{
  "best_metric": 0.9972341936836928,
  "best_model_checkpoint": "output_dev/v50-deberta-v2-xxlarge-f1-warmup-greater-patience-subword-len-500-maxlen-512-6epoch/checkpoint-5400",
  "epoch": 4.605585497305243,
  "eval_steps": 200,
  "global_step": 9400,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 8.156606851549755e-07,
      "loss": 1.051,
      "step": 10
    },
    {
      "epoch": 0.01,
      "learning_rate": 1.631321370309951e-06,
      "loss": 1.0695,
      "step": 20
    },
    {
      "epoch": 0.01,
      "learning_rate": 2.446982055464927e-06,
      "loss": 1.0459,
      "step": 30
    },
    {
      "epoch": 0.02,
      "learning_rate": 3.262642740619902e-06,
      "loss": 0.9917,
      "step": 40
    },
    {
      "epoch": 0.02,
      "learning_rate": 4.078303425774878e-06,
      "loss": 0.9952,
      "step": 50
    },
    {
      "epoch": 0.03,
      "learning_rate": 4.893964110929854e-06,
      "loss": 0.9274,
      "step": 60
    },
    {
      "epoch": 0.03,
      "learning_rate": 5.709624796084829e-06,
      "loss": 0.8327,
      "step": 70
    },
    {
      "epoch": 0.04,
      "learning_rate": 6.525285481239804e-06,
      "loss": 0.849,
      "step": 80
    },
    {
      "epoch": 0.04,
      "learning_rate": 7.3409461663947795e-06,
      "loss": 0.7667,
      "step": 90
    },
    {
      "epoch": 0.05,
      "learning_rate": 8.156606851549756e-06,
      "loss": 0.7042,
      "step": 100
    },
    {
      "epoch": 0.05,
      "learning_rate": 8.972267536704731e-06,
      "loss": 0.6697,
      "step": 110
    },
    {
      "epoch": 0.06,
      "learning_rate": 9.787928221859708e-06,
      "loss": 0.4986,
      "step": 120
    },
    {
      "epoch": 0.06,
      "learning_rate": 1.0603588907014683e-05,
      "loss": 0.4431,
      "step": 130
    },
    {
      "epoch": 0.07,
      "learning_rate": 1.1419249592169658e-05,
      "loss": 0.3337,
      "step": 140
    },
    {
      "epoch": 0.07,
      "learning_rate": 1.2234910277324634e-05,
      "loss": 0.3402,
      "step": 150
    },
    {
      "epoch": 0.08,
      "learning_rate": 1.3050570962479607e-05,
      "loss": 0.2445,
      "step": 160
    },
    {
      "epoch": 0.08,
      "learning_rate": 1.3866231647634584e-05,
      "loss": 0.2604,
      "step": 170
    },
    {
      "epoch": 0.09,
      "learning_rate": 1.4681892332789559e-05,
      "loss": 0.2642,
      "step": 180
    },
    {
      "epoch": 0.09,
      "learning_rate": 1.5497553017944536e-05,
      "loss": 0.2684,
      "step": 190
    },
    {
      "epoch": 0.1,
      "learning_rate": 1.6313213703099512e-05,
      "loss": 0.2596,
      "step": 200
    },
    {
      "epoch": 0.1,
      "eval_f1": 0.731778447890841,
      "eval_f1_0": 0.9590831557063596,
      "eval_f1_1": 0.9844054916217028,
      "eval_f1_2": 0.00856203479610941,
      "eval_f1_3": 0.9750631094391922,
      "eval_loss": 0.21930557489395142,
      "eval_runtime": 39.669,
      "eval_samples_per_second": 182.913,
      "eval_steps_per_second": 0.731,
      "step": 200
    },
    {
      "epoch": 0.1,
      "learning_rate": 1.712887438825449e-05,
      "loss": 0.2242,
      "step": 210
    },
    {
      "epoch": 0.11,
      "learning_rate": 1.7944535073409462e-05,
      "loss": 0.1821,
      "step": 220
    },
    {
      "epoch": 0.11,
      "learning_rate": 1.876019575856444e-05,
      "loss": 0.1403,
      "step": 230
    },
    {
      "epoch": 0.12,
      "learning_rate": 1.9575856443719415e-05,
      "loss": 0.1311,
      "step": 240
    },
    {
      "epoch": 0.12,
      "learning_rate": 2.0391517128874392e-05,
      "loss": 0.135,
      "step": 250
    },
    {
      "epoch": 0.13,
      "learning_rate": 2.1207177814029365e-05,
      "loss": 0.0895,
      "step": 260
    },
    {
      "epoch": 0.13,
      "learning_rate": 2.202283849918434e-05,
      "loss": 0.0792,
      "step": 270
    },
    {
      "epoch": 0.14,
      "learning_rate": 2.2838499184339315e-05,
      "loss": 0.1469,
      "step": 280
    },
    {
      "epoch": 0.14,
      "learning_rate": 2.365415986949429e-05,
      "loss": 0.1319,
      "step": 290
    },
    {
      "epoch": 0.15,
      "learning_rate": 2.4469820554649268e-05,
      "loss": 0.0819,
      "step": 300
    },
    {
      "epoch": 0.15,
      "learning_rate": 2.528548123980424e-05,
      "loss": 0.0473,
      "step": 310
    },
    {
      "epoch": 0.16,
      "learning_rate": 2.6101141924959215e-05,
      "loss": 0.0695,
      "step": 320
    },
    {
      "epoch": 0.16,
      "learning_rate": 2.6916802610114195e-05,
      "loss": 0.135,
      "step": 330
    },
    {
      "epoch": 0.17,
      "learning_rate": 2.7732463295269168e-05,
      "loss": 0.0965,
      "step": 340
    },
    {
      "epoch": 0.17,
      "learning_rate": 2.8548123980424145e-05,
      "loss": 0.0796,
      "step": 350
    },
    {
      "epoch": 0.18,
      "learning_rate": 2.9363784665579118e-05,
      "loss": 0.0971,
      "step": 360
    },
    {
      "epoch": 0.18,
      "learning_rate": 3.0179445350734098e-05,
      "loss": 0.0752,
      "step": 370
    },
    {
      "epoch": 0.19,
      "learning_rate": 3.099510603588907e-05,
      "loss": 0.0407,
      "step": 380
    },
    {
      "epoch": 0.19,
      "learning_rate": 3.181076672104405e-05,
      "loss": 0.0791,
      "step": 390
    },
    {
      "epoch": 0.2,
      "learning_rate": 3.2626427406199024e-05,
      "loss": 0.0826,
      "step": 400
    },
    {
      "epoch": 0.2,
      "eval_f1": 0.9548707347713434,
      "eval_f1_0": 0.984936030666201,
      "eval_f1_1": 0.9930152382660391,
      "eval_f1_2": 0.8450223047576755,
      "eval_f1_3": 0.9965093653954585,
      "eval_loss": 0.0709918960928917,
      "eval_runtime": 40.0435,
      "eval_samples_per_second": 181.203,
      "eval_steps_per_second": 0.724,
      "step": 400
    },
    {
      "epoch": 0.2,
      "learning_rate": 3.3442088091354e-05,
      "loss": 0.0391,
      "step": 410
    },
    {
      "epoch": 0.21,
      "learning_rate": 3.425774877650898e-05,
      "loss": 0.0529,
      "step": 420
    },
    {
      "epoch": 0.21,
      "learning_rate": 3.507340946166395e-05,
      "loss": 0.0563,
      "step": 430
    },
    {
      "epoch": 0.22,
      "learning_rate": 3.5889070146818924e-05,
      "loss": 0.0254,
      "step": 440
    },
    {
      "epoch": 0.22,
      "learning_rate": 3.67047308319739e-05,
      "loss": 0.015,
      "step": 450
    },
    {
      "epoch": 0.23,
      "learning_rate": 3.752039151712888e-05,
      "loss": 0.0678,
      "step": 460
    },
    {
      "epoch": 0.23,
      "learning_rate": 3.833605220228385e-05,
      "loss": 0.0707,
      "step": 470
    },
    {
      "epoch": 0.24,
      "learning_rate": 3.915171288743883e-05,
      "loss": 0.0258,
      "step": 480
    },
    {
      "epoch": 0.24,
      "learning_rate": 3.9967373572593804e-05,
      "loss": 0.0335,
      "step": 490
    },
    {
      "epoch": 0.24,
      "learning_rate": 4.0783034257748784e-05,
      "loss": 0.0106,
      "step": 500
    },
    {
      "epoch": 0.25,
      "learning_rate": 4.159869494290376e-05,
      "loss": 0.0574,
      "step": 510
    },
    {
      "epoch": 0.25,
      "learning_rate": 4.241435562805873e-05,
      "loss": 0.0181,
      "step": 520
    },
    {
      "epoch": 0.26,
      "learning_rate": 4.3230016313213704e-05,
      "loss": 0.0269,
      "step": 530
    },
    {
      "epoch": 0.26,
      "learning_rate": 4.404567699836868e-05,
      "loss": 0.1033,
      "step": 540
    },
    {
      "epoch": 0.27,
      "learning_rate": 4.486133768352366e-05,
      "loss": 0.0703,
      "step": 550
    },
    {
      "epoch": 0.27,
      "learning_rate": 4.567699836867863e-05,
      "loss": 0.0245,
      "step": 560
    },
    {
      "epoch": 0.28,
      "learning_rate": 4.649265905383361e-05,
      "loss": 0.0608,
      "step": 570
    },
    {
      "epoch": 0.28,
      "learning_rate": 4.730831973898858e-05,
      "loss": 0.046,
      "step": 580
    },
    {
      "epoch": 0.29,
      "learning_rate": 4.812398042414356e-05,
      "loss": 0.0681,
      "step": 590
    },
    {
      "epoch": 0.29,
      "learning_rate": 4.8939641109298537e-05,
      "loss": 0.0107,
      "step": 600
    },
    {
      "epoch": 0.29,
      "eval_f1": 0.9633730287567567,
      "eval_f1_0": 0.9880053278996667,
      "eval_f1_1": 0.9945514803020932,
      "eval_f1_2": 0.8759356995262639,
      "eval_f1_3": 0.9949996072990025,
      "eval_loss": 0.06507362425327301,
      "eval_runtime": 40.0954,
      "eval_samples_per_second": 180.969,
      "eval_steps_per_second": 0.723,
      "step": 600
    },
    {
      "epoch": 0.3,
      "learning_rate": 4.975530179445351e-05,
      "loss": 0.0525,
      "step": 610
    },
    {
      "epoch": 0.3,
      "learning_rate": 4.996991317802803e-05,
      "loss": 0.0678,
      "step": 620
    },
    {
      "epoch": 0.31,
      "learning_rate": 4.9926932003782347e-05,
      "loss": 0.1054,
      "step": 630
    },
    {
      "epoch": 0.31,
      "learning_rate": 4.9883950829536664e-05,
      "loss": 0.0076,
      "step": 640
    },
    {
      "epoch": 0.32,
      "learning_rate": 4.984096965529099e-05,
      "loss": 0.0597,
      "step": 650
    },
    {
      "epoch": 0.32,
      "learning_rate": 4.9797988481045305e-05,
      "loss": 0.0473,
      "step": 660
    },
    {
      "epoch": 0.33,
      "learning_rate": 4.975500730679962e-05,
      "loss": 0.0291,
      "step": 670
    },
    {
      "epoch": 0.33,
      "learning_rate": 4.9712026132553947e-05,
      "loss": 0.0237,
      "step": 680
    },
    {
      "epoch": 0.34,
      "learning_rate": 4.9669044958308264e-05,
      "loss": 0.0175,
      "step": 690
    },
    {
      "epoch": 0.34,
      "learning_rate": 4.962606378406258e-05,
      "loss": 0.0227,
      "step": 700
    },
    {
      "epoch": 0.35,
      "learning_rate": 4.95830826098169e-05,
      "loss": 0.0617,
      "step": 710
    },
    {
      "epoch": 0.35,
      "learning_rate": 4.954010143557122e-05,
      "loss": 0.0953,
      "step": 720
    },
    {
      "epoch": 0.36,
      "learning_rate": 4.949712026132554e-05,
      "loss": 0.0473,
      "step": 730
    },
    {
      "epoch": 0.36,
      "learning_rate": 4.945413908707986e-05,
      "loss": 0.0207,
      "step": 740
    },
    {
      "epoch": 0.37,
      "learning_rate": 4.941115791283418e-05,
      "loss": 0.0106,
      "step": 750
    },
    {
      "epoch": 0.37,
      "learning_rate": 4.93681767385885e-05,
      "loss": 0.0203,
      "step": 760
    },
    {
      "epoch": 0.38,
      "learning_rate": 4.9325195564342816e-05,
      "loss": 0.0034,
      "step": 770
    },
    {
      "epoch": 0.38,
      "learning_rate": 4.928221439009714e-05,
      "loss": 0.049,
      "step": 780
    },
    {
      "epoch": 0.39,
      "learning_rate": 4.923923321585146e-05,
      "loss": 0.057,
      "step": 790
    },
    {
      "epoch": 0.39,
      "learning_rate": 4.919625204160578e-05,
      "loss": 0.021,
      "step": 800
    },
    {
      "epoch": 0.39,
      "eval_f1": 0.9804879290964876,
      "eval_f1_0": 0.9939161573147783,
      "eval_f1_1": 0.995125377430154,
      "eval_f1_2": 0.9346414305202709,
      "eval_f1_3": 0.9982687511207473,
      "eval_loss": 0.0359017439186573,
      "eval_runtime": 40.0322,
      "eval_samples_per_second": 181.254,
      "eval_steps_per_second": 0.724,
      "step": 800
    },
    {
      "epoch": 0.4,
      "learning_rate": 4.91532708673601e-05,
      "loss": 0.0456,
      "step": 810
    },
    {
      "epoch": 0.4,
      "learning_rate": 4.911028969311442e-05,
      "loss": 0.0196,
      "step": 820
    },
    {
      "epoch": 0.41,
      "learning_rate": 4.906730851886874e-05,
      "loss": 0.0089,
      "step": 830
    },
    {
      "epoch": 0.41,
      "learning_rate": 4.902432734462306e-05,
      "loss": 0.0368,
      "step": 840
    },
    {
      "epoch": 0.42,
      "learning_rate": 4.898134617037738e-05,
      "loss": 0.0066,
      "step": 850
    },
    {
      "epoch": 0.42,
      "learning_rate": 4.89383649961317e-05,
      "loss": 0.0051,
      "step": 860
    },
    {
      "epoch": 0.43,
      "learning_rate": 4.8895383821886016e-05,
      "loss": 0.0307,
      "step": 870
    },
    {
      "epoch": 0.43,
      "learning_rate": 4.885240264764034e-05,
      "loss": 0.017,
      "step": 880
    },
    {
      "epoch": 0.44,
      "learning_rate": 4.880942147339466e-05,
      "loss": 0.0334,
      "step": 890
    },
    {
      "epoch": 0.44,
      "learning_rate": 4.8766440299148975e-05,
      "loss": 0.0075,
      "step": 900
    },
    {
      "epoch": 0.45,
      "learning_rate": 4.872345912490329e-05,
      "loss": 0.0215,
      "step": 910
    },
    {
      "epoch": 0.45,
      "learning_rate": 4.8680477950657616e-05,
      "loss": 0.0232,
      "step": 920
    },
    {
      "epoch": 0.46,
      "learning_rate": 4.863749677641193e-05,
      "loss": 0.0241,
      "step": 930
    },
    {
      "epoch": 0.46,
      "learning_rate": 4.859451560216625e-05,
      "loss": 0.0126,
      "step": 940
    },
    {
      "epoch": 0.47,
      "learning_rate": 4.8551534427920575e-05,
      "loss": 0.0031,
      "step": 950
    },
    {
      "epoch": 0.47,
      "learning_rate": 4.850855325367489e-05,
      "loss": 0.0076,
      "step": 960
    },
    {
      "epoch": 0.48,
      "learning_rate": 4.846557207942921e-05,
      "loss": 0.0368,
      "step": 970
    },
    {
      "epoch": 0.48,
      "learning_rate": 4.842259090518353e-05,
      "loss": 0.0148,
      "step": 980
    },
    {
      "epoch": 0.49,
      "learning_rate": 4.837960973093785e-05,
      "loss": 0.0192,
      "step": 990
    },
    {
      "epoch": 0.49,
      "learning_rate": 4.833662855669217e-05,
      "loss": 0.0393,
      "step": 1000
    },
    {
      "epoch": 0.49,
      "eval_f1": 0.9932672548097791,
      "eval_f1_0": 0.9980382439938018,
      "eval_f1_1": 0.9976986814529715,
      "eval_f1_2": 0.979377910960893,
      "eval_f1_3": 0.9979541828314501,
      "eval_loss": 0.012854259461164474,
      "eval_runtime": 40.1294,
      "eval_samples_per_second": 180.815,
      "eval_steps_per_second": 0.723,
      "step": 1000
    },
    {
      "epoch": 0.49,
      "learning_rate": 4.829364738244649e-05,
      "loss": 0.0181,
      "step": 1010
    },
    {
      "epoch": 0.5,
      "learning_rate": 4.825066620820081e-05,
      "loss": 0.0189,
      "step": 1020
    },
    {
      "epoch": 0.5,
      "learning_rate": 4.820768503395513e-05,
      "loss": 0.005,
      "step": 1030
    },
    {
      "epoch": 0.51,
      "learning_rate": 4.816470385970945e-05,
      "loss": 0.0198,
      "step": 1040
    },
    {
      "epoch": 0.51,
      "learning_rate": 4.8121722685463775e-05,
      "loss": 0.0122,
      "step": 1050
    },
    {
      "epoch": 0.52,
      "learning_rate": 4.807874151121809e-05,
      "loss": 0.008,
      "step": 1060
    },
    {
      "epoch": 0.52,
      "learning_rate": 4.803576033697241e-05,
      "loss": 0.0292,
      "step": 1070
    },
    {
      "epoch": 0.53,
      "learning_rate": 4.7992779162726727e-05,
      "loss": 0.0313,
      "step": 1080
    },
    {
      "epoch": 0.53,
      "learning_rate": 4.794979798848105e-05,
      "loss": 0.0221,
      "step": 1090
    },
    {
      "epoch": 0.54,
      "learning_rate": 4.790681681423537e-05,
      "loss": 0.0252,
      "step": 1100
    },
    {
      "epoch": 0.54,
      "learning_rate": 4.7863835639989685e-05,
      "loss": 0.0172,
      "step": 1110
    },
    {
      "epoch": 0.55,
      "learning_rate": 4.782085446574401e-05,
      "loss": 0.0126,
      "step": 1120
    },
    {
      "epoch": 0.55,
      "learning_rate": 4.777787329149833e-05,
      "loss": 0.023,
      "step": 1130
    },
    {
      "epoch": 0.56,
      "learning_rate": 4.7734892117252644e-05,
      "loss": 0.0117,
      "step": 1140
    },
    {
      "epoch": 0.56,
      "learning_rate": 4.769191094300697e-05,
      "loss": 0.0416,
      "step": 1150
    },
    {
      "epoch": 0.57,
      "learning_rate": 4.7648929768761285e-05,
      "loss": 0.0159,
      "step": 1160
    },
    {
      "epoch": 0.57,
      "learning_rate": 4.76059485945156e-05,
      "loss": 0.007,
      "step": 1170
    },
    {
      "epoch": 0.58,
      "learning_rate": 4.756296742026992e-05,
      "loss": 0.0268,
      "step": 1180
    },
    {
      "epoch": 0.58,
      "learning_rate": 4.7519986246024244e-05,
      "loss": 0.0169,
      "step": 1190
    },
    {
      "epoch": 0.59,
      "learning_rate": 4.747700507177856e-05,
      "loss": 0.0194,
      "step": 1200
    },
    {
      "epoch": 0.59,
      "eval_f1": 0.9899000680338335,
      "eval_f1_0": 0.997036872372344,
      "eval_f1_1": 0.9976231674621133,
      "eval_f1_2": 0.9670989467244803,
      "eval_f1_3": 0.9978412855763963,
      "eval_loss": 0.018653137609362602,
      "eval_runtime": 40.0851,
      "eval_samples_per_second": 181.015,
      "eval_steps_per_second": 0.723,
      "step": 1200
    },
    {
      "epoch": 0.59,
      "learning_rate": 4.743402389753288e-05,
      "loss": 0.0236,
      "step": 1210
    },
    {
      "epoch": 0.6,
      "learning_rate": 4.73910427232872e-05,
      "loss": 0.0401,
      "step": 1220
    },
    {
      "epoch": 0.6,
      "learning_rate": 4.734806154904152e-05,
      "loss": 0.021,
      "step": 1230
    },
    {
      "epoch": 0.61,
      "learning_rate": 4.730508037479584e-05,
      "loss": 0.0099,
      "step": 1240
    },
    {
      "epoch": 0.61,
      "learning_rate": 4.726209920055016e-05,
      "loss": 0.0082,
      "step": 1250
    },
    {
      "epoch": 0.62,
      "learning_rate": 4.721911802630448e-05,
      "loss": 0.0114,
      "step": 1260
    },
    {
      "epoch": 0.62,
      "learning_rate": 4.71761368520588e-05,
      "loss": 0.0328,
      "step": 1270
    },
    {
      "epoch": 0.63,
      "learning_rate": 4.713315567781312e-05,
      "loss": 0.0279,
      "step": 1280
    },
    {
      "epoch": 0.63,
      "learning_rate": 4.7090174503567444e-05,
      "loss": 0.0072,
      "step": 1290
    },
    {
      "epoch": 0.64,
      "learning_rate": 4.704719332932176e-05,
      "loss": 0.011,
      "step": 1300
    },
    {
      "epoch": 0.64,
      "learning_rate": 4.700421215507608e-05,
      "loss": 0.0482,
      "step": 1310
    },
    {
      "epoch": 0.65,
      "learning_rate": 4.69612309808304e-05,
      "loss": 0.0186,
      "step": 1320
    },
    {
      "epoch": 0.65,
      "learning_rate": 4.691824980658472e-05,
      "loss": 0.0369,
      "step": 1330
    },
    {
      "epoch": 0.66,
      "learning_rate": 4.687526863233904e-05,
      "loss": 0.0132,
      "step": 1340
    },
    {
      "epoch": 0.66,
      "learning_rate": 4.6832287458093355e-05,
      "loss": 0.0134,
      "step": 1350
    },
    {
      "epoch": 0.67,
      "learning_rate": 4.678930628384768e-05,
      "loss": 0.0118,
      "step": 1360
    },
    {
      "epoch": 0.67,
      "learning_rate": 4.6746325109601996e-05,
      "loss": 0.0049,
      "step": 1370
    },
    {
      "epoch": 0.68,
      "learning_rate": 4.670334393535631e-05,
      "loss": 0.0449,
      "step": 1380
    },
    {
      "epoch": 0.68,
      "learning_rate": 4.666036276111064e-05,
      "loss": 0.0026,
      "step": 1390
    },
    {
      "epoch": 0.69,
      "learning_rate": 4.6617381586864955e-05,
      "loss": 0.0076,
      "step": 1400
    },
    {
      "epoch": 0.69,
      "eval_f1": 0.9936267177377152,
      "eval_f1_0": 0.9981375904084538,
      "eval_f1_1": 0.998167602527681,
      "eval_f1_2": 0.979420808585566,
      "eval_f1_3": 0.99878086942916,
      "eval_loss": 0.012272381223738194,
      "eval_runtime": 40.0717,
      "eval_samples_per_second": 181.075,
      "eval_steps_per_second": 0.724,
      "step": 1400
    },
    {
      "epoch": 0.69,
      "learning_rate": 4.657440041261927e-05,
      "loss": 0.0081,
      "step": 1410
    },
    {
      "epoch": 0.7,
      "learning_rate": 4.6531419238373596e-05,
      "loss": 0.0082,
      "step": 1420
    },
    {
      "epoch": 0.7,
      "learning_rate": 4.648843806412791e-05,
      "loss": 0.0266,
      "step": 1430
    },
    {
      "epoch": 0.71,
      "learning_rate": 4.644545688988223e-05,
      "loss": 0.0305,
      "step": 1440
    },
    {
      "epoch": 0.71,
      "learning_rate": 4.640247571563655e-05,
      "loss": 0.0082,
      "step": 1450
    },
    {
      "epoch": 0.72,
      "learning_rate": 4.635949454139087e-05,
      "loss": 0.0091,
      "step": 1460
    },
    {
      "epoch": 0.72,
      "learning_rate": 4.631651336714519e-05,
      "loss": 0.0184,
      "step": 1470
    },
    {
      "epoch": 0.73,
      "learning_rate": 4.627353219289951e-05,
      "loss": 0.0046,
      "step": 1480
    },
    {
      "epoch": 0.73,
      "learning_rate": 4.623055101865383e-05,
      "loss": 0.0127,
      "step": 1490
    },
    {
      "epoch": 0.73,
      "learning_rate": 4.6187569844408155e-05,
      "loss": 0.0028,
      "step": 1500
    },
    {
      "epoch": 0.74,
      "learning_rate": 4.614458867016247e-05,
      "loss": 0.0209,
      "step": 1510
    },
    {
      "epoch": 0.74,
      "learning_rate": 4.6101607495916796e-05,
      "loss": 0.0195,
      "step": 1520
    },
    {
      "epoch": 0.75,
      "learning_rate": 4.605862632167111e-05,
      "loss": 0.0389,
      "step": 1530
    },
    {
      "epoch": 0.75,
      "learning_rate": 4.601564514742543e-05,
      "loss": 0.0053,
      "step": 1540
    },
    {
      "epoch": 0.76,
      "learning_rate": 4.597266397317975e-05,
      "loss": 0.0019,
      "step": 1550
    },
    {
      "epoch": 0.76,
      "learning_rate": 4.592968279893407e-05,
      "loss": 0.014,
      "step": 1560
    },
    {
      "epoch": 0.77,
      "learning_rate": 4.588670162468839e-05,
      "loss": 0.0056,
      "step": 1570
    },
    {
      "epoch": 0.77,
      "learning_rate": 4.584372045044271e-05,
      "loss": 0.0242,
      "step": 1580
    },
    {
      "epoch": 0.78,
      "learning_rate": 4.580073927619703e-05,
      "loss": 0.0053,
      "step": 1590
    },
    {
      "epoch": 0.78,
      "learning_rate": 4.575775810195135e-05,
      "loss": 0.0027,
      "step": 1600
    },
    {
      "epoch": 0.78,
      "eval_f1": 0.9949512454901512,
      "eval_f1_0": 0.9985168254870895,
      "eval_f1_1": 0.998030585857894,
      "eval_f1_2": 0.9844798851905004,
      "eval_f1_3": 0.998777685425121,
      "eval_loss": 0.009146512486040592,
      "eval_runtime": 40.0476,
      "eval_samples_per_second": 181.184,
      "eval_steps_per_second": 0.724,
      "step": 1600
    },
    {
      "epoch": 0.79,
      "learning_rate": 4.5714776927705665e-05,
      "loss": 0.0307,
      "step": 1610
    },
    {
      "epoch": 0.79,
      "learning_rate": 4.567179575345999e-05,
      "loss": 0.0026,
      "step": 1620
    },
    {
      "epoch": 0.8,
      "learning_rate": 4.562881457921431e-05,
      "loss": 0.0112,
      "step": 1630
    },
    {
      "epoch": 0.8,
      "learning_rate": 4.5585833404968624e-05,
      "loss": 0.0069,
      "step": 1640
    },
    {
      "epoch": 0.81,
      "learning_rate": 4.554285223072294e-05,
      "loss": 0.0279,
      "step": 1650
    },
    {
      "epoch": 0.81,
      "learning_rate": 4.5499871056477265e-05,
      "loss": 0.0036,
      "step": 1660
    },
    {
      "epoch": 0.82,
      "learning_rate": 4.545688988223158e-05,
      "loss": 0.0102,
      "step": 1670
    },
    {
      "epoch": 0.82,
      "learning_rate": 4.54139087079859e-05,
      "loss": 0.0103,
      "step": 1680
    },
    {
      "epoch": 0.83,
      "learning_rate": 4.5370927533740224e-05,
      "loss": 0.0028,
      "step": 1690
    },
    {
      "epoch": 0.83,
      "learning_rate": 4.532794635949454e-05,
      "loss": 0.0095,
      "step": 1700
    },
    {
      "epoch": 0.84,
      "learning_rate": 4.5284965185248865e-05,
      "loss": 0.0118,
      "step": 1710
    },
    {
      "epoch": 0.84,
      "learning_rate": 4.524198401100318e-05,
      "loss": 0.0297,
      "step": 1720
    },
    {
      "epoch": 0.85,
      "learning_rate": 4.519900283675751e-05,
      "loss": 0.0073,
      "step": 1730
    },
    {
      "epoch": 0.85,
      "learning_rate": 4.5156021662511824e-05,
      "loss": 0.0156,
      "step": 1740
    },
    {
      "epoch": 0.86,
      "learning_rate": 4.511304048826614e-05,
      "loss": 0.0067,
      "step": 1750
    },
    {
      "epoch": 0.86,
      "learning_rate": 4.5070059314020465e-05,
      "loss": 0.0036,
      "step": 1760
    },
    {
      "epoch": 0.87,
      "learning_rate": 4.502707813977478e-05,
      "loss": 0.0177,
      "step": 1770
    },
    {
      "epoch": 0.87,
      "learning_rate": 4.49840969655291e-05,
      "loss": 0.0089,
      "step": 1780
    },
    {
      "epoch": 0.88,
      "learning_rate": 4.4941115791283424e-05,
      "loss": 0.0027,
      "step": 1790
    },
    {
      "epoch": 0.88,
      "learning_rate": 4.489813461703774e-05,
      "loss": 0.0077,
      "step": 1800
    },
    {
      "epoch": 0.88,
      "eval_f1": 0.9938341293831611,
      "eval_f1_0": 0.9981911614503538,
      "eval_f1_1": 0.9984664677767968,
      "eval_f1_2": 0.9799569769109422,
      "eval_f1_3": 0.9987219113945511,
      "eval_loss": 0.012283367104828358,
      "eval_runtime": 40.1221,
      "eval_samples_per_second": 180.848,
      "eval_steps_per_second": 0.723,
      "step": 1800
    },
    {
      "epoch": 0.89,
      "learning_rate": 4.485515344279206e-05,
      "loss": 0.0039,
      "step": 1810
    },
    {
      "epoch": 0.89,
      "learning_rate": 4.4812172268546376e-05,
      "loss": 0.0322,
      "step": 1820
    },
    {
      "epoch": 0.9,
      "learning_rate": 4.47691910943007e-05,
      "loss": 0.0074,
      "step": 1830
    },
    {
      "epoch": 0.9,
      "learning_rate": 4.472620992005502e-05,
      "loss": 0.0106,
      "step": 1840
    },
    {
      "epoch": 0.91,
      "learning_rate": 4.4683228745809335e-05,
      "loss": 0.0496,
      "step": 1850
    },
    {
      "epoch": 0.91,
      "learning_rate": 4.464024757156366e-05,
      "loss": 0.0089,
      "step": 1860
    },
    {
      "epoch": 0.92,
      "learning_rate": 4.4597266397317976e-05,
      "loss": 0.0099,
      "step": 1870
    },
    {
      "epoch": 0.92,
      "learning_rate": 4.455428522307229e-05,
      "loss": 0.0198,
      "step": 1880
    },
    {
      "epoch": 0.93,
      "learning_rate": 4.451130404882662e-05,
      "loss": 0.0077,
      "step": 1890
    },
    {
      "epoch": 0.93,
      "learning_rate": 4.4468322874580935e-05,
      "loss": 0.0025,
      "step": 1900
    },
    {
      "epoch": 0.94,
      "learning_rate": 4.442534170033525e-05,
      "loss": 0.0343,
      "step": 1910
    },
    {
      "epoch": 0.94,
      "learning_rate": 4.438236052608957e-05,
      "loss": 0.0067,
      "step": 1920
    },
    {
      "epoch": 0.95,
      "learning_rate": 4.433937935184389e-05,
      "loss": 0.0036,
      "step": 1930
    },
    {
      "epoch": 0.95,
      "learning_rate": 4.429639817759821e-05,
      "loss": 0.0066,
      "step": 1940
    },
    {
      "epoch": 0.96,
      "learning_rate": 4.4253417003352535e-05,
      "loss": 0.0199,
      "step": 1950
    },
    {
      "epoch": 0.96,
      "learning_rate": 4.421043582910686e-05,
      "loss": 0.0038,
      "step": 1960
    },
    {
      "epoch": 0.97,
      "learning_rate": 4.4167454654861176e-05,
      "loss": 0.0181,
      "step": 1970
    },
    {
      "epoch": 0.97,
      "learning_rate": 4.4124473480615493e-05,
      "loss": 0.0061,
      "step": 1980
    },
    {
      "epoch": 0.98,
      "learning_rate": 4.408149230636982e-05,
      "loss": 0.0027,
      "step": 1990
    },
    {
      "epoch": 0.98,
      "learning_rate": 4.4038511132124135e-05,
      "loss": 0.0029,
      "step": 2000
    },
    {
      "epoch": 0.98,
      "eval_f1": 0.995497771242362,
      "eval_f1_0": 0.9986944836696579,
      "eval_f1_1": 0.9983975883895573,
      "eval_f1_2": 0.9860269819791043,
      "eval_f1_3": 0.9988720309311286,
      "eval_loss": 0.009064940735697746,
      "eval_runtime": 40.0581,
      "eval_samples_per_second": 181.137,
      "eval_steps_per_second": 0.724,
      "step": 2000
    },
    {
      "epoch": 0.98,
      "learning_rate": 4.399552995787845e-05,
      "loss": 0.0137,
      "step": 2010
    },
    {
      "epoch": 0.99,
      "learning_rate": 4.395254878363277e-05,
      "loss": 0.0045,
      "step": 2020
    },
    {
      "epoch": 0.99,
      "learning_rate": 4.3909567609387093e-05,
      "loss": 0.0159,
      "step": 2030
    },
    {
      "epoch": 1.0,
      "learning_rate": 4.386658643514141e-05,
      "loss": 0.0062,
      "step": 2040
    },
    {
      "epoch": 1.0,
      "learning_rate": 4.382360526089573e-05,
      "loss": 0.0127,
      "step": 2050
    },
    {
      "epoch": 1.01,
      "learning_rate": 4.378062408665005e-05,
      "loss": 0.011,
      "step": 2060
    },
    {
      "epoch": 1.01,
      "learning_rate": 4.373764291240437e-05,
      "loss": 0.0093,
      "step": 2070
    },
    {
      "epoch": 1.02,
      "learning_rate": 4.369466173815869e-05,
      "loss": 0.0054,
      "step": 2080
    },
    {
      "epoch": 1.02,
      "learning_rate": 4.3651680563913004e-05,
      "loss": 0.012,
      "step": 2090
    },
    {
      "epoch": 1.03,
      "learning_rate": 4.360869938966733e-05,
      "loss": 0.0173,
      "step": 2100
    },
    {
      "epoch": 1.03,
      "learning_rate": 4.3565718215421645e-05,
      "loss": 0.0032,
      "step": 2110
    },
    {
      "epoch": 1.04,
      "learning_rate": 4.352273704117596e-05,
      "loss": 0.025,
      "step": 2120
    },
    {
      "epoch": 1.04,
      "learning_rate": 4.347975586693029e-05,
      "loss": 0.0039,
      "step": 2130
    },
    {
      "epoch": 1.05,
      "learning_rate": 4.3436774692684604e-05,
      "loss": 0.0114,
      "step": 2140
    },
    {
      "epoch": 1.05,
      "learning_rate": 4.339379351843892e-05,
      "loss": 0.0048,
      "step": 2150
    },
    {
      "epoch": 1.06,
      "learning_rate": 4.3350812344193245e-05,
      "loss": 0.0035,
      "step": 2160
    },
    {
      "epoch": 1.06,
      "learning_rate": 4.330783116994756e-05,
      "loss": 0.0072,
      "step": 2170
    },
    {
      "epoch": 1.07,
      "learning_rate": 4.326484999570189e-05,
      "loss": 0.0025,
      "step": 2180
    },
    {
      "epoch": 1.07,
      "learning_rate": 4.3221868821456204e-05,
      "loss": 0.0037,
      "step": 2190
    },
    {
      "epoch": 1.08,
      "learning_rate": 4.317888764721053e-05,
      "loss": 0.0018,
      "step": 2200
    },
    {
      "epoch": 1.08,
      "eval_f1": 0.99474048450953,
      "eval_f1_0": 0.9984565576014481,
      "eval_f1_1": 0.9984912394857657,
      "eval_f1_2": 0.9830930443157153,
      "eval_f1_3": 0.9989210966351905,
      "eval_loss": 0.011158174835145473,
      "eval_runtime": 40.0892,
      "eval_samples_per_second": 180.997,
      "eval_steps_per_second": 0.723,
      "step": 2200
    },
    {
      "epoch": 1.08,
      "learning_rate": 4.3135906472964845e-05,
      "loss": 0.0021,
      "step": 2210
    },
    {
      "epoch": 1.09,
      "learning_rate": 4.309292529871916e-05,
      "loss": 0.0019,
      "step": 2220
    },
    {
      "epoch": 1.09,
      "learning_rate": 4.304994412447349e-05,
      "loss": 0.0181,
      "step": 2230
    },
    {
      "epoch": 1.1,
      "learning_rate": 4.3006962950227804e-05,
      "loss": 0.0045,
      "step": 2240
    },
    {
      "epoch": 1.1,
      "learning_rate": 4.296398177598212e-05,
      "loss": 0.0017,
      "step": 2250
    },
    {
      "epoch": 1.11,
      "learning_rate": 4.2921000601736446e-05,
      "loss": 0.0013,
      "step": 2260
    },
    {
      "epoch": 1.11,
      "learning_rate": 4.287801942749076e-05,
      "loss": 0.0129,
      "step": 2270
    },
    {
      "epoch": 1.12,
      "learning_rate": 4.283503825324508e-05,
      "loss": 0.0033,
      "step": 2280
    },
    {
      "epoch": 1.12,
      "learning_rate": 4.27920570789994e-05,
      "loss": 0.0047,
      "step": 2290
    },
    {
      "epoch": 1.13,
      "learning_rate": 4.274907590475372e-05,
      "loss": 0.0153,
      "step": 2300
    },
    {
      "epoch": 1.13,
      "learning_rate": 4.270609473050804e-05,
      "loss": 0.0034,
      "step": 2310
    },
    {
      "epoch": 1.14,
      "learning_rate": 4.2663113556262356e-05,
      "loss": 0.0302,
      "step": 2320
    },
    {
      "epoch": 1.14,
      "learning_rate": 4.262013238201668e-05,
      "loss": 0.0024,
      "step": 2330
    },
    {
      "epoch": 1.15,
      "learning_rate": 4.2577151207771e-05,
      "loss": 0.0023,
      "step": 2340
    },
    {
      "epoch": 1.15,
      "learning_rate": 4.2534170033525315e-05,
      "loss": 0.0097,
      "step": 2350
    },
    {
      "epoch": 1.16,
      "learning_rate": 4.249118885927964e-05,
      "loss": 0.0013,
      "step": 2360
    },
    {
      "epoch": 1.16,
      "learning_rate": 4.2448207685033956e-05,
      "loss": 0.0101,
      "step": 2370
    },
    {
      "epoch": 1.17,
      "learning_rate": 4.2405226510788273e-05,
      "loss": 0.0069,
      "step": 2380
    },
    {
      "epoch": 1.17,
      "learning_rate": 4.23622453365426e-05,
      "loss": 0.0015,
      "step": 2390
    },
    {
      "epoch": 1.18,
      "learning_rate": 4.2319264162296915e-05,
      "loss": 0.0021,
      "step": 2400
    },
    {
      "epoch": 1.18,
      "eval_f1": 0.9947890293449169,
      "eval_f1_0": 0.9984828294023925,
      "eval_f1_1": 0.9985711051198578,
      "eval_f1_2": 0.9831778722765446,
      "eval_f1_3": 0.9989243105808723,
      "eval_loss": 0.009640231728553772,
      "eval_runtime": 40.1584,
      "eval_samples_per_second": 180.685,
      "eval_steps_per_second": 0.722,
      "step": 2400
    },
    {
      "epoch": 1.18,
      "learning_rate": 4.227628298805124e-05,
      "loss": 0.0016,
      "step": 2410
    },
    {
      "epoch": 1.19,
      "learning_rate": 4.2233301813805556e-05,
      "loss": 0.0036,
      "step": 2420
    },
    {
      "epoch": 1.19,
      "learning_rate": 4.219032063955988e-05,
      "loss": 0.0031,
      "step": 2430
    },
    {
      "epoch": 1.2,
      "learning_rate": 4.21473394653142e-05,
      "loss": 0.0024,
      "step": 2440
    },
    {
      "epoch": 1.2,
      "learning_rate": 4.2104358291068515e-05,
      "loss": 0.0022,
      "step": 2450
    },
    {
      "epoch": 1.21,
      "learning_rate": 4.206137711682283e-05,
      "loss": 0.0275,
      "step": 2460
    },
    {
      "epoch": 1.21,
      "learning_rate": 4.2018395942577156e-05,
      "loss": 0.0034,
      "step": 2470
    },
    {
      "epoch": 1.22,
      "learning_rate": 4.1975414768331473e-05,
      "loss": 0.0105,
      "step": 2480
    },
    {
      "epoch": 1.22,
      "learning_rate": 4.193243359408579e-05,
      "loss": 0.0106,
      "step": 2490
    },
    {
      "epoch": 1.22,
      "learning_rate": 4.1889452419840115e-05,
      "loss": 0.0058,
      "step": 2500
    },
    {
      "epoch": 1.23,
      "learning_rate": 4.184647124559443e-05,
      "loss": 0.0018,
      "step": 2510
    },
    {
      "epoch": 1.23,
      "learning_rate": 4.180349007134875e-05,
      "loss": 0.0353,
      "step": 2520
    },
    {
      "epoch": 1.24,
      "learning_rate": 4.1760508897103074e-05,
      "loss": 0.0101,
      "step": 2530
    },
    {
      "epoch": 1.24,
      "learning_rate": 4.171752772285739e-05,
      "loss": 0.0022,
      "step": 2540
    },
    {
      "epoch": 1.25,
      "learning_rate": 4.167454654861171e-05,
      "loss": 0.0024,
      "step": 2550
    },
    {
      "epoch": 1.25,
      "learning_rate": 4.1631565374366025e-05,
      "loss": 0.0035,
      "step": 2560
    },
    {
      "epoch": 1.26,
      "learning_rate": 4.158858420012035e-05,
      "loss": 0.0033,
      "step": 2570
    },
    {
      "epoch": 1.26,
      "learning_rate": 4.154560302587467e-05,
      "loss": 0.0016,
      "step": 2580
    },
    {
      "epoch": 1.27,
      "learning_rate": 4.1502621851628984e-05,
      "loss": 0.0022,
      "step": 2590
    },
    {
      "epoch": 1.27,
      "learning_rate": 4.145964067738331e-05,
      "loss": 0.0022,
      "step": 2600
    },
    {
      "epoch": 1.27,
      "eval_f1": 0.9938231404374757,
      "eval_f1_0": 0.9981899002313855,
      "eval_f1_1": 0.9985176276714576,
      "eval_f1_2": 0.9796215262004841,
      "eval_f1_3": 0.9989635076465757,
      "eval_loss": 0.010439964011311531,
      "eval_runtime": 40.056,
      "eval_samples_per_second": 181.146,
      "eval_steps_per_second": 0.724,
      "step": 2600
    },
    {
      "epoch": 1.28,
      "learning_rate": 4.1416659503137625e-05,
      "loss": 0.0013,
      "step": 2610
    },
    {
      "epoch": 1.28,
      "learning_rate": 4.137367832889194e-05,
      "loss": 0.0016,
      "step": 2620
    },
    {
      "epoch": 1.29,
      "learning_rate": 4.133069715464627e-05,
      "loss": 0.0031,
      "step": 2630
    },
    {
      "epoch": 1.29,
      "learning_rate": 4.128771598040059e-05,
      "loss": 0.0019,
      "step": 2640
    },
    {
      "epoch": 1.3,
      "learning_rate": 4.124473480615491e-05,
      "loss": 0.0145,
      "step": 2650
    },
    {
      "epoch": 1.3,
      "learning_rate": 4.1201753631909225e-05,
      "loss": 0.0024,
      "step": 2660
    },
    {
      "epoch": 1.31,
      "learning_rate": 4.115877245766355e-05,
      "loss": 0.0054,
      "step": 2670
    },
    {
      "epoch": 1.31,
      "learning_rate": 4.111579128341787e-05,
      "loss": 0.012,
      "step": 2680
    },
    {
      "epoch": 1.32,
      "learning_rate": 4.1072810109172184e-05,
      "loss": 0.013,
      "step": 2690
    },
    {
      "epoch": 1.32,
      "learning_rate": 4.102982893492651e-05,
      "loss": 0.025,
      "step": 2700
    },
    {
      "epoch": 1.33,
      "learning_rate": 4.0986847760680826e-05,
      "loss": 0.0018,
      "step": 2710
    },
    {
      "epoch": 1.33,
      "learning_rate": 4.094386658643514e-05,
      "loss": 0.0108,
      "step": 2720
    },
    {
      "epoch": 1.34,
      "learning_rate": 4.090088541218946e-05,
      "loss": 0.0019,
      "step": 2730
    },
    {
      "epoch": 1.34,
      "learning_rate": 4.0857904237943784e-05,
      "loss": 0.0027,
      "step": 2740
    },
    {
      "epoch": 1.35,
      "learning_rate": 4.08149230636981e-05,
      "loss": 0.0022,
      "step": 2750
    },
    {
      "epoch": 1.35,
      "learning_rate": 4.077194188945242e-05,
      "loss": 0.0035,
      "step": 2760
    },
    {
      "epoch": 1.36,
      "learning_rate": 4.072896071520674e-05,
      "loss": 0.0019,
      "step": 2770
    },
    {
      "epoch": 1.36,
      "learning_rate": 4.068597954096106e-05,
      "loss": 0.0018,
      "step": 2780
    },
    {
      "epoch": 1.37,
      "learning_rate": 4.064299836671538e-05,
      "loss": 0.002,
      "step": 2790
    },
    {
      "epoch": 1.37,
      "learning_rate": 4.06000171924697e-05,
      "loss": 0.0068,
      "step": 2800
    },
    {
      "epoch": 1.37,
      "eval_f1": 0.9946561585897573,
      "eval_f1_0": 0.9984361395547772,
      "eval_f1_1": 0.9984220794670473,
      "eval_f1_2": 0.9828323083893421,
      "eval_f1_3": 0.9989341069478626,
      "eval_loss": 0.00828581303358078,
      "eval_runtime": 39.9994,
      "eval_samples_per_second": 181.403,
      "eval_steps_per_second": 0.725,
      "step": 2800
    },
    {
      "epoch": 1.38,
      "learning_rate": 4.055703601822402e-05,
      "loss": 0.004,
      "step": 2810
    },
    {
      "epoch": 1.38,
      "learning_rate": 4.0514054843978336e-05,
      "loss": 0.0405,
      "step": 2820
    },
    {
      "epoch": 1.39,
      "learning_rate": 4.0471073669732653e-05,
      "loss": 0.0129,
      "step": 2830
    },
    {
      "epoch": 1.39,
      "learning_rate": 4.042809249548698e-05,
      "loss": 0.0025,
      "step": 2840
    },
    {
      "epoch": 1.4,
      "learning_rate": 4.0385111321241295e-05,
      "loss": 0.0044,
      "step": 2850
    },
    {
      "epoch": 1.4,
      "learning_rate": 4.034213014699562e-05,
      "loss": 0.0032,
      "step": 2860
    },
    {
      "epoch": 1.41,
      "learning_rate": 4.0299148972749936e-05,
      "loss": 0.0029,
      "step": 2870
    },
    {
      "epoch": 1.41,
      "learning_rate": 4.025616779850426e-05,
      "loss": 0.0083,
      "step": 2880
    },
    {
      "epoch": 1.42,
      "learning_rate": 4.021318662425858e-05,
      "loss": 0.0037,
      "step": 2890
    },
    {
      "epoch": 1.42,
      "learning_rate": 4.01702054500129e-05,
      "loss": 0.0044,
      "step": 2900
    },
    {
      "epoch": 1.43,
      "learning_rate": 4.012722427576722e-05,
      "loss": 0.0147,
      "step": 2910
    },
    {
      "epoch": 1.43,
      "learning_rate": 4.0084243101521536e-05,
      "loss": 0.0064,
      "step": 2920
    },
    {
      "epoch": 1.44,
      "learning_rate": 4.0041261927275854e-05,
      "loss": 0.0136,
      "step": 2930
    },
    {
      "epoch": 1.44,
      "learning_rate": 3.999828075303018e-05,
      "loss": 0.0034,
      "step": 2940
    },
    {
      "epoch": 1.45,
      "learning_rate": 3.9955299578784495e-05,
      "loss": 0.004,
      "step": 2950
    },
    {
      "epoch": 1.45,
      "learning_rate": 3.991231840453881e-05,
      "loss": 0.0017,
      "step": 2960
    },
    {
      "epoch": 1.46,
      "learning_rate": 3.9869337230293136e-05,
      "loss": 0.002,
      "step": 2970
    },
    {
      "epoch": 1.46,
      "learning_rate": 3.9826356056047454e-05,
      "loss": 0.0021,
      "step": 2980
    },
    {
      "epoch": 1.46,
      "learning_rate": 3.978337488180177e-05,
      "loss": 0.0167,
      "step": 2990
    },
    {
      "epoch": 1.47,
      "learning_rate": 3.9740393707556095e-05,
      "loss": 0.0158,
      "step": 3000
    },
    {
      "epoch": 1.47,
      "eval_f1": 0.9930835179937217,
      "eval_f1_0": 0.9979343701242913,
      "eval_f1_1": 0.997910854226378,
      "eval_f1_2": 0.9775809813274636,
      "eval_f1_3": 0.9989078662967539,
      "eval_loss": 0.012813008390367031,
      "eval_runtime": 40.0833,
      "eval_samples_per_second": 181.023,
      "eval_steps_per_second": 0.723,
      "step": 3000
    },
    {
      "epoch": 1.47,
      "learning_rate": 3.969741253331041e-05,
      "loss": 0.0287,
      "step": 3010
    },
    {
      "epoch": 1.48,
      "learning_rate": 3.965443135906473e-05,
      "loss": 0.0061,
      "step": 3020
    },
    {
      "epoch": 1.48,
      "learning_rate": 3.961145018481905e-05,
      "loss": 0.0104,
      "step": 3030
    },
    {
      "epoch": 1.49,
      "learning_rate": 3.956846901057337e-05,
      "loss": 0.0134,
      "step": 3040
    },
    {
      "epoch": 1.49,
      "learning_rate": 3.952548783632769e-05,
      "loss": 0.0073,
      "step": 3050
    },
    {
      "epoch": 1.5,
      "learning_rate": 3.9482506662082005e-05,
      "loss": 0.0111,
      "step": 3060
    },
    {
      "epoch": 1.5,
      "learning_rate": 3.943952548783633e-05,
      "loss": 0.0083,
      "step": 3070
    },
    {
      "epoch": 1.51,
      "learning_rate": 3.939654431359065e-05,
      "loss": 0.0026,
      "step": 3080
    },
    {
      "epoch": 1.51,
      "learning_rate": 3.935356313934497e-05,
      "loss": 0.0024,
      "step": 3090
    },
    {
      "epoch": 1.52,
      "learning_rate": 3.931058196509929e-05,
      "loss": 0.0021,
      "step": 3100
    },
    {
      "epoch": 1.52,
      "learning_rate": 3.926760079085361e-05,
      "loss": 0.0078,
      "step": 3110
    },
    {
      "epoch": 1.53,
      "learning_rate": 3.922461961660793e-05,
      "loss": 0.0038,
      "step": 3120
    },
    {
      "epoch": 1.53,
      "learning_rate": 3.918163844236225e-05,
      "loss": 0.0041,
      "step": 3130
    },
    {
      "epoch": 1.54,
      "learning_rate": 3.913865726811657e-05,
      "loss": 0.0022,
      "step": 3140
    },
    {
      "epoch": 1.54,
      "learning_rate": 3.909567609387089e-05,
      "loss": 0.0054,
      "step": 3150
    },
    {
      "epoch": 1.55,
      "learning_rate": 3.9052694919625206e-05,
      "loss": 0.0025,
      "step": 3160
    },
    {
      "epoch": 1.55,
      "learning_rate": 3.900971374537953e-05,
      "loss": 0.0038,
      "step": 3170
    },
    {
      "epoch": 1.56,
      "learning_rate": 3.896673257113385e-05,
      "loss": 0.0042,
      "step": 3180
    },
    {
      "epoch": 1.56,
      "learning_rate": 3.8923751396888164e-05,
      "loss": 0.0029,
      "step": 3190
    },
    {
      "epoch": 1.57,
      "learning_rate": 3.888077022264248e-05,
      "loss": 0.0041,
      "step": 3200
    },
    {
      "epoch": 1.57,
      "eval_f1": 0.9947868324122469,
      "eval_f1_0": 0.9984707643374676,
      "eval_f1_1": 0.9984966870026223,
      "eval_f1_2": 0.9831837307556248,
      "eval_f1_3": 0.9989961475532726,
      "eval_loss": 0.008540821261703968,
      "eval_runtime": 40.1168,
      "eval_samples_per_second": 180.872,
      "eval_steps_per_second": 0.723,
      "step": 3200
    },
    {
      "epoch": 1.57,
      "learning_rate": 3.8837789048396806e-05,
      "loss": 0.0031,
      "step": 3210
    },
    {
      "epoch": 1.58,
      "learning_rate": 3.879480787415112e-05,
      "loss": 0.0041,
      "step": 3220
    },
    {
      "epoch": 1.58,
      "learning_rate": 3.875182669990544e-05,
      "loss": 0.0021,
      "step": 3230
    },
    {
      "epoch": 1.59,
      "learning_rate": 3.8708845525659764e-05,
      "loss": 0.0114,
      "step": 3240
    },
    {
      "epoch": 1.59,
      "learning_rate": 3.866586435141408e-05,
      "loss": 0.0079,
      "step": 3250
    },
    {
      "epoch": 1.6,
      "learning_rate": 3.86228831771684e-05,
      "loss": 0.0056,
      "step": 3260
    },
    {
      "epoch": 1.6,
      "learning_rate": 3.857990200292272e-05,
      "loss": 0.0034,
      "step": 3270
    },
    {
      "epoch": 1.61,
      "learning_rate": 3.853692082867704e-05,
      "loss": 0.0034,
      "step": 3280
    },
    {
      "epoch": 1.61,
      "learning_rate": 3.849393965443136e-05,
      "loss": 0.0062,
      "step": 3290
    },
    {
      "epoch": 1.62,
      "learning_rate": 3.8450958480185675e-05,
      "loss": 0.0022,
      "step": 3300
    },
    {
      "epoch": 1.62,
      "learning_rate": 3.840797730594e-05,
      "loss": 0.0031,
      "step": 3310
    },
    {
      "epoch": 1.63,
      "learning_rate": 3.836499613169432e-05,
      "loss": 0.003,
      "step": 3320
    },
    {
      "epoch": 1.63,
      "learning_rate": 3.832201495744864e-05,
      "loss": 0.0105,
      "step": 3330
    },
    {
      "epoch": 1.64,
      "learning_rate": 3.8279033783202964e-05,
      "loss": 0.025,
      "step": 3340
    },
    {
      "epoch": 1.64,
      "learning_rate": 3.823605260895728e-05,
      "loss": 0.0065,
      "step": 3350
    },
    {
      "epoch": 1.65,
      "learning_rate": 3.81930714347116e-05,
      "loss": 0.0048,
      "step": 3360
    },
    {
      "epoch": 1.65,
      "learning_rate": 3.815009026046592e-05,
      "loss": 0.0163,
      "step": 3370
    },
    {
      "epoch": 1.66,
      "learning_rate": 3.810710908622024e-05,
      "loss": 0.0058,
      "step": 3380
    },
    {
      "epoch": 1.66,
      "learning_rate": 3.806412791197456e-05,
      "loss": 0.0014,
      "step": 3390
    },
    {
      "epoch": 1.67,
      "learning_rate": 3.8021146737728875e-05,
      "loss": 0.0027,
      "step": 3400
    },
    {
      "epoch": 1.67,
      "eval_f1": 0.9945953876498776,
      "eval_f1_0": 0.9984039847924702,
      "eval_f1_1": 0.9986756781441227,
      "eval_f1_2": 0.9823058382737626,
      "eval_f1_3": 0.9989960493891548,
      "eval_loss": 0.00983767956495285,
      "eval_runtime": 40.0138,
      "eval_samples_per_second": 181.337,
      "eval_steps_per_second": 0.725,
      "step": 3400
    },
    {
      "epoch": 1.67,
      "learning_rate": 3.79781655634832e-05,
      "loss": 0.003,
      "step": 3410
    },
    {
      "epoch": 1.68,
      "learning_rate": 3.7935184389237516e-05,
      "loss": 0.0018,
      "step": 3420
    },
    {
      "epoch": 1.68,
      "learning_rate": 3.7892203214991834e-05,
      "loss": 0.0298,
      "step": 3430
    },
    {
      "epoch": 1.69,
      "learning_rate": 3.784922204074616e-05,
      "loss": 0.0019,
      "step": 3440
    },
    {
      "epoch": 1.69,
      "learning_rate": 3.7806240866500475e-05,
      "loss": 0.0068,
      "step": 3450
    },
    {
      "epoch": 1.7,
      "learning_rate": 3.776325969225479e-05,
      "loss": 0.0028,
      "step": 3460
    },
    {
      "epoch": 1.7,
      "learning_rate": 3.772027851800911e-05,
      "loss": 0.0199,
      "step": 3470
    },
    {
      "epoch": 1.71,
      "learning_rate": 3.7677297343763434e-05,
      "loss": 0.0028,
      "step": 3480
    },
    {
      "epoch": 1.71,
      "learning_rate": 3.763431616951775e-05,
      "loss": 0.0085,
      "step": 3490
    },
    {
      "epoch": 1.71,
      "learning_rate": 3.759133499527207e-05,
      "loss": 0.006,
      "step": 3500
    },
    {
      "epoch": 1.72,
      "learning_rate": 3.754835382102639e-05,
      "loss": 0.0012,
      "step": 3510
    },
    {
      "epoch": 1.72,
      "learning_rate": 3.750537264678071e-05,
      "loss": 0.0073,
      "step": 3520
    },
    {
      "epoch": 1.73,
      "learning_rate": 3.746239147253503e-05,
      "loss": 0.0087,
      "step": 3530
    },
    {
      "epoch": 1.73,
      "learning_rate": 3.741941029828935e-05,
      "loss": 0.003,
      "step": 3540
    },
    {
      "epoch": 1.74,
      "learning_rate": 3.737642912404367e-05,
      "loss": 0.0049,
      "step": 3550
    },
    {
      "epoch": 1.74,
      "learning_rate": 3.733344794979799e-05,
      "loss": 0.0065,
      "step": 3560
    },
    {
      "epoch": 1.75,
      "learning_rate": 3.729046677555231e-05,
      "loss": 0.0213,
      "step": 3570
    },
    {
      "epoch": 1.75,
      "learning_rate": 3.7247485601306634e-05,
      "loss": 0.0027,
      "step": 3580
    },
    {
      "epoch": 1.76,
      "learning_rate": 3.720450442706095e-05,
      "loss": 0.0024,
      "step": 3590
    },
    {
      "epoch": 1.76,
      "learning_rate": 3.716152325281527e-05,
      "loss": 0.0011,
      "step": 3600
    },
    {
      "epoch": 1.76,
      "eval_f1": 0.997014901060682,
      "eval_f1_0": 0.999131830343566,
      "eval_f1_1": 0.9986928054829864,
      "eval_f1_2": 0.9912356692327082,
      "eval_f1_3": 0.9989992991834673,
      "eval_loss": 0.005678725894540548,
      "eval_runtime": 40.0443,
      "eval_samples_per_second": 181.2,
      "eval_steps_per_second": 0.724,
      "step": 3600
    },
    {
      "epoch": 1.77,
      "learning_rate": 3.711854207856959e-05,
      "loss": 0.0041,
      "step": 3610
    },
    {
      "epoch": 1.77,
      "learning_rate": 3.707556090432391e-05,
      "loss": 0.0032,
      "step": 3620
    },
    {
      "epoch": 1.78,
      "learning_rate": 3.703257973007823e-05,
      "loss": 0.0036,
      "step": 3630
    },
    {
      "epoch": 1.78,
      "learning_rate": 3.698959855583255e-05,
      "loss": 0.0048,
      "step": 3640
    },
    {
      "epoch": 1.79,
      "learning_rate": 3.694661738158687e-05,
      "loss": 0.0011,
      "step": 3650
    },
    {
      "epoch": 1.79,
      "learning_rate": 3.6903636207341186e-05,
      "loss": 0.0057,
      "step": 3660
    },
    {
      "epoch": 1.8,
      "learning_rate": 3.68606550330955e-05,
      "loss": 0.0033,
      "step": 3670
    },
    {
      "epoch": 1.8,
      "learning_rate": 3.681767385884983e-05,
      "loss": 0.0021,
      "step": 3680
    },
    {
      "epoch": 1.81,
      "learning_rate": 3.6774692684604144e-05,
      "loss": 0.0273,
      "step": 3690
    },
    {
      "epoch": 1.81,
      "learning_rate": 3.673171151035846e-05,
      "loss": 0.0029,
      "step": 3700
    },
    {
      "epoch": 1.82,
      "learning_rate": 3.6688730336112786e-05,
      "loss": 0.0021,
      "step": 3710
    },
    {
      "epoch": 1.82,
      "learning_rate": 3.66457491618671e-05,
      "loss": 0.0013,
      "step": 3720
    },
    {
      "epoch": 1.83,
      "learning_rate": 3.660276798762142e-05,
      "loss": 0.0045,
      "step": 3730
    },
    {
      "epoch": 1.83,
      "learning_rate": 3.655978681337574e-05,
      "loss": 0.0047,
      "step": 3740
    },
    {
      "epoch": 1.84,
      "learning_rate": 3.651680563913006e-05,
      "loss": 0.0055,
      "step": 3750
    },
    {
      "epoch": 1.84,
      "learning_rate": 3.647382446488438e-05,
      "loss": 0.0017,
      "step": 3760
    },
    {
      "epoch": 1.85,
      "learning_rate": 3.64308432906387e-05,
      "loss": 0.0024,
      "step": 3770
    },
    {
      "epoch": 1.85,
      "learning_rate": 3.638786211639302e-05,
      "loss": 0.0062,
      "step": 3780
    },
    {
      "epoch": 1.86,
      "learning_rate": 3.6344880942147344e-05,
      "loss": 0.0031,
      "step": 3790
    },
    {
      "epoch": 1.86,
      "learning_rate": 3.630189976790166e-05,
      "loss": 0.0029,
      "step": 3800
    },
    {
      "epoch": 1.86,
      "eval_f1": 0.9964862061001086,
      "eval_f1_0": 0.9989784490103896,
      "eval_f1_1": 0.9986412578070586,
      "eval_f1_2": 0.9893452896107167,
      "eval_f1_3": 0.9989798279722696,
      "eval_loss": 0.007412091363221407,
      "eval_runtime": 39.9801,
      "eval_samples_per_second": 181.49,
      "eval_steps_per_second": 0.725,
      "step": 3800
    },
    {
      "epoch": 1.87,
      "learning_rate": 3.6258918593655986e-05,
      "loss": 0.0338,
      "step": 3810
    },
    {
      "epoch": 1.87,
      "learning_rate": 3.62159374194103e-05,
      "loss": 0.0168,
      "step": 3820
    },
    {
      "epoch": 1.88,
      "learning_rate": 3.617295624516462e-05,
      "loss": 0.0147,
      "step": 3830
    },
    {
      "epoch": 1.88,
      "learning_rate": 3.612997507091894e-05,
      "loss": 0.0268,
      "step": 3840
    },
    {
      "epoch": 1.89,
      "learning_rate": 3.608699389667326e-05,
      "loss": 0.0034,
      "step": 3850
    },
    {
      "epoch": 1.89,
      "learning_rate": 3.604401272242758e-05,
      "loss": 0.0044,
      "step": 3860
    },
    {
      "epoch": 1.9,
      "learning_rate": 3.6001031548181896e-05,
      "loss": 0.0031,
      "step": 3870
    },
    {
      "epoch": 1.9,
      "learning_rate": 3.595805037393622e-05,
      "loss": 0.0039,
      "step": 3880
    },
    {
      "epoch": 1.91,
      "learning_rate": 3.591506919969054e-05,
      "loss": 0.0017,
      "step": 3890
    },
    {
      "epoch": 1.91,
      "learning_rate": 3.5872088025444855e-05,
      "loss": 0.0022,
      "step": 3900
    },
    {
      "epoch": 1.92,
      "learning_rate": 3.582910685119918e-05,
      "loss": 0.0191,
      "step": 3910
    },
    {
      "epoch": 1.92,
      "learning_rate": 3.5786125676953496e-05,
      "loss": 0.0017,
      "step": 3920
    },
    {
      "epoch": 1.93,
      "learning_rate": 3.5743144502707814e-05,
      "loss": 0.0051,
      "step": 3930
    },
    {
      "epoch": 1.93,
      "learning_rate": 3.570016332846213e-05,
      "loss": 0.0038,
      "step": 3940
    },
    {
      "epoch": 1.94,
      "learning_rate": 3.5657182154216455e-05,
      "loss": 0.0016,
      "step": 3950
    },
    {
      "epoch": 1.94,
      "learning_rate": 3.561420097997077e-05,
      "loss": 0.017,
      "step": 3960
    },
    {
      "epoch": 1.95,
      "learning_rate": 3.557121980572509e-05,
      "loss": 0.0023,
      "step": 3970
    },
    {
      "epoch": 1.95,
      "learning_rate": 3.5528238631479414e-05,
      "loss": 0.0144,
      "step": 3980
    },
    {
      "epoch": 1.95,
      "learning_rate": 3.548525745723373e-05,
      "loss": 0.002,
      "step": 3990
    },
    {
      "epoch": 1.96,
      "learning_rate": 3.5442276282988055e-05,
      "loss": 0.017,
      "step": 4000
    },
    {
      "epoch": 1.96,
      "eval_f1": 0.9965576858796745,
      "eval_f1_0": 0.998999658653454,
      "eval_f1_1": 0.9986698509053197,
      "eval_f1_2": 0.9896010162249552,
      "eval_f1_3": 0.9989602177349696,
      "eval_loss": 0.006114189978688955,
      "eval_runtime": 40.0584,
      "eval_samples_per_second": 181.136,
      "eval_steps_per_second": 0.724,
      "step": 4000
    },
    {
      "epoch": 1.96,
      "learning_rate": 3.539929510874237e-05,
      "loss": 0.0021,
      "step": 4010
    },
    {
      "epoch": 1.97,
      "learning_rate": 3.5356313934496696e-05,
      "loss": 0.002,
      "step": 4020
    },
    {
      "epoch": 1.97,
      "learning_rate": 3.5313332760251014e-05,
      "loss": 0.0015,
      "step": 4030
    },
    {
      "epoch": 1.98,
      "learning_rate": 3.527035158600533e-05,
      "loss": 0.0026,
      "step": 4040
    },
    {
      "epoch": 1.98,
      "learning_rate": 3.5227370411759655e-05,
      "loss": 0.0045,
      "step": 4050
    },
    {
      "epoch": 1.99,
      "learning_rate": 3.518438923751397e-05,
      "loss": 0.001,
      "step": 4060
    },
    {
      "epoch": 1.99,
      "learning_rate": 3.514140806326829e-05,
      "loss": 0.0019,
      "step": 4070
    },
    {
      "epoch": 2.0,
      "learning_rate": 3.5098426889022614e-05,
      "loss": 0.0014,
      "step": 4080
    },
    {
      "epoch": 2.0,
      "learning_rate": 3.505544571477693e-05,
      "loss": 0.0014,
      "step": 4090
    },
    {
      "epoch": 2.01,
      "learning_rate": 3.501246454053125e-05,
      "loss": 0.0015,
      "step": 4100
    },
    {
      "epoch": 2.01,
      "learning_rate": 3.4969483366285566e-05,
      "loss": 0.0018,
      "step": 4110
    },
    {
      "epoch": 2.02,
      "learning_rate": 3.492650219203989e-05,
      "loss": 0.0014,
      "step": 4120
    },
    {
      "epoch": 2.02,
      "learning_rate": 3.488352101779421e-05,
      "loss": 0.0047,
      "step": 4130
    },
    {
      "epoch": 2.03,
      "learning_rate": 3.4840539843548524e-05,
      "loss": 0.0014,
      "step": 4140
    },
    {
      "epoch": 2.03,
      "learning_rate": 3.479755866930285e-05,
      "loss": 0.0025,
      "step": 4150
    },
    {
      "epoch": 2.04,
      "learning_rate": 3.4754577495057166e-05,
      "loss": 0.0035,
      "step": 4160
    },
    {
      "epoch": 2.04,
      "learning_rate": 3.471159632081148e-05,
      "loss": 0.0037,
      "step": 4170
    },
    {
      "epoch": 2.05,
      "learning_rate": 3.466861514656581e-05,
      "loss": 0.0023,
      "step": 4180
    },
    {
      "epoch": 2.05,
      "learning_rate": 3.4625633972320124e-05,
      "loss": 0.0018,
      "step": 4190
    },
    {
      "epoch": 2.06,
      "learning_rate": 3.458265279807444e-05,
      "loss": 0.0014,
      "step": 4200
    },
    {
      "epoch": 2.06,
      "eval_f1": 0.9951590569285889,
      "eval_f1_0": 0.9985791879302106,
      "eval_f1_1": 0.9986508472641124,
      "eval_f1_2": 0.9844199053492041,
      "eval_f1_3": 0.9989862871708285,
      "eval_loss": 0.00794305745512247,
      "eval_runtime": 40.1255,
      "eval_samples_per_second": 180.832,
      "eval_steps_per_second": 0.723,
      "step": 4200
    },
    {
      "epoch": 2.06,
      "learning_rate": 3.453967162382876e-05,
      "loss": 0.0019,
      "step": 4210
    },
    {
      "epoch": 2.07,
      "learning_rate": 3.449669044958308e-05,
      "loss": 0.0016,
      "step": 4220
    },
    {
      "epoch": 2.07,
      "learning_rate": 3.44537092753374e-05,
      "loss": 0.0017,
      "step": 4230
    },
    {
      "epoch": 2.08,
      "learning_rate": 3.4410728101091724e-05,
      "loss": 0.002,
      "step": 4240
    },
    {
      "epoch": 2.08,
      "learning_rate": 3.436774692684604e-05,
      "loss": 0.004,
      "step": 4250
    },
    {
      "epoch": 2.09,
      "learning_rate": 3.4324765752600366e-05,
      "loss": 0.0091,
      "step": 4260
    },
    {
      "epoch": 2.09,
      "learning_rate": 3.428178457835468e-05,
      "loss": 0.0016,
      "step": 4270
    },
    {
      "epoch": 2.1,
      "learning_rate": 3.423880340410901e-05,
      "loss": 0.0019,
      "step": 4280
    },
    {
      "epoch": 2.1,
      "learning_rate": 3.4195822229863324e-05,
      "loss": 0.0016,
      "step": 4290
    },
    {
      "epoch": 2.11,
      "learning_rate": 3.415284105561764e-05,
      "loss": 0.0085,
      "step": 4300
    },
    {
      "epoch": 2.11,
      "learning_rate": 3.410985988137196e-05,
      "loss": 0.011,
      "step": 4310
    },
    {
      "epoch": 2.12,
      "learning_rate": 3.406687870712628e-05,
      "loss": 0.0025,
      "step": 4320
    },
    {
      "epoch": 2.12,
      "learning_rate": 3.40238975328806e-05,
      "loss": 0.0032,
      "step": 4330
    },
    {
      "epoch": 2.13,
      "learning_rate": 3.398091635863492e-05,
      "loss": 0.0099,
      "step": 4340
    },
    {
      "epoch": 2.13,
      "learning_rate": 3.393793518438924e-05,
      "loss": 0.0016,
      "step": 4350
    },
    {
      "epoch": 2.14,
      "learning_rate": 3.389495401014356e-05,
      "loss": 0.0016,
      "step": 4360
    },
    {
      "epoch": 2.14,
      "learning_rate": 3.3851972835897876e-05,
      "loss": 0.0021,
      "step": 4370
    },
    {
      "epoch": 2.15,
      "learning_rate": 3.38089916616522e-05,
      "loss": 0.0022,
      "step": 4380
    },
    {
      "epoch": 2.15,
      "learning_rate": 3.376601048740652e-05,
      "loss": 0.0015,
      "step": 4390
    },
    {
      "epoch": 2.16,
      "learning_rate": 3.3723029313160835e-05,
      "loss": 0.0047,
      "step": 4400
    },
    {
      "epoch": 2.16,
      "eval_f1": 0.9964374764810087,
      "eval_f1_0": 0.9989611785189308,
      "eval_f1_1": 0.998629834590587,
      "eval_f1_2": 0.989156330768009,
      "eval_f1_3": 0.999002562046508,
      "eval_loss": 0.0061735427007079124,
      "eval_runtime": 40.046,
      "eval_samples_per_second": 181.192,
      "eval_steps_per_second": 0.724,
      "step": 4400
    },
    {
      "epoch": 2.16,
      "learning_rate": 3.368004813891515e-05,
      "loss": 0.0013,
      "step": 4410
    },
    {
      "epoch": 2.17,
      "learning_rate": 3.3637066964669476e-05,
      "loss": 0.003,
      "step": 4420
    },
    {
      "epoch": 2.17,
      "learning_rate": 3.3594085790423794e-05,
      "loss": 0.0011,
      "step": 4430
    },
    {
      "epoch": 2.18,
      "learning_rate": 3.355110461617811e-05,
      "loss": 0.0015,
      "step": 4440
    },
    {
      "epoch": 2.18,
      "learning_rate": 3.3508123441932435e-05,
      "loss": 0.0016,
      "step": 4450
    },
    {
      "epoch": 2.19,
      "learning_rate": 3.346514226768675e-05,
      "loss": 0.0026,
      "step": 4460
    },
    {
      "epoch": 2.19,
      "learning_rate": 3.3422161093441076e-05,
      "loss": 0.0017,
      "step": 4470
    },
    {
      "epoch": 2.2,
      "learning_rate": 3.3379179919195394e-05,
      "loss": 0.0031,
      "step": 4480
    },
    {
      "epoch": 2.2,
      "learning_rate": 3.333619874494972e-05,
      "loss": 0.0012,
      "step": 4490
    },
    {
      "epoch": 2.2,
      "learning_rate": 3.3293217570704035e-05,
      "loss": 0.0021,
      "step": 4500
    },
    {
      "epoch": 2.21,
      "learning_rate": 3.325023639645835e-05,
      "loss": 0.0078,
      "step": 4510
    },
    {
      "epoch": 2.21,
      "learning_rate": 3.3207255222212677e-05,
      "loss": 0.0018,
      "step": 4520
    },
    {
      "epoch": 2.22,
      "learning_rate": 3.3164274047966994e-05,
      "loss": 0.0027,
      "step": 4530
    },
    {
      "epoch": 2.22,
      "learning_rate": 3.312129287372131e-05,
      "loss": 0.0015,
      "step": 4540
    },
    {
      "epoch": 2.23,
      "learning_rate": 3.3078311699475635e-05,
      "loss": 0.0032,
      "step": 4550
    },
    {
      "epoch": 2.23,
      "learning_rate": 3.303533052522995e-05,
      "loss": 0.0017,
      "step": 4560
    },
    {
      "epoch": 2.24,
      "learning_rate": 3.299234935098427e-05,
      "loss": 0.0013,
      "step": 4570
    },
    {
      "epoch": 2.24,
      "learning_rate": 3.294936817673859e-05,
      "loss": 0.003,
      "step": 4580
    },
    {
      "epoch": 2.25,
      "learning_rate": 3.290638700249291e-05,
      "loss": 0.0047,
      "step": 4590
    },
    {
      "epoch": 2.25,
      "learning_rate": 3.286340582824723e-05,
      "loss": 0.0014,
      "step": 4600
    },
    {
      "epoch": 2.25,
      "eval_f1": 0.9968578679808767,
      "eval_f1_0": 0.9990990381463376,
      "eval_f1_1": 0.9989006139823642,
      "eval_f1_2": 0.9904259884253621,
      "eval_f1_3": 0.9990058313694429,
      "eval_loss": 0.006084946915507317,
      "eval_runtime": 40.1128,
      "eval_samples_per_second": 180.89,
      "eval_steps_per_second": 0.723,
      "step": 4600
    },
    {
      "epoch": 2.26,
      "learning_rate": 3.2820424654001546e-05,
      "loss": 0.0072,
      "step": 4610
    },
    {
      "epoch": 2.26,
      "learning_rate": 3.277744347975587e-05,
      "loss": 0.021,
      "step": 4620
    },
    {
      "epoch": 2.27,
      "learning_rate": 3.273446230551019e-05,
      "loss": 0.0022,
      "step": 4630
    },
    {
      "epoch": 2.27,
      "learning_rate": 3.2691481131264504e-05,
      "loss": 0.0053,
      "step": 4640
    },
    {
      "epoch": 2.28,
      "learning_rate": 3.264849995701883e-05,
      "loss": 0.0017,
      "step": 4650
    },
    {
      "epoch": 2.28,
      "learning_rate": 3.2605518782773146e-05,
      "loss": 0.0019,
      "step": 4660
    },
    {
      "epoch": 2.29,
      "learning_rate": 3.256253760852746e-05,
      "loss": 0.0039,
      "step": 4670
    },
    {
      "epoch": 2.29,
      "learning_rate": 3.251955643428179e-05,
      "loss": 0.0011,
      "step": 4680
    },
    {
      "epoch": 2.3,
      "learning_rate": 3.2476575260036104e-05,
      "loss": 0.0019,
      "step": 4690
    },
    {
      "epoch": 2.3,
      "learning_rate": 3.243359408579043e-05,
      "loss": 0.0009,
      "step": 4700
    },
    {
      "epoch": 2.31,
      "learning_rate": 3.2390612911544746e-05,
      "loss": 0.0029,
      "step": 4710
    },
    {
      "epoch": 2.31,
      "learning_rate": 3.234763173729907e-05,
      "loss": 0.0013,
      "step": 4720
    },
    {
      "epoch": 2.32,
      "learning_rate": 3.230465056305339e-05,
      "loss": 0.0016,
      "step": 4730
    },
    {
      "epoch": 2.32,
      "learning_rate": 3.2261669388807704e-05,
      "loss": 0.0028,
      "step": 4740
    },
    {
      "epoch": 2.33,
      "learning_rate": 3.221868821456203e-05,
      "loss": 0.0014,
      "step": 4750
    },
    {
      "epoch": 2.33,
      "learning_rate": 3.2175707040316346e-05,
      "loss": 0.0012,
      "step": 4760
    },
    {
      "epoch": 2.34,
      "learning_rate": 3.213272586607066e-05,
      "loss": 0.0013,
      "step": 4770
    },
    {
      "epoch": 2.34,
      "learning_rate": 3.208974469182498e-05,
      "loss": 0.0013,
      "step": 4780
    },
    {
      "epoch": 2.35,
      "learning_rate": 3.2046763517579305e-05,
      "loss": 0.0014,
      "step": 4790
    },
    {
      "epoch": 2.35,
      "learning_rate": 3.200378234333362e-05,
      "loss": 0.0014,
      "step": 4800
    },
    {
      "epoch": 2.35,
      "eval_f1": 0.9971469633109779,
      "eval_f1_0": 0.9991664057411811,
      "eval_f1_1": 0.9986450895915477,
      "eval_f1_2": 0.9917672443634182,
      "eval_f1_3": 0.9990091135477647,
      "eval_loss": 0.005138964857906103,
      "eval_runtime": 40.0699,
      "eval_samples_per_second": 181.084,
      "eval_steps_per_second": 0.724,
      "step": 4800
    },
    {
      "epoch": 2.36,
      "learning_rate": 3.196080116908794e-05,
      "loss": 0.0022,
      "step": 4810
    },
    {
      "epoch": 2.36,
      "learning_rate": 3.191781999484226e-05,
      "loss": 0.0015,
      "step": 4820
    },
    {
      "epoch": 2.37,
      "learning_rate": 3.187483882059658e-05,
      "loss": 0.0018,
      "step": 4830
    },
    {
      "epoch": 2.37,
      "learning_rate": 3.18318576463509e-05,
      "loss": 0.035,
      "step": 4840
    },
    {
      "epoch": 2.38,
      "learning_rate": 3.1788876472105215e-05,
      "loss": 0.0018,
      "step": 4850
    },
    {
      "epoch": 2.38,
      "learning_rate": 3.174589529785954e-05,
      "loss": 0.0017,
      "step": 4860
    },
    {
      "epoch": 2.39,
      "learning_rate": 3.1702914123613856e-05,
      "loss": 0.0036,
      "step": 4870
    },
    {
      "epoch": 2.39,
      "learning_rate": 3.1659932949368174e-05,
      "loss": 0.0022,
      "step": 4880
    },
    {
      "epoch": 2.4,
      "learning_rate": 3.16169517751225e-05,
      "loss": 0.0027,
      "step": 4890
    },
    {
      "epoch": 2.4,
      "learning_rate": 3.1573970600876815e-05,
      "loss": 0.0086,
      "step": 4900
    },
    {
      "epoch": 2.41,
      "learning_rate": 3.153098942663113e-05,
      "loss": 0.002,
      "step": 4910
    },
    {
      "epoch": 2.41,
      "learning_rate": 3.1488008252385457e-05,
      "loss": 0.0013,
      "step": 4920
    },
    {
      "epoch": 2.42,
      "learning_rate": 3.1445027078139774e-05,
      "loss": 0.0016,
      "step": 4930
    },
    {
      "epoch": 2.42,
      "learning_rate": 3.14020459038941e-05,
      "loss": 0.0023,
      "step": 4940
    },
    {
      "epoch": 2.43,
      "learning_rate": 3.1359064729648415e-05,
      "loss": 0.0054,
      "step": 4950
    },
    {
      "epoch": 2.43,
      "learning_rate": 3.131608355540274e-05,
      "loss": 0.0143,
      "step": 4960
    },
    {
      "epoch": 2.44,
      "learning_rate": 3.1273102381157057e-05,
      "loss": 0.0011,
      "step": 4970
    },
    {
      "epoch": 2.44,
      "learning_rate": 3.1230121206911374e-05,
      "loss": 0.0024,
      "step": 4980
    },
    {
      "epoch": 2.44,
      "learning_rate": 3.11871400326657e-05,
      "loss": 0.0023,
      "step": 4990
    },
    {
      "epoch": 2.45,
      "learning_rate": 3.1144158858420015e-05,
      "loss": 0.0018,
      "step": 5000
    },
    {
      "epoch": 2.45,
      "eval_f1": 0.9970405942808762,
      "eval_f1_0": 0.9991421792838462,
      "eval_f1_1": 0.9987689907436476,
      "eval_f1_2": 0.9912486190401824,
      "eval_f1_3": 0.999002588055829,
      "eval_loss": 0.005121254362165928,
      "eval_runtime": 40.1102,
      "eval_samples_per_second": 180.901,
      "eval_steps_per_second": 0.723,
      "step": 5000
    },
    {
      "epoch": 2.45,
      "learning_rate": 3.110117768417433e-05,
      "loss": 0.0035,
      "step": 5010
    },
    {
      "epoch": 2.46,
      "learning_rate": 3.1058196509928657e-05,
      "loss": 0.0014,
      "step": 5020
    },
    {
      "epoch": 2.46,
      "learning_rate": 3.1015215335682974e-05,
      "loss": 0.0016,
      "step": 5030
    },
    {
      "epoch": 2.47,
      "learning_rate": 3.097223416143729e-05,
      "loss": 0.0037,
      "step": 5040
    },
    {
      "epoch": 2.47,
      "learning_rate": 3.092925298719161e-05,
      "loss": 0.0008,
      "step": 5050
    },
    {
      "epoch": 2.48,
      "learning_rate": 3.088627181294593e-05,
      "loss": 0.002,
      "step": 5060
    },
    {
      "epoch": 2.48,
      "learning_rate": 3.084329063870025e-05,
      "loss": 0.0013,
      "step": 5070
    },
    {
      "epoch": 2.49,
      "learning_rate": 3.080030946445457e-05,
      "loss": 0.0161,
      "step": 5080
    },
    {
      "epoch": 2.49,
      "learning_rate": 3.075732829020889e-05,
      "loss": 0.0015,
      "step": 5090
    },
    {
      "epoch": 2.5,
      "learning_rate": 3.071434711596321e-05,
      "loss": 0.0014,
      "step": 5100
    },
    {
      "epoch": 2.5,
      "learning_rate": 3.0671365941717526e-05,
      "loss": 0.002,
      "step": 5110
    },
    {
      "epoch": 2.51,
      "learning_rate": 3.062838476747184e-05,
      "loss": 0.0013,
      "step": 5120
    },
    {
      "epoch": 2.51,
      "learning_rate": 3.058540359322617e-05,
      "loss": 0.0013,
      "step": 5130
    },
    {
      "epoch": 2.52,
      "learning_rate": 3.0542422418980484e-05,
      "loss": 0.0063,
      "step": 5140
    },
    {
      "epoch": 2.52,
      "learning_rate": 3.0499441244734805e-05,
      "loss": 0.0016,
      "step": 5150
    },
    {
      "epoch": 2.53,
      "learning_rate": 3.045646007048913e-05,
      "loss": 0.001,
      "step": 5160
    },
    {
      "epoch": 2.53,
      "learning_rate": 3.0413478896243447e-05,
      "loss": 0.001,
      "step": 5170
    },
    {
      "epoch": 2.54,
      "learning_rate": 3.0370497721997764e-05,
      "loss": 0.002,
      "step": 5180
    },
    {
      "epoch": 2.54,
      "learning_rate": 3.0327516547752088e-05,
      "loss": 0.0018,
      "step": 5190
    },
    {
      "epoch": 2.55,
      "learning_rate": 3.0284535373506405e-05,
      "loss": 0.0022,
      "step": 5200
    },
    {
      "epoch": 2.55,
      "eval_f1": 0.9967894168407214,
      "eval_f1_0": 0.999066751961815,
      "eval_f1_1": 0.9987746590355577,
      "eval_f1_2": 0.9903429269867626,
      "eval_f1_3": 0.9989733293787502,
      "eval_loss": 0.005486695561558008,
      "eval_runtime": 40.165,
      "eval_samples_per_second": 180.655,
      "eval_steps_per_second": 0.722,
      "step": 5200
    },
    {
      "epoch": 2.55,
      "learning_rate": 3.0241554199260726e-05,
      "loss": 0.0017,
      "step": 5210
    },
    {
      "epoch": 2.56,
      "learning_rate": 3.0198573025015043e-05,
      "loss": 0.0011,
      "step": 5220
    },
    {
      "epoch": 2.56,
      "learning_rate": 3.0155591850769367e-05,
      "loss": 0.0018,
      "step": 5230
    },
    {
      "epoch": 2.57,
      "learning_rate": 3.0112610676523685e-05,
      "loss": 0.0019,
      "step": 5240
    },
    {
      "epoch": 2.57,
      "learning_rate": 3.0069629502278002e-05,
      "loss": 0.0036,
      "step": 5250
    },
    {
      "epoch": 2.58,
      "learning_rate": 3.0026648328032326e-05,
      "loss": 0.0145,
      "step": 5260
    },
    {
      "epoch": 2.58,
      "learning_rate": 2.9983667153786643e-05,
      "loss": 0.0024,
      "step": 5270
    },
    {
      "epoch": 2.59,
      "learning_rate": 2.994068597954096e-05,
      "loss": 0.0054,
      "step": 5280
    },
    {
      "epoch": 2.59,
      "learning_rate": 2.9897704805295285e-05,
      "loss": 0.006,
      "step": 5290
    },
    {
      "epoch": 2.6,
      "learning_rate": 2.9854723631049602e-05,
      "loss": 0.0032,
      "step": 5300
    },
    {
      "epoch": 2.6,
      "learning_rate": 2.981174245680392e-05,
      "loss": 0.0019,
      "step": 5310
    },
    {
      "epoch": 2.61,
      "learning_rate": 2.976876128255824e-05,
      "loss": 0.0017,
      "step": 5320
    },
    {
      "epoch": 2.61,
      "learning_rate": 2.972578010831256e-05,
      "loss": 0.0038,
      "step": 5330
    },
    {
      "epoch": 2.62,
      "learning_rate": 2.968279893406688e-05,
      "loss": 0.0251,
      "step": 5340
    },
    {
      "epoch": 2.62,
      "learning_rate": 2.96398177598212e-05,
      "loss": 0.0037,
      "step": 5350
    },
    {
      "epoch": 2.63,
      "learning_rate": 2.9596836585575523e-05,
      "loss": 0.0031,
      "step": 5360
    },
    {
      "epoch": 2.63,
      "learning_rate": 2.955385541132984e-05,
      "loss": 0.0018,
      "step": 5370
    },
    {
      "epoch": 2.64,
      "learning_rate": 2.9510874237084157e-05,
      "loss": 0.0022,
      "step": 5380
    },
    {
      "epoch": 2.64,
      "learning_rate": 2.946789306283848e-05,
      "loss": 0.0013,
      "step": 5390
    },
    {
      "epoch": 2.65,
      "learning_rate": 2.94249118885928e-05,
      "loss": 0.0017,
      "step": 5400
    },
    {
      "epoch": 2.65,
      "eval_f1": 0.9972341936836928,
      "eval_f1_0": 0.9992021980230644,
      "eval_f1_1": 0.9988643691542118,
      "eval_f1_2": 0.9918904327893273,
      "eval_f1_3": 0.9989797747681677,
      "eval_loss": 0.004784774500876665,
      "eval_runtime": 40.0886,
      "eval_samples_per_second": 180.999,
      "eval_steps_per_second": 0.723,
      "step": 5400
    },
    {
      "epoch": 2.65,
      "learning_rate": 2.9381930714347116e-05,
      "loss": 0.0043,
      "step": 5410
    },
    {
      "epoch": 2.66,
      "learning_rate": 2.9338949540101433e-05,
      "loss": 0.0019,
      "step": 5420
    },
    {
      "epoch": 2.66,
      "learning_rate": 2.9295968365855757e-05,
      "loss": 0.0024,
      "step": 5430
    },
    {
      "epoch": 2.67,
      "learning_rate": 2.9252987191610075e-05,
      "loss": 0.0026,
      "step": 5440
    },
    {
      "epoch": 2.67,
      "learning_rate": 2.9210006017364395e-05,
      "loss": 0.0018,
      "step": 5450
    },
    {
      "epoch": 2.68,
      "learning_rate": 2.916702484311872e-05,
      "loss": 0.002,
      "step": 5460
    },
    {
      "epoch": 2.68,
      "learning_rate": 2.9124043668873037e-05,
      "loss": 0.0018,
      "step": 5470
    },
    {
      "epoch": 2.68,
      "learning_rate": 2.9081062494627354e-05,
      "loss": 0.0019,
      "step": 5480
    },
    {
      "epoch": 2.69,
      "learning_rate": 2.903808132038167e-05,
      "loss": 0.0051,
      "step": 5490
    },
    {
      "epoch": 2.69,
      "learning_rate": 2.8995100146135995e-05,
      "loss": 0.0028,
      "step": 5500
    },
    {
      "epoch": 2.7,
      "learning_rate": 2.8952118971890313e-05,
      "loss": 0.002,
      "step": 5510
    },
    {
      "epoch": 2.7,
      "learning_rate": 2.890913779764463e-05,
      "loss": 0.0035,
      "step": 5520
    },
    {
      "epoch": 2.71,
      "learning_rate": 2.8866156623398954e-05,
      "loss": 0.0027,
      "step": 5530
    },
    {
      "epoch": 2.71,
      "learning_rate": 2.882317544915327e-05,
      "loss": 0.0048,
      "step": 5540
    },
    {
      "epoch": 2.72,
      "learning_rate": 2.8780194274907592e-05,
      "loss": 0.0031,
      "step": 5550
    },
    {
      "epoch": 2.72,
      "learning_rate": 2.8737213100661913e-05,
      "loss": 0.0024,
      "step": 5560
    },
    {
      "epoch": 2.73,
      "learning_rate": 2.8694231926416233e-05,
      "loss": 0.0017,
      "step": 5570
    },
    {
      "epoch": 2.73,
      "learning_rate": 2.865125075217055e-05,
      "loss": 0.0028,
      "step": 5580
    },
    {
      "epoch": 2.74,
      "learning_rate": 2.8608269577924868e-05,
      "loss": 0.002,
      "step": 5590
    },
    {
      "epoch": 2.74,
      "learning_rate": 2.8565288403679192e-05,
      "loss": 0.0075,
      "step": 5600
    },
    {
      "epoch": 2.74,
      "eval_f1": 0.9964894957991478,
      "eval_f1_0": 0.9989803708861019,
      "eval_f1_1": 0.998828032845521,
      "eval_f1_2": 0.9891763973400745,
      "eval_f1_3": 0.9989731821248936,
      "eval_loss": 0.005460510030388832,
      "eval_runtime": 40.01,
      "eval_samples_per_second": 181.355,
      "eval_steps_per_second": 0.725,
      "step": 5600
    },
    {
      "epoch": 2.75,
      "learning_rate": 2.852230722943351e-05,
      "loss": 0.0028,
      "step": 5610
    },
    {
      "epoch": 2.75,
      "learning_rate": 2.8479326055187827e-05,
      "loss": 0.0027,
      "step": 5620
    },
    {
      "epoch": 2.76,
      "learning_rate": 2.843634488094215e-05,
      "loss": 0.0012,
      "step": 5630
    },
    {
      "epoch": 2.76,
      "learning_rate": 2.8393363706696468e-05,
      "loss": 0.0029,
      "step": 5640
    },
    {
      "epoch": 2.77,
      "learning_rate": 2.8350382532450785e-05,
      "loss": 0.0025,
      "step": 5650
    },
    {
      "epoch": 2.77,
      "learning_rate": 2.830740135820511e-05,
      "loss": 0.0018,
      "step": 5660
    },
    {
      "epoch": 2.78,
      "learning_rate": 2.8264420183959427e-05,
      "loss": 0.0012,
      "step": 5670
    },
    {
      "epoch": 2.78,
      "learning_rate": 2.8221439009713747e-05,
      "loss": 0.001,
      "step": 5680
    },
    {
      "epoch": 2.79,
      "learning_rate": 2.8178457835468065e-05,
      "loss": 0.0022,
      "step": 5690
    },
    {
      "epoch": 2.79,
      "learning_rate": 2.813547666122239e-05,
      "loss": 0.0013,
      "step": 5700
    },
    {
      "epoch": 2.8,
      "learning_rate": 2.8092495486976706e-05,
      "loss": 0.0022,
      "step": 5710
    },
    {
      "epoch": 2.8,
      "learning_rate": 2.8049514312731023e-05,
      "loss": 0.0025,
      "step": 5720
    },
    {
      "epoch": 2.81,
      "learning_rate": 2.8006533138485347e-05,
      "loss": 0.001,
      "step": 5730
    },
    {
      "epoch": 2.81,
      "learning_rate": 2.7963551964239665e-05,
      "loss": 0.0016,
      "step": 5740
    },
    {
      "epoch": 2.82,
      "learning_rate": 2.7920570789993982e-05,
      "loss": 0.0014,
      "step": 5750
    },
    {
      "epoch": 2.82,
      "learning_rate": 2.7877589615748306e-05,
      "loss": 0.0016,
      "step": 5760
    },
    {
      "epoch": 2.83,
      "learning_rate": 2.7834608441502623e-05,
      "loss": 0.0013,
      "step": 5770
    },
    {
      "epoch": 2.83,
      "learning_rate": 2.779162726725694e-05,
      "loss": 0.0018,
      "step": 5780
    },
    {
      "epoch": 2.84,
      "learning_rate": 2.774864609301126e-05,
      "loss": 0.004,
      "step": 5790
    },
    {
      "epoch": 2.84,
      "learning_rate": 2.7705664918765585e-05,
      "loss": 0.0011,
      "step": 5800
    },
    {
      "epoch": 2.84,
      "eval_f1": 0.9970638662701521,
      "eval_f1_0": 0.9991495242166827,
      "eval_f1_1": 0.9988298975446874,
      "eval_f1_2": 0.9912767049945479,
      "eval_f1_3": 0.9989993383246902,
      "eval_loss": 0.0054163821041584015,
      "eval_runtime": 40.0008,
      "eval_samples_per_second": 181.396,
      "eval_steps_per_second": 0.725,
      "step": 5800
    },
    {
      "epoch": 2.85,
      "learning_rate": 2.7662683744519903e-05,
      "loss": 0.0136,
      "step": 5810
    },
    {
      "epoch": 2.85,
      "learning_rate": 2.761970257027422e-05,
      "loss": 0.001,
      "step": 5820
    },
    {
      "epoch": 2.86,
      "learning_rate": 2.7576721396028544e-05,
      "loss": 0.002,
      "step": 5830
    },
    {
      "epoch": 2.86,
      "learning_rate": 2.753374022178286e-05,
      "loss": 0.0019,
      "step": 5840
    },
    {
      "epoch": 2.87,
      "learning_rate": 2.749075904753718e-05,
      "loss": 0.0017,
      "step": 5850
    },
    {
      "epoch": 2.87,
      "learning_rate": 2.7447777873291496e-05,
      "loss": 0.0051,
      "step": 5860
    },
    {
      "epoch": 2.88,
      "learning_rate": 2.740479669904582e-05,
      "loss": 0.0017,
      "step": 5870
    },
    {
      "epoch": 2.88,
      "learning_rate": 2.7361815524800137e-05,
      "loss": 0.0144,
      "step": 5880
    },
    {
      "epoch": 2.89,
      "learning_rate": 2.7318834350554458e-05,
      "loss": 0.001,
      "step": 5890
    },
    {
      "epoch": 2.89,
      "learning_rate": 2.727585317630878e-05,
      "loss": 0.0076,
      "step": 5900
    },
    {
      "epoch": 2.9,
      "learning_rate": 2.72328720020631e-05,
      "loss": 0.0011,
      "step": 5910
    },
    {
      "epoch": 2.9,
      "learning_rate": 2.7189890827817417e-05,
      "loss": 0.0055,
      "step": 5920
    },
    {
      "epoch": 2.91,
      "learning_rate": 2.714690965357174e-05,
      "loss": 0.0022,
      "step": 5930
    },
    {
      "epoch": 2.91,
      "learning_rate": 2.7103928479326058e-05,
      "loss": 0.0024,
      "step": 5940
    },
    {
      "epoch": 2.92,
      "learning_rate": 2.7060947305080375e-05,
      "loss": 0.0029,
      "step": 5950
    },
    {
      "epoch": 2.92,
      "learning_rate": 2.7017966130834693e-05,
      "loss": 0.0016,
      "step": 5960
    },
    {
      "epoch": 2.93,
      "learning_rate": 2.6974984956589017e-05,
      "loss": 0.0026,
      "step": 5970
    },
    {
      "epoch": 2.93,
      "learning_rate": 2.6932003782343334e-05,
      "loss": 0.0012,
      "step": 5980
    },
    {
      "epoch": 2.93,
      "learning_rate": 2.688902260809765e-05,
      "loss": 0.002,
      "step": 5990
    },
    {
      "epoch": 2.94,
      "learning_rate": 2.6846041433851975e-05,
      "loss": 0.0018,
      "step": 6000
    },
    {
      "epoch": 2.94,
      "eval_f1": 0.9971003887455068,
      "eval_f1_0": 0.9991635902234611,
      "eval_f1_1": 0.9988833702055778,
      "eval_f1_2": 0.9913389106814317,
      "eval_f1_3": 0.9990156838715565,
      "eval_loss": 0.00558880902826786,
      "eval_runtime": 40.1078,
      "eval_samples_per_second": 180.912,
      "eval_steps_per_second": 0.723,
      "step": 6000
    },
    {
      "epoch": 2.94,
      "learning_rate": 2.6803060259606293e-05,
      "loss": 0.0015,
      "step": 6010
    },
    {
      "epoch": 2.95,
      "learning_rate": 2.6760079085360613e-05,
      "loss": 0.001,
      "step": 6020
    },
    {
      "epoch": 2.95,
      "learning_rate": 2.6717097911114934e-05,
      "loss": 0.0018,
      "step": 6030
    },
    {
      "epoch": 2.96,
      "learning_rate": 2.6674116736869255e-05,
      "loss": 0.0015,
      "step": 6040
    },
    {
      "epoch": 2.96,
      "learning_rate": 2.6631135562623572e-05,
      "loss": 0.0014,
      "step": 6050
    },
    {
      "epoch": 2.97,
      "learning_rate": 2.658815438837789e-05,
      "loss": 0.0019,
      "step": 6060
    },
    {
      "epoch": 2.97,
      "learning_rate": 2.6545173214132213e-05,
      "loss": 0.0015,
      "step": 6070
    },
    {
      "epoch": 2.98,
      "learning_rate": 2.650219203988653e-05,
      "loss": 0.001,
      "step": 6080
    },
    {
      "epoch": 2.98,
      "learning_rate": 2.6459210865640848e-05,
      "loss": 0.0009,
      "step": 6090
    },
    {
      "epoch": 2.99,
      "learning_rate": 2.6416229691395172e-05,
      "loss": 0.0021,
      "step": 6100
    },
    {
      "epoch": 2.99,
      "learning_rate": 2.637324851714949e-05,
      "loss": 0.0035,
      "step": 6110
    },
    {
      "epoch": 3.0,
      "learning_rate": 2.6330267342903807e-05,
      "loss": 0.0126,
      "step": 6120
    },
    {
      "epoch": 3.0,
      "learning_rate": 2.628728616865813e-05,
      "loss": 0.002,
      "step": 6130
    },
    {
      "epoch": 3.01,
      "learning_rate": 2.624430499441245e-05,
      "loss": 0.0015,
      "step": 6140
    },
    {
      "epoch": 3.01,
      "learning_rate": 2.620132382016677e-05,
      "loss": 0.0014,
      "step": 6150
    },
    {
      "epoch": 3.02,
      "learning_rate": 2.6158342645921086e-05,
      "loss": 0.0015,
      "step": 6160
    },
    {
      "epoch": 3.02,
      "learning_rate": 2.611536147167541e-05,
      "loss": 0.0016,
      "step": 6170
    },
    {
      "epoch": 3.03,
      "learning_rate": 2.6072380297429727e-05,
      "loss": 0.0013,
      "step": 6180
    },
    {
      "epoch": 3.03,
      "learning_rate": 2.6029399123184045e-05,
      "loss": 0.0016,
      "step": 6190
    },
    {
      "epoch": 3.04,
      "learning_rate": 2.598641794893837e-05,
      "loss": 0.0012,
      "step": 6200
    },
    {
      "epoch": 3.04,
      "eval_f1": 0.9966069663708587,
      "eval_f1_0": 0.999013456777716,
      "eval_f1_1": 0.9987879894895666,
      "eval_f1_2": 0.9896271004616907,
      "eval_f1_3": 0.9989993187544615,
      "eval_loss": 0.0057078259997069836,
      "eval_runtime": 40.1388,
      "eval_samples_per_second": 180.773,
      "eval_steps_per_second": 0.722,
      "step": 6200
    },
    {
      "epoch": 3.04,
      "learning_rate": 2.5943436774692686e-05,
      "loss": 0.0016,
      "step": 6210
    },
    {
      "epoch": 3.05,
      "learning_rate": 2.5900455600447003e-05,
      "loss": 0.0027,
      "step": 6220
    },
    {
      "epoch": 3.05,
      "learning_rate": 2.5857474426201324e-05,
      "loss": 0.0012,
      "step": 6230
    },
    {
      "epoch": 3.06,
      "learning_rate": 2.5814493251955645e-05,
      "loss": 0.001,
      "step": 6240
    },
    {
      "epoch": 3.06,
      "learning_rate": 2.5771512077709965e-05,
      "loss": 0.0013,
      "step": 6250
    },
    {
      "epoch": 3.07,
      "learning_rate": 2.5728530903464283e-05,
      "loss": 0.0013,
      "step": 6260
    },
    {
      "epoch": 3.07,
      "learning_rate": 2.5685549729218607e-05,
      "loss": 0.0056,
      "step": 6270
    },
    {
      "epoch": 3.08,
      "learning_rate": 2.5642568554972924e-05,
      "loss": 0.0014,
      "step": 6280
    },
    {
      "epoch": 3.08,
      "learning_rate": 2.559958738072724e-05,
      "loss": 0.0022,
      "step": 6290
    },
    {
      "epoch": 3.09,
      "learning_rate": 2.5556606206481565e-05,
      "loss": 0.0017,
      "step": 6300
    },
    {
      "epoch": 3.09,
      "learning_rate": 2.5513625032235883e-05,
      "loss": 0.0016,
      "step": 6310
    },
    {
      "epoch": 3.1,
      "learning_rate": 2.54706438579902e-05,
      "loss": 0.0015,
      "step": 6320
    },
    {
      "epoch": 3.1,
      "learning_rate": 2.5427662683744517e-05,
      "loss": 0.0013,
      "step": 6330
    },
    {
      "epoch": 3.11,
      "learning_rate": 2.538468150949884e-05,
      "loss": 0.0019,
      "step": 6340
    },
    {
      "epoch": 3.11,
      "learning_rate": 2.534170033525316e-05,
      "loss": 0.0008,
      "step": 6350
    },
    {
      "epoch": 3.12,
      "learning_rate": 2.529871916100748e-05,
      "loss": 0.0023,
      "step": 6360
    },
    {
      "epoch": 3.12,
      "learning_rate": 2.52557379867618e-05,
      "loss": 0.0014,
      "step": 6370
    },
    {
      "epoch": 3.13,
      "learning_rate": 2.521275681251612e-05,
      "loss": 0.0015,
      "step": 6380
    },
    {
      "epoch": 3.13,
      "learning_rate": 2.5169775638270438e-05,
      "loss": 0.0039,
      "step": 6390
    },
    {
      "epoch": 3.14,
      "learning_rate": 2.5126794464024762e-05,
      "loss": 0.0011,
      "step": 6400
    },
    {
      "epoch": 3.14,
      "eval_f1": 0.9969391892523141,
      "eval_f1_0": 0.9991163839459859,
      "eval_f1_1": 0.9988470889948367,
      "eval_f1_2": 0.9907939196512963,
      "eval_f1_3": 0.9989993644171379,
      "eval_loss": 0.005770825315266848,
      "eval_runtime": 40.007,
      "eval_samples_per_second": 181.368,
      "eval_steps_per_second": 0.725,
      "step": 6400
    },
    {
      "epoch": 3.14,
      "learning_rate": 2.508381328977908e-05,
      "loss": 0.0015,
      "step": 6410
    },
    {
      "epoch": 3.15,
      "learning_rate": 2.5040832115533397e-05,
      "loss": 0.0012,
      "step": 6420
    },
    {
      "epoch": 3.15,
      "learning_rate": 2.4997850941287717e-05,
      "loss": 0.0017,
      "step": 6430
    },
    {
      "epoch": 3.16,
      "learning_rate": 2.4954869767042035e-05,
      "loss": 0.0014,
      "step": 6440
    },
    {
      "epoch": 3.16,
      "learning_rate": 2.4911888592796355e-05,
      "loss": 0.0016,
      "step": 6450
    },
    {
      "epoch": 3.17,
      "learning_rate": 2.4868907418550676e-05,
      "loss": 0.0024,
      "step": 6460
    },
    {
      "epoch": 3.17,
      "learning_rate": 2.4825926244304993e-05,
      "loss": 0.0015,
      "step": 6470
    },
    {
      "epoch": 3.17,
      "learning_rate": 2.4782945070059317e-05,
      "loss": 0.0014,
      "step": 6480
    },
    {
      "epoch": 3.18,
      "learning_rate": 2.4739963895813635e-05,
      "loss": 0.0009,
      "step": 6490
    },
    {
      "epoch": 3.18,
      "learning_rate": 2.4696982721567955e-05,
      "loss": 0.0015,
      "step": 6500
    },
    {
      "epoch": 3.19,
      "learning_rate": 2.4654001547322276e-05,
      "loss": 0.0015,
      "step": 6510
    },
    {
      "epoch": 3.19,
      "learning_rate": 2.4611020373076593e-05,
      "loss": 0.0017,
      "step": 6520
    },
    {
      "epoch": 3.2,
      "learning_rate": 2.4568039198830914e-05,
      "loss": 0.0018,
      "step": 6530
    },
    {
      "epoch": 3.2,
      "learning_rate": 2.452505802458523e-05,
      "loss": 0.0013,
      "step": 6540
    },
    {
      "epoch": 3.21,
      "learning_rate": 2.4482076850339552e-05,
      "loss": 0.0011,
      "step": 6550
    },
    {
      "epoch": 3.21,
      "learning_rate": 2.4439095676093873e-05,
      "loss": 0.001,
      "step": 6560
    },
    {
      "epoch": 3.22,
      "learning_rate": 2.439611450184819e-05,
      "loss": 0.0017,
      "step": 6570
    },
    {
      "epoch": 3.22,
      "learning_rate": 2.435313332760251e-05,
      "loss": 0.0163,
      "step": 6580
    },
    {
      "epoch": 3.23,
      "learning_rate": 2.431015215335683e-05,
      "loss": 0.0015,
      "step": 6590
    },
    {
      "epoch": 3.23,
      "learning_rate": 2.4267170979111152e-05,
      "loss": 0.0023,
      "step": 6600
    },
    {
      "epoch": 3.23,
      "eval_f1": 0.9968810895890741,
      "eval_f1_0": 0.999099568740818,
      "eval_f1_1": 0.9988490506094324,
      "eval_f1_2": 0.990573105437169,
      "eval_f1_3": 0.999002633568877,
      "eval_loss": 0.005641479045152664,
      "eval_runtime": 39.9844,
      "eval_samples_per_second": 181.471,
      "eval_steps_per_second": 0.725,
      "step": 6600
    },
    {
      "epoch": 3.24,
      "learning_rate": 2.4224189804865473e-05,
      "loss": 0.0012,
      "step": 6610
    },
    {
      "epoch": 3.24,
      "learning_rate": 2.418120863061979e-05,
      "loss": 0.002,
      "step": 6620
    },
    {
      "epoch": 3.25,
      "learning_rate": 2.413822745637411e-05,
      "loss": 0.0013,
      "step": 6630
    },
    {
      "epoch": 3.25,
      "learning_rate": 2.4095246282128428e-05,
      "loss": 0.0028,
      "step": 6640
    },
    {
      "epoch": 3.26,
      "learning_rate": 2.405226510788275e-05,
      "loss": 0.0012,
      "step": 6650
    },
    {
      "epoch": 3.26,
      "learning_rate": 2.4009283933637066e-05,
      "loss": 0.0017,
      "step": 6660
    },
    {
      "epoch": 3.27,
      "learning_rate": 2.3966302759391387e-05,
      "loss": 0.0047,
      "step": 6670
    },
    {
      "epoch": 3.27,
      "learning_rate": 2.3923321585145707e-05,
      "loss": 0.019,
      "step": 6680
    },
    {
      "epoch": 3.28,
      "learning_rate": 2.3880340410900025e-05,
      "loss": 0.0013,
      "step": 6690
    },
    {
      "epoch": 3.28,
      "learning_rate": 2.3837359236654345e-05,
      "loss": 0.0019,
      "step": 6700
    },
    {
      "epoch": 3.29,
      "learning_rate": 2.3794378062408666e-05,
      "loss": 0.0017,
      "step": 6710
    },
    {
      "epoch": 3.29,
      "learning_rate": 2.3751396888162987e-05,
      "loss": 0.001,
      "step": 6720
    },
    {
      "epoch": 3.3,
      "learning_rate": 2.3708415713917307e-05,
      "loss": 0.0011,
      "step": 6730
    },
    {
      "epoch": 3.3,
      "learning_rate": 2.3665434539671625e-05,
      "loss": 0.0012,
      "step": 6740
    },
    {
      "epoch": 3.31,
      "learning_rate": 2.3622453365425945e-05,
      "loss": 0.0105,
      "step": 6750
    },
    {
      "epoch": 3.31,
      "learning_rate": 2.3579472191180263e-05,
      "loss": 0.0011,
      "step": 6760
    },
    {
      "epoch": 3.32,
      "learning_rate": 2.3536491016934583e-05,
      "loss": 0.0047,
      "step": 6770
    },
    {
      "epoch": 3.32,
      "learning_rate": 2.3493509842688904e-05,
      "loss": 0.0014,
      "step": 6780
    },
    {
      "epoch": 3.33,
      "learning_rate": 2.345052866844322e-05,
      "loss": 0.0016,
      "step": 6790
    },
    {
      "epoch": 3.33,
      "learning_rate": 2.3407547494197542e-05,
      "loss": 0.0011,
      "step": 6800
    },
    {
      "epoch": 3.33,
      "eval_f1": 0.9966155469863469,
      "eval_f1_0": 0.9990164948514703,
      "eval_f1_1": 0.9988547033263068,
      "eval_f1_2": 0.9895851324743514,
      "eval_f1_3": 0.9990058572932591,
      "eval_loss": 0.005889880005270243,
      "eval_runtime": 40.0587,
      "eval_samples_per_second": 181.134,
      "eval_steps_per_second": 0.724,
      "step": 6800
    },
    {
      "epoch": 3.34,
      "learning_rate": 2.336456631995186e-05,
      "loss": 0.0011,
      "step": 6810
    },
    {
      "epoch": 3.34,
      "learning_rate": 2.3321585145706183e-05,
      "loss": 0.0016,
      "step": 6820
    },
    {
      "epoch": 3.35,
      "learning_rate": 2.3278603971460504e-05,
      "loss": 0.0205,
      "step": 6830
    },
    {
      "epoch": 3.35,
      "learning_rate": 2.323562279721482e-05,
      "loss": 0.0023,
      "step": 6840
    },
    {
      "epoch": 3.36,
      "learning_rate": 2.3192641622969142e-05,
      "loss": 0.0023,
      "step": 6850
    },
    {
      "epoch": 3.36,
      "learning_rate": 2.314966044872346e-05,
      "loss": 0.0012,
      "step": 6860
    },
    {
      "epoch": 3.37,
      "learning_rate": 2.310667927447778e-05,
      "loss": 0.0015,
      "step": 6870
    },
    {
      "epoch": 3.37,
      "learning_rate": 2.30636981002321e-05,
      "loss": 0.0014,
      "step": 6880
    },
    {
      "epoch": 3.38,
      "learning_rate": 2.3020716925986418e-05,
      "loss": 0.0015,
      "step": 6890
    },
    {
      "epoch": 3.38,
      "learning_rate": 2.297773575174074e-05,
      "loss": 0.0012,
      "step": 6900
    },
    {
      "epoch": 3.39,
      "learning_rate": 2.2934754577495056e-05,
      "loss": 0.0016,
      "step": 6910
    },
    {
      "epoch": 3.39,
      "learning_rate": 2.2891773403249377e-05,
      "loss": 0.0018,
      "step": 6920
    },
    {
      "epoch": 3.4,
      "learning_rate": 2.2848792229003697e-05,
      "loss": 0.0013,
      "step": 6930
    },
    {
      "epoch": 3.4,
      "learning_rate": 2.2805811054758018e-05,
      "loss": 0.0036,
      "step": 6940
    },
    {
      "epoch": 3.41,
      "learning_rate": 2.276282988051234e-05,
      "loss": 0.0014,
      "step": 6950
    },
    {
      "epoch": 3.41,
      "learning_rate": 2.2719848706266656e-05,
      "loss": 0.0015,
      "step": 6960
    },
    {
      "epoch": 3.41,
      "learning_rate": 2.2676867532020977e-05,
      "loss": 0.002,
      "step": 6970
    },
    {
      "epoch": 3.42,
      "learning_rate": 2.2633886357775298e-05,
      "loss": 0.0068,
      "step": 6980
    },
    {
      "epoch": 3.42,
      "learning_rate": 2.2590905183529615e-05,
      "loss": 0.0013,
      "step": 6990
    },
    {
      "epoch": 3.43,
      "learning_rate": 2.2547924009283935e-05,
      "loss": 0.0017,
      "step": 7000
    },
    {
      "epoch": 3.43,
      "eval_f1": 0.9963900572955778,
      "eval_f1_0": 0.998950158274011,
      "eval_f1_1": 0.9988624690888339,
      "eval_f1_2": 0.9887319564478889,
      "eval_f1_3": 0.9990156453715776,
      "eval_loss": 0.005993272177875042,
      "eval_runtime": 40.118,
      "eval_samples_per_second": 180.866,
      "eval_steps_per_second": 0.723,
      "step": 7000
    },
    {
      "epoch": 3.43,
      "learning_rate": 2.2504942835038253e-05,
      "loss": 0.002,
      "step": 7010
    },
    {
      "epoch": 3.44,
      "learning_rate": 2.2461961660792573e-05,
      "loss": 0.0016,
      "step": 7020
    },
    {
      "epoch": 3.44,
      "learning_rate": 2.241898048654689e-05,
      "loss": 0.0013,
      "step": 7030
    },
    {
      "epoch": 3.45,
      "learning_rate": 2.237599931230121e-05,
      "loss": 0.0023,
      "step": 7040
    },
    {
      "epoch": 3.45,
      "learning_rate": 2.2333018138055532e-05,
      "loss": 0.0024,
      "step": 7050
    },
    {
      "epoch": 3.46,
      "learning_rate": 2.2290036963809853e-05,
      "loss": 0.0012,
      "step": 7060
    },
    {
      "epoch": 3.46,
      "learning_rate": 2.2247055789564174e-05,
      "loss": 0.0016,
      "step": 7070
    },
    {
      "epoch": 3.47,
      "learning_rate": 2.220407461531849e-05,
      "loss": 0.001,
      "step": 7080
    },
    {
      "epoch": 3.47,
      "learning_rate": 2.216109344107281e-05,
      "loss": 0.001,
      "step": 7090
    },
    {
      "epoch": 3.48,
      "learning_rate": 2.2118112266827132e-05,
      "loss": 0.0011,
      "step": 7100
    },
    {
      "epoch": 3.48,
      "learning_rate": 2.207513109258145e-05,
      "loss": 0.0024,
      "step": 7110
    },
    {
      "epoch": 3.49,
      "learning_rate": 2.203214991833577e-05,
      "loss": 0.0025,
      "step": 7120
    },
    {
      "epoch": 3.49,
      "learning_rate": 2.1989168744090087e-05,
      "loss": 0.0015,
      "step": 7130
    },
    {
      "epoch": 3.5,
      "learning_rate": 2.1946187569844408e-05,
      "loss": 0.0013,
      "step": 7140
    },
    {
      "epoch": 3.5,
      "learning_rate": 2.190320639559873e-05,
      "loss": 0.0012,
      "step": 7150
    },
    {
      "epoch": 3.51,
      "learning_rate": 2.186022522135305e-05,
      "loss": 0.0011,
      "step": 7160
    },
    {
      "epoch": 3.51,
      "learning_rate": 2.181724404710737e-05,
      "loss": 0.0038,
      "step": 7170
    },
    {
      "epoch": 3.52,
      "learning_rate": 2.1774262872861688e-05,
      "loss": 0.0012,
      "step": 7180
    },
    {
      "epoch": 3.52,
      "learning_rate": 2.1731281698616008e-05,
      "loss": 0.0012,
      "step": 7190
    },
    {
      "epoch": 3.53,
      "learning_rate": 2.168830052437033e-05,
      "loss": 0.0025,
      "step": 7200
    },
    {
      "epoch": 3.53,
      "eval_f1": 0.9960830017559859,
      "eval_f1_0": 0.9988569208722053,
      "eval_f1_1": 0.9988395320079901,
      "eval_f1_2": 0.9876329855947822,
      "eval_f1_3": 0.9990025685489654,
      "eval_loss": 0.00589128490537405,
      "eval_runtime": 40.1106,
      "eval_samples_per_second": 180.9,
      "eval_steps_per_second": 0.723,
      "step": 7200
    },
    {
      "epoch": 3.53,
      "learning_rate": 2.1645319350124646e-05,
      "loss": 0.0012,
      "step": 7210
    },
    {
      "epoch": 3.54,
      "learning_rate": 2.1602338175878967e-05,
      "loss": 0.0016,
      "step": 7220
    },
    {
      "epoch": 3.54,
      "learning_rate": 2.1559357001633284e-05,
      "loss": 0.0014,
      "step": 7230
    },
    {
      "epoch": 3.55,
      "learning_rate": 2.1516375827387605e-05,
      "loss": 0.0019,
      "step": 7240
    },
    {
      "epoch": 3.55,
      "learning_rate": 2.1473394653141926e-05,
      "loss": 0.0019,
      "step": 7250
    },
    {
      "epoch": 3.56,
      "learning_rate": 2.1430413478896243e-05,
      "loss": 0.0021,
      "step": 7260
    },
    {
      "epoch": 3.56,
      "learning_rate": 2.1387432304650564e-05,
      "loss": 0.0012,
      "step": 7270
    },
    {
      "epoch": 3.57,
      "learning_rate": 2.1344451130404884e-05,
      "loss": 0.0015,
      "step": 7280
    },
    {
      "epoch": 3.57,
      "learning_rate": 2.1301469956159205e-05,
      "loss": 0.0016,
      "step": 7290
    },
    {
      "epoch": 3.58,
      "learning_rate": 2.1258488781913526e-05,
      "loss": 0.0065,
      "step": 7300
    },
    {
      "epoch": 3.58,
      "learning_rate": 2.1215507607667843e-05,
      "loss": 0.002,
      "step": 7310
    },
    {
      "epoch": 3.59,
      "learning_rate": 2.1172526433422164e-05,
      "loss": 0.0009,
      "step": 7320
    },
    {
      "epoch": 3.59,
      "learning_rate": 2.112954525917648e-05,
      "loss": 0.0012,
      "step": 7330
    },
    {
      "epoch": 3.6,
      "learning_rate": 2.10865640849308e-05,
      "loss": 0.0045,
      "step": 7340
    },
    {
      "epoch": 3.6,
      "learning_rate": 2.104358291068512e-05,
      "loss": 0.0026,
      "step": 7350
    },
    {
      "epoch": 3.61,
      "learning_rate": 2.100060173643944e-05,
      "loss": 0.0016,
      "step": 7360
    },
    {
      "epoch": 3.61,
      "learning_rate": 2.095762056219376e-05,
      "loss": 0.003,
      "step": 7370
    },
    {
      "epoch": 3.62,
      "learning_rate": 2.0914639387948078e-05,
      "loss": 0.0015,
      "step": 7380
    },
    {
      "epoch": 3.62,
      "learning_rate": 2.0871658213702398e-05,
      "loss": 0.001,
      "step": 7390
    },
    {
      "epoch": 3.63,
      "learning_rate": 2.082867703945672e-05,
      "loss": 0.0034,
      "step": 7400
    },
    {
      "epoch": 3.63,
      "eval_f1": 0.9968802208050749,
      "eval_f1_0": 0.9990961091670393,
      "eval_f1_1": 0.9988643302687371,
      "eval_f1_2": 0.9905448048299016,
      "eval_f1_3": 0.9990156389546216,
      "eval_loss": 0.005084398202598095,
      "eval_runtime": 40.1072,
      "eval_samples_per_second": 180.915,
      "eval_steps_per_second": 0.723,
      "step": 7400
    },
    {
      "epoch": 3.63,
      "learning_rate": 2.078569586521104e-05,
      "loss": 0.0013,
      "step": 7410
    },
    {
      "epoch": 3.64,
      "learning_rate": 2.074271469096536e-05,
      "loss": 0.0014,
      "step": 7420
    },
    {
      "epoch": 3.64,
      "learning_rate": 2.0699733516719678e-05,
      "loss": 0.0013,
      "step": 7430
    },
    {
      "epoch": 3.65,
      "learning_rate": 2.0656752342473998e-05,
      "loss": 0.0014,
      "step": 7440
    },
    {
      "epoch": 3.65,
      "learning_rate": 2.0613771168228316e-05,
      "loss": 0.0015,
      "step": 7450
    },
    {
      "epoch": 3.66,
      "learning_rate": 2.0570789993982636e-05,
      "loss": 0.0012,
      "step": 7460
    },
    {
      "epoch": 3.66,
      "learning_rate": 2.0527808819736957e-05,
      "loss": 0.0016,
      "step": 7470
    },
    {
      "epoch": 3.66,
      "learning_rate": 2.0484827645491274e-05,
      "loss": 0.0021,
      "step": 7480
    },
    {
      "epoch": 3.67,
      "learning_rate": 2.0441846471245595e-05,
      "loss": 0.0011,
      "step": 7490
    },
    {
      "epoch": 3.67,
      "learning_rate": 2.0398865296999916e-05,
      "loss": 0.0013,
      "step": 7500
    },
    {
      "epoch": 3.68,
      "learning_rate": 2.0355884122754236e-05,
      "loss": 0.001,
      "step": 7510
    },
    {
      "epoch": 3.68,
      "learning_rate": 2.0312902948508557e-05,
      "loss": 0.0011,
      "step": 7520
    },
    {
      "epoch": 3.69,
      "learning_rate": 2.0269921774262874e-05,
      "loss": 0.0009,
      "step": 7530
    },
    {
      "epoch": 3.69,
      "learning_rate": 2.0226940600017195e-05,
      "loss": 0.0013,
      "step": 7540
    },
    {
      "epoch": 3.7,
      "learning_rate": 2.0183959425771512e-05,
      "loss": 0.0015,
      "step": 7550
    },
    {
      "epoch": 3.7,
      "learning_rate": 2.0140978251525833e-05,
      "loss": 0.001,
      "step": 7560
    },
    {
      "epoch": 3.71,
      "learning_rate": 2.0097997077280154e-05,
      "loss": 0.0016,
      "step": 7570
    },
    {
      "epoch": 3.71,
      "learning_rate": 2.005501590303447e-05,
      "loss": 0.0016,
      "step": 7580
    },
    {
      "epoch": 3.72,
      "learning_rate": 2.001203472878879e-05,
      "loss": 0.0016,
      "step": 7590
    },
    {
      "epoch": 3.72,
      "learning_rate": 1.996905355454311e-05,
      "loss": 0.0026,
      "step": 7600
    },
    {
      "epoch": 3.72,
      "eval_f1": 0.9968933350838978,
      "eval_f1_0": 0.9990993368548533,
      "eval_f1_1": 0.9988357227650443,
      "eval_f1_2": 0.9906389554377376,
      "eval_f1_3": 0.9989993252779561,
      "eval_loss": 0.00519435154274106,
      "eval_runtime": 40.0192,
      "eval_samples_per_second": 181.313,
      "eval_steps_per_second": 0.725,
      "step": 7600
    },
    {
      "epoch": 3.73,
      "learning_rate": 1.992607238029743e-05,
      "loss": 0.0012,
      "step": 7610
    },
    {
      "epoch": 3.73,
      "learning_rate": 1.988309120605175e-05,
      "loss": 0.0014,
      "step": 7620
    },
    {
      "epoch": 3.74,
      "learning_rate": 1.984011003180607e-05,
      "loss": 0.0045,
      "step": 7630
    },
    {
      "epoch": 3.74,
      "learning_rate": 1.979712885756039e-05,
      "loss": 0.0052,
      "step": 7640
    },
    {
      "epoch": 3.75,
      "learning_rate": 1.975414768331471e-05,
      "loss": 0.0016,
      "step": 7650
    },
    {
      "epoch": 3.75,
      "learning_rate": 1.971116650906903e-05,
      "loss": 0.0013,
      "step": 7660
    },
    {
      "epoch": 3.76,
      "learning_rate": 1.966818533482335e-05,
      "loss": 0.004,
      "step": 7670
    },
    {
      "epoch": 3.76,
      "learning_rate": 1.9625204160577668e-05,
      "loss": 0.0121,
      "step": 7680
    },
    {
      "epoch": 3.77,
      "learning_rate": 1.9582222986331988e-05,
      "loss": 0.0015,
      "step": 7690
    },
    {
      "epoch": 3.77,
      "learning_rate": 1.9539241812086306e-05,
      "loss": 0.0011,
      "step": 7700
    },
    {
      "epoch": 3.78,
      "learning_rate": 1.9496260637840626e-05,
      "loss": 0.0062,
      "step": 7710
    },
    {
      "epoch": 3.78,
      "learning_rate": 1.9453279463594944e-05,
      "loss": 0.0025,
      "step": 7720
    },
    {
      "epoch": 3.79,
      "learning_rate": 1.9410298289349264e-05,
      "loss": 0.0031,
      "step": 7730
    },
    {
      "epoch": 3.79,
      "learning_rate": 1.9367317115103585e-05,
      "loss": 0.0014,
      "step": 7740
    },
    {
      "epoch": 3.8,
      "learning_rate": 1.9324335940857906e-05,
      "loss": 0.0013,
      "step": 7750
    },
    {
      "epoch": 3.8,
      "learning_rate": 1.9281354766612226e-05,
      "loss": 0.0014,
      "step": 7760
    },
    {
      "epoch": 3.81,
      "learning_rate": 1.9238373592366544e-05,
      "loss": 0.0013,
      "step": 7770
    },
    {
      "epoch": 3.81,
      "learning_rate": 1.9195392418120864e-05,
      "loss": 0.0022,
      "step": 7780
    },
    {
      "epoch": 3.82,
      "learning_rate": 1.9152411243875185e-05,
      "loss": 0.0026,
      "step": 7790
    },
    {
      "epoch": 3.82,
      "learning_rate": 1.9109430069629502e-05,
      "loss": 0.0012,
      "step": 7800
    },
    {
      "epoch": 3.82,
      "eval_f1": 0.9960833676694076,
      "eval_f1_0": 0.9988580495650717,
      "eval_f1_1": 0.9988491206711243,
      "eval_f1_2": 0.9876041361650332,
      "eval_f1_3": 0.9990221642764016,
      "eval_loss": 0.005810524336993694,
      "eval_runtime": 39.998,
      "eval_samples_per_second": 181.409,
      "eval_steps_per_second": 0.725,
      "step": 7800
    },
    {
      "epoch": 3.83,
      "learning_rate": 1.9066448895383823e-05,
      "loss": 0.0014,
      "step": 7810
    },
    {
      "epoch": 3.83,
      "learning_rate": 1.902346772113814e-05,
      "loss": 0.0036,
      "step": 7820
    },
    {
      "epoch": 3.84,
      "learning_rate": 1.898048654689246e-05,
      "loss": 0.0029,
      "step": 7830
    },
    {
      "epoch": 3.84,
      "learning_rate": 1.893750537264678e-05,
      "loss": 0.0026,
      "step": 7840
    },
    {
      "epoch": 3.85,
      "learning_rate": 1.8894524198401102e-05,
      "loss": 0.01,
      "step": 7850
    },
    {
      "epoch": 3.85,
      "learning_rate": 1.8851543024155423e-05,
      "loss": 0.0018,
      "step": 7860
    },
    {
      "epoch": 3.86,
      "learning_rate": 1.880856184990974e-05,
      "loss": 0.0018,
      "step": 7870
    },
    {
      "epoch": 3.86,
      "learning_rate": 1.876558067566406e-05,
      "loss": 0.0015,
      "step": 7880
    },
    {
      "epoch": 3.87,
      "learning_rate": 1.872259950141838e-05,
      "loss": 0.0026,
      "step": 7890
    },
    {
      "epoch": 3.87,
      "learning_rate": 1.86796183271727e-05,
      "loss": 0.0015,
      "step": 7900
    },
    {
      "epoch": 3.88,
      "learning_rate": 1.863663715292702e-05,
      "loss": 0.0037,
      "step": 7910
    },
    {
      "epoch": 3.88,
      "learning_rate": 1.8593655978681337e-05,
      "loss": 0.0014,
      "step": 7920
    },
    {
      "epoch": 3.89,
      "learning_rate": 1.8550674804435658e-05,
      "loss": 0.0022,
      "step": 7930
    },
    {
      "epoch": 3.89,
      "learning_rate": 1.8507693630189978e-05,
      "loss": 0.0011,
      "step": 7940
    },
    {
      "epoch": 3.9,
      "learning_rate": 1.8464712455944296e-05,
      "loss": 0.0018,
      "step": 7950
    },
    {
      "epoch": 3.9,
      "learning_rate": 1.8421731281698616e-05,
      "loss": 0.0017,
      "step": 7960
    },
    {
      "epoch": 3.9,
      "learning_rate": 1.8378750107452937e-05,
      "loss": 0.0017,
      "step": 7970
    },
    {
      "epoch": 3.91,
      "learning_rate": 1.8335768933207258e-05,
      "loss": 0.0011,
      "step": 7980
    },
    {
      "epoch": 3.91,
      "learning_rate": 1.829278775896158e-05,
      "loss": 0.0015,
      "step": 7990
    },
    {
      "epoch": 3.92,
      "learning_rate": 1.8249806584715896e-05,
      "loss": 0.0015,
      "step": 8000
    },
    {
      "epoch": 3.92,
      "eval_f1": 0.996516782409872,
      "eval_f1_0": 0.9989879146985331,
      "eval_f1_1": 0.9988585734450917,
      "eval_f1_2": 0.989201733474051,
      "eval_f1_3": 0.9990189080218121,
      "eval_loss": 0.005502467509359121,
      "eval_runtime": 40.0336,
      "eval_samples_per_second": 181.248,
      "eval_steps_per_second": 0.724,
      "step": 8000
    },
    {
      "epoch": 3.92,
      "learning_rate": 1.8206825410470216e-05,
      "loss": 0.001,
      "step": 8010
    },
    {
      "epoch": 3.93,
      "learning_rate": 1.8163844236224534e-05,
      "loss": 0.0017,
      "step": 8020
    },
    {
      "epoch": 3.93,
      "learning_rate": 1.8120863061978854e-05,
      "loss": 0.0015,
      "step": 8030
    },
    {
      "epoch": 3.94,
      "learning_rate": 1.807788188773317e-05,
      "loss": 0.0014,
      "step": 8040
    },
    {
      "epoch": 3.94,
      "learning_rate": 1.8034900713487492e-05,
      "loss": 0.0037,
      "step": 8050
    },
    {
      "epoch": 3.95,
      "learning_rate": 1.7991919539241813e-05,
      "loss": 0.0018,
      "step": 8060
    },
    {
      "epoch": 3.95,
      "learning_rate": 1.794893836499613e-05,
      "loss": 0.0017,
      "step": 8070
    },
    {
      "epoch": 3.96,
      "learning_rate": 1.790595719075045e-05,
      "loss": 0.0016,
      "step": 8080
    },
    {
      "epoch": 3.96,
      "learning_rate": 1.786297601650477e-05,
      "loss": 0.0019,
      "step": 8090
    },
    {
      "epoch": 3.97,
      "learning_rate": 1.7819994842259092e-05,
      "loss": 0.0055,
      "step": 8100
    },
    {
      "epoch": 3.97,
      "learning_rate": 1.7777013668013413e-05,
      "loss": 0.0024,
      "step": 8110
    },
    {
      "epoch": 3.98,
      "learning_rate": 1.773403249376773e-05,
      "loss": 0.0015,
      "step": 8120
    },
    {
      "epoch": 3.98,
      "learning_rate": 1.769105131952205e-05,
      "loss": 0.0025,
      "step": 8130
    },
    {
      "epoch": 3.99,
      "learning_rate": 1.7648070145276368e-05,
      "loss": 0.0012,
      "step": 8140
    },
    {
      "epoch": 3.99,
      "learning_rate": 1.760508897103069e-05,
      "loss": 0.0019,
      "step": 8150
    },
    {
      "epoch": 4.0,
      "learning_rate": 1.756210779678501e-05,
      "loss": 0.0013,
      "step": 8160
    },
    {
      "epoch": 4.0,
      "learning_rate": 1.7519126622539327e-05,
      "loss": 0.0025,
      "step": 8170
    },
    {
      "epoch": 4.01,
      "learning_rate": 1.7476145448293648e-05,
      "loss": 0.0018,
      "step": 8180
    },
    {
      "epoch": 4.01,
      "learning_rate": 1.743316427404797e-05,
      "loss": 0.0016,
      "step": 8190
    },
    {
      "epoch": 4.02,
      "learning_rate": 1.739018309980229e-05,
      "loss": 0.002,
      "step": 8200
    },
    {
      "epoch": 4.02,
      "eval_f1": 0.9969046595085209,
      "eval_f1_0": 0.9991035923996122,
      "eval_f1_1": 0.9988623479008607,
      "eval_f1_2": 0.9906370331116673,
      "eval_f1_3": 0.9990156646219435,
      "eval_loss": 0.005085492040961981,
      "eval_runtime": 40.0608,
      "eval_samples_per_second": 181.125,
      "eval_steps_per_second": 0.724,
      "step": 8200
    },
    {
      "epoch": 4.02,
      "learning_rate": 1.734720192555661e-05,
      "loss": 0.0017,
      "step": 8210
    },
    {
      "epoch": 4.03,
      "learning_rate": 1.7304220751310927e-05,
      "loss": 0.0015,
      "step": 8220
    },
    {
      "epoch": 4.03,
      "learning_rate": 1.7261239577065248e-05,
      "loss": 0.0011,
      "step": 8230
    },
    {
      "epoch": 4.04,
      "learning_rate": 1.7218258402819565e-05,
      "loss": 0.001,
      "step": 8240
    },
    {
      "epoch": 4.04,
      "learning_rate": 1.7175277228573886e-05,
      "loss": 0.0014,
      "step": 8250
    },
    {
      "epoch": 4.05,
      "learning_rate": 1.7132296054328206e-05,
      "loss": 0.0023,
      "step": 8260
    },
    {
      "epoch": 4.05,
      "learning_rate": 1.7089314880082524e-05,
      "loss": 0.0016,
      "step": 8270
    },
    {
      "epoch": 4.06,
      "learning_rate": 1.7046333705836844e-05,
      "loss": 0.0013,
      "step": 8280
    },
    {
      "epoch": 4.06,
      "learning_rate": 1.700335253159116e-05,
      "loss": 0.0011,
      "step": 8290
    },
    {
      "epoch": 4.07,
      "learning_rate": 1.6960371357345482e-05,
      "loss": 0.0025,
      "step": 8300
    },
    {
      "epoch": 4.07,
      "learning_rate": 1.6917390183099803e-05,
      "loss": 0.0012,
      "step": 8310
    },
    {
      "epoch": 4.08,
      "learning_rate": 1.6874409008854124e-05,
      "loss": 0.0017,
      "step": 8320
    },
    {
      "epoch": 4.08,
      "learning_rate": 1.6831427834608444e-05,
      "loss": 0.001,
      "step": 8330
    },
    {
      "epoch": 4.09,
      "learning_rate": 1.678844666036276e-05,
      "loss": 0.0015,
      "step": 8340
    },
    {
      "epoch": 4.09,
      "learning_rate": 1.6745465486117082e-05,
      "loss": 0.0014,
      "step": 8350
    },
    {
      "epoch": 4.1,
      "learning_rate": 1.6702484311871403e-05,
      "loss": 0.0015,
      "step": 8360
    },
    {
      "epoch": 4.1,
      "learning_rate": 1.665950313762572e-05,
      "loss": 0.0015,
      "step": 8370
    },
    {
      "epoch": 4.11,
      "learning_rate": 1.661652196338004e-05,
      "loss": 0.0046,
      "step": 8380
    },
    {
      "epoch": 4.11,
      "learning_rate": 1.657354078913436e-05,
      "loss": 0.0013,
      "step": 8390
    },
    {
      "epoch": 4.12,
      "learning_rate": 1.653055961488868e-05,
      "loss": 0.0014,
      "step": 8400
    },
    {
      "epoch": 4.12,
      "eval_f1": 0.9969487157284955,
      "eval_f1_0": 0.9991165791830865,
      "eval_f1_1": 0.9988661743275995,
      "eval_f1_2": 0.9907899132558434,
      "eval_f1_3": 0.9990221961474528,
      "eval_loss": 0.005065333563834429,
      "eval_runtime": 40.1621,
      "eval_samples_per_second": 180.668,
      "eval_steps_per_second": 0.722,
      "step": 8400
    },
    {
      "epoch": 4.12,
      "learning_rate": 1.6487578440642996e-05,
      "loss": 0.0011,
      "step": 8410
    },
    {
      "epoch": 4.13,
      "learning_rate": 1.6444597266397317e-05,
      "loss": 0.0011,
      "step": 8420
    },
    {
      "epoch": 4.13,
      "learning_rate": 1.6401616092151638e-05,
      "loss": 0.0011,
      "step": 8430
    },
    {
      "epoch": 4.14,
      "learning_rate": 1.635863491790596e-05,
      "loss": 0.0013,
      "step": 8440
    },
    {
      "epoch": 4.14,
      "learning_rate": 1.631565374366028e-05,
      "loss": 0.0011,
      "step": 8450
    },
    {
      "epoch": 4.15,
      "learning_rate": 1.6272672569414596e-05,
      "loss": 0.0017,
      "step": 8460
    },
    {
      "epoch": 4.15,
      "learning_rate": 1.6229691395168917e-05,
      "loss": 0.002,
      "step": 8470
    },
    {
      "epoch": 4.15,
      "learning_rate": 1.6186710220923238e-05,
      "loss": 0.0013,
      "step": 8480
    },
    {
      "epoch": 4.16,
      "learning_rate": 1.6143729046677555e-05,
      "loss": 0.0016,
      "step": 8490
    },
    {
      "epoch": 4.16,
      "learning_rate": 1.6100747872431876e-05,
      "loss": 0.0015,
      "step": 8500
    },
    {
      "epoch": 4.17,
      "learning_rate": 1.6057766698186193e-05,
      "loss": 0.0013,
      "step": 8510
    },
    {
      "epoch": 4.17,
      "learning_rate": 1.6014785523940514e-05,
      "loss": 0.0012,
      "step": 8520
    },
    {
      "epoch": 4.18,
      "learning_rate": 1.5971804349694834e-05,
      "loss": 0.0012,
      "step": 8530
    },
    {
      "epoch": 4.18,
      "learning_rate": 1.5928823175449155e-05,
      "loss": 0.0016,
      "step": 8540
    },
    {
      "epoch": 4.19,
      "learning_rate": 1.5885842001203476e-05,
      "loss": 0.0011,
      "step": 8550
    },
    {
      "epoch": 4.19,
      "learning_rate": 1.5842860826957793e-05,
      "loss": 0.0018,
      "step": 8560
    },
    {
      "epoch": 4.2,
      "learning_rate": 1.5799879652712114e-05,
      "loss": 0.0022,
      "step": 8570
    },
    {
      "epoch": 4.2,
      "learning_rate": 1.5756898478466434e-05,
      "loss": 0.0062,
      "step": 8580
    },
    {
      "epoch": 4.21,
      "learning_rate": 1.5713917304220752e-05,
      "loss": 0.0023,
      "step": 8590
    },
    {
      "epoch": 4.21,
      "learning_rate": 1.5670936129975072e-05,
      "loss": 0.0012,
      "step": 8600
    },
    {
      "epoch": 4.21,
      "eval_f1": 0.9964114001832709,
      "eval_f1_0": 0.9989568596734625,
      "eval_f1_1": 0.9988719008185853,
      "eval_f1_2": 0.9887946568417544,
      "eval_f1_3": 0.9990221833992816,
      "eval_loss": 0.005537047982215881,
      "eval_runtime": 40.0702,
      "eval_samples_per_second": 181.082,
      "eval_steps_per_second": 0.724,
      "step": 8600
    },
    {
      "epoch": 4.22,
      "learning_rate": 1.562795495572939e-05,
      "loss": 0.0013,
      "step": 8610
    },
    {
      "epoch": 4.22,
      "learning_rate": 1.558497378148371e-05,
      "loss": 0.0017,
      "step": 8620
    },
    {
      "epoch": 4.23,
      "learning_rate": 1.554199260723803e-05,
      "loss": 0.0015,
      "step": 8630
    },
    {
      "epoch": 4.23,
      "learning_rate": 1.549901143299235e-05,
      "loss": 0.0011,
      "step": 8640
    },
    {
      "epoch": 4.24,
      "learning_rate": 1.545603025874667e-05,
      "loss": 0.0018,
      "step": 8650
    },
    {
      "epoch": 4.24,
      "learning_rate": 1.541304908450099e-05,
      "loss": 0.0017,
      "step": 8660
    },
    {
      "epoch": 4.25,
      "learning_rate": 1.537006791025531e-05,
      "loss": 0.0017,
      "step": 8670
    },
    {
      "epoch": 4.25,
      "learning_rate": 1.532708673600963e-05,
      "loss": 0.0018,
      "step": 8680
    },
    {
      "epoch": 4.26,
      "learning_rate": 1.528410556176395e-05,
      "loss": 0.0014,
      "step": 8690
    },
    {
      "epoch": 4.26,
      "learning_rate": 1.5241124387518269e-05,
      "loss": 0.0014,
      "step": 8700
    },
    {
      "epoch": 4.27,
      "learning_rate": 1.5198143213272586e-05,
      "loss": 0.0014,
      "step": 8710
    },
    {
      "epoch": 4.27,
      "learning_rate": 1.5155162039026907e-05,
      "loss": 0.0012,
      "step": 8720
    },
    {
      "epoch": 4.28,
      "learning_rate": 1.5112180864781224e-05,
      "loss": 0.0014,
      "step": 8730
    },
    {
      "epoch": 4.28,
      "learning_rate": 1.5069199690535547e-05,
      "loss": 0.0008,
      "step": 8740
    },
    {
      "epoch": 4.29,
      "learning_rate": 1.5026218516289867e-05,
      "loss": 0.0016,
      "step": 8750
    },
    {
      "epoch": 4.29,
      "learning_rate": 1.4983237342044185e-05,
      "loss": 0.0012,
      "step": 8760
    },
    {
      "epoch": 4.3,
      "learning_rate": 1.4940256167798505e-05,
      "loss": 0.0095,
      "step": 8770
    },
    {
      "epoch": 4.3,
      "learning_rate": 1.4897274993552823e-05,
      "loss": 0.0054,
      "step": 8780
    },
    {
      "epoch": 4.31,
      "learning_rate": 1.4854293819307143e-05,
      "loss": 0.001,
      "step": 8790
    },
    {
      "epoch": 4.31,
      "learning_rate": 1.4811312645061464e-05,
      "loss": 0.0016,
      "step": 8800
    },
    {
      "epoch": 4.31,
      "eval_f1": 0.9963568320832532,
      "eval_f1_0": 0.9989399704103569,
      "eval_f1_1": 0.9988547817617337,
      "eval_f1_2": 0.9886103736395085,
      "eval_f1_3": 0.9990222025214137,
      "eval_loss": 0.005580628756433725,
      "eval_runtime": 40.1243,
      "eval_samples_per_second": 180.838,
      "eval_steps_per_second": 0.723,
      "step": 8800
    },
    {
      "epoch": 4.32,
      "learning_rate": 1.4768331470815783e-05,
      "loss": 0.0018,
      "step": 8810
    },
    {
      "epoch": 4.32,
      "learning_rate": 1.4725350296570104e-05,
      "loss": 0.0019,
      "step": 8820
    },
    {
      "epoch": 4.33,
      "learning_rate": 1.4682369122324421e-05,
      "loss": 0.0013,
      "step": 8830
    },
    {
      "epoch": 4.33,
      "learning_rate": 1.4639387948078742e-05,
      "loss": 0.0026,
      "step": 8840
    },
    {
      "epoch": 4.34,
      "learning_rate": 1.4596406773833062e-05,
      "loss": 0.0013,
      "step": 8850
    },
    {
      "epoch": 4.34,
      "learning_rate": 1.4553425599587381e-05,
      "loss": 0.0029,
      "step": 8860
    },
    {
      "epoch": 4.35,
      "learning_rate": 1.4510444425341702e-05,
      "loss": 0.0014,
      "step": 8870
    },
    {
      "epoch": 4.35,
      "learning_rate": 1.446746325109602e-05,
      "loss": 0.0026,
      "step": 8880
    },
    {
      "epoch": 4.36,
      "learning_rate": 1.442448207685034e-05,
      "loss": 0.0014,
      "step": 8890
    },
    {
      "epoch": 4.36,
      "learning_rate": 1.438150090260466e-05,
      "loss": 0.002,
      "step": 8900
    },
    {
      "epoch": 4.37,
      "learning_rate": 1.433851972835898e-05,
      "loss": 0.0013,
      "step": 8910
    },
    {
      "epoch": 4.37,
      "learning_rate": 1.42955385541133e-05,
      "loss": 0.0013,
      "step": 8920
    },
    {
      "epoch": 4.38,
      "learning_rate": 1.4252557379867618e-05,
      "loss": 0.0017,
      "step": 8930
    },
    {
      "epoch": 4.38,
      "learning_rate": 1.4209576205621938e-05,
      "loss": 0.005,
      "step": 8940
    },
    {
      "epoch": 4.39,
      "learning_rate": 1.4166595031376259e-05,
      "loss": 0.0012,
      "step": 8950
    },
    {
      "epoch": 4.39,
      "learning_rate": 1.4123613857130576e-05,
      "loss": 0.001,
      "step": 8960
    },
    {
      "epoch": 4.39,
      "learning_rate": 1.4080632682884897e-05,
      "loss": 0.0019,
      "step": 8970
    },
    {
      "epoch": 4.4,
      "learning_rate": 1.4037651508639216e-05,
      "loss": 0.0011,
      "step": 8980
    },
    {
      "epoch": 4.4,
      "learning_rate": 1.3994670334393537e-05,
      "loss": 0.006,
      "step": 8990
    },
    {
      "epoch": 4.41,
      "learning_rate": 1.3951689160147857e-05,
      "loss": 0.0014,
      "step": 9000
    },
    {
      "epoch": 4.41,
      "eval_f1": 0.996078784570625,
      "eval_f1_0": 0.9988563971311436,
      "eval_f1_1": 0.9988719094027207,
      "eval_f1_2": 0.9875548670281435,
      "eval_f1_3": 0.999031964720492,
      "eval_loss": 0.005990075878798962,
      "eval_runtime": 40.0655,
      "eval_samples_per_second": 181.104,
      "eval_steps_per_second": 0.724,
      "step": 9000
    },
    {
      "epoch": 4.41,
      "learning_rate": 1.3908707985902175e-05,
      "loss": 0.0016,
      "step": 9010
    },
    {
      "epoch": 4.42,
      "learning_rate": 1.3865726811656495e-05,
      "loss": 0.0014,
      "step": 9020
    },
    {
      "epoch": 4.42,
      "learning_rate": 1.3822745637410814e-05,
      "loss": 0.0017,
      "step": 9030
    },
    {
      "epoch": 4.43,
      "learning_rate": 1.3779764463165135e-05,
      "loss": 0.0011,
      "step": 9040
    },
    {
      "epoch": 4.43,
      "learning_rate": 1.3736783288919456e-05,
      "loss": 0.0015,
      "step": 9050
    },
    {
      "epoch": 4.44,
      "learning_rate": 1.3693802114673773e-05,
      "loss": 0.0014,
      "step": 9060
    },
    {
      "epoch": 4.44,
      "learning_rate": 1.3650820940428094e-05,
      "loss": 0.0014,
      "step": 9070
    },
    {
      "epoch": 4.45,
      "learning_rate": 1.3607839766182413e-05,
      "loss": 0.0042,
      "step": 9080
    },
    {
      "epoch": 4.45,
      "learning_rate": 1.3564858591936733e-05,
      "loss": 0.0012,
      "step": 9090
    },
    {
      "epoch": 4.46,
      "learning_rate": 1.352187741769105e-05,
      "loss": 0.0012,
      "step": 9100
    },
    {
      "epoch": 4.46,
      "learning_rate": 1.3478896243445371e-05,
      "loss": 0.0018,
      "step": 9110
    },
    {
      "epoch": 4.47,
      "learning_rate": 1.3435915069199692e-05,
      "loss": 0.001,
      "step": 9120
    },
    {
      "epoch": 4.47,
      "learning_rate": 1.339293389495401e-05,
      "loss": 0.0014,
      "step": 9130
    },
    {
      "epoch": 4.48,
      "learning_rate": 1.334995272070833e-05,
      "loss": 0.001,
      "step": 9140
    },
    {
      "epoch": 4.48,
      "learning_rate": 1.3306971546462649e-05,
      "loss": 0.0017,
      "step": 9150
    },
    {
      "epoch": 4.49,
      "learning_rate": 1.326399037221697e-05,
      "loss": 0.0013,
      "step": 9160
    },
    {
      "epoch": 4.49,
      "learning_rate": 1.322100919797129e-05,
      "loss": 0.0016,
      "step": 9170
    },
    {
      "epoch": 4.5,
      "learning_rate": 1.3178028023725608e-05,
      "loss": 0.0173,
      "step": 9180
    },
    {
      "epoch": 4.5,
      "learning_rate": 1.3135046849479928e-05,
      "loss": 0.0045,
      "step": 9190
    },
    {
      "epoch": 4.51,
      "learning_rate": 1.3092065675234247e-05,
      "loss": 0.0115,
      "step": 9200
    },
    {
      "epoch": 4.51,
      "eval_f1": 0.9966969586846295,
      "eval_f1_0": 0.9990406008860936,
      "eval_f1_1": 0.9988680831797161,
      "eval_f1_2": 0.9898504611596021,
      "eval_f1_3": 0.9990286895131062,
      "eval_loss": 0.005458533298224211,
      "eval_runtime": 40.1451,
      "eval_samples_per_second": 180.744,
      "eval_steps_per_second": 0.722,
      "step": 9200
    },
    {
      "epoch": 4.51,
      "learning_rate": 1.3049084500988568e-05,
      "loss": 0.0019,
      "step": 9210
    },
    {
      "epoch": 4.52,
      "learning_rate": 1.3006103326742889e-05,
      "loss": 0.0015,
      "step": 9220
    },
    {
      "epoch": 4.52,
      "learning_rate": 1.2963122152497206e-05,
      "loss": 0.0016,
      "step": 9230
    },
    {
      "epoch": 4.53,
      "learning_rate": 1.2920140978251527e-05,
      "loss": 0.0014,
      "step": 9240
    },
    {
      "epoch": 4.53,
      "learning_rate": 1.2877159804005846e-05,
      "loss": 0.0014,
      "step": 9250
    },
    {
      "epoch": 4.54,
      "learning_rate": 1.2834178629760167e-05,
      "loss": 0.0015,
      "step": 9260
    },
    {
      "epoch": 4.54,
      "learning_rate": 1.2791197455514487e-05,
      "loss": 0.0018,
      "step": 9270
    },
    {
      "epoch": 4.55,
      "learning_rate": 1.2748216281268804e-05,
      "loss": 0.002,
      "step": 9280
    },
    {
      "epoch": 4.55,
      "learning_rate": 1.2705235107023125e-05,
      "loss": 0.0017,
      "step": 9290
    },
    {
      "epoch": 4.56,
      "learning_rate": 1.2662253932777442e-05,
      "loss": 0.0013,
      "step": 9300
    },
    {
      "epoch": 4.56,
      "learning_rate": 1.2619272758531763e-05,
      "loss": 0.0014,
      "step": 9310
    },
    {
      "epoch": 4.57,
      "learning_rate": 1.2576291584286084e-05,
      "loss": 0.0015,
      "step": 9320
    },
    {
      "epoch": 4.57,
      "learning_rate": 1.2533310410040403e-05,
      "loss": 0.0014,
      "step": 9330
    },
    {
      "epoch": 4.58,
      "learning_rate": 1.2490329235794722e-05,
      "loss": 0.001,
      "step": 9340
    },
    {
      "epoch": 4.58,
      "learning_rate": 1.2447348061549043e-05,
      "loss": 0.0017,
      "step": 9350
    },
    {
      "epoch": 4.59,
      "learning_rate": 1.2404366887303361e-05,
      "loss": 0.0016,
      "step": 9360
    },
    {
      "epoch": 4.59,
      "learning_rate": 1.236138571305768e-05,
      "loss": 0.0012,
      "step": 9370
    },
    {
      "epoch": 4.6,
      "learning_rate": 1.2318404538812001e-05,
      "loss": 0.0023,
      "step": 9380
    },
    {
      "epoch": 4.6,
      "learning_rate": 1.227542336456632e-05,
      "loss": 0.0012,
      "step": 9390
    },
    {
      "epoch": 4.61,
      "learning_rate": 1.2232442190320641e-05,
      "loss": 0.0016,
      "step": 9400
    },
    {
      "epoch": 4.61,
      "eval_f1": 0.9968398701380607,
      "eval_f1_0": 0.9990837293582686,
      "eval_f1_1": 0.9988604476405178,
      "eval_f1_2": 0.990399645348218,
      "eval_f1_3": 0.9990156582052385,
      "eval_loss": 0.005362311843782663,
      "eval_runtime": 40.1111,
      "eval_samples_per_second": 180.897,
      "eval_steps_per_second": 0.723,
      "step": 9400
    },
    {
      "epoch": 4.61,
      "step": 9400,
      "total_flos": 1.2360315828007076e+18,
      "train_loss": 0.022763097804594547,
      "train_runtime": 8838.6218,
      "train_samples_per_second": 44.329,
      "train_steps_per_second": 1.386
    }
  ],
  "logging_steps": 10,
  "max_steps": 12246,
  "num_train_epochs": 6,
  "save_steps": 200,
  "total_flos": 1.2360315828007076e+18,
  "trial_name": null,
  "trial_params": null
}