944354534149,
      "epoch": 7.87,
      "learning_rate": 2.7062409309975242e-05,
      "loss": 0.359,
      "step": 21770,
      "task_loss": 0.7943992614746094
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34832853078842163,
      "epoch": 7.87,
      "learning_rate": 2.6985024170920843e-05,
      "loss": 0.3692,
      "step": 21780,
      "task_loss": 0.3599025309085846
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3316296935081482,
      "epoch": 7.87,
      "learning_rate": 2.690768035977699e-05,
      "loss": 0.3727,
      "step": 21790,
      "task_loss": 0.6260291337966919
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25265419483184814,
      "epoch": 7.88,
      "learning_rate": 2.6830378275354378e-05,
      "loss": 0.3767,
      "step": 21800,
      "task_loss": 0.29135915637016296
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4612353444099426,
      "epoch": 7.88,
      "learning_rate": 2.6753118316248545e-05,
      "loss": 0.3797,
      "step": 21810,
      "task_loss": 0.5710431337356567
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3590574264526367,
      "epoch": 7.89,
      "learning_rate": 2.6675900880837846e-05,
      "loss": 0.3856,
      "step": 21820,
      "task_loss": 0.6375476121902466
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2918437719345093,
      "epoch": 7.89,
      "learning_rate": 2.65987263672813e-05,
      "loss": 0.3624,
      "step": 21830,
      "task_loss": 0.3540777564048767
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38157540559768677,
      "epoch": 7.89,
      "learning_rate": 2.652159517351669e-05,
      "loss": 0.351,
      "step": 21840,
      "task_loss": 0.3811739385128021
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39688819646835327,
      "epoch": 7.9,
      "learning_rate": 2.644450769725837e-05,
      "loss": 0.365,
      "step": 21850,
      "task_loss": 0.533487856388092
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32374733686447144,
      "epoch": 7.9,
      "learning_rate": 2.6367464335995296e-05,
      "loss": 0.3343,
      "step": 21860,
      "task_loss": 0.4965505301952362
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36838701367378235,
      "epoch": 7.9,
      "learning_rate": 2.6290465486988934e-05,
      "loss": 0.3604,
      "step": 21870,
      "task_loss": 0.7612681984901428
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2574426531791687,
      "epoch": 7.91,
      "learning_rate": 2.6213511547271256e-05,
      "loss": 0.3493,
      "step": 21880,
      "task_loss": 0.33988291025161743
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.324415385723114,
      "epoch": 7.91,
      "learning_rate": 2.613660291364264e-05,
      "loss": 0.3632,
      "step": 21890,
      "task_loss": 0.697488009929657
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2612054944038391,
      "epoch": 7.91,
      "learning_rate": 2.6059739982669875e-05,
      "loss": 0.3544,
      "step": 21900,
      "task_loss": 0.5947055816650391
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3524448871612549,
      "epoch": 7.92,
      "learning_rate": 2.5982923150684062e-05,
      "loss": 0.3431,
      "step": 21910,
      "task_loss": 0.31323203444480896
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3077560067176819,
      "epoch": 7.92,
      "learning_rate": 2.5906152813778642e-05,
      "loss": 0.3363,
      "step": 21920,
      "task_loss": 0.5763083696365356
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32099348306655884,
      "epoch": 7.93,
      "learning_rate": 2.582942936780726e-05,
      "loss": 0.3612,
      "step": 21930,
      "task_loss": 0.7717790603637695
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43049126863479614,
      "epoch": 7.93,
      "learning_rate": 2.575275320838183e-05,
      "loss": 0.3875,
      "step": 21940,
      "task_loss": 0.670395016670227
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3461580276489258,
      "epoch": 7.93,
      "learning_rate": 2.5676124730870394e-05,
      "loss": 0.3417,
      "step": 21950,
      "task_loss": 0.7263317108154297
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2610408365726471,
      "epoch": 7.94,
      "learning_rate": 2.5599544330395164e-05,
      "loss": 0.3782,
      "step": 21960,
      "task_loss": 0.5060991048812866
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33723878860473633,
      "epoch": 7.94,
      "learning_rate": 2.5523012401830417e-05,
      "loss": 0.3645,
      "step": 21970,
      "task_loss": 0.5092849135398865
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3296668827533722,
      "epoch": 7.94,
      "learning_rate": 2.5446529339800535e-05,
      "loss": 0.3522,
      "step": 21980,
      "task_loss": 0.6965002417564392
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2845882177352905,
      "epoch": 7.95,
      "learning_rate": 2.5370095538677883e-05,
      "loss": 0.3182,
      "step": 21990,
      "task_loss": 0.4682374596595764
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31272655725479126,
      "epoch": 7.95,
      "learning_rate": 2.529371139258086e-05,
      "loss": 0.3591,
      "step": 22000,
      "task_loss": 0.5068236589431763
    },
    {
      "epoch": 7.95,
      "eval_exact_match": 82.64900662251655,
      "eval_f1": 89.68277737846533,
      "step": 22000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3210468292236328,
      "epoch": 7.95,
      "learning_rate": 2.5217377295371787e-05,
      "loss": 0.3422,
      "step": 22010,
      "task_loss": 0.3419007658958435
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5161848068237305,
      "epoch": 7.96,
      "learning_rate": 2.5141093640654972e-05,
      "loss": 0.382,
      "step": 22020,
      "task_loss": 0.9878392815589905
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3749774694442749,
      "epoch": 7.96,
      "learning_rate": 2.506486082177455e-05,
      "loss": 0.3769,
      "step": 22030,
      "task_loss": 0.583120584487915
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4009551405906677,
      "epoch": 7.97,
      "learning_rate": 2.4988679231812602e-05,
      "loss": 0.33,
      "step": 22040,
      "task_loss": 0.9236621260643005
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2605592906475067,
      "epoch": 7.97,
      "learning_rate": 2.4912549263587004e-05,
      "loss": 0.4191,
      "step": 22050,
      "task_loss": 0.47352492809295654
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2575927674770355,
      "epoch": 7.97,
      "learning_rate": 2.4836471309649488e-05,
      "loss": 0.3525,
      "step": 22060,
      "task_loss": 0.4551050066947937
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32211726903915405,
      "epoch": 7.98,
      "learning_rate": 2.4760445762283548e-05,
      "loss": 0.3314,
      "step": 22070,
      "task_loss": 0.7064181566238403
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26055270433425903,
      "epoch": 7.98,
      "learning_rate": 2.468447301350249e-05,
      "loss": 0.3222,
      "step": 22080,
      "task_loss": 0.8715969324111938
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3303118944168091,
      "epoch": 7.98,
      "learning_rate": 2.4608553455047338e-05,
      "loss": 0.3335,
      "step": 22090,
      "task_loss": 0.3880299925804138
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39064037799835205,
      "epoch": 7.99,
      "learning_rate": 2.4532687478384877e-05,
      "loss": 0.3472,
      "step": 22100,
      "task_loss": 0.5962386131286621
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41516995429992676,
      "epoch": 7.99,
      "learning_rate": 2.4456875474705573e-05,
      "loss": 0.3663,
      "step": 22110,
      "task_loss": 0.6396934986114502
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5140536427497864,
      "epoch": 7.99,
      "learning_rate": 2.4381117834921653e-05,
      "loss": 0.341,
      "step": 22120,
      "task_loss": 0.5780425071716309
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37106651067733765,
      "epoch": 8.0,
      "learning_rate": 2.4305414949664922e-05,
      "loss": 0.3482,
      "step": 22130,
      "task_loss": 0.5088555812835693
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2919508218765259,
      "epoch": 8.0,
      "learning_rate": 2.422976720928497e-05,
      "loss": 0.3278,
      "step": 22140,
      "task_loss": 0.4410659670829773
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2985595464706421,
      "epoch": 8.01,
      "learning_rate": 2.415417500384695e-05,
      "loss": 0.2867,
      "step": 22150,
      "task_loss": 0.3100461959838867
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22516822814941406,
      "epoch": 8.01,
      "learning_rate": 2.4078638723129704e-05,
      "loss": 0.3064,
      "step": 22160,
      "task_loss": 0.2526964545249939
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3517003357410431,
      "epoch": 8.01,
      "learning_rate": 2.40031587566237e-05,
      "loss": 0.3218,
      "step": 22170,
      "task_loss": 0.2854512333869934
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2611880302429199,
      "epoch": 8.02,
      "learning_rate": 2.3927735493529002e-05,
      "loss": 0.288,
      "step": 22180,
      "task_loss": 0.40842729806900024
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2626832127571106,
      "epoch": 8.02,
      "learning_rate": 2.385236932275336e-05,
      "loss": 0.3052,
      "step": 22190,
      "task_loss": 0.6702091693878174
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25344762206077576,
      "epoch": 8.02,
      "learning_rate": 2.377706063291005e-05,
      "loss": 0.2927,
      "step": 22200,
      "task_loss": 0.5836098790168762
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23019495606422424,
      "epoch": 8.03,
      "learning_rate": 2.3701809812316033e-05,
      "loss": 0.2879,
      "step": 22210,
      "task_loss": 0.5896817445755005
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2813813388347626,
      "epoch": 8.03,
      "learning_rate": 2.3626617248989822e-05,
      "loss": 0.2885,
      "step": 22220,
      "task_loss": 0.4454318881034851
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3009265959262848,
      "epoch": 8.03,
      "learning_rate": 2.3551483330649577e-05,
      "loss": 0.3059,
      "step": 22230,
      "task_loss": 0.67862468957901
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3150326907634735,
      "epoch": 8.04,
      "learning_rate": 2.3476408444711027e-05,
      "loss": 0.2978,
      "step": 22240,
      "task_loss": 0.5857815146446228
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2585512399673462,
      "epoch": 8.04,
      "learning_rate": 2.3401392978285548e-05,
      "loss": 0.3233,
      "step": 22250,
      "task_loss": 0.25251004099845886
    },
    {
      "epoch": 8.04,
      "eval_exact_match": 83.0558183538316,
      "eval_f1": 89.90000054499299,
      "step": 22250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3542407155036926,
      "epoch": 8.04,
      "learning_rate": 2.3326437318178093e-05,
      "loss": 0.3108,
      "step": 22260,
      "task_loss": 1.0562341213226318
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2739490270614624,
      "epoch": 8.05,
      "learning_rate": 2.3251541850885265e-05,
      "loss": 0.3013,
      "step": 22270,
      "task_loss": 0.9316976070404053
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26413020491600037,
      "epoch": 8.05,
      "learning_rate": 2.3176706962593264e-05,
      "loss": 0.3378,
      "step": 22280,
      "task_loss": 0.5294308662414551
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3085627853870392,
      "epoch": 8.06,
      "learning_rate": 2.310193303917596e-05,
      "loss": 0.3001,
      "step": 22290,
      "task_loss": 0.4382203221321106
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27427783608436584,
      "epoch": 8.06,
      "learning_rate": 2.302722046619281e-05,
      "loss": 0.3169,
      "step": 22300,
      "task_loss": 0.39493346214294434
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39170312881469727,
      "epoch": 8.06,
      "learning_rate": 2.2952569628887004e-05,
      "loss": 0.3369,
      "step": 22310,
      "task_loss": 0.48804953694343567
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36464059352874756,
      "epoch": 8.07,
      "learning_rate": 2.2877980912183336e-05,
      "loss": 0.2951,
      "step": 22320,
      "task_loss": 0.524968147277832
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24969807267189026,
      "epoch": 8.07,
      "learning_rate": 2.2803454700686325e-05,
      "loss": 0.2994,
      "step": 22330,
      "task_loss": 0.8763778805732727
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26518428325653076,
      "epoch": 8.07,
      "learning_rate": 2.2728991378678165e-05,
      "loss": 0.2839,
      "step": 22340,
      "task_loss": 0.5413922071456909
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.16767820715904236,
      "epoch": 8.08,
      "learning_rate": 2.2654591330116794e-05,
      "loss": 0.3052,
      "step": 22350,
      "task_loss": 0.3252837061882019
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35951513051986694,
      "epoch": 8.08,
      "learning_rate": 2.258025493863388e-05,
      "loss": 0.3146,
      "step": 22360,
      "task_loss": 0.5174765586853027
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2331804484128952,
      "epoch": 8.08,
      "learning_rate": 2.250598258753285e-05,
      "loss": 0.333,
      "step": 22370,
      "task_loss": 0.354092001914978
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2697516083717346,
      "epoch": 8.09,
      "learning_rate": 2.243177465978691e-05,
      "loss": 0.3288,
      "step": 22380,
      "task_loss": 0.4610014855861664
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28994688391685486,
      "epoch": 8.09,
      "learning_rate": 2.2357631538037123e-05,
      "loss": 0.3005,
      "step": 22390,
      "task_loss": 0.5197626352310181
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2243901789188385,
      "epoch": 8.1,
      "learning_rate": 2.2283553604590325e-05,
      "loss": 0.2993,
      "step": 22400,
      "task_loss": 0.5959091186523438
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3568100929260254,
      "epoch": 8.1,
      "learning_rate": 2.220954124141727e-05,
      "loss": 0.3233,
      "step": 22410,
      "task_loss": 0.6142816543579102
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3616940975189209,
      "epoch": 8.1,
      "learning_rate": 2.2135594830150573e-05,
      "loss": 0.3116,
      "step": 22420,
      "task_loss": 0.5814266204833984
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32821333408355713,
      "epoch": 8.11,
      "learning_rate": 2.2061714752082815e-05,
      "loss": 0.3199,
      "step": 22430,
      "task_loss": 0.38982293009757996
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28196531534194946,
      "epoch": 8.11,
      "learning_rate": 2.1987901388164503e-05,
      "loss": 0.3165,
      "step": 22440,
      "task_loss": 0.5850852727890015
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2542758584022522,
      "epoch": 8.11,
      "learning_rate": 2.1914155119002177e-05,
      "loss": 0.2977,
      "step": 22450,
      "task_loss": 0.6517289876937866
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3016597628593445,
      "epoch": 8.12,
      "learning_rate": 2.1840476324856386e-05,
      "loss": 0.2977,
      "step": 22460,
      "task_loss": 0.5211827754974365
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44591793417930603,
      "epoch": 8.12,
      "learning_rate": 2.1766865385639794e-05,
      "loss": 0.345,
      "step": 22470,
      "task_loss": 0.6531864404678345
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22515526413917542,
      "epoch": 8.12,
      "learning_rate": 2.169332268091512e-05,
      "loss": 0.2851,
      "step": 22480,
      "task_loss": 0.39858129620552063
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3161879777908325,
      "epoch": 8.13,
      "learning_rate": 2.161984858989331e-05,
      "loss": 0.312,
      "step": 22490,
      "task_loss": 0.5769674181938171
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2485162615776062,
      "epoch": 8.13,
      "learning_rate": 2.154644349143147e-05,
      "loss": 0.2835,
      "step": 22500,
      "task_loss": 0.4577278196811676
    },
    {
      "epoch": 8.13,
      "eval_exact_match": 82.90444654683066,
      "eval_f1": 89.85075823845933,
      "step": 22500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3061697483062744,
      "epoch": 8.14,
      "learning_rate": 2.1473107764030978e-05,
      "loss": 0.325,
      "step": 22510,
      "task_loss": 0.3557536005973816
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27145296335220337,
      "epoch": 8.14,
      "learning_rate": 2.1399841785835495e-05,
      "loss": 0.2845,
      "step": 22520,
      "task_loss": 0.2070748656988144
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28222960233688354,
      "epoch": 8.14,
      "learning_rate": 2.1326645934629052e-05,
      "loss": 0.3249,
      "step": 22530,
      "task_loss": 0.7930005788803101
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.377989798784256,
      "epoch": 8.15,
      "learning_rate": 2.1253520587834062e-05,
      "loss": 0.296,
      "step": 22540,
      "task_loss": 0.9438679218292236
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27759093046188354,
      "epoch": 8.15,
      "learning_rate": 2.1180466122509413e-05,
      "loss": 0.3032,
      "step": 22550,
      "task_loss": 0.33349353075027466
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3169790804386139,
      "epoch": 8.15,
      "learning_rate": 2.1107482915348477e-05,
      "loss": 0.3342,
      "step": 22560,
      "task_loss": 0.6454753875732422
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32010066509246826,
      "epoch": 8.16,
      "learning_rate": 2.1034571342677242e-05,
      "loss": 0.3413,
      "step": 22570,
      "task_loss": 0.7830992937088013
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31526443362236023,
      "epoch": 8.16,
      "learning_rate": 2.0961731780452256e-05,
      "loss": 0.3276,
      "step": 22580,
      "task_loss": 0.5468152761459351
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3077006936073303,
      "epoch": 8.16,
      "learning_rate": 2.0888964604258828e-05,
      "loss": 0.2857,
      "step": 22590,
      "task_loss": 0.5536618828773499
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24466858804225922,
      "epoch": 8.17,
      "learning_rate": 2.0816270189308964e-05,
      "loss": 0.2898,
      "step": 22600,
      "task_loss": 0.6005507707595825
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27373987436294556,
      "epoch": 8.17,
      "learning_rate": 2.0743648910439537e-05,
      "loss": 0.2916,
      "step": 22610,
      "task_loss": 0.4629909098148346
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30114513635635376,
      "epoch": 8.17,
      "learning_rate": 2.0671101142110257e-05,
      "loss": 0.3162,
      "step": 22620,
      "task_loss": 0.5511022806167603
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3324176073074341,
      "epoch": 8.18,
      "learning_rate": 2.0598627258401834e-05,
      "loss": 0.3083,
      "step": 22630,
      "task_loss": 0.8414646983146667
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33706897497177124,
      "epoch": 8.18,
      "learning_rate": 2.0526227633013956e-05,
      "loss": 0.31,
      "step": 22640,
      "task_loss": 0.5560680031776428
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3432811498641968,
      "epoch": 8.19,
      "learning_rate": 2.045390263926347e-05,
      "loss": 0.3108,
      "step": 22650,
      "task_loss": 0.7345128655433655
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3108120560646057,
      "epoch": 8.19,
      "learning_rate": 2.0381652650082324e-05,
      "loss": 0.2908,
      "step": 22660,
      "task_loss": 0.5983444452285767
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2624809741973877,
      "epoch": 8.19,
      "learning_rate": 2.0309478038015807e-05,
      "loss": 0.3283,
      "step": 22670,
      "task_loss": 0.30220234394073486
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28360921144485474,
      "epoch": 8.2,
      "learning_rate": 2.023737917522045e-05,
      "loss": 0.351,
      "step": 22680,
      "task_loss": 0.4737190008163452
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35011863708496094,
      "epoch": 8.2,
      "learning_rate": 2.016535643346224e-05,
      "loss": 0.3214,
      "step": 22690,
      "task_loss": 0.5056818723678589
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3075874447822571,
      "epoch": 8.2,
      "learning_rate": 2.0093410184114667e-05,
      "loss": 0.3411,
      "step": 22700,
      "task_loss": 0.36893951892852783
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.286554217338562,
      "epoch": 8.21,
      "learning_rate": 2.0021540798156785e-05,
      "loss": 0.303,
      "step": 22710,
      "task_loss": 0.4626897871494293
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20945802330970764,
      "epoch": 8.21,
      "learning_rate": 1.9949748646171282e-05,
      "loss": 0.3026,
      "step": 22720,
      "task_loss": 0.4665585458278656
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2661295235157013,
      "epoch": 8.21,
      "learning_rate": 1.9878034098342688e-05,
      "loss": 0.2739,
      "step": 22730,
      "task_loss": 0.44505608081817627
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3084584176540375,
      "epoch": 8.22,
      "learning_rate": 1.980639752445529e-05,
      "loss": 0.3236,
      "step": 22740,
      "task_loss": 0.47151315212249756
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37254661321640015,
      "epoch": 8.22,
      "learning_rate": 1.973483929389136e-05,
      "loss": 0.3031,
      "step": 22750,
      "task_loss": 0.5002536177635193
    },
    {
      "epoch": 8.22,
      "eval_exact_match": 83.36802270577105,
      "eval_f1": 90.14923737850407,
      "step": 22750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24313656985759735,
      "epoch": 8.23,
      "learning_rate": 1.9663359775629198e-05,
      "loss": 0.3456,
      "step": 22760,
      "task_loss": 0.3750152289867401
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2501145899295807,
      "epoch": 8.23,
      "learning_rate": 1.959195933824125e-05,
      "loss": 0.3191,
      "step": 22770,
      "task_loss": 0.7105112075805664
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3618265986442566,
      "epoch": 8.23,
      "learning_rate": 1.9520638349892158e-05,
      "loss": 0.3374,
      "step": 22780,
      "task_loss": 0.48824140429496765
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3943551182746887,
      "epoch": 8.24,
      "learning_rate": 1.9449397178336928e-05,
      "loss": 0.2946,
      "step": 22790,
      "task_loss": 0.662896990776062
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23531006276607513,
      "epoch": 8.24,
      "learning_rate": 1.9378236190919002e-05,
      "loss": 0.3176,
      "step": 22800,
      "task_loss": 0.5393174886703491
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28416210412979126,
      "epoch": 8.24,
      "learning_rate": 1.9307155754568368e-05,
      "loss": 0.2945,
      "step": 22810,
      "task_loss": 0.33785438537597656
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4986143410205841,
      "epoch": 8.25,
      "learning_rate": 1.9236156235799624e-05,
      "loss": 0.3286,
      "step": 22820,
      "task_loss": 0.6909259557723999
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36012768745422363,
      "epoch": 8.25,
      "learning_rate": 1.9165238000710218e-05,
      "loss": 0.3426,
      "step": 22830,
      "task_loss": 0.4857126474380493
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3498384952545166,
      "epoch": 8.25,
      "learning_rate": 1.9094401414978365e-05,
      "loss": 0.3115,
      "step": 22840,
      "task_loss": 0.6187173128128052
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32221004366874695,
      "epoch": 8.26,
      "learning_rate": 1.9023646843861352e-05,
      "loss": 0.313,
      "step": 22850,
      "task_loss": 0.7463681101799011
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3636491298675537,
      "epoch": 8.26,
      "learning_rate": 1.8952974652193525e-05,
      "loss": 0.2902,
      "step": 22860,
      "task_loss": 0.7195571660995483
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3010321855545044,
      "epoch": 8.27,
      "learning_rate": 1.888238520438449e-05,
      "loss": 0.3284,
      "step": 22870,
      "task_loss": 0.5021539330482483
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47702494263648987,
      "epoch": 8.27,
      "learning_rate": 1.8811878864417145e-05,
      "loss": 0.3776,
      "step": 22880,
      "task_loss": 0.7389390468597412
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3341755270957947,
      "epoch": 8.27,
      "learning_rate": 1.8741455995845906e-05,
      "loss": 0.2775,
      "step": 22890,
      "task_loss": 0.7437736988067627
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2533724904060364,
      "epoch": 8.28,
      "learning_rate": 1.8671116961794767e-05,
      "loss": 0.2914,
      "step": 22900,
      "task_loss": 0.6004878282546997
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3350604176521301,
      "epoch": 8.28,
      "learning_rate": 1.860086212495545e-05,
      "loss": 0.2999,
      "step": 22910,
      "task_loss": 0.5949532389640808
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25673776865005493,
      "epoch": 8.28,
      "learning_rate": 1.8530691847585468e-05,
      "loss": 0.308,
      "step": 22920,
      "task_loss": 0.4904293119907379
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2922332286834717,
      "epoch": 8.29,
      "learning_rate": 1.8460606491506437e-05,
      "loss": 0.3331,
      "step": 22930,
      "task_loss": 0.5136717557907104
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33091992139816284,
      "epoch": 8.29,
      "learning_rate": 1.8390606418101974e-05,
      "loss": 0.3354,
      "step": 22940,
      "task_loss": 0.9739881753921509
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36726266145706177,
      "epoch": 8.29,
      "learning_rate": 1.8320691988316008e-05,
      "loss": 0.3084,
      "step": 22950,
      "task_loss": 0.8299188017845154
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24538114666938782,
      "epoch": 8.3,
      "learning_rate": 1.8250863562650848e-05,
      "loss": 0.3162,
      "step": 22960,
      "task_loss": 0.416057288646698
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3002382516860962,
      "epoch": 8.3,
      "learning_rate": 1.8181121501165358e-05,
      "loss": 0.3206,
      "step": 22970,
      "task_loss": 0.4835646152496338
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2719339430332184,
      "epoch": 8.31,
      "learning_rate": 1.811146616347302e-05,
      "loss": 0.3197,
      "step": 22980,
      "task_loss": 0.5774111747741699
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22802424430847168,
      "epoch": 8.31,
      "learning_rate": 1.8041897908740205e-05,
      "loss": 0.2812,
      "step": 22990,
      "task_loss": 0.3885321021080017
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30409711599349976,
      "epoch": 8.31,
      "learning_rate": 1.7972417095684214e-05,
      "loss": 0.3038,
      "step": 23000,
      "task_loss": 0.4505375623703003
    },
    {
      "epoch": 8.31,
      "eval_exact_match": 83.25449385052035,
      "eval_f1": 90.06136052123999,
      "step": 23000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4253186583518982,
      "epoch": 8.32,
      "learning_rate": 1.790302408257151e-05,
      "loss": 0.3461,
      "step": 23010,
      "task_loss": 0.9150567650794983
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3353206515312195,
      "epoch": 8.32,
      "learning_rate": 1.7833719227215755e-05,
      "loss": 0.3282,
      "step": 23020,
      "task_loss": 0.7890245914459229
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3734143376350403,
      "epoch": 8.32,
      "learning_rate": 1.7764502886976142e-05,
      "loss": 0.3267,
      "step": 23030,
      "task_loss": 0.5730995535850525
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2806342542171478,
      "epoch": 8.33,
      "learning_rate": 1.769537541875536e-05,
      "loss": 0.3222,
      "step": 23040,
      "task_loss": 0.44223088026046753
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30025961995124817,
      "epoch": 8.33,
      "learning_rate": 1.7626337178997885e-05,
      "loss": 0.3062,
      "step": 23050,
      "task_loss": 0.6115245819091797
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33962395787239075,
      "epoch": 8.33,
      "learning_rate": 1.755738852368811e-05,
      "loss": 0.2956,
      "step": 23060,
      "task_loss": 0.7975297570228577
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22749751806259155,
      "epoch": 8.34,
      "learning_rate": 1.748852980834849e-05,
      "loss": 0.3171,
      "step": 23070,
      "task_loss": 0.3744935393333435
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2707931101322174,
      "epoch": 8.34,
      "learning_rate": 1.7419761388037698e-05,
      "loss": 0.295,
      "step": 23080,
      "task_loss": 0.7108138799667358
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25612032413482666,
      "epoch": 8.34,
      "learning_rate": 1.7351083617348838e-05,
      "loss": 0.3024,
      "step": 23090,
      "task_loss": 0.3751325309276581
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23818457126617432,
      "epoch": 8.35,
      "learning_rate": 1.7282496850407593e-05,
      "loss": 0.3057,
      "step": 23100,
      "task_loss": 0.8201920986175537
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27771884202957153,
      "epoch": 8.35,
      "learning_rate": 1.721400144087041e-05,
      "loss": 0.2883,
      "step": 23110,
      "task_loss": 0.4348314106464386
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3101360499858856,
      "epoch": 8.36,
      "learning_rate": 1.714559774192262e-05,
      "loss": 0.3363,
      "step": 23120,
      "task_loss": 0.5495152473449707
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29964226484298706,
      "epoch": 8.36,
      "learning_rate": 1.707728610627674e-05,
      "loss": 0.3349,
      "step": 23130,
      "task_loss": 0.6978568434715271
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28877317905426025,
      "epoch": 8.36,
      "learning_rate": 1.7009066886170497e-05,
      "loss": 0.3116,
      "step": 23140,
      "task_loss": 0.5868189334869385
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2869138717651367,
      "epoch": 8.37,
      "learning_rate": 1.6940940433365148e-05,
      "loss": 0.2882,
      "step": 23150,
      "task_loss": 0.4035617709159851
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30516546964645386,
      "epoch": 8.37,
      "learning_rate": 1.6872907099143585e-05,
      "loss": 0.3137,
      "step": 23160,
      "task_loss": 0.6182640194892883
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3285907208919525,
      "epoch": 8.37,
      "learning_rate": 1.6804967234308577e-05,
      "loss": 0.3,
      "step": 23170,
      "task_loss": 0.7113856673240662
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19711598753929138,
      "epoch": 8.38,
      "learning_rate": 1.673712118918088e-05,
      "loss": 0.3228,
      "step": 23180,
      "task_loss": 0.34729012846946716
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2672497630119324,
      "epoch": 8.38,
      "learning_rate": 1.6669369313597535e-05,
      "loss": 0.2929,
      "step": 23190,
      "task_loss": 0.33527523279190063
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30760812759399414,
      "epoch": 8.38,
      "learning_rate": 1.6601711956909998e-05,
      "loss": 0.3359,
      "step": 23200,
      "task_loss": 0.5162175893783569
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20776550471782684,
      "epoch": 8.39,
      "learning_rate": 1.653414946798235e-05,
      "loss": 0.3143,
      "step": 23210,
      "task_loss": 0.2421736717224121
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33642876148223877,
      "epoch": 8.39,
      "learning_rate": 1.646668219518948e-05,
      "loss": 0.3164,
      "step": 23220,
      "task_loss": 0.6240791082382202
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.273520827293396,
      "epoch": 8.4,
      "learning_rate": 1.639931048641538e-05,
      "loss": 0.2976,
      "step": 23230,
      "task_loss": 0.45852556824684143
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2595791220664978,
      "epoch": 8.4,
      "learning_rate": 1.6332034689051194e-05,
      "loss": 0.3234,
      "step": 23240,
      "task_loss": 0.24186216294765472
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32000231742858887,
      "epoch": 8.4,
      "learning_rate": 1.6264855149993574e-05,
      "loss": 0.3122,
      "step": 23250,
      "task_loss": 0.7240666747093201
    },
    {
      "epoch": 8.4,
      "eval_exact_match": 83.13150425733207,
      "eval_f1": 89.97983813033478,
      "step": 23250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38170820474624634,
      "epoch": 8.41,
      "learning_rate": 1.6197772215642807e-05,
      "loss": 0.319,
      "step": 23260,
      "task_loss": 0.6090672612190247
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.261313259601593,
      "epoch": 8.41,
      "learning_rate": 1.6130786231901073e-05,
      "loss": 0.3146,
      "step": 23270,
      "task_loss": 0.8908061981201172
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3164830803871155,
      "epoch": 8.41,
      "learning_rate": 1.606389754417061e-05,
      "loss": 0.3362,
      "step": 23280,
      "task_loss": 0.6108741760253906
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22788286209106445,
      "epoch": 8.42,
      "learning_rate": 1.5997106497351993e-05,
      "loss": 0.3181,
      "step": 23290,
      "task_loss": 0.3512367904186249
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29590117931365967,
      "epoch": 8.42,
      "learning_rate": 1.593041343584232e-05,
      "loss": 0.31,
      "step": 23300,
      "task_loss": 0.6037249565124512
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39731332659721375,
      "epoch": 8.42,
      "learning_rate": 1.5863818703533445e-05,
      "loss": 0.3173,
      "step": 23310,
      "task_loss": 0.5619301199913025
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37759941816329956,
      "epoch": 8.43,
      "learning_rate": 1.5797322643810196e-05,
      "loss": 0.3294,
      "step": 23320,
      "task_loss": 0.7838083505630493
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.211870938539505,
      "epoch": 8.43,
      "learning_rate": 1.5730925599548637e-05,
      "loss": 0.3146,
      "step": 23330,
      "task_loss": 0.15924398601055145
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2876982092857361,
      "epoch": 8.44,
      "learning_rate": 1.5664627913114222e-05,
      "loss": 0.3385,
      "step": 23340,
      "task_loss": 0.3038162887096405
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31377512216567993,
      "epoch": 8.44,
      "learning_rate": 1.5598429926360136e-05,
      "loss": 0.321,
      "step": 23350,
      "task_loss": 0.36857476830482483
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20846720039844513,
      "epoch": 8.44,
      "learning_rate": 1.5532331980625454e-05,
      "loss": 0.2739,
      "step": 23360,
      "task_loss": 0.42006412148475647
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.257539302110672,
      "epoch": 8.45,
      "learning_rate": 1.5466334416733425e-05,
      "loss": 0.297,
      "step": 23370,
      "task_loss": 0.5687024593353271
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3588574528694153,
      "epoch": 8.45,
      "learning_rate": 1.5400437574989648e-05,
      "loss": 0.3067,
      "step": 23380,
      "task_loss": 0.44749951362609863
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32865849137306213,
      "epoch": 8.45,
      "learning_rate": 1.5334641795180442e-05,
      "loss": 0.3075,
      "step": 23390,
      "task_loss": 0.6262214779853821
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2602709233760834,
      "epoch": 8.46,
      "learning_rate": 1.5268947416570933e-05,
      "loss": 0.3546,
      "step": 23400,
      "task_loss": 0.5079340934753418
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.15402564406394958,
      "epoch": 8.46,
      "learning_rate": 1.5203354777903448e-05,
      "loss": 0.291,
      "step": 23410,
      "task_loss": 0.20646686851978302
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2539033889770508,
      "epoch": 8.46,
      "learning_rate": 1.5137864217395681e-05,
      "loss": 0.3228,
      "step": 23420,
      "task_loss": 0.4042036831378937
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27788808941841125,
      "epoch": 8.47,
      "learning_rate": 1.5072476072739005e-05,
      "loss": 0.2676,
      "step": 23430,
      "task_loss": 0.4695894420146942
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3503713011741638,
      "epoch": 8.47,
      "learning_rate": 1.5007190681096652e-05,
      "loss": 0.3337,
      "step": 23440,
      "task_loss": 0.9511378407478333
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3843029737472534,
      "epoch": 8.47,
      "learning_rate": 1.4942008379102068e-05,
      "loss": 0.3123,
      "step": 23450,
      "task_loss": 0.4682992994785309
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34484705328941345,
      "epoch": 8.48,
      "learning_rate": 1.4876929502857115e-05,
      "loss": 0.3202,
      "step": 23460,
      "task_loss": 0.6297692060470581
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3099765181541443,
      "epoch": 8.48,
      "learning_rate": 1.4811954387930384e-05,
      "loss": 0.3222,
      "step": 23470,
      "task_loss": 0.5310537219047546
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30124685168266296,
      "epoch": 8.49,
      "learning_rate": 1.474708336935538e-05,
      "loss": 0.3038,
      "step": 23480,
      "task_loss": 0.5357911586761475
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23196589946746826,
      "epoch": 8.49,
      "learning_rate": 1.4682316781628928e-05,
      "loss": 0.3089,
      "step": 23490,
      "task_loss": 0.565646231174469
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24483175575733185,
      "epoch": 8.49,
      "learning_rate": 1.4617654958709308e-05,
      "loss": 0.3079,
      "step": 23500,
      "task_loss": 0.6693049669265747
    },
    {
      "epoch": 8.49,
      "eval_exact_match": 83.12204351939451,
      "eval_f1": 89.96558516117871,
      "step": 23500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24072255194187164,
      "epoch": 8.5,
      "learning_rate": 1.4553098234014623e-05,
      "loss": 0.3041,
      "step": 23510,
      "task_loss": 0.2527986466884613
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24449646472930908,
      "epoch": 8.5,
      "learning_rate": 1.4488646940421064e-05,
      "loss": 0.3331,
      "step": 23520,
      "task_loss": 0.4520113468170166
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3243837356567383,
      "epoch": 8.5,
      "learning_rate": 1.4424301410261182e-05,
      "loss": 0.306,
      "step": 23530,
      "task_loss": 0.5300151705741882
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32603827118873596,
      "epoch": 8.51,
      "learning_rate": 1.4360061975322142e-05,
      "loss": 0.2902,
      "step": 23540,
      "task_loss": 0.27277934551239014
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3202453851699829,
      "epoch": 8.51,
      "learning_rate": 1.4295928966844085e-05,
      "loss": 0.3045,
      "step": 23550,
      "task_loss": 0.5317361354827881
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24685320258140564,
      "epoch": 8.51,
      "learning_rate": 1.423190271551837e-05,
      "loss": 0.3274,
      "step": 23560,
      "task_loss": 0.22007986903190613
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35863107442855835,
      "epoch": 8.52,
      "learning_rate": 1.4167983551485887e-05,
      "loss": 0.339,
      "step": 23570,
      "task_loss": 0.37125977873802185
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3166862428188324,
      "epoch": 8.52,
      "learning_rate": 1.4104171804335311e-05,
      "loss": 0.3295,
      "step": 23580,
      "task_loss": 0.4220460057258606
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.330414354801178,
      "epoch": 8.53,
      "learning_rate": 1.404046780310151e-05,
      "loss": 0.3021,
      "step": 23590,
      "task_loss": 0.3584241271018982
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2974003553390503,
      "epoch": 8.53,
      "learning_rate": 1.39768718762637e-05,
      "loss": 0.3255,
      "step": 23600,
      "task_loss": 0.7118889093399048
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22618696093559265,
      "epoch": 8.53,
      "learning_rate": 1.391338435174388e-05,
      "loss": 0.3214,
      "step": 23610,
      "task_loss": 0.2164674550294876
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2954291105270386,
      "epoch": 8.54,
      "learning_rate": 1.3850005556905072e-05,
      "loss": 0.3052,
      "step": 23620,
      "task_loss": 0.5197908282279968
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2527273893356323,
      "epoch": 8.54,
      "learning_rate": 1.3786735818549667e-05,
      "loss": 0.3171,
      "step": 23630,
      "task_loss": 0.652328372001648
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39712145924568176,
      "epoch": 8.54,
      "learning_rate": 1.372357546291769e-05,
      "loss": 0.3434,
      "step": 23640,
      "task_loss": 1.187708854675293
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47445833683013916,
      "epoch": 8.55,
      "learning_rate": 1.3660524815685188e-05,
      "loss": 0.3131,
      "step": 23650,
      "task_loss": 0.677756130695343
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2589554190635681,
      "epoch": 8.55,
      "learning_rate": 1.359758420196249e-05,
      "loss": 0.3015,
      "step": 23660,
      "task_loss": 0.3724703788757324
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30475327372550964,
      "epoch": 8.55,
      "learning_rate": 1.3534753946292586e-05,
      "loss": 0.3147,
      "step": 23670,
      "task_loss": 0.464124470949173
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26423659920692444,
      "epoch": 8.56,
      "learning_rate": 1.347203437264936e-05,
      "loss": 0.319,
      "step": 23680,
      "task_loss": 0.6123193502426147
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2449067085981369,
      "epoch": 8.56,
      "learning_rate": 1.3409425804436078e-05,
      "loss": 0.2941,
      "step": 23690,
      "task_loss": 0.26298606395721436
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3464253842830658,
      "epoch": 8.57,
      "learning_rate": 1.3346928564483535e-05,
      "loss": 0.3161,
      "step": 23700,
      "task_loss": 0.5386533737182617
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2706737816333771,
      "epoch": 8.57,
      "learning_rate": 1.3284542975048519e-05,
      "loss": 0.3168,
      "step": 23710,
      "task_loss": 0.3305785655975342
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38617292046546936,
      "epoch": 8.57,
      "learning_rate": 1.3222269357812115e-05,
      "loss": 0.3332,
      "step": 23720,
      "task_loss": 0.7377265691757202
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2689724564552307,
      "epoch": 8.58,
      "learning_rate": 1.3160108033878046e-05,
      "loss": 0.3306,
      "step": 23730,
      "task_loss": 0.4797666668891907
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.314256489276886,
      "epoch": 8.58,
      "learning_rate": 1.3098059323770975e-05,
      "loss": 0.3069,
      "step": 23740,
      "task_loss": 0.6989715695381165
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28853997588157654,
      "epoch": 8.58,
      "learning_rate": 1.3036123547434927e-05,
      "loss": 0.3023,
      "step": 23750,
      "task_loss": 0.3831864297389984
    },
    {
      "epoch": 8.58,
      "eval_exact_match": 83.23557237464522,
      "eval_f1": 90.0526820682477,
      "step": 23750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2678840756416321,
      "epoch": 8.59,
      "learning_rate": 1.2974301024231595e-05,
      "loss": 0.2999,
      "step": 23760,
      "task_loss": 0.6748077869415283
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4937059283256531,
      "epoch": 8.59,
      "learning_rate": 1.2912592072938709e-05,
      "loss": 0.307,
      "step": 23770,
      "task_loss": 0.5111285448074341
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26543956995010376,
      "epoch": 8.59,
      "learning_rate": 1.2850997011748333e-05,
      "loss": 0.3042,
      "step": 23780,
      "task_loss": 0.5223565101623535
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2755920886993408,
      "epoch": 8.6,
      "learning_rate": 1.2789516158265369e-05,
      "loss": 0.2859,
      "step": 23790,
      "task_loss": 0.7235616445541382
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2730027437210083,
      "epoch": 8.6,
      "learning_rate": 1.272814982950573e-05,
      "loss": 0.2873,
      "step": 23800,
      "task_loss": 0.4539136588573456
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3317798972129822,
      "epoch": 8.6,
      "learning_rate": 1.2666898341894853e-05,
      "loss": 0.3073,
      "step": 23810,
      "task_loss": 0.7397780418395996
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3097071945667267,
      "epoch": 8.61,
      "learning_rate": 1.2605762011266012e-05,
      "loss": 0.3264,
      "step": 23820,
      "task_loss": 0.5371835231781006
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35412904620170593,
      "epoch": 8.61,
      "learning_rate": 1.2544741152858692e-05,
      "loss": 0.33,
      "step": 23830,
      "task_loss": 0.9191930294036865
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38428613543510437,
      "epoch": 8.62,
      "learning_rate": 1.248383608131694e-05,
      "loss": 0.3249,
      "step": 23840,
      "task_loss": 0.49637895822525024
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3155321478843689,
      "epoch": 8.62,
      "learning_rate": 1.2423047110687794e-05,
      "loss": 0.2899,
      "step": 23850,
      "task_loss": 0.28766974806785583
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3072446882724762,
      "epoch": 8.62,
      "learning_rate": 1.2362374554419625e-05,
      "loss": 0.2689,
      "step": 23860,
      "task_loss": 0.4855462908744812
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27834153175354004,
      "epoch": 8.63,
      "learning_rate": 1.2301818725360548e-05,
      "loss": 0.2806,
      "step": 23870,
      "task_loss": 0.6370406746864319
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3155345916748047,
      "epoch": 8.63,
      "learning_rate": 1.2241379935756749e-05,
      "loss": 0.3611,
      "step": 23880,
      "task_loss": 0.5455687046051025
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25810787081718445,
      "epoch": 8.63,
      "learning_rate": 1.2181058497250998e-05,
      "loss": 0.3393,
      "step": 23890,
      "task_loss": 0.3866387605667114
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2348720282316208,
      "epoch": 8.64,
      "learning_rate": 1.2120854720880877e-05,
      "loss": 0.2784,
      "step": 23900,
      "task_loss": 0.5155615210533142
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2619689106941223,
      "epoch": 8.64,
      "learning_rate": 1.206076891707731e-05,
      "loss": 0.2989,
      "step": 23910,
      "task_loss": 0.4420399069786072
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27206704020500183,
      "epoch": 8.64,
      "learning_rate": 1.2000801395662918e-05,
      "loss": 0.342,
      "step": 23920,
      "task_loss": 0.17631283402442932
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2513728141784668,
      "epoch": 8.65,
      "learning_rate": 1.1940952465850405e-05,
      "loss": 0.291,
      "step": 23930,
      "task_loss": 0.3972375988960266
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3371734023094177,
      "epoch": 8.65,
      "learning_rate": 1.1881222436240966e-05,
      "loss": 0.3306,
      "step": 23940,
      "task_loss": 0.607552170753479
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43905913829803467,
      "epoch": 8.66,
      "learning_rate": 1.1821611614822724e-05,
      "loss": 0.3306,
      "step": 23950,
      "task_loss": 0.5458712577819824
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23583155870437622,
      "epoch": 8.66,
      "learning_rate": 1.176212030896912e-05,
      "loss": 0.3205,
      "step": 23960,
      "task_loss": 0.6100504398345947
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46240583062171936,
      "epoch": 8.66,
      "learning_rate": 1.1702748825437348e-05,
      "loss": 0.3571,
      "step": 23970,
      "task_loss": 0.3039376735687256
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3308100998401642,
      "epoch": 8.67,
      "learning_rate": 1.16434974703667e-05,
      "loss": 0.3045,
      "step": 23980,
      "task_loss": 0.6687793731689453
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20676454901695251,
      "epoch": 8.67,
      "learning_rate": 1.1584366549277138e-05,
      "loss": 0.3027,
      "step": 23990,
      "task_loss": 0.1708330363035202
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3004925847053528,
      "epoch": 8.67,
      "learning_rate": 1.1525356367067541e-05,
      "loss": 0.3253,
      "step": 24000,
      "task_loss": 0.639447033405304
    },
    {
      "epoch": 8.67,
      "eval_exact_match": 83.1693472090823,
      "eval_f1": 90.00836238528001,
      "step": 24000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3217807412147522,
      "epoch": 8.68,
      "learning_rate": 1.1466467228014262e-05,
      "loss": 0.2779,
      "step": 24010,
      "task_loss": 0.3356829285621643
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3157300055027008,
      "epoch": 8.68,
      "learning_rate": 1.14076994357695e-05,
      "loss": 0.2987,
      "step": 24020,
      "task_loss": 0.46271902322769165
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33147913217544556,
      "epoch": 8.68,
      "learning_rate": 1.134905329335976e-05,
      "loss": 0.3338,
      "step": 24030,
      "task_loss": 0.4586483836174011
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2759827971458435,
      "epoch": 8.69,
      "learning_rate": 1.1290529103184282e-05,
      "loss": 0.2986,
      "step": 24040,
      "task_loss": 0.5049731731414795
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27000918984413147,
      "epoch": 8.69,
      "learning_rate": 1.1232127167013436e-05,
      "loss": 0.3024,
      "step": 24050,
      "task_loss": 0.619734525680542
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28138256072998047,
      "epoch": 8.7,
      "learning_rate": 1.1173847785987288e-05,
      "loss": 0.2856,
      "step": 24060,
      "task_loss": 0.572986900806427
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31417810916900635,
      "epoch": 8.7,
      "learning_rate": 1.1115691260613889e-05,
      "loss": 0.3172,
      "step": 24070,
      "task_loss": 0.5405772924423218
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24283945560455322,
      "epoch": 8.7,
      "learning_rate": 1.1057657890767854e-05,
      "loss": 0.287,
      "step": 24080,
      "task_loss": 0.6355692744255066
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2398536056280136,
      "epoch": 8.71,
      "learning_rate": 1.0999747975688754e-05,
      "loss": 0.2763,
      "step": 24090,
      "task_loss": 0.46709010004997253
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30363500118255615,
      "epoch": 8.71,
      "learning_rate": 1.094196181397959e-05,
      "loss": 0.327,
      "step": 24100,
      "task_loss": 0.44533640146255493
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2811916470527649,
      "epoch": 8.71,
      "learning_rate": 1.0884299703605228e-05,
      "loss": 0.292,
      "step": 24110,
      "task_loss": 0.5359020233154297
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3132500648498535,
      "epoch": 8.72,
      "learning_rate": 1.0826761941890906e-05,
      "loss": 0.276,
      "step": 24120,
      "task_loss": 0.5441325902938843
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3040134906768799,
      "epoch": 8.72,
      "learning_rate": 1.0769348825520669e-05,
      "loss": 0.329,
      "step": 24130,
      "task_loss": 0.2830539345741272
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3059019446372986,
      "epoch": 8.72,
      "learning_rate": 1.0712060650535867e-05,
      "loss": 0.2936,
      "step": 24140,
      "task_loss": 0.47051605582237244
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3694162666797638,
      "epoch": 8.73,
      "learning_rate": 1.0654897712333557e-05,
      "loss": 0.3272,
      "step": 24150,
      "task_loss": 0.48830804228782654
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28213226795196533,
      "epoch": 8.73,
      "learning_rate": 1.059786030566512e-05,
      "loss": 0.2719,
      "step": 24160,
      "task_loss": 0.2837570905685425
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2804946303367615,
      "epoch": 8.74,
      "learning_rate": 1.0540948724634564e-05,
      "loss": 0.3319,
      "step": 24170,
      "task_loss": 0.25393009185791016
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36412930488586426,
      "epoch": 8.74,
      "learning_rate": 1.0484163262697156e-05,
      "loss": 0.3162,
      "step": 24180,
      "task_loss": 0.8942012190818787
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25897452235221863,
      "epoch": 8.74,
      "learning_rate": 1.042750421265783e-05,
      "loss": 0.2955,
      "step": 24190,
      "task_loss": 0.5445088148117065
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2680971324443817,
      "epoch": 8.75,
      "learning_rate": 1.0370971866669712e-05,
      "loss": 0.2892,
      "step": 24200,
      "task_loss": 0.4629053473472595
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2193024605512619,
      "epoch": 8.75,
      "learning_rate": 1.0314566516232566e-05,
      "loss": 0.2714,
      "step": 24210,
      "task_loss": 0.27407723665237427
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22114166617393494,
      "epoch": 8.75,
      "learning_rate": 1.025828845219136e-05,
      "loss": 0.3139,
      "step": 24220,
      "task_loss": 0.6401352882385254
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30140265822410583,
      "epoch": 8.76,
      "learning_rate": 1.0202137964734709e-05,
      "loss": 0.296,
      "step": 24230,
      "task_loss": 0.4010269343852997
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29186466336250305,
      "epoch": 8.76,
      "learning_rate": 1.0146115343393427e-05,
      "loss": 0.3189,
      "step": 24240,
      "task_loss": 0.40033257007598877
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2807198762893677,
      "epoch": 8.76,
      "learning_rate": 1.0090220877038949e-05,
      "loss": 0.3235,
      "step": 24250,
      "task_loss": 0.6275461316108704
    },
    {
      "epoch": 8.76,
      "eval_exact_match": 83.37748344370861,
      "eval_f1": 90.13053426752283,
      "step": 24250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24098724126815796,
      "epoch": 8.77,
      "learning_rate": 1.0034454853881984e-05,
      "loss": 0.2883,
      "step": 24260,
      "task_loss": 0.41819271445274353
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2524189352989197,
      "epoch": 8.77,
      "learning_rate": 9.978817561470873e-06,
      "loss": 0.2971,
      "step": 24270,
      "task_loss": 0.790593147277832
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2279432862997055,
      "epoch": 8.77,
      "learning_rate": 9.923309286690207e-06,
      "loss": 0.2953,
      "step": 24280,
      "task_loss": 0.2802739441394806
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32530757784843445,
      "epoch": 8.78,
      "learning_rate": 9.86793031575932e-06,
      "loss": 0.3165,
      "step": 24290,
      "task_loss": 0.5283281207084656
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30216020345687866,
      "epoch": 8.78,
      "learning_rate": 9.812680934230822e-06,
      "loss": 0.3341,
      "step": 24300,
      "task_loss": 0.9180983304977417
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24517136812210083,
      "epoch": 8.79,
      "learning_rate": 9.757561426989076e-06,
      "loss": 0.3064,
      "step": 24310,
      "task_loss": 0.44735443592071533
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27164602279663086,
      "epoch": 8.79,
      "learning_rate": 9.702572078248811e-06,
      "loss": 0.2865,
      "step": 24320,
      "task_loss": 0.6187409162521362
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30350732803344727,
      "epoch": 8.79,
      "learning_rate": 9.647713171553596e-06,
      "loss": 0.3185,
      "step": 24330,
      "task_loss": 0.5968219637870789
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25562575459480286,
      "epoch": 8.8,
      "learning_rate": 9.59298498977442e-06,
      "loss": 0.284,
      "step": 24340,
      "task_loss": 0.49374791979789734
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25753918290138245,
      "epoch": 8.8,
      "learning_rate": 9.53838781510815e-06,
      "loss": 0.3393,
      "step": 24350,
      "task_loss": 0.4400380253791809
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3407808542251587,
      "epoch": 8.8,
      "learning_rate": 9.483921929076223e-06,
      "loss": 0.2922,
      "step": 24360,
      "task_loss": 0.7328530550003052
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32406896352767944,
      "epoch": 8.81,
      "learning_rate": 9.429587612523036e-06,
      "loss": 0.3012,
      "step": 24370,
      "task_loss": 0.496076762676239
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3123983144760132,
      "epoch": 8.81,
      "learning_rate": 9.3753851456146e-06,
      "loss": 0.3108,
      "step": 24380,
      "task_loss": 0.6676946878433228
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31041258573532104,
      "epoch": 8.81,
      "learning_rate": 9.321314807837065e-06,
      "loss": 0.2882,
      "step": 24390,
      "task_loss": 0.697533369064331
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3565478026866913,
      "epoch": 8.82,
      "learning_rate": 9.267376877995281e-06,
      "loss": 0.3524,
      "step": 24400,
      "task_loss": 0.5232845544815063
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29742079973220825,
      "epoch": 8.82,
      "learning_rate": 9.213571634211337e-06,
      "loss": 0.3312,
      "step": 24410,
      "task_loss": 0.7807422876358032
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33375900983810425,
      "epoch": 8.83,
      "learning_rate": 9.159899353923163e-06,
      "loss": 0.3086,
      "step": 24420,
      "task_loss": 0.9461276531219482
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.300223708152771,
      "epoch": 8.83,
      "learning_rate": 9.106360313883099e-06,
      "loss": 0.3211,
      "step": 24430,
      "task_loss": 0.4010380506515503
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27285024523735046,
      "epoch": 8.83,
      "learning_rate": 9.052954790156437e-06,
      "loss": 0.3262,
      "step": 24440,
      "task_loss": 0.3871135711669922
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2731539011001587,
      "epoch": 8.84,
      "learning_rate": 8.999683058119992e-06,
      "loss": 0.3058,
      "step": 24450,
      "task_loss": 0.4444922208786011
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3398137390613556,
      "epoch": 8.84,
      "learning_rate": 8.946545392460764e-06,
      "loss": 0.3225,
      "step": 24460,
      "task_loss": 0.4049307703971863
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33151674270629883,
      "epoch": 8.84,
      "learning_rate": 8.893542067174394e-06,
      "loss": 0.2843,
      "step": 24470,
      "task_loss": 0.3135819733142853
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26082944869995117,
      "epoch": 8.85,
      "learning_rate": 8.84067335556387e-06,
      "loss": 0.3005,
      "step": 24480,
      "task_loss": 0.6256157755851746
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3421831429004669,
      "epoch": 8.85,
      "learning_rate": 8.787939530238035e-06,
      "loss": 0.319,
      "step": 24490,
      "task_loss": 0.47051504254341125
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44341349601745605,
      "epoch": 8.85,
      "learning_rate": 8.735340863110246e-06,
      "loss": 0.3367,
      "step": 24500,
      "task_loss": 0.28469258546829224
    },
    {
      "epoch": 8.85,
      "eval_exact_match": 83.4247871333964,
      "eval_f1": 90.13270154934425,
      "step": 24500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2990521788597107,
      "epoch": 8.86,
      "learning_rate": 8.68287762539689e-06,
      "loss": 0.3001,
      "step": 24510,
      "task_loss": 0.48367151618003845
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3240818977355957,
      "epoch": 8.86,
      "learning_rate": 8.630550087616077e-06,
      "loss": 0.2951,
      "step": 24520,
      "task_loss": 0.6060233116149902
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.346133828163147,
      "epoch": 8.87,
      "learning_rate": 8.578358519586184e-06,
      "loss": 0.306,
      "step": 24530,
      "task_loss": 0.7388773560523987
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.198516845703125,
      "epoch": 8.87,
      "learning_rate": 8.526303190424489e-06,
      "loss": 0.3196,
      "step": 24540,
      "task_loss": 0.34029442071914673
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21414974331855774,
      "epoch": 8.87,
      "learning_rate": 8.474384368545771e-06,
      "loss": 0.3066,
      "step": 24550,
      "task_loss": 0.6532096862792969
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3364071547985077,
      "epoch": 8.88,
      "learning_rate": 8.422602321660946e-06,
      "loss": 0.2868,
      "step": 24560,
      "task_loss": 0.7407635450363159
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24205178022384644,
      "epoch": 8.88,
      "learning_rate": 8.37095731677564e-06,
      "loss": 0.2937,
      "step": 24570,
      "task_loss": 0.36204320192337036
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32036536931991577,
      "epoch": 8.88,
      "learning_rate": 8.319449620188875e-06,
      "loss": 0.3009,
      "step": 24580,
      "task_loss": 0.5532072186470032
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3431103229522705,
      "epoch": 8.89,
      "learning_rate": 8.268079497491656e-06,
      "loss": 0.2938,
      "step": 24590,
      "task_loss": 0.47085040807724
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26091110706329346,
      "epoch": 8.89,
      "learning_rate": 8.216847213565618e-06,
      "loss": 0.2954,
      "step": 24600,
      "task_loss": 0.6554282903671265
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.284633994102478,
      "epoch": 8.89,
      "learning_rate": 8.165753032581614e-06,
      "loss": 0.3042,
      "step": 24610,
      "task_loss": 0.5411564111709595
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23999324440956116,
      "epoch": 8.9,
      "learning_rate": 8.114797217998467e-06,
      "loss": 0.3104,
      "step": 24620,
      "task_loss": 0.3989071846008301
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38341277837753296,
      "epoch": 8.9,
      "learning_rate": 8.063980032561454e-06,
      "loss": 0.3384,
      "step": 24630,
      "task_loss": 0.959318995475769
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21845731139183044,
      "epoch": 8.9,
      "learning_rate": 8.013301738301086e-06,
      "loss": 0.2954,
      "step": 24640,
      "task_loss": 0.33747684955596924
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3176514506340027,
      "epoch": 8.91,
      "learning_rate": 7.96276259653169e-06,
      "loss": 0.3091,
      "step": 24650,
      "task_loss": 0.33057931065559387
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2908194065093994,
      "epoch": 8.91,
      "learning_rate": 7.912362867850083e-06,
      "loss": 0.2927,
      "step": 24660,
      "task_loss": 0.41401976346969604
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3124035596847534,
      "epoch": 8.92,
      "learning_rate": 7.8621028121342e-06,
      "loss": 0.3021,
      "step": 24670,
      "task_loss": 0.8065996170043945
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21487663686275482,
      "epoch": 8.92,
      "learning_rate": 7.811982688541792e-06,
      "loss": 0.3237,
      "step": 24680,
      "task_loss": 0.5148287415504456
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3609391748905182,
      "epoch": 8.92,
      "learning_rate": 7.76200275550907e-06,
      "loss": 0.3003,
      "step": 24690,
      "task_loss": 0.5247024297714233
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26552635431289673,
      "epoch": 8.93,
      "learning_rate": 7.71216327074939e-06,
      "loss": 0.301,
      "step": 24700,
      "task_loss": 0.49575549364089966
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2935408055782318,
      "epoch": 8.93,
      "learning_rate": 7.662464491251865e-06,
      "loss": 0.2868,
      "step": 24710,
      "task_loss": 0.6105607748031616
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25080007314682007,
      "epoch": 8.93,
      "learning_rate": 7.612906673280149e-06,
      "loss": 0.3025,
      "step": 24720,
      "task_loss": 0.2942613363265991
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2175687700510025,
      "epoch": 8.94,
      "learning_rate": 7.563490072370988e-06,
      "loss": 0.2828,
      "step": 24730,
      "task_loss": 0.4805278778076172
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3156837224960327,
      "epoch": 8.94,
      "learning_rate": 7.5142149433330064e-06,
      "loss": 0.275,
      "step": 24740,
      "task_loss": 0.5296868085861206
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24178892374038696,
      "epoch": 8.94,
      "learning_rate": 7.46508154024534e-06,
      "loss": 0.2847,
      "step": 24750,
      "task_loss": 0.38275861740112305
    },
    {
      "epoch": 8.94,
      "eval_exact_match": 83.40586565752129,
      "eval_f1": 90.19943208724408,
      "step": 24750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23588743805885315,
      "epoch": 8.95,
      "learning_rate": 7.41609011645634e-06,
      "loss": 0.3342,
      "step": 24760,
      "task_loss": 0.388775110244751
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22134511172771454,
      "epoch": 8.95,
      "learning_rate": 7.367240924582245e-06,
      "loss": 0.2849,
      "step": 24770,
      "task_loss": 0.3853006362915039
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2299889624118805,
      "epoch": 8.96,
      "learning_rate": 7.3185342165059285e-06,
      "loss": 0.2984,
      "step": 24780,
      "task_loss": 0.22366122901439667
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5092062950134277,
      "epoch": 8.96,
      "learning_rate": 7.269970243375551e-06,
      "loss": 0.2925,
      "step": 24790,
      "task_loss": 0.64637291431427
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3082968294620514,
      "epoch": 8.96,
      "learning_rate": 7.221549255603302e-06,
      "loss": 0.3162,
      "step": 24800,
      "task_loss": 0.43850719928741455
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30956393480300903,
      "epoch": 8.97,
      "learning_rate": 7.1732715028640445e-06,
      "loss": 0.3237,
      "step": 24810,
      "task_loss": 0.5990317463874817
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24433210492134094,
      "epoch": 8.97,
      "learning_rate": 7.125137234094145e-06,
      "loss": 0.3109,
      "step": 24820,
      "task_loss": 0.46226024627685547
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3623145818710327,
      "epoch": 8.97,
      "learning_rate": 7.077146697490052e-06,
      "loss": 0.3295,
      "step": 24830,
      "task_loss": 0.4708067774772644
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2837136387825012,
      "epoch": 8.98,
      "learning_rate": 7.02930014050712e-06,
      "loss": 0.3078,
      "step": 24840,
      "task_loss": 0.6401948928833008
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2711445093154907,
      "epoch": 8.98,
      "learning_rate": 6.98159780985828e-06,
      "loss": 0.3127,
      "step": 24850,
      "task_loss": 0.38128766417503357
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32651424407958984,
      "epoch": 8.98,
      "learning_rate": 6.934039951512807e-06,
      "loss": 0.3118,
      "step": 24860,
      "task_loss": 0.5000793933868408
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3723224401473999,
      "epoch": 8.99,
      "learning_rate": 6.886626810694988e-06,
      "loss": 0.3275,
      "step": 24870,
      "task_loss": 0.5765509009361267
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.570521891117096,
      "epoch": 8.99,
      "learning_rate": 6.839358631882928e-06,
      "loss": 0.3048,
      "step": 24880,
      "task_loss": 0.7745019197463989
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35303252935409546,
      "epoch": 9.0,
      "learning_rate": 6.792235658807258e-06,
      "loss": 0.2959,
      "step": 24890,
      "task_loss": 0.6391245126724243
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3144342005252838,
      "epoch": 9.0,
      "learning_rate": 6.745258134449877e-06,
      "loss": 0.3054,
      "step": 24900,
      "task_loss": 0.3757985830307007
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33729472756385803,
      "epoch": 9.0,
      "learning_rate": 6.698426301042671e-06,
      "loss": 0.3118,
      "step": 24910,
      "task_loss": 0.5654165744781494
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2673114538192749,
      "epoch": 9.01,
      "learning_rate": 6.651740400066356e-06,
      "loss": 0.2841,
      "step": 24920,
      "task_loss": 0.7930485606193542
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3094874918460846,
      "epoch": 9.01,
      "learning_rate": 6.605200672249114e-06,
      "loss": 0.3033,
      "step": 24930,
      "task_loss": 0.5905535817146301
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2626067101955414,
      "epoch": 9.01,
      "learning_rate": 6.558807357565434e-06,
      "loss": 0.2958,
      "step": 24940,
      "task_loss": 0.5565435886383057
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20787197351455688,
      "epoch": 9.02,
      "learning_rate": 6.512560695234854e-06,
      "loss": 0.2749,
      "step": 24950,
      "task_loss": 0.6173804402351379
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4326663315296173,
      "epoch": 9.02,
      "learning_rate": 6.4664609237207205e-06,
      "loss": 0.2847,
      "step": 24960,
      "task_loss": 0.4918290674686432
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3277257978916168,
      "epoch": 9.02,
      "learning_rate": 6.42050828072894e-06,
      "loss": 0.2905,
      "step": 24970,
      "task_loss": 0.5493934154510498
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26245617866516113,
      "epoch": 9.03,
      "learning_rate": 6.3747030032068e-06,
      "loss": 0.2992,
      "step": 24980,
      "task_loss": 0.6874579787254333
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30544644594192505,
      "epoch": 9.03,
      "learning_rate": 6.32904532734172e-06,
      "loss": 0.2894,
      "step": 24990,
      "task_loss": 0.7642111778259277
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38944071531295776,
      "epoch": 9.04,
      "learning_rate": 6.283535488560031e-06,
      "loss": 0.2805,
      "step": 25000,
      "task_loss": 0.9234870672225952
    },
    {
      "epoch": 9.04,
      "eval_exact_match": 83.20719016083254,
      "eval_f1": 90.12546160298142,
      "step": 25000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29679083824157715,
      "epoch": 9.04,
      "learning_rate": 6.2381737215257446e-06,
      "loss": 0.2977,
      "step": 25010,
      "task_loss": 0.505827784538269
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34737223386764526,
      "epoch": 9.04,
      "learning_rate": 6.192960260139417e-06,
      "loss": 0.2884,
      "step": 25020,
      "task_loss": 0.4295722246170044
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3087272644042969,
      "epoch": 9.05,
      "learning_rate": 6.147895337536835e-06,
      "loss": 0.3117,
      "step": 25030,
      "task_loss": 0.35688644647598267
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25100329518318176,
      "epoch": 9.05,
      "learning_rate": 6.102979186087896e-06,
      "loss": 0.2811,
      "step": 25040,
      "task_loss": 0.38825923204421997
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26149922609329224,
      "epoch": 9.05,
      "learning_rate": 6.058212037395389e-06,
      "loss": 0.3192,
      "step": 25050,
      "task_loss": 0.3461078107357025
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25997209548950195,
      "epoch": 9.06,
      "learning_rate": 6.013594122293796e-06,
      "loss": 0.2923,
      "step": 25060,
      "task_loss": 0.4676539897918701
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3166268765926361,
      "epoch": 9.06,
      "learning_rate": 5.969125670848064e-06,
      "loss": 0.2811,
      "step": 25070,
      "task_loss": 0.2946656048297882
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22155921161174774,
      "epoch": 9.06,
      "learning_rate": 5.924806912352495e-06,
      "loss": 0.2831,
      "step": 25080,
      "task_loss": 0.49417394399642944
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2537713348865509,
      "epoch": 9.07,
      "learning_rate": 5.880638075329512e-06,
      "loss": 0.2797,
      "step": 25090,
      "task_loss": 0.48220908641815186
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20896217226982117,
      "epoch": 9.07,
      "learning_rate": 5.836619387528503e-06,
      "loss": 0.2584,
      "step": 25100,
      "task_loss": 0.2613734304904938
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22877725958824158,
      "epoch": 9.07,
      "learning_rate": 5.792751075924592e-06,
      "loss": 0.2966,
      "step": 25110,
      "task_loss": 0.4536246359348297
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35032469034194946,
      "epoch": 9.08,
      "learning_rate": 5.7490333667175805e-06,
      "loss": 0.2964,
      "step": 25120,
      "task_loss": 0.9190288782119751
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2878870964050293,
      "epoch": 9.08,
      "learning_rate": 5.705466485330657e-06,
      "loss": 0.2988,
      "step": 25130,
      "task_loss": 0.7237658500671387
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28640902042388916,
      "epoch": 9.09,
      "learning_rate": 5.662050656409317e-06,
      "loss": 0.282,
      "step": 25140,
      "task_loss": 0.24912115931510925
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24825415015220642,
      "epoch": 9.09,
      "learning_rate": 5.618786103820175e-06,
      "loss": 0.2743,
      "step": 25150,
      "task_loss": 0.7251200079917908
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3168322443962097,
      "epoch": 9.09,
      "learning_rate": 5.575673050649818e-06,
      "loss": 0.283,
      "step": 25160,
      "task_loss": 0.9995952844619751
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24792785942554474,
      "epoch": 9.1,
      "learning_rate": 5.532711719203632e-06,
      "loss": 0.3235,
      "step": 25170,
      "task_loss": 0.4345579147338867
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.269896924495697,
      "epoch": 9.1,
      "learning_rate": 5.494176426081241e-06,
      "loss": 0.3099,
      "step": 25180,
      "task_loss": 0.5822728872299194
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3608018457889557,
      "epoch": 9.1,
      "learning_rate": 5.45150397555954e-06,
      "loss": 0.2885,
      "step": 25190,
      "task_loss": 0.6332502365112305
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27320295572280884,
      "epoch": 9.11,
      "learning_rate": 5.4089838870195325e-06,
      "loss": 0.2992,
      "step": 25200,
      "task_loss": 0.3529553711414337
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22571894526481628,
      "epoch": 9.11,
      "learning_rate": 5.366616379709066e-06,
      "loss": 0.2927,
      "step": 25210,
      "task_loss": 0.5368704795837402
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27815741300582886,
      "epoch": 9.11,
      "learning_rate": 5.324401672089289e-06,
      "loss": 0.3008,
      "step": 25220,
      "task_loss": 0.5218074321746826
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2736639380455017,
      "epoch": 9.12,
      "learning_rate": 5.282339981833406e-06,
      "loss": 0.2941,
      "step": 25230,
      "task_loss": 0.5212745070457458
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30182701349258423,
      "epoch": 9.12,
      "learning_rate": 5.240431525825641e-06,
      "loss": 0.303,
      "step": 25240,
      "task_loss": 0.42352715134620667
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28600355982780457,
      "epoch": 9.13,
      "learning_rate": 5.198676520160087e-06,
      "loss": 0.2664,
      "step": 25250,
      "task_loss": 0.4605422914028168
    },
    {
      "epoch": 9.13,
      "eval_exact_match": 83.32071901608326,
      "eval_f1": 90.1759308609037,
      "step": 25250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22900843620300293,
      "epoch": 9.13,
      "learning_rate": 5.157075180139602e-06,
      "loss": 0.286,
      "step": 25260,
      "task_loss": 0.6918904781341553
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23345553874969482,
      "epoch": 9.13,
      "learning_rate": 5.1156277202746694e-06,
      "loss": 0.2836,
      "step": 25270,
      "task_loss": 0.3756367862224579
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29328298568725586,
      "epoch": 9.14,
      "learning_rate": 5.074334354282335e-06,
      "loss": 0.2912,
      "step": 25280,
      "task_loss": 0.5100090503692627
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26017504930496216,
      "epoch": 9.14,
      "learning_rate": 5.033195295085081e-06,
      "loss": 0.2888,
      "step": 25290,
      "task_loss": 0.4601510465145111
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3635603189468384,
      "epoch": 9.14,
      "learning_rate": 4.992210754809733e-06,
      "loss": 0.2877,
      "step": 25300,
      "task_loss": 0.943723738193512
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2453012764453888,
      "epoch": 9.15,
      "learning_rate": 4.951380944786361e-06,
      "loss": 0.2797,
      "step": 25310,
      "task_loss": 0.5946983695030212
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31875067949295044,
      "epoch": 9.15,
      "learning_rate": 4.910706075547198e-06,
      "loss": 0.2872,
      "step": 25320,
      "task_loss": 0.6306192874908447
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24196487665176392,
      "epoch": 9.15,
      "learning_rate": 4.8701863568255366e-06,
      "loss": 0.2564,
      "step": 25330,
      "task_loss": 0.8180104494094849
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2831239700317383,
      "epoch": 9.16,
      "learning_rate": 4.829821997554683e-06,
      "loss": 0.2927,
      "step": 25340,
      "task_loss": 0.5410017967224121
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3321933448314667,
      "epoch": 9.16,
      "learning_rate": 4.789613205866839e-06,
      "loss": 0.2908,
      "step": 25350,
      "task_loss": 0.8938983678817749
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2835544943809509,
      "epoch": 9.17,
      "learning_rate": 4.749560189092066e-06,
      "loss": 0.2755,
      "step": 25360,
      "task_loss": 0.45705628395080566
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2626034617424011,
      "epoch": 9.17,
      "learning_rate": 4.709663153757165e-06,
      "loss": 0.2861,
      "step": 25370,
      "task_loss": 0.3554117977619171
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24724841117858887,
      "epoch": 9.17,
      "learning_rate": 4.669922305584701e-06,
      "loss": 0.2859,
      "step": 25380,
      "task_loss": 0.2859405279159546
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2531455457210541,
      "epoch": 9.18,
      "learning_rate": 4.630337849491818e-06,
      "loss": 0.2832,
      "step": 25390,
      "task_loss": 0.3663128614425659
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34095948934555054,
      "epoch": 9.18,
      "learning_rate": 4.5909099895892995e-06,
      "loss": 0.2919,
      "step": 25400,
      "task_loss": 0.6796036958694458
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2302282750606537,
      "epoch": 9.18,
      "learning_rate": 4.551638929180444e-06,
      "loss": 0.2915,
      "step": 25410,
      "task_loss": 0.39458438754081726
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.220722034573555,
      "epoch": 9.19,
      "learning_rate": 4.512524870760054e-06,
      "loss": 0.2622,
      "step": 25420,
      "task_loss": 0.15429294109344482
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3225393295288086,
      "epoch": 9.19,
      "learning_rate": 4.473568016013349e-06,
      "loss": 0.2863,
      "step": 25430,
      "task_loss": 0.43470489978790283
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18860642611980438,
      "epoch": 9.19,
      "learning_rate": 4.4347685658149885e-06,
      "loss": 0.2572,
      "step": 25440,
      "task_loss": 0.4267352819442749
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34222182631492615,
      "epoch": 9.2,
      "learning_rate": 4.396126720227975e-06,
      "loss": 0.3,
      "step": 25450,
      "task_loss": 0.569180428981781
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28997889161109924,
      "epoch": 9.2,
      "learning_rate": 4.357642678502669e-06,
      "loss": 0.2775,
      "step": 25460,
      "task_loss": 0.6120558977127075
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2590046525001526,
      "epoch": 9.2,
      "learning_rate": 4.319316639075711e-06,
      "loss": 0.2804,
      "step": 25470,
      "task_loss": 0.45842182636260986
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.282612144947052,
      "epoch": 9.21,
      "learning_rate": 4.281148799569073e-06,
      "loss": 0.2986,
      "step": 25480,
      "task_loss": 0.5861254930496216
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19793371856212616,
      "epoch": 9.21,
      "learning_rate": 4.243139356788951e-06,
      "loss": 0.2715,
      "step": 25490,
      "task_loss": 0.5529229640960693
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25628167390823364,
      "epoch": 9.22,
      "learning_rate": 4.205288506724823e-06,
      "loss": 0.2758,
      "step": 25500,
      "task_loss": 0.30487939715385437
    },
    {
      "epoch": 9.22,
      "eval_exact_match": 83.37748344370861,
      "eval_f1": 90.1936862438257,
      "step": 25500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22346541285514832,
      "epoch": 9.22,
      "learning_rate": 4.167596444548396e-06,
      "loss": 0.2788,
      "step": 25510,
      "task_loss": 0.3134952187538147
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3261803686618805,
      "epoch": 9.22,
      "learning_rate": 4.130063364612621e-06,
      "loss": 0.3254,
      "step": 25520,
      "task_loss": 0.47811222076416016
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22242692112922668,
      "epoch": 9.23,
      "learning_rate": 4.0926894604506725e-06,
      "loss": 0.3134,
      "step": 25530,
      "task_loss": 0.42973819375038147
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2093992978334427,
      "epoch": 9.23,
      "learning_rate": 4.055474924774975e-06,
      "loss": 0.2835,
      "step": 25540,
      "task_loss": 0.3554569482803345
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.223178893327713,
      "epoch": 9.23,
      "learning_rate": 4.01841994947618e-06,
      "loss": 0.2766,
      "step": 25550,
      "task_loss": 0.5469294786453247
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26629751920700073,
      "epoch": 9.24,
      "learning_rate": 3.981524725622215e-06,
      "loss": 0.2989,
      "step": 25560,
      "task_loss": 0.5456836223602295
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4269055426120758,
      "epoch": 9.24,
      "learning_rate": 3.944789443457237e-06,
      "loss": 0.3092,
      "step": 25570,
      "task_loss": 0.5398896336555481
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2995172142982483,
      "epoch": 9.24,
      "learning_rate": 3.908214292400743e-06,
      "loss": 0.291,
      "step": 25580,
      "task_loss": 0.6619876623153687
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23310492932796478,
      "epoch": 9.25,
      "learning_rate": 3.871799461046491e-06,
      "loss": 0.2687,
      "step": 25590,
      "task_loss": 0.3361056447029114
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2728113532066345,
      "epoch": 9.25,
      "learning_rate": 3.8355451371616e-06,
      "loss": 0.2857,
      "step": 25600,
      "task_loss": 0.4070819616317749
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21583805978298187,
      "epoch": 9.26,
      "learning_rate": 3.799451507685557e-06,
      "loss": 0.2933,
      "step": 25610,
      "task_loss": 0.3874829411506653
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3354247808456421,
      "epoch": 9.26,
      "learning_rate": 3.7635187587292616e-06,
      "loss": 0.2835,
      "step": 25620,
      "task_loss": 0.5514976978302002
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3161737322807312,
      "epoch": 9.26,
      "learning_rate": 3.727747075574033e-06,
      "loss": 0.2812,
      "step": 25630,
      "task_loss": 0.7555432319641113
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23402747511863708,
      "epoch": 9.27,
      "learning_rate": 3.6921366426707064e-06,
      "loss": 0.2882,
      "step": 25640,
      "task_loss": 0.5421460270881653
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28667300939559937,
      "epoch": 9.27,
      "learning_rate": 3.656687643638646e-06,
      "loss": 0.2625,
      "step": 25650,
      "task_loss": 0.6737666130065918
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3086755871772766,
      "epoch": 9.27,
      "learning_rate": 3.6214002612648187e-06,
      "loss": 0.2956,
      "step": 25660,
      "task_loss": 0.5526766777038574
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24900871515274048,
      "epoch": 9.28,
      "learning_rate": 3.5862746775028274e-06,
      "loss": 0.269,
      "step": 25670,
      "task_loss": 0.7501120567321777
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25524961948394775,
      "epoch": 9.28,
      "learning_rate": 3.5513110734719947e-06,
      "loss": 0.2781,
      "step": 25680,
      "task_loss": 0.4351167678833008
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24399134516716003,
      "epoch": 9.28,
      "learning_rate": 3.5165096294564294e-06,
      "loss": 0.2996,
      "step": 25690,
      "task_loss": 0.3778623342514038
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26403456926345825,
      "epoch": 9.29,
      "learning_rate": 3.481870524904063e-06,
      "loss": 0.2898,
      "step": 25700,
      "task_loss": 0.4827141761779785
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31666100025177,
      "epoch": 9.29,
      "learning_rate": 3.4473939384257713e-06,
      "loss": 0.2671,
      "step": 25710,
      "task_loss": 0.6182833909988403
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2807222604751587,
      "epoch": 9.3,
      "learning_rate": 3.4130800477944353e-06,
      "loss": 0.2924,
      "step": 25720,
      "task_loss": 0.3987352252006531
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3156668543815613,
      "epoch": 9.3,
      "learning_rate": 3.378929029944014e-06,
      "loss": 0.2552,
      "step": 25730,
      "task_loss": 0.339354932308197
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2318783700466156,
      "epoch": 9.3,
      "learning_rate": 3.3449410609686222e-06,
      "loss": 0.2721,
      "step": 25740,
      "task_loss": 0.5828352570533752
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33963948488235474,
      "epoch": 9.31,
      "learning_rate": 3.311116316121681e-06,
      "loss": 0.2969,
      "step": 25750,
      "task_loss": 0.6039541363716125
    },
    {
      "epoch": 9.31,
      "eval_exact_match": 83.45316934720908,
      "eval_f1": 90.26675509384134,
      "step": 25750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2563812732696533,
      "epoch": 9.31,
      "learning_rate": 3.27745496981493e-06,
      "loss": 0.3062,
      "step": 25760,
      "task_loss": 0.2901385724544525
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3281647562980652,
      "epoch": 9.31,
      "learning_rate": 3.2439571956175955e-06,
      "loss": 0.2789,
      "step": 25770,
      "task_loss": 0.7631877660751343
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21847324073314667,
      "epoch": 9.32,
      "learning_rate": 3.2106231662554687e-06,
      "loss": 0.2985,
      "step": 25780,
      "task_loss": 0.2929229140281677
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2040231078863144,
      "epoch": 9.32,
      "learning_rate": 3.17745305361002e-06,
      "loss": 0.263,
      "step": 25790,
      "task_loss": 0.2177676260471344
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26135972142219543,
      "epoch": 9.32,
      "learning_rate": 3.144447028717492e-06,
      "loss": 0.2908,
      "step": 25800,
      "task_loss": 0.380513995885849
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.313922643661499,
      "epoch": 9.33,
      "learning_rate": 3.1116052617680475e-06,
      "loss": 0.312,
      "step": 25810,
      "task_loss": 0.41387584805488586
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3513936400413513,
      "epoch": 9.33,
      "learning_rate": 3.078927922104888e-06,
      "loss": 0.2748,
      "step": 25820,
      "task_loss": 0.607092022895813
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2572995126247406,
      "epoch": 9.34,
      "learning_rate": 3.04641517822337e-06,
      "loss": 0.2937,
      "step": 25830,
      "task_loss": 0.3874950110912323
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.336165189743042,
      "epoch": 9.34,
      "learning_rate": 3.0140671977701126e-06,
      "loss": 0.2795,
      "step": 25840,
      "task_loss": 0.7848570346832275
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2801818549633026,
      "epoch": 9.34,
      "learning_rate": 2.9818841475422074e-06,
      "loss": 0.2711,
      "step": 25850,
      "task_loss": 0.5263223648071289
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2951273024082184,
      "epoch": 9.35,
      "learning_rate": 2.9498661934862625e-06,
      "loss": 0.2934,
      "step": 25860,
      "task_loss": 0.39920687675476074
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3156875967979431,
      "epoch": 9.35,
      "learning_rate": 2.9180135006976253e-06,
      "loss": 0.2879,
      "step": 25870,
      "task_loss": 0.4999083876609802
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.325722873210907,
      "epoch": 9.35,
      "learning_rate": 2.8863262334194892e-06,
      "loss": 0.2734,
      "step": 25880,
      "task_loss": 0.49259310960769653
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3251590132713318,
      "epoch": 9.36,
      "learning_rate": 2.854804555042066e-06,
      "loss": 0.291,
      "step": 25890,
      "task_loss": 0.6317105293273926
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2367677092552185,
      "epoch": 9.36,
      "learning_rate": 2.823448628101714e-06,
      "loss": 0.268,
      "step": 25900,
      "task_loss": 0.24415592849254608
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.381255567073822,
      "epoch": 9.36,
      "learning_rate": 2.792258614280147e-06,
      "loss": 0.2871,
      "step": 25910,
      "task_loss": 0.8889896869659424
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26195400953292847,
      "epoch": 9.37,
      "learning_rate": 2.761234674403565e-06,
      "loss": 0.2644,
      "step": 25920,
      "task_loss": 0.5699973702430725
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.307630717754364,
      "epoch": 9.37,
      "learning_rate": 2.730376968441837e-06,
      "loss": 0.2978,
      "step": 25930,
      "task_loss": 0.7901893258094788
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24636954069137573,
      "epoch": 9.37,
      "learning_rate": 2.699685655507652e-06,
      "loss": 0.2725,
      "step": 25940,
      "task_loss": 0.4935848116874695
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27287501096725464,
      "epoch": 9.38,
      "learning_rate": 2.6691608938557673e-06,
      "loss": 0.2744,
      "step": 25950,
      "task_loss": 0.5872029066085815
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24528878927230835,
      "epoch": 9.38,
      "learning_rate": 2.638802840882099e-06,
      "loss": 0.2937,
      "step": 25960,
      "task_loss": 0.5012997388839722
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2681114971637726,
      "epoch": 9.39,
      "learning_rate": 2.608611653122982e-06,
      "loss": 0.2798,
      "step": 25970,
      "task_loss": 0.5551661849021912
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25442802906036377,
      "epoch": 9.39,
      "learning_rate": 2.5785874862543364e-06,
      "loss": 0.2823,
      "step": 25980,
      "task_loss": 0.28142935037612915
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3172297477722168,
      "epoch": 9.39,
      "learning_rate": 2.548730495090867e-06,
      "loss": 0.273,
      "step": 25990,
      "task_loss": 0.5310195088386536
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23420450091362,
      "epoch": 9.4,
      "learning_rate": 2.5190408335852423e-06,
      "loss": 0.2762,
      "step": 26000,
      "task_loss": 0.34213316440582275
    },
    {
      "epoch": 9.4,
      "eval_exact_match": 83.3112582781457,
      "eval_f1": 90.12102540358967,
      "step": 26000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23456361889839172,
      "epoch": 9.4,
      "learning_rate": 2.4895186548273553e-06,
      "loss": 0.2888,
      "step": 26010,
      "task_loss": 0.6904394626617432
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24155323207378387,
      "epoch": 9.4,
      "learning_rate": 2.4601641110434842e-06,
      "loss": 0.272,
      "step": 26020,
      "task_loss": 0.5414040088653564
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25628653168678284,
      "epoch": 9.41,
      "learning_rate": 2.430977353595531e-06,
      "loss": 0.2877,
      "step": 26030,
      "task_loss": 0.47668391466140747
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26848334074020386,
      "epoch": 9.41,
      "learning_rate": 2.401958532980213e-06,
      "loss": 0.264,
      "step": 26040,
      "task_loss": 0.5960483551025391
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3336072564125061,
      "epoch": 9.41,
      "learning_rate": 2.3731077988283558e-06,
      "loss": 0.305,
      "step": 26050,
      "task_loss": 0.7050433158874512
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39169514179229736,
      "epoch": 9.42,
      "learning_rate": 2.34442529990403e-06,
      "loss": 0.269,
      "step": 26060,
      "task_loss": 1.0993900299072266
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25674140453338623,
      "epoch": 9.42,
      "learning_rate": 2.315911184103842e-06,
      "loss": 0.2943,
      "step": 26070,
      "task_loss": 0.6436038017272949
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26511308550834656,
      "epoch": 9.43,
      "learning_rate": 2.287565598456169e-06,
      "loss": 0.2789,
      "step": 26080,
      "task_loss": 0.13820317387580872
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24257254600524902,
      "epoch": 9.43,
      "learning_rate": 2.259388689120386e-06,
      "loss": 0.2852,
      "step": 26090,
      "task_loss": 0.5551714897155762
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25758957862854004,
      "epoch": 9.43,
      "learning_rate": 2.2313806013861007e-06,
      "loss": 0.281,
      "step": 26100,
      "task_loss": 0.46998971700668335
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28022924065589905,
      "epoch": 9.44,
      "learning_rate": 2.2035414796724365e-06,
      "loss": 0.2816,
      "step": 26110,
      "task_loss": 0.5369085073471069
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3347325325012207,
      "epoch": 9.44,
      "learning_rate": 2.1758714675272663e-06,
      "loss": 0.2728,
      "step": 26120,
      "task_loss": 0.9577845931053162
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37050122022628784,
      "epoch": 9.44,
      "learning_rate": 2.1483707076264827e-06,
      "loss": 0.2957,
      "step": 26130,
      "task_loss": 0.3568909168243408
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27206090092658997,
      "epoch": 9.45,
      "learning_rate": 2.121039341773235e-06,
      "loss": 0.2914,
      "step": 26140,
      "task_loss": 0.5149966478347778
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2517910599708557,
      "epoch": 9.45,
      "learning_rate": 2.0938775108972628e-06,
      "loss": 0.2741,
      "step": 26150,
      "task_loss": 0.27969563007354736
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1978009045124054,
      "epoch": 9.45,
      "learning_rate": 2.0668853550540775e-06,
      "loss": 0.2689,
      "step": 26160,
      "task_loss": 0.6869149208068848
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35537177324295044,
      "epoch": 9.46,
      "learning_rate": 2.0400630134243162e-06,
      "loss": 0.2731,
      "step": 26170,
      "task_loss": 0.6466042399406433
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32208681106567383,
      "epoch": 9.46,
      "learning_rate": 2.013410624312997e-06,
      "loss": 0.3048,
      "step": 26180,
      "task_loss": 0.41794082522392273
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32221776247024536,
      "epoch": 9.47,
      "learning_rate": 1.9869283251488022e-06,
      "loss": 0.285,
      "step": 26190,
      "task_loss": 0.5621099472045898
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28082454204559326,
      "epoch": 9.47,
      "learning_rate": 1.9606162524833615e-06,
      "loss": 0.2701,
      "step": 26200,
      "task_loss": 0.492424339056015
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.293172687292099,
      "epoch": 9.47,
      "learning_rate": 1.9344745419905726e-06,
      "loss": 0.2903,
      "step": 26210,
      "task_loss": 0.8020915985107422
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3128081262111664,
      "epoch": 9.48,
      "learning_rate": 1.9085033284658917e-06,
      "loss": 0.3234,
      "step": 26220,
      "task_loss": 0.9452996253967285
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22008506953716278,
      "epoch": 9.48,
      "learning_rate": 1.8827027458256239e-06,
      "loss": 0.2711,
      "step": 26230,
      "task_loss": 0.38822752237319946
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2757604420185089,
      "epoch": 9.48,
      "learning_rate": 1.857072927106244e-06,
      "loss": 0.2846,
      "step": 26240,
      "task_loss": 0.554851233959198
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33274781703948975,
      "epoch": 9.49,
      "learning_rate": 1.831614004463733e-06,
      "loss": 0.301,
      "step": 26250,
      "task_loss": 0.6326380968093872
    },
    {
      "epoch": 9.49,
      "eval_exact_match": 83.4247871333964,
      "eval_f1": 90.22614199574525,
      "step": 26250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.210701122879982,
      "epoch": 9.49,
      "learning_rate": 1.806326109172838e-06,
      "loss": 0.278,
      "step": 26260,
      "task_loss": 0.2917744815349579
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38611477613449097,
      "epoch": 9.49,
      "learning_rate": 1.781209371626461e-06,
      "loss": 0.3125,
      "step": 26270,
      "task_loss": 0.5859373807907104
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2683292031288147,
      "epoch": 9.5,
      "learning_rate": 1.75626392133494e-06,
      "loss": 0.2873,
      "step": 26280,
      "task_loss": 0.4396800994873047
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3382301926612854,
      "epoch": 9.5,
      "learning_rate": 1.7314898869254157e-06,
      "loss": 0.2832,
      "step": 26290,
      "task_loss": 0.4854893684387207
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20767495036125183,
      "epoch": 9.5,
      "learning_rate": 1.7068873961411294e-06,
      "loss": 0.2781,
      "step": 26300,
      "task_loss": 0.23974767327308655
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26005885004997253,
      "epoch": 9.51,
      "learning_rate": 1.6824565758408027e-06,
      "loss": 0.286,
      "step": 26310,
      "task_loss": 0.3799484968185425
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2298750877380371,
      "epoch": 9.51,
      "learning_rate": 1.6581975519979562e-06,
      "loss": 0.2722,
      "step": 26320,
      "task_loss": 0.4086710810661316
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2837666869163513,
      "epoch": 9.52,
      "learning_rate": 1.6341104497002816e-06,
      "loss": 0.2824,
      "step": 26330,
      "task_loss": 0.403903067111969
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19592724740505219,
      "epoch": 9.52,
      "learning_rate": 1.6101953931489643e-06,
      "loss": 0.2891,
      "step": 26340,
      "task_loss": 0.30998939275741577
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23085609078407288,
      "epoch": 9.52,
      "learning_rate": 1.5864525056580967e-06,
      "loss": 0.3152,
      "step": 26350,
      "task_loss": 0.6044930219650269
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3257802724838257,
      "epoch": 9.53,
      "learning_rate": 1.5628819096539764e-06,
      "loss": 0.3009,
      "step": 26360,
      "task_loss": 0.5781923532485962
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22241540253162384,
      "epoch": 9.53,
      "learning_rate": 1.5394837266745246e-06,
      "loss": 0.2875,
      "step": 26370,
      "task_loss": 0.3078111410140991
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28470808267593384,
      "epoch": 9.53,
      "learning_rate": 1.5162580773686438e-06,
      "loss": 0.3048,
      "step": 26380,
      "task_loss": 0.3852382302284241
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2786301076412201,
      "epoch": 9.54,
      "learning_rate": 1.4932050814955942e-06,
      "loss": 0.2919,
      "step": 26390,
      "task_loss": 0.42377230525016785
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20077863335609436,
      "epoch": 9.54,
      "learning_rate": 1.470324857924365e-06,
      "loss": 0.2532,
      "step": 26400,
      "task_loss": 0.4952772259712219
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27380576729774475,
      "epoch": 9.54,
      "learning_rate": 1.4476175246330906e-06,
      "loss": 0.269,
      "step": 26410,
      "task_loss": 0.581963300704956
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28494900465011597,
      "epoch": 9.55,
      "learning_rate": 1.4250831987084155e-06,
      "loss": 0.2935,
      "step": 26420,
      "task_loss": 0.604192852973938
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23649415373802185,
      "epoch": 9.55,
      "learning_rate": 1.4027219963449087e-06,
      "loss": 0.2837,
      "step": 26430,
      "task_loss": 0.4841282367706299
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.284720242023468,
      "epoch": 9.56,
      "learning_rate": 1.3805340328444287e-06,
      "loss": 0.2888,
      "step": 26440,
      "task_loss": 0.30555108189582825
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26956403255462646,
      "epoch": 9.56,
      "learning_rate": 1.358519422615601e-06,
      "loss": 0.2874,
      "step": 26450,
      "task_loss": 0.7349599599838257
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2146541029214859,
      "epoch": 9.56,
      "learning_rate": 1.3366782791731396e-06,
      "loss": 0.2741,
      "step": 26460,
      "task_loss": 0.4127127528190613
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3042886257171631,
      "epoch": 9.57,
      "learning_rate": 1.315010715137333e-06,
      "loss": 0.2722,
      "step": 26470,
      "task_loss": 0.5704532861709595
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2715678811073303,
      "epoch": 9.57,
      "learning_rate": 1.2935168422334307e-06,
      "loss": 0.2912,
      "step": 26480,
      "task_loss": 0.40685567259788513
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23535844683647156,
      "epoch": 9.57,
      "learning_rate": 1.2721967712910648e-06,
      "loss": 0.2938,
      "step": 26490,
      "task_loss": 0.571018636226654
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34679874777793884,
      "epoch": 9.58,
      "learning_rate": 1.2510506122436886e-06,
      "loss": 0.2925,
      "step": 26500,
      "task_loss": 0.7452425956726074
    },
    {
      "epoch": 9.58,
      "eval_exact_match": 83.37748344370861,
      "eval_f1": 90.14017214184005,
      "step": 26500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25748512148857117,
      "epoch": 9.58,
      "learning_rate": 1.2300784741280078e-06,
      "loss": 0.2815,
      "step": 26510,
      "task_loss": 0.6882176399230957
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.314084529876709,
      "epoch": 9.58,
      "learning_rate": 1.209280465083427e-06,
      "loss": 0.2916,
      "step": 26520,
      "task_loss": 0.6169145107269287
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3948596715927124,
      "epoch": 9.59,
      "learning_rate": 1.1886566923514634e-06,
      "loss": 0.2843,
      "step": 26530,
      "task_loss": 0.5022493600845337
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2946526110172272,
      "epoch": 9.59,
      "learning_rate": 1.1682072622752342e-06,
      "loss": 0.3098,
      "step": 26540,
      "task_loss": 0.4513667821884155
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26810207962989807,
      "epoch": 9.6,
      "learning_rate": 1.147932280298877e-06,
      "loss": 0.2904,
      "step": 26550,
      "task_loss": 0.5581741333007812
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3102302551269531,
      "epoch": 9.6,
      "learning_rate": 1.127831850967007e-06,
      "loss": 0.2916,
      "step": 26560,
      "task_loss": 0.4799581468105316
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2536352872848511,
      "epoch": 9.6,
      "learning_rate": 1.1079060779242022e-06,
      "loss": 0.2907,
      "step": 26570,
      "task_loss": 0.4057456851005554
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30852973461151123,
      "epoch": 9.61,
      "learning_rate": 1.0881550639144531e-06,
      "loss": 0.2985,
      "step": 26580,
      "task_loss": 0.35658955574035645
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2384970337152481,
      "epoch": 9.61,
      "learning_rate": 1.0685789107806264e-06,
      "loss": 0.27,
      "step": 26590,
      "task_loss": 0.4211677014827728
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2926676273345947,
      "epoch": 9.61,
      "learning_rate": 1.0491777194639433e-06,
      "loss": 0.2998,
      "step": 26600,
      "task_loss": 0.9406899809837341
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30535075068473816,
      "epoch": 9.62,
      "learning_rate": 1.029951590003487e-06,
      "loss": 0.2847,
      "step": 26610,
      "task_loss": 0.4420982897281647
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25047576427459717,
      "epoch": 9.62,
      "learning_rate": 1.0109006215356314e-06,
      "loss": 0.2606,
      "step": 26620,
      "task_loss": 0.855073094367981
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2909805178642273,
      "epoch": 9.62,
      "learning_rate": 9.920249122935775e-07,
      "loss": 0.293,
      "step": 26630,
      "task_loss": 0.49562349915504456
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2831922769546509,
      "epoch": 9.63,
      "learning_rate": 9.733245596068346e-07,
      "loss": 0.2753,
      "step": 26640,
      "task_loss": 0.673780083656311
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2511312961578369,
      "epoch": 9.63,
      "learning_rate": 9.547996599007135e-07,
      "loss": 0.2762,
      "step": 26650,
      "task_loss": 0.47640103101730347
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3380993604660034,
      "epoch": 9.63,
      "learning_rate": 9.364503086958147e-07,
      "loss": 0.3117,
      "step": 26660,
      "task_loss": 0.5715634822845459
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21443752944469452,
      "epoch": 9.64,
      "learning_rate": 9.182766006075692e-07,
      "loss": 0.3208,
      "step": 26670,
      "task_loss": 0.2922282814979553
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2359142154455185,
      "epoch": 9.64,
      "learning_rate": 9.002786293457259e-07,
      "loss": 0.2814,
      "step": 26680,
      "task_loss": 0.5400431156158447
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22956624627113342,
      "epoch": 9.65,
      "learning_rate": 8.824564877138851e-07,
      "loss": 0.2725,
      "step": 26690,
      "task_loss": 0.6306565999984741
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2651234269142151,
      "epoch": 9.65,
      "learning_rate": 8.648102676089857e-07,
      "loss": 0.2841,
      "step": 26700,
      "task_loss": 0.4753642678260803
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3443555235862732,
      "epoch": 9.65,
      "learning_rate": 8.473400600208857e-07,
      "loss": 0.2948,
      "step": 26710,
      "task_loss": 0.5741276741027832
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2716384530067444,
      "epoch": 9.66,
      "learning_rate": 8.317674383652068e-07,
      "loss": 0.3119,
      "step": 26720,
      "task_loss": 0.25244268774986267
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29673638939857483,
      "epoch": 9.66,
      "learning_rate": 8.146319019851694e-07,
      "loss": 0.2915,
      "step": 26730,
      "task_loss": 0.7287982702255249
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2692734897136688,
      "epoch": 9.66,
      "learning_rate": 7.976726368584502e-07,
      "loss": 0.298,
      "step": 26740,
      "task_loss": 0.32548055052757263
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2941330671310425,
      "epoch": 9.67,
      "learning_rate": 7.808897304327305e-07,
      "loss": 0.3058,
      "step": 26750,
      "task_loss": 0.7823320031166077
    },
    {
      "epoch": 9.67,
      "eval_exact_match": 83.44370860927152,
      "eval_f1": 90.1998081355494,
      "step": 26750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3193681240081787,
      "epoch": 9.67,
      "learning_rate": 7.642832692463141e-07,
      "loss": 0.2897,
      "step": 26760,
      "task_loss": 0.8019513487815857
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22801771759986877,
      "epoch": 9.67,
      "learning_rate": 7.478533389276997e-07,
      "loss": 0.2885,
      "step": 26770,
      "task_loss": 0.4875749945640564
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2520209550857544,
      "epoch": 9.68,
      "learning_rate": 7.316000241951303e-07,
      "loss": 0.2847,
      "step": 26780,
      "task_loss": 0.4516429603099823
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2385767251253128,
      "epoch": 9.68,
      "learning_rate": 7.155234088561735e-07,
      "loss": 0.288,
      "step": 26790,
      "task_loss": 0.7741764187812805
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3399713337421417,
      "epoch": 9.69,
      "learning_rate": 6.996235758072472e-07,
      "loss": 0.2673,
      "step": 26800,
      "task_loss": 0.5370351672172546
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.202695831656456,
      "epoch": 9.69,
      "learning_rate": 6.839006070332509e-07,
      "loss": 0.2775,
      "step": 26810,
      "task_loss": 0.36401206254959106
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.255326509475708,
      "epoch": 9.69,
      "learning_rate": 6.683545836070914e-07,
      "loss": 0.2813,
      "step": 26820,
      "task_loss": 0.7015193700790405
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23634222149848938,
      "epoch": 9.7,
      "learning_rate": 6.529855856892941e-07,
      "loss": 0.2704,
      "step": 26830,
      "task_loss": 0.3981613516807556
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25996315479278564,
      "epoch": 9.7,
      "learning_rate": 6.377936925275834e-07,
      "loss": 0.2703,
      "step": 26840,
      "task_loss": 0.31600648164749146
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.294119656085968,
      "epoch": 9.7,
      "learning_rate": 6.22778982456471e-07,
      "loss": 0.2867,
      "step": 26850,
      "task_loss": 0.4647236764431
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3896469175815582,
      "epoch": 9.71,
      "learning_rate": 6.07941532896844e-07,
      "loss": 0.2964,
      "step": 26860,
      "task_loss": 0.46653464436531067
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24614590406417847,
      "epoch": 9.71,
      "learning_rate": 5.932814203555835e-07,
      "loss": 0.2757,
      "step": 26870,
      "task_loss": 0.26867130398750305
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24889053404331207,
      "epoch": 9.71,
      "learning_rate": 5.787987204251615e-07,
      "loss": 0.2827,
      "step": 26880,
      "task_loss": 1.2295489311218262
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2833463251590729,
      "epoch": 9.72,
      "learning_rate": 5.644935077832475e-07,
      "loss": 0.2791,
      "step": 26890,
      "task_loss": 0.6843339800834656
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28050845861434937,
      "epoch": 9.72,
      "learning_rate": 5.50365856192328e-07,
      "loss": 0.2909,
      "step": 26900,
      "task_loss": 1.0125864744186401
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22105228900909424,
      "epoch": 9.73,
      "learning_rate": 5.3641583849933e-07,
      "loss": 0.2665,
      "step": 26910,
      "task_loss": 0.577481746673584
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29488325119018555,
      "epoch": 9.73,
      "learning_rate": 5.22643526635228e-07,
      "loss": 0.2781,
      "step": 26920,
      "task_loss": 0.3614484965801239
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29972535371780396,
      "epoch": 9.73,
      "learning_rate": 5.090489916146983e-07,
      "loss": 0.2706,
      "step": 26930,
      "task_loss": 0.7378089427947998
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2527899444103241,
      "epoch": 9.74,
      "learning_rate": 4.956323035357346e-07,
      "loss": 0.2986,
      "step": 26940,
      "task_loss": 0.4819590449333191
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2659050226211548,
      "epoch": 9.74,
      "learning_rate": 4.823935315792937e-07,
      "loss": 0.304,
      "step": 26950,
      "task_loss": 0.5571587681770325
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2077215313911438,
      "epoch": 9.74,
      "learning_rate": 4.693327440089273e-07,
      "loss": 0.2783,
      "step": 26960,
      "task_loss": 0.5672307014465332
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35510140657424927,
      "epoch": 9.75,
      "learning_rate": 4.5645000817045476e-07,
      "loss": 0.3091,
      "step": 26970,
      "task_loss": 0.4848492443561554
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39507466554641724,
      "epoch": 9.75,
      "learning_rate": 4.437453904915944e-07,
      "loss": 0.2982,
      "step": 26980,
      "task_loss": 0.914400577545166
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3037624955177307,
      "epoch": 9.75,
      "learning_rate": 4.312189564816293e-07,
      "loss": 0.266,
      "step": 26990,
      "task_loss": 0.8692284822463989
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29725557565689087,
      "epoch": 9.76,
      "learning_rate": 4.1887077073105733e-07,
      "loss": 0.2947,
      "step": 27000,
      "task_loss": 0.7238430976867676
    },
    {
      "epoch": 9.76,
      "eval_exact_match": 83.04635761589404,
      "eval_f1": 89.96929111407182,
      "step": 27000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2712564468383789,
      "epoch": 9.76,
      "learning_rate": 4.067008969112845e-07,
      "loss": 0.2767,
      "step": 27010,
      "task_loss": 0.7396634817123413
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28446534276008606,
      "epoch": 9.77,
      "learning_rate": 3.9470939777426715e-07,
      "loss": 0.273,
      "step": 27020,
      "task_loss": 0.5197400450706482
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2189192771911621,
      "epoch": 9.77,
      "learning_rate": 3.828963351522052e-07,
      "loss": 0.2674,
      "step": 27030,
      "task_loss": 0.5614264607429504
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31505560874938965,
      "epoch": 9.77,
      "learning_rate": 3.712617699572312e-07,
      "loss": 0.2818,
      "step": 27040,
      "task_loss": 0.6665478944778442
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3059966564178467,
      "epoch": 9.78,
      "learning_rate": 3.5980576218107237e-07,
      "loss": 0.3109,
      "step": 27050,
      "task_loss": 0.7996760010719299
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2522026300430298,
      "epoch": 9.78,
      "learning_rate": 3.4852837089475116e-07,
      "loss": 0.2912,
      "step": 27060,
      "task_loss": 0.3180551826953888
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28468310832977295,
      "epoch": 9.78,
      "learning_rate": 3.3742965424829017e-07,
      "loss": 0.2834,
      "step": 27070,
      "task_loss": 0.4297701418399811
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33653613924980164,
      "epoch": 9.79,
      "learning_rate": 3.265096694704128e-07,
      "loss": 0.285,
      "step": 27080,
      "task_loss": 0.666775107383728
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3369113802909851,
      "epoch": 9.79,
      "learning_rate": 3.157684728682247e-07,
      "loss": 0.278,
      "step": 27090,
      "task_loss": 0.5345579385757446
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2958570420742035,
      "epoch": 9.79,
      "learning_rate": 3.0520611982694955e-07,
      "loss": 0.2678,
      "step": 27100,
      "task_loss": 0.4423734247684479
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29316452145576477,
      "epoch": 9.8,
      "learning_rate": 2.9482266480963355e-07,
      "loss": 0.3016,
      "step": 27110,
      "task_loss": 0.5462176203727722
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29860496520996094,
      "epoch": 9.8,
      "learning_rate": 2.846181613568621e-07,
      "loss": 0.2853,
      "step": 27120,
      "task_loss": 0.36607810854911804
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3323076367378235,
      "epoch": 9.8,
      "learning_rate": 2.7459266208648354e-07,
      "loss": 0.2869,
      "step": 27130,
      "task_loss": 0.6542870998382568
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2639560401439667,
      "epoch": 9.81,
      "learning_rate": 2.6474621869334134e-07,
      "loss": 0.2994,
      "step": 27140,
      "task_loss": 0.6767532229423523
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28091293573379517,
      "epoch": 9.81,
      "learning_rate": 2.5507888194901735e-07,
      "loss": 0.2785,
      "step": 27150,
      "task_loss": 0.632643461227417
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2712363004684448,
      "epoch": 9.82,
      "learning_rate": 2.455907017015446e-07,
      "loss": 0.2911,
      "step": 27160,
      "task_loss": 0.640754222869873
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26335710287094116,
      "epoch": 9.82,
      "learning_rate": 2.3628172687516988e-07,
      "loss": 0.2924,
      "step": 27170,
      "task_loss": 0.45733070373535156
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27504533529281616,
      "epoch": 9.82,
      "learning_rate": 2.2715200547010147e-07,
      "loss": 0.2924,
      "step": 27180,
      "task_loss": 0.5883292555809021
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2938588857650757,
      "epoch": 9.83,
      "learning_rate": 2.1820158456225257e-07,
      "loss": 0.3078,
      "step": 27190,
      "task_loss": 0.45251211524009705
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2968611717224121,
      "epoch": 9.83,
      "learning_rate": 2.094305103029964e-07,
      "loss": 0.2776,
      "step": 27200,
      "task_loss": 0.3613067865371704
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18803226947784424,
      "epoch": 9.83,
      "learning_rate": 2.008388279189488e-07,
      "loss": 0.2865,
      "step": 27210,
      "task_loss": 0.317171573638916
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2504810094833374,
      "epoch": 9.84,
      "learning_rate": 1.9242658171171154e-07,
      "loss": 0.2758,
      "step": 27220,
      "task_loss": 0.5608529448509216
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2654745876789093,
      "epoch": 9.84,
      "learning_rate": 1.8419381505765497e-07,
      "loss": 0.2829,
      "step": 27230,
      "task_loss": 0.4092085361480713
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27439069747924805,
      "epoch": 9.84,
      "learning_rate": 1.7614057040769235e-07,
      "loss": 0.2687,
      "step": 27240,
      "task_loss": 0.5781913995742798
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3405267000198364,
      "epoch": 9.85,
      "learning_rate": 1.6826688928706256e-07,
      "loss": 0.3083,
      "step": 27250,
      "task_loss": 0.7811588048934937
    },
    {
      "epoch": 9.85,
      "eval_exact_match": 83.23557237464522,
      "eval_f1": 90.1943930477886,
      "step": 27250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3025619685649872,
      "epoch": 9.85,
      "learning_rate": 1.6057281229510831e-07,
      "loss": 0.3215,
      "step": 27260,
      "task_loss": 0.4894791841506958
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25182005763053894,
      "epoch": 9.86,
      "learning_rate": 1.5305837910507824e-07,
      "loss": 0.2655,
      "step": 27270,
      "task_loss": 0.25380706787109375
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24280399084091187,
      "epoch": 9.86,
      "learning_rate": 1.4572362846392073e-07,
      "loss": 0.2681,
      "step": 27280,
      "task_loss": 0.8004210591316223
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25001049041748047,
      "epoch": 9.86,
      "learning_rate": 1.3856859819207422e-07,
      "loss": 0.2768,
      "step": 27290,
      "task_loss": 0.3485010266304016
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2553676664829254,
      "epoch": 9.87,
      "learning_rate": 1.3159332518328458e-07,
      "loss": 0.2877,
      "step": 27300,
      "task_loss": 0.37561753392219543
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32698652148246765,
      "epoch": 9.87,
      "learning_rate": 1.2479784540439907e-07,
      "loss": 0.2789,
      "step": 27310,
      "task_loss": 0.47436389327049255
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22853092849254608,
      "epoch": 9.87,
      "learning_rate": 1.1818219389520323e-07,
      "loss": 0.2661,
      "step": 27320,
      "task_loss": 0.34670090675354004
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.17037039995193481,
      "epoch": 9.88,
      "learning_rate": 1.1174640476821096e-07,
      "loss": 0.2788,
      "step": 27330,
      "task_loss": 0.2407091110944748
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.257743775844574,
      "epoch": 9.88,
      "learning_rate": 1.054905112085247e-07,
      "loss": 0.2676,
      "step": 27340,
      "task_loss": 0.41856345534324646
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2541770339012146,
      "epoch": 9.88,
      "learning_rate": 9.941454547362948e-08,
      "loss": 0.2946,
      "step": 27350,
      "task_loss": 0.6853160262107849
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25304538011550903,
      "epoch": 9.89,
      "learning_rate": 9.3518538893253e-08,
      "loss": 0.2595,
      "step": 27360,
      "task_loss": 0.4963826537132263
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2942352294921875,
      "epoch": 9.89,
      "learning_rate": 8.780252186918302e-08,
      "loss": 0.2936,
      "step": 27370,
      "task_loss": 0.8194656372070312
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2384113222360611,
      "epoch": 9.9,
      "learning_rate": 8.226652387512745e-08,
      "loss": 0.276,
      "step": 27380,
      "task_loss": 0.5101650953292847
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21360494196414948,
      "epoch": 9.9,
      "learning_rate": 7.691057345655505e-08,
      "loss": 0.2589,
      "step": 27390,
      "task_loss": 0.39360785484313965
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33178994059562683,
      "epoch": 9.9,
      "learning_rate": 7.17346982305439e-08,
      "loss": 0.2801,
      "step": 27400,
      "task_loss": 0.43100279569625854
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20473375916481018,
      "epoch": 9.91,
      "learning_rate": 6.673892488564925e-08,
      "loss": 0.2998,
      "step": 27410,
      "task_loss": 0.714508056640625
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39567437767982483,
      "epoch": 9.91,
      "learning_rate": 6.192327918175588e-08,
      "loss": 0.2809,
      "step": 27420,
      "task_loss": 0.7333629131317139
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35162118077278137,
      "epoch": 9.91,
      "learning_rate": 5.7287785949957645e-08,
      "loss": 0.2931,
      "step": 27430,
      "task_loss": 0.6849603652954102
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25811824202537537,
      "epoch": 9.92,
      "learning_rate": 5.283246909240979e-08,
      "loss": 0.2797,
      "step": 27440,
      "task_loss": 0.2360648214817047
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2470182627439499,
      "epoch": 9.92,
      "learning_rate": 4.855735158223184e-08,
      "loss": 0.2769,
      "step": 27450,
      "task_loss": 0.27092844247817993
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30194318294525146,
      "epoch": 9.92,
      "learning_rate": 4.446245546335991e-08,
      "loss": 0.2753,
      "step": 27460,
      "task_loss": 0.637370765209198
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23251381516456604,
      "epoch": 9.93,
      "learning_rate": 4.054780185045736e-08,
      "loss": 0.2846,
      "step": 27470,
      "task_loss": 0.4610733985900879
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26140350103378296,
      "epoch": 9.93,
      "learning_rate": 3.6813410928798194e-08,
      "loss": 0.2864,
      "step": 27480,
      "task_loss": 0.3604086637496948
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2288418412208557,
      "epoch": 9.93,
      "learning_rate": 3.3259301954154385e-08,
      "loss": 0.3032,
      "step": 27490,
      "task_loss": 0.36844557523727417
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30113106966018677,
      "epoch": 9.94,
      "learning_rate": 2.9885493252706504e-08,
      "loss": 0.311,
      "step": 27500,
      "task_loss": 0.8197970390319824
    },
    {
      "epoch": 9.94,
      "eval_exact_match": 83.40586565752129,
      "eval_f1": 90.22397249750207,
      "step": 27500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2746572494506836,
      "epoch": 9.94,
      "learning_rate": 2.6692002220946585e-08,
      "loss": 0.2967,
      "step": 27510,
      "task_loss": 0.5205618143081665
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2536380887031555,
      "epoch": 9.95,
      "learning_rate": 2.367884532559261e-08,
      "loss": 0.2916,
      "step": 27520,
      "task_loss": 0.2964051365852356
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22983843088150024,
      "epoch": 9.95,
      "learning_rate": 2.084603810349139e-08,
      "loss": 0.2668,
      "step": 27530,
      "task_loss": 0.4493524134159088
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21424216032028198,
      "epoch": 9.95,
      "learning_rate": 1.8193595161548612e-08,
      "loss": 0.2512,
      "step": 27540,
      "task_loss": 0.4849116802215576
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2508390247821808,
      "epoch": 9.96,
      "learning_rate": 1.5721530176655007e-08,
      "loss": 0.3112,
      "step": 27550,
      "task_loss": 0.2029593586921692
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23758484423160553,
      "epoch": 9.96,
      "learning_rate": 1.3429855895604747e-08,
      "loss": 0.2829,
      "step": 27560,
      "task_loss": 0.41130393743515015
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2728580832481384,
      "epoch": 9.96,
      "learning_rate": 1.1318584135048826e-08,
      "loss": 0.2873,
      "step": 27570,
      "task_loss": 0.4645345211029053
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3066297769546509,
      "epoch": 9.97,
      "learning_rate": 9.387725781405674e-09,
      "loss": 0.2714,
      "step": 27580,
      "task_loss": 0.41699323058128357
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23650574684143066,
      "epoch": 9.97,
      "learning_rate": 7.637290790830087e-09,
      "loss": 0.2506,
      "step": 27590,
      "task_loss": 0.28398048877716064
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28961479663848877,
      "epoch": 9.97,
      "learning_rate": 6.067288189162689e-09,
      "loss": 0.2773,
      "step": 27600,
      "task_loss": 0.5457533001899719
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27585506439208984,
      "epoch": 9.98,
      "learning_rate": 4.677726071856125e-09,
      "loss": 0.2972,
      "step": 27610,
      "task_loss": 0.5272984504699707
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25609445571899414,
      "epoch": 9.98,
      "learning_rate": 3.4686116039633893e-09,
      "loss": 0.2935,
      "step": 27620,
      "task_loss": 0.41631197929382324
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29504984617233276,
      "epoch": 9.99,
      "learning_rate": 2.4399510200873164e-09,
      "loss": 0.2935,
      "step": 27630,
      "task_loss": 0.48990002274513245
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26198610663414,
      "epoch": 9.99,
      "learning_rate": 1.5917496243417205e-09,
      "loss": 0.2849,
      "step": 27640,
      "task_loss": 0.4885336458683014
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3299500644207001,
      "epoch": 9.99,
      "learning_rate": 9.240117903436261e-10,
      "loss": 0.3169,
      "step": 27650,
      "task_loss": 0.3545738160610199
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2477174997329712,
      "epoch": 10.0,
      "learning_rate": 4.3674096117052304e-10,
      "loss": 0.2932,
      "step": 27660,
      "task_loss": 0.4664233326911926
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18182365596294403,
      "epoch": 10.0,
      "learning_rate": 1.299396493603666e-10,
      "loss": 0.2592,
      "step": 27670,
      "task_loss": 0.4754747748374939
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27542853355407715,
      "epoch": 10.0,
      "learning_rate": 3.6094368766059754e-12,
      "loss": 0.295,
      "step": 27680,
      "task_loss": 0.7415885925292969
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2828420400619507,
      "epoch": 10.01,
      "learning_rate": 6.999994224902488e-05,
      "loss": 0.2729,
      "step": 27690,
      "task_loss": 0.44293585419654846
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24220460653305054,
      "epoch": 10.01,
      "learning_rate": 6.999970763601507e-05,
      "loss": 0.2814,
      "step": 27700,
      "task_loss": 0.429747611284256
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25033819675445557,
      "epoch": 10.01,
      "learning_rate": 6.999929255274344e-05,
      "loss": 0.3027,
      "step": 27710,
      "task_loss": 0.3507382273674011
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29220086336135864,
      "epoch": 10.02,
      "learning_rate": 6.999869700135032e-05,
      "loss": 0.3413,
      "step": 27720,
      "task_loss": 0.4566412568092346
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2963969111442566,
      "epoch": 10.02,
      "learning_rate": 6.999792098490654e-05,
      "loss": 0.309,
      "step": 27730,
      "task_loss": 0.6837878227233887
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36075472831726074,
      "epoch": 10.03,
      "learning_rate": 6.999696450741354e-05,
      "loss": 0.3257,
      "step": 27740,
      "task_loss": 0.6962229609489441
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33492541313171387,
      "epoch": 10.03,
      "learning_rate": 6.99958275738032e-05,
      "loss": 0.332,
      "step": 27750,
      "task_loss": 0.8062801361083984
    },
    {
      "epoch": 10.03,
      "eval_exact_match": 82.8476821192053,
      "eval_f1": 89.69458878455617,
      "step": 27750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23911315202713013,
      "epoch": 10.03,
      "learning_rate": 6.999451018993795e-05,
      "loss": 0.3091,
      "step": 27760,
      "task_loss": 0.32358914613723755
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31843769550323486,
      "epoch": 10.04,
      "learning_rate": 6.999301236261068e-05,
      "loss": 0.3318,
      "step": 27770,
      "task_loss": 0.6448853015899658
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24498042464256287,
      "epoch": 10.04,
      "learning_rate": 6.999133409954468e-05,
      "loss": 0.2847,
      "step": 27780,
      "task_loss": 0.6785096526145935
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28559374809265137,
      "epoch": 10.04,
      "learning_rate": 6.998947540939361e-05,
      "loss": 0.3198,
      "step": 27790,
      "task_loss": 0.3165036141872406
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2844002842903137,
      "epoch": 10.05,
      "learning_rate": 6.998743630174154e-05,
      "loss": 0.35,
      "step": 27800,
      "task_loss": 0.4596233665943146
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3088065981864929,
      "epoch": 10.05,
      "learning_rate": 6.998521678710278e-05,
      "loss": 0.3219,
      "step": 27810,
      "task_loss": 0.4042125642299652
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42176491022109985,
      "epoch": 10.05,
      "learning_rate": 6.998281687692191e-05,
      "loss": 0.3226,
      "step": 27820,
      "task_loss": 0.6242938041687012
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26688888669013977,
      "epoch": 10.06,
      "learning_rate": 6.998023658357363e-05,
      "loss": 0.2935,
      "step": 27830,
      "task_loss": 0.7084911465644836
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2304016351699829,
      "epoch": 10.06,
      "learning_rate": 6.997747592036285e-05,
      "loss": 0.2946,
      "step": 27840,
      "task_loss": 0.3190610706806183
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3269502818584442,
      "epoch": 10.07,
      "learning_rate": 6.997453490152445e-05,
      "loss": 0.3034,
      "step": 27850,
      "task_loss": 0.8847068548202515
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.273052453994751,
      "epoch": 10.07,
      "learning_rate": 6.997141354222332e-05,
      "loss": 0.302,
      "step": 27860,
      "task_loss": 0.6260406970977783
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3863045573234558,
      "epoch": 10.07,
      "learning_rate": 6.996811185855423e-05,
      "loss": 0.3105,
      "step": 27870,
      "task_loss": 0.745919406414032
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3247497081756592,
      "epoch": 10.08,
      "learning_rate": 6.996462986754179e-05,
      "loss": 0.323,
      "step": 27880,
      "task_loss": 0.9325064420700073
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30745017528533936,
      "epoch": 10.08,
      "learning_rate": 6.99609675871403e-05,
      "loss": 0.3525,
      "step": 27890,
      "task_loss": 0.42076602578163147
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2294982671737671,
      "epoch": 10.08,
      "learning_rate": 6.995712503623371e-05,
      "loss": 0.3174,
      "step": 27900,
      "task_loss": 0.40610066056251526
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29557597637176514,
      "epoch": 10.09,
      "learning_rate": 6.995310223463551e-05,
      "loss": 0.3273,
      "step": 27910,
      "task_loss": 0.42172640562057495
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41919246315956116,
      "epoch": 10.09,
      "learning_rate": 6.994889920308862e-05,
      "loss": 0.3145,
      "step": 27920,
      "task_loss": 0.9535088539123535
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32514023780822754,
      "epoch": 10.09,
      "learning_rate": 6.994451596326527e-05,
      "loss": 0.3451,
      "step": 27930,
      "task_loss": 0.29906806349754333
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23976783454418182,
      "epoch": 10.1,
      "learning_rate": 6.993995253776693e-05,
      "loss": 0.2963,
      "step": 27940,
      "task_loss": 0.37582865357398987
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2302740216255188,
      "epoch": 10.1,
      "learning_rate": 6.993520895012417e-05,
      "loss": 0.2894,
      "step": 27950,
      "task_loss": 0.5257495641708374
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.305722177028656,
      "epoch": 10.1,
      "learning_rate": 6.993028522479649e-05,
      "loss": 0.3383,
      "step": 27960,
      "task_loss": 0.5389013886451721
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29677560925483704,
      "epoch": 10.11,
      "learning_rate": 6.992518138717231e-05,
      "loss": 0.3484,
      "step": 27970,
      "task_loss": 0.7256383895874023
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29055553674697876,
      "epoch": 10.11,
      "learning_rate": 6.99198974635687e-05,
      "loss": 0.3162,
      "step": 27980,
      "task_loss": 0.2578037679195404
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26083576679229736,
      "epoch": 10.12,
      "learning_rate": 6.991443348123137e-05,
      "loss": 0.35,
      "step": 27990,
      "task_loss": 0.4990968108177185
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2811985909938812,
      "epoch": 10.12,
      "learning_rate": 6.990878946833444e-05,
      "loss": 0.2998,
      "step": 28000,
      "task_loss": 0.43009185791015625
    },
    {
      "epoch": 10.12,
      "eval_exact_match": 83.0085146641438,
      "eval_f1": 89.82292999619031,
      "step": 28000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3266202211380005,
      "epoch": 10.12,
      "learning_rate": 6.990296545398035e-05,
      "loss": 0.324,
      "step": 28010,
      "task_loss": 0.6813063621520996
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28243324160575867,
      "epoch": 10.13,
      "learning_rate": 6.989696146819965e-05,
      "loss": 0.3062,
      "step": 28020,
      "task_loss": 0.74167400598526
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44650954008102417,
      "epoch": 10.13,
      "learning_rate": 6.989077754195093e-05,
      "loss": 0.3442,
      "step": 28030,
      "task_loss": 0.6010991930961609
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31492531299591064,
      "epoch": 10.13,
      "learning_rate": 6.98844137071206e-05,
      "loss": 0.3425,
      "step": 28040,
      "task_loss": 0.3457045257091522
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4411157965660095,
      "epoch": 10.14,
      "learning_rate": 6.987786999652272e-05,
      "loss": 0.3247,
      "step": 28050,
      "task_loss": 0.5268443822860718
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3450314402580261,
      "epoch": 10.14,
      "learning_rate": 6.987114644389885e-05,
      "loss": 0.3357,
      "step": 28060,
      "task_loss": 0.4746524691581726
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31496644020080566,
      "epoch": 10.14,
      "learning_rate": 6.986424308391791e-05,
      "loss": 0.3108,
      "step": 28070,
      "task_loss": 0.5983226299285889
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36821407079696655,
      "epoch": 10.15,
      "learning_rate": 6.985715995217594e-05,
      "loss": 0.3272,
      "step": 28080,
      "task_loss": 0.38681572675704956
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2646081745624542,
      "epoch": 10.15,
      "learning_rate": 6.984989708519595e-05,
      "loss": 0.3579,
      "step": 28090,
      "task_loss": 0.4529436528682709
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39518582820892334,
      "epoch": 10.16,
      "learning_rate": 6.984245452042771e-05,
      "loss": 0.3484,
      "step": 28100,
      "task_loss": 0.5414221882820129
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3264884352684021,
      "epoch": 10.16,
      "learning_rate": 6.983483229624761e-05,
      "loss": 0.3151,
      "step": 28110,
      "task_loss": 0.29908621311187744
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3241349458694458,
      "epoch": 10.16,
      "learning_rate": 6.982703045195837e-05,
      "loss": 0.3256,
      "step": 28120,
      "task_loss": 0.4549922049045563
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39608556032180786,
      "epoch": 10.17,
      "learning_rate": 6.981904902778897e-05,
      "loss": 0.3149,
      "step": 28130,
      "task_loss": 0.4623832702636719
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2852123975753784,
      "epoch": 10.17,
      "learning_rate": 6.981088806489427e-05,
      "loss": 0.3348,
      "step": 28140,
      "task_loss": 0.5782081484794617
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.295036256313324,
      "epoch": 10.17,
      "learning_rate": 6.980254760535496e-05,
      "loss": 0.3309,
      "step": 28150,
      "task_loss": 0.6902420520782471
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35321053862571716,
      "epoch": 10.18,
      "learning_rate": 6.979402769217725e-05,
      "loss": 0.361,
      "step": 28160,
      "task_loss": 0.7698351144790649
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38389283418655396,
      "epoch": 10.18,
      "learning_rate": 6.978532836929269e-05,
      "loss": 0.3518,
      "step": 28170,
      "task_loss": 0.5905847549438477
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3887339234352112,
      "epoch": 10.18,
      "learning_rate": 6.977644968155789e-05,
      "loss": 0.3568,
      "step": 28180,
      "task_loss": 0.321150004863739
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28333669900894165,
      "epoch": 10.19,
      "learning_rate": 6.976739167475437e-05,
      "loss": 0.3561,
      "step": 28190,
      "task_loss": 0.30792325735092163
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34358274936676025,
      "epoch": 10.19,
      "learning_rate": 6.975815439558825e-05,
      "loss": 0.3405,
      "step": 28200,
      "task_loss": 0.562443733215332
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21040789783000946,
      "epoch": 10.2,
      "learning_rate": 6.974873789169005e-05,
      "loss": 0.3179,
      "step": 28210,
      "task_loss": 0.5361858606338501
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3476879596710205,
      "epoch": 10.2,
      "learning_rate": 6.973914221161442e-05,
      "loss": 0.3577,
      "step": 28220,
      "task_loss": 0.46973469853401184
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32371383905410767,
      "epoch": 10.2,
      "learning_rate": 6.972936740483992e-05,
      "loss": 0.3243,
      "step": 28230,
      "task_loss": 0.969584584236145
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27451419830322266,
      "epoch": 10.21,
      "learning_rate": 6.971941352176876e-05,
      "loss": 0.3535,
      "step": 28240,
      "task_loss": 0.4802190661430359
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38729017972946167,
      "epoch": 10.21,
      "learning_rate": 6.970928061372649e-05,
      "loss": 0.3805,
      "step": 28250,
      "task_loss": 0.5118781328201294
    },
    {
      "epoch": 10.21,
      "eval_exact_match": 82.44087038789026,
      "eval_f1": 89.28424231167489,
      "step": 28250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3107261657714844,
      "epoch": 10.21,
      "learning_rate": 6.969896873296178e-05,
      "loss": 0.3785,
      "step": 28260,
      "task_loss": 0.37636032700538635
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.264718234539032,
      "epoch": 10.22,
      "learning_rate": 6.968847793264618e-05,
      "loss": 0.3562,
      "step": 28270,
      "task_loss": 0.523037314414978
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30193042755126953,
      "epoch": 10.22,
      "learning_rate": 6.967780826687376e-05,
      "loss": 0.3533,
      "step": 28280,
      "task_loss": 0.3143378794193268
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4219399094581604,
      "epoch": 10.22,
      "learning_rate": 6.966695979066094e-05,
      "loss": 0.3557,
      "step": 28290,
      "task_loss": 0.5991357564926147
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2808760404586792,
      "epoch": 10.23,
      "learning_rate": 6.965593255994606e-05,
      "loss": 0.3481,
      "step": 28300,
      "task_loss": 0.2749364376068115
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31335628032684326,
      "epoch": 10.23,
      "learning_rate": 6.964472663158928e-05,
      "loss": 0.3115,
      "step": 28310,
      "task_loss": 0.4930295944213867
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46561485528945923,
      "epoch": 10.23,
      "learning_rate": 6.96333420633721e-05,
      "loss": 0.3376,
      "step": 28320,
      "task_loss": 0.6846115589141846
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27835899591445923,
      "epoch": 10.24,
      "learning_rate": 6.962177891399719e-05,
      "loss": 0.3513,
      "step": 28330,
      "task_loss": 0.5056399703025818
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2896972894668579,
      "epoch": 10.24,
      "learning_rate": 6.961003724308804e-05,
      "loss": 0.3194,
      "step": 28340,
      "task_loss": 0.43239104747772217
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3265209197998047,
      "epoch": 10.25,
      "learning_rate": 6.959811711118866e-05,
      "loss": 0.3258,
      "step": 28350,
      "task_loss": 0.7637419700622559
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.344051331281662,
      "epoch": 10.25,
      "learning_rate": 6.958601857976325e-05,
      "loss": 0.3481,
      "step": 28360,
      "task_loss": 0.33167892694473267
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39247578382492065,
      "epoch": 10.25,
      "learning_rate": 6.957374171119591e-05,
      "loss": 0.363,
      "step": 28370,
      "task_loss": 0.6352831125259399
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33645737171173096,
      "epoch": 10.26,
      "learning_rate": 6.956128656879031e-05,
      "loss": 0.3583,
      "step": 28380,
      "task_loss": 0.49702680110931396
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.369404673576355,
      "epoch": 10.26,
      "learning_rate": 6.954865321676934e-05,
      "loss": 0.3321,
      "step": 28390,
      "task_loss": 0.4059206247329712
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31144052743911743,
      "epoch": 10.26,
      "learning_rate": 6.953584172027481e-05,
      "loss": 0.3918,
      "step": 28400,
      "task_loss": 0.5401674509048462
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3088327646255493,
      "epoch": 10.27,
      "learning_rate": 6.95228521453671e-05,
      "loss": 0.3069,
      "step": 28410,
      "task_loss": 0.7411280870437622
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3105813264846802,
      "epoch": 10.27,
      "learning_rate": 6.950968455902485e-05,
      "loss": 0.3334,
      "step": 28420,
      "task_loss": 0.5137057900428772
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3664456009864807,
      "epoch": 10.27,
      "learning_rate": 6.949633902914455e-05,
      "loss": 0.3259,
      "step": 28430,
      "task_loss": 0.42102259397506714
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27014023065567017,
      "epoch": 10.28,
      "learning_rate": 6.948281562454023e-05,
      "loss": 0.3174,
      "step": 28440,
      "task_loss": 0.765162467956543
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30423638224601746,
      "epoch": 10.28,
      "learning_rate": 6.946911441494312e-05,
      "loss": 0.3587,
      "step": 28450,
      "task_loss": 0.5088251829147339
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2934759855270386,
      "epoch": 10.29,
      "learning_rate": 6.945523547100128e-05,
      "loss": 0.3603,
      "step": 28460,
      "task_loss": 0.7545541524887085
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3176019787788391,
      "epoch": 10.29,
      "learning_rate": 6.944117886427917e-05,
      "loss": 0.3805,
      "step": 28470,
      "task_loss": 0.49498802423477173
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27311158180236816,
      "epoch": 10.29,
      "learning_rate": 6.942694466725742e-05,
      "loss": 0.3732,
      "step": 28480,
      "task_loss": 0.46992021799087524
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30763930082321167,
      "epoch": 10.3,
      "learning_rate": 6.941253295333232e-05,
      "loss": 0.3442,
      "step": 28490,
      "task_loss": 0.9897069334983826
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32085344195365906,
      "epoch": 10.3,
      "learning_rate": 6.939794379681553e-05,
      "loss": 0.3693,
      "step": 28500,
      "task_loss": 0.38556602597236633
    },
    {
      "epoch": 10.3,
      "eval_exact_match": 82.76253547776727,
      "eval_f1": 89.61271925469204,
      "step": 28500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4487529397010803,
      "epoch": 10.3,
      "learning_rate": 6.938317727293362e-05,
      "loss": 0.3326,
      "step": 28510,
      "task_loss": 0.825667142868042
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3878217339515686,
      "epoch": 10.31,
      "learning_rate": 6.936823345782777e-05,
      "loss": 0.3751,
      "step": 28520,
      "task_loss": 0.8324292302131653
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24599125981330872,
      "epoch": 10.31,
      "learning_rate": 6.935311242855331e-05,
      "loss": 0.3319,
      "step": 28530,
      "task_loss": 0.2502692639827728
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3209283947944641,
      "epoch": 10.31,
      "learning_rate": 6.933781426307934e-05,
      "loss": 0.3166,
      "step": 28540,
      "task_loss": 0.6298089027404785
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3517735004425049,
      "epoch": 10.32,
      "learning_rate": 6.932233904028838e-05,
      "loss": 0.3164,
      "step": 28550,
      "task_loss": 0.541971743106842
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29791098833084106,
      "epoch": 10.32,
      "learning_rate": 6.930668683997586e-05,
      "loss": 0.3842,
      "step": 28560,
      "task_loss": 0.3178185820579529
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2982218861579895,
      "epoch": 10.33,
      "learning_rate": 6.929085774284978e-05,
      "loss": 0.3407,
      "step": 28570,
      "task_loss": 0.3605462312698364
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.279826283454895,
      "epoch": 10.33,
      "learning_rate": 6.92748518305303e-05,
      "loss": 0.3561,
      "step": 28580,
      "task_loss": 0.41762399673461914
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4041915535926819,
      "epoch": 10.33,
      "learning_rate": 6.925866918554928e-05,
      "loss": 0.3758,
      "step": 28590,
      "task_loss": 0.7959795594215393
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3129355311393738,
      "epoch": 10.34,
      "learning_rate": 6.924230989134988e-05,
      "loss": 0.3674,
      "step": 28600,
      "task_loss": 0.2740345299243927
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40068450570106506,
      "epoch": 10.34,
      "learning_rate": 6.922577403228613e-05,
      "loss": 0.3409,
      "step": 28610,
      "task_loss": 0.4845631718635559
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3121311366558075,
      "epoch": 10.34,
      "learning_rate": 6.920906169362248e-05,
      "loss": 0.3437,
      "step": 28620,
      "task_loss": 0.6060510277748108
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3488480746746063,
      "epoch": 10.35,
      "learning_rate": 6.919217296153334e-05,
      "loss": 0.3565,
      "step": 28630,
      "task_loss": 0.5007976293563843
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31296420097351074,
      "epoch": 10.35,
      "learning_rate": 6.917510792310273e-05,
      "loss": 0.3155,
      "step": 28640,
      "task_loss": 0.6471648812294006
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33371689915657043,
      "epoch": 10.35,
      "learning_rate": 6.915786666632371e-05,
      "loss": 0.3836,
      "step": 28650,
      "task_loss": 0.538357138633728
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2839885950088501,
      "epoch": 10.36,
      "learning_rate": 6.914044928009801e-05,
      "loss": 0.3373,
      "step": 28660,
      "task_loss": 0.298797070980072
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.409714013338089,
      "epoch": 10.36,
      "learning_rate": 6.91228558542355e-05,
      "loss": 0.314,
      "step": 28670,
      "task_loss": 0.4804989695549011
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34721839427948,
      "epoch": 10.37,
      "learning_rate": 6.910508647945382e-05,
      "loss": 0.3392,
      "step": 28680,
      "task_loss": 0.4981377124786377
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3223315477371216,
      "epoch": 10.37,
      "learning_rate": 6.908714124737785e-05,
      "loss": 0.3647,
      "step": 28690,
      "task_loss": 0.8685950040817261
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4292945861816406,
      "epoch": 10.37,
      "learning_rate": 6.906902025053921e-05,
      "loss": 0.3661,
      "step": 28700,
      "task_loss": 0.7423338294029236
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3805582523345947,
      "epoch": 10.38,
      "learning_rate": 6.905072358237589e-05,
      "loss": 0.3568,
      "step": 28710,
      "task_loss": 0.8040531277656555
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29263046383857727,
      "epoch": 10.38,
      "learning_rate": 6.903225133723164e-05,
      "loss": 0.325,
      "step": 28720,
      "task_loss": 0.5416374206542969
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4065228998661041,
      "epoch": 10.38,
      "learning_rate": 6.901360361035558e-05,
      "loss": 0.3613,
      "step": 28730,
      "task_loss": 0.29483121633529663
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2587207853794098,
      "epoch": 10.39,
      "learning_rate": 6.899478049790166e-05,
      "loss": 0.3002,
      "step": 28740,
      "task_loss": 0.4107722342014313
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31920528411865234,
      "epoch": 10.39,
      "learning_rate": 6.897578209692816e-05,
      "loss": 0.3557,
      "step": 28750,
      "task_loss": 0.5641241073608398
    },
    {
      "epoch": 10.39,
      "eval_exact_match": 82.47871333964049,
      "eval_f1": 89.37729603350506,
      "step": 28750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.420917809009552,
      "epoch": 10.39,
      "learning_rate": 6.895660850539724e-05,
      "loss": 0.385,
      "step": 28760,
      "task_loss": 0.7262535095214844
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.304683119058609,
      "epoch": 10.4,
      "learning_rate": 6.89372598221744e-05,
      "loss": 0.3361,
      "step": 28770,
      "task_loss": 0.34669065475463867
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3875262141227722,
      "epoch": 10.4,
      "learning_rate": 6.891773614702792e-05,
      "loss": 0.3481,
      "step": 28780,
      "task_loss": 0.563037633895874
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3265758156776428,
      "epoch": 10.4,
      "learning_rate": 6.889803758062846e-05,
      "loss": 0.365,
      "step": 28790,
      "task_loss": 0.550947904586792
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26651397347450256,
      "epoch": 10.41,
      "learning_rate": 6.887816422454846e-05,
      "loss": 0.3395,
      "step": 28800,
      "task_loss": 0.2152198851108551
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3168576657772064,
      "epoch": 10.41,
      "learning_rate": 6.885811618126159e-05,
      "loss": 0.3296,
      "step": 28810,
      "task_loss": 0.5101230144500732
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32384443283081055,
      "epoch": 10.42,
      "learning_rate": 6.883789355414233e-05,
      "loss": 0.314,
      "step": 28820,
      "task_loss": 0.4653700590133667
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35834062099456787,
      "epoch": 10.42,
      "learning_rate": 6.881749644746535e-05,
      "loss": 0.3611,
      "step": 28830,
      "task_loss": 0.37349605560302734
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2610001266002655,
      "epoch": 10.42,
      "learning_rate": 6.879692496640498e-05,
      "loss": 0.3326,
      "step": 28840,
      "task_loss": 0.7551101446151733
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3156958818435669,
      "epoch": 10.43,
      "learning_rate": 6.877617921703468e-05,
      "loss": 0.3077,
      "step": 28850,
      "task_loss": 0.5917555093765259
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5137543082237244,
      "epoch": 10.43,
      "learning_rate": 6.875525930632653e-05,
      "loss": 0.3528,
      "step": 28860,
      "task_loss": 0.5057504177093506
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2843777537345886,
      "epoch": 10.43,
      "learning_rate": 6.873416534215064e-05,
      "loss": 0.3364,
      "step": 28870,
      "task_loss": 0.560868501663208
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2687450051307678,
      "epoch": 10.44,
      "learning_rate": 6.871289743327455e-05,
      "loss": 0.3915,
      "step": 28880,
      "task_loss": 0.39322781562805176
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4328731894493103,
      "epoch": 10.44,
      "learning_rate": 6.869145568936275e-05,
      "loss": 0.3473,
      "step": 28890,
      "task_loss": 1.012038230895996
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.385903537273407,
      "epoch": 10.44,
      "learning_rate": 6.866984022097612e-05,
      "loss": 0.371,
      "step": 28900,
      "task_loss": 0.6390801668167114
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32300782203674316,
      "epoch": 10.45,
      "learning_rate": 6.864805113957123e-05,
      "loss": 0.3252,
      "step": 28910,
      "task_loss": 0.3792077302932739
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23234272003173828,
      "epoch": 10.45,
      "learning_rate": 6.862608855749995e-05,
      "loss": 0.3353,
      "step": 28920,
      "task_loss": 0.5149093866348267
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2324933558702469,
      "epoch": 10.46,
      "learning_rate": 6.860395258800871e-05,
      "loss": 0.3615,
      "step": 28930,
      "task_loss": 0.41214847564697266
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2846713066101074,
      "epoch": 10.46,
      "learning_rate": 6.858164334523806e-05,
      "loss": 0.3071,
      "step": 28940,
      "task_loss": 0.3972511887550354
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.329755038022995,
      "epoch": 10.46,
      "learning_rate": 6.85591609442219e-05,
      "loss": 0.3385,
      "step": 28950,
      "task_loss": 0.5056531429290771
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27635079622268677,
      "epoch": 10.47,
      "learning_rate": 6.853650550088709e-05,
      "loss": 0.3617,
      "step": 28960,
      "task_loss": 0.3319745659828186
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24840697646141052,
      "epoch": 10.47,
      "learning_rate": 6.85136771320527e-05,
      "loss": 0.3252,
      "step": 28970,
      "task_loss": 0.514814019203186
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28714248538017273,
      "epoch": 10.47,
      "learning_rate": 6.849067595542946e-05,
      "loss": 0.3363,
      "step": 28980,
      "task_loss": 0.4475414752960205
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34752196073532104,
      "epoch": 10.48,
      "learning_rate": 6.846750208961921e-05,
      "loss": 0.3385,
      "step": 28990,
      "task_loss": 0.3642599582672119
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31241142749786377,
      "epoch": 10.48,
      "learning_rate": 6.844415565411416e-05,
      "loss": 0.3484,
      "step": 29000,
      "task_loss": 0.38981834053993225
    },
    {
      "epoch": 10.48,
      "eval_exact_match": 82.66792809839167,
      "eval_f1": 89.50805048559363,
      "step": 29000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43216896057128906,
      "epoch": 10.48,
      "learning_rate": 6.84206367692964e-05,
      "loss": 0.3545,
      "step": 29010,
      "task_loss": 0.4858664274215698
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4173024296760559,
      "epoch": 10.49,
      "learning_rate": 6.83969455564372e-05,
      "loss": 0.3654,
      "step": 29020,
      "task_loss": 0.564123272895813
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2556716203689575,
      "epoch": 10.49,
      "learning_rate": 6.837308213769643e-05,
      "loss": 0.352,
      "step": 29030,
      "task_loss": 0.6481924653053284
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2759532332420349,
      "epoch": 10.5,
      "learning_rate": 6.834904663612188e-05,
      "loss": 0.3315,
      "step": 29040,
      "task_loss": 0.5259366035461426
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3986692428588867,
      "epoch": 10.5,
      "learning_rate": 6.832483917564871e-05,
      "loss": 0.3774,
      "step": 29050,
      "task_loss": 0.8505685329437256
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.353090763092041,
      "epoch": 10.5,
      "learning_rate": 6.830045988109869e-05,
      "loss": 0.3574,
      "step": 29060,
      "task_loss": 0.39217841625213623
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46030208468437195,
      "epoch": 10.51,
      "learning_rate": 6.827590887817969e-05,
      "loss": 0.3201,
      "step": 29070,
      "task_loss": 0.6554713249206543
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30041784048080444,
      "epoch": 10.51,
      "learning_rate": 6.825118629348493e-05,
      "loss": 0.3514,
      "step": 29080,
      "task_loss": 0.2941610813140869
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4227467179298401,
      "epoch": 10.51,
      "learning_rate": 6.822629225449237e-05,
      "loss": 0.361,
      "step": 29090,
      "task_loss": 0.520158052444458
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32006293535232544,
      "epoch": 10.52,
      "learning_rate": 6.820122688956404e-05,
      "loss": 0.3093,
      "step": 29100,
      "task_loss": 0.8387197256088257
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4144279956817627,
      "epoch": 10.52,
      "learning_rate": 6.817599032794539e-05,
      "loss": 0.4,
      "step": 29110,
      "task_loss": 0.5196925401687622
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36042001843452454,
      "epoch": 10.52,
      "learning_rate": 6.815058269976462e-05,
      "loss": 0.3619,
      "step": 29120,
      "task_loss": 0.47894126176834106
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2716860771179199,
      "epoch": 10.53,
      "learning_rate": 6.8125004136032e-05,
      "loss": 0.3666,
      "step": 29130,
      "task_loss": 0.32598868012428284
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3188129663467407,
      "epoch": 10.53,
      "learning_rate": 6.809925476863924e-05,
      "loss": 0.3423,
      "step": 29140,
      "task_loss": 0.672299861907959
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26386719942092896,
      "epoch": 10.53,
      "learning_rate": 6.807333473035868e-05,
      "loss": 0.3565,
      "step": 29150,
      "task_loss": 0.4383131265640259
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27759337425231934,
      "epoch": 10.54,
      "learning_rate": 6.80498608827427e-05,
      "loss": 0.3565,
      "step": 29160,
      "task_loss": 0.5900576114654541
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32816818356513977,
      "epoch": 10.54,
      "learning_rate": 6.802361693871463e-05,
      "loss": 0.3393,
      "step": 29170,
      "task_loss": 0.5355024933815002
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41895249485969543,
      "epoch": 10.55,
      "learning_rate": 6.799720271381286e-05,
      "loss": 0.3376,
      "step": 29180,
      "task_loss": 0.832222044467926
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33841216564178467,
      "epoch": 10.55,
      "learning_rate": 6.797061834423806e-05,
      "loss": 0.3537,
      "step": 29190,
      "task_loss": 0.6101768016815186
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28902673721313477,
      "epoch": 10.55,
      "learning_rate": 6.794386396706813e-05,
      "loss": 0.3269,
      "step": 29200,
      "task_loss": 0.2802298069000244
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.336454302072525,
      "epoch": 10.56,
      "learning_rate": 6.791693972025767e-05,
      "loss": 0.3511,
      "step": 29210,
      "task_loss": 0.5762617588043213
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31067967414855957,
      "epoch": 10.56,
      "learning_rate": 6.788984574263712e-05,
      "loss": 0.3879,
      "step": 29220,
      "task_loss": 0.5875623226165771
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32384225726127625,
      "epoch": 10.56,
      "learning_rate": 6.786258217391215e-05,
      "loss": 0.3467,
      "step": 29230,
      "task_loss": 0.5701868534088135
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37212711572647095,
      "epoch": 10.57,
      "learning_rate": 6.78351491546629e-05,
      "loss": 0.3791,
      "step": 29240,
      "task_loss": 0.7481470108032227
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3019322156906128,
      "epoch": 10.57,
      "learning_rate": 6.780754682634324e-05,
      "loss": 0.3331,
      "step": 29250,
      "task_loss": 0.5364646911621094
    },
    {
      "epoch": 10.57,
      "eval_exact_match": 82.82876064333018,
      "eval_f1": 89.56640654091615,
      "step": 29250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43777012825012207,
      "epoch": 10.57,
      "learning_rate": 6.777977533128004e-05,
      "loss": 0.3748,
      "step": 29260,
      "task_loss": 0.9455623626708984
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4065621495246887,
      "epoch": 10.58,
      "learning_rate": 6.775183481267248e-05,
      "loss": 0.3454,
      "step": 29270,
      "task_loss": 0.46658486127853394
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5213733911514282,
      "epoch": 10.58,
      "learning_rate": 6.772372541459127e-05,
      "loss": 0.3868,
      "step": 29280,
      "task_loss": 0.6847914457321167
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26462650299072266,
      "epoch": 10.59,
      "learning_rate": 6.769544728197792e-05,
      "loss": 0.3352,
      "step": 29290,
      "task_loss": 0.2257605642080307
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2530308961868286,
      "epoch": 10.59,
      "learning_rate": 6.766700056064398e-05,
      "loss": 0.3316,
      "step": 29300,
      "task_loss": 0.6150197982788086
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30782073736190796,
      "epoch": 10.59,
      "learning_rate": 6.763838539727032e-05,
      "loss": 0.3285,
      "step": 29310,
      "task_loss": 0.5591270923614502
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4068589210510254,
      "epoch": 10.6,
      "learning_rate": 6.760960193940634e-05,
      "loss": 0.4125,
      "step": 29320,
      "task_loss": 0.6238890886306763
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3537275195121765,
      "epoch": 10.6,
      "learning_rate": 6.758065033546923e-05,
      "loss": 0.3695,
      "step": 29330,
      "task_loss": 0.33286938071250916
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3277410864830017,
      "epoch": 10.6,
      "learning_rate": 6.755153073474321e-05,
      "loss": 0.3383,
      "step": 29340,
      "task_loss": 0.6027262806892395
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3822060227394104,
      "epoch": 10.61,
      "learning_rate": 6.752224328737871e-05,
      "loss": 0.3328,
      "step": 29350,
      "task_loss": 0.5942087173461914
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5594038367271423,
      "epoch": 10.61,
      "learning_rate": 6.749278814439167e-05,
      "loss": 0.402,
      "step": 29360,
      "task_loss": 0.8234184980392456
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38665011525154114,
      "epoch": 10.61,
      "learning_rate": 6.74631654576627e-05,
      "loss": 0.3533,
      "step": 29370,
      "task_loss": 0.835229218006134
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4332032799720764,
      "epoch": 10.62,
      "learning_rate": 6.743337537993633e-05,
      "loss": 0.3692,
      "step": 29380,
      "task_loss": 0.5134978890419006
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34394896030426025,
      "epoch": 10.62,
      "learning_rate": 6.740341806482025e-05,
      "loss": 0.3678,
      "step": 29390,
      "task_loss": 0.38100263476371765
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3152540326118469,
      "epoch": 10.63,
      "learning_rate": 6.737329366678442e-05,
      "loss": 0.3637,
      "step": 29400,
      "task_loss": 0.8232011198997498
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44516491889953613,
      "epoch": 10.63,
      "learning_rate": 6.734300234116038e-05,
      "loss": 0.3467,
      "step": 29410,
      "task_loss": 0.7560466527938843
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28295183181762695,
      "epoch": 10.63,
      "learning_rate": 6.731254424414039e-05,
      "loss": 0.3554,
      "step": 29420,
      "task_loss": 0.6221187114715576
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2875472605228424,
      "epoch": 10.64,
      "learning_rate": 6.728191953277663e-05,
      "loss": 0.338,
      "step": 29430,
      "task_loss": 0.6836280822753906
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26682746410369873,
      "epoch": 10.64,
      "learning_rate": 6.725112836498043e-05,
      "loss": 0.3611,
      "step": 29440,
      "task_loss": 0.6012932062149048
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32244402170181274,
      "epoch": 10.64,
      "learning_rate": 6.722017089952138e-05,
      "loss": 0.3602,
      "step": 29450,
      "task_loss": 0.39122363924980164
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36310338973999023,
      "epoch": 10.65,
      "learning_rate": 6.71890472960266e-05,
      "loss": 0.3738,
      "step": 29460,
      "task_loss": 0.5470183491706848
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3462553918361664,
      "epoch": 10.65,
      "learning_rate": 6.715775771497985e-05,
      "loss": 0.367,
      "step": 29470,
      "task_loss": 0.5989048480987549
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35961276292800903,
      "epoch": 10.65,
      "learning_rate": 6.712630231772072e-05,
      "loss": 0.3563,
      "step": 29480,
      "task_loss": 0.6379199028015137
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30847465991973877,
      "epoch": 10.66,
      "learning_rate": 6.709468126644384e-05,
      "loss": 0.3509,
      "step": 29490,
      "task_loss": 0.7017759084701538
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3288659453392029,
      "epoch": 10.66,
      "learning_rate": 6.706289472419797e-05,
      "loss": 0.3393,
      "step": 29500,
      "task_loss": 0.6862675547599792
    },
    {
      "epoch": 10.66,
      "eval_exact_match": 82.18543046357615,
      "eval_f1": 89.15288807465497,
      "step": 29500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39969784021377563,
      "epoch": 10.66,
      "learning_rate": 6.703094285488522e-05,
      "loss": 0.3586,
      "step": 29510,
      "task_loss": 0.5755327939987183
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33747461438179016,
      "epoch": 10.67,
      "learning_rate": 6.699882582326016e-05,
      "loss": 0.3234,
      "step": 29520,
      "task_loss": 0.8560225963592529
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6065595149993896,
      "epoch": 10.67,
      "learning_rate": 6.6966543794929e-05,
      "loss": 0.3863,
      "step": 29530,
      "task_loss": 0.9350003004074097
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2929680347442627,
      "epoch": 10.68,
      "learning_rate": 6.693409693634875e-05,
      "loss": 0.337,
      "step": 29540,
      "task_loss": 0.4647179841995239
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2577773332595825,
      "epoch": 10.68,
      "learning_rate": 6.69014854148263e-05,
      "loss": 0.3784,
      "step": 29550,
      "task_loss": 0.2627255618572235
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3602387309074402,
      "epoch": 10.68,
      "learning_rate": 6.686870939851766e-05,
      "loss": 0.3755,
      "step": 29560,
      "task_loss": 0.4507540464401245
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34514838457107544,
      "epoch": 10.69,
      "learning_rate": 6.683576905642695e-05,
      "loss": 0.3542,
      "step": 29570,
      "task_loss": 0.6163294315338135
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3738999664783478,
      "epoch": 10.69,
      "learning_rate": 6.680266455840568e-05,
      "loss": 0.3251,
      "step": 29580,
      "task_loss": 0.40096384286880493
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37798354029655457,
      "epoch": 10.69,
      "learning_rate": 6.676939607515175e-05,
      "loss": 0.3634,
      "step": 29590,
      "task_loss": 0.8661670684814453
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36543959379196167,
      "epoch": 10.7,
      "learning_rate": 6.673596377820867e-05,
      "loss": 0.3391,
      "step": 29600,
      "task_loss": 0.6424458026885986
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39285749197006226,
      "epoch": 10.7,
      "learning_rate": 6.67023678399646e-05,
      "loss": 0.3876,
      "step": 29610,
      "task_loss": 0.49506786465644836
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35833126306533813,
      "epoch": 10.7,
      "learning_rate": 6.666860843365148e-05,
      "loss": 0.3749,
      "step": 29620,
      "task_loss": 0.40317094326019287
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32828640937805176,
      "epoch": 10.71,
      "learning_rate": 6.663468573334417e-05,
      "loss": 0.3612,
      "step": 29630,
      "task_loss": 0.4777117371559143
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44851982593536377,
      "epoch": 10.71,
      "learning_rate": 6.660059991395954e-05,
      "loss": 0.3595,
      "step": 29640,
      "task_loss": 0.5143475532531738
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4900306761264801,
      "epoch": 10.72,
      "learning_rate": 6.656635115125549e-05,
      "loss": 0.3457,
      "step": 29650,
      "task_loss": 0.8377711772918701
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29536741971969604,
      "epoch": 10.72,
      "learning_rate": 6.65319396218302e-05,
      "loss": 0.3508,
      "step": 29660,
      "task_loss": 0.6646140813827515
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2504414916038513,
      "epoch": 10.72,
      "learning_rate": 6.649736550312107e-05,
      "loss": 0.401,
      "step": 29670,
      "task_loss": 0.3634515702724457
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4538690745830536,
      "epoch": 10.73,
      "learning_rate": 6.646262897340388e-05,
      "loss": 0.3476,
      "step": 29680,
      "task_loss": 0.3508909344673157
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3044033646583557,
      "epoch": 10.73,
      "learning_rate": 6.642773021179186e-05,
      "loss": 0.3712,
      "step": 29690,
      "task_loss": 0.4467262625694275
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32180851697921753,
      "epoch": 10.73,
      "learning_rate": 6.639266939823477e-05,
      "loss": 0.3639,
      "step": 29700,
      "task_loss": 0.6727160215377808
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3948644995689392,
      "epoch": 10.74,
      "learning_rate": 6.635744671351794e-05,
      "loss": 0.3736,
      "step": 29710,
      "task_loss": 0.30169928073883057
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5408639907836914,
      "epoch": 10.74,
      "learning_rate": 6.632206233926139e-05,
      "loss": 0.4084,
      "step": 29720,
      "task_loss": 0.41215962171554565
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3110385537147522,
      "epoch": 10.74,
      "learning_rate": 6.628651645791885e-05,
      "loss": 0.3506,
      "step": 29730,
      "task_loss": 0.32455557584762573
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4405866861343384,
      "epoch": 10.75,
      "learning_rate": 6.625080925277681e-05,
      "loss": 0.3797,
      "step": 29740,
      "task_loss": 0.5964502096176147
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3487781882286072,
      "epoch": 10.75,
      "learning_rate": 6.621494090795367e-05,
      "loss": 0.3952,
      "step": 29750,
      "task_loss": 0.4532567262649536
    },
    {
      "epoch": 10.75,
      "eval_exact_match": 82.44087038789026,
      "eval_f1": 89.45854412702076,
      "step": 29750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28924113512039185,
      "epoch": 10.76,
      "learning_rate": 6.617891160839865e-05,
      "loss": 0.3319,
      "step": 29760,
      "task_loss": 0.6428242921829224
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3873733580112457,
      "epoch": 10.76,
      "learning_rate": 6.614272153989095e-05,
      "loss": 0.3655,
      "step": 29770,
      "task_loss": 0.48227518796920776
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3506860136985779,
      "epoch": 10.76,
      "learning_rate": 6.61063708890387e-05,
      "loss": 0.3383,
      "step": 29780,
      "task_loss": 0.4459092915058136
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29523494839668274,
      "epoch": 10.77,
      "learning_rate": 6.606985984327813e-05,
      "loss": 0.3294,
      "step": 29790,
      "task_loss": 0.5016797780990601
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3241945803165436,
      "epoch": 10.77,
      "learning_rate": 6.603318859087243e-05,
      "loss": 0.355,
      "step": 29800,
      "task_loss": 0.35577547550201416
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30834487080574036,
      "epoch": 10.77,
      "learning_rate": 6.599635732091092e-05,
      "loss": 0.3293,
      "step": 29810,
      "task_loss": 0.6570316553115845
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4538179636001587,
      "epoch": 10.78,
      "learning_rate": 6.595936622330802e-05,
      "loss": 0.3694,
      "step": 29820,
      "task_loss": 1.0713361501693726
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3451743721961975,
      "epoch": 10.78,
      "learning_rate": 6.592221548880224e-05,
      "loss": 0.3395,
      "step": 29830,
      "task_loss": 0.3584628403186798
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5398263931274414,
      "epoch": 10.78,
      "learning_rate": 6.58849053089553e-05,
      "loss": 0.3693,
      "step": 29840,
      "task_loss": 0.9976974725723267
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4781019687652588,
      "epoch": 10.79,
      "learning_rate": 6.584743587615102e-05,
      "loss": 0.3554,
      "step": 29850,
      "task_loss": 0.7893520593643188
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27053534984588623,
      "epoch": 10.79,
      "learning_rate": 6.580980738359438e-05,
      "loss": 0.383,
      "step": 29860,
      "task_loss": 0.35242271423339844
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3637334704399109,
      "epoch": 10.8,
      "learning_rate": 6.577202002531056e-05,
      "loss": 0.3162,
      "step": 29870,
      "task_loss": 0.5553810596466064
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3257104754447937,
      "epoch": 10.8,
      "learning_rate": 6.573407399614388e-05,
      "loss": 0.3478,
      "step": 29880,
      "task_loss": 0.5849871635437012
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29262852668762207,
      "epoch": 10.8,
      "learning_rate": 6.569596949175681e-05,
      "loss": 0.3707,
      "step": 29890,
      "task_loss": 0.6378922462463379
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28541576862335205,
      "epoch": 10.81,
      "learning_rate": 6.5657706708629e-05,
      "loss": 0.3268,
      "step": 29900,
      "task_loss": 0.3432478904724121
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.455171138048172,
      "epoch": 10.81,
      "learning_rate": 6.561928584405624e-05,
      "loss": 0.3743,
      "step": 29910,
      "task_loss": 0.7067415714263916
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4353346824645996,
      "epoch": 10.81,
      "learning_rate": 6.558070709614942e-05,
      "loss": 0.3513,
      "step": 29920,
      "task_loss": 0.6395390629768372
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.313425213098526,
      "epoch": 10.82,
      "learning_rate": 6.55419706638335e-05,
      "loss": 0.3437,
      "step": 29930,
      "task_loss": 1.3602879047393799
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3399485647678375,
      "epoch": 10.82,
      "learning_rate": 6.550307674684662e-05,
      "loss": 0.3678,
      "step": 29940,
      "task_loss": 0.3591665029525757
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3630330562591553,
      "epoch": 10.82,
      "learning_rate": 6.546402554573885e-05,
      "loss": 0.3401,
      "step": 29950,
      "task_loss": 0.3910972774028778
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47807592153549194,
      "epoch": 10.83,
      "learning_rate": 6.54248172618713e-05,
      "loss": 0.3686,
      "step": 29960,
      "task_loss": 0.49961668252944946
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3700833320617676,
      "epoch": 10.83,
      "learning_rate": 6.538545209741511e-05,
      "loss": 0.3815,
      "step": 29970,
      "task_loss": 0.7051327228546143
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28954607248306274,
      "epoch": 10.83,
      "learning_rate": 6.534593025535028e-05,
      "loss": 0.3727,
      "step": 29980,
      "task_loss": 0.5030256509780884
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3162761926651001,
      "epoch": 10.84,
      "learning_rate": 6.530625193946472e-05,
      "loss": 0.3309,
      "step": 29990,
      "task_loss": 0.4224050045013428
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43589645624160767,
      "epoch": 10.84,
      "learning_rate": 6.526641735435317e-05,
      "loss": 0.3636,
      "step": 30000,
      "task_loss": 0.6760845184326172
    },
    {
      "epoch": 10.84,
      "eval_exact_match": 82.28949858088932,
      "eval_f1": 89.30660591571832,
      "step": 30000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47988027334213257,
      "epoch": 10.85,
      "learning_rate": 6.522642670541613e-05,
      "loss": 0.3823,
      "step": 30010,
      "task_loss": 0.8904911279678345
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2862057089805603,
      "epoch": 10.85,
      "learning_rate": 6.518628019885888e-05,
      "loss": 0.3367,
      "step": 30020,
      "task_loss": 0.6241217851638794
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3560543656349182,
      "epoch": 10.85,
      "learning_rate": 6.514597804169025e-05,
      "loss": 0.3666,
      "step": 30030,
      "task_loss": 0.6103063821792603
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6194554567337036,
      "epoch": 10.86,
      "learning_rate": 6.510552044172176e-05,
      "loss": 0.412,
      "step": 30040,
      "task_loss": 0.8107380270957947
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3812370300292969,
      "epoch": 10.86,
      "learning_rate": 6.506490760756639e-05,
      "loss": 0.3513,
      "step": 30050,
      "task_loss": 0.5243598818778992
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3137257695198059,
      "epoch": 10.86,
      "learning_rate": 6.502413974863753e-05,
      "loss": 0.3283,
      "step": 30060,
      "task_loss": 0.411754310131073
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2372526228427887,
      "epoch": 10.87,
      "learning_rate": 6.498321707514802e-05,
      "loss": 0.3206,
      "step": 30070,
      "task_loss": 0.6143356561660767
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36465486884117126,
      "epoch": 10.87,
      "learning_rate": 6.494213979810891e-05,
      "loss": 0.3486,
      "step": 30080,
      "task_loss": 0.6931987404823303
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5073741674423218,
      "epoch": 10.87,
      "learning_rate": 6.490090812932844e-05,
      "loss": 0.3895,
      "step": 30090,
      "task_loss": 0.9303778409957886
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2504512071609497,
      "epoch": 10.88,
      "learning_rate": 6.485952228141097e-05,
      "loss": 0.3479,
      "step": 30100,
      "task_loss": 0.41266775131225586
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3641784191131592,
      "epoch": 10.88,
      "learning_rate": 6.481798246775586e-05,
      "loss": 0.3401,
      "step": 30110,
      "task_loss": 0.5535559058189392
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46111181378364563,
      "epoch": 10.89,
      "learning_rate": 6.477628890255634e-05,
      "loss": 0.3689,
      "step": 30120,
      "task_loss": 0.7662212252616882
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36023229360580444,
      "epoch": 10.89,
      "learning_rate": 6.473444180079845e-05,
      "loss": 0.3828,
      "step": 30130,
      "task_loss": 0.5160745978355408
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33962923288345337,
      "epoch": 10.89,
      "learning_rate": 6.469244137825993e-05,
      "loss": 0.3318,
      "step": 30140,
      "task_loss": 0.8273515701293945
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3233642280101776,
      "epoch": 10.9,
      "learning_rate": 6.465028785150908e-05,
      "loss": 0.3746,
      "step": 30150,
      "task_loss": 0.4669772982597351
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39144349098205566,
      "epoch": 10.9,
      "learning_rate": 6.460798143790366e-05,
      "loss": 0.332,
      "step": 30160,
      "task_loss": 0.5270528793334961
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44592922925949097,
      "epoch": 10.9,
      "learning_rate": 6.456552235558976e-05,
      "loss": 0.3701,
      "step": 30170,
      "task_loss": 1.0220377445220947
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3141802251338959,
      "epoch": 10.91,
      "learning_rate": 6.452291082350068e-05,
      "loss": 0.3544,
      "step": 30180,
      "task_loss": 0.2899113893508911
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28677743673324585,
      "epoch": 10.91,
      "learning_rate": 6.448014706135582e-05,
      "loss": 0.3924,
      "step": 30190,
      "task_loss": 0.26135581731796265
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32510054111480713,
      "epoch": 10.91,
      "learning_rate": 6.443723128965951e-05,
      "loss": 0.3469,
      "step": 30200,
      "task_loss": 0.88126540184021
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3127478361129761,
      "epoch": 10.92,
      "learning_rate": 6.439416372969992e-05,
      "loss": 0.3498,
      "step": 30210,
      "task_loss": 0.8732039332389832
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2943447232246399,
      "epoch": 10.92,
      "learning_rate": 6.435094460354784e-05,
      "loss": 0.3535,
      "step": 30220,
      "task_loss": 0.645883321762085
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3853033185005188,
      "epoch": 10.93,
      "learning_rate": 6.430757413405562e-05,
      "loss": 0.3697,
      "step": 30230,
      "task_loss": 0.7212446928024292
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3296411633491516,
      "epoch": 10.93,
      "learning_rate": 6.426405254485603e-05,
      "loss": 0.3912,
      "step": 30240,
      "task_loss": 0.6896770000457764
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31863152980804443,
      "epoch": 10.93,
      "learning_rate": 6.422038006036097e-05,
      "loss": 0.3574,
      "step": 30250,
      "task_loss": 0.39320269227027893
    },
    {
      "epoch": 10.93,
      "eval_exact_match": 82.65846736045411,
      "eval_f1": 89.52884792493833,
      "step": 30250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4146077036857605,
      "epoch": 10.94,
      "learning_rate": 6.417655690576046e-05,
      "loss": 0.4061,
      "step": 30260,
      "task_loss": 0.8556808233261108
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36979159712791443,
      "epoch": 10.94,
      "learning_rate": 6.413258330702145e-05,
      "loss": 0.377,
      "step": 30270,
      "task_loss": 0.4253246486186981
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30879682302474976,
      "epoch": 10.94,
      "learning_rate": 6.408845949088657e-05,
      "loss": 0.3524,
      "step": 30280,
      "task_loss": 0.4455922245979309
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32917648553848267,
      "epoch": 10.95,
      "learning_rate": 6.404418568487308e-05,
      "loss": 0.3711,
      "step": 30290,
      "task_loss": 0.7015597820281982
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33461427688598633,
      "epoch": 10.95,
      "learning_rate": 6.39997621172716e-05,
      "loss": 0.3623,
      "step": 30300,
      "task_loss": 0.3214136064052582
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3308192491531372,
      "epoch": 10.95,
      "learning_rate": 6.395518901714497e-05,
      "loss": 0.3402,
      "step": 30310,
      "task_loss": 0.5261917114257812
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3672281503677368,
      "epoch": 10.96,
      "learning_rate": 6.391046661432711e-05,
      "loss": 0.3823,
      "step": 30320,
      "task_loss": 0.616168737411499
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4320986270904541,
      "epoch": 10.96,
      "learning_rate": 6.386559513942175e-05,
      "loss": 0.3622,
      "step": 30330,
      "task_loss": 0.6258487701416016
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38706454634666443,
      "epoch": 10.96,
      "learning_rate": 6.38205748238013e-05,
      "loss": 0.3706,
      "step": 30340,
      "task_loss": 0.9309755563735962
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35249724984169006,
      "epoch": 10.97,
      "learning_rate": 6.377540589960567e-05,
      "loss": 0.3604,
      "step": 30350,
      "task_loss": 0.4586070775985718
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31285613775253296,
      "epoch": 10.97,
      "learning_rate": 6.373008859974099e-05,
      "loss": 0.3438,
      "step": 30360,
      "task_loss": 0.6232818365097046
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3567231297492981,
      "epoch": 10.98,
      "learning_rate": 6.36846231578785e-05,
      "loss": 0.3905,
      "step": 30370,
      "task_loss": 0.9104863405227661
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21704626083374023,
      "epoch": 10.98,
      "learning_rate": 6.363900980845333e-05,
      "loss": 0.3709,
      "step": 30380,
      "task_loss": 0.4989585876464844
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35998252034187317,
      "epoch": 10.98,
      "learning_rate": 6.359324878666324e-05,
      "loss": 0.3687,
      "step": 30390,
      "task_loss": 0.31220775842666626
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24193412065505981,
      "epoch": 10.99,
      "learning_rate": 6.354734032846744e-05,
      "loss": 0.374,
      "step": 30400,
      "task_loss": 0.6692792177200317
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3294738829135895,
      "epoch": 10.99,
      "learning_rate": 6.350128467058539e-05,
      "loss": 0.3461,
      "step": 30410,
      "task_loss": 0.4882463812828064
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35924404859542847,
      "epoch": 10.99,
      "learning_rate": 6.345508205049552e-05,
      "loss": 0.3747,
      "step": 30420,
      "task_loss": 0.5941590070724487
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3159148693084717,
      "epoch": 11.0,
      "learning_rate": 6.340873270643411e-05,
      "loss": 0.3673,
      "step": 30430,
      "task_loss": 0.6530588865280151
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40394383668899536,
      "epoch": 11.0,
      "learning_rate": 6.336223687739394e-05,
      "loss": 0.3517,
      "step": 30440,
      "task_loss": 0.5973001718521118
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2577652335166931,
      "epoch": 11.0,
      "learning_rate": 6.331559480312315e-05,
      "loss": 0.3301,
      "step": 30450,
      "task_loss": 0.4629456698894501
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18141883611679077,
      "epoch": 11.01,
      "learning_rate": 6.326880672412396e-05,
      "loss": 0.3134,
      "step": 30460,
      "task_loss": 0.34045690298080444
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33387744426727295,
      "epoch": 11.01,
      "learning_rate": 6.322187288165144e-05,
      "loss": 0.3303,
      "step": 30470,
      "task_loss": 0.6234238743782043
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30332377552986145,
      "epoch": 11.02,
      "learning_rate": 6.317479351771226e-05,
      "loss": 0.3449,
      "step": 30480,
      "task_loss": 0.3982439637184143
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2568157911300659,
      "epoch": 11.02,
      "learning_rate": 6.312756887506345e-05,
      "loss": 0.3402,
      "step": 30490,
      "task_loss": 0.42569881677627563
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34609246253967285,
      "epoch": 11.02,
      "learning_rate": 6.308019919721113e-05,
      "loss": 0.3115,
      "step": 30500,
      "task_loss": 0.29865169525146484
    },
    {
      "epoch": 11.02,
      "eval_exact_match": 82.52601702932829,
      "eval_f1": 89.58780871354091,
      "step": 30500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30228573083877563,
      "epoch": 11.03,
      "learning_rate": 6.303268472840934e-05,
      "loss": 0.3476,
      "step": 30510,
      "task_loss": 0.5901381969451904
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39503413438796997,
      "epoch": 11.03,
      "learning_rate": 6.29850257136586e-05,
      "loss": 0.3726,
      "step": 30520,
      "task_loss": 0.6809341907501221
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34292078018188477,
      "epoch": 11.03,
      "learning_rate": 6.293722239870485e-05,
      "loss": 0.3374,
      "step": 30530,
      "task_loss": 0.6341568827629089
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2626557946205139,
      "epoch": 11.04,
      "learning_rate": 6.288927503003805e-05,
      "loss": 0.3437,
      "step": 30540,
      "task_loss": 0.8239531517028809
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25144004821777344,
      "epoch": 11.04,
      "learning_rate": 6.284118385489095e-05,
      "loss": 0.3219,
      "step": 30550,
      "task_loss": 0.44134482741355896
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3792293667793274,
      "epoch": 11.04,
      "learning_rate": 6.279294912123784e-05,
      "loss": 0.3665,
      "step": 30560,
      "task_loss": 0.43918779492378235
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5096983313560486,
      "epoch": 11.05,
      "learning_rate": 6.27445710777932e-05,
      "loss": 0.3373,
      "step": 30570,
      "task_loss": 0.5208146572113037
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4361113905906677,
      "epoch": 11.05,
      "learning_rate": 6.269604997401051e-05,
      "loss": 0.3685,
      "step": 30580,
      "task_loss": 0.9114473462104797
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3823733329772949,
      "epoch": 11.06,
      "learning_rate": 6.264738606008087e-05,
      "loss": 0.3577,
      "step": 30590,
      "task_loss": 0.5972809195518494
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2892557382583618,
      "epoch": 11.06,
      "learning_rate": 6.259857958693182e-05,
      "loss": 0.3116,
      "step": 30600,
      "task_loss": 0.4515109956264496
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.225788414478302,
      "epoch": 11.06,
      "learning_rate": 6.254963080622591e-05,
      "loss": 0.3125,
      "step": 30610,
      "task_loss": 0.40221279859542847
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31782832741737366,
      "epoch": 11.07,
      "learning_rate": 6.250053997035956e-05,
      "loss": 0.3688,
      "step": 30620,
      "task_loss": 0.5184889435768127
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33337557315826416,
      "epoch": 11.07,
      "learning_rate": 6.245130733246159e-05,
      "loss": 0.3247,
      "step": 30630,
      "task_loss": 0.5070463418960571
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33691734075546265,
      "epoch": 11.07,
      "learning_rate": 6.240193314639205e-05,
      "loss": 0.328,
      "step": 30640,
      "task_loss": 0.4712425470352173
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3276863694190979,
      "epoch": 11.08,
      "learning_rate": 6.235241766674084e-05,
      "loss": 0.3553,
      "step": 30650,
      "task_loss": 0.6992507576942444
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4350316822528839,
      "epoch": 11.08,
      "learning_rate": 6.230276114882642e-05,
      "loss": 0.369,
      "step": 30660,
      "task_loss": 0.7506589889526367
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32287871837615967,
      "epoch": 11.08,
      "learning_rate": 6.225296384869451e-05,
      "loss": 0.3442,
      "step": 30670,
      "task_loss": 0.5573731660842896
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.275474488735199,
      "epoch": 11.09,
      "learning_rate": 6.220302602311674e-05,
      "loss": 0.3197,
      "step": 30680,
      "task_loss": 0.7555659413337708
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21844923496246338,
      "epoch": 11.09,
      "learning_rate": 6.21529479295893e-05,
      "loss": 0.3129,
      "step": 30690,
      "task_loss": 0.4042033851146698
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3810504376888275,
      "epoch": 11.1,
      "learning_rate": 6.21027298263317e-05,
      "loss": 0.3421,
      "step": 30700,
      "task_loss": 0.4075656235218048
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2908656597137451,
      "epoch": 11.1,
      "learning_rate": 6.205237197228537e-05,
      "loss": 0.3134,
      "step": 30710,
      "task_loss": 0.87837815284729
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.321632444858551,
      "epoch": 11.1,
      "learning_rate": 6.200187462711232e-05,
      "loss": 0.3111,
      "step": 30720,
      "task_loss": 0.4441836476325989
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25605639815330505,
      "epoch": 11.11,
      "learning_rate": 6.195123805119386e-05,
      "loss": 0.3282,
      "step": 30730,
      "task_loss": 0.5329892039299011
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3184843063354492,
      "epoch": 11.11,
      "learning_rate": 6.19004625056292e-05,
      "loss": 0.3756,
      "step": 30740,
      "task_loss": 0.49759477376937866
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2887711226940155,
      "epoch": 11.11,
      "learning_rate": 6.184954825223412e-05,
      "loss": 0.3403,
      "step": 30750,
      "task_loss": 0.390583336353302
    },
    {
      "epoch": 11.11,
      "eval_exact_match": 82.57332071901608,
      "eval_f1": 89.61192290821421,
      "step": 30750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4305284321308136,
      "epoch": 11.12,
      "learning_rate": 6.179849555353966e-05,
      "loss": 0.3243,
      "step": 30760,
      "task_loss": 0.6925809383392334
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3064586818218231,
      "epoch": 11.12,
      "learning_rate": 6.174730467279065e-05,
      "loss": 0.3077,
      "step": 30770,
      "task_loss": 0.49236008524894714
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37597641348838806,
      "epoch": 11.12,
      "learning_rate": 6.169597587394453e-05,
      "loss": 0.311,
      "step": 30780,
      "task_loss": 1.0985610485076904
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25383347272872925,
      "epoch": 11.13,
      "learning_rate": 6.164450942166983e-05,
      "loss": 0.302,
      "step": 30790,
      "task_loss": 0.2893102169036865
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29239368438720703,
      "epoch": 11.13,
      "learning_rate": 6.159290558134487e-05,
      "loss": 0.3136,
      "step": 30800,
      "task_loss": 0.43395623564720154
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3338686227798462,
      "epoch": 11.13,
      "learning_rate": 6.154116461905642e-05,
      "loss": 0.2958,
      "step": 30810,
      "task_loss": 0.33142733573913574
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4973904490470886,
      "epoch": 11.14,
      "learning_rate": 6.14892868015983e-05,
      "loss": 0.3597,
      "step": 30820,
      "task_loss": 0.8295789957046509
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3147459030151367,
      "epoch": 11.14,
      "learning_rate": 6.143727239646995e-05,
      "loss": 0.3328,
      "step": 30830,
      "task_loss": 0.5006129145622253
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37410253286361694,
      "epoch": 11.15,
      "learning_rate": 6.138512167187514e-05,
      "loss": 0.322,
      "step": 30840,
      "task_loss": 0.4317512512207031
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26965123414993286,
      "epoch": 11.15,
      "learning_rate": 6.133283489672054e-05,
      "loss": 0.3166,
      "step": 30850,
      "task_loss": 0.7003058791160583
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3355432450771332,
      "epoch": 11.15,
      "learning_rate": 6.128041234061437e-05,
      "loss": 0.3247,
      "step": 30860,
      "task_loss": 0.6356285810470581
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3858422636985779,
      "epoch": 11.16,
      "learning_rate": 6.122785427386493e-05,
      "loss": 0.3423,
      "step": 30870,
      "task_loss": 0.6549724340438843
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28326427936553955,
      "epoch": 11.16,
      "learning_rate": 6.117516096747929e-05,
      "loss": 0.3442,
      "step": 30880,
      "task_loss": 0.4158441424369812
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24142545461654663,
      "epoch": 11.16,
      "learning_rate": 6.112233269316187e-05,
      "loss": 0.2961,
      "step": 30890,
      "task_loss": 0.7818669080734253
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3277119994163513,
      "epoch": 11.17,
      "learning_rate": 6.106936972331298e-05,
      "loss": 0.3308,
      "step": 30900,
      "task_loss": 0.7212725877761841
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26108160614967346,
      "epoch": 11.17,
      "learning_rate": 6.101627233102756e-05,
      "loss": 0.3228,
      "step": 30910,
      "task_loss": 0.43422454595565796
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28188270330429077,
      "epoch": 11.17,
      "learning_rate": 6.096304079009358e-05,
      "loss": 0.3063,
      "step": 30920,
      "task_loss": 0.39963170886039734
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3516286015510559,
      "epoch": 11.18,
      "learning_rate": 6.090967537499077e-05,
      "loss": 0.3139,
      "step": 30930,
      "task_loss": 0.4946768879890442
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4153703451156616,
      "epoch": 11.18,
      "learning_rate": 6.085617636088917e-05,
      "loss": 0.3593,
      "step": 30940,
      "task_loss": 0.6110119223594666
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3178633451461792,
      "epoch": 11.19,
      "learning_rate": 6.080254402364767e-05,
      "loss": 0.3265,
      "step": 30950,
      "task_loss": 0.6084519624710083
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29074785113334656,
      "epoch": 11.19,
      "learning_rate": 6.074877863981264e-05,
      "loss": 0.3544,
      "step": 30960,
      "task_loss": 0.6702461242675781
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3396759331226349,
      "epoch": 11.19,
      "learning_rate": 6.06948804866165e-05,
      "loss": 0.3124,
      "step": 30970,
      "task_loss": 0.5206983685493469
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25989753007888794,
      "epoch": 11.2,
      "learning_rate": 6.0640849841976206e-05,
      "loss": 0.3126,
      "step": 30980,
      "task_loss": 0.25708910822868347
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2514250874519348,
      "epoch": 11.2,
      "learning_rate": 6.058668698449197e-05,
      "loss": 0.3122,
      "step": 30990,
      "task_loss": 0.6891814470291138
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25748398900032043,
      "epoch": 11.2,
      "learning_rate": 6.0532392193445684e-05,
      "loss": 0.3463,
      "step": 31000,
      "task_loss": 0.2810893654823303
    },
    {
      "epoch": 11.2,
      "eval_exact_match": 82.61116367076632,
      "eval_f1": 89.39437638190044,
      "step": 31000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3203161358833313,
      "epoch": 11.21,
      "learning_rate": 6.047796574879955e-05,
      "loss": 0.3683,
      "step": 31010,
      "task_loss": 0.6424096822738647
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2515981197357178,
      "epoch": 11.21,
      "learning_rate": 6.0423407931194616e-05,
      "loss": 0.3159,
      "step": 31020,
      "task_loss": 0.38904130458831787
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21264244616031647,
      "epoch": 11.21,
      "learning_rate": 6.036871902194934e-05,
      "loss": 0.3023,
      "step": 31030,
      "task_loss": 0.4546867609024048
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36801475286483765,
      "epoch": 11.22,
      "learning_rate": 6.031389930305813e-05,
      "loss": 0.3127,
      "step": 31040,
      "task_loss": 0.9397456645965576
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2525968551635742,
      "epoch": 11.22,
      "learning_rate": 6.025894905718988e-05,
      "loss": 0.3171,
      "step": 31050,
      "task_loss": 0.3194689154624939
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3747779130935669,
      "epoch": 11.23,
      "learning_rate": 6.0203868567686544e-05,
      "loss": 0.3413,
      "step": 31060,
      "task_loss": 0.3353426456451416
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3535149097442627,
      "epoch": 11.23,
      "learning_rate": 6.014865811856164e-05,
      "loss": 0.3619,
      "step": 31070,
      "task_loss": 0.5516100525856018
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26662665605545044,
      "epoch": 11.23,
      "learning_rate": 6.009331799449884e-05,
      "loss": 0.3077,
      "step": 31080,
      "task_loss": 0.58455491065979
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31872040033340454,
      "epoch": 11.24,
      "learning_rate": 6.003784848085037e-05,
      "loss": 0.3497,
      "step": 31090,
      "task_loss": 0.7595039010047913
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.348061740398407,
      "epoch": 11.24,
      "learning_rate": 5.998224986363576e-05,
      "loss": 0.337,
      "step": 31100,
      "task_loss": 0.7282963395118713
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4052058160305023,
      "epoch": 11.24,
      "learning_rate": 5.992652242954014e-05,
      "loss": 0.3312,
      "step": 31110,
      "task_loss": 0.7850685119628906
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22310039401054382,
      "epoch": 11.25,
      "learning_rate": 5.987066646591291e-05,
      "loss": 0.3091,
      "step": 31120,
      "task_loss": 0.4512239396572113
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3533927798271179,
      "epoch": 11.25,
      "learning_rate": 5.9814682260766194e-05,
      "loss": 0.3151,
      "step": 31130,
      "task_loss": 0.6712297201156616
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2797766923904419,
      "epoch": 11.25,
      "learning_rate": 5.975857010277339e-05,
      "loss": 0.3103,
      "step": 31140,
      "task_loss": 0.7752123475074768
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25427502393722534,
      "epoch": 11.26,
      "learning_rate": 5.9702330281267646e-05,
      "loss": 0.3234,
      "step": 31150,
      "task_loss": 0.39408910274505615
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31670671701431274,
      "epoch": 11.26,
      "learning_rate": 5.9645963086240405e-05,
      "loss": 0.3193,
      "step": 31160,
      "task_loss": 0.3765723705291748
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38054102659225464,
      "epoch": 11.26,
      "learning_rate": 5.958946880833986e-05,
      "loss": 0.3457,
      "step": 31170,
      "task_loss": 0.794228196144104
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38272950053215027,
      "epoch": 11.27,
      "learning_rate": 5.953284773886952e-05,
      "loss": 0.3371,
      "step": 31180,
      "task_loss": 0.8740808367729187
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4635079801082611,
      "epoch": 11.27,
      "learning_rate": 5.947610016978664e-05,
      "loss": 0.3692,
      "step": 31190,
      "task_loss": 0.7308275699615479
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5193721055984497,
      "epoch": 11.28,
      "learning_rate": 5.94192263937008e-05,
      "loss": 0.3636,
      "step": 31200,
      "task_loss": 0.47289326786994934
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42167362570762634,
      "epoch": 11.28,
      "learning_rate": 5.936222670387228e-05,
      "loss": 0.3795,
      "step": 31210,
      "task_loss": 0.5789846777915955
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.17998720705509186,
      "epoch": 11.28,
      "learning_rate": 5.930510139421068e-05,
      "loss": 0.3465,
      "step": 31220,
      "task_loss": 0.12794144451618195
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2598450183868408,
      "epoch": 11.29,
      "learning_rate": 5.924785075927328e-05,
      "loss": 0.295,
      "step": 31230,
      "task_loss": 0.22558437287807465
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31792667508125305,
      "epoch": 11.29,
      "learning_rate": 5.919047509426362e-05,
      "loss": 0.3208,
      "step": 31240,
      "task_loss": 0.40455853939056396
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27053987979888916,
      "epoch": 11.29,
      "learning_rate": 5.913297469502991e-05,
      "loss": 0.3305,
      "step": 31250,
      "task_loss": 0.8206726312637329
    },
    {
      "epoch": 11.29,
      "eval_exact_match": 82.97067171239357,
      "eval_f1": 89.6492007370016,
      "step": 31250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3249356746673584,
      "epoch": 11.3,
      "learning_rate": 5.907534985806355e-05,
      "loss": 0.3421,
      "step": 31260,
      "task_loss": 1.1308776140213013
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3523188829421997,
      "epoch": 11.3,
      "learning_rate": 5.901760088049758e-05,
      "loss": 0.3354,
      "step": 31270,
      "task_loss": 0.3868599534034729
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34525448083877563,
      "epoch": 11.3,
      "learning_rate": 5.895972806010516e-05,
      "loss": 0.3391,
      "step": 31280,
      "task_loss": 0.7810451984405518
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.297976553440094,
      "epoch": 11.31,
      "learning_rate": 5.890173169529798e-05,
      "loss": 0.2762,
      "step": 31290,
      "task_loss": 0.4372379779815674
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.252549409866333,
      "epoch": 11.31,
      "learning_rate": 5.884361208512483e-05,
      "loss": 0.3142,
      "step": 31300,
      "task_loss": 0.6390154361724854
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24922886490821838,
      "epoch": 11.32,
      "learning_rate": 5.878536952926994e-05,
      "loss": 0.3699,
      "step": 31310,
      "task_loss": 0.2871250510215759
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20339345932006836,
      "epoch": 11.32,
      "learning_rate": 5.872700432805154e-05,
      "loss": 0.2908,
      "step": 31320,
      "task_loss": 0.5091993808746338
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24334312975406647,
      "epoch": 11.32,
      "learning_rate": 5.866851678242021e-05,
      "loss": 0.3197,
      "step": 31330,
      "task_loss": 0.23363611102104187
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26076406240463257,
      "epoch": 11.33,
      "learning_rate": 5.86099071939574e-05,
      "loss": 0.3054,
      "step": 31340,
      "task_loss": 0.5186704397201538
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20574674010276794,
      "epoch": 11.33,
      "learning_rate": 5.855117586487388e-05,
      "loss": 0.3113,
      "step": 31350,
      "task_loss": 0.47972744703292847
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21276235580444336,
      "epoch": 11.33,
      "learning_rate": 5.849232309800812e-05,
      "loss": 0.3232,
      "step": 31360,
      "task_loss": 0.5652778148651123
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2867165505886078,
      "epoch": 11.34,
      "learning_rate": 5.8433349196824764e-05,
      "loss": 0.3078,
      "step": 31370,
      "task_loss": 0.45484477281570435
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47911810874938965,
      "epoch": 11.34,
      "learning_rate": 5.83742544654131e-05,
      "loss": 0.3217,
      "step": 31380,
      "task_loss": 0.4997098445892334
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2617582082748413,
      "epoch": 11.34,
      "learning_rate": 5.831503920848542e-05,
      "loss": 0.3012,
      "step": 31390,
      "task_loss": 0.45006659626960754
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33783483505249023,
      "epoch": 11.35,
      "learning_rate": 5.825570373137551e-05,
      "loss": 0.3159,
      "step": 31400,
      "task_loss": 0.509252667427063
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29730767011642456,
      "epoch": 11.35,
      "learning_rate": 5.819624834003702e-05,
      "loss": 0.3595,
      "step": 31410,
      "task_loss": 0.6149415373802185
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3289676308631897,
      "epoch": 11.36,
      "learning_rate": 5.8136673341041975e-05,
      "loss": 0.3254,
      "step": 31420,
      "task_loss": 0.471310019493103
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21676236391067505,
      "epoch": 11.36,
      "learning_rate": 5.807697904157908e-05,
      "loss": 0.3152,
      "step": 31430,
      "task_loss": 0.29653164744377136
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3164607584476471,
      "epoch": 11.36,
      "learning_rate": 5.801716574945222e-05,
      "loss": 0.3203,
      "step": 31440,
      "task_loss": 0.6140503883361816
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3576650619506836,
      "epoch": 11.37,
      "learning_rate": 5.795723377307885e-05,
      "loss": 0.3473,
      "step": 31450,
      "task_loss": 0.5120994448661804
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28827202320098877,
      "epoch": 11.37,
      "learning_rate": 5.789718342148839e-05,
      "loss": 0.317,
      "step": 31460,
      "task_loss": 0.6643293499946594
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27452167868614197,
      "epoch": 11.37,
      "learning_rate": 5.783701500432064e-05,
      "loss": 0.3256,
      "step": 31470,
      "task_loss": 0.4244112968444824
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3297208547592163,
      "epoch": 11.38,
      "learning_rate": 5.777672883182419e-05,
      "loss": 0.3106,
      "step": 31480,
      "task_loss": 0.4949456453323364
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28091195225715637,
      "epoch": 11.38,
      "learning_rate": 5.771632521485482e-05,
      "loss": 0.3386,
      "step": 31490,
      "task_loss": 0.4345235526561737
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22229045629501343,
      "epoch": 11.38,
      "learning_rate": 5.76558044648739e-05,
      "loss": 0.3329,
      "step": 31500,
      "task_loss": 0.3153355121612549
    },
    {
      "epoch": 11.38,
      "eval_exact_match": 82.72469252601702,
      "eval_f1": 89.59696763050977,
      "step": 31500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22494566440582275,
      "epoch": 11.39,
      "learning_rate": 5.7595166893946776e-05,
      "loss": 0.325,
      "step": 31510,
      "task_loss": 0.46599242091178894
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.302801251411438,
      "epoch": 11.39,
      "learning_rate": 5.7534412814741126e-05,
      "loss": 0.3226,
      "step": 31520,
      "task_loss": 0.5004099607467651
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24632063508033752,
      "epoch": 11.4,
      "learning_rate": 5.747354254052542e-05,
      "loss": 0.3322,
      "step": 31530,
      "task_loss": 0.31375330686569214
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3572053015232086,
      "epoch": 11.4,
      "learning_rate": 5.741255638516727e-05,
      "loss": 0.313,
      "step": 31540,
      "task_loss": 0.5052705407142639
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2660755217075348,
      "epoch": 11.4,
      "learning_rate": 5.7351454663131803e-05,
      "loss": 0.289,
      "step": 31550,
      "task_loss": 0.6535313129425049
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2776443064212799,
      "epoch": 11.41,
      "learning_rate": 5.729023768948003e-05,
      "loss": 0.3646,
      "step": 31560,
      "task_loss": 0.2539638578891754
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.437635600566864,
      "epoch": 11.41,
      "learning_rate": 5.7228905779867264e-05,
      "loss": 0.3812,
      "step": 31570,
      "task_loss": 0.6456525325775146
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29459071159362793,
      "epoch": 11.41,
      "learning_rate": 5.7167459250541455e-05,
      "loss": 0.4021,
      "step": 31580,
      "task_loss": 0.6220089197158813
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2666949927806854,
      "epoch": 11.42,
      "learning_rate": 5.710589841834156e-05,
      "loss": 0.3197,
      "step": 31590,
      "task_loss": 0.3189229965209961
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3018212914466858,
      "epoch": 11.42,
      "learning_rate": 5.704422360069595e-05,
      "loss": 0.3131,
      "step": 31600,
      "task_loss": 0.5461843013763428
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20376259088516235,
      "epoch": 11.42,
      "learning_rate": 5.69824351156207e-05,
      "loss": 0.3089,
      "step": 31610,
      "task_loss": 0.4694615602493286
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30408358573913574,
      "epoch": 11.43,
      "learning_rate": 5.692053328171803e-05,
      "loss": 0.3011,
      "step": 31620,
      "task_loss": 0.26619333028793335
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2672536075115204,
      "epoch": 11.43,
      "learning_rate": 5.685851841817462e-05,
      "loss": 0.3006,
      "step": 31630,
      "task_loss": 0.3129875063896179
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39254266023635864,
      "epoch": 11.43,
      "learning_rate": 5.679639084475993e-05,
      "loss": 0.3127,
      "step": 31640,
      "task_loss": 0.5990224480628967
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3391401171684265,
      "epoch": 11.44,
      "learning_rate": 5.6734150881824656e-05,
      "loss": 0.3575,
      "step": 31650,
      "task_loss": 0.7453941106796265
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38718274235725403,
      "epoch": 11.44,
      "learning_rate": 5.667179885029895e-05,
      "loss": 0.3365,
      "step": 31660,
      "task_loss": 0.9288395047187805
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43893879652023315,
      "epoch": 11.45,
      "learning_rate": 5.660933507169086e-05,
      "loss": 0.3261,
      "step": 31670,
      "task_loss": 0.6180750131607056
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32078641653060913,
      "epoch": 11.45,
      "learning_rate": 5.654675986808465e-05,
      "loss": 0.3462,
      "step": 31680,
      "task_loss": 0.8182356953620911
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32552042603492737,
      "epoch": 11.45,
      "learning_rate": 5.6484073562139083e-05,
      "loss": 0.3571,
      "step": 31690,
      "task_loss": 0.369057297706604
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29520851373672485,
      "epoch": 11.46,
      "learning_rate": 5.642127647708586e-05,
      "loss": 0.3246,
      "step": 31700,
      "task_loss": 0.45395147800445557
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3191979229450226,
      "epoch": 11.46,
      "learning_rate": 5.635836893672784e-05,
      "loss": 0.3323,
      "step": 31710,
      "task_loss": 0.4409838318824768
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22453919053077698,
      "epoch": 11.46,
      "learning_rate": 5.6295351265437475e-05,
      "loss": 0.3511,
      "step": 31720,
      "task_loss": 0.3913407623767853
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24045340716838837,
      "epoch": 11.47,
      "learning_rate": 5.6232223788155074e-05,
      "loss": 0.3016,
      "step": 31730,
      "task_loss": 0.4056016206741333
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28492245078086853,
      "epoch": 11.47,
      "learning_rate": 5.616898683038712e-05,
      "loss": 0.3164,
      "step": 31740,
      "task_loss": 0.5923171043395996
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22334977984428406,
      "epoch": 11.47,
      "learning_rate": 5.610564071820462e-05,
      "loss": 0.3048,
      "step": 31750,
      "task_loss": 0.5784961581230164
    },
    {
      "epoch": 11.47,
      "eval_exact_match": 82.30842005676443,
      "eval_f1": 89.281748835677,
      "step": 31750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25941359996795654,
      "epoch": 11.48,
      "learning_rate": 5.6042185778241474e-05,
      "loss": 0.3379,
      "step": 31760,
      "task_loss": 0.6838859915733337
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30071139335632324,
      "epoch": 11.48,
      "learning_rate": 5.5978622337692656e-05,
      "loss": 0.3302,
      "step": 31770,
      "task_loss": 0.4620378911495209
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2648431062698364,
      "epoch": 11.49,
      "learning_rate": 5.5914950724312644e-05,
      "loss": 0.3303,
      "step": 31780,
      "task_loss": 0.5686721801757812
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34629106521606445,
      "epoch": 11.49,
      "learning_rate": 5.5851171266413694e-05,
      "loss": 0.323,
      "step": 31790,
      "task_loss": 0.44669267535209656
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2935123145580292,
      "epoch": 11.49,
      "learning_rate": 5.578728429286414e-05,
      "loss": 0.3134,
      "step": 31800,
      "task_loss": 0.4752408564090729
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28745779395103455,
      "epoch": 11.5,
      "learning_rate": 5.5723290133086686e-05,
      "loss": 0.3296,
      "step": 31810,
      "task_loss": 0.6825571060180664
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3426585793495178,
      "epoch": 11.5,
      "learning_rate": 5.5659189117056755e-05,
      "loss": 0.317,
      "step": 31820,
      "task_loss": 0.5306519269943237
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2709750533103943,
      "epoch": 11.5,
      "learning_rate": 5.5594981575300746e-05,
      "loss": 0.3061,
      "step": 31830,
      "task_loss": 0.470102995634079
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2657717168331146,
      "epoch": 11.51,
      "learning_rate": 5.5530667838894306e-05,
      "loss": 0.2997,
      "step": 31840,
      "task_loss": 0.36683228611946106
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3502514958381653,
      "epoch": 11.51,
      "learning_rate": 5.546624823946071e-05,
      "loss": 0.3257,
      "step": 31850,
      "task_loss": 0.5826776623725891
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26688486337661743,
      "epoch": 11.51,
      "learning_rate": 5.5401723109169074e-05,
      "loss": 0.2944,
      "step": 31860,
      "task_loss": 0.4791288375854492
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2857460379600525,
      "epoch": 11.52,
      "learning_rate": 5.5337092780732664e-05,
      "loss": 0.3741,
      "step": 31870,
      "task_loss": 0.541092574596405
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4241938591003418,
      "epoch": 11.52,
      "learning_rate": 5.5272357587407176e-05,
      "loss": 0.3506,
      "step": 31880,
      "task_loss": 0.7306923866271973
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3634362816810608,
      "epoch": 11.53,
      "learning_rate": 5.520751786298905e-05,
      "loss": 0.3177,
      "step": 31890,
      "task_loss": 0.46968650817871094
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2074853479862213,
      "epoch": 11.53,
      "learning_rate": 5.51425739418137e-05,
      "loss": 0.3507,
      "step": 31900,
      "task_loss": 0.3567825257778168
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32588648796081543,
      "epoch": 11.53,
      "learning_rate": 5.507752615875383e-05,
      "loss": 0.3368,
      "step": 31910,
      "task_loss": 0.8394519686698914
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23712632060050964,
      "epoch": 11.54,
      "learning_rate": 5.501237484921767e-05,
      "loss": 0.3403,
      "step": 31920,
      "task_loss": 0.6089857816696167
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4756302237510681,
      "epoch": 11.54,
      "learning_rate": 5.494712034914728e-05,
      "loss": 0.3444,
      "step": 31930,
      "task_loss": 0.49841880798339844
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32165443897247314,
      "epoch": 11.54,
      "learning_rate": 5.488176299501683e-05,
      "loss": 0.3121,
      "step": 31940,
      "task_loss": 0.2868342995643616
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3431640863418579,
      "epoch": 11.55,
      "learning_rate": 5.4816303123830796e-05,
      "loss": 0.3307,
      "step": 31950,
      "task_loss": 0.5774615406990051
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32206475734710693,
      "epoch": 11.55,
      "learning_rate": 5.4750741073122284e-05,
      "loss": 0.3721,
      "step": 31960,
      "task_loss": 0.7175636887550354
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2849663197994232,
      "epoch": 11.55,
      "learning_rate": 5.4685077180951276e-05,
      "loss": 0.3497,
      "step": 31970,
      "task_loss": 0.49566948413848877
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2988675832748413,
      "epoch": 11.56,
      "learning_rate": 5.461931178590289e-05,
      "loss": 0.317,
      "step": 31980,
      "task_loss": 0.7831475734710693
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28928619623184204,
      "epoch": 11.56,
      "learning_rate": 5.4553445227085605e-05,
      "loss": 0.3441,
      "step": 31990,
      "task_loss": 0.38043415546417236
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3057170510292053,
      "epoch": 11.56,
      "learning_rate": 5.4487477844129556e-05,
      "loss": 0.2937,
      "step": 32000,
      "task_loss": 0.5053399801254272
    },
    {
      "epoch": 11.56,
      "eval_exact_match": 82.34626300851467,
      "eval_f1": 89.30547427778144,
      "step": 32000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27672266960144043,
      "epoch": 11.57,
      "learning_rate": 5.442140997718475e-05,
      "loss": 0.3506,
      "step": 32010,
      "task_loss": 0.31775251030921936
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44757071137428284,
      "epoch": 11.57,
      "learning_rate": 5.4355241966919324e-05,
      "loss": 0.3183,
      "step": 32020,
      "task_loss": 0.7788525819778442
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27162885665893555,
      "epoch": 11.58,
      "learning_rate": 5.428897415451778e-05,
      "loss": 0.3686,
      "step": 32030,
      "task_loss": 0.6087551712989807
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3032134175300598,
      "epoch": 11.58,
      "learning_rate": 5.422260688167926e-05,
      "loss": 0.3345,
      "step": 32040,
      "task_loss": 0.33256518840789795
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3117225170135498,
      "epoch": 11.58,
      "learning_rate": 5.415614049061573e-05,
      "loss": 0.3204,
      "step": 32050,
      "task_loss": 0.4947884678840637
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3810727894306183,
      "epoch": 11.59,
      "learning_rate": 5.408957532405025e-05,
      "loss": 0.3141,
      "step": 32060,
      "task_loss": 0.6038593053817749
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30839699506759644,
      "epoch": 11.59,
      "learning_rate": 5.4022911725215214e-05,
      "loss": 0.3569,
      "step": 32070,
      "task_loss": 0.44963526725769043
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2700961232185364,
      "epoch": 11.59,
      "learning_rate": 5.395615003785054e-05,
      "loss": 0.2955,
      "step": 32080,
      "task_loss": 0.42938604950904846
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27627208828926086,
      "epoch": 11.6,
      "learning_rate": 5.388929060620194e-05,
      "loss": 0.3578,
      "step": 32090,
      "task_loss": 0.7172139883041382
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.217061847448349,
      "epoch": 11.6,
      "learning_rate": 5.3822333775019146e-05,
      "loss": 0.3172,
      "step": 32100,
      "task_loss": 0.2939155697822571
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3464398980140686,
      "epoch": 11.6,
      "learning_rate": 5.375527988955407e-05,
      "loss": 0.2948,
      "step": 32110,
      "task_loss": 0.6745935678482056
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26883465051651,
      "epoch": 11.61,
      "learning_rate": 5.36881292955591e-05,
      "loss": 0.3402,
      "step": 32120,
      "task_loss": 0.5124499797821045
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29246842861175537,
      "epoch": 11.61,
      "learning_rate": 5.36208823392853e-05,
      "loss": 0.3084,
      "step": 32130,
      "task_loss": 0.4192900061607361
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34540215134620667,
      "epoch": 11.62,
      "learning_rate": 5.3553539367480557e-05,
      "loss": 0.3292,
      "step": 32140,
      "task_loss": 0.6095007658004761
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2868419885635376,
      "epoch": 11.62,
      "learning_rate": 5.34861007273879e-05,
      "loss": 0.336,
      "step": 32150,
      "task_loss": 0.6032599210739136
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3358624577522278,
      "epoch": 11.62,
      "learning_rate": 5.341856676674362e-05,
      "loss": 0.2955,
      "step": 32160,
      "task_loss": 0.44446879625320435
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3103525638580322,
      "epoch": 11.63,
      "learning_rate": 5.335093783377554e-05,
      "loss": 0.3357,
      "step": 32170,
      "task_loss": 0.4964514672756195
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2463071644306183,
      "epoch": 11.63,
      "learning_rate": 5.328321427720118e-05,
      "loss": 0.328,
      "step": 32180,
      "task_loss": 0.960116982460022
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38493576645851135,
      "epoch": 11.63,
      "learning_rate": 5.321539644622596e-05,
      "loss": 0.3408,
      "step": 32190,
      "task_loss": 0.4280901253223419
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28553837537765503,
      "epoch": 11.64,
      "learning_rate": 5.314748469054143e-05,
      "loss": 0.3255,
      "step": 32200,
      "task_loss": 0.525431752204895
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2808997631072998,
      "epoch": 11.64,
      "learning_rate": 5.307947936032344e-05,
      "loss": 0.3177,
      "step": 32210,
      "task_loss": 0.23566016554832458
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4643576145172119,
      "epoch": 11.64,
      "learning_rate": 5.301138080623034e-05,
      "loss": 0.3763,
      "step": 32220,
      "task_loss": 0.7409594058990479
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2858960032463074,
      "epoch": 11.65,
      "learning_rate": 5.294318937940116e-05,
      "loss": 0.3012,
      "step": 32230,
      "task_loss": 0.6493170261383057
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30860576033592224,
      "epoch": 11.65,
      "learning_rate": 5.287490543145385e-05,
      "loss": 0.3199,
      "step": 32240,
      "task_loss": 0.3269897401332855
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5268182158470154,
      "epoch": 11.66,
      "learning_rate": 5.280652931448339e-05,
      "loss": 0.3442,
      "step": 32250,
      "task_loss": 0.6542600989341736
    },
    {
      "epoch": 11.66,
      "eval_exact_match": 83.08420056764427,
      "eval_f1": 89.76224592743083,
      "step": 32250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33490172028541565,
      "epoch": 11.66,
      "learning_rate": 5.273806138106004e-05,
      "loss": 0.3572,
      "step": 32260,
      "task_loss": 0.6601828932762146
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34004050493240356,
      "epoch": 11.66,
      "learning_rate": 5.266950198422749e-05,
      "loss": 0.3248,
      "step": 32270,
      "task_loss": 0.6301026344299316
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.295710027217865,
      "epoch": 11.67,
      "learning_rate": 5.260085147750104e-05,
      "loss": 0.315,
      "step": 32280,
      "task_loss": 0.3331633508205414
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35243725776672363,
      "epoch": 11.67,
      "learning_rate": 5.253211021486578e-05,
      "loss": 0.3147,
      "step": 32290,
      "task_loss": 0.6616904735565186
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30269598960876465,
      "epoch": 11.67,
      "learning_rate": 5.246327855077479e-05,
      "loss": 0.3256,
      "step": 32300,
      "task_loss": 0.541825532913208
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28160130977630615,
      "epoch": 11.68,
      "learning_rate": 5.239435684014727e-05,
      "loss": 0.3064,
      "step": 32310,
      "task_loss": 0.45904257893562317
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46022456884384155,
      "epoch": 11.68,
      "learning_rate": 5.232534543836673e-05,
      "loss": 0.3411,
      "step": 32320,
      "task_loss": 0.5846478939056396
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2371748834848404,
      "epoch": 11.68,
      "learning_rate": 5.225624470127917e-05,
      "loss": 0.33,
      "step": 32330,
      "task_loss": 0.49647220969200134
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2978493571281433,
      "epoch": 11.69,
      "learning_rate": 5.218705498519123e-05,
      "loss": 0.3187,
      "step": 32340,
      "task_loss": 0.6566059589385986
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25985074043273926,
      "epoch": 11.69,
      "learning_rate": 5.211777664686834e-05,
      "loss": 0.3058,
      "step": 32350,
      "task_loss": 0.6025593280792236
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30825042724609375,
      "epoch": 11.69,
      "learning_rate": 5.2048410043532935e-05,
      "loss": 0.3376,
      "step": 32360,
      "task_loss": 0.31907224655151367
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3601059317588806,
      "epoch": 11.7,
      "learning_rate": 5.197895553286251e-05,
      "loss": 0.3641,
      "step": 32370,
      "task_loss": 0.500454843044281
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25591498613357544,
      "epoch": 11.7,
      "learning_rate": 5.190941347298791e-05,
      "loss": 0.34,
      "step": 32380,
      "task_loss": 0.4937874972820282
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29595816135406494,
      "epoch": 11.71,
      "learning_rate": 5.1839784222491365e-05,
      "loss": 0.3345,
      "step": 32390,
      "task_loss": 0.30738627910614014
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26647788286209106,
      "epoch": 11.71,
      "learning_rate": 5.1770068140404696e-05,
      "loss": 0.351,
      "step": 32400,
      "task_loss": 0.4921760559082031
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.286870539188385,
      "epoch": 11.71,
      "learning_rate": 5.170026558620748e-05,
      "loss": 0.2885,
      "step": 32410,
      "task_loss": 0.5846974849700928
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3173612952232361,
      "epoch": 11.72,
      "learning_rate": 5.1630376919825164e-05,
      "loss": 0.3107,
      "step": 32420,
      "task_loss": 0.6225560307502747
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31489187479019165,
      "epoch": 11.72,
      "learning_rate": 5.15604025016272e-05,
      "loss": 0.3051,
      "step": 32430,
      "task_loss": 0.4998015761375427
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3452304005622864,
      "epoch": 11.72,
      "learning_rate": 5.149034269242522e-05,
      "loss": 0.3372,
      "step": 32440,
      "task_loss": 0.6053261756896973
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26079899072647095,
      "epoch": 11.73,
      "learning_rate": 5.1420197853471154e-05,
      "loss": 0.3191,
      "step": 32450,
      "task_loss": 0.5432778596878052
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3432601988315582,
      "epoch": 11.73,
      "learning_rate": 5.13499683464554e-05,
      "loss": 0.3461,
      "step": 32460,
      "task_loss": 0.9408321976661682
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22599393129348755,
      "epoch": 11.73,
      "learning_rate": 5.1279654533504885e-05,
      "loss": 0.2994,
      "step": 32470,
      "task_loss": 0.27330297231674194
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33337026834487915,
      "epoch": 11.74,
      "learning_rate": 5.12092567771813e-05,
      "loss": 0.3018,
      "step": 32480,
      "task_loss": 0.4547075927257538
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3493223786354065,
      "epoch": 11.74,
      "learning_rate": 5.1138775440479126e-05,
      "loss": 0.325,
      "step": 32490,
      "task_loss": 0.41317543387413025
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4307718575000763,
      "epoch": 11.75,
      "learning_rate": 5.106821088682384e-05,
      "loss": 0.3321,
      "step": 32500,
      "task_loss": 1.1518049240112305
    },
    {
      "epoch": 11.75,
      "eval_exact_match": 82.68684957426679,
      "eval_f1": 89.72324307540484,
      "step": 32500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4119787812232971,
      "epoch": 11.75,
      "learning_rate": 5.0997563480070024e-05,
      "loss": 0.3619,
      "step": 32510,
      "task_loss": 0.5112029314041138
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.255260705947876,
      "epoch": 11.75,
      "learning_rate": 5.093391027566244e-05,
      "loss": 0.3288,
      "step": 32520,
      "task_loss": 1.0041379928588867
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3400477170944214,
      "epoch": 11.76,
      "learning_rate": 5.086310645196928e-05,
      "loss": 0.3147,
      "step": 32530,
      "task_loss": 0.41630157828330994
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.293648362159729,
      "epoch": 11.76,
      "learning_rate": 5.079222083276504e-05,
      "loss": 0.3078,
      "step": 32540,
      "task_loss": 0.4691644608974457
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3460580110549927,
      "epoch": 11.76,
      "learning_rate": 5.072125378355978e-05,
      "loss": 0.3015,
      "step": 32550,
      "task_loss": 0.6114804744720459
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2778604030609131,
      "epoch": 11.77,
      "learning_rate": 5.0650205670283475e-05,
      "loss": 0.2822,
      "step": 32560,
      "task_loss": 0.5171606540679932
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2882919907569885,
      "epoch": 11.77,
      "learning_rate": 5.057907685928408e-05,
      "loss": 0.3564,
      "step": 32570,
      "task_loss": 0.25201505422592163
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2875009775161743,
      "epoch": 11.77,
      "learning_rate": 5.050786771732567e-05,
      "loss": 0.3119,
      "step": 32580,
      "task_loss": 0.5717591047286987
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3220732808113098,
      "epoch": 11.78,
      "learning_rate": 5.043657861158653e-05,
      "loss": 0.3817,
      "step": 32590,
      "task_loss": 0.6619873046875
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29478880763053894,
      "epoch": 11.78,
      "learning_rate": 5.036520990965726e-05,
      "loss": 0.3383,
      "step": 32600,
      "task_loss": 0.4968605041503906
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2925856113433838,
      "epoch": 11.79,
      "learning_rate": 5.029376197953888e-05,
      "loss": 0.3088,
      "step": 32610,
      "task_loss": 0.5669705271720886
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3187752366065979,
      "epoch": 11.79,
      "learning_rate": 5.022223518964095e-05,
      "loss": 0.3481,
      "step": 32620,
      "task_loss": 0.3745642900466919
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3223860263824463,
      "epoch": 11.79,
      "learning_rate": 5.015062990877964e-05,
      "loss": 0.3286,
      "step": 32630,
      "task_loss": 0.4044720530509949
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29211077094078064,
      "epoch": 11.8,
      "learning_rate": 5.007894650617588e-05,
      "loss": 0.3262,
      "step": 32640,
      "task_loss": 0.3009093999862671
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28116658329963684,
      "epoch": 11.8,
      "learning_rate": 5.0007185351453374e-05,
      "loss": 0.3063,
      "step": 32650,
      "task_loss": 0.6478732824325562
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28009843826293945,
      "epoch": 11.8,
      "learning_rate": 4.9935346814636785e-05,
      "loss": 0.3234,
      "step": 32660,
      "task_loss": 0.6960561275482178
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42953410744667053,
      "epoch": 11.81,
      "learning_rate": 4.9863431266149745e-05,
      "loss": 0.3186,
      "step": 32670,
      "task_loss": 0.5555198192596436
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36337924003601074,
      "epoch": 11.81,
      "learning_rate": 4.979143907681301e-05,
      "loss": 0.3635,
      "step": 32680,
      "task_loss": 0.5978571176528931
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24997106194496155,
      "epoch": 11.81,
      "learning_rate": 4.9719370617842544e-05,
      "loss": 0.3219,
      "step": 32690,
      "task_loss": 0.2275729477405548
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35133838653564453,
      "epoch": 11.82,
      "learning_rate": 4.964722626084752e-05,
      "loss": 0.3278,
      "step": 32700,
      "task_loss": 0.3992236852645874
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23231764137744904,
      "epoch": 11.82,
      "learning_rate": 4.9575006377828535e-05,
      "loss": 0.3368,
      "step": 32710,
      "task_loss": 0.3263181447982788
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21223102509975433,
      "epoch": 11.83,
      "learning_rate": 4.9502711341175553e-05,
      "loss": 0.2913,
      "step": 32720,
      "task_loss": 0.4191708564758301
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41620898246765137,
      "epoch": 11.83,
      "learning_rate": 4.9430341523666124e-05,
      "loss": 0.3523,
      "step": 32730,
      "task_loss": 0.41136446595191956
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40111666917800903,
      "epoch": 11.83,
      "learning_rate": 4.935789729846335e-05,
      "loss": 0.3663,
      "step": 32740,
      "task_loss": 0.567755937576294
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2847909927368164,
      "epoch": 11.84,
      "learning_rate": 4.928537903911402e-05,
      "loss": 0.326,
      "step": 32750,
      "task_loss": 0.5023891925811768
    },
    {
      "epoch": 11.84,
      "eval_exact_match": 83.04635761589404,
      "eval_f1": 89.84907706561316,
      "step": 32750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30441999435424805,
      "epoch": 11.84,
      "learning_rate": 4.921278711954666e-05,
      "loss": 0.3179,
      "step": 32760,
      "task_loss": 0.5857506990432739
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24043527245521545,
      "epoch": 11.84,
      "learning_rate": 4.9140121914069626e-05,
      "loss": 0.3814,
      "step": 32770,
      "task_loss": 0.7202674150466919
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3235805332660675,
      "epoch": 11.85,
      "learning_rate": 4.9067383797369146e-05,
      "loss": 0.331,
      "step": 32780,
      "task_loss": 1.0481606721878052
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23604023456573486,
      "epoch": 11.85,
      "learning_rate": 4.899457314450742e-05,
      "loss": 0.3057,
      "step": 32790,
      "task_loss": 0.3260385990142822
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31892240047454834,
      "epoch": 11.85,
      "learning_rate": 4.8921690330920637e-05,
      "loss": 0.3378,
      "step": 32800,
      "task_loss": 0.5068680047988892
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30003371834754944,
      "epoch": 11.86,
      "learning_rate": 4.884873573241711e-05,
      "loss": 0.3413,
      "step": 32810,
      "task_loss": 0.8090519905090332
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3101547956466675,
      "epoch": 11.86,
      "learning_rate": 4.8775709725175277e-05,
      "loss": 0.3439,
      "step": 32820,
      "task_loss": 0.572382926940918
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2574950158596039,
      "epoch": 11.86,
      "learning_rate": 4.870261268574178e-05,
      "loss": 0.3328,
      "step": 32830,
      "task_loss": 0.7044030427932739
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27755796909332275,
      "epoch": 11.87,
      "learning_rate": 4.8629444991029544e-05,
      "loss": 0.3105,
      "step": 32840,
      "task_loss": 0.3686292767524719
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2532276511192322,
      "epoch": 11.87,
      "learning_rate": 4.8556207018315793e-05,
      "loss": 0.3486,
      "step": 32850,
      "task_loss": 0.602851390838623
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23359063267707825,
      "epoch": 11.88,
      "learning_rate": 4.848289914524017e-05,
      "loss": 0.3187,
      "step": 32860,
      "task_loss": 0.5139017105102539
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33160653710365295,
      "epoch": 11.88,
      "learning_rate": 4.840952174980269e-05,
      "loss": 0.3345,
      "step": 32870,
      "task_loss": 0.6995959281921387
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3129425644874573,
      "epoch": 11.88,
      "learning_rate": 4.8336075210361884e-05,
      "loss": 0.3458,
      "step": 32880,
      "task_loss": 0.4224226474761963
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3381350040435791,
      "epoch": 11.89,
      "learning_rate": 4.826255990563282e-05,
      "loss": 0.3631,
      "step": 32890,
      "task_loss": 0.5300081968307495
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32694685459136963,
      "epoch": 11.89,
      "learning_rate": 4.8188976214685125e-05,
      "loss": 0.3714,
      "step": 32900,
      "task_loss": 0.3771377205848694
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35713815689086914,
      "epoch": 11.89,
      "learning_rate": 4.811532451694104e-05,
      "loss": 0.3172,
      "step": 32910,
      "task_loss": 0.4956726133823395
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2920774519443512,
      "epoch": 11.9,
      "learning_rate": 4.8041605192173505e-05,
      "loss": 0.3005,
      "step": 32920,
      "task_loss": 0.30127841234207153
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4189726710319519,
      "epoch": 11.9,
      "learning_rate": 4.796781862050413e-05,
      "loss": 0.3277,
      "step": 32930,
      "task_loss": 0.741553544998169
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3866938054561615,
      "epoch": 11.9,
      "learning_rate": 4.789396518240132e-05,
      "loss": 0.3125,
      "step": 32940,
      "task_loss": 0.5937274694442749
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26409783959388733,
      "epoch": 11.91,
      "learning_rate": 4.782004525867819e-05,
      "loss": 0.3068,
      "step": 32950,
      "task_loss": 0.2822565734386444
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2546120285987854,
      "epoch": 11.91,
      "learning_rate": 4.774605923049076e-05,
      "loss": 0.3082,
      "step": 32960,
      "task_loss": 0.7707274556159973
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5309508442878723,
      "epoch": 11.92,
      "learning_rate": 4.767200747933586e-05,
      "loss": 0.3719,
      "step": 32970,
      "task_loss": 0.5996285676956177
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.350615531206131,
      "epoch": 11.92,
      "learning_rate": 4.759789038704922e-05,
      "loss": 0.3197,
      "step": 32980,
      "task_loss": 0.44299453496932983
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2523714303970337,
      "epoch": 11.92,
      "learning_rate": 4.7523708335803504e-05,
      "loss": 0.3624,
      "step": 32990,
      "task_loss": 0.663590669631958
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24829280376434326,
      "epoch": 11.93,
      "learning_rate": 4.7449461708106276e-05,
      "loss": 0.294,
      "step": 33000,
      "task_loss": 0.3095482289791107
    },
    {
      "epoch": 11.93,
      "eval_exact_match": 82.90444654683066,
      "eval_f1": 89.76255523912135,
      "step": 33000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3185913562774658,
      "epoch": 11.93,
      "learning_rate": 4.7375150886798146e-05,
      "loss": 0.2944,
      "step": 33010,
      "task_loss": 0.35980331897735596
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28336402773857117,
      "epoch": 11.93,
      "learning_rate": 4.73007762550507e-05,
      "loss": 0.3057,
      "step": 33020,
      "task_loss": 0.38741469383239746
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.329383909702301,
      "epoch": 11.94,
      "learning_rate": 4.722633819636453e-05,
      "loss": 0.3433,
      "step": 33030,
      "task_loss": 0.597511351108551
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24982626736164093,
      "epoch": 11.94,
      "learning_rate": 4.715183709456731e-05,
      "loss": 0.3088,
      "step": 33040,
      "task_loss": 0.361911416053772
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40126579999923706,
      "epoch": 11.94,
      "learning_rate": 4.707727333381177e-05,
      "loss": 0.3145,
      "step": 33050,
      "task_loss": 0.30749595165252686
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2451639622449875,
      "epoch": 11.95,
      "learning_rate": 4.7002647298573725e-05,
      "loss": 0.3047,
      "step": 33060,
      "task_loss": 0.4949933886528015
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22195953130722046,
      "epoch": 11.95,
      "learning_rate": 4.692795937365013e-05,
      "loss": 0.3003,
      "step": 33070,
      "task_loss": 0.4203670620918274
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2641502618789673,
      "epoch": 11.96,
      "learning_rate": 4.685320994415701e-05,
      "loss": 0.3094,
      "step": 33080,
      "task_loss": 0.35039326548576355
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39277034997940063,
      "epoch": 11.96,
      "learning_rate": 4.677839939552759e-05,
      "loss": 0.3251,
      "step": 33090,
      "task_loss": 0.46645432710647583
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32009559869766235,
      "epoch": 11.96,
      "learning_rate": 4.670352811351019e-05,
      "loss": 0.3152,
      "step": 33100,
      "task_loss": 0.6483212113380432
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2656344175338745,
      "epoch": 11.97,
      "learning_rate": 4.662859648416633e-05,
      "loss": 0.3348,
      "step": 33110,
      "task_loss": 0.69025719165802
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3949262499809265,
      "epoch": 11.97,
      "learning_rate": 4.655360489386869e-05,
      "loss": 0.3289,
      "step": 33120,
      "task_loss": 0.7546541690826416
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32530757784843445,
      "epoch": 11.97,
      "learning_rate": 4.647855372929912e-05,
      "loss": 0.3205,
      "step": 33130,
      "task_loss": 0.35394996404647827
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21367156505584717,
      "epoch": 11.98,
      "learning_rate": 4.640344337744667e-05,
      "loss": 0.3238,
      "step": 33140,
      "task_loss": 0.46600231528282166
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2911337614059448,
      "epoch": 11.98,
      "learning_rate": 4.6328274225605556e-05,
      "loss": 0.3662,
      "step": 33150,
      "task_loss": 0.4676685035228729
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28591835498809814,
      "epoch": 11.98,
      "learning_rate": 4.625304666137321e-05,
      "loss": 0.3161,
      "step": 33160,
      "task_loss": 0.36528101563453674
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42281341552734375,
      "epoch": 11.99,
      "learning_rate": 4.617776107264826e-05,
      "loss": 0.3452,
      "step": 33170,
      "task_loss": 0.449604868888855
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35083454847335815,
      "epoch": 11.99,
      "learning_rate": 4.6102417847628494e-05,
      "loss": 0.3073,
      "step": 33180,
      "task_loss": 0.4036543667316437
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2771892547607422,
      "epoch": 11.99,
      "learning_rate": 4.602701737480895e-05,
      "loss": 0.3069,
      "step": 33190,
      "task_loss": 0.48936066031455994
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30712735652923584,
      "epoch": 12.0,
      "learning_rate": 4.595156004297978e-05,
      "loss": 0.2947,
      "step": 33200,
      "task_loss": 0.6743454337120056
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2192678451538086,
      "epoch": 12.0,
      "learning_rate": 4.5876046241224384e-05,
      "loss": 0.3159,
      "step": 33210,
      "task_loss": 0.6000691652297974
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30531418323516846,
      "epoch": 12.01,
      "learning_rate": 4.580047635891733e-05,
      "loss": 0.2964,
      "step": 33220,
      "task_loss": 0.49903973937034607
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2905973196029663,
      "epoch": 12.01,
      "learning_rate": 4.572485078572231e-05,
      "loss": 0.2861,
      "step": 33230,
      "task_loss": 0.5353321433067322
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25237756967544556,
      "epoch": 12.01,
      "learning_rate": 4.564916991159023e-05,
      "loss": 0.257,
      "step": 33240,
      "task_loss": 0.4064103364944458
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27284255623817444,
      "epoch": 12.02,
      "learning_rate": 4.5573434126757125e-05,
      "loss": 0.2949,
      "step": 33250,
      "task_loss": 0.46959832310676575
    },
    {
      "epoch": 12.02,
      "eval_exact_match": 83.06527909176916,
      "eval_f1": 89.86909660147846,
      "step": 33250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3078652620315552,
      "epoch": 12.02,
      "learning_rate": 4.5497643821742164e-05,
      "loss": 0.2734,
      "step": 33260,
      "task_loss": 0.5497879981994629
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3155045509338379,
      "epoch": 12.02,
      "learning_rate": 4.5421799387345656e-05,
      "loss": 0.2881,
      "step": 33270,
      "task_loss": 0.45438718795776367
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28334158658981323,
      "epoch": 12.03,
      "learning_rate": 4.534590121464699e-05,
      "loss": 0.2867,
      "step": 33280,
      "task_loss": 0.575757622718811
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22371214628219604,
      "epoch": 12.03,
      "learning_rate": 4.526994969500268e-05,
      "loss": 0.2901,
      "step": 33290,
      "task_loss": 0.511568546295166
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2883744239807129,
      "epoch": 12.03,
      "learning_rate": 4.5193945220044304e-05,
      "loss": 0.2849,
      "step": 33300,
      "task_loss": 0.4415782392024994
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3361637592315674,
      "epoch": 12.04,
      "learning_rate": 4.5117888181676476e-05,
      "loss": 0.3318,
      "step": 33310,
      "task_loss": 0.5499709844589233
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31543105840682983,
      "epoch": 12.04,
      "learning_rate": 4.5041778972074884e-05,
      "loss": 0.3009,
      "step": 33320,
      "task_loss": 0.6422604322433472
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34402894973754883,
      "epoch": 12.05,
      "learning_rate": 4.4965617983684185e-05,
      "loss": 0.3017,
      "step": 33330,
      "task_loss": 0.5380134582519531
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2677575945854187,
      "epoch": 12.05,
      "learning_rate": 4.4889405609216065e-05,
      "loss": 0.313,
      "step": 33340,
      "task_loss": 0.6588090658187866
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26906073093414307,
      "epoch": 12.05,
      "learning_rate": 4.481314224164713e-05,
      "loss": 0.2905,
      "step": 33350,
      "task_loss": 0.5632128715515137
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25012487173080444,
      "epoch": 12.06,
      "learning_rate": 4.473682827421697e-05,
      "loss": 0.3141,
      "step": 33360,
      "task_loss": 0.5552922487258911
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2697087526321411,
      "epoch": 12.06,
      "learning_rate": 4.4660464100426054e-05,
      "loss": 0.298,
      "step": 33370,
      "task_loss": 0.8239359855651855
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24572470784187317,
      "epoch": 12.06,
      "learning_rate": 4.458405011403374e-05,
      "loss": 0.3245,
      "step": 33380,
      "task_loss": 0.5354356169700623
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27262991666793823,
      "epoch": 12.07,
      "learning_rate": 4.450758670905624e-05,
      "loss": 0.2973,
      "step": 33390,
      "task_loss": 0.6752634048461914
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3405141830444336,
      "epoch": 12.07,
      "learning_rate": 4.4431074279764585e-05,
      "loss": 0.2833,
      "step": 33400,
      "task_loss": 0.645380973815918
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26243701577186584,
      "epoch": 12.07,
      "learning_rate": 4.4354513220682575e-05,
      "loss": 0.2856,
      "step": 33410,
      "task_loss": 0.2894069254398346
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2766551673412323,
      "epoch": 12.08,
      "learning_rate": 4.427790392658479e-05,
      "loss": 0.2967,
      "step": 33420,
      "task_loss": 0.4259956181049347
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.295553982257843,
      "epoch": 12.08,
      "learning_rate": 4.4201246792494514e-05,
      "loss": 0.2812,
      "step": 33430,
      "task_loss": 0.4812880754470825
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27736884355545044,
      "epoch": 12.09,
      "learning_rate": 4.41245422136817e-05,
      "loss": 0.2961,
      "step": 33440,
      "task_loss": 0.7917148470878601
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29041779041290283,
      "epoch": 12.09,
      "learning_rate": 4.404779058566094e-05,
      "loss": 0.2964,
      "step": 33450,
      "task_loss": 0.6089922189712524
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22682823240756989,
      "epoch": 12.09,
      "learning_rate": 4.397099230418947e-05,
      "loss": 0.2697,
      "step": 33460,
      "task_loss": 0.293168842792511
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3263562321662903,
      "epoch": 12.1,
      "learning_rate": 4.3894147765265014e-05,
      "loss": 0.2857,
      "step": 33470,
      "task_loss": 0.5664174556732178
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24178797006607056,
      "epoch": 12.1,
      "learning_rate": 4.381725736512387e-05,
      "loss": 0.304,
      "step": 33480,
      "task_loss": 0.3365327715873718
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22949285805225372,
      "epoch": 12.1,
      "learning_rate": 4.374032150023881e-05,
      "loss": 0.2844,
      "step": 33490,
      "task_loss": 0.2510908842086792
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3991518020629883,
      "epoch": 12.11,
      "learning_rate": 4.366334056731701e-05,
      "loss": 0.3076,
      "step": 33500,
      "task_loss": 0.7375933527946472
    },
    {
      "epoch": 12.11,
      "eval_exact_match": 82.7057710501419,
      "eval_f1": 89.65179404829924,
      "step": 33500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24318742752075195,
      "epoch": 12.11,
      "learning_rate": 4.358631496329804e-05,
      "loss": 0.2705,
      "step": 33510,
      "task_loss": 0.706390917301178
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3634289801120758,
      "epoch": 12.11,
      "learning_rate": 4.350924508535184e-05,
      "loss": 0.2967,
      "step": 33520,
      "task_loss": 0.8388668298721313
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2415129542350769,
      "epoch": 12.12,
      "learning_rate": 4.3432131330876594e-05,
      "loss": 0.2609,
      "step": 33530,
      "task_loss": 0.35680413246154785
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23396481573581696,
      "epoch": 12.12,
      "learning_rate": 4.335497409749677e-05,
      "loss": 0.3141,
      "step": 33540,
      "task_loss": 0.7150663137435913
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3096352219581604,
      "epoch": 12.13,
      "learning_rate": 4.327777378306099e-05,
      "loss": 0.2872,
      "step": 33550,
      "task_loss": 0.46862053871154785
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2673993408679962,
      "epoch": 12.13,
      "learning_rate": 4.3200530785640036e-05,
      "loss": 0.2759,
      "step": 33560,
      "task_loss": 0.5018470883369446
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21115058660507202,
      "epoch": 12.13,
      "learning_rate": 4.312324550352479e-05,
      "loss": 0.2761,
      "step": 33570,
      "task_loss": 0.22329533100128174
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3349929749965668,
      "epoch": 12.14,
      "learning_rate": 4.3045918335224154e-05,
      "loss": 0.3091,
      "step": 33580,
      "task_loss": 0.651825487613678
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24410584568977356,
      "epoch": 12.14,
      "learning_rate": 4.296854967946301e-05,
      "loss": 0.2903,
      "step": 33590,
      "task_loss": 0.5682234764099121
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3624010980129242,
      "epoch": 12.14,
      "learning_rate": 4.289113993518015e-05,
      "loss": 0.2888,
      "step": 33600,
      "task_loss": 0.6332154870033264
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25731930136680603,
      "epoch": 12.15,
      "learning_rate": 4.2813689501526246e-05,
      "loss": 0.3136,
      "step": 33610,
      "task_loss": 0.5591039061546326
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.256285697221756,
      "epoch": 12.15,
      "learning_rate": 4.2736198777861805e-05,
      "loss": 0.285,
      "step": 33620,
      "task_loss": 0.45655739307403564
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21870087087154388,
      "epoch": 12.15,
      "learning_rate": 4.265866816375503e-05,
      "loss": 0.2756,
      "step": 33630,
      "task_loss": 0.47334593534469604
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20090892910957336,
      "epoch": 12.16,
      "learning_rate": 4.258109805897985e-05,
      "loss": 0.2775,
      "step": 33640,
      "task_loss": 0.42821335792541504
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28893759846687317,
      "epoch": 12.16,
      "learning_rate": 4.2503488863513794e-05,
      "loss": 0.2808,
      "step": 33650,
      "task_loss": 0.40764930844306946
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21737822890281677,
      "epoch": 12.16,
      "learning_rate": 4.242584097753599e-05,
      "loss": 0.266,
      "step": 33660,
      "task_loss": 0.5878552198410034
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2558310329914093,
      "epoch": 12.17,
      "learning_rate": 4.2348154801425044e-05,
      "loss": 0.2763,
      "step": 33670,
      "task_loss": 0.48859918117523193
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26445093750953674,
      "epoch": 12.17,
      "learning_rate": 4.227043073575699e-05,
      "loss": 0.2772,
      "step": 33680,
      "task_loss": 0.4119749367237091
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23118102550506592,
      "epoch": 12.18,
      "learning_rate": 4.219266918130327e-05,
      "loss": 0.2699,
      "step": 33690,
      "task_loss": 0.5234097242355347
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22412827610969543,
      "epoch": 12.18,
      "learning_rate": 4.211487053902858e-05,
      "loss": 0.2676,
      "step": 33700,
      "task_loss": 0.4862504005432129
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24602150917053223,
      "epoch": 12.18,
      "learning_rate": 4.203703521008891e-05,
      "loss": 0.2841,
      "step": 33710,
      "task_loss": 0.6615269184112549
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2593457102775574,
      "epoch": 12.19,
      "learning_rate": 4.195916359582938e-05,
      "loss": 0.2693,
      "step": 33720,
      "task_loss": 0.8488281965255737
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27883416414260864,
      "epoch": 12.19,
      "learning_rate": 4.188125609778219e-05,
      "loss": 0.3318,
      "step": 33730,
      "task_loss": 0.54170823097229
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2291126549243927,
      "epoch": 12.19,
      "learning_rate": 4.180331311766464e-05,
      "loss": 0.3152,
      "step": 33740,
      "task_loss": 0.2940157651901245
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2450833022594452,
      "epoch": 12.2,
      "learning_rate": 4.1725335057376915e-05,
      "loss": 0.2688,
      "step": 33750,
      "task_loss": 0.35134434700012207
    },
    {
      "epoch": 12.2,
      "eval_exact_match": 83.12204351939451,
      "eval_f1": 89.85422216477365,
      "step": 33750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2731408178806305,
      "epoch": 12.2,
      "learning_rate": 4.164732231900013e-05,
      "loss": 0.2812,
      "step": 33760,
      "task_loss": 0.45388272404670715
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23454083502292633,
      "epoch": 12.2,
      "learning_rate": 4.15692753047942e-05,
      "loss": 0.2674,
      "step": 33770,
      "task_loss": 0.18816009163856506
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25454145669937134,
      "epoch": 12.21,
      "learning_rate": 4.149119441719576e-05,
      "loss": 0.3194,
      "step": 33780,
      "task_loss": 0.46342360973358154
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21274858713150024,
      "epoch": 12.21,
      "learning_rate": 4.141308005881614e-05,
      "loss": 0.2808,
      "step": 33790,
      "task_loss": 0.5964909195899963
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2354692816734314,
      "epoch": 12.22,
      "learning_rate": 4.133493263243922e-05,
      "loss": 0.2998,
      "step": 33800,
      "task_loss": 0.20641979575157166
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19088518619537354,
      "epoch": 12.22,
      "learning_rate": 4.1256752541019415e-05,
      "loss": 0.2725,
      "step": 33810,
      "task_loss": 0.28886133432388306
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3029336631298065,
      "epoch": 12.22,
      "learning_rate": 4.1178540187679585e-05,
      "loss": 0.2913,
      "step": 33820,
      "task_loss": 0.6475502252578735
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2550453841686249,
      "epoch": 12.23,
      "learning_rate": 4.1100295975708904e-05,
      "loss": 0.2773,
      "step": 33830,
      "task_loss": 0.3866943120956421
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26052963733673096,
      "epoch": 12.23,
      "learning_rate": 4.102202030856085e-05,
      "loss": 0.2875,
      "step": 33840,
      "task_loss": 0.4841282367706299
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36123284697532654,
      "epoch": 12.23,
      "learning_rate": 4.0943713589851066e-05,
      "loss": 0.2902,
      "step": 33850,
      "task_loss": 0.39048629999160767
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25504371523857117,
      "epoch": 12.24,
      "learning_rate": 4.086537622335534e-05,
      "loss": 0.2961,
      "step": 33860,
      "task_loss": 0.6866843700408936
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23795445263385773,
      "epoch": 12.24,
      "learning_rate": 4.0787008613007484e-05,
      "loss": 0.2726,
      "step": 33870,
      "task_loss": 0.35257768630981445
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23061011731624603,
      "epoch": 12.24,
      "learning_rate": 4.070861116289723e-05,
      "loss": 0.2754,
      "step": 33880,
      "task_loss": 0.5739815831184387
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35582464933395386,
      "epoch": 12.25,
      "learning_rate": 4.063018427726821e-05,
      "loss": 0.3302,
      "step": 33890,
      "task_loss": 0.5090781450271606
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19990065693855286,
      "epoch": 12.25,
      "learning_rate": 4.05517283605158e-05,
      "loss": 0.2506,
      "step": 33900,
      "task_loss": 0.41246160864830017
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19301462173461914,
      "epoch": 12.26,
      "learning_rate": 4.047324381718511e-05,
      "loss": 0.2695,
      "step": 33910,
      "task_loss": 0.27115631103515625
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2891654968261719,
      "epoch": 12.26,
      "learning_rate": 4.039473105196883e-05,
      "loss": 0.2885,
      "step": 33920,
      "task_loss": 0.4743727445602417
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21699106693267822,
      "epoch": 12.26,
      "learning_rate": 4.031619046970517e-05,
      "loss": 0.2759,
      "step": 33930,
      "task_loss": 0.4584755301475525
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2696276903152466,
      "epoch": 12.27,
      "learning_rate": 4.02376224753758e-05,
      "loss": 0.2789,
      "step": 33940,
      "task_loss": 0.42498350143432617
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3191455602645874,
      "epoch": 12.27,
      "learning_rate": 4.01590274741037e-05,
      "loss": 0.3073,
      "step": 33950,
      "task_loss": 0.5300410985946655
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23330631852149963,
      "epoch": 12.27,
      "learning_rate": 4.008040587115112e-05,
      "loss": 0.2813,
      "step": 33960,
      "task_loss": 0.3184712529182434
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2462092638015747,
      "epoch": 12.28,
      "learning_rate": 4.000175807191752e-05,
      "loss": 0.2358,
      "step": 33970,
      "task_loss": 0.5075340270996094
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27791628241539,
      "epoch": 12.28,
      "learning_rate": 3.992308448193736e-05,
      "loss": 0.275,
      "step": 33980,
      "task_loss": 0.5956826210021973
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28359711170196533,
      "epoch": 12.28,
      "learning_rate": 3.984438550687815e-05,
      "loss": 0.2697,
      "step": 33990,
      "task_loss": 0.415661484003067
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28219592571258545,
      "epoch": 12.29,
      "learning_rate": 3.976566155253826e-05,
      "loss": 0.3357,
      "step": 34000,
      "task_loss": 0.6862603425979614
    },
    {
      "epoch": 12.29,
      "eval_exact_match": 83.07473982970671,
      "eval_f1": 89.81688188992857,
      "step": 34000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23105570673942566,
      "epoch": 12.29,
      "learning_rate": 3.9686913024844855e-05,
      "loss": 0.2561,
      "step": 34010,
      "task_loss": 0.22716933488845825
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23710978031158447,
      "epoch": 12.29,
      "learning_rate": 3.960814032985186e-05,
      "loss": 0.2952,
      "step": 34020,
      "task_loss": 0.43777596950531006
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3221627175807953,
      "epoch": 12.3,
      "learning_rate": 3.952934387373775e-05,
      "loss": 0.3072,
      "step": 34030,
      "task_loss": 0.644108235836029
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31833896040916443,
      "epoch": 12.3,
      "learning_rate": 3.945052406280356e-05,
      "loss": 0.2708,
      "step": 34040,
      "task_loss": 0.3173179626464844
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20409566164016724,
      "epoch": 12.31,
      "learning_rate": 3.937168130347074e-05,
      "loss": 0.3017,
      "step": 34050,
      "task_loss": 0.2745077610015869
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21003592014312744,
      "epoch": 12.31,
      "learning_rate": 3.9292816002279055e-05,
      "loss": 0.2715,
      "step": 34060,
      "task_loss": 0.48984411358833313
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24261227250099182,
      "epoch": 12.31,
      "learning_rate": 3.921392856588455e-05,
      "loss": 0.2858,
      "step": 34070,
      "task_loss": 0.5235458016395569
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2940651476383209,
      "epoch": 12.32,
      "learning_rate": 3.9135019401057337e-05,
      "loss": 0.2886,
      "step": 34080,
      "task_loss": 0.642257571220398
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36813968420028687,
      "epoch": 12.32,
      "learning_rate": 3.9056088914679635e-05,
      "loss": 0.2887,
      "step": 34090,
      "task_loss": 0.4939306974411011
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2847978472709656,
      "epoch": 12.32,
      "learning_rate": 3.897713751374355e-05,
      "loss": 0.275,
      "step": 34100,
      "task_loss": 0.32528945803642273
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23599255084991455,
      "epoch": 12.33,
      "learning_rate": 3.889816560534907e-05,
      "loss": 0.2762,
      "step": 34110,
      "task_loss": 0.4346529543399811
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23587912321090698,
      "epoch": 12.33,
      "learning_rate": 3.881917359670191e-05,
      "loss": 0.3107,
      "step": 34120,
      "task_loss": 0.9650452136993408
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35129642486572266,
      "epoch": 12.33,
      "learning_rate": 3.874016189511141e-05,
      "loss": 0.2857,
      "step": 34130,
      "task_loss": 0.754157543182373
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21880073845386505,
      "epoch": 12.34,
      "learning_rate": 3.8661130907988505e-05,
      "loss": 0.2876,
      "step": 34140,
      "task_loss": 0.7854479551315308
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23840254545211792,
      "epoch": 12.34,
      "learning_rate": 3.85820810428435e-05,
      "loss": 0.2877,
      "step": 34150,
      "task_loss": 0.46722444891929626
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28145772218704224,
      "epoch": 12.35,
      "learning_rate": 3.850301270728412e-05,
      "loss": 0.2828,
      "step": 34160,
      "task_loss": 0.25351595878601074
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24102526903152466,
      "epoch": 12.35,
      "learning_rate": 3.8423926309013267e-05,
      "loss": 0.3016,
      "step": 34170,
      "task_loss": 0.599195659160614
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2735949456691742,
      "epoch": 12.35,
      "learning_rate": 3.834482225582701e-05,
      "loss": 0.2608,
      "step": 34180,
      "task_loss": 0.5106882452964783
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30498093366622925,
      "epoch": 12.36,
      "learning_rate": 3.826570095561245e-05,
      "loss": 0.2833,
      "step": 34190,
      "task_loss": 0.5841542482376099
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26933300495147705,
      "epoch": 12.36,
      "learning_rate": 3.8186562816345615e-05,
      "loss": 0.2951,
      "step": 34200,
      "task_loss": 0.6943976879119873
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23551398515701294,
      "epoch": 12.36,
      "learning_rate": 3.8107408246089365e-05,
      "loss": 0.2878,
      "step": 34210,
      "task_loss": 0.5216627717018127
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.16981089115142822,
      "epoch": 12.37,
      "learning_rate": 3.802823765299128e-05,
      "loss": 0.2675,
      "step": 34220,
      "task_loss": 0.4077721834182739
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32236146926879883,
      "epoch": 12.37,
      "learning_rate": 3.794905144528156e-05,
      "loss": 0.2892,
      "step": 34230,
      "task_loss": 0.38724401593208313
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3376367688179016,
      "epoch": 12.37,
      "learning_rate": 3.7869850031270916e-05,
      "loss": 0.3007,
      "step": 34240,
      "task_loss": 0.293702632188797
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3073074519634247,
      "epoch": 12.38,
      "learning_rate": 3.7790633819348476e-05,
      "loss": 0.3085,
      "step": 34250,
      "task_loss": 0.5922950506210327
    },
    {
      "epoch": 12.38,
      "eval_exact_match": 83.11258278145695,
      "eval_f1": 89.88753195084642,
      "step": 34250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35342901945114136,
      "epoch": 12.38,
      "learning_rate": 3.771140321797967e-05,
      "loss": 0.3038,
      "step": 34260,
      "task_loss": 0.5709468126296997
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.329917311668396,
      "epoch": 12.39,
      "learning_rate": 3.7632158635704116e-05,
      "loss": 0.2948,
      "step": 34270,
      "task_loss": 0.670756995677948
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2711179554462433,
      "epoch": 12.39,
      "learning_rate": 3.755290048113352e-05,
      "loss": 0.2785,
      "step": 34280,
      "task_loss": 0.5022716522216797
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27370622754096985,
      "epoch": 12.39,
      "learning_rate": 3.747362916294959e-05,
      "loss": 0.2923,
      "step": 34290,
      "task_loss": 0.7954187393188477
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31736844778060913,
      "epoch": 12.4,
      "learning_rate": 3.7394345089901914e-05,
      "loss": 0.2757,
      "step": 34300,
      "task_loss": 0.5828770399093628
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2652580440044403,
      "epoch": 12.4,
      "learning_rate": 3.731504867080579e-05,
      "loss": 0.2822,
      "step": 34310,
      "task_loss": 0.7285773754119873
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31226301193237305,
      "epoch": 12.4,
      "learning_rate": 3.723574031454026e-05,
      "loss": 0.278,
      "step": 34320,
      "task_loss": 0.802804708480835
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33275431394577026,
      "epoch": 12.41,
      "learning_rate": 3.715642043004586e-05,
      "loss": 0.2922,
      "step": 34330,
      "task_loss": 0.6281599402427673
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28947165608406067,
      "epoch": 12.41,
      "learning_rate": 3.7077089426322586e-05,
      "loss": 0.3,
      "step": 34340,
      "task_loss": 0.38332265615463257
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28193265199661255,
      "epoch": 12.41,
      "learning_rate": 3.699774771242778e-05,
      "loss": 0.2769,
      "step": 34350,
      "task_loss": 0.4479888677597046
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22736775875091553,
      "epoch": 12.42,
      "learning_rate": 3.691839569747399e-05,
      "loss": 0.2843,
      "step": 34360,
      "task_loss": 0.5466663837432861
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23163321614265442,
      "epoch": 12.42,
      "learning_rate": 3.68390337906269e-05,
      "loss": 0.2998,
      "step": 34370,
      "task_loss": 0.5594272613525391
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22144640982151031,
      "epoch": 12.43,
      "learning_rate": 3.675966240110316e-05,
      "loss": 0.2973,
      "step": 34380,
      "task_loss": 0.4239853024482727
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24784286320209503,
      "epoch": 12.43,
      "learning_rate": 3.668028193816837e-05,
      "loss": 0.2847,
      "step": 34390,
      "task_loss": 0.42308953404426575
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.245379239320755,
      "epoch": 12.43,
      "learning_rate": 3.6600892811134887e-05,
      "loss": 0.2732,
      "step": 34400,
      "task_loss": 0.4906768500804901
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24077722430229187,
      "epoch": 12.44,
      "learning_rate": 3.652149542935974e-05,
      "loss": 0.2933,
      "step": 34410,
      "task_loss": 0.7623639702796936
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27252477407455444,
      "epoch": 12.44,
      "learning_rate": 3.644209020224254e-05,
      "loss": 0.2908,
      "step": 34420,
      "task_loss": 0.45984989404678345
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2691285014152527,
      "epoch": 12.44,
      "learning_rate": 3.6362677539223316e-05,
      "loss": 0.3059,
      "step": 34430,
      "task_loss": 0.5970430970191956
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23063981533050537,
      "epoch": 12.45,
      "learning_rate": 3.628325784978048e-05,
      "loss": 0.2843,
      "step": 34440,
      "task_loss": 0.48113882541656494
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3235628306865692,
      "epoch": 12.45,
      "learning_rate": 3.620383154342866e-05,
      "loss": 0.2779,
      "step": 34450,
      "task_loss": 0.42377451062202454
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26905784010887146,
      "epoch": 12.45,
      "learning_rate": 3.612439902971659e-05,
      "loss": 0.2975,
      "step": 34460,
      "task_loss": 0.45585575699806213
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26199162006378174,
      "epoch": 12.46,
      "learning_rate": 3.604496071822503e-05,
      "loss": 0.296,
      "step": 34470,
      "task_loss": 0.3905210494995117
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24511463940143585,
      "epoch": 12.46,
      "learning_rate": 3.596551701856461e-05,
      "loss": 0.2921,
      "step": 34480,
      "task_loss": 0.36547398567199707
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23199574649333954,
      "epoch": 12.46,
      "learning_rate": 3.5886068340373774e-05,
      "loss": 0.2606,
      "step": 34490,
      "task_loss": 0.40434420108795166
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20520880818367004,
      "epoch": 12.47,
      "learning_rate": 3.580661509331662e-05,
      "loss": 0.2554,
      "step": 34500,
      "task_loss": 0.35260075330734253
    },
    {
      "epoch": 12.47,
      "eval_exact_match": 83.46263008514664,
      "eval_f1": 90.16885980331539,
      "step": 34500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2748667597770691,
      "epoch": 12.47,
      "learning_rate": 3.572715768708081e-05,
      "loss": 0.2598,
      "step": 34510,
      "task_loss": 0.41463881731033325
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2754099369049072,
      "epoch": 12.48,
      "learning_rate": 3.564769653137545e-05,
      "loss": 0.3053,
      "step": 34520,
      "task_loss": 0.4715232253074646
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21734142303466797,
      "epoch": 12.48,
      "learning_rate": 3.556823203592897e-05,
      "loss": 0.2774,
      "step": 34530,
      "task_loss": 0.37835949659347534
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25816744565963745,
      "epoch": 12.48,
      "learning_rate": 3.548876461048703e-05,
      "loss": 0.2882,
      "step": 34540,
      "task_loss": 0.409110426902771
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3097730875015259,
      "epoch": 12.49,
      "learning_rate": 3.5409294664810414e-05,
      "loss": 0.2758,
      "step": 34550,
      "task_loss": 0.6100963354110718
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28760984539985657,
      "epoch": 12.49,
      "learning_rate": 3.5329822608672863e-05,
      "loss": 0.3113,
      "step": 34560,
      "task_loss": 0.49667757749557495
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.558922290802002,
      "epoch": 12.49,
      "learning_rate": 3.52582962923918e-05,
      "loss": 0.3384,
      "step": 34570,
      "task_loss": 0.5127227306365967
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27977892756462097,
      "epoch": 12.5,
      "learning_rate": 3.517882135534249e-05,
      "loss": 0.2669,
      "step": 34580,
      "task_loss": 0.39960137009620667
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.253059446811676,
      "epoch": 12.5,
      "learning_rate": 3.509934549623014e-05,
      "loss": 0.2768,
      "step": 34590,
      "task_loss": 0.5317621231079102
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2835026681423187,
      "epoch": 12.5,
      "learning_rate": 3.501986912485901e-05,
      "loss": 0.2904,
      "step": 34600,
      "task_loss": 0.4835074841976166
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25000596046447754,
      "epoch": 12.51,
      "learning_rate": 3.494039265103597e-05,
      "loss": 0.2765,
      "step": 34610,
      "task_loss": 0.3324766755104065
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21621635556221008,
      "epoch": 12.51,
      "learning_rate": 3.486091648456848e-05,
      "loss": 0.2912,
      "step": 34620,
      "task_loss": 0.3933340907096863
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.17932358384132385,
      "epoch": 12.52,
      "learning_rate": 3.478144103526236e-05,
      "loss": 0.269,
      "step": 34630,
      "task_loss": 0.4845110774040222
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23999449610710144,
      "epoch": 12.52,
      "learning_rate": 3.470196671291976e-05,
      "loss": 0.2881,
      "step": 34640,
      "task_loss": 0.4917372763156891
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.405902624130249,
      "epoch": 12.52,
      "learning_rate": 3.4622493927337014e-05,
      "loss": 0.2936,
      "step": 34650,
      "task_loss": 0.43812593817710876
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2771126925945282,
      "epoch": 12.53,
      "learning_rate": 3.4543023088302515e-05,
      "loss": 0.297,
      "step": 34660,
      "task_loss": 0.2483316957950592
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23978747427463531,
      "epoch": 12.53,
      "learning_rate": 3.4463554605594636e-05,
      "loss": 0.3009,
      "step": 34670,
      "task_loss": 0.6507888436317444
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2043863832950592,
      "epoch": 12.53,
      "learning_rate": 3.438408888897961e-05,
      "loss": 0.282,
      "step": 34680,
      "task_loss": 0.5260913372039795
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2058839201927185,
      "epoch": 12.54,
      "learning_rate": 3.4304626348209364e-05,
      "loss": 0.2655,
      "step": 34690,
      "task_loss": 0.3453657031059265
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39385679364204407,
      "epoch": 12.54,
      "learning_rate": 3.4225167393019515e-05,
      "loss": 0.2896,
      "step": 34700,
      "task_loss": 0.8374224305152893
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28174975514411926,
      "epoch": 12.54,
      "learning_rate": 3.4145712433127126e-05,
      "loss": 0.2754,
      "step": 34710,
      "task_loss": 0.2922405004501343
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22185012698173523,
      "epoch": 12.55,
      "learning_rate": 3.4066261878228706e-05,
      "loss": 0.2613,
      "step": 34720,
      "task_loss": 0.3246780335903168
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2959524989128113,
      "epoch": 12.55,
      "learning_rate": 3.398681613799803e-05,
      "loss": 0.2727,
      "step": 34730,
      "task_loss": 0.3286215662956238
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.17410224676132202,
      "epoch": 12.56,
      "learning_rate": 3.390737562208405e-05,
      "loss": 0.2764,
      "step": 34740,
      "task_loss": 0.3103204667568207
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26146069169044495,
      "epoch": 12.56,
      "learning_rate": 3.3827940740108764e-05,
      "loss": 0.3019,
      "step": 34750,
      "task_loss": 0.49389928579330444
    },
    {
      "epoch": 12.56,
      "eval_exact_match": 83.30179754020814,
      "eval_f1": 89.99643864393832,
      "step": 34750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.17524585127830505,
      "epoch": 12.56,
      "learning_rate": 3.374851190166516e-05,
      "loss": 0.2547,
      "step": 34760,
      "task_loss": 0.23321254551410675
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3621963858604431,
      "epoch": 12.57,
      "learning_rate": 3.366908951631501e-05,
      "loss": 0.3203,
      "step": 34770,
      "task_loss": 0.8130776882171631
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26339295506477356,
      "epoch": 12.57,
      "learning_rate": 3.3589673993586866e-05,
      "loss": 0.2794,
      "step": 34780,
      "task_loss": 0.4487806558609009
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2658938765525818,
      "epoch": 12.57,
      "learning_rate": 3.351026574297384e-05,
      "loss": 0.2782,
      "step": 34790,
      "task_loss": 0.472459614276886
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22098635137081146,
      "epoch": 12.58,
      "learning_rate": 3.343086517393162e-05,
      "loss": 0.2921,
      "step": 34800,
      "task_loss": 0.45475247502326965
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2744857370853424,
      "epoch": 12.58,
      "learning_rate": 3.335147269587619e-05,
      "loss": 0.2564,
      "step": 34810,
      "task_loss": 0.5082106590270996
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22003915905952454,
      "epoch": 12.58,
      "learning_rate": 3.327208871818189e-05,
      "loss": 0.2594,
      "step": 34820,
      "task_loss": 0.3585473895072937
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34640657901763916,
      "epoch": 12.59,
      "learning_rate": 3.3192713650179204e-05,
      "loss": 0.3076,
      "step": 34830,
      "task_loss": 0.5335635542869568
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26574617624282837,
      "epoch": 12.59,
      "learning_rate": 3.311334790115267e-05,
      "loss": 0.2817,
      "step": 34840,
      "task_loss": 0.4539814591407776
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25745895504951477,
      "epoch": 12.59,
      "learning_rate": 3.303399188033876e-05,
      "loss": 0.2834,
      "step": 34850,
      "task_loss": 0.32244840264320374
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21593958139419556,
      "epoch": 12.6,
      "learning_rate": 3.295464599692384e-05,
      "loss": 0.2987,
      "step": 34860,
      "task_loss": 0.6063805818557739
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25682130455970764,
      "epoch": 12.6,
      "learning_rate": 3.287531066004193e-05,
      "loss": 0.2643,
      "step": 34870,
      "task_loss": 0.38414889574050903
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2547285556793213,
      "epoch": 12.61,
      "learning_rate": 3.279598627877273e-05,
      "loss": 0.2656,
      "step": 34880,
      "task_loss": 0.31116461753845215
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2546624541282654,
      "epoch": 12.61,
      "learning_rate": 3.2716673262139395e-05,
      "loss": 0.2939,
      "step": 34890,
      "task_loss": 0.37165433168411255
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.289786696434021,
      "epoch": 12.61,
      "learning_rate": 3.2637372019106546e-05,
      "loss": 0.265,
      "step": 34900,
      "task_loss": 0.6424096822738647
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3912149667739868,
      "epoch": 12.62,
      "learning_rate": 3.255808295857803e-05,
      "loss": 0.2967,
      "step": 34910,
      "task_loss": 0.8441682457923889
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2348523736000061,
      "epoch": 12.62,
      "learning_rate": 3.247880648939492e-05,
      "loss": 0.2771,
      "step": 34920,
      "task_loss": 0.5961621999740601
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30086714029312134,
      "epoch": 12.62,
      "learning_rate": 3.239954302033335e-05,
      "loss": 0.2532,
      "step": 34930,
      "task_loss": 0.5938539505004883
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24177315831184387,
      "epoch": 12.63,
      "learning_rate": 3.232029296010241e-05,
      "loss": 0.266,
      "step": 34940,
      "task_loss": 0.607329249382019
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2449285387992859,
      "epoch": 12.63,
      "learning_rate": 3.224105671734206e-05,
      "loss": 0.2816,
      "step": 34950,
      "task_loss": 0.5142256617546082
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3181091248989105,
      "epoch": 12.63,
      "learning_rate": 3.2161834700621026e-05,
      "loss": 0.2875,
      "step": 34960,
      "task_loss": 0.6336065530776978
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23876434564590454,
      "epoch": 12.64,
      "learning_rate": 3.2082627318434634e-05,
      "loss": 0.2967,
      "step": 34970,
      "task_loss": 0.4739626944065094
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2478228509426117,
      "epoch": 12.64,
      "learning_rate": 3.2003434979202836e-05,
      "loss": 0.2556,
      "step": 34980,
      "task_loss": 0.5977648496627808
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3044435977935791,
      "epoch": 12.65,
      "learning_rate": 3.19242580912679e-05,
      "loss": 0.2845,
      "step": 34990,
      "task_loss": 0.6682428121566772
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25522321462631226,
      "epoch": 12.65,
      "learning_rate": 3.1845097062892545e-05,
      "loss": 0.2961,
      "step": 35000,
      "task_loss": 0.6363422870635986
    },
    {
      "epoch": 12.65,
      "eval_exact_match": 83.68968779564806,
      "eval_f1": 90.17109381218279,
      "step": 35000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3207133710384369,
      "epoch": 12.65,
      "learning_rate": 3.1765952302257604e-05,
      "loss": 0.2852,
      "step": 35010,
      "task_loss": 0.6862952709197998
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2543371319770813,
      "epoch": 12.66,
      "learning_rate": 3.1686824217460105e-05,
      "loss": 0.2625,
      "step": 35020,
      "task_loss": 0.505595326423645
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2681245505809784,
      "epoch": 12.66,
      "learning_rate": 3.160771321651105e-05,
      "loss": 0.2769,
      "step": 35030,
      "task_loss": 0.6629229187965393
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2507745027542114,
      "epoch": 12.66,
      "learning_rate": 3.152861970733336e-05,
      "loss": 0.2827,
      "step": 35040,
      "task_loss": 0.600731611251831
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19638881087303162,
      "epoch": 12.67,
      "learning_rate": 3.144954409775978e-05,
      "loss": 0.2881,
      "step": 35050,
      "task_loss": 0.39548712968826294
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2732619047164917,
      "epoch": 12.67,
      "learning_rate": 3.1370486795530724e-05,
      "loss": 0.2787,
      "step": 35060,
      "task_loss": 0.4209481477737427
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21758723258972168,
      "epoch": 12.67,
      "learning_rate": 3.129144820829223e-05,
      "loss": 0.289,
      "step": 35070,
      "task_loss": 0.27617648243904114
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26905715465545654,
      "epoch": 12.68,
      "learning_rate": 3.1212428743593856e-05,
      "loss": 0.2781,
      "step": 35080,
      "task_loss": 0.3222195506095886
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.309938907623291,
      "epoch": 12.68,
      "learning_rate": 3.113342880888649e-05,
      "loss": 0.3007,
      "step": 35090,
      "task_loss": 0.4127185642719269
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23867590725421906,
      "epoch": 12.69,
      "learning_rate": 3.10544488115204e-05,
      "loss": 0.2763,
      "step": 35100,
      "task_loss": 0.7365835905075073
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26832306385040283,
      "epoch": 12.69,
      "learning_rate": 3.097548915874299e-05,
      "loss": 0.2743,
      "step": 35110,
      "task_loss": 0.39295315742492676
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4001271724700928,
      "epoch": 12.69,
      "learning_rate": 3.08965502576968e-05,
      "loss": 0.3228,
      "step": 35120,
      "task_loss": 1.1282687187194824
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25665977597236633,
      "epoch": 12.7,
      "learning_rate": 3.081763251541732e-05,
      "loss": 0.2797,
      "step": 35130,
      "task_loss": 0.37274888157844543
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.238399475812912,
      "epoch": 12.7,
      "learning_rate": 3.0746624974438324e-05,
      "loss": 0.303,
      "step": 35140,
      "task_loss": 0.32780641317367554
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20166674256324768,
      "epoch": 12.7,
      "learning_rate": 3.0667748554806034e-05,
      "loss": 0.2461,
      "step": 35150,
      "task_loss": 0.15192970633506775
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20919165015220642,
      "epoch": 12.71,
      "learning_rate": 3.058889447371902e-05,
      "loss": 0.2623,
      "step": 35160,
      "task_loss": 0.45256832242012024
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24076633155345917,
      "epoch": 12.71,
      "learning_rate": 3.0510063137775374e-05,
      "loss": 0.2748,
      "step": 35170,
      "task_loss": 0.4359937608242035
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24580487608909607,
      "epoch": 12.71,
      "learning_rate": 3.0431254953456025e-05,
      "loss": 0.3164,
      "step": 35180,
      "task_loss": 0.3960850238800049
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3059123754501343,
      "epoch": 12.72,
      "learning_rate": 3.0352470327122443e-05,
      "loss": 0.2723,
      "step": 35190,
      "task_loss": 0.7698179483413696
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2135799527168274,
      "epoch": 12.72,
      "learning_rate": 3.0273709665014673e-05,
      "loss": 0.2579,
      "step": 35200,
      "task_loss": 0.4562651515007019
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2941896915435791,
      "epoch": 12.72,
      "learning_rate": 3.0194973373249165e-05,
      "loss": 0.2959,
      "step": 35210,
      "task_loss": 0.7255433797836304
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26296886801719666,
      "epoch": 12.73,
      "learning_rate": 3.011626185781672e-05,
      "loss": 0.278,
      "step": 35220,
      "task_loss": 0.3990821838378906
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22505085170269012,
      "epoch": 12.73,
      "learning_rate": 3.0037575524580362e-05,
      "loss": 0.2853,
      "step": 35230,
      "task_loss": 0.31899845600128174
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.355307400226593,
      "epoch": 12.74,
      "learning_rate": 2.9958914779273314e-05,
      "loss": 0.2866,
      "step": 35240,
      "task_loss": 1.150566577911377
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20547062158584595,
      "epoch": 12.74,
      "learning_rate": 2.988028002749679e-05,
      "loss": 0.2836,
      "step": 35250,
      "task_loss": 0.3349851667881012
    },
    {
      "epoch": 12.74,
      "eval_exact_match": 83.08420056764427,
      "eval_f1": 89.95081892512363,
      "step": 35250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31211623549461365,
      "epoch": 12.74,
      "learning_rate": 2.980167167471807e-05,
      "loss": 0.2894,
      "step": 35260,
      "task_loss": 0.522201657295227
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20951145887374878,
      "epoch": 12.75,
      "learning_rate": 2.9723090126268194e-05,
      "loss": 0.2801,
      "step": 35270,
      "task_loss": 0.3398115634918213
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25808507204055786,
      "epoch": 12.75,
      "learning_rate": 2.9644535787340092e-05,
      "loss": 0.2669,
      "step": 35280,
      "task_loss": 0.536829948425293
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2925964295864105,
      "epoch": 12.75,
      "learning_rate": 2.9566009062986328e-05,
      "loss": 0.3058,
      "step": 35290,
      "task_loss": 0.49623316526412964
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2753320336341858,
      "epoch": 12.76,
      "learning_rate": 2.9487510358117113e-05,
      "loss": 0.2797,
      "step": 35300,
      "task_loss": 0.265572190284729
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.193668395280838,
      "epoch": 12.76,
      "learning_rate": 2.9409040077498153e-05,
      "loss": 0.2522,
      "step": 35310,
      "task_loss": 0.2860652804374695
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2943766713142395,
      "epoch": 12.76,
      "learning_rate": 2.9330598625748617e-05,
      "loss": 0.2603,
      "step": 35320,
      "task_loss": 0.687859296798706
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22939196228981018,
      "epoch": 12.77,
      "learning_rate": 2.925218640733899e-05,
      "loss": 0.2641,
      "step": 35330,
      "task_loss": 0.977037787437439
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.262288898229599,
      "epoch": 12.77,
      "learning_rate": 2.9173803826589055e-05,
      "loss": 0.2829,
      "step": 35340,
      "task_loss": 0.4137219786643982
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2971470355987549,
      "epoch": 12.78,
      "learning_rate": 2.909545128766573e-05,
      "loss": 0.2854,
      "step": 35350,
      "task_loss": 0.44294825196266174
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27984118461608887,
      "epoch": 12.78,
      "learning_rate": 2.9017129194581096e-05,
      "loss": 0.26,
      "step": 35360,
      "task_loss": 0.3375154733657837
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2255486100912094,
      "epoch": 12.78,
      "learning_rate": 2.893883795119015e-05,
      "loss": 0.2937,
      "step": 35370,
      "task_loss": 0.46902692317962646
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.299260675907135,
      "epoch": 12.79,
      "learning_rate": 2.8860577961188912e-05,
      "loss": 0.2452,
      "step": 35380,
      "task_loss": 0.5150102972984314
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2901751399040222,
      "epoch": 12.79,
      "learning_rate": 2.8782349628112184e-05,
      "loss": 0.2674,
      "step": 35390,
      "task_loss": 1.1621662378311157
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2164711356163025,
      "epoch": 12.79,
      "learning_rate": 2.870415335533157e-05,
      "loss": 0.2524,
      "step": 35400,
      "task_loss": 0.32596057653427124
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20497137308120728,
      "epoch": 12.8,
      "learning_rate": 2.8625989546053335e-05,
      "loss": 0.2764,
      "step": 35410,
      "task_loss": 0.3471028506755829
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26700571179389954,
      "epoch": 12.8,
      "learning_rate": 2.8547858603316378e-05,
      "loss": 0.3007,
      "step": 35420,
      "task_loss": 0.29583919048309326
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27736860513687134,
      "epoch": 12.8,
      "learning_rate": 2.846976092999009e-05,
      "loss": 0.2877,
      "step": 35430,
      "task_loss": 0.562250018119812
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23967581987380981,
      "epoch": 12.81,
      "learning_rate": 2.8391696928772366e-05,
      "loss": 0.2601,
      "step": 35440,
      "task_loss": 0.7210532426834106
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25436583161354065,
      "epoch": 12.81,
      "learning_rate": 2.831366700218741e-05,
      "loss": 0.2864,
      "step": 35450,
      "task_loss": 0.4490795135498047
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25366324186325073,
      "epoch": 12.82,
      "learning_rate": 2.8235671552583812e-05,
      "loss": 0.274,
      "step": 35460,
      "task_loss": 0.7202346324920654
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24929600954055786,
      "epoch": 12.82,
      "learning_rate": 2.8157710982132277e-05,
      "loss": 0.3073,
      "step": 35470,
      "task_loss": 0.7886189222335815
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21746213734149933,
      "epoch": 12.82,
      "learning_rate": 2.8079785692823776e-05,
      "loss": 0.2845,
      "step": 35480,
      "task_loss": 0.2715526819229126
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24432429671287537,
      "epoch": 12.83,
      "learning_rate": 2.800189608646727e-05,
      "loss": 0.256,
      "step": 35490,
      "task_loss": 0.4983975291252136
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28279516100883484,
      "epoch": 12.83,
      "learning_rate": 2.7924042564687786e-05,
      "loss": 0.2715,
      "step": 35500,
      "task_loss": 0.5286850333213806
    },
    {
      "epoch": 12.83,
      "eval_exact_match": 83.34910122989594,
      "eval_f1": 90.11317455780872,
      "step": 35500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22306914627552032,
      "epoch": 12.83,
      "learning_rate": 2.7846225528924238e-05,
      "loss": 0.2723,
      "step": 35510,
      "task_loss": 0.4120517373085022
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.252077579498291,
      "epoch": 12.84,
      "learning_rate": 2.7768445380427447e-05,
      "loss": 0.2739,
      "step": 35520,
      "task_loss": 0.3027448356151581
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2874928116798401,
      "epoch": 12.84,
      "learning_rate": 2.7690702520258e-05,
      "loss": 0.2884,
      "step": 35530,
      "task_loss": 0.663945198059082
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3042541742324829,
      "epoch": 12.84,
      "learning_rate": 2.7612997349284244e-05,
      "loss": 0.2923,
      "step": 35540,
      "task_loss": 0.8597861528396606
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34958600997924805,
      "epoch": 12.85,
      "learning_rate": 2.753533026818014e-05,
      "loss": 0.2926,
      "step": 35550,
      "task_loss": 0.43330618739128113
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.265541136264801,
      "epoch": 12.85,
      "learning_rate": 2.7457701677423323e-05,
      "loss": 0.2763,
      "step": 35560,
      "task_loss": 0.4258151054382324
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25206151604652405,
      "epoch": 12.86,
      "learning_rate": 2.7380111977292854e-05,
      "loss": 0.2988,
      "step": 35570,
      "task_loss": 0.5166367888450623
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26590320467948914,
      "epoch": 12.86,
      "learning_rate": 2.7302561567867363e-05,
      "loss": 0.2942,
      "step": 35580,
      "task_loss": 0.9321345686912537
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19803495705127716,
      "epoch": 12.86,
      "learning_rate": 2.722505084902282e-05,
      "loss": 0.2596,
      "step": 35590,
      "task_loss": 0.3814893960952759
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21219594776630402,
      "epoch": 12.87,
      "learning_rate": 2.7147580220430556e-05,
      "loss": 0.2698,
      "step": 35600,
      "task_loss": 0.26240330934524536
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.274894118309021,
      "epoch": 12.87,
      "learning_rate": 2.7070150081555185e-05,
      "loss": 0.2635,
      "step": 35610,
      "task_loss": 0.6794742345809937
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29060205817222595,
      "epoch": 12.87,
      "learning_rate": 2.699276083165254e-05,
      "loss": 0.3269,
      "step": 35620,
      "task_loss": 0.4442636966705322
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2644478380680084,
      "epoch": 12.88,
      "learning_rate": 2.6915412869767616e-05,
      "loss": 0.2703,
      "step": 35630,
      "task_loss": 0.33713334798812866
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2862852215766907,
      "epoch": 12.88,
      "learning_rate": 2.6838106594732518e-05,
      "loss": 0.2643,
      "step": 35640,
      "task_loss": 0.5014687180519104
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2821693420410156,
      "epoch": 12.88,
      "learning_rate": 2.6760842405164378e-05,
      "loss": 0.2717,
      "step": 35650,
      "task_loss": 0.4597575068473816
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2885741591453552,
      "epoch": 12.89,
      "learning_rate": 2.6683620699463366e-05,
      "loss": 0.3032,
      "step": 35660,
      "task_loss": 0.7140188217163086
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23422183096408844,
      "epoch": 12.89,
      "learning_rate": 2.6606441875810548e-05,
      "loss": 0.2679,
      "step": 35670,
      "task_loss": 0.4431329369544983
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2877368927001953,
      "epoch": 12.89,
      "learning_rate": 2.6529306332165898e-05,
      "loss": 0.286,
      "step": 35680,
      "task_loss": 0.5499070286750793
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3035220503807068,
      "epoch": 12.9,
      "learning_rate": 2.645221446626621e-05,
      "loss": 0.291,
      "step": 35690,
      "task_loss": 0.6294026970863342
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26873719692230225,
      "epoch": 12.9,
      "learning_rate": 2.6375166675623076e-05,
      "loss": 0.306,
      "step": 35700,
      "task_loss": 0.4272887706756592
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26753750443458557,
      "epoch": 12.91,
      "learning_rate": 2.62981633575208e-05,
      "loss": 0.2603,
      "step": 35710,
      "task_loss": 0.4566969871520996
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22760161757469177,
      "epoch": 12.91,
      "learning_rate": 2.62212049090144e-05,
      "loss": 0.2906,
      "step": 35720,
      "task_loss": 0.3315356373786926
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22385570406913757,
      "epoch": 12.91,
      "learning_rate": 2.6144291726927492e-05,
      "loss": 0.2608,
      "step": 35730,
      "task_loss": 0.3490475118160248
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22863799333572388,
      "epoch": 12.92,
      "learning_rate": 2.606742420785035e-05,
      "loss": 0.2873,
      "step": 35740,
      "task_loss": 0.3199641704559326
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2786450982093811,
      "epoch": 12.92,
      "learning_rate": 2.5990602748137694e-05,
      "loss": 0.2993,
      "step": 35750,
      "task_loss": 0.9228396415710449
    },
    {
      "epoch": 12.92,
      "eval_exact_match": 83.40586565752129,
      "eval_f1": 90.15130214122858,
      "step": 35750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22297132015228271,
      "epoch": 12.92,
      "learning_rate": 2.5913827743906845e-05,
      "loss": 0.2731,
      "step": 35760,
      "task_loss": 0.8088899850845337
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3312681317329407,
      "epoch": 12.93,
      "learning_rate": 2.5837099591035526e-05,
      "loss": 0.2676,
      "step": 35770,
      "task_loss": 0.6634341478347778
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24431252479553223,
      "epoch": 12.93,
      "learning_rate": 2.57604186851599e-05,
      "loss": 0.2808,
      "step": 35780,
      "task_loss": 0.480110228061676
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2923346161842346,
      "epoch": 12.93,
      "learning_rate": 2.5683785421672488e-05,
      "loss": 0.2755,
      "step": 35790,
      "task_loss": 0.431102454662323
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2945399582386017,
      "epoch": 12.94,
      "learning_rate": 2.560720019572019e-05,
      "loss": 0.2777,
      "step": 35800,
      "task_loss": 0.495644211769104
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36957889795303345,
      "epoch": 12.94,
      "learning_rate": 2.5530663402202158e-05,
      "loss": 0.2792,
      "step": 35810,
      "task_loss": 0.6453205347061157
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.327834814786911,
      "epoch": 12.95,
      "learning_rate": 2.5454175435767846e-05,
      "loss": 0.2772,
      "step": 35820,
      "task_loss": 0.5545930862426758
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30460476875305176,
      "epoch": 12.95,
      "learning_rate": 2.5377736690814943e-05,
      "loss": 0.2708,
      "step": 35830,
      "task_loss": 0.7154297828674316
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29110848903656006,
      "epoch": 12.95,
      "learning_rate": 2.5301347561487302e-05,
      "loss": 0.2833,
      "step": 35840,
      "task_loss": 0.4856996238231659
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22446209192276,
      "epoch": 12.96,
      "learning_rate": 2.5225008441672982e-05,
      "loss": 0.2434,
      "step": 35850,
      "task_loss": 0.5817791223526001
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22967638075351715,
      "epoch": 12.96,
      "learning_rate": 2.514871972500213e-05,
      "loss": 0.2544,
      "step": 35860,
      "task_loss": 0.41800975799560547
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3124304413795471,
      "epoch": 12.96,
      "learning_rate": 2.507248180484505e-05,
      "loss": 0.289,
      "step": 35870,
      "task_loss": 0.35523521900177
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.378414511680603,
      "epoch": 12.97,
      "learning_rate": 2.4996295074310068e-05,
      "loss": 0.2855,
      "step": 35880,
      "task_loss": 0.6406650543212891
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21744538843631744,
      "epoch": 12.97,
      "learning_rate": 2.492015992624161e-05,
      "loss": 0.2866,
      "step": 35890,
      "task_loss": 0.428682804107666
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23970133066177368,
      "epoch": 12.97,
      "learning_rate": 2.484407675321807e-05,
      "loss": 0.292,
      "step": 35900,
      "task_loss": 0.4487767517566681
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3047142028808594,
      "epoch": 12.98,
      "learning_rate": 2.4768045947549914e-05,
      "loss": 0.3119,
      "step": 35910,
      "task_loss": 0.6567196846008301
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3284294009208679,
      "epoch": 12.98,
      "learning_rate": 2.46920679012775e-05,
      "loss": 0.2771,
      "step": 35920,
      "task_loss": 0.5150258541107178
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23500069975852966,
      "epoch": 12.99,
      "learning_rate": 2.4616143006169216e-05,
      "loss": 0.27,
      "step": 35930,
      "task_loss": 0.28987038135528564
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21058207750320435,
      "epoch": 12.99,
      "learning_rate": 2.4540271653719334e-05,
      "loss": 0.2711,
      "step": 35940,
      "task_loss": 0.4640544652938843
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23709851503372192,
      "epoch": 12.99,
      "learning_rate": 2.4464454235146077e-05,
      "loss": 0.2738,
      "step": 35950,
      "task_loss": 0.6334000825881958
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22366857528686523,
      "epoch": 13.0,
      "learning_rate": 2.4388691141389522e-05,
      "loss": 0.2613,
      "step": 35960,
      "task_loss": 0.36642172932624817
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2464757263660431,
      "epoch": 13.0,
      "learning_rate": 2.431298276310969e-05,
      "loss": 0.2681,
      "step": 35970,
      "task_loss": 0.5318572521209717
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21534934639930725,
      "epoch": 13.0,
      "learning_rate": 2.423732949068441e-05,
      "loss": 0.2772,
      "step": 35980,
      "task_loss": 0.5838292241096497
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20216085016727448,
      "epoch": 13.01,
      "learning_rate": 2.4161731714207415e-05,
      "loss": 0.2649,
      "step": 35990,
      "task_loss": 0.29948505759239197
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24598738551139832,
      "epoch": 13.01,
      "learning_rate": 2.4086189823486237e-05,
      "loss": 0.2785,
      "step": 36000,
      "task_loss": 0.5142921209335327
    },
    {
      "epoch": 13.01,
      "eval_exact_match": 83.43424787133397,
      "eval_f1": 90.10950263383882,
      "step": 36000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25713568925857544,
      "epoch": 13.01,
      "learning_rate": 2.401070420804031e-05,
      "loss": 0.2739,
      "step": 36010,
      "task_loss": 0.5379408597946167
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25871193408966064,
      "epoch": 13.02,
      "learning_rate": 2.393527525709879e-05,
      "loss": 0.2571,
      "step": 36020,
      "task_loss": 0.6447278261184692
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19045813381671906,
      "epoch": 13.02,
      "learning_rate": 2.3859903359598768e-05,
      "loss": 0.2508,
      "step": 36030,
      "task_loss": 0.358597069978714
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2523159384727478,
      "epoch": 13.02,
      "learning_rate": 2.378458890418306e-05,
      "loss": 0.2541,
      "step": 36040,
      "task_loss": 0.5766875743865967
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.16886655986309052,
      "epoch": 13.03,
      "learning_rate": 2.370933227919834e-05,
      "loss": 0.2527,
      "step": 36050,
      "task_loss": 0.5645083785057068
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24453911185264587,
      "epoch": 13.03,
      "learning_rate": 2.3634133872693063e-05,
      "loss": 0.2473,
      "step": 36060,
      "task_loss": 0.7919536828994751
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19830533862113953,
      "epoch": 13.04,
      "learning_rate": 2.3558994072415506e-05,
      "loss": 0.2636,
      "step": 36070,
      "task_loss": 0.5828542709350586
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3160579204559326,
      "epoch": 13.04,
      "learning_rate": 2.3483913265811735e-05,
      "loss": 0.3015,
      "step": 36080,
      "task_loss": 0.5136288404464722
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2591038942337036,
      "epoch": 13.04,
      "learning_rate": 2.3408891840023662e-05,
      "loss": 0.2589,
      "step": 36090,
      "task_loss": 0.4600667953491211
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20899376273155212,
      "epoch": 13.05,
      "learning_rate": 2.3333930181886945e-05,
      "loss": 0.2379,
      "step": 36100,
      "task_loss": 0.19576582312583923
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20569036900997162,
      "epoch": 13.05,
      "learning_rate": 2.3259028677929147e-05,
      "loss": 0.2558,
      "step": 36110,
      "task_loss": 0.40065017342567444
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2914265990257263,
      "epoch": 13.05,
      "learning_rate": 2.3184187714367566e-05,
      "loss": 0.2405,
      "step": 36120,
      "task_loss": 0.456284761428833
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23679134249687195,
      "epoch": 13.06,
      "learning_rate": 2.3109407677107403e-05,
      "loss": 0.2539,
      "step": 36130,
      "task_loss": 0.49874114990234375
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22600065171718597,
      "epoch": 13.06,
      "learning_rate": 2.303468895173967e-05,
      "loss": 0.2804,
      "step": 36140,
      "task_loss": 0.49588102102279663
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36351796984672546,
      "epoch": 13.06,
      "learning_rate": 2.2960031923539245e-05,
      "loss": 0.2586,
      "step": 36150,
      "task_loss": 0.7948037385940552
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.284883052110672,
      "epoch": 13.07,
      "learning_rate": 2.2885436977462856e-05,
      "loss": 0.2464,
      "step": 36160,
      "task_loss": 0.5250895619392395
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18825049698352814,
      "epoch": 13.07,
      "learning_rate": 2.281090449814715e-05,
      "loss": 0.2421,
      "step": 36170,
      "task_loss": 0.3618870973587036
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2766856849193573,
      "epoch": 13.08,
      "learning_rate": 2.2736434869906633e-05,
      "loss": 0.2754,
      "step": 36180,
      "task_loss": 0.2450508177280426
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20527908205986023,
      "epoch": 13.08,
      "learning_rate": 2.266202847673176e-05,
      "loss": 0.2416,
      "step": 36190,
      "task_loss": 0.37746256589889526
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20512312650680542,
      "epoch": 13.08,
      "learning_rate": 2.2587685702286905e-05,
      "loss": 0.2368,
      "step": 36200,
      "task_loss": 0.5858408808708191
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26339060068130493,
      "epoch": 13.09,
      "learning_rate": 2.2513406929908436e-05,
      "loss": 0.2462,
      "step": 36210,
      "task_loss": 0.5640926361083984
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26149576902389526,
      "epoch": 13.09,
      "learning_rate": 2.2439192542602643e-05,
      "loss": 0.253,
      "step": 36220,
      "task_loss": 1.1567996740341187
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22706861793994904,
      "epoch": 13.09,
      "learning_rate": 2.2365042923043904e-05,
      "loss": 0.2413,
      "step": 36230,
      "task_loss": 0.4056210219860077
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22760742902755737,
      "epoch": 13.1,
      "learning_rate": 2.2290958453572563e-05,
      "loss": 0.2546,
      "step": 36240,
      "task_loss": 0.624477744102478
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23840272426605225,
      "epoch": 13.1,
      "learning_rate": 2.2216939516193066e-05,
      "loss": 0.2624,
      "step": 36250,
      "task_loss": 0.6678675413131714
    },
    {
      "epoch": 13.1,
      "eval_exact_match": 83.44370860927152,
      "eval_f1": 90.08635960567796,
      "step": 36250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22495925426483154,
      "epoch": 13.1,
      "learning_rate": 2.2142986492571942e-05,
      "loss": 0.2348,
      "step": 36260,
      "task_loss": 0.35559767484664917
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2641688287258148,
      "epoch": 13.11,
      "learning_rate": 2.2069099764035857e-05,
      "loss": 0.2598,
      "step": 36270,
      "task_loss": 0.4362841546535492
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1987636238336563,
      "epoch": 13.11,
      "learning_rate": 2.1995279711569616e-05,
      "loss": 0.2754,
      "step": 36280,
      "task_loss": 0.5346006155014038
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21387869119644165,
      "epoch": 13.12,
      "learning_rate": 2.192152671581425e-05,
      "loss": 0.262,
      "step": 36290,
      "task_loss": 0.45050013065338135
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24132338166236877,
      "epoch": 13.12,
      "learning_rate": 2.1847841157064977e-05,
      "loss": 0.2405,
      "step": 36300,
      "task_loss": 0.3253955841064453
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26122480630874634,
      "epoch": 13.12,
      "learning_rate": 2.1774223415269346e-05,
      "loss": 0.2751,
      "step": 36310,
      "task_loss": 0.5988302826881409
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22946585714817047,
      "epoch": 13.13,
      "learning_rate": 2.1700673870025175e-05,
      "loss": 0.2816,
      "step": 36320,
      "task_loss": 0.4920123815536499
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24815085530281067,
      "epoch": 13.13,
      "learning_rate": 2.1627192900578653e-05,
      "loss": 0.285,
      "step": 36330,
      "task_loss": 0.44152703881263733
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2004362791776657,
      "epoch": 13.13,
      "learning_rate": 2.1553780885822373e-05,
      "loss": 0.2194,
      "step": 36340,
      "task_loss": 0.41171136498451233
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19157975912094116,
      "epoch": 13.14,
      "learning_rate": 2.1480438204293358e-05,
      "loss": 0.2551,
      "step": 36350,
      "task_loss": 0.6984323263168335
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.17804864048957825,
      "epoch": 13.14,
      "learning_rate": 2.1407165234171147e-05,
      "loss": 0.2508,
      "step": 36360,
      "task_loss": 0.849889874458313
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3071947693824768,
      "epoch": 13.14,
      "learning_rate": 2.1333962353275808e-05,
      "loss": 0.2493,
      "step": 36370,
      "task_loss": 0.6484819650650024
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2442978024482727,
      "epoch": 13.15,
      "learning_rate": 2.1260829939066002e-05,
      "loss": 0.2629,
      "step": 36380,
      "task_loss": 0.4738425016403198
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41016677021980286,
      "epoch": 13.15,
      "learning_rate": 2.118776836863708e-05,
      "loss": 0.2816,
      "step": 36390,
      "task_loss": 0.7150113582611084
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2693639397621155,
      "epoch": 13.16,
      "learning_rate": 2.1114778018719025e-05,
      "loss": 0.2863,
      "step": 36400,
      "task_loss": 0.9608858823776245
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22125914692878723,
      "epoch": 13.16,
      "learning_rate": 2.104185926567466e-05,
      "loss": 0.2354,
      "step": 36410,
      "task_loss": 0.38470879197120667
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19842985272407532,
      "epoch": 13.16,
      "learning_rate": 2.096901248549757e-05,
      "loss": 0.2564,
      "step": 36420,
      "task_loss": 0.4488826394081116
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25292330980300903,
      "epoch": 13.17,
      "learning_rate": 2.0896238053810277e-05,
      "loss": 0.268,
      "step": 36430,
      "task_loss": 0.5951564311981201
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2672657370567322,
      "epoch": 13.17,
      "learning_rate": 2.082353634586219e-05,
      "loss": 0.2637,
      "step": 36440,
      "task_loss": 0.6640857458114624
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2842154800891876,
      "epoch": 13.17,
      "learning_rate": 2.0750907736527796e-05,
      "loss": 0.2534,
      "step": 36450,
      "task_loss": 0.653637170791626
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2182607650756836,
      "epoch": 13.18,
      "learning_rate": 2.06783526003046e-05,
      "loss": 0.2511,
      "step": 36460,
      "task_loss": 0.5978858470916748
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22036978602409363,
      "epoch": 13.18,
      "learning_rate": 2.06058713113113e-05,
      "loss": 0.2466,
      "step": 36470,
      "task_loss": 0.27851811051368713
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20994271337985992,
      "epoch": 13.18,
      "learning_rate": 2.0533464243285776e-05,
      "loss": 0.2533,
      "step": 36480,
      "task_loss": 0.39874762296676636
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23912584781646729,
      "epoch": 13.19,
      "learning_rate": 2.046113176958325e-05,
      "loss": 0.252,
      "step": 36490,
      "task_loss": 0.5322191715240479
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23245467245578766,
      "epoch": 13.19,
      "learning_rate": 2.038887426317424e-05,
      "loss": 0.2353,
      "step": 36500,
      "task_loss": 0.4917253851890564
    },
    {
      "epoch": 13.19,
      "eval_exact_match": 83.519394512772,
      "eval_f1": 90.19268259004296,
      "step": 36500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2391589879989624,
      "epoch": 13.19,
      "learning_rate": 2.0316692096642786e-05,
      "loss": 0.2526,
      "step": 36510,
      "task_loss": 0.7205773591995239
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18963998556137085,
      "epoch": 13.2,
      "learning_rate": 2.0244585642184383e-05,
      "loss": 0.2644,
      "step": 36520,
      "task_loss": 0.45464566349983215
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3253783583641052,
      "epoch": 13.2,
      "learning_rate": 2.017255527160416e-05,
      "loss": 0.2724,
      "step": 36530,
      "task_loss": 0.5648168921470642
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23560191690921783,
      "epoch": 13.21,
      "learning_rate": 2.0100601356314944e-05,
      "loss": 0.247,
      "step": 36540,
      "task_loss": 0.2432674914598465
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22225362062454224,
      "epoch": 13.21,
      "learning_rate": 2.0028724267335317e-05,
      "loss": 0.2637,
      "step": 36550,
      "task_loss": 0.35249823331832886
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26490503549575806,
      "epoch": 13.21,
      "learning_rate": 1.9956924375287686e-05,
      "loss": 0.2337,
      "step": 36560,
      "task_loss": 0.7395399808883667
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29515647888183594,
      "epoch": 13.22,
      "learning_rate": 1.9885202050396498e-05,
      "loss": 0.2518,
      "step": 36570,
      "task_loss": 0.35637781023979187
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.17328497767448425,
      "epoch": 13.22,
      "learning_rate": 1.9813557662486124e-05,
      "loss": 0.223,
      "step": 36580,
      "task_loss": 0.549172580242157
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21613357961177826,
      "epoch": 13.22,
      "learning_rate": 1.9741991580979143e-05,
      "loss": 0.2387,
      "step": 36590,
      "task_loss": 0.46144798398017883
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18344077467918396,
      "epoch": 13.23,
      "learning_rate": 1.967050417489433e-05,
      "loss": 0.2439,
      "step": 36600,
      "task_loss": 0.4379880428314209
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2288159728050232,
      "epoch": 13.23,
      "learning_rate": 1.9599095812844813e-05,
      "loss": 0.2528,
      "step": 36610,
      "task_loss": 0.6495665311813354
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.16957059502601624,
      "epoch": 13.23,
      "learning_rate": 1.9527766863036083e-05,
      "loss": 0.2552,
      "step": 36620,
      "task_loss": 0.3471338748931885
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2566898465156555,
      "epoch": 13.24,
      "learning_rate": 1.9456517693264205e-05,
      "loss": 0.2451,
      "step": 36630,
      "task_loss": 0.4642353057861328
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24621951580047607,
      "epoch": 13.24,
      "learning_rate": 1.9385348670913862e-05,
      "loss": 0.2651,
      "step": 36640,
      "task_loss": 0.24718859791755676
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22665154933929443,
      "epoch": 13.25,
      "learning_rate": 1.9314260162956464e-05,
      "loss": 0.2416,
      "step": 36650,
      "task_loss": 0.3895531892776489
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25208577513694763,
      "epoch": 13.25,
      "learning_rate": 1.924325253594826e-05,
      "loss": 0.263,
      "step": 36660,
      "task_loss": 0.4521622359752655
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28293052315711975,
      "epoch": 13.25,
      "learning_rate": 1.9172326156028467e-05,
      "loss": 0.2326,
      "step": 36670,
      "task_loss": 0.6860798597335815
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2210230976343155,
      "epoch": 13.26,
      "learning_rate": 1.910148138891732e-05,
      "loss": 0.2396,
      "step": 36680,
      "task_loss": 0.4906100928783417
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20675742626190186,
      "epoch": 13.26,
      "learning_rate": 1.9030718599914283e-05,
      "loss": 0.2274,
      "step": 36690,
      "task_loss": 0.3917931318283081
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2826964855194092,
      "epoch": 13.26,
      "learning_rate": 1.896003815389608e-05,
      "loss": 0.2658,
      "step": 36700,
      "task_loss": 0.39695435762405396
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24488209187984467,
      "epoch": 13.27,
      "learning_rate": 1.888944041531488e-05,
      "loss": 0.2542,
      "step": 36710,
      "task_loss": 0.6131412982940674
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.17799416184425354,
      "epoch": 13.27,
      "learning_rate": 1.8818925748196324e-05,
      "loss": 0.2525,
      "step": 36720,
      "task_loss": 0.38853368163108826
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31300222873687744,
      "epoch": 13.27,
      "learning_rate": 1.874849451613779e-05,
      "loss": 0.2527,
      "step": 36730,
      "task_loss": 0.44057413935661316
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22045227885246277,
      "epoch": 13.28,
      "learning_rate": 1.8678147082306366e-05,
      "loss": 0.259,
      "step": 36740,
      "task_loss": 0.5224936008453369
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25457197427749634,
      "epoch": 13.28,
      "learning_rate": 1.8607883809437084e-05,
      "loss": 0.2566,
      "step": 36750,
      "task_loss": 0.47354400157928467
    },
    {
      "epoch": 13.28,
      "eval_exact_match": 83.19772942289498,
      "eval_f1": 89.94332125061486,
      "step": 36750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.214283749461174,
      "epoch": 13.29,
      "learning_rate": 1.853770505983101e-05,
      "loss": 0.2223,
      "step": 36760,
      "task_loss": 0.3899107873439789
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22692391276359558,
      "epoch": 13.29,
      "learning_rate": 1.846761119535338e-05,
      "loss": 0.2566,
      "step": 36770,
      "task_loss": 0.5070915222167969
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2641848027706146,
      "epoch": 13.29,
      "learning_rate": 1.839760257743171e-05,
      "loss": 0.2649,
      "step": 36780,
      "task_loss": 0.8205592632293701
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25266072154045105,
      "epoch": 13.3,
      "learning_rate": 1.832767956705399e-05,
      "loss": 0.2525,
      "step": 36790,
      "task_loss": 0.5840563178062439
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27707502245903015,
      "epoch": 13.3,
      "learning_rate": 1.825784252476677e-05,
      "loss": 0.2384,
      "step": 36800,
      "task_loss": 0.6803827285766602
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20863491296768188,
      "epoch": 13.3,
      "learning_rate": 1.818809181067334e-05,
      "loss": 0.281,
      "step": 36810,
      "task_loss": 0.24492724239826202
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2772062122821808,
      "epoch": 13.31,
      "learning_rate": 1.8118427784431795e-05,
      "loss": 0.2565,
      "step": 36820,
      "task_loss": 0.2199469804763794
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19231802225112915,
      "epoch": 13.31,
      "learning_rate": 1.8048850805253334e-05,
      "loss": 0.2663,
      "step": 36830,
      "task_loss": 0.6367592811584473
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24778011441230774,
      "epoch": 13.31,
      "learning_rate": 1.7979361231900222e-05,
      "loss": 0.2265,
      "step": 36840,
      "task_loss": 0.48328888416290283
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22561031579971313,
      "epoch": 13.32,
      "learning_rate": 1.7909959422684075e-05,
      "loss": 0.2561,
      "step": 36850,
      "task_loss": 0.49358463287353516
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3352190852165222,
      "epoch": 13.32,
      "learning_rate": 1.7840645735463956e-05,
      "loss": 0.2427,
      "step": 36860,
      "task_loss": 0.6199913024902344
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3952946662902832,
      "epoch": 13.32,
      "learning_rate": 1.7771420527644567e-05,
      "loss": 0.2558,
      "step": 36870,
      "task_loss": 0.5466555953025818
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22066161036491394,
      "epoch": 13.33,
      "learning_rate": 1.770228415617432e-05,
      "loss": 0.2415,
      "step": 36880,
      "task_loss": 0.6449267864227295
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22198528051376343,
      "epoch": 13.33,
      "learning_rate": 1.763323697754362e-05,
      "loss": 0.2611,
      "step": 36890,
      "task_loss": 0.6034165620803833
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24511726200580597,
      "epoch": 13.34,
      "learning_rate": 1.7564279347782926e-05,
      "loss": 0.2325,
      "step": 36900,
      "task_loss": 0.346041738986969
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20405146479606628,
      "epoch": 13.34,
      "learning_rate": 1.7495411622460984e-05,
      "loss": 0.2366,
      "step": 36910,
      "task_loss": 0.2541355788707733
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3636261820793152,
      "epoch": 13.34,
      "learning_rate": 1.7426634156682904e-05,
      "loss": 0.2816,
      "step": 36920,
      "task_loss": 0.7391115427017212
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1870489865541458,
      "epoch": 13.35,
      "learning_rate": 1.735794730508848e-05,
      "loss": 0.2518,
      "step": 36930,
      "task_loss": 0.22427958250045776
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28781723976135254,
      "epoch": 13.35,
      "learning_rate": 1.728935142185018e-05,
      "loss": 0.24,
      "step": 36940,
      "task_loss": 0.36266857385635376
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24645911157131195,
      "epoch": 13.35,
      "learning_rate": 1.7220846860671454e-05,
      "loss": 0.2527,
      "step": 36950,
      "task_loss": 0.7112646102905273
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23091328144073486,
      "epoch": 13.36,
      "learning_rate": 1.7152433974784854e-05,
      "loss": 0.2435,
      "step": 36960,
      "task_loss": 0.5005269646644592
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2623830735683441,
      "epoch": 13.36,
      "learning_rate": 1.708411311695024e-05,
      "loss": 0.2452,
      "step": 36970,
      "task_loss": 0.438732385635376
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2850179970264435,
      "epoch": 13.36,
      "learning_rate": 1.70158846394529e-05,
      "loss": 0.2597,
      "step": 36980,
      "task_loss": 0.5956829786300659
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2658255100250244,
      "epoch": 13.37,
      "learning_rate": 1.6947748894101824e-05,
      "loss": 0.2444,
      "step": 36990,
      "task_loss": 0.4191124737262726
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21543040871620178,
      "epoch": 13.37,
      "learning_rate": 1.6879706232227825e-05,
      "loss": 0.2556,
      "step": 37000,
      "task_loss": 0.8701019883155823
    },
    {
      "epoch": 13.37,
      "eval_exact_match": 83.40586565752129,
      "eval_f1": 90.1025124401472,
      "step": 37000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20009112358093262,
      "epoch": 13.38,
      "learning_rate": 1.6811757004681757e-05,
      "loss": 0.2638,
      "step": 37010,
      "task_loss": 0.5392252802848816
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20085132122039795,
      "epoch": 13.38,
      "learning_rate": 1.674390156183265e-05,
      "loss": 0.2408,
      "step": 37020,
      "task_loss": 0.40361395478248596
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22304938733577728,
      "epoch": 13.38,
      "learning_rate": 1.6676140253566046e-05,
      "loss": 0.2363,
      "step": 37030,
      "task_loss": 0.27312329411506653
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18347741663455963,
      "epoch": 13.39,
      "learning_rate": 1.6608473429281984e-05,
      "loss": 0.2462,
      "step": 37040,
      "task_loss": 0.39412713050842285
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1991209089756012,
      "epoch": 13.39,
      "learning_rate": 1.6540901437893383e-05,
      "loss": 0.2617,
      "step": 37050,
      "task_loss": 0.39125919342041016
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.306510865688324,
      "epoch": 13.39,
      "learning_rate": 1.6473424627824154e-05,
      "loss": 0.2626,
      "step": 37060,
      "task_loss": 0.8488194942474365
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26255011558532715,
      "epoch": 13.4,
      "learning_rate": 1.6406043347007433e-05,
      "loss": 0.2727,
      "step": 37070,
      "task_loss": 0.9364755749702454
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21056485176086426,
      "epoch": 13.4,
      "learning_rate": 1.633875794288373e-05,
      "loss": 0.2605,
      "step": 37080,
      "task_loss": 0.48968979716300964
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2805178165435791,
      "epoch": 13.4,
      "learning_rate": 1.6271568762399232e-05,
      "loss": 0.2535,
      "step": 37090,
      "task_loss": 0.5518473386764526
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21690964698791504,
      "epoch": 13.41,
      "learning_rate": 1.620447615200395e-05,
      "loss": 0.2596,
      "step": 37100,
      "task_loss": 0.42174243927001953
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2182358354330063,
      "epoch": 13.41,
      "learning_rate": 1.613748045764995e-05,
      "loss": 0.2438,
      "step": 37110,
      "task_loss": 0.3501666188240051
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25687599182128906,
      "epoch": 13.42,
      "learning_rate": 1.607058202478953e-05,
      "loss": 0.2505,
      "step": 37120,
      "task_loss": 0.27550244331359863
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23267127573490143,
      "epoch": 13.42,
      "learning_rate": 1.600378119837355e-05,
      "loss": 0.2358,
      "step": 37130,
      "task_loss": 0.7091718912124634
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.14930732548236847,
      "epoch": 13.42,
      "learning_rate": 1.59370783228495e-05,
      "loss": 0.2473,
      "step": 37140,
      "task_loss": 0.22773753106594086
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1943025439977646,
      "epoch": 13.43,
      "learning_rate": 1.5870473742159846e-05,
      "loss": 0.236,
      "step": 37150,
      "task_loss": 0.4039733409881592
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2754412591457367,
      "epoch": 13.43,
      "learning_rate": 1.5803967799740204e-05,
      "loss": 0.2567,
      "step": 37160,
      "task_loss": 1.006929874420166
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20370972156524658,
      "epoch": 13.43,
      "learning_rate": 1.573756083851759e-05,
      "loss": 0.2439,
      "step": 37170,
      "task_loss": 0.3426205515861511
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.17631080746650696,
      "epoch": 13.44,
      "learning_rate": 1.5671253200908604e-05,
      "loss": 0.2271,
      "step": 37180,
      "task_loss": 0.44608959555625916
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23708629608154297,
      "epoch": 13.44,
      "learning_rate": 1.560504522881773e-05,
      "loss": 0.2556,
      "step": 37190,
      "task_loss": 0.6119314432144165
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.347969651222229,
      "epoch": 13.44,
      "learning_rate": 1.5538937263635538e-05,
      "loss": 0.2768,
      "step": 37200,
      "task_loss": 0.7669745683670044
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2737374007701874,
      "epoch": 13.45,
      "learning_rate": 1.5472929646236938e-05,
      "loss": 0.2327,
      "step": 37210,
      "task_loss": 0.8010777831077576
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2608548700809479,
      "epoch": 13.45,
      "learning_rate": 1.5407022716979363e-05,
      "loss": 0.2563,
      "step": 37220,
      "task_loss": 0.33176928758621216
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21440593898296356,
      "epoch": 13.46,
      "learning_rate": 1.534121681570116e-05,
      "loss": 0.255,
      "step": 37230,
      "task_loss": 0.3645903468132019
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23056039214134216,
      "epoch": 13.46,
      "learning_rate": 1.5275512281719637e-05,
      "loss": 0.2443,
      "step": 37240,
      "task_loss": 0.5096204280853271
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2819928228855133,
      "epoch": 13.46,
      "learning_rate": 1.5209909453829482e-05,
      "loss": 0.2712,
      "step": 37250,
      "task_loss": 0.2542950510978699
    },
    {
      "epoch": 13.46,
      "eval_exact_match": 83.4247871333964,
      "eval_f1": 90.14533748801033,
      "step": 37250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21100404858589172,
      "epoch": 13.47,
      "learning_rate": 1.5144408670300936e-05,
      "loss": 0.2543,
      "step": 37260,
      "task_loss": 0.6692283749580383
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23726926743984222,
      "epoch": 13.47,
      "learning_rate": 1.5079010268878078e-05,
      "loss": 0.2438,
      "step": 37270,
      "task_loss": 0.6562150120735168
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24360351264476776,
      "epoch": 13.47,
      "learning_rate": 1.5013714586777031e-05,
      "loss": 0.2696,
      "step": 37280,
      "task_loss": 0.3701857328414917
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20092684030532837,
      "epoch": 13.48,
      "learning_rate": 1.4948521960684298e-05,
      "loss": 0.2565,
      "step": 37290,
      "task_loss": 0.5324305295944214
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24796535074710846,
      "epoch": 13.48,
      "learning_rate": 1.4883432726754985e-05,
      "loss": 0.2497,
      "step": 37300,
      "task_loss": 0.6669043302536011
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24593818187713623,
      "epoch": 13.48,
      "learning_rate": 1.4818447220611084e-05,
      "loss": 0.2446,
      "step": 37310,
      "task_loss": 0.5198768377304077
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18586906790733337,
      "epoch": 13.49,
      "learning_rate": 1.475356577733968e-05,
      "loss": 0.2344,
      "step": 37320,
      "task_loss": 0.40673136711120605
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24557605385780334,
      "epoch": 13.49,
      "learning_rate": 1.4688788731491351e-05,
      "loss": 0.2487,
      "step": 37330,
      "task_loss": 0.43017250299453735
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22308820486068726,
      "epoch": 13.49,
      "learning_rate": 1.4624116417078305e-05,
      "loss": 0.2509,
      "step": 37340,
      "task_loss": 0.2053220272064209
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1951208859682083,
      "epoch": 13.5,
      "learning_rate": 1.4559549167572745e-05,
      "loss": 0.2265,
      "step": 37350,
      "task_loss": 0.3179614543914795
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2037278413772583,
      "epoch": 13.5,
      "learning_rate": 1.4495087315905124e-05,
      "loss": 0.2395,
      "step": 37360,
      "task_loss": 0.2637331187725067
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2836878299713135,
      "epoch": 13.51,
      "learning_rate": 1.4430731194462441e-05,
      "loss": 0.2714,
      "step": 37370,
      "task_loss": 0.6579056978225708
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32002729177474976,
      "epoch": 13.51,
      "learning_rate": 1.4366481135086468e-05,
      "loss": 0.2596,
      "step": 37380,
      "task_loss": 0.3903340697288513
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21812787652015686,
      "epoch": 13.51,
      "learning_rate": 1.4302337469072133e-05,
      "loss": 0.2482,
      "step": 37390,
      "task_loss": 0.3896595239639282
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24210324883460999,
      "epoch": 13.52,
      "learning_rate": 1.423830052716575e-05,
      "loss": 0.2454,
      "step": 37400,
      "task_loss": 0.29697751998901367
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1834373027086258,
      "epoch": 13.52,
      "learning_rate": 1.4174370639563332e-05,
      "loss": 0.2527,
      "step": 37410,
      "task_loss": 0.2785026729106903
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23967260122299194,
      "epoch": 13.52,
      "learning_rate": 1.4110548135908838e-05,
      "loss": 0.2604,
      "step": 37420,
      "task_loss": 0.4717600643634796
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18912403285503387,
      "epoch": 13.53,
      "learning_rate": 1.4046833345292616e-05,
      "loss": 0.2177,
      "step": 37430,
      "task_loss": 0.38669267296791077
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2931283116340637,
      "epoch": 13.53,
      "learning_rate": 1.398322659624951e-05,
      "loss": 0.2575,
      "step": 37440,
      "task_loss": 0.4614885449409485
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24810287356376648,
      "epoch": 13.53,
      "learning_rate": 1.3919728216757317e-05,
      "loss": 0.2462,
      "step": 37450,
      "task_loss": 0.48419344425201416
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2596108913421631,
      "epoch": 13.54,
      "learning_rate": 1.3856338534235039e-05,
      "loss": 0.2684,
      "step": 37460,
      "task_loss": 0.7017297148704529
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2619935870170593,
      "epoch": 13.54,
      "learning_rate": 1.37930578755412e-05,
      "loss": 0.2436,
      "step": 37470,
      "task_loss": 0.4745601415634155
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.17720942199230194,
      "epoch": 13.55,
      "learning_rate": 1.3729886566972142e-05,
      "loss": 0.245,
      "step": 37480,
      "task_loss": 0.16972871124744415
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2728440463542938,
      "epoch": 13.55,
      "learning_rate": 1.3666824934260375e-05,
      "loss": 0.2331,
      "step": 37490,
      "task_loss": 0.8913913369178772
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2932123839855194,
      "epoch": 13.55,
      "learning_rate": 1.3603873302572893e-05,
      "loss": 0.2365,
      "step": 37500,
      "task_loss": 0.40376877784729004
    },
    {
      "epoch": 13.55,
      "eval_exact_match": 83.50993377483444,
      "eval_f1": 90.26178759077712,
      "step": 37500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20828589797019958,
      "epoch": 13.56,
      "learning_rate": 1.3541031996509478e-05,
      "loss": 0.256,
      "step": 37510,
      "task_loss": 0.26341885328292847
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.16295726597309113,
      "epoch": 13.56,
      "learning_rate": 1.347830134010101e-05,
      "loss": 0.2389,
      "step": 37520,
      "task_loss": 0.3301553428173065
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25007620453834534,
      "epoch": 13.56,
      "learning_rate": 1.3415681656807893e-05,
      "loss": 0.2506,
      "step": 37530,
      "task_loss": 0.49313557147979736
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24634864926338196,
      "epoch": 13.57,
      "learning_rate": 1.335317326951823e-05,
      "loss": 0.2445,
      "step": 37540,
      "task_loss": 0.4683687090873718
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1813734769821167,
      "epoch": 13.57,
      "learning_rate": 1.3290776500546303e-05,
      "loss": 0.2337,
      "step": 37550,
      "task_loss": 0.4950636625289917
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2741466164588928,
      "epoch": 13.57,
      "learning_rate": 1.3228491671630832e-05,
      "loss": 0.2403,
      "step": 37560,
      "task_loss": 0.7804323434829712
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2687833905220032,
      "epoch": 13.58,
      "learning_rate": 1.3166319103933353e-05,
      "loss": 0.2493,
      "step": 37570,
      "task_loss": 0.4127853810787201
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2195703685283661,
      "epoch": 13.58,
      "learning_rate": 1.3104259118036505e-05,
      "loss": 0.2727,
      "step": 37580,
      "task_loss": 0.16775909066200256
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2721908688545227,
      "epoch": 13.59,
      "learning_rate": 1.3042312033942458e-05,
      "loss": 0.2317,
      "step": 37590,
      "task_loss": 0.5137249231338501
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1676669418811798,
      "epoch": 13.59,
      "learning_rate": 1.2980478171071208e-05,
      "loss": 0.2515,
      "step": 37600,
      "task_loss": 0.4399087727069855
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22237111628055573,
      "epoch": 13.59,
      "learning_rate": 1.2918757848258958e-05,
      "loss": 0.2403,
      "step": 37610,
      "task_loss": 0.5158141851425171
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2540138065814972,
      "epoch": 13.6,
      "learning_rate": 1.285715138375641e-05,
      "loss": 0.2341,
      "step": 37620,
      "task_loss": 0.6373366117477417
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26354658603668213,
      "epoch": 13.6,
      "learning_rate": 1.2795659095227255e-05,
      "loss": 0.2724,
      "step": 37630,
      "task_loss": 0.64971923828125
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19850783050060272,
      "epoch": 13.6,
      "learning_rate": 1.2734281299746384e-05,
      "loss": 0.2667,
      "step": 37640,
      "task_loss": 0.29709264636039734
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24133870005607605,
      "epoch": 13.61,
      "learning_rate": 1.2673018313798351e-05,
      "loss": 0.2627,
      "step": 37650,
      "task_loss": 0.3527446687221527
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2671836018562317,
      "epoch": 13.61,
      "learning_rate": 1.261187045327572e-05,
      "loss": 0.2677,
      "step": 37660,
      "task_loss": 0.8679643869400024
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2491600662469864,
      "epoch": 13.61,
      "learning_rate": 1.2550838033477418e-05,
      "loss": 0.2418,
      "step": 37670,
      "task_loss": 0.40323901176452637
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21075092256069183,
      "epoch": 13.62,
      "learning_rate": 1.2489921369107137e-05,
      "loss": 0.2454,
      "step": 37680,
      "task_loss": 0.5236786603927612
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30563992261886597,
      "epoch": 13.62,
      "learning_rate": 1.242912077427165e-05,
      "loss": 0.2314,
      "step": 37690,
      "task_loss": 0.5004376173019409
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2149924635887146,
      "epoch": 13.62,
      "learning_rate": 1.2368436562479312e-05,
      "loss": 0.2544,
      "step": 37700,
      "task_loss": 0.5571039319038391
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22124919295310974,
      "epoch": 13.63,
      "learning_rate": 1.2307869046638296e-05,
      "loss": 0.2382,
      "step": 37710,
      "task_loss": 0.41436147689819336
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21285533905029297,
      "epoch": 13.63,
      "learning_rate": 1.2247418539055088e-05,
      "loss": 0.2447,
      "step": 37720,
      "task_loss": 0.5282078385353088
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22626446187496185,
      "epoch": 13.64,
      "learning_rate": 1.2187085351432828e-05,
      "loss": 0.2537,
      "step": 37730,
      "task_loss": 0.5426703691482544
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2785342335700989,
      "epoch": 13.64,
      "learning_rate": 1.2126869794869742e-05,
      "loss": 0.2479,
      "step": 37740,
      "task_loss": 0.4130212068557739
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2820085287094116,
      "epoch": 13.64,
      "learning_rate": 1.2066772179857454e-05,
      "loss": 0.2727,
      "step": 37750,
      "task_loss": 0.482336163520813
    },
    {
      "epoch": 13.64,
      "eval_exact_match": 83.50047303689688,
      "eval_f1": 90.1728300271499,
      "step": 37750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2469613254070282,
      "epoch": 13.65,
      "learning_rate": 1.2006792816279496e-05,
      "loss": 0.2344,
      "step": 37760,
      "task_loss": 0.39518874883651733
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.17427587509155273,
      "epoch": 13.65,
      "learning_rate": 1.1946932013409631e-05,
      "loss": 0.2599,
      "step": 37770,
      "task_loss": 0.5201607346534729
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29734861850738525,
      "epoch": 13.65,
      "learning_rate": 1.1887190079910299e-05,
      "loss": 0.2529,
      "step": 37780,
      "task_loss": 0.47351279854774475
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1882125586271286,
      "epoch": 13.66,
      "learning_rate": 1.1827567323830968e-05,
      "loss": 0.2428,
      "step": 37790,
      "task_loss": 0.2635359764099121
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2885553240776062,
      "epoch": 13.66,
      "learning_rate": 1.1768064052606658e-05,
      "loss": 0.2525,
      "step": 37800,
      "task_loss": 0.5406814813613892
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19153067469596863,
      "epoch": 13.66,
      "learning_rate": 1.1708680573056214e-05,
      "loss": 0.2589,
      "step": 37810,
      "task_loss": 0.31528323888778687
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24100527167320251,
      "epoch": 13.67,
      "learning_rate": 1.1649417191380835e-05,
      "loss": 0.2487,
      "step": 37820,
      "task_loss": 0.6156647205352783
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.17599420249462128,
      "epoch": 13.67,
      "learning_rate": 1.1590274213162445e-05,
      "loss": 0.2461,
      "step": 37830,
      "task_loss": 0.36050450801849365
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2299099862575531,
      "epoch": 13.68,
      "learning_rate": 1.153125194336214e-05,
      "loss": 0.2828,
      "step": 37840,
      "task_loss": 0.4206226170063019
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20498183369636536,
      "epoch": 13.68,
      "learning_rate": 1.147235068631857e-05,
      "loss": 0.2303,
      "step": 37850,
      "task_loss": 0.47948789596557617
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2277766764163971,
      "epoch": 13.68,
      "learning_rate": 1.1413570745746428e-05,
      "loss": 0.25,
      "step": 37860,
      "task_loss": 0.5457226037979126
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26872360706329346,
      "epoch": 13.69,
      "learning_rate": 1.135491242473486e-05,
      "loss": 0.2548,
      "step": 37870,
      "task_loss": 0.5940477848052979
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.16519543528556824,
      "epoch": 13.69,
      "learning_rate": 1.1296376025745887e-05,
      "loss": 0.2321,
      "step": 37880,
      "task_loss": 0.3459252417087555
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26993200182914734,
      "epoch": 13.69,
      "learning_rate": 1.123796185061287e-05,
      "loss": 0.2517,
      "step": 37890,
      "task_loss": 0.8263578414916992
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23767806589603424,
      "epoch": 13.7,
      "learning_rate": 1.1179670200538954e-05,
      "loss": 0.2378,
      "step": 37900,
      "task_loss": 0.47348955273628235
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2323693037033081,
      "epoch": 13.7,
      "learning_rate": 1.1121501376095457e-05,
      "loss": 0.2371,
      "step": 37910,
      "task_loss": 0.2555083632469177
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21224288642406464,
      "epoch": 13.7,
      "learning_rate": 1.1063455677220416e-05,
      "loss": 0.2315,
      "step": 37920,
      "task_loss": 0.27007701992988586
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21018406748771667,
      "epoch": 13.71,
      "learning_rate": 1.1005533403216972e-05,
      "loss": 0.235,
      "step": 37930,
      "task_loss": 0.5078712105751038
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25440073013305664,
      "epoch": 13.71,
      "learning_rate": 1.0947734852751854e-05,
      "loss": 0.2425,
      "step": 37940,
      "task_loss": 0.3627524673938751
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2233213484287262,
      "epoch": 13.72,
      "learning_rate": 1.08900603238538e-05,
      "loss": 0.2278,
      "step": 37950,
      "task_loss": 0.5547107458114624
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2389487624168396,
      "epoch": 13.72,
      "learning_rate": 1.083251011391211e-05,
      "loss": 0.2451,
      "step": 37960,
      "task_loss": 0.47202402353286743
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2538294196128845,
      "epoch": 13.72,
      "learning_rate": 1.0775084519674985e-05,
      "loss": 0.2364,
      "step": 37970,
      "task_loss": 0.4768025279045105
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2578142285346985,
      "epoch": 13.73,
      "learning_rate": 1.071778383724812e-05,
      "loss": 0.2361,
      "step": 37980,
      "task_loss": 0.3769608438014984
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20831502974033356,
      "epoch": 13.73,
      "learning_rate": 1.0660608362093095e-05,
      "loss": 0.2343,
      "step": 37990,
      "task_loss": 0.616707444190979
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1966831386089325,
      "epoch": 13.73,
      "learning_rate": 1.0603558389025896e-05,
      "loss": 0.2402,
      "step": 38000,
      "task_loss": 0.3687472343444824
    },
    {
      "epoch": 13.73,
      "eval_exact_match": 83.59508041627247,
      "eval_f1": 90.24016460930066,
      "step": 38000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2341974973678589,
      "epoch": 13.74,
      "learning_rate": 1.0546634212215348e-05,
      "loss": 0.2361,
      "step": 38010,
      "task_loss": 0.31787946820259094
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18781892955303192,
      "epoch": 13.74,
      "learning_rate": 1.0489836125181665e-05,
      "loss": 0.2314,
      "step": 38020,
      "task_loss": 0.30610185861587524
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32764458656311035,
      "epoch": 13.74,
      "learning_rate": 1.0433164420794874e-05,
      "loss": 0.2738,
      "step": 38030,
      "task_loss": 0.8996621370315552
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19176249206066132,
      "epoch": 13.75,
      "learning_rate": 1.0376619391273353e-05,
      "loss": 0.2427,
      "step": 38040,
      "task_loss": 0.2879910469055176
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22764107584953308,
      "epoch": 13.75,
      "learning_rate": 1.0320201328182257e-05,
      "loss": 0.2905,
      "step": 38050,
      "task_loss": 0.39458274841308594
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23433086276054382,
      "epoch": 13.75,
      "learning_rate": 1.0263910522432138e-05,
      "loss": 0.2425,
      "step": 38060,
      "task_loss": 0.8387042284011841
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2206321656703949,
      "epoch": 13.76,
      "learning_rate": 1.020774726427728e-05,
      "loss": 0.25,
      "step": 38070,
      "task_loss": 0.3945035934448242
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19005993008613586,
      "epoch": 13.76,
      "learning_rate": 1.0151711843314342e-05,
      "loss": 0.2536,
      "step": 38080,
      "task_loss": 0.741138219833374
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30447882413864136,
      "epoch": 13.77,
      "learning_rate": 1.0095804548480798e-05,
      "loss": 0.2431,
      "step": 38090,
      "task_loss": 0.5937259197235107
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24016176164150238,
      "epoch": 13.77,
      "learning_rate": 1.004002566805347e-05,
      "loss": 0.2601,
      "step": 38100,
      "task_loss": 0.4847050905227661
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21476280689239502,
      "epoch": 13.77,
      "learning_rate": 9.984375489647e-06,
      "loss": 0.236,
      "step": 38110,
      "task_loss": 0.262297123670578
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23751616477966309,
      "epoch": 13.78,
      "learning_rate": 9.92885430021243e-06,
      "loss": 0.2438,
      "step": 38120,
      "task_loss": 0.323070228099823
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1588648557662964,
      "epoch": 13.78,
      "learning_rate": 9.87346238603569e-06,
      "loss": 0.2267,
      "step": 38130,
      "task_loss": 0.26954102516174316
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.227637380361557,
      "epoch": 13.78,
      "learning_rate": 9.81820003273612e-06,
      "loss": 0.241,
      "step": 38140,
      "task_loss": 0.6161648631095886
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.17858195304870605,
      "epoch": 13.79,
      "learning_rate": 9.763067525264964e-06,
      "loss": 0.2348,
      "step": 38150,
      "task_loss": 0.34298527240753174
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1912699043750763,
      "epoch": 13.79,
      "learning_rate": 9.708065147904013e-06,
      "loss": 0.2339,
      "step": 38160,
      "task_loss": 0.35443249344825745
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2191619575023651,
      "epoch": 13.79,
      "learning_rate": 9.653193184263991e-06,
      "loss": 0.2252,
      "step": 38170,
      "task_loss": 0.5474572777748108
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23868389427661896,
      "epoch": 13.8,
      "learning_rate": 9.598451917283206e-06,
      "loss": 0.2393,
      "step": 38180,
      "task_loss": 0.30752283334732056
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2384638637304306,
      "epoch": 13.8,
      "learning_rate": 9.543841629226033e-06,
      "loss": 0.2517,
      "step": 38190,
      "task_loss": 0.5427389144897461
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2237512767314911,
      "epoch": 13.81,
      "learning_rate": 9.489362601681498e-06,
      "loss": 0.2463,
      "step": 38200,
      "task_loss": 0.23124629259109497
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23062996566295624,
      "epoch": 13.81,
      "learning_rate": 9.435015115561758e-06,
      "loss": 0.2568,
      "step": 38210,
      "task_loss": 0.16488313674926758
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21428482234477997,
      "epoch": 13.81,
      "learning_rate": 9.380799451100741e-06,
      "loss": 0.2449,
      "step": 38220,
      "task_loss": 0.6556562185287476
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22192302346229553,
      "epoch": 13.82,
      "learning_rate": 9.326715887852645e-06,
      "loss": 0.2381,
      "step": 38230,
      "task_loss": 0.637799859046936
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24508652091026306,
      "epoch": 13.82,
      "learning_rate": 9.272764704690518e-06,
      "loss": 0.2477,
      "step": 38240,
      "task_loss": 0.5043495893478394
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20607826113700867,
      "epoch": 13.82,
      "learning_rate": 9.218946179804766e-06,
      "loss": 0.2682,
      "step": 38250,
      "task_loss": 0.5137103796005249
    },
    {
      "epoch": 13.82,
      "eval_exact_match": 83.41532639545885,
      "eval_f1": 90.16543784803585,
      "step": 38250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2131115049123764,
      "epoch": 13.83,
      "learning_rate": 9.16526059070185e-06,
      "loss": 0.2364,
      "step": 38260,
      "task_loss": 0.7902463674545288
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21433007717132568,
      "epoch": 13.83,
      "learning_rate": 9.11170821420269e-06,
      "loss": 0.2245,
      "step": 38270,
      "task_loss": 0.7640265226364136
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2725834250450134,
      "epoch": 13.83,
      "learning_rate": 9.058289326441349e-06,
      "loss": 0.2431,
      "step": 38280,
      "task_loss": 0.3009580969810486
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22229403257369995,
      "epoch": 13.84,
      "learning_rate": 9.005004202863581e-06,
      "loss": 0.2366,
      "step": 38290,
      "task_loss": 0.2828536927700043
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20776572823524475,
      "epoch": 13.84,
      "learning_rate": 8.951853118225413e-06,
      "loss": 0.2613,
      "step": 38300,
      "task_loss": 0.49406182765960693
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1888459324836731,
      "epoch": 13.85,
      "learning_rate": 8.898836346591686e-06,
      "loss": 0.2452,
      "step": 38310,
      "task_loss": 0.24637502431869507
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1972496658563614,
      "epoch": 13.85,
      "learning_rate": 8.845954161334712e-06,
      "loss": 0.2501,
      "step": 38320,
      "task_loss": 0.23441748321056366
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25202658772468567,
      "epoch": 13.85,
      "learning_rate": 8.793206835132823e-06,
      "loss": 0.2476,
      "step": 38330,
      "task_loss": 0.30959269404411316
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18595066666603088,
      "epoch": 13.86,
      "learning_rate": 8.74059463996898e-06,
      "loss": 0.2561,
      "step": 38340,
      "task_loss": 0.45879602432250977
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22673186659812927,
      "epoch": 13.86,
      "learning_rate": 8.688117847129323e-06,
      "loss": 0.264,
      "step": 38350,
      "task_loss": 0.5123938322067261
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24467988312244415,
      "epoch": 13.86,
      "learning_rate": 8.635776727201879e-06,
      "loss": 0.2449,
      "step": 38360,
      "task_loss": 0.344679057598114
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.17541030049324036,
      "epoch": 13.87,
      "learning_rate": 8.583571550075038e-06,
      "loss": 0.2452,
      "step": 38370,
      "task_loss": 0.3095884919166565
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18970157206058502,
      "epoch": 13.87,
      "learning_rate": 8.531502584936257e-06,
      "loss": 0.2627,
      "step": 38380,
      "task_loss": 0.2229578197002411
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19921694695949554,
      "epoch": 13.87,
      "learning_rate": 8.479570100270628e-06,
      "loss": 0.2512,
      "step": 38390,
      "task_loss": 0.7031322717666626
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19549816846847534,
      "epoch": 13.88,
      "learning_rate": 8.427774363859511e-06,
      "loss": 0.2552,
      "step": 38400,
      "task_loss": 0.4025685787200928
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29793527722358704,
      "epoch": 13.88,
      "learning_rate": 8.37611564277913e-06,
      "loss": 0.2567,
      "step": 38410,
      "task_loss": 0.4364060163497925
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27970990538597107,
      "epoch": 13.89,
      "learning_rate": 8.324594203399223e-06,
      "loss": 0.2352,
      "step": 38420,
      "task_loss": 0.619552731513977
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20510753989219666,
      "epoch": 13.89,
      "learning_rate": 8.273210311381659e-06,
      "loss": 0.2461,
      "step": 38430,
      "task_loss": 0.30276721715927124
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22325491905212402,
      "epoch": 13.89,
      "learning_rate": 8.221964231679075e-06,
      "loss": 0.2481,
      "step": 38440,
      "task_loss": 0.5327010750770569
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1987735629081726,
      "epoch": 13.9,
      "learning_rate": 8.170856228533461e-06,
      "loss": 0.2206,
      "step": 38450,
      "task_loss": 0.49047574400901794
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21051108837127686,
      "epoch": 13.9,
      "learning_rate": 8.119886565474906e-06,
      "loss": 0.233,
      "step": 38460,
      "task_loss": 0.3335927724838257
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19253046810626984,
      "epoch": 13.9,
      "learning_rate": 8.069055505320102e-06,
      "loss": 0.2326,
      "step": 38470,
      "task_loss": 0.495675653219223
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23543712496757507,
      "epoch": 13.91,
      "learning_rate": 8.018363310171097e-06,
      "loss": 0.2497,
      "step": 38480,
      "task_loss": 0.45676204562187195
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2220216989517212,
      "epoch": 13.91,
      "learning_rate": 7.9678102414139e-06,
      "loss": 0.2253,
      "step": 38490,
      "task_loss": 0.6177787780761719
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2542474865913391,
      "epoch": 13.91,
      "learning_rate": 7.917396559717134e-06,
      "loss": 0.2341,
      "step": 38500,
      "task_loss": 0.5604841113090515
    },
    {
      "epoch": 13.91,
      "eval_exact_match": 83.73699148533585,
      "eval_f1": 90.33564700005304,
      "step": 38500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2272586077451706,
      "epoch": 13.92,
      "learning_rate": 7.867122525030685e-06,
      "loss": 0.2341,
      "step": 38510,
      "task_loss": 0.45142441987991333
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19398754835128784,
      "epoch": 13.92,
      "learning_rate": 7.816988396584382e-06,
      "loss": 0.2454,
      "step": 38520,
      "task_loss": 0.5788732767105103
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21237385272979736,
      "epoch": 13.92,
      "learning_rate": 7.766994432886658e-06,
      "loss": 0.2452,
      "step": 38530,
      "task_loss": 0.3880312442779541
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21336814761161804,
      "epoch": 13.93,
      "learning_rate": 7.7171408917232e-06,
      "loss": 0.2534,
      "step": 38540,
      "task_loss": 0.5350017547607422
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24495704472064972,
      "epoch": 13.93,
      "learning_rate": 7.667428030155615e-06,
      "loss": 0.2345,
      "step": 38550,
      "task_loss": 0.38775384426116943
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2679603695869446,
      "epoch": 13.94,
      "learning_rate": 7.61785610452016e-06,
      "loss": 0.2355,
      "step": 38560,
      "task_loss": 0.4116300046443939
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.15576696395874023,
      "epoch": 13.94,
      "learning_rate": 7.568425370426332e-06,
      "loss": 0.2455,
      "step": 38570,
      "task_loss": 0.24840381741523743
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25543665885925293,
      "epoch": 13.94,
      "learning_rate": 7.5191360827556245e-06,
      "loss": 0.2583,
      "step": 38580,
      "task_loss": 0.5544813871383667
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2729243040084839,
      "epoch": 13.95,
      "learning_rate": 7.4699884956601825e-06,
      "loss": 0.2514,
      "step": 38590,
      "task_loss": 0.45332229137420654
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27071091532707214,
      "epoch": 13.95,
      "learning_rate": 7.420982862561493e-06,
      "loss": 0.2566,
      "step": 38600,
      "task_loss": 0.4697257876396179
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26173168420791626,
      "epoch": 13.95,
      "learning_rate": 7.372119436149068e-06,
      "loss": 0.224,
      "step": 38610,
      "task_loss": 0.1902489811182022
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23809251189231873,
      "epoch": 13.96,
      "learning_rate": 7.323398468379176e-06,
      "loss": 0.2403,
      "step": 38620,
      "task_loss": 0.5104325413703918
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3130069673061371,
      "epoch": 13.96,
      "learning_rate": 7.2748202104735e-06,
      "loss": 0.2752,
      "step": 38630,
      "task_loss": 1.0560616254806519
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19314908981323242,
      "epoch": 13.96,
      "learning_rate": 7.226384912917893e-06,
      "loss": 0.2569,
      "step": 38640,
      "task_loss": 0.5038644075393677
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2326485961675644,
      "epoch": 13.97,
      "learning_rate": 7.178092825461004e-06,
      "loss": 0.2354,
      "step": 38650,
      "task_loss": 0.5128936171531677
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1915554702281952,
      "epoch": 13.97,
      "learning_rate": 7.129944197113112e-06,
      "loss": 0.248,
      "step": 38660,
      "task_loss": 0.3978360891342163
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1979401707649231,
      "epoch": 13.98,
      "learning_rate": 7.081939276144708e-06,
      "loss": 0.2195,
      "step": 38670,
      "task_loss": 0.4705093801021576
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23233872652053833,
      "epoch": 13.98,
      "learning_rate": 7.034078310085315e-06,
      "loss": 0.23,
      "step": 38680,
      "task_loss": 0.6098400354385376
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2868359088897705,
      "epoch": 13.98,
      "learning_rate": 6.986361545722173e-06,
      "loss": 0.257,
      "step": 38690,
      "task_loss": 0.7146321535110474
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2127329409122467,
      "epoch": 13.99,
      "learning_rate": 6.938789229098967e-06,
      "loss": 0.2325,
      "step": 38700,
      "task_loss": 0.2952830195426941
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23154892027378082,
      "epoch": 13.99,
      "learning_rate": 6.8913616055145456e-06,
      "loss": 0.2365,
      "step": 38710,
      "task_loss": 0.494489848613739
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22295670211315155,
      "epoch": 13.99,
      "learning_rate": 6.8440789195216915e-06,
      "loss": 0.2562,
      "step": 38720,
      "task_loss": 0.3631395101547241
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2322646975517273,
      "epoch": 14.0,
      "learning_rate": 6.796941414925831e-06,
      "loss": 0.2093,
      "step": 38730,
      "task_loss": 0.459478497505188
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2091185450553894,
      "epoch": 14.0,
      "learning_rate": 6.7499493347838e-06,
      "loss": 0.2628,
      "step": 38740,
      "task_loss": 0.42174839973449707
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.16904908418655396,
      "epoch": 14.0,
      "learning_rate": 6.703102921402538e-06,
      "loss": 0.2227,
      "step": 38750,
      "task_loss": 0.30236685276031494
    },
    {
      "epoch": 14.0,
      "eval_exact_match": 83.50993377483444,
      "eval_f1": 90.17927551421984,
      "step": 38750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.16442182660102844,
      "epoch": 14.01,
      "learning_rate": 6.656402416337944e-06,
      "loss": 0.2263,
      "step": 38760,
      "task_loss": 0.2523331344127655
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25732508301734924,
      "epoch": 14.01,
      "learning_rate": 6.609848060393494e-06,
      "loss": 0.2272,
      "step": 38770,
      "task_loss": 0.5731315612792969
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21109287440776825,
      "epoch": 14.02,
      "learning_rate": 6.563440093619115e-06,
      "loss": 0.2242,
      "step": 38780,
      "task_loss": 0.5353798866271973
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18337416648864746,
      "epoch": 14.02,
      "learning_rate": 6.5171787553098855e-06,
      "loss": 0.228,
      "step": 38790,
      "task_loss": 0.34574657678604126
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23161965608596802,
      "epoch": 14.02,
      "learning_rate": 6.471064284004835e-06,
      "loss": 0.2217,
      "step": 38800,
      "task_loss": 0.3575626015663147
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18752753734588623,
      "epoch": 14.03,
      "learning_rate": 6.4250969174856635e-06,
      "loss": 0.2242,
      "step": 38810,
      "task_loss": 0.2804476022720337
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.164092019200325,
      "epoch": 14.03,
      "learning_rate": 6.379276892775579e-06,
      "loss": 0.2367,
      "step": 38820,
      "task_loss": 0.18792039155960083
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23862093687057495,
      "epoch": 14.03,
      "learning_rate": 6.3336044461380385e-06,
      "loss": 0.2302,
      "step": 38830,
      "task_loss": 0.6268182992935181
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2213876098394394,
      "epoch": 14.04,
      "learning_rate": 6.288079813075543e-06,
      "loss": 0.2261,
      "step": 38840,
      "task_loss": 0.4968934655189514
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23776161670684814,
      "epoch": 14.04,
      "learning_rate": 6.242703228328384e-06,
      "loss": 0.2257,
      "step": 38850,
      "task_loss": 0.5843216180801392
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18511147797107697,
      "epoch": 14.04,
      "learning_rate": 6.197474925873525e-06,
      "loss": 0.2269,
      "step": 38860,
      "task_loss": 0.2638187110424042
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1563921570777893,
      "epoch": 14.05,
      "learning_rate": 6.1523951389232715e-06,
      "loss": 0.22,
      "step": 38870,
      "task_loss": 0.43536078929901123
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18232986330986023,
      "epoch": 14.05,
      "learning_rate": 6.107464099924179e-06,
      "loss": 0.2148,
      "step": 38880,
      "task_loss": 0.36110395193099976
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2118646204471588,
      "epoch": 14.05,
      "learning_rate": 6.062682040555794e-06,
      "loss": 0.2408,
      "step": 38890,
      "task_loss": 0.8396679162979126
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.17698946595191956,
      "epoch": 14.06,
      "learning_rate": 6.018049191729475e-06,
      "loss": 0.2141,
      "step": 38900,
      "task_loss": 0.2846156060695648
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22236953675746918,
      "epoch": 14.06,
      "learning_rate": 5.9735657835871894e-06,
      "loss": 0.2393,
      "step": 38910,
      "task_loss": 0.6163221001625061
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21342119574546814,
      "epoch": 14.07,
      "learning_rate": 5.929232045500351e-06,
      "loss": 0.2244,
      "step": 38920,
      "task_loss": 0.48823344707489014
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2061123251914978,
      "epoch": 14.07,
      "learning_rate": 5.885048206068627e-06,
      "loss": 0.2401,
      "step": 38930,
      "task_loss": 0.4645582437515259
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22864103317260742,
      "epoch": 14.07,
      "learning_rate": 5.841014493118753e-06,
      "loss": 0.2497,
      "step": 38940,
      "task_loss": 0.8002902269363403
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1768096685409546,
      "epoch": 14.08,
      "learning_rate": 5.797131133703335e-06,
      "loss": 0.2108,
      "step": 38950,
      "task_loss": 0.3029367923736572
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1883765161037445,
      "epoch": 14.08,
      "learning_rate": 5.753398354099773e-06,
      "loss": 0.2286,
      "step": 38960,
      "task_loss": 0.6003336906433105
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.16090941429138184,
      "epoch": 14.08,
      "learning_rate": 5.7098163798089505e-06,
      "loss": 0.2301,
      "step": 38970,
      "task_loss": 1.0162906646728516
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21309834718704224,
      "epoch": 14.09,
      "learning_rate": 5.666385435554192e-06,
      "loss": 0.2746,
      "step": 38980,
      "task_loss": 0.6796290874481201
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20901435613632202,
      "epoch": 14.09,
      "learning_rate": 5.623105745280054e-06,
      "loss": 0.2181,
      "step": 38990,
      "task_loss": 0.4081113934516907
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2516651749610901,
      "epoch": 14.09,
      "learning_rate": 5.5799775321511795e-06,
      "loss": 0.2362,
      "step": 39000,
      "task_loss": 0.3016287684440613
    },
    {
      "epoch": 14.09,
      "eval_exact_match": 83.56669820245979,
      "eval_f1": 90.22665241010446,
      "step": 39000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21509164571762085,
      "epoch": 14.1,
      "learning_rate": 5.53700101855112e-06,
      "loss": 0.2353,
      "step": 39010,
      "task_loss": 0.5538349747657776
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24842707812786102,
      "epoch": 14.1,
      "learning_rate": 5.494176426081241e-06,
      "loss": 0.2412,
      "step": 39020,
      "task_loss": 0.48312896490097046
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25581663846969604,
      "epoch": 14.11,
      "learning_rate": 5.45150397555954e-06,
      "loss": 0.2317,
      "step": 39030,
      "task_loss": 0.7214528918266296
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22178195416927338,
      "epoch": 14.11,
      "learning_rate": 5.4089838870195325e-06,
      "loss": 0.2353,
      "step": 39040,
      "task_loss": 0.23669970035552979
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2143976390361786,
      "epoch": 14.11,
      "learning_rate": 5.366616379709066e-06,
      "loss": 0.2262,
      "step": 39050,
      "task_loss": 0.7643693685531616
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23287276923656464,
      "epoch": 14.12,
      "learning_rate": 5.324401672089289e-06,
      "loss": 0.2219,
      "step": 39060,
      "task_loss": 0.35092318058013916
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19542303681373596,
      "epoch": 14.12,
      "learning_rate": 5.282339981833406e-06,
      "loss": 0.2282,
      "step": 39070,
      "task_loss": 0.43859049677848816
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19342200458049774,
      "epoch": 14.12,
      "learning_rate": 5.240431525825641e-06,
      "loss": 0.238,
      "step": 39080,
      "task_loss": 0.4503197968006134
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2664431035518646,
      "epoch": 14.13,
      "learning_rate": 5.198676520160087e-06,
      "loss": 0.2364,
      "step": 39090,
      "task_loss": 0.7081995010375977
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2214469313621521,
      "epoch": 14.13,
      "learning_rate": 5.157075180139602e-06,
      "loss": 0.2593,
      "step": 39100,
      "task_loss": 0.6777288913726807
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19831112027168274,
      "epoch": 14.13,
      "learning_rate": 5.1156277202746694e-06,
      "loss": 0.2219,
      "step": 39110,
      "task_loss": 0.3114470839500427
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2304876744747162,
      "epoch": 14.14,
      "learning_rate": 5.074334354282335e-06,
      "loss": 0.2282,
      "step": 39120,
      "task_loss": 0.4329706132411957
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2472088485956192,
      "epoch": 14.14,
      "learning_rate": 5.033195295085081e-06,
      "loss": 0.2383,
      "step": 39130,
      "task_loss": 0.4525589048862457
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19407299160957336,
      "epoch": 14.15,
      "learning_rate": 4.992210754809733e-06,
      "loss": 0.2222,
      "step": 39140,
      "task_loss": 0.46681374311447144
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21169891953468323,
      "epoch": 14.15,
      "learning_rate": 4.951380944786361e-06,
      "loss": 0.2294,
      "step": 39150,
      "task_loss": 0.44812092185020447
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18998125195503235,
      "epoch": 14.15,
      "learning_rate": 4.910706075547198e-06,
      "loss": 0.203,
      "step": 39160,
      "task_loss": 0.23759853839874268
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23228806257247925,
      "epoch": 14.16,
      "learning_rate": 4.8701863568255366e-06,
      "loss": 0.2103,
      "step": 39170,
      "task_loss": 0.4678202271461487
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22845202684402466,
      "epoch": 14.16,
      "learning_rate": 4.829821997554683e-06,
      "loss": 0.2306,
      "step": 39180,
      "task_loss": 0.26507970690727234
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22435474395751953,
      "epoch": 14.16,
      "learning_rate": 4.789613205866839e-06,
      "loss": 0.2266,
      "step": 39190,
      "task_loss": 0.41273343563079834
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20098178088665009,
      "epoch": 14.17,
      "learning_rate": 4.749560189092066e-06,
      "loss": 0.2104,
      "step": 39200,
      "task_loss": 0.5112037658691406
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.15663176774978638,
      "epoch": 14.17,
      "learning_rate": 4.709663153757165e-06,
      "loss": 0.2332,
      "step": 39210,
      "task_loss": 0.496734082698822
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21804989874362946,
      "epoch": 14.17,
      "learning_rate": 4.669922305584701e-06,
      "loss": 0.2141,
      "step": 39220,
      "task_loss": 0.4116891026496887
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.15763679146766663,
      "epoch": 14.18,
      "learning_rate": 4.630337849491818e-06,
      "loss": 0.2189,
      "step": 39230,
      "task_loss": 0.22102904319763184
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2064918875694275,
      "epoch": 14.18,
      "learning_rate": 4.5909099895892995e-06,
      "loss": 0.227,
      "step": 39240,
      "task_loss": 0.31075674295425415
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21560850739479065,
      "epoch": 14.19,
      "learning_rate": 4.551638929180444e-06,
      "loss": 0.221,
      "step": 39250,
      "task_loss": 0.34478896856307983
    },
    {
      "epoch": 14.19,
      "eval_exact_match": 83.55723746452223,
      "eval_f1": 90.20349881769643,
      "step": 39250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1906319260597229,
      "epoch": 14.19,
      "learning_rate": 4.512524870760054e-06,
      "loss": 0.2139,
      "step": 39260,
      "task_loss": 0.35769015550613403
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25359436869621277,
      "epoch": 14.19,
      "learning_rate": 4.473568016013349e-06,
      "loss": 0.2418,
      "step": 39270,
      "task_loss": 0.5408420562744141
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1744784563779831,
      "epoch": 14.2,
      "learning_rate": 4.4347685658149885e-06,
      "loss": 0.2174,
      "step": 39280,
      "task_loss": 0.6951953768730164
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24684374034404755,
      "epoch": 14.2,
      "learning_rate": 4.396126720227975e-06,
      "loss": 0.2248,
      "step": 39290,
      "task_loss": 0.6458311080932617
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.15859639644622803,
      "epoch": 14.2,
      "learning_rate": 4.357642678502669e-06,
      "loss": 0.2302,
      "step": 39300,
      "task_loss": 0.617123007774353
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19645214080810547,
      "epoch": 14.21,
      "learning_rate": 4.319316639075711e-06,
      "loss": 0.2244,
      "step": 39310,
      "task_loss": 0.3386276066303253
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23448264598846436,
      "epoch": 14.21,
      "learning_rate": 4.281148799569073e-06,
      "loss": 0.2319,
      "step": 39320,
      "task_loss": 0.33614152669906616
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1803082823753357,
      "epoch": 14.21,
      "learning_rate": 4.243139356788951e-06,
      "loss": 0.2391,
      "step": 39330,
      "task_loss": 0.28528082370758057
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19339770078659058,
      "epoch": 14.22,
      "learning_rate": 4.205288506724823e-06,
      "loss": 0.2206,
      "step": 39340,
      "task_loss": 0.5189188718795776
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20033058524131775,
      "epoch": 14.22,
      "learning_rate": 4.167596444548396e-06,
      "loss": 0.2131,
      "step": 39350,
      "task_loss": 0.5193696022033691
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19192567467689514,
      "epoch": 14.22,
      "learning_rate": 4.130063364612621e-06,
      "loss": 0.2354,
      "step": 39360,
      "task_loss": 0.2575511336326599
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.16454698145389557,
      "epoch": 14.23,
      "learning_rate": 4.0926894604506725e-06,
      "loss": 0.2141,
      "step": 39370,
      "task_loss": 0.344363272190094
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24127215147018433,
      "epoch": 14.23,
      "learning_rate": 4.055474924774975e-06,
      "loss": 0.2321,
      "step": 39380,
      "task_loss": 0.5481629371643066
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2271345555782318,
      "epoch": 14.24,
      "learning_rate": 4.01841994947618e-06,
      "loss": 0.228,
      "step": 39390,
      "task_loss": 0.5968689918518066
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23536965250968933,
      "epoch": 14.24,
      "learning_rate": 3.981524725622215e-06,
      "loss": 0.22,
      "step": 39400,
      "task_loss": 0.4662066102027893
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22390343248844147,
      "epoch": 14.24,
      "learning_rate": 3.944789443457237e-06,
      "loss": 0.2461,
      "step": 39410,
      "task_loss": 0.687394917011261
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23100188374519348,
      "epoch": 14.25,
      "learning_rate": 3.908214292400743e-06,
      "loss": 0.2261,
      "step": 39420,
      "task_loss": 1.0438381433486938
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20339612662792206,
      "epoch": 14.25,
      "learning_rate": 3.871799461046491e-06,
      "loss": 0.2311,
      "step": 39430,
      "task_loss": 0.3669714629650116
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1861114203929901,
      "epoch": 14.25,
      "learning_rate": 3.8355451371616e-06,
      "loss": 0.2234,
      "step": 39440,
      "task_loss": 0.5127718448638916
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.17946402728557587,
      "epoch": 14.26,
      "learning_rate": 3.799451507685557e-06,
      "loss": 0.2174,
      "step": 39450,
      "task_loss": 0.5029610991477966
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2114979326725006,
      "epoch": 14.26,
      "learning_rate": 3.7635187587292616e-06,
      "loss": 0.2317,
      "step": 39460,
      "task_loss": 0.777336597442627
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24398508667945862,
      "epoch": 14.26,
      "learning_rate": 3.727747075574033e-06,
      "loss": 0.2516,
      "step": 39470,
      "task_loss": 0.46529749035835266
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23544840514659882,
      "epoch": 14.27,
      "learning_rate": 3.6921366426707064e-06,
      "loss": 0.2317,
      "step": 39480,
      "task_loss": 0.6463897228240967
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23673115670681,
      "epoch": 14.27,
      "learning_rate": 3.656687643638646e-06,
      "loss": 0.2184,
      "step": 39490,
      "task_loss": 0.5956941843032837
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21421056985855103,
      "epoch": 14.28,
      "learning_rate": 3.6214002612648187e-06,
      "loss": 0.2351,
      "step": 39500,
      "task_loss": 0.561495304107666
    },
    {
      "epoch": 14.28,
      "eval_exact_match": 83.55723746452223,
      "eval_f1": 90.22059674354702,
      "step": 39500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21173754334449768,
      "epoch": 14.28,
      "learning_rate": 3.5862746775028274e-06,
      "loss": 0.2371,
      "step": 39510,
      "task_loss": 0.5109820365905762
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2444034218788147,
      "epoch": 14.28,
      "learning_rate": 3.5513110734719947e-06,
      "loss": 0.2274,
      "step": 39520,
      "task_loss": 0.3416980504989624
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22895483672618866,
      "epoch": 14.29,
      "learning_rate": 3.5165096294564294e-06,
      "loss": 0.2376,
      "step": 39530,
      "task_loss": 0.4999857246875763
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19510109722614288,
      "epoch": 14.29,
      "learning_rate": 3.481870524904063e-06,
      "loss": 0.2148,
      "step": 39540,
      "task_loss": 0.38726818561553955
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18519558012485504,
      "epoch": 14.29,
      "learning_rate": 3.4473939384257713e-06,
      "loss": 0.2241,
      "step": 39550,
      "task_loss": 0.305391401052475
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2279016077518463,
      "epoch": 14.3,
      "learning_rate": 3.4130800477944353e-06,
      "loss": 0.2413,
      "step": 39560,
      "task_loss": 0.4362182021141052
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19982075691223145,
      "epoch": 14.3,
      "learning_rate": 3.378929029944014e-06,
      "loss": 0.2107,
      "step": 39570,
      "task_loss": 0.4617334008216858
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.17380177974700928,
      "epoch": 14.3,
      "learning_rate": 3.3449410609686222e-06,
      "loss": 0.225,
      "step": 39580,
      "task_loss": 0.5341837406158447
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21998389065265656,
      "epoch": 14.31,
      "learning_rate": 3.311116316121681e-06,
      "loss": 0.2425,
      "step": 39590,
      "task_loss": 0.48477378487586975
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2229992151260376,
      "epoch": 14.31,
      "learning_rate": 3.27745496981493e-06,
      "loss": 0.2219,
      "step": 39600,
      "task_loss": 0.3780628442764282
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.280860960483551,
      "epoch": 14.32,
      "learning_rate": 3.2439571956175955e-06,
      "loss": 0.236,
      "step": 39610,
      "task_loss": 0.7838178873062134
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.16504058241844177,
      "epoch": 14.32,
      "learning_rate": 3.2106231662554687e-06,
      "loss": 0.2116,
      "step": 39620,
      "task_loss": 0.38846850395202637
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18678365647792816,
      "epoch": 14.32,
      "learning_rate": 3.17745305361002e-06,
      "loss": 0.2178,
      "step": 39630,
      "task_loss": 0.3168988525867462
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24787138402462006,
      "epoch": 14.33,
      "learning_rate": 3.144447028717492e-06,
      "loss": 0.2207,
      "step": 39640,
      "task_loss": 0.3311496376991272
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20401303470134735,
      "epoch": 14.33,
      "learning_rate": 3.1116052617680475e-06,
      "loss": 0.2311,
      "step": 39650,
      "task_loss": 0.2993205189704895
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2255595624446869,
      "epoch": 14.33,
      "learning_rate": 3.078927922104888e-06,
      "loss": 0.2449,
      "step": 39660,
      "task_loss": 0.6406166553497314
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26740413904190063,
      "epoch": 14.34,
      "learning_rate": 3.04641517822337e-06,
      "loss": 0.2378,
      "step": 39670,
      "task_loss": 0.8494973182678223
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.190795436501503,
      "epoch": 14.34,
      "learning_rate": 3.0140671977701126e-06,
      "loss": 0.2478,
      "step": 39680,
      "task_loss": 0.43718254566192627
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24205990135669708,
      "epoch": 14.34,
      "learning_rate": 2.9818841475422074e-06,
      "loss": 0.2441,
      "step": 39690,
      "task_loss": 0.549262285232544
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24730156362056732,
      "epoch": 14.35,
      "learning_rate": 2.9498661934862625e-06,
      "loss": 0.2497,
      "step": 39700,
      "task_loss": 0.48019862174987793
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20119762420654297,
      "epoch": 14.35,
      "learning_rate": 2.9180135006976253e-06,
      "loss": 0.2274,
      "step": 39710,
      "task_loss": 0.28718143701553345
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22375862300395966,
      "epoch": 14.35,
      "learning_rate": 2.8863262334194892e-06,
      "loss": 0.2073,
      "step": 39720,
      "task_loss": 0.5566632747650146
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2002100944519043,
      "epoch": 14.36,
      "learning_rate": 2.854804555042066e-06,
      "loss": 0.2189,
      "step": 39730,
      "task_loss": 0.43000146746635437
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19941216707229614,
      "epoch": 14.36,
      "learning_rate": 2.823448628101714e-06,
      "loss": 0.219,
      "step": 39740,
      "task_loss": 0.43014416098594666
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20073619484901428,
      "epoch": 14.37,
      "learning_rate": 2.792258614280147e-06,
      "loss": 0.2455,
      "step": 39750,
      "task_loss": 0.4905133843421936
    },
    {
      "epoch": 14.37,
      "eval_exact_match": 83.61400189214758,
      "eval_f1": 90.27336271124364,
      "step": 39750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18684911727905273,
      "epoch": 14.37,
      "learning_rate": 2.761234674403565e-06,
      "loss": 0.225,
      "step": 39760,
      "task_loss": 0.3729088306427002
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2193862795829773,
      "epoch": 14.37,
      "learning_rate": 2.730376968441837e-06,
      "loss": 0.2194,
      "step": 39770,
      "task_loss": 0.48270559310913086
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19356009364128113,
      "epoch": 14.38,
      "learning_rate": 2.699685655507652e-06,
      "loss": 0.2208,
      "step": 39780,
      "task_loss": 0.43202751874923706
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2428002506494522,
      "epoch": 14.38,
      "learning_rate": 2.6691608938557673e-06,
      "loss": 0.2174,
      "step": 39790,
      "task_loss": 0.57557213306427
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18401771783828735,
      "epoch": 14.38,
      "learning_rate": 2.638802840882099e-06,
      "loss": 0.2166,
      "step": 39800,
      "task_loss": 0.38599061965942383
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25447237491607666,
      "epoch": 14.39,
      "learning_rate": 2.608611653122982e-06,
      "loss": 0.2425,
      "step": 39810,
      "task_loss": 0.964231014251709
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2148197889328003,
      "epoch": 14.39,
      "learning_rate": 2.5785874862543364e-06,
      "loss": 0.2342,
      "step": 39820,
      "task_loss": 0.5468176603317261
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.275215744972229,
      "epoch": 14.39,
      "learning_rate": 2.548730495090867e-06,
      "loss": 0.2414,
      "step": 39830,
      "task_loss": 0.6998399496078491
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20677562057971954,
      "epoch": 14.4,
      "learning_rate": 2.5190408335852423e-06,
      "loss": 0.2275,
      "step": 39840,
      "task_loss": 0.8021308183670044
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2932314872741699,
      "epoch": 14.4,
      "learning_rate": 2.4895186548273553e-06,
      "loss": 0.2277,
      "step": 39850,
      "task_loss": 0.35569363832473755
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.16546630859375,
      "epoch": 14.41,
      "learning_rate": 2.4601641110434842e-06,
      "loss": 0.2139,
      "step": 39860,
      "task_loss": 0.3520973324775696
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20243391394615173,
      "epoch": 14.41,
      "learning_rate": 2.430977353595531e-06,
      "loss": 0.2541,
      "step": 39870,
      "task_loss": 0.41768544912338257
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21307632327079773,
      "epoch": 14.41,
      "learning_rate": 2.401958532980213e-06,
      "loss": 0.2169,
      "step": 39880,
      "task_loss": 0.7245582342147827
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22703363001346588,
      "epoch": 14.42,
      "learning_rate": 2.3731077988283558e-06,
      "loss": 0.2314,
      "step": 39890,
      "task_loss": 0.3949422240257263
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28492501378059387,
      "epoch": 14.42,
      "learning_rate": 2.34442529990403e-06,
      "loss": 0.2391,
      "step": 39900,
      "task_loss": 0.8998227715492249
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22178995609283447,
      "epoch": 14.42,
      "learning_rate": 2.315911184103842e-06,
      "loss": 0.2465,
      "step": 39910,
      "task_loss": 0.5418390035629272
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18975073099136353,
      "epoch": 14.43,
      "learning_rate": 2.287565598456169e-06,
      "loss": 0.2399,
      "step": 39920,
      "task_loss": 0.41155844926834106
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.16683003306388855,
      "epoch": 14.43,
      "learning_rate": 2.259388689120386e-06,
      "loss": 0.225,
      "step": 39930,
      "task_loss": 0.4172460436820984
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19753775000572205,
      "epoch": 14.43,
      "learning_rate": 2.2313806013861007e-06,
      "loss": 0.223,
      "step": 39940,
      "task_loss": 0.580410361289978
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.15868158638477325,
      "epoch": 14.44,
      "learning_rate": 2.2035414796724365e-06,
      "loss": 0.2266,
      "step": 39950,
      "task_loss": 0.234841987490654
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1994379162788391,
      "epoch": 14.44,
      "learning_rate": 2.1758714675272663e-06,
      "loss": 0.2301,
      "step": 39960,
      "task_loss": 0.28453922271728516
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22968558967113495,
      "epoch": 14.45,
      "learning_rate": 2.1483707076264827e-06,
      "loss": 0.2385,
      "step": 39970,
      "task_loss": 0.34896284341812134
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2558647394180298,
      "epoch": 14.45,
      "learning_rate": 2.121039341773235e-06,
      "loss": 0.2324,
      "step": 39980,
      "task_loss": 0.5354092121124268
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1978209763765335,
      "epoch": 14.45,
      "learning_rate": 2.0938775108972628e-06,
      "loss": 0.2397,
      "step": 39990,
      "task_loss": 0.3964638411998749
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21577095985412598,
      "epoch": 14.46,
      "learning_rate": 2.0668853550540775e-06,
      "loss": 0.2277,
      "step": 40000,
      "task_loss": 0.24615420401096344
    },
    {
      "epoch": 14.46,
      "eval_exact_match": 83.37748344370861,
      "eval_f1": 90.24861808055506,
      "step": 40000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2140657901763916,
      "epoch": 14.46,
      "learning_rate": 2.0400630134243162e-06,
      "loss": 0.2277,
      "step": 40010,
      "task_loss": 0.3943939805030823
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28893810510635376,
      "epoch": 14.46,
      "learning_rate": 2.013410624312997e-06,
      "loss": 0.2403,
      "step": 40020,
      "task_loss": 0.6239956617355347
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21812288463115692,
      "epoch": 14.47,
      "learning_rate": 1.9869283251488022e-06,
      "loss": 0.2467,
      "step": 40030,
      "task_loss": 0.9183149337768555
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2515392601490021,
      "epoch": 14.47,
      "learning_rate": 1.9606162524833615e-06,
      "loss": 0.241,
      "step": 40040,
      "task_loss": 0.2775781750679016
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20658659934997559,
      "epoch": 14.47,
      "learning_rate": 1.9344745419905726e-06,
      "loss": 0.2436,
      "step": 40050,
      "task_loss": 0.43849068880081177
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18757006525993347,
      "epoch": 14.48,
      "learning_rate": 1.9085033284658917e-06,
      "loss": 0.2378,
      "step": 40060,
      "task_loss": 0.36795151233673096
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2144419550895691,
      "epoch": 14.48,
      "learning_rate": 1.8827027458256239e-06,
      "loss": 0.2404,
      "step": 40070,
      "task_loss": 0.2985188961029053
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22410696744918823,
      "epoch": 14.49,
      "learning_rate": 1.857072927106244e-06,
      "loss": 0.2361,
      "step": 40080,
      "task_loss": 0.7707335948944092
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2690439820289612,
      "epoch": 14.49,
      "learning_rate": 1.831614004463733e-06,
      "loss": 0.2433,
      "step": 40090,
      "task_loss": 0.44754502177238464
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2019033432006836,
      "epoch": 14.49,
      "learning_rate": 1.806326109172838e-06,
      "loss": 0.2378,
      "step": 40100,
      "task_loss": 0.4766843914985657
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2467237263917923,
      "epoch": 14.5,
      "learning_rate": 1.781209371626461e-06,
      "loss": 0.2516,
      "step": 40110,
      "task_loss": 0.5889577865600586
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20924049615859985,
      "epoch": 14.5,
      "learning_rate": 1.75626392133494e-06,
      "loss": 0.2475,
      "step": 40120,
      "task_loss": 0.3500496745109558
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23063108325004578,
      "epoch": 14.5,
      "learning_rate": 1.7314898869254157e-06,
      "loss": 0.2097,
      "step": 40130,
      "task_loss": 0.8268368244171143
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19039425253868103,
      "epoch": 14.51,
      "learning_rate": 1.7068873961411294e-06,
      "loss": 0.2395,
      "step": 40140,
      "task_loss": 0.4907156527042389
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1660790741443634,
      "epoch": 14.51,
      "learning_rate": 1.6824565758408027e-06,
      "loss": 0.2432,
      "step": 40150,
      "task_loss": 0.385088711977005
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2549957036972046,
      "epoch": 14.51,
      "learning_rate": 1.6581975519979562e-06,
      "loss": 0.2305,
      "step": 40160,
      "task_loss": 0.37437817454338074
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22340743243694305,
      "epoch": 14.52,
      "learning_rate": 1.6341104497002816e-06,
      "loss": 0.2425,
      "step": 40170,
      "task_loss": 0.4963937997817993
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23421865701675415,
      "epoch": 14.52,
      "learning_rate": 1.6101953931489643e-06,
      "loss": 0.2153,
      "step": 40180,
      "task_loss": 0.5987656116485596
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20578370988368988,
      "epoch": 14.52,
      "learning_rate": 1.5864525056580967e-06,
      "loss": 0.2197,
      "step": 40190,
      "task_loss": 0.49276524782180786
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21253164112567902,
      "epoch": 14.53,
      "learning_rate": 1.5628819096539764e-06,
      "loss": 0.2373,
      "step": 40200,
      "task_loss": 0.39236021041870117
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25942492485046387,
      "epoch": 14.53,
      "learning_rate": 1.5394837266745246e-06,
      "loss": 0.2421,
      "step": 40210,
      "task_loss": 0.8538446426391602
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.197534441947937,
      "epoch": 14.54,
      "learning_rate": 1.5162580773686438e-06,
      "loss": 0.2345,
      "step": 40220,
      "task_loss": 0.390971302986145
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2024911344051361,
      "epoch": 14.54,
      "learning_rate": 1.4932050814955942e-06,
      "loss": 0.2328,
      "step": 40230,
      "task_loss": 0.24561244249343872
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23663444817066193,
      "epoch": 14.54,
      "learning_rate": 1.470324857924365e-06,
      "loss": 0.2561,
      "step": 40240,
      "task_loss": 0.2838215231895447
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.17373546957969666,
      "epoch": 14.55,
      "learning_rate": 1.4476175246330906e-06,
      "loss": 0.2219,
      "step": 40250,
      "task_loss": 0.3289911150932312
    },
    {
      "epoch": 14.55,
      "eval_exact_match": 83.72753074739829,
      "eval_f1": 90.29893044950633,
      "step": 40250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19693242013454437,
      "epoch": 14.55,
      "learning_rate": 1.4250831987084155e-06,
      "loss": 0.232,
      "step": 40260,
      "task_loss": 0.5237109661102295
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25203973054885864,
      "epoch": 14.55,
      "learning_rate": 1.4027219963449087e-06,
      "loss": 0.2456,
      "step": 40270,
      "task_loss": 0.6143960356712341
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19381505250930786,
      "epoch": 14.56,
      "learning_rate": 1.3805340328444287e-06,
      "loss": 0.2183,
      "step": 40280,
      "task_loss": 0.33310258388519287
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18629619479179382,
      "epoch": 14.56,
      "learning_rate": 1.358519422615601e-06,
      "loss": 0.2311,
      "step": 40290,
      "task_loss": 0.2998196482658386
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23643043637275696,
      "epoch": 14.56,
      "learning_rate": 1.3366782791731396e-06,
      "loss": 0.2306,
      "step": 40300,
      "task_loss": 0.6539130210876465
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.17593146860599518,
      "epoch": 14.57,
      "learning_rate": 1.315010715137333e-06,
      "loss": 0.2361,
      "step": 40310,
      "task_loss": 0.36985522508621216
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24259349703788757,
      "epoch": 14.57,
      "learning_rate": 1.2935168422334307e-06,
      "loss": 0.226,
      "step": 40320,
      "task_loss": 0.6864180564880371
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2658407688140869,
      "epoch": 14.58,
      "learning_rate": 1.2721967712910648e-06,
      "loss": 0.2228,
      "step": 40330,
      "task_loss": 0.5224426984786987
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2879309058189392,
      "epoch": 14.58,
      "learning_rate": 1.2510506122436886e-06,
      "loss": 0.2457,
      "step": 40340,
      "task_loss": 0.31292012333869934
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20499932765960693,
      "epoch": 14.58,
      "learning_rate": 1.2300784741280078e-06,
      "loss": 0.241,
      "step": 40350,
      "task_loss": 0.533927857875824
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21396301686763763,
      "epoch": 14.59,
      "learning_rate": 1.209280465083427e-06,
      "loss": 0.2439,
      "step": 40360,
      "task_loss": 0.6710692048072815
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22768867015838623,
      "epoch": 14.59,
      "learning_rate": 1.1886566923514634e-06,
      "loss": 0.2508,
      "step": 40370,
      "task_loss": 0.5868658423423767
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.16053186357021332,
      "epoch": 14.59,
      "learning_rate": 1.1682072622752342e-06,
      "loss": 0.2185,
      "step": 40380,
      "task_loss": 0.35523098707199097
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.17393621802330017,
      "epoch": 14.6,
      "learning_rate": 1.147932280298877e-06,
      "loss": 0.2215,
      "step": 40390,
      "task_loss": 0.25129181146621704
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18951040506362915,
      "epoch": 14.6,
      "learning_rate": 1.127831850967007e-06,
      "loss": 0.2484,
      "step": 40400,
      "task_loss": 0.7178854942321777
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31038203835487366,
      "epoch": 14.6,
      "learning_rate": 1.1079060779242022e-06,
      "loss": 0.2424,
      "step": 40410,
      "task_loss": 0.8680423498153687
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19648787379264832,
      "epoch": 14.61,
      "learning_rate": 1.0881550639144531e-06,
      "loss": 0.2384,
      "step": 40420,
      "task_loss": 0.40625467896461487
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23141324520111084,
      "epoch": 14.61,
      "learning_rate": 1.0685789107806264e-06,
      "loss": 0.2243,
      "step": 40430,
      "task_loss": 0.5935995578765869
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24341002106666565,
      "epoch": 14.62,
      "learning_rate": 1.0491777194639433e-06,
      "loss": 0.2549,
      "step": 40440,
      "task_loss": 0.7823227643966675
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20551814138889313,
      "epoch": 14.62,
      "learning_rate": 1.029951590003487e-06,
      "loss": 0.2144,
      "step": 40450,
      "task_loss": 0.535841166973114
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19820314645767212,
      "epoch": 14.62,
      "learning_rate": 1.0109006215356314e-06,
      "loss": 0.2259,
      "step": 40460,
      "task_loss": 0.4061843454837799
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20060187578201294,
      "epoch": 14.63,
      "learning_rate": 9.920249122935775e-07,
      "loss": 0.2362,
      "step": 40470,
      "task_loss": 0.34745022654533386
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19940611720085144,
      "epoch": 14.63,
      "learning_rate": 9.733245596068346e-07,
      "loss": 0.2439,
      "step": 40480,
      "task_loss": 0.47528618574142456
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.16198787093162537,
      "epoch": 14.63,
      "learning_rate": 9.547996599007135e-07,
      "loss": 0.2246,
      "step": 40490,
      "task_loss": 0.47620028257369995
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.265262633562088,
      "epoch": 14.64,
      "learning_rate": 9.364503086958147e-07,
      "loss": 0.2355,
      "step": 40500,
      "task_loss": 0.40639764070510864
    },
    {
      "epoch": 14.64,
      "eval_exact_match": 83.68968779564806,
      "eval_f1": 90.2170241386278,
      "step": 40500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22631597518920898,
      "epoch": 14.64,
      "learning_rate": 9.182766006075692e-07,
      "loss": 0.2346,
      "step": 40510,
      "task_loss": 0.6658477187156677
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27636265754699707,
      "epoch": 14.64,
      "learning_rate": 9.002786293457259e-07,
      "loss": 0.2253,
      "step": 40520,
      "task_loss": 0.48663219809532166
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.231238454580307,
      "epoch": 14.65,
      "learning_rate": 8.824564877138851e-07,
      "loss": 0.2379,
      "step": 40530,
      "task_loss": 0.5162807703018188
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23019526898860931,
      "epoch": 14.65,
      "learning_rate": 8.648102676089857e-07,
      "loss": 0.2215,
      "step": 40540,
      "task_loss": 0.7763354778289795
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20232398808002472,
      "epoch": 14.65,
      "learning_rate": 8.473400600208857e-07,
      "loss": 0.22,
      "step": 40550,
      "task_loss": 0.3750525712966919
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21048077940940857,
      "epoch": 14.66,
      "learning_rate": 8.30045955031845e-07,
      "loss": 0.236,
      "step": 40560,
      "task_loss": 0.27062106132507324
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22069762647151947,
      "epoch": 14.66,
      "learning_rate": 8.129280418160867e-07,
      "loss": 0.2234,
      "step": 40570,
      "task_loss": 0.44164735078811646
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20091667771339417,
      "epoch": 14.67,
      "learning_rate": 7.959864086393225e-07,
      "loss": 0.2289,
      "step": 40580,
      "task_loss": 0.5987957715988159
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21585312485694885,
      "epoch": 14.67,
      "learning_rate": 7.792211428583184e-07,
      "loss": 0.2481,
      "step": 40590,
      "task_loss": 0.4686650037765503
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19087418913841248,
      "epoch": 14.67,
      "learning_rate": 7.626323309204157e-07,
      "loss": 0.2306,
      "step": 40600,
      "task_loss": 0.6071935892105103
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2141316682100296,
      "epoch": 14.68,
      "learning_rate": 7.462200583631045e-07,
      "loss": 0.2395,
      "step": 40610,
      "task_loss": 0.5493015050888062
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24774694442749023,
      "epoch": 14.68,
      "learning_rate": 7.299844098135915e-07,
      "loss": 0.2227,
      "step": 40620,
      "task_loss": 0.5905795097351074
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19672465324401855,
      "epoch": 14.68,
      "learning_rate": 7.139254689883305e-07,
      "loss": 0.2171,
      "step": 40630,
      "task_loss": 0.7069876194000244
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20999035239219666,
      "epoch": 14.69,
      "learning_rate": 6.980433186926221e-07,
      "loss": 0.2293,
      "step": 40640,
      "task_loss": 0.24222975969314575
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23579585552215576,
      "epoch": 14.69,
      "learning_rate": 6.823380408201817e-07,
      "loss": 0.243,
      "step": 40650,
      "task_loss": 0.5637131929397583
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21097102761268616,
      "epoch": 14.69,
      "learning_rate": 6.668097163526936e-07,
      "loss": 0.2402,
      "step": 40660,
      "task_loss": 0.7529999017715454
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2499391883611679,
      "epoch": 14.7,
      "learning_rate": 6.514584253594218e-07,
      "loss": 0.2128,
      "step": 40670,
      "task_loss": 0.42130064964294434
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1964101791381836,
      "epoch": 14.7,
      "learning_rate": 6.362842469967905e-07,
      "loss": 0.238,
      "step": 40680,
      "task_loss": 0.38533705472946167
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1837211400270462,
      "epoch": 14.71,
      "learning_rate": 6.212872595079643e-07,
      "loss": 0.2251,
      "step": 40690,
      "task_loss": 0.638512372970581
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19221729040145874,
      "epoch": 14.71,
      "learning_rate": 6.064675402224444e-07,
      "loss": 0.2334,
      "step": 40700,
      "task_loss": 0.5338441133499146
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23053930699825287,
      "epoch": 14.71,
      "learning_rate": 5.918251655556994e-07,
      "loss": 0.2217,
      "step": 40710,
      "task_loss": 0.5507243275642395
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2665104269981384,
      "epoch": 14.72,
      "learning_rate": 5.773602110087295e-07,
      "loss": 0.2192,
      "step": 40720,
      "task_loss": 0.5469443202018738
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20541295409202576,
      "epoch": 14.72,
      "learning_rate": 5.630727511677097e-07,
      "loss": 0.2415,
      "step": 40730,
      "task_loss": 0.3799017667770386
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.186172217130661,
      "epoch": 14.72,
      "learning_rate": 5.489628597035817e-07,
      "loss": 0.2563,
      "step": 40740,
      "task_loss": 0.44071048498153687
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2783205509185791,
      "epoch": 14.73,
      "learning_rate": 5.350306093716961e-07,
      "loss": 0.2394,
      "step": 40750,
      "task_loss": 0.34275901317596436
    },
    {
      "epoch": 14.73,
      "eval_exact_match": 83.60454115421003,
      "eval_f1": 90.33495064433302,
      "step": 40750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2505665123462677,
      "epoch": 14.73,
      "learning_rate": 5.212760720114123e-07,
      "loss": 0.2319,
      "step": 40760,
      "task_loss": 0.45389094948768616
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24596086144447327,
      "epoch": 14.73,
      "learning_rate": 5.076993185457568e-07,
      "loss": 0.2276,
      "step": 40770,
      "task_loss": 0.7891265153884888
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2095947414636612,
      "epoch": 14.74,
      "learning_rate": 4.943004189810379e-07,
      "loss": 0.2466,
      "step": 40780,
      "task_loss": 0.546643853187561
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2007887363433838,
      "epoch": 14.74,
      "learning_rate": 4.810794424064813e-07,
      "loss": 0.2357,
      "step": 40790,
      "task_loss": 0.5920835733413696
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2600013315677643,
      "epoch": 14.75,
      "learning_rate": 4.6803645699389105e-07,
      "loss": 0.2341,
      "step": 40800,
      "task_loss": 0.507832407951355
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20566104352474213,
      "epoch": 14.75,
      "learning_rate": 4.551715299972852e-07,
      "loss": 0.2377,
      "step": 40810,
      "task_loss": 0.4447738528251648
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.17498670518398285,
      "epoch": 14.75,
      "learning_rate": 4.424847277525534e-07,
      "loss": 0.2181,
      "step": 40820,
      "task_loss": 0.2954162657260895
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23140192031860352,
      "epoch": 14.76,
      "learning_rate": 4.299761156771148e-07,
      "loss": 0.2491,
      "step": 40830,
      "task_loss": 0.5240877866744995
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.237901508808136,
      "epoch": 14.76,
      "learning_rate": 4.1764575826957634e-07,
      "loss": 0.2321,
      "step": 40840,
      "task_loss": 0.5805107355117798
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2360258400440216,
      "epoch": 14.76,
      "learning_rate": 4.054937191094143e-07,
      "loss": 0.226,
      "step": 40850,
      "task_loss": 0.4690331816673279
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.198323056101799,
      "epoch": 14.77,
      "learning_rate": 3.935200608566241e-07,
      "loss": 0.2309,
      "step": 40860,
      "task_loss": 0.4538593292236328
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18780264258384705,
      "epoch": 14.77,
      "learning_rate": 3.817248452514138e-07,
      "loss": 0.2387,
      "step": 40870,
      "task_loss": 0.43486571311950684
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2032957375049591,
      "epoch": 14.77,
      "learning_rate": 3.701081331138772e-07,
      "loss": 0.2278,
      "step": 40880,
      "task_loss": 0.46985048055648804
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21978767216205597,
      "epoch": 14.78,
      "learning_rate": 3.586699843436991e-07,
      "loss": 0.2466,
      "step": 40890,
      "task_loss": 0.5232217311859131
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2598414421081543,
      "epoch": 14.78,
      "learning_rate": 3.474104579198128e-07,
      "loss": 0.2476,
      "step": 40900,
      "task_loss": 0.3403189480304718
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1761493980884552,
      "epoch": 14.78,
      "learning_rate": 3.363296119001246e-07,
      "loss": 0.2419,
      "step": 40910,
      "task_loss": 0.2819960117340088
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21891748905181885,
      "epoch": 14.79,
      "learning_rate": 3.254275034212028e-07,
      "loss": 0.2313,
      "step": 40920,
      "task_loss": 0.6238331198692322
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19511379301548004,
      "epoch": 14.79,
      "learning_rate": 3.1470418869799007e-07,
      "loss": 0.2242,
      "step": 40930,
      "task_loss": 0.3012145161628723
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21911239624023438,
      "epoch": 14.8,
      "learning_rate": 3.041597230234888e-07,
      "loss": 0.2193,
      "step": 40940,
      "task_loss": 0.4811599850654602
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3620220422744751,
      "epoch": 14.8,
      "learning_rate": 2.9379416076852405e-07,
      "loss": 0.2397,
      "step": 40950,
      "task_loss": 0.6693317890167236
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2984312176704407,
      "epoch": 14.8,
      "learning_rate": 2.836075553814171e-07,
      "loss": 0.2383,
      "step": 40960,
      "task_loss": 0.8876379728317261
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20343519747257233,
      "epoch": 14.81,
      "learning_rate": 2.735999593877253e-07,
      "loss": 0.21,
      "step": 40970,
      "task_loss": 0.48944130539894104
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1964031308889389,
      "epoch": 14.81,
      "learning_rate": 2.6377142438998134e-07,
      "loss": 0.2216,
      "step": 40980,
      "task_loss": 0.39222803711891174
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24534821510314941,
      "epoch": 14.81,
      "learning_rate": 2.5412200106742166e-07,
      "loss": 0.2391,
      "step": 40990,
      "task_loss": 0.7029440402984619
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2845875322818756,
      "epoch": 14.82,
      "learning_rate": 2.4465173917571023e-07,
      "loss": 0.2691,
      "step": 41000,
      "task_loss": 0.5345797538757324
    },
    {
      "epoch": 14.82,
      "eval_exact_match": 83.4247871333964,
      "eval_f1": 90.21943083788699,
      "step": 41000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20766814053058624,
      "epoch": 14.82,
      "learning_rate": 2.3536068754670568e-07,
      "loss": 0.2328,
      "step": 41010,
      "task_loss": 0.43465369939804077
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22689062356948853,
      "epoch": 14.82,
      "learning_rate": 2.262488940881968e-07,
      "loss": 0.2223,
      "step": 41020,
      "task_loss": 0.578322172164917
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.16850513219833374,
      "epoch": 14.83,
      "learning_rate": 2.1731640578365016e-07,
      "loss": 0.2158,
      "step": 41030,
      "task_loss": 0.36379289627075195
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3516436517238617,
      "epoch": 14.83,
      "learning_rate": 2.0856326869198082e-07,
      "loss": 0.25,
      "step": 41040,
      "task_loss": 0.681757390499115
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20471343398094177,
      "epoch": 14.84,
      "learning_rate": 1.999895279473074e-07,
      "loss": 0.2079,
      "step": 41050,
      "task_loss": 0.4940853714942932
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24832800030708313,
      "epoch": 14.84,
      "learning_rate": 1.9159522775871906e-07,
      "loss": 0.2473,
      "step": 41060,
      "task_loss": 0.5383921265602112
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1713516116142273,
      "epoch": 14.84,
      "learning_rate": 1.83380411410054e-07,
      "loss": 0.2302,
      "step": 41070,
      "task_loss": 0.4822900593280792
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2330705225467682,
      "epoch": 14.85,
      "learning_rate": 1.7534512125966237e-07,
      "loss": 0.2271,
      "step": 41080,
      "task_loss": 0.46958887577056885
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1865522861480713,
      "epoch": 14.85,
      "learning_rate": 1.6748939874020818e-07,
      "loss": 0.2261,
      "step": 41090,
      "task_loss": 0.3367389142513275
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22266581654548645,
      "epoch": 14.85,
      "learning_rate": 1.598132843584321e-07,
      "loss": 0.2236,
      "step": 41100,
      "task_loss": 0.2846643328666687
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2285633385181427,
      "epoch": 14.86,
      "learning_rate": 1.5231681769496517e-07,
      "loss": 0.2501,
      "step": 41110,
      "task_loss": 0.41894304752349854
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18297666311264038,
      "epoch": 14.86,
      "learning_rate": 1.4500003740410715e-07,
      "loss": 0.2222,
      "step": 41120,
      "task_loss": 0.27748382091522217
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27647221088409424,
      "epoch": 14.86,
      "learning_rate": 1.3786298121364392e-07,
      "loss": 0.2177,
      "step": 41130,
      "task_loss": 0.48041605949401855
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2500835955142975,
      "epoch": 14.87,
      "learning_rate": 1.3090568592462603e-07,
      "loss": 0.2289,
      "step": 41140,
      "task_loss": 0.9432734251022339
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25542983412742615,
      "epoch": 14.87,
      "learning_rate": 1.24128187411221e-07,
      "loss": 0.233,
      "step": 41150,
      "task_loss": 0.25100335478782654
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.198891282081604,
      "epoch": 14.88,
      "learning_rate": 1.1753052062048018e-07,
      "loss": 0.2451,
      "step": 41160,
      "task_loss": 0.2514464557170868
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21026000380516052,
      "epoch": 14.88,
      "learning_rate": 1.1111271957219503e-07,
      "loss": 0.2173,
      "step": 41170,
      "task_loss": 0.7069941759109497
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19374729692935944,
      "epoch": 14.88,
      "learning_rate": 1.0487481735870662e-07,
      "loss": 0.2266,
      "step": 41180,
      "task_loss": 0.5104742646217346
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24391800165176392,
      "epoch": 14.89,
      "learning_rate": 9.881684614473861e-08,
      "loss": 0.23,
      "step": 41190,
      "task_loss": 0.6051548719406128
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21831145882606506,
      "epoch": 14.89,
      "learning_rate": 9.29388371672224e-08,
      "loss": 0.2269,
      "step": 41200,
      "task_loss": 0.44926178455352783
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18416348099708557,
      "epoch": 14.89,
      "learning_rate": 8.724082073515326e-08,
      "loss": 0.2154,
      "step": 41210,
      "task_loss": 0.4953691065311432
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20730862021446228,
      "epoch": 14.9,
      "learning_rate": 8.172282622941562e-08,
      "loss": 0.2198,
      "step": 41220,
      "task_loss": 0.3815509080886841
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21018816530704498,
      "epoch": 14.9,
      "learning_rate": 7.638488210265081e-08,
      "loss": 0.2203,
      "step": 41230,
      "task_loss": 0.5777589678764343
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1664731502532959,
      "epoch": 14.9,
      "learning_rate": 7.122701587908619e-08,
      "loss": 0.226,
      "step": 41240,
      "task_loss": 0.28747373819351196
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18536382913589478,
      "epoch": 14.91,
      "learning_rate": 6.62492541544224e-08,
      "loss": 0.2528,
      "step": 41250,
      "task_loss": 0.2921023368835449
    },
    {
      "epoch": 14.91,
      "eval_exact_match": 83.59508041627247,
      "eval_f1": 90.3394868872591,
      "step": 41250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23831024765968323,
      "epoch": 14.91,
      "learning_rate": 6.145162259566628e-08,
      "loss": 0.2462,
      "step": 41260,
      "task_loss": 0.546221137046814
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1984724998474121,
      "epoch": 14.92,
      "learning_rate": 5.683414594102209e-08,
      "loss": 0.2253,
      "step": 41270,
      "task_loss": 0.4562731683254242
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20618247985839844,
      "epoch": 14.92,
      "learning_rate": 5.2396847999751635e-08,
      "loss": 0.2312,
      "step": 41280,
      "task_loss": 0.5624033212661743
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2884097099304199,
      "epoch": 14.92,
      "learning_rate": 4.813975165205763e-08,
      "loss": 0.2308,
      "step": 41290,
      "task_loss": 0.6284477114677429
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20076477527618408,
      "epoch": 14.93,
      "learning_rate": 4.406287884895943e-08,
      "loss": 0.2219,
      "step": 41300,
      "task_loss": 0.31397148966789246
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19199396669864655,
      "epoch": 14.93,
      "learning_rate": 4.016625061218026e-08,
      "loss": 0.2335,
      "step": 41310,
      "task_loss": 0.2553127408027649
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19816777110099792,
      "epoch": 14.93,
      "learning_rate": 3.644988703405016e-08,
      "loss": 0.2214,
      "step": 41320,
      "task_loss": 0.3272134065628052
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21667525172233582,
      "epoch": 14.94,
      "learning_rate": 3.2913807277385464e-08,
      "loss": 0.2311,
      "step": 41330,
      "task_loss": 0.6902732849121094
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2011464536190033,
      "epoch": 14.94,
      "learning_rate": 2.9558029575407204e-08,
      "loss": 0.2092,
      "step": 41340,
      "task_loss": 0.38788002729415894
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19852641224861145,
      "epoch": 14.94,
      "learning_rate": 2.6382571231628456e-08,
      "loss": 0.2389,
      "step": 41350,
      "task_loss": 0.48264265060424805
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18369928002357483,
      "epoch": 14.95,
      "learning_rate": 2.3387448619784367e-08,
      "loss": 0.2286,
      "step": 41360,
      "task_loss": 0.4098126292228699
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.258108526468277,
      "epoch": 14.95,
      "learning_rate": 2.0572677183731122e-08,
      "loss": 0.2408,
      "step": 41370,
      "task_loss": 0.47995632886886597
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20995879173278809,
      "epoch": 14.95,
      "learning_rate": 1.7938271437379913e-08,
      "loss": 0.2241,
      "step": 41380,
      "task_loss": 0.8705676198005676
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1764739751815796,
      "epoch": 14.96,
      "learning_rate": 1.5484244964611427e-08,
      "loss": 0.2331,
      "step": 41390,
      "task_loss": 0.258797287940979
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20735256373882294,
      "epoch": 14.96,
      "learning_rate": 1.3210610419213673e-08,
      "loss": 0.2153,
      "step": 41400,
      "task_loss": 0.7630883455276489
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22570012509822845,
      "epoch": 14.97,
      "learning_rate": 1.1117379524804293e-08,
      "loss": 0.2368,
      "step": 41410,
      "task_loss": 0.5106080174446106
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19130569696426392,
      "epoch": 14.97,
      "learning_rate": 9.204563074791671e-09,
      "loss": 0.2303,
      "step": 41420,
      "task_loss": 0.5977288484573364
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30169206857681274,
      "epoch": 14.97,
      "learning_rate": 7.47217093229724e-09,
      "loss": 0.2319,
      "step": 41430,
      "task_loss": 0.6631218194961548
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22515720129013062,
      "epoch": 14.98,
      "learning_rate": 5.9202120301204966e-09,
      "loss": 0.2271,
      "step": 41440,
      "task_loss": 0.4315108358860016
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2272672951221466,
      "epoch": 14.98,
      "learning_rate": 4.548694370680728e-09,
      "loss": 0.2382,
      "step": 41450,
      "task_loss": 0.8516280651092529
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18287137150764465,
      "epoch": 14.98,
      "learning_rate": 3.3576250259859106e-09,
      "loss": 0.2305,
      "step": 41460,
      "task_loss": 0.19684766232967377
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22013694047927856,
      "epoch": 14.99,
      "learning_rate": 2.3470101375860917e-09,
      "loss": 0.228,
      "step": 41470,
      "task_loss": 0.7428228855133057
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27389436960220337,
      "epoch": 14.99,
      "learning_rate": 1.5168549165578415e-09,
      "loss": 0.2314,
      "step": 41480,
      "task_loss": 0.6218991279602051
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.17496860027313232,
      "epoch": 14.99,
      "learning_rate": 8.671636434576246e-10,
      "loss": 0.2295,
      "step": 41490,
      "task_loss": 0.33579161763191223
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22415469586849213,
      "epoch": 15.0,
      "learning_rate": 4.3674096117052304e-10,
      "loss": 0.2327,
      "step": 41500,
      "task_loss": 0.2707256078720093
    },
    {
      "epoch": 15.0,
      "eval_exact_match": 83.68968779564806,
      "eval_f1": 90.37926101768134,
      "step": 41500
    }
  ],
  "max_steps": 41505,
  "num_train_epochs": 15,
  "total_flos": 3.16732419781632e+16,
  "trial_name": null,
  "trial_params": null
}