{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 14.817491868449585,
  "global_step": 41000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "compression_loss": 0.0,
      "distillation_loss": 9.49934196472168,
      "epoch": 0.0,
      "learning_rate": 5.99999721557765e-05,
      "loss": 12.0065,
      "step": 10,
      "task_loss": 3.6033935546875
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 5.30767822265625,
      "epoch": 0.01,
      "learning_rate": 5.999980199682008e-05,
      "loss": 6.9237,
      "step": 20,
      "task_loss": 1.9013495445251465
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 2.916347026824951,
      "epoch": 0.01,
      "learning_rate": 5.9999477148796615e-05,
      "loss": 3.7725,
      "step": 30,
      "task_loss": 1.5464043617248535
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 2.6741557121276855,
      "epoch": 0.01,
      "learning_rate": 5.999899761338114e-05,
      "loss": 2.769,
      "step": 40,
      "task_loss": 1.156141996383667
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 2.1403422355651855,
      "epoch": 0.02,
      "learning_rate": 5.999836339304631e-05,
      "loss": 2.5443,
      "step": 50,
      "task_loss": 1.239074945449829
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 2.6367697715759277,
      "epoch": 0.02,
      "learning_rate": 5.9997574491062356e-05,
      "loss": 2.3399,
      "step": 60,
      "task_loss": 1.2156091928482056
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.863443374633789,
      "epoch": 0.03,
      "learning_rate": 5.999663091149714e-05,
      "loss": 2.2202,
      "step": 70,
      "task_loss": 1.239349126815796
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 2.0775623321533203,
      "epoch": 0.03,
      "learning_rate": 5.9995532659216076e-05,
      "loss": 2.0928,
      "step": 80,
      "task_loss": 1.497615098953247
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 2.05897855758667,
      "epoch": 0.03,
      "learning_rate": 5.9994279739882106e-05,
      "loss": 2.1269,
      "step": 90,
      "task_loss": 1.0416953563690186
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.7927725315093994,
      "epoch": 0.04,
      "learning_rate": 5.999287215995571e-05,
      "loss": 1.8854,
      "step": 100,
      "task_loss": 0.6938618421554565
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.8927876949310303,
      "epoch": 0.04,
      "learning_rate": 5.9991309926694865e-05,
      "loss": 1.8586,
      "step": 110,
      "task_loss": 1.012183666229248
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 2.0477404594421387,
      "epoch": 0.04,
      "learning_rate": 5.998959304815494e-05,
      "loss": 1.8545,
      "step": 120,
      "task_loss": 1.4201068878173828
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.5746378898620605,
      "epoch": 0.05,
      "learning_rate": 5.998772153318875e-05,
      "loss": 1.9129,
      "step": 130,
      "task_loss": 0.9690418243408203
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.8912169933319092,
      "epoch": 0.05,
      "learning_rate": 5.9985695391446455e-05,
      "loss": 1.76,
      "step": 140,
      "task_loss": 0.9980369806289673
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.6582715511322021,
      "epoch": 0.05,
      "learning_rate": 5.9983514633375525e-05,
      "loss": 1.8208,
      "step": 150,
      "task_loss": 0.9474058747291565
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.706885814666748,
      "epoch": 0.06,
      "learning_rate": 5.9981179270220685e-05,
      "loss": 1.6574,
      "step": 160,
      "task_loss": 0.8353824615478516
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 2.296937942504883,
      "epoch": 0.06,
      "learning_rate": 5.997868931402385e-05,
      "loss": 1.9116,
      "step": 170,
      "task_loss": 1.2023546695709229
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.843885064125061,
      "epoch": 0.07,
      "learning_rate": 5.997604477762407e-05,
      "loss": 1.75,
      "step": 180,
      "task_loss": 1.366844892501831
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.7979713678359985,
      "epoch": 0.07,
      "learning_rate": 5.997324567465745e-05,
      "loss": 1.7307,
      "step": 190,
      "task_loss": 0.9796158075332642
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 2.4352352619171143,
      "epoch": 0.07,
      "learning_rate": 5.997029201955714e-05,
      "loss": 1.6291,
      "step": 200,
      "task_loss": 1.3526904582977295
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.3992118835449219,
      "epoch": 0.08,
      "learning_rate": 5.996718382755315e-05,
      "loss": 1.6391,
      "step": 210,
      "task_loss": 0.7047762870788574
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.648317575454712,
      "epoch": 0.08,
      "learning_rate": 5.9963921114672385e-05,
      "loss": 1.5126,
      "step": 220,
      "task_loss": 1.3377418518066406
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.356798768043518,
      "epoch": 0.08,
      "learning_rate": 5.996050389773846e-05,
      "loss": 1.5781,
      "step": 230,
      "task_loss": 0.6622216105461121
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.5281479358673096,
      "epoch": 0.09,
      "learning_rate": 5.995693219437173e-05,
      "loss": 1.6074,
      "step": 240,
      "task_loss": 1.0915801525115967
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.4357974529266357,
      "epoch": 0.09,
      "learning_rate": 5.995320602298906e-05,
      "loss": 1.6211,
      "step": 250,
      "task_loss": 1.0162770748138428
    },
    {
      "epoch": 0.09,
      "eval_exact_match": 75.68590350047303,
      "eval_f1": 85.28012898809695,
      "step": 250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.83945894241333,
      "epoch": 0.09,
      "learning_rate": 5.9949325402803876e-05,
      "loss": 1.5693,
      "step": 260,
      "task_loss": 1.0681490898132324
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.4410802125930786,
      "epoch": 0.1,
      "learning_rate": 5.9945290353825946e-05,
      "loss": 1.5176,
      "step": 270,
      "task_loss": 0.9294320940971375
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.778608798980713,
      "epoch": 0.1,
      "learning_rate": 5.994110089686134e-05,
      "loss": 1.5048,
      "step": 280,
      "task_loss": 1.2308306694030762
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1298151016235352,
      "epoch": 0.1,
      "learning_rate": 5.993675705351232e-05,
      "loss": 1.5331,
      "step": 290,
      "task_loss": 0.5350617170333862
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.5367448329925537,
      "epoch": 0.11,
      "learning_rate": 5.993225884617717e-05,
      "loss": 1.5269,
      "step": 300,
      "task_loss": 0.7491666078567505
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.389266848564148,
      "epoch": 0.11,
      "learning_rate": 5.9927606298050194e-05,
      "loss": 1.4866,
      "step": 310,
      "task_loss": 0.8327745199203491
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.6235179901123047,
      "epoch": 0.12,
      "learning_rate": 5.9922799433121474e-05,
      "loss": 1.5065,
      "step": 320,
      "task_loss": 0.8440191149711609
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.564889669418335,
      "epoch": 0.12,
      "learning_rate": 5.9917838276176835e-05,
      "loss": 1.5237,
      "step": 330,
      "task_loss": 1.3443149328231812
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.447906494140625,
      "epoch": 0.12,
      "learning_rate": 5.9912722852797664e-05,
      "loss": 1.4332,
      "step": 340,
      "task_loss": 0.8888688087463379
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.415635108947754,
      "epoch": 0.13,
      "learning_rate": 5.99074531893608e-05,
      "loss": 1.4817,
      "step": 350,
      "task_loss": 0.8752025365829468
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.707700252532959,
      "epoch": 0.13,
      "learning_rate": 5.990202931303841e-05,
      "loss": 1.4789,
      "step": 360,
      "task_loss": 1.3695316314697266
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2083714008331299,
      "epoch": 0.13,
      "learning_rate": 5.989645125179782e-05,
      "loss": 1.5234,
      "step": 370,
      "task_loss": 0.7723703384399414
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.7325220108032227,
      "epoch": 0.14,
      "learning_rate": 5.989071903440139e-05,
      "loss": 1.499,
      "step": 380,
      "task_loss": 1.0906333923339844
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2659305334091187,
      "epoch": 0.14,
      "learning_rate": 5.988483269040635e-05,
      "loss": 1.4476,
      "step": 390,
      "task_loss": 0.84165358543396
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.5708644390106201,
      "epoch": 0.14,
      "learning_rate": 5.987879225016469e-05,
      "loss": 1.4519,
      "step": 400,
      "task_loss": 0.8206751346588135
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.410794734954834,
      "epoch": 0.15,
      "learning_rate": 5.987259774482292e-05,
      "loss": 1.4455,
      "step": 410,
      "task_loss": 0.7794616222381592
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.5559226274490356,
      "epoch": 0.15,
      "learning_rate": 5.986624920632203e-05,
      "loss": 1.5825,
      "step": 420,
      "task_loss": 1.1558645963668823
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.5958247184753418,
      "epoch": 0.16,
      "learning_rate": 5.9859746667397186e-05,
      "loss": 1.4642,
      "step": 430,
      "task_loss": 1.1452465057373047
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0762372016906738,
      "epoch": 0.16,
      "learning_rate": 5.985309016157769e-05,
      "loss": 1.4096,
      "step": 440,
      "task_loss": 0.7453725337982178
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.3428254127502441,
      "epoch": 0.16,
      "learning_rate": 5.9846279723186714e-05,
      "loss": 1.3398,
      "step": 450,
      "task_loss": 0.6058871746063232
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.6145907640457153,
      "epoch": 0.17,
      "learning_rate": 5.983931538734117e-05,
      "loss": 1.4124,
      "step": 460,
      "task_loss": 1.0071673393249512
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.4324638843536377,
      "epoch": 0.17,
      "learning_rate": 5.983219718995151e-05,
      "loss": 1.3421,
      "step": 470,
      "task_loss": 0.8209178447723389
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.294832468032837,
      "epoch": 0.17,
      "learning_rate": 5.982492516772156e-05,
      "loss": 1.3454,
      "step": 480,
      "task_loss": 1.0602484941482544
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.3344955444335938,
      "epoch": 0.18,
      "learning_rate": 5.9817499358148303e-05,
      "loss": 1.3674,
      "step": 490,
      "task_loss": 1.2323447465896606
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.31858229637146,
      "epoch": 0.18,
      "learning_rate": 5.980991979952172e-05,
      "loss": 1.4034,
      "step": 500,
      "task_loss": 0.8082805871963501
    },
    {
      "epoch": 0.18,
      "eval_exact_match": 76.43330179754021,
      "eval_f1": 85.8826809188487,
      "step": 500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.117767095565796,
      "epoch": 0.18,
      "learning_rate": 5.980218653092457e-05,
      "loss": 1.4255,
      "step": 510,
      "task_loss": 0.8356488943099976
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.4943193197250366,
      "epoch": 0.19,
      "learning_rate": 5.9794299592232164e-05,
      "loss": 1.3749,
      "step": 520,
      "task_loss": 0.6115718483924866
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.5185236930847168,
      "epoch": 0.19,
      "learning_rate": 5.978625902411224e-05,
      "loss": 1.37,
      "step": 530,
      "task_loss": 0.9306137561798096
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.3840581178665161,
      "epoch": 0.2,
      "learning_rate": 5.977806486802464e-05,
      "loss": 1.2457,
      "step": 540,
      "task_loss": 1.2043516635894775
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.6848634481430054,
      "epoch": 0.2,
      "learning_rate": 5.97697171662212e-05,
      "loss": 1.3547,
      "step": 550,
      "task_loss": 1.2931828498840332
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0846725702285767,
      "epoch": 0.2,
      "learning_rate": 5.9761215961745495e-05,
      "loss": 1.1797,
      "step": 560,
      "task_loss": 0.9227917194366455
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2260918617248535,
      "epoch": 0.21,
      "learning_rate": 5.9752561298432565e-05,
      "loss": 1.1848,
      "step": 570,
      "task_loss": 0.8526758551597595
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.9792821407318115,
      "epoch": 0.21,
      "learning_rate": 5.974375322090879e-05,
      "loss": 1.3254,
      "step": 580,
      "task_loss": 1.3365485668182373
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0318635702133179,
      "epoch": 0.21,
      "learning_rate": 5.973479177459156e-05,
      "loss": 1.249,
      "step": 590,
      "task_loss": 0.5771055817604065
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.460340142250061,
      "epoch": 0.22,
      "learning_rate": 5.972567700568911e-05,
      "loss": 1.4128,
      "step": 600,
      "task_loss": 0.593090295791626
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.5659215450286865,
      "epoch": 0.22,
      "learning_rate": 5.9716408961200254e-05,
      "loss": 1.3882,
      "step": 610,
      "task_loss": 1.2778445482254028
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.173753023147583,
      "epoch": 0.22,
      "learning_rate": 5.970698768891416e-05,
      "loss": 1.191,
      "step": 620,
      "task_loss": 0.5027570128440857
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.417959451675415,
      "epoch": 0.23,
      "learning_rate": 5.9697413237410065e-05,
      "loss": 1.4367,
      "step": 630,
      "task_loss": 0.7332655787467957
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.5544065237045288,
      "epoch": 0.23,
      "learning_rate": 5.968768565605706e-05,
      "loss": 1.3264,
      "step": 640,
      "task_loss": 0.8936142921447754
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.5732910633087158,
      "epoch": 0.23,
      "learning_rate": 5.9677804995013815e-05,
      "loss": 1.3731,
      "step": 650,
      "task_loss": 1.1154710054397583
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.3809391260147095,
      "epoch": 0.24,
      "learning_rate": 5.966777130522836e-05,
      "loss": 1.4006,
      "step": 660,
      "task_loss": 0.9069052338600159
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0899161100387573,
      "epoch": 0.24,
      "learning_rate": 5.965758463843777e-05,
      "loss": 1.2987,
      "step": 670,
      "task_loss": 0.8597618341445923
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.151376724243164,
      "epoch": 0.25,
      "learning_rate": 5.9647245047167914e-05,
      "loss": 1.305,
      "step": 680,
      "task_loss": 0.8611929416656494
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.3947758674621582,
      "epoch": 0.25,
      "learning_rate": 5.96367525847332e-05,
      "loss": 1.2468,
      "step": 690,
      "task_loss": 1.21012544631958
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.192865252494812,
      "epoch": 0.25,
      "learning_rate": 5.962610730523631e-05,
      "loss": 1.3553,
      "step": 700,
      "task_loss": 1.0180736780166626
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9713619947433472,
      "epoch": 0.26,
      "learning_rate": 5.961530926356787e-05,
      "loss": 1.3005,
      "step": 710,
      "task_loss": 1.0774257183074951
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.5023868083953857,
      "epoch": 0.26,
      "learning_rate": 5.960435851540622e-05,
      "loss": 1.3108,
      "step": 720,
      "task_loss": 1.0222753286361694
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.454642415046692,
      "epoch": 0.26,
      "learning_rate": 5.9593255117217096e-05,
      "loss": 1.3155,
      "step": 730,
      "task_loss": 0.8456030488014221
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9722533226013184,
      "epoch": 0.27,
      "learning_rate": 5.958199912625336e-05,
      "loss": 1.3135,
      "step": 740,
      "task_loss": 0.7934273481369019
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.4297173023223877,
      "epoch": 0.27,
      "learning_rate": 5.957059060055468e-05,
      "loss": 1.2445,
      "step": 750,
      "task_loss": 0.8182862997055054
    },
    {
      "epoch": 0.27,
      "eval_exact_match": 77.19962157048249,
      "eval_f1": 86.33401538349308,
      "step": 750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.6894798278808594,
      "epoch": 0.27,
      "learning_rate": 5.955902959894726e-05,
      "loss": 1.2795,
      "step": 760,
      "task_loss": 1.0540611743927002
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2129619121551514,
      "epoch": 0.28,
      "learning_rate": 5.954731618104348e-05,
      "loss": 1.1677,
      "step": 770,
      "task_loss": 1.3634464740753174
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2262523174285889,
      "epoch": 0.28,
      "learning_rate": 5.953545040724171e-05,
      "loss": 1.2771,
      "step": 780,
      "task_loss": 0.7795358896255493
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.224480390548706,
      "epoch": 0.29,
      "learning_rate": 5.952343233872584e-05,
      "loss": 1.2326,
      "step": 790,
      "task_loss": 0.8323132991790771
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2114362716674805,
      "epoch": 0.29,
      "learning_rate": 5.951126203746507e-05,
      "loss": 1.247,
      "step": 800,
      "task_loss": 0.9133867621421814
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1349849700927734,
      "epoch": 0.29,
      "learning_rate": 5.949893956621358e-05,
      "loss": 1.4212,
      "step": 810,
      "task_loss": 0.7154116630554199
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.6616898775100708,
      "epoch": 0.3,
      "learning_rate": 5.9486464988510165e-05,
      "loss": 1.2637,
      "step": 820,
      "task_loss": 1.1065783500671387
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.537660837173462,
      "epoch": 0.3,
      "learning_rate": 5.9473838368677945e-05,
      "loss": 1.3276,
      "step": 830,
      "task_loss": 1.0488613843917847
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.110896110534668,
      "epoch": 0.3,
      "learning_rate": 5.9461059771824025e-05,
      "loss": 1.2912,
      "step": 840,
      "task_loss": 0.5585556030273438
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.865274965763092,
      "epoch": 0.31,
      "learning_rate": 5.944812926383914e-05,
      "loss": 1.2452,
      "step": 850,
      "task_loss": 0.7370738387107849
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.265105962753296,
      "epoch": 0.31,
      "learning_rate": 5.943504691139734e-05,
      "loss": 1.2056,
      "step": 860,
      "task_loss": 1.171694278717041
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2932755947113037,
      "epoch": 0.31,
      "learning_rate": 5.942181278195563e-05,
      "loss": 1.2814,
      "step": 870,
      "task_loss": 0.8684425354003906
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1378037929534912,
      "epoch": 0.32,
      "learning_rate": 5.940842694375364e-05,
      "loss": 1.1371,
      "step": 880,
      "task_loss": 0.4964973032474518
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.041467547416687,
      "epoch": 0.32,
      "learning_rate": 5.9394889465813246e-05,
      "loss": 1.3367,
      "step": 890,
      "task_loss": 0.9691845178604126
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2773613929748535,
      "epoch": 0.33,
      "learning_rate": 5.9381200417938226e-05,
      "loss": 1.2445,
      "step": 900,
      "task_loss": 0.8152469992637634
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2654523849487305,
      "epoch": 0.33,
      "learning_rate": 5.936735987071393e-05,
      "loss": 1.3448,
      "step": 910,
      "task_loss": 0.8703353404998779
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.325675368309021,
      "epoch": 0.33,
      "learning_rate": 5.935336789550688e-05,
      "loss": 1.2264,
      "step": 920,
      "task_loss": 0.905992865562439
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1226941347122192,
      "epoch": 0.34,
      "learning_rate": 5.9339224564464384e-05,
      "loss": 1.2136,
      "step": 930,
      "task_loss": 0.8291558623313904
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 2.0841875076293945,
      "epoch": 0.34,
      "learning_rate": 5.932492995051423e-05,
      "loss": 1.3336,
      "step": 940,
      "task_loss": 1.5568690299987793
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2928574085235596,
      "epoch": 0.34,
      "learning_rate": 5.9310484127364235e-05,
      "loss": 1.2545,
      "step": 950,
      "task_loss": 0.6856122612953186
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.5862665176391602,
      "epoch": 0.35,
      "learning_rate": 5.9295887169501945e-05,
      "loss": 1.2767,
      "step": 960,
      "task_loss": 0.9553361535072327
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1855202913284302,
      "epoch": 0.35,
      "learning_rate": 5.9281139152194166e-05,
      "loss": 1.2932,
      "step": 970,
      "task_loss": 1.2013435363769531
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.276527762413025,
      "epoch": 0.35,
      "learning_rate": 5.9266240151486647e-05,
      "loss": 1.2817,
      "step": 980,
      "task_loss": 0.5954961776733398
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2427090406417847,
      "epoch": 0.36,
      "learning_rate": 5.925119024420363e-05,
      "loss": 1.108,
      "step": 990,
      "task_loss": 0.9799728989601135
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2510955333709717,
      "epoch": 0.36,
      "learning_rate": 5.9235989507947515e-05,
      "loss": 1.2554,
      "step": 1000,
      "task_loss": 1.1080634593963623
    },
    {
      "epoch": 0.36,
      "eval_exact_match": 77.90917691579943,
      "eval_f1": 86.79589371678661,
      "step": 1000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.3496589660644531,
      "epoch": 0.37,
      "learning_rate": 5.922063802109839e-05,
      "loss": 1.3452,
      "step": 1010,
      "task_loss": 0.8422398567199707
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.35902738571167,
      "epoch": 0.37,
      "learning_rate": 5.92051358628137e-05,
      "loss": 1.1973,
      "step": 1020,
      "task_loss": 1.0076375007629395
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9760231971740723,
      "epoch": 0.37,
      "learning_rate": 5.918948311302778e-05,
      "loss": 1.3059,
      "step": 1030,
      "task_loss": 0.6299121975898743
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.3105099201202393,
      "epoch": 0.38,
      "learning_rate": 5.917367985245147e-05,
      "loss": 1.242,
      "step": 1040,
      "task_loss": 1.0138111114501953
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1091604232788086,
      "epoch": 0.38,
      "learning_rate": 5.91577261625717e-05,
      "loss": 1.1236,
      "step": 1050,
      "task_loss": 1.2920585870742798
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.125967264175415,
      "epoch": 0.38,
      "learning_rate": 5.914162212565106e-05,
      "loss": 1.1758,
      "step": 1060,
      "task_loss": 0.6104905605316162
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1518621444702148,
      "epoch": 0.39,
      "learning_rate": 5.9125367824727376e-05,
      "loss": 1.2793,
      "step": 1070,
      "task_loss": 1.1348037719726562
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.4855108261108398,
      "epoch": 0.39,
      "learning_rate": 5.9108963343613286e-05,
      "loss": 1.2306,
      "step": 1080,
      "task_loss": 0.7915964722633362
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2316033840179443,
      "epoch": 0.39,
      "learning_rate": 5.909240876689582e-05,
      "loss": 1.3181,
      "step": 1090,
      "task_loss": 1.3942852020263672
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0107777118682861,
      "epoch": 0.4,
      "learning_rate": 5.9075704179935926e-05,
      "loss": 1.2463,
      "step": 1100,
      "task_loss": 0.9370169639587402
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1373622417449951,
      "epoch": 0.4,
      "learning_rate": 5.9058849668868097e-05,
      "loss": 1.2639,
      "step": 1110,
      "task_loss": 0.5862497091293335
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1231906414031982,
      "epoch": 0.4,
      "learning_rate": 5.904184532059984e-05,
      "loss": 1.1012,
      "step": 1120,
      "task_loss": 0.7552740573883057
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.4255497455596924,
      "epoch": 0.41,
      "learning_rate": 5.90246912228113e-05,
      "loss": 1.2399,
      "step": 1130,
      "task_loss": 1.2231707572937012
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.3374457359313965,
      "epoch": 0.41,
      "learning_rate": 5.9007387463954766e-05,
      "loss": 1.2758,
      "step": 1140,
      "task_loss": 0.9943552017211914
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0493338108062744,
      "epoch": 0.42,
      "learning_rate": 5.898993413325424e-05,
      "loss": 1.1956,
      "step": 1150,
      "task_loss": 0.7132070064544678
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.372544288635254,
      "epoch": 0.42,
      "learning_rate": 5.897233132070498e-05,
      "loss": 1.3824,
      "step": 1160,
      "task_loss": 0.6550334095954895
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7538251876831055,
      "epoch": 0.42,
      "learning_rate": 5.8954579117072986e-05,
      "loss": 1.1745,
      "step": 1170,
      "task_loss": 0.6548653841018677
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.3968942165374756,
      "epoch": 0.43,
      "learning_rate": 5.89366776138946e-05,
      "loss": 1.1757,
      "step": 1180,
      "task_loss": 1.0903644561767578
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8698831796646118,
      "epoch": 0.43,
      "learning_rate": 5.8918626903475984e-05,
      "loss": 1.1275,
      "step": 1190,
      "task_loss": 0.9105096459388733
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0546157360076904,
      "epoch": 0.43,
      "learning_rate": 5.8900427078892686e-05,
      "loss": 1.2295,
      "step": 1200,
      "task_loss": 1.0799559354782104
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9607287645339966,
      "epoch": 0.44,
      "learning_rate": 5.888207823398911e-05,
      "loss": 1.071,
      "step": 1210,
      "task_loss": 0.8502562046051025
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.7898484468460083,
      "epoch": 0.44,
      "learning_rate": 5.886358046337806e-05,
      "loss": 1.1555,
      "step": 1220,
      "task_loss": 1.1786938905715942
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9928610324859619,
      "epoch": 0.44,
      "learning_rate": 5.884493386244029e-05,
      "loss": 1.0441,
      "step": 1230,
      "task_loss": 0.6678194999694824
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.371687412261963,
      "epoch": 0.45,
      "learning_rate": 5.882613852732392e-05,
      "loss": 1.1489,
      "step": 1240,
      "task_loss": 0.8427919149398804
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.259211778640747,
      "epoch": 0.45,
      "learning_rate": 5.880719455494403e-05,
      "loss": 1.1599,
      "step": 1250,
      "task_loss": 0.7002037763595581
    },
    {
      "epoch": 0.45,
      "eval_exact_match": 78.89309366130558,
      "eval_f1": 87.30477997622836,
      "step": 1250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2537944316864014,
      "epoch": 0.46,
      "learning_rate": 5.87881020429821e-05,
      "loss": 1.2433,
      "step": 1260,
      "task_loss": 0.9072924852371216
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.542639970779419,
      "epoch": 0.46,
      "learning_rate": 5.8768861089885557e-05,
      "loss": 1.1954,
      "step": 1270,
      "task_loss": 0.7959115505218506
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.015520453453064,
      "epoch": 0.46,
      "learning_rate": 5.8749471794867215e-05,
      "loss": 1.0924,
      "step": 1280,
      "task_loss": 0.7978045344352722
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0034399032592773,
      "epoch": 0.47,
      "learning_rate": 5.87299342579048e-05,
      "loss": 1.1828,
      "step": 1290,
      "task_loss": 0.656456470489502
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1494542360305786,
      "epoch": 0.47,
      "learning_rate": 5.8710248579740436e-05,
      "loss": 1.1708,
      "step": 1300,
      "task_loss": 0.9302552342414856
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.231022834777832,
      "epoch": 0.47,
      "learning_rate": 5.8690414861880074e-05,
      "loss": 1.2117,
      "step": 1310,
      "task_loss": 1.1827170848846436
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1523957252502441,
      "epoch": 0.48,
      "learning_rate": 5.867043320659306e-05,
      "loss": 1.0301,
      "step": 1320,
      "task_loss": 1.0149617195129395
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0345340967178345,
      "epoch": 0.48,
      "learning_rate": 5.865030371691151e-05,
      "loss": 1.0718,
      "step": 1330,
      "task_loss": 0.6616455912590027
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1779638528823853,
      "epoch": 0.48,
      "learning_rate": 5.863002649662984e-05,
      "loss": 1.1689,
      "step": 1340,
      "task_loss": 1.1839759349822998
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.4716678857803345,
      "epoch": 0.49,
      "learning_rate": 5.860960165030423e-05,
      "loss": 1.1607,
      "step": 1350,
      "task_loss": 1.0619685649871826
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9555326700210571,
      "epoch": 0.49,
      "learning_rate": 5.858902928325205e-05,
      "loss": 1.1595,
      "step": 1360,
      "task_loss": 0.7998465299606323
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0534662008285522,
      "epoch": 0.5,
      "learning_rate": 5.856830950155133e-05,
      "loss": 1.1149,
      "step": 1370,
      "task_loss": 0.737253725528717
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.4515752792358398,
      "epoch": 0.5,
      "learning_rate": 5.854744241204025e-05,
      "loss": 1.1515,
      "step": 1380,
      "task_loss": 1.2615107297897339
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1234278678894043,
      "epoch": 0.5,
      "learning_rate": 5.852642812231654e-05,
      "loss": 1.1093,
      "step": 1390,
      "task_loss": 1.0702711343765259
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.302065134048462,
      "epoch": 0.51,
      "learning_rate": 5.850526674073692e-05,
      "loss": 1.1776,
      "step": 1400,
      "task_loss": 0.933853268623352
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9075888395309448,
      "epoch": 0.51,
      "learning_rate": 5.848395837641663e-05,
      "loss": 0.9961,
      "step": 1410,
      "task_loss": 0.7664486169815063
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0795106887817383,
      "epoch": 0.51,
      "learning_rate": 5.8462503139228716e-05,
      "loss": 1.1341,
      "step": 1420,
      "task_loss": 0.921210765838623
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2469671964645386,
      "epoch": 0.52,
      "learning_rate": 5.844090113980363e-05,
      "loss": 1.1763,
      "step": 1430,
      "task_loss": 0.5581917762756348
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1218199729919434,
      "epoch": 0.52,
      "learning_rate": 5.8419152489528536e-05,
      "loss": 1.1253,
      "step": 1440,
      "task_loss": 0.9810710549354553
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1303911209106445,
      "epoch": 0.52,
      "learning_rate": 5.839725730054677e-05,
      "loss": 1.1801,
      "step": 1450,
      "task_loss": 0.9915053248405457
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9778804779052734,
      "epoch": 0.53,
      "learning_rate": 5.837521568575732e-05,
      "loss": 1.1791,
      "step": 1460,
      "task_loss": 0.6165598630905151
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9693446159362793,
      "epoch": 0.53,
      "learning_rate": 5.835302775881414e-05,
      "loss": 1.0693,
      "step": 1470,
      "task_loss": 0.7351759076118469
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2816810607910156,
      "epoch": 0.53,
      "learning_rate": 5.8330693634125654e-05,
      "loss": 1.1872,
      "step": 1480,
      "task_loss": 0.8726886510848999
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8330242037773132,
      "epoch": 0.54,
      "learning_rate": 5.830821342685411e-05,
      "loss": 1.0149,
      "step": 1490,
      "task_loss": 0.8250100612640381
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.049477219581604,
      "epoch": 0.54,
      "learning_rate": 5.8285587252915027e-05,
      "loss": 1.1407,
      "step": 1500,
      "task_loss": 1.0598646402359009
    },
    {
      "epoch": 0.54,
      "eval_exact_match": 79.03500473036897,
      "eval_f1": 87.58001380842249,
      "step": 1500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.89173424243927,
      "epoch": 0.55,
      "learning_rate": 5.8262815228976555e-05,
      "loss": 1.2533,
      "step": 1510,
      "task_loss": 0.8165568113327026
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0344847440719604,
      "epoch": 0.55,
      "learning_rate": 5.823989747245893e-05,
      "loss": 1.0947,
      "step": 1520,
      "task_loss": 0.7850322723388672
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1379281282424927,
      "epoch": 0.55,
      "learning_rate": 5.821683410153377e-05,
      "loss": 1.1351,
      "step": 1530,
      "task_loss": 1.0284912586212158
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.4056248664855957,
      "epoch": 0.56,
      "learning_rate": 5.8193625235123606e-05,
      "loss": 1.2041,
      "step": 1540,
      "task_loss": 1.2799898386001587
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1627094745635986,
      "epoch": 0.56,
      "learning_rate": 5.817027099290115e-05,
      "loss": 1.0303,
      "step": 1550,
      "task_loss": 1.232445240020752
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2967350482940674,
      "epoch": 0.56,
      "learning_rate": 5.814677149528872e-05,
      "loss": 1.1817,
      "step": 1560,
      "task_loss": 1.5413718223571777
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.879601776599884,
      "epoch": 0.57,
      "learning_rate": 5.812312686345763e-05,
      "loss": 1.0928,
      "step": 1570,
      "task_loss": 0.42186737060546875
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1352674961090088,
      "epoch": 0.57,
      "learning_rate": 5.809933721932756e-05,
      "loss": 1.1005,
      "step": 1580,
      "task_loss": 1.1184422969818115
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.5701181888580322,
      "epoch": 0.57,
      "learning_rate": 5.8075402685565904e-05,
      "loss": 1.2745,
      "step": 1590,
      "task_loss": 0.9984648823738098
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1425367593765259,
      "epoch": 0.58,
      "learning_rate": 5.805132338558717e-05,
      "loss": 1.2407,
      "step": 1600,
      "task_loss": 0.7507245540618896
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9696254730224609,
      "epoch": 0.58,
      "learning_rate": 5.8027099443552336e-05,
      "loss": 1.2035,
      "step": 1610,
      "task_loss": 0.4481644630432129
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1334104537963867,
      "epoch": 0.59,
      "learning_rate": 5.800273098436818e-05,
      "loss": 1.1432,
      "step": 1620,
      "task_loss": 0.890143096446991
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0847598314285278,
      "epoch": 0.59,
      "learning_rate": 5.7978218133686684e-05,
      "loss": 1.1767,
      "step": 1630,
      "task_loss": 1.5157380104064941
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.137290358543396,
      "epoch": 0.59,
      "learning_rate": 5.795356101790434e-05,
      "loss": 1.1079,
      "step": 1640,
      "task_loss": 1.1900906562805176
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9380532503128052,
      "epoch": 0.6,
      "learning_rate": 5.792875976416152e-05,
      "loss": 1.1299,
      "step": 1650,
      "task_loss": 0.44667813181877136
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.4331817626953125,
      "epoch": 0.6,
      "learning_rate": 5.7903814500341845e-05,
      "loss": 1.0832,
      "step": 1660,
      "task_loss": 1.059866189956665
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.4325172901153564,
      "epoch": 0.6,
      "learning_rate": 5.7878725355071475e-05,
      "loss": 1.1974,
      "step": 1670,
      "task_loss": 1.4602153301239014
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9795972108840942,
      "epoch": 0.61,
      "learning_rate": 5.785349245771848e-05,
      "loss": 0.9908,
      "step": 1680,
      "task_loss": 0.6619421243667603
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9228504300117493,
      "epoch": 0.61,
      "learning_rate": 5.782811593839217e-05,
      "loss": 1.0195,
      "step": 1690,
      "task_loss": 0.7317041158676147
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.483938217163086,
      "epoch": 0.61,
      "learning_rate": 5.78025959279424e-05,
      "loss": 1.1531,
      "step": 1700,
      "task_loss": 1.3933497667312622
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2623841762542725,
      "epoch": 0.62,
      "learning_rate": 5.777693255795894e-05,
      "loss": 1.1414,
      "step": 1710,
      "task_loss": 1.1138461828231812
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1957447528839111,
      "epoch": 0.62,
      "learning_rate": 5.775112596077074e-05,
      "loss": 1.1567,
      "step": 1720,
      "task_loss": 0.7848667502403259
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0970215797424316,
      "epoch": 0.63,
      "learning_rate": 5.772517626944531e-05,
      "loss": 1.1256,
      "step": 1730,
      "task_loss": 1.2455201148986816
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2585291862487793,
      "epoch": 0.63,
      "learning_rate": 5.769908361778797e-05,
      "loss": 1.0587,
      "step": 1740,
      "task_loss": 1.0000256299972534
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.362696886062622,
      "epoch": 0.63,
      "learning_rate": 5.76728481403412e-05,
      "loss": 1.0007,
      "step": 1750,
      "task_loss": 0.9608546495437622
    },
    {
      "epoch": 0.63,
      "eval_exact_match": 79.28098391674551,
      "eval_f1": 87.6563371777527,
      "step": 1750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8324779272079468,
      "epoch": 0.64,
      "learning_rate": 5.764646997238398e-05,
      "loss": 1.134,
      "step": 1760,
      "task_loss": 0.6974785923957825
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.3671345710754395,
      "epoch": 0.64,
      "learning_rate": 5.761994924993097e-05,
      "loss": 1.0616,
      "step": 1770,
      "task_loss": 0.9882352948188782
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.3772859573364258,
      "epoch": 0.64,
      "learning_rate": 5.759328610973195e-05,
      "loss": 1.2092,
      "step": 1780,
      "task_loss": 1.0786349773406982
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.871800422668457,
      "epoch": 0.65,
      "learning_rate": 5.756648068927106e-05,
      "loss": 1.0171,
      "step": 1790,
      "task_loss": 0.830288827419281
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.5783218145370483,
      "epoch": 0.65,
      "learning_rate": 5.753953312676602e-05,
      "loss": 1.255,
      "step": 1800,
      "task_loss": 1.0050296783447266
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.3028333187103271,
      "epoch": 0.65,
      "learning_rate": 5.751244356116755e-05,
      "loss": 1.0499,
      "step": 1810,
      "task_loss": 1.112877607345581
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.003974437713623,
      "epoch": 0.66,
      "learning_rate": 5.7485212132158555e-05,
      "loss": 1.0624,
      "step": 1820,
      "task_loss": 0.4431191086769104
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9127657413482666,
      "epoch": 0.66,
      "learning_rate": 5.7457838980153446e-05,
      "loss": 1.1427,
      "step": 1830,
      "task_loss": 0.7105373740196228
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.3443561792373657,
      "epoch": 0.66,
      "learning_rate": 5.7430324246297385e-05,
      "loss": 1.0933,
      "step": 1840,
      "task_loss": 1.1131591796875
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2187169790267944,
      "epoch": 0.67,
      "learning_rate": 5.74026680724656e-05,
      "loss": 1.1155,
      "step": 1850,
      "task_loss": 1.3410766124725342
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7574527263641357,
      "epoch": 0.67,
      "learning_rate": 5.737487060126263e-05,
      "loss": 1.0802,
      "step": 1860,
      "task_loss": 0.42934900522232056
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1824791431427002,
      "epoch": 0.68,
      "learning_rate": 5.734693197602157e-05,
      "loss": 1.1564,
      "step": 1870,
      "task_loss": 0.9754879474639893
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1144288778305054,
      "epoch": 0.68,
      "learning_rate": 5.731885234080337e-05,
      "loss": 1.0964,
      "step": 1880,
      "task_loss": 0.7791981101036072
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.101729154586792,
      "epoch": 0.68,
      "learning_rate": 5.729063184039608e-05,
      "loss": 1.073,
      "step": 1890,
      "task_loss": 0.9119380712509155
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7584141492843628,
      "epoch": 0.69,
      "learning_rate": 5.7262270620314076e-05,
      "loss": 1.0943,
      "step": 1900,
      "task_loss": 0.6965209245681763
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7631694078445435,
      "epoch": 0.69,
      "learning_rate": 5.7233768826797364e-05,
      "loss": 1.0534,
      "step": 1910,
      "task_loss": 0.8611739873886108
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8627794981002808,
      "epoch": 0.69,
      "learning_rate": 5.720512660681077e-05,
      "loss": 1.1024,
      "step": 1920,
      "task_loss": 0.5158926248550415
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1938802003860474,
      "epoch": 0.7,
      "learning_rate": 5.717634410804319e-05,
      "loss": 1.0845,
      "step": 1930,
      "task_loss": 1.100459337234497
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9154514074325562,
      "epoch": 0.7,
      "learning_rate": 5.71474214789069e-05,
      "loss": 1.1354,
      "step": 1940,
      "task_loss": 0.6148239374160767
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.3251101970672607,
      "epoch": 0.7,
      "learning_rate": 5.711835886853667e-05,
      "loss": 1.2525,
      "step": 1950,
      "task_loss": 0.8512697219848633
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8448692560195923,
      "epoch": 0.71,
      "learning_rate": 5.708915642678912e-05,
      "loss": 1.0824,
      "step": 1960,
      "task_loss": 0.6256545782089233
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6884149312973022,
      "epoch": 0.71,
      "learning_rate": 5.705981430424184e-05,
      "loss": 0.9549,
      "step": 1970,
      "task_loss": 0.3767446279525757
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.3547323942184448,
      "epoch": 0.72,
      "learning_rate": 5.7030332652192685e-05,
      "loss": 1.0335,
      "step": 1980,
      "task_loss": 1.198007345199585
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.630234956741333,
      "epoch": 0.72,
      "learning_rate": 5.700071162265897e-05,
      "loss": 1.0725,
      "step": 1990,
      "task_loss": 0.7066274881362915
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0253219604492188,
      "epoch": 0.72,
      "learning_rate": 5.697095136837669e-05,
      "loss": 1.1779,
      "step": 2000,
      "task_loss": 1.2108477354049683
    },
    {
      "epoch": 0.72,
      "eval_exact_match": 79.67833491012298,
      "eval_f1": 87.87620766435882,
      "step": 2000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.978434145450592,
      "epoch": 0.73,
      "learning_rate": 5.6941052042799705e-05,
      "loss": 1.008,
      "step": 2010,
      "task_loss": 1.1101996898651123
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.021071195602417,
      "epoch": 0.73,
      "learning_rate": 5.6911013800099014e-05,
      "loss": 1.0764,
      "step": 2020,
      "task_loss": 0.9951556921005249
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0078377723693848,
      "epoch": 0.73,
      "learning_rate": 5.688083679516188e-05,
      "loss": 1.1377,
      "step": 2030,
      "task_loss": 0.8278811573982239
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.949942409992218,
      "epoch": 0.74,
      "learning_rate": 5.6850521183591096e-05,
      "loss": 1.0232,
      "step": 2040,
      "task_loss": 1.2979423999786377
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1799646615982056,
      "epoch": 0.74,
      "learning_rate": 5.6820067121704145e-05,
      "loss": 1.1882,
      "step": 2050,
      "task_loss": 0.8740980625152588
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.05833101272583,
      "epoch": 0.74,
      "learning_rate": 5.678947476653241e-05,
      "loss": 1.0395,
      "step": 2060,
      "task_loss": 0.6604486703872681
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.5961391925811768,
      "epoch": 0.75,
      "learning_rate": 5.675874427582037e-05,
      "loss": 1.1881,
      "step": 2070,
      "task_loss": 1.4038045406341553
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8555396795272827,
      "epoch": 0.75,
      "learning_rate": 5.6727875808024764e-05,
      "loss": 1.0642,
      "step": 2080,
      "task_loss": 0.6731501221656799
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9887014627456665,
      "epoch": 0.76,
      "learning_rate": 5.669686952231379e-05,
      "loss": 1.0166,
      "step": 2090,
      "task_loss": 0.642594039440155
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7086762189865112,
      "epoch": 0.76,
      "learning_rate": 5.6665725578566285e-05,
      "loss": 1.0289,
      "step": 2100,
      "task_loss": 0.5841311812400818
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1380422115325928,
      "epoch": 0.76,
      "learning_rate": 5.663444413737089e-05,
      "loss": 1.1594,
      "step": 2110,
      "task_loss": 0.7255545258522034
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8511086702346802,
      "epoch": 0.77,
      "learning_rate": 5.660302536002525e-05,
      "loss": 1.2183,
      "step": 2120,
      "task_loss": 0.7066528797149658
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0993759632110596,
      "epoch": 0.77,
      "learning_rate": 5.657146940853514e-05,
      "loss": 1.008,
      "step": 2130,
      "task_loss": 0.6117959022521973
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.3931245803833008,
      "epoch": 0.77,
      "learning_rate": 5.6539776445613644e-05,
      "loss": 1.1547,
      "step": 2140,
      "task_loss": 0.8887019157409668
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7988817691802979,
      "epoch": 0.78,
      "learning_rate": 5.6507946634680345e-05,
      "loss": 0.9199,
      "step": 2150,
      "task_loss": 0.5578777194023132
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.683596134185791,
      "epoch": 0.78,
      "learning_rate": 5.647598013986046e-05,
      "loss": 1.0362,
      "step": 2160,
      "task_loss": 0.5091426968574524
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0852502584457397,
      "epoch": 0.78,
      "learning_rate": 5.644387712598397e-05,
      "loss": 1.0067,
      "step": 2170,
      "task_loss": 1.3019379377365112
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2326734066009521,
      "epoch": 0.79,
      "learning_rate": 5.641163775858481e-05,
      "loss": 1.0256,
      "step": 2180,
      "task_loss": 1.3112915754318237
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7502934336662292,
      "epoch": 0.79,
      "learning_rate": 5.637926220390003e-05,
      "loss": 1.1512,
      "step": 2190,
      "task_loss": 0.5557284951210022
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.5017701387405396,
      "epoch": 0.8,
      "learning_rate": 5.6346750628868835e-05,
      "loss": 1.1388,
      "step": 2200,
      "task_loss": 1.1129333972930908
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8690292835235596,
      "epoch": 0.8,
      "learning_rate": 5.631410320113186e-05,
      "loss": 1.0063,
      "step": 2210,
      "task_loss": 0.8763360977172852
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0584996938705444,
      "epoch": 0.8,
      "learning_rate": 5.628132008903022e-05,
      "loss": 1.0087,
      "step": 2220,
      "task_loss": 0.5723036527633667
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.3691318035125732,
      "epoch": 0.81,
      "learning_rate": 5.624840146160466e-05,
      "loss": 1.0832,
      "step": 2230,
      "task_loss": 1.132154941558838
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2483141422271729,
      "epoch": 0.81,
      "learning_rate": 5.62153474885947e-05,
      "loss": 1.1432,
      "step": 2240,
      "task_loss": 0.8840174674987793
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1159393787384033,
      "epoch": 0.81,
      "learning_rate": 5.618215834043773e-05,
      "loss": 1.1353,
      "step": 2250,
      "task_loss": 0.7122650146484375
    },
    {
      "epoch": 0.81,
      "eval_exact_match": 79.82970671712394,
      "eval_f1": 87.98350972516127,
      "step": 2250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.971450686454773,
      "epoch": 0.82,
      "learning_rate": 5.614883418826817e-05,
      "loss": 0.9947,
      "step": 2260,
      "task_loss": 1.4125139713287354
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.3253803253173828,
      "epoch": 0.82,
      "learning_rate": 5.6115375203916534e-05,
      "loss": 1.0477,
      "step": 2270,
      "task_loss": 1.3677072525024414
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0992299318313599,
      "epoch": 0.82,
      "learning_rate": 5.6081781559908616e-05,
      "loss": 1.0079,
      "step": 2280,
      "task_loss": 0.9259253144264221
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9781779646873474,
      "epoch": 0.83,
      "learning_rate": 5.604805342946454e-05,
      "loss": 1.0634,
      "step": 2290,
      "task_loss": 0.6949033737182617
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9337608218193054,
      "epoch": 0.83,
      "learning_rate": 5.601419098649789e-05,
      "loss": 1.0581,
      "step": 2300,
      "task_loss": 0.7444375157356262
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.783240556716919,
      "epoch": 0.83,
      "learning_rate": 5.5980194405614785e-05,
      "loss": 1.016,
      "step": 2310,
      "task_loss": 0.6494660377502441
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2547146081924438,
      "epoch": 0.84,
      "learning_rate": 5.5946063862113056e-05,
      "loss": 1.0121,
      "step": 2320,
      "task_loss": 1.5860847234725952
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.198014497756958,
      "epoch": 0.84,
      "learning_rate": 5.591179953198124e-05,
      "loss": 0.9877,
      "step": 2330,
      "task_loss": 0.8051793575286865
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6500747799873352,
      "epoch": 0.85,
      "learning_rate": 5.587740159189777e-05,
      "loss": 1.06,
      "step": 2340,
      "task_loss": 0.9356144666671753
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9654289484024048,
      "epoch": 0.85,
      "learning_rate": 5.584287021922997e-05,
      "loss": 1.2073,
      "step": 2350,
      "task_loss": 0.7282505035400391
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.4243820905685425,
      "epoch": 0.85,
      "learning_rate": 5.5808205592033224e-05,
      "loss": 1.0644,
      "step": 2360,
      "task_loss": 1.230690360069275
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.6079635620117188,
      "epoch": 0.86,
      "learning_rate": 5.5773407889049986e-05,
      "loss": 1.07,
      "step": 2370,
      "task_loss": 1.2446208000183105
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.937513530254364,
      "epoch": 0.86,
      "learning_rate": 5.573847728970894e-05,
      "loss": 1.0576,
      "step": 2380,
      "task_loss": 0.9011927247047424
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1456040143966675,
      "epoch": 0.86,
      "learning_rate": 5.570341397412398e-05,
      "loss": 1.1172,
      "step": 2390,
      "task_loss": 1.1431217193603516
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9871031045913696,
      "epoch": 0.87,
      "learning_rate": 5.5668218123093375e-05,
      "loss": 1.0467,
      "step": 2400,
      "task_loss": 1.1605985164642334
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9994000196456909,
      "epoch": 0.87,
      "learning_rate": 5.563288991809875e-05,
      "loss": 1.2362,
      "step": 2410,
      "task_loss": 0.7171075940132141
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.3609836101531982,
      "epoch": 0.87,
      "learning_rate": 5.559742954130421e-05,
      "loss": 1.147,
      "step": 2420,
      "task_loss": 1.1994080543518066
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9730468988418579,
      "epoch": 0.88,
      "learning_rate": 5.556183717555539e-05,
      "loss": 0.9454,
      "step": 2430,
      "task_loss": 0.6406092047691345
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.6829783916473389,
      "epoch": 0.88,
      "learning_rate": 5.552611300437849e-05,
      "loss": 1.0889,
      "step": 2440,
      "task_loss": 1.5461273193359375
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.061495065689087,
      "epoch": 0.89,
      "learning_rate": 5.549025721197935e-05,
      "loss": 1.0418,
      "step": 2450,
      "task_loss": 0.6560171842575073
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8872650861740112,
      "epoch": 0.89,
      "learning_rate": 5.54542699832425e-05,
      "loss": 0.9725,
      "step": 2460,
      "task_loss": 0.8630092740058899
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9294049739837646,
      "epoch": 0.89,
      "learning_rate": 5.5418151503730185e-05,
      "loss": 0.9812,
      "step": 2470,
      "task_loss": 0.8304383158683777
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1200015544891357,
      "epoch": 0.9,
      "learning_rate": 5.538190195968143e-05,
      "loss": 0.8769,
      "step": 2480,
      "task_loss": 0.8949685096740723
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1685872077941895,
      "epoch": 0.9,
      "learning_rate": 5.534552153801109e-05,
      "loss": 1.06,
      "step": 2490,
      "task_loss": 0.9758301973342896
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.159787893295288,
      "epoch": 0.9,
      "learning_rate": 5.530901042630882e-05,
      "loss": 1.0887,
      "step": 2500,
      "task_loss": 0.6261301040649414
    },
    {
      "epoch": 0.9,
      "eval_exact_match": 80.66225165562913,
      "eval_f1": 88.41245606075655,
      "step": 2500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9179141521453857,
      "epoch": 0.91,
      "learning_rate": 5.527236881283822e-05,
      "loss": 1.0642,
      "step": 2510,
      "task_loss": 0.7565745711326599
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0321706533432007,
      "epoch": 0.91,
      "learning_rate": 5.523559688653577e-05,
      "loss": 1.0146,
      "step": 2520,
      "task_loss": 0.9108026027679443
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8541651368141174,
      "epoch": 0.91,
      "learning_rate": 5.519869483700986e-05,
      "loss": 0.937,
      "step": 2530,
      "task_loss": 0.887135922908783
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9746845960617065,
      "epoch": 0.92,
      "learning_rate": 5.516166285453989e-05,
      "loss": 0.9061,
      "step": 2540,
      "task_loss": 1.0225167274475098
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2189579010009766,
      "epoch": 0.92,
      "learning_rate": 5.5124501130075204e-05,
      "loss": 1.0058,
      "step": 2550,
      "task_loss": 1.3460352420806885
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8514137864112854,
      "epoch": 0.93,
      "learning_rate": 5.508720985523416e-05,
      "loss": 1.0001,
      "step": 2560,
      "task_loss": 0.9920178651809692
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7871240377426147,
      "epoch": 0.93,
      "learning_rate": 5.504978922230309e-05,
      "loss": 1.011,
      "step": 2570,
      "task_loss": 0.8328629732131958
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8508641719818115,
      "epoch": 0.93,
      "learning_rate": 5.501223942423539e-05,
      "loss": 0.9933,
      "step": 2580,
      "task_loss": 0.5792371034622192
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9439315795898438,
      "epoch": 0.94,
      "learning_rate": 5.4974560654650424e-05,
      "loss": 1.0669,
      "step": 2590,
      "task_loss": 0.7067208290100098
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9048341512680054,
      "epoch": 0.94,
      "learning_rate": 5.4936753107832604e-05,
      "loss": 1.0051,
      "step": 2600,
      "task_loss": 0.78888338804245
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8586219549179077,
      "epoch": 0.94,
      "learning_rate": 5.489881697873035e-05,
      "loss": 0.9638,
      "step": 2610,
      "task_loss": 0.556171178817749
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9864273071289062,
      "epoch": 0.95,
      "learning_rate": 5.486075246295511e-05,
      "loss": 1.0877,
      "step": 2620,
      "task_loss": 0.6050583124160767
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.968224048614502,
      "epoch": 0.95,
      "learning_rate": 5.482255975678033e-05,
      "loss": 1.0996,
      "step": 2630,
      "task_loss": 1.148437261581421
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8083895444869995,
      "epoch": 0.95,
      "learning_rate": 5.4784239057140426e-05,
      "loss": 1.0497,
      "step": 2640,
      "task_loss": 1.3358066082000732
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0225319862365723,
      "epoch": 0.96,
      "learning_rate": 5.474579056162983e-05,
      "loss": 1.0384,
      "step": 2650,
      "task_loss": 0.7764133810997009
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1559662818908691,
      "epoch": 0.96,
      "learning_rate": 5.470721446850191e-05,
      "loss": 1.0839,
      "step": 2660,
      "task_loss": 0.9642658233642578
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7244634032249451,
      "epoch": 0.96,
      "learning_rate": 5.466851097666796e-05,
      "loss": 1.0632,
      "step": 2670,
      "task_loss": 0.8971454501152039
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.644745409488678,
      "epoch": 0.97,
      "learning_rate": 5.462968028569621e-05,
      "loss": 1.0227,
      "step": 2680,
      "task_loss": 0.508349597454071
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6737414002418518,
      "epoch": 0.97,
      "learning_rate": 5.459072259581078e-05,
      "loss": 1.0816,
      "step": 2690,
      "task_loss": 0.4952778220176697
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9121610522270203,
      "epoch": 0.98,
      "learning_rate": 5.455163810789059e-05,
      "loss": 0.9735,
      "step": 2700,
      "task_loss": 0.49527156352996826
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9224643707275391,
      "epoch": 0.98,
      "learning_rate": 5.451242702346841e-05,
      "loss": 1.0023,
      "step": 2710,
      "task_loss": 0.9647371768951416
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8482234477996826,
      "epoch": 0.98,
      "learning_rate": 5.447308954472978e-05,
      "loss": 1.0031,
      "step": 2720,
      "task_loss": 0.7513360977172852
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9275264739990234,
      "epoch": 0.99,
      "learning_rate": 5.443362587451196e-05,
      "loss": 1.0583,
      "step": 2730,
      "task_loss": 0.7999730110168457
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 2.1053638458251953,
      "epoch": 0.99,
      "learning_rate": 5.4394036216302914e-05,
      "loss": 1.1403,
      "step": 2740,
      "task_loss": 1.4788403511047363
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2016551494598389,
      "epoch": 0.99,
      "learning_rate": 5.4354320774240234e-05,
      "loss": 0.8911,
      "step": 2750,
      "task_loss": 0.817990243434906
    },
    {
      "epoch": 0.99,
      "eval_exact_match": 80.32166508987702,
      "eval_f1": 88.34998419616365,
      "step": 2750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.011232614517212,
      "epoch": 1.0,
      "learning_rate": 5.431447975311009e-05,
      "loss": 1.0087,
      "step": 2760,
      "task_loss": 0.7390681505203247
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7077229022979736,
      "epoch": 1.0,
      "learning_rate": 5.427451335834618e-05,
      "loss": 0.8541,
      "step": 2770,
      "task_loss": 0.632800817489624
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.930655837059021,
      "epoch": 1.0,
      "learning_rate": 5.4234421796028677e-05,
      "loss": 0.8487,
      "step": 2780,
      "task_loss": 0.698851466178894
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7608747482299805,
      "epoch": 1.01,
      "learning_rate": 5.419420527288317e-05,
      "loss": 0.7985,
      "step": 2790,
      "task_loss": 0.4454770088195801
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.916797399520874,
      "epoch": 1.01,
      "learning_rate": 5.415386399627955e-05,
      "loss": 0.8375,
      "step": 2800,
      "task_loss": 0.9108859896659851
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0602134466171265,
      "epoch": 1.02,
      "learning_rate": 5.411339817423103e-05,
      "loss": 0.8843,
      "step": 2810,
      "task_loss": 1.0451345443725586
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8087363839149475,
      "epoch": 1.02,
      "learning_rate": 5.4072808015392984e-05,
      "loss": 0.8028,
      "step": 2820,
      "task_loss": 0.8453482389450073
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7011986374855042,
      "epoch": 1.02,
      "learning_rate": 5.403209372906192e-05,
      "loss": 0.8078,
      "step": 2830,
      "task_loss": 0.6264968514442444
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8319069147109985,
      "epoch": 1.03,
      "learning_rate": 5.3991255525174396e-05,
      "loss": 0.8133,
      "step": 2840,
      "task_loss": 0.9526384472846985
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0532593727111816,
      "epoch": 1.03,
      "learning_rate": 5.395029361430591e-05,
      "loss": 0.798,
      "step": 2850,
      "task_loss": 1.0793445110321045
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1473658084869385,
      "epoch": 1.03,
      "learning_rate": 5.3909208207669864e-05,
      "loss": 0.8647,
      "step": 2860,
      "task_loss": 0.8664270639419556
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2761093378067017,
      "epoch": 1.04,
      "learning_rate": 5.3867999517116424e-05,
      "loss": 0.8462,
      "step": 2870,
      "task_loss": 0.9831790924072266
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8598681688308716,
      "epoch": 1.04,
      "learning_rate": 5.382666775513145e-05,
      "loss": 0.7603,
      "step": 2880,
      "task_loss": 0.5603002309799194
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5312719941139221,
      "epoch": 1.04,
      "learning_rate": 5.378521313483541e-05,
      "loss": 0.7213,
      "step": 2890,
      "task_loss": 0.5749783515930176
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8793454766273499,
      "epoch": 1.05,
      "learning_rate": 5.3743635869982276e-05,
      "loss": 0.872,
      "step": 2900,
      "task_loss": 0.5542865991592407
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6059684753417969,
      "epoch": 1.05,
      "learning_rate": 5.370193617495839e-05,
      "loss": 0.7783,
      "step": 2910,
      "task_loss": 1.0900731086730957
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6760936379432678,
      "epoch": 1.06,
      "learning_rate": 5.3660114264781405e-05,
      "loss": 0.7037,
      "step": 2920,
      "task_loss": 0.3957022428512573
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9233886003494263,
      "epoch": 1.06,
      "learning_rate": 5.361817035509917e-05,
      "loss": 0.8746,
      "step": 2930,
      "task_loss": 1.043057918548584
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6635424494743347,
      "epoch": 1.06,
      "learning_rate": 5.357610466218856e-05,
      "loss": 0.8787,
      "step": 2940,
      "task_loss": 0.712294340133667
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6662827730178833,
      "epoch": 1.07,
      "learning_rate": 5.3533917402954456e-05,
      "loss": 0.7418,
      "step": 2950,
      "task_loss": 0.5668191313743591
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7443856000900269,
      "epoch": 1.07,
      "learning_rate": 5.349160879492854e-05,
      "loss": 0.8513,
      "step": 2960,
      "task_loss": 0.8575096726417542
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9767590761184692,
      "epoch": 1.07,
      "learning_rate": 5.3449179056268235e-05,
      "loss": 0.7926,
      "step": 2970,
      "task_loss": 0.8986884951591492
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0205166339874268,
      "epoch": 1.08,
      "learning_rate": 5.340662840575553e-05,
      "loss": 0.8059,
      "step": 2980,
      "task_loss": 1.0653899908065796
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7430188655853271,
      "epoch": 1.08,
      "learning_rate": 5.336395706279589e-05,
      "loss": 0.818,
      "step": 2990,
      "task_loss": 0.44489508867263794
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6891306638717651,
      "epoch": 1.08,
      "learning_rate": 5.33211652474171e-05,
      "loss": 0.8024,
      "step": 3000,
      "task_loss": 0.9418755769729614
    },
    {
      "epoch": 1.08,
      "eval_exact_match": 80.1608325449385,
      "eval_f1": 88.15996436894764,
      "step": 3000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.876007080078125,
      "epoch": 1.09,
      "learning_rate": 5.327825318026816e-05,
      "loss": 0.7745,
      "step": 3010,
      "task_loss": 0.6779855489730835
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8885762095451355,
      "epoch": 1.09,
      "learning_rate": 5.323522108261813e-05,
      "loss": 0.8159,
      "step": 3020,
      "task_loss": 0.6267703175544739
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.872785747051239,
      "epoch": 1.1,
      "learning_rate": 5.319206917635494e-05,
      "loss": 0.8425,
      "step": 3030,
      "task_loss": 0.7042969465255737
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7279747128486633,
      "epoch": 1.1,
      "learning_rate": 5.314879768398437e-05,
      "loss": 0.8151,
      "step": 3040,
      "task_loss": 0.6893462538719177
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8534276485443115,
      "epoch": 1.1,
      "learning_rate": 5.310540682862876e-05,
      "loss": 0.9091,
      "step": 3050,
      "task_loss": 1.0376639366149902
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9732213020324707,
      "epoch": 1.11,
      "learning_rate": 5.306189683402595e-05,
      "loss": 0.8051,
      "step": 3060,
      "task_loss": 1.1177102327346802
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8836879730224609,
      "epoch": 1.11,
      "learning_rate": 5.3018267924528124e-05,
      "loss": 0.8283,
      "step": 3070,
      "task_loss": 1.0020360946655273
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6803722381591797,
      "epoch": 1.11,
      "learning_rate": 5.29745203251006e-05,
      "loss": 0.8084,
      "step": 3080,
      "task_loss": 0.797760546207428
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9686804413795471,
      "epoch": 1.12,
      "learning_rate": 5.293065426132069e-05,
      "loss": 0.8542,
      "step": 3090,
      "task_loss": 0.6910236477851868
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6933484077453613,
      "epoch": 1.12,
      "learning_rate": 5.28866699593766e-05,
      "loss": 0.7601,
      "step": 3100,
      "task_loss": 0.6895896196365356
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9216572046279907,
      "epoch": 1.12,
      "learning_rate": 5.284256764606617e-05,
      "loss": 0.7732,
      "step": 3110,
      "task_loss": 0.7823605537414551
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.677863597869873,
      "epoch": 1.13,
      "learning_rate": 5.279834754879575e-05,
      "loss": 0.7433,
      "step": 3120,
      "task_loss": 0.5516537427902222
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7393268346786499,
      "epoch": 1.13,
      "learning_rate": 5.275400989557905e-05,
      "loss": 0.8662,
      "step": 3130,
      "task_loss": 0.4517717659473419
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9628137350082397,
      "epoch": 1.13,
      "learning_rate": 5.270955491503589e-05,
      "loss": 0.8509,
      "step": 3140,
      "task_loss": 0.9863475561141968
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8606938719749451,
      "epoch": 1.14,
      "learning_rate": 5.266498283639113e-05,
      "loss": 0.8131,
      "step": 3150,
      "task_loss": 0.9285368919372559
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9628182649612427,
      "epoch": 1.14,
      "learning_rate": 5.2620293889473387e-05,
      "loss": 0.8319,
      "step": 3160,
      "task_loss": 1.1551557779312134
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8328995704650879,
      "epoch": 1.15,
      "learning_rate": 5.257548830471388e-05,
      "loss": 0.8528,
      "step": 3170,
      "task_loss": 0.7933187484741211
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6802824139595032,
      "epoch": 1.15,
      "learning_rate": 5.25305663131453e-05,
      "loss": 0.8096,
      "step": 3180,
      "task_loss": 0.6108599305152893
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7289232015609741,
      "epoch": 1.15,
      "learning_rate": 5.2485528146400505e-05,
      "loss": 0.8256,
      "step": 3190,
      "task_loss": 1.2673640251159668
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7009760141372681,
      "epoch": 1.16,
      "learning_rate": 5.244037403671146e-05,
      "loss": 0.7034,
      "step": 3200,
      "task_loss": 0.6759889125823975
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6823523044586182,
      "epoch": 1.16,
      "learning_rate": 5.2395104216907926e-05,
      "loss": 0.8028,
      "step": 3210,
      "task_loss": 0.513982892036438
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5610631704330444,
      "epoch": 1.16,
      "learning_rate": 5.234971892041632e-05,
      "loss": 0.7435,
      "step": 3220,
      "task_loss": 0.5406886339187622
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.918226420879364,
      "epoch": 1.17,
      "learning_rate": 5.230421838125847e-05,
      "loss": 0.8205,
      "step": 3230,
      "task_loss": 1.1968251466751099
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8683892488479614,
      "epoch": 1.17,
      "learning_rate": 5.22586028340505e-05,
      "loss": 0.7288,
      "step": 3240,
      "task_loss": 0.45106273889541626
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9409477710723877,
      "epoch": 1.17,
      "learning_rate": 5.2212872514001466e-05,
      "loss": 0.8127,
      "step": 3250,
      "task_loss": 0.9111529588699341
    },
    {
      "epoch": 1.17,
      "eval_exact_match": 80.80416272469253,
      "eval_f1": 88.54957233403387,
      "step": 3250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5494179725646973,
      "epoch": 1.18,
      "learning_rate": 5.2167027656912314e-05,
      "loss": 0.7938,
      "step": 3260,
      "task_loss": 0.5460392236709595
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6427707672119141,
      "epoch": 1.18,
      "learning_rate": 5.212106849917452e-05,
      "loss": 0.777,
      "step": 3270,
      "task_loss": 0.673181414604187
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8692964315414429,
      "epoch": 1.19,
      "learning_rate": 5.2074995277768956e-05,
      "loss": 0.8644,
      "step": 3280,
      "task_loss": 0.860024094581604
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.891358494758606,
      "epoch": 1.19,
      "learning_rate": 5.202880823026465e-05,
      "loss": 0.8198,
      "step": 3290,
      "task_loss": 0.9117802381515503
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6767855882644653,
      "epoch": 1.19,
      "learning_rate": 5.198250759481754e-05,
      "loss": 0.8081,
      "step": 3300,
      "task_loss": 0.46542519330978394
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6959289908409119,
      "epoch": 1.2,
      "learning_rate": 5.193609361016929e-05,
      "loss": 0.7415,
      "step": 3310,
      "task_loss": 0.86083984375
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0413265228271484,
      "epoch": 1.2,
      "learning_rate": 5.188956651564602e-05,
      "loss": 0.8289,
      "step": 3320,
      "task_loss": 1.4512548446655273
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7771068215370178,
      "epoch": 1.2,
      "learning_rate": 5.184292655115703e-05,
      "loss": 0.8058,
      "step": 3330,
      "task_loss": 0.5583703517913818
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1072239875793457,
      "epoch": 1.21,
      "learning_rate": 5.1796173957193705e-05,
      "loss": 0.8859,
      "step": 3340,
      "task_loss": 1.0872106552124023
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.532997727394104,
      "epoch": 1.21,
      "learning_rate": 5.174930897482812e-05,
      "loss": 0.6922,
      "step": 3350,
      "task_loss": 0.584416389465332
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8668481111526489,
      "epoch": 1.21,
      "learning_rate": 5.1702331845711894e-05,
      "loss": 0.7652,
      "step": 3360,
      "task_loss": 1.1341835260391235
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8142274618148804,
      "epoch": 1.22,
      "learning_rate": 5.165524281207489e-05,
      "loss": 0.796,
      "step": 3370,
      "task_loss": 1.4565706253051758
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4797128438949585,
      "epoch": 1.22,
      "learning_rate": 5.1608042116724006e-05,
      "loss": 0.8104,
      "step": 3380,
      "task_loss": 0.45489227771759033
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8843653202056885,
      "epoch": 1.23,
      "learning_rate": 5.1560730003041896e-05,
      "loss": 0.9012,
      "step": 3390,
      "task_loss": 1.0831022262573242
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5749521255493164,
      "epoch": 1.23,
      "learning_rate": 5.1513306714985725e-05,
      "loss": 0.693,
      "step": 3400,
      "task_loss": 0.51276695728302
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.558767557144165,
      "epoch": 1.23,
      "learning_rate": 5.146577249708593e-05,
      "loss": 0.807,
      "step": 3410,
      "task_loss": 0.6122146844863892
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4809660017490387,
      "epoch": 1.24,
      "learning_rate": 5.1418127594444906e-05,
      "loss": 0.7465,
      "step": 3420,
      "task_loss": 0.35081037878990173
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6747417449951172,
      "epoch": 1.24,
      "learning_rate": 5.1370372252735796e-05,
      "loss": 0.7898,
      "step": 3430,
      "task_loss": 0.8822129964828491
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6882492303848267,
      "epoch": 1.24,
      "learning_rate": 5.1322506718201206e-05,
      "loss": 0.7297,
      "step": 3440,
      "task_loss": 0.6773614883422852
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8323681354522705,
      "epoch": 1.25,
      "learning_rate": 5.1274531237651915e-05,
      "loss": 0.8202,
      "step": 3450,
      "task_loss": 1.0216150283813477
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9863547682762146,
      "epoch": 1.25,
      "learning_rate": 5.122644605846565e-05,
      "loss": 0.8462,
      "step": 3460,
      "task_loss": 1.3056061267852783
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7683500647544861,
      "epoch": 1.25,
      "learning_rate": 5.117825142858575e-05,
      "loss": 0.7162,
      "step": 3470,
      "task_loss": 0.978795051574707
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9813886880874634,
      "epoch": 1.26,
      "learning_rate": 5.1129947596519946e-05,
      "loss": 0.7575,
      "step": 3480,
      "task_loss": 0.8878324031829834
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.039483666419983,
      "epoch": 1.26,
      "learning_rate": 5.1081534811339035e-05,
      "loss": 0.7397,
      "step": 3490,
      "task_loss": 0.7671569585800171
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8197644948959351,
      "epoch": 1.26,
      "learning_rate": 5.10330133226756e-05,
      "loss": 0.8642,
      "step": 3500,
      "task_loss": 0.5664966702461243
    },
    {
      "epoch": 1.26,
      "eval_exact_match": 80.7379375591296,
      "eval_f1": 88.59756699218569,
      "step": 3500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7040039300918579,
      "epoch": 1.27,
      "learning_rate": 5.0984383380722776e-05,
      "loss": 0.779,
      "step": 3510,
      "task_loss": 1.0531933307647705
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8066257834434509,
      "epoch": 1.27,
      "learning_rate": 5.0935645236232885e-05,
      "loss": 0.7653,
      "step": 3520,
      "task_loss": 0.7610957622528076
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0015265941619873,
      "epoch": 1.28,
      "learning_rate": 5.088679914051619e-05,
      "loss": 0.7878,
      "step": 3530,
      "task_loss": 0.990369439125061
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5788767337799072,
      "epoch": 1.28,
      "learning_rate": 5.0837845345439585e-05,
      "loss": 0.8144,
      "step": 3540,
      "task_loss": 0.6293957829475403
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6693412065505981,
      "epoch": 1.28,
      "learning_rate": 5.078878410342531e-05,
      "loss": 0.7938,
      "step": 3550,
      "task_loss": 0.5734085440635681
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8092402219772339,
      "epoch": 1.29,
      "learning_rate": 5.073961566744961e-05,
      "loss": 0.7598,
      "step": 3560,
      "task_loss": 0.8599215745925903
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7224042415618896,
      "epoch": 1.29,
      "learning_rate": 5.069034029104148e-05,
      "loss": 0.7476,
      "step": 3570,
      "task_loss": 0.6502309441566467
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5168977975845337,
      "epoch": 1.29,
      "learning_rate": 5.064095822828134e-05,
      "loss": 0.7329,
      "step": 3580,
      "task_loss": 0.6600958704948425
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7091388702392578,
      "epoch": 1.3,
      "learning_rate": 5.05914697337997e-05,
      "loss": 0.8189,
      "step": 3590,
      "task_loss": 0.43178802728652954
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6240181922912598,
      "epoch": 1.3,
      "learning_rate": 5.054187506277588e-05,
      "loss": 0.7433,
      "step": 3600,
      "task_loss": 0.5513472557067871
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.044684886932373,
      "epoch": 1.3,
      "learning_rate": 5.0492174470936685e-05,
      "loss": 0.8185,
      "step": 3610,
      "task_loss": 0.6776243448257446
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6448076963424683,
      "epoch": 1.31,
      "learning_rate": 5.044236821455508e-05,
      "loss": 0.7551,
      "step": 3620,
      "task_loss": 0.62340247631073
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0077701807022095,
      "epoch": 1.31,
      "learning_rate": 5.039245655044887e-05,
      "loss": 0.8528,
      "step": 3630,
      "task_loss": 0.9782606363296509
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.989992082118988,
      "epoch": 1.32,
      "learning_rate": 5.0342439735979367e-05,
      "loss": 0.8235,
      "step": 3640,
      "task_loss": 0.947731614112854
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6484651565551758,
      "epoch": 1.32,
      "learning_rate": 5.029231802905011e-05,
      "loss": 0.8563,
      "step": 3650,
      "task_loss": 0.8938379883766174
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9811206459999084,
      "epoch": 1.32,
      "learning_rate": 5.024209168810545e-05,
      "loss": 0.8731,
      "step": 3660,
      "task_loss": 0.7048652172088623
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9399855136871338,
      "epoch": 1.33,
      "learning_rate": 5.0191760972129305e-05,
      "loss": 0.8083,
      "step": 3670,
      "task_loss": 0.49602335691452026
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6922057271003723,
      "epoch": 1.33,
      "learning_rate": 5.014132614064376e-05,
      "loss": 0.7944,
      "step": 3680,
      "task_loss": 0.680495023727417
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.685440182685852,
      "epoch": 1.33,
      "learning_rate": 5.009078745370778e-05,
      "loss": 0.8661,
      "step": 3690,
      "task_loss": 0.7724747657775879
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8706653714179993,
      "epoch": 1.34,
      "learning_rate": 5.004014517191581e-05,
      "loss": 0.7439,
      "step": 3700,
      "task_loss": 0.8229076862335205
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8825914263725281,
      "epoch": 1.34,
      "learning_rate": 4.998939955639651e-05,
      "loss": 0.7976,
      "step": 3710,
      "task_loss": 0.4416514039039612
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1340502500534058,
      "epoch": 1.34,
      "learning_rate": 4.993855086881132e-05,
      "loss": 0.8594,
      "step": 3720,
      "task_loss": 1.094773769378662
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7764121294021606,
      "epoch": 1.35,
      "learning_rate": 4.988759937135318e-05,
      "loss": 0.7709,
      "step": 3730,
      "task_loss": 1.3073153495788574
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9309306740760803,
      "epoch": 1.35,
      "learning_rate": 4.983654532674513e-05,
      "loss": 0.7804,
      "step": 3740,
      "task_loss": 0.9863795638084412
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6571683287620544,
      "epoch": 1.36,
      "learning_rate": 4.9785388998239e-05,
      "loss": 0.838,
      "step": 3750,
      "task_loss": 1.0076504945755005
    },
    {
      "epoch": 1.36,
      "eval_exact_match": 80.55818353831599,
      "eval_f1": 88.49080617027154,
      "step": 3750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7349100112915039,
      "epoch": 1.36,
      "learning_rate": 4.9734130649614034e-05,
      "loss": 0.8363,
      "step": 3760,
      "task_loss": 0.6905062794685364
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6799863576889038,
      "epoch": 1.36,
      "learning_rate": 4.96827705451755e-05,
      "loss": 0.7666,
      "step": 3770,
      "task_loss": 0.8045142292976379
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.96070396900177,
      "epoch": 1.37,
      "learning_rate": 4.963130894975336e-05,
      "loss": 0.8555,
      "step": 3780,
      "task_loss": 0.9497197270393372
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7121753692626953,
      "epoch": 1.37,
      "learning_rate": 4.9579746128700925e-05,
      "loss": 0.8341,
      "step": 3790,
      "task_loss": 0.6066538095474243
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7798603177070618,
      "epoch": 1.37,
      "learning_rate": 4.952808234789343e-05,
      "loss": 0.8343,
      "step": 3800,
      "task_loss": 1.0046956539154053
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8791223764419556,
      "epoch": 1.38,
      "learning_rate": 4.9476317873726695e-05,
      "loss": 0.7689,
      "step": 3810,
      "task_loss": 0.8858872652053833
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5737714767456055,
      "epoch": 1.38,
      "learning_rate": 4.942445297311577e-05,
      "loss": 0.7408,
      "step": 3820,
      "task_loss": 0.548050045967102
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8374737501144409,
      "epoch": 1.38,
      "learning_rate": 4.9372487913493503e-05,
      "loss": 0.7491,
      "step": 3830,
      "task_loss": 0.9598158597946167
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8589112758636475,
      "epoch": 1.39,
      "learning_rate": 4.9320422962809235e-05,
      "loss": 0.7824,
      "step": 3840,
      "task_loss": 1.1727523803710938
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9275731444358826,
      "epoch": 1.39,
      "learning_rate": 4.926825838952736e-05,
      "loss": 0.7659,
      "step": 3850,
      "task_loss": 1.0407650470733643
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5879838466644287,
      "epoch": 1.4,
      "learning_rate": 4.921599446262594e-05,
      "loss": 0.7324,
      "step": 3860,
      "task_loss": 0.4965895414352417
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7945241928100586,
      "epoch": 1.4,
      "learning_rate": 4.916363145159537e-05,
      "loss": 0.8039,
      "step": 3870,
      "task_loss": 0.9777019023895264
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6736628413200378,
      "epoch": 1.4,
      "learning_rate": 4.911116962643693e-05,
      "loss": 0.7911,
      "step": 3880,
      "task_loss": 0.9778854250907898
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5995903015136719,
      "epoch": 1.41,
      "learning_rate": 4.9058609257661456e-05,
      "loss": 0.7825,
      "step": 3890,
      "task_loss": 0.6814795732498169
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7362528443336487,
      "epoch": 1.41,
      "learning_rate": 4.9005950616287865e-05,
      "loss": 0.9157,
      "step": 3900,
      "task_loss": 0.9911819100379944
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7359405159950256,
      "epoch": 1.41,
      "learning_rate": 4.895319397384182e-05,
      "loss": 0.7861,
      "step": 3910,
      "task_loss": 0.6138168573379517
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0004394054412842,
      "epoch": 1.42,
      "learning_rate": 4.8900339602354324e-05,
      "loss": 0.7014,
      "step": 3920,
      "task_loss": 1.2659577131271362
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.566203773021698,
      "epoch": 1.42,
      "learning_rate": 4.884738777436027e-05,
      "loss": 0.7089,
      "step": 3930,
      "task_loss": 0.6886886358261108
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6759481430053711,
      "epoch": 1.42,
      "learning_rate": 4.87943387628971e-05,
      "loss": 0.78,
      "step": 3940,
      "task_loss": 0.7935036420822144
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6697784066200256,
      "epoch": 1.43,
      "learning_rate": 4.874119284150336e-05,
      "loss": 0.7388,
      "step": 3950,
      "task_loss": 0.602413535118103
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8146666288375854,
      "epoch": 1.43,
      "learning_rate": 4.868795028421728e-05,
      "loss": 0.8598,
      "step": 3960,
      "task_loss": 0.9612588882446289
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.995978057384491,
      "epoch": 1.43,
      "learning_rate": 4.86346113655754e-05,
      "loss": 0.8505,
      "step": 3970,
      "task_loss": 0.7699757814407349
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8782490491867065,
      "epoch": 1.44,
      "learning_rate": 4.8581176360611114e-05,
      "loss": 0.8492,
      "step": 3980,
      "task_loss": 0.879914402961731
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8815962076187134,
      "epoch": 1.44,
      "learning_rate": 4.852764554485328e-05,
      "loss": 0.8009,
      "step": 3990,
      "task_loss": 0.573132336139679
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8560042381286621,
      "epoch": 1.45,
      "learning_rate": 4.8474019194324777e-05,
      "loss": 0.842,
      "step": 4000,
      "task_loss": 1.0868232250213623
    },
    {
      "epoch": 1.45,
      "eval_exact_match": 80.65279091769158,
      "eval_f1": 88.59402264707015,
      "step": 4000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4224177896976471,
      "epoch": 1.45,
      "learning_rate": 4.8420297585541116e-05,
      "loss": 0.7367,
      "step": 4010,
      "task_loss": 0.4378919005393982
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5388481616973877,
      "epoch": 1.45,
      "learning_rate": 4.836648099550896e-05,
      "loss": 0.7675,
      "step": 4020,
      "task_loss": 0.44602516293525696
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8967180252075195,
      "epoch": 1.46,
      "learning_rate": 4.8312569701724754e-05,
      "loss": 0.7752,
      "step": 4030,
      "task_loss": 0.7708736062049866
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8984067440032959,
      "epoch": 1.46,
      "learning_rate": 4.8258563982173244e-05,
      "loss": 0.9208,
      "step": 4040,
      "task_loss": 0.9523441791534424
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6791067123413086,
      "epoch": 1.46,
      "learning_rate": 4.82044641153261e-05,
      "loss": 0.7119,
      "step": 4050,
      "task_loss": 0.957298219203949
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5926834940910339,
      "epoch": 1.47,
      "learning_rate": 4.8150270380140414e-05,
      "loss": 0.8366,
      "step": 4060,
      "task_loss": 0.86881023645401
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8107943534851074,
      "epoch": 1.47,
      "learning_rate": 4.809598305605732e-05,
      "loss": 0.7996,
      "step": 4070,
      "task_loss": 0.8448442220687866
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0969449281692505,
      "epoch": 1.47,
      "learning_rate": 4.8041602423000505e-05,
      "loss": 0.8194,
      "step": 4080,
      "task_loss": 1.0413978099822998
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8810731172561646,
      "epoch": 1.48,
      "learning_rate": 4.79871287613748e-05,
      "loss": 0.8494,
      "step": 4090,
      "task_loss": 0.8740400075912476
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5368291735649109,
      "epoch": 1.48,
      "learning_rate": 4.793256235206473e-05,
      "loss": 0.7184,
      "step": 4100,
      "task_loss": 0.465472012758255
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8813285827636719,
      "epoch": 1.49,
      "learning_rate": 4.787790347643305e-05,
      "loss": 0.7444,
      "step": 4110,
      "task_loss": 0.8482736349105835
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8207933902740479,
      "epoch": 1.49,
      "learning_rate": 4.782315241631929e-05,
      "loss": 0.7815,
      "step": 4120,
      "task_loss": 0.8252826929092407
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7789289355278015,
      "epoch": 1.49,
      "learning_rate": 4.776830945403833e-05,
      "loss": 0.7304,
      "step": 4130,
      "task_loss": 1.0397837162017822
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0559360980987549,
      "epoch": 1.5,
      "learning_rate": 4.771337487237894e-05,
      "loss": 0.741,
      "step": 4140,
      "task_loss": 1.10592520236969
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9466273784637451,
      "epoch": 1.5,
      "learning_rate": 4.7658348954602285e-05,
      "loss": 0.7681,
      "step": 4150,
      "task_loss": 0.8297045230865479
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6521314382553101,
      "epoch": 1.5,
      "learning_rate": 4.76032319844405e-05,
      "loss": 0.8071,
      "step": 4160,
      "task_loss": 0.7315813302993774
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.860609769821167,
      "epoch": 1.51,
      "learning_rate": 4.754802424609521e-05,
      "loss": 0.8721,
      "step": 4170,
      "task_loss": 0.8967682123184204
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6362369060516357,
      "epoch": 1.51,
      "learning_rate": 4.74927260242361e-05,
      "loss": 0.7326,
      "step": 4180,
      "task_loss": 0.5996143221855164
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7511730194091797,
      "epoch": 1.51,
      "learning_rate": 4.7437337603999376e-05,
      "loss": 0.7683,
      "step": 4190,
      "task_loss": 0.43765193223953247
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8653720617294312,
      "epoch": 1.52,
      "learning_rate": 4.7381859270986374e-05,
      "loss": 0.8353,
      "step": 4200,
      "task_loss": 1.0502616167068481
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6381678581237793,
      "epoch": 1.52,
      "learning_rate": 4.732629131126202e-05,
      "loss": 0.7639,
      "step": 4210,
      "task_loss": 0.5875723361968994
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7945320010185242,
      "epoch": 1.53,
      "learning_rate": 4.727063401135341e-05,
      "loss": 0.7863,
      "step": 4220,
      "task_loss": 1.58515465259552
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6202449798583984,
      "epoch": 1.53,
      "learning_rate": 4.7214887658248296e-05,
      "loss": 0.7453,
      "step": 4230,
      "task_loss": 0.8521810173988342
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7749199867248535,
      "epoch": 1.53,
      "learning_rate": 4.715905253939361e-05,
      "loss": 0.8201,
      "step": 4240,
      "task_loss": 0.766315221786499
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.729579746723175,
      "epoch": 1.54,
      "learning_rate": 4.710312894269402e-05,
      "loss": 0.8902,
      "step": 4250,
      "task_loss": 0.6951254606246948
    },
    {
      "epoch": 1.54,
      "eval_exact_match": 81.32450331125828,
      "eval_f1": 88.97617818833929,
      "step": 4250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9440791606903076,
      "epoch": 1.54,
      "learning_rate": 4.704711715651038e-05,
      "loss": 0.7623,
      "step": 4260,
      "task_loss": 1.0249818563461304
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8210508227348328,
      "epoch": 1.54,
      "learning_rate": 4.699101746965829e-05,
      "loss": 0.8148,
      "step": 4270,
      "task_loss": 0.9564526081085205
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7049989104270935,
      "epoch": 1.55,
      "learning_rate": 4.6934830171406636e-05,
      "loss": 0.7944,
      "step": 4280,
      "task_loss": 0.941627025604248
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6714322566986084,
      "epoch": 1.55,
      "learning_rate": 4.687855555147597e-05,
      "loss": 0.7231,
      "step": 4290,
      "task_loss": 1.0258185863494873
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8125296831130981,
      "epoch": 1.55,
      "learning_rate": 4.682219390003719e-05,
      "loss": 0.6968,
      "step": 4300,
      "task_loss": 0.7712159156799316
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9884834885597229,
      "epoch": 1.56,
      "learning_rate": 4.676574550770991e-05,
      "loss": 0.7995,
      "step": 4310,
      "task_loss": 0.8624418377876282
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.195237398147583,
      "epoch": 1.56,
      "learning_rate": 4.6709210665561035e-05,
      "loss": 0.8232,
      "step": 4320,
      "task_loss": 0.8453245162963867
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.87236487865448,
      "epoch": 1.56,
      "learning_rate": 4.6652589665103204e-05,
      "loss": 0.7264,
      "step": 4330,
      "task_loss": 0.9678608179092407
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7140805721282959,
      "epoch": 1.57,
      "learning_rate": 4.659588279829335e-05,
      "loss": 0.8526,
      "step": 4340,
      "task_loss": 0.8649274706840515
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8390883207321167,
      "epoch": 1.57,
      "learning_rate": 4.653909035753114e-05,
      "loss": 0.7946,
      "step": 4350,
      "task_loss": 0.8922809362411499
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6814520955085754,
      "epoch": 1.58,
      "learning_rate": 4.648221263565751e-05,
      "loss": 0.7372,
      "step": 4360,
      "task_loss": 0.740043580532074
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.922714114189148,
      "epoch": 1.58,
      "learning_rate": 4.642524992595309e-05,
      "loss": 0.8215,
      "step": 4370,
      "task_loss": 0.5661073327064514
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.942134439945221,
      "epoch": 1.58,
      "learning_rate": 4.636820252213679e-05,
      "loss": 0.7335,
      "step": 4380,
      "task_loss": 0.8119366765022278
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0408191680908203,
      "epoch": 1.59,
      "learning_rate": 4.63110707183642e-05,
      "loss": 0.8429,
      "step": 4390,
      "task_loss": 0.9176040887832642
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9637014865875244,
      "epoch": 1.59,
      "learning_rate": 4.6253854809226115e-05,
      "loss": 0.75,
      "step": 4400,
      "task_loss": 1.2698314189910889
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9256500005722046,
      "epoch": 1.59,
      "learning_rate": 4.6196555089747e-05,
      "loss": 0.8744,
      "step": 4410,
      "task_loss": 0.9903556704521179
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8625351190567017,
      "epoch": 1.6,
      "learning_rate": 4.6139171855383473e-05,
      "loss": 0.7773,
      "step": 4420,
      "task_loss": 0.6395310163497925
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5867180824279785,
      "epoch": 1.6,
      "learning_rate": 4.608170540202279e-05,
      "loss": 0.7754,
      "step": 4430,
      "task_loss": 0.774559736251831
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5111645460128784,
      "epoch": 1.6,
      "learning_rate": 4.602415602598132e-05,
      "loss": 0.8272,
      "step": 4440,
      "task_loss": 0.4568556547164917
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6454306840896606,
      "epoch": 1.61,
      "learning_rate": 4.5966524024002976e-05,
      "loss": 0.831,
      "step": 4450,
      "task_loss": 0.40431562066078186
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5694042444229126,
      "epoch": 1.61,
      "learning_rate": 4.590880969325774e-05,
      "loss": 0.7259,
      "step": 4460,
      "task_loss": 0.7366609573364258
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6696236729621887,
      "epoch": 1.62,
      "learning_rate": 4.585101333134013e-05,
      "loss": 0.7484,
      "step": 4470,
      "task_loss": 0.5174720883369446
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6274579167366028,
      "epoch": 1.62,
      "learning_rate": 4.5793135236267626e-05,
      "loss": 0.7402,
      "step": 4480,
      "task_loss": 0.747776985168457
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7382104396820068,
      "epoch": 1.62,
      "learning_rate": 4.573517570647912e-05,
      "loss": 0.7127,
      "step": 4490,
      "task_loss": 0.7014381885528564
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7991716861724854,
      "epoch": 1.63,
      "learning_rate": 4.567713504083346e-05,
      "loss": 0.7075,
      "step": 4500,
      "task_loss": 0.9051458835601807
    },
    {
      "epoch": 1.63,
      "eval_exact_match": 81.36234626300852,
      "eval_f1": 88.86679422849299,
      "step": 4500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7607914209365845,
      "epoch": 1.63,
      "learning_rate": 4.5619013538607814e-05,
      "loss": 0.7915,
      "step": 4510,
      "task_loss": 0.7439800500869751
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7172999382019043,
      "epoch": 1.63,
      "learning_rate": 4.556081149949621e-05,
      "loss": 0.7687,
      "step": 4520,
      "task_loss": 0.8847224116325378
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7143849730491638,
      "epoch": 1.64,
      "learning_rate": 4.5502529223607936e-05,
      "loss": 0.815,
      "step": 4530,
      "task_loss": 0.8012955784797668
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.046454668045044,
      "epoch": 1.64,
      "learning_rate": 4.5444167011466e-05,
      "loss": 0.8533,
      "step": 4540,
      "task_loss": 0.7980386018753052
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.649091362953186,
      "epoch": 1.64,
      "learning_rate": 4.5385725164005585e-05,
      "loss": 0.7456,
      "step": 4550,
      "task_loss": 0.49612122774124146
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7305445075035095,
      "epoch": 1.65,
      "learning_rate": 4.5327203982572514e-05,
      "loss": 0.7565,
      "step": 4560,
      "task_loss": 0.6083104610443115
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2499300241470337,
      "epoch": 1.65,
      "learning_rate": 4.526860376892167e-05,
      "loss": 0.9337,
      "step": 4570,
      "task_loss": 1.3941344022750854
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0218126773834229,
      "epoch": 1.66,
      "learning_rate": 4.5209924825215474e-05,
      "loss": 0.7354,
      "step": 4580,
      "task_loss": 0.9629853963851929
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7129672169685364,
      "epoch": 1.66,
      "learning_rate": 4.515116745402229e-05,
      "loss": 0.7491,
      "step": 4590,
      "task_loss": 0.6554641127586365
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7573808431625366,
      "epoch": 1.66,
      "learning_rate": 4.509233195831487e-05,
      "loss": 0.8387,
      "step": 4600,
      "task_loss": 0.813507080078125
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7598761320114136,
      "epoch": 1.67,
      "learning_rate": 4.5033418641468816e-05,
      "loss": 0.7462,
      "step": 4610,
      "task_loss": 0.8632051348686218
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4978368580341339,
      "epoch": 1.67,
      "learning_rate": 4.497442780726101e-05,
      "loss": 0.6874,
      "step": 4620,
      "task_loss": 0.6175663471221924
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6145604848861694,
      "epoch": 1.67,
      "learning_rate": 4.4915359759868034e-05,
      "loss": 0.713,
      "step": 4630,
      "task_loss": 0.6943603157997131
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7277992963790894,
      "epoch": 1.68,
      "learning_rate": 4.485621480386459e-05,
      "loss": 0.8655,
      "step": 4640,
      "task_loss": 0.6401335000991821
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5838598608970642,
      "epoch": 1.68,
      "learning_rate": 4.479699324422195e-05,
      "loss": 0.6839,
      "step": 4650,
      "task_loss": 0.43302324414253235
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5886027216911316,
      "epoch": 1.68,
      "learning_rate": 4.473769538630642e-05,
      "loss": 0.6724,
      "step": 4660,
      "task_loss": 0.7869935631752014
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8098095655441284,
      "epoch": 1.69,
      "learning_rate": 4.4678321535877664e-05,
      "loss": 0.7717,
      "step": 4670,
      "task_loss": 1.0552672147750854
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6143741607666016,
      "epoch": 1.69,
      "learning_rate": 4.4618871999087255e-05,
      "loss": 0.825,
      "step": 4680,
      "task_loss": 0.6005398035049438
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.659191370010376,
      "epoch": 1.69,
      "learning_rate": 4.455934708247696e-05,
      "loss": 0.7628,
      "step": 4690,
      "task_loss": 0.5167052745819092
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.686668872833252,
      "epoch": 1.7,
      "learning_rate": 4.449974709297729e-05,
      "loss": 0.7712,
      "step": 4700,
      "task_loss": 0.7116361856460571
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6677272915840149,
      "epoch": 1.7,
      "learning_rate": 4.4440072337905815e-05,
      "loss": 0.7906,
      "step": 4710,
      "task_loss": 0.5613526701927185
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45507311820983887,
      "epoch": 1.71,
      "learning_rate": 4.438032312496565e-05,
      "loss": 0.8587,
      "step": 4720,
      "task_loss": 0.553399920463562
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.2598834037780762,
      "epoch": 1.71,
      "learning_rate": 4.432049976224383e-05,
      "loss": 0.9358,
      "step": 4730,
      "task_loss": 1.3447682857513428
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7433596849441528,
      "epoch": 1.71,
      "learning_rate": 4.4260602558209715e-05,
      "loss": 0.7002,
      "step": 4740,
      "task_loss": 0.9043145179748535
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8138759136199951,
      "epoch": 1.72,
      "learning_rate": 4.420063182171345e-05,
      "loss": 0.8739,
      "step": 4750,
      "task_loss": 0.9384928941726685
    },
    {
      "epoch": 1.72,
      "eval_exact_match": 81.15421002838221,
      "eval_f1": 88.77749919883216,
      "step": 4750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6397891044616699,
      "epoch": 1.72,
      "learning_rate": 4.414058786198431e-05,
      "loss": 0.7098,
      "step": 4760,
      "task_loss": 0.641806423664093
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9567407965660095,
      "epoch": 1.72,
      "learning_rate": 4.408047098862914e-05,
      "loss": 0.7674,
      "step": 4770,
      "task_loss": 0.8469678163528442
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1057536602020264,
      "epoch": 1.73,
      "learning_rate": 4.402028151163076e-05,
      "loss": 0.7887,
      "step": 4780,
      "task_loss": 0.9880897998809814
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.706554114818573,
      "epoch": 1.73,
      "learning_rate": 4.396001974134634e-05,
      "loss": 0.7421,
      "step": 4790,
      "task_loss": 0.516491711139679
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7376039028167725,
      "epoch": 1.73,
      "learning_rate": 4.389968598850585e-05,
      "loss": 0.7885,
      "step": 4800,
      "task_loss": 0.7677445411682129
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9835151433944702,
      "epoch": 1.74,
      "learning_rate": 4.38392805642104e-05,
      "loss": 0.7134,
      "step": 4810,
      "task_loss": 0.8267749547958374
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9274640083312988,
      "epoch": 1.74,
      "learning_rate": 4.3778803779930655e-05,
      "loss": 0.7763,
      "step": 4820,
      "task_loss": 0.7708688974380493
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.807561993598938,
      "epoch": 1.75,
      "learning_rate": 4.371825594750528e-05,
      "loss": 0.663,
      "step": 4830,
      "task_loss": 1.1051645278930664
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8475966453552246,
      "epoch": 1.75,
      "learning_rate": 4.365763737913924e-05,
      "loss": 0.6866,
      "step": 4840,
      "task_loss": 0.7194445133209229
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.662378191947937,
      "epoch": 1.75,
      "learning_rate": 4.359694838740225e-05,
      "loss": 0.7215,
      "step": 4850,
      "task_loss": 0.9244500398635864
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8736042976379395,
      "epoch": 1.76,
      "learning_rate": 4.353618928522718e-05,
      "loss": 0.6807,
      "step": 4860,
      "task_loss": 0.5476257801055908
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9758906960487366,
      "epoch": 1.76,
      "learning_rate": 4.3475360385908385e-05,
      "loss": 0.7897,
      "step": 4870,
      "task_loss": 1.1290806531906128
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5853473544120789,
      "epoch": 1.76,
      "learning_rate": 4.3414462003100126e-05,
      "loss": 0.7186,
      "step": 4880,
      "task_loss": 0.6802440881729126
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8614143133163452,
      "epoch": 1.77,
      "learning_rate": 4.335349445081493e-05,
      "loss": 0.77,
      "step": 4890,
      "task_loss": 0.674217939376831
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6680730581283569,
      "epoch": 1.77,
      "learning_rate": 4.329245804342201e-05,
      "loss": 0.7647,
      "step": 4900,
      "task_loss": 0.5699944496154785
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8439807891845703,
      "epoch": 1.77,
      "learning_rate": 4.3231353095645604e-05,
      "loss": 0.7738,
      "step": 4910,
      "task_loss": 0.8963417410850525
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0166106224060059,
      "epoch": 1.78,
      "learning_rate": 4.317017992256337e-05,
      "loss": 0.7552,
      "step": 4920,
      "task_loss": 0.746208667755127
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.114891529083252,
      "epoch": 1.78,
      "learning_rate": 4.310893883960476e-05,
      "loss": 0.7747,
      "step": 4930,
      "task_loss": 1.3238515853881836
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9112187623977661,
      "epoch": 1.79,
      "learning_rate": 4.304763016254939e-05,
      "loss": 0.742,
      "step": 4940,
      "task_loss": 1.0595370531082153
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6720353364944458,
      "epoch": 1.79,
      "learning_rate": 4.298625420752541e-05,
      "loss": 0.7768,
      "step": 4950,
      "task_loss": 0.7741559147834778
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.14748215675354,
      "epoch": 1.79,
      "learning_rate": 4.29248112910079e-05,
      "loss": 0.7824,
      "step": 4960,
      "task_loss": 0.944894552230835
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9729048013687134,
      "epoch": 1.8,
      "learning_rate": 4.286330172981718e-05,
      "loss": 0.8742,
      "step": 4970,
      "task_loss": 1.025977611541748
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.939768373966217,
      "epoch": 1.8,
      "learning_rate": 4.2801725841117244e-05,
      "loss": 0.7605,
      "step": 4980,
      "task_loss": 0.6698074340820312
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5564023852348328,
      "epoch": 1.8,
      "learning_rate": 4.274008394241407e-05,
      "loss": 0.8018,
      "step": 4990,
      "task_loss": 0.4944656491279602
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8198148012161255,
      "epoch": 1.81,
      "learning_rate": 4.267837635155402e-05,
      "loss": 0.7426,
      "step": 5000,
      "task_loss": 0.6710585355758667
    },
    {
      "epoch": 1.81,
      "eval_exact_match": 81.48533585619678,
      "eval_f1": 88.84117960148326,
      "step": 5000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1788697242736816,
      "epoch": 1.81,
      "learning_rate": 4.2616603386722185e-05,
      "loss": 0.7101,
      "step": 5010,
      "task_loss": 0.7871378660202026
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8743175268173218,
      "epoch": 1.81,
      "learning_rate": 4.2554765366440736e-05,
      "loss": 0.7741,
      "step": 5020,
      "task_loss": 1.2785630226135254
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9149342775344849,
      "epoch": 1.82,
      "learning_rate": 4.249286260956732e-05,
      "loss": 0.8054,
      "step": 5030,
      "task_loss": 0.8568088412284851
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6473352909088135,
      "epoch": 1.82,
      "learning_rate": 4.2430895435293335e-05,
      "loss": 0.8112,
      "step": 5040,
      "task_loss": 0.44709980487823486
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.807868242263794,
      "epoch": 1.83,
      "learning_rate": 4.2368864163142396e-05,
      "loss": 0.7528,
      "step": 5050,
      "task_loss": 0.8173571825027466
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8056721091270447,
      "epoch": 1.83,
      "learning_rate": 4.230676911296859e-05,
      "loss": 0.7581,
      "step": 5060,
      "task_loss": 1.0143595933914185
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8686327934265137,
      "epoch": 1.83,
      "learning_rate": 4.2244610604954874e-05,
      "loss": 0.7872,
      "step": 5070,
      "task_loss": 0.8795457482337952
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8215399980545044,
      "epoch": 1.84,
      "learning_rate": 4.218238895961143e-05,
      "loss": 0.8153,
      "step": 5080,
      "task_loss": 0.6292321085929871
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6530864238739014,
      "epoch": 1.84,
      "learning_rate": 4.212010449777397e-05,
      "loss": 0.7122,
      "step": 5090,
      "task_loss": 0.908234715461731
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6864101886749268,
      "epoch": 1.84,
      "learning_rate": 4.2057757540602134e-05,
      "loss": 0.7769,
      "step": 5100,
      "task_loss": 0.761084794998169
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9541764855384827,
      "epoch": 1.85,
      "learning_rate": 4.199534840957779e-05,
      "loss": 0.7948,
      "step": 5110,
      "task_loss": 0.8573384284973145
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5462681651115417,
      "epoch": 1.85,
      "learning_rate": 4.193287742650341e-05,
      "loss": 0.6368,
      "step": 5120,
      "task_loss": 0.4982544779777527
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6552547216415405,
      "epoch": 1.85,
      "learning_rate": 4.1870344913500387e-05,
      "loss": 0.8185,
      "step": 5130,
      "task_loss": 1.1584696769714355
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7229572534561157,
      "epoch": 1.86,
      "learning_rate": 4.180775119300738e-05,
      "loss": 0.7605,
      "step": 5140,
      "task_loss": 0.7089170813560486
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43640923500061035,
      "epoch": 1.86,
      "learning_rate": 4.174509658777867e-05,
      "loss": 0.7827,
      "step": 5150,
      "task_loss": 0.9247764945030212
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.969160258769989,
      "epoch": 1.86,
      "learning_rate": 4.168238142088247e-05,
      "loss": 0.7693,
      "step": 5160,
      "task_loss": 0.8934771418571472
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7915940284729004,
      "epoch": 1.87,
      "learning_rate": 4.1619606015699254e-05,
      "loss": 0.7791,
      "step": 5170,
      "task_loss": 0.7979614734649658
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5941789150238037,
      "epoch": 1.87,
      "learning_rate": 4.155677069592015e-05,
      "loss": 0.7402,
      "step": 5180,
      "task_loss": 0.8741079568862915
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.63032066822052,
      "epoch": 1.88,
      "learning_rate": 4.149387578554516e-05,
      "loss": 0.8416,
      "step": 5190,
      "task_loss": 0.6060689687728882
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7978523373603821,
      "epoch": 1.88,
      "learning_rate": 4.143092160888162e-05,
      "loss": 0.7329,
      "step": 5200,
      "task_loss": 1.083559274673462
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7854770421981812,
      "epoch": 1.88,
      "learning_rate": 4.1367908490542424e-05,
      "loss": 0.6658,
      "step": 5210,
      "task_loss": 0.6225592494010925
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5963532328605652,
      "epoch": 1.89,
      "learning_rate": 4.1304836755444396e-05,
      "loss": 0.768,
      "step": 5220,
      "task_loss": 0.6593186855316162
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8601577281951904,
      "epoch": 1.89,
      "learning_rate": 4.1241706728806614e-05,
      "loss": 0.6606,
      "step": 5230,
      "task_loss": 0.8627996444702148
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6065561175346375,
      "epoch": 1.89,
      "learning_rate": 4.1178518736148726e-05,
      "loss": 0.6936,
      "step": 5240,
      "task_loss": 0.49574264883995056
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8394991159439087,
      "epoch": 1.9,
      "learning_rate": 4.111527310328926e-05,
      "loss": 0.809,
      "step": 5250,
      "task_loss": 0.6712428331375122
    },
    {
      "epoch": 1.9,
      "eval_exact_match": 81.60832544938505,
      "eval_f1": 88.89492666071773,
      "step": 5250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8298504948616028,
      "epoch": 1.9,
      "learning_rate": 4.105197015634399e-05,
      "loss": 0.7804,
      "step": 5260,
      "task_loss": 0.8732647895812988
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6441981196403503,
      "epoch": 1.9,
      "learning_rate": 4.0988610221724165e-05,
      "loss": 0.7308,
      "step": 5270,
      "task_loss": 0.6386420726776123
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.327653408050537,
      "epoch": 1.91,
      "learning_rate": 4.092519362613494e-05,
      "loss": 0.8782,
      "step": 5280,
      "task_loss": 1.0381779670715332
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6743829250335693,
      "epoch": 1.91,
      "learning_rate": 4.08617206965736e-05,
      "loss": 0.7572,
      "step": 5290,
      "task_loss": 0.6540873050689697
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.1734296083450317,
      "epoch": 1.92,
      "learning_rate": 4.079819176032791e-05,
      "loss": 0.7435,
      "step": 5300,
      "task_loss": 0.9436138868331909
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.97495037317276,
      "epoch": 1.92,
      "learning_rate": 4.073460714497443e-05,
      "loss": 0.7854,
      "step": 5310,
      "task_loss": 1.2246060371398926
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8801750540733337,
      "epoch": 1.92,
      "learning_rate": 4.067096717837681e-05,
      "loss": 0.854,
      "step": 5320,
      "task_loss": 0.799035370349884
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.0151453018188477,
      "epoch": 1.93,
      "learning_rate": 4.060727218868413e-05,
      "loss": 0.7464,
      "step": 5330,
      "task_loss": 0.9292905330657959
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8881112337112427,
      "epoch": 1.93,
      "learning_rate": 4.054352250432917e-05,
      "loss": 0.6831,
      "step": 5340,
      "task_loss": 0.8586100339889526
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9265180230140686,
      "epoch": 1.93,
      "learning_rate": 4.047971845402674e-05,
      "loss": 0.811,
      "step": 5350,
      "task_loss": 0.8724481463432312
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.686070442199707,
      "epoch": 1.94,
      "learning_rate": 4.0415860366771986e-05,
      "loss": 0.754,
      "step": 5360,
      "task_loss": 0.8045363426208496
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9585437774658203,
      "epoch": 1.94,
      "learning_rate": 4.0351948571838665e-05,
      "loss": 0.7982,
      "step": 5370,
      "task_loss": 0.6699115037918091
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8560003638267517,
      "epoch": 1.94,
      "learning_rate": 4.0287983398777485e-05,
      "loss": 0.7212,
      "step": 5380,
      "task_loss": 0.46158933639526367
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8578959107398987,
      "epoch": 1.95,
      "learning_rate": 4.02239651774144e-05,
      "loss": 0.8158,
      "step": 5390,
      "task_loss": 0.8129777908325195
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7851342558860779,
      "epoch": 1.95,
      "learning_rate": 4.015989423784887e-05,
      "loss": 0.7846,
      "step": 5400,
      "task_loss": 0.6260675191879272
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6622868180274963,
      "epoch": 1.96,
      "learning_rate": 4.009577091045222e-05,
      "loss": 0.6604,
      "step": 5410,
      "task_loss": 0.5617303252220154
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6715734601020813,
      "epoch": 1.96,
      "learning_rate": 4.003159552586588e-05,
      "loss": 0.7204,
      "step": 5420,
      "task_loss": 0.5333033204078674
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6720329523086548,
      "epoch": 1.96,
      "learning_rate": 3.996736841499972e-05,
      "loss": 0.77,
      "step": 5430,
      "task_loss": 0.5945698618888855
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7461095452308655,
      "epoch": 1.97,
      "learning_rate": 3.990308990903031e-05,
      "loss": 0.7432,
      "step": 5440,
      "task_loss": 0.89726722240448
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6053512096405029,
      "epoch": 1.97,
      "learning_rate": 3.983876033939925e-05,
      "loss": 0.7936,
      "step": 5450,
      "task_loss": 0.46585309505462646
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5714142322540283,
      "epoch": 1.97,
      "learning_rate": 3.977438003781144e-05,
      "loss": 0.7128,
      "step": 5460,
      "task_loss": 0.5288292169570923
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.734135627746582,
      "epoch": 1.98,
      "learning_rate": 3.970994933623334e-05,
      "loss": 0.7645,
      "step": 5470,
      "task_loss": 0.9016965627670288
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7678335905075073,
      "epoch": 1.98,
      "learning_rate": 3.9645468566891326e-05,
      "loss": 0.7166,
      "step": 5480,
      "task_loss": 0.5996376276016235
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6975915431976318,
      "epoch": 1.98,
      "learning_rate": 3.958093806226994e-05,
      "loss": 0.7699,
      "step": 5490,
      "task_loss": 0.5111979246139526
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6483110785484314,
      "epoch": 1.99,
      "learning_rate": 3.951635815511014e-05,
      "loss": 0.6991,
      "step": 5500,
      "task_loss": 1.0150556564331055
    },
    {
      "epoch": 1.99,
      "eval_exact_match": 81.40964995269631,
      "eval_f1": 88.93790583717866,
      "step": 5500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6153757572174072,
      "epoch": 1.99,
      "learning_rate": 3.9451729178407676e-05,
      "loss": 0.7601,
      "step": 5510,
      "task_loss": 0.49160200357437134
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6790378093719482,
      "epoch": 1.99,
      "learning_rate": 3.9387051465411245e-05,
      "loss": 0.7548,
      "step": 5520,
      "task_loss": 0.7084499597549438
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.648993730545044,
      "epoch": 2.0,
      "learning_rate": 3.93223253496209e-05,
      "loss": 0.8321,
      "step": 5530,
      "task_loss": 0.7858644723892212
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6678321361541748,
      "epoch": 2.0,
      "learning_rate": 3.925755116478628e-05,
      "loss": 0.644,
      "step": 5540,
      "task_loss": 0.40885019302368164
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5563902854919434,
      "epoch": 2.01,
      "learning_rate": 3.919272924490484e-05,
      "loss": 0.584,
      "step": 5550,
      "task_loss": 0.6156301498413086
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.48990246653556824,
      "epoch": 2.01,
      "learning_rate": 3.91278599242202e-05,
      "loss": 0.597,
      "step": 5560,
      "task_loss": 0.5672547221183777
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9053139686584473,
      "epoch": 2.01,
      "learning_rate": 3.9062943537220394e-05,
      "loss": 0.6525,
      "step": 5570,
      "task_loss": 1.018937587738037
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.643484354019165,
      "epoch": 2.02,
      "learning_rate": 3.899798041863615e-05,
      "loss": 0.6214,
      "step": 5580,
      "task_loss": 0.5043469667434692
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9523886442184448,
      "epoch": 2.02,
      "learning_rate": 3.8932970903439134e-05,
      "loss": 0.6711,
      "step": 5590,
      "task_loss": 1.3548312187194824
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.502714216709137,
      "epoch": 2.02,
      "learning_rate": 3.886791532684028e-05,
      "loss": 0.5996,
      "step": 5600,
      "task_loss": 1.051832675933838
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5625650882720947,
      "epoch": 2.03,
      "learning_rate": 3.880281402428802e-05,
      "loss": 0.5483,
      "step": 5610,
      "task_loss": 0.5134047269821167
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6442995071411133,
      "epoch": 2.03,
      "learning_rate": 3.8737667331466554e-05,
      "loss": 0.6055,
      "step": 5620,
      "task_loss": 0.7023986577987671
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5948624610900879,
      "epoch": 2.03,
      "learning_rate": 3.8672475584294126e-05,
      "loss": 0.6392,
      "step": 5630,
      "task_loss": 0.9365327954292297
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9390955567359924,
      "epoch": 2.04,
      "learning_rate": 3.860723911892134e-05,
      "loss": 0.6544,
      "step": 5640,
      "task_loss": 1.2662063837051392
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.612360954284668,
      "epoch": 2.04,
      "learning_rate": 3.8541958271729304e-05,
      "loss": 0.5981,
      "step": 5650,
      "task_loss": 0.834581732749939
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5364567041397095,
      "epoch": 2.05,
      "learning_rate": 3.847663337932806e-05,
      "loss": 0.577,
      "step": 5660,
      "task_loss": 0.5253503918647766
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7174457907676697,
      "epoch": 2.05,
      "learning_rate": 3.841126477855469e-05,
      "loss": 0.6112,
      "step": 5670,
      "task_loss": 0.977493405342102
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7490948438644409,
      "epoch": 2.05,
      "learning_rate": 3.834585280647169e-05,
      "loss": 0.6091,
      "step": 5680,
      "task_loss": 0.9850953817367554
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5830691456794739,
      "epoch": 2.06,
      "learning_rate": 3.82803978003652e-05,
      "loss": 0.6088,
      "step": 5690,
      "task_loss": 0.850995659828186
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5099242925643921,
      "epoch": 2.06,
      "learning_rate": 3.821490009774321e-05,
      "loss": 0.6365,
      "step": 5700,
      "task_loss": 0.6015151739120483
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.781893253326416,
      "epoch": 2.06,
      "learning_rate": 3.814936003633393e-05,
      "loss": 0.6158,
      "step": 5710,
      "task_loss": 0.7863786220550537
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4905242323875427,
      "epoch": 2.07,
      "learning_rate": 3.808377795408394e-05,
      "loss": 0.5719,
      "step": 5720,
      "task_loss": 0.5809369087219238
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6399200558662415,
      "epoch": 2.07,
      "learning_rate": 3.80181541891565e-05,
      "loss": 0.6363,
      "step": 5730,
      "task_loss": 0.7770982384681702
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7491112947463989,
      "epoch": 2.07,
      "learning_rate": 3.7952489079929826e-05,
      "loss": 0.6044,
      "step": 5740,
      "task_loss": 0.6825676560401917
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5090913772583008,
      "epoch": 2.08,
      "learning_rate": 3.7886782964995304e-05,
      "loss": 0.6148,
      "step": 5750,
      "task_loss": 1.064110279083252
    },
    {
      "epoch": 2.08,
      "eval_exact_match": 81.86376537369915,
      "eval_f1": 89.1819885255817,
      "step": 5750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5572813749313354,
      "epoch": 2.08,
      "learning_rate": 3.782103618315575e-05,
      "loss": 0.6215,
      "step": 5760,
      "task_loss": 0.6721667051315308
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.674677848815918,
      "epoch": 2.09,
      "learning_rate": 3.775524907342367e-05,
      "loss": 0.5723,
      "step": 5770,
      "task_loss": 0.8653391003608704
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45973920822143555,
      "epoch": 2.09,
      "learning_rate": 3.768942197501955e-05,
      "loss": 0.6199,
      "step": 5780,
      "task_loss": 0.8878629207611084
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7025313377380371,
      "epoch": 2.09,
      "learning_rate": 3.7623555227370017e-05,
      "loss": 0.5915,
      "step": 5790,
      "task_loss": 0.6296009421348572
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6605579853057861,
      "epoch": 2.1,
      "learning_rate": 3.755764917010618e-05,
      "loss": 0.5127,
      "step": 5800,
      "task_loss": 0.660987377166748
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5550613403320312,
      "epoch": 2.1,
      "learning_rate": 3.749170414306184e-05,
      "loss": 0.6212,
      "step": 5810,
      "task_loss": 0.5019931793212891
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5897912383079529,
      "epoch": 2.1,
      "learning_rate": 3.7425720486271726e-05,
      "loss": 0.6035,
      "step": 5820,
      "task_loss": 0.6660510301589966
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5896255970001221,
      "epoch": 2.11,
      "learning_rate": 3.735969853996976e-05,
      "loss": 0.5614,
      "step": 5830,
      "task_loss": 0.5327462553977966
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6656285524368286,
      "epoch": 2.11,
      "learning_rate": 3.72936386445873e-05,
      "loss": 0.5375,
      "step": 5840,
      "task_loss": 0.6203778386116028
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5610683560371399,
      "epoch": 2.11,
      "learning_rate": 3.722754114075137e-05,
      "loss": 0.6618,
      "step": 5850,
      "task_loss": 0.43519946932792664
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7016445398330688,
      "epoch": 2.12,
      "learning_rate": 3.716140636928295e-05,
      "loss": 0.648,
      "step": 5860,
      "task_loss": 0.7705813646316528
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.496102511882782,
      "epoch": 2.12,
      "learning_rate": 3.709523467119514e-05,
      "loss": 0.5707,
      "step": 5870,
      "task_loss": 0.6810226440429688
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5184519290924072,
      "epoch": 2.13,
      "learning_rate": 3.7029026387691464e-05,
      "loss": 0.5762,
      "step": 5880,
      "task_loss": 0.5692154169082642
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7326685786247253,
      "epoch": 2.13,
      "learning_rate": 3.696278186016411e-05,
      "loss": 0.6707,
      "step": 5890,
      "task_loss": 0.6973918676376343
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6160091161727905,
      "epoch": 2.13,
      "learning_rate": 3.6896501430192134e-05,
      "loss": 0.617,
      "step": 5900,
      "task_loss": 0.6182623505592346
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5769556164741516,
      "epoch": 2.14,
      "learning_rate": 3.6830185439539726e-05,
      "loss": 0.598,
      "step": 5910,
      "task_loss": 0.6873341202735901
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46607810258865356,
      "epoch": 2.14,
      "learning_rate": 3.676383423015442e-05,
      "loss": 0.5983,
      "step": 5920,
      "task_loss": 0.8500930666923523
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46026360988616943,
      "epoch": 2.14,
      "learning_rate": 3.6697448144165357e-05,
      "loss": 0.5742,
      "step": 5930,
      "task_loss": 0.8118923902511597
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6993731260299683,
      "epoch": 2.15,
      "learning_rate": 3.663102752388155e-05,
      "loss": 0.6208,
      "step": 5940,
      "task_loss": 0.3766775131225586
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5157681703567505,
      "epoch": 2.15,
      "learning_rate": 3.656457271179003e-05,
      "loss": 0.5637,
      "step": 5950,
      "task_loss": 0.42299726605415344
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7117334604263306,
      "epoch": 2.15,
      "learning_rate": 3.6498084050554164e-05,
      "loss": 0.691,
      "step": 5960,
      "task_loss": 0.8287671804428101
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41603410243988037,
      "epoch": 2.16,
      "learning_rate": 3.643156188301183e-05,
      "loss": 0.5911,
      "step": 5970,
      "task_loss": 0.4762575626373291
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5428121089935303,
      "epoch": 2.16,
      "learning_rate": 3.636500655217371e-05,
      "loss": 0.5762,
      "step": 5980,
      "task_loss": 0.6505820751190186
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3971426486968994,
      "epoch": 2.16,
      "learning_rate": 3.629841840122147e-05,
      "loss": 0.6486,
      "step": 5990,
      "task_loss": 0.5191299915313721
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7955690622329712,
      "epoch": 2.17,
      "learning_rate": 3.6231797773505994e-05,
      "loss": 0.6275,
      "step": 6000,
      "task_loss": 0.656362771987915
    },
    {
      "epoch": 2.17,
      "eval_exact_match": 81.8070009460738,
      "eval_f1": 89.1233677854583,
      "step": 6000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6281618475914001,
      "epoch": 2.17,
      "learning_rate": 3.616514501254567e-05,
      "loss": 0.5901,
      "step": 6010,
      "task_loss": 0.6404191255569458
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5767370462417603,
      "epoch": 2.18,
      "learning_rate": 3.6098460462024506e-05,
      "loss": 0.6066,
      "step": 6020,
      "task_loss": 0.506596565246582
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7444403171539307,
      "epoch": 2.18,
      "learning_rate": 3.6031744465790495e-05,
      "loss": 0.6459,
      "step": 6030,
      "task_loss": 0.8166441917419434
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6288504600524902,
      "epoch": 2.18,
      "learning_rate": 3.5964997367853755e-05,
      "loss": 0.5967,
      "step": 6040,
      "task_loss": 0.7653200030326843
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6448060274124146,
      "epoch": 2.19,
      "learning_rate": 3.589821951238474e-05,
      "loss": 0.5961,
      "step": 6050,
      "task_loss": 1.2137736082077026
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6648958921432495,
      "epoch": 2.19,
      "learning_rate": 3.5831411243712555e-05,
      "loss": 0.5949,
      "step": 6060,
      "task_loss": 0.5861155986785889
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6277318596839905,
      "epoch": 2.19,
      "learning_rate": 3.5764572906323075e-05,
      "loss": 0.5954,
      "step": 6070,
      "task_loss": 0.6117310523986816
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5213579535484314,
      "epoch": 2.2,
      "learning_rate": 3.569770484485726e-05,
      "loss": 0.5722,
      "step": 6080,
      "task_loss": 0.9827824831008911
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4223155379295349,
      "epoch": 2.2,
      "learning_rate": 3.563080740410932e-05,
      "loss": 0.6018,
      "step": 6090,
      "task_loss": 0.6889990568161011
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.562755286693573,
      "epoch": 2.2,
      "learning_rate": 3.556388092902494e-05,
      "loss": 0.6329,
      "step": 6100,
      "task_loss": 0.5587911605834961
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5433652997016907,
      "epoch": 2.21,
      "learning_rate": 3.549692576469955e-05,
      "loss": 0.602,
      "step": 6110,
      "task_loss": 0.4544098377227783
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4736385941505432,
      "epoch": 2.21,
      "learning_rate": 3.542994225637648e-05,
      "loss": 0.58,
      "step": 6120,
      "task_loss": 0.4461681544780731
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6638973951339722,
      "epoch": 2.22,
      "learning_rate": 3.536293074944522e-05,
      "loss": 0.6859,
      "step": 6130,
      "task_loss": 0.5688951015472412
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7005724906921387,
      "epoch": 2.22,
      "learning_rate": 3.529589158943965e-05,
      "loss": 0.625,
      "step": 6140,
      "task_loss": 0.5859721899032593
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6913973093032837,
      "epoch": 2.22,
      "learning_rate": 3.522882512203621e-05,
      "loss": 0.6089,
      "step": 6150,
      "task_loss": 1.2206425666809082
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7091802358627319,
      "epoch": 2.23,
      "learning_rate": 3.516173169305216e-05,
      "loss": 0.6443,
      "step": 6160,
      "task_loss": 0.4974295496940613
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7010023593902588,
      "epoch": 2.23,
      "learning_rate": 3.5094611648443773e-05,
      "loss": 0.654,
      "step": 6170,
      "task_loss": 0.5480866432189941
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6212794780731201,
      "epoch": 2.23,
      "learning_rate": 3.5027465334304586e-05,
      "loss": 0.6342,
      "step": 6180,
      "task_loss": 0.7250664830207825
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5260363221168518,
      "epoch": 2.24,
      "learning_rate": 3.496029309686356e-05,
      "loss": 0.5235,
      "step": 6190,
      "task_loss": 0.6064286231994629
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.546513020992279,
      "epoch": 2.24,
      "learning_rate": 3.4893095282483346e-05,
      "loss": 0.5609,
      "step": 6200,
      "task_loss": 0.3039868474006653
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5084608793258667,
      "epoch": 2.24,
      "learning_rate": 3.482587223765847e-05,
      "loss": 0.6002,
      "step": 6210,
      "task_loss": 0.36580145359039307
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7696444392204285,
      "epoch": 2.25,
      "learning_rate": 3.475862430901355e-05,
      "loss": 0.6239,
      "step": 6220,
      "task_loss": 1.048229694366455
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38577717542648315,
      "epoch": 2.25,
      "learning_rate": 3.469135184330153e-05,
      "loss": 0.5504,
      "step": 6230,
      "task_loss": 0.4784127473831177
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7590066194534302,
      "epoch": 2.26,
      "learning_rate": 3.462405518740186e-05,
      "loss": 0.6185,
      "step": 6240,
      "task_loss": 0.6026236414909363
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7041105031967163,
      "epoch": 2.26,
      "learning_rate": 3.455673468831872e-05,
      "loss": 0.6119,
      "step": 6250,
      "task_loss": 0.9708863496780396
    },
    {
      "epoch": 2.26,
      "eval_exact_match": 81.5042573320719,
      "eval_f1": 88.78182464091005,
      "step": 6250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.612851619720459,
      "epoch": 2.26,
      "learning_rate": 3.448939069317926e-05,
      "loss": 0.5884,
      "step": 6260,
      "task_loss": 1.315657615661621
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5629523992538452,
      "epoch": 2.27,
      "learning_rate": 3.442202354923175e-05,
      "loss": 0.5514,
      "step": 6270,
      "task_loss": 0.5583246946334839
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7930963039398193,
      "epoch": 2.27,
      "learning_rate": 3.435463360384383e-05,
      "loss": 0.6183,
      "step": 6280,
      "task_loss": 0.959362268447876
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8136299848556519,
      "epoch": 2.27,
      "learning_rate": 3.428722120450074e-05,
      "loss": 0.644,
      "step": 6290,
      "task_loss": 0.8650078773498535
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6255238056182861,
      "epoch": 2.28,
      "learning_rate": 3.421978669880346e-05,
      "loss": 0.594,
      "step": 6300,
      "task_loss": 1.1138728857040405
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.700169026851654,
      "epoch": 2.28,
      "learning_rate": 3.4152330434466994e-05,
      "loss": 0.5617,
      "step": 6310,
      "task_loss": 0.5409881472587585
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5610506534576416,
      "epoch": 2.28,
      "learning_rate": 3.408485275931851e-05,
      "loss": 0.6282,
      "step": 6320,
      "task_loss": 0.4196653366088867
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.504428505897522,
      "epoch": 2.29,
      "learning_rate": 3.40173540212956e-05,
      "loss": 0.5916,
      "step": 6330,
      "task_loss": 0.4185686707496643
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5590067505836487,
      "epoch": 2.29,
      "learning_rate": 3.394983456844446e-05,
      "loss": 0.6516,
      "step": 6340,
      "task_loss": 0.6237858533859253
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6239622235298157,
      "epoch": 2.29,
      "learning_rate": 3.3882294748918074e-05,
      "loss": 0.6038,
      "step": 6350,
      "task_loss": 0.54815673828125
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6079332828521729,
      "epoch": 2.3,
      "learning_rate": 3.381473491097448e-05,
      "loss": 0.5802,
      "step": 6360,
      "task_loss": 0.738688051700592
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6828926801681519,
      "epoch": 2.3,
      "learning_rate": 3.374715540297492e-05,
      "loss": 0.5867,
      "step": 6370,
      "task_loss": 0.9791734218597412
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5750732421875,
      "epoch": 2.31,
      "learning_rate": 3.367955657338205e-05,
      "loss": 0.5862,
      "step": 6380,
      "task_loss": 0.6127207279205322
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5833463668823242,
      "epoch": 2.31,
      "learning_rate": 3.3611938770758186e-05,
      "loss": 0.6614,
      "step": 6390,
      "task_loss": 0.5630807876586914
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6061692237854004,
      "epoch": 2.31,
      "learning_rate": 3.354430234376344e-05,
      "loss": 0.5972,
      "step": 6400,
      "task_loss": 0.6525106430053711
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6266899704933167,
      "epoch": 2.32,
      "learning_rate": 3.347664764115398e-05,
      "loss": 0.537,
      "step": 6410,
      "task_loss": 0.9735203385353088
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7700897455215454,
      "epoch": 2.32,
      "learning_rate": 3.340897501178019e-05,
      "loss": 0.6051,
      "step": 6420,
      "task_loss": 0.7603612542152405
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.541512131690979,
      "epoch": 2.32,
      "learning_rate": 3.334128480458492e-05,
      "loss": 0.6205,
      "step": 6430,
      "task_loss": 0.633751392364502
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47812074422836304,
      "epoch": 2.33,
      "learning_rate": 3.3273577368601644e-05,
      "loss": 0.6212,
      "step": 6440,
      "task_loss": 0.5502046346664429
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.574849545955658,
      "epoch": 2.33,
      "learning_rate": 3.320585305295264e-05,
      "loss": 0.5781,
      "step": 6450,
      "task_loss": 0.6506343483924866
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7085522413253784,
      "epoch": 2.33,
      "learning_rate": 3.3138112206847296e-05,
      "loss": 0.5527,
      "step": 6460,
      "task_loss": 0.9275124073028564
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44274505972862244,
      "epoch": 2.34,
      "learning_rate": 3.307035517958015e-05,
      "loss": 0.5773,
      "step": 6470,
      "task_loss": 0.32708221673965454
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3988785147666931,
      "epoch": 2.34,
      "learning_rate": 3.3002582320529247e-05,
      "loss": 0.6215,
      "step": 6480,
      "task_loss": 0.6397969126701355
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5385051965713501,
      "epoch": 2.35,
      "learning_rate": 3.293479397915423e-05,
      "loss": 0.5694,
      "step": 6490,
      "task_loss": 0.43270617723464966
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5102781653404236,
      "epoch": 2.35,
      "learning_rate": 3.286699050499458e-05,
      "loss": 0.5186,
      "step": 6500,
      "task_loss": 0.4291302561759949
    },
    {
      "epoch": 2.35,
      "eval_exact_match": 81.9678334910123,
      "eval_f1": 89.18938641074499,
      "step": 6500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5587887763977051,
      "epoch": 2.35,
      "learning_rate": 3.279917224766781e-05,
      "loss": 0.5673,
      "step": 6510,
      "task_loss": 0.38369613885879517
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4702199697494507,
      "epoch": 2.36,
      "learning_rate": 3.273133955686768e-05,
      "loss": 0.6397,
      "step": 6520,
      "task_loss": 0.7154326438903809
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5072304010391235,
      "epoch": 2.36,
      "learning_rate": 3.2663492782362315e-05,
      "loss": 0.597,
      "step": 6530,
      "task_loss": 0.7844618558883667
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45684850215911865,
      "epoch": 2.36,
      "learning_rate": 3.259563227399253e-05,
      "loss": 0.5534,
      "step": 6540,
      "task_loss": 0.30611854791641235
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8612761497497559,
      "epoch": 2.37,
      "learning_rate": 3.252775838166991e-05,
      "loss": 0.6271,
      "step": 6550,
      "task_loss": 0.923941969871521
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5912673473358154,
      "epoch": 2.37,
      "learning_rate": 3.2459871455375075e-05,
      "loss": 0.5532,
      "step": 6560,
      "task_loss": 0.7655602097511292
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6388885974884033,
      "epoch": 2.37,
      "learning_rate": 3.239197184515584e-05,
      "loss": 0.5633,
      "step": 6570,
      "task_loss": 1.067481517791748
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.51146000623703,
      "epoch": 2.38,
      "learning_rate": 3.232405990112543e-05,
      "loss": 0.5882,
      "step": 6580,
      "task_loss": 0.6557249426841736
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5170506238937378,
      "epoch": 2.38,
      "learning_rate": 3.225613597346068e-05,
      "loss": 0.5887,
      "step": 6590,
      "task_loss": 0.25539761781692505
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4798586368560791,
      "epoch": 2.39,
      "learning_rate": 3.218820041240017e-05,
      "loss": 0.6124,
      "step": 6600,
      "task_loss": 0.6159725785255432
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5324605703353882,
      "epoch": 2.39,
      "learning_rate": 3.212025356824251e-05,
      "loss": 0.5845,
      "step": 6610,
      "task_loss": 0.5922216176986694
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40312764048576355,
      "epoch": 2.39,
      "learning_rate": 3.20522957913445e-05,
      "loss": 0.5713,
      "step": 6620,
      "task_loss": 0.4075431227684021
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5648410320281982,
      "epoch": 2.4,
      "learning_rate": 3.198432743211925e-05,
      "loss": 0.6679,
      "step": 6630,
      "task_loss": 0.5671501755714417
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4727499186992645,
      "epoch": 2.4,
      "learning_rate": 3.191634884103451e-05,
      "loss": 0.631,
      "step": 6640,
      "task_loss": 0.527313768863678
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5744331479072571,
      "epoch": 2.4,
      "learning_rate": 3.184836036861074e-05,
      "loss": 0.566,
      "step": 6650,
      "task_loss": 0.8066692352294922
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5417715311050415,
      "epoch": 2.41,
      "learning_rate": 3.178036236541936e-05,
      "loss": 0.5379,
      "step": 6660,
      "task_loss": 0.3417711555957794
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4441830515861511,
      "epoch": 2.41,
      "learning_rate": 3.171235518208096e-05,
      "loss": 0.5847,
      "step": 6670,
      "task_loss": 0.5670948028564453
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6484647393226624,
      "epoch": 2.41,
      "learning_rate": 3.164433916926342e-05,
      "loss": 0.6377,
      "step": 6680,
      "task_loss": 0.8303718566894531
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4492584764957428,
      "epoch": 2.42,
      "learning_rate": 3.15763146776802e-05,
      "loss": 0.5715,
      "step": 6690,
      "task_loss": 0.5020931959152222
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7698655128479004,
      "epoch": 2.42,
      "learning_rate": 3.1508282058088424e-05,
      "loss": 0.6297,
      "step": 6700,
      "task_loss": 0.8664661645889282
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5374277234077454,
      "epoch": 2.43,
      "learning_rate": 3.144024166128718e-05,
      "loss": 0.5943,
      "step": 6710,
      "task_loss": 0.8371603488922119
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44918084144592285,
      "epoch": 2.43,
      "learning_rate": 3.137219383811562e-05,
      "loss": 0.5709,
      "step": 6720,
      "task_loss": 0.6834157109260559
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5904335975646973,
      "epoch": 2.43,
      "learning_rate": 3.130413893945121e-05,
      "loss": 0.5909,
      "step": 6730,
      "task_loss": 0.6250230073928833
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5855615139007568,
      "epoch": 2.44,
      "learning_rate": 3.12360773162079e-05,
      "loss": 0.5643,
      "step": 6740,
      "task_loss": 0.7704566121101379
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5405184626579285,
      "epoch": 2.44,
      "learning_rate": 3.1168009319334276e-05,
      "loss": 0.5897,
      "step": 6750,
      "task_loss": 0.7256063222885132
    },
    {
      "epoch": 2.44,
      "eval_exact_match": 81.759697256386,
      "eval_f1": 89.06270353602996,
      "step": 6750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8909243941307068,
      "epoch": 2.44,
      "learning_rate": 3.1099935299811844e-05,
      "loss": 0.7078,
      "step": 6760,
      "task_loss": 0.8761190176010132
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5284467935562134,
      "epoch": 2.45,
      "learning_rate": 3.103185560865314e-05,
      "loss": 0.5888,
      "step": 6770,
      "task_loss": 0.5950831770896912
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6189389228820801,
      "epoch": 2.45,
      "learning_rate": 3.096377059689994e-05,
      "loss": 0.5407,
      "step": 6780,
      "task_loss": 0.8411575555801392
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5810912847518921,
      "epoch": 2.45,
      "learning_rate": 3.089568061562146e-05,
      "loss": 0.6317,
      "step": 6790,
      "task_loss": 0.706413745880127
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7708436250686646,
      "epoch": 2.46,
      "learning_rate": 3.082758601591253e-05,
      "loss": 0.6281,
      "step": 6800,
      "task_loss": 0.8470696806907654
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7277690172195435,
      "epoch": 2.46,
      "learning_rate": 3.075948714889181e-05,
      "loss": 0.647,
      "step": 6810,
      "task_loss": 0.6179013848304749
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43784040212631226,
      "epoch": 2.46,
      "learning_rate": 3.069138436569997e-05,
      "loss": 0.6279,
      "step": 6820,
      "task_loss": 0.5682166814804077
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6058757305145264,
      "epoch": 2.47,
      "learning_rate": 3.062327801749784e-05,
      "loss": 0.5572,
      "step": 6830,
      "task_loss": 0.6574521064758301
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45435667037963867,
      "epoch": 2.47,
      "learning_rate": 3.055516845546468e-05,
      "loss": 0.5397,
      "step": 6840,
      "task_loss": 0.5073586106300354
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5500397682189941,
      "epoch": 2.48,
      "learning_rate": 3.0487056030796262e-05,
      "loss": 0.586,
      "step": 6850,
      "task_loss": 0.6582145094871521
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46652621030807495,
      "epoch": 2.48,
      "learning_rate": 3.0418941094703176e-05,
      "loss": 0.5826,
      "step": 6860,
      "task_loss": 0.367644339799881
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.583972692489624,
      "epoch": 2.48,
      "learning_rate": 3.0350823998408926e-05,
      "loss": 0.5876,
      "step": 6870,
      "task_loss": 1.2166520357131958
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.73996502161026,
      "epoch": 2.49,
      "learning_rate": 3.028270509314817e-05,
      "loss": 0.6818,
      "step": 6880,
      "task_loss": 0.9112734198570251
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38398513197898865,
      "epoch": 2.49,
      "learning_rate": 3.0214584730164898e-05,
      "loss": 0.5749,
      "step": 6890,
      "task_loss": 0.844524085521698
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40584832429885864,
      "epoch": 2.49,
      "learning_rate": 3.014646326071059e-05,
      "loss": 0.5586,
      "step": 6900,
      "task_loss": 0.5562839508056641
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5574773550033569,
      "epoch": 2.5,
      "learning_rate": 3.0078341036042457e-05,
      "loss": 0.644,
      "step": 6910,
      "task_loss": 0.594291090965271
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5767573118209839,
      "epoch": 2.5,
      "learning_rate": 3.0010218407421613e-05,
      "loss": 0.5991,
      "step": 6920,
      "task_loss": 0.8964383006095886
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5444501638412476,
      "epoch": 2.5,
      "learning_rate": 2.9942095726111204e-05,
      "loss": 0.5118,
      "step": 6930,
      "task_loss": 0.5208333730697632
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6177530288696289,
      "epoch": 2.51,
      "learning_rate": 2.9873973343374728e-05,
      "loss": 0.6443,
      "step": 6940,
      "task_loss": 0.48882049322128296
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5032823085784912,
      "epoch": 2.51,
      "learning_rate": 2.9805851610474044e-05,
      "loss": 0.5943,
      "step": 6950,
      "task_loss": 0.5606658458709717
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6921489238739014,
      "epoch": 2.52,
      "learning_rate": 2.9737730878667765e-05,
      "loss": 0.6433,
      "step": 6960,
      "task_loss": 0.5478000044822693
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40710246562957764,
      "epoch": 2.52,
      "learning_rate": 2.9669611499209254e-05,
      "loss": 0.644,
      "step": 6970,
      "task_loss": 0.2990033030509949
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5349779725074768,
      "epoch": 2.52,
      "learning_rate": 2.960149382334496e-05,
      "loss": 0.5927,
      "step": 6980,
      "task_loss": 0.5776247978210449
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6616565585136414,
      "epoch": 2.53,
      "learning_rate": 2.9533378202312512e-05,
      "loss": 0.5778,
      "step": 6990,
      "task_loss": 0.7887768745422363
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8387987613677979,
      "epoch": 2.53,
      "learning_rate": 2.9465264987338966e-05,
      "loss": 0.6088,
      "step": 7000,
      "task_loss": 0.5467420220375061
    },
    {
      "epoch": 2.53,
      "eval_exact_match": 82.07190160832545,
      "eval_f1": 89.27410906455074,
      "step": 7000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47956252098083496,
      "epoch": 2.53,
      "learning_rate": 2.939715452963896e-05,
      "loss": 0.5898,
      "step": 7010,
      "task_loss": 0.6686305999755859
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4508477747440338,
      "epoch": 2.54,
      "learning_rate": 2.9329047180412914e-05,
      "loss": 0.5906,
      "step": 7020,
      "task_loss": 0.5521897077560425
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47773054242134094,
      "epoch": 2.54,
      "learning_rate": 2.9260943290845216e-05,
      "loss": 0.5783,
      "step": 7030,
      "task_loss": 0.7009822130203247
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5153251886367798,
      "epoch": 2.54,
      "learning_rate": 2.919284321210245e-05,
      "loss": 0.5826,
      "step": 7040,
      "task_loss": 0.7742196321487427
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5433927178382874,
      "epoch": 2.55,
      "learning_rate": 2.9124747295331482e-05,
      "loss": 0.5769,
      "step": 7050,
      "task_loss": 0.5901714563369751
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7135424613952637,
      "epoch": 2.55,
      "learning_rate": 2.9056655891657793e-05,
      "loss": 0.6148,
      "step": 7060,
      "task_loss": 0.9341310262680054
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7931948304176331,
      "epoch": 2.56,
      "learning_rate": 2.8988569352183534e-05,
      "loss": 0.6634,
      "step": 7070,
      "task_loss": 1.1662893295288086
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30930817127227783,
      "epoch": 2.56,
      "learning_rate": 2.8920488027985812e-05,
      "loss": 0.5629,
      "step": 7080,
      "task_loss": 0.2516777217388153
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6308006644248962,
      "epoch": 2.56,
      "learning_rate": 2.8852412270114817e-05,
      "loss": 0.6637,
      "step": 7090,
      "task_loss": 0.7488775253295898
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8185533285140991,
      "epoch": 2.57,
      "learning_rate": 2.8784342429592058e-05,
      "loss": 0.6215,
      "step": 7100,
      "task_loss": 0.9923599362373352
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7322626113891602,
      "epoch": 2.57,
      "learning_rate": 2.871627885740851e-05,
      "loss": 0.5773,
      "step": 7110,
      "task_loss": 0.755662739276886
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5336388349533081,
      "epoch": 2.57,
      "learning_rate": 2.8648221904522858e-05,
      "loss": 0.574,
      "step": 7120,
      "task_loss": 0.44107335805892944
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5362550020217896,
      "epoch": 2.58,
      "learning_rate": 2.8580171921859606e-05,
      "loss": 0.6477,
      "step": 7130,
      "task_loss": 0.5944764614105225
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5030221939086914,
      "epoch": 2.58,
      "learning_rate": 2.851212926030738e-05,
      "loss": 0.6169,
      "step": 7140,
      "task_loss": 0.8135521411895752
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.742468535900116,
      "epoch": 2.58,
      "learning_rate": 2.8450897414439317e-05,
      "loss": 0.623,
      "step": 7150,
      "task_loss": 0.9540405869483948
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7146216630935669,
      "epoch": 2.59,
      "learning_rate": 2.838286962955943e-05,
      "loss": 0.6073,
      "step": 7160,
      "task_loss": 0.8092219233512878
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5073987245559692,
      "epoch": 2.59,
      "learning_rate": 2.8314850183147516e-05,
      "loss": 0.5979,
      "step": 7170,
      "task_loss": 0.6714034080505371
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5770305395126343,
      "epoch": 2.59,
      "learning_rate": 2.8246839425934724e-05,
      "loss": 0.4854,
      "step": 7180,
      "task_loss": 0.33490344882011414
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5794994831085205,
      "epoch": 2.6,
      "learning_rate": 2.817883770860737e-05,
      "loss": 0.6159,
      "step": 7190,
      "task_loss": 0.6019021272659302
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5001416206359863,
      "epoch": 2.6,
      "learning_rate": 2.81108453818052e-05,
      "loss": 0.5719,
      "step": 7200,
      "task_loss": 0.525328516960144
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6738238334655762,
      "epoch": 2.61,
      "learning_rate": 2.8042862796119482e-05,
      "loss": 0.6662,
      "step": 7210,
      "task_loss": 0.6700210571289062
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.48849350214004517,
      "epoch": 2.61,
      "learning_rate": 2.7974890302091327e-05,
      "loss": 0.5327,
      "step": 7220,
      "task_loss": 0.6737467050552368
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.593249499797821,
      "epoch": 2.61,
      "learning_rate": 2.7906928250209743e-05,
      "loss": 0.6226,
      "step": 7230,
      "task_loss": 0.7583106160163879
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6394388675689697,
      "epoch": 2.62,
      "learning_rate": 2.783897699090994e-05,
      "loss": 0.5993,
      "step": 7240,
      "task_loss": 0.9611802101135254
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.814205527305603,
      "epoch": 2.62,
      "learning_rate": 2.7771036874571443e-05,
      "loss": 0.5924,
      "step": 7250,
      "task_loss": 0.9047456979751587
    },
    {
      "epoch": 2.62,
      "eval_exact_match": 81.57048249763481,
      "eval_f1": 88.98593489499765,
      "step": 7250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5149030685424805,
      "epoch": 2.62,
      "learning_rate": 2.770310825151635e-05,
      "loss": 0.6186,
      "step": 7260,
      "task_loss": 0.5792198777198792
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6406042575836182,
      "epoch": 2.63,
      "learning_rate": 2.763519147200748e-05,
      "loss": 0.6658,
      "step": 7270,
      "task_loss": 0.5885680913925171
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.702430009841919,
      "epoch": 2.63,
      "learning_rate": 2.7567286886246593e-05,
      "loss": 0.5986,
      "step": 7280,
      "task_loss": 0.6880907416343689
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5257836580276489,
      "epoch": 2.63,
      "learning_rate": 2.749939484437255e-05,
      "loss": 0.5546,
      "step": 7290,
      "task_loss": 0.5597729086875916
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.583315372467041,
      "epoch": 2.64,
      "learning_rate": 2.7431515696459577e-05,
      "loss": 0.5803,
      "step": 7300,
      "task_loss": 0.6037495732307434
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5714931488037109,
      "epoch": 2.64,
      "learning_rate": 2.736364979251535e-05,
      "loss": 0.5847,
      "step": 7310,
      "task_loss": 0.6359392404556274
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5830150842666626,
      "epoch": 2.65,
      "learning_rate": 2.7295797482479327e-05,
      "loss": 0.5466,
      "step": 7320,
      "task_loss": 0.561113715171814
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7011891603469849,
      "epoch": 2.65,
      "learning_rate": 2.7227959116220803e-05,
      "loss": 0.674,
      "step": 7330,
      "task_loss": 1.1329269409179688
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.028478980064392,
      "epoch": 2.65,
      "learning_rate": 2.7160135043537236e-05,
      "loss": 0.6032,
      "step": 7340,
      "task_loss": 0.9812748432159424
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8232661485671997,
      "epoch": 2.66,
      "learning_rate": 2.7092325614152328e-05,
      "loss": 0.5801,
      "step": 7350,
      "task_loss": 0.5196477770805359
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47645464539527893,
      "epoch": 2.66,
      "learning_rate": 2.7024531177714316e-05,
      "loss": 0.5197,
      "step": 7360,
      "task_loss": 0.3743845224380493
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5509925484657288,
      "epoch": 2.66,
      "learning_rate": 2.6956752083794094e-05,
      "loss": 0.6336,
      "step": 7370,
      "task_loss": 0.6567373871803284
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6547894477844238,
      "epoch": 2.67,
      "learning_rate": 2.688898868188348e-05,
      "loss": 0.632,
      "step": 7380,
      "task_loss": 0.632176399230957
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.624300479888916,
      "epoch": 2.67,
      "learning_rate": 2.682124132139334e-05,
      "loss": 0.5946,
      "step": 7390,
      "task_loss": 0.5805683135986328
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5678684115409851,
      "epoch": 2.67,
      "learning_rate": 2.675351035165188e-05,
      "loss": 0.5989,
      "step": 7400,
      "task_loss": 0.570992112159729
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8260520696640015,
      "epoch": 2.68,
      "learning_rate": 2.668579612190271e-05,
      "loss": 0.5988,
      "step": 7410,
      "task_loss": 0.6632480621337891
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5764724612236023,
      "epoch": 2.68,
      "learning_rate": 2.6618098981303204e-05,
      "loss": 0.6097,
      "step": 7420,
      "task_loss": 0.6332680583000183
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5330133438110352,
      "epoch": 2.69,
      "learning_rate": 2.655041927892257e-05,
      "loss": 0.5412,
      "step": 7430,
      "task_loss": 0.6299136281013489
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5107343196868896,
      "epoch": 2.69,
      "learning_rate": 2.6482757363740117e-05,
      "loss": 0.4791,
      "step": 7440,
      "task_loss": 0.7189974188804626
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.48826032876968384,
      "epoch": 2.69,
      "learning_rate": 2.6415113584643424e-05,
      "loss": 0.562,
      "step": 7450,
      "task_loss": 0.5355274677276611
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 1.131540060043335,
      "epoch": 2.7,
      "learning_rate": 2.6347488290426573e-05,
      "loss": 0.6699,
      "step": 7460,
      "task_loss": 0.9076927900314331
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5723243355751038,
      "epoch": 2.7,
      "learning_rate": 2.627988182978831e-05,
      "loss": 0.5461,
      "step": 7470,
      "task_loss": 0.6198363304138184
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4197266101837158,
      "epoch": 2.7,
      "learning_rate": 2.6212294551330293e-05,
      "loss": 0.5525,
      "step": 7480,
      "task_loss": 0.34966617822647095
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46035024523735046,
      "epoch": 2.71,
      "learning_rate": 2.6144726803555232e-05,
      "loss": 0.6712,
      "step": 7490,
      "task_loss": 0.39124152064323425
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46283459663391113,
      "epoch": 2.71,
      "learning_rate": 2.6077178934865193e-05,
      "loss": 0.5959,
      "step": 7500,
      "task_loss": 0.8845962882041931
    },
    {
      "epoch": 2.71,
      "eval_exact_match": 82.52601702932829,
      "eval_f1": 89.60594328334614,
      "step": 7500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5695556402206421,
      "epoch": 2.71,
      "learning_rate": 2.6009651293559663e-05,
      "loss": 0.5596,
      "step": 7510,
      "task_loss": 0.4987695515155792
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6206524968147278,
      "epoch": 2.72,
      "learning_rate": 2.59421442278339e-05,
      "loss": 0.6555,
      "step": 7520,
      "task_loss": 0.6371747255325317
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.431548535823822,
      "epoch": 2.72,
      "learning_rate": 2.5874658085777014e-05,
      "loss": 0.535,
      "step": 7530,
      "task_loss": 0.4861009418964386
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4274836778640747,
      "epoch": 2.72,
      "learning_rate": 2.580719321537026e-05,
      "loss": 0.5952,
      "step": 7540,
      "task_loss": 0.7763680219650269
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43078041076660156,
      "epoch": 2.73,
      "learning_rate": 2.5739749964485183e-05,
      "loss": 0.6147,
      "step": 7550,
      "task_loss": 0.37037426233291626
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5834618806838989,
      "epoch": 2.73,
      "learning_rate": 2.5672328680881876e-05,
      "loss": 0.6496,
      "step": 7560,
      "task_loss": 0.8014817833900452
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8053905963897705,
      "epoch": 2.74,
      "learning_rate": 2.5604929712207137e-05,
      "loss": 0.6696,
      "step": 7570,
      "task_loss": 0.8470737934112549
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7601429224014282,
      "epoch": 2.74,
      "learning_rate": 2.5537553405992723e-05,
      "loss": 0.5971,
      "step": 7580,
      "task_loss": 1.2324957847595215
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46784117817878723,
      "epoch": 2.74,
      "learning_rate": 2.547020010965351e-05,
      "loss": 0.558,
      "step": 7590,
      "task_loss": 0.6156980991363525
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8725177049636841,
      "epoch": 2.75,
      "learning_rate": 2.5402870170485775e-05,
      "loss": 0.6351,
      "step": 7600,
      "task_loss": 0.8051373958587646
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5979294776916504,
      "epoch": 2.75,
      "learning_rate": 2.533556393566528e-05,
      "loss": 0.583,
      "step": 7610,
      "task_loss": 0.5045502781867981
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.422203004360199,
      "epoch": 2.75,
      "learning_rate": 2.5268281752245642e-05,
      "loss": 0.5823,
      "step": 7620,
      "task_loss": 0.6445773243904114
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5774694085121155,
      "epoch": 2.76,
      "learning_rate": 2.520102396715641e-05,
      "loss": 0.6053,
      "step": 7630,
      "task_loss": 0.4244164824485779
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4557744264602661,
      "epoch": 2.76,
      "learning_rate": 2.513379092720134e-05,
      "loss": 0.566,
      "step": 7640,
      "task_loss": 0.6151514053344727
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5988390445709229,
      "epoch": 2.76,
      "learning_rate": 2.5066582979056587e-05,
      "loss": 0.5673,
      "step": 7650,
      "task_loss": 1.005041480064392
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8742297887802124,
      "epoch": 2.77,
      "learning_rate": 2.4999400469268948e-05,
      "loss": 0.6041,
      "step": 7660,
      "task_loss": 0.8448403477668762
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42271387577056885,
      "epoch": 2.77,
      "learning_rate": 2.493224374425402e-05,
      "loss": 0.5586,
      "step": 7670,
      "task_loss": 0.44839856028556824
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3397362232208252,
      "epoch": 2.78,
      "learning_rate": 2.486511315029447e-05,
      "loss": 0.5687,
      "step": 7680,
      "task_loss": 0.346187949180603
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3857961893081665,
      "epoch": 2.78,
      "learning_rate": 2.47980090335382e-05,
      "loss": 0.5529,
      "step": 7690,
      "task_loss": 0.4324136972427368
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5957242250442505,
      "epoch": 2.78,
      "learning_rate": 2.4730931739996625e-05,
      "loss": 0.6139,
      "step": 7700,
      "task_loss": 0.9946978092193604
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7779924869537354,
      "epoch": 2.79,
      "learning_rate": 2.4663881615542794e-05,
      "loss": 0.6516,
      "step": 7710,
      "task_loss": 0.8503498435020447
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4511723816394806,
      "epoch": 2.79,
      "learning_rate": 2.4596859005909728e-05,
      "loss": 0.6176,
      "step": 7720,
      "task_loss": 0.6886883974075317
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49279749393463135,
      "epoch": 2.79,
      "learning_rate": 2.4529864256688515e-05,
      "loss": 0.5525,
      "step": 7730,
      "task_loss": 0.6275307536125183
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4928208589553833,
      "epoch": 2.8,
      "learning_rate": 2.4462897713326633e-05,
      "loss": 0.5676,
      "step": 7740,
      "task_loss": 0.784211277961731
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6046674847602844,
      "epoch": 2.8,
      "learning_rate": 2.4395959721126073e-05,
      "loss": 0.5396,
      "step": 7750,
      "task_loss": 0.780253529548645
    },
    {
      "epoch": 2.8,
      "eval_exact_match": 82.2705771050142,
      "eval_f1": 89.53313416894864,
      "step": 7750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.83915114402771,
      "epoch": 2.8,
      "learning_rate": 2.432905062524165e-05,
      "loss": 0.5805,
      "step": 7760,
      "task_loss": 1.0383943319320679
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5021464228630066,
      "epoch": 2.81,
      "learning_rate": 2.426217077067916e-05,
      "loss": 0.544,
      "step": 7770,
      "task_loss": 0.6696529388427734
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6492622494697571,
      "epoch": 2.81,
      "learning_rate": 2.419532050229361e-05,
      "loss": 0.6244,
      "step": 7780,
      "task_loss": 0.7537462115287781
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5134307146072388,
      "epoch": 2.82,
      "learning_rate": 2.412850016478747e-05,
      "loss": 0.6005,
      "step": 7790,
      "task_loss": 0.8974807262420654
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4806973338127136,
      "epoch": 2.82,
      "learning_rate": 2.4061710102708885e-05,
      "loss": 0.5571,
      "step": 7800,
      "task_loss": 0.3275541663169861
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.535984218120575,
      "epoch": 2.82,
      "learning_rate": 2.3994950660449844e-05,
      "loss": 0.542,
      "step": 7810,
      "task_loss": 0.9069480895996094
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5129786133766174,
      "epoch": 2.83,
      "learning_rate": 2.3928222182244508e-05,
      "loss": 0.5683,
      "step": 7820,
      "task_loss": 0.3967326581478119
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5826820135116577,
      "epoch": 2.83,
      "learning_rate": 2.3861525012167334e-05,
      "loss": 0.581,
      "step": 7830,
      "task_loss": 0.7916654348373413
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8494963645935059,
      "epoch": 2.83,
      "learning_rate": 2.379485949413137e-05,
      "loss": 0.6048,
      "step": 7840,
      "task_loss": 1.0144283771514893
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5546177625656128,
      "epoch": 2.84,
      "learning_rate": 2.3728225971886433e-05,
      "loss": 0.581,
      "step": 7850,
      "task_loss": 0.45552924275398254
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.581791877746582,
      "epoch": 2.84,
      "learning_rate": 2.366162478901738e-05,
      "loss": 0.5915,
      "step": 7860,
      "task_loss": 0.5242971777915955
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4776926040649414,
      "epoch": 2.84,
      "learning_rate": 2.359505628894229e-05,
      "loss": 0.5925,
      "step": 7870,
      "task_loss": 0.7510772943496704
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4774831235408783,
      "epoch": 2.85,
      "learning_rate": 2.3528520814910756e-05,
      "loss": 0.612,
      "step": 7880,
      "task_loss": 0.41770559549331665
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5438613891601562,
      "epoch": 2.85,
      "learning_rate": 2.346201871000203e-05,
      "loss": 0.5752,
      "step": 7890,
      "task_loss": 0.6803189516067505
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6210540533065796,
      "epoch": 2.86,
      "learning_rate": 2.3395550317123355e-05,
      "loss": 0.5516,
      "step": 7900,
      "task_loss": 0.5259459614753723
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7965078353881836,
      "epoch": 2.86,
      "learning_rate": 2.3329115979008098e-05,
      "loss": 0.6016,
      "step": 7910,
      "task_loss": 0.9792982935905457
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7713618278503418,
      "epoch": 2.86,
      "learning_rate": 2.3262716038214055e-05,
      "loss": 0.5892,
      "step": 7920,
      "task_loss": 0.8038817048072815
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6315901875495911,
      "epoch": 2.87,
      "learning_rate": 2.319635083712164e-05,
      "loss": 0.5492,
      "step": 7930,
      "task_loss": 0.800554096698761
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7543363571166992,
      "epoch": 2.87,
      "learning_rate": 2.3130020717932155e-05,
      "loss": 0.5724,
      "step": 7940,
      "task_loss": 0.6426233649253845
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7032104134559631,
      "epoch": 2.87,
      "learning_rate": 2.306372602266599e-05,
      "loss": 0.6558,
      "step": 7950,
      "task_loss": 0.6244052648544312
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8306782245635986,
      "epoch": 2.88,
      "learning_rate": 2.29974670931609e-05,
      "loss": 0.6546,
      "step": 7960,
      "task_loss": 0.9088537096977234
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.951589822769165,
      "epoch": 2.88,
      "learning_rate": 2.2931244271070183e-05,
      "loss": 0.6349,
      "step": 7970,
      "task_loss": 1.1932218074798584
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.524719774723053,
      "epoch": 2.88,
      "learning_rate": 2.2865057897861014e-05,
      "loss": 0.5249,
      "step": 7980,
      "task_loss": 0.6670576333999634
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5125932693481445,
      "epoch": 2.89,
      "learning_rate": 2.2798908314812546e-05,
      "loss": 0.5753,
      "step": 7990,
      "task_loss": 0.8458143472671509
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6288058161735535,
      "epoch": 2.89,
      "learning_rate": 2.273279586301431e-05,
      "loss": 0.5709,
      "step": 8000,
      "task_loss": 0.5643374919891357
    },
    {
      "epoch": 2.89,
      "eval_exact_match": 82.39356669820246,
      "eval_f1": 89.43276664538762,
      "step": 8000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42751121520996094,
      "epoch": 2.89,
      "learning_rate": 2.2666720883364317e-05,
      "loss": 0.4948,
      "step": 8010,
      "task_loss": 0.41463011503219604
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6234784126281738,
      "epoch": 2.9,
      "learning_rate": 2.26006837165674e-05,
      "loss": 0.5213,
      "step": 8020,
      "task_loss": 0.8985073566436768
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35889315605163574,
      "epoch": 2.9,
      "learning_rate": 2.2534684703133375e-05,
      "loss": 0.593,
      "step": 8030,
      "task_loss": 0.40073755383491516
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3940616250038147,
      "epoch": 2.91,
      "learning_rate": 2.2468724183375365e-05,
      "loss": 0.608,
      "step": 8040,
      "task_loss": 0.7963597774505615
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5547673106193542,
      "epoch": 2.91,
      "learning_rate": 2.240280249740798e-05,
      "loss": 0.5817,
      "step": 8050,
      "task_loss": 0.8905491828918457
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5244134664535522,
      "epoch": 2.91,
      "learning_rate": 2.233691998514561e-05,
      "loss": 0.5501,
      "step": 8060,
      "task_loss": 0.6890424489974976
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5362120270729065,
      "epoch": 2.92,
      "learning_rate": 2.2271076986300626e-05,
      "loss": 0.5356,
      "step": 8070,
      "task_loss": 0.6972655057907104
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5454753637313843,
      "epoch": 2.92,
      "learning_rate": 2.2205273840381697e-05,
      "loss": 0.5795,
      "step": 8080,
      "task_loss": 0.44935011863708496
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5678622722625732,
      "epoch": 2.92,
      "learning_rate": 2.213951088669194e-05,
      "loss": 0.5637,
      "step": 8090,
      "task_loss": 0.5042102336883545
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8571547269821167,
      "epoch": 2.93,
      "learning_rate": 2.2073788464327286e-05,
      "loss": 0.6529,
      "step": 8100,
      "task_loss": 0.9279406666755676
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4891819357872009,
      "epoch": 2.93,
      "learning_rate": 2.2008106912174626e-05,
      "loss": 0.5716,
      "step": 8110,
      "task_loss": 0.863795280456543
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9461628198623657,
      "epoch": 2.93,
      "learning_rate": 2.1942466568910142e-05,
      "loss": 0.6349,
      "step": 8120,
      "task_loss": 0.8306816220283508
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49649107456207275,
      "epoch": 2.94,
      "learning_rate": 2.18768677729975e-05,
      "loss": 0.6054,
      "step": 8130,
      "task_loss": 0.5083526372909546
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6678476929664612,
      "epoch": 2.94,
      "learning_rate": 2.1811310862686177e-05,
      "loss": 0.5735,
      "step": 8140,
      "task_loss": 0.7546992301940918
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.657577633857727,
      "epoch": 2.95,
      "learning_rate": 2.1745796176009618e-05,
      "loss": 0.632,
      "step": 8150,
      "task_loss": 0.7159422636032104
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6389050483703613,
      "epoch": 2.95,
      "learning_rate": 2.1680324050783598e-05,
      "loss": 0.6101,
      "step": 8160,
      "task_loss": 0.4867662787437439
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3546678125858307,
      "epoch": 2.95,
      "learning_rate": 2.161489482460439e-05,
      "loss": 0.5069,
      "step": 8170,
      "task_loss": 0.4260442852973938
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49699175357818604,
      "epoch": 2.96,
      "learning_rate": 2.1549508834847124e-05,
      "loss": 0.5537,
      "step": 8180,
      "task_loss": 0.6937697529792786
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6141911149024963,
      "epoch": 2.96,
      "learning_rate": 2.1484166418663904e-05,
      "loss": 0.5792,
      "step": 8190,
      "task_loss": 0.7282106876373291
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7464488744735718,
      "epoch": 2.96,
      "learning_rate": 2.1418867912982233e-05,
      "loss": 0.6254,
      "step": 8200,
      "task_loss": 0.807386040687561
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.712665855884552,
      "epoch": 2.97,
      "learning_rate": 2.135361365450315e-05,
      "loss": 0.6006,
      "step": 8210,
      "task_loss": 1.1894166469573975
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6086498498916626,
      "epoch": 2.97,
      "learning_rate": 2.1288403979699563e-05,
      "loss": 0.5498,
      "step": 8220,
      "task_loss": 0.7981095910072327
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.646178126335144,
      "epoch": 2.97,
      "learning_rate": 2.122323922481447e-05,
      "loss": 0.5831,
      "step": 8230,
      "task_loss": 0.7211898565292358
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.522659420967102,
      "epoch": 2.98,
      "learning_rate": 2.1158119725859282e-05,
      "loss": 0.5215,
      "step": 8240,
      "task_loss": 0.5190514326095581
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5700294971466064,
      "epoch": 2.98,
      "learning_rate": 2.1093045818612006e-05,
      "loss": 0.5412,
      "step": 8250,
      "task_loss": 0.9010043144226074
    },
    {
      "epoch": 2.98,
      "eval_exact_match": 82.32734153263955,
      "eval_f1": 89.52267203340489,
      "step": 8250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5628690123558044,
      "epoch": 2.99,
      "learning_rate": 2.102801783861561e-05,
      "loss": 0.5405,
      "step": 8260,
      "task_loss": 0.7926698327064514
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41890618205070496,
      "epoch": 2.99,
      "learning_rate": 2.0963036121176206e-05,
      "loss": 0.5562,
      "step": 8270,
      "task_loss": 0.4853794574737549
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6962077021598816,
      "epoch": 2.99,
      "learning_rate": 2.0898101001361418e-05,
      "loss": 0.5628,
      "step": 8280,
      "task_loss": 0.6412039995193481
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.629459023475647,
      "epoch": 3.0,
      "learning_rate": 2.083321281399851e-05,
      "loss": 0.6054,
      "step": 8290,
      "task_loss": 0.4869459867477417
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5349444150924683,
      "epoch": 3.0,
      "learning_rate": 2.0768371893672834e-05,
      "loss": 0.5765,
      "step": 8300,
      "task_loss": 0.7212353944778442
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4701830744743347,
      "epoch": 3.0,
      "learning_rate": 2.0703578574725962e-05,
      "loss": 0.4958,
      "step": 8310,
      "task_loss": 0.6365481615066528
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37688684463500977,
      "epoch": 3.01,
      "learning_rate": 2.0638833191254032e-05,
      "loss": 0.5041,
      "step": 8320,
      "task_loss": 0.551572859287262
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5466549396514893,
      "epoch": 3.01,
      "learning_rate": 2.0574136077106028e-05,
      "loss": 0.5433,
      "step": 8330,
      "task_loss": 0.3961504399776459
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.522490382194519,
      "epoch": 3.01,
      "learning_rate": 2.0509487565882006e-05,
      "loss": 0.5166,
      "step": 8340,
      "task_loss": 0.9316266775131226
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.549529492855072,
      "epoch": 3.02,
      "learning_rate": 2.0444887990931454e-05,
      "loss": 0.5098,
      "step": 8350,
      "task_loss": 0.7794489860534668
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4621466398239136,
      "epoch": 3.02,
      "learning_rate": 2.038033768535147e-05,
      "loss": 0.4768,
      "step": 8360,
      "task_loss": 0.4732331335544586
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6309295892715454,
      "epoch": 3.02,
      "learning_rate": 2.0315836981985175e-05,
      "loss": 0.5065,
      "step": 8370,
      "task_loss": 1.0547935962677002
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31558752059936523,
      "epoch": 3.03,
      "learning_rate": 2.025138621341985e-05,
      "loss": 0.5306,
      "step": 8380,
      "task_loss": 0.29626554250717163
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6012607216835022,
      "epoch": 3.03,
      "learning_rate": 2.0186985711985356e-05,
      "loss": 0.502,
      "step": 8390,
      "task_loss": 1.0542261600494385
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.427781879901886,
      "epoch": 3.04,
      "learning_rate": 2.012263580975231e-05,
      "loss": 0.4638,
      "step": 8400,
      "task_loss": 0.7197548747062683
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5939445495605469,
      "epoch": 3.04,
      "learning_rate": 2.005833683853047e-05,
      "loss": 0.4944,
      "step": 8410,
      "task_loss": 0.7222771048545837
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.603698194026947,
      "epoch": 3.04,
      "learning_rate": 1.999408912986694e-05,
      "loss": 0.5148,
      "step": 8420,
      "task_loss": 0.7187478542327881
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3711569011211395,
      "epoch": 3.05,
      "learning_rate": 1.9929893015044516e-05,
      "loss": 0.4703,
      "step": 8430,
      "task_loss": 0.48197200894355774
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3467344641685486,
      "epoch": 3.05,
      "learning_rate": 1.986574882507994e-05,
      "loss": 0.5674,
      "step": 8440,
      "task_loss": 0.2590097486972809
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38601693511009216,
      "epoch": 3.05,
      "learning_rate": 1.9801656890722253e-05,
      "loss": 0.4763,
      "step": 8450,
      "task_loss": 0.44248801469802856
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3432322144508362,
      "epoch": 3.06,
      "learning_rate": 1.973761754245098e-05,
      "loss": 0.4624,
      "step": 8460,
      "task_loss": 0.3968687653541565
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37857741117477417,
      "epoch": 3.06,
      "learning_rate": 1.9673631110474576e-05,
      "loss": 0.4573,
      "step": 8470,
      "task_loss": 0.5128192901611328
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6123422980308533,
      "epoch": 3.06,
      "learning_rate": 1.9609697924728577e-05,
      "loss": 0.4654,
      "step": 8480,
      "task_loss": 0.9716970324516296
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6166114807128906,
      "epoch": 3.07,
      "learning_rate": 1.9545818314873996e-05,
      "loss": 0.5385,
      "step": 8490,
      "task_loss": 0.4100545644760132
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5452293157577515,
      "epoch": 3.07,
      "learning_rate": 1.9481992610295572e-05,
      "loss": 0.5531,
      "step": 8500,
      "task_loss": 0.3822357654571533
    },
    {
      "epoch": 3.07,
      "eval_exact_match": 82.11920529801324,
      "eval_f1": 89.28049962544765,
      "step": 8500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7301047444343567,
      "epoch": 3.08,
      "learning_rate": 1.941822114010011e-05,
      "loss": 0.486,
      "step": 8510,
      "task_loss": 0.9077910780906677
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7157807350158691,
      "epoch": 3.08,
      "learning_rate": 1.9354504233114756e-05,
      "loss": 0.5203,
      "step": 8520,
      "task_loss": 0.7050743103027344
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5539354681968689,
      "epoch": 3.08,
      "learning_rate": 1.92908422178853e-05,
      "loss": 0.4978,
      "step": 8530,
      "task_loss": 0.8747463226318359
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3879743814468384,
      "epoch": 3.09,
      "learning_rate": 1.9227235422674496e-05,
      "loss": 0.4905,
      "step": 8540,
      "task_loss": 0.4578686058521271
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47939151525497437,
      "epoch": 3.09,
      "learning_rate": 1.9163684175460395e-05,
      "loss": 0.5013,
      "step": 8550,
      "task_loss": 0.5444331765174866
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6490421891212463,
      "epoch": 3.09,
      "learning_rate": 1.9100188803934563e-05,
      "loss": 0.5392,
      "step": 8560,
      "task_loss": 0.6370527744293213
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5446510314941406,
      "epoch": 3.1,
      "learning_rate": 1.9036749635500517e-05,
      "loss": 0.518,
      "step": 8570,
      "task_loss": 0.7046805620193481
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5223091840744019,
      "epoch": 3.1,
      "learning_rate": 1.897336699727192e-05,
      "loss": 0.5085,
      "step": 8580,
      "task_loss": 0.8283286094665527
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5764731764793396,
      "epoch": 3.1,
      "learning_rate": 1.8910041216070985e-05,
      "loss": 0.5382,
      "step": 8590,
      "task_loss": 0.5241403579711914
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44843530654907227,
      "epoch": 3.11,
      "learning_rate": 1.884677261842672e-05,
      "loss": 0.5134,
      "step": 8600,
      "task_loss": 0.47629907727241516
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5873540639877319,
      "epoch": 3.11,
      "learning_rate": 1.8783561530573297e-05,
      "loss": 0.4634,
      "step": 8610,
      "task_loss": 0.49033936858177185
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5004246234893799,
      "epoch": 3.12,
      "learning_rate": 1.8720408278448332e-05,
      "loss": 0.4896,
      "step": 8620,
      "task_loss": 0.4907362759113312
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3840811252593994,
      "epoch": 3.12,
      "learning_rate": 1.8657313187691257e-05,
      "loss": 0.5091,
      "step": 8630,
      "task_loss": 0.6038985252380371
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3785248398780823,
      "epoch": 3.12,
      "learning_rate": 1.8594276583641534e-05,
      "loss": 0.52,
      "step": 8640,
      "task_loss": 0.36846432089805603
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5661005973815918,
      "epoch": 3.13,
      "learning_rate": 1.8531298791337127e-05,
      "loss": 0.5173,
      "step": 8650,
      "task_loss": 0.5859448909759521
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5267096757888794,
      "epoch": 3.13,
      "learning_rate": 1.8468380135512688e-05,
      "loss": 0.4774,
      "step": 8660,
      "task_loss": 0.985637903213501
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5141612887382507,
      "epoch": 3.13,
      "learning_rate": 1.8405520940597985e-05,
      "loss": 0.5026,
      "step": 8670,
      "task_loss": 0.5452569723129272
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4393519163131714,
      "epoch": 3.14,
      "learning_rate": 1.834272153071614e-05,
      "loss": 0.488,
      "step": 8680,
      "task_loss": 0.5240843892097473
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4073109030723572,
      "epoch": 3.14,
      "learning_rate": 1.827998222968205e-05,
      "loss": 0.4844,
      "step": 8690,
      "task_loss": 0.4562574625015259
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4827924966812134,
      "epoch": 3.14,
      "learning_rate": 1.8217303361000625e-05,
      "loss": 0.5465,
      "step": 8700,
      "task_loss": 0.5933099985122681
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5707128047943115,
      "epoch": 3.15,
      "learning_rate": 1.8154685247865215e-05,
      "loss": 0.4999,
      "step": 8710,
      "task_loss": 0.6060525178909302
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4365973472595215,
      "epoch": 3.15,
      "learning_rate": 1.809212821315584e-05,
      "loss": 0.4649,
      "step": 8720,
      "task_loss": 0.8164606094360352
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3885442614555359,
      "epoch": 3.16,
      "learning_rate": 1.802963257943764e-05,
      "loss": 0.4712,
      "step": 8730,
      "task_loss": 0.46396589279174805
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5628999471664429,
      "epoch": 3.16,
      "learning_rate": 1.7967198668959077e-05,
      "loss": 0.4756,
      "step": 8740,
      "task_loss": 0.540549635887146
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6016356945037842,
      "epoch": 3.16,
      "learning_rate": 1.7904826803650426e-05,
      "loss": 0.5062,
      "step": 8750,
      "task_loss": 0.46754416823387146
    },
    {
      "epoch": 3.16,
      "eval_exact_match": 82.5449385052034,
      "eval_f1": 89.74640245653393,
      "step": 8750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6363502740859985,
      "epoch": 3.17,
      "learning_rate": 1.7842517305121973e-05,
      "loss": 0.5155,
      "step": 8760,
      "task_loss": 0.5670214891433716
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6069682836532593,
      "epoch": 3.17,
      "learning_rate": 1.778027049466246e-05,
      "loss": 0.4794,
      "step": 8770,
      "task_loss": 0.5283325910568237
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5104897022247314,
      "epoch": 3.17,
      "learning_rate": 1.7718086693237365e-05,
      "loss": 0.4731,
      "step": 8780,
      "task_loss": 0.604925274848938
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5205997228622437,
      "epoch": 3.18,
      "learning_rate": 1.7655966221487286e-05,
      "loss": 0.5561,
      "step": 8790,
      "task_loss": 0.7673854231834412
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43104785680770874,
      "epoch": 3.18,
      "learning_rate": 1.759390939972625e-05,
      "loss": 0.4787,
      "step": 8800,
      "task_loss": 0.4828697443008423
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39554059505462646,
      "epoch": 3.18,
      "learning_rate": 1.753191654794012e-05,
      "loss": 0.4671,
      "step": 8810,
      "task_loss": 0.33180779218673706
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5428735613822937,
      "epoch": 3.19,
      "learning_rate": 1.7469987985784853e-05,
      "loss": 0.4931,
      "step": 8820,
      "task_loss": 1.0688862800598145
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40573301911354065,
      "epoch": 3.19,
      "learning_rate": 1.7408124032584977e-05,
      "loss": 0.5403,
      "step": 8830,
      "task_loss": 0.7760818004608154
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39982184767723083,
      "epoch": 3.19,
      "learning_rate": 1.7346325007331814e-05,
      "loss": 0.4537,
      "step": 8840,
      "task_loss": 0.48054105043411255
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5107048153877258,
      "epoch": 3.2,
      "learning_rate": 1.7284591228681924e-05,
      "loss": 0.4919,
      "step": 8850,
      "task_loss": 0.7338130474090576
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28509578108787537,
      "epoch": 3.2,
      "learning_rate": 1.722292301495543e-05,
      "loss": 0.4804,
      "step": 8860,
      "task_loss": 0.32352930307388306
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46888577938079834,
      "epoch": 3.21,
      "learning_rate": 1.716132068413439e-05,
      "loss": 0.5633,
      "step": 8870,
      "task_loss": 0.8163585662841797
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4140506684780121,
      "epoch": 3.21,
      "learning_rate": 1.70997845538611e-05,
      "loss": 0.5328,
      "step": 8880,
      "task_loss": 0.5608940124511719
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4530109763145447,
      "epoch": 3.21,
      "learning_rate": 1.7038314941436593e-05,
      "loss": 0.5011,
      "step": 8890,
      "task_loss": 0.6756296157836914
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45567673444747925,
      "epoch": 3.22,
      "learning_rate": 1.697691216381882e-05,
      "loss": 0.51,
      "step": 8900,
      "task_loss": 0.7713099122047424
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.602544903755188,
      "epoch": 3.22,
      "learning_rate": 1.691557653762117e-05,
      "loss": 0.5366,
      "step": 8910,
      "task_loss": 0.8172224164009094
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6989117860794067,
      "epoch": 3.22,
      "learning_rate": 1.685430837911074e-05,
      "loss": 0.5546,
      "step": 8920,
      "task_loss": 0.6363785862922668
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4007064700126648,
      "epoch": 3.23,
      "learning_rate": 1.6793108004206788e-05,
      "loss": 0.4931,
      "step": 8930,
      "task_loss": 0.4201916456222534
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4066169261932373,
      "epoch": 3.23,
      "learning_rate": 1.6731975728478994e-05,
      "loss": 0.4984,
      "step": 8940,
      "task_loss": 0.5423001050949097
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5652238130569458,
      "epoch": 3.23,
      "learning_rate": 1.667091186714594e-05,
      "loss": 0.5475,
      "step": 8950,
      "task_loss": 0.619628369808197
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5001813769340515,
      "epoch": 3.24,
      "learning_rate": 1.6609916735073432e-05,
      "loss": 0.4621,
      "step": 8960,
      "task_loss": 0.90772545337677
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3538113236427307,
      "epoch": 3.24,
      "learning_rate": 1.654899064677289e-05,
      "loss": 0.5043,
      "step": 8970,
      "task_loss": 0.443660706281662
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.48391348123550415,
      "epoch": 3.25,
      "learning_rate": 1.648813391639968e-05,
      "loss": 0.5245,
      "step": 8980,
      "task_loss": 0.739136815071106
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45259201526641846,
      "epoch": 3.25,
      "learning_rate": 1.6427346857751618e-05,
      "loss": 0.4996,
      "step": 8990,
      "task_loss": 0.3973962962627411
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4466731548309326,
      "epoch": 3.25,
      "learning_rate": 1.6366629784267172e-05,
      "loss": 0.4757,
      "step": 9000,
      "task_loss": 0.7052876949310303
    },
    {
      "epoch": 3.25,
      "eval_exact_match": 82.45979186376537,
      "eval_f1": 89.55562803988978,
      "step": 9000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5209038257598877,
      "epoch": 3.26,
      "learning_rate": 1.6305983009024018e-05,
      "loss": 0.4895,
      "step": 9010,
      "task_loss": 0.9570509195327759
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5868513584136963,
      "epoch": 3.26,
      "learning_rate": 1.624540684473731e-05,
      "loss": 0.5143,
      "step": 9020,
      "task_loss": 0.8736361861228943
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4569542706012726,
      "epoch": 3.26,
      "learning_rate": 1.6184901603758137e-05,
      "loss": 0.4714,
      "step": 9030,
      "task_loss": 0.41086018085479736
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3990710973739624,
      "epoch": 3.27,
      "learning_rate": 1.612446759807184e-05,
      "loss": 0.47,
      "step": 9040,
      "task_loss": 0.8022845387458801
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5180515646934509,
      "epoch": 3.27,
      "learning_rate": 1.606410513929649e-05,
      "loss": 0.5147,
      "step": 9050,
      "task_loss": 0.989239513874054
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41727307438850403,
      "epoch": 3.27,
      "learning_rate": 1.600381453868123e-05,
      "loss": 0.51,
      "step": 9060,
      "task_loss": 0.6814265251159668
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6537793278694153,
      "epoch": 3.28,
      "learning_rate": 1.594359610710467e-05,
      "loss": 0.4618,
      "step": 9070,
      "task_loss": 1.0158696174621582
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31821686029434204,
      "epoch": 3.28,
      "learning_rate": 1.5883450155073262e-05,
      "loss": 0.4612,
      "step": 9080,
      "task_loss": 0.2660524845123291
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5743929147720337,
      "epoch": 3.29,
      "learning_rate": 1.5823376992719805e-05,
      "loss": 0.5242,
      "step": 9090,
      "task_loss": 0.4032166600227356
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42509496212005615,
      "epoch": 3.29,
      "learning_rate": 1.5763376929801693e-05,
      "loss": 0.5119,
      "step": 9100,
      "task_loss": 0.5546834468841553
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3368126153945923,
      "epoch": 3.29,
      "learning_rate": 1.5703450275699436e-05,
      "loss": 0.5213,
      "step": 9110,
      "task_loss": 0.5160356163978577
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46083498001098633,
      "epoch": 3.3,
      "learning_rate": 1.5643597339415015e-05,
      "loss": 0.5059,
      "step": 9120,
      "task_loss": 0.8579056262969971
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43022578954696655,
      "epoch": 3.3,
      "learning_rate": 1.5583818429570307e-05,
      "loss": 0.5503,
      "step": 9130,
      "task_loss": 0.37180399894714355
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4419669508934021,
      "epoch": 3.3,
      "learning_rate": 1.5524113854405446e-05,
      "loss": 0.5065,
      "step": 9140,
      "task_loss": 0.5681513547897339
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4496462941169739,
      "epoch": 3.31,
      "learning_rate": 1.546448392177732e-05,
      "loss": 0.5143,
      "step": 9150,
      "task_loss": 0.5917080640792847
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4904141128063202,
      "epoch": 3.31,
      "learning_rate": 1.54049289391579e-05,
      "loss": 0.5138,
      "step": 9160,
      "task_loss": 0.5543509721755981
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44269993901252747,
      "epoch": 3.31,
      "learning_rate": 1.5345449213632725e-05,
      "loss": 0.5018,
      "step": 9170,
      "task_loss": 0.5917487144470215
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4732619524002075,
      "epoch": 3.32,
      "learning_rate": 1.528604505189922e-05,
      "loss": 0.4905,
      "step": 9180,
      "task_loss": 0.40163156390190125
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5729936361312866,
      "epoch": 3.32,
      "learning_rate": 1.5226716760265265e-05,
      "loss": 0.4991,
      "step": 9190,
      "task_loss": 0.8118762969970703
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4613145589828491,
      "epoch": 3.32,
      "learning_rate": 1.5167464644647453e-05,
      "loss": 0.4727,
      "step": 9200,
      "task_loss": 0.69233238697052
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6504011154174805,
      "epoch": 3.33,
      "learning_rate": 1.5108289010569618e-05,
      "loss": 0.5479,
      "step": 9210,
      "task_loss": 0.8994011282920837
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6478760242462158,
      "epoch": 3.33,
      "learning_rate": 1.504919016316124e-05,
      "loss": 0.475,
      "step": 9220,
      "task_loss": 0.603376030921936
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7874225974082947,
      "epoch": 3.34,
      "learning_rate": 1.4990168407155853e-05,
      "loss": 0.4952,
      "step": 9230,
      "task_loss": 0.5990192890167236
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34686243534088135,
      "epoch": 3.34,
      "learning_rate": 1.4931224046889456e-05,
      "loss": 0.4821,
      "step": 9240,
      "task_loss": 0.4154958128929138
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4504637122154236,
      "epoch": 3.34,
      "learning_rate": 1.4872357386299005e-05,
      "loss": 0.5222,
      "step": 9250,
      "task_loss": 0.7193425297737122
    },
    {
      "epoch": 3.34,
      "eval_exact_match": 81.93945127719962,
      "eval_f1": 89.31012659211146,
      "step": 9250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.632270872592926,
      "epoch": 3.35,
      "learning_rate": 1.4813568728920795e-05,
      "loss": 0.5471,
      "step": 9260,
      "task_loss": 0.6672326326370239
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.453116774559021,
      "epoch": 3.35,
      "learning_rate": 1.4754858377888926e-05,
      "loss": 0.5107,
      "step": 9270,
      "task_loss": 0.31693387031555176
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5828425288200378,
      "epoch": 3.35,
      "learning_rate": 1.4696226635933674e-05,
      "loss": 0.4849,
      "step": 9280,
      "task_loss": 0.9231215119361877
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4946627616882324,
      "epoch": 3.36,
      "learning_rate": 1.4637673805380065e-05,
      "loss": 0.5012,
      "step": 9290,
      "task_loss": 0.6342258453369141
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41230204701423645,
      "epoch": 3.36,
      "learning_rate": 1.4579200188146144e-05,
      "loss": 0.482,
      "step": 9300,
      "task_loss": 0.5047462582588196
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5421258807182312,
      "epoch": 3.36,
      "learning_rate": 1.4520806085741555e-05,
      "loss": 0.5333,
      "step": 9310,
      "task_loss": 0.9924592971801758
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4269621968269348,
      "epoch": 3.37,
      "learning_rate": 1.4462491799265932e-05,
      "loss": 0.4559,
      "step": 9320,
      "task_loss": 0.7256890535354614
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5573655366897583,
      "epoch": 3.37,
      "learning_rate": 1.4404257629407353e-05,
      "loss": 0.5239,
      "step": 9330,
      "task_loss": 0.46378791332244873
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3567560911178589,
      "epoch": 3.38,
      "learning_rate": 1.4346103876440755e-05,
      "loss": 0.5374,
      "step": 9340,
      "task_loss": 0.4994860887527466
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45855897665023804,
      "epoch": 3.38,
      "learning_rate": 1.4288030840226461e-05,
      "loss": 0.5311,
      "step": 9350,
      "task_loss": 0.6456738710403442
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39899706840515137,
      "epoch": 3.38,
      "learning_rate": 1.4230038820208569e-05,
      "loss": 0.4944,
      "step": 9360,
      "task_loss": 0.5475515723228455
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4496174454689026,
      "epoch": 3.39,
      "learning_rate": 1.4172128115413446e-05,
      "loss": 0.4966,
      "step": 9370,
      "task_loss": 0.5722992420196533
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.48511573672294617,
      "epoch": 3.39,
      "learning_rate": 1.4114299024448127e-05,
      "loss": 0.4852,
      "step": 9380,
      "task_loss": 0.5806999206542969
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3434864580631256,
      "epoch": 3.39,
      "learning_rate": 1.4056551845498899e-05,
      "loss": 0.473,
      "step": 9390,
      "task_loss": 0.28214818239212036
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.520751953125,
      "epoch": 3.4,
      "learning_rate": 1.3998886876329597e-05,
      "loss": 0.4578,
      "step": 9400,
      "task_loss": 0.6089061498641968
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49395525455474854,
      "epoch": 3.4,
      "learning_rate": 1.3941304414280207e-05,
      "loss": 0.461,
      "step": 9410,
      "task_loss": 0.9712688326835632
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33728528022766113,
      "epoch": 3.4,
      "learning_rate": 1.3883804756265264e-05,
      "loss": 0.4602,
      "step": 9420,
      "task_loss": 0.5431016087532043
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47305771708488464,
      "epoch": 3.41,
      "learning_rate": 1.382638819877235e-05,
      "loss": 0.448,
      "step": 9430,
      "task_loss": 0.8087161183357239
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47628888487815857,
      "epoch": 3.41,
      "learning_rate": 1.3769055037860525e-05,
      "loss": 0.5081,
      "step": 9440,
      "task_loss": 0.7864973545074463
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7246733903884888,
      "epoch": 3.42,
      "learning_rate": 1.3711805569158852e-05,
      "loss": 0.5485,
      "step": 9450,
      "task_loss": 0.620114803314209
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41885077953338623,
      "epoch": 3.42,
      "learning_rate": 1.3654640087864848e-05,
      "loss": 0.4701,
      "step": 9460,
      "task_loss": 0.7940152883529663
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33186885714530945,
      "epoch": 3.42,
      "learning_rate": 1.3597558888742954e-05,
      "loss": 0.4616,
      "step": 9470,
      "task_loss": 0.5817943811416626
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3579007387161255,
      "epoch": 3.43,
      "learning_rate": 1.3540562266123025e-05,
      "loss": 0.4819,
      "step": 9480,
      "task_loss": 0.36475780606269836
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32448476552963257,
      "epoch": 3.43,
      "learning_rate": 1.3483650513898832e-05,
      "loss": 0.469,
      "step": 9490,
      "task_loss": 0.7694838047027588
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5285438299179077,
      "epoch": 3.43,
      "learning_rate": 1.3426823925526479e-05,
      "loss": 0.5106,
      "step": 9500,
      "task_loss": 0.8708394765853882
    },
    {
      "epoch": 3.43,
      "eval_exact_match": 82.31788079470199,
      "eval_f1": 89.50688669947229,
      "step": 9500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4109702706336975,
      "epoch": 3.44,
      "learning_rate": 1.3370082794022974e-05,
      "loss": 0.5511,
      "step": 9510,
      "task_loss": 0.45803433656692505
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3952751159667969,
      "epoch": 3.44,
      "learning_rate": 1.3313427411964676e-05,
      "loss": 0.4921,
      "step": 9520,
      "task_loss": 0.49687451124191284
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5060081481933594,
      "epoch": 3.44,
      "learning_rate": 1.3256858071485795e-05,
      "loss": 0.4775,
      "step": 9530,
      "task_loss": 0.594099223613739
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.426449716091156,
      "epoch": 3.45,
      "learning_rate": 1.3200375064276841e-05,
      "loss": 0.492,
      "step": 9540,
      "task_loss": 0.5237439274787903
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5308140516281128,
      "epoch": 3.45,
      "learning_rate": 1.3143978681583236e-05,
      "loss": 0.5023,
      "step": 9550,
      "task_loss": 1.1491860151290894
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6118658781051636,
      "epoch": 3.46,
      "learning_rate": 1.3087669214203657e-05,
      "loss": 0.5216,
      "step": 9560,
      "task_loss": 0.6933400630950928
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4912259578704834,
      "epoch": 3.46,
      "learning_rate": 1.303144695248867e-05,
      "loss": 0.4434,
      "step": 9570,
      "task_loss": 0.86939936876297
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7513962388038635,
      "epoch": 3.46,
      "learning_rate": 1.2975312186339156e-05,
      "loss": 0.5468,
      "step": 9580,
      "task_loss": 1.1207659244537354
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6973851919174194,
      "epoch": 3.47,
      "learning_rate": 1.2919265205204863e-05,
      "loss": 0.4847,
      "step": 9590,
      "task_loss": 0.9924426078796387
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5232359766960144,
      "epoch": 3.47,
      "learning_rate": 1.2863306298082847e-05,
      "loss": 0.5701,
      "step": 9600,
      "task_loss": 1.4724518060684204
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49260246753692627,
      "epoch": 3.47,
      "learning_rate": 1.2807435753516059e-05,
      "loss": 0.5398,
      "step": 9610,
      "task_loss": 1.156559944152832
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4523448348045349,
      "epoch": 3.48,
      "learning_rate": 1.2751653859591815e-05,
      "loss": 0.4651,
      "step": 9620,
      "task_loss": 0.5585622787475586
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40095406770706177,
      "epoch": 3.48,
      "learning_rate": 1.2695960903940331e-05,
      "loss": 0.5763,
      "step": 9630,
      "task_loss": 0.6868075132369995
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39510098099708557,
      "epoch": 3.48,
      "learning_rate": 1.2640357173733184e-05,
      "loss": 0.4757,
      "step": 9640,
      "task_loss": 0.5561718940734863
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6113804578781128,
      "epoch": 3.49,
      "learning_rate": 1.258484295568194e-05,
      "loss": 0.4769,
      "step": 9650,
      "task_loss": 0.5155743956565857
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38609778881073,
      "epoch": 3.49,
      "learning_rate": 1.252941853603655e-05,
      "loss": 0.4734,
      "step": 9660,
      "task_loss": 0.5266010761260986
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5123001933097839,
      "epoch": 3.49,
      "learning_rate": 1.2474084200583964e-05,
      "loss": 0.4694,
      "step": 9670,
      "task_loss": 0.6387827396392822
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4852139353752136,
      "epoch": 3.5,
      "learning_rate": 1.2418840234646627e-05,
      "loss": 0.4779,
      "step": 9680,
      "task_loss": 0.48199230432510376
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5093953013420105,
      "epoch": 3.5,
      "learning_rate": 1.2363686923081015e-05,
      "loss": 0.5113,
      "step": 9690,
      "task_loss": 0.34967291355133057
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4891633689403534,
      "epoch": 3.51,
      "learning_rate": 1.2308624550276123e-05,
      "loss": 0.5272,
      "step": 9700,
      "task_loss": 0.5109860897064209
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37727656960487366,
      "epoch": 3.51,
      "learning_rate": 1.2253653400152073e-05,
      "loss": 0.5071,
      "step": 9710,
      "task_loss": 0.5218365788459778
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43080103397369385,
      "epoch": 3.51,
      "learning_rate": 1.2198773756158604e-05,
      "loss": 0.5266,
      "step": 9720,
      "task_loss": 0.6511129140853882
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42583030462265015,
      "epoch": 3.52,
      "learning_rate": 1.2143985901273618e-05,
      "loss": 0.5077,
      "step": 9730,
      "task_loss": 0.5220778584480286
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5337873697280884,
      "epoch": 3.52,
      "learning_rate": 1.2089290118001697e-05,
      "loss": 0.4677,
      "step": 9740,
      "task_loss": 0.7937961220741272
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6372555494308472,
      "epoch": 3.52,
      "learning_rate": 1.2034686688372724e-05,
      "loss": 0.6156,
      "step": 9750,
      "task_loss": 0.818684458732605
    },
    {
      "epoch": 3.52,
      "eval_exact_match": 82.30842005676443,
      "eval_f1": 89.48111611207813,
      "step": 9750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6067251563072205,
      "epoch": 3.53,
      "learning_rate": 1.1980175893940316e-05,
      "loss": 0.5211,
      "step": 9760,
      "task_loss": 0.7992770671844482
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5251370072364807,
      "epoch": 3.53,
      "learning_rate": 1.192575801578047e-05,
      "loss": 0.491,
      "step": 9770,
      "task_loss": 0.5564746856689453
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46255046129226685,
      "epoch": 3.53,
      "learning_rate": 1.1871433334490063e-05,
      "loss": 0.5013,
      "step": 9780,
      "task_loss": 0.593856930732727
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6018144488334656,
      "epoch": 3.54,
      "learning_rate": 1.181720213018543e-05,
      "loss": 0.5316,
      "step": 9790,
      "task_loss": 0.5894225239753723
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49758362770080566,
      "epoch": 3.54,
      "learning_rate": 1.1763064682500879e-05,
      "loss": 0.5543,
      "step": 9800,
      "task_loss": 0.5570957660675049
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.536584734916687,
      "epoch": 3.55,
      "learning_rate": 1.1709021270587304e-05,
      "loss": 0.4794,
      "step": 9810,
      "task_loss": 0.5450813174247742
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3794777989387512,
      "epoch": 3.55,
      "learning_rate": 1.1655072173110707e-05,
      "loss": 0.4492,
      "step": 9820,
      "task_loss": 0.3681301474571228
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4257237911224365,
      "epoch": 3.55,
      "learning_rate": 1.1601217668250789e-05,
      "loss": 0.4717,
      "step": 9830,
      "task_loss": 0.7739952802658081
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5841401815414429,
      "epoch": 3.56,
      "learning_rate": 1.1547458033699452e-05,
      "loss": 0.4669,
      "step": 9840,
      "task_loss": 0.7634145021438599
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44450539350509644,
      "epoch": 3.56,
      "learning_rate": 1.1493793546659497e-05,
      "loss": 0.4667,
      "step": 9850,
      "task_loss": 0.6069711446762085
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5876635313034058,
      "epoch": 3.56,
      "learning_rate": 1.1440224483843031e-05,
      "loss": 0.4881,
      "step": 9860,
      "task_loss": 0.7760765552520752
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5074859857559204,
      "epoch": 3.57,
      "learning_rate": 1.1386751121470161e-05,
      "loss": 0.5055,
      "step": 9870,
      "task_loss": 0.43628257513046265
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40634429454803467,
      "epoch": 3.57,
      "learning_rate": 1.1333373735267528e-05,
      "loss": 0.5235,
      "step": 9880,
      "task_loss": 0.5263395309448242
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5098382234573364,
      "epoch": 3.57,
      "learning_rate": 1.1280092600466898e-05,
      "loss": 0.4973,
      "step": 9890,
      "task_loss": 0.7390273809432983
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49760282039642334,
      "epoch": 3.58,
      "learning_rate": 1.1226907991803694e-05,
      "loss": 0.4334,
      "step": 9900,
      "task_loss": 0.48946425318717957
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3809628188610077,
      "epoch": 3.58,
      "learning_rate": 1.1173820183515652e-05,
      "loss": 0.491,
      "step": 9910,
      "task_loss": 0.5488647222518921
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5812700986862183,
      "epoch": 3.59,
      "learning_rate": 1.112082944934137e-05,
      "loss": 0.5098,
      "step": 9920,
      "task_loss": 0.656021237373352
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32435011863708496,
      "epoch": 3.59,
      "learning_rate": 1.1067936062518893e-05,
      "loss": 0.4879,
      "step": 9930,
      "task_loss": 0.2461182326078415
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7479100227355957,
      "epoch": 3.59,
      "learning_rate": 1.1015140295784286e-05,
      "loss": 0.5034,
      "step": 9940,
      "task_loss": 1.272243618965149
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6995210647583008,
      "epoch": 3.6,
      "learning_rate": 1.0962442421370318e-05,
      "loss": 0.4988,
      "step": 9950,
      "task_loss": 1.0233780145645142
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.48262107372283936,
      "epoch": 3.6,
      "learning_rate": 1.0909842711004913e-05,
      "loss": 0.472,
      "step": 9960,
      "task_loss": 0.5522131323814392
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5361160039901733,
      "epoch": 3.6,
      "learning_rate": 1.0857341435909876e-05,
      "loss": 0.4965,
      "step": 9970,
      "task_loss": 0.7532282471656799
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4412672221660614,
      "epoch": 3.61,
      "learning_rate": 1.0804938866799441e-05,
      "loss": 0.5342,
      "step": 9980,
      "task_loss": 0.8714137077331543
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49830132722854614,
      "epoch": 3.61,
      "learning_rate": 1.075263527387888e-05,
      "loss": 0.5097,
      "step": 9990,
      "task_loss": 0.6957769989967346
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5538281798362732,
      "epoch": 3.61,
      "learning_rate": 1.0700430926843094e-05,
      "loss": 0.4753,
      "step": 10000,
      "task_loss": 0.6752587556838989
    },
    {
      "epoch": 3.61,
      "eval_exact_match": 82.52601702932829,
      "eval_f1": 89.68187205330648,
      "step": 10000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5624034404754639,
      "epoch": 3.62,
      "learning_rate": 1.0648326094875253e-05,
      "loss": 0.4598,
      "step": 10010,
      "task_loss": 0.692568302154541
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4360852837562561,
      "epoch": 3.62,
      "learning_rate": 1.0596321046645394e-05,
      "loss": 0.4773,
      "step": 10020,
      "task_loss": 0.4527958631515503
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4201253056526184,
      "epoch": 3.62,
      "learning_rate": 1.0544416050309043e-05,
      "loss": 0.4736,
      "step": 10030,
      "task_loss": 0.4331403374671936
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4714534282684326,
      "epoch": 3.63,
      "learning_rate": 1.0492611373505785e-05,
      "loss": 0.4866,
      "step": 10040,
      "task_loss": 0.6479029655456543
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5169629454612732,
      "epoch": 3.63,
      "learning_rate": 1.0440907283357999e-05,
      "loss": 0.4508,
      "step": 10050,
      "task_loss": 0.6727709770202637
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4062344431877136,
      "epoch": 3.64,
      "learning_rate": 1.0389304046469323e-05,
      "loss": 0.4978,
      "step": 10060,
      "task_loss": 0.5695604085922241
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5002131462097168,
      "epoch": 3.64,
      "learning_rate": 1.033780192892341e-05,
      "loss": 0.4988,
      "step": 10070,
      "task_loss": 0.6775874495506287
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4262256622314453,
      "epoch": 3.64,
      "learning_rate": 1.0286401196282502e-05,
      "loss": 0.5113,
      "step": 10080,
      "task_loss": 0.5431799292564392
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3821599781513214,
      "epoch": 3.65,
      "learning_rate": 1.0235102113586063e-05,
      "loss": 0.4595,
      "step": 10090,
      "task_loss": 0.6419239044189453
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5522717237472534,
      "epoch": 3.65,
      "learning_rate": 1.01839049453494e-05,
      "loss": 0.481,
      "step": 10100,
      "task_loss": 0.4990994334220886
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5535731315612793,
      "epoch": 3.65,
      "learning_rate": 1.0132809955562337e-05,
      "loss": 0.4963,
      "step": 10110,
      "task_loss": 0.6528249979019165
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5421990156173706,
      "epoch": 3.66,
      "learning_rate": 1.0081817407687819e-05,
      "loss": 0.519,
      "step": 10120,
      "task_loss": 0.9474392533302307
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5861812829971313,
      "epoch": 3.66,
      "learning_rate": 1.0030927564660585e-05,
      "loss": 0.4887,
      "step": 10130,
      "task_loss": 0.5796793103218079
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37179023027420044,
      "epoch": 3.66,
      "learning_rate": 9.980140688885745e-06,
      "loss": 0.4896,
      "step": 10140,
      "task_loss": 0.537672221660614
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.513761579990387,
      "epoch": 3.67,
      "learning_rate": 9.929457042237548e-06,
      "loss": 0.5025,
      "step": 10150,
      "task_loss": 0.6180034279823303
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5233842134475708,
      "epoch": 3.67,
      "learning_rate": 9.878876886057894e-06,
      "loss": 0.4689,
      "step": 10160,
      "task_loss": 0.5639371871948242
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5454587936401367,
      "epoch": 3.68,
      "learning_rate": 9.828400481155083e-06,
      "loss": 0.5063,
      "step": 10170,
      "task_loss": 0.8569517731666565
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5449087023735046,
      "epoch": 3.68,
      "learning_rate": 9.77802808780243e-06,
      "loss": 0.5119,
      "step": 10180,
      "task_loss": 0.47911345958709717
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5373599529266357,
      "epoch": 3.68,
      "learning_rate": 9.727759965736939e-06,
      "loss": 0.4877,
      "step": 10190,
      "task_loss": 0.7237974405288696
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3350166082382202,
      "epoch": 3.69,
      "learning_rate": 9.677596374157957e-06,
      "loss": 0.4623,
      "step": 10200,
      "task_loss": 0.43222576379776
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47038358449935913,
      "epoch": 3.69,
      "learning_rate": 9.627537571725803e-06,
      "loss": 0.5102,
      "step": 10210,
      "task_loss": 0.5583560466766357
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6115131974220276,
      "epoch": 3.69,
      "learning_rate": 9.577583816560534e-06,
      "loss": 0.5384,
      "step": 10220,
      "task_loss": 0.6348150968551636
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.537125825881958,
      "epoch": 3.7,
      "learning_rate": 9.527735366240478e-06,
      "loss": 0.4803,
      "step": 10230,
      "task_loss": 0.3411647081375122
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38711780309677124,
      "epoch": 3.7,
      "learning_rate": 9.477992477801019e-06,
      "loss": 0.4531,
      "step": 10240,
      "task_loss": 0.2967953681945801
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.419807493686676,
      "epoch": 3.7,
      "learning_rate": 9.428355407733219e-06,
      "loss": 0.4769,
      "step": 10250,
      "task_loss": 0.7556804418563843
    },
    {
      "epoch": 3.7,
      "eval_exact_match": 82.36518448438979,
      "eval_f1": 89.59782459360834,
      "step": 10250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4036286771297455,
      "epoch": 3.71,
      "learning_rate": 9.378824411982507e-06,
      "loss": 0.4836,
      "step": 10260,
      "task_loss": 1.1075952053070068
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.637839674949646,
      "epoch": 3.71,
      "learning_rate": 9.329399745947339e-06,
      "loss": 0.4865,
      "step": 10270,
      "task_loss": 1.0219868421554565
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.387042760848999,
      "epoch": 3.72,
      "learning_rate": 9.28008166447792e-06,
      "loss": 0.4587,
      "step": 10280,
      "task_loss": 0.7554829120635986
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.519973635673523,
      "epoch": 3.72,
      "learning_rate": 9.230870421874859e-06,
      "loss": 0.5099,
      "step": 10290,
      "task_loss": 0.7394765615463257
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4160851836204529,
      "epoch": 3.72,
      "learning_rate": 9.181766271887887e-06,
      "loss": 0.4725,
      "step": 10300,
      "task_loss": 0.6241825819015503
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5004531741142273,
      "epoch": 3.73,
      "learning_rate": 9.132769467714479e-06,
      "loss": 0.4985,
      "step": 10310,
      "task_loss": 0.6873956322669983
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3454711437225342,
      "epoch": 3.73,
      "learning_rate": 9.083880261998675e-06,
      "loss": 0.4592,
      "step": 10320,
      "task_loss": 0.4680253863334656
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4951745271682739,
      "epoch": 3.73,
      "learning_rate": 9.035098906829628e-06,
      "loss": 0.462,
      "step": 10330,
      "task_loss": 0.4511879086494446
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7201694250106812,
      "epoch": 3.74,
      "learning_rate": 8.986425653740421e-06,
      "loss": 0.5548,
      "step": 10340,
      "task_loss": 0.6285726428031921
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6020553112030029,
      "epoch": 3.74,
      "learning_rate": 8.937860753706713e-06,
      "loss": 0.4861,
      "step": 10350,
      "task_loss": 0.7205938696861267
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40691256523132324,
      "epoch": 3.74,
      "learning_rate": 8.88940445714547e-06,
      "loss": 0.4598,
      "step": 10360,
      "task_loss": 0.44143691658973694
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5514005422592163,
      "epoch": 3.75,
      "learning_rate": 8.841057013913629e-06,
      "loss": 0.473,
      "step": 10370,
      "task_loss": 1.1116174459457397
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6847667694091797,
      "epoch": 3.75,
      "learning_rate": 8.79281867330688e-06,
      "loss": 0.5555,
      "step": 10380,
      "task_loss": 1.004784107208252
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5778000354766846,
      "epoch": 3.75,
      "learning_rate": 8.744689684058324e-06,
      "loss": 0.4873,
      "step": 10390,
      "task_loss": 0.9104892015457153
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41177552938461304,
      "epoch": 3.76,
      "learning_rate": 8.696670294337223e-06,
      "loss": 0.4595,
      "step": 10400,
      "task_loss": 0.8145915865898132
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34814390540122986,
      "epoch": 3.76,
      "learning_rate": 8.648760751747671e-06,
      "loss": 0.5126,
      "step": 10410,
      "task_loss": 0.520122766494751
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41130712628364563,
      "epoch": 3.77,
      "learning_rate": 8.600961303327416e-06,
      "loss": 0.4893,
      "step": 10420,
      "task_loss": 0.43144863843917847
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34951236844062805,
      "epoch": 3.77,
      "learning_rate": 8.553272195546463e-06,
      "loss": 0.459,
      "step": 10430,
      "task_loss": 0.6265913248062134
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4187074899673462,
      "epoch": 3.77,
      "learning_rate": 8.505693674305892e-06,
      "loss": 0.4915,
      "step": 10440,
      "task_loss": 0.4581117033958435
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4234219193458557,
      "epoch": 3.78,
      "learning_rate": 8.458225984936562e-06,
      "loss": 0.4749,
      "step": 10450,
      "task_loss": 0.49265989661216736
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5711984634399414,
      "epoch": 3.78,
      "learning_rate": 8.410869372197849e-06,
      "loss": 0.4633,
      "step": 10460,
      "task_loss": 0.7391009330749512
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6687216758728027,
      "epoch": 3.78,
      "learning_rate": 8.363624080276351e-06,
      "loss": 0.516,
      "step": 10470,
      "task_loss": 0.6271883249282837
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4507598876953125,
      "epoch": 3.79,
      "learning_rate": 8.316490352784698e-06,
      "loss": 0.4851,
      "step": 10480,
      "task_loss": 0.7498517036437988
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.376162052154541,
      "epoch": 3.79,
      "learning_rate": 8.269468432760227e-06,
      "loss": 0.4669,
      "step": 10490,
      "task_loss": 0.399840384721756
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39421695470809937,
      "epoch": 3.79,
      "learning_rate": 8.222558562663788e-06,
      "loss": 0.5207,
      "step": 10500,
      "task_loss": 0.44343698024749756
    },
    {
      "epoch": 3.79,
      "eval_exact_match": 82.3841059602649,
      "eval_f1": 89.54316499056209,
      "step": 10500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4426266551017761,
      "epoch": 3.8,
      "learning_rate": 8.175760984378414e-06,
      "loss": 0.4776,
      "step": 10510,
      "task_loss": 0.5466393232345581
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4493921101093292,
      "epoch": 3.8,
      "learning_rate": 8.129075939208192e-06,
      "loss": 0.4836,
      "step": 10520,
      "task_loss": 0.5781248211860657
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44374221563339233,
      "epoch": 3.81,
      "learning_rate": 8.082503667876887e-06,
      "loss": 0.5203,
      "step": 10530,
      "task_loss": 0.3778543472290039
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43476569652557373,
      "epoch": 3.81,
      "learning_rate": 8.036044410526801e-06,
      "loss": 0.4235,
      "step": 10540,
      "task_loss": 0.45925962924957275
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6843395829200745,
      "epoch": 3.81,
      "learning_rate": 7.989698406717485e-06,
      "loss": 0.476,
      "step": 10550,
      "task_loss": 0.8984243869781494
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4233591556549072,
      "epoch": 3.82,
      "learning_rate": 7.943465895424528e-06,
      "loss": 0.4784,
      "step": 10560,
      "task_loss": 0.6609106063842773
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35403668880462646,
      "epoch": 3.82,
      "learning_rate": 7.897347115038289e-06,
      "loss": 0.4811,
      "step": 10570,
      "task_loss": 0.6570682525634766
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45888036489486694,
      "epoch": 3.82,
      "learning_rate": 7.851342303362713e-06,
      "loss": 0.4827,
      "step": 10580,
      "task_loss": 0.8530523777008057
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.538689374923706,
      "epoch": 3.83,
      "learning_rate": 7.805451697614085e-06,
      "loss": 0.485,
      "step": 10590,
      "task_loss": 0.692087709903717
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4368409812450409,
      "epoch": 3.83,
      "learning_rate": 7.759675534419805e-06,
      "loss": 0.546,
      "step": 10600,
      "task_loss": 0.5097652077674866
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4115678071975708,
      "epoch": 3.83,
      "learning_rate": 7.714014049817136e-06,
      "loss": 0.4959,
      "step": 10610,
      "task_loss": 0.3430953919887543
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3076716959476471,
      "epoch": 3.84,
      "learning_rate": 7.668467479252084e-06,
      "loss": 0.4085,
      "step": 10620,
      "task_loss": 0.44386640191078186
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37531307339668274,
      "epoch": 3.84,
      "learning_rate": 7.623036057578053e-06,
      "loss": 0.4485,
      "step": 10630,
      "task_loss": 0.4283076524734497
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4523438811302185,
      "epoch": 3.85,
      "learning_rate": 7.5777200190547465e-06,
      "loss": 0.5024,
      "step": 10640,
      "task_loss": 0.8270664215087891
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40491363406181335,
      "epoch": 3.85,
      "learning_rate": 7.532519597346889e-06,
      "loss": 0.5433,
      "step": 10650,
      "task_loss": 0.5985238552093506
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4389151334762573,
      "epoch": 3.85,
      "learning_rate": 7.487435025523069e-06,
      "loss": 0.4752,
      "step": 10660,
      "task_loss": 0.7840195894241333
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4390360713005066,
      "epoch": 3.86,
      "learning_rate": 7.442466536054479e-06,
      "loss": 0.4214,
      "step": 10670,
      "task_loss": 1.005985140800476
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4800419211387634,
      "epoch": 3.86,
      "learning_rate": 7.397614360813781e-06,
      "loss": 0.4512,
      "step": 10680,
      "task_loss": 0.5574986338615417
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5794166326522827,
      "epoch": 3.86,
      "learning_rate": 7.352878731073872e-06,
      "loss": 0.4872,
      "step": 10690,
      "task_loss": 0.7646304368972778
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38002362847328186,
      "epoch": 3.87,
      "learning_rate": 7.308259877506705e-06,
      "loss": 0.5005,
      "step": 10700,
      "task_loss": 0.547897219657898
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.377015620470047,
      "epoch": 3.87,
      "learning_rate": 7.26375803018209e-06,
      "loss": 0.4254,
      "step": 10710,
      "task_loss": 0.2385042905807495
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5748364925384521,
      "epoch": 3.87,
      "learning_rate": 7.219373418566526e-06,
      "loss": 0.5412,
      "step": 10720,
      "task_loss": 0.6794745326042175
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34330910444259644,
      "epoch": 3.88,
      "learning_rate": 7.175106271521979e-06,
      "loss": 0.4973,
      "step": 10730,
      "task_loss": 0.5638716220855713
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5264981985092163,
      "epoch": 3.88,
      "learning_rate": 7.130956817304751e-06,
      "loss": 0.4934,
      "step": 10740,
      "task_loss": 0.7932864427566528
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4208604693412781,
      "epoch": 3.89,
      "learning_rate": 7.0869252835642775e-06,
      "loss": 0.407,
      "step": 10750,
      "task_loss": 0.8152490854263306
    },
    {
      "epoch": 3.89,
      "eval_exact_match": 82.64900662251655,
      "eval_f1": 89.68607828698693,
      "step": 10750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.846092939376831,
      "epoch": 3.89,
      "learning_rate": 7.043011897341959e-06,
      "loss": 0.5306,
      "step": 10760,
      "task_loss": 1.4170434474945068
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5525998473167419,
      "epoch": 3.89,
      "learning_rate": 6.999216885069956e-06,
      "loss": 0.4667,
      "step": 10770,
      "task_loss": 0.9994626045227051
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4946367144584656,
      "epoch": 3.9,
      "learning_rate": 6.955540472570116e-06,
      "loss": 0.4602,
      "step": 10780,
      "task_loss": 0.696071982383728
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.529938817024231,
      "epoch": 3.9,
      "learning_rate": 6.911982885052676e-06,
      "loss": 0.4975,
      "step": 10790,
      "task_loss": 0.6633660197257996
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5889250040054321,
      "epoch": 3.9,
      "learning_rate": 6.868544347115216e-06,
      "loss": 0.5195,
      "step": 10800,
      "task_loss": 0.9271608591079712
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5101796388626099,
      "epoch": 3.91,
      "learning_rate": 6.82522508274145e-06,
      "loss": 0.4942,
      "step": 10810,
      "task_loss": 0.4550142288208008
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41657310724258423,
      "epoch": 3.91,
      "learning_rate": 6.7820253153000725e-06,
      "loss": 0.453,
      "step": 10820,
      "task_loss": 0.6524810791015625
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42488449811935425,
      "epoch": 3.91,
      "learning_rate": 6.7389452675436e-06,
      "loss": 0.4542,
      "step": 10830,
      "task_loss": 0.4262767732143402
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4936716854572296,
      "epoch": 3.92,
      "learning_rate": 6.695985161607251e-06,
      "loss": 0.4998,
      "step": 10840,
      "task_loss": 0.932105541229248
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6871744990348816,
      "epoch": 3.92,
      "learning_rate": 6.6531452190077755e-06,
      "loss": 0.5045,
      "step": 10850,
      "task_loss": 1.0812252759933472
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3607264459133148,
      "epoch": 3.92,
      "learning_rate": 6.610425660642335e-06,
      "loss": 0.4217,
      "step": 10860,
      "task_loss": 0.3593289852142334
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7296729683876038,
      "epoch": 3.93,
      "learning_rate": 6.567826706787313e-06,
      "loss": 0.4889,
      "step": 10870,
      "task_loss": 0.8488539457321167
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5310675501823425,
      "epoch": 3.93,
      "learning_rate": 6.525348577097271e-06,
      "loss": 0.5384,
      "step": 10880,
      "task_loss": 0.5912023782730103
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4117245376110077,
      "epoch": 3.94,
      "learning_rate": 6.482991490603705e-06,
      "loss": 0.4646,
      "step": 10890,
      "task_loss": 0.33494138717651367
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39375028014183044,
      "epoch": 3.94,
      "learning_rate": 6.440755665714006e-06,
      "loss": 0.5194,
      "step": 10900,
      "task_loss": 0.352378249168396
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5667198896408081,
      "epoch": 3.94,
      "learning_rate": 6.398641320210292e-06,
      "loss": 0.4856,
      "step": 10910,
      "task_loss": 0.6015273332595825
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.408721923828125,
      "epoch": 3.95,
      "learning_rate": 6.356648671248292e-06,
      "loss": 0.5528,
      "step": 10920,
      "task_loss": 0.5410779714584351
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4362407326698303,
      "epoch": 3.95,
      "learning_rate": 6.314777935356211e-06,
      "loss": 0.4703,
      "step": 10930,
      "task_loss": 0.7694103717803955
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42711496353149414,
      "epoch": 3.95,
      "learning_rate": 6.273029328433653e-06,
      "loss": 0.4835,
      "step": 10940,
      "task_loss": 0.690986692905426
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49706920981407166,
      "epoch": 3.96,
      "learning_rate": 6.231403065750473e-06,
      "loss": 0.4767,
      "step": 10950,
      "task_loss": 0.5795883536338806
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44640737771987915,
      "epoch": 3.96,
      "learning_rate": 6.189899361945688e-06,
      "loss": 0.4741,
      "step": 10960,
      "task_loss": 0.7339274883270264
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5700512528419495,
      "epoch": 3.96,
      "learning_rate": 6.148518431026324e-06,
      "loss": 0.4344,
      "step": 10970,
      "task_loss": 0.7164508700370789
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3941676616668701,
      "epoch": 3.97,
      "learning_rate": 6.10726048636641e-06,
      "loss": 0.4584,
      "step": 10980,
      "task_loss": 0.9847264885902405
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5191637277603149,
      "epoch": 3.97,
      "learning_rate": 6.066125740705759e-06,
      "loss": 0.4964,
      "step": 10990,
      "task_loss": 0.5576518774032593
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4768173098564148,
      "epoch": 3.98,
      "learning_rate": 6.02511440614896e-06,
      "loss": 0.4876,
      "step": 11000,
      "task_loss": 0.7509768009185791
    },
    {
      "epoch": 3.98,
      "eval_exact_match": 82.52601702932829,
      "eval_f1": 89.64100301643371,
      "step": 11000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5284594297409058,
      "epoch": 3.98,
      "learning_rate": 5.984226694164241e-06,
      "loss": 0.5149,
      "step": 11010,
      "task_loss": 0.827378511428833
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.495005339384079,
      "epoch": 3.98,
      "learning_rate": 5.9434628155824066e-06,
      "loss": 0.5033,
      "step": 11020,
      "task_loss": 0.8550190925598145
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43676552176475525,
      "epoch": 3.99,
      "learning_rate": 5.902822980595704e-06,
      "loss": 0.508,
      "step": 11030,
      "task_loss": 0.7428649663925171
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.48685529828071594,
      "epoch": 3.99,
      "learning_rate": 5.8623073987567965e-06,
      "loss": 0.4647,
      "step": 11040,
      "task_loss": 0.7491306066513062
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42806217074394226,
      "epoch": 3.99,
      "learning_rate": 5.821916278977651e-06,
      "loss": 0.4953,
      "step": 11050,
      "task_loss": 0.973045825958252
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46617090702056885,
      "epoch": 4.0,
      "learning_rate": 5.781649829528467e-06,
      "loss": 0.4884,
      "step": 11060,
      "task_loss": 0.6378223299980164
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5613276958465576,
      "epoch": 4.0,
      "learning_rate": 5.741508258036575e-06,
      "loss": 0.549,
      "step": 11070,
      "task_loss": 0.9049055576324463
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3290178179740906,
      "epoch": 4.0,
      "learning_rate": 5.701491771485448e-06,
      "loss": 0.4582,
      "step": 11080,
      "task_loss": 0.7116979360580444
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5921218395233154,
      "epoch": 4.01,
      "learning_rate": 5.661600576213527e-06,
      "loss": 0.4359,
      "step": 11090,
      "task_loss": 0.6299700140953064
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44644230604171753,
      "epoch": 4.01,
      "learning_rate": 5.62183487791323e-06,
      "loss": 0.5159,
      "step": 11100,
      "task_loss": 0.5687563419342041
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38395532965660095,
      "epoch": 4.02,
      "learning_rate": 5.582194881629875e-06,
      "loss": 0.4807,
      "step": 11110,
      "task_loss": 0.677031934261322
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39592987298965454,
      "epoch": 4.02,
      "learning_rate": 5.542680791760618e-06,
      "loss": 0.4683,
      "step": 11120,
      "task_loss": 0.31899696588516235
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42465940117836,
      "epoch": 4.02,
      "learning_rate": 5.5032928120533776e-06,
      "loss": 0.4734,
      "step": 11130,
      "task_loss": 1.0328463315963745
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5964382886886597,
      "epoch": 4.03,
      "learning_rate": 5.464031145605829e-06,
      "loss": 0.4592,
      "step": 11140,
      "task_loss": 0.6344958543777466
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38633590936660767,
      "epoch": 4.03,
      "learning_rate": 5.424895994864332e-06,
      "loss": 0.4756,
      "step": 11150,
      "task_loss": 0.6650869846343994
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36032289266586304,
      "epoch": 4.03,
      "learning_rate": 5.385887561622884e-06,
      "loss": 0.4349,
      "step": 11160,
      "task_loss": 0.5552083849906921
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4399193525314331,
      "epoch": 4.04,
      "learning_rate": 5.347006047022067e-06,
      "loss": 0.4414,
      "step": 11170,
      "task_loss": 0.614956796169281
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.285051554441452,
      "epoch": 4.04,
      "learning_rate": 5.3082516515480715e-06,
      "loss": 0.43,
      "step": 11180,
      "task_loss": 0.4062471389770508
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3828444480895996,
      "epoch": 4.04,
      "learning_rate": 5.269624575031573e-06,
      "loss": 0.464,
      "step": 11190,
      "task_loss": 0.5335018634796143
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4207053780555725,
      "epoch": 4.05,
      "learning_rate": 5.231125016646768e-06,
      "loss": 0.4634,
      "step": 11200,
      "task_loss": 0.5893898010253906
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4219140410423279,
      "epoch": 4.05,
      "learning_rate": 5.192753174910334e-06,
      "loss": 0.4549,
      "step": 11210,
      "task_loss": 0.4798952341079712
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4892703890800476,
      "epoch": 4.05,
      "learning_rate": 5.154509247680397e-06,
      "loss": 0.4225,
      "step": 11220,
      "task_loss": 0.4158070981502533
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5082884430885315,
      "epoch": 4.06,
      "learning_rate": 5.116393432155484e-06,
      "loss": 0.4373,
      "step": 11230,
      "task_loss": 0.8525248765945435
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45040473341941833,
      "epoch": 4.06,
      "learning_rate": 5.078405924873568e-06,
      "loss": 0.4378,
      "step": 11240,
      "task_loss": 0.8633182644844055
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4001632332801819,
      "epoch": 4.07,
      "learning_rate": 5.040546921711011e-06,
      "loss": 0.4203,
      "step": 11250,
      "task_loss": 0.9423239827156067
    },
    {
      "epoch": 4.07,
      "eval_exact_match": 82.61116367076632,
      "eval_f1": 89.66884691314182,
      "step": 11250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33361560106277466,
      "epoch": 4.07,
      "learning_rate": 5.002816617881575e-06,
      "loss": 0.4725,
      "step": 11260,
      "task_loss": 0.45950907468795776
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3659554123878479,
      "epoch": 4.07,
      "learning_rate": 4.965215207935365e-06,
      "loss": 0.4997,
      "step": 11270,
      "task_loss": 0.5733004808425903
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41657739877700806,
      "epoch": 4.08,
      "learning_rate": 4.931484303514092e-06,
      "loss": 0.4386,
      "step": 11280,
      "task_loss": 0.5786995887756348
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4608357846736908,
      "epoch": 4.08,
      "learning_rate": 4.8941283255505295e-06,
      "loss": 0.4184,
      "step": 11290,
      "task_loss": 0.7504065036773682
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42020905017852783,
      "epoch": 4.08,
      "learning_rate": 4.856901801903594e-06,
      "loss": 0.4813,
      "step": 11300,
      "task_loss": 0.8636704683303833
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6672564744949341,
      "epoch": 4.09,
      "learning_rate": 4.8198049245257615e-06,
      "loss": 0.4454,
      "step": 11310,
      "task_loss": 1.0013171434402466
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4561736583709717,
      "epoch": 4.09,
      "learning_rate": 4.782837884701011e-06,
      "loss": 0.4985,
      "step": 11320,
      "task_loss": 0.4815346598625183
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4631105065345764,
      "epoch": 4.09,
      "learning_rate": 4.746000873043818e-06,
      "loss": 0.4496,
      "step": 11330,
      "task_loss": 0.5157427787780762
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4157012403011322,
      "epoch": 4.1,
      "learning_rate": 4.709294079498207e-06,
      "loss": 0.4962,
      "step": 11340,
      "task_loss": 0.353486031293869
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41181379556655884,
      "epoch": 4.1,
      "learning_rate": 4.672717693336749e-06,
      "loss": 0.4367,
      "step": 11350,
      "task_loss": 0.79695725440979
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4717938303947449,
      "epoch": 4.11,
      "learning_rate": 4.6362719031596e-06,
      "loss": 0.4518,
      "step": 11360,
      "task_loss": 0.6884281635284424
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.430459201335907,
      "epoch": 4.11,
      "learning_rate": 4.5999568968934854e-06,
      "loss": 0.4505,
      "step": 11370,
      "task_loss": 0.7808297872543335
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42840248346328735,
      "epoch": 4.11,
      "learning_rate": 4.5637728617908196e-06,
      "loss": 0.4285,
      "step": 11380,
      "task_loss": 1.0069183111190796
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38663750886917114,
      "epoch": 4.12,
      "learning_rate": 4.527719984428635e-06,
      "loss": 0.4425,
      "step": 11390,
      "task_loss": 0.5139755010604858
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5585314035415649,
      "epoch": 4.12,
      "learning_rate": 4.491798450707693e-06,
      "loss": 0.465,
      "step": 11400,
      "task_loss": 0.5836506485939026
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42551636695861816,
      "epoch": 4.12,
      "learning_rate": 4.456008445851504e-06,
      "loss": 0.4585,
      "step": 11410,
      "task_loss": 0.4914487600326538
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5012587904930115,
      "epoch": 4.13,
      "learning_rate": 4.420350154405373e-06,
      "loss": 0.4651,
      "step": 11420,
      "task_loss": 0.9073992371559143
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49892178177833557,
      "epoch": 4.13,
      "learning_rate": 4.384823760235432e-06,
      "loss": 0.4523,
      "step": 11430,
      "task_loss": 0.6013965606689453
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4007934331893921,
      "epoch": 4.13,
      "learning_rate": 4.3494294465277165e-06,
      "loss": 0.4745,
      "step": 11440,
      "task_loss": 0.5818295478820801
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4963274896144867,
      "epoch": 4.14,
      "learning_rate": 4.314167395787213e-06,
      "loss": 0.4859,
      "step": 11450,
      "task_loss": 0.552923858165741
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46873417496681213,
      "epoch": 4.14,
      "learning_rate": 4.279037789836915e-06,
      "loss": 0.5081,
      "step": 11460,
      "task_loss": 0.44107770919799805
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6212577819824219,
      "epoch": 4.15,
      "learning_rate": 4.244040809816882e-06,
      "loss": 0.5153,
      "step": 11470,
      "task_loss": 0.7792346477508545
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3834245800971985,
      "epoch": 4.15,
      "learning_rate": 4.209176636183313e-06,
      "loss": 0.4428,
      "step": 11480,
      "task_loss": 0.35695308446884155
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4868013858795166,
      "epoch": 4.15,
      "learning_rate": 4.174445448707604e-06,
      "loss": 0.4634,
      "step": 11490,
      "task_loss": 0.5800308585166931
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4813215434551239,
      "epoch": 4.16,
      "learning_rate": 4.139847426475443e-06,
      "loss": 0.4672,
      "step": 11500,
      "task_loss": 0.6517878174781799
    },
    {
      "epoch": 4.16,
      "eval_exact_match": 82.51655629139073,
      "eval_f1": 89.63512158721852,
      "step": 11500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45381027460098267,
      "epoch": 4.16,
      "learning_rate": 4.105382747885863e-06,
      "loss": 0.4824,
      "step": 11510,
      "task_loss": 0.8077258467674255
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47750887274742126,
      "epoch": 4.16,
      "learning_rate": 4.071051590650343e-06,
      "loss": 0.4829,
      "step": 11520,
      "task_loss": 0.5288639664649963
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.417965829372406,
      "epoch": 4.17,
      "learning_rate": 4.036854131791856e-06,
      "loss": 0.4444,
      "step": 11530,
      "task_loss": 0.3982042670249939
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4850964844226837,
      "epoch": 4.17,
      "learning_rate": 4.002790547644029e-06,
      "loss": 0.4159,
      "step": 11540,
      "task_loss": 0.42431122064590454
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4070306420326233,
      "epoch": 4.17,
      "learning_rate": 3.96886101385013e-06,
      "loss": 0.4419,
      "step": 11550,
      "task_loss": 0.5825086832046509
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.358112633228302,
      "epoch": 4.18,
      "learning_rate": 3.935065705362258e-06,
      "loss": 0.4052,
      "step": 11560,
      "task_loss": 0.5550183057785034
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49228614568710327,
      "epoch": 4.18,
      "learning_rate": 3.9014047964403805e-06,
      "loss": 0.431,
      "step": 11570,
      "task_loss": 0.819994330406189
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5448801517486572,
      "epoch": 4.19,
      "learning_rate": 3.867878460651475e-06,
      "loss": 0.4863,
      "step": 11580,
      "task_loss": 0.5822548866271973
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44271165132522583,
      "epoch": 4.19,
      "learning_rate": 3.834486870868585e-06,
      "loss": 0.4002,
      "step": 11590,
      "task_loss": 0.501520574092865
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40002530813217163,
      "epoch": 4.19,
      "learning_rate": 3.8012301992699906e-06,
      "loss": 0.4177,
      "step": 11600,
      "task_loss": 0.6586291790008545
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3801167607307434,
      "epoch": 4.2,
      "learning_rate": 3.768108617338265e-06,
      "loss": 0.4417,
      "step": 11610,
      "task_loss": 0.7325814962387085
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4070740342140198,
      "epoch": 4.2,
      "learning_rate": 3.735122295859431e-06,
      "loss": 0.4995,
      "step": 11620,
      "task_loss": 0.35257917642593384
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4214155673980713,
      "epoch": 4.2,
      "learning_rate": 3.7022714049220387e-06,
      "loss": 0.4334,
      "step": 11630,
      "task_loss": 0.5078192949295044
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.444728285074234,
      "epoch": 4.21,
      "learning_rate": 3.669556113916349e-06,
      "loss": 0.4291,
      "step": 11640,
      "task_loss": 0.6134557127952576
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6096001863479614,
      "epoch": 4.21,
      "learning_rate": 3.6369765915333876e-06,
      "loss": 0.4695,
      "step": 11650,
      "task_loss": 0.7519853115081787
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4547508955001831,
      "epoch": 4.21,
      "learning_rate": 3.6045330057641344e-06,
      "loss": 0.4402,
      "step": 11660,
      "task_loss": 0.5918570756912231
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5541609525680542,
      "epoch": 4.22,
      "learning_rate": 3.5722255238986255e-06,
      "loss": 0.464,
      "step": 11670,
      "task_loss": 0.573555588722229
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3390219807624817,
      "epoch": 4.22,
      "learning_rate": 3.5400543125251037e-06,
      "loss": 0.4442,
      "step": 11680,
      "task_loss": 0.5026131272315979
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34992164373397827,
      "epoch": 4.22,
      "learning_rate": 3.5112168706785487e-06,
      "loss": 0.4619,
      "step": 11690,
      "task_loss": 0.9039815664291382
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42350032925605774,
      "epoch": 4.23,
      "learning_rate": 3.4793050296731365e-06,
      "loss": 0.4914,
      "step": 11700,
      "task_loss": 0.5576711893081665
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.432574987411499,
      "epoch": 4.23,
      "learning_rate": 3.447529938289038e-06,
      "loss": 0.455,
      "step": 11710,
      "task_loss": 0.30313971638679504
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3797840476036072,
      "epoch": 4.24,
      "learning_rate": 3.4158917603692885e-06,
      "loss": 0.4161,
      "step": 11720,
      "task_loss": 0.35903841257095337
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.405576229095459,
      "epoch": 4.24,
      "learning_rate": 3.384390659050979e-06,
      "loss": 0.4451,
      "step": 11730,
      "task_loss": 0.584861695766449
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.51385897397995,
      "epoch": 4.24,
      "learning_rate": 3.353026796764378e-06,
      "loss": 0.4533,
      "step": 11740,
      "task_loss": 0.6303707361221313
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4929048418998718,
      "epoch": 4.25,
      "learning_rate": 3.321800335232118e-06,
      "loss": 0.5089,
      "step": 11750,
      "task_loss": 0.545328676700592
    },
    {
      "epoch": 4.25,
      "eval_exact_match": 82.50709555345317,
      "eval_f1": 89.55435656315157,
      "step": 11750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3543594181537628,
      "epoch": 4.25,
      "learning_rate": 3.2907114354683175e-06,
      "loss": 0.4477,
      "step": 11760,
      "task_loss": 0.4058670699596405
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6473177075386047,
      "epoch": 4.25,
      "learning_rate": 3.2597602577778065e-06,
      "loss": 0.4899,
      "step": 11770,
      "task_loss": 0.8452356457710266
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.48143479228019714,
      "epoch": 4.26,
      "learning_rate": 3.2289469617552613e-06,
      "loss": 0.4521,
      "step": 11780,
      "task_loss": 0.6364303231239319
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3805115818977356,
      "epoch": 4.26,
      "learning_rate": 3.198271706284409e-06,
      "loss": 0.4724,
      "step": 11790,
      "task_loss": 0.5545732975006104
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42327409982681274,
      "epoch": 4.26,
      "learning_rate": 3.1677346495371616e-06,
      "loss": 0.4401,
      "step": 11800,
      "task_loss": 0.3729352653026581
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49267518520355225,
      "epoch": 4.27,
      "learning_rate": 3.1373359489728783e-06,
      "loss": 0.4854,
      "step": 11810,
      "task_loss": 0.7921924591064453
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4918190538883209,
      "epoch": 4.27,
      "learning_rate": 3.107075761337458e-06,
      "loss": 0.4366,
      "step": 11820,
      "task_loss": 0.6186762452125549
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3143673539161682,
      "epoch": 4.28,
      "learning_rate": 3.076954242662615e-06,
      "loss": 0.4444,
      "step": 11830,
      "task_loss": 0.41375160217285156
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4199334979057312,
      "epoch": 4.28,
      "learning_rate": 3.0469715482650264e-06,
      "loss": 0.4666,
      "step": 11840,
      "task_loss": 0.3213299512863159
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45714980363845825,
      "epoch": 4.28,
      "learning_rate": 3.01712783274555e-06,
      "loss": 0.4897,
      "step": 11850,
      "task_loss": 0.6414152383804321
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4583025276660919,
      "epoch": 4.29,
      "learning_rate": 2.987423249988411e-06,
      "loss": 0.4416,
      "step": 11860,
      "task_loss": 0.6744894981384277
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4339994788169861,
      "epoch": 4.29,
      "learning_rate": 2.9578579531604335e-06,
      "loss": 0.5102,
      "step": 11870,
      "task_loss": 0.7543153762817383
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5392136573791504,
      "epoch": 4.29,
      "learning_rate": 2.9284320947102227e-06,
      "loss": 0.4107,
      "step": 11880,
      "task_loss": 0.8216215372085571
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3788997232913971,
      "epoch": 4.3,
      "learning_rate": 2.899145826367412e-06,
      "loss": 0.4469,
      "step": 11890,
      "task_loss": 1.0387654304504395
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.596594512462616,
      "epoch": 4.3,
      "learning_rate": 2.869999299141829e-06,
      "loss": 0.4655,
      "step": 11900,
      "task_loss": 0.8393262624740601
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5400757789611816,
      "epoch": 4.3,
      "learning_rate": 2.8409926633227947e-06,
      "loss": 0.4444,
      "step": 11910,
      "task_loss": 0.5543546676635742
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3381040394306183,
      "epoch": 4.31,
      "learning_rate": 2.8121260684782567e-06,
      "loss": 0.4609,
      "step": 11920,
      "task_loss": 0.7791619300842285
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4478145241737366,
      "epoch": 4.31,
      "learning_rate": 2.7833996634540914e-06,
      "loss": 0.4413,
      "step": 11930,
      "task_loss": 0.5504406094551086
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3867071866989136,
      "epoch": 4.32,
      "learning_rate": 2.7548135963733057e-06,
      "loss": 0.4504,
      "step": 11940,
      "task_loss": 0.35446274280548096
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4211112856864929,
      "epoch": 4.32,
      "learning_rate": 2.726368014635275e-06,
      "loss": 0.3832,
      "step": 11950,
      "task_loss": 0.7330440282821655
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.450991690158844,
      "epoch": 4.32,
      "learning_rate": 2.6980630649149797e-06,
      "loss": 0.4979,
      "step": 11960,
      "task_loss": 0.9241546392440796
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.427333265542984,
      "epoch": 4.33,
      "learning_rate": 2.669898893162257e-06,
      "loss": 0.4731,
      "step": 11970,
      "task_loss": 0.7334015369415283
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6288331747055054,
      "epoch": 4.33,
      "learning_rate": 2.641875644601047e-06,
      "loss": 0.4594,
      "step": 11980,
      "task_loss": 0.7425243854522705
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41228044033050537,
      "epoch": 4.33,
      "learning_rate": 2.6139934637286546e-06,
      "loss": 0.4227,
      "step": 11990,
      "task_loss": 0.4954274892807007
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39298591017723083,
      "epoch": 4.34,
      "learning_rate": 2.586252494314961e-06,
      "loss": 0.461,
      "step": 12000,
      "task_loss": 0.3948507010936737
    },
    {
      "epoch": 4.34,
      "eval_exact_match": 82.52601702932829,
      "eval_f1": 89.65617956341971,
      "step": 12000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38348010182380676,
      "epoch": 4.34,
      "learning_rate": 2.558652879401753e-06,
      "loss": 0.4474,
      "step": 12010,
      "task_loss": 0.36755043268203735
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44406384229660034,
      "epoch": 4.34,
      "learning_rate": 2.531194761301907e-06,
      "loss": 0.3773,
      "step": 12020,
      "task_loss": 0.4231364130973816
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3030942678451538,
      "epoch": 4.35,
      "learning_rate": 2.503878281598726e-06,
      "loss": 0.4116,
      "step": 12030,
      "task_loss": 0.23304599523544312
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42190343141555786,
      "epoch": 4.35,
      "learning_rate": 2.476703581145162e-06,
      "loss": 0.4454,
      "step": 12040,
      "task_loss": 0.5864576101303101
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3552584648132324,
      "epoch": 4.35,
      "learning_rate": 2.4496708000631094e-06,
      "loss": 0.4174,
      "step": 12050,
      "task_loss": 0.28264883160591125
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3596591651439667,
      "epoch": 4.36,
      "learning_rate": 2.4227800777426746e-06,
      "loss": 0.3788,
      "step": 12060,
      "task_loss": 0.5582944750785828
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4927133619785309,
      "epoch": 4.36,
      "learning_rate": 2.396031552841462e-06,
      "loss": 0.4512,
      "step": 12070,
      "task_loss": 0.9744982123374939
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4307158589363098,
      "epoch": 4.37,
      "learning_rate": 2.369425363283865e-06,
      "loss": 0.4468,
      "step": 12080,
      "task_loss": 0.7413559556007385
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5698738098144531,
      "epoch": 4.37,
      "learning_rate": 2.3429616462603477e-06,
      "loss": 0.4575,
      "step": 12090,
      "task_loss": 0.5306029915809631
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44272828102111816,
      "epoch": 4.37,
      "learning_rate": 2.316640538226721e-06,
      "loss": 0.4357,
      "step": 12100,
      "task_loss": 0.6407321691513062
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3783756196498871,
      "epoch": 4.38,
      "learning_rate": 2.290462174903486e-06,
      "loss": 0.4917,
      "step": 12110,
      "task_loss": 0.6652238368988037
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4510440230369568,
      "epoch": 4.38,
      "learning_rate": 2.2644266912750733e-06,
      "loss": 0.4379,
      "step": 12120,
      "task_loss": 0.5491682291030884
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6077476739883423,
      "epoch": 4.38,
      "learning_rate": 2.238534221589196e-06,
      "loss": 0.4289,
      "step": 12130,
      "task_loss": 0.8977972269058228
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5110172033309937,
      "epoch": 4.39,
      "learning_rate": 2.212784899356136e-06,
      "loss": 0.4378,
      "step": 12140,
      "task_loss": 0.40151098370552063
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4892664849758148,
      "epoch": 4.39,
      "learning_rate": 2.187178857348061e-06,
      "loss": 0.4869,
      "step": 12150,
      "task_loss": 0.7194719314575195
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3257119953632355,
      "epoch": 4.39,
      "learning_rate": 2.1617162275983217e-06,
      "loss": 0.3906,
      "step": 12160,
      "task_loss": 0.4725964069366455
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46231818199157715,
      "epoch": 4.4,
      "learning_rate": 2.1363971414008097e-06,
      "loss": 0.4395,
      "step": 12170,
      "task_loss": 0.3184768557548523
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6130329966545105,
      "epoch": 4.4,
      "learning_rate": 2.1112217293092405e-06,
      "loss": 0.4499,
      "step": 12180,
      "task_loss": 0.6699998378753662
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.48698151111602783,
      "epoch": 4.41,
      "learning_rate": 2.0861901211365177e-06,
      "loss": 0.431,
      "step": 12190,
      "task_loss": 0.5479468703269958
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6351044178009033,
      "epoch": 4.41,
      "learning_rate": 2.0613024459540076e-06,
      "loss": 0.4908,
      "step": 12200,
      "task_loss": 0.5488524436950684
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47335565090179443,
      "epoch": 4.41,
      "learning_rate": 2.0365588320909576e-06,
      "loss": 0.4721,
      "step": 12210,
      "task_loss": 0.5685126781463623
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36355626583099365,
      "epoch": 4.42,
      "learning_rate": 2.0119594071337433e-06,
      "loss": 0.3822,
      "step": 12220,
      "task_loss": 0.5155311822891235
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5141717195510864,
      "epoch": 4.42,
      "learning_rate": 1.98750429792528e-06,
      "loss": 0.4493,
      "step": 12230,
      "task_loss": 0.6456845998764038
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3986155092716217,
      "epoch": 4.42,
      "learning_rate": 1.9631936305643294e-06,
      "loss": 0.4575,
      "step": 12240,
      "task_loss": 0.7241479158401489
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4736446142196655,
      "epoch": 4.43,
      "learning_rate": 1.9390275304048755e-06,
      "loss": 0.4109,
      "step": 12250,
      "task_loss": 0.35844236612319946
    },
    {
      "epoch": 4.43,
      "eval_exact_match": 82.72469252601702,
      "eval_f1": 89.86566403216932,
      "step": 12250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46747684478759766,
      "epoch": 4.43,
      "learning_rate": 1.915006122055445e-06,
      "loss": 0.4611,
      "step": 12260,
      "task_loss": 0.9917988777160645
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31288278102874756,
      "epoch": 4.43,
      "learning_rate": 1.891129529378508e-06,
      "loss": 0.4293,
      "step": 12270,
      "task_loss": 0.5523203015327454
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4253838360309601,
      "epoch": 4.44,
      "learning_rate": 1.867397875489799e-06,
      "loss": 0.4312,
      "step": 12280,
      "task_loss": 0.4688680171966553
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3778352737426758,
      "epoch": 4.44,
      "learning_rate": 1.8438112827577068e-06,
      "loss": 0.4365,
      "step": 12290,
      "task_loss": 0.704608142375946
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43521445989608765,
      "epoch": 4.45,
      "learning_rate": 1.8203698728026386e-06,
      "loss": 0.4443,
      "step": 12300,
      "task_loss": 0.73112952709198
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3996022641658783,
      "epoch": 4.45,
      "learning_rate": 1.7970737664963832e-06,
      "loss": 0.4243,
      "step": 12310,
      "task_loss": 0.5031775832176208
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47173169255256653,
      "epoch": 4.45,
      "learning_rate": 1.7739230839614962e-06,
      "loss": 0.4668,
      "step": 12320,
      "task_loss": 0.6459155082702637
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38863933086395264,
      "epoch": 4.46,
      "learning_rate": 1.7509179445706858e-06,
      "loss": 0.4154,
      "step": 12330,
      "task_loss": 0.5454069375991821
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5435532331466675,
      "epoch": 4.46,
      "learning_rate": 1.7280584669461808e-06,
      "loss": 0.4891,
      "step": 12340,
      "task_loss": 0.40724098682403564
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44464439153671265,
      "epoch": 4.46,
      "learning_rate": 1.7053447689591473e-06,
      "loss": 0.4487,
      "step": 12350,
      "task_loss": 0.31382906436920166
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3412976861000061,
      "epoch": 4.47,
      "learning_rate": 1.6827769677290294e-06,
      "loss": 0.4525,
      "step": 12360,
      "task_loss": 0.5036389827728271
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3674119710922241,
      "epoch": 4.47,
      "learning_rate": 1.6603551796230232e-06,
      "loss": 0.4865,
      "step": 12370,
      "task_loss": 0.5364322662353516
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3727283477783203,
      "epoch": 4.47,
      "learning_rate": 1.6380795202553866e-06,
      "loss": 0.4124,
      "step": 12380,
      "task_loss": 0.6606693863868713
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5182126760482788,
      "epoch": 4.48,
      "learning_rate": 1.615950104486924e-06,
      "loss": 0.4459,
      "step": 12390,
      "task_loss": 0.6920464038848877
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4384608864784241,
      "epoch": 4.48,
      "learning_rate": 1.5939670464243362e-06,
      "loss": 0.486,
      "step": 12400,
      "task_loss": 0.5016846656799316
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3930404484272003,
      "epoch": 4.49,
      "learning_rate": 1.572130459419674e-06,
      "loss": 0.4514,
      "step": 12410,
      "task_loss": 0.5308805704116821
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5489310622215271,
      "epoch": 4.49,
      "learning_rate": 1.5504404560697093e-06,
      "loss": 0.4661,
      "step": 12420,
      "task_loss": 0.7806792259216309
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6114611625671387,
      "epoch": 4.49,
      "learning_rate": 1.5288971482153957e-06,
      "loss": 0.4664,
      "step": 12430,
      "task_loss": 0.9652453660964966
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32875922322273254,
      "epoch": 4.5,
      "learning_rate": 1.5075006469412778e-06,
      "loss": 0.4655,
      "step": 12440,
      "task_loss": 0.2918616533279419
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3842153549194336,
      "epoch": 4.5,
      "learning_rate": 1.486251062574916e-06,
      "loss": 0.4458,
      "step": 12450,
      "task_loss": 0.4864546060562134
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5039974451065063,
      "epoch": 4.5,
      "learning_rate": 1.4651485046862933e-06,
      "loss": 0.4366,
      "step": 12460,
      "task_loss": 0.816196858882904
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3842373192310333,
      "epoch": 4.51,
      "learning_rate": 1.4441930820873195e-06,
      "loss": 0.4098,
      "step": 12470,
      "task_loss": 0.7222305536270142
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5461684465408325,
      "epoch": 4.51,
      "learning_rate": 1.4233849028311808e-06,
      "loss": 0.491,
      "step": 12480,
      "task_loss": 0.6540833711624146
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38882261514663696,
      "epoch": 4.51,
      "learning_rate": 1.4027240742118542e-06,
      "loss": 0.4521,
      "step": 12490,
      "task_loss": 1.0828897953033447
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3728446364402771,
      "epoch": 4.52,
      "learning_rate": 1.3822107027635178e-06,
      "loss": 0.4369,
      "step": 12500,
      "task_loss": 0.6404829025268555
    },
    {
      "epoch": 4.52,
      "eval_exact_match": 82.67738883632923,
      "eval_f1": 89.75424790980458,
      "step": 12500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8121786713600159,
      "epoch": 4.52,
      "learning_rate": 1.3618448942600182e-06,
      "loss": 0.4755,
      "step": 12510,
      "task_loss": 0.6036579012870789
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3688916862010956,
      "epoch": 4.52,
      "learning_rate": 1.3416267537143035e-06,
      "loss": 0.4518,
      "step": 12520,
      "task_loss": 0.3554149866104126
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5656636953353882,
      "epoch": 4.53,
      "learning_rate": 1.3215563853779112e-06,
      "loss": 0.4684,
      "step": 12530,
      "task_loss": 0.6867916584014893
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3812546730041504,
      "epoch": 4.53,
      "learning_rate": 1.3016338927404047e-06,
      "loss": 0.4268,
      "step": 12540,
      "task_loss": 0.5564362406730652
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3769798278808594,
      "epoch": 4.54,
      "learning_rate": 1.2818593785288645e-06,
      "loss": 0.4809,
      "step": 12550,
      "task_loss": 0.5600254535675049
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5981583595275879,
      "epoch": 4.54,
      "learning_rate": 1.262232944707321e-06,
      "loss": 0.4021,
      "step": 12560,
      "task_loss": 0.5983515977859497
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35735106468200684,
      "epoch": 4.54,
      "learning_rate": 1.2427546924762823e-06,
      "loss": 0.3927,
      "step": 12570,
      "task_loss": 0.3935161828994751
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4289091229438782,
      "epoch": 4.55,
      "learning_rate": 1.2234247222721573e-06,
      "loss": 0.4543,
      "step": 12580,
      "task_loss": 0.548987090587616
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42020517587661743,
      "epoch": 4.55,
      "learning_rate": 1.2042431337667704e-06,
      "loss": 0.4656,
      "step": 12590,
      "task_loss": 0.9573279619216919
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43121659755706787,
      "epoch": 4.55,
      "learning_rate": 1.1852100258668507e-06,
      "loss": 0.4789,
      "step": 12600,
      "task_loss": 0.45250964164733887
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49477794766426086,
      "epoch": 4.56,
      "learning_rate": 1.1663254967134973e-06,
      "loss": 0.4642,
      "step": 12610,
      "task_loss": 1.1415183544158936
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38228267431259155,
      "epoch": 4.56,
      "learning_rate": 1.1475896436816947e-06,
      "loss": 0.4493,
      "step": 12620,
      "task_loss": 0.3687947392463684
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.400796115398407,
      "epoch": 4.56,
      "learning_rate": 1.1290025633797973e-06,
      "loss": 0.4358,
      "step": 12630,
      "task_loss": 0.3746778964996338
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3107837438583374,
      "epoch": 4.57,
      "learning_rate": 1.1105643516490438e-06,
      "loss": 0.4588,
      "step": 12640,
      "task_loss": 0.5554065108299255
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4274769425392151,
      "epoch": 4.57,
      "learning_rate": 1.0922751035630595e-06,
      "loss": 0.4397,
      "step": 12650,
      "task_loss": 0.5595571994781494
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5575343370437622,
      "epoch": 4.58,
      "learning_rate": 1.0741349134273448e-06,
      "loss": 0.4558,
      "step": 12660,
      "task_loss": 0.8480510711669922
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4240317940711975,
      "epoch": 4.58,
      "learning_rate": 1.0561438747788377e-06,
      "loss": 0.4562,
      "step": 12670,
      "task_loss": 0.5816398859024048
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4905956983566284,
      "epoch": 4.58,
      "learning_rate": 1.0383020803853682e-06,
      "loss": 0.4599,
      "step": 12680,
      "task_loss": 0.7364996671676636
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5033512115478516,
      "epoch": 4.59,
      "learning_rate": 1.0206096222452321e-06,
      "loss": 0.4555,
      "step": 12690,
      "task_loss": 0.5982179641723633
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3769073486328125,
      "epoch": 4.59,
      "learning_rate": 1.0030665915866944e-06,
      "loss": 0.4675,
      "step": 12700,
      "task_loss": 0.6038902997970581
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5291876196861267,
      "epoch": 4.59,
      "learning_rate": 9.856730788675228e-07,
      "loss": 0.4573,
      "step": 12710,
      "task_loss": 0.5451856851577759
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3201839327812195,
      "epoch": 4.6,
      "learning_rate": 9.68429173774512e-07,
      "loss": 0.4138,
      "step": 12720,
      "task_loss": 0.4877792000770569
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4193733334541321,
      "epoch": 4.6,
      "learning_rate": 9.513349652230407e-07,
      "loss": 0.4541,
      "step": 12730,
      "task_loss": 0.6895813345909119
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3674892783164978,
      "epoch": 4.6,
      "learning_rate": 9.343905413565878e-07,
      "loss": 0.4354,
      "step": 12740,
      "task_loss": 0.5178918838500977
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49814528226852417,
      "epoch": 4.61,
      "learning_rate": 9.175959895463138e-07,
      "loss": 0.4446,
      "step": 12750,
      "task_loss": 1.101963996887207
    },
    {
      "epoch": 4.61,
      "eval_exact_match": 82.61116367076632,
      "eval_f1": 89.66291770374835,
      "step": 12750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4390224516391754,
      "epoch": 4.61,
      "learning_rate": 9.009513963905602e-07,
      "loss": 0.435,
      "step": 12760,
      "task_loss": 0.6204954385757446
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35951822996139526,
      "epoch": 4.62,
      "learning_rate": 8.844568477144644e-07,
      "loss": 0.4543,
      "step": 12770,
      "task_loss": 0.5188806056976318
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5767937898635864,
      "epoch": 4.62,
      "learning_rate": 8.681124285694486e-07,
      "loss": 0.442,
      "step": 12780,
      "task_loss": 0.7311397790908813
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4548581838607788,
      "epoch": 4.62,
      "learning_rate": 8.519182232328415e-07,
      "loss": 0.4579,
      "step": 12790,
      "task_loss": 0.5679984092712402
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5455349087715149,
      "epoch": 4.63,
      "learning_rate": 8.358743152074111e-07,
      "loss": 0.451,
      "step": 12800,
      "task_loss": 0.6041518449783325
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33094823360443115,
      "epoch": 4.63,
      "learning_rate": 8.199807872209452e-07,
      "loss": 0.39,
      "step": 12810,
      "task_loss": 0.2264111191034317
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38882723450660706,
      "epoch": 4.63,
      "learning_rate": 8.042377212258123e-07,
      "loss": 0.4509,
      "step": 12820,
      "task_loss": 0.685081958770752
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5015546083450317,
      "epoch": 4.64,
      "learning_rate": 7.886451983985576e-07,
      "loss": 0.4549,
      "step": 12830,
      "task_loss": 0.4656003713607788
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5418596267700195,
      "epoch": 4.64,
      "learning_rate": 7.73203299139471e-07,
      "loss": 0.491,
      "step": 12840,
      "task_loss": 0.481580913066864
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6147862672805786,
      "epoch": 4.64,
      "learning_rate": 7.579121030721837e-07,
      "loss": 0.4705,
      "step": 12850,
      "task_loss": 0.7846791744232178
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39587533473968506,
      "epoch": 4.65,
      "learning_rate": 7.427716890432346e-07,
      "loss": 0.464,
      "step": 12860,
      "task_loss": 0.40945491194725037
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3885082006454468,
      "epoch": 4.65,
      "learning_rate": 7.277821351216984e-07,
      "loss": 0.4694,
      "step": 12870,
      "task_loss": 0.507175087928772
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4322596788406372,
      "epoch": 4.65,
      "learning_rate": 7.129435185987487e-07,
      "loss": 0.4253,
      "step": 12880,
      "task_loss": 0.6645572185516357
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4641607999801636,
      "epoch": 4.66,
      "learning_rate": 6.982559159872881e-07,
      "loss": 0.435,
      "step": 12890,
      "task_loss": 0.5848242044448853
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42792952060699463,
      "epoch": 4.66,
      "learning_rate": 6.837194030215288e-07,
      "loss": 0.4249,
      "step": 12900,
      "task_loss": 0.37192124128341675
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36941879987716675,
      "epoch": 4.67,
      "learning_rate": 6.693340546566263e-07,
      "loss": 0.4517,
      "step": 12910,
      "task_loss": 0.48606473207473755
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5151842832565308,
      "epoch": 4.67,
      "learning_rate": 6.550999450682693e-07,
      "loss": 0.474,
      "step": 12920,
      "task_loss": 0.6494313478469849
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41002902388572693,
      "epoch": 4.67,
      "learning_rate": 6.410171476523141e-07,
      "loss": 0.4543,
      "step": 12930,
      "task_loss": 0.6349010467529297
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4869542717933655,
      "epoch": 4.68,
      "learning_rate": 6.270857350243974e-07,
      "loss": 0.4688,
      "step": 12940,
      "task_loss": 0.5892748832702637
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4189644455909729,
      "epoch": 4.68,
      "learning_rate": 6.133057790195773e-07,
      "loss": 0.4564,
      "step": 12950,
      "task_loss": 0.7822315096855164
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5754233598709106,
      "epoch": 4.68,
      "learning_rate": 5.996773506919262e-07,
      "loss": 0.457,
      "step": 12960,
      "task_loss": 0.5739059448242188
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5854544639587402,
      "epoch": 4.69,
      "learning_rate": 5.862005203142151e-07,
      "loss": 0.4339,
      "step": 12970,
      "task_loss": 0.8967189788818359
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4532425105571747,
      "epoch": 4.69,
      "learning_rate": 5.728753573775069e-07,
      "loss": 0.449,
      "step": 12980,
      "task_loss": 0.6138538122177124
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3805669844150543,
      "epoch": 4.69,
      "learning_rate": 5.597019305908235e-07,
      "loss": 0.4256,
      "step": 12990,
      "task_loss": 0.49195724725723267
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5049669742584229,
      "epoch": 4.7,
      "learning_rate": 5.466803078807859e-07,
      "loss": 0.4906,
      "step": 13000,
      "task_loss": 0.5219841003417969
    },
    {
      "epoch": 4.7,
      "eval_exact_match": 82.639545884579,
      "eval_f1": 89.75210989680092,
      "step": 13000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.327298104763031,
      "epoch": 4.7,
      "learning_rate": 5.33810556391261e-07,
      "loss": 0.461,
      "step": 13010,
      "task_loss": 0.7025531530380249
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4368177652359009,
      "epoch": 4.71,
      "learning_rate": 5.210927424830092e-07,
      "loss": 0.4248,
      "step": 13020,
      "task_loss": 0.5124354958534241
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3765318989753723,
      "epoch": 4.71,
      "learning_rate": 5.085269317333574e-07,
      "loss": 0.4322,
      "step": 13030,
      "task_loss": 0.44112464785575867
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41373926401138306,
      "epoch": 4.71,
      "learning_rate": 4.961131889358528e-07,
      "loss": 0.4575,
      "step": 13040,
      "task_loss": 1.0655057430267334
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42176538705825806,
      "epoch": 4.72,
      "learning_rate": 4.838515780999264e-07,
      "loss": 0.4261,
      "step": 13050,
      "task_loss": 0.6866442561149597
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4533381462097168,
      "epoch": 4.72,
      "learning_rate": 4.717421624505669e-07,
      "loss": 0.5023,
      "step": 13060,
      "task_loss": 0.5832757353782654
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5116744041442871,
      "epoch": 4.72,
      "learning_rate": 4.597850044279972e-07,
      "loss": 0.4487,
      "step": 13070,
      "task_loss": 0.9590458869934082
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.48838239908218384,
      "epoch": 4.73,
      "learning_rate": 4.4798016568733837e-07,
      "loss": 0.43,
      "step": 13080,
      "task_loss": 0.8301137089729309
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4273003339767456,
      "epoch": 4.73,
      "learning_rate": 4.3632770709831293e-07,
      "loss": 0.4554,
      "step": 13090,
      "task_loss": 0.5610930919647217
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47295600175857544,
      "epoch": 4.73,
      "learning_rate": 4.248276887449154e-07,
      "loss": 0.4865,
      "step": 13100,
      "task_loss": 0.6153036952018738
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47993203997612,
      "epoch": 4.74,
      "learning_rate": 4.1348016992510895e-07,
      "loss": 0.4574,
      "step": 13110,
      "task_loss": 0.5558942556381226
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39272961020469666,
      "epoch": 4.74,
      "learning_rate": 4.0228520915050915e-07,
      "loss": 0.4293,
      "step": 13120,
      "task_loss": 0.4004030227661133
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6470414400100708,
      "epoch": 4.75,
      "learning_rate": 3.912428641461041e-07,
      "loss": 0.4491,
      "step": 13130,
      "task_loss": 0.7483572959899902
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3939260244369507,
      "epoch": 4.75,
      "learning_rate": 3.8035319184993813e-07,
      "loss": 0.4283,
      "step": 13140,
      "task_loss": 0.3680633306503296
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.558957576751709,
      "epoch": 4.75,
      "learning_rate": 3.6961624841282516e-07,
      "loss": 0.4946,
      "step": 13150,
      "task_loss": 0.9486407041549683
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4826211929321289,
      "epoch": 4.76,
      "learning_rate": 3.590320891980492e-07,
      "loss": 0.435,
      "step": 13160,
      "task_loss": 0.8779466152191162
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42434561252593994,
      "epoch": 4.76,
      "learning_rate": 3.4860076878110103e-07,
      "loss": 0.4788,
      "step": 13170,
      "task_loss": 0.6615882515907288
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4156636595726013,
      "epoch": 4.76,
      "learning_rate": 3.383223409493719e-07,
      "loss": 0.4392,
      "step": 13180,
      "task_loss": 0.5900173187255859
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39588701725006104,
      "epoch": 4.77,
      "learning_rate": 3.281968587018902e-07,
      "loss": 0.4215,
      "step": 13190,
      "task_loss": 0.5002977848052979
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4153065085411072,
      "epoch": 4.77,
      "learning_rate": 3.1822437424905536e-07,
      "loss": 0.4643,
      "step": 13200,
      "task_loss": 0.506074845790863
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4411586821079254,
      "epoch": 4.77,
      "learning_rate": 3.084049390123478e-07,
      "loss": 0.4208,
      "step": 13210,
      "task_loss": 0.49065306782722473
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5167171359062195,
      "epoch": 4.78,
      "learning_rate": 2.9873860362407244e-07,
      "loss": 0.459,
      "step": 13220,
      "task_loss": 0.6649248003959656
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3746606707572937,
      "epoch": 4.78,
      "learning_rate": 2.892254179271059e-07,
      "loss": 0.4396,
      "step": 13230,
      "task_loss": 0.3278374671936035
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.482988178730011,
      "epoch": 4.78,
      "learning_rate": 2.798654309746396e-07,
      "loss": 0.4777,
      "step": 13240,
      "task_loss": 0.7824634313583374
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3645211458206177,
      "epoch": 4.79,
      "learning_rate": 2.706586910299069e-07,
      "loss": 0.423,
      "step": 13250,
      "task_loss": 0.4886607527732849
    },
    {
      "epoch": 4.79,
      "eval_exact_match": 82.56385998107852,
      "eval_f1": 89.66224189133082,
      "step": 13250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4217372536659241,
      "epoch": 4.79,
      "learning_rate": 2.616052455659568e-07,
      "loss": 0.4264,
      "step": 13260,
      "task_loss": 0.5518814325332642
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29746928811073303,
      "epoch": 4.8,
      "learning_rate": 2.5270514126540025e-07,
      "loss": 0.4416,
      "step": 13270,
      "task_loss": 0.2353348284959793
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40688103437423706,
      "epoch": 4.8,
      "learning_rate": 2.4395842402016756e-07,
      "loss": 0.4465,
      "step": 13280,
      "task_loss": 0.3618243336677551
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42863696813583374,
      "epoch": 4.8,
      "learning_rate": 2.3536513893127166e-07,
      "loss": 0.423,
      "step": 13290,
      "task_loss": 0.7000741958618164
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6176955699920654,
      "epoch": 4.81,
      "learning_rate": 2.2692533030857832e-07,
      "loss": 0.4629,
      "step": 13300,
      "task_loss": 0.5734840631484985
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.48465853929519653,
      "epoch": 4.81,
      "learning_rate": 2.1863904167058634e-07,
      "loss": 0.4697,
      "step": 13310,
      "task_loss": 1.253157377243042
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5344210863113403,
      "epoch": 4.81,
      "learning_rate": 2.1050631574418112e-07,
      "loss": 0.4755,
      "step": 13320,
      "task_loss": 0.5893040895462036
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.673054575920105,
      "epoch": 4.82,
      "learning_rate": 2.0252719446443135e-07,
      "loss": 0.5033,
      "step": 13330,
      "task_loss": 0.8126822710037231
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34842216968536377,
      "epoch": 4.82,
      "learning_rate": 1.9470171897437273e-07,
      "loss": 0.444,
      "step": 13340,
      "task_loss": 0.5029393434524536
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5423133969306946,
      "epoch": 4.82,
      "learning_rate": 1.8702992962478792e-07,
      "loss": 0.4286,
      "step": 13350,
      "task_loss": 0.41326332092285156
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43724340200424194,
      "epoch": 4.83,
      "learning_rate": 1.7951186597399693e-07,
      "loss": 0.5312,
      "step": 13360,
      "task_loss": 0.6493289470672607
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38860416412353516,
      "epoch": 4.83,
      "learning_rate": 1.7214756678767042e-07,
      "loss": 0.4399,
      "step": 13370,
      "task_loss": 0.623653769493103
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40182918310165405,
      "epoch": 4.84,
      "learning_rate": 1.649370700386099e-07,
      "loss": 0.4528,
      "step": 13380,
      "task_loss": 0.3763955235481262
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34956735372543335,
      "epoch": 4.84,
      "learning_rate": 1.578804129065614e-07,
      "loss": 0.4706,
      "step": 13390,
      "task_loss": 0.48613160848617554
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5024135112762451,
      "epoch": 4.84,
      "learning_rate": 1.5097763177802205e-07,
      "loss": 0.463,
      "step": 13400,
      "task_loss": 0.7361462712287903
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3231491148471832,
      "epoch": 4.85,
      "learning_rate": 1.4422876224605365e-07,
      "loss": 0.4349,
      "step": 13410,
      "task_loss": 0.4247916638851166
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6355755925178528,
      "epoch": 4.85,
      "learning_rate": 1.3763383911009287e-07,
      "loss": 0.4766,
      "step": 13420,
      "task_loss": 0.9337074160575867
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5704454183578491,
      "epoch": 4.85,
      "learning_rate": 1.3119289637578135e-07,
      "loss": 0.461,
      "step": 13430,
      "task_loss": 1.1079574823379517
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.48450767993927,
      "epoch": 4.86,
      "learning_rate": 1.249059672547892e-07,
      "loss": 0.4929,
      "step": 13440,
      "task_loss": 0.8369407653808594
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.419539213180542,
      "epoch": 4.86,
      "learning_rate": 1.1877308416463506e-07,
      "loss": 0.4397,
      "step": 13450,
      "task_loss": 0.6441231369972229
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4916451573371887,
      "epoch": 4.86,
      "learning_rate": 1.1279427872852965e-07,
      "loss": 0.4731,
      "step": 13460,
      "task_loss": 0.6786421537399292
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41480889916419983,
      "epoch": 4.87,
      "learning_rate": 1.0696958177519922e-07,
      "loss": 0.4472,
      "step": 13470,
      "task_loss": 0.6673970222473145
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4530647397041321,
      "epoch": 4.87,
      "learning_rate": 1.0129902333874563e-07,
      "loss": 0.4267,
      "step": 13480,
      "task_loss": 1.1996808052062988
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37082067131996155,
      "epoch": 4.88,
      "learning_rate": 9.578263265846655e-08,
      "loss": 0.4798,
      "step": 13490,
      "task_loss": 0.3775365352630615
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25928795337677,
      "epoch": 4.88,
      "learning_rate": 9.042043817873547e-08,
      "loss": 0.4649,
      "step": 13500,
      "task_loss": 0.15395194292068481
    },
    {
      "epoch": 4.88,
      "eval_exact_match": 82.80983916745507,
      "eval_f1": 89.66839121824178,
      "step": 13500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5914658308029175,
      "epoch": 4.88,
      "learning_rate": 8.521246754882528e-08,
      "loss": 0.4885,
      "step": 13510,
      "task_loss": 0.7391992807388306
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38663071393966675,
      "epoch": 4.89,
      "learning_rate": 8.01587476227883e-08,
      "loss": 0.3928,
      "step": 13520,
      "task_loss": 0.24446828663349152
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4410104751586914,
      "epoch": 4.89,
      "learning_rate": 7.525930445929974e-08,
      "loss": 0.4671,
      "step": 13530,
      "task_loss": 0.339232861995697
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4469359517097473,
      "epoch": 4.89,
      "learning_rate": 7.051416332153781e-08,
      "loss": 0.4552,
      "step": 13540,
      "task_loss": 0.6987203359603882
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5052058100700378,
      "epoch": 4.9,
      "learning_rate": 6.592334867704719e-08,
      "loss": 0.4769,
      "step": 13550,
      "task_loss": 0.6112401485443115
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4418012499809265,
      "epoch": 4.9,
      "learning_rate": 6.148688419760906e-08,
      "loss": 0.4485,
      "step": 13560,
      "task_loss": 0.6320434212684631
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4680473804473877,
      "epoch": 4.9,
      "learning_rate": 5.7204792759127936e-08,
      "loss": 0.4803,
      "step": 13570,
      "task_loss": 0.8190406560897827
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37423837184906006,
      "epoch": 4.91,
      "learning_rate": 5.307709644150505e-08,
      "loss": 0.4563,
      "step": 13580,
      "task_loss": 0.4952273368835449
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5851823091506958,
      "epoch": 4.91,
      "learning_rate": 4.910381652853513e-08,
      "loss": 0.4465,
      "step": 13590,
      "task_loss": 0.7107750177383423
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3553691506385803,
      "epoch": 4.92,
      "learning_rate": 4.528497350777983e-08,
      "loss": 0.4799,
      "step": 13600,
      "task_loss": 0.5710075497627258
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3894813358783722,
      "epoch": 4.92,
      "learning_rate": 4.162058707048444e-08,
      "loss": 0.4332,
      "step": 13610,
      "task_loss": 0.5974895358085632
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37952518463134766,
      "epoch": 4.92,
      "learning_rate": 3.8110676111451357e-08,
      "loss": 0.4155,
      "step": 13620,
      "task_loss": 0.8900310397148132
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4508509933948517,
      "epoch": 4.93,
      "learning_rate": 3.4755258728963455e-08,
      "loss": 0.4591,
      "step": 13630,
      "task_loss": 0.43794721364974976
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40247878432273865,
      "epoch": 4.93,
      "learning_rate": 3.155435222468417e-08,
      "loss": 0.4891,
      "step": 13640,
      "task_loss": 0.7496728897094727
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47712230682373047,
      "epoch": 4.93,
      "learning_rate": 2.8507973103560903e-08,
      "loss": 0.4467,
      "step": 13650,
      "task_loss": 0.9178839325904846
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4095407724380493,
      "epoch": 4.94,
      "learning_rate": 2.5616137073748436e-08,
      "loss": 0.4325,
      "step": 13660,
      "task_loss": 0.5698050856590271
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5157955884933472,
      "epoch": 4.94,
      "learning_rate": 2.2878859046525648e-08,
      "loss": 0.4647,
      "step": 13670,
      "task_loss": 0.4228496551513672
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3722189664840698,
      "epoch": 4.94,
      "learning_rate": 2.029615313622224e-08,
      "loss": 0.462,
      "step": 13680,
      "task_loss": 0.534970760345459
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4381513297557831,
      "epoch": 4.95,
      "learning_rate": 1.786803266013548e-08,
      "loss": 0.428,
      "step": 13690,
      "task_loss": 0.7130058407783508
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43109041452407837,
      "epoch": 4.95,
      "learning_rate": 1.559451013847024e-08,
      "loss": 0.4274,
      "step": 13700,
      "task_loss": 0.47240149974823
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5169810056686401,
      "epoch": 4.95,
      "learning_rate": 1.3475597294275722e-08,
      "loss": 0.421,
      "step": 13710,
      "task_loss": 0.9785813689231873
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4849017262458801,
      "epoch": 4.96,
      "learning_rate": 1.1511305053375498e-08,
      "loss": 0.4566,
      "step": 13720,
      "task_loss": 0.9021215438842773
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4900825619697571,
      "epoch": 4.96,
      "learning_rate": 9.701643544327565e-09,
      "loss": 0.5105,
      "step": 13730,
      "task_loss": 0.8299789428710938
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3275885581970215,
      "epoch": 4.97,
      "learning_rate": 8.046622098347723e-09,
      "loss": 0.4322,
      "step": 13740,
      "task_loss": 0.4920736849308014
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4971184730529785,
      "epoch": 4.97,
      "learning_rate": 6.5462492492829316e-09,
      "loss": 0.4324,
      "step": 13750,
      "task_loss": 0.6894551515579224
    },
    {
      "epoch": 4.97,
      "eval_exact_match": 82.4314096499527,
      "eval_f1": 89.51060111859181,
      "step": 13750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2823774218559265,
      "epoch": 4.97,
      "learning_rate": 5.2005327335680196e-09,
      "loss": 0.4704,
      "step": 13760,
      "task_loss": 0.2854418456554413
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4087982773780823,
      "epoch": 4.98,
      "learning_rate": 4.0094794901623934e-09,
      "loss": 0.4415,
      "step": 13770,
      "task_loss": 0.6700844168663025
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41539302468299866,
      "epoch": 4.98,
      "learning_rate": 2.973095660540048e-09,
      "loss": 0.4678,
      "step": 13780,
      "task_loss": 0.6645501852035522
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6088934540748596,
      "epoch": 4.98,
      "learning_rate": 2.0913865886462714e-09,
      "loss": 0.4834,
      "step": 13790,
      "task_loss": 0.9101829528808594
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.562792181968689,
      "epoch": 4.99,
      "learning_rate": 1.364356820864332e-09,
      "loss": 0.4619,
      "step": 13800,
      "task_loss": 0.7332584261894226
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3973432183265686,
      "epoch": 4.99,
      "learning_rate": 7.920101060088225e-10,
      "loss": 0.4212,
      "step": 13810,
      "task_loss": 0.6076244711875916
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3704431653022766,
      "epoch": 4.99,
      "learning_rate": 3.743493952890198e-10,
      "loss": 0.4176,
      "step": 13820,
      "task_loss": 0.6046047806739807
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4479011595249176,
      "epoch": 5.0,
      "learning_rate": 1.1137684230888567e-10,
      "loss": 0.491,
      "step": 13830,
      "task_loss": 0.41174694895744324
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6051190495491028,
      "epoch": 5.0,
      "learning_rate": 3.0938030370908365e-12,
      "loss": 0.4473,
      "step": 13840,
      "task_loss": 0.9124219417572021
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36170339584350586,
      "epoch": 5.01,
      "learning_rate": 5.9999950499164185e-05,
      "loss": 0.409,
      "step": 13850,
      "task_loss": 0.36121177673339844
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.537726104259491,
      "epoch": 5.01,
      "learning_rate": 5.999974940229864e-05,
      "loss": 0.4383,
      "step": 13860,
      "task_loss": 0.6074551343917847
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5061550140380859,
      "epoch": 5.01,
      "learning_rate": 5.999939361663725e-05,
      "loss": 0.431,
      "step": 13870,
      "task_loss": 0.4759042263031006
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4849228262901306,
      "epoch": 5.02,
      "learning_rate": 5.999888314401456e-05,
      "loss": 0.4848,
      "step": 13880,
      "task_loss": 0.6583101749420166
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4311343729496002,
      "epoch": 5.02,
      "learning_rate": 5.9998217987062765e-05,
      "loss": 0.4236,
      "step": 13890,
      "task_loss": 0.6400418281555176
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6894408464431763,
      "epoch": 5.02,
      "learning_rate": 5.999739814921161e-05,
      "loss": 0.4667,
      "step": 13900,
      "task_loss": 0.9435676336288452
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6396845579147339,
      "epoch": 5.03,
      "learning_rate": 5.9996423634688465e-05,
      "loss": 0.5015,
      "step": 13910,
      "task_loss": 0.7678850293159485
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39257675409317017,
      "epoch": 5.03,
      "learning_rate": 5.999529444851826e-05,
      "loss": 0.4825,
      "step": 13920,
      "task_loss": 0.5393377542495728
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3359961211681366,
      "epoch": 5.03,
      "learning_rate": 5.999401059652345e-05,
      "loss": 0.4539,
      "step": 13930,
      "task_loss": 0.7673816680908203
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39863264560699463,
      "epoch": 5.04,
      "learning_rate": 5.999257208532401e-05,
      "loss": 0.483,
      "step": 13940,
      "task_loss": 1.0004619359970093
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38244152069091797,
      "epoch": 5.04,
      "learning_rate": 5.99909789223374e-05,
      "loss": 0.4579,
      "step": 13950,
      "task_loss": 0.3268355429172516
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40854525566101074,
      "epoch": 5.05,
      "learning_rate": 5.998923111577847e-05,
      "loss": 0.4941,
      "step": 13960,
      "task_loss": 0.5148742198944092
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3848707675933838,
      "epoch": 5.05,
      "learning_rate": 5.998732867465954e-05,
      "loss": 0.4979,
      "step": 13970,
      "task_loss": 0.5055022835731506
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7005259990692139,
      "epoch": 5.05,
      "learning_rate": 5.998527160879021e-05,
      "loss": 0.5703,
      "step": 13980,
      "task_loss": 0.7080357074737549
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4665435552597046,
      "epoch": 5.06,
      "learning_rate": 5.998305992877741e-05,
      "loss": 0.4853,
      "step": 13990,
      "task_loss": 0.7121034860610962
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4767329692840576,
      "epoch": 5.06,
      "learning_rate": 5.99806936460253e-05,
      "loss": 0.4968,
      "step": 14000,
      "task_loss": 0.7108612060546875
    },
    {
      "epoch": 5.06,
      "eval_exact_match": 82.15704824976348,
      "eval_f1": 89.43257264133547,
      "step": 14000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3834998607635498,
      "epoch": 5.06,
      "learning_rate": 5.997817277273525e-05,
      "loss": 0.4928,
      "step": 14010,
      "task_loss": 0.45428478717803955
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7256457805633545,
      "epoch": 5.07,
      "learning_rate": 5.997549732190571e-05,
      "loss": 0.525,
      "step": 14020,
      "task_loss": 0.8021411895751953
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5225358009338379,
      "epoch": 5.07,
      "learning_rate": 5.997266730733221e-05,
      "loss": 0.537,
      "step": 14030,
      "task_loss": 0.5008682012557983
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6417495012283325,
      "epoch": 5.07,
      "learning_rate": 5.996968274360725e-05,
      "loss": 0.5379,
      "step": 14040,
      "task_loss": 0.9481985569000244
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4337199330329895,
      "epoch": 5.08,
      "learning_rate": 5.996654364612026e-05,
      "loss": 0.5181,
      "step": 14050,
      "task_loss": 0.4173719882965088
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5246778726577759,
      "epoch": 5.08,
      "learning_rate": 5.996325003105747e-05,
      "loss": 0.5021,
      "step": 14060,
      "task_loss": 1.0055538415908813
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41551411151885986,
      "epoch": 5.08,
      "learning_rate": 5.9959801915401866e-05,
      "loss": 0.477,
      "step": 14070,
      "task_loss": 0.45110952854156494
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5802291631698608,
      "epoch": 5.09,
      "learning_rate": 5.99561993169331e-05,
      "loss": 0.5067,
      "step": 14080,
      "task_loss": 0.8387892246246338
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4913982152938843,
      "epoch": 5.09,
      "learning_rate": 5.995244225422738e-05,
      "loss": 0.5391,
      "step": 14090,
      "task_loss": 0.5127557516098022
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.465750515460968,
      "epoch": 5.1,
      "learning_rate": 5.994853074665739e-05,
      "loss": 0.4964,
      "step": 14100,
      "task_loss": 0.8444522619247437
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6058213710784912,
      "epoch": 5.1,
      "learning_rate": 5.9944464814392157e-05,
      "loss": 0.5417,
      "step": 14110,
      "task_loss": 0.7191112637519836
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4102645516395569,
      "epoch": 5.1,
      "learning_rate": 5.9940244478397005e-05,
      "loss": 0.4553,
      "step": 14120,
      "task_loss": 0.39053791761398315
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6364562511444092,
      "epoch": 5.11,
      "learning_rate": 5.9935869760433415e-05,
      "loss": 0.5512,
      "step": 14130,
      "task_loss": 0.624964714050293
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43969202041625977,
      "epoch": 5.11,
      "learning_rate": 5.99313406830589e-05,
      "loss": 0.4893,
      "step": 14140,
      "task_loss": 0.36895930767059326
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4144252836704254,
      "epoch": 5.11,
      "learning_rate": 5.99266572696269e-05,
      "loss": 0.5299,
      "step": 14150,
      "task_loss": 0.5147717595100403
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3856452703475952,
      "epoch": 5.12,
      "learning_rate": 5.992181954428667e-05,
      "loss": 0.499,
      "step": 14160,
      "task_loss": 0.5266073346138
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5156189799308777,
      "epoch": 5.12,
      "learning_rate": 5.9916827531983165e-05,
      "loss": 0.4847,
      "step": 14170,
      "task_loss": 0.4401143789291382
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4977889657020569,
      "epoch": 5.12,
      "learning_rate": 5.991168125845685e-05,
      "loss": 0.4558,
      "step": 14180,
      "task_loss": 0.5201467275619507
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4669112265110016,
      "epoch": 5.13,
      "learning_rate": 5.990638075024366e-05,
      "loss": 0.4877,
      "step": 14190,
      "task_loss": 0.5265516042709351
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32335522770881653,
      "epoch": 5.13,
      "learning_rate": 5.990092603467481e-05,
      "loss": 0.5017,
      "step": 14200,
      "task_loss": 0.3526550233364105
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41797494888305664,
      "epoch": 5.14,
      "learning_rate": 5.989531713987662e-05,
      "loss": 0.4715,
      "step": 14210,
      "task_loss": 0.4137646555900574
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4363604187965393,
      "epoch": 5.14,
      "learning_rate": 5.988955409477045e-05,
      "loss": 0.5498,
      "step": 14220,
      "task_loss": 0.6443294286727905
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6083539724349976,
      "epoch": 5.14,
      "learning_rate": 5.98836369290725e-05,
      "loss": 0.5433,
      "step": 14230,
      "task_loss": 0.6782771348953247
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43323948979377747,
      "epoch": 5.15,
      "learning_rate": 5.987756567329367e-05,
      "loss": 0.4889,
      "step": 14240,
      "task_loss": 0.440801739692688
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5146920680999756,
      "epoch": 5.15,
      "learning_rate": 5.9871340358739385e-05,
      "loss": 0.4781,
      "step": 14250,
      "task_loss": 0.48200464248657227
    },
    {
      "epoch": 5.15,
      "eval_exact_match": 82.10974456007568,
      "eval_f1": 89.35325173039678,
      "step": 14250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4706941545009613,
      "epoch": 5.15,
      "learning_rate": 5.9864961017509476e-05,
      "loss": 0.5426,
      "step": 14260,
      "task_loss": 0.9112202525138855
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4482850432395935,
      "epoch": 5.16,
      "learning_rate": 5.985842768249795e-05,
      "loss": 0.4626,
      "step": 14270,
      "task_loss": 0.46659618616104126
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47765567898750305,
      "epoch": 5.16,
      "learning_rate": 5.98517403873929e-05,
      "loss": 0.5386,
      "step": 14280,
      "task_loss": 0.7766150832176208
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6142876148223877,
      "epoch": 5.16,
      "learning_rate": 5.984489916667626e-05,
      "loss": 0.5518,
      "step": 14290,
      "task_loss": 0.7481546998023987
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5589594841003418,
      "epoch": 5.17,
      "learning_rate": 5.983790405562367e-05,
      "loss": 0.4997,
      "step": 14300,
      "task_loss": 0.8356878757476807
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5612353086471558,
      "epoch": 5.17,
      "learning_rate": 5.983075509030426e-05,
      "loss": 0.5829,
      "step": 14310,
      "task_loss": 0.865998387336731
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8121389150619507,
      "epoch": 5.18,
      "learning_rate": 5.982345230758051e-05,
      "loss": 0.5064,
      "step": 14320,
      "task_loss": 1.0466325283050537
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5815309286117554,
      "epoch": 5.18,
      "learning_rate": 5.981599574510802e-05,
      "loss": 0.5482,
      "step": 14330,
      "task_loss": 0.8093647956848145
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49722617864608765,
      "epoch": 5.18,
      "learning_rate": 5.980838544133534e-05,
      "loss": 0.5367,
      "step": 14340,
      "task_loss": 0.6952990889549255
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5625362396240234,
      "epoch": 5.19,
      "learning_rate": 5.980062143550375e-05,
      "loss": 0.5873,
      "step": 14350,
      "task_loss": 0.5973076820373535
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8071750402450562,
      "epoch": 5.19,
      "learning_rate": 5.979270376764708e-05,
      "loss": 0.5362,
      "step": 14360,
      "task_loss": 0.801953911781311
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5167639255523682,
      "epoch": 5.19,
      "learning_rate": 5.9784632478591475e-05,
      "loss": 0.5556,
      "step": 14370,
      "task_loss": 0.4926624298095703
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5415273308753967,
      "epoch": 5.2,
      "learning_rate": 5.977640760995523e-05,
      "loss": 0.5254,
      "step": 14380,
      "task_loss": 0.8653746843338013
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5194505453109741,
      "epoch": 5.2,
      "learning_rate": 5.9768029204148513e-05,
      "loss": 0.5942,
      "step": 14390,
      "task_loss": 0.769846498966217
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5619365572929382,
      "epoch": 5.2,
      "learning_rate": 5.9759497304373234e-05,
      "loss": 0.5572,
      "step": 14400,
      "task_loss": 0.6389682292938232
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6544053554534912,
      "epoch": 5.21,
      "learning_rate": 5.9750811954622714e-05,
      "loss": 0.5505,
      "step": 14410,
      "task_loss": 0.6094022989273071
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9797204732894897,
      "epoch": 5.21,
      "learning_rate": 5.974197319968153e-05,
      "loss": 0.5554,
      "step": 14420,
      "task_loss": 0.9586498141288757
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4406765103340149,
      "epoch": 5.22,
      "learning_rate": 5.973298108512531e-05,
      "loss": 0.5066,
      "step": 14430,
      "task_loss": 0.542404055595398
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5614707469940186,
      "epoch": 5.22,
      "learning_rate": 5.972383565732038e-05,
      "loss": 0.5418,
      "step": 14440,
      "task_loss": 0.6197214126586914
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.529656708240509,
      "epoch": 5.22,
      "learning_rate": 5.971453696342367e-05,
      "loss": 0.5157,
      "step": 14450,
      "task_loss": 0.7224357724189758
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49263083934783936,
      "epoch": 5.23,
      "learning_rate": 5.9705085051382344e-05,
      "loss": 0.4777,
      "step": 14460,
      "task_loss": 0.3264583349227905
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4123901426792145,
      "epoch": 5.23,
      "learning_rate": 5.969547996993367e-05,
      "loss": 0.4791,
      "step": 14470,
      "task_loss": 0.6501359343528748
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5078509449958801,
      "epoch": 5.23,
      "learning_rate": 5.9685721768604656e-05,
      "loss": 0.6034,
      "step": 14480,
      "task_loss": 0.6295353174209595
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3895311951637268,
      "epoch": 5.24,
      "learning_rate": 5.967581049771188e-05,
      "loss": 0.4871,
      "step": 14490,
      "task_loss": 0.34086859226226807
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5105054974555969,
      "epoch": 5.24,
      "learning_rate": 5.966574620836119e-05,
      "loss": 0.4953,
      "step": 14500,
      "task_loss": 0.6598679423332214
    },
    {
      "epoch": 5.24,
      "eval_exact_match": 81.95837275307474,
      "eval_f1": 89.23343113339236,
      "step": 14500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5998846292495728,
      "epoch": 5.24,
      "learning_rate": 5.9655528952447426e-05,
      "loss": 0.5945,
      "step": 14510,
      "task_loss": 0.5025860071182251
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6489298939704895,
      "epoch": 5.25,
      "learning_rate": 5.9645158782654216e-05,
      "loss": 0.5164,
      "step": 14520,
      "task_loss": 0.7763657569885254
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.480061411857605,
      "epoch": 5.25,
      "learning_rate": 5.9634635752453644e-05,
      "loss": 0.4953,
      "step": 14530,
      "task_loss": 0.626976728439331
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5622721910476685,
      "epoch": 5.25,
      "learning_rate": 5.962395991610599e-05,
      "loss": 0.5129,
      "step": 14540,
      "task_loss": 0.6075005531311035
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4024551510810852,
      "epoch": 5.26,
      "learning_rate": 5.961313132865944e-05,
      "loss": 0.5174,
      "step": 14550,
      "task_loss": 0.892763614654541
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.48125165700912476,
      "epoch": 5.26,
      "learning_rate": 5.960215004594984e-05,
      "loss": 0.5541,
      "step": 14560,
      "task_loss": 0.6092730164527893
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38604000210762024,
      "epoch": 5.27,
      "learning_rate": 5.959101612460038e-05,
      "loss": 0.4897,
      "step": 14570,
      "task_loss": 0.7240785360336304
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.516532301902771,
      "epoch": 5.27,
      "learning_rate": 5.957972962202131e-05,
      "loss": 0.5445,
      "step": 14580,
      "task_loss": 0.7919718027114868
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.537828266620636,
      "epoch": 5.27,
      "learning_rate": 5.9568290596409616e-05,
      "loss": 0.5246,
      "step": 14590,
      "task_loss": 0.6629247069358826
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49570807814598083,
      "epoch": 5.28,
      "learning_rate": 5.955669910674877e-05,
      "loss": 0.5261,
      "step": 14600,
      "task_loss": 0.3058420419692993
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6100714206695557,
      "epoch": 5.28,
      "learning_rate": 5.95449552128084e-05,
      "loss": 0.524,
      "step": 14610,
      "task_loss": 1.2158831357955933
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5747349262237549,
      "epoch": 5.28,
      "learning_rate": 5.9533058975143956e-05,
      "loss": 0.5039,
      "step": 14620,
      "task_loss": 0.4315424859523773
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5765594840049744,
      "epoch": 5.29,
      "learning_rate": 5.9521010455096446e-05,
      "loss": 0.5584,
      "step": 14630,
      "task_loss": 0.831333339214325
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6396365761756897,
      "epoch": 5.29,
      "learning_rate": 5.950880971479208e-05,
      "loss": 0.5345,
      "step": 14640,
      "task_loss": 0.8963925242424011
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.594346821308136,
      "epoch": 5.29,
      "learning_rate": 5.9496456817141996e-05,
      "loss": 0.5928,
      "step": 14650,
      "task_loss": 1.1544039249420166
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4301066994667053,
      "epoch": 5.3,
      "learning_rate": 5.948395182584189e-05,
      "loss": 0.4974,
      "step": 14660,
      "task_loss": 0.5060799717903137
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46475663781166077,
      "epoch": 5.3,
      "learning_rate": 5.9471294805371684e-05,
      "loss": 0.5466,
      "step": 14670,
      "task_loss": 0.6515345573425293
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43622440099716187,
      "epoch": 5.31,
      "learning_rate": 5.9458485820995235e-05,
      "loss": 0.5506,
      "step": 14680,
      "task_loss": 0.7932460308074951
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4005524516105652,
      "epoch": 5.31,
      "learning_rate": 5.9445524938759986e-05,
      "loss": 0.4265,
      "step": 14690,
      "task_loss": 0.30546653270721436
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6993730068206787,
      "epoch": 5.31,
      "learning_rate": 5.9432412225496586e-05,
      "loss": 0.4971,
      "step": 14700,
      "task_loss": 0.7546347379684448
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8860001564025879,
      "epoch": 5.32,
      "learning_rate": 5.941914774881862e-05,
      "loss": 0.5512,
      "step": 14710,
      "task_loss": 0.7079243063926697
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6870177984237671,
      "epoch": 5.32,
      "learning_rate": 5.940573157712217e-05,
      "loss": 0.5731,
      "step": 14720,
      "task_loss": 0.8747744560241699
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5104433298110962,
      "epoch": 5.32,
      "learning_rate": 5.9392163779585525e-05,
      "loss": 0.5506,
      "step": 14730,
      "task_loss": 1.362548589706421
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40730664134025574,
      "epoch": 5.33,
      "learning_rate": 5.937844442616883e-05,
      "loss": 0.518,
      "step": 14740,
      "task_loss": 0.5095131397247314
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5867040157318115,
      "epoch": 5.33,
      "learning_rate": 5.936457358761367e-05,
      "loss": 0.5429,
      "step": 14750,
      "task_loss": 0.6059485673904419
    },
    {
      "epoch": 5.33,
      "eval_exact_match": 81.59886471144749,
      "eval_f1": 88.89803036593699,
      "step": 14750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39228177070617676,
      "epoch": 5.33,
      "learning_rate": 5.935055133544276e-05,
      "loss": 0.5219,
      "step": 14760,
      "task_loss": 0.5120740532875061
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4774217903614044,
      "epoch": 5.34,
      "learning_rate": 5.933637774195955e-05,
      "loss": 0.5376,
      "step": 14770,
      "task_loss": 0.6802091002464294
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6107138395309448,
      "epoch": 5.34,
      "learning_rate": 5.932205288024784e-05,
      "loss": 0.5498,
      "step": 14780,
      "task_loss": 0.6426236033439636
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6453840732574463,
      "epoch": 5.35,
      "learning_rate": 5.930757682417145e-05,
      "loss": 0.5544,
      "step": 14790,
      "task_loss": 0.845929741859436
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6528950929641724,
      "epoch": 5.35,
      "learning_rate": 5.929294964837378e-05,
      "loss": 0.514,
      "step": 14800,
      "task_loss": 0.9554112553596497
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6820924878120422,
      "epoch": 5.35,
      "learning_rate": 5.9278171428277475e-05,
      "loss": 0.6105,
      "step": 14810,
      "task_loss": 0.9136593341827393
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4658266603946686,
      "epoch": 5.36,
      "learning_rate": 5.926324224008401e-05,
      "loss": 0.5843,
      "step": 14820,
      "task_loss": 0.46199867129325867
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7525842189788818,
      "epoch": 5.36,
      "learning_rate": 5.9248162160773286e-05,
      "loss": 0.5121,
      "step": 14830,
      "task_loss": 0.7043994665145874
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5841355323791504,
      "epoch": 5.36,
      "learning_rate": 5.923293126810328e-05,
      "loss": 0.5481,
      "step": 14840,
      "task_loss": 0.6969549655914307
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6712848544120789,
      "epoch": 5.37,
      "learning_rate": 5.921754964060959e-05,
      "loss": 0.588,
      "step": 14850,
      "task_loss": 0.48286545276641846
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5721571445465088,
      "epoch": 5.37,
      "learning_rate": 5.9202017357605045e-05,
      "loss": 0.5374,
      "step": 14860,
      "task_loss": 0.847301721572876
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47767120599746704,
      "epoch": 5.37,
      "learning_rate": 5.918633449917934e-05,
      "loss": 0.5735,
      "step": 14870,
      "task_loss": 0.924615204334259
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47853583097457886,
      "epoch": 5.38,
      "learning_rate": 5.917050114619856e-05,
      "loss": 0.5108,
      "step": 14880,
      "task_loss": 0.7048487663269043
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47235754132270813,
      "epoch": 5.38,
      "learning_rate": 5.9154517380304794e-05,
      "loss": 0.5083,
      "step": 14890,
      "task_loss": 0.7272685766220093
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5390362739562988,
      "epoch": 5.38,
      "learning_rate": 5.9138383283915716e-05,
      "loss": 0.5101,
      "step": 14900,
      "task_loss": 0.7805506587028503
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42291706800460815,
      "epoch": 5.39,
      "learning_rate": 5.912209894022415e-05,
      "loss": 0.5247,
      "step": 14910,
      "task_loss": 0.4546523988246918
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6531658172607422,
      "epoch": 5.39,
      "learning_rate": 5.910566443319765e-05,
      "loss": 0.5639,
      "step": 14920,
      "task_loss": 0.6974441409111023
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5830737948417664,
      "epoch": 5.4,
      "learning_rate": 5.908907984757806e-05,
      "loss": 0.5388,
      "step": 14930,
      "task_loss": 0.6396021842956543
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5406855344772339,
      "epoch": 5.4,
      "learning_rate": 5.9072345268881085e-05,
      "loss": 0.5475,
      "step": 14940,
      "task_loss": 0.7636957168579102
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7267158031463623,
      "epoch": 5.4,
      "learning_rate": 5.905715597527462e-05,
      "loss": 0.5783,
      "step": 14950,
      "task_loss": 0.6421911120414734
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5212079286575317,
      "epoch": 5.41,
      "learning_rate": 5.9040136648095986e-05,
      "loss": 0.5195,
      "step": 14960,
      "task_loss": 0.9378068447113037
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41629695892333984,
      "epoch": 5.41,
      "learning_rate": 5.902296758020755e-05,
      "loss": 0.4998,
      "step": 14970,
      "task_loss": 0.8873476982116699
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5604190826416016,
      "epoch": 5.41,
      "learning_rate": 5.900564886013882e-05,
      "loss": 0.5128,
      "step": 14980,
      "task_loss": 0.5965309143066406
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6193476915359497,
      "epoch": 5.42,
      "learning_rate": 5.8988180577190914e-05,
      "loss": 0.5533,
      "step": 14990,
      "task_loss": 0.8003284335136414
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3812726140022278,
      "epoch": 5.42,
      "learning_rate": 5.8970562821436184e-05,
      "loss": 0.5454,
      "step": 15000,
      "task_loss": 0.5890715718269348
    },
    {
      "epoch": 5.42,
      "eval_exact_match": 82.06244087038789,
      "eval_f1": 89.37476650775072,
      "step": 15000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4163830280303955,
      "epoch": 5.42,
      "learning_rate": 5.895279568371772e-05,
      "loss": 0.5529,
      "step": 15010,
      "task_loss": 0.8374828696250916
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5462802648544312,
      "epoch": 5.43,
      "learning_rate": 5.8934879255648834e-05,
      "loss": 0.5532,
      "step": 15020,
      "task_loss": 0.939515233039856
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4633031487464905,
      "epoch": 5.43,
      "learning_rate": 5.8916813629612655e-05,
      "loss": 0.5588,
      "step": 15030,
      "task_loss": 0.38705766201019287
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5264443755149841,
      "epoch": 5.44,
      "learning_rate": 5.8898598898761656e-05,
      "loss": 0.4643,
      "step": 15040,
      "task_loss": 0.5500625967979431
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4520212411880493,
      "epoch": 5.44,
      "learning_rate": 5.8880235157017074e-05,
      "loss": 0.4999,
      "step": 15050,
      "task_loss": 0.41272589564323425
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4026266038417816,
      "epoch": 5.44,
      "learning_rate": 5.886172249906856e-05,
      "loss": 0.4997,
      "step": 15060,
      "task_loss": 0.30943921208381653
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7537382245063782,
      "epoch": 5.45,
      "learning_rate": 5.8843061020373594e-05,
      "loss": 0.5524,
      "step": 15070,
      "task_loss": 0.9021023511886597
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4830913245677948,
      "epoch": 5.45,
      "learning_rate": 5.882425081715705e-05,
      "loss": 0.5472,
      "step": 15080,
      "task_loss": 0.48394033312797546
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45110899209976196,
      "epoch": 5.45,
      "learning_rate": 5.8805291986410646e-05,
      "loss": 0.5265,
      "step": 15090,
      "task_loss": 0.49047964811325073
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39607346057891846,
      "epoch": 5.46,
      "learning_rate": 5.878618462589249e-05,
      "loss": 0.5501,
      "step": 15100,
      "task_loss": 0.5193868279457092
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.475017786026001,
      "epoch": 5.46,
      "learning_rate": 5.876692883412656e-05,
      "loss": 0.4851,
      "step": 15110,
      "task_loss": 0.7011887431144714
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6903802156448364,
      "epoch": 5.46,
      "learning_rate": 5.87475247104022e-05,
      "loss": 0.5406,
      "step": 15120,
      "task_loss": 0.7297502756118774
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4474151134490967,
      "epoch": 5.47,
      "learning_rate": 5.872797235477359e-05,
      "loss": 0.5009,
      "step": 15130,
      "task_loss": 0.7269125580787659
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38013216853141785,
      "epoch": 5.47,
      "learning_rate": 5.8708271868059255e-05,
      "loss": 0.5268,
      "step": 15140,
      "task_loss": 0.738182544708252
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.503143310546875,
      "epoch": 5.48,
      "learning_rate": 5.868842335184153e-05,
      "loss": 0.5503,
      "step": 15150,
      "task_loss": 1.0320183038711548
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5389174222946167,
      "epoch": 5.48,
      "learning_rate": 5.866842690846602e-05,
      "loss": 0.5421,
      "step": 15160,
      "task_loss": 0.44373756647109985
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4785197377204895,
      "epoch": 5.48,
      "learning_rate": 5.864828264104114e-05,
      "loss": 0.5715,
      "step": 15170,
      "task_loss": 0.76427161693573
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40499380230903625,
      "epoch": 5.49,
      "learning_rate": 5.862799065343749e-05,
      "loss": 0.529,
      "step": 15180,
      "task_loss": 0.8471218347549438
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34510934352874756,
      "epoch": 5.49,
      "learning_rate": 5.860755105028738e-05,
      "loss": 0.4904,
      "step": 15190,
      "task_loss": 0.42475417256355286
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6029459238052368,
      "epoch": 5.49,
      "learning_rate": 5.858696393698428e-05,
      "loss": 0.5395,
      "step": 15200,
      "task_loss": 0.6130906343460083
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6409763097763062,
      "epoch": 5.5,
      "learning_rate": 5.8566229419682275e-05,
      "loss": 0.5447,
      "step": 15210,
      "task_loss": 0.3819064199924469
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3984764814376831,
      "epoch": 5.5,
      "learning_rate": 5.8545347605295516e-05,
      "loss": 0.571,
      "step": 15220,
      "task_loss": 0.6237053871154785
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44126883149147034,
      "epoch": 5.5,
      "learning_rate": 5.852431860149763e-05,
      "loss": 0.5209,
      "step": 15230,
      "task_loss": 0.5754768252372742
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4907762110233307,
      "epoch": 5.51,
      "learning_rate": 5.850314251672126e-05,
      "loss": 0.552,
      "step": 15240,
      "task_loss": 0.34236305952072144
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.552409291267395,
      "epoch": 5.51,
      "learning_rate": 5.848181946015741e-05,
      "loss": 0.5916,
      "step": 15250,
      "task_loss": 0.6347699165344238
    },
    {
      "epoch": 5.51,
      "eval_exact_match": 82.09082308420057,
      "eval_f1": 89.29649475160107,
      "step": 15250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4228856861591339,
      "epoch": 5.51,
      "learning_rate": 5.846034954175493e-05,
      "loss": 0.5081,
      "step": 15260,
      "task_loss": 0.397327721118927
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39670825004577637,
      "epoch": 5.52,
      "learning_rate": 5.8438732872219934e-05,
      "loss": 0.5519,
      "step": 15270,
      "task_loss": 0.37487858533859253
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5530394911766052,
      "epoch": 5.52,
      "learning_rate": 5.8416969563015246e-05,
      "loss": 0.494,
      "step": 15280,
      "task_loss": 0.7215574979782104
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5375179648399353,
      "epoch": 5.53,
      "learning_rate": 5.83950597263598e-05,
      "loss": 0.4888,
      "step": 15290,
      "task_loss": 0.5879909992218018
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6529196500778198,
      "epoch": 5.53,
      "learning_rate": 5.837300347522809e-05,
      "loss": 0.5204,
      "step": 15300,
      "task_loss": 1.0084820985794067
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4595330059528351,
      "epoch": 5.53,
      "learning_rate": 5.8350800923349556e-05,
      "loss": 0.5129,
      "step": 15310,
      "task_loss": 0.5193516612052917
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5306529998779297,
      "epoch": 5.54,
      "learning_rate": 5.8328452185208034e-05,
      "loss": 0.5032,
      "step": 15320,
      "task_loss": 0.6539807915687561
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35273468494415283,
      "epoch": 5.54,
      "learning_rate": 5.830595737604112e-05,
      "loss": 0.5855,
      "step": 15330,
      "task_loss": 0.36837586760520935
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5520907640457153,
      "epoch": 5.54,
      "learning_rate": 5.828331661183961e-05,
      "loss": 0.5686,
      "step": 15340,
      "task_loss": 0.6090099215507507
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5058830380439758,
      "epoch": 5.55,
      "learning_rate": 5.826053000934692e-05,
      "loss": 0.516,
      "step": 15350,
      "task_loss": 0.4280429780483246
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5331200361251831,
      "epoch": 5.55,
      "learning_rate": 5.8237597686058406e-05,
      "loss": 0.5455,
      "step": 15360,
      "task_loss": 0.46258166432380676
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49761390686035156,
      "epoch": 5.55,
      "learning_rate": 5.821451976022086e-05,
      "loss": 0.5249,
      "step": 15370,
      "task_loss": 0.8308390974998474
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.662971019744873,
      "epoch": 5.56,
      "learning_rate": 5.819129635083182e-05,
      "loss": 0.5582,
      "step": 15380,
      "task_loss": 0.6953208446502686
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6624317765235901,
      "epoch": 5.56,
      "learning_rate": 5.8167927577638994e-05,
      "loss": 0.5488,
      "step": 15390,
      "task_loss": 0.9981436729431152
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5271157026290894,
      "epoch": 5.57,
      "learning_rate": 5.8144413561139636e-05,
      "loss": 0.5159,
      "step": 15400,
      "task_loss": 0.5372161865234375
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7370679378509521,
      "epoch": 5.57,
      "learning_rate": 5.812075442257992e-05,
      "loss": 0.5604,
      "step": 15410,
      "task_loss": 1.0659924745559692
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6820415258407593,
      "epoch": 5.57,
      "learning_rate": 5.8096950283954324e-05,
      "loss": 0.5597,
      "step": 15420,
      "task_loss": 0.9119454622268677
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39658281207084656,
      "epoch": 5.58,
      "learning_rate": 5.8073001268004986e-05,
      "loss": 0.5198,
      "step": 15430,
      "task_loss": 0.8047652840614319
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42770448327064514,
      "epoch": 5.58,
      "learning_rate": 5.804890749822109e-05,
      "loss": 0.4929,
      "step": 15440,
      "task_loss": 1.5828943252563477
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3862127363681793,
      "epoch": 5.58,
      "learning_rate": 5.8024669098838224e-05,
      "loss": 0.5447,
      "step": 15450,
      "task_loss": 0.45715686678886414
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4763340353965759,
      "epoch": 5.59,
      "learning_rate": 5.8000286194837706e-05,
      "loss": 0.5788,
      "step": 15460,
      "task_loss": 0.5869323015213013
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6561774015426636,
      "epoch": 5.59,
      "learning_rate": 5.7975758911945995e-05,
      "loss": 0.5451,
      "step": 15470,
      "task_loss": 0.7080704569816589
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40479183197021484,
      "epoch": 5.59,
      "learning_rate": 5.795108737663401e-05,
      "loss": 0.5185,
      "step": 15480,
      "task_loss": 0.6346590518951416
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.734893262386322,
      "epoch": 5.6,
      "learning_rate": 5.79262717161165e-05,
      "loss": 0.5102,
      "step": 15490,
      "task_loss": 0.9744662046432495
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4641631841659546,
      "epoch": 5.6,
      "learning_rate": 5.7901312058351334e-05,
      "loss": 0.5243,
      "step": 15500,
      "task_loss": 0.44881314039230347
    },
    {
      "epoch": 5.6,
      "eval_exact_match": 82.25165562913908,
      "eval_f1": 89.3428073483407,
      "step": 15500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44037681818008423,
      "epoch": 5.61,
      "learning_rate": 5.78762085320389e-05,
      "loss": 0.4944,
      "step": 15510,
      "task_loss": 0.3445591330528259
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6916956901550293,
      "epoch": 5.61,
      "learning_rate": 5.7850961266621434e-05,
      "loss": 0.5736,
      "step": 15520,
      "task_loss": 0.8610516786575317
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41196784377098083,
      "epoch": 5.61,
      "learning_rate": 5.782557039228231e-05,
      "loss": 0.5498,
      "step": 15530,
      "task_loss": 0.3979591131210327
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.48362213373184204,
      "epoch": 5.62,
      "learning_rate": 5.780003603994543e-05,
      "loss": 0.5389,
      "step": 15540,
      "task_loss": 0.5863582491874695
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.508769154548645,
      "epoch": 5.62,
      "learning_rate": 5.77743583412745e-05,
      "loss": 0.4957,
      "step": 15550,
      "task_loss": 0.9604640007019043
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5353500247001648,
      "epoch": 5.62,
      "learning_rate": 5.774853742867237e-05,
      "loss": 0.5351,
      "step": 15560,
      "task_loss": 0.4285392165184021
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6010510921478271,
      "epoch": 5.63,
      "learning_rate": 5.772257343528033e-05,
      "loss": 0.5232,
      "step": 15570,
      "task_loss": 0.6189045906066895
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5353314876556396,
      "epoch": 5.63,
      "learning_rate": 5.7696466494977485e-05,
      "loss": 0.5602,
      "step": 15580,
      "task_loss": 0.7458304762840271
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7870986461639404,
      "epoch": 5.63,
      "learning_rate": 5.767021674237998e-05,
      "loss": 0.5116,
      "step": 15590,
      "task_loss": 0.6424393057823181
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.443065881729126,
      "epoch": 5.64,
      "learning_rate": 5.7643824312840375e-05,
      "loss": 0.4851,
      "step": 15600,
      "task_loss": 0.6533344984054565
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3721536099910736,
      "epoch": 5.64,
      "learning_rate": 5.76172893424469e-05,
      "loss": 0.5041,
      "step": 15610,
      "task_loss": 0.5399680733680725
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7717987298965454,
      "epoch": 5.65,
      "learning_rate": 5.75906119680228e-05,
      "loss": 0.5318,
      "step": 15620,
      "task_loss": 1.1942107677459717
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49710613489151,
      "epoch": 5.65,
      "learning_rate": 5.756379232712559e-05,
      "loss": 0.5569,
      "step": 15630,
      "task_loss": 0.30331313610076904
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42363065481185913,
      "epoch": 5.65,
      "learning_rate": 5.753683055804634e-05,
      "loss": 0.5337,
      "step": 15640,
      "task_loss": 0.6503758430480957
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6108335256576538,
      "epoch": 5.66,
      "learning_rate": 5.750972679980902e-05,
      "loss": 0.572,
      "step": 15650,
      "task_loss": 0.9395371079444885
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40696799755096436,
      "epoch": 5.66,
      "learning_rate": 5.7482481192169704e-05,
      "loss": 0.5518,
      "step": 15660,
      "task_loss": 0.6489532589912415
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45866596698760986,
      "epoch": 5.66,
      "learning_rate": 5.745509387561591e-05,
      "loss": 0.5466,
      "step": 15670,
      "task_loss": 0.814210832118988
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5743582248687744,
      "epoch": 5.67,
      "learning_rate": 5.742756499136585e-05,
      "loss": 0.5609,
      "step": 15680,
      "task_loss": 0.7212681174278259
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4180968999862671,
      "epoch": 5.67,
      "learning_rate": 5.739989468136772e-05,
      "loss": 0.6103,
      "step": 15690,
      "task_loss": 0.3201150894165039
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38410964608192444,
      "epoch": 5.67,
      "learning_rate": 5.7372083088298935e-05,
      "loss": 0.5811,
      "step": 15700,
      "task_loss": 0.44139623641967773
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6996259093284607,
      "epoch": 5.68,
      "learning_rate": 5.734413035556541e-05,
      "loss": 0.5444,
      "step": 15710,
      "task_loss": 0.8911246061325073
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5756803750991821,
      "epoch": 5.68,
      "learning_rate": 5.731603662730085e-05,
      "loss": 0.5422,
      "step": 15720,
      "task_loss": 0.8120925426483154
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5546524524688721,
      "epoch": 5.68,
      "learning_rate": 5.7287802048365964e-05,
      "loss": 0.5102,
      "step": 15730,
      "task_loss": 0.6311538219451904
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8656566143035889,
      "epoch": 5.69,
      "learning_rate": 5.725942676434773e-05,
      "loss": 0.5599,
      "step": 15740,
      "task_loss": 0.7161672115325928
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4685131311416626,
      "epoch": 5.69,
      "learning_rate": 5.723091092155865e-05,
      "loss": 0.5562,
      "step": 15750,
      "task_loss": 0.3831646144390106
    },
    {
      "epoch": 5.69,
      "eval_exact_match": 82.02459791863765,
      "eval_f1": 89.1522756044948,
      "step": 15750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7688452005386353,
      "epoch": 5.7,
      "learning_rate": 5.720225466703601e-05,
      "loss": 0.6282,
      "step": 15760,
      "task_loss": 0.5565420389175415
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47321581840515137,
      "epoch": 5.7,
      "learning_rate": 5.7173458148541094e-05,
      "loss": 0.6024,
      "step": 15770,
      "task_loss": 0.37659692764282227
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4902139902114868,
      "epoch": 5.7,
      "learning_rate": 5.714452151455841e-05,
      "loss": 0.6281,
      "step": 15780,
      "task_loss": 0.9982061386108398
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44813448190689087,
      "epoch": 5.71,
      "learning_rate": 5.711544491429501e-05,
      "loss": 0.5142,
      "step": 15790,
      "task_loss": 0.7714956998825073
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4338916838169098,
      "epoch": 5.71,
      "learning_rate": 5.708622849767961e-05,
      "loss": 0.5278,
      "step": 15800,
      "task_loss": 0.4820694327354431
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4372040629386902,
      "epoch": 5.71,
      "learning_rate": 5.705687241536186e-05,
      "loss": 0.5699,
      "step": 15810,
      "task_loss": 0.41295021772384644
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3634551465511322,
      "epoch": 5.72,
      "learning_rate": 5.702737681871161e-05,
      "loss": 0.4913,
      "step": 15820,
      "task_loss": 0.31981927156448364
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47144439816474915,
      "epoch": 5.72,
      "learning_rate": 5.699774185981806e-05,
      "loss": 0.5732,
      "step": 15830,
      "task_loss": 0.3408505916595459
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.664949893951416,
      "epoch": 5.72,
      "learning_rate": 5.696796769148905e-05,
      "loss": 0.5635,
      "step": 15840,
      "task_loss": 0.5504066348075867
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5689000487327576,
      "epoch": 5.73,
      "learning_rate": 5.693805446725018e-05,
      "loss": 0.6337,
      "step": 15850,
      "task_loss": 0.7448662519454956
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5832275152206421,
      "epoch": 5.73,
      "learning_rate": 5.69080023413441e-05,
      "loss": 0.5428,
      "step": 15860,
      "task_loss": 0.5787875056266785
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4936749339103699,
      "epoch": 5.74,
      "learning_rate": 5.687781146872967e-05,
      "loss": 0.5589,
      "step": 15870,
      "task_loss": 0.5635548830032349
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4124121069908142,
      "epoch": 5.74,
      "learning_rate": 5.6847482005081207e-05,
      "loss": 0.5222,
      "step": 15880,
      "task_loss": 1.0899685621261597
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.390001118183136,
      "epoch": 5.74,
      "learning_rate": 5.681701410678759e-05,
      "loss": 0.4966,
      "step": 15890,
      "task_loss": 0.5182036757469177
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38561293482780457,
      "epoch": 5.75,
      "learning_rate": 5.678640793095156e-05,
      "loss": 0.5762,
      "step": 15900,
      "task_loss": 0.7452099323272705
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7303899526596069,
      "epoch": 5.75,
      "learning_rate": 5.6755663635388865e-05,
      "loss": 0.5726,
      "step": 15910,
      "task_loss": 0.6906106472015381
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5508267283439636,
      "epoch": 5.75,
      "learning_rate": 5.6724781378627425e-05,
      "loss": 0.6166,
      "step": 15920,
      "task_loss": 0.4653738737106323
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3965685963630676,
      "epoch": 5.76,
      "learning_rate": 5.669376131990653e-05,
      "loss": 0.5519,
      "step": 15930,
      "task_loss": 0.5775076746940613
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8321218490600586,
      "epoch": 5.76,
      "learning_rate": 5.666260361917604e-05,
      "loss": 0.5691,
      "step": 15940,
      "task_loss": 0.9377269148826599
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.712748110294342,
      "epoch": 5.76,
      "learning_rate": 5.6631308437095546e-05,
      "loss": 0.5579,
      "step": 15950,
      "task_loss": 0.767340898513794
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8006727695465088,
      "epoch": 5.77,
      "learning_rate": 5.6599875935033516e-05,
      "loss": 0.6169,
      "step": 15960,
      "task_loss": 0.9040712714195251
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5734758973121643,
      "epoch": 5.77,
      "learning_rate": 5.656830627506651e-05,
      "loss": 0.5505,
      "step": 15970,
      "task_loss": 0.7179234027862549
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5710708498954773,
      "epoch": 5.78,
      "learning_rate": 5.6536599619978306e-05,
      "loss": 0.5686,
      "step": 15980,
      "task_loss": 0.7058345079421997
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49838507175445557,
      "epoch": 5.78,
      "learning_rate": 5.650475613325907e-05,
      "loss": 0.5444,
      "step": 15990,
      "task_loss": 0.4772479236125946
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6272715330123901,
      "epoch": 5.78,
      "learning_rate": 5.647277597910456e-05,
      "loss": 0.5258,
      "step": 16000,
      "task_loss": 0.5559104681015015
    },
    {
      "epoch": 5.78,
      "eval_exact_match": 81.86376537369915,
      "eval_f1": 89.00954211887367,
      "step": 16000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45378392934799194,
      "epoch": 5.79,
      "learning_rate": 5.644065932241517e-05,
      "loss": 0.5763,
      "step": 16010,
      "task_loss": 0.5300541520118713
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.493353933095932,
      "epoch": 5.79,
      "learning_rate": 5.6408406328795195e-05,
      "loss": 0.5179,
      "step": 16020,
      "task_loss": 0.4974842071533203
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7274896502494812,
      "epoch": 5.79,
      "learning_rate": 5.6376017164551915e-05,
      "loss": 0.59,
      "step": 16030,
      "task_loss": 0.7033414840698242
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4151965379714966,
      "epoch": 5.8,
      "learning_rate": 5.6343491996694754e-05,
      "loss": 0.5246,
      "step": 16040,
      "task_loss": 0.7589873671531677
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40979576110839844,
      "epoch": 5.8,
      "learning_rate": 5.631083099293441e-05,
      "loss": 0.531,
      "step": 16050,
      "task_loss": 0.45069462060928345
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45870572328567505,
      "epoch": 5.8,
      "learning_rate": 5.627803432168202e-05,
      "loss": 0.5516,
      "step": 16060,
      "task_loss": 0.5535123348236084
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47782111167907715,
      "epoch": 5.81,
      "learning_rate": 5.6245102152048215e-05,
      "loss": 0.4949,
      "step": 16070,
      "task_loss": 0.8046015501022339
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5272634029388428,
      "epoch": 5.81,
      "learning_rate": 5.621203465384236e-05,
      "loss": 0.5679,
      "step": 16080,
      "task_loss": 0.5130640268325806
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40280604362487793,
      "epoch": 5.81,
      "learning_rate": 5.617883199757159e-05,
      "loss": 0.4614,
      "step": 16090,
      "task_loss": 0.4750958979129791
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6462051868438721,
      "epoch": 5.82,
      "learning_rate": 5.6145494354439964e-05,
      "loss": 0.5555,
      "step": 16100,
      "task_loss": 0.9935194253921509
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4317300319671631,
      "epoch": 5.82,
      "learning_rate": 5.611202189634758e-05,
      "loss": 0.5154,
      "step": 16110,
      "task_loss": 0.36278975009918213
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5830767154693604,
      "epoch": 5.83,
      "learning_rate": 5.6078414795889694e-05,
      "loss": 0.5193,
      "step": 16120,
      "task_loss": 0.2859882414340973
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39172929525375366,
      "epoch": 5.83,
      "learning_rate": 5.604467322635582e-05,
      "loss": 0.4694,
      "step": 16130,
      "task_loss": 0.44212770462036133
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6151372790336609,
      "epoch": 5.83,
      "learning_rate": 5.6010797361728816e-05,
      "loss": 0.4802,
      "step": 16140,
      "task_loss": 0.8025726079940796
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.612571656703949,
      "epoch": 5.84,
      "learning_rate": 5.597678737668405e-05,
      "loss": 0.5408,
      "step": 16150,
      "task_loss": 0.6253871917724609
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34145790338516235,
      "epoch": 5.84,
      "learning_rate": 5.594264344658843e-05,
      "loss": 0.4514,
      "step": 16160,
      "task_loss": 0.4706138074398041
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.56342613697052,
      "epoch": 5.84,
      "learning_rate": 5.5908365747499555e-05,
      "loss": 0.6194,
      "step": 16170,
      "task_loss": 0.40212613344192505
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4632798433303833,
      "epoch": 5.85,
      "learning_rate": 5.587395445616475e-05,
      "loss": 0.5395,
      "step": 16180,
      "task_loss": 0.6225385665893555
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5093396902084351,
      "epoch": 5.85,
      "learning_rate": 5.583940975002023e-05,
      "loss": 0.5351,
      "step": 16190,
      "task_loss": 0.6102060079574585
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3960115313529968,
      "epoch": 5.85,
      "learning_rate": 5.5804731807190085e-05,
      "loss": 0.5119,
      "step": 16200,
      "task_loss": 0.3458710014820099
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7457716464996338,
      "epoch": 5.86,
      "learning_rate": 5.576992080648548e-05,
      "loss": 0.4945,
      "step": 16210,
      "task_loss": 1.0185015201568604
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7339679002761841,
      "epoch": 5.86,
      "learning_rate": 5.57349769274036e-05,
      "loss": 0.6524,
      "step": 16220,
      "task_loss": 0.9575223922729492
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.48004066944122314,
      "epoch": 5.87,
      "learning_rate": 5.569990035012688e-05,
      "loss": 0.5535,
      "step": 16230,
      "task_loss": 0.6020759344100952
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37398168444633484,
      "epoch": 5.87,
      "learning_rate": 5.566469125552193e-05,
      "loss": 0.4726,
      "step": 16240,
      "task_loss": 0.555140495300293
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49550220370292664,
      "epoch": 5.87,
      "learning_rate": 5.562934982513867e-05,
      "loss": 0.6053,
      "step": 16250,
      "task_loss": 0.5213397741317749
    },
    {
      "epoch": 5.87,
      "eval_exact_match": 82.0151371807001,
      "eval_f1": 89.09085178965859,
      "step": 16250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49956950545310974,
      "epoch": 5.88,
      "learning_rate": 5.559387624120942e-05,
      "loss": 0.487,
      "step": 16260,
      "task_loss": 0.537997841835022
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43931645154953003,
      "epoch": 5.88,
      "learning_rate": 5.5558270686647886e-05,
      "loss": 0.5127,
      "step": 16270,
      "task_loss": 0.6900993585586548
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49603772163391113,
      "epoch": 5.88,
      "learning_rate": 5.55225333450483e-05,
      "loss": 0.4836,
      "step": 16280,
      "task_loss": 0.4945855438709259
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5353456139564514,
      "epoch": 5.89,
      "learning_rate": 5.5486664400684396e-05,
      "loss": 0.6114,
      "step": 16290,
      "task_loss": 0.5326354503631592
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6305358409881592,
      "epoch": 5.89,
      "learning_rate": 5.5450664038508526e-05,
      "loss": 0.5082,
      "step": 16300,
      "task_loss": 0.6449335813522339
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.48550814390182495,
      "epoch": 5.89,
      "learning_rate": 5.541453244415065e-05,
      "loss": 0.418,
      "step": 16310,
      "task_loss": 0.724381685256958
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5628600120544434,
      "epoch": 5.9,
      "learning_rate": 5.5378269803917425e-05,
      "loss": 0.509,
      "step": 16320,
      "task_loss": 0.6630486249923706
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5237171053886414,
      "epoch": 5.9,
      "learning_rate": 5.534187630479123e-05,
      "loss": 0.5697,
      "step": 16330,
      "task_loss": 0.5837256908416748
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4340498745441437,
      "epoch": 5.91,
      "learning_rate": 5.530535213442916e-05,
      "loss": 0.6226,
      "step": 16340,
      "task_loss": 0.5448073148727417
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.557012677192688,
      "epoch": 5.91,
      "learning_rate": 5.5268697481162134e-05,
      "loss": 0.5754,
      "step": 16350,
      "task_loss": 0.3911644220352173
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5198036432266235,
      "epoch": 5.91,
      "learning_rate": 5.523191253399388e-05,
      "loss": 0.521,
      "step": 16360,
      "task_loss": 0.6677054166793823
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.9684406518936157,
      "epoch": 5.92,
      "learning_rate": 5.519499748259993e-05,
      "loss": 0.5403,
      "step": 16370,
      "task_loss": 0.9720715284347534
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5757874250411987,
      "epoch": 5.92,
      "learning_rate": 5.515795251732672e-05,
      "loss": 0.4946,
      "step": 16380,
      "task_loss": 0.8197060227394104
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5475500822067261,
      "epoch": 5.92,
      "learning_rate": 5.5120777829190543e-05,
      "loss": 0.4821,
      "step": 16390,
      "task_loss": 0.47533178329467773
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4618098735809326,
      "epoch": 5.93,
      "learning_rate": 5.50834736098766e-05,
      "loss": 0.5336,
      "step": 16400,
      "task_loss": 0.6744464039802551
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5418887734413147,
      "epoch": 5.93,
      "learning_rate": 5.5046040051737976e-05,
      "loss": 0.5763,
      "step": 16410,
      "task_loss": 0.4931833744049072
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5653811693191528,
      "epoch": 5.93,
      "learning_rate": 5.500847734779469e-05,
      "loss": 0.504,
      "step": 16420,
      "task_loss": 0.5599584579467773
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.384804904460907,
      "epoch": 5.94,
      "learning_rate": 5.497078569173267e-05,
      "loss": 0.5194,
      "step": 16430,
      "task_loss": 1.2085001468658447
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6292977333068848,
      "epoch": 5.94,
      "learning_rate": 5.4932965277902784e-05,
      "loss": 0.4963,
      "step": 16440,
      "task_loss": 0.8269913792610168
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.611187756061554,
      "epoch": 5.95,
      "learning_rate": 5.4895016301319786e-05,
      "loss": 0.5997,
      "step": 16450,
      "task_loss": 0.6286755204200745
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5451836585998535,
      "epoch": 5.95,
      "learning_rate": 5.4856938957661377e-05,
      "loss": 0.6417,
      "step": 16460,
      "task_loss": 0.48253700137138367
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36945682764053345,
      "epoch": 5.95,
      "learning_rate": 5.481873344326713e-05,
      "loss": 0.4661,
      "step": 16470,
      "task_loss": 0.24678575992584229
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6069915294647217,
      "epoch": 5.96,
      "learning_rate": 5.478039995513753e-05,
      "loss": 0.5643,
      "step": 16480,
      "task_loss": 1.2377862930297852
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5773873925209045,
      "epoch": 5.96,
      "learning_rate": 5.474193869093293e-05,
      "loss": 0.5338,
      "step": 16490,
      "task_loss": 0.890839159488678
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5319252014160156,
      "epoch": 5.96,
      "learning_rate": 5.4703349848972554e-05,
      "loss": 0.5673,
      "step": 16500,
      "task_loss": 0.6719323992729187
    },
    {
      "epoch": 5.96,
      "eval_exact_match": 82.07190160832545,
      "eval_f1": 89.23505876854658,
      "step": 16500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4537544846534729,
      "epoch": 5.97,
      "learning_rate": 5.466463362823343e-05,
      "loss": 0.5421,
      "step": 16510,
      "task_loss": 0.8447656035423279
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3538123071193695,
      "epoch": 5.97,
      "learning_rate": 5.4625790228349424e-05,
      "loss": 0.5304,
      "step": 16520,
      "task_loss": 0.41902589797973633
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5340615510940552,
      "epoch": 5.97,
      "learning_rate": 5.458681984961015e-05,
      "loss": 0.5857,
      "step": 16530,
      "task_loss": 0.5699688196182251
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5087348222732544,
      "epoch": 5.98,
      "learning_rate": 5.4547722692960005e-05,
      "loss": 0.4812,
      "step": 16540,
      "task_loss": 0.8394026756286621
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4404750168323517,
      "epoch": 5.98,
      "learning_rate": 5.450849895999707e-05,
      "loss": 0.5653,
      "step": 16550,
      "task_loss": 0.5041899085044861
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2825411558151245,
      "epoch": 5.98,
      "learning_rate": 5.44691488529721e-05,
      "loss": 0.499,
      "step": 16560,
      "task_loss": 0.20059698820114136
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6841438412666321,
      "epoch": 5.99,
      "learning_rate": 5.442967257478748e-05,
      "loss": 0.5434,
      "step": 16570,
      "task_loss": 0.7802176475524902
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.8112098574638367,
      "epoch": 5.99,
      "learning_rate": 5.4390070328996166e-05,
      "loss": 0.6064,
      "step": 16580,
      "task_loss": 0.8238754272460938
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42704349756240845,
      "epoch": 6.0,
      "learning_rate": 5.435034231980067e-05,
      "loss": 0.5143,
      "step": 16590,
      "task_loss": 0.6214593052864075
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4533708691596985,
      "epoch": 6.0,
      "learning_rate": 5.431048875205195e-05,
      "loss": 0.5248,
      "step": 16600,
      "task_loss": 0.4769049882888794
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5088812112808228,
      "epoch": 6.0,
      "learning_rate": 5.427050983124843e-05,
      "loss": 0.4649,
      "step": 16610,
      "task_loss": 0.9433021545410156
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3258495628833771,
      "epoch": 6.01,
      "learning_rate": 5.423040576353483e-05,
      "loss": 0.4545,
      "step": 16620,
      "task_loss": 0.5422816276550293
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5611860752105713,
      "epoch": 6.01,
      "learning_rate": 5.4190176755701236e-05,
      "loss": 0.4954,
      "step": 16630,
      "task_loss": 0.4522503614425659
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37038782238960266,
      "epoch": 6.01,
      "learning_rate": 5.414982301518194e-05,
      "loss": 0.486,
      "step": 16640,
      "task_loss": 0.4855346083641052
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4425680339336395,
      "epoch": 6.02,
      "learning_rate": 5.410934475005439e-05,
      "loss": 0.435,
      "step": 16650,
      "task_loss": 0.4484761953353882
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3988327980041504,
      "epoch": 6.02,
      "learning_rate": 5.4068742169038124e-05,
      "loss": 0.4155,
      "step": 16660,
      "task_loss": 0.4220232367515564
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43297821283340454,
      "epoch": 6.02,
      "learning_rate": 5.402801548149372e-05,
      "loss": 0.4468,
      "step": 16670,
      "task_loss": 1.0445358753204346
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5155505537986755,
      "epoch": 6.03,
      "learning_rate": 5.398716489742166e-05,
      "loss": 0.4967,
      "step": 16680,
      "task_loss": 0.43968135118484497
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3766743838787079,
      "epoch": 6.03,
      "learning_rate": 5.3946190627461314e-05,
      "loss": 0.4799,
      "step": 16690,
      "task_loss": 0.3420065939426422
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34114527702331543,
      "epoch": 6.04,
      "learning_rate": 5.390509288288977e-05,
      "loss": 0.4431,
      "step": 16700,
      "task_loss": 0.38847649097442627
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3855423331260681,
      "epoch": 6.04,
      "learning_rate": 5.386387187562082e-05,
      "loss": 0.4372,
      "step": 16710,
      "task_loss": 0.21255548298358917
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3481213450431824,
      "epoch": 6.04,
      "learning_rate": 5.3822527818203865e-05,
      "loss": 0.4527,
      "step": 16720,
      "task_loss": 0.4471372365951538
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44838201999664307,
      "epoch": 6.05,
      "learning_rate": 5.378106092382275e-05,
      "loss": 0.4386,
      "step": 16730,
      "task_loss": 0.743310809135437
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.644596517086029,
      "epoch": 6.05,
      "learning_rate": 5.3739471406294725e-05,
      "loss": 0.473,
      "step": 16740,
      "task_loss": 0.6514346599578857
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2982652187347412,
      "epoch": 6.05,
      "learning_rate": 5.369775948006933e-05,
      "loss": 0.4327,
      "step": 16750,
      "task_loss": 0.28472357988357544
    },
    {
      "epoch": 6.05,
      "eval_exact_match": 81.88268684957427,
      "eval_f1": 89.02189324828987,
      "step": 16750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4026915431022644,
      "epoch": 6.06,
      "learning_rate": 5.365592536022728e-05,
      "loss": 0.4691,
      "step": 16760,
      "task_loss": 0.688092827796936
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.408430278301239,
      "epoch": 6.06,
      "learning_rate": 5.361396926247936e-05,
      "loss": 0.4222,
      "step": 16770,
      "task_loss": 0.34195762872695923
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5037705302238464,
      "epoch": 6.06,
      "learning_rate": 5.3571891403165345e-05,
      "loss": 0.4151,
      "step": 16780,
      "task_loss": 0.6692633032798767
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33279550075531006,
      "epoch": 6.07,
      "learning_rate": 5.3529691999252797e-05,
      "loss": 0.4216,
      "step": 16790,
      "task_loss": 0.763120174407959
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3310759961605072,
      "epoch": 6.07,
      "learning_rate": 5.348737126833605e-05,
      "loss": 0.4252,
      "step": 16800,
      "task_loss": 0.38220885396003723
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4700435698032379,
      "epoch": 6.08,
      "learning_rate": 5.344492942863501e-05,
      "loss": 0.4713,
      "step": 16810,
      "task_loss": 0.7272550463676453
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49290114641189575,
      "epoch": 6.08,
      "learning_rate": 5.340236669899409e-05,
      "loss": 0.4825,
      "step": 16820,
      "task_loss": 0.7776498794555664
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4069538712501526,
      "epoch": 6.08,
      "learning_rate": 5.3359683298881014e-05,
      "loss": 0.4863,
      "step": 16830,
      "task_loss": 0.6780685782432556
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4466571509838104,
      "epoch": 6.09,
      "learning_rate": 5.331687944838578e-05,
      "loss": 0.4775,
      "step": 16840,
      "task_loss": 0.6824575662612915
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6595938205718994,
      "epoch": 6.09,
      "learning_rate": 5.327395536821941e-05,
      "loss": 0.4408,
      "step": 16850,
      "task_loss": 1.1058135032653809
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39724981784820557,
      "epoch": 6.09,
      "learning_rate": 5.323091127971289e-05,
      "loss": 0.4776,
      "step": 16860,
      "task_loss": 0.5060951709747314
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4856202006340027,
      "epoch": 6.1,
      "learning_rate": 5.318774740481603e-05,
      "loss": 0.4653,
      "step": 16870,
      "task_loss": 0.4803681969642639
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.401211142539978,
      "epoch": 6.1,
      "learning_rate": 5.314446396609628e-05,
      "loss": 0.4307,
      "step": 16880,
      "task_loss": 0.3891042470932007
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3375917673110962,
      "epoch": 6.1,
      "learning_rate": 5.310106118673761e-05,
      "loss": 0.458,
      "step": 16890,
      "task_loss": 0.7722040414810181
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3473578989505768,
      "epoch": 6.11,
      "learning_rate": 5.305753929053932e-05,
      "loss": 0.4412,
      "step": 16900,
      "task_loss": 0.7287085056304932
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.680449366569519,
      "epoch": 6.11,
      "learning_rate": 5.301389850191497e-05,
      "loss": 0.4862,
      "step": 16910,
      "task_loss": 0.8735339045524597
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42973363399505615,
      "epoch": 6.11,
      "learning_rate": 5.297013904589114e-05,
      "loss": 0.4532,
      "step": 16920,
      "task_loss": 0.4185020625591278
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5793424844741821,
      "epoch": 6.12,
      "learning_rate": 5.292626114810628e-05,
      "loss": 0.4858,
      "step": 16930,
      "task_loss": 0.8029798269271851
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3389170169830322,
      "epoch": 6.12,
      "learning_rate": 5.28822650348096e-05,
      "loss": 0.4769,
      "step": 16940,
      "task_loss": 0.3712247610092163
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4885207414627075,
      "epoch": 6.13,
      "learning_rate": 5.2838150932859857e-05,
      "loss": 0.4678,
      "step": 16950,
      "task_loss": 0.6370507478713989
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.316753625869751,
      "epoch": 6.13,
      "learning_rate": 5.279391906972418e-05,
      "loss": 0.4271,
      "step": 16960,
      "task_loss": 0.5604386329650879
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41553622484207153,
      "epoch": 6.13,
      "learning_rate": 5.274956967347694e-05,
      "loss": 0.4632,
      "step": 16970,
      "task_loss": 0.49455636739730835
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3349422514438629,
      "epoch": 6.14,
      "learning_rate": 5.270510297279855e-05,
      "loss": 0.4382,
      "step": 16980,
      "task_loss": 0.6961212158203125
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4425978660583496,
      "epoch": 6.14,
      "learning_rate": 5.266051919697425e-05,
      "loss": 0.4478,
      "step": 16990,
      "task_loss": 0.5341584086418152
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.570014238357544,
      "epoch": 6.14,
      "learning_rate": 5.2615818575892986e-05,
      "loss": 0.4476,
      "step": 17000,
      "task_loss": 0.5794532895088196
    },
    {
      "epoch": 6.14,
      "eval_exact_match": 82.07190160832545,
      "eval_f1": 89.09099844346169,
      "step": 17000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5344833135604858,
      "epoch": 6.15,
      "learning_rate": 5.2571001340046186e-05,
      "loss": 0.4784,
      "step": 17010,
      "task_loss": 0.5975182056427002
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43689098954200745,
      "epoch": 6.15,
      "learning_rate": 5.2526067720526606e-05,
      "loss": 0.4314,
      "step": 17020,
      "task_loss": 0.8588857650756836
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36046090722084045,
      "epoch": 6.15,
      "learning_rate": 5.2481017949027086e-05,
      "loss": 0.4663,
      "step": 17030,
      "task_loss": 0.5059746503829956
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4314071536064148,
      "epoch": 6.16,
      "learning_rate": 5.2435852257839397e-05,
      "loss": 0.4958,
      "step": 17040,
      "task_loss": 0.6044473648071289
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4972867965698242,
      "epoch": 6.16,
      "learning_rate": 5.239057087985303e-05,
      "loss": 0.3913,
      "step": 17050,
      "task_loss": 0.7851539850234985
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4286130368709564,
      "epoch": 6.17,
      "learning_rate": 5.234517404855399e-05,
      "loss": 0.4531,
      "step": 17060,
      "task_loss": 0.44648587703704834
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4931001365184784,
      "epoch": 6.17,
      "learning_rate": 5.229966199802363e-05,
      "loss": 0.4906,
      "step": 17070,
      "task_loss": 0.4792635142803192
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4335430860519409,
      "epoch": 6.17,
      "learning_rate": 5.225403496293736e-05,
      "loss": 0.4316,
      "step": 17080,
      "task_loss": 0.2784348130226135
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5062400102615356,
      "epoch": 6.18,
      "learning_rate": 5.2208293178563524e-05,
      "loss": 0.4188,
      "step": 17090,
      "task_loss": 0.8841861486434937
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42178788781166077,
      "epoch": 6.18,
      "learning_rate": 5.216243688076215e-05,
      "loss": 0.4311,
      "step": 17100,
      "task_loss": 0.5392696261405945
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.48605018854141235,
      "epoch": 6.18,
      "learning_rate": 5.211646630598372e-05,
      "loss": 0.492,
      "step": 17110,
      "task_loss": 0.7885700464248657
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43214869499206543,
      "epoch": 6.19,
      "learning_rate": 5.207038169126799e-05,
      "loss": 0.4787,
      "step": 17120,
      "task_loss": 0.8961579203605652
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4661581516265869,
      "epoch": 6.19,
      "learning_rate": 5.2024183274242715e-05,
      "loss": 0.4191,
      "step": 17130,
      "task_loss": 0.6008694171905518
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5632531642913818,
      "epoch": 6.19,
      "learning_rate": 5.197787129312247e-05,
      "loss": 0.4638,
      "step": 17140,
      "task_loss": 0.6885539293289185
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.343812495470047,
      "epoch": 6.2,
      "learning_rate": 5.193144598670741e-05,
      "loss": 0.4286,
      "step": 17150,
      "task_loss": 0.5359995365142822
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4356578588485718,
      "epoch": 6.2,
      "learning_rate": 5.188490759438202e-05,
      "loss": 0.4945,
      "step": 17160,
      "task_loss": 0.7381937503814697
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3725375533103943,
      "epoch": 6.21,
      "learning_rate": 5.18382563561139e-05,
      "loss": 0.4391,
      "step": 17170,
      "task_loss": 0.5693670511245728
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38266217708587646,
      "epoch": 6.21,
      "learning_rate": 5.179149251245253e-05,
      "loss": 0.4688,
      "step": 17180,
      "task_loss": 0.5379250049591064
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5168706178665161,
      "epoch": 6.21,
      "learning_rate": 5.174461630452802e-05,
      "loss": 0.4385,
      "step": 17190,
      "task_loss": 0.6097722053527832
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3769886791706085,
      "epoch": 6.22,
      "learning_rate": 5.169762797404983e-05,
      "loss": 0.4674,
      "step": 17200,
      "task_loss": 0.3455650210380554
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.590087890625,
      "epoch": 6.22,
      "learning_rate": 5.165052776330562e-05,
      "loss": 0.4498,
      "step": 17210,
      "task_loss": 0.5600070357322693
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5225914716720581,
      "epoch": 6.22,
      "learning_rate": 5.16033159151599e-05,
      "loss": 0.4534,
      "step": 17220,
      "task_loss": 1.0604463815689087
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4769077003002167,
      "epoch": 6.23,
      "learning_rate": 5.1555992673052845e-05,
      "loss": 0.4587,
      "step": 17230,
      "task_loss": 0.7149950265884399
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35209062695503235,
      "epoch": 6.23,
      "learning_rate": 5.150855828099901e-05,
      "loss": 0.449,
      "step": 17240,
      "task_loss": 0.9201719164848328
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2952001988887787,
      "epoch": 6.23,
      "learning_rate": 5.146101298358604e-05,
      "loss": 0.4543,
      "step": 17250,
      "task_loss": 0.3284863233566284
    },
    {
      "epoch": 6.23,
      "eval_exact_match": 82.06244087038789,
      "eval_f1": 89.28534999216927,
      "step": 17250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45337045192718506,
      "epoch": 6.24,
      "learning_rate": 5.1413357025973514e-05,
      "loss": 0.3967,
      "step": 17260,
      "task_loss": 0.4615285396575928
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5968666076660156,
      "epoch": 6.24,
      "learning_rate": 5.1365590653891554e-05,
      "loss": 0.4762,
      "step": 17270,
      "task_loss": 0.6431958675384521
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47852158546447754,
      "epoch": 6.25,
      "learning_rate": 5.131771411363964e-05,
      "loss": 0.4462,
      "step": 17280,
      "task_loss": 0.8368129730224609
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4786396026611328,
      "epoch": 6.25,
      "learning_rate": 5.1269727652085315e-05,
      "loss": 0.4775,
      "step": 17290,
      "task_loss": 0.828134298324585
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2864540219306946,
      "epoch": 6.25,
      "learning_rate": 5.1221631516662914e-05,
      "loss": 0.4184,
      "step": 17300,
      "task_loss": 0.5259373188018799
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6025077104568481,
      "epoch": 6.26,
      "learning_rate": 5.117342595537228e-05,
      "loss": 0.4859,
      "step": 17310,
      "task_loss": 0.6346871256828308
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37674152851104736,
      "epoch": 6.26,
      "learning_rate": 5.1125111216777496e-05,
      "loss": 0.4247,
      "step": 17320,
      "task_loss": 0.42210888862609863
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39654457569122314,
      "epoch": 6.26,
      "learning_rate": 5.10766875500056e-05,
      "loss": 0.4151,
      "step": 17330,
      "task_loss": 0.45702940225601196
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4761405289173126,
      "epoch": 6.27,
      "learning_rate": 5.10281552047453e-05,
      "loss": 0.4388,
      "step": 17340,
      "task_loss": 0.564466118812561
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4193245470523834,
      "epoch": 6.27,
      "learning_rate": 5.09795144312457e-05,
      "loss": 0.4997,
      "step": 17350,
      "task_loss": 0.3380679190158844
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3163435161113739,
      "epoch": 6.27,
      "learning_rate": 5.0930765480314974e-05,
      "loss": 0.4753,
      "step": 17360,
      "task_loss": 0.28184810280799866
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4957866668701172,
      "epoch": 6.28,
      "learning_rate": 5.0881908603319105e-05,
      "loss": 0.4457,
      "step": 17370,
      "task_loss": 0.8675940036773682
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3084317445755005,
      "epoch": 6.28,
      "learning_rate": 5.083294405218059e-05,
      "loss": 0.4188,
      "step": 17380,
      "task_loss": 0.35364723205566406
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41741424798965454,
      "epoch": 6.28,
      "learning_rate": 5.07838720793771e-05,
      "loss": 0.5694,
      "step": 17390,
      "task_loss": 0.5752847194671631
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3092053532600403,
      "epoch": 6.29,
      "learning_rate": 5.073469293794025e-05,
      "loss": 0.4577,
      "step": 17400,
      "task_loss": 0.5759758949279785
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40555286407470703,
      "epoch": 6.29,
      "learning_rate": 5.0685406881454215e-05,
      "loss": 0.4448,
      "step": 17410,
      "task_loss": 0.7658182382583618
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3874298334121704,
      "epoch": 6.3,
      "learning_rate": 5.0636014164054486e-05,
      "loss": 0.4463,
      "step": 17420,
      "task_loss": 0.6377713680267334
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.686197817325592,
      "epoch": 6.3,
      "learning_rate": 5.0586515040426506e-05,
      "loss": 0.4981,
      "step": 17430,
      "task_loss": 0.7246427536010742
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37397500872612,
      "epoch": 6.3,
      "learning_rate": 5.053690976580443e-05,
      "loss": 0.4869,
      "step": 17440,
      "task_loss": 0.693129301071167
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4268683195114136,
      "epoch": 6.31,
      "learning_rate": 5.0487198595969706e-05,
      "loss": 0.4614,
      "step": 17450,
      "task_loss": 0.3804856538772583
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45956653356552124,
      "epoch": 6.31,
      "learning_rate": 5.0437381787249864e-05,
      "loss": 0.4549,
      "step": 17460,
      "task_loss": 0.8962446451187134
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4504271149635315,
      "epoch": 6.31,
      "learning_rate": 5.03874595965171e-05,
      "loss": 0.5028,
      "step": 17470,
      "task_loss": 0.4768270254135132
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4447179436683655,
      "epoch": 6.32,
      "learning_rate": 5.033743228118704e-05,
      "loss": 0.4214,
      "step": 17480,
      "task_loss": 0.49524736404418945
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6099137663841248,
      "epoch": 6.32,
      "learning_rate": 5.028730009921733e-05,
      "loss": 0.5059,
      "step": 17490,
      "task_loss": 0.5523144006729126
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5245805978775024,
      "epoch": 6.32,
      "learning_rate": 5.023706330910635e-05,
      "loss": 0.4337,
      "step": 17500,
      "task_loss": 0.48322606086730957
    },
    {
      "epoch": 6.32,
      "eval_exact_match": 82.15704824976348,
      "eval_f1": 89.3212773932139,
      "step": 17500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6089495420455933,
      "epoch": 6.33,
      "learning_rate": 5.0186722169891904e-05,
      "loss": 0.4971,
      "step": 17510,
      "task_loss": 1.3261122703552246
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42009443044662476,
      "epoch": 6.33,
      "learning_rate": 5.013627694114982e-05,
      "loss": 0.4783,
      "step": 17520,
      "task_loss": 0.6886181831359863
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3833591938018799,
      "epoch": 6.34,
      "learning_rate": 5.008572788299266e-05,
      "loss": 0.4261,
      "step": 17530,
      "task_loss": 0.498070627450943
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4751524329185486,
      "epoch": 6.34,
      "learning_rate": 5.0035075256068377e-05,
      "loss": 0.4438,
      "step": 17540,
      "task_loss": 0.8459548950195312
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5234730839729309,
      "epoch": 6.34,
      "learning_rate": 4.998431932155894e-05,
      "loss": 0.4572,
      "step": 17550,
      "task_loss": 0.5055698752403259
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4609374403953552,
      "epoch": 6.35,
      "learning_rate": 4.993346034117901e-05,
      "loss": 0.4893,
      "step": 17560,
      "task_loss": 0.8448480367660522
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3884456157684326,
      "epoch": 6.35,
      "learning_rate": 4.9882498577174596e-05,
      "loss": 0.4746,
      "step": 17570,
      "task_loss": 0.4255600571632385
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5293577909469604,
      "epoch": 6.35,
      "learning_rate": 4.98314342923217e-05,
      "loss": 0.4517,
      "step": 17580,
      "task_loss": 0.5622125864028931
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45298856496810913,
      "epoch": 6.36,
      "learning_rate": 4.9780267749924934e-05,
      "loss": 0.4589,
      "step": 17590,
      "task_loss": 0.5631417036056519
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5164973735809326,
      "epoch": 6.36,
      "learning_rate": 4.9728999213816197e-05,
      "loss": 0.4505,
      "step": 17600,
      "task_loss": 0.5395184755325317
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36571723222732544,
      "epoch": 6.36,
      "learning_rate": 4.9677628948353304e-05,
      "loss": 0.4707,
      "step": 17610,
      "task_loss": 0.5570918321609497
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.408827006816864,
      "epoch": 6.37,
      "learning_rate": 4.962615721841863e-05,
      "loss": 0.43,
      "step": 17620,
      "task_loss": 0.5332542657852173
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.383866548538208,
      "epoch": 6.37,
      "learning_rate": 4.9574584289417695e-05,
      "loss": 0.4391,
      "step": 17630,
      "task_loss": 0.4717841148376465
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3984270393848419,
      "epoch": 6.38,
      "learning_rate": 4.952291042727788e-05,
      "loss": 0.4708,
      "step": 17640,
      "task_loss": 0.685359001159668
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5164908170700073,
      "epoch": 6.38,
      "learning_rate": 4.9471135898447e-05,
      "loss": 0.5033,
      "step": 17650,
      "task_loss": 0.6516079902648926
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5407741069793701,
      "epoch": 6.38,
      "learning_rate": 4.9419260969891926e-05,
      "loss": 0.4793,
      "step": 17660,
      "task_loss": 0.7795162200927734
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3706479072570801,
      "epoch": 6.39,
      "learning_rate": 4.936728590909724e-05,
      "loss": 0.4658,
      "step": 17670,
      "task_loss": 0.6481907367706299
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3029181957244873,
      "epoch": 6.39,
      "learning_rate": 4.9320422962809235e-05,
      "loss": 0.4577,
      "step": 17680,
      "task_loss": 0.4625958800315857
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3568081259727478,
      "epoch": 6.39,
      "learning_rate": 4.926825838952736e-05,
      "loss": 0.4723,
      "step": 17690,
      "task_loss": 0.5693346261978149
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.387743204832077,
      "epoch": 6.4,
      "learning_rate": 4.921599446262594e-05,
      "loss": 0.4436,
      "step": 17700,
      "task_loss": 0.5526187419891357
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28367936611175537,
      "epoch": 6.4,
      "learning_rate": 4.916363145159537e-05,
      "loss": 0.459,
      "step": 17710,
      "task_loss": 0.28557759523391724
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5074309706687927,
      "epoch": 6.4,
      "learning_rate": 4.911116962643693e-05,
      "loss": 0.4521,
      "step": 17720,
      "task_loss": 0.7370518445968628
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42953354120254517,
      "epoch": 6.41,
      "learning_rate": 4.9058609257661456e-05,
      "loss": 0.4991,
      "step": 17730,
      "task_loss": 0.4267071485519409
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42934396862983704,
      "epoch": 6.41,
      "learning_rate": 4.9005950616287865e-05,
      "loss": 0.4038,
      "step": 17740,
      "task_loss": 0.5813796520233154
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.425101637840271,
      "epoch": 6.41,
      "learning_rate": 4.895319397384182e-05,
      "loss": 0.4605,
      "step": 17750,
      "task_loss": 0.6897194385528564
    },
    {
      "epoch": 6.41,
      "eval_exact_match": 82.23273415326395,
      "eval_f1": 89.3422644153395,
      "step": 17750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4546336531639099,
      "epoch": 6.42,
      "learning_rate": 4.8900339602354324e-05,
      "loss": 0.4959,
      "step": 17760,
      "task_loss": 0.6193094253540039
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5773555040359497,
      "epoch": 6.42,
      "learning_rate": 4.884738777436027e-05,
      "loss": 0.5355,
      "step": 17770,
      "task_loss": 0.8573752641677856
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4337666630744934,
      "epoch": 6.43,
      "learning_rate": 4.87943387628971e-05,
      "loss": 0.4405,
      "step": 17780,
      "task_loss": 0.5380644202232361
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5035710334777832,
      "epoch": 6.43,
      "learning_rate": 4.874119284150336e-05,
      "loss": 0.4727,
      "step": 17790,
      "task_loss": 0.7339311838150024
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39960795640945435,
      "epoch": 6.43,
      "learning_rate": 4.868795028421728e-05,
      "loss": 0.4856,
      "step": 17800,
      "task_loss": 0.28274768590927124
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42659035325050354,
      "epoch": 6.44,
      "learning_rate": 4.86346113655754e-05,
      "loss": 0.4335,
      "step": 17810,
      "task_loss": 0.7971694469451904
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.509783148765564,
      "epoch": 6.44,
      "learning_rate": 4.8581176360611114e-05,
      "loss": 0.474,
      "step": 17820,
      "task_loss": 0.7136898636817932
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5505311489105225,
      "epoch": 6.44,
      "learning_rate": 4.852764554485328e-05,
      "loss": 0.4862,
      "step": 17830,
      "task_loss": 0.5040520429611206
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3778039515018463,
      "epoch": 6.45,
      "learning_rate": 4.8474019194324777e-05,
      "loss": 0.4955,
      "step": 17840,
      "task_loss": 0.4302838444709778
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38377246260643005,
      "epoch": 6.45,
      "learning_rate": 4.8420297585541116e-05,
      "loss": 0.4369,
      "step": 17850,
      "task_loss": 0.646184504032135
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47774505615234375,
      "epoch": 6.45,
      "learning_rate": 4.836648099550896e-05,
      "loss": 0.4616,
      "step": 17860,
      "task_loss": 0.9298821091651917
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43539664149284363,
      "epoch": 6.46,
      "learning_rate": 4.8312569701724754e-05,
      "loss": 0.432,
      "step": 17870,
      "task_loss": 0.5792849063873291
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4134996831417084,
      "epoch": 6.46,
      "learning_rate": 4.8258563982173244e-05,
      "loss": 0.4159,
      "step": 17880,
      "task_loss": 0.7238243818283081
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37172645330429077,
      "epoch": 6.47,
      "learning_rate": 4.82044641153261e-05,
      "loss": 0.4143,
      "step": 17890,
      "task_loss": 0.4079241156578064
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44674891233444214,
      "epoch": 6.47,
      "learning_rate": 4.8150270380140414e-05,
      "loss": 0.4914,
      "step": 17900,
      "task_loss": 0.38229721784591675
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5314511060714722,
      "epoch": 6.47,
      "learning_rate": 4.809598305605732e-05,
      "loss": 0.4783,
      "step": 17910,
      "task_loss": 0.935704231262207
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5900159478187561,
      "epoch": 6.48,
      "learning_rate": 4.8041602423000505e-05,
      "loss": 0.4613,
      "step": 17920,
      "task_loss": 0.8068616986274719
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3018052577972412,
      "epoch": 6.48,
      "learning_rate": 4.79871287613748e-05,
      "loss": 0.4789,
      "step": 17930,
      "task_loss": 0.6351147890090942
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44554102420806885,
      "epoch": 6.48,
      "learning_rate": 4.793256235206473e-05,
      "loss": 0.4141,
      "step": 17940,
      "task_loss": 0.5927938222885132
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40561315417289734,
      "epoch": 6.49,
      "learning_rate": 4.787790347643305e-05,
      "loss": 0.4314,
      "step": 17950,
      "task_loss": 0.4989495277404785
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5982292294502258,
      "epoch": 6.49,
      "learning_rate": 4.782315241631929e-05,
      "loss": 0.483,
      "step": 17960,
      "task_loss": 1.0493371486663818
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46976473927497864,
      "epoch": 6.49,
      "learning_rate": 4.776830945403833e-05,
      "loss": 0.4004,
      "step": 17970,
      "task_loss": 0.8933740854263306
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36101314425468445,
      "epoch": 6.5,
      "learning_rate": 4.771337487237894e-05,
      "loss": 0.4111,
      "step": 17980,
      "task_loss": 0.6764540672302246
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38094139099121094,
      "epoch": 6.5,
      "learning_rate": 4.7658348954602285e-05,
      "loss": 0.4418,
      "step": 17990,
      "task_loss": 0.8435592651367188
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31571972370147705,
      "epoch": 6.51,
      "learning_rate": 4.76032319844405e-05,
      "loss": 0.4856,
      "step": 18000,
      "task_loss": 0.4472782015800476
    },
    {
      "epoch": 6.51,
      "eval_exact_match": 82.02459791863765,
      "eval_f1": 89.19706431714104,
      "step": 18000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3489418029785156,
      "epoch": 6.51,
      "learning_rate": 4.754802424609521e-05,
      "loss": 0.4659,
      "step": 18010,
      "task_loss": 0.3588530421257019
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47085440158843994,
      "epoch": 6.51,
      "learning_rate": 4.74927260242361e-05,
      "loss": 0.4414,
      "step": 18020,
      "task_loss": 0.8189518451690674
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49984803795814514,
      "epoch": 6.52,
      "learning_rate": 4.7437337603999376e-05,
      "loss": 0.423,
      "step": 18030,
      "task_loss": 0.3147749900817871
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34126660227775574,
      "epoch": 6.52,
      "learning_rate": 4.7381859270986374e-05,
      "loss": 0.4315,
      "step": 18040,
      "task_loss": 0.3727048337459564
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.48777493834495544,
      "epoch": 6.52,
      "learning_rate": 4.732629131126202e-05,
      "loss": 0.4487,
      "step": 18050,
      "task_loss": 0.4514695405960083
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44182878732681274,
      "epoch": 6.53,
      "learning_rate": 4.727063401135341e-05,
      "loss": 0.4873,
      "step": 18060,
      "task_loss": 0.5817041397094727
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49510326981544495,
      "epoch": 6.53,
      "learning_rate": 4.7214887658248296e-05,
      "loss": 0.4561,
      "step": 18070,
      "task_loss": 0.5546749830245972
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33820611238479614,
      "epoch": 6.53,
      "learning_rate": 4.715905253939361e-05,
      "loss": 0.4116,
      "step": 18080,
      "task_loss": 0.45475584268569946
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35808318853378296,
      "epoch": 6.54,
      "learning_rate": 4.710312894269402e-05,
      "loss": 0.4597,
      "step": 18090,
      "task_loss": 0.5782244205474854
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3415292203426361,
      "epoch": 6.54,
      "learning_rate": 4.704711715651038e-05,
      "loss": 0.4549,
      "step": 18100,
      "task_loss": 0.5431405901908875
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.350860595703125,
      "epoch": 6.54,
      "learning_rate": 4.699101746965829e-05,
      "loss": 0.4275,
      "step": 18110,
      "task_loss": 0.34007489681243896
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4397695064544678,
      "epoch": 6.55,
      "learning_rate": 4.6934830171406636e-05,
      "loss": 0.4853,
      "step": 18120,
      "task_loss": 0.5518962144851685
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4137437045574188,
      "epoch": 6.55,
      "learning_rate": 4.687855555147597e-05,
      "loss": 0.4352,
      "step": 18130,
      "task_loss": 0.4553642272949219
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45643168687820435,
      "epoch": 6.56,
      "learning_rate": 4.682219390003719e-05,
      "loss": 0.4659,
      "step": 18140,
      "task_loss": 0.5998508930206299
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5669125914573669,
      "epoch": 6.56,
      "learning_rate": 4.676574550770991e-05,
      "loss": 0.4963,
      "step": 18150,
      "task_loss": 0.7089822292327881
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45733779668807983,
      "epoch": 6.56,
      "learning_rate": 4.6709210665561035e-05,
      "loss": 0.4983,
      "step": 18160,
      "task_loss": 0.47134196758270264
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4086102843284607,
      "epoch": 6.57,
      "learning_rate": 4.6652589665103204e-05,
      "loss": 0.4618,
      "step": 18170,
      "task_loss": 0.4143943786621094
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5107593536376953,
      "epoch": 6.57,
      "learning_rate": 4.659588279829335e-05,
      "loss": 0.4485,
      "step": 18180,
      "task_loss": 0.5046135187149048
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3629281222820282,
      "epoch": 6.57,
      "learning_rate": 4.653909035753114e-05,
      "loss": 0.431,
      "step": 18190,
      "task_loss": 0.6013154983520508
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4908069372177124,
      "epoch": 6.58,
      "learning_rate": 4.648221263565751e-05,
      "loss": 0.4443,
      "step": 18200,
      "task_loss": 0.5517657995223999
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36872684955596924,
      "epoch": 6.58,
      "learning_rate": 4.642524992595309e-05,
      "loss": 0.4264,
      "step": 18210,
      "task_loss": 0.3264022171497345
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5466926097869873,
      "epoch": 6.58,
      "learning_rate": 4.636820252213679e-05,
      "loss": 0.4298,
      "step": 18220,
      "task_loss": 0.9148642420768738
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5774287581443787,
      "epoch": 6.59,
      "learning_rate": 4.63110707183642e-05,
      "loss": 0.4806,
      "step": 18230,
      "task_loss": 0.44062086939811707
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3985748887062073,
      "epoch": 6.59,
      "learning_rate": 4.6253854809226115e-05,
      "loss": 0.4406,
      "step": 18240,
      "task_loss": 0.4817260801792145
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2817017734050751,
      "epoch": 6.6,
      "learning_rate": 4.6196555089747e-05,
      "loss": 0.425,
      "step": 18250,
      "task_loss": 0.5607198476791382
    },
    {
      "epoch": 6.6,
      "eval_exact_match": 82.28003784295176,
      "eval_f1": 89.29688426511828,
      "step": 18250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38135290145874023,
      "epoch": 6.6,
      "learning_rate": 4.6139171855383473e-05,
      "loss": 0.4354,
      "step": 18260,
      "task_loss": 0.7876352667808533
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40633225440979004,
      "epoch": 6.6,
      "learning_rate": 4.608170540202279e-05,
      "loss": 0.4319,
      "step": 18270,
      "task_loss": 0.35161352157592773
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43270987272262573,
      "epoch": 6.61,
      "learning_rate": 4.602415602598132e-05,
      "loss": 0.4343,
      "step": 18280,
      "task_loss": 0.5352205634117126
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4512401223182678,
      "epoch": 6.61,
      "learning_rate": 4.5966524024002976e-05,
      "loss": 0.49,
      "step": 18290,
      "task_loss": 0.3575047552585602
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5585800409317017,
      "epoch": 6.61,
      "learning_rate": 4.590880969325774e-05,
      "loss": 0.4595,
      "step": 18300,
      "task_loss": 0.6749407052993774
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39090049266815186,
      "epoch": 6.62,
      "learning_rate": 4.585101333134013e-05,
      "loss": 0.491,
      "step": 18310,
      "task_loss": 0.5440311431884766
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7067247033119202,
      "epoch": 6.62,
      "learning_rate": 4.5793135236267626e-05,
      "loss": 0.5154,
      "step": 18320,
      "task_loss": 0.8724928498268127
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36196187138557434,
      "epoch": 6.62,
      "learning_rate": 4.573517570647912e-05,
      "loss": 0.4102,
      "step": 18330,
      "task_loss": 0.6686896681785583
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.620430052280426,
      "epoch": 6.63,
      "learning_rate": 4.567713504083346e-05,
      "loss": 0.4554,
      "step": 18340,
      "task_loss": 0.8327699899673462
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40227943658828735,
      "epoch": 6.63,
      "learning_rate": 4.5619013538607814e-05,
      "loss": 0.5117,
      "step": 18350,
      "task_loss": 0.7482736110687256
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47123730182647705,
      "epoch": 6.64,
      "learning_rate": 4.556081149949621e-05,
      "loss": 0.456,
      "step": 18360,
      "task_loss": 0.5184352993965149
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3152047395706177,
      "epoch": 6.64,
      "learning_rate": 4.5502529223607936e-05,
      "loss": 0.4324,
      "step": 18370,
      "task_loss": 0.5678387880325317
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45514464378356934,
      "epoch": 6.64,
      "learning_rate": 4.5444167011466e-05,
      "loss": 0.4807,
      "step": 18380,
      "task_loss": 0.612543523311615
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3955601453781128,
      "epoch": 6.65,
      "learning_rate": 4.5385725164005585e-05,
      "loss": 0.4173,
      "step": 18390,
      "task_loss": 0.455796480178833
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.396673321723938,
      "epoch": 6.65,
      "learning_rate": 4.5327203982572514e-05,
      "loss": 0.4673,
      "step": 18400,
      "task_loss": 0.5756609439849854
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3982319235801697,
      "epoch": 6.65,
      "learning_rate": 4.526860376892167e-05,
      "loss": 0.4573,
      "step": 18410,
      "task_loss": 0.7715901136398315
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5171875953674316,
      "epoch": 6.66,
      "learning_rate": 4.5209924825215474e-05,
      "loss": 0.4395,
      "step": 18420,
      "task_loss": 0.8181601166725159
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4994076192378998,
      "epoch": 6.66,
      "learning_rate": 4.515116745402229e-05,
      "loss": 0.4798,
      "step": 18430,
      "task_loss": 0.5784910917282104
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44892919063568115,
      "epoch": 6.66,
      "learning_rate": 4.509233195831487e-05,
      "loss": 0.4927,
      "step": 18440,
      "task_loss": 0.552018940448761
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6276687383651733,
      "epoch": 6.67,
      "learning_rate": 4.5033418641468816e-05,
      "loss": 0.4696,
      "step": 18450,
      "task_loss": 0.43832433223724365
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5821511745452881,
      "epoch": 6.67,
      "learning_rate": 4.497442780726101e-05,
      "loss": 0.5283,
      "step": 18460,
      "task_loss": 0.7905117869377136
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4565989673137665,
      "epoch": 6.68,
      "learning_rate": 4.4915359759868034e-05,
      "loss": 0.4865,
      "step": 18470,
      "task_loss": 0.676786482334137
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4760429859161377,
      "epoch": 6.68,
      "learning_rate": 4.485621480386459e-05,
      "loss": 0.4953,
      "step": 18480,
      "task_loss": 0.6198767423629761
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3461025059223175,
      "epoch": 6.68,
      "learning_rate": 4.479699324422195e-05,
      "loss": 0.4824,
      "step": 18490,
      "task_loss": 0.5995087623596191
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40634581446647644,
      "epoch": 6.69,
      "learning_rate": 4.473769538630642e-05,
      "loss": 0.4957,
      "step": 18500,
      "task_loss": 0.4016622006893158
    },
    {
      "epoch": 6.69,
      "eval_exact_match": 82.42194891201514,
      "eval_f1": 89.54146936438647,
      "step": 18500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.386337012052536,
      "epoch": 6.69,
      "learning_rate": 4.4678321535877664e-05,
      "loss": 0.4908,
      "step": 18510,
      "task_loss": 0.4025079607963562
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.48946017026901245,
      "epoch": 6.69,
      "learning_rate": 4.4618871999087255e-05,
      "loss": 0.4565,
      "step": 18520,
      "task_loss": 0.6569441556930542
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3229324221611023,
      "epoch": 6.7,
      "learning_rate": 4.455934708247696e-05,
      "loss": 0.4545,
      "step": 18530,
      "task_loss": 0.5276088714599609
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5329983830451965,
      "epoch": 6.7,
      "learning_rate": 4.449974709297729e-05,
      "loss": 0.49,
      "step": 18540,
      "task_loss": 0.4105309545993805
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5233172178268433,
      "epoch": 6.7,
      "learning_rate": 4.4440072337905815e-05,
      "loss": 0.4177,
      "step": 18550,
      "task_loss": 0.7585268616676331
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5894426107406616,
      "epoch": 6.71,
      "learning_rate": 4.438032312496565e-05,
      "loss": 0.4819,
      "step": 18560,
      "task_loss": 0.8770580291748047
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46357041597366333,
      "epoch": 6.71,
      "learning_rate": 4.432049976224383e-05,
      "loss": 0.5125,
      "step": 18570,
      "task_loss": 0.943016529083252
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3945166766643524,
      "epoch": 6.71,
      "learning_rate": 4.4260602558209715e-05,
      "loss": 0.4636,
      "step": 18580,
      "task_loss": 0.4682915210723877
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5548945665359497,
      "epoch": 6.72,
      "learning_rate": 4.420063182171345e-05,
      "loss": 0.4443,
      "step": 18590,
      "task_loss": 0.2853912115097046
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5039093494415283,
      "epoch": 6.72,
      "learning_rate": 4.414058786198431e-05,
      "loss": 0.4431,
      "step": 18600,
      "task_loss": 0.5872310400009155
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4797692894935608,
      "epoch": 6.73,
      "learning_rate": 4.408047098862914e-05,
      "loss": 0.5034,
      "step": 18610,
      "task_loss": 0.5820059776306152
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.361568421125412,
      "epoch": 6.73,
      "learning_rate": 4.402028151163076e-05,
      "loss": 0.45,
      "step": 18620,
      "task_loss": 0.9699276089668274
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3789913058280945,
      "epoch": 6.73,
      "learning_rate": 4.396001974134634e-05,
      "loss": 0.4786,
      "step": 18630,
      "task_loss": 0.5539405941963196
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5141831636428833,
      "epoch": 6.74,
      "learning_rate": 4.389968598850585e-05,
      "loss": 0.4835,
      "step": 18640,
      "task_loss": 1.006083607673645
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37998679280281067,
      "epoch": 6.74,
      "learning_rate": 4.38392805642104e-05,
      "loss": 0.4461,
      "step": 18650,
      "task_loss": 0.8515222668647766
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4921640157699585,
      "epoch": 6.74,
      "learning_rate": 4.3778803779930655e-05,
      "loss": 0.4936,
      "step": 18660,
      "task_loss": 0.9974778890609741
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4045303463935852,
      "epoch": 6.75,
      "learning_rate": 4.371825594750528e-05,
      "loss": 0.438,
      "step": 18670,
      "task_loss": 0.9138513803482056
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4021495580673218,
      "epoch": 6.75,
      "learning_rate": 4.365763737913924e-05,
      "loss": 0.4834,
      "step": 18680,
      "task_loss": 0.5079997777938843
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5624068975448608,
      "epoch": 6.75,
      "learning_rate": 4.359694838740225e-05,
      "loss": 0.4978,
      "step": 18690,
      "task_loss": 0.5158367156982422
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4766731858253479,
      "epoch": 6.76,
      "learning_rate": 4.353618928522718e-05,
      "loss": 0.5086,
      "step": 18700,
      "task_loss": 0.3754899501800537
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3693751096725464,
      "epoch": 6.76,
      "learning_rate": 4.3475360385908385e-05,
      "loss": 0.4476,
      "step": 18710,
      "task_loss": 0.4243190884590149
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4787391126155853,
      "epoch": 6.77,
      "learning_rate": 4.3414462003100126e-05,
      "loss": 0.4959,
      "step": 18720,
      "task_loss": 0.8037554025650024
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5594469308853149,
      "epoch": 6.77,
      "learning_rate": 4.335349445081493e-05,
      "loss": 0.5292,
      "step": 18730,
      "task_loss": 0.5515763759613037
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49907881021499634,
      "epoch": 6.77,
      "learning_rate": 4.329245804342201e-05,
      "loss": 0.4717,
      "step": 18740,
      "task_loss": 0.6888168454170227
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3339458107948303,
      "epoch": 6.78,
      "learning_rate": 4.3231353095645604e-05,
      "loss": 0.413,
      "step": 18750,
      "task_loss": 0.3833264112472534
    },
    {
      "epoch": 6.78,
      "eval_exact_match": 82.639545884579,
      "eval_f1": 89.63747801805509,
      "step": 18750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44776710867881775,
      "epoch": 6.78,
      "learning_rate": 4.317017992256337e-05,
      "loss": 0.4343,
      "step": 18760,
      "task_loss": 0.36081254482269287
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5654640197753906,
      "epoch": 6.78,
      "learning_rate": 4.310893883960476e-05,
      "loss": 0.4711,
      "step": 18770,
      "task_loss": 0.7916117906570435
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47904425859451294,
      "epoch": 6.79,
      "learning_rate": 4.304763016254939e-05,
      "loss": 0.4494,
      "step": 18780,
      "task_loss": 0.6122930645942688
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5881616473197937,
      "epoch": 6.79,
      "learning_rate": 4.298625420752541e-05,
      "loss": 0.4592,
      "step": 18790,
      "task_loss": 0.9475069642066956
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49008673429489136,
      "epoch": 6.79,
      "learning_rate": 4.29248112910079e-05,
      "loss": 0.4432,
      "step": 18800,
      "task_loss": 0.3713112473487854
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5173578262329102,
      "epoch": 6.8,
      "learning_rate": 4.286330172981718e-05,
      "loss": 0.437,
      "step": 18810,
      "task_loss": 0.8030617237091064
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5325472950935364,
      "epoch": 6.8,
      "learning_rate": 4.2801725841117244e-05,
      "loss": 0.4291,
      "step": 18820,
      "task_loss": 0.7275118827819824
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41245532035827637,
      "epoch": 6.81,
      "learning_rate": 4.274008394241407e-05,
      "loss": 0.4659,
      "step": 18830,
      "task_loss": 0.6393299102783203
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3188970685005188,
      "epoch": 6.81,
      "learning_rate": 4.267837635155402e-05,
      "loss": 0.4185,
      "step": 18840,
      "task_loss": 0.23499251902103424
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.442760169506073,
      "epoch": 6.81,
      "learning_rate": 4.2616603386722185e-05,
      "loss": 0.4602,
      "step": 18850,
      "task_loss": 0.6652241945266724
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46997952461242676,
      "epoch": 6.82,
      "learning_rate": 4.2554765366440736e-05,
      "loss": 0.4335,
      "step": 18860,
      "task_loss": 0.46291276812553406
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4596181809902191,
      "epoch": 6.82,
      "learning_rate": 4.249286260956732e-05,
      "loss": 0.4364,
      "step": 18870,
      "task_loss": 0.43191826343536377
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3417094945907593,
      "epoch": 6.82,
      "learning_rate": 4.2430895435293335e-05,
      "loss": 0.4439,
      "step": 18880,
      "task_loss": 0.45981624722480774
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42703115940093994,
      "epoch": 6.83,
      "learning_rate": 4.2368864163142396e-05,
      "loss": 0.4596,
      "step": 18890,
      "task_loss": 0.5582728385925293
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4443010091781616,
      "epoch": 6.83,
      "learning_rate": 4.230676911296859e-05,
      "loss": 0.443,
      "step": 18900,
      "task_loss": 0.6267966032028198
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4458591938018799,
      "epoch": 6.83,
      "learning_rate": 4.2244610604954874e-05,
      "loss": 0.4731,
      "step": 18910,
      "task_loss": 0.81801837682724
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38323748111724854,
      "epoch": 6.84,
      "learning_rate": 4.218238895961143e-05,
      "loss": 0.477,
      "step": 18920,
      "task_loss": 1.1212983131408691
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7608765363693237,
      "epoch": 6.84,
      "learning_rate": 4.212010449777397e-05,
      "loss": 0.509,
      "step": 18930,
      "task_loss": 0.9213683009147644
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38869547843933105,
      "epoch": 6.84,
      "learning_rate": 4.2057757540602134e-05,
      "loss": 0.4825,
      "step": 18940,
      "task_loss": 0.5943039655685425
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.355135053396225,
      "epoch": 6.85,
      "learning_rate": 4.199534840957779e-05,
      "loss": 0.4582,
      "step": 18950,
      "task_loss": 0.9915359020233154
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4248897135257721,
      "epoch": 6.85,
      "learning_rate": 4.193287742650341e-05,
      "loss": 0.4378,
      "step": 18960,
      "task_loss": 0.6342843174934387
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4118010401725769,
      "epoch": 6.86,
      "learning_rate": 4.1870344913500387e-05,
      "loss": 0.4238,
      "step": 18970,
      "task_loss": 0.6452910900115967
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39560335874557495,
      "epoch": 6.86,
      "learning_rate": 4.180775119300738e-05,
      "loss": 0.4288,
      "step": 18980,
      "task_loss": 0.6294822096824646
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3214435577392578,
      "epoch": 6.86,
      "learning_rate": 4.174509658777867e-05,
      "loss": 0.4134,
      "step": 18990,
      "task_loss": 0.2638058662414551
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5299546122550964,
      "epoch": 6.87,
      "learning_rate": 4.168238142088247e-05,
      "loss": 0.459,
      "step": 19000,
      "task_loss": 0.5968713760375977
    },
    {
      "epoch": 6.87,
      "eval_exact_match": 82.23273415326395,
      "eval_f1": 89.16111651632492,
      "step": 19000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40809687972068787,
      "epoch": 6.87,
      "learning_rate": 4.1619606015699254e-05,
      "loss": 0.4248,
      "step": 19010,
      "task_loss": 0.7575485706329346
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5880671739578247,
      "epoch": 6.87,
      "learning_rate": 4.155677069592015e-05,
      "loss": 0.4437,
      "step": 19020,
      "task_loss": 0.7457284927368164
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3408096432685852,
      "epoch": 6.88,
      "learning_rate": 4.149387578554516e-05,
      "loss": 0.4678,
      "step": 19030,
      "task_loss": 0.8817156553268433
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.374206006526947,
      "epoch": 6.88,
      "learning_rate": 4.143092160888162e-05,
      "loss": 0.4213,
      "step": 19040,
      "task_loss": 0.5227565169334412
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3797913193702698,
      "epoch": 6.88,
      "learning_rate": 4.1367908490542424e-05,
      "loss": 0.409,
      "step": 19050,
      "task_loss": 0.44340643286705017
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3299967646598816,
      "epoch": 6.89,
      "learning_rate": 4.1304836755444396e-05,
      "loss": 0.4099,
      "step": 19060,
      "task_loss": 0.6511759757995605
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6297036409378052,
      "epoch": 6.89,
      "learning_rate": 4.1241706728806614e-05,
      "loss": 0.4631,
      "step": 19070,
      "task_loss": 0.9518231749534607
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27615126967430115,
      "epoch": 6.9,
      "learning_rate": 4.1178518736148726e-05,
      "loss": 0.4206,
      "step": 19080,
      "task_loss": 0.2932681739330292
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3559025526046753,
      "epoch": 6.9,
      "learning_rate": 4.111527310328926e-05,
      "loss": 0.4576,
      "step": 19090,
      "task_loss": 0.3928833603858948
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6219375133514404,
      "epoch": 6.9,
      "learning_rate": 4.105197015634399e-05,
      "loss": 0.4262,
      "step": 19100,
      "task_loss": 0.6940836906433105
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34312307834625244,
      "epoch": 6.91,
      "learning_rate": 4.0988610221724165e-05,
      "loss": 0.4914,
      "step": 19110,
      "task_loss": 0.5049322247505188
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5630295276641846,
      "epoch": 6.91,
      "learning_rate": 4.092519362613494e-05,
      "loss": 0.4919,
      "step": 19120,
      "task_loss": 0.8349260091781616
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3788623809814453,
      "epoch": 6.91,
      "learning_rate": 4.08617206965736e-05,
      "loss": 0.4542,
      "step": 19130,
      "task_loss": 0.4850049614906311
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4747163653373718,
      "epoch": 6.92,
      "learning_rate": 4.079819176032791e-05,
      "loss": 0.5041,
      "step": 19140,
      "task_loss": 0.7553776502609253
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47813260555267334,
      "epoch": 6.92,
      "learning_rate": 4.073460714497443e-05,
      "loss": 0.4269,
      "step": 19150,
      "task_loss": 0.6414305567741394
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34697532653808594,
      "epoch": 6.92,
      "learning_rate": 4.067096717837681e-05,
      "loss": 0.4815,
      "step": 19160,
      "task_loss": 0.41636407375335693
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4089597165584564,
      "epoch": 6.93,
      "learning_rate": 4.060727218868413e-05,
      "loss": 0.4731,
      "step": 19170,
      "task_loss": 0.5453172326087952
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.504246711730957,
      "epoch": 6.93,
      "learning_rate": 4.054352250432917e-05,
      "loss": 0.4974,
      "step": 19180,
      "task_loss": 0.4027463495731354
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4585569500923157,
      "epoch": 6.94,
      "learning_rate": 4.047971845402674e-05,
      "loss": 0.4536,
      "step": 19190,
      "task_loss": 0.49631595611572266
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3411811590194702,
      "epoch": 6.94,
      "learning_rate": 4.0415860366771986e-05,
      "loss": 0.4268,
      "step": 19200,
      "task_loss": 0.20603856444358826
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3674144744873047,
      "epoch": 6.94,
      "learning_rate": 4.0351948571838665e-05,
      "loss": 0.4593,
      "step": 19210,
      "task_loss": 0.7646399736404419
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5466402173042297,
      "epoch": 6.95,
      "learning_rate": 4.0287983398777485e-05,
      "loss": 0.512,
      "step": 19220,
      "task_loss": 1.0787473917007446
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3193405270576477,
      "epoch": 6.95,
      "learning_rate": 4.02239651774144e-05,
      "loss": 0.4307,
      "step": 19230,
      "task_loss": 0.415851354598999
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3150692582130432,
      "epoch": 6.95,
      "learning_rate": 4.015989423784887e-05,
      "loss": 0.4961,
      "step": 19240,
      "task_loss": 0.33591964840888977
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.658031165599823,
      "epoch": 6.96,
      "learning_rate": 4.009577091045222e-05,
      "loss": 0.4653,
      "step": 19250,
      "task_loss": 1.2693332433700562
    },
    {
      "epoch": 6.96,
      "eval_exact_match": 82.82876064333018,
      "eval_f1": 89.67725624937343,
      "step": 19250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4980925917625427,
      "epoch": 6.96,
      "learning_rate": 4.003159552586588e-05,
      "loss": 0.4392,
      "step": 19260,
      "task_loss": 0.6456292867660522
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4320065379142761,
      "epoch": 6.96,
      "learning_rate": 3.996736841499972e-05,
      "loss": 0.4594,
      "step": 19270,
      "task_loss": 0.566118597984314
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4771836996078491,
      "epoch": 6.97,
      "learning_rate": 3.990308990903031e-05,
      "loss": 0.4181,
      "step": 19280,
      "task_loss": 0.8952730298042297
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4210527837276459,
      "epoch": 6.97,
      "learning_rate": 3.983876033939925e-05,
      "loss": 0.465,
      "step": 19290,
      "task_loss": 0.7912062406539917
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43877533078193665,
      "epoch": 6.98,
      "learning_rate": 3.977438003781144e-05,
      "loss": 0.4059,
      "step": 19300,
      "task_loss": 0.8035497069358826
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3345944285392761,
      "epoch": 6.98,
      "learning_rate": 3.970994933623334e-05,
      "loss": 0.4686,
      "step": 19310,
      "task_loss": 0.6707329750061035
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6642370820045471,
      "epoch": 6.98,
      "learning_rate": 3.9645468566891326e-05,
      "loss": 0.4572,
      "step": 19320,
      "task_loss": 1.2499730587005615
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2801411747932434,
      "epoch": 6.99,
      "learning_rate": 3.958093806226994e-05,
      "loss": 0.4667,
      "step": 19330,
      "task_loss": 0.9053614139556885
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6855303049087524,
      "epoch": 6.99,
      "learning_rate": 3.951635815511014e-05,
      "loss": 0.4742,
      "step": 19340,
      "task_loss": 0.8596073985099792
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26637864112854004,
      "epoch": 6.99,
      "learning_rate": 3.9451729178407676e-05,
      "loss": 0.4094,
      "step": 19350,
      "task_loss": 0.24470390379428864
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5194672346115112,
      "epoch": 7.0,
      "learning_rate": 3.9387051465411245e-05,
      "loss": 0.4304,
      "step": 19360,
      "task_loss": 0.4968796968460083
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33181172609329224,
      "epoch": 7.0,
      "learning_rate": 3.93223253496209e-05,
      "loss": 0.4214,
      "step": 19370,
      "task_loss": 0.5487986207008362
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3815804421901703,
      "epoch": 7.0,
      "learning_rate": 3.925755116478628e-05,
      "loss": 0.3593,
      "step": 19380,
      "task_loss": 0.4483642876148224
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33547675609588623,
      "epoch": 7.01,
      "learning_rate": 3.919272924490484e-05,
      "loss": 0.4288,
      "step": 19390,
      "task_loss": 0.5775176286697388
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36918535828590393,
      "epoch": 7.01,
      "learning_rate": 3.91278599242202e-05,
      "loss": 0.3726,
      "step": 19400,
      "task_loss": 0.6206930875778198
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4281493127346039,
      "epoch": 7.01,
      "learning_rate": 3.9062943537220394e-05,
      "loss": 0.3883,
      "step": 19410,
      "task_loss": 0.4776683449745178
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3090469241142273,
      "epoch": 7.02,
      "learning_rate": 3.899798041863615e-05,
      "loss": 0.3848,
      "step": 19420,
      "task_loss": 0.5897292494773865
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37905633449554443,
      "epoch": 7.02,
      "learning_rate": 3.8932970903439134e-05,
      "loss": 0.4084,
      "step": 19430,
      "task_loss": 0.6165533661842346
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42853739857673645,
      "epoch": 7.03,
      "learning_rate": 3.886791532684028e-05,
      "loss": 0.452,
      "step": 19440,
      "task_loss": 0.6996277570724487
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44645875692367554,
      "epoch": 7.03,
      "learning_rate": 3.880281402428802e-05,
      "loss": 0.3729,
      "step": 19450,
      "task_loss": 0.6104946732521057
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43968871235847473,
      "epoch": 7.03,
      "learning_rate": 3.8737667331466554e-05,
      "loss": 0.4113,
      "step": 19460,
      "task_loss": 1.4070665836334229
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32989373803138733,
      "epoch": 7.04,
      "learning_rate": 3.8672475584294126e-05,
      "loss": 0.3745,
      "step": 19470,
      "task_loss": 0.7486312389373779
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3704969584941864,
      "epoch": 7.04,
      "learning_rate": 3.860723911892134e-05,
      "loss": 0.3999,
      "step": 19480,
      "task_loss": 0.90283203125
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41549235582351685,
      "epoch": 7.04,
      "learning_rate": 3.8541958271729304e-05,
      "loss": 0.398,
      "step": 19490,
      "task_loss": 0.6458536386489868
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3486079275608063,
      "epoch": 7.05,
      "learning_rate": 3.847663337932806e-05,
      "loss": 0.3815,
      "step": 19500,
      "task_loss": 0.5868500471115112
    },
    {
      "epoch": 7.05,
      "eval_exact_match": 82.79091769157995,
      "eval_f1": 89.71267092502406,
      "step": 19500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34197574853897095,
      "epoch": 7.05,
      "learning_rate": 3.841126477855469e-05,
      "loss": 0.4191,
      "step": 19510,
      "task_loss": 0.7607687711715698
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39318233728408813,
      "epoch": 7.05,
      "learning_rate": 3.834585280647169e-05,
      "loss": 0.368,
      "step": 19520,
      "task_loss": 0.4875791072845459
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28019875288009644,
      "epoch": 7.06,
      "learning_rate": 3.82803978003652e-05,
      "loss": 0.4371,
      "step": 19530,
      "task_loss": 0.4103456437587738
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40750378370285034,
      "epoch": 7.06,
      "learning_rate": 3.821490009774321e-05,
      "loss": 0.38,
      "step": 19540,
      "task_loss": 0.7055575251579285
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.349598228931427,
      "epoch": 7.07,
      "learning_rate": 3.814936003633393e-05,
      "loss": 0.4285,
      "step": 19550,
      "task_loss": 0.4073978066444397
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46198201179504395,
      "epoch": 7.07,
      "learning_rate": 3.808377795408394e-05,
      "loss": 0.3817,
      "step": 19560,
      "task_loss": 0.9234828948974609
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40304064750671387,
      "epoch": 7.07,
      "learning_rate": 3.80181541891565e-05,
      "loss": 0.3844,
      "step": 19570,
      "task_loss": 0.6997017860412598
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4080723822116852,
      "epoch": 7.08,
      "learning_rate": 3.7952489079929826e-05,
      "loss": 0.3802,
      "step": 19580,
      "task_loss": 0.5474996566772461
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30232366919517517,
      "epoch": 7.08,
      "learning_rate": 3.7886782964995304e-05,
      "loss": 0.3964,
      "step": 19590,
      "task_loss": 0.6640204191207886
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4523887634277344,
      "epoch": 7.08,
      "learning_rate": 3.782103618315575e-05,
      "loss": 0.4024,
      "step": 19600,
      "task_loss": 0.5164095163345337
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4048759639263153,
      "epoch": 7.09,
      "learning_rate": 3.775524907342367e-05,
      "loss": 0.4032,
      "step": 19610,
      "task_loss": 0.7620391845703125
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3249843120574951,
      "epoch": 7.09,
      "learning_rate": 3.768942197501955e-05,
      "loss": 0.3736,
      "step": 19620,
      "task_loss": 0.4513287842273712
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3706258237361908,
      "epoch": 7.09,
      "learning_rate": 3.7623555227370017e-05,
      "loss": 0.3994,
      "step": 19630,
      "task_loss": 0.6358516216278076
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41293206810951233,
      "epoch": 7.1,
      "learning_rate": 3.755764917010618e-05,
      "loss": 0.3806,
      "step": 19640,
      "task_loss": 0.4972880482673645
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4146842360496521,
      "epoch": 7.1,
      "learning_rate": 3.749170414306184e-05,
      "loss": 0.4166,
      "step": 19650,
      "task_loss": 0.4463941156864166
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30623093247413635,
      "epoch": 7.11,
      "learning_rate": 3.7425720486271726e-05,
      "loss": 0.3611,
      "step": 19660,
      "task_loss": 0.24231675267219543
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5473644733428955,
      "epoch": 7.11,
      "learning_rate": 3.735969853996976e-05,
      "loss": 0.4418,
      "step": 19670,
      "task_loss": 0.8937996625900269
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3078344464302063,
      "epoch": 7.11,
      "learning_rate": 3.72936386445873e-05,
      "loss": 0.3904,
      "step": 19680,
      "task_loss": 0.4181814193725586
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3279660940170288,
      "epoch": 7.12,
      "learning_rate": 3.722754114075137e-05,
      "loss": 0.393,
      "step": 19690,
      "task_loss": 0.41047608852386475
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30717504024505615,
      "epoch": 7.12,
      "learning_rate": 3.716140636928295e-05,
      "loss": 0.3757,
      "step": 19700,
      "task_loss": 0.40464282035827637
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37334302067756653,
      "epoch": 7.12,
      "learning_rate": 3.709523467119514e-05,
      "loss": 0.3663,
      "step": 19710,
      "task_loss": 0.7159653902053833
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29783928394317627,
      "epoch": 7.13,
      "learning_rate": 3.7029026387691464e-05,
      "loss": 0.3925,
      "step": 19720,
      "task_loss": 0.36046409606933594
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33988526463508606,
      "epoch": 7.13,
      "learning_rate": 3.696278186016411e-05,
      "loss": 0.3777,
      "step": 19730,
      "task_loss": 0.3282432556152344
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.520359218120575,
      "epoch": 7.13,
      "learning_rate": 3.6896501430192134e-05,
      "loss": 0.3646,
      "step": 19740,
      "task_loss": 0.7145998477935791
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3622116446495056,
      "epoch": 7.14,
      "learning_rate": 3.6830185439539726e-05,
      "loss": 0.4161,
      "step": 19750,
      "task_loss": 0.8804785013198853
    },
    {
      "epoch": 7.14,
      "eval_exact_match": 82.51655629139073,
      "eval_f1": 89.55102082659974,
      "step": 19750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35625964403152466,
      "epoch": 7.14,
      "learning_rate": 3.676383423015442e-05,
      "loss": 0.3932,
      "step": 19760,
      "task_loss": 0.3122161626815796
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3781408965587616,
      "epoch": 7.14,
      "learning_rate": 3.6697448144165357e-05,
      "loss": 0.3812,
      "step": 19770,
      "task_loss": 0.32182449102401733
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5030004978179932,
      "epoch": 7.15,
      "learning_rate": 3.663102752388155e-05,
      "loss": 0.3671,
      "step": 19780,
      "task_loss": 0.6021063327789307
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3663809299468994,
      "epoch": 7.15,
      "learning_rate": 3.656457271179003e-05,
      "loss": 0.4523,
      "step": 19790,
      "task_loss": 0.5944013595581055
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3215387165546417,
      "epoch": 7.16,
      "learning_rate": 3.6498084050554164e-05,
      "loss": 0.403,
      "step": 19800,
      "task_loss": 0.6697392463684082
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.487884521484375,
      "epoch": 7.16,
      "learning_rate": 3.643156188301183e-05,
      "loss": 0.3921,
      "step": 19810,
      "task_loss": 0.4826103448867798
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31640633940696716,
      "epoch": 7.16,
      "learning_rate": 3.636500655217371e-05,
      "loss": 0.4237,
      "step": 19820,
      "task_loss": 0.24287016689777374
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4092869162559509,
      "epoch": 7.17,
      "learning_rate": 3.629841840122147e-05,
      "loss": 0.3917,
      "step": 19830,
      "task_loss": 0.7395460605621338
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3161425292491913,
      "epoch": 7.17,
      "learning_rate": 3.6231797773505994e-05,
      "loss": 0.3842,
      "step": 19840,
      "task_loss": 0.4677432179450989
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45856085419654846,
      "epoch": 7.17,
      "learning_rate": 3.616514501254567e-05,
      "loss": 0.3748,
      "step": 19850,
      "task_loss": 0.6139996647834778
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4803004860877991,
      "epoch": 7.18,
      "learning_rate": 3.6098460462024506e-05,
      "loss": 0.4153,
      "step": 19860,
      "task_loss": 0.6763205528259277
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4521372318267822,
      "epoch": 7.18,
      "learning_rate": 3.6031744465790495e-05,
      "loss": 0.4046,
      "step": 19870,
      "task_loss": 0.5324259996414185
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34895187616348267,
      "epoch": 7.18,
      "learning_rate": 3.5964997367853755e-05,
      "loss": 0.3662,
      "step": 19880,
      "task_loss": 0.3761727213859558
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3079893887042999,
      "epoch": 7.19,
      "learning_rate": 3.589821951238474e-05,
      "loss": 0.3581,
      "step": 19890,
      "task_loss": 0.48900896310806274
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.375984251499176,
      "epoch": 7.19,
      "learning_rate": 3.5831411243712555e-05,
      "loss": 0.4284,
      "step": 19900,
      "task_loss": 0.38719305396080017
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4510999023914337,
      "epoch": 7.2,
      "learning_rate": 3.5764572906323075e-05,
      "loss": 0.4194,
      "step": 19910,
      "task_loss": 0.7633450031280518
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4013802409172058,
      "epoch": 7.2,
      "learning_rate": 3.569770484485726e-05,
      "loss": 0.4213,
      "step": 19920,
      "task_loss": 0.39382585883140564
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4381757974624634,
      "epoch": 7.2,
      "learning_rate": 3.563080740410932e-05,
      "loss": 0.405,
      "step": 19930,
      "task_loss": 0.4762646555900574
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36635851860046387,
      "epoch": 7.21,
      "learning_rate": 3.556388092902494e-05,
      "loss": 0.4212,
      "step": 19940,
      "task_loss": 0.4292539358139038
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4319227933883667,
      "epoch": 7.21,
      "learning_rate": 3.549692576469955e-05,
      "loss": 0.4012,
      "step": 19950,
      "task_loss": 0.705822229385376
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2866457402706146,
      "epoch": 7.21,
      "learning_rate": 3.542994225637648e-05,
      "loss": 0.422,
      "step": 19960,
      "task_loss": 0.3992174565792084
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43599292635917664,
      "epoch": 7.22,
      "learning_rate": 3.536293074944522e-05,
      "loss": 0.369,
      "step": 19970,
      "task_loss": 0.9676334857940674
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.48719683289527893,
      "epoch": 7.22,
      "learning_rate": 3.529589158943965e-05,
      "loss": 0.386,
      "step": 19980,
      "task_loss": 0.6881073713302612
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3196744918823242,
      "epoch": 7.22,
      "learning_rate": 3.522882512203621e-05,
      "loss": 0.4417,
      "step": 19990,
      "task_loss": 0.5994395017623901
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3680725693702698,
      "epoch": 7.23,
      "learning_rate": 3.516173169305216e-05,
      "loss": 0.3802,
      "step": 20000,
      "task_loss": 0.521813154220581
    },
    {
      "epoch": 7.23,
      "eval_exact_match": 82.36518448438979,
      "eval_f1": 89.49724844867848,
      "step": 20000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39387252926826477,
      "epoch": 7.23,
      "learning_rate": 3.5094611648443773e-05,
      "loss": 0.4165,
      "step": 20010,
      "task_loss": 0.5132284164428711
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3651062548160553,
      "epoch": 7.24,
      "learning_rate": 3.5034181137980914e-05,
      "loss": 0.3956,
      "step": 20020,
      "task_loss": 0.5812886357307434
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2903411388397217,
      "epoch": 7.24,
      "learning_rate": 3.496701147728595e-05,
      "loss": 0.3921,
      "step": 20030,
      "task_loss": 0.5770473480224609
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.321413516998291,
      "epoch": 7.24,
      "learning_rate": 3.489981620500956e-05,
      "loss": 0.4927,
      "step": 20040,
      "task_loss": 0.3058428168296814
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36443406343460083,
      "epoch": 7.25,
      "learning_rate": 3.483259566763317e-05,
      "loss": 0.3565,
      "step": 20050,
      "task_loss": 0.5052045583724976
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3698991537094116,
      "epoch": 7.25,
      "learning_rate": 3.476535021176848e-05,
      "loss": 0.3733,
      "step": 20060,
      "task_loss": 0.6003679037094116
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34568488597869873,
      "epoch": 7.25,
      "learning_rate": 3.4698080184155674e-05,
      "loss": 0.3943,
      "step": 20070,
      "task_loss": 0.5871903896331787
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.373020738363266,
      "epoch": 7.26,
      "learning_rate": 3.463078593166162e-05,
      "loss": 0.394,
      "step": 20080,
      "task_loss": 0.3602370023727417
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33881717920303345,
      "epoch": 7.26,
      "learning_rate": 3.456346780127812e-05,
      "loss": 0.4087,
      "step": 20090,
      "task_loss": 0.5115830898284912
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5135064125061035,
      "epoch": 7.26,
      "learning_rate": 3.44961261401201e-05,
      "loss": 0.3805,
      "step": 20100,
      "task_loss": 0.4804335832595825
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43249940872192383,
      "epoch": 7.27,
      "learning_rate": 3.4428761295423806e-05,
      "loss": 0.3812,
      "step": 20110,
      "task_loss": 0.7190282940864563
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3258872926235199,
      "epoch": 7.27,
      "learning_rate": 3.436137361454502e-05,
      "loss": 0.4134,
      "step": 20120,
      "task_loss": 0.613054633140564
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3143617510795593,
      "epoch": 7.28,
      "learning_rate": 3.42939634449573e-05,
      "loss": 0.3406,
      "step": 20130,
      "task_loss": 0.3883563280105591
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34468793869018555,
      "epoch": 7.28,
      "learning_rate": 3.422653113425013e-05,
      "loss": 0.3866,
      "step": 20140,
      "task_loss": 0.5091181993484497
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4208349585533142,
      "epoch": 7.28,
      "learning_rate": 3.415907703012719e-05,
      "loss": 0.3896,
      "step": 20150,
      "task_loss": 0.5660240650177002
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3679567873477936,
      "epoch": 7.29,
      "learning_rate": 3.4091601480404535e-05,
      "loss": 0.3947,
      "step": 20160,
      "task_loss": 0.6140090227127075
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43558749556541443,
      "epoch": 7.29,
      "learning_rate": 3.402410483300877e-05,
      "loss": 0.4325,
      "step": 20170,
      "task_loss": 0.9385803937911987
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4312199354171753,
      "epoch": 7.29,
      "learning_rate": 3.395658743597531e-05,
      "loss": 0.3422,
      "step": 20180,
      "task_loss": 0.6504771709442139
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.333647757768631,
      "epoch": 7.3,
      "learning_rate": 3.388904963744656e-05,
      "loss": 0.4073,
      "step": 20190,
      "task_loss": 0.4446883201599121
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3883225917816162,
      "epoch": 7.3,
      "learning_rate": 3.382149178567012e-05,
      "loss": 0.403,
      "step": 20200,
      "task_loss": 0.622755229473114
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32313084602355957,
      "epoch": 7.3,
      "learning_rate": 3.3753914228997e-05,
      "loss": 0.4027,
      "step": 20210,
      "task_loss": 0.6717528104782104
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3080373704433441,
      "epoch": 7.31,
      "learning_rate": 3.3686317315879785e-05,
      "loss": 0.3907,
      "step": 20220,
      "task_loss": 0.4064512848854065
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32508981227874756,
      "epoch": 7.31,
      "learning_rate": 3.361870139487092e-05,
      "loss": 0.3981,
      "step": 20230,
      "task_loss": 0.2644849419593811
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2956303358078003,
      "epoch": 7.31,
      "learning_rate": 3.3551066814620796e-05,
      "loss": 0.3662,
      "step": 20240,
      "task_loss": 0.3841955065727234
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3077021837234497,
      "epoch": 7.32,
      "learning_rate": 3.3483413923876075e-05,
      "loss": 0.3706,
      "step": 20250,
      "task_loss": 0.4486422538757324
    },
    {
      "epoch": 7.32,
      "eval_exact_match": 82.73415326395458,
      "eval_f1": 89.86822273734353,
      "step": 20250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35297924280166626,
      "epoch": 7.32,
      "learning_rate": 3.3415743071477814e-05,
      "loss": 0.3836,
      "step": 20260,
      "task_loss": 0.4488530457019806
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4628410041332245,
      "epoch": 7.33,
      "learning_rate": 3.334805460635966e-05,
      "loss": 0.4589,
      "step": 20270,
      "task_loss": 0.9875482320785522
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.416703999042511,
      "epoch": 7.33,
      "learning_rate": 3.328034887754613e-05,
      "loss": 0.3884,
      "step": 20280,
      "task_loss": 0.3616025447845459
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3806164264678955,
      "epoch": 7.33,
      "learning_rate": 3.321262623415071e-05,
      "loss": 0.4103,
      "step": 20290,
      "task_loss": 0.4707099199295044
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4472636580467224,
      "epoch": 7.34,
      "learning_rate": 3.3144887025374125e-05,
      "loss": 0.4004,
      "step": 20300,
      "task_loss": 0.48857200145721436
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3303232789039612,
      "epoch": 7.34,
      "learning_rate": 3.307713160050252e-05,
      "loss": 0.3606,
      "step": 20310,
      "task_loss": 0.4707809388637543
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4367483854293823,
      "epoch": 7.34,
      "learning_rate": 3.300936030890564e-05,
      "loss": 0.4182,
      "step": 20320,
      "task_loss": 0.7906453609466553
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28419390320777893,
      "epoch": 7.35,
      "learning_rate": 3.294157350003507e-05,
      "loss": 0.3636,
      "step": 20330,
      "task_loss": 0.2948117256164551
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.339834988117218,
      "epoch": 7.35,
      "learning_rate": 3.287377152342236e-05,
      "loss": 0.3916,
      "step": 20340,
      "task_loss": 0.34666329622268677
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2876173257827759,
      "epoch": 7.35,
      "learning_rate": 3.280595472867733e-05,
      "loss": 0.4121,
      "step": 20350,
      "task_loss": 0.3110664188861847
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.317335307598114,
      "epoch": 7.36,
      "learning_rate": 3.273812346548617e-05,
      "loss": 0.3872,
      "step": 20360,
      "task_loss": 0.4784393012523651
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4668826460838318,
      "epoch": 7.36,
      "learning_rate": 3.2670278083609685e-05,
      "loss": 0.3571,
      "step": 20370,
      "task_loss": 0.6749038100242615
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3103598654270172,
      "epoch": 7.37,
      "learning_rate": 3.260241893288147e-05,
      "loss": 0.3929,
      "step": 20380,
      "task_loss": 0.7942910194396973
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.332332581281662,
      "epoch": 7.37,
      "learning_rate": 3.253454636320613e-05,
      "loss": 0.3819,
      "step": 20390,
      "task_loss": 0.4826880097389221
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40551626682281494,
      "epoch": 7.37,
      "learning_rate": 3.246666072455746e-05,
      "loss": 0.3616,
      "step": 20400,
      "task_loss": 0.43429872393608093
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3160107135772705,
      "epoch": 7.38,
      "learning_rate": 3.2398762366976647e-05,
      "loss": 0.3852,
      "step": 20410,
      "task_loss": 0.43068432807922363
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4586879014968872,
      "epoch": 7.38,
      "learning_rate": 3.2330851640570427e-05,
      "loss": 0.4322,
      "step": 20420,
      "task_loss": 0.702597975730896
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33801770210266113,
      "epoch": 7.38,
      "learning_rate": 3.226292889550938e-05,
      "loss": 0.4115,
      "step": 20430,
      "task_loss": 1.104683756828308
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4458228051662445,
      "epoch": 7.39,
      "learning_rate": 3.219499448202598e-05,
      "loss": 0.3881,
      "step": 20440,
      "task_loss": 0.5195741057395935
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3442004323005676,
      "epoch": 7.39,
      "learning_rate": 3.212704875041293e-05,
      "loss": 0.3945,
      "step": 20450,
      "task_loss": 0.6913373470306396
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.420304536819458,
      "epoch": 7.39,
      "learning_rate": 3.205909205102128e-05,
      "loss": 0.4567,
      "step": 20460,
      "task_loss": 0.4324960708618164
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43511003255844116,
      "epoch": 7.4,
      "learning_rate": 3.19911247342586e-05,
      "loss": 0.3947,
      "step": 20470,
      "task_loss": 0.8736984729766846
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4031512141227722,
      "epoch": 7.4,
      "learning_rate": 3.192314715058724e-05,
      "loss": 0.4193,
      "step": 20480,
      "task_loss": 0.5991595983505249
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5769762992858887,
      "epoch": 7.41,
      "learning_rate": 3.185515965052248e-05,
      "loss": 0.4721,
      "step": 20490,
      "task_loss": 0.5272440314292908
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32588744163513184,
      "epoch": 7.41,
      "learning_rate": 3.1787162584630735e-05,
      "loss": 0.3777,
      "step": 20500,
      "task_loss": 0.5834881067276001
    },
    {
      "epoch": 7.41,
      "eval_exact_match": 82.74361400189214,
      "eval_f1": 89.73354102474433,
      "step": 20500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34094029664993286,
      "epoch": 7.41,
      "learning_rate": 3.1719156303527744e-05,
      "loss": 0.3924,
      "step": 20510,
      "task_loss": 0.6706292629241943
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47249582409858704,
      "epoch": 7.42,
      "learning_rate": 3.1651141157876755e-05,
      "loss": 0.4,
      "step": 20520,
      "task_loss": 0.5581311583518982
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3450727164745331,
      "epoch": 7.42,
      "learning_rate": 3.1583117498386725e-05,
      "loss": 0.4057,
      "step": 20530,
      "task_loss": 0.6178710460662842
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3991689383983612,
      "epoch": 7.42,
      "learning_rate": 3.1515085675810534e-05,
      "loss": 0.3887,
      "step": 20540,
      "task_loss": 0.5536105632781982
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36258408427238464,
      "epoch": 7.43,
      "learning_rate": 3.144704604094312e-05,
      "loss": 0.3549,
      "step": 20550,
      "task_loss": 0.48555445671081543
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4089629650115967,
      "epoch": 7.43,
      "learning_rate": 3.137899894461973e-05,
      "loss": 0.3655,
      "step": 20560,
      "task_loss": 0.2968985438346863
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45736321806907654,
      "epoch": 7.43,
      "learning_rate": 3.131094473771406e-05,
      "loss": 0.3927,
      "step": 20570,
      "task_loss": 0.938363790512085
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35800325870513916,
      "epoch": 7.44,
      "learning_rate": 3.12428837711365e-05,
      "loss": 0.4001,
      "step": 20580,
      "task_loss": 0.7353466749191284
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4800975024700165,
      "epoch": 7.44,
      "learning_rate": 3.117481639583228e-05,
      "loss": 0.4031,
      "step": 20590,
      "task_loss": 0.35843586921691895
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2761758267879486,
      "epoch": 7.44,
      "learning_rate": 3.110674296277967e-05,
      "loss": 0.3579,
      "step": 20600,
      "task_loss": 0.5030902028083801
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5137525200843811,
      "epoch": 7.45,
      "learning_rate": 3.103866382298818e-05,
      "loss": 0.3511,
      "step": 20610,
      "task_loss": 0.6685386896133423
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45088014006614685,
      "epoch": 7.45,
      "learning_rate": 3.097057932749675e-05,
      "loss": 0.4005,
      "step": 20620,
      "task_loss": 0.695102334022522
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3517519533634186,
      "epoch": 7.46,
      "learning_rate": 3.0902489827371924e-05,
      "loss": 0.3538,
      "step": 20630,
      "task_loss": 0.6277685165405273
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4504373073577881,
      "epoch": 7.46,
      "learning_rate": 3.083439567370609e-05,
      "loss": 0.4344,
      "step": 20640,
      "task_loss": 0.8747158050537109
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5259716510772705,
      "epoch": 7.46,
      "learning_rate": 3.076629721761557e-05,
      "loss": 0.4092,
      "step": 20650,
      "task_loss": 0.684150218963623
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33155426383018494,
      "epoch": 7.47,
      "learning_rate": 3.069819481023892e-05,
      "loss": 0.3818,
      "step": 20660,
      "task_loss": 0.4055613875389099
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3492274284362793,
      "epoch": 7.47,
      "learning_rate": 3.0630088802735046e-05,
      "loss": 0.4271,
      "step": 20670,
      "task_loss": 0.657691478729248
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47741013765335083,
      "epoch": 7.47,
      "learning_rate": 3.056197954628143e-05,
      "loss": 0.4267,
      "step": 20680,
      "task_loss": 0.8212167620658875
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2485727220773697,
      "epoch": 7.48,
      "learning_rate": 3.0493867392072296e-05,
      "loss": 0.3732,
      "step": 20690,
      "task_loss": 0.37689927220344543
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47409436106681824,
      "epoch": 7.48,
      "learning_rate": 3.0425752691316814e-05,
      "loss": 0.4012,
      "step": 20700,
      "task_loss": 0.818795382976532
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3214324116706848,
      "epoch": 7.48,
      "learning_rate": 3.03576357952373e-05,
      "loss": 0.3844,
      "step": 20710,
      "task_loss": 0.9982783198356628
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49584317207336426,
      "epoch": 7.49,
      "learning_rate": 3.028951705506736e-05,
      "loss": 0.4084,
      "step": 20720,
      "task_loss": 0.7639179229736328
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2755817770957947,
      "epoch": 7.49,
      "learning_rate": 3.0221396822050114e-05,
      "loss": 0.3612,
      "step": 20730,
      "task_loss": 0.32425567507743835
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3090718388557434,
      "epoch": 7.5,
      "learning_rate": 3.0153275447436423e-05,
      "loss": 0.3943,
      "step": 20740,
      "task_loss": 0.2633516192436218
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41136348247528076,
      "epoch": 7.5,
      "learning_rate": 3.0085153282482977e-05,
      "loss": 0.368,
      "step": 20750,
      "task_loss": 0.3509225845336914
    },
    {
      "epoch": 7.5,
      "eval_exact_match": 82.76253547776727,
      "eval_f1": 89.78687617165846,
      "step": 20750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2749996483325958,
      "epoch": 7.5,
      "learning_rate": 3.0017030678450578e-05,
      "loss": 0.3833,
      "step": 20760,
      "task_loss": 0.2450357973575592
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3595663905143738,
      "epoch": 7.51,
      "learning_rate": 2.9948907986602262e-05,
      "loss": 0.3604,
      "step": 20770,
      "task_loss": 0.8052037954330444
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3556064963340759,
      "epoch": 7.51,
      "learning_rate": 2.988078555820156e-05,
      "loss": 0.3699,
      "step": 20780,
      "task_loss": 0.8142600059509277
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3247981071472168,
      "epoch": 7.51,
      "learning_rate": 2.98126637445106e-05,
      "loss": 0.4256,
      "step": 20790,
      "task_loss": 0.5774292945861816
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39211082458496094,
      "epoch": 7.52,
      "learning_rate": 2.9744542896788375e-05,
      "loss": 0.3755,
      "step": 20800,
      "task_loss": 0.44028204679489136
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3495498597621918,
      "epoch": 7.52,
      "learning_rate": 2.967642336628887e-05,
      "loss": 0.3667,
      "step": 20810,
      "task_loss": 0.7542967796325684
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5816289186477661,
      "epoch": 7.52,
      "learning_rate": 2.9608305504259303e-05,
      "loss": 0.464,
      "step": 20820,
      "task_loss": 0.7060995101928711
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38165950775146484,
      "epoch": 7.53,
      "learning_rate": 2.9540189661938263e-05,
      "loss": 0.3788,
      "step": 20830,
      "task_loss": 0.652009129524231
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2923566699028015,
      "epoch": 7.53,
      "learning_rate": 2.9472076190553952e-05,
      "loss": 0.3983,
      "step": 20840,
      "task_loss": 0.5197100043296814
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4252726435661316,
      "epoch": 7.54,
      "learning_rate": 2.9403965441322315e-05,
      "loss": 0.3834,
      "step": 20850,
      "task_loss": 0.48940667510032654
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33503347635269165,
      "epoch": 7.54,
      "learning_rate": 2.9335857765445304e-05,
      "loss": 0.3915,
      "step": 20860,
      "task_loss": 1.0881128311157227
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3840358257293701,
      "epoch": 7.54,
      "learning_rate": 2.9267753514108968e-05,
      "loss": 0.3975,
      "step": 20870,
      "task_loss": 0.38832956552505493
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5155704021453857,
      "epoch": 7.55,
      "learning_rate": 2.9199653038481753e-05,
      "loss": 0.3877,
      "step": 20880,
      "task_loss": 0.7354077100753784
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3932916820049286,
      "epoch": 7.55,
      "learning_rate": 2.9131556689712597e-05,
      "loss": 0.3647,
      "step": 20890,
      "task_loss": 0.3581256866455078
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.48106008768081665,
      "epoch": 7.55,
      "learning_rate": 2.9063464818929186e-05,
      "loss": 0.4336,
      "step": 20900,
      "task_loss": 0.7806867361068726
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30515527725219727,
      "epoch": 7.56,
      "learning_rate": 2.8995377777236085e-05,
      "loss": 0.3727,
      "step": 20910,
      "task_loss": 0.5825417637825012
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.347696989774704,
      "epoch": 7.56,
      "learning_rate": 2.8927295915712994e-05,
      "loss": 0.384,
      "step": 20920,
      "task_loss": 0.5347048044204712
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31037092208862305,
      "epoch": 7.56,
      "learning_rate": 2.885921958541287e-05,
      "loss": 0.3555,
      "step": 20930,
      "task_loss": 0.301509827375412
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42140477895736694,
      "epoch": 7.57,
      "learning_rate": 2.8791149137360176e-05,
      "loss": 0.3656,
      "step": 20940,
      "task_loss": 0.45308709144592285
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44975745677948,
      "epoch": 7.57,
      "learning_rate": 2.872308492254901e-05,
      "loss": 0.4172,
      "step": 20950,
      "task_loss": 0.6830534934997559
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35245752334594727,
      "epoch": 7.57,
      "learning_rate": 2.865502729194139e-05,
      "loss": 0.3779,
      "step": 20960,
      "task_loss": 0.3799573481082916
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36846715211868286,
      "epoch": 7.58,
      "learning_rate": 2.8586976596465308e-05,
      "loss": 0.3959,
      "step": 20970,
      "task_loss": 0.5989868640899658
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33940085768699646,
      "epoch": 7.58,
      "learning_rate": 2.8518933187013053e-05,
      "loss": 0.3268,
      "step": 20980,
      "task_loss": 0.5030498504638672
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4226445257663727,
      "epoch": 7.59,
      "learning_rate": 2.8450897414439317e-05,
      "loss": 0.3917,
      "step": 20990,
      "task_loss": 0.596397876739502
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39020735025405884,
      "epoch": 7.59,
      "learning_rate": 2.838286962955943e-05,
      "loss": 0.3701,
      "step": 21000,
      "task_loss": 0.7234203815460205
    },
    {
      "epoch": 7.59,
      "eval_exact_match": 82.58278145695364,
      "eval_f1": 89.55032112084302,
      "step": 21000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5218778848648071,
      "epoch": 7.59,
      "learning_rate": 2.8314850183147516e-05,
      "loss": 0.3957,
      "step": 21010,
      "task_loss": 0.6053802967071533
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.348112016916275,
      "epoch": 7.6,
      "learning_rate": 2.8246839425934724e-05,
      "loss": 0.377,
      "step": 21020,
      "task_loss": 0.6146667003631592
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3650135397911072,
      "epoch": 7.6,
      "learning_rate": 2.817883770860737e-05,
      "loss": 0.3546,
      "step": 21030,
      "task_loss": 0.37088972330093384
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3635476231575012,
      "epoch": 7.6,
      "learning_rate": 2.81108453818052e-05,
      "loss": 0.3664,
      "step": 21040,
      "task_loss": 0.4450761079788208
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4935896098613739,
      "epoch": 7.61,
      "learning_rate": 2.8042862796119482e-05,
      "loss": 0.4058,
      "step": 21050,
      "task_loss": 0.5729918479919434
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3340076804161072,
      "epoch": 7.61,
      "learning_rate": 2.7974890302091327e-05,
      "loss": 0.4005,
      "step": 21060,
      "task_loss": 0.7589318752288818
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27468156814575195,
      "epoch": 7.61,
      "learning_rate": 2.7906928250209743e-05,
      "loss": 0.4049,
      "step": 21070,
      "task_loss": 0.6373112797737122
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6113014817237854,
      "epoch": 7.62,
      "learning_rate": 2.783897699090994e-05,
      "loss": 0.4058,
      "step": 21080,
      "task_loss": 1.0075609683990479
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3902479410171509,
      "epoch": 7.62,
      "learning_rate": 2.7771036874571443e-05,
      "loss": 0.4069,
      "step": 21090,
      "task_loss": 0.6849322319030762
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31239408254623413,
      "epoch": 7.63,
      "learning_rate": 2.770310825151635e-05,
      "loss": 0.3778,
      "step": 21100,
      "task_loss": 0.46959567070007324
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33670780062675476,
      "epoch": 7.63,
      "learning_rate": 2.763519147200748e-05,
      "loss": 0.3693,
      "step": 21110,
      "task_loss": 0.45546576380729675
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4706672430038452,
      "epoch": 7.63,
      "learning_rate": 2.7567286886246593e-05,
      "loss": 0.3903,
      "step": 21120,
      "task_loss": 0.5884417295455933
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3736233115196228,
      "epoch": 7.64,
      "learning_rate": 2.749939484437255e-05,
      "loss": 0.4197,
      "step": 21130,
      "task_loss": 0.3115198016166687
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39010971784591675,
      "epoch": 7.64,
      "learning_rate": 2.7431515696459577e-05,
      "loss": 0.4015,
      "step": 21140,
      "task_loss": 0.5224123001098633
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29932332038879395,
      "epoch": 7.64,
      "learning_rate": 2.736364979251535e-05,
      "loss": 0.3853,
      "step": 21150,
      "task_loss": 0.4086366891860962
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42739713191986084,
      "epoch": 7.65,
      "learning_rate": 2.7295797482479327e-05,
      "loss": 0.4023,
      "step": 21160,
      "task_loss": 0.9876140356063843
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3685208857059479,
      "epoch": 7.65,
      "learning_rate": 2.7227959116220803e-05,
      "loss": 0.3945,
      "step": 21170,
      "task_loss": 0.5753424167633057
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47203847765922546,
      "epoch": 7.65,
      "learning_rate": 2.7160135043537236e-05,
      "loss": 0.3561,
      "step": 21180,
      "task_loss": 0.7574573755264282
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3211335837841034,
      "epoch": 7.66,
      "learning_rate": 2.7092325614152328e-05,
      "loss": 0.4059,
      "step": 21190,
      "task_loss": 0.621513843536377
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4134974479675293,
      "epoch": 7.66,
      "learning_rate": 2.7024531177714316e-05,
      "loss": 0.3693,
      "step": 21200,
      "task_loss": 0.39401504397392273
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29084134101867676,
      "epoch": 7.67,
      "learning_rate": 2.6956752083794094e-05,
      "loss": 0.3972,
      "step": 21210,
      "task_loss": 0.49650153517723083
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5855056047439575,
      "epoch": 7.67,
      "learning_rate": 2.688898868188348e-05,
      "loss": 0.425,
      "step": 21220,
      "task_loss": 0.6136555075645447
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4040588140487671,
      "epoch": 7.67,
      "learning_rate": 2.682124132139334e-05,
      "loss": 0.3688,
      "step": 21230,
      "task_loss": 0.5768276453018188
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3762333393096924,
      "epoch": 7.68,
      "learning_rate": 2.675351035165188e-05,
      "loss": 0.41,
      "step": 21240,
      "task_loss": 0.6931823492050171
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5697497725486755,
      "epoch": 7.68,
      "learning_rate": 2.668579612190271e-05,
      "loss": 0.4016,
      "step": 21250,
      "task_loss": 0.7728415131568909
    },
    {
      "epoch": 7.68,
      "eval_exact_match": 83.15042573320719,
      "eval_f1": 90.07272671513242,
      "step": 21250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39408552646636963,
      "epoch": 7.68,
      "learning_rate": 2.6618098981303204e-05,
      "loss": 0.3603,
      "step": 21260,
      "task_loss": 0.6829339265823364
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4046620726585388,
      "epoch": 7.69,
      "learning_rate": 2.655041927892257e-05,
      "loss": 0.3888,
      "step": 21270,
      "task_loss": 0.6104872822761536
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32940006256103516,
      "epoch": 7.69,
      "learning_rate": 2.6482757363740117e-05,
      "loss": 0.3446,
      "step": 21280,
      "task_loss": 0.4668557047843933
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4409995675086975,
      "epoch": 7.69,
      "learning_rate": 2.6415113584643424e-05,
      "loss": 0.4242,
      "step": 21290,
      "task_loss": 0.6875514984130859
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2878825068473816,
      "epoch": 7.7,
      "learning_rate": 2.6347488290426573e-05,
      "loss": 0.3609,
      "step": 21300,
      "task_loss": 0.7109676599502563
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3507508635520935,
      "epoch": 7.7,
      "learning_rate": 2.627988182978831e-05,
      "loss": 0.4358,
      "step": 21310,
      "task_loss": 0.6243581771850586
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3260875344276428,
      "epoch": 7.71,
      "learning_rate": 2.6212294551330293e-05,
      "loss": 0.4248,
      "step": 21320,
      "task_loss": 0.5162962675094604
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39942505955696106,
      "epoch": 7.71,
      "learning_rate": 2.6144726803555232e-05,
      "loss": 0.4051,
      "step": 21330,
      "task_loss": 0.8916963338851929
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4103178083896637,
      "epoch": 7.71,
      "learning_rate": 2.6077178934865193e-05,
      "loss": 0.3919,
      "step": 21340,
      "task_loss": 0.9658797979354858
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2807236313819885,
      "epoch": 7.72,
      "learning_rate": 2.6009651293559663e-05,
      "loss": 0.3637,
      "step": 21350,
      "task_loss": 0.9017109870910645
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23867225646972656,
      "epoch": 7.72,
      "learning_rate": 2.59421442278339e-05,
      "loss": 0.3577,
      "step": 21360,
      "task_loss": 0.29913216829299927
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2474673092365265,
      "epoch": 7.72,
      "learning_rate": 2.5874658085777014e-05,
      "loss": 0.3767,
      "step": 21370,
      "task_loss": 0.40880587697029114
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.441650390625,
      "epoch": 7.73,
      "learning_rate": 2.580719321537026e-05,
      "loss": 0.3848,
      "step": 21380,
      "task_loss": 1.2267597913742065
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4035109877586365,
      "epoch": 7.73,
      "learning_rate": 2.5739749964485183e-05,
      "loss": 0.401,
      "step": 21390,
      "task_loss": 0.6787598133087158
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3127707839012146,
      "epoch": 7.73,
      "learning_rate": 2.5672328680881876e-05,
      "loss": 0.3999,
      "step": 21400,
      "task_loss": 0.43720167875289917
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.7328001856803894,
      "epoch": 7.74,
      "learning_rate": 2.5604929712207137e-05,
      "loss": 0.4533,
      "step": 21410,
      "task_loss": 0.9978082776069641
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3034532368183136,
      "epoch": 7.74,
      "learning_rate": 2.5537553405992723e-05,
      "loss": 0.3571,
      "step": 21420,
      "task_loss": 0.33651202917099
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4111151695251465,
      "epoch": 7.74,
      "learning_rate": 2.547020010965351e-05,
      "loss": 0.3843,
      "step": 21430,
      "task_loss": 0.40663081407546997
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46831807494163513,
      "epoch": 7.75,
      "learning_rate": 2.5402870170485775e-05,
      "loss": 0.3962,
      "step": 21440,
      "task_loss": 0.5977736711502075
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30742383003234863,
      "epoch": 7.75,
      "learning_rate": 2.533556393566528e-05,
      "loss": 0.3719,
      "step": 21450,
      "task_loss": 0.30203139781951904
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3444630801677704,
      "epoch": 7.76,
      "learning_rate": 2.5268281752245642e-05,
      "loss": 0.3626,
      "step": 21460,
      "task_loss": 0.5389599800109863
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33669689297676086,
      "epoch": 7.76,
      "learning_rate": 2.520102396715641e-05,
      "loss": 0.3656,
      "step": 21470,
      "task_loss": 0.5374178290367126
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38278454542160034,
      "epoch": 7.76,
      "learning_rate": 2.513379092720134e-05,
      "loss": 0.3485,
      "step": 21480,
      "task_loss": 0.6707999110221863
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3585819900035858,
      "epoch": 7.77,
      "learning_rate": 2.5066582979056587e-05,
      "loss": 0.361,
      "step": 21490,
      "task_loss": 0.27591443061828613
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24885764718055725,
      "epoch": 7.77,
      "learning_rate": 2.4999400469268948e-05,
      "loss": 0.3919,
      "step": 21500,
      "task_loss": 0.283647745847702
    },
    {
      "epoch": 7.77,
      "eval_exact_match": 83.03689687795648,
      "eval_f1": 90.03524090780901,
      "step": 21500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29895299673080444,
      "epoch": 7.77,
      "learning_rate": 2.493224374425402e-05,
      "loss": 0.3684,
      "step": 21510,
      "task_loss": 0.42891183495521545
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4570951461791992,
      "epoch": 7.78,
      "learning_rate": 2.486511315029447e-05,
      "loss": 0.3693,
      "step": 21520,
      "task_loss": 0.3669062554836273
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2935447692871094,
      "epoch": 7.78,
      "learning_rate": 2.47980090335382e-05,
      "loss": 0.3801,
      "step": 21530,
      "task_loss": 0.5963490009307861
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34089094400405884,
      "epoch": 7.78,
      "learning_rate": 2.4730931739996625e-05,
      "loss": 0.3634,
      "step": 21540,
      "task_loss": 0.725817084312439
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2687324285507202,
      "epoch": 7.79,
      "learning_rate": 2.4663881615542794e-05,
      "loss": 0.3626,
      "step": 21550,
      "task_loss": 0.6789768934249878
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4002509117126465,
      "epoch": 7.79,
      "learning_rate": 2.4596859005909728e-05,
      "loss": 0.4158,
      "step": 21560,
      "task_loss": 0.43731197714805603
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33176618814468384,
      "epoch": 7.8,
      "learning_rate": 2.4529864256688515e-05,
      "loss": 0.3768,
      "step": 21570,
      "task_loss": 0.32815253734588623
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3556719720363617,
      "epoch": 7.8,
      "learning_rate": 2.4462897713326633e-05,
      "loss": 0.3839,
      "step": 21580,
      "task_loss": 0.42700621485710144
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39052969217300415,
      "epoch": 7.8,
      "learning_rate": 2.4395959721126073e-05,
      "loss": 0.4138,
      "step": 21590,
      "task_loss": 0.3304116129875183
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2801629900932312,
      "epoch": 7.81,
      "learning_rate": 2.432905062524165e-05,
      "loss": 0.3815,
      "step": 21600,
      "task_loss": 0.41076505184173584
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29806920886039734,
      "epoch": 7.81,
      "learning_rate": 2.426217077067916e-05,
      "loss": 0.4044,
      "step": 21610,
      "task_loss": 0.513504147529602
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3114551901817322,
      "epoch": 7.81,
      "learning_rate": 2.419532050229361e-05,
      "loss": 0.375,
      "step": 21620,
      "task_loss": 0.46315598487854004
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2388986349105835,
      "epoch": 7.82,
      "learning_rate": 2.412850016478747e-05,
      "loss": 0.3385,
      "step": 21630,
      "task_loss": 0.29883062839508057
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2899854779243469,
      "epoch": 7.82,
      "learning_rate": 2.4061710102708885e-05,
      "loss": 0.3532,
      "step": 21640,
      "task_loss": 0.5468807816505432
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5108985304832458,
      "epoch": 7.82,
      "learning_rate": 2.3994950660449844e-05,
      "loss": 0.394,
      "step": 21650,
      "task_loss": 0.38056838512420654
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33521783351898193,
      "epoch": 7.83,
      "learning_rate": 2.3928222182244508e-05,
      "loss": 0.3658,
      "step": 21660,
      "task_loss": 0.3311670422554016
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4127667546272278,
      "epoch": 7.83,
      "learning_rate": 2.3861525012167334e-05,
      "loss": 0.4362,
      "step": 21670,
      "task_loss": 0.7293896079063416
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45115426182746887,
      "epoch": 7.84,
      "learning_rate": 2.379485949413137e-05,
      "loss": 0.4158,
      "step": 21680,
      "task_loss": 0.564669668674469
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38690847158432007,
      "epoch": 7.84,
      "learning_rate": 2.3728225971886433e-05,
      "loss": 0.3499,
      "step": 21690,
      "task_loss": 0.669680118560791
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4596896767616272,
      "epoch": 7.84,
      "learning_rate": 2.366162478901738e-05,
      "loss": 0.3607,
      "step": 21700,
      "task_loss": 0.5676548480987549
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36195582151412964,
      "epoch": 7.85,
      "learning_rate": 2.359505628894229e-05,
      "loss": 0.3731,
      "step": 21710,
      "task_loss": 0.8088686466217041
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4107312858104706,
      "epoch": 7.85,
      "learning_rate": 2.3528520814910756e-05,
      "loss": 0.3817,
      "step": 21720,
      "task_loss": 0.7661799788475037
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44738367199897766,
      "epoch": 7.85,
      "learning_rate": 2.346201871000203e-05,
      "loss": 0.4273,
      "step": 21730,
      "task_loss": 0.3972681760787964
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38369739055633545,
      "epoch": 7.86,
      "learning_rate": 2.3395550317123355e-05,
      "loss": 0.3842,
      "step": 21740,
      "task_loss": 0.83580482006073
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26634058356285095,
      "epoch": 7.86,
      "learning_rate": 2.3329115979008098e-05,
      "loss": 0.3654,
      "step": 21750,
      "task_loss": 0.47266823053359985
    },
    {
      "epoch": 7.86,
      "eval_exact_match": 82.86660359508042,
      "eval_f1": 89.9055183315166,
      "step": 21750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3641436696052551,
      "epoch": 7.86,
      "learning_rate": 2.3262716038214055e-05,
      "loss": 0.3793,
      "step": 21760,
      "task_loss": 0.44877809286117554
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5512512922286987,
      "epoch": 7.87,
      "learning_rate": 2.319635083712164e-05,
      "loss": 0.397,
      "step": 21770,
      "task_loss": 1.344954013824463
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34335392713546753,
      "epoch": 7.87,
      "learning_rate": 2.3130020717932155e-05,
      "loss": 0.4084,
      "step": 21780,
      "task_loss": 0.48081788420677185
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.352206826210022,
      "epoch": 7.87,
      "learning_rate": 2.306372602266599e-05,
      "loss": 0.3501,
      "step": 21790,
      "task_loss": 0.6583988666534424
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46836334466934204,
      "epoch": 7.88,
      "learning_rate": 2.29974670931609e-05,
      "loss": 0.4028,
      "step": 21800,
      "task_loss": 0.6334257125854492
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36540892720222473,
      "epoch": 7.88,
      "learning_rate": 2.2931244271070183e-05,
      "loss": 0.3821,
      "step": 21810,
      "task_loss": 0.39173829555511475
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46499329805374146,
      "epoch": 7.89,
      "learning_rate": 2.2865057897861014e-05,
      "loss": 0.3748,
      "step": 21820,
      "task_loss": 0.4624268412590027
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5395506024360657,
      "epoch": 7.89,
      "learning_rate": 2.2798908314812546e-05,
      "loss": 0.3834,
      "step": 21830,
      "task_loss": 0.9067790508270264
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33177289366722107,
      "epoch": 7.89,
      "learning_rate": 2.273279586301431e-05,
      "loss": 0.348,
      "step": 21840,
      "task_loss": 0.4920841157436371
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3163832426071167,
      "epoch": 7.9,
      "learning_rate": 2.2666720883364317e-05,
      "loss": 0.3463,
      "step": 21850,
      "task_loss": 0.6887298822402954
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39988842606544495,
      "epoch": 7.9,
      "learning_rate": 2.26006837165674e-05,
      "loss": 0.3928,
      "step": 21860,
      "task_loss": 0.6893279552459717
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35277271270751953,
      "epoch": 7.9,
      "learning_rate": 2.2534684703133375e-05,
      "loss": 0.3875,
      "step": 21870,
      "task_loss": 0.3911385238170624
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2951965630054474,
      "epoch": 7.91,
      "learning_rate": 2.2468724183375365e-05,
      "loss": 0.3776,
      "step": 21880,
      "task_loss": 0.3761315941810608
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3309294879436493,
      "epoch": 7.91,
      "learning_rate": 2.240280249740798e-05,
      "loss": 0.376,
      "step": 21890,
      "task_loss": 0.522919774055481
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5310372114181519,
      "epoch": 7.91,
      "learning_rate": 2.233691998514561e-05,
      "loss": 0.4073,
      "step": 21900,
      "task_loss": 1.1187676191329956
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33567559719085693,
      "epoch": 7.92,
      "learning_rate": 2.2271076986300626e-05,
      "loss": 0.4021,
      "step": 21910,
      "task_loss": 0.7209323644638062
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46349138021469116,
      "epoch": 7.92,
      "learning_rate": 2.2205273840381697e-05,
      "loss": 0.3781,
      "step": 21920,
      "task_loss": 0.9030737280845642
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5550073981285095,
      "epoch": 7.93,
      "learning_rate": 2.213951088669194e-05,
      "loss": 0.407,
      "step": 21930,
      "task_loss": 0.8301783800125122
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4752506613731384,
      "epoch": 7.93,
      "learning_rate": 2.2073788464327286e-05,
      "loss": 0.3842,
      "step": 21940,
      "task_loss": 0.5708151459693909
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30064135789871216,
      "epoch": 7.93,
      "learning_rate": 2.2008106912174626e-05,
      "loss": 0.3408,
      "step": 21950,
      "task_loss": 0.5206139087677002
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4654524326324463,
      "epoch": 7.94,
      "learning_rate": 2.1942466568910142e-05,
      "loss": 0.3693,
      "step": 21960,
      "task_loss": 0.6942557096481323
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47742265462875366,
      "epoch": 7.94,
      "learning_rate": 2.18768677729975e-05,
      "loss": 0.3627,
      "step": 21970,
      "task_loss": 0.5125290155410767
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45493432879447937,
      "epoch": 7.94,
      "learning_rate": 2.1811310862686177e-05,
      "loss": 0.4222,
      "step": 21980,
      "task_loss": 0.7441220879554749
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2649460434913635,
      "epoch": 7.95,
      "learning_rate": 2.1745796176009618e-05,
      "loss": 0.3469,
      "step": 21990,
      "task_loss": 0.3829834759235382
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41353529691696167,
      "epoch": 7.95,
      "learning_rate": 2.1680324050783598e-05,
      "loss": 0.3821,
      "step": 22000,
      "task_loss": 0.5880372524261475
    },
    {
      "epoch": 7.95,
      "eval_exact_match": 83.01797540208136,
      "eval_f1": 89.93481086824103,
      "step": 22000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33622804284095764,
      "epoch": 7.95,
      "learning_rate": 2.161489482460439e-05,
      "loss": 0.3528,
      "step": 22010,
      "task_loss": 0.4933108687400818
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46350884437561035,
      "epoch": 7.96,
      "learning_rate": 2.1549508834847124e-05,
      "loss": 0.386,
      "step": 22020,
      "task_loss": 0.8076874613761902
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27121564745903015,
      "epoch": 7.96,
      "learning_rate": 2.1484166418663904e-05,
      "loss": 0.3791,
      "step": 22030,
      "task_loss": 0.5695164799690247
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31125015020370483,
      "epoch": 7.97,
      "learning_rate": 2.1418867912982233e-05,
      "loss": 0.3831,
      "step": 22040,
      "task_loss": 0.41965681314468384
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41130608320236206,
      "epoch": 7.97,
      "learning_rate": 2.135361365450315e-05,
      "loss": 0.3684,
      "step": 22050,
      "task_loss": 0.39245760440826416
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3798828721046448,
      "epoch": 7.97,
      "learning_rate": 2.1288403979699563e-05,
      "loss": 0.3601,
      "step": 22060,
      "task_loss": 0.4798113703727722
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32262685894966125,
      "epoch": 7.98,
      "learning_rate": 2.122323922481447e-05,
      "loss": 0.3987,
      "step": 22070,
      "task_loss": 0.7292386293411255
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24890893697738647,
      "epoch": 7.98,
      "learning_rate": 2.1158119725859282e-05,
      "loss": 0.3586,
      "step": 22080,
      "task_loss": 0.5080243945121765
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4089392423629761,
      "epoch": 7.98,
      "learning_rate": 2.1093045818612006e-05,
      "loss": 0.347,
      "step": 22090,
      "task_loss": 0.41398802399635315
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4324132800102234,
      "epoch": 7.99,
      "learning_rate": 2.102801783861561e-05,
      "loss": 0.3558,
      "step": 22100,
      "task_loss": 0.5894123315811157
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47795045375823975,
      "epoch": 7.99,
      "learning_rate": 2.0963036121176206e-05,
      "loss": 0.3962,
      "step": 22110,
      "task_loss": 0.7632349729537964
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4554988145828247,
      "epoch": 7.99,
      "learning_rate": 2.0898101001361418e-05,
      "loss": 0.4285,
      "step": 22120,
      "task_loss": 0.838047981262207
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4030979871749878,
      "epoch": 8.0,
      "learning_rate": 2.083321281399851e-05,
      "loss": 0.4023,
      "step": 22130,
      "task_loss": 0.6322430372238159
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46217092871665955,
      "epoch": 8.0,
      "learning_rate": 2.0768371893672834e-05,
      "loss": 0.3616,
      "step": 22140,
      "task_loss": 0.5198900103569031
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3689584732055664,
      "epoch": 8.01,
      "learning_rate": 2.0703578574725962e-05,
      "loss": 0.3295,
      "step": 22150,
      "task_loss": 0.5603437423706055
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3704279661178589,
      "epoch": 8.01,
      "learning_rate": 2.0638833191254032e-05,
      "loss": 0.3244,
      "step": 22160,
      "task_loss": 0.4228596091270447
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4226457476615906,
      "epoch": 8.01,
      "learning_rate": 2.0574136077106028e-05,
      "loss": 0.3405,
      "step": 22170,
      "task_loss": 0.663974404335022
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2713632583618164,
      "epoch": 8.02,
      "learning_rate": 2.0509487565882006e-05,
      "loss": 0.3215,
      "step": 22180,
      "task_loss": 0.1982375979423523
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18484729528427124,
      "epoch": 8.02,
      "learning_rate": 2.0444887990931454e-05,
      "loss": 0.3104,
      "step": 22190,
      "task_loss": 0.37546926736831665
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2333574742078781,
      "epoch": 8.02,
      "learning_rate": 2.0386790489299768e-05,
      "loss": 0.326,
      "step": 22200,
      "task_loss": 0.40536680817604065
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3073742091655731,
      "epoch": 8.03,
      "learning_rate": 2.0322284810741438e-05,
      "loss": 0.3456,
      "step": 22210,
      "task_loss": 0.6669155359268188
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31559669971466064,
      "epoch": 8.03,
      "learning_rate": 2.0257829033736913e-05,
      "loss": 0.3496,
      "step": 22220,
      "task_loss": 0.4252294898033142
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3236352205276489,
      "epoch": 8.03,
      "learning_rate": 2.0193423490641865e-05,
      "loss": 0.346,
      "step": 22230,
      "task_loss": 0.31383252143859863
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28318703174591064,
      "epoch": 8.04,
      "learning_rate": 2.012906851355292e-05,
      "loss": 0.3354,
      "step": 22240,
      "task_loss": 0.45789241790771484
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2784233093261719,
      "epoch": 8.04,
      "learning_rate": 2.0064764434306e-05,
      "loss": 0.3764,
      "step": 22250,
      "task_loss": 0.5130239725112915
    },
    {
      "epoch": 8.04,
      "eval_exact_match": 82.99905392620624,
      "eval_f1": 89.85647992545962,
      "step": 22250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35418766736984253,
      "epoch": 8.04,
      "learning_rate": 2.0000511584474526e-05,
      "loss": 0.3849,
      "step": 22260,
      "task_loss": 0.6872138977050781
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3738144338130951,
      "epoch": 8.05,
      "learning_rate": 1.9936310295367844e-05,
      "loss": 0.3544,
      "step": 22270,
      "task_loss": 1.0411014556884766
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3013811707496643,
      "epoch": 8.05,
      "learning_rate": 1.9872160898029345e-05,
      "loss": 0.3503,
      "step": 22280,
      "task_loss": 1.0129387378692627
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37890174984931946,
      "epoch": 8.06,
      "learning_rate": 1.9808063723234918e-05,
      "loss": 0.3773,
      "step": 22290,
      "task_loss": 0.4624231457710266
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3464246988296509,
      "epoch": 8.06,
      "learning_rate": 1.9744019101491147e-05,
      "loss": 0.3466,
      "step": 22300,
      "task_loss": 0.8617278337478638
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3005381226539612,
      "epoch": 8.06,
      "learning_rate": 1.968002736303364e-05,
      "loss": 0.3255,
      "step": 22310,
      "task_loss": 0.4841776490211487
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38545113801956177,
      "epoch": 8.07,
      "learning_rate": 1.9616088837825307e-05,
      "loss": 0.3392,
      "step": 22320,
      "task_loss": 0.839179277420044
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2746634781360626,
      "epoch": 8.07,
      "learning_rate": 1.9552203855554702e-05,
      "loss": 0.3239,
      "step": 22330,
      "task_loss": 0.5544817447662354
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3266233503818512,
      "epoch": 8.07,
      "learning_rate": 1.948837274563426e-05,
      "loss": 0.3442,
      "step": 22340,
      "task_loss": 0.3120766282081604
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2832903563976288,
      "epoch": 8.08,
      "learning_rate": 1.9424595837198654e-05,
      "loss": 0.3225,
      "step": 22350,
      "task_loss": 0.5503286719322205
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2415936142206192,
      "epoch": 8.08,
      "learning_rate": 1.936087345910306e-05,
      "loss": 0.3388,
      "step": 22360,
      "task_loss": 0.4857402443885803
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40143972635269165,
      "epoch": 8.08,
      "learning_rate": 1.929720593992152e-05,
      "loss": 0.332,
      "step": 22370,
      "task_loss": 0.700149655342102
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3912586569786072,
      "epoch": 8.09,
      "learning_rate": 1.9233593607945127e-05,
      "loss": 0.3277,
      "step": 22380,
      "task_loss": 0.6662502288818359
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24969267845153809,
      "epoch": 8.09,
      "learning_rate": 1.917003679118049e-05,
      "loss": 0.3495,
      "step": 22390,
      "task_loss": 0.33860787749290466
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4209668040275574,
      "epoch": 8.1,
      "learning_rate": 1.9106535817347912e-05,
      "loss": 0.3384,
      "step": 22400,
      "task_loss": 0.3553839325904846
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2749325633049011,
      "epoch": 8.1,
      "learning_rate": 1.9043091013879773e-05,
      "loss": 0.3445,
      "step": 22410,
      "task_loss": 0.48762959241867065
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3994620144367218,
      "epoch": 8.1,
      "learning_rate": 1.897970270791881e-05,
      "loss": 0.3522,
      "step": 22420,
      "task_loss": 0.5795477628707886
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3599868714809418,
      "epoch": 8.11,
      "learning_rate": 1.891637122631645e-05,
      "loss": 0.3621,
      "step": 22430,
      "task_loss": 0.8509477376937866
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29197603464126587,
      "epoch": 8.11,
      "learning_rate": 1.88530968956311e-05,
      "loss": 0.3364,
      "step": 22440,
      "task_loss": 0.39719563722610474
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24607793986797333,
      "epoch": 8.11,
      "learning_rate": 1.8789880042126502e-05,
      "loss": 0.3425,
      "step": 22450,
      "task_loss": 0.2874683737754822
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3210759162902832,
      "epoch": 8.12,
      "learning_rate": 1.8726720991769983e-05,
      "loss": 0.3485,
      "step": 22460,
      "task_loss": 0.6553330421447754
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34106212854385376,
      "epoch": 8.12,
      "learning_rate": 1.8663620070230873e-05,
      "loss": 0.3472,
      "step": 22470,
      "task_loss": 0.47652125358581543
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35027986764907837,
      "epoch": 8.12,
      "learning_rate": 1.8600577602878722e-05,
      "loss": 0.3235,
      "step": 22480,
      "task_loss": 0.4614375829696655
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36589890718460083,
      "epoch": 8.13,
      "learning_rate": 1.8537593914781706e-05,
      "loss": 0.3435,
      "step": 22490,
      "task_loss": 0.7123420834541321
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38203728199005127,
      "epoch": 8.13,
      "learning_rate": 1.847466933070489e-05,
      "loss": 0.3821,
      "step": 22500,
      "task_loss": 0.4685204327106476
    },
    {
      "epoch": 8.13,
      "eval_exact_match": 82.86660359508042,
      "eval_f1": 89.73513596202447,
      "step": 22500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.373041033744812,
      "epoch": 8.14,
      "learning_rate": 1.8411804175108595e-05,
      "loss": 0.3285,
      "step": 22510,
      "task_loss": 0.7189540863037109
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26004570722579956,
      "epoch": 8.14,
      "learning_rate": 1.8348998772146698e-05,
      "loss": 0.338,
      "step": 22520,
      "task_loss": 0.36693066358566284
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34126996994018555,
      "epoch": 8.14,
      "learning_rate": 1.828625344566498e-05,
      "loss": 0.3419,
      "step": 22530,
      "task_loss": 0.47086426615715027
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3550693988800049,
      "epoch": 8.15,
      "learning_rate": 1.822356851919943e-05,
      "loss": 0.3521,
      "step": 22540,
      "task_loss": 0.7392463684082031
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3915178179740906,
      "epoch": 8.15,
      "learning_rate": 1.816094431597464e-05,
      "loss": 0.363,
      "step": 22550,
      "task_loss": 0.5709539651870728
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41465628147125244,
      "epoch": 8.15,
      "learning_rate": 1.8098381158902023e-05,
      "loss": 0.3423,
      "step": 22560,
      "task_loss": 0.49347808957099915
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2688755393028259,
      "epoch": 8.16,
      "learning_rate": 1.803587937057828e-05,
      "loss": 0.3412,
      "step": 22570,
      "task_loss": 0.5061269402503967
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3253447115421295,
      "epoch": 8.16,
      "learning_rate": 1.7973439273283633e-05,
      "loss": 0.3537,
      "step": 22580,
      "task_loss": 0.4826359748840332
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4960862994194031,
      "epoch": 8.16,
      "learning_rate": 1.791106118898024e-05,
      "loss": 0.3537,
      "step": 22590,
      "task_loss": 0.66794753074646
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.308212548494339,
      "epoch": 8.17,
      "learning_rate": 1.7848745439310454e-05,
      "loss": 0.3229,
      "step": 22600,
      "task_loss": 0.4701980650424957
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3355109393596649,
      "epoch": 8.17,
      "learning_rate": 1.7786492345595258e-05,
      "loss": 0.31,
      "step": 22610,
      "task_loss": 0.3176344037055969
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37522727251052856,
      "epoch": 8.17,
      "learning_rate": 1.7724302228832518e-05,
      "loss": 0.3228,
      "step": 22620,
      "task_loss": 0.7154313325881958
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30252838134765625,
      "epoch": 8.18,
      "learning_rate": 1.7662175409695403e-05,
      "loss": 0.3255,
      "step": 22630,
      "task_loss": 0.3516947627067566
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2813971936702728,
      "epoch": 8.18,
      "learning_rate": 1.760011220853067e-05,
      "loss": 0.3469,
      "step": 22640,
      "task_loss": 0.3032105267047882
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3111063241958618,
      "epoch": 8.19,
      "learning_rate": 1.7538112945357074e-05,
      "loss": 0.3466,
      "step": 22650,
      "task_loss": 0.6328559517860413
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3012014627456665,
      "epoch": 8.19,
      "learning_rate": 1.747617793986364e-05,
      "loss": 0.3383,
      "step": 22660,
      "task_loss": 0.7136552333831787
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4168001413345337,
      "epoch": 8.19,
      "learning_rate": 1.7414307511408105e-05,
      "loss": 0.33,
      "step": 22670,
      "task_loss": 0.7946159839630127
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3220928907394409,
      "epoch": 8.2,
      "learning_rate": 1.7352501979015185e-05,
      "loss": 0.3734,
      "step": 22680,
      "task_loss": 0.6289501786231995
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3598100543022156,
      "epoch": 8.2,
      "learning_rate": 1.7290761661374998e-05,
      "loss": 0.3356,
      "step": 22690,
      "task_loss": 0.5484864711761475
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39641427993774414,
      "epoch": 8.2,
      "learning_rate": 1.722908687684138e-05,
      "loss": 0.3569,
      "step": 22700,
      "task_loss": 0.34807297587394714
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3675978183746338,
      "epoch": 8.21,
      "learning_rate": 1.7167477943430275e-05,
      "loss": 0.3319,
      "step": 22710,
      "task_loss": 0.27945399284362793
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37253180146217346,
      "epoch": 8.21,
      "learning_rate": 1.710593517881802e-05,
      "loss": 0.3363,
      "step": 22720,
      "task_loss": 0.5771284103393555
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46653449535369873,
      "epoch": 8.21,
      "learning_rate": 1.7044458900339855e-05,
      "loss": 0.3938,
      "step": 22730,
      "task_loss": 0.6149200201034546
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32485833764076233,
      "epoch": 8.22,
      "learning_rate": 1.6983049424988106e-05,
      "loss": 0.3088,
      "step": 22740,
      "task_loss": 0.6905471086502075
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32625168561935425,
      "epoch": 8.22,
      "learning_rate": 1.6921707069410698e-05,
      "loss": 0.3635,
      "step": 22750,
      "task_loss": 0.6236306428909302
    },
    {
      "epoch": 8.22,
      "eval_exact_match": 82.82876064333018,
      "eval_f1": 89.90657086280171,
      "step": 22750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36564576625823975,
      "epoch": 8.23,
      "learning_rate": 1.686043214990943e-05,
      "loss": 0.3627,
      "step": 22760,
      "task_loss": 0.611585259437561
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.323543906211853,
      "epoch": 8.23,
      "learning_rate": 1.6799224982438413e-05,
      "loss": 0.3304,
      "step": 22770,
      "task_loss": 0.7676356434822083
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29762110114097595,
      "epoch": 8.23,
      "learning_rate": 1.6738085882602358e-05,
      "loss": 0.3379,
      "step": 22780,
      "task_loss": 0.8042917251586914
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3024720549583435,
      "epoch": 8.24,
      "learning_rate": 1.6677015165655034e-05,
      "loss": 0.3291,
      "step": 22790,
      "task_loss": 0.48665961623191833
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4022981524467468,
      "epoch": 8.24,
      "learning_rate": 1.6616013146497598e-05,
      "loss": 0.3783,
      "step": 22800,
      "task_loss": 0.40880414843559265
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4667377173900604,
      "epoch": 8.24,
      "learning_rate": 1.6555080139676972e-05,
      "loss": 0.3679,
      "step": 22810,
      "task_loss": 0.6365650296211243
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3593342900276184,
      "epoch": 8.25,
      "learning_rate": 1.6494216459384225e-05,
      "loss": 0.3606,
      "step": 22820,
      "task_loss": 0.582703709602356
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31626957654953003,
      "epoch": 8.25,
      "learning_rate": 1.6433422419452973e-05,
      "loss": 0.3391,
      "step": 22830,
      "task_loss": 0.41589945554733276
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2823561429977417,
      "epoch": 8.25,
      "learning_rate": 1.6372698333357705e-05,
      "loss": 0.3464,
      "step": 22840,
      "task_loss": 0.6627315282821655
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3531639873981476,
      "epoch": 8.26,
      "learning_rate": 1.6312044514212245e-05,
      "loss": 0.3758,
      "step": 22850,
      "task_loss": 0.809797465801239
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3731263279914856,
      "epoch": 8.26,
      "learning_rate": 1.625146127476807e-05,
      "loss": 0.3786,
      "step": 22860,
      "task_loss": 0.7332509160041809
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2990288734436035,
      "epoch": 8.27,
      "learning_rate": 1.6190948927412758e-05,
      "loss": 0.3298,
      "step": 22870,
      "task_loss": 0.38085365295410156
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2953346371650696,
      "epoch": 8.27,
      "learning_rate": 1.613050778416828e-05,
      "loss": 0.3321,
      "step": 22880,
      "task_loss": 0.7593106031417847
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30572813749313354,
      "epoch": 8.27,
      "learning_rate": 1.607013815668954e-05,
      "loss": 0.3467,
      "step": 22890,
      "task_loss": 0.40899914503097534
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3036983907222748,
      "epoch": 8.28,
      "learning_rate": 1.6009840356262603e-05,
      "loss": 0.3755,
      "step": 22900,
      "task_loss": 0.4268397092819214
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37761154770851135,
      "epoch": 8.28,
      "learning_rate": 1.594961469380322e-05,
      "loss": 0.3332,
      "step": 22910,
      "task_loss": 0.41044527292251587
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2791655659675598,
      "epoch": 8.28,
      "learning_rate": 1.5889461479855153e-05,
      "loss": 0.3374,
      "step": 22920,
      "task_loss": 0.546173095703125
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30385681986808777,
      "epoch": 8.29,
      "learning_rate": 1.5829381024588614e-05,
      "loss": 0.3145,
      "step": 22930,
      "task_loss": 0.5504984855651855
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31653857231140137,
      "epoch": 8.29,
      "learning_rate": 1.576937363779861e-05,
      "loss": 0.3167,
      "step": 22940,
      "task_loss": 0.4941592812538147
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3911159634590149,
      "epoch": 8.29,
      "learning_rate": 1.570943962890342e-05,
      "loss": 0.3725,
      "step": 22950,
      "task_loss": 0.6922523975372314
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3799233138561249,
      "epoch": 8.3,
      "learning_rate": 1.5649579306942943e-05,
      "loss": 0.3422,
      "step": 22960,
      "task_loss": 0.794413685798645
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33418601751327515,
      "epoch": 8.3,
      "learning_rate": 1.558979298057715e-05,
      "loss": 0.3407,
      "step": 22970,
      "task_loss": 0.6918811798095703
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.48559606075286865,
      "epoch": 8.31,
      "learning_rate": 1.55300809580844e-05,
      "loss": 0.3429,
      "step": 22980,
      "task_loss": 0.6242055892944336
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3377108573913574,
      "epoch": 8.31,
      "learning_rate": 1.5470443547360003e-05,
      "loss": 0.3321,
      "step": 22990,
      "task_loss": 0.6736431121826172
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2609238624572754,
      "epoch": 8.31,
      "learning_rate": 1.5410881055914478e-05,
      "loss": 0.3416,
      "step": 23000,
      "task_loss": 0.43551382422447205
    },
    {
      "epoch": 8.31,
      "eval_exact_match": 83.0558183538316,
      "eval_f1": 89.9465346529477,
      "step": 23000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31109219789505005,
      "epoch": 8.32,
      "learning_rate": 1.5351393790872065e-05,
      "loss": 0.342,
      "step": 23010,
      "task_loss": 0.5256096124649048
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33370792865753174,
      "epoch": 8.32,
      "learning_rate": 1.529198205896911e-05,
      "loss": 0.3416,
      "step": 23020,
      "task_loss": 0.42889517545700073
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.323059618473053,
      "epoch": 8.32,
      "learning_rate": 1.5232646166552486e-05,
      "loss": 0.3426,
      "step": 23030,
      "task_loss": 0.44258755445480347
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36747613549232483,
      "epoch": 8.33,
      "learning_rate": 1.517338641957799e-05,
      "loss": 0.373,
      "step": 23040,
      "task_loss": 0.6945732235908508
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3251588046550751,
      "epoch": 8.33,
      "learning_rate": 1.5114203123608817e-05,
      "loss": 0.3609,
      "step": 23050,
      "task_loss": 0.5549402236938477
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29842013120651245,
      "epoch": 8.33,
      "learning_rate": 1.5055096583813937e-05,
      "loss": 0.3394,
      "step": 23060,
      "task_loss": 0.3050425946712494
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3361862599849701,
      "epoch": 8.34,
      "learning_rate": 1.499606710496656e-05,
      "loss": 0.3192,
      "step": 23070,
      "task_loss": 1.0102906227111816
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3161877989768982,
      "epoch": 8.34,
      "learning_rate": 1.4937114991442491e-05,
      "loss": 0.3482,
      "step": 23080,
      "task_loss": 0.4595767557621002
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39450502395629883,
      "epoch": 8.34,
      "learning_rate": 1.48782405472187e-05,
      "loss": 0.3569,
      "step": 23090,
      "task_loss": 0.48541557788848877
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4245753586292267,
      "epoch": 8.35,
      "learning_rate": 1.4819444075871585e-05,
      "loss": 0.3323,
      "step": 23100,
      "task_loss": 0.6281993389129639
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29610657691955566,
      "epoch": 8.35,
      "learning_rate": 1.4760725880575534e-05,
      "loss": 0.3332,
      "step": 23110,
      "task_loss": 0.373093843460083
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.353634774684906,
      "epoch": 8.36,
      "learning_rate": 1.4702086264101305e-05,
      "loss": 0.3962,
      "step": 23120,
      "task_loss": 0.8853891491889954
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29741644859313965,
      "epoch": 8.36,
      "learning_rate": 1.4643525528814492e-05,
      "loss": 0.2991,
      "step": 23130,
      "task_loss": 0.7390473484992981
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37111955881118774,
      "epoch": 8.36,
      "learning_rate": 1.4585043976673916e-05,
      "loss": 0.334,
      "step": 23140,
      "task_loss": 0.7053943872451782
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3489043712615967,
      "epoch": 8.37,
      "learning_rate": 1.4526641909230136e-05,
      "loss": 0.3145,
      "step": 23150,
      "task_loss": 0.5620697736740112
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35167962312698364,
      "epoch": 8.37,
      "learning_rate": 1.446831962762385e-05,
      "loss": 0.3207,
      "step": 23160,
      "task_loss": 0.5666888356208801
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3380488157272339,
      "epoch": 8.37,
      "learning_rate": 1.4410077432584366e-05,
      "loss": 0.3348,
      "step": 23170,
      "task_loss": 0.8974223732948303
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29322537779808044,
      "epoch": 8.38,
      "learning_rate": 1.435191562442799e-05,
      "loss": 0.3445,
      "step": 23180,
      "task_loss": 0.5745025277137756
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31768181920051575,
      "epoch": 8.38,
      "learning_rate": 1.4293834503056611e-05,
      "loss": 0.3563,
      "step": 23190,
      "task_loss": 0.39405155181884766
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41235482692718506,
      "epoch": 8.38,
      "learning_rate": 1.4235834367955988e-05,
      "loss": 0.3429,
      "step": 23200,
      "task_loss": 1.0112557411193848
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27087026834487915,
      "epoch": 8.39,
      "learning_rate": 1.417791551819433e-05,
      "loss": 0.3267,
      "step": 23210,
      "task_loss": 0.3743332028388977
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49186843633651733,
      "epoch": 8.39,
      "learning_rate": 1.4120078252420704e-05,
      "loss": 0.3509,
      "step": 23220,
      "task_loss": 0.6274521350860596
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3783082962036133,
      "epoch": 8.4,
      "learning_rate": 1.4062322868863515e-05,
      "loss": 0.3594,
      "step": 23230,
      "task_loss": 0.38089364767074585
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21668830513954163,
      "epoch": 8.4,
      "learning_rate": 1.4004649665328914e-05,
      "loss": 0.316,
      "step": 23240,
      "task_loss": 0.3113134503364563
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32121360301971436,
      "epoch": 8.4,
      "learning_rate": 1.3947058939199343e-05,
      "loss": 0.3575,
      "step": 23250,
      "task_loss": 0.3753986656665802
    },
    {
      "epoch": 8.4,
      "eval_exact_match": 83.12204351939451,
      "eval_f1": 89.97779373367567,
      "step": 23250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32461658120155334,
      "epoch": 8.41,
      "learning_rate": 1.388955098743196e-05,
      "loss": 0.3335,
      "step": 23260,
      "task_loss": 0.4626964330673218
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31419169902801514,
      "epoch": 8.41,
      "learning_rate": 1.3832126106557103e-05,
      "loss": 0.3385,
      "step": 23270,
      "task_loss": 0.5672830939292908
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28030550479888916,
      "epoch": 8.41,
      "learning_rate": 1.377478459267674e-05,
      "loss": 0.3333,
      "step": 23280,
      "task_loss": 0.6520091891288757
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3894043564796448,
      "epoch": 8.42,
      "learning_rate": 1.3717526741463045e-05,
      "loss": 0.3725,
      "step": 23290,
      "task_loss": 0.9674755334854126
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3570999801158905,
      "epoch": 8.42,
      "learning_rate": 1.3660352848156717e-05,
      "loss": 0.3313,
      "step": 23300,
      "task_loss": 0.33604758977890015
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3844010829925537,
      "epoch": 8.42,
      "learning_rate": 1.3603263207565584e-05,
      "loss": 0.3143,
      "step": 23310,
      "task_loss": 0.5411736369132996
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2638304829597473,
      "epoch": 8.43,
      "learning_rate": 1.3546258114063033e-05,
      "loss": 0.3217,
      "step": 23320,
      "task_loss": 0.6530275344848633
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3618456721305847,
      "epoch": 8.43,
      "learning_rate": 1.3489337861586507e-05,
      "loss": 0.341,
      "step": 23330,
      "task_loss": 0.48471835255622864
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4442026615142822,
      "epoch": 8.44,
      "learning_rate": 1.3432502743635948e-05,
      "loss": 0.3354,
      "step": 23340,
      "task_loss": 0.5538107752799988
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28205060958862305,
      "epoch": 8.44,
      "learning_rate": 1.3375753053272343e-05,
      "loss": 0.3587,
      "step": 23350,
      "task_loss": 0.4763302206993103
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3960496783256531,
      "epoch": 8.44,
      "learning_rate": 1.3319089083116176e-05,
      "loss": 0.3454,
      "step": 23360,
      "task_loss": 0.461125910282135
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2497769594192505,
      "epoch": 8.45,
      "learning_rate": 1.326251112534595e-05,
      "loss": 0.353,
      "step": 23370,
      "task_loss": 0.6359678506851196
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4250258803367615,
      "epoch": 8.45,
      "learning_rate": 1.32060194716966e-05,
      "loss": 0.3319,
      "step": 23380,
      "task_loss": 0.5756826400756836
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4019028842449188,
      "epoch": 8.45,
      "learning_rate": 1.3149614413458139e-05,
      "loss": 0.3462,
      "step": 23390,
      "task_loss": 0.9020615220069885
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3332751989364624,
      "epoch": 8.46,
      "learning_rate": 1.3093296241473975e-05,
      "loss": 0.3435,
      "step": 23400,
      "task_loss": 0.681817889213562
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2737824618816376,
      "epoch": 8.46,
      "learning_rate": 1.3037065246139558e-05,
      "loss": 0.3495,
      "step": 23410,
      "task_loss": 0.40974706411361694
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3022178113460541,
      "epoch": 8.46,
      "learning_rate": 1.2980921717400803e-05,
      "loss": 0.3158,
      "step": 23420,
      "task_loss": 0.33448854088783264
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31435340642929077,
      "epoch": 8.47,
      "learning_rate": 1.292486594475264e-05,
      "loss": 0.3092,
      "step": 23430,
      "task_loss": 0.4791732430458069
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.311881422996521,
      "epoch": 8.47,
      "learning_rate": 1.2868898217237457e-05,
      "loss": 0.3531,
      "step": 23440,
      "task_loss": 0.34755033254623413
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28101277351379395,
      "epoch": 8.47,
      "learning_rate": 1.2813018823443685e-05,
      "loss": 0.3431,
      "step": 23450,
      "task_loss": 0.3975358009338379
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36834776401519775,
      "epoch": 8.48,
      "learning_rate": 1.2757228051504276e-05,
      "loss": 0.3747,
      "step": 23460,
      "task_loss": 0.6111078262329102
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32710838317871094,
      "epoch": 8.48,
      "learning_rate": 1.2701526189095216e-05,
      "loss": 0.3374,
      "step": 23470,
      "task_loss": 0.6565884351730347
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27171874046325684,
      "epoch": 8.49,
      "learning_rate": 1.2645913523434012e-05,
      "loss": 0.3174,
      "step": 23480,
      "task_loss": 0.7133381962776184
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28032228350639343,
      "epoch": 8.49,
      "learning_rate": 1.2590390341278302e-05,
      "loss": 0.3107,
      "step": 23490,
      "task_loss": 0.2568185031414032
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33454135060310364,
      "epoch": 8.49,
      "learning_rate": 1.2534956928924263e-05,
      "loss": 0.3542,
      "step": 23500,
      "task_loss": 0.3434966206550598
    },
    {
      "epoch": 8.49,
      "eval_exact_match": 83.09366130558183,
      "eval_f1": 89.88354014153792,
      "step": 23500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2474634200334549,
      "epoch": 8.5,
      "learning_rate": 1.2479613572205212e-05,
      "loss": 0.3277,
      "step": 23510,
      "task_loss": 0.43857425451278687
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36067289113998413,
      "epoch": 8.5,
      "learning_rate": 1.2424360556490109e-05,
      "loss": 0.3517,
      "step": 23520,
      "task_loss": 0.5712844133377075
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2622072398662567,
      "epoch": 8.5,
      "learning_rate": 1.2369198166682094e-05,
      "loss": 0.321,
      "step": 23530,
      "task_loss": 0.4479938745498657
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24256464838981628,
      "epoch": 8.51,
      "learning_rate": 1.2314126687216974e-05,
      "loss": 0.3231,
      "step": 23540,
      "task_loss": 0.3161258101463318
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24894395470619202,
      "epoch": 8.51,
      "learning_rate": 1.2259146402061829e-05,
      "loss": 0.3464,
      "step": 23550,
      "task_loss": 0.3210480809211731
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3175157904624939,
      "epoch": 8.51,
      "learning_rate": 1.2204257594713501e-05,
      "loss": 0.3401,
      "step": 23560,
      "task_loss": 0.35496532917022705
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2891998589038849,
      "epoch": 8.52,
      "learning_rate": 1.2149460548197143e-05,
      "loss": 0.3373,
      "step": 23570,
      "task_loss": 0.4902755618095398
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38131245970726013,
      "epoch": 8.52,
      "learning_rate": 1.2094755545064719e-05,
      "loss": 0.3375,
      "step": 23580,
      "task_loss": 0.6391329765319824
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3201846480369568,
      "epoch": 8.53,
      "learning_rate": 1.2040142867393671e-05,
      "loss": 0.3589,
      "step": 23590,
      "task_loss": 0.46474406123161316
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32470813393592834,
      "epoch": 8.53,
      "learning_rate": 1.1985622796785295e-05,
      "loss": 0.3063,
      "step": 23600,
      "task_loss": 0.8694692850112915
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3770723044872284,
      "epoch": 8.53,
      "learning_rate": 1.1931195614363417e-05,
      "loss": 0.3198,
      "step": 23610,
      "task_loss": 0.585353434085846
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3081551492214203,
      "epoch": 8.54,
      "learning_rate": 1.1876861600772893e-05,
      "loss": 0.3371,
      "step": 23620,
      "task_loss": 0.6682502627372742
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41357192397117615,
      "epoch": 8.54,
      "learning_rate": 1.1822621036178173e-05,
      "loss": 0.359,
      "step": 23630,
      "task_loss": 0.5059230327606201
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31322067975997925,
      "epoch": 8.54,
      "learning_rate": 1.1768474200261838e-05,
      "loss": 0.3108,
      "step": 23640,
      "task_loss": 0.473344087600708
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2495288848876953,
      "epoch": 8.55,
      "learning_rate": 1.1714421372223179e-05,
      "loss": 0.3173,
      "step": 23650,
      "task_loss": 0.38251379132270813
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21137635409832,
      "epoch": 8.55,
      "learning_rate": 1.1660462830776766e-05,
      "loss": 0.3165,
      "step": 23660,
      "task_loss": 0.6774376630783081
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25261959433555603,
      "epoch": 8.55,
      "learning_rate": 1.1606598854150983e-05,
      "loss": 0.3302,
      "step": 23670,
      "task_loss": 0.6455027461051941
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38040322065353394,
      "epoch": 8.56,
      "learning_rate": 1.1552829720086581e-05,
      "loss": 0.366,
      "step": 23680,
      "task_loss": 0.4775557518005371
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3587028384208679,
      "epoch": 8.56,
      "learning_rate": 1.1499155705835338e-05,
      "loss": 0.3393,
      "step": 23690,
      "task_loss": 0.6567662358283997
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4207734167575836,
      "epoch": 8.57,
      "learning_rate": 1.1445577088158486e-05,
      "loss": 0.347,
      "step": 23700,
      "task_loss": 0.6872299909591675
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3458840250968933,
      "epoch": 8.57,
      "learning_rate": 1.1392094143325404e-05,
      "loss": 0.3251,
      "step": 23710,
      "task_loss": 0.5791058540344238
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2893916368484497,
      "epoch": 8.57,
      "learning_rate": 1.1338707147112145e-05,
      "loss": 0.3497,
      "step": 23720,
      "task_loss": 0.3764035403728485
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3328968584537506,
      "epoch": 8.58,
      "learning_rate": 1.1285416374800018e-05,
      "loss": 0.3678,
      "step": 23730,
      "task_loss": 0.46390044689178467
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2898051142692566,
      "epoch": 8.58,
      "learning_rate": 1.1232222101174148e-05,
      "loss": 0.3194,
      "step": 23740,
      "task_loss": 0.8645293712615967
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2928909659385681,
      "epoch": 8.58,
      "learning_rate": 1.1179124600522107e-05,
      "loss": 0.3093,
      "step": 23750,
      "task_loss": 0.472074955701828
    },
    {
      "epoch": 8.58,
      "eval_exact_match": 83.28287606433302,
      "eval_f1": 90.04093742477677,
      "step": 23750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3210963010787964,
      "epoch": 8.59,
      "learning_rate": 1.1126124146632464e-05,
      "loss": 0.3379,
      "step": 23760,
      "task_loss": 0.5641353130340576
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32252565026283264,
      "epoch": 8.59,
      "learning_rate": 1.1073221012793393e-05,
      "loss": 0.3445,
      "step": 23770,
      "task_loss": 0.6479802131652832
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.290593683719635,
      "epoch": 8.59,
      "learning_rate": 1.102041547179121e-05,
      "loss": 0.3177,
      "step": 23780,
      "task_loss": 0.4076792597770691
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30375468730926514,
      "epoch": 8.6,
      "learning_rate": 1.0967707795909077e-05,
      "loss": 0.3838,
      "step": 23790,
      "task_loss": 0.4609778821468353
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2941964268684387,
      "epoch": 8.6,
      "learning_rate": 1.0915098256925474e-05,
      "loss": 0.3539,
      "step": 23800,
      "task_loss": 0.5235286355018616
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33607012033462524,
      "epoch": 8.6,
      "learning_rate": 1.0862587126112873e-05,
      "loss": 0.3128,
      "step": 23810,
      "task_loss": 0.6910580396652222
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.321811318397522,
      "epoch": 8.61,
      "learning_rate": 1.0810174674236335e-05,
      "loss": 0.2864,
      "step": 23820,
      "task_loss": 0.5703722834587097
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37454020977020264,
      "epoch": 8.61,
      "learning_rate": 1.0757861171552074e-05,
      "loss": 0.3266,
      "step": 23830,
      "task_loss": 0.5501848459243774
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32955604791641235,
      "epoch": 8.62,
      "learning_rate": 1.0705646887806119e-05,
      "loss": 0.3291,
      "step": 23840,
      "task_loss": 0.6074403524398804
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3324754238128662,
      "epoch": 8.62,
      "learning_rate": 1.0653532092232843e-05,
      "loss": 0.3458,
      "step": 23850,
      "task_loss": 0.34126532077789307
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33597028255462646,
      "epoch": 8.62,
      "learning_rate": 1.0601517053553697e-05,
      "loss": 0.3562,
      "step": 23860,
      "task_loss": 0.7301780581474304
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3327868580818176,
      "epoch": 8.63,
      "learning_rate": 1.0549602039975683e-05,
      "loss": 0.3614,
      "step": 23870,
      "task_loss": 0.520077645778656
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2888182997703552,
      "epoch": 8.63,
      "learning_rate": 1.0497787319190076e-05,
      "loss": 0.3371,
      "step": 23880,
      "task_loss": 0.4426148235797882
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25183236598968506,
      "epoch": 8.63,
      "learning_rate": 1.0446073158370996e-05,
      "loss": 0.3579,
      "step": 23890,
      "task_loss": 0.7299506664276123
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29746419191360474,
      "epoch": 8.64,
      "learning_rate": 1.0394459824174065e-05,
      "loss": 0.3397,
      "step": 23900,
      "task_loss": 0.5756563544273376
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3072645664215088,
      "epoch": 8.64,
      "learning_rate": 1.0342947582734962e-05,
      "loss": 0.3284,
      "step": 23910,
      "task_loss": 0.5769457817077637
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35788294672966003,
      "epoch": 8.64,
      "learning_rate": 1.029153669966814e-05,
      "loss": 0.3455,
      "step": 23920,
      "task_loss": 0.6118526458740234
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2843881845474243,
      "epoch": 8.65,
      "learning_rate": 1.02402274400654e-05,
      "loss": 0.34,
      "step": 23930,
      "task_loss": 0.4474141299724579
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3228676915168762,
      "epoch": 8.65,
      "learning_rate": 1.0189020068494543e-05,
      "loss": 0.3477,
      "step": 23940,
      "task_loss": 0.43117979168891907
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32196763157844543,
      "epoch": 8.66,
      "learning_rate": 1.0137914848997974e-05,
      "loss": 0.3313,
      "step": 23950,
      "task_loss": 0.8009005784988403
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.321696400642395,
      "epoch": 8.66,
      "learning_rate": 1.0086912045091422e-05,
      "loss": 0.3255,
      "step": 23960,
      "task_loss": 0.3024410009384155
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34478169679641724,
      "epoch": 8.66,
      "learning_rate": 1.0036011919762471e-05,
      "loss": 0.3573,
      "step": 23970,
      "task_loss": 0.5820472836494446
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30829232931137085,
      "epoch": 8.67,
      "learning_rate": 9.98521473546929e-06,
      "loss": 0.3552,
      "step": 23980,
      "task_loss": 0.7267586588859558
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3902008533477783,
      "epoch": 8.67,
      "learning_rate": 9.93452075413924e-06,
      "loss": 0.3605,
      "step": 23990,
      "task_loss": 0.5337940454483032
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3022475242614746,
      "epoch": 8.67,
      "learning_rate": 9.88393023716755e-06,
      "loss": 0.3119,
      "step": 24000,
      "task_loss": 0.2621081471443176
    },
    {
      "epoch": 8.67,
      "eval_exact_match": 83.20719016083254,
      "eval_f1": 89.99360400964093,
      "step": 24000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3147343397140503,
      "epoch": 8.68,
      "learning_rate": 9.833443445415917e-06,
      "loss": 0.3734,
      "step": 24010,
      "task_loss": 0.4403849244117737
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3691180348396301,
      "epoch": 8.68,
      "learning_rate": 9.783060639211225e-06,
      "loss": 0.3333,
      "step": 24020,
      "task_loss": 0.7338327169418335
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34679627418518066,
      "epoch": 8.68,
      "learning_rate": 9.732782078344166e-06,
      "loss": 0.3361,
      "step": 24030,
      "task_loss": 0.6814612746238708
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24168024957180023,
      "epoch": 8.69,
      "learning_rate": 9.682608022067903e-06,
      "loss": 0.3084,
      "step": 24040,
      "task_loss": 0.3496478199958801
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25266405940055847,
      "epoch": 8.69,
      "learning_rate": 9.632538729096749e-06,
      "loss": 0.3407,
      "step": 24050,
      "task_loss": 0.8575865030288696
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2826876640319824,
      "epoch": 8.7,
      "learning_rate": 9.58257445760482e-06,
      "loss": 0.3108,
      "step": 24060,
      "task_loss": 0.5487475395202637
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3644624948501587,
      "epoch": 8.7,
      "learning_rate": 9.532715465224678e-06,
      "loss": 0.3213,
      "step": 24070,
      "task_loss": 0.26308804750442505
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37130007147789,
      "epoch": 8.7,
      "learning_rate": 9.482962009046073e-06,
      "loss": 0.3427,
      "step": 24080,
      "task_loss": 0.6274560689926147
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34187111258506775,
      "epoch": 8.71,
      "learning_rate": 9.433314345614548e-06,
      "loss": 0.3268,
      "step": 24090,
      "task_loss": 0.6488524675369263
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3473740220069885,
      "epoch": 8.71,
      "learning_rate": 9.383772730930163e-06,
      "loss": 0.3466,
      "step": 24100,
      "task_loss": 0.5115723609924316
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34605729579925537,
      "epoch": 8.71,
      "learning_rate": 9.334337420446116e-06,
      "loss": 0.3619,
      "step": 24110,
      "task_loss": 0.5676177740097046
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27803748846054077,
      "epoch": 8.72,
      "learning_rate": 9.285008669067524e-06,
      "loss": 0.3174,
      "step": 24120,
      "task_loss": 1.14910089969635
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31176722049713135,
      "epoch": 8.72,
      "learning_rate": 9.23578673114999e-06,
      "loss": 0.3318,
      "step": 24130,
      "task_loss": 0.6742380857467651
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.412189245223999,
      "epoch": 8.72,
      "learning_rate": 9.18667186049839e-06,
      "loss": 0.3511,
      "step": 24140,
      "task_loss": 0.9542233943939209
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30980244278907776,
      "epoch": 8.73,
      "learning_rate": 9.13766431036551e-06,
      "loss": 0.354,
      "step": 24150,
      "task_loss": 0.554745078086853
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3011392056941986,
      "epoch": 8.73,
      "learning_rate": 9.088764333450769e-06,
      "loss": 0.3053,
      "step": 24160,
      "task_loss": 0.3188933730125427
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31323549151420593,
      "epoch": 8.74,
      "learning_rate": 9.039972181898872e-06,
      "loss": 0.3359,
      "step": 24170,
      "task_loss": 0.7727498412132263
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.375557541847229,
      "epoch": 8.74,
      "learning_rate": 8.991288107298571e-06,
      "loss": 0.3463,
      "step": 24180,
      "task_loss": 0.512423574924469
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33595043420791626,
      "epoch": 8.74,
      "learning_rate": 8.942712360681321e-06,
      "loss": 0.337,
      "step": 24190,
      "task_loss": 0.27191513776779175
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33838656544685364,
      "epoch": 8.75,
      "learning_rate": 8.894245192520018e-06,
      "loss": 0.3529,
      "step": 24200,
      "task_loss": 0.5304921865463257
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3112967610359192,
      "epoch": 8.75,
      "learning_rate": 8.845886852727651e-06,
      "loss": 0.3491,
      "step": 24210,
      "task_loss": 0.6719498634338379
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24464872479438782,
      "epoch": 8.75,
      "learning_rate": 8.79763759065612e-06,
      "loss": 0.3141,
      "step": 24220,
      "task_loss": 0.7396454215049744
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2790934443473816,
      "epoch": 8.76,
      "learning_rate": 8.749497655094812e-06,
      "loss": 0.3297,
      "step": 24230,
      "task_loss": 0.21234217286109924
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2472725212574005,
      "epoch": 8.76,
      "learning_rate": 8.701467294269437e-06,
      "loss": 0.3184,
      "step": 24240,
      "task_loss": 0.30988842248916626
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2908933162689209,
      "epoch": 8.76,
      "learning_rate": 8.653546755840686e-06,
      "loss": 0.3048,
      "step": 24250,
      "task_loss": 0.3451705873012543
    },
    {
      "epoch": 8.76,
      "eval_exact_match": 83.28287606433302,
      "eval_f1": 90.04894867144425,
      "step": 24250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3388180434703827,
      "epoch": 8.77,
      "learning_rate": 8.605736286902976e-06,
      "loss": 0.3563,
      "step": 24260,
      "task_loss": 0.44518616795539856
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3412635624408722,
      "epoch": 8.77,
      "learning_rate": 8.558036133983143e-06,
      "loss": 0.3404,
      "step": 24270,
      "task_loss": 1.031734585762024
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2612943649291992,
      "epoch": 8.77,
      "learning_rate": 8.510446543039228e-06,
      "loss": 0.3304,
      "step": 24280,
      "task_loss": 0.7055977582931519
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28402096033096313,
      "epoch": 8.78,
      "learning_rate": 8.462967759459164e-06,
      "loss": 0.3342,
      "step": 24290,
      "task_loss": 0.5784477591514587
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39517825841903687,
      "epoch": 8.78,
      "learning_rate": 8.415600028059531e-06,
      "loss": 0.3588,
      "step": 24300,
      "task_loss": 0.700530469417572
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3836681544780731,
      "epoch": 8.79,
      "learning_rate": 8.368343593084255e-06,
      "loss": 0.3471,
      "step": 24310,
      "task_loss": 0.4598133862018585
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30892038345336914,
      "epoch": 8.79,
      "learning_rate": 8.32119869820344e-06,
      "loss": 0.3385,
      "step": 24320,
      "task_loss": 0.5151838660240173
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.358475923538208,
      "epoch": 8.79,
      "learning_rate": 8.274165586511993e-06,
      "loss": 0.3616,
      "step": 24330,
      "task_loss": 0.7977138757705688
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4102436900138855,
      "epoch": 8.8,
      "learning_rate": 8.227244500528464e-06,
      "loss": 0.386,
      "step": 24340,
      "task_loss": 0.9865103960037231
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25304460525512695,
      "epoch": 8.8,
      "learning_rate": 8.180435682193744e-06,
      "loss": 0.3081,
      "step": 24350,
      "task_loss": 0.3072149455547333
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2997356057167053,
      "epoch": 8.8,
      "learning_rate": 8.133739372869856e-06,
      "loss": 0.3325,
      "step": 24360,
      "task_loss": 0.4807734489440918
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30522841215133667,
      "epoch": 8.81,
      "learning_rate": 8.08715581333865e-06,
      "loss": 0.3412,
      "step": 24370,
      "task_loss": 0.40793800354003906
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3938402235507965,
      "epoch": 8.81,
      "learning_rate": 8.040685243800636e-06,
      "loss": 0.334,
      "step": 24380,
      "task_loss": 0.4832313358783722
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38264673948287964,
      "epoch": 8.81,
      "learning_rate": 7.994327903873696e-06,
      "loss": 0.3277,
      "step": 24390,
      "task_loss": 0.7113062143325806
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2893344759941101,
      "epoch": 8.82,
      "learning_rate": 7.948084032591872e-06,
      "loss": 0.3337,
      "step": 24400,
      "task_loss": 0.28811150789260864
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18665549159049988,
      "epoch": 8.82,
      "learning_rate": 7.901953868404086e-06,
      "loss": 0.3474,
      "step": 24410,
      "task_loss": 0.5828348398208618
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2941107153892517,
      "epoch": 8.83,
      "learning_rate": 7.855937649173017e-06,
      "loss": 0.3248,
      "step": 24420,
      "task_loss": 0.45032960176467896
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2906193137168884,
      "epoch": 8.83,
      "learning_rate": 7.810035612173736e-06,
      "loss": 0.3679,
      "step": 24430,
      "task_loss": 0.7031286954879761
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39351212978363037,
      "epoch": 8.83,
      "learning_rate": 7.764247994092586e-06,
      "loss": 0.3464,
      "step": 24440,
      "task_loss": 0.3175731897354126
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2753757834434509,
      "epoch": 8.84,
      "learning_rate": 7.718575031025928e-06,
      "loss": 0.3265,
      "step": 24450,
      "task_loss": 0.5854820013046265
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3078010082244873,
      "epoch": 8.84,
      "learning_rate": 7.673016958478925e-06,
      "loss": 0.3592,
      "step": 24460,
      "task_loss": 0.7642509341239929
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.286521852016449,
      "epoch": 8.84,
      "learning_rate": 7.627574011364303e-06,
      "loss": 0.3358,
      "step": 24470,
      "task_loss": 0.549680233001709
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2735077142715454,
      "epoch": 8.85,
      "learning_rate": 7.582246424001182e-06,
      "loss": 0.3227,
      "step": 24480,
      "task_loss": 0.30459415912628174
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33238837122917175,
      "epoch": 8.85,
      "learning_rate": 7.5370344301138494e-06,
      "loss": 0.3252,
      "step": 24490,
      "task_loss": 0.2824766933917999
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29000821709632874,
      "epoch": 8.85,
      "learning_rate": 7.491938262830554e-06,
      "loss": 0.2903,
      "step": 24500,
      "task_loss": 0.8501275777816772
    },
    {
      "epoch": 8.85,
      "eval_exact_match": 83.41532639545885,
      "eval_f1": 89.98208448301646,
      "step": 24500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3185809850692749,
      "epoch": 8.86,
      "learning_rate": 7.446958154682279e-06,
      "loss": 0.3237,
      "step": 24510,
      "task_loss": 0.9856406450271606
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3409211039543152,
      "epoch": 8.86,
      "learning_rate": 7.402094337601612e-06,
      "loss": 0.329,
      "step": 24520,
      "task_loss": 0.4515220522880554
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22276683151721954,
      "epoch": 8.87,
      "learning_rate": 7.357347042921463e-06,
      "loss": 0.3529,
      "step": 24530,
      "task_loss": 0.22912217676639557
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4352920949459076,
      "epoch": 8.87,
      "learning_rate": 7.312716501373936e-06,
      "loss": 0.3475,
      "step": 24540,
      "task_loss": 0.6683515310287476
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2936110496520996,
      "epoch": 8.87,
      "learning_rate": 7.26820294308911e-06,
      "loss": 0.3313,
      "step": 24550,
      "task_loss": 0.36033982038497925
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3248167037963867,
      "epoch": 8.88,
      "learning_rate": 7.223806597593867e-06,
      "loss": 0.3343,
      "step": 24560,
      "task_loss": 0.5283554792404175
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35730212926864624,
      "epoch": 8.88,
      "learning_rate": 7.179527693810683e-06,
      "loss": 0.3508,
      "step": 24570,
      "task_loss": 1.0318670272827148
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5101932287216187,
      "epoch": 8.88,
      "learning_rate": 7.135366460056477e-06,
      "loss": 0.3355,
      "step": 24580,
      "task_loss": 0.7086995840072632
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3750256896018982,
      "epoch": 8.89,
      "learning_rate": 7.091323124041422e-06,
      "loss": 0.3953,
      "step": 24590,
      "task_loss": 0.8333643674850464
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2565135657787323,
      "epoch": 8.89,
      "learning_rate": 7.047397912867779e-06,
      "loss": 0.3184,
      "step": 24600,
      "task_loss": 0.31886714696884155
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3164982199668884,
      "epoch": 8.89,
      "learning_rate": 7.003591053028682e-06,
      "loss": 0.3167,
      "step": 24610,
      "task_loss": 0.41132819652557373
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40736451745033264,
      "epoch": 8.9,
      "learning_rate": 6.959902770407063e-06,
      "loss": 0.3695,
      "step": 24620,
      "task_loss": 0.8851571083068848
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27555185556411743,
      "epoch": 8.9,
      "learning_rate": 6.916333290274374e-06,
      "loss": 0.3252,
      "step": 24630,
      "task_loss": 0.5711570978164673
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2528682053089142,
      "epoch": 8.9,
      "learning_rate": 6.872882837289513e-06,
      "loss": 0.3101,
      "step": 24640,
      "task_loss": 0.5852243900299072
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3240984380245209,
      "epoch": 8.91,
      "learning_rate": 6.829551635497628e-06,
      "loss": 0.3337,
      "step": 24650,
      "task_loss": 0.9316442012786865
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23819328844547272,
      "epoch": 8.91,
      "learning_rate": 6.786339908328972e-06,
      "loss": 0.3461,
      "step": 24660,
      "task_loss": 0.5214189291000366
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45258206129074097,
      "epoch": 8.92,
      "learning_rate": 6.743247878597731e-06,
      "loss": 0.3531,
      "step": 24670,
      "task_loss": 0.5347813367843628
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30123329162597656,
      "epoch": 8.92,
      "learning_rate": 6.7002757685009e-06,
      "loss": 0.3493,
      "step": 24680,
      "task_loss": 0.7872360944747925
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2336387187242508,
      "epoch": 8.92,
      "learning_rate": 6.657423799617136e-06,
      "loss": 0.3438,
      "step": 24690,
      "task_loss": 0.49200066924095154
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35450315475463867,
      "epoch": 8.93,
      "learning_rate": 6.614692192905601e-06,
      "loss": 0.3303,
      "step": 24700,
      "task_loss": 0.516528844833374
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.300314337015152,
      "epoch": 8.93,
      "learning_rate": 6.572081168704813e-06,
      "loss": 0.3105,
      "step": 24710,
      "task_loss": 0.5251568555831909
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31300753355026245,
      "epoch": 8.93,
      "learning_rate": 6.529590946731566e-06,
      "loss": 0.3362,
      "step": 24720,
      "task_loss": 0.43648332357406616
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35132792592048645,
      "epoch": 8.94,
      "learning_rate": 6.487221746079714e-06,
      "loss": 0.3731,
      "step": 24730,
      "task_loss": 0.6553832292556763
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22395357489585876,
      "epoch": 8.94,
      "learning_rate": 6.444973785219107e-06,
      "loss": 0.3393,
      "step": 24740,
      "task_loss": 0.3021842837333679
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.315970242023468,
      "epoch": 8.94,
      "learning_rate": 6.402847281994443e-06,
      "loss": 0.2966,
      "step": 24750,
      "task_loss": 0.5357693433761597
    },
    {
      "epoch": 8.94,
      "eval_exact_match": 83.23557237464522,
      "eval_f1": 90.0436443008999,
      "step": 24750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3409693241119385,
      "epoch": 8.95,
      "learning_rate": 6.360842453624137e-06,
      "loss": 0.3445,
      "step": 24760,
      "task_loss": 0.9291391372680664
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38368433713912964,
      "epoch": 8.95,
      "learning_rate": 6.318959516699202e-06,
      "loss": 0.3723,
      "step": 24770,
      "task_loss": 0.6492195129394531
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25230035185813904,
      "epoch": 8.96,
      "learning_rate": 6.277198687182152e-06,
      "loss": 0.3391,
      "step": 24780,
      "task_loss": 0.8180399537086487
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3488203287124634,
      "epoch": 8.96,
      "learning_rate": 6.235560180405858e-06,
      "loss": 0.3587,
      "step": 24790,
      "task_loss": 0.5325171947479248
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4679383635520935,
      "epoch": 8.96,
      "learning_rate": 6.1940442110724805e-06,
      "loss": 0.3328,
      "step": 24800,
      "task_loss": 0.6240459680557251
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39593812823295593,
      "epoch": 8.97,
      "learning_rate": 6.1526509932522904e-06,
      "loss": 0.3521,
      "step": 24810,
      "task_loss": 0.3513973355293274
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39898785948753357,
      "epoch": 8.97,
      "learning_rate": 6.111380740382668e-06,
      "loss": 0.3396,
      "step": 24820,
      "task_loss": 0.6773884892463684
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33439552783966064,
      "epoch": 8.97,
      "learning_rate": 6.070233665266894e-06,
      "loss": 0.323,
      "step": 24830,
      "task_loss": 0.5181454420089722
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23859478533267975,
      "epoch": 8.98,
      "learning_rate": 6.029209980073129e-06,
      "loss": 0.3325,
      "step": 24840,
      "task_loss": 0.2097017616033554
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2746514678001404,
      "epoch": 8.98,
      "learning_rate": 5.988309896333292e-06,
      "loss": 0.3399,
      "step": 24850,
      "task_loss": 0.3311827480792999
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3430408239364624,
      "epoch": 8.98,
      "learning_rate": 5.947533624941972e-06,
      "loss": 0.337,
      "step": 24860,
      "task_loss": 0.6250514984130859
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30967992544174194,
      "epoch": 8.99,
      "learning_rate": 5.906881376155325e-06,
      "loss": 0.3168,
      "step": 24870,
      "task_loss": 0.4880766272544861
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34915846586227417,
      "epoch": 8.99,
      "learning_rate": 5.866353359590022e-06,
      "loss": 0.3472,
      "step": 24880,
      "task_loss": 0.41612374782562256
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.326454758644104,
      "epoch": 9.0,
      "learning_rate": 5.825949784222142e-06,
      "loss": 0.3226,
      "step": 24890,
      "task_loss": 0.3217955529689789
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32856497168540955,
      "epoch": 9.0,
      "learning_rate": 5.785670858386115e-06,
      "loss": 0.3293,
      "step": 24900,
      "task_loss": 0.561482310295105
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29845374822616577,
      "epoch": 9.0,
      "learning_rate": 5.745516789773605e-06,
      "loss": 0.334,
      "step": 24910,
      "task_loss": 0.6011270880699158
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26393237709999084,
      "epoch": 9.01,
      "learning_rate": 5.705487785432524e-06,
      "loss": 0.3155,
      "step": 24920,
      "task_loss": 0.40632134675979614
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32185572385787964,
      "epoch": 9.01,
      "learning_rate": 5.665584051765853e-06,
      "loss": 0.2986,
      "step": 24930,
      "task_loss": 1.1531330347061157
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2393801510334015,
      "epoch": 9.01,
      "learning_rate": 5.62580579453067e-06,
      "loss": 0.3166,
      "step": 24940,
      "task_loss": 0.6275138854980469
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2692786753177643,
      "epoch": 9.02,
      "learning_rate": 5.586153218837046e-06,
      "loss": 0.3394,
      "step": 24950,
      "task_loss": 0.508584201335907
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5188926458358765,
      "epoch": 9.02,
      "learning_rate": 5.546626529147002e-06,
      "loss": 0.3341,
      "step": 24960,
      "task_loss": 0.7853003740310669
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.287403404712677,
      "epoch": 9.02,
      "learning_rate": 5.507225929273426e-06,
      "loss": 0.2979,
      "step": 24970,
      "task_loss": 0.2948596477508545
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2628954350948334,
      "epoch": 9.03,
      "learning_rate": 5.4679516223790685e-06,
      "loss": 0.3202,
      "step": 24980,
      "task_loss": 0.34732797741889954
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21323145925998688,
      "epoch": 9.03,
      "learning_rate": 5.428803810975462e-06,
      "loss": 0.3072,
      "step": 24990,
      "task_loss": 0.38041430711746216
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3073531687259674,
      "epoch": 9.04,
      "learning_rate": 5.389782696921894e-06,
      "loss": 0.3485,
      "step": 25000,
      "task_loss": 0.7490738034248352
    },
    {
      "epoch": 9.04,
      "eval_exact_match": 83.09366130558183,
      "eval_f1": 89.96078940144051,
      "step": 25000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3114092946052551,
      "epoch": 9.04,
      "learning_rate": 5.3508884814243295e-06,
      "loss": 0.3131,
      "step": 25010,
      "task_loss": 0.709526538848877
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19256383180618286,
      "epoch": 9.04,
      "learning_rate": 5.31212136503445e-06,
      "loss": 0.3359,
      "step": 25020,
      "task_loss": 0.2658892273902893
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3078978955745697,
      "epoch": 9.05,
      "learning_rate": 5.2734815476485186e-06,
      "loss": 0.3117,
      "step": 25030,
      "task_loss": 0.3521191477775574
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24306003749370575,
      "epoch": 9.05,
      "learning_rate": 5.23496922850644e-06,
      "loss": 0.333,
      "step": 25040,
      "task_loss": 0.3750509023666382
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2680177688598633,
      "epoch": 9.05,
      "learning_rate": 5.1965846061906805e-06,
      "loss": 0.309,
      "step": 25050,
      "task_loss": 0.37960493564605713
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2974720597267151,
      "epoch": 9.06,
      "learning_rate": 5.158327878625265e-06,
      "loss": 0.313,
      "step": 25060,
      "task_loss": 0.557094156742096
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.292024165391922,
      "epoch": 9.06,
      "learning_rate": 5.120199243074735e-06,
      "loss": 0.3009,
      "step": 25070,
      "task_loss": 0.4877658486366272
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.321384072303772,
      "epoch": 9.06,
      "learning_rate": 5.082198896143158e-06,
      "loss": 0.3207,
      "step": 25080,
      "task_loss": 0.7667458653450012
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35278797149658203,
      "epoch": 9.07,
      "learning_rate": 5.044327033773109e-06,
      "loss": 0.3105,
      "step": 25090,
      "task_loss": 0.5815832614898682
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2632955312728882,
      "epoch": 9.07,
      "learning_rate": 5.0065838512446464e-06,
      "loss": 0.325,
      "step": 25100,
      "task_loss": 0.34696850180625916
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3683552145957947,
      "epoch": 9.07,
      "learning_rate": 4.9689695431742876e-06,
      "loss": 0.3271,
      "step": 25110,
      "task_loss": 0.5845701694488525
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2908926010131836,
      "epoch": 9.08,
      "learning_rate": 4.931484303514092e-06,
      "loss": 0.3412,
      "step": 25120,
      "task_loss": 0.5417877435684204
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.341631144285202,
      "epoch": 9.08,
      "learning_rate": 4.8941283255505295e-06,
      "loss": 0.3059,
      "step": 25130,
      "task_loss": 0.5817184448242188
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2900015115737915,
      "epoch": 9.09,
      "learning_rate": 4.856901801903594e-06,
      "loss": 0.3146,
      "step": 25140,
      "task_loss": 0.41760894656181335
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2963072657585144,
      "epoch": 9.09,
      "learning_rate": 4.8198049245257615e-06,
      "loss": 0.3642,
      "step": 25150,
      "task_loss": 0.4821188449859619
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.414448082447052,
      "epoch": 9.09,
      "learning_rate": 4.782837884701011e-06,
      "loss": 0.3301,
      "step": 25160,
      "task_loss": 0.5216705203056335
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24655842781066895,
      "epoch": 9.1,
      "learning_rate": 4.746000873043818e-06,
      "loss": 0.2995,
      "step": 25170,
      "task_loss": 0.814030647277832
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3046429455280304,
      "epoch": 9.1,
      "learning_rate": 4.709294079498207e-06,
      "loss": 0.3073,
      "step": 25180,
      "task_loss": 0.6938967704772949
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38174983859062195,
      "epoch": 9.1,
      "learning_rate": 4.672717693336749e-06,
      "loss": 0.3216,
      "step": 25190,
      "task_loss": 0.9253899455070496
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43978115916252136,
      "epoch": 9.11,
      "learning_rate": 4.6362719031596e-06,
      "loss": 0.3133,
      "step": 25200,
      "task_loss": 0.6062501072883606
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3401195704936981,
      "epoch": 9.11,
      "learning_rate": 4.5999568968934854e-06,
      "loss": 0.3333,
      "step": 25210,
      "task_loss": 0.42778706550598145
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30857905745506287,
      "epoch": 9.11,
      "learning_rate": 4.5637728617908196e-06,
      "loss": 0.3275,
      "step": 25220,
      "task_loss": 0.6282418966293335
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3164075016975403,
      "epoch": 9.12,
      "learning_rate": 4.527719984428635e-06,
      "loss": 0.281,
      "step": 25230,
      "task_loss": 0.4652021527290344
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36166757345199585,
      "epoch": 9.12,
      "learning_rate": 4.491798450707693e-06,
      "loss": 0.3245,
      "step": 25240,
      "task_loss": 0.38490480184555054
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2693134546279907,
      "epoch": 9.13,
      "learning_rate": 4.456008445851504e-06,
      "loss": 0.3044,
      "step": 25250,
      "task_loss": 0.4064496159553528
    },
    {
      "epoch": 9.13,
      "eval_exact_match": 83.34910122989594,
      "eval_f1": 90.04081962102016,
      "step": 25250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5317705869674683,
      "epoch": 9.13,
      "learning_rate": 4.420350154405373e-06,
      "loss": 0.3251,
      "step": 25260,
      "task_loss": 0.7308588027954102
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3306049108505249,
      "epoch": 9.13,
      "learning_rate": 4.384823760235432e-06,
      "loss": 0.3609,
      "step": 25270,
      "task_loss": 0.578725278377533
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3114280700683594,
      "epoch": 9.14,
      "learning_rate": 4.3494294465277165e-06,
      "loss": 0.3169,
      "step": 25280,
      "task_loss": 0.4894099533557892
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25917303562164307,
      "epoch": 9.14,
      "learning_rate": 4.314167395787213e-06,
      "loss": 0.2984,
      "step": 25290,
      "task_loss": 0.49306440353393555
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24381668865680695,
      "epoch": 9.14,
      "learning_rate": 4.279037789836915e-06,
      "loss": 0.3322,
      "step": 25300,
      "task_loss": 0.25567013025283813
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28523534536361694,
      "epoch": 9.15,
      "learning_rate": 4.244040809816882e-06,
      "loss": 0.3257,
      "step": 25310,
      "task_loss": 0.6707668304443359
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3397090435028076,
      "epoch": 9.15,
      "learning_rate": 4.209176636183313e-06,
      "loss": 0.3319,
      "step": 25320,
      "task_loss": 0.5605945587158203
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28864046931266785,
      "epoch": 9.15,
      "learning_rate": 4.174445448707604e-06,
      "loss": 0.3127,
      "step": 25330,
      "task_loss": 0.5202219486236572
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32487910985946655,
      "epoch": 9.16,
      "learning_rate": 4.139847426475443e-06,
      "loss": 0.3338,
      "step": 25340,
      "task_loss": 0.44852226972579956
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3215107321739197,
      "epoch": 9.16,
      "learning_rate": 4.105382747885863e-06,
      "loss": 0.3108,
      "step": 25350,
      "task_loss": 0.5720582008361816
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2655651569366455,
      "epoch": 9.17,
      "learning_rate": 4.071051590650343e-06,
      "loss": 0.3315,
      "step": 25360,
      "task_loss": 0.8129599094390869
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28251737356185913,
      "epoch": 9.17,
      "learning_rate": 4.036854131791856e-06,
      "loss": 0.2958,
      "step": 25370,
      "task_loss": 0.7780440449714661
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3596683144569397,
      "epoch": 9.17,
      "learning_rate": 4.002790547644029e-06,
      "loss": 0.3233,
      "step": 25380,
      "task_loss": 0.7711316347122192
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32312220335006714,
      "epoch": 9.18,
      "learning_rate": 3.96886101385013e-06,
      "loss": 0.3196,
      "step": 25390,
      "task_loss": 0.5299769043922424
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2755924463272095,
      "epoch": 9.18,
      "learning_rate": 3.935065705362258e-06,
      "loss": 0.3004,
      "step": 25400,
      "task_loss": 0.8743758201599121
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19421358406543732,
      "epoch": 9.18,
      "learning_rate": 3.9014047964403805e-06,
      "loss": 0.301,
      "step": 25410,
      "task_loss": 0.40134578943252563
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3234480023384094,
      "epoch": 9.19,
      "learning_rate": 3.867878460651475e-06,
      "loss": 0.319,
      "step": 25420,
      "task_loss": 0.5269187688827515
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35207459330558777,
      "epoch": 9.19,
      "learning_rate": 3.834486870868585e-06,
      "loss": 0.3288,
      "step": 25430,
      "task_loss": 0.5806119441986084
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24220982193946838,
      "epoch": 9.19,
      "learning_rate": 3.8012301992699906e-06,
      "loss": 0.3152,
      "step": 25440,
      "task_loss": 0.3459148406982422
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35971611738204956,
      "epoch": 9.2,
      "learning_rate": 3.768108617338265e-06,
      "loss": 0.3115,
      "step": 25450,
      "task_loss": 0.9299747347831726
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3165518641471863,
      "epoch": 9.2,
      "learning_rate": 3.735122295859431e-06,
      "loss": 0.2978,
      "step": 25460,
      "task_loss": 0.8229607343673706
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3041304349899292,
      "epoch": 9.2,
      "learning_rate": 3.7022714049220387e-06,
      "loss": 0.3304,
      "step": 25470,
      "task_loss": 0.6964329481124878
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2823697328567505,
      "epoch": 9.21,
      "learning_rate": 3.669556113916349e-06,
      "loss": 0.3331,
      "step": 25480,
      "task_loss": 0.44915440678596497
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3072703182697296,
      "epoch": 9.21,
      "learning_rate": 3.6369765915333876e-06,
      "loss": 0.3241,
      "step": 25490,
      "task_loss": 0.34987956285476685
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2890113294124603,
      "epoch": 9.22,
      "learning_rate": 3.6045330057641344e-06,
      "loss": 0.3328,
      "step": 25500,
      "task_loss": 0.8481899499893188
    },
    {
      "epoch": 9.22,
      "eval_exact_match": 83.29233680227058,
      "eval_f1": 90.02785814666667,
      "step": 25500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29243725538253784,
      "epoch": 9.22,
      "learning_rate": 3.5722255238986255e-06,
      "loss": 0.3104,
      "step": 25510,
      "task_loss": 0.5860704183578491
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25151023268699646,
      "epoch": 9.22,
      "learning_rate": 3.5400543125251037e-06,
      "loss": 0.2972,
      "step": 25520,
      "task_loss": 0.5240755081176758
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2432774305343628,
      "epoch": 9.23,
      "learning_rate": 3.508019537529148e-06,
      "loss": 0.3093,
      "step": 25530,
      "task_loss": 0.18137642741203308
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2751896381378174,
      "epoch": 9.23,
      "learning_rate": 3.4761213640928357e-06,
      "loss": 0.3322,
      "step": 25540,
      "task_loss": 0.48760008811950684
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.404532253742218,
      "epoch": 9.23,
      "learning_rate": 3.4443599566938687e-06,
      "loss": 0.3474,
      "step": 25550,
      "task_loss": 0.6081807017326355
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2790014445781708,
      "epoch": 9.24,
      "learning_rate": 3.412735479104756e-06,
      "loss": 0.2977,
      "step": 25560,
      "task_loss": 0.5392050743103027
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28828808665275574,
      "epoch": 9.24,
      "learning_rate": 3.3812480943919176e-06,
      "loss": 0.3254,
      "step": 25570,
      "task_loss": 0.46198827028274536
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28925055265426636,
      "epoch": 9.24,
      "learning_rate": 3.349897964914923e-06,
      "loss": 0.31,
      "step": 25580,
      "task_loss": 0.7260234355926514
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29103171825408936,
      "epoch": 9.25,
      "learning_rate": 3.318685252325564e-06,
      "loss": 0.3192,
      "step": 25590,
      "task_loss": 0.7414866089820862
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30816346406936646,
      "epoch": 9.25,
      "learning_rate": 3.287610117567086e-06,
      "loss": 0.3264,
      "step": 25600,
      "task_loss": 0.8291851282119751
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32630425691604614,
      "epoch": 9.26,
      "learning_rate": 3.256672720873335e-06,
      "loss": 0.3499,
      "step": 25610,
      "task_loss": 0.6567143201828003
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3918631970882416,
      "epoch": 9.26,
      "learning_rate": 3.2258732217679388e-06,
      "loss": 0.3278,
      "step": 25620,
      "task_loss": 0.5021631717681885
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3493354916572571,
      "epoch": 9.26,
      "learning_rate": 3.195211779063457e-06,
      "loss": 0.3106,
      "step": 25630,
      "task_loss": 0.4202045798301697
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2900541424751282,
      "epoch": 9.27,
      "learning_rate": 3.164688550860606e-06,
      "loss": 0.3223,
      "step": 25640,
      "task_loss": 0.37236881256103516
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30690574645996094,
      "epoch": 9.27,
      "learning_rate": 3.1343036945474114e-06,
      "loss": 0.3456,
      "step": 25650,
      "task_loss": 0.6118491888046265
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3027680218219757,
      "epoch": 9.27,
      "learning_rate": 3.1040573667984164e-06,
      "loss": 0.339,
      "step": 25660,
      "task_loss": 0.5271258354187012
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26531606912612915,
      "epoch": 9.28,
      "learning_rate": 3.073949723573852e-06,
      "loss": 0.3036,
      "step": 25670,
      "task_loss": 0.37352457642555237
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19538961350917816,
      "epoch": 9.28,
      "learning_rate": 3.043980920118853e-06,
      "loss": 0.3216,
      "step": 25680,
      "task_loss": 0.29823341965675354
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3124743103981018,
      "epoch": 9.28,
      "learning_rate": 3.014151110962654e-06,
      "loss": 0.3246,
      "step": 25690,
      "task_loss": 0.2712612748146057
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33055031299591064,
      "epoch": 9.29,
      "learning_rate": 2.9844604499177686e-06,
      "loss": 0.3172,
      "step": 25700,
      "task_loss": 0.5724518299102783
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.307403564453125,
      "epoch": 9.29,
      "learning_rate": 2.954909090079233e-06,
      "loss": 0.3294,
      "step": 25710,
      "task_loss": 0.3865295648574829
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2786659002304077,
      "epoch": 9.3,
      "learning_rate": 2.925497183823802e-06,
      "loss": 0.3085,
      "step": 25720,
      "task_loss": 0.5146081447601318
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30434009432792664,
      "epoch": 9.3,
      "learning_rate": 2.8962248828091555e-06,
      "loss": 0.3249,
      "step": 25730,
      "task_loss": 0.5385090112686157
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.287394642829895,
      "epoch": 9.3,
      "learning_rate": 2.867092337973105e-06,
      "loss": 0.314,
      "step": 25740,
      "task_loss": 0.7510730028152466
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2954567074775696,
      "epoch": 9.31,
      "learning_rate": 2.83809969953287e-06,
      "loss": 0.3224,
      "step": 25750,
      "task_loss": 0.4567757248878479
    },
    {
      "epoch": 9.31,
      "eval_exact_match": 83.20719016083254,
      "eval_f1": 90.01758651781414,
      "step": 25750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.370003879070282,
      "epoch": 9.31,
      "learning_rate": 2.809247116984226e-06,
      "loss": 0.3028,
      "step": 25760,
      "task_loss": 0.6199772357940674
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2925732433795929,
      "epoch": 9.31,
      "learning_rate": 2.7805347391007963e-06,
      "loss": 0.3122,
      "step": 25770,
      "task_loss": 0.4094921350479126
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3006725311279297,
      "epoch": 9.32,
      "learning_rate": 2.7519627139332594e-06,
      "loss": 0.336,
      "step": 25780,
      "task_loss": 0.9209330081939697
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3157416582107544,
      "epoch": 9.32,
      "learning_rate": 2.723531188808589e-06,
      "loss": 0.3277,
      "step": 25790,
      "task_loss": 0.8077236413955688
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28249573707580566,
      "epoch": 9.32,
      "learning_rate": 2.695240310329279e-06,
      "loss": 0.3234,
      "step": 25800,
      "task_loss": 0.369956374168396
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2779986262321472,
      "epoch": 9.33,
      "learning_rate": 2.6670902243726127e-06,
      "loss": 0.3028,
      "step": 25810,
      "task_loss": 0.7005301713943481
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26586776971817017,
      "epoch": 9.33,
      "learning_rate": 2.6390810760899043e-06,
      "loss": 0.3389,
      "step": 25820,
      "task_loss": 0.5821472406387329
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2681348919868469,
      "epoch": 9.34,
      "learning_rate": 2.6112130099057465e-06,
      "loss": 0.3343,
      "step": 25830,
      "task_loss": 0.43781211972236633
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28912442922592163,
      "epoch": 9.34,
      "learning_rate": 2.5834861695172394e-06,
      "loss": 0.2981,
      "step": 25840,
      "task_loss": 0.5437923669815063
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2632666528224945,
      "epoch": 9.34,
      "learning_rate": 2.555900697893321e-06,
      "loss": 0.3046,
      "step": 25850,
      "task_loss": 0.8168158531188965
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4091622233390808,
      "epoch": 9.35,
      "learning_rate": 2.52845673727394e-06,
      "loss": 0.3188,
      "step": 25860,
      "task_loss": 0.5111245512962341
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26104307174682617,
      "epoch": 9.35,
      "learning_rate": 2.501154429169393e-06,
      "loss": 0.3321,
      "step": 25870,
      "task_loss": 0.3140479326248169
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3605012893676758,
      "epoch": 9.35,
      "learning_rate": 2.4739939143595626e-06,
      "loss": 0.3232,
      "step": 25880,
      "task_loss": 0.478623628616333
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38676559925079346,
      "epoch": 9.36,
      "learning_rate": 2.4469753328932e-06,
      "loss": 0.3311,
      "step": 25890,
      "task_loss": 0.5500181317329407
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26764678955078125,
      "epoch": 9.36,
      "learning_rate": 2.4200988240871837e-06,
      "loss": 0.352,
      "step": 25900,
      "task_loss": 0.387978196144104
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2940652370452881,
      "epoch": 9.36,
      "learning_rate": 2.3933645265258407e-06,
      "loss": 0.3374,
      "step": 25910,
      "task_loss": 0.5337430238723755
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23789635300636292,
      "epoch": 9.37,
      "learning_rate": 2.366772578060199e-06,
      "loss": 0.3429,
      "step": 25920,
      "task_loss": 0.5305555462837219
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2854630947113037,
      "epoch": 9.37,
      "learning_rate": 2.3403231158072893e-06,
      "loss": 0.341,
      "step": 25930,
      "task_loss": 0.6268982887268066
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2674894630908966,
      "epoch": 9.37,
      "learning_rate": 2.3140162761494166e-06,
      "loss": 0.3222,
      "step": 25940,
      "task_loss": 0.48702147603034973
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.302157998085022,
      "epoch": 9.38,
      "learning_rate": 2.287852194733515e-06,
      "loss": 0.3122,
      "step": 25950,
      "task_loss": 0.3152402639389038
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43852946162223816,
      "epoch": 9.38,
      "learning_rate": 2.2618310064703707e-06,
      "loss": 0.3381,
      "step": 25960,
      "task_loss": 0.6149742603302002
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27004188299179077,
      "epoch": 9.39,
      "learning_rate": 2.235952845533985e-06,
      "loss": 0.3265,
      "step": 25970,
      "task_loss": 0.5028146505355835
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.507034957408905,
      "epoch": 9.39,
      "learning_rate": 2.21021784536086e-06,
      "loss": 0.317,
      "step": 25980,
      "task_loss": 0.6118849515914917
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24345892667770386,
      "epoch": 9.39,
      "learning_rate": 2.184626138649315e-06,
      "loss": 0.3307,
      "step": 25990,
      "task_loss": 0.4160289764404297
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30164164304733276,
      "epoch": 9.4,
      "learning_rate": 2.1591778573587795e-06,
      "loss": 0.2989,
      "step": 26000,
      "task_loss": 0.6134577989578247
    },
    {
      "epoch": 9.4,
      "eval_exact_match": 83.24503311258277,
      "eval_f1": 90.07899218127025,
      "step": 26000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3065416216850281,
      "epoch": 9.4,
      "learning_rate": 2.133873132709162e-06,
      "loss": 0.2929,
      "step": 26010,
      "task_loss": 0.5050661563873291
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20736584067344666,
      "epoch": 9.4,
      "learning_rate": 2.1087120951801298e-06,
      "loss": 0.3064,
      "step": 26020,
      "task_loss": 0.6142216920852661
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30738502740859985,
      "epoch": 9.41,
      "learning_rate": 2.0836948745104554e-06,
      "loss": 0.3022,
      "step": 26030,
      "task_loss": 0.42188823223114014
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34054452180862427,
      "epoch": 9.41,
      "learning_rate": 2.058821599697326e-06,
      "loss": 0.3338,
      "step": 26040,
      "task_loss": 0.4355012774467468
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.287036657333374,
      "epoch": 9.41,
      "learning_rate": 2.0340923989957337e-06,
      "loss": 0.3057,
      "step": 26050,
      "task_loss": 0.6002825498580933
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34066513180732727,
      "epoch": 9.42,
      "learning_rate": 2.0095073999177405e-06,
      "loss": 0.3077,
      "step": 26060,
      "task_loss": 0.42962759733200073
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26722410321235657,
      "epoch": 9.42,
      "learning_rate": 1.985066729231865e-06,
      "loss": 0.3017,
      "step": 26070,
      "task_loss": 0.2320828139781952
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2443457543849945,
      "epoch": 9.43,
      "learning_rate": 1.960770512962431e-06,
      "loss": 0.2898,
      "step": 26080,
      "task_loss": 0.3011275827884674
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2698020935058594,
      "epoch": 9.43,
      "learning_rate": 1.9366188763889026e-06,
      "loss": 0.3416,
      "step": 26090,
      "task_loss": 0.8731992244720459
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3889712989330292,
      "epoch": 9.43,
      "learning_rate": 1.9126119440452293e-06,
      "loss": 0.3149,
      "step": 26100,
      "task_loss": 0.667431116104126
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41749489307403564,
      "epoch": 9.44,
      "learning_rate": 1.8887498397192316e-06,
      "loss": 0.3004,
      "step": 26110,
      "task_loss": 1.0217292308807373
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2434772253036499,
      "epoch": 9.44,
      "learning_rate": 1.8650326864519428e-06,
      "loss": 0.3241,
      "step": 26120,
      "task_loss": 0.27343904972076416
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2682727873325348,
      "epoch": 9.44,
      "learning_rate": 1.8414606065369855e-06,
      "loss": 0.3026,
      "step": 26130,
      "task_loss": 0.7200096845626831
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30232781171798706,
      "epoch": 9.45,
      "learning_rate": 1.818033721519916e-06,
      "loss": 0.3156,
      "step": 26140,
      "task_loss": 0.41904714703559875
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31530797481536865,
      "epoch": 9.45,
      "learning_rate": 1.7970737664963832e-06,
      "loss": 0.302,
      "step": 26150,
      "task_loss": 0.4437169134616852
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28844863176345825,
      "epoch": 9.45,
      "learning_rate": 1.7739230839614962e-06,
      "loss": 0.2845,
      "step": 26160,
      "task_loss": 0.5498650074005127
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38394349813461304,
      "epoch": 9.46,
      "learning_rate": 1.7509179445706858e-06,
      "loss": 0.321,
      "step": 26170,
      "task_loss": 0.843754768371582
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29749709367752075,
      "epoch": 9.46,
      "learning_rate": 1.7280584669461808e-06,
      "loss": 0.3067,
      "step": 26180,
      "task_loss": 0.2624616026878357
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2989083528518677,
      "epoch": 9.47,
      "learning_rate": 1.7053447689591473e-06,
      "loss": 0.3149,
      "step": 26190,
      "task_loss": 0.6814029812812805
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24795949459075928,
      "epoch": 9.47,
      "learning_rate": 1.6827769677290294e-06,
      "loss": 0.2966,
      "step": 26200,
      "task_loss": 0.6931198239326477
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31588611006736755,
      "epoch": 9.47,
      "learning_rate": 1.6603551796230232e-06,
      "loss": 0.3074,
      "step": 26210,
      "task_loss": 0.7078069448471069
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2802581191062927,
      "epoch": 9.48,
      "learning_rate": 1.6380795202553866e-06,
      "loss": 0.312,
      "step": 26220,
      "task_loss": 0.7879334688186646
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3153989911079407,
      "epoch": 9.48,
      "learning_rate": 1.615950104486924e-06,
      "loss": 0.3211,
      "step": 26230,
      "task_loss": 0.593709409236908
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2652713358402252,
      "epoch": 9.48,
      "learning_rate": 1.5939670464243362e-06,
      "loss": 0.3022,
      "step": 26240,
      "task_loss": 0.5030221939086914
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2979765236377716,
      "epoch": 9.49,
      "learning_rate": 1.572130459419674e-06,
      "loss": 0.3224,
      "step": 26250,
      "task_loss": 0.3495621085166931
    },
    {
      "epoch": 9.49,
      "eval_exact_match": 83.15988647114474,
      "eval_f1": 90.06487572380685,
      "step": 26250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30114901065826416,
      "epoch": 9.49,
      "learning_rate": 1.5504404560697093e-06,
      "loss": 0.3202,
      "step": 26260,
      "task_loss": 0.4952106177806854
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30627989768981934,
      "epoch": 9.49,
      "learning_rate": 1.5288971482153957e-06,
      "loss": 0.324,
      "step": 26270,
      "task_loss": 0.6019161939620972
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27386802434921265,
      "epoch": 9.5,
      "learning_rate": 1.5075006469412778e-06,
      "loss": 0.3126,
      "step": 26280,
      "task_loss": 0.6094711422920227
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3032553195953369,
      "epoch": 9.5,
      "learning_rate": 1.486251062574916e-06,
      "loss": 0.2942,
      "step": 26290,
      "task_loss": 0.475771963596344
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2948029637336731,
      "epoch": 9.5,
      "learning_rate": 1.4651485046862933e-06,
      "loss": 0.3229,
      "step": 26300,
      "task_loss": 0.44787514209747314
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3109568953514099,
      "epoch": 9.51,
      "learning_rate": 1.4441930820873195e-06,
      "loss": 0.3074,
      "step": 26310,
      "task_loss": 0.4297507107257843
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2993125319480896,
      "epoch": 9.51,
      "learning_rate": 1.4233849028311808e-06,
      "loss": 0.3203,
      "step": 26320,
      "task_loss": 0.41013583540916443
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3109281659126282,
      "epoch": 9.52,
      "learning_rate": 1.4027240742118542e-06,
      "loss": 0.31,
      "step": 26330,
      "task_loss": 0.48224520683288574
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.320933073759079,
      "epoch": 9.52,
      "learning_rate": 1.3822107027635178e-06,
      "loss": 0.3106,
      "step": 26340,
      "task_loss": 0.3480777144432068
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40795719623565674,
      "epoch": 9.52,
      "learning_rate": 1.3618448942600182e-06,
      "loss": 0.3197,
      "step": 26350,
      "task_loss": 0.6005345582962036
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2971874475479126,
      "epoch": 9.53,
      "learning_rate": 1.3416267537143035e-06,
      "loss": 0.3062,
      "step": 26360,
      "task_loss": 0.45181214809417725
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3023059368133545,
      "epoch": 9.53,
      "learning_rate": 1.3215563853779112e-06,
      "loss": 0.2991,
      "step": 26370,
      "task_loss": 0.5861034989356995
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34587642550468445,
      "epoch": 9.53,
      "learning_rate": 1.3016338927404047e-06,
      "loss": 0.3595,
      "step": 26380,
      "task_loss": 0.5939334034919739
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28393590450286865,
      "epoch": 9.54,
      "learning_rate": 1.2818593785288645e-06,
      "loss": 0.3235,
      "step": 26390,
      "task_loss": 0.6647424697875977
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35412973165512085,
      "epoch": 9.54,
      "learning_rate": 1.262232944707321e-06,
      "loss": 0.3574,
      "step": 26400,
      "task_loss": 0.7059704065322876
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22645823657512665,
      "epoch": 9.54,
      "learning_rate": 1.2427546924762823e-06,
      "loss": 0.2852,
      "step": 26410,
      "task_loss": 0.5531166791915894
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2105254828929901,
      "epoch": 9.55,
      "learning_rate": 1.2234247222721573e-06,
      "loss": 0.2749,
      "step": 26420,
      "task_loss": 0.22269278764724731
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39202356338500977,
      "epoch": 9.55,
      "learning_rate": 1.2042431337667704e-06,
      "loss": 0.3171,
      "step": 26430,
      "task_loss": 0.6872595548629761
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2564923167228699,
      "epoch": 9.56,
      "learning_rate": 1.1852100258668507e-06,
      "loss": 0.3454,
      "step": 26440,
      "task_loss": 0.50974440574646
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3733943998813629,
      "epoch": 9.56,
      "learning_rate": 1.1663254967134973e-06,
      "loss": 0.3358,
      "step": 26450,
      "task_loss": 0.4759913980960846
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34948068857192993,
      "epoch": 9.56,
      "learning_rate": 1.1475896436816947e-06,
      "loss": 0.3324,
      "step": 26460,
      "task_loss": 0.8849142789840698
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4483783543109894,
      "epoch": 9.57,
      "learning_rate": 1.1290025633797973e-06,
      "loss": 0.3227,
      "step": 26470,
      "task_loss": 1.083753228187561
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3329492211341858,
      "epoch": 9.57,
      "learning_rate": 1.1105643516490438e-06,
      "loss": 0.2947,
      "step": 26480,
      "task_loss": 0.70804762840271
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3900536298751831,
      "epoch": 9.57,
      "learning_rate": 1.0922751035630595e-06,
      "loss": 0.3534,
      "step": 26490,
      "task_loss": 0.8057717084884644
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38479679822921753,
      "epoch": 9.58,
      "learning_rate": 1.0741349134273448e-06,
      "loss": 0.324,
      "step": 26500,
      "task_loss": 0.5944019556045532
    },
    {
      "epoch": 9.58,
      "eval_exact_match": 83.19772942289498,
      "eval_f1": 89.94716840298422,
      "step": 26500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31899452209472656,
      "epoch": 9.58,
      "learning_rate": 1.0561438747788377e-06,
      "loss": 0.3297,
      "step": 26510,
      "task_loss": 0.6771483421325684
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2978406250476837,
      "epoch": 9.58,
      "learning_rate": 1.0383020803853682e-06,
      "loss": 0.3061,
      "step": 26520,
      "task_loss": 0.3691001534461975
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20832061767578125,
      "epoch": 9.59,
      "learning_rate": 1.0206096222452321e-06,
      "loss": 0.3142,
      "step": 26530,
      "task_loss": 0.4117594361305237
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3379862904548645,
      "epoch": 9.59,
      "learning_rate": 1.0030665915866944e-06,
      "loss": 0.3164,
      "step": 26540,
      "task_loss": 0.4368290305137634
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24347561597824097,
      "epoch": 9.6,
      "learning_rate": 9.856730788675228e-07,
      "loss": 0.3003,
      "step": 26550,
      "task_loss": 0.31799739599227905
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2570602595806122,
      "epoch": 9.6,
      "learning_rate": 9.68429173774512e-07,
      "loss": 0.3272,
      "step": 26560,
      "task_loss": 0.6868414282798767
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29512378573417664,
      "epoch": 9.6,
      "learning_rate": 9.513349652230407e-07,
      "loss": 0.3032,
      "step": 26570,
      "task_loss": 0.7735044956207275
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.216871976852417,
      "epoch": 9.61,
      "learning_rate": 9.343905413565878e-07,
      "loss": 0.307,
      "step": 26580,
      "task_loss": 0.2056950479745865
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2929142713546753,
      "epoch": 9.61,
      "learning_rate": 9.175959895463138e-07,
      "loss": 0.3184,
      "step": 26590,
      "task_loss": 0.566573977470398
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3073238134384155,
      "epoch": 9.61,
      "learning_rate": 9.009513963905602e-07,
      "loss": 0.3183,
      "step": 26600,
      "task_loss": 0.6305901408195496
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2523428201675415,
      "epoch": 9.62,
      "learning_rate": 8.844568477144644e-07,
      "loss": 0.3194,
      "step": 26610,
      "task_loss": 0.4756616950035095
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40382206439971924,
      "epoch": 9.62,
      "learning_rate": 8.681124285694486e-07,
      "loss": 0.3168,
      "step": 26620,
      "task_loss": 0.5824015736579895
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25531330704689026,
      "epoch": 9.62,
      "learning_rate": 8.519182232328415e-07,
      "loss": 0.3202,
      "step": 26630,
      "task_loss": 0.5715607404708862
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3254801630973816,
      "epoch": 9.63,
      "learning_rate": 8.358743152074111e-07,
      "loss": 0.3626,
      "step": 26640,
      "task_loss": 0.4549102187156677
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25074678659439087,
      "epoch": 9.63,
      "learning_rate": 8.199807872209452e-07,
      "loss": 0.3284,
      "step": 26650,
      "task_loss": 0.33278608322143555
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23102590441703796,
      "epoch": 9.63,
      "learning_rate": 8.042377212258123e-07,
      "loss": 0.2935,
      "step": 26660,
      "task_loss": 0.5513738989830017
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47359585762023926,
      "epoch": 9.64,
      "learning_rate": 7.886451983985576e-07,
      "loss": 0.308,
      "step": 26670,
      "task_loss": 0.6858689785003662
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29826098680496216,
      "epoch": 9.64,
      "learning_rate": 7.73203299139471e-07,
      "loss": 0.3315,
      "step": 26680,
      "task_loss": 0.3836401104927063
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3404310345649719,
      "epoch": 9.65,
      "learning_rate": 7.579121030721837e-07,
      "loss": 0.3221,
      "step": 26690,
      "task_loss": 0.625718355178833
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27862346172332764,
      "epoch": 9.65,
      "learning_rate": 7.427716890432346e-07,
      "loss": 0.2999,
      "step": 26700,
      "task_loss": 0.26387932896614075
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30959299206733704,
      "epoch": 9.65,
      "learning_rate": 7.277821351216984e-07,
      "loss": 0.3153,
      "step": 26710,
      "task_loss": 0.7208259105682373
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3370169401168823,
      "epoch": 9.66,
      "learning_rate": 7.129435185987487e-07,
      "loss": 0.2994,
      "step": 26720,
      "task_loss": 0.43289870023727417
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36385589838027954,
      "epoch": 9.66,
      "learning_rate": 6.982559159872881e-07,
      "loss": 0.328,
      "step": 26730,
      "task_loss": 0.5257211923599243
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2223806381225586,
      "epoch": 9.66,
      "learning_rate": 6.837194030215288e-07,
      "loss": 0.3035,
      "step": 26740,
      "task_loss": 0.39960020780563354
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3810596168041229,
      "epoch": 9.67,
      "learning_rate": 6.693340546566263e-07,
      "loss": 0.3517,
      "step": 26750,
      "task_loss": 0.6381791830062866
    },
    {
      "epoch": 9.67,
      "eval_exact_match": 83.3112582781457,
      "eval_f1": 90.07123919343461,
      "step": 26750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2689059376716614,
      "epoch": 9.67,
      "learning_rate": 6.550999450682693e-07,
      "loss": 0.2918,
      "step": 26760,
      "task_loss": 0.48646509647369385
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22755245864391327,
      "epoch": 9.67,
      "learning_rate": 6.410171476523141e-07,
      "loss": 0.2787,
      "step": 26770,
      "task_loss": 0.27000099420547485
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2955330014228821,
      "epoch": 9.68,
      "learning_rate": 6.270857350243974e-07,
      "loss": 0.3318,
      "step": 26780,
      "task_loss": 0.44846802949905396
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3734288513660431,
      "epoch": 9.68,
      "learning_rate": 6.133057790195773e-07,
      "loss": 0.3282,
      "step": 26790,
      "task_loss": 0.5023435950279236
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3733587861061096,
      "epoch": 9.69,
      "learning_rate": 5.996773506919262e-07,
      "loss": 0.3225,
      "step": 26800,
      "task_loss": 0.8738093972206116
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3161894977092743,
      "epoch": 9.69,
      "learning_rate": 5.862005203142151e-07,
      "loss": 0.3023,
      "step": 26810,
      "task_loss": 0.5742534399032593
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2755066156387329,
      "epoch": 9.69,
      "learning_rate": 5.728753573775069e-07,
      "loss": 0.2993,
      "step": 26820,
      "task_loss": 0.5461592078208923
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28466445207595825,
      "epoch": 9.7,
      "learning_rate": 5.597019305908235e-07,
      "loss": 0.283,
      "step": 26830,
      "task_loss": 0.4959646463394165
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2563675343990326,
      "epoch": 9.7,
      "learning_rate": 5.466803078807859e-07,
      "loss": 0.3423,
      "step": 26840,
      "task_loss": 0.4334827959537506
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26834434270858765,
      "epoch": 9.7,
      "learning_rate": 5.33810556391261e-07,
      "loss": 0.2885,
      "step": 26850,
      "task_loss": 0.8743736147880554
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2733617424964905,
      "epoch": 9.71,
      "learning_rate": 5.210927424830092e-07,
      "loss": 0.2869,
      "step": 26860,
      "task_loss": 0.5635610222816467
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3465721309185028,
      "epoch": 9.71,
      "learning_rate": 5.085269317333574e-07,
      "loss": 0.2863,
      "step": 26870,
      "task_loss": 0.5172914266586304
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24393369257450104,
      "epoch": 9.71,
      "learning_rate": 4.961131889358528e-07,
      "loss": 0.2913,
      "step": 26880,
      "task_loss": 0.544761598110199
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.304984450340271,
      "epoch": 9.72,
      "learning_rate": 4.838515780999264e-07,
      "loss": 0.3326,
      "step": 26890,
      "task_loss": 0.49490851163864136
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31869620084762573,
      "epoch": 9.72,
      "learning_rate": 4.717421624505669e-07,
      "loss": 0.308,
      "step": 26900,
      "task_loss": 0.4404585361480713
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26529282331466675,
      "epoch": 9.73,
      "learning_rate": 4.597850044279972e-07,
      "loss": 0.3086,
      "step": 26910,
      "task_loss": 0.6393947601318359
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3640422821044922,
      "epoch": 9.73,
      "learning_rate": 4.4798016568733837e-07,
      "loss": 0.3077,
      "step": 26920,
      "task_loss": 0.7868601083755493
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2870105504989624,
      "epoch": 9.73,
      "learning_rate": 4.3632770709831293e-07,
      "loss": 0.3128,
      "step": 26930,
      "task_loss": 0.6231682300567627
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2648237347602844,
      "epoch": 9.74,
      "learning_rate": 4.248276887449154e-07,
      "loss": 0.2981,
      "step": 26940,
      "task_loss": 0.6860414147377014
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30803900957107544,
      "epoch": 9.74,
      "learning_rate": 4.1348016992510895e-07,
      "loss": 0.3192,
      "step": 26950,
      "task_loss": 0.6218051910400391
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30648428201675415,
      "epoch": 9.74,
      "learning_rate": 4.0228520915050915e-07,
      "loss": 0.319,
      "step": 26960,
      "task_loss": 0.6169790625572205
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29523447155952454,
      "epoch": 9.75,
      "learning_rate": 3.912428641461041e-07,
      "loss": 0.3132,
      "step": 26970,
      "task_loss": 0.9747434854507446
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23645155131816864,
      "epoch": 9.75,
      "learning_rate": 3.8035319184993813e-07,
      "loss": 0.2862,
      "step": 26980,
      "task_loss": 0.47767388820648193
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30601510405540466,
      "epoch": 9.75,
      "learning_rate": 3.6961624841282516e-07,
      "loss": 0.3058,
      "step": 26990,
      "task_loss": 0.3921341598033905
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3483054041862488,
      "epoch": 9.76,
      "learning_rate": 3.590320891980492e-07,
      "loss": 0.3432,
      "step": 27000,
      "task_loss": 0.7122894525527954
    },
    {
      "epoch": 9.76,
      "eval_exact_match": 83.19772942289498,
      "eval_f1": 89.93300652023898,
      "step": 27000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32802069187164307,
      "epoch": 9.76,
      "learning_rate": 3.4860076878110103e-07,
      "loss": 0.2999,
      "step": 27010,
      "task_loss": 0.7228002548217773
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.286615788936615,
      "epoch": 9.77,
      "learning_rate": 3.383223409493719e-07,
      "loss": 0.3082,
      "step": 27020,
      "task_loss": 0.4433107376098633
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3046417832374573,
      "epoch": 9.77,
      "learning_rate": 3.281968587018902e-07,
      "loss": 0.3324,
      "step": 27030,
      "task_loss": 0.54556804895401
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.271384596824646,
      "epoch": 9.77,
      "learning_rate": 3.1822437424905536e-07,
      "loss": 0.3071,
      "step": 27040,
      "task_loss": 0.44324588775634766
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28623712062835693,
      "epoch": 9.78,
      "learning_rate": 3.084049390123478e-07,
      "loss": 0.3245,
      "step": 27050,
      "task_loss": 0.7106075882911682
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31551676988601685,
      "epoch": 9.78,
      "learning_rate": 2.9873860362407244e-07,
      "loss": 0.3025,
      "step": 27060,
      "task_loss": 0.47528547048568726
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28199514746665955,
      "epoch": 9.78,
      "learning_rate": 2.892254179271059e-07,
      "loss": 0.3273,
      "step": 27070,
      "task_loss": 0.40891680121421814
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3044975996017456,
      "epoch": 9.79,
      "learning_rate": 2.798654309746396e-07,
      "loss": 0.2912,
      "step": 27080,
      "task_loss": 0.4443657398223877
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25702059268951416,
      "epoch": 9.79,
      "learning_rate": 2.706586910299069e-07,
      "loss": 0.2979,
      "step": 27090,
      "task_loss": 0.4479978084564209
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2585090100765228,
      "epoch": 9.79,
      "learning_rate": 2.616052455659568e-07,
      "loss": 0.3239,
      "step": 27100,
      "task_loss": 0.38295847177505493
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3074435293674469,
      "epoch": 9.8,
      "learning_rate": 2.5270514126540025e-07,
      "loss": 0.2986,
      "step": 27110,
      "task_loss": 0.5449491739273071
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3268190026283264,
      "epoch": 9.8,
      "learning_rate": 2.4395842402016756e-07,
      "loss": 0.3059,
      "step": 27120,
      "task_loss": 0.49111929535865784
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2732863426208496,
      "epoch": 9.8,
      "learning_rate": 2.3536513893127166e-07,
      "loss": 0.302,
      "step": 27130,
      "task_loss": 0.46516484022140503
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3503320813179016,
      "epoch": 9.81,
      "learning_rate": 2.2692533030857832e-07,
      "loss": 0.3421,
      "step": 27140,
      "task_loss": 0.8032338619232178
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4571889638900757,
      "epoch": 9.81,
      "learning_rate": 2.1863904167058634e-07,
      "loss": 0.3358,
      "step": 27150,
      "task_loss": 0.7194252014160156
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28522342443466187,
      "epoch": 9.82,
      "learning_rate": 2.1050631574418112e-07,
      "loss": 0.3302,
      "step": 27160,
      "task_loss": 0.594832181930542
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2757059335708618,
      "epoch": 9.82,
      "learning_rate": 2.0252719446443135e-07,
      "loss": 0.3168,
      "step": 27170,
      "task_loss": 0.7021136283874512
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35544660687446594,
      "epoch": 9.82,
      "learning_rate": 1.9470171897437273e-07,
      "loss": 0.2886,
      "step": 27180,
      "task_loss": 0.5380356907844543
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26072177290916443,
      "epoch": 9.83,
      "learning_rate": 1.8702992962478792e-07,
      "loss": 0.3091,
      "step": 27190,
      "task_loss": 0.5083783864974976
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2878473997116089,
      "epoch": 9.83,
      "learning_rate": 1.7951186597399693e-07,
      "loss": 0.3031,
      "step": 27200,
      "task_loss": 0.4856301248073578
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3149654269218445,
      "epoch": 9.83,
      "learning_rate": 1.7214756678767042e-07,
      "loss": 0.3054,
      "step": 27210,
      "task_loss": 0.7618269324302673
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.292486310005188,
      "epoch": 9.84,
      "learning_rate": 1.649370700386099e-07,
      "loss": 0.2823,
      "step": 27220,
      "task_loss": 0.2056034803390503
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30289211869239807,
      "epoch": 9.84,
      "learning_rate": 1.578804129065614e-07,
      "loss": 0.3203,
      "step": 27230,
      "task_loss": 0.478914737701416
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.248458594083786,
      "epoch": 9.84,
      "learning_rate": 1.5097763177802205e-07,
      "loss": 0.3146,
      "step": 27240,
      "task_loss": 0.2979784309864044
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34328991174697876,
      "epoch": 9.85,
      "learning_rate": 1.4422876224605365e-07,
      "loss": 0.335,
      "step": 27250,
      "task_loss": 0.7470317482948303
    },
    {
      "epoch": 9.85,
      "eval_exact_match": 83.09366130558183,
      "eval_f1": 89.97577505817762,
      "step": 27250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3134426176548004,
      "epoch": 9.85,
      "learning_rate": 1.3763383911009287e-07,
      "loss": 0.3113,
      "step": 27260,
      "task_loss": 0.5885034799575806
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3346371352672577,
      "epoch": 9.86,
      "learning_rate": 1.3119289637578135e-07,
      "loss": 0.3353,
      "step": 27270,
      "task_loss": 0.4166927635669708
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3193387985229492,
      "epoch": 9.86,
      "learning_rate": 1.249059672547892e-07,
      "loss": 0.3067,
      "step": 27280,
      "task_loss": 0.41660696268081665
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2841346859931946,
      "epoch": 9.86,
      "learning_rate": 1.1877308416463506e-07,
      "loss": 0.2894,
      "step": 27290,
      "task_loss": 0.28571581840515137
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2764800786972046,
      "epoch": 9.87,
      "learning_rate": 1.1279427872852965e-07,
      "loss": 0.3167,
      "step": 27300,
      "task_loss": 0.3838335871696472
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3414124846458435,
      "epoch": 9.87,
      "learning_rate": 1.0696958177519922e-07,
      "loss": 0.3141,
      "step": 27310,
      "task_loss": 0.6372711658477783
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23128125071525574,
      "epoch": 9.87,
      "learning_rate": 1.0129902333874563e-07,
      "loss": 0.2906,
      "step": 27320,
      "task_loss": 0.5092746019363403
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4284539818763733,
      "epoch": 9.88,
      "learning_rate": 9.578263265846655e-08,
      "loss": 0.328,
      "step": 27330,
      "task_loss": 0.3266478180885315
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2824637293815613,
      "epoch": 9.88,
      "learning_rate": 9.042043817873547e-08,
      "loss": 0.3189,
      "step": 27340,
      "task_loss": 0.4688670337200165
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40793055295944214,
      "epoch": 9.88,
      "learning_rate": 8.521246754882528e-08,
      "loss": 0.3287,
      "step": 27350,
      "task_loss": 0.45810216665267944
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24485069513320923,
      "epoch": 9.89,
      "learning_rate": 8.01587476227883e-08,
      "loss": 0.3213,
      "step": 27360,
      "task_loss": 0.4046506881713867
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2961767911911011,
      "epoch": 9.89,
      "learning_rate": 7.525930445929974e-08,
      "loss": 0.3041,
      "step": 27370,
      "task_loss": 1.140902042388916
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2372909039258957,
      "epoch": 9.9,
      "learning_rate": 7.051416332153781e-08,
      "loss": 0.2999,
      "step": 27380,
      "task_loss": 0.7308323979377747
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34356093406677246,
      "epoch": 9.9,
      "learning_rate": 6.592334867704719e-08,
      "loss": 0.3112,
      "step": 27390,
      "task_loss": 0.5305260419845581
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23644784092903137,
      "epoch": 9.9,
      "learning_rate": 6.148688419760906e-08,
      "loss": 0.314,
      "step": 27400,
      "task_loss": 0.43768665194511414
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29776376485824585,
      "epoch": 9.91,
      "learning_rate": 5.7204792759127936e-08,
      "loss": 0.3048,
      "step": 27410,
      "task_loss": 0.7872896194458008
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3038090467453003,
      "epoch": 9.91,
      "learning_rate": 5.307709644150505e-08,
      "loss": 0.2959,
      "step": 27420,
      "task_loss": 0.8730282783508301
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3143189549446106,
      "epoch": 9.91,
      "learning_rate": 4.910381652853513e-08,
      "loss": 0.3617,
      "step": 27430,
      "task_loss": 0.34668901562690735
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2636554539203644,
      "epoch": 9.92,
      "learning_rate": 4.528497350777983e-08,
      "loss": 0.3384,
      "step": 27440,
      "task_loss": 0.6265454292297363
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32342687249183655,
      "epoch": 9.92,
      "learning_rate": 4.162058707048444e-08,
      "loss": 0.3056,
      "step": 27450,
      "task_loss": 0.8568477630615234
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3245423436164856,
      "epoch": 9.92,
      "learning_rate": 3.8110676111451357e-08,
      "loss": 0.3082,
      "step": 27460,
      "task_loss": 0.5641300678253174
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25845956802368164,
      "epoch": 9.93,
      "learning_rate": 3.4755258728963455e-08,
      "loss": 0.2974,
      "step": 27470,
      "task_loss": 0.5490345358848572
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2910792827606201,
      "epoch": 9.93,
      "learning_rate": 3.155435222468417e-08,
      "loss": 0.3161,
      "step": 27480,
      "task_loss": 0.5065259337425232
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20717984437942505,
      "epoch": 9.93,
      "learning_rate": 2.8507973103560903e-08,
      "loss": 0.3401,
      "step": 27490,
      "task_loss": 0.6428301334381104
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27313390374183655,
      "epoch": 9.94,
      "learning_rate": 2.5616137073748436e-08,
      "loss": 0.3013,
      "step": 27500,
      "task_loss": 0.7006524801254272
    },
    {
      "epoch": 9.94,
      "eval_exact_match": 83.30179754020814,
      "eval_f1": 90.00136438136511,
      "step": 27500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4248490631580353,
      "epoch": 9.94,
      "learning_rate": 2.2878859046525648e-08,
      "loss": 0.3218,
      "step": 27510,
      "task_loss": 0.45274466276168823
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2887372672557831,
      "epoch": 9.95,
      "learning_rate": 2.029615313622224e-08,
      "loss": 0.3149,
      "step": 27520,
      "task_loss": 0.3250759541988373
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3086904287338257,
      "epoch": 9.95,
      "learning_rate": 1.786803266013548e-08,
      "loss": 0.3138,
      "step": 27530,
      "task_loss": 0.5649579167366028
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27589845657348633,
      "epoch": 9.95,
      "learning_rate": 1.559451013847024e-08,
      "loss": 0.3084,
      "step": 27540,
      "task_loss": 0.6957834362983704
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2580502927303314,
      "epoch": 9.96,
      "learning_rate": 1.3475597294275722e-08,
      "loss": 0.2937,
      "step": 27550,
      "task_loss": 0.3605492115020752
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3019343316555023,
      "epoch": 9.96,
      "learning_rate": 1.1511305053375498e-08,
      "loss": 0.3341,
      "step": 27560,
      "task_loss": 0.6453779935836792
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22303266823291779,
      "epoch": 9.96,
      "learning_rate": 9.701643544327565e-09,
      "loss": 0.2955,
      "step": 27570,
      "task_loss": 0.48411405086517334
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29931584000587463,
      "epoch": 9.97,
      "learning_rate": 8.046622098347723e-09,
      "loss": 0.3183,
      "step": 27580,
      "task_loss": 0.4647744596004486
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30010122060775757,
      "epoch": 9.97,
      "learning_rate": 6.5462492492829316e-09,
      "loss": 0.3054,
      "step": 27590,
      "task_loss": 0.8768385648727417
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23750996589660645,
      "epoch": 9.97,
      "learning_rate": 5.2005327335680196e-09,
      "loss": 0.3119,
      "step": 27600,
      "task_loss": 0.33597642183303833
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2556949853897095,
      "epoch": 9.98,
      "learning_rate": 4.0094794901623934e-09,
      "loss": 0.3085,
      "step": 27610,
      "task_loss": 0.19064223766326904
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27123573422431946,
      "epoch": 9.98,
      "learning_rate": 2.973095660540048e-09,
      "loss": 0.2973,
      "step": 27620,
      "task_loss": 0.4306481182575226
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30214256048202515,
      "epoch": 9.99,
      "learning_rate": 2.0913865886462714e-09,
      "loss": 0.3515,
      "step": 27630,
      "task_loss": 0.6253249645233154
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25861015915870667,
      "epoch": 9.99,
      "learning_rate": 1.364356820864332e-09,
      "loss": 0.3419,
      "step": 27640,
      "task_loss": 0.4915951192378998
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2867693603038788,
      "epoch": 9.99,
      "learning_rate": 7.920101060088225e-10,
      "loss": 0.3529,
      "step": 27650,
      "task_loss": 0.38751643896102905
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2829661965370178,
      "epoch": 10.0,
      "learning_rate": 3.743493952890198e-10,
      "loss": 0.2897,
      "step": 27660,
      "task_loss": 0.3813509941101074
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1939045637845993,
      "epoch": 10.0,
      "learning_rate": 1.1137684230888567e-10,
      "loss": 0.3115,
      "step": 27670,
      "task_loss": 0.2901547849178314
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2560853064060211,
      "epoch": 10.0,
      "learning_rate": 3.0938030370908365e-12,
      "loss": 0.3379,
      "step": 27680,
      "task_loss": 0.46912720799446106
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2885116934776306,
      "epoch": 10.01,
      "learning_rate": 5.9999950499164185e-05,
      "loss": 0.3083,
      "step": 27690,
      "task_loss": 0.39502573013305664
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3206736445426941,
      "epoch": 10.01,
      "learning_rate": 5.999974940229864e-05,
      "loss": 0.3122,
      "step": 27700,
      "task_loss": 0.4612738788127899
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33336344361305237,
      "epoch": 10.01,
      "learning_rate": 5.999939361663725e-05,
      "loss": 0.3242,
      "step": 27710,
      "task_loss": 0.4723026752471924
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3171226382255554,
      "epoch": 10.02,
      "learning_rate": 5.999888314401456e-05,
      "loss": 0.3177,
      "step": 27720,
      "task_loss": 0.48637786507606506
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26646995544433594,
      "epoch": 10.02,
      "learning_rate": 5.9998217987062765e-05,
      "loss": 0.3087,
      "step": 27730,
      "task_loss": 0.5959435701370239
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35628658533096313,
      "epoch": 10.03,
      "learning_rate": 5.999739814921161e-05,
      "loss": 0.3672,
      "step": 27740,
      "task_loss": 0.9028927683830261
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35965442657470703,
      "epoch": 10.03,
      "learning_rate": 5.9996423634688465e-05,
      "loss": 0.3404,
      "step": 27750,
      "task_loss": 0.8516654968261719
    },
    {
      "epoch": 10.03,
      "eval_exact_match": 82.55439924314096,
      "eval_f1": 89.54112319609712,
      "step": 27750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2923398017883301,
      "epoch": 10.03,
      "learning_rate": 5.999529444851826e-05,
      "loss": 0.3326,
      "step": 27760,
      "task_loss": 0.4506998658180237
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39279308915138245,
      "epoch": 10.04,
      "learning_rate": 5.999401059652345e-05,
      "loss": 0.3135,
      "step": 27770,
      "task_loss": 0.7872811555862427
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.368738055229187,
      "epoch": 10.04,
      "learning_rate": 5.999257208532401e-05,
      "loss": 0.3256,
      "step": 27780,
      "task_loss": 0.42682385444641113
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3173050284385681,
      "epoch": 10.04,
      "learning_rate": 5.99909789223374e-05,
      "loss": 0.3742,
      "step": 27790,
      "task_loss": 0.6968157291412354
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3928011655807495,
      "epoch": 10.05,
      "learning_rate": 5.998923111577847e-05,
      "loss": 0.3723,
      "step": 27800,
      "task_loss": 0.42035573720932007
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3175928294658661,
      "epoch": 10.05,
      "learning_rate": 5.998732867465954e-05,
      "loss": 0.3445,
      "step": 27810,
      "task_loss": 0.7033043503761292
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28419843316078186,
      "epoch": 10.05,
      "learning_rate": 5.998527160879021e-05,
      "loss": 0.3125,
      "step": 27820,
      "task_loss": 0.340492844581604
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3261581063270569,
      "epoch": 10.06,
      "learning_rate": 5.998305992877741e-05,
      "loss": 0.3205,
      "step": 27830,
      "task_loss": 0.5372412204742432
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36411145329475403,
      "epoch": 10.06,
      "learning_rate": 5.99806936460253e-05,
      "loss": 0.3273,
      "step": 27840,
      "task_loss": 0.5985209345817566
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38092899322509766,
      "epoch": 10.07,
      "learning_rate": 5.997817277273525e-05,
      "loss": 0.3682,
      "step": 27850,
      "task_loss": 0.6099511384963989
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2801337242126465,
      "epoch": 10.07,
      "learning_rate": 5.997549732190571e-05,
      "loss": 0.3201,
      "step": 27860,
      "task_loss": 0.6733810901641846
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3363659381866455,
      "epoch": 10.07,
      "learning_rate": 5.997266730733221e-05,
      "loss": 0.3784,
      "step": 27870,
      "task_loss": 0.6295250654220581
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45163094997406006,
      "epoch": 10.08,
      "learning_rate": 5.996968274360725e-05,
      "loss": 0.366,
      "step": 27880,
      "task_loss": 0.9575765132904053
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3112249970436096,
      "epoch": 10.08,
      "learning_rate": 5.996654364612026e-05,
      "loss": 0.3643,
      "step": 27890,
      "task_loss": 0.45321786403656006
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32966169714927673,
      "epoch": 10.08,
      "learning_rate": 5.996325003105747e-05,
      "loss": 0.3322,
      "step": 27900,
      "task_loss": 0.330741822719574
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3450155556201935,
      "epoch": 10.09,
      "learning_rate": 5.9959801915401866e-05,
      "loss": 0.3285,
      "step": 27910,
      "task_loss": 0.3840063810348511
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4525710344314575,
      "epoch": 10.09,
      "learning_rate": 5.99561993169331e-05,
      "loss": 0.3503,
      "step": 27920,
      "task_loss": 0.7056595087051392
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36409515142440796,
      "epoch": 10.09,
      "learning_rate": 5.995244225422738e-05,
      "loss": 0.347,
      "step": 27930,
      "task_loss": 0.591515302658081
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4213048815727234,
      "epoch": 10.1,
      "learning_rate": 5.994853074665739e-05,
      "loss": 0.3146,
      "step": 27940,
      "task_loss": 0.4202615022659302
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38822439312934875,
      "epoch": 10.1,
      "learning_rate": 5.9944464814392157e-05,
      "loss": 0.3537,
      "step": 27950,
      "task_loss": 0.646776556968689
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38637375831604004,
      "epoch": 10.1,
      "learning_rate": 5.9940244478397005e-05,
      "loss": 0.3575,
      "step": 27960,
      "task_loss": 0.4372355341911316
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3227386474609375,
      "epoch": 10.11,
      "learning_rate": 5.9935869760433415e-05,
      "loss": 0.3577,
      "step": 27970,
      "task_loss": 0.44878262281417847
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.322928786277771,
      "epoch": 10.11,
      "learning_rate": 5.99313406830589e-05,
      "loss": 0.3376,
      "step": 27980,
      "task_loss": 0.7455002665519714
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40133416652679443,
      "epoch": 10.12,
      "learning_rate": 5.99266572696269e-05,
      "loss": 0.3633,
      "step": 27990,
      "task_loss": 0.8214153051376343
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31198346614837646,
      "epoch": 10.12,
      "learning_rate": 5.992181954428667e-05,
      "loss": 0.3622,
      "step": 28000,
      "task_loss": 0.37712812423706055
    },
    {
      "epoch": 10.12,
      "eval_exact_match": 82.68684957426679,
      "eval_f1": 89.56443195465796,
      "step": 28000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24220602214336395,
      "epoch": 10.12,
      "learning_rate": 5.9916827531983165e-05,
      "loss": 0.3767,
      "step": 28010,
      "task_loss": 0.3110200762748718
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24042615294456482,
      "epoch": 10.13,
      "learning_rate": 5.991168125845685e-05,
      "loss": 0.3189,
      "step": 28020,
      "task_loss": 0.5726457238197327
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37541890144348145,
      "epoch": 10.13,
      "learning_rate": 5.990638075024366e-05,
      "loss": 0.3511,
      "step": 28030,
      "task_loss": 0.4889417290687561
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28294020891189575,
      "epoch": 10.13,
      "learning_rate": 5.990092603467481e-05,
      "loss": 0.3717,
      "step": 28040,
      "task_loss": 0.3334283232688904
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3576716184616089,
      "epoch": 10.14,
      "learning_rate": 5.989531713987662e-05,
      "loss": 0.3557,
      "step": 28050,
      "task_loss": 0.8734741806983948
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4389265179634094,
      "epoch": 10.14,
      "learning_rate": 5.988955409477045e-05,
      "loss": 0.3551,
      "step": 28060,
      "task_loss": 0.5746753215789795
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4083833694458008,
      "epoch": 10.14,
      "learning_rate": 5.98836369290725e-05,
      "loss": 0.3335,
      "step": 28070,
      "task_loss": 0.6252074241638184
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4334869384765625,
      "epoch": 10.15,
      "learning_rate": 5.987756567329367e-05,
      "loss": 0.3735,
      "step": 28080,
      "task_loss": 0.5796908140182495
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2732439339160919,
      "epoch": 10.15,
      "learning_rate": 5.9871340358739385e-05,
      "loss": 0.3554,
      "step": 28090,
      "task_loss": 0.26658836007118225
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3145688772201538,
      "epoch": 10.16,
      "learning_rate": 5.9864961017509476e-05,
      "loss": 0.3682,
      "step": 28100,
      "task_loss": 0.4686429500579834
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29603311419487,
      "epoch": 10.16,
      "learning_rate": 5.985842768249795e-05,
      "loss": 0.3878,
      "step": 28110,
      "task_loss": 0.2740791440010071
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4538814425468445,
      "epoch": 10.16,
      "learning_rate": 5.98517403873929e-05,
      "loss": 0.3946,
      "step": 28120,
      "task_loss": 0.6446802616119385
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3252262473106384,
      "epoch": 10.17,
      "learning_rate": 5.984489916667626e-05,
      "loss": 0.3301,
      "step": 28130,
      "task_loss": 0.6380620002746582
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3742297887802124,
      "epoch": 10.17,
      "learning_rate": 5.983790405562367e-05,
      "loss": 0.3774,
      "step": 28140,
      "task_loss": 0.592261016368866
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38454627990722656,
      "epoch": 10.17,
      "learning_rate": 5.983075509030426e-05,
      "loss": 0.3663,
      "step": 28150,
      "task_loss": 0.7711337804794312
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3592371344566345,
      "epoch": 10.18,
      "learning_rate": 5.982345230758051e-05,
      "loss": 0.3388,
      "step": 28160,
      "task_loss": 0.875689685344696
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29503804445266724,
      "epoch": 10.18,
      "learning_rate": 5.981599574510802e-05,
      "loss": 0.3679,
      "step": 28170,
      "task_loss": 0.408102810382843
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32741719484329224,
      "epoch": 10.18,
      "learning_rate": 5.980838544133534e-05,
      "loss": 0.3239,
      "step": 28180,
      "task_loss": 0.49565500020980835
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3471035957336426,
      "epoch": 10.19,
      "learning_rate": 5.980062143550375e-05,
      "loss": 0.363,
      "step": 28190,
      "task_loss": 0.9158684015274048
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5277878046035767,
      "epoch": 10.19,
      "learning_rate": 5.979270376764708e-05,
      "loss": 0.3955,
      "step": 28200,
      "task_loss": 0.8009511232376099
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30897581577301025,
      "epoch": 10.2,
      "learning_rate": 5.9784632478591475e-05,
      "loss": 0.3419,
      "step": 28210,
      "task_loss": 0.5038349032402039
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3534333407878876,
      "epoch": 10.2,
      "learning_rate": 5.977640760995523e-05,
      "loss": 0.3718,
      "step": 28220,
      "task_loss": 0.5224078297615051
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3860829472541809,
      "epoch": 10.2,
      "learning_rate": 5.9768029204148513e-05,
      "loss": 0.362,
      "step": 28230,
      "task_loss": 0.5800853967666626
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3779979348182678,
      "epoch": 10.21,
      "learning_rate": 5.9759497304373234e-05,
      "loss": 0.4041,
      "step": 28240,
      "task_loss": 1.0521299839019775
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36808979511260986,
      "epoch": 10.21,
      "learning_rate": 5.9750811954622714e-05,
      "loss": 0.3753,
      "step": 28250,
      "task_loss": 0.42708760499954224
    },
    {
      "epoch": 10.21,
      "eval_exact_match": 82.14758751182592,
      "eval_f1": 89.47133153334977,
      "step": 28250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3896569013595581,
      "epoch": 10.21,
      "learning_rate": 5.974197319968153e-05,
      "loss": 0.3771,
      "step": 28260,
      "task_loss": 0.5128267407417297
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38547414541244507,
      "epoch": 10.22,
      "learning_rate": 5.973298108512531e-05,
      "loss": 0.3917,
      "step": 28270,
      "task_loss": 0.6350282430648804
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2885977029800415,
      "epoch": 10.22,
      "learning_rate": 5.972383565732038e-05,
      "loss": 0.349,
      "step": 28280,
      "task_loss": 0.38273125886917114
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3749898374080658,
      "epoch": 10.22,
      "learning_rate": 5.9715473728427166e-05,
      "loss": 0.3714,
      "step": 28290,
      "task_loss": 0.29564017057418823
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36325937509536743,
      "epoch": 10.23,
      "learning_rate": 5.970603713602016e-05,
      "loss": 0.3571,
      "step": 28300,
      "task_loss": 0.5103201866149902
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4782273471355438,
      "epoch": 10.23,
      "learning_rate": 5.9696447369296534e-05,
      "loss": 0.3782,
      "step": 28310,
      "task_loss": 0.811413586139679
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25456976890563965,
      "epoch": 10.23,
      "learning_rate": 5.968670447770433e-05,
      "loss": 0.3548,
      "step": 28320,
      "task_loss": 0.374428927898407
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33408117294311523,
      "epoch": 10.24,
      "learning_rate": 5.96768085114812e-05,
      "loss": 0.3829,
      "step": 28330,
      "task_loss": 0.4150253236293793
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32442450523376465,
      "epoch": 10.24,
      "learning_rate": 5.966675952165403e-05,
      "loss": 0.3675,
      "step": 28340,
      "task_loss": 0.7664307355880737
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4130185842514038,
      "epoch": 10.25,
      "learning_rate": 5.965655756003884e-05,
      "loss": 0.3821,
      "step": 28350,
      "task_loss": 0.812406063079834
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3772509694099426,
      "epoch": 10.25,
      "learning_rate": 5.964620267924034e-05,
      "loss": 0.3437,
      "step": 28360,
      "task_loss": 0.6886706352233887
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3586800694465637,
      "epoch": 10.25,
      "learning_rate": 5.963569493265179e-05,
      "loss": 0.3691,
      "step": 28370,
      "task_loss": 0.726294994354248
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33081382513046265,
      "epoch": 10.26,
      "learning_rate": 5.9625034374454664e-05,
      "loss": 0.3461,
      "step": 28380,
      "task_loss": 0.8060112595558167
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39789003133773804,
      "epoch": 10.26,
      "learning_rate": 5.961422105961839e-05,
      "loss": 0.3707,
      "step": 28390,
      "task_loss": 0.5250519514083862
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4983270764350891,
      "epoch": 10.26,
      "learning_rate": 5.9603255043900047e-05,
      "loss": 0.3661,
      "step": 28400,
      "task_loss": 0.4081471264362335
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.256054162979126,
      "epoch": 10.27,
      "learning_rate": 5.95921363838441e-05,
      "loss": 0.373,
      "step": 28410,
      "task_loss": 0.4245930314064026
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26962295174598694,
      "epoch": 10.27,
      "learning_rate": 5.95808651367821e-05,
      "loss": 0.3395,
      "step": 28420,
      "task_loss": 0.26497402787208557
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2624947726726532,
      "epoch": 10.27,
      "learning_rate": 5.956944136083239e-05,
      "loss": 0.3957,
      "step": 28430,
      "task_loss": 0.25945717096328735
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34345728158950806,
      "epoch": 10.28,
      "learning_rate": 5.95578651148998e-05,
      "loss": 0.3927,
      "step": 28440,
      "task_loss": 0.41743308305740356
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34728893637657166,
      "epoch": 10.28,
      "learning_rate": 5.954613645867534e-05,
      "loss": 0.3566,
      "step": 28450,
      "task_loss": 0.5428339838981628
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3047831058502197,
      "epoch": 10.29,
      "learning_rate": 5.9534255452635915e-05,
      "loss": 0.386,
      "step": 28460,
      "task_loss": 0.3285978436470032
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44751378893852234,
      "epoch": 10.29,
      "learning_rate": 5.9522222158043986e-05,
      "loss": 0.3781,
      "step": 28470,
      "task_loss": 0.32941681146621704
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3645317554473877,
      "epoch": 10.29,
      "learning_rate": 5.9510036636947265e-05,
      "loss": 0.381,
      "step": 28480,
      "task_loss": 0.7505057454109192
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3114251494407654,
      "epoch": 10.3,
      "learning_rate": 5.9497698952178385e-05,
      "loss": 0.3428,
      "step": 28490,
      "task_loss": 0.47057342529296875
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3543356657028198,
      "epoch": 10.3,
      "learning_rate": 5.948520916735462e-05,
      "loss": 0.4157,
      "step": 28500,
      "task_loss": 0.4492913484573364
    },
    {
      "epoch": 10.3,
      "eval_exact_match": 82.74361400189214,
      "eval_f1": 89.51669420473074,
      "step": 28500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3564971089363098,
      "epoch": 10.3,
      "learning_rate": 5.9472567346877475e-05,
      "loss": 0.3685,
      "step": 28510,
      "task_loss": 0.3423532545566559
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28164559602737427,
      "epoch": 10.31,
      "learning_rate": 5.9459773555932445e-05,
      "loss": 0.3434,
      "step": 28520,
      "task_loss": 0.7573295831680298
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24590784311294556,
      "epoch": 10.31,
      "learning_rate": 5.9446827860488616e-05,
      "loss": 0.3907,
      "step": 28530,
      "task_loss": 0.570406436920166
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34086698293685913,
      "epoch": 10.31,
      "learning_rate": 5.943373032729835e-05,
      "loss": 0.3792,
      "step": 28540,
      "task_loss": 0.24719828367233276
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37996360659599304,
      "epoch": 10.32,
      "learning_rate": 5.942048102389691e-05,
      "loss": 0.3728,
      "step": 28550,
      "task_loss": 0.6601479053497314
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2580659091472626,
      "epoch": 10.32,
      "learning_rate": 5.940708001860219e-05,
      "loss": 0.3737,
      "step": 28560,
      "task_loss": 0.340899795293808
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43038126826286316,
      "epoch": 10.33,
      "learning_rate": 5.9393527380514265e-05,
      "loss": 0.3748,
      "step": 28570,
      "task_loss": 0.684032678604126
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30722588300704956,
      "epoch": 10.33,
      "learning_rate": 5.937982317951509e-05,
      "loss": 0.3499,
      "step": 28580,
      "task_loss": 0.2970711290836334
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38105618953704834,
      "epoch": 10.33,
      "learning_rate": 5.936596748626814e-05,
      "loss": 0.3456,
      "step": 28590,
      "task_loss": 0.3015686273574829
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4224506616592407,
      "epoch": 10.34,
      "learning_rate": 5.935196037221803e-05,
      "loss": 0.3923,
      "step": 28600,
      "task_loss": 0.6144254207611084
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2678179442882538,
      "epoch": 10.34,
      "learning_rate": 5.933780190959015e-05,
      "loss": 0.3617,
      "step": 28610,
      "task_loss": 0.404130220413208
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40702712535858154,
      "epoch": 10.34,
      "learning_rate": 5.932349217139029e-05,
      "loss": 0.3927,
      "step": 28620,
      "task_loss": 0.5161640048027039
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39445358514785767,
      "epoch": 10.35,
      "learning_rate": 5.930903123140427e-05,
      "loss": 0.3783,
      "step": 28630,
      "task_loss": 0.7414616346359253
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30149519443511963,
      "epoch": 10.35,
      "learning_rate": 5.9294419164197565e-05,
      "loss": 0.3494,
      "step": 28640,
      "task_loss": 0.5813414454460144
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43674802780151367,
      "epoch": 10.35,
      "learning_rate": 5.927965604511491e-05,
      "loss": 0.3455,
      "step": 28650,
      "task_loss": 0.9160444736480713
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4207167625427246,
      "epoch": 10.36,
      "learning_rate": 5.9264741950279895e-05,
      "loss": 0.4298,
      "step": 28660,
      "task_loss": 0.6424660682678223
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38436251878738403,
      "epoch": 10.36,
      "learning_rate": 5.924967695659463e-05,
      "loss": 0.3763,
      "step": 28670,
      "task_loss": 0.8374863862991333
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25715458393096924,
      "epoch": 10.37,
      "learning_rate": 5.923446114173928e-05,
      "loss": 0.3494,
      "step": 28680,
      "task_loss": 0.34484949707984924
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3070823550224304,
      "epoch": 10.37,
      "learning_rate": 5.9219094584171695e-05,
      "loss": 0.3655,
      "step": 28690,
      "task_loss": 0.45581191778182983
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.392423152923584,
      "epoch": 10.37,
      "learning_rate": 5.9203577363127016e-05,
      "loss": 0.3727,
      "step": 28700,
      "task_loss": 0.5320243835449219
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44943130016326904,
      "epoch": 10.38,
      "learning_rate": 5.918790955861727e-05,
      "loss": 0.3844,
      "step": 28710,
      "task_loss": 0.5158875584602356
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36688247323036194,
      "epoch": 10.38,
      "learning_rate": 5.91720912514309e-05,
      "loss": 0.3352,
      "step": 28720,
      "task_loss": 0.4710213243961334
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30681973695755005,
      "epoch": 10.38,
      "learning_rate": 5.915612252313243e-05,
      "loss": 0.3831,
      "step": 28730,
      "task_loss": 0.39453649520874023
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3728558421134949,
      "epoch": 10.39,
      "learning_rate": 5.914000345606199e-05,
      "loss": 0.4136,
      "step": 28740,
      "task_loss": 0.4394037425518036
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.388075053691864,
      "epoch": 10.39,
      "learning_rate": 5.9123734133334916e-05,
      "loss": 0.3804,
      "step": 28750,
      "task_loss": 0.9567576050758362
    },
    {
      "epoch": 10.39,
      "eval_exact_match": 82.3841059602649,
      "eval_f1": 89.3657178878086,
      "step": 28750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4076917767524719,
      "epoch": 10.39,
      "learning_rate": 5.91073146388413e-05,
      "loss": 0.3567,
      "step": 28760,
      "task_loss": 0.8818804025650024
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3763560950756073,
      "epoch": 10.4,
      "learning_rate": 5.909074505724557e-05,
      "loss": 0.3752,
      "step": 28770,
      "task_loss": 0.5184281468391418
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41304516792297363,
      "epoch": 10.4,
      "learning_rate": 5.907402547398609e-05,
      "loss": 0.3807,
      "step": 28780,
      "task_loss": 0.6674418449401855
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5048227906227112,
      "epoch": 10.4,
      "learning_rate": 5.905715597527462e-05,
      "loss": 0.3916,
      "step": 28790,
      "task_loss": 1.2223567962646484
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45865944027900696,
      "epoch": 10.41,
      "learning_rate": 5.9040136648095986e-05,
      "loss": 0.3883,
      "step": 28800,
      "task_loss": 0.4730258882045746
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3826584219932556,
      "epoch": 10.41,
      "learning_rate": 5.902296758020755e-05,
      "loss": 0.4247,
      "step": 28810,
      "task_loss": 0.3842507004737854
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3068469762802124,
      "epoch": 10.42,
      "learning_rate": 5.900564886013882e-05,
      "loss": 0.3572,
      "step": 28820,
      "task_loss": 0.348866730928421
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30432772636413574,
      "epoch": 10.42,
      "learning_rate": 5.8988180577190914e-05,
      "loss": 0.3867,
      "step": 28830,
      "task_loss": 0.32357197999954224
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25718772411346436,
      "epoch": 10.42,
      "learning_rate": 5.8970562821436184e-05,
      "loss": 0.327,
      "step": 28840,
      "task_loss": 0.38030368089675903
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3453293442726135,
      "epoch": 10.43,
      "learning_rate": 5.895279568371772e-05,
      "loss": 0.336,
      "step": 28850,
      "task_loss": 0.5080013275146484
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35937240719795227,
      "epoch": 10.43,
      "learning_rate": 5.8934879255648834e-05,
      "loss": 0.39,
      "step": 28860,
      "task_loss": 0.3022608458995819
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30319833755493164,
      "epoch": 10.43,
      "learning_rate": 5.8916813629612655e-05,
      "loss": 0.4019,
      "step": 28870,
      "task_loss": 0.4994625151157379
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37237548828125,
      "epoch": 10.44,
      "learning_rate": 5.8898598898761656e-05,
      "loss": 0.4214,
      "step": 28880,
      "task_loss": 0.6839677095413208
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4111763834953308,
      "epoch": 10.44,
      "learning_rate": 5.8880235157017074e-05,
      "loss": 0.4021,
      "step": 28890,
      "task_loss": 0.46749764680862427
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34317219257354736,
      "epoch": 10.44,
      "learning_rate": 5.886172249906856e-05,
      "loss": 0.3653,
      "step": 28900,
      "task_loss": 0.7069752216339111
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47758495807647705,
      "epoch": 10.45,
      "learning_rate": 5.8843061020373594e-05,
      "loss": 0.3454,
      "step": 28910,
      "task_loss": 0.5334576964378357
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3918111324310303,
      "epoch": 10.45,
      "learning_rate": 5.882425081715705e-05,
      "loss": 0.4134,
      "step": 28920,
      "task_loss": 0.47633877396583557
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42572399973869324,
      "epoch": 10.46,
      "learning_rate": 5.8805291986410646e-05,
      "loss": 0.3451,
      "step": 28930,
      "task_loss": 0.44892263412475586
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3907076120376587,
      "epoch": 10.46,
      "learning_rate": 5.878618462589249e-05,
      "loss": 0.417,
      "step": 28940,
      "task_loss": 0.8796771168708801
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49205639958381653,
      "epoch": 10.46,
      "learning_rate": 5.876692883412656e-05,
      "loss": 0.374,
      "step": 28950,
      "task_loss": 0.7883171439170837
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23329684138298035,
      "epoch": 10.47,
      "learning_rate": 5.87475247104022e-05,
      "loss": 0.3493,
      "step": 28960,
      "task_loss": 0.2832324504852295
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4535112977027893,
      "epoch": 10.47,
      "learning_rate": 5.872797235477359e-05,
      "loss": 0.3867,
      "step": 28970,
      "task_loss": 0.45803922414779663
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.397172749042511,
      "epoch": 10.47,
      "learning_rate": 5.8708271868059255e-05,
      "loss": 0.3743,
      "step": 28980,
      "task_loss": 0.587207019329071
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33365657925605774,
      "epoch": 10.48,
      "learning_rate": 5.868842335184153e-05,
      "loss": 0.3598,
      "step": 28990,
      "task_loss": 0.5650234818458557
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.395450234413147,
      "epoch": 10.48,
      "learning_rate": 5.866842690846602e-05,
      "loss": 0.4087,
      "step": 29000,
      "task_loss": 0.5595653057098389
    },
    {
      "epoch": 10.48,
      "eval_exact_match": 82.42194891201514,
      "eval_f1": 89.48886603082968,
      "step": 29000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3053317666053772,
      "epoch": 10.48,
      "learning_rate": 5.864828264104114e-05,
      "loss": 0.3498,
      "step": 29010,
      "task_loss": 0.34997349977493286
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3626655340194702,
      "epoch": 10.49,
      "learning_rate": 5.862799065343749e-05,
      "loss": 0.3818,
      "step": 29020,
      "task_loss": 0.5937673449516296
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42667001485824585,
      "epoch": 10.49,
      "learning_rate": 5.860755105028738e-05,
      "loss": 0.3867,
      "step": 29030,
      "task_loss": 0.9323544502258301
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42904749512672424,
      "epoch": 10.5,
      "learning_rate": 5.858696393698428e-05,
      "loss": 0.3985,
      "step": 29040,
      "task_loss": 0.46694380044937134
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39898544549942017,
      "epoch": 10.5,
      "learning_rate": 5.8566229419682275e-05,
      "loss": 0.4042,
      "step": 29050,
      "task_loss": 0.5771615505218506
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20150551199913025,
      "epoch": 10.5,
      "learning_rate": 5.8545347605295516e-05,
      "loss": 0.3512,
      "step": 29060,
      "task_loss": 0.5410712957382202
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3072255253791809,
      "epoch": 10.51,
      "learning_rate": 5.852431860149763e-05,
      "loss": 0.3717,
      "step": 29070,
      "task_loss": 0.4550820589065552
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4038892388343811,
      "epoch": 10.51,
      "learning_rate": 5.850314251672126e-05,
      "loss": 0.398,
      "step": 29080,
      "task_loss": 0.5738189220428467
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3132917284965515,
      "epoch": 10.51,
      "learning_rate": 5.848181946015741e-05,
      "loss": 0.3839,
      "step": 29090,
      "task_loss": 0.5845606923103333
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4155728816986084,
      "epoch": 10.52,
      "learning_rate": 5.846034954175493e-05,
      "loss": 0.3925,
      "step": 29100,
      "task_loss": 0.6973046660423279
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31962496042251587,
      "epoch": 10.52,
      "learning_rate": 5.8438732872219934e-05,
      "loss": 0.3677,
      "step": 29110,
      "task_loss": 0.6435040235519409
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34306031465530396,
      "epoch": 10.52,
      "learning_rate": 5.8416969563015246e-05,
      "loss": 0.3737,
      "step": 29120,
      "task_loss": 0.42822667956352234
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25659072399139404,
      "epoch": 10.53,
      "learning_rate": 5.83950597263598e-05,
      "loss": 0.3953,
      "step": 29130,
      "task_loss": 0.5244855880737305
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3888118863105774,
      "epoch": 10.53,
      "learning_rate": 5.837300347522809e-05,
      "loss": 0.4479,
      "step": 29140,
      "task_loss": 0.445207417011261
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3718984127044678,
      "epoch": 10.53,
      "learning_rate": 5.8350800923349556e-05,
      "loss": 0.3726,
      "step": 29150,
      "task_loss": 0.8107948303222656
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3401542007923126,
      "epoch": 10.54,
      "learning_rate": 5.8328452185208034e-05,
      "loss": 0.3725,
      "step": 29160,
      "task_loss": 0.29308414459228516
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3953520655632019,
      "epoch": 10.54,
      "learning_rate": 5.830595737604112e-05,
      "loss": 0.3662,
      "step": 29170,
      "task_loss": 0.6135333776473999
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33224552869796753,
      "epoch": 10.55,
      "learning_rate": 5.828331661183961e-05,
      "loss": 0.4135,
      "step": 29180,
      "task_loss": 0.6807444095611572
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46813124418258667,
      "epoch": 10.55,
      "learning_rate": 5.826053000934692e-05,
      "loss": 0.4053,
      "step": 29190,
      "task_loss": 0.48048609495162964
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3617834448814392,
      "epoch": 10.55,
      "learning_rate": 5.8237597686058406e-05,
      "loss": 0.3684,
      "step": 29200,
      "task_loss": 0.684455156326294
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37988534569740295,
      "epoch": 10.56,
      "learning_rate": 5.821451976022086e-05,
      "loss": 0.3826,
      "step": 29210,
      "task_loss": 0.44116660952568054
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3315032124519348,
      "epoch": 10.56,
      "learning_rate": 5.819129635083182e-05,
      "loss": 0.4355,
      "step": 29220,
      "task_loss": 0.8068733215332031
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32461968064308167,
      "epoch": 10.56,
      "learning_rate": 5.8167927577638994e-05,
      "loss": 0.4045,
      "step": 29230,
      "task_loss": 0.49032747745513916
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4072458744049072,
      "epoch": 10.57,
      "learning_rate": 5.8144413561139636e-05,
      "loss": 0.3736,
      "step": 29240,
      "task_loss": 0.5367380976676941
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4317589998245239,
      "epoch": 10.57,
      "learning_rate": 5.812075442257992e-05,
      "loss": 0.4024,
      "step": 29250,
      "task_loss": 0.4210602045059204
    },
    {
      "epoch": 10.57,
      "eval_exact_match": 82.28949858088932,
      "eval_f1": 89.47220420737969,
      "step": 29250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30034422874450684,
      "epoch": 10.57,
      "learning_rate": 5.8096950283954324e-05,
      "loss": 0.3691,
      "step": 29260,
      "task_loss": 0.6281427145004272
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.6280544996261597,
      "epoch": 10.58,
      "learning_rate": 5.8073001268004986e-05,
      "loss": 0.433,
      "step": 29270,
      "task_loss": 0.9039506316184998
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3547070026397705,
      "epoch": 10.58,
      "learning_rate": 5.804890749822109e-05,
      "loss": 0.379,
      "step": 29280,
      "task_loss": 0.6923799514770508
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4179234504699707,
      "epoch": 10.59,
      "learning_rate": 5.8024669098838224e-05,
      "loss": 0.4061,
      "step": 29290,
      "task_loss": 0.5038123726844788
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4669201970100403,
      "epoch": 10.59,
      "learning_rate": 5.8000286194837706e-05,
      "loss": 0.3888,
      "step": 29300,
      "task_loss": 0.8348492383956909
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.354667067527771,
      "epoch": 10.59,
      "learning_rate": 5.7975758911945995e-05,
      "loss": 0.4108,
      "step": 29310,
      "task_loss": 0.4626341164112091
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2784455418586731,
      "epoch": 10.6,
      "learning_rate": 5.795108737663401e-05,
      "loss": 0.3752,
      "step": 29320,
      "task_loss": 0.5510279536247253
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32893824577331543,
      "epoch": 10.6,
      "learning_rate": 5.79262717161165e-05,
      "loss": 0.3742,
      "step": 29330,
      "task_loss": 0.7631270885467529
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3505815863609314,
      "epoch": 10.6,
      "learning_rate": 5.7901312058351334e-05,
      "loss": 0.3496,
      "step": 29340,
      "task_loss": 0.40224701166152954
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4556928277015686,
      "epoch": 10.61,
      "learning_rate": 5.78762085320389e-05,
      "loss": 0.3705,
      "step": 29350,
      "task_loss": 0.6299753189086914
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40033918619155884,
      "epoch": 10.61,
      "learning_rate": 5.7850961266621434e-05,
      "loss": 0.3959,
      "step": 29360,
      "task_loss": 0.7346347570419312
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43684858083724976,
      "epoch": 10.61,
      "learning_rate": 5.782557039228231e-05,
      "loss": 0.3916,
      "step": 29370,
      "task_loss": 0.645737886428833
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4691216051578522,
      "epoch": 10.62,
      "learning_rate": 5.780003603994543e-05,
      "loss": 0.4144,
      "step": 29380,
      "task_loss": 0.3999781012535095
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46305355429649353,
      "epoch": 10.62,
      "learning_rate": 5.77743583412745e-05,
      "loss": 0.4086,
      "step": 29390,
      "task_loss": 0.7214027643203735
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3314850926399231,
      "epoch": 10.63,
      "learning_rate": 5.774853742867237e-05,
      "loss": 0.378,
      "step": 29400,
      "task_loss": 0.6907333731651306
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33742809295654297,
      "epoch": 10.63,
      "learning_rate": 5.772257343528033e-05,
      "loss": 0.3629,
      "step": 29410,
      "task_loss": 0.6695260405540466
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36523616313934326,
      "epoch": 10.63,
      "learning_rate": 5.7696466494977485e-05,
      "loss": 0.4139,
      "step": 29420,
      "task_loss": 0.6513307094573975
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4402333199977875,
      "epoch": 10.64,
      "learning_rate": 5.767021674237998e-05,
      "loss": 0.4005,
      "step": 29430,
      "task_loss": 0.6502874493598938
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2928636968135834,
      "epoch": 10.64,
      "learning_rate": 5.7643824312840375e-05,
      "loss": 0.4047,
      "step": 29440,
      "task_loss": 0.4543102979660034
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3670118749141693,
      "epoch": 10.64,
      "learning_rate": 5.76172893424469e-05,
      "loss": 0.412,
      "step": 29450,
      "task_loss": 0.467820942401886
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5537813901901245,
      "epoch": 10.65,
      "learning_rate": 5.75906119680228e-05,
      "loss": 0.3999,
      "step": 29460,
      "task_loss": 0.6507628560066223
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27481332421302795,
      "epoch": 10.65,
      "learning_rate": 5.756379232712559e-05,
      "loss": 0.3434,
      "step": 29470,
      "task_loss": 0.5384325385093689
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38477927446365356,
      "epoch": 10.65,
      "learning_rate": 5.753683055804634e-05,
      "loss": 0.3771,
      "step": 29480,
      "task_loss": 0.5689960718154907
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31757038831710815,
      "epoch": 10.66,
      "learning_rate": 5.750972679980902e-05,
      "loss": 0.4164,
      "step": 29490,
      "task_loss": 0.6483100056648254
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44253355264663696,
      "epoch": 10.66,
      "learning_rate": 5.7482481192169704e-05,
      "loss": 0.376,
      "step": 29500,
      "task_loss": 0.624772310256958
    },
    {
      "epoch": 10.66,
      "eval_exact_match": 82.56385998107852,
      "eval_f1": 89.60784348253303,
      "step": 29500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44806426763534546,
      "epoch": 10.66,
      "learning_rate": 5.745509387561591e-05,
      "loss": 0.3933,
      "step": 29510,
      "task_loss": 0.5298267602920532
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3329993486404419,
      "epoch": 10.67,
      "learning_rate": 5.742756499136585e-05,
      "loss": 0.3558,
      "step": 29520,
      "task_loss": 0.5523538589477539
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3392196297645569,
      "epoch": 10.67,
      "learning_rate": 5.739989468136772e-05,
      "loss": 0.3664,
      "step": 29530,
      "task_loss": 0.44542059302330017
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3531342148780823,
      "epoch": 10.68,
      "learning_rate": 5.7372083088298935e-05,
      "loss": 0.3518,
      "step": 29540,
      "task_loss": 0.43197929859161377
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4389444887638092,
      "epoch": 10.68,
      "learning_rate": 5.734413035556541e-05,
      "loss": 0.3831,
      "step": 29550,
      "task_loss": 0.48105964064598083
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35376453399658203,
      "epoch": 10.68,
      "learning_rate": 5.731603662730085e-05,
      "loss": 0.391,
      "step": 29560,
      "task_loss": 0.582084059715271
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38745567202568054,
      "epoch": 10.69,
      "learning_rate": 5.7287802048365964e-05,
      "loss": 0.3715,
      "step": 29570,
      "task_loss": 0.6971489191055298
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3742871582508087,
      "epoch": 10.69,
      "learning_rate": 5.725942676434773e-05,
      "loss": 0.3688,
      "step": 29580,
      "task_loss": 0.3916323482990265
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5940446853637695,
      "epoch": 10.69,
      "learning_rate": 5.723091092155865e-05,
      "loss": 0.453,
      "step": 29590,
      "task_loss": 0.8099455833435059
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3112245798110962,
      "epoch": 10.7,
      "learning_rate": 5.720225466703601e-05,
      "loss": 0.3629,
      "step": 29600,
      "task_loss": 0.631229043006897
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32478272914886475,
      "epoch": 10.7,
      "learning_rate": 5.7173458148541094e-05,
      "loss": 0.4228,
      "step": 29610,
      "task_loss": 0.4904170632362366
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41761791706085205,
      "epoch": 10.7,
      "learning_rate": 5.714452151455841e-05,
      "loss": 0.3698,
      "step": 29620,
      "task_loss": 1.1223629713058472
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3478744626045227,
      "epoch": 10.71,
      "learning_rate": 5.711544491429501e-05,
      "loss": 0.3702,
      "step": 29630,
      "task_loss": 0.19475889205932617
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41987693309783936,
      "epoch": 10.71,
      "learning_rate": 5.708622849767961e-05,
      "loss": 0.4604,
      "step": 29640,
      "task_loss": 0.6271519660949707
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2809098958969116,
      "epoch": 10.72,
      "learning_rate": 5.705687241536186e-05,
      "loss": 0.4281,
      "step": 29650,
      "task_loss": 0.6296674013137817
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2818308472633362,
      "epoch": 10.72,
      "learning_rate": 5.702737681871161e-05,
      "loss": 0.3584,
      "step": 29660,
      "task_loss": 0.39925143122673035
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35148537158966064,
      "epoch": 10.72,
      "learning_rate": 5.699774185981806e-05,
      "loss": 0.3706,
      "step": 29670,
      "task_loss": 0.5052469372749329
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35409435629844666,
      "epoch": 10.73,
      "learning_rate": 5.696796769148905e-05,
      "loss": 0.418,
      "step": 29680,
      "task_loss": 0.5041142702102661
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3384610414505005,
      "epoch": 10.73,
      "learning_rate": 5.693805446725018e-05,
      "loss": 0.4002,
      "step": 29690,
      "task_loss": 0.8012888431549072
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41421744227409363,
      "epoch": 10.73,
      "learning_rate": 5.69080023413441e-05,
      "loss": 0.398,
      "step": 29700,
      "task_loss": 0.4893695116043091
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30970844626426697,
      "epoch": 10.74,
      "learning_rate": 5.687781146872967e-05,
      "loss": 0.3823,
      "step": 29710,
      "task_loss": 0.433634489774704
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20200911164283752,
      "epoch": 10.74,
      "learning_rate": 5.6847482005081207e-05,
      "loss": 0.3594,
      "step": 29720,
      "task_loss": 0.3214665651321411
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.366502583026886,
      "epoch": 10.74,
      "learning_rate": 5.681701410678759e-05,
      "loss": 0.4093,
      "step": 29730,
      "task_loss": 0.535211443901062
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3267229199409485,
      "epoch": 10.75,
      "learning_rate": 5.678640793095156e-05,
      "loss": 0.3556,
      "step": 29740,
      "task_loss": 0.5452145934104919
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39769360423088074,
      "epoch": 10.75,
      "learning_rate": 5.6755663635388865e-05,
      "loss": 0.3696,
      "step": 29750,
      "task_loss": 0.32848280668258667
    },
    {
      "epoch": 10.75,
      "eval_exact_match": 82.94228949858089,
      "eval_f1": 89.87294387331669,
      "step": 29750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3788045048713684,
      "epoch": 10.76,
      "learning_rate": 5.6724781378627425e-05,
      "loss": 0.3952,
      "step": 29760,
      "task_loss": 0.6843670606613159
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46630746126174927,
      "epoch": 10.76,
      "learning_rate": 5.669376131990653e-05,
      "loss": 0.402,
      "step": 29770,
      "task_loss": 0.6446393132209778
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35927873849868774,
      "epoch": 10.76,
      "learning_rate": 5.666260361917604e-05,
      "loss": 0.3987,
      "step": 29780,
      "task_loss": 0.5452622771263123
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44025394320487976,
      "epoch": 10.77,
      "learning_rate": 5.6631308437095546e-05,
      "loss": 0.3841,
      "step": 29790,
      "task_loss": 0.5754637122154236
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3944118916988373,
      "epoch": 10.77,
      "learning_rate": 5.6599875935033516e-05,
      "loss": 0.3838,
      "step": 29800,
      "task_loss": 0.34058427810668945
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45717811584472656,
      "epoch": 10.77,
      "learning_rate": 5.656830627506651e-05,
      "loss": 0.3812,
      "step": 29810,
      "task_loss": 0.5898969173431396
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3032148778438568,
      "epoch": 10.78,
      "learning_rate": 5.6536599619978306e-05,
      "loss": 0.3668,
      "step": 29820,
      "task_loss": 0.6649981141090393
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40282613039016724,
      "epoch": 10.78,
      "learning_rate": 5.650475613325907e-05,
      "loss": 0.3864,
      "step": 29830,
      "task_loss": 0.959244966506958
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25969162583351135,
      "epoch": 10.78,
      "learning_rate": 5.647277597910456e-05,
      "loss": 0.3826,
      "step": 29840,
      "task_loss": 0.3896484673023224
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3874444365501404,
      "epoch": 10.79,
      "learning_rate": 5.644065932241517e-05,
      "loss": 0.3751,
      "step": 29850,
      "task_loss": 0.5907013416290283
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.355577677488327,
      "epoch": 10.79,
      "learning_rate": 5.6408406328795195e-05,
      "loss": 0.3604,
      "step": 29860,
      "task_loss": 0.539420485496521
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33990269899368286,
      "epoch": 10.8,
      "learning_rate": 5.6376017164551915e-05,
      "loss": 0.3428,
      "step": 29870,
      "task_loss": 0.47408655285835266
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.295065313577652,
      "epoch": 10.8,
      "learning_rate": 5.6343491996694754e-05,
      "loss": 0.4273,
      "step": 29880,
      "task_loss": 0.3917389512062073
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3400767147541046,
      "epoch": 10.8,
      "learning_rate": 5.631083099293441e-05,
      "loss": 0.3782,
      "step": 29890,
      "task_loss": 0.29112324118614197
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30475717782974243,
      "epoch": 10.81,
      "learning_rate": 5.627803432168202e-05,
      "loss": 0.3963,
      "step": 29900,
      "task_loss": 0.6054189205169678
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38359367847442627,
      "epoch": 10.81,
      "learning_rate": 5.6245102152048215e-05,
      "loss": 0.3931,
      "step": 29910,
      "task_loss": 0.5336320400238037
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38667869567871094,
      "epoch": 10.81,
      "learning_rate": 5.621203465384236e-05,
      "loss": 0.3865,
      "step": 29920,
      "task_loss": 0.47453516721725464
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2894802391529083,
      "epoch": 10.82,
      "learning_rate": 5.617883199757159e-05,
      "loss": 0.393,
      "step": 29930,
      "task_loss": 0.37323397397994995
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3388747572898865,
      "epoch": 10.82,
      "learning_rate": 5.6145494354439964e-05,
      "loss": 0.3999,
      "step": 29940,
      "task_loss": 0.6223235130310059
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3649522066116333,
      "epoch": 10.82,
      "learning_rate": 5.611202189634758e-05,
      "loss": 0.3712,
      "step": 29950,
      "task_loss": 0.5318911075592041
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46085262298583984,
      "epoch": 10.83,
      "learning_rate": 5.6078414795889694e-05,
      "loss": 0.3899,
      "step": 29960,
      "task_loss": 0.5634099245071411
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4994548559188843,
      "epoch": 10.83,
      "learning_rate": 5.604467322635582e-05,
      "loss": 0.3865,
      "step": 29970,
      "task_loss": 0.6656855344772339
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3881114721298218,
      "epoch": 10.83,
      "learning_rate": 5.6010797361728816e-05,
      "loss": 0.4178,
      "step": 29980,
      "task_loss": 0.48665136098861694
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3745848834514618,
      "epoch": 10.84,
      "learning_rate": 5.597678737668405e-05,
      "loss": 0.3859,
      "step": 29990,
      "task_loss": 0.7532287836074829
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34191638231277466,
      "epoch": 10.84,
      "learning_rate": 5.594264344658843e-05,
      "loss": 0.3605,
      "step": 30000,
      "task_loss": 1.0175566673278809
    },
    {
      "epoch": 10.84,
      "eval_exact_match": 82.94228949858089,
      "eval_f1": 89.72641922575288,
      "step": 30000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3670957684516907,
      "epoch": 10.85,
      "learning_rate": 5.5908365747499555e-05,
      "loss": 0.3526,
      "step": 30010,
      "task_loss": 0.5466479659080505
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3133726716041565,
      "epoch": 10.85,
      "learning_rate": 5.587395445616475e-05,
      "loss": 0.3785,
      "step": 30020,
      "task_loss": 0.41326943039894104
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38958606123924255,
      "epoch": 10.85,
      "learning_rate": 5.583940975002023e-05,
      "loss": 0.3621,
      "step": 30030,
      "task_loss": 0.5244191884994507
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34110361337661743,
      "epoch": 10.86,
      "learning_rate": 5.5804731807190085e-05,
      "loss": 0.3703,
      "step": 30040,
      "task_loss": 0.4226912260055542
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3145621418952942,
      "epoch": 10.86,
      "learning_rate": 5.576992080648548e-05,
      "loss": 0.431,
      "step": 30050,
      "task_loss": 0.4115636944770813
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4243555963039398,
      "epoch": 10.86,
      "learning_rate": 5.57349769274036e-05,
      "loss": 0.3942,
      "step": 30060,
      "task_loss": 0.42534300684928894
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4849332571029663,
      "epoch": 10.87,
      "learning_rate": 5.569990035012688e-05,
      "loss": 0.3818,
      "step": 30070,
      "task_loss": 0.6155873537063599
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19062018394470215,
      "epoch": 10.87,
      "learning_rate": 5.566469125552193e-05,
      "loss": 0.398,
      "step": 30080,
      "task_loss": 0.18269407749176025
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3313145339488983,
      "epoch": 10.87,
      "learning_rate": 5.562934982513867e-05,
      "loss": 0.3657,
      "step": 30090,
      "task_loss": 0.5402665138244629
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4278787076473236,
      "epoch": 10.88,
      "learning_rate": 5.559387624120942e-05,
      "loss": 0.3884,
      "step": 30100,
      "task_loss": 0.4952021837234497
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32850050926208496,
      "epoch": 10.88,
      "learning_rate": 5.5558270686647886e-05,
      "loss": 0.3914,
      "step": 30110,
      "task_loss": 0.5060421824455261
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5403910279273987,
      "epoch": 10.89,
      "learning_rate": 5.55225333450483e-05,
      "loss": 0.3696,
      "step": 30120,
      "task_loss": 0.9411796927452087
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3275948464870453,
      "epoch": 10.89,
      "learning_rate": 5.5486664400684396e-05,
      "loss": 0.4234,
      "step": 30130,
      "task_loss": 0.7505064010620117
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2951309084892273,
      "epoch": 10.89,
      "learning_rate": 5.5450664038508526e-05,
      "loss": 0.403,
      "step": 30140,
      "task_loss": 0.41326138377189636
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5003862977027893,
      "epoch": 10.9,
      "learning_rate": 5.541453244415065e-05,
      "loss": 0.3756,
      "step": 30150,
      "task_loss": 0.6409727334976196
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28051674365997314,
      "epoch": 10.9,
      "learning_rate": 5.5378269803917425e-05,
      "loss": 0.3686,
      "step": 30160,
      "task_loss": 0.228929340839386
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3465709388256073,
      "epoch": 10.9,
      "learning_rate": 5.534187630479123e-05,
      "loss": 0.3693,
      "step": 30170,
      "task_loss": 0.365300714969635
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41735100746154785,
      "epoch": 10.91,
      "learning_rate": 5.530535213442916e-05,
      "loss": 0.3856,
      "step": 30180,
      "task_loss": 0.6544727683067322
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3385262191295624,
      "epoch": 10.91,
      "learning_rate": 5.5268697481162134e-05,
      "loss": 0.3742,
      "step": 30190,
      "task_loss": 0.4145801365375519
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2984004616737366,
      "epoch": 10.91,
      "learning_rate": 5.523191253399388e-05,
      "loss": 0.3715,
      "step": 30200,
      "task_loss": 0.7145050764083862
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3871958255767822,
      "epoch": 10.92,
      "learning_rate": 5.519499748259993e-05,
      "loss": 0.3578,
      "step": 30210,
      "task_loss": 0.8310413360595703
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5353628396987915,
      "epoch": 10.92,
      "learning_rate": 5.515795251732672e-05,
      "loss": 0.4082,
      "step": 30220,
      "task_loss": 0.6640851497650146
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4021126627922058,
      "epoch": 10.93,
      "learning_rate": 5.5120777829190543e-05,
      "loss": 0.3779,
      "step": 30230,
      "task_loss": 0.7048424482345581
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34633779525756836,
      "epoch": 10.93,
      "learning_rate": 5.50834736098766e-05,
      "loss": 0.3843,
      "step": 30240,
      "task_loss": 0.9538134336471558
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3343544006347656,
      "epoch": 10.93,
      "learning_rate": 5.5046040051737976e-05,
      "loss": 0.3904,
      "step": 30250,
      "task_loss": 0.4619567394256592
    },
    {
      "epoch": 10.93,
      "eval_exact_match": 82.96121097445601,
      "eval_f1": 89.80941939893061,
      "step": 30250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.453721821308136,
      "epoch": 10.94,
      "learning_rate": 5.500847734779469e-05,
      "loss": 0.4285,
      "step": 30260,
      "task_loss": 0.5236670970916748
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2954332232475281,
      "epoch": 10.94,
      "learning_rate": 5.497078569173267e-05,
      "loss": 0.3821,
      "step": 30270,
      "task_loss": 0.24834580719470978
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3309508264064789,
      "epoch": 10.94,
      "learning_rate": 5.4932965277902784e-05,
      "loss": 0.4257,
      "step": 30280,
      "task_loss": 0.7076093554496765
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39211854338645935,
      "epoch": 10.95,
      "learning_rate": 5.4895016301319786e-05,
      "loss": 0.3832,
      "step": 30290,
      "task_loss": 0.6536873579025269
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3901174068450928,
      "epoch": 10.95,
      "learning_rate": 5.4856938957661377e-05,
      "loss": 0.4079,
      "step": 30300,
      "task_loss": 0.6007660627365112
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4000302851200104,
      "epoch": 10.95,
      "learning_rate": 5.481873344326713e-05,
      "loss": 0.443,
      "step": 30310,
      "task_loss": 0.7138227224349976
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4617716073989868,
      "epoch": 10.96,
      "learning_rate": 5.478039995513753e-05,
      "loss": 0.4003,
      "step": 30320,
      "task_loss": 0.7568254470825195
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4206737279891968,
      "epoch": 10.96,
      "learning_rate": 5.474193869093293e-05,
      "loss": 0.39,
      "step": 30330,
      "task_loss": 0.7496936917304993
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34480810165405273,
      "epoch": 10.96,
      "learning_rate": 5.4703349848972554e-05,
      "loss": 0.3514,
      "step": 30340,
      "task_loss": 0.44453567266464233
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2850736379623413,
      "epoch": 10.97,
      "learning_rate": 5.466463362823343e-05,
      "loss": 0.3435,
      "step": 30350,
      "task_loss": 0.6559290289878845
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3789885342121124,
      "epoch": 10.97,
      "learning_rate": 5.4625790228349424e-05,
      "loss": 0.3775,
      "step": 30360,
      "task_loss": 1.417602300643921
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37427616119384766,
      "epoch": 10.98,
      "learning_rate": 5.458681984961015e-05,
      "loss": 0.4092,
      "step": 30370,
      "task_loss": 0.6762957572937012
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3822694420814514,
      "epoch": 10.98,
      "learning_rate": 5.4547722692960005e-05,
      "loss": 0.373,
      "step": 30380,
      "task_loss": 0.40661516785621643
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28816527128219604,
      "epoch": 10.98,
      "learning_rate": 5.450849895999707e-05,
      "loss": 0.3938,
      "step": 30390,
      "task_loss": 0.5435937643051147
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43996739387512207,
      "epoch": 10.99,
      "learning_rate": 5.44691488529721e-05,
      "loss": 0.4486,
      "step": 30400,
      "task_loss": 0.7547143697738647
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33609914779663086,
      "epoch": 10.99,
      "learning_rate": 5.442967257478748e-05,
      "loss": 0.3759,
      "step": 30410,
      "task_loss": 0.41951659321784973
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34954917430877686,
      "epoch": 10.99,
      "learning_rate": 5.4390070328996166e-05,
      "loss": 0.3761,
      "step": 30420,
      "task_loss": 0.5150220990180969
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40067267417907715,
      "epoch": 11.0,
      "learning_rate": 5.435034231980067e-05,
      "loss": 0.3763,
      "step": 30430,
      "task_loss": 0.7470228672027588
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2949705123901367,
      "epoch": 11.0,
      "learning_rate": 5.431048875205195e-05,
      "loss": 0.4451,
      "step": 30440,
      "task_loss": 0.6367085576057434
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29878896474838257,
      "epoch": 11.0,
      "learning_rate": 5.427050983124843e-05,
      "loss": 0.3842,
      "step": 30450,
      "task_loss": 0.4922108054161072
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44266384840011597,
      "epoch": 11.01,
      "learning_rate": 5.423040576353483e-05,
      "loss": 0.3627,
      "step": 30460,
      "task_loss": 0.9143210053443909
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29100504517555237,
      "epoch": 11.01,
      "learning_rate": 5.4190176755701236e-05,
      "loss": 0.3293,
      "step": 30470,
      "task_loss": 0.46341586112976074
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38211023807525635,
      "epoch": 11.02,
      "learning_rate": 5.414982301518194e-05,
      "loss": 0.349,
      "step": 30480,
      "task_loss": 0.6259920001029968
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28083837032318115,
      "epoch": 11.02,
      "learning_rate": 5.410934475005439e-05,
      "loss": 0.3214,
      "step": 30490,
      "task_loss": 0.5484700202941895
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3642995059490204,
      "epoch": 11.02,
      "learning_rate": 5.4068742169038124e-05,
      "loss": 0.3605,
      "step": 30500,
      "task_loss": 0.8151582479476929
    },
    {
      "epoch": 11.02,
      "eval_exact_match": 82.61116367076632,
      "eval_f1": 89.53006867143404,
      "step": 30500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29346784949302673,
      "epoch": 11.03,
      "learning_rate": 5.402801548149372e-05,
      "loss": 0.3385,
      "step": 30510,
      "task_loss": 0.5374734997749329
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4059075117111206,
      "epoch": 11.03,
      "learning_rate": 5.398716489742166e-05,
      "loss": 0.3405,
      "step": 30520,
      "task_loss": 0.6881134510040283
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36759153008461,
      "epoch": 11.03,
      "learning_rate": 5.3946190627461314e-05,
      "loss": 0.3453,
      "step": 30530,
      "task_loss": 0.667883038520813
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49028345942497253,
      "epoch": 11.04,
      "learning_rate": 5.390509288288977e-05,
      "loss": 0.3501,
      "step": 30540,
      "task_loss": 0.9341843128204346
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25190505385398865,
      "epoch": 11.04,
      "learning_rate": 5.386387187562082e-05,
      "loss": 0.2926,
      "step": 30550,
      "task_loss": 0.5309723615646362
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.287572979927063,
      "epoch": 11.04,
      "learning_rate": 5.3822527818203865e-05,
      "loss": 0.3514,
      "step": 30560,
      "task_loss": 0.4287169575691223
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25738686323165894,
      "epoch": 11.05,
      "learning_rate": 5.378106092382275e-05,
      "loss": 0.3649,
      "step": 30570,
      "task_loss": 0.6980565786361694
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2965894043445587,
      "epoch": 11.05,
      "learning_rate": 5.3739471406294725e-05,
      "loss": 0.3187,
      "step": 30580,
      "task_loss": 0.43664753437042236
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3387247323989868,
      "epoch": 11.06,
      "learning_rate": 5.369775948006933e-05,
      "loss": 0.335,
      "step": 30590,
      "task_loss": 0.5256164073944092
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3025720715522766,
      "epoch": 11.06,
      "learning_rate": 5.365592536022728e-05,
      "loss": 0.3239,
      "step": 30600,
      "task_loss": 0.5954160094261169
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23759621381759644,
      "epoch": 11.06,
      "learning_rate": 5.361396926247936e-05,
      "loss": 0.3192,
      "step": 30610,
      "task_loss": 0.5290464162826538
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3050764203071594,
      "epoch": 11.07,
      "learning_rate": 5.3571891403165345e-05,
      "loss": 0.3963,
      "step": 30620,
      "task_loss": 0.6192272901535034
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37906011939048767,
      "epoch": 11.07,
      "learning_rate": 5.3529691999252797e-05,
      "loss": 0.367,
      "step": 30630,
      "task_loss": 0.738808274269104
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28784462809562683,
      "epoch": 11.07,
      "learning_rate": 5.348737126833605e-05,
      "loss": 0.3624,
      "step": 30640,
      "task_loss": 0.6045219898223877
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3597484529018402,
      "epoch": 11.08,
      "learning_rate": 5.344492942863501e-05,
      "loss": 0.3306,
      "step": 30650,
      "task_loss": 0.46364304423332214
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28757065534591675,
      "epoch": 11.08,
      "learning_rate": 5.340236669899409e-05,
      "loss": 0.3167,
      "step": 30660,
      "task_loss": 0.5059828162193298
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4648352861404419,
      "epoch": 11.08,
      "learning_rate": 5.336395706279589e-05,
      "loss": 0.369,
      "step": 30670,
      "task_loss": 1.1628811359405518
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3188636302947998,
      "epoch": 11.09,
      "learning_rate": 5.33211652474171e-05,
      "loss": 0.2976,
      "step": 30680,
      "task_loss": 0.8980743885040283
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39204472303390503,
      "epoch": 11.09,
      "learning_rate": 5.327825318026816e-05,
      "loss": 0.3738,
      "step": 30690,
      "task_loss": 0.52778559923172
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3116000294685364,
      "epoch": 11.1,
      "learning_rate": 5.323522108261813e-05,
      "loss": 0.3696,
      "step": 30700,
      "task_loss": 0.4787866771221161
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32855403423309326,
      "epoch": 11.1,
      "learning_rate": 5.319206917635494e-05,
      "loss": 0.3564,
      "step": 30710,
      "task_loss": 0.4829432964324951
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33017492294311523,
      "epoch": 11.1,
      "learning_rate": 5.314879768398437e-05,
      "loss": 0.3741,
      "step": 30720,
      "task_loss": 0.5279051065444946
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2766365706920624,
      "epoch": 11.11,
      "learning_rate": 5.310540682862876e-05,
      "loss": 0.3449,
      "step": 30730,
      "task_loss": 0.4771583378314972
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4740075469017029,
      "epoch": 11.11,
      "learning_rate": 5.306189683402595e-05,
      "loss": 0.3884,
      "step": 30740,
      "task_loss": 0.8812153935432434
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3995136320590973,
      "epoch": 11.11,
      "learning_rate": 5.3018267924528124e-05,
      "loss": 0.3768,
      "step": 30750,
      "task_loss": 0.6199120283126831
    },
    {
      "epoch": 11.11,
      "eval_exact_match": 82.72469252601702,
      "eval_f1": 89.62041686562053,
      "step": 30750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27484428882598877,
      "epoch": 11.12,
      "learning_rate": 5.29745203251006e-05,
      "loss": 0.3488,
      "step": 30760,
      "task_loss": 0.3609499931335449
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37529778480529785,
      "epoch": 11.12,
      "learning_rate": 5.293065426132069e-05,
      "loss": 0.3489,
      "step": 30770,
      "task_loss": 0.41195905208587646
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35930025577545166,
      "epoch": 11.12,
      "learning_rate": 5.28866699593766e-05,
      "loss": 0.3429,
      "step": 30780,
      "task_loss": 0.5321624279022217
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.46943289041519165,
      "epoch": 11.13,
      "learning_rate": 5.284256764606617e-05,
      "loss": 0.3324,
      "step": 30790,
      "task_loss": 0.6085718870162964
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2779424488544464,
      "epoch": 11.13,
      "learning_rate": 5.279834754879575e-05,
      "loss": 0.353,
      "step": 30800,
      "task_loss": 0.33586573600769043
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38808631896972656,
      "epoch": 11.13,
      "learning_rate": 5.275400989557905e-05,
      "loss": 0.346,
      "step": 30810,
      "task_loss": 1.155708909034729
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.412808895111084,
      "epoch": 11.14,
      "learning_rate": 5.270955491503589e-05,
      "loss": 0.3386,
      "step": 30820,
      "task_loss": 0.46611833572387695
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2601992189884186,
      "epoch": 11.14,
      "learning_rate": 5.266498283639113e-05,
      "loss": 0.3435,
      "step": 30830,
      "task_loss": 0.25323614478111267
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47554054856300354,
      "epoch": 11.15,
      "learning_rate": 5.2620293889473387e-05,
      "loss": 0.3747,
      "step": 30840,
      "task_loss": 0.6405948400497437
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38526564836502075,
      "epoch": 11.15,
      "learning_rate": 5.257548830471388e-05,
      "loss": 0.3412,
      "step": 30850,
      "task_loss": 0.32339832186698914
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27890390157699585,
      "epoch": 11.15,
      "learning_rate": 5.25305663131453e-05,
      "loss": 0.3417,
      "step": 30860,
      "task_loss": 0.5287346243858337
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3683282136917114,
      "epoch": 11.16,
      "learning_rate": 5.2485528146400505e-05,
      "loss": 0.3392,
      "step": 30870,
      "task_loss": 0.29345154762268066
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40986067056655884,
      "epoch": 11.16,
      "learning_rate": 5.244037403671146e-05,
      "loss": 0.3427,
      "step": 30880,
      "task_loss": 0.7018733620643616
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3679884970188141,
      "epoch": 11.16,
      "learning_rate": 5.2395104216907926e-05,
      "loss": 0.3656,
      "step": 30890,
      "task_loss": 0.49413740634918213
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3055950999259949,
      "epoch": 11.17,
      "learning_rate": 5.234971892041632e-05,
      "loss": 0.3658,
      "step": 30900,
      "task_loss": 0.9225125312805176
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3304484784603119,
      "epoch": 11.17,
      "learning_rate": 5.230421838125847e-05,
      "loss": 0.3447,
      "step": 30910,
      "task_loss": 0.45518964529037476
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.44711834192276,
      "epoch": 11.17,
      "learning_rate": 5.2263169557434734e-05,
      "loss": 0.3952,
      "step": 30920,
      "task_loss": 0.3594067394733429
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2703867256641388,
      "epoch": 11.18,
      "learning_rate": 5.221745070406853e-05,
      "loss": 0.3152,
      "step": 30930,
      "task_loss": 0.4056801497936249
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2892513573169708,
      "epoch": 11.18,
      "learning_rate": 5.21716172900555e-05,
      "loss": 0.3801,
      "step": 30940,
      "task_loss": 0.38734328746795654
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5101137161254883,
      "epoch": 11.19,
      "learning_rate": 5.2125669551728134e-05,
      "loss": 0.3908,
      "step": 30950,
      "task_loss": 0.9189091920852661
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2777116894721985,
      "epoch": 11.19,
      "learning_rate": 5.207960772600845e-05,
      "loss": 0.3743,
      "step": 30960,
      "task_loss": 0.5410131216049194
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3328074812889099,
      "epoch": 11.19,
      "learning_rate": 5.203343205040667e-05,
      "loss": 0.3374,
      "step": 30970,
      "task_loss": 0.3766447603702545
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4837460517883301,
      "epoch": 11.2,
      "learning_rate": 5.198714276302013e-05,
      "loss": 0.3569,
      "step": 30980,
      "task_loss": 0.7223513722419739
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29842859506607056,
      "epoch": 11.2,
      "learning_rate": 5.194074010253195e-05,
      "loss": 0.3412,
      "step": 30990,
      "task_loss": 0.35386359691619873
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36791229248046875,
      "epoch": 11.2,
      "learning_rate": 5.189422430820986e-05,
      "loss": 0.3493,
      "step": 31000,
      "task_loss": 0.574812650680542
    },
    {
      "epoch": 11.2,
      "eval_exact_match": 82.3841059602649,
      "eval_f1": 89.51719049738145,
      "step": 31000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32887816429138184,
      "epoch": 11.21,
      "learning_rate": 5.1847595619904946e-05,
      "loss": 0.3514,
      "step": 31010,
      "task_loss": 0.3748660683631897
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2734318971633911,
      "epoch": 11.21,
      "learning_rate": 5.180085427805038e-05,
      "loss": 0.3117,
      "step": 31020,
      "task_loss": 0.5383692979812622
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4769822061061859,
      "epoch": 11.21,
      "learning_rate": 5.175400052366024e-05,
      "loss": 0.3758,
      "step": 31030,
      "task_loss": 1.0459678173065186
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4481627941131592,
      "epoch": 11.22,
      "learning_rate": 5.170703459832827e-05,
      "loss": 0.3415,
      "step": 31040,
      "task_loss": 0.712192714214325
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4470297396183014,
      "epoch": 11.22,
      "learning_rate": 5.165995674422654e-05,
      "loss": 0.3709,
      "step": 31050,
      "task_loss": 0.6727391481399536
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3790176510810852,
      "epoch": 11.23,
      "learning_rate": 5.1612767204104316e-05,
      "loss": 0.3639,
      "step": 31060,
      "task_loss": 0.6856481432914734
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3655795156955719,
      "epoch": 11.23,
      "learning_rate": 5.156546622128673e-05,
      "loss": 0.3633,
      "step": 31070,
      "task_loss": 0.962065577507019
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31923535466194153,
      "epoch": 11.23,
      "learning_rate": 5.1518054039673544e-05,
      "loss": 0.3486,
      "step": 31080,
      "task_loss": 0.6942268013954163
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3935302495956421,
      "epoch": 11.24,
      "learning_rate": 5.1470530903737924e-05,
      "loss": 0.3564,
      "step": 31090,
      "task_loss": 0.3999423384666443
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34719374775886536,
      "epoch": 11.24,
      "learning_rate": 5.142289705852514e-05,
      "loss": 0.3651,
      "step": 31100,
      "task_loss": 0.6013666391372681
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34161674976348877,
      "epoch": 11.24,
      "learning_rate": 5.1375152749651315e-05,
      "loss": 0.3532,
      "step": 31110,
      "task_loss": 0.5428796410560608
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3403222858905792,
      "epoch": 11.25,
      "learning_rate": 5.132729822330215e-05,
      "loss": 0.3699,
      "step": 31120,
      "task_loss": 0.34771913290023804
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27695637941360474,
      "epoch": 11.25,
      "learning_rate": 5.1279333726231676e-05,
      "loss": 0.3289,
      "step": 31130,
      "task_loss": 0.3937012553215027
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3218735456466675,
      "epoch": 11.25,
      "learning_rate": 5.123125950576098e-05,
      "loss": 0.3141,
      "step": 31140,
      "task_loss": 0.4355730414390564
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26944002509117126,
      "epoch": 11.26,
      "learning_rate": 5.11830758097769e-05,
      "loss": 0.3284,
      "step": 31150,
      "task_loss": 0.3624120354652405
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3208332061767578,
      "epoch": 11.26,
      "learning_rate": 5.113478288673078e-05,
      "loss": 0.3409,
      "step": 31160,
      "task_loss": 0.4535531997680664
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25942161679267883,
      "epoch": 11.26,
      "learning_rate": 5.108638098563717e-05,
      "loss": 0.3449,
      "step": 31170,
      "task_loss": 0.2241821587085724
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4324563145637512,
      "epoch": 11.27,
      "learning_rate": 5.1037870356072545e-05,
      "loss": 0.3713,
      "step": 31180,
      "task_loss": 0.6393023729324341
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4508543014526367,
      "epoch": 11.27,
      "learning_rate": 5.0989251248174015e-05,
      "loss": 0.3414,
      "step": 31190,
      "task_loss": 0.34567761421203613
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32019156217575073,
      "epoch": 11.28,
      "learning_rate": 5.094052391263807e-05,
      "loss": 0.3471,
      "step": 31200,
      "task_loss": 0.3136403262615204
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3434525728225708,
      "epoch": 11.28,
      "learning_rate": 5.0891688600719215e-05,
      "loss": 0.3478,
      "step": 31210,
      "task_loss": 0.7961330413818359
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3131732642650604,
      "epoch": 11.28,
      "learning_rate": 5.084274556422876e-05,
      "loss": 0.3376,
      "step": 31220,
      "task_loss": 0.5231847167015076
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37508511543273926,
      "epoch": 11.29,
      "learning_rate": 5.0793695055533446e-05,
      "loss": 0.3595,
      "step": 31230,
      "task_loss": 0.5824639797210693
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43298447132110596,
      "epoch": 11.29,
      "learning_rate": 5.074453732755419e-05,
      "loss": 0.3633,
      "step": 31240,
      "task_loss": 0.7249249219894409
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3672495484352112,
      "epoch": 11.29,
      "learning_rate": 5.069527263376478e-05,
      "loss": 0.3543,
      "step": 31250,
      "task_loss": 0.9668033123016357
    },
    {
      "epoch": 11.29,
      "eval_exact_match": 82.45033112582782,
      "eval_f1": 89.4760442654279,
      "step": 31250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3237718343734741,
      "epoch": 11.3,
      "learning_rate": 5.0645901228190525e-05,
      "loss": 0.3553,
      "step": 31260,
      "task_loss": 0.513779878616333
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36057525873184204,
      "epoch": 11.3,
      "learning_rate": 5.059642336540702e-05,
      "loss": 0.3848,
      "step": 31270,
      "task_loss": 0.2133713960647583
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.294028639793396,
      "epoch": 11.3,
      "learning_rate": 5.054683930053873e-05,
      "loss": 0.3499,
      "step": 31280,
      "task_loss": 0.3805010914802551
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24722647666931152,
      "epoch": 11.31,
      "learning_rate": 5.0497149289257796e-05,
      "loss": 0.3333,
      "step": 31290,
      "task_loss": 0.3370016813278198
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4259105324745178,
      "epoch": 11.31,
      "learning_rate": 5.044735358778261e-05,
      "loss": 0.3769,
      "step": 31300,
      "task_loss": 0.6986902952194214
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26834627985954285,
      "epoch": 11.32,
      "learning_rate": 5.039745245287655e-05,
      "loss": 0.3284,
      "step": 31310,
      "task_loss": 0.721771240234375
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3485707640647888,
      "epoch": 11.32,
      "learning_rate": 5.0347446141846656e-05,
      "loss": 0.345,
      "step": 31320,
      "task_loss": 0.5485772490501404
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38015708327293396,
      "epoch": 11.32,
      "learning_rate": 5.029733491254229e-05,
      "loss": 0.3858,
      "step": 31330,
      "task_loss": 0.9029421806335449
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3657987713813782,
      "epoch": 11.33,
      "learning_rate": 5.024711902335379e-05,
      "loss": 0.3339,
      "step": 31340,
      "task_loss": 0.5325937271118164
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28831857442855835,
      "epoch": 11.33,
      "learning_rate": 5.019679873321117e-05,
      "loss": 0.3236,
      "step": 31350,
      "task_loss": 0.4304049015045166
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.373060405254364,
      "epoch": 11.33,
      "learning_rate": 5.014637430158276e-05,
      "loss": 0.381,
      "step": 31360,
      "task_loss": 0.4949566423892975
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4010653495788574,
      "epoch": 11.34,
      "learning_rate": 5.0095845988473896e-05,
      "loss": 0.3504,
      "step": 31370,
      "task_loss": 0.5861020088195801
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3670312762260437,
      "epoch": 11.34,
      "learning_rate": 5.0045214054425545e-05,
      "loss": 0.3616,
      "step": 31380,
      "task_loss": 0.5232160091400146
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2984731197357178,
      "epoch": 11.34,
      "learning_rate": 4.9994478760512994e-05,
      "loss": 0.3671,
      "step": 31390,
      "task_loss": 0.5723747611045837
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3843269348144531,
      "epoch": 11.35,
      "learning_rate": 4.9943640368344464e-05,
      "loss": 0.3533,
      "step": 31400,
      "task_loss": 0.664061427116394
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2829972505569458,
      "epoch": 11.35,
      "learning_rate": 4.989269914005981e-05,
      "loss": 0.3357,
      "step": 31410,
      "task_loss": 0.541733980178833
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2264384627342224,
      "epoch": 11.36,
      "learning_rate": 4.984165533832913e-05,
      "loss": 0.3432,
      "step": 31420,
      "task_loss": 0.3720897436141968
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3180156350135803,
      "epoch": 11.36,
      "learning_rate": 4.979050922635144e-05,
      "loss": 0.365,
      "step": 31430,
      "task_loss": 0.7879629135131836
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32890772819519043,
      "epoch": 11.36,
      "learning_rate": 4.973926106785329e-05,
      "loss": 0.3623,
      "step": 31440,
      "task_loss": 0.5077107548713684
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35433271527290344,
      "epoch": 11.37,
      "learning_rate": 4.96879111270874e-05,
      "loss": 0.4012,
      "step": 31450,
      "task_loss": 0.822060227394104
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28777164220809937,
      "epoch": 11.37,
      "learning_rate": 4.9636459668831357e-05,
      "loss": 0.3549,
      "step": 31460,
      "task_loss": 0.5379538536071777
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34426411986351013,
      "epoch": 11.37,
      "learning_rate": 4.958490695838617e-05,
      "loss": 0.3264,
      "step": 31470,
      "task_loss": 0.3982798457145691
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4349164366722107,
      "epoch": 11.38,
      "learning_rate": 4.953325326157494e-05,
      "loss": 0.3671,
      "step": 31480,
      "task_loss": 0.3434685468673706
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33134931325912476,
      "epoch": 11.38,
      "learning_rate": 4.948149884474152e-05,
      "loss": 0.3144,
      "step": 31490,
      "task_loss": 0.5727837085723877
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30214446783065796,
      "epoch": 11.38,
      "learning_rate": 4.942964397474906e-05,
      "loss": 0.3607,
      "step": 31500,
      "task_loss": 0.4910222291946411
    },
    {
      "epoch": 11.38,
      "eval_exact_match": 82.57332071901608,
      "eval_f1": 89.62623525440266,
      "step": 31500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2808883786201477,
      "epoch": 11.39,
      "learning_rate": 4.937768891897872e-05,
      "loss": 0.3645,
      "step": 31510,
      "task_loss": 0.4864407777786255
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3554116487503052,
      "epoch": 11.39,
      "learning_rate": 4.9325633945328245e-05,
      "loss": 0.3451,
      "step": 31520,
      "task_loss": 0.6188965439796448
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31889331340789795,
      "epoch": 11.4,
      "learning_rate": 4.927347932221058e-05,
      "loss": 0.3408,
      "step": 31530,
      "task_loss": 0.4872661232948303
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4375583529472351,
      "epoch": 11.4,
      "learning_rate": 4.922122531855249e-05,
      "loss": 0.3419,
      "step": 31540,
      "task_loss": 0.7491167187690735
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26411664485931396,
      "epoch": 11.4,
      "learning_rate": 4.916887220379319e-05,
      "loss": 0.3321,
      "step": 31550,
      "task_loss": 0.3491521179676056
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2768267095088959,
      "epoch": 11.41,
      "learning_rate": 4.911642024788296e-05,
      "loss": 0.3543,
      "step": 31560,
      "task_loss": 0.4296256899833679
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3377191722393036,
      "epoch": 11.41,
      "learning_rate": 4.90638697212817e-05,
      "loss": 0.3446,
      "step": 31570,
      "task_loss": 0.6762540936470032
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29912063479423523,
      "epoch": 11.41,
      "learning_rate": 4.901122089495762e-05,
      "loss": 0.3145,
      "step": 31580,
      "task_loss": 0.6021625995635986
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.41637420654296875,
      "epoch": 11.42,
      "learning_rate": 4.895847404038576e-05,
      "loss": 0.3784,
      "step": 31590,
      "task_loss": 0.5565124154090881
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42575061321258545,
      "epoch": 11.42,
      "learning_rate": 4.890562942954664e-05,
      "loss": 0.3568,
      "step": 31600,
      "task_loss": 0.6394255757331848
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38120606541633606,
      "epoch": 11.42,
      "learning_rate": 4.885268733492484e-05,
      "loss": 0.3373,
      "step": 31610,
      "task_loss": 0.5133061408996582
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29386794567108154,
      "epoch": 11.43,
      "learning_rate": 4.87996480295076e-05,
      "loss": 0.3409,
      "step": 31620,
      "task_loss": 0.475729376077652
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3488690257072449,
      "epoch": 11.43,
      "learning_rate": 4.8746511786783434e-05,
      "loss": 0.3525,
      "step": 31630,
      "task_loss": 0.7107182145118713
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32025083899497986,
      "epoch": 11.43,
      "learning_rate": 4.8693278880740654e-05,
      "loss": 0.3548,
      "step": 31640,
      "task_loss": 0.4914471507072449
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3770049214363098,
      "epoch": 11.44,
      "learning_rate": 4.863994958586604e-05,
      "loss": 0.3559,
      "step": 31650,
      "task_loss": 0.6165828704833984
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26748329401016235,
      "epoch": 11.44,
      "learning_rate": 4.858652417714335e-05,
      "loss": 0.3236,
      "step": 31660,
      "task_loss": 0.08374525606632233
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27789509296417236,
      "epoch": 11.45,
      "learning_rate": 4.853300293005198e-05,
      "loss": 0.3334,
      "step": 31670,
      "task_loss": 0.46393516659736633
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3311939239501953,
      "epoch": 11.45,
      "learning_rate": 4.847938612056547e-05,
      "loss": 0.3449,
      "step": 31680,
      "task_loss": 0.7730482816696167
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.395574152469635,
      "epoch": 11.45,
      "learning_rate": 4.842567402515011e-05,
      "loss": 0.3831,
      "step": 31690,
      "task_loss": 1.0190998315811157
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33000457286834717,
      "epoch": 11.46,
      "learning_rate": 4.837186692076353e-05,
      "loss": 0.3674,
      "step": 31700,
      "task_loss": 1.0033828020095825
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29676058888435364,
      "epoch": 11.46,
      "learning_rate": 4.831796508485326e-05,
      "loss": 0.419,
      "step": 31710,
      "task_loss": 0.2888988256454468
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33335405588150024,
      "epoch": 11.46,
      "learning_rate": 4.8263968795355294e-05,
      "loss": 0.3478,
      "step": 31720,
      "task_loss": 0.5477027893066406
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2935606837272644,
      "epoch": 11.47,
      "learning_rate": 4.820987833069265e-05,
      "loss": 0.3257,
      "step": 31730,
      "task_loss": 0.8460813164710999
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.355922669172287,
      "epoch": 11.47,
      "learning_rate": 4.8155693969773935e-05,
      "loss": 0.3805,
      "step": 31740,
      "task_loss": 0.4853879511356354
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30375123023986816,
      "epoch": 11.47,
      "learning_rate": 4.8101415991991965e-05,
      "loss": 0.3192,
      "step": 31750,
      "task_loss": 0.6044105291366577
    },
    {
      "epoch": 11.47,
      "eval_exact_match": 82.60170293282876,
      "eval_f1": 89.55296899531102,
      "step": 31750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39717626571655273,
      "epoch": 11.48,
      "learning_rate": 4.804704467722223e-05,
      "loss": 0.3636,
      "step": 31760,
      "task_loss": 0.6207161545753479
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49976447224617004,
      "epoch": 11.48,
      "learning_rate": 4.799258030582152e-05,
      "loss": 0.3847,
      "step": 31770,
      "task_loss": 0.6690294742584229
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3473104238510132,
      "epoch": 11.49,
      "learning_rate": 4.793802315862644e-05,
      "loss": 0.3781,
      "step": 31780,
      "task_loss": 0.3822641968727112
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2778242826461792,
      "epoch": 11.49,
      "learning_rate": 4.788337351695199e-05,
      "loss": 0.3443,
      "step": 31790,
      "task_loss": 0.7948096990585327
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3013157248497009,
      "epoch": 11.49,
      "learning_rate": 4.78286316625901e-05,
      "loss": 0.3315,
      "step": 31800,
      "task_loss": 0.8763753175735474
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28930824995040894,
      "epoch": 11.5,
      "learning_rate": 4.777379787780818e-05,
      "loss": 0.361,
      "step": 31810,
      "task_loss": 0.3076488971710205
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38833189010620117,
      "epoch": 11.5,
      "learning_rate": 4.771887244534766e-05,
      "loss": 0.3049,
      "step": 31820,
      "task_loss": 0.661001443862915
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3576197326183319,
      "epoch": 11.5,
      "learning_rate": 4.766385564842255e-05,
      "loss": 0.3281,
      "step": 31830,
      "task_loss": 0.7011573314666748
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25735771656036377,
      "epoch": 11.51,
      "learning_rate": 4.760874777071794e-05,
      "loss": 0.3408,
      "step": 31840,
      "task_loss": 0.612379789352417
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2917349934577942,
      "epoch": 11.51,
      "learning_rate": 4.755354909638858e-05,
      "loss": 0.3483,
      "step": 31850,
      "task_loss": 0.38969987630844116
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2821897864341736,
      "epoch": 11.51,
      "learning_rate": 4.74982599100574e-05,
      "loss": 0.3279,
      "step": 31860,
      "task_loss": 0.46879321336746216
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38328707218170166,
      "epoch": 11.52,
      "learning_rate": 4.744288049681405e-05,
      "loss": 0.3868,
      "step": 31870,
      "task_loss": 0.33758458495140076
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3531540036201477,
      "epoch": 11.52,
      "learning_rate": 4.738741114221338e-05,
      "loss": 0.3493,
      "step": 31880,
      "task_loss": 0.6903041005134583
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5110961198806763,
      "epoch": 11.53,
      "learning_rate": 4.733185213227406e-05,
      "loss": 0.3551,
      "step": 31890,
      "task_loss": 0.7856677770614624
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.320343017578125,
      "epoch": 11.53,
      "learning_rate": 4.7276203753477014e-05,
      "loss": 0.3306,
      "step": 31900,
      "task_loss": 0.22347959876060486
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3419710099697113,
      "epoch": 11.53,
      "learning_rate": 4.7220466292763996e-05,
      "loss": 0.3564,
      "step": 31910,
      "task_loss": 0.6967019438743591
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3632848560810089,
      "epoch": 11.54,
      "learning_rate": 4.7164640037536104e-05,
      "loss": 0.3353,
      "step": 31920,
      "task_loss": 0.6160529851913452
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36113396286964417,
      "epoch": 11.54,
      "learning_rate": 4.7108725275652284e-05,
      "loss": 0.3068,
      "step": 31930,
      "task_loss": 0.5054973363876343
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.315176784992218,
      "epoch": 11.54,
      "learning_rate": 4.7052722295427844e-05,
      "loss": 0.3345,
      "step": 31940,
      "task_loss": 0.4096016585826874
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27597475051879883,
      "epoch": 11.55,
      "learning_rate": 4.6996631385632974e-05,
      "loss": 0.3571,
      "step": 31950,
      "task_loss": 0.5770153999328613
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3908612132072449,
      "epoch": 11.55,
      "learning_rate": 4.6940452835491274e-05,
      "loss": 0.3732,
      "step": 31960,
      "task_loss": 0.758826732635498
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2762240171432495,
      "epoch": 11.55,
      "learning_rate": 4.688418693467824e-05,
      "loss": 0.3733,
      "step": 31970,
      "task_loss": 0.26867347955703735
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4848494529724121,
      "epoch": 11.56,
      "learning_rate": 4.682783397331978e-05,
      "loss": 0.3885,
      "step": 31980,
      "task_loss": 0.517472505569458
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3382634222507477,
      "epoch": 11.56,
      "learning_rate": 4.677139424199072e-05,
      "loss": 0.3462,
      "step": 31990,
      "task_loss": 0.5025910139083862
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28954753279685974,
      "epoch": 11.56,
      "learning_rate": 4.671486803171327e-05,
      "loss": 0.376,
      "step": 32000,
      "task_loss": 0.3863941431045532
    },
    {
      "epoch": 11.56,
      "eval_exact_match": 82.68684957426679,
      "eval_f1": 89.58686709901265,
      "step": 32000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34954577684402466,
      "epoch": 11.57,
      "learning_rate": 4.665825563395559e-05,
      "loss": 0.3468,
      "step": 32010,
      "task_loss": 0.48891741037368774
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24595540761947632,
      "epoch": 11.57,
      "learning_rate": 4.6601557340630255e-05,
      "loss": 0.3359,
      "step": 32020,
      "task_loss": 0.3322390615940094
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3430486023426056,
      "epoch": 11.58,
      "learning_rate": 4.654477344409271e-05,
      "loss": 0.376,
      "step": 32030,
      "task_loss": 0.38788700103759766
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35017162561416626,
      "epoch": 11.58,
      "learning_rate": 4.648790423713982e-05,
      "loss": 0.3646,
      "step": 32040,
      "task_loss": 0.512260913848877
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37731289863586426,
      "epoch": 11.58,
      "learning_rate": 4.643095001300836e-05,
      "loss": 0.3785,
      "step": 32050,
      "task_loss": 0.5028886198997498
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3077854514122009,
      "epoch": 11.59,
      "learning_rate": 4.637391106537343e-05,
      "loss": 0.3383,
      "step": 32060,
      "task_loss": 0.4493955969810486
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2899521291255951,
      "epoch": 11.59,
      "learning_rate": 4.631678768834706e-05,
      "loss": 0.339,
      "step": 32070,
      "task_loss": 0.5738735198974609
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.42707955837249756,
      "epoch": 11.59,
      "learning_rate": 4.625958017647657e-05,
      "loss": 0.3544,
      "step": 32080,
      "task_loss": 0.5231696367263794
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34892863035202026,
      "epoch": 11.6,
      "learning_rate": 4.620228882474312e-05,
      "loss": 0.3458,
      "step": 32090,
      "task_loss": 0.6890476942062378
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29222404956817627,
      "epoch": 11.6,
      "learning_rate": 4.61449139285602e-05,
      "loss": 0.3261,
      "step": 32100,
      "task_loss": 0.31315627694129944
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23572243750095367,
      "epoch": 11.6,
      "learning_rate": 4.608745578377208e-05,
      "loss": 0.346,
      "step": 32110,
      "task_loss": 0.3648180663585663
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29230839014053345,
      "epoch": 11.61,
      "learning_rate": 4.602991468665224e-05,
      "loss": 0.3495,
      "step": 32120,
      "task_loss": 0.45920950174331665
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32993003726005554,
      "epoch": 11.61,
      "learning_rate": 4.597229093390195e-05,
      "loss": 0.3658,
      "step": 32130,
      "task_loss": 0.6793662309646606
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45977842807769775,
      "epoch": 11.62,
      "learning_rate": 4.591458482264866e-05,
      "loss": 0.3503,
      "step": 32140,
      "task_loss": 0.6161090731620789
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2830163836479187,
      "epoch": 11.62,
      "learning_rate": 4.585679665044447e-05,
      "loss": 0.3435,
      "step": 32150,
      "task_loss": 0.5471289157867432
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32702142000198364,
      "epoch": 11.62,
      "learning_rate": 4.579892671526463e-05,
      "loss": 0.3666,
      "step": 32160,
      "task_loss": 0.3628914952278137
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38225293159484863,
      "epoch": 11.63,
      "learning_rate": 4.5740975315505996e-05,
      "loss": 0.3323,
      "step": 32170,
      "task_loss": 0.8414508104324341
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5750049352645874,
      "epoch": 11.63,
      "learning_rate": 4.5682942749985466e-05,
      "loss": 0.3754,
      "step": 32180,
      "task_loss": 0.532424807548523
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3244228959083557,
      "epoch": 11.63,
      "learning_rate": 4.562482931793846e-05,
      "loss": 0.3462,
      "step": 32190,
      "task_loss": 0.7852087616920471
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39638885855674744,
      "epoch": 11.64,
      "learning_rate": 4.5566635319017376e-05,
      "loss": 0.3636,
      "step": 32200,
      "task_loss": 0.9977461099624634
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25993040204048157,
      "epoch": 11.64,
      "learning_rate": 4.550836105329005e-05,
      "loss": 0.3724,
      "step": 32210,
      "task_loss": 0.39915382862091064
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27179035544395447,
      "epoch": 11.64,
      "learning_rate": 4.545000682123818e-05,
      "loss": 0.3409,
      "step": 32220,
      "task_loss": 0.43176907300949097
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29763340950012207,
      "epoch": 11.65,
      "learning_rate": 4.539157292375581e-05,
      "loss": 0.3388,
      "step": 32230,
      "task_loss": 0.5615702867507935
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3821195662021637,
      "epoch": 11.65,
      "learning_rate": 4.5333059662147775e-05,
      "loss": 0.3492,
      "step": 32240,
      "task_loss": 0.5893045663833618
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3731650710105896,
      "epoch": 11.66,
      "learning_rate": 4.527446733812812e-05,
      "loss": 0.3697,
      "step": 32250,
      "task_loss": 0.7143180966377258
    },
    {
      "epoch": 11.66,
      "eval_exact_match": 82.80037842951751,
      "eval_f1": 89.72915392440778,
      "step": 32250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3984615206718445,
      "epoch": 11.66,
      "learning_rate": 4.521579625381856e-05,
      "loss": 0.3311,
      "step": 32260,
      "task_loss": 0.6453840732574463
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3678897023200989,
      "epoch": 11.66,
      "learning_rate": 4.515704671174696e-05,
      "loss": 0.3751,
      "step": 32270,
      "task_loss": 0.7683279514312744
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3974093198776245,
      "epoch": 11.67,
      "learning_rate": 4.509821901484569e-05,
      "loss": 0.3603,
      "step": 32280,
      "task_loss": 0.5658100247383118
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2496042549610138,
      "epoch": 11.67,
      "learning_rate": 4.503931346645016e-05,
      "loss": 0.3419,
      "step": 32290,
      "task_loss": 0.3385689854621887
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27575236558914185,
      "epoch": 11.67,
      "learning_rate": 4.498033037029717e-05,
      "loss": 0.3434,
      "step": 32300,
      "task_loss": 0.25666338205337524
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3864726424217224,
      "epoch": 11.68,
      "learning_rate": 4.49212700305234e-05,
      "loss": 0.3361,
      "step": 32310,
      "task_loss": 0.4089565873146057
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3019556999206543,
      "epoch": 11.68,
      "learning_rate": 4.486213275166382e-05,
      "loss": 0.38,
      "step": 32320,
      "task_loss": 0.3994622230529785
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22430866956710815,
      "epoch": 11.68,
      "learning_rate": 4.480291883865013e-05,
      "loss": 0.3506,
      "step": 32330,
      "task_loss": 0.40575936436653137
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31596285104751587,
      "epoch": 11.69,
      "learning_rate": 4.474362859680918e-05,
      "loss": 0.3414,
      "step": 32340,
      "task_loss": 0.5828789472579956
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35511207580566406,
      "epoch": 11.69,
      "learning_rate": 4.468426233186139e-05,
      "loss": 0.3421,
      "step": 32350,
      "task_loss": 0.6438747644424438
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3512818217277527,
      "epoch": 11.69,
      "learning_rate": 4.4624820349919185e-05,
      "loss": 0.3627,
      "step": 32360,
      "task_loss": 0.7819494009017944
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5698807239532471,
      "epoch": 11.7,
      "learning_rate": 4.456530295748541e-05,
      "loss": 0.375,
      "step": 32370,
      "task_loss": 1.0448747873306274
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36504194140434265,
      "epoch": 11.7,
      "learning_rate": 4.450571046145175e-05,
      "loss": 0.3415,
      "step": 32380,
      "task_loss": 0.3004506826400757
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3651716411113739,
      "epoch": 11.71,
      "learning_rate": 4.444604316909717e-05,
      "loss": 0.366,
      "step": 32390,
      "task_loss": 0.4216272234916687
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3165162205696106,
      "epoch": 11.71,
      "learning_rate": 4.438630138808628e-05,
      "loss": 0.3705,
      "step": 32400,
      "task_loss": 0.45583444833755493
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24389581382274628,
      "epoch": 11.71,
      "learning_rate": 4.4326485426467796e-05,
      "loss": 0.3477,
      "step": 32410,
      "task_loss": 0.5839788913726807
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34841057658195496,
      "epoch": 11.72,
      "learning_rate": 4.426659559267294e-05,
      "loss": 0.3751,
      "step": 32420,
      "task_loss": 0.4604908227920532
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2817525565624237,
      "epoch": 11.72,
      "learning_rate": 4.420663219551381e-05,
      "loss": 0.3287,
      "step": 32430,
      "task_loss": 0.4700966477394104
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4372870624065399,
      "epoch": 11.72,
      "learning_rate": 4.414659554418186e-05,
      "loss": 0.3625,
      "step": 32440,
      "task_loss": 1.404410481452942
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29228758811950684,
      "epoch": 11.73,
      "learning_rate": 4.408648594824626e-05,
      "loss": 0.3602,
      "step": 32450,
      "task_loss": 0.46710091829299927
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3614046573638916,
      "epoch": 11.73,
      "learning_rate": 4.402630371765228e-05,
      "loss": 0.3403,
      "step": 32460,
      "task_loss": 0.9066839814186096
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4615110158920288,
      "epoch": 11.73,
      "learning_rate": 4.396604916271977e-05,
      "loss": 0.4101,
      "step": 32470,
      "task_loss": 0.623814582824707
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3592243194580078,
      "epoch": 11.74,
      "learning_rate": 4.3905722594141444e-05,
      "loss": 0.3632,
      "step": 32480,
      "task_loss": 0.442016065120697
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2413320541381836,
      "epoch": 11.74,
      "learning_rate": 4.384532432298139e-05,
      "loss": 0.3191,
      "step": 32490,
      "task_loss": 0.45909905433654785
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30714505910873413,
      "epoch": 11.75,
      "learning_rate": 4.37848546606734e-05,
      "loss": 0.3199,
      "step": 32500,
      "task_loss": 0.4241373538970947
    },
    {
      "epoch": 11.75,
      "eval_exact_match": 82.95175023651845,
      "eval_f1": 89.72995053074105,
      "step": 32500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34651437401771545,
      "epoch": 11.75,
      "learning_rate": 4.372431391901939e-05,
      "loss": 0.324,
      "step": 32510,
      "task_loss": 0.47919905185699463
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3726678788661957,
      "epoch": 11.75,
      "learning_rate": 4.3663702410187786e-05,
      "loss": 0.3411,
      "step": 32520,
      "task_loss": 0.3535993993282318
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25239893794059753,
      "epoch": 11.76,
      "learning_rate": 4.36030204467119e-05,
      "loss": 0.3611,
      "step": 32530,
      "task_loss": 0.40208229422569275
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3526584506034851,
      "epoch": 11.76,
      "learning_rate": 4.354226834148834e-05,
      "loss": 0.3648,
      "step": 32540,
      "task_loss": 0.5418181419372559
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.442671537399292,
      "epoch": 11.76,
      "learning_rate": 4.34814464077754e-05,
      "loss": 0.3529,
      "step": 32550,
      "task_loss": 0.5203021764755249
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3775908350944519,
      "epoch": 11.77,
      "learning_rate": 4.3420554959191415e-05,
      "loss": 0.3386,
      "step": 32560,
      "task_loss": 0.4405955672264099
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39824825525283813,
      "epoch": 11.77,
      "learning_rate": 4.335959430971318e-05,
      "loss": 0.3693,
      "step": 32570,
      "task_loss": 0.5429366230964661
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3338078260421753,
      "epoch": 11.77,
      "learning_rate": 4.3298564773674284e-05,
      "loss": 0.3439,
      "step": 32580,
      "task_loss": 0.3427931070327759
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27158862352371216,
      "epoch": 11.78,
      "learning_rate": 4.3237466665763554e-05,
      "loss": 0.3539,
      "step": 32590,
      "task_loss": 0.9125320911407471
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28898924589157104,
      "epoch": 11.78,
      "learning_rate": 4.317630030102336e-05,
      "loss": 0.3263,
      "step": 32600,
      "task_loss": 0.6775388717651367
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4036529064178467,
      "epoch": 11.79,
      "learning_rate": 4.311506599484807e-05,
      "loss": 0.3472,
      "step": 32610,
      "task_loss": 0.6710329055786133
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26878538727760315,
      "epoch": 11.79,
      "learning_rate": 4.3053764062982354e-05,
      "loss": 0.3293,
      "step": 32620,
      "task_loss": 0.46130526065826416
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2734406590461731,
      "epoch": 11.79,
      "learning_rate": 4.299239482151957e-05,
      "loss": 0.3186,
      "step": 32630,
      "task_loss": 0.4390842318534851
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30930137634277344,
      "epoch": 11.8,
      "learning_rate": 4.293095858690018e-05,
      "loss": 0.3809,
      "step": 32640,
      "task_loss": 0.4176652431488037
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24716287851333618,
      "epoch": 11.8,
      "learning_rate": 4.286945567591006e-05,
      "loss": 0.3327,
      "step": 32650,
      "task_loss": 0.5557129979133606
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38499587774276733,
      "epoch": 11.8,
      "learning_rate": 4.2807886405678904e-05,
      "loss": 0.3916,
      "step": 32660,
      "task_loss": 0.5830663442611694
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36196571588516235,
      "epoch": 11.81,
      "learning_rate": 4.274625109367857e-05,
      "loss": 0.35,
      "step": 32670,
      "task_loss": 0.6983354687690735
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.330838143825531,
      "epoch": 11.81,
      "learning_rate": 4.268455005772145e-05,
      "loss": 0.3793,
      "step": 32680,
      "task_loss": 0.48391640186309814
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4563385844230652,
      "epoch": 11.81,
      "learning_rate": 4.262278361595885e-05,
      "loss": 0.3782,
      "step": 32690,
      "task_loss": 0.7140378952026367
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3044314682483673,
      "epoch": 11.82,
      "learning_rate": 4.25609520868793e-05,
      "loss": 0.3713,
      "step": 32700,
      "task_loss": 0.5128190517425537
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30874133110046387,
      "epoch": 11.82,
      "learning_rate": 4.249905578930695e-05,
      "loss": 0.3483,
      "step": 32710,
      "task_loss": 0.3443719148635864
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3269858658313751,
      "epoch": 11.83,
      "learning_rate": 4.2437095042399935e-05,
      "loss": 0.3142,
      "step": 32720,
      "task_loss": 0.5143279433250427
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3339177370071411,
      "epoch": 11.83,
      "learning_rate": 4.2375070165648695e-05,
      "loss": 0.3226,
      "step": 32730,
      "task_loss": 0.38503554463386536
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35822218656539917,
      "epoch": 11.83,
      "learning_rate": 4.231298147887436e-05,
      "loss": 0.3998,
      "step": 32740,
      "task_loss": 0.5768170356750488
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32746079564094543,
      "epoch": 11.84,
      "learning_rate": 4.225082930222706e-05,
      "loss": 0.342,
      "step": 32750,
      "task_loss": 0.8048028945922852
    },
    {
      "epoch": 11.84,
      "eval_exact_match": 82.92336802270577,
      "eval_f1": 89.59989360234766,
      "step": 32750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4053548276424408,
      "epoch": 11.84,
      "learning_rate": 4.218861395618434e-05,
      "loss": 0.322,
      "step": 32760,
      "task_loss": 0.5083299875259399
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34768015146255493,
      "epoch": 11.84,
      "learning_rate": 4.212633576154943e-05,
      "loss": 0.3578,
      "step": 32770,
      "task_loss": 0.8128582239151001
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33670443296432495,
      "epoch": 11.85,
      "learning_rate": 4.206399503944965e-05,
      "loss": 0.3744,
      "step": 32780,
      "task_loss": 0.5336966514587402
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37081536650657654,
      "epoch": 11.85,
      "learning_rate": 4.200159211133473e-05,
      "loss": 0.3778,
      "step": 32790,
      "task_loss": 0.44150346517562866
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38825803995132446,
      "epoch": 11.85,
      "learning_rate": 4.193912729897513e-05,
      "loss": 0.3604,
      "step": 32800,
      "task_loss": 0.6093878149986267
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30060669779777527,
      "epoch": 11.86,
      "learning_rate": 4.187660092446045e-05,
      "loss": 0.3095,
      "step": 32810,
      "task_loss": 0.5632023811340332
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27266132831573486,
      "epoch": 11.86,
      "learning_rate": 4.18140133101977e-05,
      "loss": 0.3467,
      "step": 32820,
      "task_loss": 0.47467291355133057
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3054879307746887,
      "epoch": 11.86,
      "learning_rate": 4.1751364778909655e-05,
      "loss": 0.3512,
      "step": 32830,
      "task_loss": 0.48678484559059143
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32673510909080505,
      "epoch": 11.87,
      "learning_rate": 4.1688655653633224e-05,
      "loss": 0.3892,
      "step": 32840,
      "task_loss": 0.4992232024669647
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3742809295654297,
      "epoch": 11.87,
      "learning_rate": 4.162588625771773e-05,
      "loss": 0.3305,
      "step": 32850,
      "task_loss": 0.8625702857971191
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.423870325088501,
      "epoch": 11.88,
      "learning_rate": 4.156305691482328e-05,
      "loss": 0.3461,
      "step": 32860,
      "task_loss": 0.5573567748069763
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45885270833969116,
      "epoch": 11.88,
      "learning_rate": 4.150016794891912e-05,
      "loss": 0.3838,
      "step": 32870,
      "task_loss": 0.5527869462966919
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23280248045921326,
      "epoch": 11.88,
      "learning_rate": 4.143721968428188e-05,
      "loss": 0.3861,
      "step": 32880,
      "task_loss": 0.41633132100105286
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3545595705509186,
      "epoch": 11.89,
      "learning_rate": 4.1374212445494e-05,
      "loss": 0.336,
      "step": 32890,
      "task_loss": 0.5708621740341187
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3292989730834961,
      "epoch": 11.89,
      "learning_rate": 4.131114655744196e-05,
      "loss": 0.3401,
      "step": 32900,
      "task_loss": 0.6946372985839844
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.43926069140434265,
      "epoch": 11.89,
      "learning_rate": 4.124802234531472e-05,
      "loss": 0.3562,
      "step": 32910,
      "task_loss": 0.786125123500824
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3756178021430969,
      "epoch": 11.9,
      "learning_rate": 4.118484013460192e-05,
      "loss": 0.3703,
      "step": 32920,
      "task_loss": 0.4792003035545349
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5044987797737122,
      "epoch": 11.9,
      "learning_rate": 4.112160025109231e-05,
      "loss": 0.3538,
      "step": 32930,
      "task_loss": 0.7191258072853088
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2991417944431305,
      "epoch": 11.9,
      "learning_rate": 4.1058303020871964e-05,
      "loss": 0.3681,
      "step": 32940,
      "task_loss": 0.38584643602371216
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3430123031139374,
      "epoch": 11.91,
      "learning_rate": 4.099494877032271e-05,
      "loss": 0.3623,
      "step": 32950,
      "task_loss": 0.599341630935669
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3016449809074402,
      "epoch": 11.91,
      "learning_rate": 4.093153782612035e-05,
      "loss": 0.3564,
      "step": 32960,
      "task_loss": 0.49827057123184204
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26733893156051636,
      "epoch": 11.92,
      "learning_rate": 4.086807051523305e-05,
      "loss": 0.3167,
      "step": 32970,
      "task_loss": 0.39013808965682983
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30923688411712646,
      "epoch": 11.92,
      "learning_rate": 4.08045471649196e-05,
      "loss": 0.3393,
      "step": 32980,
      "task_loss": 0.455595463514328
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31686368584632874,
      "epoch": 11.92,
      "learning_rate": 4.074096810272776e-05,
      "loss": 0.3616,
      "step": 32990,
      "task_loss": 0.6364704966545105
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36765047907829285,
      "epoch": 11.93,
      "learning_rate": 4.067733365649255e-05,
      "loss": 0.3513,
      "step": 33000,
      "task_loss": 0.4931289255619049
    },
    {
      "epoch": 11.93,
      "eval_exact_match": 82.8476821192053,
      "eval_f1": 89.65882395133605,
      "step": 33000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.422896146774292,
      "epoch": 11.93,
      "learning_rate": 4.061364415433457e-05,
      "loss": 0.3568,
      "step": 33010,
      "task_loss": 0.4513043165206909
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3466213643550873,
      "epoch": 11.93,
      "learning_rate": 4.0549899924658306e-05,
      "loss": 0.3668,
      "step": 33020,
      "task_loss": 0.47706955671310425
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35656729340553284,
      "epoch": 11.94,
      "learning_rate": 4.0486101296150455e-05,
      "loss": 0.3532,
      "step": 33030,
      "task_loss": 0.7009148597717285
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.40172943472862244,
      "epoch": 11.94,
      "learning_rate": 4.042224859777819e-05,
      "loss": 0.3756,
      "step": 33040,
      "task_loss": 0.6676845550537109
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28709399700164795,
      "epoch": 11.94,
      "learning_rate": 4.03583421587875e-05,
      "loss": 0.3515,
      "step": 33050,
      "task_loss": 0.5011777877807617
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26920461654663086,
      "epoch": 11.95,
      "learning_rate": 4.029438230870147e-05,
      "loss": 0.343,
      "step": 33060,
      "task_loss": 0.8642024993896484
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.330618292093277,
      "epoch": 11.95,
      "learning_rate": 4.02303693773186e-05,
      "loss": 0.3859,
      "step": 33070,
      "task_loss": 0.4772784411907196
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3623301386833191,
      "epoch": 11.96,
      "learning_rate": 4.016630369471107e-05,
      "loss": 0.3878,
      "step": 33080,
      "task_loss": 0.6183096766471863
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4232385456562042,
      "epoch": 11.96,
      "learning_rate": 4.010218559122312e-05,
      "loss": 0.3692,
      "step": 33090,
      "task_loss": 0.5985369682312012
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4274103343486786,
      "epoch": 11.96,
      "learning_rate": 4.003801539746922e-05,
      "loss": 0.36,
      "step": 33100,
      "task_loss": 0.637836217880249
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31120359897613525,
      "epoch": 11.97,
      "learning_rate": 3.997379344433248e-05,
      "loss": 0.3336,
      "step": 33110,
      "task_loss": 0.4128878116607666
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35576552152633667,
      "epoch": 11.97,
      "learning_rate": 3.990952006296288e-05,
      "loss": 0.3998,
      "step": 33120,
      "task_loss": 1.0541032552719116
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.399566113948822,
      "epoch": 11.97,
      "learning_rate": 3.98451955847756e-05,
      "loss": 0.3525,
      "step": 33130,
      "task_loss": 0.8153650760650635
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2788093388080597,
      "epoch": 11.98,
      "learning_rate": 3.978082034144926e-05,
      "loss": 0.3263,
      "step": 33140,
      "task_loss": 0.2653265595436096
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3298949599266052,
      "epoch": 11.98,
      "learning_rate": 3.971639466492429e-05,
      "loss": 0.3387,
      "step": 33150,
      "task_loss": 0.5124013423919678
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2869107127189636,
      "epoch": 11.98,
      "learning_rate": 3.9651918887401116e-05,
      "loss": 0.3311,
      "step": 33160,
      "task_loss": 0.3431087136268616
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.45431074500083923,
      "epoch": 11.99,
      "learning_rate": 3.958739334133853e-05,
      "loss": 0.3839,
      "step": 33170,
      "task_loss": 1.1194489002227783
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34249794483184814,
      "epoch": 11.99,
      "learning_rate": 3.952281835945196e-05,
      "loss": 0.3492,
      "step": 33180,
      "task_loss": 0.924396812915802
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4952884912490845,
      "epoch": 11.99,
      "learning_rate": 3.9458194274711716e-05,
      "loss": 0.3321,
      "step": 33190,
      "task_loss": 0.9026622772216797
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34919559955596924,
      "epoch": 12.0,
      "learning_rate": 3.939352142034133e-05,
      "loss": 0.3531,
      "step": 33200,
      "task_loss": 0.6708678007125854
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30084478855133057,
      "epoch": 12.0,
      "learning_rate": 3.932880012981575e-05,
      "loss": 0.3387,
      "step": 33210,
      "task_loss": 0.6755636930465698
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2988947331905365,
      "epoch": 12.01,
      "learning_rate": 3.926403073685974e-05,
      "loss": 0.289,
      "step": 33220,
      "task_loss": 0.42057114839553833
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33406442403793335,
      "epoch": 12.01,
      "learning_rate": 3.919921357544606e-05,
      "loss": 0.3132,
      "step": 33230,
      "task_loss": 0.5716870427131653
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4645366072654724,
      "epoch": 12.01,
      "learning_rate": 3.91343489797938e-05,
      "loss": 0.3505,
      "step": 33240,
      "task_loss": 0.8373175263404846
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23620998859405518,
      "epoch": 12.02,
      "learning_rate": 3.906943728436662e-05,
      "loss": 0.3047,
      "step": 33250,
      "task_loss": 0.5662074089050293
    },
    {
      "epoch": 12.02,
      "eval_exact_match": 82.88552507095554,
      "eval_f1": 89.60898535282445,
      "step": 33250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35468730330467224,
      "epoch": 12.02,
      "learning_rate": 3.900447882387106e-05,
      "loss": 0.3293,
      "step": 33260,
      "task_loss": 0.5851520895957947
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3405912518501282,
      "epoch": 12.02,
      "learning_rate": 3.8939473933254764e-05,
      "loss": 0.316,
      "step": 33270,
      "task_loss": 0.5842773914337158
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29812487959861755,
      "epoch": 12.03,
      "learning_rate": 3.8874422947704814e-05,
      "loss": 0.3099,
      "step": 33280,
      "task_loss": 0.353479266166687
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28859126567840576,
      "epoch": 12.03,
      "learning_rate": 3.8809326202645984e-05,
      "loss": 0.3409,
      "step": 33290,
      "task_loss": 0.20747502148151398
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28653669357299805,
      "epoch": 12.03,
      "learning_rate": 3.874418403373896e-05,
      "loss": 0.3129,
      "step": 33300,
      "task_loss": 0.6832212209701538
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33109143376350403,
      "epoch": 12.04,
      "learning_rate": 3.867899677687868e-05,
      "loss": 0.3081,
      "step": 33310,
      "task_loss": 0.6277453899383545
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3134967088699341,
      "epoch": 12.04,
      "learning_rate": 3.8613764768192565e-05,
      "loss": 0.2934,
      "step": 33320,
      "task_loss": 0.6228721737861633
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24643732607364655,
      "epoch": 12.05,
      "learning_rate": 3.8548488344038775e-05,
      "loss": 0.294,
      "step": 33330,
      "task_loss": 0.4370988607406616
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27333444356918335,
      "epoch": 12.05,
      "learning_rate": 3.8483167841004506e-05,
      "loss": 0.2789,
      "step": 33340,
      "task_loss": 0.3062555193901062
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2717025876045227,
      "epoch": 12.05,
      "learning_rate": 3.8417803595904245e-05,
      "loss": 0.3134,
      "step": 33350,
      "task_loss": 0.8133729100227356
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35719507932662964,
      "epoch": 12.06,
      "learning_rate": 3.835239594577801e-05,
      "loss": 0.3114,
      "step": 33360,
      "task_loss": 0.638519287109375
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2419562190771103,
      "epoch": 12.06,
      "learning_rate": 3.8286945227889645e-05,
      "loss": 0.3095,
      "step": 33370,
      "task_loss": 0.40532201528549194
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21433952450752258,
      "epoch": 12.06,
      "learning_rate": 3.8221451779725066e-05,
      "loss": 0.3122,
      "step": 33380,
      "task_loss": 0.5453925728797913
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5214677453041077,
      "epoch": 12.07,
      "learning_rate": 3.815591593899051e-05,
      "loss": 0.3375,
      "step": 33390,
      "task_loss": 0.7860684394836426
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25638237595558167,
      "epoch": 12.07,
      "learning_rate": 3.80903380436108e-05,
      "loss": 0.2805,
      "step": 33400,
      "task_loss": 0.2972143292427063
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2858333885669708,
      "epoch": 12.07,
      "learning_rate": 3.802471843172762e-05,
      "loss": 0.3611,
      "step": 33410,
      "task_loss": 0.7213091850280762
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24611973762512207,
      "epoch": 12.08,
      "learning_rate": 3.795905744169777e-05,
      "loss": 0.3028,
      "step": 33420,
      "task_loss": 0.353307843208313
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39217904210090637,
      "epoch": 12.08,
      "learning_rate": 3.789335541209138e-05,
      "loss": 0.3213,
      "step": 33430,
      "task_loss": 0.5517551898956299
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2604614496231079,
      "epoch": 12.09,
      "learning_rate": 3.782761268169021e-05,
      "loss": 0.3053,
      "step": 33440,
      "task_loss": 0.36389464139938354
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24966314435005188,
      "epoch": 12.09,
      "learning_rate": 3.7761829589485886e-05,
      "loss": 0.3254,
      "step": 33450,
      "task_loss": 0.5294799208641052
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20053541660308838,
      "epoch": 12.09,
      "learning_rate": 3.769600647467815e-05,
      "loss": 0.2726,
      "step": 33460,
      "task_loss": 0.25484806299209595
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22263674437999725,
      "epoch": 12.1,
      "learning_rate": 3.7630143676673115e-05,
      "loss": 0.3163,
      "step": 33470,
      "task_loss": 0.4103575050830841
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2307630330324173,
      "epoch": 12.1,
      "learning_rate": 3.756424153508152e-05,
      "loss": 0.2862,
      "step": 33480,
      "task_loss": 0.2939828634262085
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27595192193984985,
      "epoch": 12.1,
      "learning_rate": 3.749830038971697e-05,
      "loss": 0.3034,
      "step": 33490,
      "task_loss": 0.576181173324585
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26516520977020264,
      "epoch": 12.11,
      "learning_rate": 3.743892029168103e-05,
      "loss": 0.3007,
      "step": 33500,
      "task_loss": 0.4581066966056824
    },
    {
      "epoch": 12.11,
      "eval_exact_match": 82.83822138126774,
      "eval_f1": 89.7388509210768,
      "step": 33500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33838415145874023,
      "epoch": 12.11,
      "learning_rate": 3.737290597605322e-05,
      "loss": 0.3144,
      "step": 33510,
      "task_loss": 0.6935557126998901
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27680984139442444,
      "epoch": 12.11,
      "learning_rate": 3.730685364324294e-05,
      "loss": 0.2987,
      "step": 33520,
      "task_loss": 0.5029197931289673
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3041922450065613,
      "epoch": 12.12,
      "learning_rate": 3.724076363383821e-05,
      "loss": 0.2933,
      "step": 33530,
      "task_loss": 0.7183068990707397
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3190416991710663,
      "epoch": 12.12,
      "learning_rate": 3.7174636288621354e-05,
      "loss": 0.3141,
      "step": 33540,
      "task_loss": 0.7233753204345703
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3400724232196808,
      "epoch": 12.13,
      "learning_rate": 3.710847194856721e-05,
      "loss": 0.3335,
      "step": 33550,
      "task_loss": 0.4207264184951782
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22366857528686523,
      "epoch": 12.13,
      "learning_rate": 3.704227095484135e-05,
      "loss": 0.3038,
      "step": 33560,
      "task_loss": 0.3199766278266907
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.313113272190094,
      "epoch": 12.13,
      "learning_rate": 3.697603364879837e-05,
      "loss": 0.2895,
      "step": 33570,
      "task_loss": 0.9284161329269409
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28061339259147644,
      "epoch": 12.14,
      "learning_rate": 3.6909760371980076e-05,
      "loss": 0.312,
      "step": 33580,
      "task_loss": 0.40282338857650757
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.304638534784317,
      "epoch": 12.14,
      "learning_rate": 3.684345146611379e-05,
      "loss": 0.3121,
      "step": 33590,
      "task_loss": 0.33339840173721313
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2917397916316986,
      "epoch": 12.14,
      "learning_rate": 3.677710727311052e-05,
      "loss": 0.3095,
      "step": 33600,
      "task_loss": 0.5817675590515137
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25793755054473877,
      "epoch": 12.15,
      "learning_rate": 3.6710728135063216e-05,
      "loss": 0.3059,
      "step": 33610,
      "task_loss": 0.40517011284828186
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2668699622154236,
      "epoch": 12.15,
      "learning_rate": 3.6644314394245064e-05,
      "loss": 0.3173,
      "step": 33620,
      "task_loss": 0.5295995473861694
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2723209261894226,
      "epoch": 12.15,
      "learning_rate": 3.657786639310763e-05,
      "loss": 0.3043,
      "step": 33630,
      "task_loss": 0.49763721227645874
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2693908214569092,
      "epoch": 12.16,
      "learning_rate": 3.651138447427916e-05,
      "loss": 0.2886,
      "step": 33640,
      "task_loss": 0.4400332570075989
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.5167297124862671,
      "epoch": 12.16,
      "learning_rate": 3.644486898056278e-05,
      "loss": 0.3416,
      "step": 33650,
      "task_loss": 0.9263020753860474
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2826722264289856,
      "epoch": 12.16,
      "learning_rate": 3.637832025493472e-05,
      "loss": 0.3065,
      "step": 33660,
      "task_loss": 0.5701964497566223
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29284870624542236,
      "epoch": 12.17,
      "learning_rate": 3.631173864054263e-05,
      "loss": 0.3013,
      "step": 33670,
      "task_loss": 0.5212148427963257
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2942213714122772,
      "epoch": 12.17,
      "learning_rate": 3.624512448070367e-05,
      "loss": 0.3022,
      "step": 33680,
      "task_loss": 0.4938852787017822
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2890348434448242,
      "epoch": 12.18,
      "learning_rate": 3.617847811890286e-05,
      "loss": 0.3044,
      "step": 33690,
      "task_loss": 0.3784128427505493
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2744063138961792,
      "epoch": 12.18,
      "learning_rate": 3.6111799898791254e-05,
      "loss": 0.3688,
      "step": 33700,
      "task_loss": 0.38440898060798645
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32158753275871277,
      "epoch": 12.18,
      "learning_rate": 3.604509016418417e-05,
      "loss": 0.3187,
      "step": 33710,
      "task_loss": 0.6017839908599854
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2738969624042511,
      "epoch": 12.19,
      "learning_rate": 3.597834925905944e-05,
      "loss": 0.3011,
      "step": 33720,
      "task_loss": 0.653806746006012
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31608352065086365,
      "epoch": 12.19,
      "learning_rate": 3.59115775275556e-05,
      "loss": 0.3181,
      "step": 33730,
      "task_loss": 0.9595538377761841
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3204280138015747,
      "epoch": 12.19,
      "learning_rate": 3.584477531397016e-05,
      "loss": 0.3341,
      "step": 33740,
      "task_loss": 0.6531627774238586
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2891020178794861,
      "epoch": 12.2,
      "learning_rate": 3.577794296275781e-05,
      "loss": 0.3018,
      "step": 33750,
      "task_loss": 0.5011224150657654
    },
    {
      "epoch": 12.2,
      "eval_exact_match": 82.99905392620624,
      "eval_f1": 89.93215898637722,
      "step": 33750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27659574151039124,
      "epoch": 12.2,
      "learning_rate": 3.57110808185286e-05,
      "loss": 0.3311,
      "step": 33760,
      "task_loss": 0.37247174978256226
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22908209264278412,
      "epoch": 12.2,
      "learning_rate": 3.564418922604625e-05,
      "loss": 0.331,
      "step": 33770,
      "task_loss": 0.4124646484851837
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2758791148662567,
      "epoch": 12.21,
      "learning_rate": 3.5577268530226296e-05,
      "loss": 0.3204,
      "step": 33780,
      "task_loss": 0.4327734708786011
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28634101152420044,
      "epoch": 12.21,
      "learning_rate": 3.551031907613435e-05,
      "loss": 0.3361,
      "step": 33790,
      "task_loss": 0.5430434942245483
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2788851261138916,
      "epoch": 12.22,
      "learning_rate": 3.544334120898431e-05,
      "loss": 0.3317,
      "step": 33800,
      "task_loss": 0.3666086792945862
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2842410206794739,
      "epoch": 12.22,
      "learning_rate": 3.5376335274136567e-05,
      "loss": 0.3154,
      "step": 33810,
      "task_loss": 0.3724209666252136
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27617254853248596,
      "epoch": 12.22,
      "learning_rate": 3.530930161709628e-05,
      "loss": 0.2959,
      "step": 33820,
      "task_loss": 0.8962105512619019
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28928452730178833,
      "epoch": 12.23,
      "learning_rate": 3.5242240583511506e-05,
      "loss": 0.319,
      "step": 33830,
      "task_loss": 0.5773432850837708
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34535133838653564,
      "epoch": 12.23,
      "learning_rate": 3.5175152519171495e-05,
      "loss": 0.3082,
      "step": 33840,
      "task_loss": 0.5388073921203613
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30456995964050293,
      "epoch": 12.23,
      "learning_rate": 3.5108037770004867e-05,
      "loss": 0.2933,
      "step": 33850,
      "task_loss": 0.4720555543899536
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2527344524860382,
      "epoch": 12.24,
      "learning_rate": 3.5040896682077835e-05,
      "loss": 0.3107,
      "step": 33860,
      "task_loss": 0.48990774154663086
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2734881639480591,
      "epoch": 12.24,
      "learning_rate": 3.497372960159241e-05,
      "loss": 0.3044,
      "step": 33870,
      "task_loss": 0.3934549391269684
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2221376895904541,
      "epoch": 12.24,
      "learning_rate": 3.490653687488467e-05,
      "loss": 0.3214,
      "step": 33880,
      "task_loss": 0.2999269366264343
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3240625262260437,
      "epoch": 12.25,
      "learning_rate": 3.4839318848422894e-05,
      "loss": 0.2987,
      "step": 33890,
      "task_loss": 0.44523006677627563
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37814152240753174,
      "epoch": 12.25,
      "learning_rate": 3.477207586880583e-05,
      "loss": 0.3273,
      "step": 33900,
      "task_loss": 0.7390934228897095
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28805142641067505,
      "epoch": 12.26,
      "learning_rate": 3.47048082827609e-05,
      "loss": 0.2712,
      "step": 33910,
      "task_loss": 0.47077667713165283
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24550151824951172,
      "epoch": 12.26,
      "learning_rate": 3.463751643714238e-05,
      "loss": 0.2862,
      "step": 33920,
      "task_loss": 0.31704485416412354
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37711456418037415,
      "epoch": 12.26,
      "learning_rate": 3.457020067892969e-05,
      "loss": 0.3118,
      "step": 33930,
      "task_loss": 0.8705230951309204
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2998274862766266,
      "epoch": 12.27,
      "learning_rate": 3.450286135522547e-05,
      "loss": 0.3065,
      "step": 33940,
      "task_loss": 0.455272376537323
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2974916100502014,
      "epoch": 12.27,
      "learning_rate": 3.443549881325395e-05,
      "loss": 0.3035,
      "step": 33950,
      "task_loss": 0.31812000274658203
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3739232122898102,
      "epoch": 12.27,
      "learning_rate": 3.4368113400359036e-05,
      "loss": 0.318,
      "step": 33960,
      "task_loss": 1.0148547887802124
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2662932276725769,
      "epoch": 12.28,
      "learning_rate": 3.430070546400257e-05,
      "loss": 0.3136,
      "step": 33970,
      "task_loss": 0.4627422094345093
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32645705342292786,
      "epoch": 12.28,
      "learning_rate": 3.423327535176256e-05,
      "loss": 0.3071,
      "step": 33980,
      "task_loss": 0.5508854389190674
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3119924068450928,
      "epoch": 12.28,
      "learning_rate": 3.4165823411331315e-05,
      "loss": 0.3126,
      "step": 33990,
      "task_loss": 0.5757681727409363
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23790660500526428,
      "epoch": 12.29,
      "learning_rate": 3.409834999051374e-05,
      "loss": 0.2965,
      "step": 34000,
      "task_loss": 0.7489010691642761
    },
    {
      "epoch": 12.29,
      "eval_exact_match": 82.72469252601702,
      "eval_f1": 89.51422301711462,
      "step": 34000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24514292180538177,
      "epoch": 12.29,
      "learning_rate": 3.403085543722547e-05,
      "loss": 0.3226,
      "step": 34010,
      "task_loss": 0.48626774549484253
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.38149791955947876,
      "epoch": 12.29,
      "learning_rate": 3.396334009949112e-05,
      "loss": 0.2977,
      "step": 34020,
      "task_loss": 0.31508100032806396
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3336488604545593,
      "epoch": 12.3,
      "learning_rate": 3.389580432544248e-05,
      "loss": 0.3386,
      "step": 34030,
      "task_loss": 0.5556856393814087
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20330330729484558,
      "epoch": 12.3,
      "learning_rate": 3.3828248463316716e-05,
      "loss": 0.3154,
      "step": 34040,
      "task_loss": 0.5865062475204468
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25431859493255615,
      "epoch": 12.31,
      "learning_rate": 3.376067286145455e-05,
      "loss": 0.3229,
      "step": 34050,
      "task_loss": 0.32260221242904663
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.47123128175735474,
      "epoch": 12.31,
      "learning_rate": 3.3693077868298524e-05,
      "loss": 0.3253,
      "step": 34060,
      "task_loss": 0.4305189251899719
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27410605549812317,
      "epoch": 12.31,
      "learning_rate": 3.362546383239115e-05,
      "loss": 0.3286,
      "step": 34070,
      "task_loss": 0.6105645895004272
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31597283482551575,
      "epoch": 12.32,
      "learning_rate": 3.355783110237313e-05,
      "loss": 0.2906,
      "step": 34080,
      "task_loss": 0.6624294519424438
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2440500557422638,
      "epoch": 12.32,
      "learning_rate": 3.349018002698157e-05,
      "loss": 0.2856,
      "step": 34090,
      "task_loss": 0.4859399199485779
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2693696618080139,
      "epoch": 12.32,
      "learning_rate": 3.342251095504816e-05,
      "loss": 0.3057,
      "step": 34100,
      "task_loss": 0.5549313426017761
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29348957538604736,
      "epoch": 12.33,
      "learning_rate": 3.335482423549738e-05,
      "loss": 0.3031,
      "step": 34110,
      "task_loss": 0.5355805158615112
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3276856541633606,
      "epoch": 12.33,
      "learning_rate": 3.3287120217344734e-05,
      "loss": 0.3146,
      "step": 34120,
      "task_loss": 1.2835772037506104
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36418938636779785,
      "epoch": 12.33,
      "learning_rate": 3.321939924969489e-05,
      "loss": 0.3095,
      "step": 34130,
      "task_loss": 0.4305654466152191
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.48510342836380005,
      "epoch": 12.34,
      "learning_rate": 3.315166168173994e-05,
      "loss": 0.3438,
      "step": 34140,
      "task_loss": 0.6050916314125061
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24135936796665192,
      "epoch": 12.34,
      "learning_rate": 3.308390786275757e-05,
      "loss": 0.2987,
      "step": 34150,
      "task_loss": 0.5380813479423523
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2821900546550751,
      "epoch": 12.35,
      "learning_rate": 3.3016138142109226e-05,
      "loss": 0.3237,
      "step": 34160,
      "task_loss": 0.503202497959137
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32364422082901,
      "epoch": 12.35,
      "learning_rate": 3.29483528692384e-05,
      "loss": 0.3045,
      "step": 34170,
      "task_loss": 0.6083263754844666
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2627584934234619,
      "epoch": 12.35,
      "learning_rate": 3.2880552393668754e-05,
      "loss": 0.3165,
      "step": 34180,
      "task_loss": 0.13859528303146362
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2632201910018921,
      "epoch": 12.36,
      "learning_rate": 3.2812737065002324e-05,
      "loss": 0.3,
      "step": 34190,
      "task_loss": 0.669399082660675
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25493764877319336,
      "epoch": 12.36,
      "learning_rate": 3.274490723291776e-05,
      "loss": 0.2976,
      "step": 34200,
      "task_loss": 0.6728479862213135
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33259302377700806,
      "epoch": 12.36,
      "learning_rate": 3.267706324716848e-05,
      "loss": 0.3314,
      "step": 34210,
      "task_loss": 0.9753580689430237
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31175535917282104,
      "epoch": 12.37,
      "learning_rate": 3.2609205457580886e-05,
      "loss": 0.3139,
      "step": 34220,
      "task_loss": 0.3914549946784973
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3147822618484497,
      "epoch": 12.37,
      "learning_rate": 3.254133421405257e-05,
      "loss": 0.2919,
      "step": 34230,
      "task_loss": 0.8125878572463989
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24442313611507416,
      "epoch": 12.37,
      "learning_rate": 3.247344986655048e-05,
      "loss": 0.3332,
      "step": 34240,
      "task_loss": 0.5116071105003357
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2911091446876526,
      "epoch": 12.38,
      "learning_rate": 3.2405552765109136e-05,
      "loss": 0.3303,
      "step": 34250,
      "task_loss": 1.3424382209777832
    },
    {
      "epoch": 12.38,
      "eval_exact_match": 82.68684957426679,
      "eval_f1": 89.51533730161242,
      "step": 34250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3006014823913574,
      "epoch": 12.38,
      "learning_rate": 3.2337643259828825e-05,
      "loss": 0.2999,
      "step": 34260,
      "task_loss": 0.33499881625175476
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2646566927433014,
      "epoch": 12.39,
      "learning_rate": 3.226972170087379e-05,
      "loss": 0.2945,
      "step": 34270,
      "task_loss": 0.45225051045417786
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34713464975357056,
      "epoch": 12.39,
      "learning_rate": 3.220178843847045e-05,
      "loss": 0.3283,
      "step": 34280,
      "task_loss": 0.6201308369636536
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28564396500587463,
      "epoch": 12.39,
      "learning_rate": 3.213384382290552e-05,
      "loss": 0.2633,
      "step": 34290,
      "task_loss": 0.5016683340072632
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24384666979312897,
      "epoch": 12.4,
      "learning_rate": 3.20658882045243e-05,
      "loss": 0.3113,
      "step": 34300,
      "task_loss": 0.4746541976928711
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23822401463985443,
      "epoch": 12.4,
      "learning_rate": 3.199792193372881e-05,
      "loss": 0.2974,
      "step": 34310,
      "task_loss": 0.3712196946144104
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2704951763153076,
      "epoch": 12.4,
      "learning_rate": 3.192994536097598e-05,
      "loss": 0.2739,
      "step": 34320,
      "task_loss": 0.5020608901977539
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25836482644081116,
      "epoch": 12.41,
      "learning_rate": 3.186195883677591e-05,
      "loss": 0.3205,
      "step": 34330,
      "task_loss": 0.4176018238067627
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22267907857894897,
      "epoch": 12.41,
      "learning_rate": 3.1793962711689954e-05,
      "loss": 0.3038,
      "step": 34340,
      "task_loss": 0.36904817819595337
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24134516716003418,
      "epoch": 12.41,
      "learning_rate": 3.172595733632902e-05,
      "loss": 0.3062,
      "step": 34350,
      "task_loss": 0.5529676079750061
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2891293168067932,
      "epoch": 12.42,
      "learning_rate": 3.1657943061351665e-05,
      "loss": 0.2842,
      "step": 34360,
      "task_loss": 0.40239569544792175
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25356581807136536,
      "epoch": 12.42,
      "learning_rate": 3.158992023746236e-05,
      "loss": 0.3037,
      "step": 34370,
      "task_loss": 0.5014570355415344
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32125765085220337,
      "epoch": 12.43,
      "learning_rate": 3.15218892154097e-05,
      "loss": 0.3079,
      "step": 34380,
      "task_loss": 0.4384068250656128
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3878769874572754,
      "epoch": 12.43,
      "learning_rate": 3.145385034598447e-05,
      "loss": 0.3303,
      "step": 34390,
      "task_loss": 0.8171356916427612
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28572502732276917,
      "epoch": 12.43,
      "learning_rate": 3.138580398001798e-05,
      "loss": 0.3081,
      "step": 34400,
      "task_loss": 0.39703404903411865
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23337849974632263,
      "epoch": 12.44,
      "learning_rate": 3.131775046838017e-05,
      "loss": 0.2909,
      "step": 34410,
      "task_loss": 0.33157849311828613
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3653300404548645,
      "epoch": 12.44,
      "learning_rate": 3.124969016197782e-05,
      "loss": 0.3113,
      "step": 34420,
      "task_loss": 0.7859286069869995
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2919321358203888,
      "epoch": 12.44,
      "learning_rate": 3.1181623411752774e-05,
      "loss": 0.3266,
      "step": 34430,
      "task_loss": 0.43368104100227356
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20942334830760956,
      "epoch": 12.45,
      "learning_rate": 3.111355056868007e-05,
      "loss": 0.2836,
      "step": 34440,
      "task_loss": 0.4100843667984009
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3563976287841797,
      "epoch": 12.45,
      "learning_rate": 3.104547198376618e-05,
      "loss": 0.32,
      "step": 34450,
      "task_loss": 0.28328973054885864
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2783935070037842,
      "epoch": 12.45,
      "learning_rate": 3.09773880080472e-05,
      "loss": 0.3252,
      "step": 34460,
      "task_loss": 0.48669523000717163
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28177738189697266,
      "epoch": 12.46,
      "learning_rate": 3.0909298992586974e-05,
      "loss": 0.3014,
      "step": 34470,
      "task_loss": 0.44701170921325684
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2627007067203522,
      "epoch": 12.46,
      "learning_rate": 3.084120528847538e-05,
      "loss": 0.3006,
      "step": 34480,
      "task_loss": 0.44245171546936035
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3353618383407593,
      "epoch": 12.46,
      "learning_rate": 3.077310724682645e-05,
      "loss": 0.3143,
      "step": 34490,
      "task_loss": 0.4516112208366394
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32477229833602905,
      "epoch": 12.47,
      "learning_rate": 3.070500521877658e-05,
      "loss": 0.301,
      "step": 34500,
      "task_loss": 0.6717529892921448
    },
    {
      "epoch": 12.47,
      "eval_exact_match": 82.96121097445601,
      "eval_f1": 89.7702056217889,
      "step": 34500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27970433235168457,
      "epoch": 12.47,
      "learning_rate": 3.063689955548274e-05,
      "loss": 0.2976,
      "step": 34510,
      "task_loss": 0.5146293640136719
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2640838921070099,
      "epoch": 12.48,
      "learning_rate": 3.056879060812061e-05,
      "loss": 0.3187,
      "step": 34520,
      "task_loss": 0.29868653416633606
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2819470167160034,
      "epoch": 12.48,
      "learning_rate": 3.0500678727882856e-05,
      "loss": 0.2936,
      "step": 34530,
      "task_loss": 0.43330201506614685
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3299875855445862,
      "epoch": 12.48,
      "learning_rate": 3.043256426597721e-05,
      "loss": 0.3005,
      "step": 34540,
      "task_loss": 0.4487469792366028
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31510305404663086,
      "epoch": 12.49,
      "learning_rate": 3.036444757362475e-05,
      "loss": 0.3098,
      "step": 34550,
      "task_loss": 0.3750832676887512
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31631648540496826,
      "epoch": 12.49,
      "learning_rate": 3.029632900205806e-05,
      "loss": 0.3136,
      "step": 34560,
      "task_loss": 0.29234394431114197
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25344857573509216,
      "epoch": 12.49,
      "learning_rate": 3.0228208902519374e-05,
      "loss": 0.3155,
      "step": 34570,
      "task_loss": 0.4813919961452484
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23850274085998535,
      "epoch": 12.5,
      "learning_rate": 3.016008762625886e-05,
      "loss": 0.3022,
      "step": 34580,
      "task_loss": 0.5706331729888916
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23559576272964478,
      "epoch": 12.5,
      "learning_rate": 3.00919655245327e-05,
      "loss": 0.2813,
      "step": 34590,
      "task_loss": 0.3056916892528534
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3584052324295044,
      "epoch": 12.5,
      "learning_rate": 3.002384294860138e-05,
      "loss": 0.3686,
      "step": 34600,
      "task_loss": 0.783943235874176
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2597944736480713,
      "epoch": 12.51,
      "learning_rate": 2.9955720249727796e-05,
      "loss": 0.334,
      "step": 34610,
      "task_loss": 0.4639328718185425
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2866797149181366,
      "epoch": 12.51,
      "learning_rate": 2.9887597779175498e-05,
      "loss": 0.3025,
      "step": 34620,
      "task_loss": 0.8007729053497314
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3565661311149597,
      "epoch": 12.52,
      "learning_rate": 2.9819475888206846e-05,
      "loss": 0.3179,
      "step": 34630,
      "task_loss": 0.7241135239601135
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30475327372550964,
      "epoch": 12.52,
      "learning_rate": 2.975135492808122e-05,
      "loss": 0.3257,
      "step": 34640,
      "task_loss": 0.4078003466129303
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3345334827899933,
      "epoch": 12.52,
      "learning_rate": 2.9683235250053192e-05,
      "loss": 0.3087,
      "step": 34650,
      "task_loss": 0.4602591395378113
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.49751919507980347,
      "epoch": 12.53,
      "learning_rate": 2.9615117205370738e-05,
      "loss": 0.3164,
      "step": 34660,
      "task_loss": 0.6007466316223145
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2072526514530182,
      "epoch": 12.53,
      "learning_rate": 2.9547001145273384e-05,
      "loss": 0.3216,
      "step": 34670,
      "task_loss": 0.7436428070068359
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34686729311943054,
      "epoch": 12.53,
      "learning_rate": 2.9478887420990473e-05,
      "loss": 0.3079,
      "step": 34680,
      "task_loss": 0.9251314401626587
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27641862630844116,
      "epoch": 12.54,
      "learning_rate": 2.9410776383739235e-05,
      "loss": 0.3094,
      "step": 34690,
      "task_loss": 0.45291346311569214
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2968369126319885,
      "epoch": 12.54,
      "learning_rate": 2.934266838472311e-05,
      "loss": 0.2977,
      "step": 34700,
      "task_loss": 0.524625301361084
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32808536291122437,
      "epoch": 12.54,
      "learning_rate": 2.9274563775129826e-05,
      "loss": 0.3131,
      "step": 34710,
      "task_loss": 0.7311720848083496
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37095907330513,
      "epoch": 12.55,
      "learning_rate": 2.920646290612966e-05,
      "loss": 0.2997,
      "step": 34720,
      "task_loss": 0.9131951332092285
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23286473751068115,
      "epoch": 12.55,
      "learning_rate": 2.913836612887359e-05,
      "loss": 0.3163,
      "step": 34730,
      "task_loss": 0.5092307329177856
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2292613983154297,
      "epoch": 12.56,
      "learning_rate": 2.9070273794491498e-05,
      "loss": 0.2968,
      "step": 34740,
      "task_loss": 0.39329832792282104
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24411755800247192,
      "epoch": 12.56,
      "learning_rate": 2.900218625409036e-05,
      "loss": 0.3213,
      "step": 34750,
      "task_loss": 0.35126519203186035
    },
    {
      "epoch": 12.56,
      "eval_exact_match": 83.33964049195838,
      "eval_f1": 90.00245261288558,
      "step": 34750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4118926227092743,
      "epoch": 12.56,
      "learning_rate": 2.8934103858752427e-05,
      "loss": 0.332,
      "step": 34760,
      "task_loss": 0.562553882598877
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2720871865749359,
      "epoch": 12.57,
      "learning_rate": 2.8866026959533416e-05,
      "loss": 0.2967,
      "step": 34770,
      "task_loss": 0.33152109384536743
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2610456645488739,
      "epoch": 12.57,
      "learning_rate": 2.8797955907460733e-05,
      "loss": 0.2883,
      "step": 34780,
      "task_loss": 0.835556149482727
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24886228144168854,
      "epoch": 12.57,
      "learning_rate": 2.8729891053531584e-05,
      "loss": 0.2994,
      "step": 34790,
      "task_loss": 0.4246406555175781
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31593984365463257,
      "epoch": 12.58,
      "learning_rate": 2.866183274871127e-05,
      "loss": 0.3201,
      "step": 34800,
      "task_loss": 0.6031840443611145
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.289052814245224,
      "epoch": 12.58,
      "learning_rate": 2.859378134393128e-05,
      "loss": 0.2935,
      "step": 34810,
      "task_loss": 0.4136536717414856
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25019770860671997,
      "epoch": 12.58,
      "learning_rate": 2.8525737190087548e-05,
      "loss": 0.3172,
      "step": 34820,
      "task_loss": 0.5181039571762085
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2552999258041382,
      "epoch": 12.59,
      "learning_rate": 2.8457700638038606e-05,
      "loss": 0.2592,
      "step": 34830,
      "task_loss": 0.4570186734199524
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25203728675842285,
      "epoch": 12.59,
      "learning_rate": 2.8389672038603805e-05,
      "loss": 0.2763,
      "step": 34840,
      "task_loss": 0.5494889616966248
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3487963080406189,
      "epoch": 12.59,
      "learning_rate": 2.8321651742561465e-05,
      "loss": 0.3192,
      "step": 34850,
      "task_loss": 0.5809800624847412
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36887413263320923,
      "epoch": 12.6,
      "learning_rate": 2.8253640100647124e-05,
      "loss": 0.3091,
      "step": 34860,
      "task_loss": 0.24406704306602478
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3776524066925049,
      "epoch": 12.6,
      "learning_rate": 2.818563746355165e-05,
      "loss": 0.3065,
      "step": 34870,
      "task_loss": 0.6552656888961792
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27506256103515625,
      "epoch": 12.61,
      "learning_rate": 2.8117644181919547e-05,
      "loss": 0.3169,
      "step": 34880,
      "task_loss": 0.5139168500900269
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.35786789655685425,
      "epoch": 12.61,
      "learning_rate": 2.8049660606346994e-05,
      "loss": 0.3498,
      "step": 34890,
      "task_loss": 0.5621675252914429
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3235039710998535,
      "epoch": 12.61,
      "learning_rate": 2.79816870873802e-05,
      "loss": 0.2973,
      "step": 34900,
      "task_loss": 0.4850617051124573
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2791781425476074,
      "epoch": 12.62,
      "learning_rate": 2.7913723975513473e-05,
      "loss": 0.3168,
      "step": 34910,
      "task_loss": 0.44785845279693604
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3183351457118988,
      "epoch": 12.62,
      "learning_rate": 2.7845771621187476e-05,
      "loss": 0.3185,
      "step": 34920,
      "task_loss": 0.26022273302078247
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24664825201034546,
      "epoch": 12.62,
      "learning_rate": 2.7777830374787396e-05,
      "loss": 0.2517,
      "step": 34930,
      "task_loss": 0.3348398208618164
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2722611427307129,
      "epoch": 12.63,
      "learning_rate": 2.770990058664115e-05,
      "loss": 0.3454,
      "step": 34940,
      "task_loss": 0.5424426794052124
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.280570924282074,
      "epoch": 12.63,
      "learning_rate": 2.7641982607017554e-05,
      "loss": 0.2948,
      "step": 34950,
      "task_loss": 0.5146224498748779
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2772490680217743,
      "epoch": 12.63,
      "learning_rate": 2.757407678612457e-05,
      "loss": 0.2776,
      "step": 34960,
      "task_loss": 0.44447261095046997
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24304187297821045,
      "epoch": 12.64,
      "learning_rate": 2.7506183474107415e-05,
      "loss": 0.274,
      "step": 34970,
      "task_loss": 0.3981664776802063
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2865305542945862,
      "epoch": 12.64,
      "learning_rate": 2.743830302104688e-05,
      "loss": 0.3251,
      "step": 34980,
      "task_loss": 0.4353870749473572
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2837364673614502,
      "epoch": 12.65,
      "learning_rate": 2.7370435776957363e-05,
      "loss": 0.2872,
      "step": 34990,
      "task_loss": 0.5949020385742188
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3170028626918793,
      "epoch": 12.65,
      "learning_rate": 2.7302582091785228e-05,
      "loss": 0.3451,
      "step": 35000,
      "task_loss": 0.6154573559761047
    },
    {
      "epoch": 12.65,
      "eval_exact_match": 83.18826868495742,
      "eval_f1": 89.9089136876254,
      "step": 35000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23709344863891602,
      "epoch": 12.65,
      "learning_rate": 2.723474231540687e-05,
      "loss": 0.2959,
      "step": 35010,
      "task_loss": 0.532070517539978
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2065018117427826,
      "epoch": 12.66,
      "learning_rate": 2.7166916797627012e-05,
      "loss": 0.2954,
      "step": 35020,
      "task_loss": 0.36863160133361816
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34484899044036865,
      "epoch": 12.66,
      "learning_rate": 2.7099105888176807e-05,
      "loss": 0.3376,
      "step": 35030,
      "task_loss": 0.5684521794319153
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20965680480003357,
      "epoch": 12.66,
      "learning_rate": 2.7031309936712125e-05,
      "loss": 0.2896,
      "step": 35040,
      "task_loss": 0.5758606195449829
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2636858820915222,
      "epoch": 12.67,
      "learning_rate": 2.6963529292811673e-05,
      "loss": 0.2785,
      "step": 35050,
      "task_loss": 0.4598173499107361
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23789474368095398,
      "epoch": 12.67,
      "learning_rate": 2.689576430597526e-05,
      "loss": 0.3038,
      "step": 35060,
      "task_loss": 0.490090548992157
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29482781887054443,
      "epoch": 12.67,
      "learning_rate": 2.682801532562192e-05,
      "loss": 0.2974,
      "step": 35070,
      "task_loss": 0.9217762351036072
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25767529010772705,
      "epoch": 12.68,
      "learning_rate": 2.6760282701088214e-05,
      "loss": 0.3068,
      "step": 35080,
      "task_loss": 0.642894983291626
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2678022086620331,
      "epoch": 12.68,
      "learning_rate": 2.66925667816263e-05,
      "loss": 0.3014,
      "step": 35090,
      "task_loss": 0.6086078882217407
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2076597809791565,
      "epoch": 12.69,
      "learning_rate": 2.662486791640225e-05,
      "loss": 0.2907,
      "step": 35100,
      "task_loss": 0.23006045818328857
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33956581354141235,
      "epoch": 12.69,
      "learning_rate": 2.655718645449416e-05,
      "loss": 0.297,
      "step": 35110,
      "task_loss": 0.8203099966049194
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2641342282295227,
      "epoch": 12.69,
      "learning_rate": 2.648952274489042e-05,
      "loss": 0.3053,
      "step": 35120,
      "task_loss": 0.3075989782810211
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24056673049926758,
      "epoch": 12.7,
      "learning_rate": 2.6421877136487866e-05,
      "loss": 0.2915,
      "step": 35130,
      "task_loss": 1.0190541744232178
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21946954727172852,
      "epoch": 12.7,
      "learning_rate": 2.6354249978089998e-05,
      "loss": 0.2835,
      "step": 35140,
      "task_loss": 0.5367399454116821
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3418148458003998,
      "epoch": 12.7,
      "learning_rate": 2.6286641618405177e-05,
      "loss": 0.3303,
      "step": 35150,
      "task_loss": 0.3500261902809143
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.283519446849823,
      "epoch": 12.71,
      "learning_rate": 2.6219052406044874e-05,
      "loss": 0.3565,
      "step": 35160,
      "task_loss": 0.4854230284690857
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3122413158416748,
      "epoch": 12.71,
      "learning_rate": 2.615148268952175e-05,
      "loss": 0.3211,
      "step": 35170,
      "task_loss": 0.6190265417098999
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2543804943561554,
      "epoch": 12.71,
      "learning_rate": 2.6083932817248023e-05,
      "loss": 0.3016,
      "step": 35180,
      "task_loss": 0.2701253890991211
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3085487484931946,
      "epoch": 12.72,
      "learning_rate": 2.6016403137533524e-05,
      "loss": 0.3235,
      "step": 35190,
      "task_loss": 0.6589609980583191
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28512585163116455,
      "epoch": 12.72,
      "learning_rate": 2.594889399858401e-05,
      "loss": 0.3019,
      "step": 35200,
      "task_loss": 0.3173820972442627
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2618717551231384,
      "epoch": 12.72,
      "learning_rate": 2.5881405748499288e-05,
      "loss": 0.3216,
      "step": 35210,
      "task_loss": 0.3481888771057129
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24822674691677094,
      "epoch": 12.73,
      "learning_rate": 2.5813938735271474e-05,
      "loss": 0.3037,
      "step": 35220,
      "task_loss": 0.6111527681350708
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33503878116607666,
      "epoch": 12.73,
      "learning_rate": 2.5746493306783173e-05,
      "loss": 0.2805,
      "step": 35230,
      "task_loss": 0.28723621368408203
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21374112367630005,
      "epoch": 12.74,
      "learning_rate": 2.56790698108057e-05,
      "loss": 0.2953,
      "step": 35240,
      "task_loss": 0.32919079065322876
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26212984323501587,
      "epoch": 12.74,
      "learning_rate": 2.5611668594997254e-05,
      "loss": 0.2846,
      "step": 35250,
      "task_loss": 0.3463735580444336
    },
    {
      "epoch": 12.74,
      "eval_exact_match": 83.28287606433302,
      "eval_f1": 90.13822188130554,
      "step": 35250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4006456732749939,
      "epoch": 12.74,
      "learning_rate": 2.5544290006901206e-05,
      "loss": 0.2915,
      "step": 35260,
      "task_loss": 0.35442763566970825
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24394173920154572,
      "epoch": 12.75,
      "learning_rate": 2.5476934393944168e-05,
      "loss": 0.3141,
      "step": 35270,
      "task_loss": 0.5410995483398438
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27239710092544556,
      "epoch": 12.75,
      "learning_rate": 2.5409602103434368e-05,
      "loss": 0.3226,
      "step": 35280,
      "task_loss": 0.3332737386226654
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26623234152793884,
      "epoch": 12.75,
      "learning_rate": 2.5342293482559712e-05,
      "loss": 0.314,
      "step": 35290,
      "task_loss": 0.45213043689727783
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28726357221603394,
      "epoch": 12.76,
      "learning_rate": 2.5275008878386102e-05,
      "loss": 0.2934,
      "step": 35300,
      "task_loss": 0.5393993854522705
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3673667311668396,
      "epoch": 12.76,
      "learning_rate": 2.5207748637855563e-05,
      "loss": 0.2806,
      "step": 35310,
      "task_loss": 0.4093773663043976
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.334319144487381,
      "epoch": 12.76,
      "learning_rate": 2.5140513107784532e-05,
      "loss": 0.3143,
      "step": 35320,
      "task_loss": 1.018317461013794
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28693056106567383,
      "epoch": 12.77,
      "learning_rate": 2.5073302634861993e-05,
      "loss": 0.2935,
      "step": 35330,
      "task_loss": 0.6353425979614258
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2879865765571594,
      "epoch": 12.77,
      "learning_rate": 2.5006117565647766e-05,
      "loss": 0.2832,
      "step": 35340,
      "task_loss": 0.2560500502586365
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2619374394416809,
      "epoch": 12.78,
      "learning_rate": 2.4938958246570628e-05,
      "loss": 0.3187,
      "step": 35350,
      "task_loss": 0.5173212289810181
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26497191190719604,
      "epoch": 12.78,
      "learning_rate": 2.4871825023926654e-05,
      "loss": 0.2858,
      "step": 35360,
      "task_loss": 0.5506118535995483
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2945820689201355,
      "epoch": 12.78,
      "learning_rate": 2.4804718243877275e-05,
      "loss": 0.2953,
      "step": 35370,
      "task_loss": 0.6132495403289795
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3165275752544403,
      "epoch": 12.79,
      "learning_rate": 2.4737638252447642e-05,
      "loss": 0.3079,
      "step": 35380,
      "task_loss": 0.4461279511451721
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3094812035560608,
      "epoch": 12.79,
      "learning_rate": 2.4670585395524735e-05,
      "loss": 0.3315,
      "step": 35390,
      "task_loss": 0.4002085328102112
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.39851170778274536,
      "epoch": 12.79,
      "learning_rate": 2.4603560018855634e-05,
      "loss": 0.3087,
      "step": 35400,
      "task_loss": 0.39463096857070923
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3484683632850647,
      "epoch": 12.8,
      "learning_rate": 2.4536562468045718e-05,
      "loss": 0.2993,
      "step": 35410,
      "task_loss": 0.7016239166259766
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27489396929740906,
      "epoch": 12.8,
      "learning_rate": 2.4469593088556896e-05,
      "loss": 0.2889,
      "step": 35420,
      "task_loss": 0.5056446194648743
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25945621728897095,
      "epoch": 12.8,
      "learning_rate": 2.4402652225705794e-05,
      "loss": 0.3146,
      "step": 35430,
      "task_loss": 0.49554070830345154
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3142120838165283,
      "epoch": 12.81,
      "learning_rate": 2.4335740224662032e-05,
      "loss": 0.2934,
      "step": 35440,
      "task_loss": 0.4447091817855835
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33970868587493896,
      "epoch": 12.81,
      "learning_rate": 2.4268857430446355e-05,
      "loss": 0.3232,
      "step": 35450,
      "task_loss": 0.654474675655365
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3706507682800293,
      "epoch": 12.82,
      "learning_rate": 2.4202004187928985e-05,
      "loss": 0.2983,
      "step": 35460,
      "task_loss": 0.5285034775733948
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2316771298646927,
      "epoch": 12.82,
      "learning_rate": 2.413518084182767e-05,
      "loss": 0.3351,
      "step": 35470,
      "task_loss": 0.49286916851997375
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2711471915245056,
      "epoch": 12.82,
      "learning_rate": 2.4068387736706095e-05,
      "loss": 0.2861,
      "step": 35480,
      "task_loss": 0.4698110818862915
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2512504756450653,
      "epoch": 12.83,
      "learning_rate": 2.400162521697195e-05,
      "loss": 0.294,
      "step": 35490,
      "task_loss": 0.32535380125045776
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3419429659843445,
      "epoch": 12.83,
      "learning_rate": 2.3934893626875248e-05,
      "loss": 0.3093,
      "step": 35500,
      "task_loss": 0.4717838764190674
    },
    {
      "epoch": 12.83,
      "eval_exact_match": 83.36802270577105,
      "eval_f1": 89.98336633831875,
      "step": 35500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26069217920303345,
      "epoch": 12.83,
      "learning_rate": 2.3868193310506494e-05,
      "loss": 0.3388,
      "step": 35510,
      "task_loss": 0.7599429488182068
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22465810179710388,
      "epoch": 12.84,
      "learning_rate": 2.3801524611794955e-05,
      "loss": 0.2862,
      "step": 35520,
      "task_loss": 0.5255027413368225
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3821221590042114,
      "epoch": 12.84,
      "learning_rate": 2.373488787450686e-05,
      "loss": 0.2972,
      "step": 35530,
      "task_loss": 0.5774866938591003
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2761344909667969,
      "epoch": 12.84,
      "learning_rate": 2.3668283442243642e-05,
      "loss": 0.2888,
      "step": 35540,
      "task_loss": 0.48153436183929443
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30200979113578796,
      "epoch": 12.85,
      "learning_rate": 2.3601711658440125e-05,
      "loss": 0.3074,
      "step": 35550,
      "task_loss": 0.48879027366638184
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27535152435302734,
      "epoch": 12.85,
      "learning_rate": 2.353517286636285e-05,
      "loss": 0.291,
      "step": 35560,
      "task_loss": 0.36273616552352905
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36965012550354004,
      "epoch": 12.86,
      "learning_rate": 2.3468667409108163e-05,
      "loss": 0.3443,
      "step": 35570,
      "task_loss": 0.4493734836578369
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3609243631362915,
      "epoch": 12.86,
      "learning_rate": 2.34021956296006e-05,
      "loss": 0.3018,
      "step": 35580,
      "task_loss": 0.5398873686790466
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2682161331176758,
      "epoch": 12.86,
      "learning_rate": 2.333575787059099e-05,
      "loss": 0.2933,
      "step": 35590,
      "task_loss": 0.4194197952747345
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28163596987724304,
      "epoch": 12.87,
      "learning_rate": 2.3269354474654765e-05,
      "loss": 0.3208,
      "step": 35600,
      "task_loss": 0.28600168228149414
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19686777889728546,
      "epoch": 12.87,
      "learning_rate": 2.320298578419016e-05,
      "loss": 0.2912,
      "step": 35610,
      "task_loss": 0.33698779344558716
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3002614378929138,
      "epoch": 12.87,
      "learning_rate": 2.3136652141416464e-05,
      "loss": 0.3026,
      "step": 35620,
      "task_loss": 0.7368478775024414
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27924036979675293,
      "epoch": 12.88,
      "learning_rate": 2.3070353888372245e-05,
      "loss": 0.2995,
      "step": 35630,
      "task_loss": 0.7203919887542725
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32518601417541504,
      "epoch": 12.88,
      "learning_rate": 2.300409136691359e-05,
      "loss": 0.2785,
      "step": 35640,
      "task_loss": 0.8650093078613281
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31950339674949646,
      "epoch": 12.88,
      "learning_rate": 2.2937864918712325e-05,
      "loss": 0.3148,
      "step": 35650,
      "task_loss": 0.5481790900230408
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3315272927284241,
      "epoch": 12.89,
      "learning_rate": 2.287167488525432e-05,
      "loss": 0.2794,
      "step": 35660,
      "task_loss": 0.5987747311592102
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3169107437133789,
      "epoch": 12.89,
      "learning_rate": 2.2805521607837617e-05,
      "loss": 0.3253,
      "step": 35670,
      "task_loss": 0.4451930522918701
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28908562660217285,
      "epoch": 12.89,
      "learning_rate": 2.2739405427570775e-05,
      "loss": 0.3056,
      "step": 35680,
      "task_loss": 0.5222287178039551
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2468404620885849,
      "epoch": 12.9,
      "learning_rate": 2.267332668537104e-05,
      "loss": 0.3056,
      "step": 35690,
      "task_loss": 0.6599795818328857
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3190767168998718,
      "epoch": 12.9,
      "learning_rate": 2.260728572196264e-05,
      "loss": 0.3096,
      "step": 35700,
      "task_loss": 0.4459420442581177
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33378738164901733,
      "epoch": 12.91,
      "learning_rate": 2.2541282877874973e-05,
      "loss": 0.331,
      "step": 35710,
      "task_loss": 0.7016539573669434
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.36465150117874146,
      "epoch": 12.91,
      "learning_rate": 2.2475318493440918e-05,
      "loss": 0.3183,
      "step": 35720,
      "task_loss": 0.37818336486816406
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2557653784751892,
      "epoch": 12.91,
      "learning_rate": 2.2409392908794996e-05,
      "loss": 0.3124,
      "step": 35730,
      "task_loss": 0.4792822599411011
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26674747467041016,
      "epoch": 12.92,
      "learning_rate": 2.2343506463871732e-05,
      "loss": 0.2902,
      "step": 35740,
      "task_loss": 0.28261226415634155
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29377883672714233,
      "epoch": 12.92,
      "learning_rate": 2.227765949840374e-05,
      "loss": 0.3053,
      "step": 35750,
      "task_loss": 0.5766406059265137
    },
    {
      "epoch": 12.92,
      "eval_exact_match": 83.22611163670766,
      "eval_f1": 90.02511649594013,
      "step": 35750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23586252331733704,
      "epoch": 12.92,
      "learning_rate": 2.2211852351920155e-05,
      "loss": 0.275,
      "step": 35760,
      "task_loss": 0.46377095580101013
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25810280442237854,
      "epoch": 12.93,
      "learning_rate": 2.214608536374474e-05,
      "loss": 0.3178,
      "step": 35770,
      "task_loss": 0.4173620641231537
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2911950349807739,
      "epoch": 12.93,
      "learning_rate": 2.2080358872994204e-05,
      "loss": 0.3009,
      "step": 35780,
      "task_loss": 0.7365570664405823
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3325164318084717,
      "epoch": 12.93,
      "learning_rate": 2.201467321857642e-05,
      "loss": 0.3085,
      "step": 35790,
      "task_loss": 0.7347078323364258
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21316677331924438,
      "epoch": 12.94,
      "learning_rate": 2.1949028739188735e-05,
      "loss": 0.3124,
      "step": 35800,
      "task_loss": 0.2687242031097412
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3000262379646301,
      "epoch": 12.94,
      "learning_rate": 2.188342577331614e-05,
      "loss": 0.3113,
      "step": 35810,
      "task_loss": 0.494425892829895
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22712643444538116,
      "epoch": 12.95,
      "learning_rate": 2.1817864659229586e-05,
      "loss": 0.2945,
      "step": 35820,
      "task_loss": 0.3827125132083893
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2773548364639282,
      "epoch": 12.95,
      "learning_rate": 2.175234573498424e-05,
      "loss": 0.3116,
      "step": 35830,
      "task_loss": 0.37510883808135986
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34838151931762695,
      "epoch": 12.95,
      "learning_rate": 2.1686869338417688e-05,
      "loss": 0.2949,
      "step": 35840,
      "task_loss": 0.8953801989555359
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28336194157600403,
      "epoch": 12.96,
      "learning_rate": 2.162143580714827e-05,
      "loss": 0.3152,
      "step": 35850,
      "task_loss": 0.4684157371520996
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27176231145858765,
      "epoch": 12.96,
      "learning_rate": 2.1556045478573257e-05,
      "loss": 0.2948,
      "step": 35860,
      "task_loss": 0.5649899244308472
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3015035390853882,
      "epoch": 12.96,
      "learning_rate": 2.149069868986719e-05,
      "loss": 0.2936,
      "step": 35870,
      "task_loss": 0.7198494672775269
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2527949810028076,
      "epoch": 12.97,
      "learning_rate": 2.1425395777980063e-05,
      "loss": 0.3064,
      "step": 35880,
      "task_loss": 0.45133841037750244
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2298678755760193,
      "epoch": 12.97,
      "learning_rate": 2.136013707963567e-05,
      "loss": 0.2759,
      "step": 35890,
      "task_loss": 0.46662408113479614
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2663559913635254,
      "epoch": 12.97,
      "learning_rate": 2.1294922931329776e-05,
      "loss": 0.3132,
      "step": 35900,
      "task_loss": 0.3944147825241089
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28284984827041626,
      "epoch": 12.98,
      "learning_rate": 2.12297536693285e-05,
      "loss": 0.2986,
      "step": 35910,
      "task_loss": 0.6652886867523193
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.316769540309906,
      "epoch": 12.98,
      "learning_rate": 2.116462962966643e-05,
      "loss": 0.3136,
      "step": 35920,
      "task_loss": 0.6186679601669312
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26545512676239014,
      "epoch": 12.99,
      "learning_rate": 2.1099551148145046e-05,
      "loss": 0.3145,
      "step": 35930,
      "task_loss": 0.32127320766448975
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3045346140861511,
      "epoch": 12.99,
      "learning_rate": 2.103451856033086e-05,
      "loss": 0.3368,
      "step": 35940,
      "task_loss": 0.5651464462280273
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3307114243507385,
      "epoch": 12.99,
      "learning_rate": 2.0969532201553784e-05,
      "loss": 0.3048,
      "step": 35950,
      "task_loss": 0.7461162209510803
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23095571994781494,
      "epoch": 13.0,
      "learning_rate": 2.0904592406905308e-05,
      "loss": 0.3029,
      "step": 35960,
      "task_loss": 0.5822739601135254
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2591657042503357,
      "epoch": 13.0,
      "learning_rate": 2.083969951123688e-05,
      "loss": 0.3287,
      "step": 35970,
      "task_loss": 0.45596399903297424
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2697155773639679,
      "epoch": 13.0,
      "learning_rate": 2.077485384915807e-05,
      "loss": 0.3053,
      "step": 35980,
      "task_loss": 0.38982054591178894
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25783872604370117,
      "epoch": 13.01,
      "learning_rate": 2.071005575503493e-05,
      "loss": 0.2639,
      "step": 35990,
      "task_loss": 0.4523618221282959
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2397623360157013,
      "epoch": 13.01,
      "learning_rate": 2.0645305562988208e-05,
      "loss": 0.3041,
      "step": 36000,
      "task_loss": 0.26546239852905273
    },
    {
      "epoch": 13.01,
      "eval_exact_match": 83.08420056764427,
      "eval_f1": 89.96743931947285,
      "step": 36000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29155221581459045,
      "epoch": 13.01,
      "learning_rate": 2.0580603606891696e-05,
      "loss": 0.3011,
      "step": 36010,
      "task_loss": 0.455023854970932
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25799697637557983,
      "epoch": 13.02,
      "learning_rate": 2.0515950220370395e-05,
      "loss": 0.2589,
      "step": 36020,
      "task_loss": 0.9454857110977173
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24174681305885315,
      "epoch": 13.02,
      "learning_rate": 2.0451345736798946e-05,
      "loss": 0.2773,
      "step": 36030,
      "task_loss": 0.7607046365737915
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32882970571517944,
      "epoch": 13.02,
      "learning_rate": 2.0386790489299768e-05,
      "loss": 0.2842,
      "step": 36040,
      "task_loss": 0.4914087653160095
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37801340222358704,
      "epoch": 13.03,
      "learning_rate": 2.0322284810741438e-05,
      "loss": 0.2852,
      "step": 36050,
      "task_loss": 0.872128963470459
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29121482372283936,
      "epoch": 13.03,
      "learning_rate": 2.0257829033736913e-05,
      "loss": 0.2849,
      "step": 36060,
      "task_loss": 0.5817961096763611
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30120038986206055,
      "epoch": 13.04,
      "learning_rate": 2.0193423490641865e-05,
      "loss": 0.2789,
      "step": 36070,
      "task_loss": 0.8409297466278076
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2733917534351349,
      "epoch": 13.04,
      "learning_rate": 2.012906851355292e-05,
      "loss": 0.2832,
      "step": 36080,
      "task_loss": 0.7088493704795837
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34920522570610046,
      "epoch": 13.04,
      "learning_rate": 2.0064764434306e-05,
      "loss": 0.2774,
      "step": 36090,
      "task_loss": 0.8275156617164612
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2293846607208252,
      "epoch": 13.05,
      "learning_rate": 2.0000511584474526e-05,
      "loss": 0.2758,
      "step": 36100,
      "task_loss": 0.28478482365608215
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21923667192459106,
      "epoch": 13.05,
      "learning_rate": 1.9936310295367844e-05,
      "loss": 0.2872,
      "step": 36110,
      "task_loss": 0.6393476128578186
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23265597224235535,
      "epoch": 13.05,
      "learning_rate": 1.9872160898029345e-05,
      "loss": 0.2657,
      "step": 36120,
      "task_loss": 0.3859668970108032
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3707078695297241,
      "epoch": 13.06,
      "learning_rate": 1.9808063723234918e-05,
      "loss": 0.275,
      "step": 36130,
      "task_loss": 0.7055017948150635
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2488030344247818,
      "epoch": 13.06,
      "learning_rate": 1.9744019101491147e-05,
      "loss": 0.3085,
      "step": 36140,
      "task_loss": 0.3865114450454712
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2739602327346802,
      "epoch": 13.06,
      "learning_rate": 1.968002736303364e-05,
      "loss": 0.3111,
      "step": 36150,
      "task_loss": 0.4336225986480713
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24673353135585785,
      "epoch": 13.07,
      "learning_rate": 1.9616088837825307e-05,
      "loss": 0.277,
      "step": 36160,
      "task_loss": 0.626995325088501
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24530833959579468,
      "epoch": 13.07,
      "learning_rate": 1.9552203855554702e-05,
      "loss": 0.2805,
      "step": 36170,
      "task_loss": 0.31983304023742676
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.37968188524246216,
      "epoch": 13.08,
      "learning_rate": 1.948837274563426e-05,
      "loss": 0.2758,
      "step": 36180,
      "task_loss": 0.33091604709625244
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20261365175247192,
      "epoch": 13.08,
      "learning_rate": 1.9424595837198654e-05,
      "loss": 0.2734,
      "step": 36190,
      "task_loss": 0.5553550124168396
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2725498080253601,
      "epoch": 13.08,
      "learning_rate": 1.936087345910306e-05,
      "loss": 0.2921,
      "step": 36200,
      "task_loss": 0.3602065443992615
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22300642728805542,
      "epoch": 13.09,
      "learning_rate": 1.929720593992152e-05,
      "loss": 0.2541,
      "step": 36210,
      "task_loss": 0.37333545088768005
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2175058275461197,
      "epoch": 13.09,
      "learning_rate": 1.9233593607945127e-05,
      "loss": 0.2579,
      "step": 36220,
      "task_loss": 0.2774814963340759
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24256110191345215,
      "epoch": 13.09,
      "learning_rate": 1.917003679118049e-05,
      "loss": 0.2731,
      "step": 36230,
      "task_loss": 0.3967154920101166
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20085860788822174,
      "epoch": 13.1,
      "learning_rate": 1.9106535817347912e-05,
      "loss": 0.2741,
      "step": 36240,
      "task_loss": 0.39077961444854736
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25992414355278015,
      "epoch": 13.1,
      "learning_rate": 1.9043091013879773e-05,
      "loss": 0.2689,
      "step": 36250,
      "task_loss": 0.6975187063217163
    },
    {
      "epoch": 13.1,
      "eval_exact_match": 83.15042573320719,
      "eval_f1": 90.02105443305646,
      "step": 36250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27337756752967834,
      "epoch": 13.1,
      "learning_rate": 1.897970270791881e-05,
      "loss": 0.2923,
      "step": 36260,
      "task_loss": 0.6318705081939697
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25147658586502075,
      "epoch": 13.11,
      "learning_rate": 1.891637122631645e-05,
      "loss": 0.2631,
      "step": 36270,
      "task_loss": 0.4380658268928528
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24042510986328125,
      "epoch": 13.11,
      "learning_rate": 1.88530968956311e-05,
      "loss": 0.2943,
      "step": 36280,
      "task_loss": 0.4693765640258789
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2659061551094055,
      "epoch": 13.12,
      "learning_rate": 1.8789880042126502e-05,
      "loss": 0.2875,
      "step": 36290,
      "task_loss": 0.5884963274002075
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28706324100494385,
      "epoch": 13.12,
      "learning_rate": 1.8726720991769983e-05,
      "loss": 0.2792,
      "step": 36300,
      "task_loss": 0.6943851709365845
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27372080087661743,
      "epoch": 13.12,
      "learning_rate": 1.8663620070230873e-05,
      "loss": 0.3017,
      "step": 36310,
      "task_loss": 0.6589608192443848
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26573216915130615,
      "epoch": 13.13,
      "learning_rate": 1.8600577602878722e-05,
      "loss": 0.2827,
      "step": 36320,
      "task_loss": 0.418401300907135
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2142913043498993,
      "epoch": 13.13,
      "learning_rate": 1.8537593914781706e-05,
      "loss": 0.2583,
      "step": 36330,
      "task_loss": 1.0775578022003174
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33948519825935364,
      "epoch": 13.13,
      "learning_rate": 1.847466933070489e-05,
      "loss": 0.2959,
      "step": 36340,
      "task_loss": 0.8234223127365112
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2795005738735199,
      "epoch": 13.14,
      "learning_rate": 1.8411804175108595e-05,
      "loss": 0.2732,
      "step": 36350,
      "task_loss": 0.662077784538269
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.287386417388916,
      "epoch": 13.14,
      "learning_rate": 1.8348998772146698e-05,
      "loss": 0.2564,
      "step": 36360,
      "task_loss": 0.8269932270050049
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23500356078147888,
      "epoch": 13.14,
      "learning_rate": 1.828625344566498e-05,
      "loss": 0.2576,
      "step": 36370,
      "task_loss": 0.5284491777420044
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2579915523529053,
      "epoch": 13.15,
      "learning_rate": 1.822356851919943e-05,
      "loss": 0.2789,
      "step": 36380,
      "task_loss": 0.5953745245933533
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27623817324638367,
      "epoch": 13.15,
      "learning_rate": 1.816094431597464e-05,
      "loss": 0.2708,
      "step": 36390,
      "task_loss": 0.582027792930603
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23820829391479492,
      "epoch": 13.16,
      "learning_rate": 1.8098381158902023e-05,
      "loss": 0.2726,
      "step": 36400,
      "task_loss": 0.6655611395835876
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24597719311714172,
      "epoch": 13.16,
      "learning_rate": 1.803587937057828e-05,
      "loss": 0.2771,
      "step": 36410,
      "task_loss": 0.2989462912082672
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2530842423439026,
      "epoch": 13.16,
      "learning_rate": 1.7973439273283633e-05,
      "loss": 0.2647,
      "step": 36420,
      "task_loss": 0.3660958409309387
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2751024663448334,
      "epoch": 13.17,
      "learning_rate": 1.791106118898024e-05,
      "loss": 0.273,
      "step": 36430,
      "task_loss": 0.7918814420700073
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21775180101394653,
      "epoch": 13.17,
      "learning_rate": 1.7848745439310454e-05,
      "loss": 0.251,
      "step": 36440,
      "task_loss": 0.6314330697059631
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23425433039665222,
      "epoch": 13.17,
      "learning_rate": 1.7786492345595258e-05,
      "loss": 0.2819,
      "step": 36450,
      "task_loss": 0.5131993293762207
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24818137288093567,
      "epoch": 13.18,
      "learning_rate": 1.7724302228832518e-05,
      "loss": 0.2778,
      "step": 36460,
      "task_loss": 0.5289952754974365
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2537691295146942,
      "epoch": 13.18,
      "learning_rate": 1.7662175409695403e-05,
      "loss": 0.3113,
      "step": 36470,
      "task_loss": 0.515838623046875
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4750899076461792,
      "epoch": 13.18,
      "learning_rate": 1.760011220853067e-05,
      "loss": 0.3064,
      "step": 36480,
      "task_loss": 0.7261573672294617
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26949992775917053,
      "epoch": 13.19,
      "learning_rate": 1.7538112945357074e-05,
      "loss": 0.2587,
      "step": 36490,
      "task_loss": 0.7647272348403931
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22216796875,
      "epoch": 13.19,
      "learning_rate": 1.747617793986364e-05,
      "loss": 0.273,
      "step": 36500,
      "task_loss": 0.3125956952571869
    },
    {
      "epoch": 13.19,
      "eval_exact_match": 83.4247871333964,
      "eval_f1": 90.10237204205706,
      "step": 36500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2604176998138428,
      "epoch": 13.19,
      "learning_rate": 1.7414307511408105e-05,
      "loss": 0.2708,
      "step": 36510,
      "task_loss": 0.7602921724319458
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32731837034225464,
      "epoch": 13.2,
      "learning_rate": 1.7352501979015185e-05,
      "loss": 0.2674,
      "step": 36520,
      "task_loss": 0.29305794835090637
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30296820402145386,
      "epoch": 13.2,
      "learning_rate": 1.7290761661374998e-05,
      "loss": 0.2957,
      "step": 36530,
      "task_loss": 0.5144740343093872
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23465166985988617,
      "epoch": 13.21,
      "learning_rate": 1.722908687684138e-05,
      "loss": 0.254,
      "step": 36540,
      "task_loss": 0.3443576693534851
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28396230936050415,
      "epoch": 13.21,
      "learning_rate": 1.7167477943430275e-05,
      "loss": 0.2687,
      "step": 36550,
      "task_loss": 0.6265298128128052
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24885854125022888,
      "epoch": 13.21,
      "learning_rate": 1.710593517881802e-05,
      "loss": 0.2967,
      "step": 36560,
      "task_loss": 0.49757009744644165
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2213917374610901,
      "epoch": 13.22,
      "learning_rate": 1.7044458900339855e-05,
      "loss": 0.2624,
      "step": 36570,
      "task_loss": 0.45970025658607483
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23916515707969666,
      "epoch": 13.22,
      "learning_rate": 1.6983049424988106e-05,
      "loss": 0.2659,
      "step": 36580,
      "task_loss": 0.7370936870574951
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2173445075750351,
      "epoch": 13.22,
      "learning_rate": 1.6921707069410698e-05,
      "loss": 0.2602,
      "step": 36590,
      "task_loss": 0.3339085578918457
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25310006737709045,
      "epoch": 13.23,
      "learning_rate": 1.686043214990943e-05,
      "loss": 0.2725,
      "step": 36600,
      "task_loss": 0.5834509134292603
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2566218674182892,
      "epoch": 13.23,
      "learning_rate": 1.6799224982438413e-05,
      "loss": 0.2721,
      "step": 36610,
      "task_loss": 0.3749829828739166
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2179889976978302,
      "epoch": 13.23,
      "learning_rate": 1.6738085882602358e-05,
      "loss": 0.3091,
      "step": 36620,
      "task_loss": 0.4416264295578003
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25685155391693115,
      "epoch": 13.24,
      "learning_rate": 1.6677015165655034e-05,
      "loss": 0.2846,
      "step": 36630,
      "task_loss": 0.3986138701438904
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3165907859802246,
      "epoch": 13.24,
      "learning_rate": 1.6616013146497598e-05,
      "loss": 0.3063,
      "step": 36640,
      "task_loss": 0.36311423778533936
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21515513956546783,
      "epoch": 13.25,
      "learning_rate": 1.6555080139676972e-05,
      "loss": 0.2745,
      "step": 36650,
      "task_loss": 0.3444489538669586
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28814902901649475,
      "epoch": 13.25,
      "learning_rate": 1.6494216459384225e-05,
      "loss": 0.2534,
      "step": 36660,
      "task_loss": 0.6024624705314636
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24540041387081146,
      "epoch": 13.25,
      "learning_rate": 1.6433422419452973e-05,
      "loss": 0.2658,
      "step": 36670,
      "task_loss": 0.6741258502006531
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22696809470653534,
      "epoch": 13.26,
      "learning_rate": 1.6372698333357705e-05,
      "loss": 0.266,
      "step": 36680,
      "task_loss": 0.4832145571708679
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31842273473739624,
      "epoch": 13.26,
      "learning_rate": 1.6312044514212245e-05,
      "loss": 0.2864,
      "step": 36690,
      "task_loss": 0.5585686564445496
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28510862588882446,
      "epoch": 13.26,
      "learning_rate": 1.625146127476807e-05,
      "loss": 0.2635,
      "step": 36700,
      "task_loss": 0.5174915790557861
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26577115058898926,
      "epoch": 13.27,
      "learning_rate": 1.6190948927412758e-05,
      "loss": 0.2706,
      "step": 36710,
      "task_loss": 0.7755802869796753
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21998828649520874,
      "epoch": 13.27,
      "learning_rate": 1.613050778416828e-05,
      "loss": 0.2745,
      "step": 36720,
      "task_loss": 0.4503459632396698
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34111788868904114,
      "epoch": 13.27,
      "learning_rate": 1.607013815668954e-05,
      "loss": 0.3072,
      "step": 36730,
      "task_loss": 0.6373797655105591
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18357953429222107,
      "epoch": 13.28,
      "learning_rate": 1.6009840356262603e-05,
      "loss": 0.2499,
      "step": 36740,
      "task_loss": 0.23571716248989105
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19039833545684814,
      "epoch": 13.28,
      "learning_rate": 1.594961469380322e-05,
      "loss": 0.2918,
      "step": 36750,
      "task_loss": 0.5455808043479919
    },
    {
      "epoch": 13.28,
      "eval_exact_match": 82.98959318826869,
      "eval_f1": 89.86127481224247,
      "step": 36750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2975670099258423,
      "epoch": 13.29,
      "learning_rate": 1.5889461479855153e-05,
      "loss": 0.2895,
      "step": 36760,
      "task_loss": 0.6404368281364441
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2604735493659973,
      "epoch": 13.29,
      "learning_rate": 1.5829381024588614e-05,
      "loss": 0.2847,
      "step": 36770,
      "task_loss": 0.5445793867111206
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23690715432167053,
      "epoch": 13.29,
      "learning_rate": 1.576937363779861e-05,
      "loss": 0.2696,
      "step": 36780,
      "task_loss": 0.5591181516647339
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2577180862426758,
      "epoch": 13.3,
      "learning_rate": 1.570943962890342e-05,
      "loss": 0.2738,
      "step": 36790,
      "task_loss": 0.5541560649871826
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2612329125404358,
      "epoch": 13.3,
      "learning_rate": 1.5649579306942943e-05,
      "loss": 0.27,
      "step": 36800,
      "task_loss": 0.48154765367507935
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25569772720336914,
      "epoch": 13.3,
      "learning_rate": 1.558979298057715e-05,
      "loss": 0.2397,
      "step": 36810,
      "task_loss": 0.5663192868232727
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24582552909851074,
      "epoch": 13.31,
      "learning_rate": 1.55300809580844e-05,
      "loss": 0.2712,
      "step": 36820,
      "task_loss": 0.39500463008880615
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26157790422439575,
      "epoch": 13.31,
      "learning_rate": 1.5470443547360003e-05,
      "loss": 0.2886,
      "step": 36830,
      "task_loss": 0.5957146883010864
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3169706463813782,
      "epoch": 13.31,
      "learning_rate": 1.5410881055914478e-05,
      "loss": 0.3012,
      "step": 36840,
      "task_loss": 0.5682312846183777
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2886982858181,
      "epoch": 13.32,
      "learning_rate": 1.5351393790872065e-05,
      "loss": 0.2565,
      "step": 36850,
      "task_loss": 0.5027368664741516
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2950834631919861,
      "epoch": 13.32,
      "learning_rate": 1.529198205896911e-05,
      "loss": 0.2677,
      "step": 36860,
      "task_loss": 0.46899813413619995
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22698701918125153,
      "epoch": 13.32,
      "learning_rate": 1.5232646166552486e-05,
      "loss": 0.2933,
      "step": 36870,
      "task_loss": 0.334617555141449
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2680467665195465,
      "epoch": 13.33,
      "learning_rate": 1.517338641957799e-05,
      "loss": 0.2625,
      "step": 36880,
      "task_loss": 0.32778269052505493
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18022421002388,
      "epoch": 13.33,
      "learning_rate": 1.5114203123608817e-05,
      "loss": 0.3041,
      "step": 36890,
      "task_loss": 0.47484010457992554
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2711172103881836,
      "epoch": 13.34,
      "learning_rate": 1.5055096583813937e-05,
      "loss": 0.2648,
      "step": 36900,
      "task_loss": 0.6222798824310303
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33530765771865845,
      "epoch": 13.34,
      "learning_rate": 1.499606710496656e-05,
      "loss": 0.2891,
      "step": 36910,
      "task_loss": 0.5925235152244568
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33584457635879517,
      "epoch": 13.34,
      "learning_rate": 1.4937114991442491e-05,
      "loss": 0.2857,
      "step": 36920,
      "task_loss": 0.4298863410949707
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28718897700309753,
      "epoch": 13.35,
      "learning_rate": 1.48782405472187e-05,
      "loss": 0.3004,
      "step": 36930,
      "task_loss": 0.34843045473098755
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3173147737979889,
      "epoch": 13.35,
      "learning_rate": 1.4819444075871585e-05,
      "loss": 0.2667,
      "step": 36940,
      "task_loss": 0.42160385847091675
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27303415536880493,
      "epoch": 13.35,
      "learning_rate": 1.4760725880575534e-05,
      "loss": 0.2751,
      "step": 36950,
      "task_loss": 0.508614718914032
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23166467249393463,
      "epoch": 13.36,
      "learning_rate": 1.4702086264101305e-05,
      "loss": 0.2968,
      "step": 36960,
      "task_loss": 0.6358349323272705
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2538238763809204,
      "epoch": 13.36,
      "learning_rate": 1.4643525528814492e-05,
      "loss": 0.299,
      "step": 36970,
      "task_loss": 0.6361325979232788
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.16733571887016296,
      "epoch": 13.36,
      "learning_rate": 1.4585043976673916e-05,
      "loss": 0.2509,
      "step": 36980,
      "task_loss": 0.2583733797073364
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30147695541381836,
      "epoch": 13.37,
      "learning_rate": 1.4526641909230136e-05,
      "loss": 0.3064,
      "step": 36990,
      "task_loss": 0.45961910486221313
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33842676877975464,
      "epoch": 13.37,
      "learning_rate": 1.446831962762385e-05,
      "loss": 0.265,
      "step": 37000,
      "task_loss": 0.7715430855751038
    },
    {
      "epoch": 13.37,
      "eval_exact_match": 83.28287606433302,
      "eval_f1": 90.0359133389191,
      "step": 37000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24696779251098633,
      "epoch": 13.38,
      "learning_rate": 1.4410077432584366e-05,
      "loss": 0.2908,
      "step": 37010,
      "task_loss": 0.3677746057510376
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2108057141304016,
      "epoch": 13.38,
      "learning_rate": 1.435191562442799e-05,
      "loss": 0.257,
      "step": 37020,
      "task_loss": 0.38845571875572205
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18310099840164185,
      "epoch": 13.38,
      "learning_rate": 1.4293834503056611e-05,
      "loss": 0.2736,
      "step": 37030,
      "task_loss": 0.3544849753379822
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26856935024261475,
      "epoch": 13.39,
      "learning_rate": 1.4235834367955988e-05,
      "loss": 0.2669,
      "step": 37040,
      "task_loss": 0.813249945640564
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27939939498901367,
      "epoch": 13.39,
      "learning_rate": 1.417791551819433e-05,
      "loss": 0.2738,
      "step": 37050,
      "task_loss": 0.4761803448200226
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2665058970451355,
      "epoch": 13.39,
      "learning_rate": 1.4120078252420704e-05,
      "loss": 0.2915,
      "step": 37060,
      "task_loss": 0.35540011525154114
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2548675835132599,
      "epoch": 13.4,
      "learning_rate": 1.4062322868863515e-05,
      "loss": 0.3536,
      "step": 37070,
      "task_loss": 0.5708368420600891
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2190103977918625,
      "epoch": 13.4,
      "learning_rate": 1.4004649665328914e-05,
      "loss": 0.2515,
      "step": 37080,
      "task_loss": 0.5469504594802856
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1838226318359375,
      "epoch": 13.4,
      "learning_rate": 1.3947058939199343e-05,
      "loss": 0.2486,
      "step": 37090,
      "task_loss": 0.3871685862541199
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1938687115907669,
      "epoch": 13.41,
      "learning_rate": 1.388955098743196e-05,
      "loss": 0.2762,
      "step": 37100,
      "task_loss": 0.6385823488235474
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26518717408180237,
      "epoch": 13.41,
      "learning_rate": 1.3832126106557103e-05,
      "loss": 0.288,
      "step": 37110,
      "task_loss": 0.784468412399292
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20624689757823944,
      "epoch": 13.42,
      "learning_rate": 1.377478459267674e-05,
      "loss": 0.2598,
      "step": 37120,
      "task_loss": 0.3261072635650635
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1749204695224762,
      "epoch": 13.42,
      "learning_rate": 1.3717526741463045e-05,
      "loss": 0.2694,
      "step": 37130,
      "task_loss": 0.3123036324977875
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3422029912471771,
      "epoch": 13.42,
      "learning_rate": 1.3660352848156717e-05,
      "loss": 0.2659,
      "step": 37140,
      "task_loss": 0.37898117303848267
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21857774257659912,
      "epoch": 13.43,
      "learning_rate": 1.3603263207565584e-05,
      "loss": 0.2627,
      "step": 37150,
      "task_loss": 0.3548792600631714
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3651176691055298,
      "epoch": 13.43,
      "learning_rate": 1.3546258114063033e-05,
      "loss": 0.2737,
      "step": 37160,
      "task_loss": 0.5343203544616699
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24481065571308136,
      "epoch": 13.43,
      "learning_rate": 1.3489337861586507e-05,
      "loss": 0.2748,
      "step": 37170,
      "task_loss": 0.40071243047714233
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2896316349506378,
      "epoch": 13.44,
      "learning_rate": 1.3432502743635948e-05,
      "loss": 0.2827,
      "step": 37180,
      "task_loss": 0.4923239052295685
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23948924243450165,
      "epoch": 13.44,
      "learning_rate": 1.3375753053272343e-05,
      "loss": 0.274,
      "step": 37190,
      "task_loss": 0.47201618552207947
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23711787164211273,
      "epoch": 13.44,
      "learning_rate": 1.3319089083116176e-05,
      "loss": 0.2608,
      "step": 37200,
      "task_loss": 0.49309107661247253
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33351051807403564,
      "epoch": 13.45,
      "learning_rate": 1.326251112534595e-05,
      "loss": 0.2907,
      "step": 37210,
      "task_loss": 0.7690310478210449
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30798155069351196,
      "epoch": 13.45,
      "learning_rate": 1.32060194716966e-05,
      "loss": 0.2686,
      "step": 37220,
      "task_loss": 0.9489891529083252
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2473529577255249,
      "epoch": 13.46,
      "learning_rate": 1.3149614413458139e-05,
      "loss": 0.2617,
      "step": 37230,
      "task_loss": 0.7801364660263062
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3227131962776184,
      "epoch": 13.46,
      "learning_rate": 1.3093296241473975e-05,
      "loss": 0.2915,
      "step": 37240,
      "task_loss": 0.5432208180427551
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24519944190979004,
      "epoch": 13.46,
      "learning_rate": 1.3037065246139558e-05,
      "loss": 0.2713,
      "step": 37250,
      "task_loss": 0.3368620276451111
    },
    {
      "epoch": 13.46,
      "eval_exact_match": 83.4247871333964,
      "eval_f1": 90.01338222798685,
      "step": 37250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31522136926651,
      "epoch": 13.47,
      "learning_rate": 1.2980921717400803e-05,
      "loss": 0.2731,
      "step": 37260,
      "task_loss": 0.7898993492126465
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27051877975463867,
      "epoch": 13.47,
      "learning_rate": 1.292486594475264e-05,
      "loss": 0.2496,
      "step": 37270,
      "task_loss": 0.5739043354988098
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33544689416885376,
      "epoch": 13.47,
      "learning_rate": 1.2868898217237457e-05,
      "loss": 0.2742,
      "step": 37280,
      "task_loss": 0.44129830598831177
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20862698554992676,
      "epoch": 13.48,
      "learning_rate": 1.2813018823443685e-05,
      "loss": 0.2452,
      "step": 37290,
      "task_loss": 0.27721577882766724
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24522057175636292,
      "epoch": 13.48,
      "learning_rate": 1.2757228051504276e-05,
      "loss": 0.2757,
      "step": 37300,
      "task_loss": 0.5336194038391113
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2989899516105652,
      "epoch": 13.48,
      "learning_rate": 1.2701526189095216e-05,
      "loss": 0.2576,
      "step": 37310,
      "task_loss": 0.4906134009361267
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3206063210964203,
      "epoch": 13.49,
      "learning_rate": 1.2645913523434012e-05,
      "loss": 0.2517,
      "step": 37320,
      "task_loss": 0.6886972188949585
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2796739339828491,
      "epoch": 13.49,
      "learning_rate": 1.2590390341278302e-05,
      "loss": 0.2925,
      "step": 37330,
      "task_loss": 0.3325718641281128
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3092641830444336,
      "epoch": 13.49,
      "learning_rate": 1.2534956928924263e-05,
      "loss": 0.2733,
      "step": 37340,
      "task_loss": 0.41676968336105347
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23041902482509613,
      "epoch": 13.5,
      "learning_rate": 1.2479613572205212e-05,
      "loss": 0.2796,
      "step": 37350,
      "task_loss": 0.4734857678413391
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2079801857471466,
      "epoch": 13.5,
      "learning_rate": 1.2424360556490109e-05,
      "loss": 0.2577,
      "step": 37360,
      "task_loss": 0.3639516532421112
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24474698305130005,
      "epoch": 13.51,
      "learning_rate": 1.2369198166682094e-05,
      "loss": 0.276,
      "step": 37370,
      "task_loss": 0.4654710292816162
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2466435730457306,
      "epoch": 13.51,
      "learning_rate": 1.2314126687216974e-05,
      "loss": 0.2784,
      "step": 37380,
      "task_loss": 0.3153854012489319
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2671513855457306,
      "epoch": 13.51,
      "learning_rate": 1.2259146402061829e-05,
      "loss": 0.2773,
      "step": 37390,
      "task_loss": 0.6717286109924316
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31550562381744385,
      "epoch": 13.52,
      "learning_rate": 1.2204257594713501e-05,
      "loss": 0.2574,
      "step": 37400,
      "task_loss": 0.389668345451355
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2816436290740967,
      "epoch": 13.52,
      "learning_rate": 1.2149460548197143e-05,
      "loss": 0.2817,
      "step": 37410,
      "task_loss": 0.5112844705581665
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2804926037788391,
      "epoch": 13.52,
      "learning_rate": 1.2094755545064719e-05,
      "loss": 0.2612,
      "step": 37420,
      "task_loss": 0.4115559458732605
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3953973054885864,
      "epoch": 13.53,
      "learning_rate": 1.2040142867393671e-05,
      "loss": 0.2852,
      "step": 37430,
      "task_loss": 0.3008333444595337
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3041832149028778,
      "epoch": 13.53,
      "learning_rate": 1.1985622796785295e-05,
      "loss": 0.2876,
      "step": 37440,
      "task_loss": 0.303343802690506
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19918349385261536,
      "epoch": 13.53,
      "learning_rate": 1.1931195614363417e-05,
      "loss": 0.2762,
      "step": 37450,
      "task_loss": 0.2883751392364502
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3426469564437866,
      "epoch": 13.54,
      "learning_rate": 1.1876861600772893e-05,
      "loss": 0.3045,
      "step": 37460,
      "task_loss": 0.582063615322113
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1825534999370575,
      "epoch": 13.54,
      "learning_rate": 1.1822621036178173e-05,
      "loss": 0.2802,
      "step": 37470,
      "task_loss": 0.3473503589630127
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2241651713848114,
      "epoch": 13.55,
      "learning_rate": 1.1768474200261838e-05,
      "loss": 0.2738,
      "step": 37480,
      "task_loss": 0.5678466558456421
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.288207471370697,
      "epoch": 13.55,
      "learning_rate": 1.1714421372223179e-05,
      "loss": 0.275,
      "step": 37490,
      "task_loss": 0.5218504667282104
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24978035688400269,
      "epoch": 13.55,
      "learning_rate": 1.1660462830776766e-05,
      "loss": 0.2603,
      "step": 37500,
      "task_loss": 0.4382033050060272
    },
    {
      "epoch": 13.55,
      "eval_exact_match": 83.32071901608326,
      "eval_f1": 89.90574640557605,
      "step": 37500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21151039004325867,
      "epoch": 13.56,
      "learning_rate": 1.1606598854150983e-05,
      "loss": 0.2647,
      "step": 37510,
      "task_loss": 0.38994455337524414
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2562404274940491,
      "epoch": 13.56,
      "learning_rate": 1.1552829720086581e-05,
      "loss": 0.2566,
      "step": 37520,
      "task_loss": 0.6831725835800171
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21861696243286133,
      "epoch": 13.56,
      "learning_rate": 1.1499155705835338e-05,
      "loss": 0.2516,
      "step": 37530,
      "task_loss": 0.24222993850708008
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2870420217514038,
      "epoch": 13.57,
      "learning_rate": 1.1445577088158486e-05,
      "loss": 0.2781,
      "step": 37540,
      "task_loss": 0.4957524240016937
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25489604473114014,
      "epoch": 13.57,
      "learning_rate": 1.1392094143325404e-05,
      "loss": 0.2559,
      "step": 37550,
      "task_loss": 0.3785157799720764
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23829561471939087,
      "epoch": 13.57,
      "learning_rate": 1.1338707147112145e-05,
      "loss": 0.2596,
      "step": 37560,
      "task_loss": 0.49412262439727783
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2603868544101715,
      "epoch": 13.58,
      "learning_rate": 1.1285416374800018e-05,
      "loss": 0.2809,
      "step": 37570,
      "task_loss": 0.42271482944488525
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2052895724773407,
      "epoch": 13.58,
      "learning_rate": 1.1232222101174148e-05,
      "loss": 0.2781,
      "step": 37580,
      "task_loss": 0.18088670074939728
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3106672763824463,
      "epoch": 13.59,
      "learning_rate": 1.1179124600522107e-05,
      "loss": 0.273,
      "step": 37590,
      "task_loss": 0.551022469997406
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29829642176628113,
      "epoch": 13.59,
      "learning_rate": 1.1126124146632464e-05,
      "loss": 0.2803,
      "step": 37600,
      "task_loss": 0.4130031168460846
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29676663875579834,
      "epoch": 13.59,
      "learning_rate": 1.1073221012793393e-05,
      "loss": 0.29,
      "step": 37610,
      "task_loss": 0.2675638198852539
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31227415800094604,
      "epoch": 13.6,
      "learning_rate": 1.102041547179121e-05,
      "loss": 0.2634,
      "step": 37620,
      "task_loss": 0.5230624675750732
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22656656801700592,
      "epoch": 13.6,
      "learning_rate": 1.0967707795909077e-05,
      "loss": 0.264,
      "step": 37630,
      "task_loss": 0.4179735481739044
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2572036385536194,
      "epoch": 13.6,
      "learning_rate": 1.0915098256925474e-05,
      "loss": 0.2931,
      "step": 37640,
      "task_loss": 0.5777921676635742
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22712549567222595,
      "epoch": 13.61,
      "learning_rate": 1.0862587126112873e-05,
      "loss": 0.261,
      "step": 37650,
      "task_loss": 0.4374583065509796
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3408048748970032,
      "epoch": 13.61,
      "learning_rate": 1.0810174674236335e-05,
      "loss": 0.3011,
      "step": 37660,
      "task_loss": 0.6254220008850098
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2701667547225952,
      "epoch": 13.61,
      "learning_rate": 1.0757861171552074e-05,
      "loss": 0.2735,
      "step": 37670,
      "task_loss": 0.7898134589195251
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2562268078327179,
      "epoch": 13.62,
      "learning_rate": 1.0705646887806119e-05,
      "loss": 0.2414,
      "step": 37680,
      "task_loss": 0.7383904457092285
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2719931900501251,
      "epoch": 13.62,
      "learning_rate": 1.0653532092232843e-05,
      "loss": 0.2863,
      "step": 37690,
      "task_loss": 0.7737250924110413
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27693867683410645,
      "epoch": 13.62,
      "learning_rate": 1.0601517053553697e-05,
      "loss": 0.3029,
      "step": 37700,
      "task_loss": 0.4310859739780426
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2542468011379242,
      "epoch": 13.63,
      "learning_rate": 1.0549602039975683e-05,
      "loss": 0.2914,
      "step": 37710,
      "task_loss": 0.4154024124145508
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2769840955734253,
      "epoch": 13.63,
      "learning_rate": 1.0497787319190076e-05,
      "loss": 0.2727,
      "step": 37720,
      "task_loss": 0.5301538705825806
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23731346428394318,
      "epoch": 13.64,
      "learning_rate": 1.0446073158370996e-05,
      "loss": 0.287,
      "step": 37730,
      "task_loss": 0.6987727880477905
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3362783193588257,
      "epoch": 13.64,
      "learning_rate": 1.0394459824174065e-05,
      "loss": 0.2774,
      "step": 37740,
      "task_loss": 1.005785584449768
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.33067774772644043,
      "epoch": 13.64,
      "learning_rate": 1.0342947582734962e-05,
      "loss": 0.2619,
      "step": 37750,
      "task_loss": 0.7543010711669922
    },
    {
      "epoch": 13.64,
      "eval_exact_match": 83.12204351939451,
      "eval_f1": 89.86274805388804,
      "step": 37750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2364063858985901,
      "epoch": 13.65,
      "learning_rate": 1.029153669966814e-05,
      "loss": 0.2825,
      "step": 37760,
      "task_loss": 0.892463207244873
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25400644540786743,
      "epoch": 13.65,
      "learning_rate": 1.02402274400654e-05,
      "loss": 0.2813,
      "step": 37770,
      "task_loss": 0.4577575623989105
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18836143612861633,
      "epoch": 13.65,
      "learning_rate": 1.0189020068494543e-05,
      "loss": 0.2781,
      "step": 37780,
      "task_loss": 0.47670626640319824
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3060493469238281,
      "epoch": 13.66,
      "learning_rate": 1.0137914848997974e-05,
      "loss": 0.2802,
      "step": 37790,
      "task_loss": 0.5032558441162109
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21806681156158447,
      "epoch": 13.66,
      "learning_rate": 1.0086912045091422e-05,
      "loss": 0.2802,
      "step": 37800,
      "task_loss": 0.6232914328575134
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3554821014404297,
      "epoch": 13.66,
      "learning_rate": 1.0036011919762471e-05,
      "loss": 0.2783,
      "step": 37810,
      "task_loss": 1.0437252521514893
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.17857030034065247,
      "epoch": 13.67,
      "learning_rate": 9.98521473546929e-06,
      "loss": 0.2551,
      "step": 37820,
      "task_loss": 0.293536901473999
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2957502007484436,
      "epoch": 13.67,
      "learning_rate": 9.93452075413924e-06,
      "loss": 0.2708,
      "step": 37830,
      "task_loss": 0.6464605331420898
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.266282856464386,
      "epoch": 13.68,
      "learning_rate": 9.88393023716755e-06,
      "loss": 0.2543,
      "step": 37840,
      "task_loss": 0.849335789680481
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.233785018324852,
      "epoch": 13.68,
      "learning_rate": 9.833443445415917e-06,
      "loss": 0.2606,
      "step": 37850,
      "task_loss": 1.1107016801834106
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24783793091773987,
      "epoch": 13.68,
      "learning_rate": 9.783060639211225e-06,
      "loss": 0.3038,
      "step": 37860,
      "task_loss": 0.616372287273407
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24006393551826477,
      "epoch": 13.69,
      "learning_rate": 9.732782078344166e-06,
      "loss": 0.2648,
      "step": 37870,
      "task_loss": 0.7905951142311096
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.4207942485809326,
      "epoch": 13.69,
      "learning_rate": 9.682608022067903e-06,
      "loss": 0.3115,
      "step": 37880,
      "task_loss": 1.0007433891296387
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2837132215499878,
      "epoch": 13.69,
      "learning_rate": 9.632538729096749e-06,
      "loss": 0.25,
      "step": 37890,
      "task_loss": 0.38253968954086304
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23701059818267822,
      "epoch": 13.7,
      "learning_rate": 9.58257445760482e-06,
      "loss": 0.2581,
      "step": 37900,
      "task_loss": 0.3679601550102234
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24502034485340118,
      "epoch": 13.7,
      "learning_rate": 9.532715465224678e-06,
      "loss": 0.2914,
      "step": 37910,
      "task_loss": 0.435784250497818
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28633737564086914,
      "epoch": 13.7,
      "learning_rate": 9.482962009046073e-06,
      "loss": 0.2968,
      "step": 37920,
      "task_loss": 0.4898184537887573
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2632879912853241,
      "epoch": 13.71,
      "learning_rate": 9.433314345614548e-06,
      "loss": 0.2611,
      "step": 37930,
      "task_loss": 0.6624996066093445
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1981096863746643,
      "epoch": 13.71,
      "learning_rate": 9.383772730930163e-06,
      "loss": 0.2542,
      "step": 37940,
      "task_loss": 0.4208466112613678
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25894618034362793,
      "epoch": 13.72,
      "learning_rate": 9.334337420446116e-06,
      "loss": 0.3009,
      "step": 37950,
      "task_loss": 0.699665367603302
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32491040229797363,
      "epoch": 13.72,
      "learning_rate": 9.285008669067524e-06,
      "loss": 0.2866,
      "step": 37960,
      "task_loss": 0.42975303530693054
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23986268043518066,
      "epoch": 13.72,
      "learning_rate": 9.23578673114999e-06,
      "loss": 0.265,
      "step": 37970,
      "task_loss": 0.4938848614692688
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2247878909111023,
      "epoch": 13.73,
      "learning_rate": 9.18667186049839e-06,
      "loss": 0.2574,
      "step": 37980,
      "task_loss": 0.4352090060710907
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3065471649169922,
      "epoch": 13.73,
      "learning_rate": 9.14256022874918e-06,
      "loss": 0.2714,
      "step": 37990,
      "task_loss": 0.6363136768341064
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27550625801086426,
      "epoch": 13.73,
      "learning_rate": 9.09364948315695e-06,
      "loss": 0.2897,
      "step": 38000,
      "task_loss": 0.688079833984375
    },
    {
      "epoch": 13.73,
      "eval_exact_match": 83.15988647114474,
      "eval_f1": 89.95458217564638,
      "step": 38000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22112786769866943,
      "epoch": 13.74,
      "learning_rate": 9.044846537738092e-06,
      "loss": 0.2722,
      "step": 38010,
      "task_loss": 0.36097025871276855
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20671901106834412,
      "epoch": 13.74,
      "learning_rate": 8.996151644137009e-06,
      "loss": 0.2845,
      "step": 38020,
      "task_loss": 0.3453246057033539
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21796171367168427,
      "epoch": 13.74,
      "learning_rate": 8.947565053440948e-06,
      "loss": 0.2786,
      "step": 38030,
      "task_loss": 0.25846922397613525
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2829256057739258,
      "epoch": 13.75,
      "learning_rate": 8.89908701617872e-06,
      "loss": 0.2858,
      "step": 38040,
      "task_loss": 0.7385716438293457
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28261667490005493,
      "epoch": 13.75,
      "learning_rate": 8.850717782319354e-06,
      "loss": 0.2765,
      "step": 38050,
      "task_loss": 0.36890166997909546
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23844432830810547,
      "epoch": 13.75,
      "learning_rate": 8.802457601270925e-06,
      "loss": 0.2739,
      "step": 38060,
      "task_loss": 0.4469187259674072
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2905295789241791,
      "epoch": 13.76,
      "learning_rate": 8.754306721879127e-06,
      "loss": 0.2875,
      "step": 38070,
      "task_loss": 0.6334402561187744
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25112390518188477,
      "epoch": 13.76,
      "learning_rate": 8.706265392426096e-06,
      "loss": 0.2441,
      "step": 38080,
      "task_loss": 0.2148028463125229
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18571755290031433,
      "epoch": 13.77,
      "learning_rate": 8.658333860629078e-06,
      "loss": 0.2691,
      "step": 38090,
      "task_loss": 0.4117751121520996
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29254400730133057,
      "epoch": 13.77,
      "learning_rate": 8.610512373639182e-06,
      "loss": 0.2785,
      "step": 38100,
      "task_loss": 0.7656481862068176
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31580865383148193,
      "epoch": 13.77,
      "learning_rate": 8.562801178040054e-06,
      "loss": 0.2624,
      "step": 38110,
      "task_loss": 0.42948007583618164
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28428199887275696,
      "epoch": 13.78,
      "learning_rate": 8.515200519846676e-06,
      "loss": 0.2668,
      "step": 38120,
      "task_loss": 0.7075586915016174
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20711860060691833,
      "epoch": 13.78,
      "learning_rate": 8.46771064450404e-06,
      "loss": 0.2627,
      "step": 38130,
      "task_loss": 0.5416389107704163
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24480155110359192,
      "epoch": 13.78,
      "learning_rate": 8.420331796885931e-06,
      "loss": 0.2857,
      "step": 38140,
      "task_loss": 0.6288049221038818
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18393155932426453,
      "epoch": 13.79,
      "learning_rate": 8.373064221293584e-06,
      "loss": 0.2625,
      "step": 38150,
      "task_loss": 0.23807212710380554
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23062025010585785,
      "epoch": 13.79,
      "learning_rate": 8.325908161454554e-06,
      "loss": 0.2629,
      "step": 38160,
      "task_loss": 0.3707573115825653
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24218732118606567,
      "epoch": 13.79,
      "learning_rate": 8.278863860521317e-06,
      "loss": 0.2774,
      "step": 38170,
      "task_loss": 0.48844727873802185
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.16815409064292908,
      "epoch": 13.8,
      "learning_rate": 8.231931561070113e-06,
      "loss": 0.2581,
      "step": 38180,
      "task_loss": 0.25500476360321045
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24209289252758026,
      "epoch": 13.8,
      "learning_rate": 8.185111505099662e-06,
      "loss": 0.2689,
      "step": 38190,
      "task_loss": 0.7029504776000977
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23915712535381317,
      "epoch": 13.81,
      "learning_rate": 8.138403934029924e-06,
      "loss": 0.2745,
      "step": 38200,
      "task_loss": 0.34845811128616333
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25952520966529846,
      "epoch": 13.81,
      "learning_rate": 8.091809088700822e-06,
      "loss": 0.2516,
      "step": 38210,
      "task_loss": 0.8645590543746948
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25939249992370605,
      "epoch": 13.81,
      "learning_rate": 8.045327209371062e-06,
      "loss": 0.3073,
      "step": 38220,
      "task_loss": 0.5040200352668762
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2844274044036865,
      "epoch": 13.82,
      "learning_rate": 7.998958535716842e-06,
      "loss": 0.2743,
      "step": 38230,
      "task_loss": 0.5570114254951477
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27089565992355347,
      "epoch": 13.82,
      "learning_rate": 7.95270330683064e-06,
      "loss": 0.2819,
      "step": 38240,
      "task_loss": 0.4997580647468567
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3141476511955261,
      "epoch": 13.82,
      "learning_rate": 7.906561761219948e-06,
      "loss": 0.2522,
      "step": 38250,
      "task_loss": 1.0347281694412231
    },
    {
      "epoch": 13.82,
      "eval_exact_match": 83.3112582781457,
      "eval_f1": 90.02387764179856,
      "step": 38250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2614123225212097,
      "epoch": 13.83,
      "learning_rate": 7.860534136806123e-06,
      "loss": 0.2562,
      "step": 38260,
      "task_loss": 0.4036688804626465
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23341913521289825,
      "epoch": 13.83,
      "learning_rate": 7.814620670923052e-06,
      "loss": 0.2655,
      "step": 38270,
      "task_loss": 0.3119831681251526
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34925973415374756,
      "epoch": 13.83,
      "learning_rate": 7.768821600316012e-06,
      "loss": 0.2829,
      "step": 38280,
      "task_loss": 0.6085324287414551
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27616429328918457,
      "epoch": 13.84,
      "learning_rate": 7.723137161140407e-06,
      "loss": 0.2932,
      "step": 38290,
      "task_loss": 0.49988824129104614
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25335800647735596,
      "epoch": 13.84,
      "learning_rate": 7.677567588960583e-06,
      "loss": 0.2753,
      "step": 38300,
      "task_loss": 0.4380761384963989
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25001299381256104,
      "epoch": 13.85,
      "learning_rate": 7.632113118748554e-06,
      "loss": 0.2591,
      "step": 38310,
      "task_loss": 0.5579447746276855
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28439947962760925,
      "epoch": 13.85,
      "learning_rate": 7.586773984882869e-06,
      "loss": 0.2777,
      "step": 38320,
      "task_loss": 0.4338080883026123
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3441578149795532,
      "epoch": 13.85,
      "learning_rate": 7.541550421147339e-06,
      "loss": 0.2936,
      "step": 38330,
      "task_loss": 0.6278230547904968
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29420727491378784,
      "epoch": 13.86,
      "learning_rate": 7.496442660729885e-06,
      "loss": 0.2811,
      "step": 38340,
      "task_loss": 0.472691148519516
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28853827714920044,
      "epoch": 13.86,
      "learning_rate": 7.451450936221256e-06,
      "loss": 0.2634,
      "step": 38350,
      "task_loss": 0.36157310009002686
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2415386438369751,
      "epoch": 13.86,
      "learning_rate": 7.4065754796139465e-06,
      "loss": 0.2792,
      "step": 38360,
      "task_loss": 0.5524470806121826
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23241202533245087,
      "epoch": 13.87,
      "learning_rate": 7.3618165223008746e-06,
      "loss": 0.2783,
      "step": 38370,
      "task_loss": 0.27023470401763916
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20150120556354523,
      "epoch": 13.87,
      "learning_rate": 7.31717429507429e-06,
      "loss": 0.2511,
      "step": 38380,
      "task_loss": 0.30635228753089905
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20298174023628235,
      "epoch": 13.87,
      "learning_rate": 7.272649028124523e-06,
      "loss": 0.2488,
      "step": 38390,
      "task_loss": 0.5815449953079224
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21220584213733673,
      "epoch": 13.88,
      "learning_rate": 7.228240951038831e-06,
      "loss": 0.2501,
      "step": 38400,
      "task_loss": 0.5942157506942749
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23308081924915314,
      "epoch": 13.88,
      "learning_rate": 7.183950292800176e-06,
      "loss": 0.2973,
      "step": 38410,
      "task_loss": 0.3363405466079712
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3139108419418335,
      "epoch": 13.89,
      "learning_rate": 7.139777281786092e-06,
      "loss": 0.2944,
      "step": 38420,
      "task_loss": 0.492350697517395
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25158751010894775,
      "epoch": 13.89,
      "learning_rate": 7.095722145767483e-06,
      "loss": 0.2719,
      "step": 38430,
      "task_loss": 0.5319156646728516
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2320326417684555,
      "epoch": 13.89,
      "learning_rate": 7.051785111907449e-06,
      "loss": 0.2772,
      "step": 38440,
      "task_loss": 0.39209455251693726
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21443305909633636,
      "epoch": 13.9,
      "learning_rate": 7.007966406760085e-06,
      "loss": 0.2823,
      "step": 38450,
      "task_loss": 0.20772241055965424
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26291579008102417,
      "epoch": 13.9,
      "learning_rate": 6.964266256269405e-06,
      "loss": 0.2433,
      "step": 38460,
      "task_loss": 0.31175604462623596
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2669048011302948,
      "epoch": 13.9,
      "learning_rate": 6.920684885768056e-06,
      "loss": 0.2731,
      "step": 38470,
      "task_loss": 0.7086321115493774
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18979918956756592,
      "epoch": 13.91,
      "learning_rate": 6.877222519976244e-06,
      "loss": 0.2844,
      "step": 38480,
      "task_loss": 0.2133408784866333
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28547948598861694,
      "epoch": 13.91,
      "learning_rate": 6.833879383000548e-06,
      "loss": 0.2716,
      "step": 38490,
      "task_loss": 0.504301905632019
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1712222397327423,
      "epoch": 13.91,
      "learning_rate": 6.790655698332759e-06,
      "loss": 0.2693,
      "step": 38500,
      "task_loss": 0.39858031272888184
    },
    {
      "epoch": 13.91,
      "eval_exact_match": 83.33964049195838,
      "eval_f1": 90.03027688318437,
      "step": 38500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20968683063983917,
      "epoch": 13.92,
      "learning_rate": 6.747551688848714e-06,
      "loss": 0.2652,
      "step": 38510,
      "task_loss": 0.19403359293937683
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31207019090652466,
      "epoch": 13.92,
      "learning_rate": 6.704567576807191e-06,
      "loss": 0.276,
      "step": 38520,
      "task_loss": 0.6474623680114746
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2539207935333252,
      "epoch": 13.92,
      "learning_rate": 6.661703583848728e-06,
      "loss": 0.239,
      "step": 38530,
      "task_loss": 0.2865660488605499
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22099719941616058,
      "epoch": 13.93,
      "learning_rate": 6.61895993099449e-06,
      "loss": 0.2723,
      "step": 38540,
      "task_loss": 0.3822336792945862
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18672585487365723,
      "epoch": 13.93,
      "learning_rate": 6.576336838645102e-06,
      "loss": 0.2508,
      "step": 38550,
      "task_loss": 0.49510657787323
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22289982438087463,
      "epoch": 13.94,
      "learning_rate": 6.533834526579591e-06,
      "loss": 0.2732,
      "step": 38560,
      "task_loss": 0.4035664200782776
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19927158951759338,
      "epoch": 13.94,
      "learning_rate": 6.4914532139541505e-06,
      "loss": 0.2728,
      "step": 38570,
      "task_loss": 0.21784666180610657
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25364750623703003,
      "epoch": 13.94,
      "learning_rate": 6.449193119301083e-06,
      "loss": 0.2487,
      "step": 38580,
      "task_loss": 0.5803826451301575
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27665311098098755,
      "epoch": 13.95,
      "learning_rate": 6.407054460527651e-06,
      "loss": 0.2664,
      "step": 38590,
      "task_loss": 0.6068881154060364
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27936381101608276,
      "epoch": 13.95,
      "learning_rate": 6.365037454914958e-06,
      "loss": 0.2908,
      "step": 38600,
      "task_loss": 0.3354106843471527
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30930522084236145,
      "epoch": 13.95,
      "learning_rate": 6.323142319116794e-06,
      "loss": 0.2636,
      "step": 38610,
      "task_loss": 0.5985520482063293
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29672688245773315,
      "epoch": 13.96,
      "learning_rate": 6.2813692691585755e-06,
      "loss": 0.2853,
      "step": 38620,
      "task_loss": 0.8483359813690186
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2600293755531311,
      "epoch": 13.96,
      "learning_rate": 6.239718520436195e-06,
      "loss": 0.2688,
      "step": 38630,
      "task_loss": 0.6306997537612915
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26006078720092773,
      "epoch": 13.96,
      "learning_rate": 6.198190287714922e-06,
      "loss": 0.2678,
      "step": 38640,
      "task_loss": 0.49998217821121216
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19626963138580322,
      "epoch": 13.97,
      "learning_rate": 6.156784785128284e-06,
      "loss": 0.2674,
      "step": 38650,
      "task_loss": 0.5400117635726929
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25180870294570923,
      "epoch": 13.97,
      "learning_rate": 6.115502226176989e-06,
      "loss": 0.261,
      "step": 38660,
      "task_loss": 0.32281187176704407
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20361928641796112,
      "epoch": 13.98,
      "learning_rate": 6.074342823727773e-06,
      "loss": 0.266,
      "step": 38670,
      "task_loss": 0.43746501207351685
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29030585289001465,
      "epoch": 13.98,
      "learning_rate": 6.0333067900123664e-06,
      "loss": 0.2883,
      "step": 38680,
      "task_loss": 0.5268990993499756
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23285171389579773,
      "epoch": 13.98,
      "learning_rate": 5.992394336626352e-06,
      "loss": 0.2654,
      "step": 38690,
      "task_loss": 0.3039249777793884
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.34479501843452454,
      "epoch": 13.99,
      "learning_rate": 5.951605674528116e-06,
      "loss": 0.2851,
      "step": 38700,
      "task_loss": 0.7995741963386536
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2292272448539734,
      "epoch": 13.99,
      "learning_rate": 5.910941014037681e-06,
      "loss": 0.2899,
      "step": 38710,
      "task_loss": 0.6517189145088196
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23348368704319,
      "epoch": 13.99,
      "learning_rate": 5.870400564835748e-06,
      "loss": 0.2841,
      "step": 38720,
      "task_loss": 0.2811996340751648
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21506094932556152,
      "epoch": 14.0,
      "learning_rate": 5.8299845359624806e-06,
      "loss": 0.3014,
      "step": 38730,
      "task_loss": 0.37734872102737427
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21314091980457306,
      "epoch": 14.0,
      "learning_rate": 5.789693135816522e-06,
      "loss": 0.2496,
      "step": 38740,
      "task_loss": 0.38001203536987305
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2878772020339966,
      "epoch": 14.0,
      "learning_rate": 5.7495265721538805e-06,
      "loss": 0.2594,
      "step": 38750,
      "task_loss": 0.5972883701324463
    },
    {
      "epoch": 14.0,
      "eval_exact_match": 83.13150425733207,
      "eval_f1": 89.9718447883911,
      "step": 38750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20778994262218475,
      "epoch": 14.01,
      "learning_rate": 5.709485052086874e-06,
      "loss": 0.2567,
      "step": 38760,
      "task_loss": 0.30834150314331055
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20597517490386963,
      "epoch": 14.01,
      "learning_rate": 5.669568782083027e-06,
      "loss": 0.2411,
      "step": 38770,
      "task_loss": 0.4118894338607788
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18355685472488403,
      "epoch": 14.02,
      "learning_rate": 5.629777967964059e-06,
      "loss": 0.2841,
      "step": 38780,
      "task_loss": 0.43749743700027466
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29135215282440186,
      "epoch": 14.02,
      "learning_rate": 5.5901128149047855e-06,
      "loss": 0.2679,
      "step": 38790,
      "task_loss": 0.6453703045845032
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25022754073143005,
      "epoch": 14.02,
      "learning_rate": 5.550573527432088e-06,
      "loss": 0.2557,
      "step": 38800,
      "task_loss": 0.7301109433174133
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3784531354904175,
      "epoch": 14.03,
      "learning_rate": 5.511160309423795e-06,
      "loss": 0.2771,
      "step": 38810,
      "task_loss": 1.0786703824996948
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22921237349510193,
      "epoch": 14.03,
      "learning_rate": 5.471873364107755e-06,
      "loss": 0.2695,
      "step": 38820,
      "task_loss": 0.885239839553833
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27633237838745117,
      "epoch": 14.03,
      "learning_rate": 5.432712894060635e-06,
      "loss": 0.2674,
      "step": 38830,
      "task_loss": 0.9187273979187012
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22725428640842438,
      "epoch": 14.04,
      "learning_rate": 5.3936791012070005e-06,
      "loss": 0.2401,
      "step": 38840,
      "task_loss": 0.5373157858848572
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22272822260856628,
      "epoch": 14.04,
      "learning_rate": 5.354772186818211e-06,
      "loss": 0.2611,
      "step": 38850,
      "task_loss": 0.3665705621242523
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.17846062779426575,
      "epoch": 14.04,
      "learning_rate": 5.315992351511408e-06,
      "loss": 0.2476,
      "step": 38860,
      "task_loss": 0.30240577459335327
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2393602430820465,
      "epoch": 14.05,
      "learning_rate": 5.277339795248451e-06,
      "loss": 0.2667,
      "step": 38870,
      "task_loss": 0.34489595890045166
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20027129352092743,
      "epoch": 14.05,
      "learning_rate": 5.238814717334919e-06,
      "loss": 0.2618,
      "step": 38880,
      "task_loss": 0.476666122674942
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28304845094680786,
      "epoch": 14.05,
      "learning_rate": 5.2004173164190746e-06,
      "loss": 0.2651,
      "step": 38890,
      "task_loss": 0.8417985439300537
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27796077728271484,
      "epoch": 14.06,
      "learning_rate": 5.16214779049083e-06,
      "loss": 0.2548,
      "step": 38900,
      "task_loss": 0.38383617997169495
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24468521773815155,
      "epoch": 14.06,
      "learning_rate": 5.124006336880716e-06,
      "loss": 0.2648,
      "step": 38910,
      "task_loss": 0.6111380457878113
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2580738961696625,
      "epoch": 14.07,
      "learning_rate": 5.085993152258912e-06,
      "loss": 0.2666,
      "step": 38920,
      "task_loss": 0.5941046476364136
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21132975816726685,
      "epoch": 14.07,
      "learning_rate": 5.048108432634165e-06,
      "loss": 0.2466,
      "step": 38930,
      "task_loss": 0.42760226130485535
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26778656244277954,
      "epoch": 14.07,
      "learning_rate": 5.010352373352838e-06,
      "loss": 0.2682,
      "step": 38940,
      "task_loss": 0.5522580146789551
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2655670642852783,
      "epoch": 14.08,
      "learning_rate": 4.972725169097863e-06,
      "loss": 0.2532,
      "step": 38950,
      "task_loss": 0.710306704044342
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21947705745697021,
      "epoch": 14.08,
      "learning_rate": 4.935227013887771e-06,
      "loss": 0.2299,
      "step": 38960,
      "task_loss": 0.4309462904930115
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22299286723136902,
      "epoch": 14.08,
      "learning_rate": 4.89785810107564e-06,
      "loss": 0.28,
      "step": 38970,
      "task_loss": 0.21669545769691467
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22194522619247437,
      "epoch": 14.09,
      "learning_rate": 4.8606186233481585e-06,
      "loss": 0.2666,
      "step": 38980,
      "task_loss": 0.44074350595474243
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2186507284641266,
      "epoch": 14.09,
      "learning_rate": 4.823508772724598e-06,
      "loss": 0.2435,
      "step": 38990,
      "task_loss": 0.41085922718048096
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22034092247486115,
      "epoch": 14.09,
      "learning_rate": 4.786528740555834e-06,
      "loss": 0.2503,
      "step": 39000,
      "task_loss": 0.5862570405006409
    },
    {
      "epoch": 14.09,
      "eval_exact_match": 83.65184484389782,
      "eval_f1": 90.2313400355417,
      "step": 39000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.210486501455307,
      "epoch": 14.1,
      "learning_rate": 4.749678717523327e-06,
      "loss": 0.2644,
      "step": 39010,
      "task_loss": 0.37090587615966797
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2620948255062103,
      "epoch": 14.1,
      "learning_rate": 4.7129588936382125e-06,
      "loss": 0.2383,
      "step": 39020,
      "task_loss": 0.5931086540222168
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22131627798080444,
      "epoch": 14.11,
      "learning_rate": 4.676369458240229e-06,
      "loss": 0.2592,
      "step": 39030,
      "task_loss": 0.47717371582984924
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20887985825538635,
      "epoch": 14.11,
      "learning_rate": 4.639910599996822e-06,
      "loss": 0.2599,
      "step": 39040,
      "task_loss": 0.3629271984100342
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2679504156112671,
      "epoch": 14.11,
      "learning_rate": 4.6035825069021165e-06,
      "loss": 0.2778,
      "step": 39050,
      "task_loss": 0.27707189321517944
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23027104139328003,
      "epoch": 14.12,
      "learning_rate": 4.56738536627599e-06,
      "loss": 0.2934,
      "step": 39060,
      "task_loss": 0.5485252141952515
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19164533913135529,
      "epoch": 14.12,
      "learning_rate": 4.531319364763053e-06,
      "loss": 0.2653,
      "step": 39070,
      "task_loss": 0.44614046812057495
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2256176471710205,
      "epoch": 14.12,
      "learning_rate": 4.495384688331747e-06,
      "loss": 0.2532,
      "step": 39080,
      "task_loss": 0.36074715852737427
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28316569328308105,
      "epoch": 14.13,
      "learning_rate": 4.459581522273347e-06,
      "loss": 0.2566,
      "step": 39090,
      "task_loss": 0.43225523829460144
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1975765973329544,
      "epoch": 14.13,
      "learning_rate": 4.423910051201023e-06,
      "loss": 0.2412,
      "step": 39100,
      "task_loss": 0.46456313133239746
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.31945353746414185,
      "epoch": 14.13,
      "learning_rate": 4.388370459048854e-06,
      "loss": 0.2579,
      "step": 39110,
      "task_loss": 0.428020179271698
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21575048565864563,
      "epoch": 14.14,
      "learning_rate": 4.352962929070954e-06,
      "loss": 0.2495,
      "step": 39120,
      "task_loss": 0.5014318227767944
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23357847332954407,
      "epoch": 14.14,
      "learning_rate": 4.317687643840437e-06,
      "loss": 0.2673,
      "step": 39130,
      "task_loss": 0.34480270743370056
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2436312735080719,
      "epoch": 14.15,
      "learning_rate": 4.2825447852485375e-06,
      "loss": 0.2711,
      "step": 39140,
      "task_loss": 0.42724448442459106
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.227299302816391,
      "epoch": 14.15,
      "learning_rate": 4.247534534503656e-06,
      "loss": 0.2647,
      "step": 39150,
      "task_loss": 0.5893539190292358
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2549132704734802,
      "epoch": 14.15,
      "learning_rate": 4.212657072130425e-06,
      "loss": 0.259,
      "step": 39160,
      "task_loss": 0.4166608452796936
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2273089587688446,
      "epoch": 14.16,
      "learning_rate": 4.177912577968752e-06,
      "loss": 0.2439,
      "step": 39170,
      "task_loss": 0.30703485012054443
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27625080943107605,
      "epoch": 14.16,
      "learning_rate": 4.143301231172939e-06,
      "loss": 0.2615,
      "step": 39180,
      "task_loss": 0.5613874793052673
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2191055864095688,
      "epoch": 14.16,
      "learning_rate": 4.108823210210733e-06,
      "loss": 0.247,
      "step": 39190,
      "task_loss": 0.4466383755207062
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.17954121530056,
      "epoch": 14.17,
      "learning_rate": 4.0744786928624065e-06,
      "loss": 0.2321,
      "step": 39200,
      "task_loss": 0.3059490919113159
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20783843100070953,
      "epoch": 14.17,
      "learning_rate": 4.040267856219822e-06,
      "loss": 0.235,
      "step": 39210,
      "task_loss": 0.41658928990364075
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2153421938419342,
      "epoch": 14.17,
      "learning_rate": 4.006190876685586e-06,
      "loss": 0.2544,
      "step": 39220,
      "task_loss": 0.39804089069366455
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24418562650680542,
      "epoch": 14.18,
      "learning_rate": 3.9722479299720385e-06,
      "loss": 0.2599,
      "step": 39230,
      "task_loss": 0.6925318837165833
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2603607177734375,
      "epoch": 14.18,
      "learning_rate": 3.938439191100437e-06,
      "loss": 0.2733,
      "step": 39240,
      "task_loss": 0.6092582941055298
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23545420169830322,
      "epoch": 14.19,
      "learning_rate": 3.904764834400003e-06,
      "loss": 0.2586,
      "step": 39250,
      "task_loss": 0.40439313650131226
    },
    {
      "epoch": 14.19,
      "eval_exact_match": 83.64238410596026,
      "eval_f1": 90.21788917699949,
      "step": 39250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2226608842611313,
      "epoch": 14.19,
      "learning_rate": 3.8712250335070584e-06,
      "loss": 0.2538,
      "step": 39260,
      "task_loss": 0.7255889177322388
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.15965428948402405,
      "epoch": 14.19,
      "learning_rate": 3.837819961364074e-06,
      "loss": 0.2443,
      "step": 39270,
      "task_loss": 0.4545014500617981
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26116734743118286,
      "epoch": 14.2,
      "learning_rate": 3.8045497902188455e-06,
      "loss": 0.2566,
      "step": 39280,
      "task_loss": 0.4564734697341919
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26344382762908936,
      "epoch": 14.2,
      "learning_rate": 3.771414691623565e-06,
      "loss": 0.251,
      "step": 39290,
      "task_loss": 1.0150420665740967
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19946512579917908,
      "epoch": 14.2,
      "learning_rate": 3.7384148364339423e-06,
      "loss": 0.2671,
      "step": 39300,
      "task_loss": 0.4203642010688782
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2514525353908539,
      "epoch": 14.21,
      "learning_rate": 3.7055503948083112e-06,
      "loss": 0.2652,
      "step": 39310,
      "task_loss": 0.6718653440475464
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27354827523231506,
      "epoch": 14.21,
      "learning_rate": 3.6728215362068086e-06,
      "loss": 0.2628,
      "step": 39320,
      "task_loss": 0.5433547496795654
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2853875756263733,
      "epoch": 14.21,
      "learning_rate": 3.640228429390429e-06,
      "loss": 0.2409,
      "step": 39330,
      "task_loss": 0.5214777588844299
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2823842465877533,
      "epoch": 14.22,
      "learning_rate": 3.6077712424201805e-06,
      "loss": 0.231,
      "step": 39340,
      "task_loss": 0.6327749490737915
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2311672866344452,
      "epoch": 14.22,
      "learning_rate": 3.5754501426562393e-06,
      "loss": 0.2442,
      "step": 39350,
      "task_loss": 0.5876401662826538
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2697458863258362,
      "epoch": 14.22,
      "learning_rate": 3.5432652967570623e-06,
      "loss": 0.2538,
      "step": 39360,
      "task_loss": 0.5169529914855957
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21121466159820557,
      "epoch": 14.23,
      "learning_rate": 3.5112168706785487e-06,
      "loss": 0.2612,
      "step": 39370,
      "task_loss": 0.3359443247318268
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1496531367301941,
      "epoch": 14.23,
      "learning_rate": 3.4793050296731365e-06,
      "loss": 0.2507,
      "step": 39380,
      "task_loss": 0.36530622839927673
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2131895124912262,
      "epoch": 14.24,
      "learning_rate": 3.447529938289038e-06,
      "loss": 0.2535,
      "step": 39390,
      "task_loss": 0.32073259353637695
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24818633496761322,
      "epoch": 14.24,
      "learning_rate": 3.4158917603692885e-06,
      "loss": 0.2409,
      "step": 39400,
      "task_loss": 0.3351534307003021
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23910020291805267,
      "epoch": 14.24,
      "learning_rate": 3.384390659050979e-06,
      "loss": 0.2667,
      "step": 39410,
      "task_loss": 0.5424809455871582
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2993757724761963,
      "epoch": 14.25,
      "learning_rate": 3.353026796764378e-06,
      "loss": 0.2607,
      "step": 39420,
      "task_loss": 0.5096408724784851
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2651025056838989,
      "epoch": 14.25,
      "learning_rate": 3.321800335232118e-06,
      "loss": 0.2468,
      "step": 39430,
      "task_loss": 0.3386688232421875
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24977470934391022,
      "epoch": 14.25,
      "learning_rate": 3.2907114354683175e-06,
      "loss": 0.2505,
      "step": 39440,
      "task_loss": 0.3789813816547394
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23737601935863495,
      "epoch": 14.26,
      "learning_rate": 3.2597602577778065e-06,
      "loss": 0.239,
      "step": 39450,
      "task_loss": 0.4992079734802246
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20579993724822998,
      "epoch": 14.26,
      "learning_rate": 3.2289469617552613e-06,
      "loss": 0.2446,
      "step": 39460,
      "task_loss": 0.5167773962020874
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18076182901859283,
      "epoch": 14.26,
      "learning_rate": 3.198271706284409e-06,
      "loss": 0.2441,
      "step": 39470,
      "task_loss": 0.6046625375747681
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19197119772434235,
      "epoch": 14.27,
      "learning_rate": 3.1677346495371616e-06,
      "loss": 0.248,
      "step": 39480,
      "task_loss": 0.26388877630233765
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22745785117149353,
      "epoch": 14.27,
      "learning_rate": 3.1373359489728783e-06,
      "loss": 0.2513,
      "step": 39490,
      "task_loss": 0.3964017629623413
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.29309511184692383,
      "epoch": 14.28,
      "learning_rate": 3.107075761337458e-06,
      "loss": 0.2644,
      "step": 39500,
      "task_loss": 0.5477885007858276
    },
    {
      "epoch": 14.28,
      "eval_exact_match": 83.14096499526963,
      "eval_f1": 90.00285618886593,
      "step": 39500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19775673747062683,
      "epoch": 14.28,
      "learning_rate": 3.076954242662615e-06,
      "loss": 0.2366,
      "step": 39510,
      "task_loss": 0.5608834028244019
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3160851001739502,
      "epoch": 14.28,
      "learning_rate": 3.0469715482650264e-06,
      "loss": 0.2756,
      "step": 39520,
      "task_loss": 0.6379892826080322
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2946835458278656,
      "epoch": 14.29,
      "learning_rate": 3.01712783274555e-06,
      "loss": 0.2552,
      "step": 39530,
      "task_loss": 0.5279828906059265
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24564538896083832,
      "epoch": 14.29,
      "learning_rate": 2.987423249988411e-06,
      "loss": 0.2377,
      "step": 39540,
      "task_loss": 0.45563802123069763
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.315203458070755,
      "epoch": 14.29,
      "learning_rate": 2.9578579531604335e-06,
      "loss": 0.2574,
      "step": 39550,
      "task_loss": 0.4949073791503906
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2737749218940735,
      "epoch": 14.3,
      "learning_rate": 2.9284320947102227e-06,
      "loss": 0.2593,
      "step": 39560,
      "task_loss": 0.41951417922973633
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21220695972442627,
      "epoch": 14.3,
      "learning_rate": 2.899145826367412e-06,
      "loss": 0.2522,
      "step": 39570,
      "task_loss": 0.40739068388938904
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18755120038986206,
      "epoch": 14.3,
      "learning_rate": 2.869999299141829e-06,
      "loss": 0.265,
      "step": 39580,
      "task_loss": 0.7724450826644897
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21923671662807465,
      "epoch": 14.31,
      "learning_rate": 2.8409926633227947e-06,
      "loss": 0.2635,
      "step": 39590,
      "task_loss": 0.39632830023765564
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22970925271511078,
      "epoch": 14.31,
      "learning_rate": 2.8121260684782567e-06,
      "loss": 0.246,
      "step": 39600,
      "task_loss": 0.34349530935287476
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2519911229610443,
      "epoch": 14.32,
      "learning_rate": 2.7833996634540914e-06,
      "loss": 0.2908,
      "step": 39610,
      "task_loss": 0.5939441919326782
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.16203315556049347,
      "epoch": 14.32,
      "learning_rate": 2.7548135963733057e-06,
      "loss": 0.2331,
      "step": 39620,
      "task_loss": 0.20400336384773254
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23254115879535675,
      "epoch": 14.32,
      "learning_rate": 2.726368014635275e-06,
      "loss": 0.2454,
      "step": 39630,
      "task_loss": 0.6872649192810059
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22584383189678192,
      "epoch": 14.33,
      "learning_rate": 2.6980630649149797e-06,
      "loss": 0.2486,
      "step": 39640,
      "task_loss": 0.4566018581390381
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.175010085105896,
      "epoch": 14.33,
      "learning_rate": 2.669898893162257e-06,
      "loss": 0.2542,
      "step": 39650,
      "task_loss": 0.36080285906791687
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2514447569847107,
      "epoch": 14.33,
      "learning_rate": 2.641875644601047e-06,
      "loss": 0.2486,
      "step": 39660,
      "task_loss": 0.7753199338912964
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2958347201347351,
      "epoch": 14.34,
      "learning_rate": 2.6139934637286546e-06,
      "loss": 0.2493,
      "step": 39670,
      "task_loss": 0.58829665184021
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3366328477859497,
      "epoch": 14.34,
      "learning_rate": 2.586252494314961e-06,
      "loss": 0.2687,
      "step": 39680,
      "task_loss": 0.6414390802383423
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2517812252044678,
      "epoch": 14.34,
      "learning_rate": 2.558652879401753e-06,
      "loss": 0.2972,
      "step": 39690,
      "task_loss": 0.31333988904953003
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23453620076179504,
      "epoch": 14.35,
      "learning_rate": 2.531194761301907e-06,
      "loss": 0.2559,
      "step": 39700,
      "task_loss": 0.5426726937294006
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2716495990753174,
      "epoch": 14.35,
      "learning_rate": 2.503878281598726e-06,
      "loss": 0.2516,
      "step": 39710,
      "task_loss": 0.7860714197158813
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18287578225135803,
      "epoch": 14.35,
      "learning_rate": 2.476703581145162e-06,
      "loss": 0.2595,
      "step": 39720,
      "task_loss": 0.48666173219680786
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24697816371917725,
      "epoch": 14.36,
      "learning_rate": 2.4496708000631094e-06,
      "loss": 0.2729,
      "step": 39730,
      "task_loss": 0.5118921995162964
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3017914891242981,
      "epoch": 14.36,
      "learning_rate": 2.4227800777426746e-06,
      "loss": 0.262,
      "step": 39740,
      "task_loss": 0.6162266135215759
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27042317390441895,
      "epoch": 14.37,
      "learning_rate": 2.396031552841462e-06,
      "loss": 0.2544,
      "step": 39750,
      "task_loss": 1.038049340248108
    },
    {
      "epoch": 14.37,
      "eval_exact_match": 83.23557237464522,
      "eval_f1": 89.99843780981955,
      "step": 39750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21385975182056427,
      "epoch": 14.37,
      "learning_rate": 2.369425363283865e-06,
      "loss": 0.2569,
      "step": 39760,
      "task_loss": 0.37484338879585266
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2144310027360916,
      "epoch": 14.37,
      "learning_rate": 2.3429616462603477e-06,
      "loss": 0.253,
      "step": 39770,
      "task_loss": 0.592105507850647
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18579715490341187,
      "epoch": 14.38,
      "learning_rate": 2.316640538226721e-06,
      "loss": 0.2585,
      "step": 39780,
      "task_loss": 0.5446082949638367
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20289719104766846,
      "epoch": 14.38,
      "learning_rate": 2.290462174903486e-06,
      "loss": 0.2512,
      "step": 39790,
      "task_loss": 0.3528170883655548
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23953914642333984,
      "epoch": 14.38,
      "learning_rate": 2.2644266912750733e-06,
      "loss": 0.2547,
      "step": 39800,
      "task_loss": 0.43491581082344055
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22709573805332184,
      "epoch": 14.39,
      "learning_rate": 2.238534221589196e-06,
      "loss": 0.2312,
      "step": 39810,
      "task_loss": 0.40255218744277954
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27763283252716064,
      "epoch": 14.39,
      "learning_rate": 2.212784899356136e-06,
      "loss": 0.2786,
      "step": 39820,
      "task_loss": 0.830498218536377
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2324332594871521,
      "epoch": 14.39,
      "learning_rate": 2.187178857348061e-06,
      "loss": 0.2695,
      "step": 39830,
      "task_loss": 0.6241767406463623
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2514335513114929,
      "epoch": 14.4,
      "learning_rate": 2.1617162275983217e-06,
      "loss": 0.2693,
      "step": 39840,
      "task_loss": 0.6369235515594482
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3284478187561035,
      "epoch": 14.4,
      "learning_rate": 2.1363971414008097e-06,
      "loss": 0.2641,
      "step": 39850,
      "task_loss": 0.7467869520187378
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28706008195877075,
      "epoch": 14.41,
      "learning_rate": 2.1112217293092405e-06,
      "loss": 0.2763,
      "step": 39860,
      "task_loss": 0.45901063084602356
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22485694289207458,
      "epoch": 14.41,
      "learning_rate": 2.0861901211365177e-06,
      "loss": 0.2772,
      "step": 39870,
      "task_loss": 0.40547189116477966
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3032034635543823,
      "epoch": 14.41,
      "learning_rate": 2.0613024459540076e-06,
      "loss": 0.2529,
      "step": 39880,
      "task_loss": 0.6667108535766602
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3540099263191223,
      "epoch": 14.42,
      "learning_rate": 2.0365588320909576e-06,
      "loss": 0.2639,
      "step": 39890,
      "task_loss": 0.9680027961730957
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25109708309173584,
      "epoch": 14.42,
      "learning_rate": 2.0119594071337433e-06,
      "loss": 0.251,
      "step": 39900,
      "task_loss": 0.5497735738754272
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21858660876750946,
      "epoch": 14.42,
      "learning_rate": 1.98750429792528e-06,
      "loss": 0.2511,
      "step": 39910,
      "task_loss": 0.22567731142044067
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2529667913913727,
      "epoch": 14.43,
      "learning_rate": 1.9631936305643294e-06,
      "loss": 0.2542,
      "step": 39920,
      "task_loss": 0.6540744304656982
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2905096411705017,
      "epoch": 14.43,
      "learning_rate": 1.9390275304048755e-06,
      "loss": 0.2536,
      "step": 39930,
      "task_loss": 0.6890720129013062
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21479448676109314,
      "epoch": 14.43,
      "learning_rate": 1.915006122055445e-06,
      "loss": 0.2467,
      "step": 39940,
      "task_loss": 0.41722339391708374
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22225235402584076,
      "epoch": 14.44,
      "learning_rate": 1.891129529378508e-06,
      "loss": 0.2533,
      "step": 39950,
      "task_loss": 0.3307686448097229
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24084031581878662,
      "epoch": 14.44,
      "learning_rate": 1.867397875489799e-06,
      "loss": 0.2514,
      "step": 39960,
      "task_loss": 0.584737241268158
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23029622435569763,
      "epoch": 14.45,
      "learning_rate": 1.8438112827577068e-06,
      "loss": 0.2666,
      "step": 39970,
      "task_loss": 0.8088452816009521
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2523832619190216,
      "epoch": 14.45,
      "learning_rate": 1.8203698728026386e-06,
      "loss": 0.2525,
      "step": 39980,
      "task_loss": 0.4110991060733795
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.250593900680542,
      "epoch": 14.45,
      "learning_rate": 1.7970737664963832e-06,
      "loss": 0.2674,
      "step": 39990,
      "task_loss": 0.4489569067955017
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24846208095550537,
      "epoch": 14.46,
      "learning_rate": 1.7739230839614962e-06,
      "loss": 0.273,
      "step": 40000,
      "task_loss": 0.5540671944618225
    },
    {
      "epoch": 14.46,
      "eval_exact_match": 83.58561967833491,
      "eval_f1": 90.22663507791502,
      "step": 40000
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24182166159152985,
      "epoch": 14.46,
      "learning_rate": 1.7509179445706858e-06,
      "loss": 0.2589,
      "step": 40010,
      "task_loss": 0.34240439534187317
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19622673094272614,
      "epoch": 14.46,
      "learning_rate": 1.7280584669461808e-06,
      "loss": 0.2581,
      "step": 40020,
      "task_loss": 0.298112690448761
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23341301083564758,
      "epoch": 14.47,
      "learning_rate": 1.7053447689591473e-06,
      "loss": 0.2296,
      "step": 40030,
      "task_loss": 0.5251190662384033
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18002331256866455,
      "epoch": 14.47,
      "learning_rate": 1.6827769677290294e-06,
      "loss": 0.2664,
      "step": 40040,
      "task_loss": 0.36037880182266235
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2654402256011963,
      "epoch": 14.47,
      "learning_rate": 1.6603551796230232e-06,
      "loss": 0.254,
      "step": 40050,
      "task_loss": 0.7449613809585571
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2447563111782074,
      "epoch": 14.48,
      "learning_rate": 1.6380795202553866e-06,
      "loss": 0.2711,
      "step": 40060,
      "task_loss": 0.5265692472457886
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27043670415878296,
      "epoch": 14.48,
      "learning_rate": 1.615950104486924e-06,
      "loss": 0.2571,
      "step": 40070,
      "task_loss": 0.5612872838973999
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27174508571624756,
      "epoch": 14.49,
      "learning_rate": 1.5939670464243362e-06,
      "loss": 0.2778,
      "step": 40080,
      "task_loss": 0.26313579082489014
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23366330564022064,
      "epoch": 14.49,
      "learning_rate": 1.572130459419674e-06,
      "loss": 0.2423,
      "step": 40090,
      "task_loss": 0.49717170000076294
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22687816619873047,
      "epoch": 14.49,
      "learning_rate": 1.5504404560697093e-06,
      "loss": 0.2412,
      "step": 40100,
      "task_loss": 0.37348806858062744
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24465885758399963,
      "epoch": 14.5,
      "learning_rate": 1.5288971482153957e-06,
      "loss": 0.2711,
      "step": 40110,
      "task_loss": 0.3896139860153198
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2719659209251404,
      "epoch": 14.5,
      "learning_rate": 1.5075006469412778e-06,
      "loss": 0.2532,
      "step": 40120,
      "task_loss": 0.400590181350708
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23399712145328522,
      "epoch": 14.5,
      "learning_rate": 1.486251062574916e-06,
      "loss": 0.267,
      "step": 40130,
      "task_loss": 0.3120698034763336
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24053291976451874,
      "epoch": 14.51,
      "learning_rate": 1.4651485046862933e-06,
      "loss": 0.2828,
      "step": 40140,
      "task_loss": 0.2387630045413971
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21442466974258423,
      "epoch": 14.51,
      "learning_rate": 1.4441930820873195e-06,
      "loss": 0.2591,
      "step": 40150,
      "task_loss": 0.3812238574028015
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3263953924179077,
      "epoch": 14.51,
      "learning_rate": 1.4233849028311808e-06,
      "loss": 0.2577,
      "step": 40160,
      "task_loss": 0.6377623677253723
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3390432596206665,
      "epoch": 14.52,
      "learning_rate": 1.4027240742118542e-06,
      "loss": 0.259,
      "step": 40170,
      "task_loss": 0.5427772402763367
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2371370792388916,
      "epoch": 14.52,
      "learning_rate": 1.3822107027635178e-06,
      "loss": 0.2576,
      "step": 40180,
      "task_loss": 0.49518412351608276
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21536654233932495,
      "epoch": 14.52,
      "learning_rate": 1.3618448942600182e-06,
      "loss": 0.2747,
      "step": 40190,
      "task_loss": 0.5360449552536011
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2513597011566162,
      "epoch": 14.53,
      "learning_rate": 1.3416267537143035e-06,
      "loss": 0.2498,
      "step": 40200,
      "task_loss": 0.5347855091094971
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23619955778121948,
      "epoch": 14.53,
      "learning_rate": 1.3215563853779112e-06,
      "loss": 0.2524,
      "step": 40210,
      "task_loss": 0.4691365361213684
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24675723910331726,
      "epoch": 14.54,
      "learning_rate": 1.3016338927404047e-06,
      "loss": 0.2325,
      "step": 40220,
      "task_loss": 0.5890612602233887
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2255321443080902,
      "epoch": 14.54,
      "learning_rate": 1.2818593785288645e-06,
      "loss": 0.2666,
      "step": 40230,
      "task_loss": 0.4460175037384033
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2343456894159317,
      "epoch": 14.54,
      "learning_rate": 1.262232944707321e-06,
      "loss": 0.2699,
      "step": 40240,
      "task_loss": 0.2739362120628357
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24591998755931854,
      "epoch": 14.55,
      "learning_rate": 1.2427546924762823e-06,
      "loss": 0.2648,
      "step": 40250,
      "task_loss": 0.5649697780609131
    },
    {
      "epoch": 14.55,
      "eval_exact_match": 83.50993377483444,
      "eval_f1": 90.18612000703668,
      "step": 40250
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28279536962509155,
      "epoch": 14.55,
      "learning_rate": 1.2234247222721573e-06,
      "loss": 0.2494,
      "step": 40260,
      "task_loss": 0.7293515205383301
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.275890976190567,
      "epoch": 14.55,
      "learning_rate": 1.2042431337667704e-06,
      "loss": 0.2645,
      "step": 40270,
      "task_loss": 0.5459344983100891
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3029371201992035,
      "epoch": 14.56,
      "learning_rate": 1.1852100258668507e-06,
      "loss": 0.2717,
      "step": 40280,
      "task_loss": 0.5998193025588989
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2221463918685913,
      "epoch": 14.56,
      "learning_rate": 1.1663254967134973e-06,
      "loss": 0.2615,
      "step": 40290,
      "task_loss": 0.6425062417984009
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2234063744544983,
      "epoch": 14.56,
      "learning_rate": 1.1475896436816947e-06,
      "loss": 0.2291,
      "step": 40300,
      "task_loss": 0.43196287751197815
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2834321856498718,
      "epoch": 14.57,
      "learning_rate": 1.1290025633797973e-06,
      "loss": 0.2739,
      "step": 40310,
      "task_loss": 0.6205756664276123
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19937190413475037,
      "epoch": 14.57,
      "learning_rate": 1.1105643516490438e-06,
      "loss": 0.2409,
      "step": 40320,
      "task_loss": 0.2715055048465729
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2276768982410431,
      "epoch": 14.58,
      "learning_rate": 1.0922751035630595e-06,
      "loss": 0.2689,
      "step": 40330,
      "task_loss": 0.4545585513114929
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2939984202384949,
      "epoch": 14.58,
      "learning_rate": 1.0741349134273448e-06,
      "loss": 0.2416,
      "step": 40340,
      "task_loss": 0.833946704864502
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18050755560398102,
      "epoch": 14.58,
      "learning_rate": 1.0561438747788377e-06,
      "loss": 0.2358,
      "step": 40350,
      "task_loss": 0.5802313685417175
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.17724663019180298,
      "epoch": 14.59,
      "learning_rate": 1.0383020803853682e-06,
      "loss": 0.2495,
      "step": 40360,
      "task_loss": 0.6058967709541321
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20336687564849854,
      "epoch": 14.59,
      "learning_rate": 1.0206096222452321e-06,
      "loss": 0.2634,
      "step": 40370,
      "task_loss": 0.6848382949829102
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1893875151872635,
      "epoch": 14.59,
      "learning_rate": 1.0030665915866944e-06,
      "loss": 0.2509,
      "step": 40380,
      "task_loss": 0.7192158699035645
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2207334339618683,
      "epoch": 14.6,
      "learning_rate": 9.856730788675228e-07,
      "loss": 0.2711,
      "step": 40390,
      "task_loss": 0.6762465238571167
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2893259525299072,
      "epoch": 14.6,
      "learning_rate": 9.68429173774512e-07,
      "loss": 0.2469,
      "step": 40400,
      "task_loss": 0.8992417454719543
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25874942541122437,
      "epoch": 14.6,
      "learning_rate": 9.513349652230407e-07,
      "loss": 0.278,
      "step": 40410,
      "task_loss": 0.3645339608192444
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.21643847227096558,
      "epoch": 14.61,
      "learning_rate": 9.343905413565878e-07,
      "loss": 0.2496,
      "step": 40420,
      "task_loss": 0.4934597611427307
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20562595129013062,
      "epoch": 14.61,
      "learning_rate": 9.175959895463138e-07,
      "loss": 0.263,
      "step": 40430,
      "task_loss": 0.4362022578716278
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2967050075531006,
      "epoch": 14.62,
      "learning_rate": 9.009513963905602e-07,
      "loss": 0.2735,
      "step": 40440,
      "task_loss": 1.1383086442947388
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1844678521156311,
      "epoch": 14.62,
      "learning_rate": 8.844568477144644e-07,
      "loss": 0.248,
      "step": 40450,
      "task_loss": 0.38792985677719116
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25160741806030273,
      "epoch": 14.62,
      "learning_rate": 8.681124285694486e-07,
      "loss": 0.2813,
      "step": 40460,
      "task_loss": 0.45143911242485046
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.296841561794281,
      "epoch": 14.63,
      "learning_rate": 8.519182232328415e-07,
      "loss": 0.2561,
      "step": 40470,
      "task_loss": 0.6558849811553955
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2747701108455658,
      "epoch": 14.63,
      "learning_rate": 8.358743152074111e-07,
      "loss": 0.2519,
      "step": 40480,
      "task_loss": 0.3261622190475464
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1828879565000534,
      "epoch": 14.63,
      "learning_rate": 8.199807872209452e-07,
      "loss": 0.2396,
      "step": 40490,
      "task_loss": 0.2521137595176697
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23122264444828033,
      "epoch": 14.64,
      "learning_rate": 8.042377212258123e-07,
      "loss": 0.284,
      "step": 40500,
      "task_loss": 0.34270986914634705
    },
    {
      "epoch": 14.64,
      "eval_exact_match": 83.66130558183538,
      "eval_f1": 90.23675007307362,
      "step": 40500
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.32861417531967163,
      "epoch": 14.64,
      "learning_rate": 7.886451983985576e-07,
      "loss": 0.2855,
      "step": 40510,
      "task_loss": 0.5027445554733276
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.28874510526657104,
      "epoch": 14.64,
      "learning_rate": 7.73203299139471e-07,
      "loss": 0.3079,
      "step": 40520,
      "task_loss": 0.6847516894340515
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24555762112140656,
      "epoch": 14.65,
      "learning_rate": 7.579121030721837e-07,
      "loss": 0.254,
      "step": 40530,
      "task_loss": 0.565000057220459
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24243924021720886,
      "epoch": 14.65,
      "learning_rate": 7.427716890432346e-07,
      "loss": 0.2675,
      "step": 40540,
      "task_loss": 0.38796621561050415
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2042592316865921,
      "epoch": 14.65,
      "learning_rate": 7.277821351216984e-07,
      "loss": 0.2418,
      "step": 40550,
      "task_loss": 0.22988463938236237
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22411584854125977,
      "epoch": 14.66,
      "learning_rate": 7.129435185987487e-07,
      "loss": 0.2477,
      "step": 40560,
      "task_loss": 0.4908447861671448
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1977735459804535,
      "epoch": 14.66,
      "learning_rate": 6.982559159872881e-07,
      "loss": 0.255,
      "step": 40570,
      "task_loss": 0.4119817018508911
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1954713761806488,
      "epoch": 14.67,
      "learning_rate": 6.837194030215288e-07,
      "loss": 0.2546,
      "step": 40580,
      "task_loss": 0.5656400918960571
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26718318462371826,
      "epoch": 14.67,
      "learning_rate": 6.693340546566263e-07,
      "loss": 0.2438,
      "step": 40590,
      "task_loss": 0.43914592266082764
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23150411248207092,
      "epoch": 14.67,
      "learning_rate": 6.550999450682693e-07,
      "loss": 0.2456,
      "step": 40600,
      "task_loss": 0.3785105347633362
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.30644112825393677,
      "epoch": 14.68,
      "learning_rate": 6.410171476523141e-07,
      "loss": 0.2634,
      "step": 40610,
      "task_loss": 0.5664448738098145
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.265768438577652,
      "epoch": 14.68,
      "learning_rate": 6.270857350243974e-07,
      "loss": 0.2696,
      "step": 40620,
      "task_loss": 0.4568207561969757
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.25326547026634216,
      "epoch": 14.68,
      "learning_rate": 6.133057790195773e-07,
      "loss": 0.2648,
      "step": 40630,
      "task_loss": 0.4748091697692871
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20044632256031036,
      "epoch": 14.69,
      "learning_rate": 5.996773506919262e-07,
      "loss": 0.2603,
      "step": 40640,
      "task_loss": 0.3128846287727356
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.204111248254776,
      "epoch": 14.69,
      "learning_rate": 5.862005203142151e-07,
      "loss": 0.2759,
      "step": 40650,
      "task_loss": 0.6005997657775879
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.253496915102005,
      "epoch": 14.69,
      "learning_rate": 5.728753573775069e-07,
      "loss": 0.2404,
      "step": 40660,
      "task_loss": 0.39388108253479004
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2911907732486725,
      "epoch": 14.7,
      "learning_rate": 5.597019305908235e-07,
      "loss": 0.2716,
      "step": 40670,
      "task_loss": 0.5800839066505432
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2877765893936157,
      "epoch": 14.7,
      "learning_rate": 5.466803078807859e-07,
      "loss": 0.2683,
      "step": 40680,
      "task_loss": 0.4347482919692993
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.26411518454551697,
      "epoch": 14.71,
      "learning_rate": 5.33810556391261e-07,
      "loss": 0.2645,
      "step": 40690,
      "task_loss": 0.559246301651001
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.27388328313827515,
      "epoch": 14.71,
      "learning_rate": 5.210927424830092e-07,
      "loss": 0.2675,
      "step": 40700,
      "task_loss": 1.1375054121017456
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2683138847351074,
      "epoch": 14.71,
      "learning_rate": 5.085269317333574e-07,
      "loss": 0.2511,
      "step": 40710,
      "task_loss": 0.8775959014892578
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2723231613636017,
      "epoch": 14.72,
      "learning_rate": 4.961131889358528e-07,
      "loss": 0.2553,
      "step": 40720,
      "task_loss": 0.2622547149658203
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.24217526614665985,
      "epoch": 14.72,
      "learning_rate": 4.838515780999264e-07,
      "loss": 0.2746,
      "step": 40730,
      "task_loss": 0.6096961498260498
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2224244624376297,
      "epoch": 14.72,
      "learning_rate": 4.717421624505669e-07,
      "loss": 0.2642,
      "step": 40740,
      "task_loss": 0.5393527746200562
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2478673756122589,
      "epoch": 14.73,
      "learning_rate": 4.597850044279972e-07,
      "loss": 0.2461,
      "step": 40750,
      "task_loss": 0.46678709983825684
    },
    {
      "epoch": 14.73,
      "eval_exact_match": 83.37748344370861,
      "eval_f1": 90.06774265907103,
      "step": 40750
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20128917694091797,
      "epoch": 14.73,
      "learning_rate": 4.4798016568733837e-07,
      "loss": 0.2472,
      "step": 40760,
      "task_loss": 0.41552314162254333
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2630329430103302,
      "epoch": 14.73,
      "learning_rate": 4.3632770709831293e-07,
      "loss": 0.2575,
      "step": 40770,
      "task_loss": 0.25791311264038086
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1728416383266449,
      "epoch": 14.74,
      "learning_rate": 4.248276887449154e-07,
      "loss": 0.2627,
      "step": 40780,
      "task_loss": 0.49050474166870117
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2832787036895752,
      "epoch": 14.74,
      "learning_rate": 4.1348016992510895e-07,
      "loss": 0.2755,
      "step": 40790,
      "task_loss": 0.6007558107376099
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2201354056596756,
      "epoch": 14.75,
      "learning_rate": 4.0228520915050915e-07,
      "loss": 0.2605,
      "step": 40800,
      "task_loss": 0.5979535579681396
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2848866283893585,
      "epoch": 14.75,
      "learning_rate": 3.912428641461041e-07,
      "loss": 0.2484,
      "step": 40810,
      "task_loss": 0.6617125272750854
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20860302448272705,
      "epoch": 14.75,
      "learning_rate": 3.8035319184993813e-07,
      "loss": 0.2533,
      "step": 40820,
      "task_loss": 0.6819936037063599
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2802797555923462,
      "epoch": 14.76,
      "learning_rate": 3.6961624841282516e-07,
      "loss": 0.2516,
      "step": 40830,
      "task_loss": 0.6139797568321228
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.1864747554063797,
      "epoch": 14.76,
      "learning_rate": 3.590320891980492e-07,
      "loss": 0.2661,
      "step": 40840,
      "task_loss": 0.2509109675884247
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2201501429080963,
      "epoch": 14.76,
      "learning_rate": 3.4860076878110103e-07,
      "loss": 0.2353,
      "step": 40850,
      "task_loss": 0.45859768986701965
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2391672134399414,
      "epoch": 14.77,
      "learning_rate": 3.383223409493719e-07,
      "loss": 0.2631,
      "step": 40860,
      "task_loss": 0.41706597805023193
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.183127760887146,
      "epoch": 14.77,
      "learning_rate": 3.281968587018902e-07,
      "loss": 0.2552,
      "step": 40870,
      "task_loss": 0.5521631836891174
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22473424673080444,
      "epoch": 14.77,
      "learning_rate": 3.1822437424905536e-07,
      "loss": 0.2545,
      "step": 40880,
      "task_loss": 0.358811616897583
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.20832541584968567,
      "epoch": 14.78,
      "learning_rate": 3.084049390123478e-07,
      "loss": 0.2506,
      "step": 40890,
      "task_loss": 0.40031999349594116
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18158313632011414,
      "epoch": 14.78,
      "learning_rate": 2.9873860362407244e-07,
      "loss": 0.2465,
      "step": 40900,
      "task_loss": 0.39136573672294617
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2597076892852783,
      "epoch": 14.78,
      "learning_rate": 2.892254179271059e-07,
      "loss": 0.2613,
      "step": 40910,
      "task_loss": 0.5301886796951294
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2583584189414978,
      "epoch": 14.79,
      "learning_rate": 2.798654309746396e-07,
      "loss": 0.2467,
      "step": 40920,
      "task_loss": 0.5840583443641663
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.18077127635478973,
      "epoch": 14.79,
      "learning_rate": 2.706586910299069e-07,
      "loss": 0.2714,
      "step": 40930,
      "task_loss": 0.4754323959350586
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2263426035642624,
      "epoch": 14.8,
      "learning_rate": 2.616052455659568e-07,
      "loss": 0.2638,
      "step": 40940,
      "task_loss": 0.34484654664993286
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.3615092635154724,
      "epoch": 14.8,
      "learning_rate": 2.5270514126540025e-07,
      "loss": 0.2714,
      "step": 40950,
      "task_loss": 0.6930694580078125
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.22011780738830566,
      "epoch": 14.8,
      "learning_rate": 2.4395842402016756e-07,
      "loss": 0.253,
      "step": 40960,
      "task_loss": 0.5014408230781555
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2508123815059662,
      "epoch": 14.81,
      "learning_rate": 2.3536513893127166e-07,
      "loss": 0.2388,
      "step": 40970,
      "task_loss": 0.49518001079559326
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.2500151991844177,
      "epoch": 14.81,
      "learning_rate": 2.2692533030857832e-07,
      "loss": 0.2742,
      "step": 40980,
      "task_loss": 0.4772796630859375
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.19878563284873962,
      "epoch": 14.81,
      "learning_rate": 2.1863904167058634e-07,
      "loss": 0.2475,
      "step": 40990,
      "task_loss": 0.3303414583206177
    },
    {
      "compression_loss": 0.0,
      "distillation_loss": 0.23328831791877747,
      "epoch": 14.82,
      "learning_rate": 2.1050631574418112e-07,
      "loss": 0.2655,
      "step": 41000,
      "task_loss": 0.6365581154823303
    },
    {
      "epoch": 14.82,
      "eval_exact_match": 83.66130558183538,
      "eval_f1": 90.30829859300735,
      "step": 41000
    }
  ],
  "max_steps": 41505,
  "num_train_epochs": 15,
  "total_flos": 3.128973983327232e+16,
  "trial_name": null,
  "trial_params": null
}