{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9999745398070118,
  "eval_steps": 500,
  "global_step": 9819,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 2.1875,
      "learning_rate": 3.389830508474576e-07,
      "loss": 2.0274,
      "step": 1
    },
    {
      "epoch": 0.0,
      "grad_norm": 2.875,
      "learning_rate": 1.6949152542372882e-06,
      "loss": 1.6129,
      "step": 5
    },
    {
      "epoch": 0.0,
      "grad_norm": 9.5625,
      "learning_rate": 3.3898305084745763e-06,
      "loss": 1.5432,
      "step": 10
    },
    {
      "epoch": 0.0,
      "grad_norm": 3.265625,
      "learning_rate": 5.084745762711865e-06,
      "loss": 1.6908,
      "step": 15
    },
    {
      "epoch": 0.0,
      "grad_norm": 2.0625,
      "learning_rate": 6.779661016949153e-06,
      "loss": 1.6712,
      "step": 20
    },
    {
      "epoch": 0.0,
      "grad_norm": 3.0625,
      "learning_rate": 8.47457627118644e-06,
      "loss": 1.7115,
      "step": 25
    },
    {
      "epoch": 0.0,
      "grad_norm": 3.328125,
      "learning_rate": 1.016949152542373e-05,
      "loss": 1.6835,
      "step": 30
    },
    {
      "epoch": 0.0,
      "grad_norm": 2.390625,
      "learning_rate": 1.1864406779661018e-05,
      "loss": 1.7974,
      "step": 35
    },
    {
      "epoch": 0.0,
      "grad_norm": 4.4375,
      "learning_rate": 1.3559322033898305e-05,
      "loss": 1.631,
      "step": 40
    },
    {
      "epoch": 0.0,
      "grad_norm": 11.3125,
      "learning_rate": 1.5254237288135596e-05,
      "loss": 1.3636,
      "step": 45
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.359375,
      "learning_rate": 1.694915254237288e-05,
      "loss": 1.674,
      "step": 50
    },
    {
      "epoch": 0.01,
      "grad_norm": 7.84375,
      "learning_rate": 1.864406779661017e-05,
      "loss": 1.7198,
      "step": 55
    },
    {
      "epoch": 0.01,
      "grad_norm": 10.8125,
      "learning_rate": 2.033898305084746e-05,
      "loss": 1.5772,
      "step": 60
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.203125,
      "learning_rate": 2.2033898305084748e-05,
      "loss": 1.5545,
      "step": 65
    },
    {
      "epoch": 0.01,
      "grad_norm": 4.125,
      "learning_rate": 2.3728813559322036e-05,
      "loss": 1.4847,
      "step": 70
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.34375,
      "learning_rate": 2.5423728813559322e-05,
      "loss": 1.5674,
      "step": 75
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.8671875,
      "learning_rate": 2.711864406779661e-05,
      "loss": 1.6044,
      "step": 80
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.953125,
      "learning_rate": 2.88135593220339e-05,
      "loss": 1.7277,
      "step": 85
    },
    {
      "epoch": 0.01,
      "grad_norm": 10.0625,
      "learning_rate": 3.050847457627119e-05,
      "loss": 1.4924,
      "step": 90
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.75,
      "learning_rate": 3.2203389830508473e-05,
      "loss": 1.5469,
      "step": 95
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.609375,
      "learning_rate": 3.389830508474576e-05,
      "loss": 1.7402,
      "step": 100
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.3203125,
      "learning_rate": 3.559322033898305e-05,
      "loss": 1.3229,
      "step": 105
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.4140625,
      "learning_rate": 3.728813559322034e-05,
      "loss": 1.3506,
      "step": 110
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.7578125,
      "learning_rate": 3.898305084745763e-05,
      "loss": 1.3828,
      "step": 115
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.328125,
      "learning_rate": 4.067796610169492e-05,
      "loss": 1.5638,
      "step": 120
    },
    {
      "epoch": 0.01,
      "grad_norm": 11.4375,
      "learning_rate": 4.2372881355932206e-05,
      "loss": 1.337,
      "step": 125
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.21875,
      "learning_rate": 4.4067796610169495e-05,
      "loss": 1.2208,
      "step": 130
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.0,
      "learning_rate": 4.5762711864406784e-05,
      "loss": 1.3564,
      "step": 135
    },
    {
      "epoch": 0.01,
      "grad_norm": 4.5625,
      "learning_rate": 4.745762711864407e-05,
      "loss": 1.1207,
      "step": 140
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.4375,
      "learning_rate": 4.915254237288136e-05,
      "loss": 1.6007,
      "step": 145
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.8671875,
      "learning_rate": 5.0847457627118643e-05,
      "loss": 1.4556,
      "step": 150
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.75,
      "learning_rate": 5.254237288135594e-05,
      "loss": 1.3288,
      "step": 155
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.1796875,
      "learning_rate": 5.423728813559322e-05,
      "loss": 1.3321,
      "step": 160
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.6875,
      "learning_rate": 5.593220338983051e-05,
      "loss": 1.316,
      "step": 165
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.5625,
      "learning_rate": 5.76271186440678e-05,
      "loss": 1.0569,
      "step": 170
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.375,
      "learning_rate": 5.932203389830509e-05,
      "loss": 1.2836,
      "step": 175
    },
    {
      "epoch": 0.02,
      "grad_norm": 3.578125,
      "learning_rate": 6.101694915254238e-05,
      "loss": 1.2244,
      "step": 180
    },
    {
      "epoch": 0.02,
      "grad_norm": 4.0625,
      "learning_rate": 6.271186440677966e-05,
      "loss": 1.3131,
      "step": 185
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.9296875,
      "learning_rate": 6.440677966101695e-05,
      "loss": 1.2183,
      "step": 190
    },
    {
      "epoch": 0.02,
      "grad_norm": 3.609375,
      "learning_rate": 6.610169491525424e-05,
      "loss": 1.1306,
      "step": 195
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.3203125,
      "learning_rate": 6.779661016949152e-05,
      "loss": 1.101,
      "step": 200
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.125,
      "learning_rate": 6.949152542372882e-05,
      "loss": 1.1969,
      "step": 205
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.71875,
      "learning_rate": 7.11864406779661e-05,
      "loss": 1.2251,
      "step": 210
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.1484375,
      "learning_rate": 7.288135593220338e-05,
      "loss": 1.2799,
      "step": 215
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.3984375,
      "learning_rate": 7.457627118644068e-05,
      "loss": 1.1388,
      "step": 220
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.1796875,
      "learning_rate": 7.627118644067796e-05,
      "loss": 1.3403,
      "step": 225
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.03125,
      "learning_rate": 7.796610169491526e-05,
      "loss": 1.064,
      "step": 230
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.171875,
      "learning_rate": 7.966101694915254e-05,
      "loss": 1.1494,
      "step": 235
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.90234375,
      "learning_rate": 8.135593220338983e-05,
      "loss": 1.1034,
      "step": 240
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.953125,
      "learning_rate": 8.305084745762712e-05,
      "loss": 0.981,
      "step": 245
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.6015625,
      "learning_rate": 8.474576271186441e-05,
      "loss": 1.13,
      "step": 250
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.8828125,
      "learning_rate": 8.644067796610171e-05,
      "loss": 1.153,
      "step": 255
    },
    {
      "epoch": 0.03,
      "grad_norm": 3.953125,
      "learning_rate": 8.813559322033899e-05,
      "loss": 1.047,
      "step": 260
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.1640625,
      "learning_rate": 8.983050847457629e-05,
      "loss": 1.0913,
      "step": 265
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.92578125,
      "learning_rate": 9.152542372881357e-05,
      "loss": 1.2967,
      "step": 270
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.046875,
      "learning_rate": 9.322033898305085e-05,
      "loss": 1.0959,
      "step": 275
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.546875,
      "learning_rate": 9.491525423728815e-05,
      "loss": 1.2349,
      "step": 280
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.953125,
      "learning_rate": 9.661016949152543e-05,
      "loss": 1.1364,
      "step": 285
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.4765625,
      "learning_rate": 9.830508474576272e-05,
      "loss": 1.0485,
      "step": 290
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0001,
      "loss": 1.1364,
      "step": 295
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.453125,
      "learning_rate": 9.999993199499282e-05,
      "loss": 1.1016,
      "step": 300
    },
    {
      "epoch": 0.03,
      "grad_norm": 3.1875,
      "learning_rate": 9.999972798015626e-05,
      "loss": 1.31,
      "step": 305
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.0625,
      "learning_rate": 9.999938795604529e-05,
      "loss": 1.3549,
      "step": 310
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.828125,
      "learning_rate": 9.999891192358482e-05,
      "loss": 1.004,
      "step": 315
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.7734375,
      "learning_rate": 9.999829988406979e-05,
      "loss": 1.0005,
      "step": 320
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.3515625,
      "learning_rate": 9.999755183916503e-05,
      "loss": 1.1688,
      "step": 325
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.99609375,
      "learning_rate": 9.999666779090541e-05,
      "loss": 1.0039,
      "step": 330
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.8359375,
      "learning_rate": 9.99956477416957e-05,
      "loss": 1.2665,
      "step": 335
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.91796875,
      "learning_rate": 9.999449169431064e-05,
      "loss": 1.1508,
      "step": 340
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.734375,
      "learning_rate": 9.99931996518949e-05,
      "loss": 1.04,
      "step": 345
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.0390625,
      "learning_rate": 9.99917716179631e-05,
      "loss": 1.1615,
      "step": 350
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.77734375,
      "learning_rate": 9.999020759639978e-05,
      "loss": 1.098,
      "step": 355
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.421875,
      "learning_rate": 9.998850759145942e-05,
      "loss": 1.1783,
      "step": 360
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.453125,
      "learning_rate": 9.998667160776634e-05,
      "loss": 1.1064,
      "step": 365
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.1015625,
      "learning_rate": 9.998469965031477e-05,
      "loss": 1.1819,
      "step": 370
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.5859375,
      "learning_rate": 9.998259172446887e-05,
      "loss": 1.0785,
      "step": 375
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.5859375,
      "learning_rate": 9.998034783596259e-05,
      "loss": 1.2747,
      "step": 380
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.8828125,
      "learning_rate": 9.997796799089978e-05,
      "loss": 1.1299,
      "step": 385
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.83984375,
      "learning_rate": 9.997545219575406e-05,
      "loss": 1.2208,
      "step": 390
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.7890625,
      "learning_rate": 9.997280045736894e-05,
      "loss": 1.1041,
      "step": 395
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.8671875,
      "learning_rate": 9.997001278295766e-05,
      "loss": 1.188,
      "step": 400
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.90625,
      "learning_rate": 9.996708918010323e-05,
      "loss": 1.1673,
      "step": 405
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.87890625,
      "learning_rate": 9.996402965675849e-05,
      "loss": 1.1214,
      "step": 410
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.0625,
      "learning_rate": 9.996083422124589e-05,
      "loss": 1.038,
      "step": 415
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.87109375,
      "learning_rate": 9.995750288225769e-05,
      "loss": 1.09,
      "step": 420
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.94921875,
      "learning_rate": 9.995403564885581e-05,
      "loss": 0.9791,
      "step": 425
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.78125,
      "learning_rate": 9.995043253047181e-05,
      "loss": 1.1455,
      "step": 430
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.77734375,
      "learning_rate": 9.99466935369069e-05,
      "loss": 1.2065,
      "step": 435
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.984375,
      "learning_rate": 9.994281867833185e-05,
      "loss": 1.0559,
      "step": 440
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.8984375,
      "learning_rate": 9.99388079652871e-05,
      "loss": 0.9766,
      "step": 445
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.40625,
      "learning_rate": 9.993466140868258e-05,
      "loss": 1.1912,
      "step": 450
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.93359375,
      "learning_rate": 9.993037901979773e-05,
      "loss": 1.1181,
      "step": 455
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.671875,
      "learning_rate": 9.992596081028153e-05,
      "loss": 1.1538,
      "step": 460
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.0234375,
      "learning_rate": 9.992140679215241e-05,
      "loss": 1.1806,
      "step": 465
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.03125,
      "learning_rate": 9.991671697779817e-05,
      "loss": 1.1783,
      "step": 470
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.1953125,
      "learning_rate": 9.991189137997607e-05,
      "loss": 1.1097,
      "step": 475
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.8984375,
      "learning_rate": 9.990693001181271e-05,
      "loss": 1.1734,
      "step": 480
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.859375,
      "learning_rate": 9.990183288680399e-05,
      "loss": 0.9929,
      "step": 485
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.82421875,
      "learning_rate": 9.989660001881512e-05,
      "loss": 1.1162,
      "step": 490
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.79296875,
      "learning_rate": 9.989123142208052e-05,
      "loss": 0.9353,
      "step": 495
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.8984375,
      "learning_rate": 9.988572711120388e-05,
      "loss": 1.1636,
      "step": 500
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.03125,
      "learning_rate": 9.988008710115803e-05,
      "loss": 1.1438,
      "step": 505
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.03125,
      "learning_rate": 9.987431140728491e-05,
      "loss": 1.1095,
      "step": 510
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.84375,
      "learning_rate": 9.986840004529558e-05,
      "loss": 1.1126,
      "step": 515
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.015625,
      "learning_rate": 9.986235303127011e-05,
      "loss": 1.041,
      "step": 520
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.82421875,
      "learning_rate": 9.98561703816576e-05,
      "loss": 1.2251,
      "step": 525
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.91796875,
      "learning_rate": 9.984985211327611e-05,
      "loss": 1.5059,
      "step": 530
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.94921875,
      "learning_rate": 9.984339824331255e-05,
      "loss": 1.0779,
      "step": 535
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.015625,
      "learning_rate": 9.98368087893228e-05,
      "loss": 0.9696,
      "step": 540
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.81640625,
      "learning_rate": 9.983008376923144e-05,
      "loss": 1.1885,
      "step": 545
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.74609375,
      "learning_rate": 9.982322320133191e-05,
      "loss": 0.9718,
      "step": 550
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.734375,
      "learning_rate": 9.98162271042863e-05,
      "loss": 1.0219,
      "step": 555
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.75,
      "learning_rate": 9.98090954971254e-05,
      "loss": 0.9908,
      "step": 560
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.96875,
      "learning_rate": 9.980182839924864e-05,
      "loss": 1.2804,
      "step": 565
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.7109375,
      "learning_rate": 9.979442583042396e-05,
      "loss": 1.0956,
      "step": 570
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.796875,
      "learning_rate": 9.978688781078783e-05,
      "loss": 1.1121,
      "step": 575
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.9140625,
      "learning_rate": 9.977921436084516e-05,
      "loss": 1.1813,
      "step": 580
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.0625,
      "learning_rate": 9.977140550146928e-05,
      "loss": 1.1323,
      "step": 585
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.859375,
      "learning_rate": 9.976346125390187e-05,
      "loss": 1.0552,
      "step": 590
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.78515625,
      "learning_rate": 9.975538163975285e-05,
      "loss": 1.0293,
      "step": 595
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.81640625,
      "learning_rate": 9.974716668100041e-05,
      "loss": 1.0693,
      "step": 600
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.78515625,
      "learning_rate": 9.973881639999086e-05,
      "loss": 1.055,
      "step": 605
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.796875,
      "learning_rate": 9.973033081943865e-05,
      "loss": 1.1065,
      "step": 610
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.1171875,
      "learning_rate": 9.972170996242627e-05,
      "loss": 1.1045,
      "step": 615
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.921875,
      "learning_rate": 9.971295385240415e-05,
      "loss": 1.1502,
      "step": 620
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.0,
      "learning_rate": 9.97040625131907e-05,
      "loss": 1.1024,
      "step": 625
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.6796875,
      "learning_rate": 9.96950359689721e-05,
      "loss": 1.0256,
      "step": 630
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.83203125,
      "learning_rate": 9.968587424430239e-05,
      "loss": 1.1097,
      "step": 635
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.74609375,
      "learning_rate": 9.967657736410329e-05,
      "loss": 1.1827,
      "step": 640
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.76953125,
      "learning_rate": 9.966714535366416e-05,
      "loss": 0.9866,
      "step": 645
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.640625,
      "learning_rate": 9.965757823864197e-05,
      "loss": 1.047,
      "step": 650
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.75390625,
      "learning_rate": 9.964787604506119e-05,
      "loss": 1.0826,
      "step": 655
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.7890625,
      "learning_rate": 9.963803879931372e-05,
      "loss": 1.0919,
      "step": 660
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.0078125,
      "learning_rate": 9.962806652815887e-05,
      "loss": 1.138,
      "step": 665
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.8203125,
      "learning_rate": 9.961795925872319e-05,
      "loss": 1.3509,
      "step": 670
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.7265625,
      "learning_rate": 9.960771701850046e-05,
      "loss": 1.2172,
      "step": 675
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.8046875,
      "learning_rate": 9.959733983535166e-05,
      "loss": 1.0813,
      "step": 680
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.80078125,
      "learning_rate": 9.958682773750478e-05,
      "loss": 0.9785,
      "step": 685
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.8515625,
      "learning_rate": 9.957618075355483e-05,
      "loss": 1.1987,
      "step": 690
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.85546875,
      "learning_rate": 9.956539891246378e-05,
      "loss": 1.1459,
      "step": 695
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.87890625,
      "learning_rate": 9.955448224356035e-05,
      "loss": 1.1929,
      "step": 700
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.609375,
      "learning_rate": 9.95434307765401e-05,
      "loss": 1.1577,
      "step": 705
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.6640625,
      "learning_rate": 9.95322445414652e-05,
      "loss": 0.9736,
      "step": 710
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.9375,
      "learning_rate": 9.952092356876447e-05,
      "loss": 1.0244,
      "step": 715
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.828125,
      "learning_rate": 9.950946788923323e-05,
      "loss": 1.0721,
      "step": 720
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.83984375,
      "learning_rate": 9.94978775340332e-05,
      "loss": 1.0923,
      "step": 725
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.84375,
      "learning_rate": 9.94861525346925e-05,
      "loss": 1.1278,
      "step": 730
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.765625,
      "learning_rate": 9.947429292310544e-05,
      "loss": 1.1433,
      "step": 735
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.8671875,
      "learning_rate": 9.946229873153257e-05,
      "loss": 1.0413,
      "step": 740
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.80859375,
      "learning_rate": 9.945016999260046e-05,
      "loss": 0.9796,
      "step": 745
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.75,
      "learning_rate": 9.943790673930174e-05,
      "loss": 1.111,
      "step": 750
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.71875,
      "learning_rate": 9.942550900499492e-05,
      "loss": 0.9821,
      "step": 755
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.76171875,
      "learning_rate": 9.941297682340428e-05,
      "loss": 1.1518,
      "step": 760
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.125,
      "learning_rate": 9.940031022861991e-05,
      "loss": 1.2209,
      "step": 765
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.77734375,
      "learning_rate": 9.938750925509744e-05,
      "loss": 1.141,
      "step": 770
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.71484375,
      "learning_rate": 9.937457393765812e-05,
      "loss": 1.1641,
      "step": 775
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.8515625,
      "learning_rate": 9.936150431148858e-05,
      "loss": 1.1908,
      "step": 780
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.76171875,
      "learning_rate": 9.934830041214085e-05,
      "loss": 1.1927,
      "step": 785
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.76953125,
      "learning_rate": 9.933496227553213e-05,
      "loss": 1.1013,
      "step": 790
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7734375,
      "learning_rate": 9.932148993794489e-05,
      "loss": 1.1508,
      "step": 795
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.80078125,
      "learning_rate": 9.930788343602653e-05,
      "loss": 1.0958,
      "step": 800
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.73828125,
      "learning_rate": 9.929414280678948e-05,
      "loss": 1.0411,
      "step": 805
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.81640625,
      "learning_rate": 9.928026808761101e-05,
      "loss": 1.135,
      "step": 810
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.83984375,
      "learning_rate": 9.926625931623314e-05,
      "loss": 1.149,
      "step": 815
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.703125,
      "learning_rate": 9.92521165307625e-05,
      "loss": 1.1497,
      "step": 820
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.703125,
      "learning_rate": 9.923783976967033e-05,
      "loss": 1.0114,
      "step": 825
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.83984375,
      "learning_rate": 9.922342907179229e-05,
      "loss": 1.2451,
      "step": 830
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.84375,
      "learning_rate": 9.920888447632834e-05,
      "loss": 1.1813,
      "step": 835
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.734375,
      "learning_rate": 9.919420602284268e-05,
      "loss": 1.0433,
      "step": 840
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.71484375,
      "learning_rate": 9.917939375126368e-05,
      "loss": 1.1203,
      "step": 845
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.90234375,
      "learning_rate": 9.916444770188366e-05,
      "loss": 1.0648,
      "step": 850
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.83203125,
      "learning_rate": 9.914936791535889e-05,
      "loss": 1.1393,
      "step": 855
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.78125,
      "learning_rate": 9.91341544327094e-05,
      "loss": 1.0386,
      "step": 860
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.859375,
      "learning_rate": 9.911880729531889e-05,
      "loss": 1.1063,
      "step": 865
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.89453125,
      "learning_rate": 9.910332654493466e-05,
      "loss": 1.0915,
      "step": 870
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.73828125,
      "learning_rate": 9.908771222366746e-05,
      "loss": 1.056,
      "step": 875
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.7421875,
      "learning_rate": 9.907196437399138e-05,
      "loss": 1.0099,
      "step": 880
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.86328125,
      "learning_rate": 9.90560830387437e-05,
      "loss": 0.9894,
      "step": 885
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.82421875,
      "learning_rate": 9.904006826112483e-05,
      "loss": 1.0191,
      "step": 890
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.671875,
      "learning_rate": 9.902392008469821e-05,
      "loss": 1.0701,
      "step": 895
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.703125,
      "learning_rate": 9.900763855339008e-05,
      "loss": 1.25,
      "step": 900
    },
    {
      "epoch": 0.09,
      "grad_norm": 5.71875,
      "learning_rate": 9.899122371148948e-05,
      "loss": 1.0782,
      "step": 905
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.80859375,
      "learning_rate": 9.897467560364806e-05,
      "loss": 1.1018,
      "step": 910
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.77734375,
      "learning_rate": 9.895799427487999e-05,
      "loss": 1.0424,
      "step": 915
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.421875,
      "learning_rate": 9.894117977056185e-05,
      "loss": 1.062,
      "step": 920
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.82421875,
      "learning_rate": 9.892423213643242e-05,
      "loss": 1.156,
      "step": 925
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.9921875,
      "learning_rate": 9.890715141859268e-05,
      "loss": 1.0613,
      "step": 930
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.8125,
      "learning_rate": 9.888993766350559e-05,
      "loss": 1.079,
      "step": 935
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.73046875,
      "learning_rate": 9.887259091799604e-05,
      "loss": 1.129,
      "step": 940
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.95703125,
      "learning_rate": 9.885511122925062e-05,
      "loss": 1.232,
      "step": 945
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.71484375,
      "learning_rate": 9.883749864481759e-05,
      "loss": 1.0411,
      "step": 950
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.7109375,
      "learning_rate": 9.881975321260672e-05,
      "loss": 0.9664,
      "step": 955
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.76953125,
      "learning_rate": 9.880187498088915e-05,
      "loss": 1.2187,
      "step": 960
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.7890625,
      "learning_rate": 9.878386399829723e-05,
      "loss": 1.1002,
      "step": 965
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.765625,
      "learning_rate": 9.876572031382445e-05,
      "loss": 0.9465,
      "step": 970
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.8828125,
      "learning_rate": 9.874744397682524e-05,
      "loss": 1.1099,
      "step": 975
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.78125,
      "learning_rate": 9.872903503701494e-05,
      "loss": 1.0486,
      "step": 980
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.71875,
      "learning_rate": 9.871049354446953e-05,
      "loss": 1.0597,
      "step": 985
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.6328125,
      "learning_rate": 9.869181954962559e-05,
      "loss": 1.0353,
      "step": 990
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.98046875,
      "learning_rate": 9.867301310328011e-05,
      "loss": 1.0108,
      "step": 995
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.80859375,
      "learning_rate": 9.865407425659041e-05,
      "loss": 1.1462,
      "step": 1000
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.77734375,
      "learning_rate": 9.863500306107394e-05,
      "loss": 1.0011,
      "step": 1005
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.921875,
      "learning_rate": 9.861579956860816e-05,
      "loss": 1.1095,
      "step": 1010
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.0,
      "learning_rate": 9.859646383143043e-05,
      "loss": 1.2169,
      "step": 1015
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.74609375,
      "learning_rate": 9.857699590213783e-05,
      "loss": 1.0282,
      "step": 1020
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.953125,
      "learning_rate": 9.855739583368702e-05,
      "loss": 1.3305,
      "step": 1025
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.7109375,
      "learning_rate": 9.85376636793941e-05,
      "loss": 1.1528,
      "step": 1030
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.60546875,
      "learning_rate": 9.851779949293452e-05,
      "loss": 0.9676,
      "step": 1035
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.88671875,
      "learning_rate": 9.84978033283428e-05,
      "loss": 1.1296,
      "step": 1040
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.80859375,
      "learning_rate": 9.847767524001252e-05,
      "loss": 1.09,
      "step": 1045
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.7890625,
      "learning_rate": 9.845741528269614e-05,
      "loss": 1.11,
      "step": 1050
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.75,
      "learning_rate": 9.843702351150479e-05,
      "loss": 1.1153,
      "step": 1055
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.72265625,
      "learning_rate": 9.841649998190817e-05,
      "loss": 1.2189,
      "step": 1060
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.75390625,
      "learning_rate": 9.839584474973438e-05,
      "loss": 1.1517,
      "step": 1065
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.7265625,
      "learning_rate": 9.83750578711698e-05,
      "loss": 0.9702,
      "step": 1070
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.75,
      "learning_rate": 9.835413940275891e-05,
      "loss": 0.9542,
      "step": 1075
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.8359375,
      "learning_rate": 9.833308940140411e-05,
      "loss": 1.1009,
      "step": 1080
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.265625,
      "learning_rate": 9.831190792436564e-05,
      "loss": 1.1224,
      "step": 1085
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.80859375,
      "learning_rate": 9.829059502926137e-05,
      "loss": 0.9884,
      "step": 1090
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.73046875,
      "learning_rate": 9.82691507740666e-05,
      "loss": 1.0683,
      "step": 1095
    },
    {
      "epoch": 0.11,
      "grad_norm": 2.15625,
      "learning_rate": 9.824757521711405e-05,
      "loss": 1.1777,
      "step": 1100
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.83203125,
      "learning_rate": 9.822586841709351e-05,
      "loss": 1.1444,
      "step": 1105
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.74609375,
      "learning_rate": 9.820403043305187e-05,
      "loss": 1.0494,
      "step": 1110
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.75,
      "learning_rate": 9.818206132439278e-05,
      "loss": 0.9759,
      "step": 1115
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.78515625,
      "learning_rate": 9.815996115087664e-05,
      "loss": 1.0381,
      "step": 1120
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.88671875,
      "learning_rate": 9.813772997262034e-05,
      "loss": 1.1601,
      "step": 1125
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.0625,
      "learning_rate": 9.811536785009714e-05,
      "loss": 1.1123,
      "step": 1130
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.8515625,
      "learning_rate": 9.809287484413649e-05,
      "loss": 0.9873,
      "step": 1135
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.69140625,
      "learning_rate": 9.807025101592387e-05,
      "loss": 0.9711,
      "step": 1140
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.7734375,
      "learning_rate": 9.804749642700061e-05,
      "loss": 1.142,
      "step": 1145
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.7890625,
      "learning_rate": 9.802461113926378e-05,
      "loss": 1.1325,
      "step": 1150
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.5703125,
      "learning_rate": 9.800159521496595e-05,
      "loss": 1.0743,
      "step": 1155
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.83984375,
      "learning_rate": 9.7978448716715e-05,
      "loss": 1.0507,
      "step": 1160
    },
    {
      "epoch": 0.12,
      "grad_norm": 2.375,
      "learning_rate": 9.795517170747407e-05,
      "loss": 1.0658,
      "step": 1165
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.71875,
      "learning_rate": 9.793176425056128e-05,
      "loss": 1.0853,
      "step": 1170
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.81640625,
      "learning_rate": 9.790822640964961e-05,
      "loss": 1.1156,
      "step": 1175
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.95703125,
      "learning_rate": 9.788455824876671e-05,
      "loss": 1.2031,
      "step": 1180
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.6953125,
      "learning_rate": 9.786075983229469e-05,
      "loss": 0.9827,
      "step": 1185
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.7578125,
      "learning_rate": 9.783683122497003e-05,
      "loss": 1.1374,
      "step": 1190
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.765625,
      "learning_rate": 9.781277249188332e-05,
      "loss": 1.0057,
      "step": 1195
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.8203125,
      "learning_rate": 9.778858369847914e-05,
      "loss": 1.1041,
      "step": 1200
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.6640625,
      "learning_rate": 9.776426491055587e-05,
      "loss": 1.1098,
      "step": 1205
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.734375,
      "learning_rate": 9.773981619426546e-05,
      "loss": 0.9285,
      "step": 1210
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.78125,
      "learning_rate": 9.771523761611332e-05,
      "loss": 1.228,
      "step": 1215
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.77734375,
      "learning_rate": 9.76905292429581e-05,
      "loss": 1.005,
      "step": 1220
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.60546875,
      "learning_rate": 9.766569114201154e-05,
      "loss": 0.9875,
      "step": 1225
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.80859375,
      "learning_rate": 9.764072338083826e-05,
      "loss": 1.0508,
      "step": 1230
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.90625,
      "learning_rate": 9.761562602735552e-05,
      "loss": 1.2457,
      "step": 1235
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.78515625,
      "learning_rate": 9.759039914983319e-05,
      "loss": 0.8978,
      "step": 1240
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.73046875,
      "learning_rate": 9.756504281689343e-05,
      "loss": 1.0165,
      "step": 1245
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.8125,
      "learning_rate": 9.753955709751053e-05,
      "loss": 1.3444,
      "step": 1250
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.78125,
      "learning_rate": 9.751394206101074e-05,
      "loss": 1.1455,
      "step": 1255
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.75390625,
      "learning_rate": 9.748819777707212e-05,
      "loss": 1.0634,
      "step": 1260
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.91796875,
      "learning_rate": 9.746232431572426e-05,
      "loss": 1.2258,
      "step": 1265
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.80859375,
      "learning_rate": 9.743632174734816e-05,
      "loss": 1.0882,
      "step": 1270
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.546875,
      "learning_rate": 9.741019014267601e-05,
      "loss": 1.1057,
      "step": 1275
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6796875,
      "learning_rate": 9.738392957279103e-05,
      "loss": 1.1838,
      "step": 1280
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.73828125,
      "learning_rate": 9.735754010912719e-05,
      "loss": 1.0147,
      "step": 1285
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.66796875,
      "learning_rate": 9.733102182346915e-05,
      "loss": 0.979,
      "step": 1290
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.83203125,
      "learning_rate": 9.730437478795195e-05,
      "loss": 1.1407,
      "step": 1295
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.75390625,
      "learning_rate": 9.727759907506085e-05,
      "loss": 1.0828,
      "step": 1300
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.74609375,
      "learning_rate": 9.725069475763118e-05,
      "loss": 0.9176,
      "step": 1305
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.82421875,
      "learning_rate": 9.722366190884805e-05,
      "loss": 0.9957,
      "step": 1310
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.75,
      "learning_rate": 9.719650060224622e-05,
      "loss": 1.0143,
      "step": 1315
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.734375,
      "learning_rate": 9.71692109117099e-05,
      "loss": 1.0765,
      "step": 1320
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.76953125,
      "learning_rate": 9.71417929114725e-05,
      "loss": 1.1173,
      "step": 1325
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.8984375,
      "learning_rate": 9.711424667611648e-05,
      "loss": 1.078,
      "step": 1330
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.7890625,
      "learning_rate": 9.708657228057312e-05,
      "loss": 1.0597,
      "step": 1335
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.765625,
      "learning_rate": 9.70587698001223e-05,
      "loss": 1.0769,
      "step": 1340
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.65234375,
      "learning_rate": 9.703083931039236e-05,
      "loss": 1.0842,
      "step": 1345
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.77734375,
      "learning_rate": 9.700278088735982e-05,
      "loss": 1.1953,
      "step": 1350
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.68359375,
      "learning_rate": 9.697459460734921e-05,
      "loss": 0.9097,
      "step": 1355
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.77734375,
      "learning_rate": 9.694628054703285e-05,
      "loss": 1.0923,
      "step": 1360
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.75390625,
      "learning_rate": 9.691783878343063e-05,
      "loss": 0.996,
      "step": 1365
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.3359375,
      "learning_rate": 9.68892693939099e-05,
      "loss": 0.946,
      "step": 1370
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.70703125,
      "learning_rate": 9.68605724561851e-05,
      "loss": 0.892,
      "step": 1375
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.65625,
      "learning_rate": 9.683174804831763e-05,
      "loss": 1.0465,
      "step": 1380
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.0859375,
      "learning_rate": 9.680279624871567e-05,
      "loss": 1.1461,
      "step": 1385
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.76171875,
      "learning_rate": 9.67737171361339e-05,
      "loss": 1.0644,
      "step": 1390
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.83203125,
      "learning_rate": 9.674451078967334e-05,
      "loss": 1.1142,
      "step": 1395
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.7890625,
      "learning_rate": 9.67151772887811e-05,
      "loss": 0.9974,
      "step": 1400
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.859375,
      "learning_rate": 9.668571671325018e-05,
      "loss": 1.3387,
      "step": 1405
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.7734375,
      "learning_rate": 9.665612914321925e-05,
      "loss": 0.9411,
      "step": 1410
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.7421875,
      "learning_rate": 9.66264146591724e-05,
      "loss": 1.0943,
      "step": 1415
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.8125,
      "learning_rate": 9.659657334193901e-05,
      "loss": 1.1811,
      "step": 1420
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.77734375,
      "learning_rate": 9.656660527269344e-05,
      "loss": 1.3589,
      "step": 1425
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.8046875,
      "learning_rate": 9.65365105329548e-05,
      "loss": 1.1464,
      "step": 1430
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.83203125,
      "learning_rate": 9.650628920458686e-05,
      "loss": 1.0347,
      "step": 1435
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.73828125,
      "learning_rate": 9.647594136979768e-05,
      "loss": 1.0573,
      "step": 1440
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.8046875,
      "learning_rate": 9.64454671111394e-05,
      "loss": 1.162,
      "step": 1445
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.75390625,
      "learning_rate": 9.641486651150815e-05,
      "loss": 1.03,
      "step": 1450
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.89453125,
      "learning_rate": 9.63841396541437e-05,
      "loss": 1.0686,
      "step": 1455
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.9765625,
      "learning_rate": 9.635328662262921e-05,
      "loss": 1.2671,
      "step": 1460
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.65625,
      "learning_rate": 9.632230750089114e-05,
      "loss": 1.0885,
      "step": 1465
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.73828125,
      "learning_rate": 9.62912023731989e-05,
      "loss": 0.9341,
      "step": 1470
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.73046875,
      "learning_rate": 9.625997132416467e-05,
      "loss": 1.1212,
      "step": 1475
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.72265625,
      "learning_rate": 9.622861443874314e-05,
      "loss": 1.0372,
      "step": 1480
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.86328125,
      "learning_rate": 9.619713180223135e-05,
      "loss": 1.1321,
      "step": 1485
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.9453125,
      "learning_rate": 9.616552350026835e-05,
      "loss": 1.179,
      "step": 1490
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.71875,
      "learning_rate": 9.613378961883506e-05,
      "loss": 1.1647,
      "step": 1495
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.69921875,
      "learning_rate": 9.610193024425399e-05,
      "loss": 1.0691,
      "step": 1500
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.796875,
      "learning_rate": 9.606994546318904e-05,
      "loss": 1.169,
      "step": 1505
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.921875,
      "learning_rate": 9.603783536264519e-05,
      "loss": 1.156,
      "step": 1510
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.66796875,
      "learning_rate": 9.600560002996837e-05,
      "loss": 1.1447,
      "step": 1515
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.76171875,
      "learning_rate": 9.597323955284511e-05,
      "loss": 1.0639,
      "step": 1520
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.71484375,
      "learning_rate": 9.594075401930242e-05,
      "loss": 1.0137,
      "step": 1525
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.72265625,
      "learning_rate": 9.590814351770747e-05,
      "loss": 1.1508,
      "step": 1530
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.7109375,
      "learning_rate": 9.587540813676732e-05,
      "loss": 1.2712,
      "step": 1535
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.2578125,
      "learning_rate": 9.584254796552877e-05,
      "loss": 1.2424,
      "step": 1540
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.69921875,
      "learning_rate": 9.580956309337807e-05,
      "loss": 1.0838,
      "step": 1545
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.70703125,
      "learning_rate": 9.577645361004068e-05,
      "loss": 1.1275,
      "step": 1550
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.82421875,
      "learning_rate": 9.574321960558104e-05,
      "loss": 1.1691,
      "step": 1555
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.734375,
      "learning_rate": 9.57098611704023e-05,
      "loss": 0.967,
      "step": 1560
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.78515625,
      "learning_rate": 9.567637839524604e-05,
      "loss": 1.0619,
      "step": 1565
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.75390625,
      "learning_rate": 9.564277137119217e-05,
      "loss": 1.0978,
      "step": 1570
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.83984375,
      "learning_rate": 9.56090401896585e-05,
      "loss": 1.1419,
      "step": 1575
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.69921875,
      "learning_rate": 9.557518494240059e-05,
      "loss": 1.0186,
      "step": 1580
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.7734375,
      "learning_rate": 9.554120572151152e-05,
      "loss": 1.0404,
      "step": 1585
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.796875,
      "learning_rate": 9.550710261942155e-05,
      "loss": 1.0433,
      "step": 1590
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.80078125,
      "learning_rate": 9.547287572889796e-05,
      "loss": 0.9927,
      "step": 1595
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.828125,
      "learning_rate": 9.543852514304476e-05,
      "loss": 1.1879,
      "step": 1600
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.0078125,
      "learning_rate": 9.54040509553024e-05,
      "loss": 1.1554,
      "step": 1605
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.7734375,
      "learning_rate": 9.53694532594476e-05,
      "loss": 1.1748,
      "step": 1610
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6640625,
      "learning_rate": 9.5334732149593e-05,
      "loss": 1.0011,
      "step": 1615
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.78515625,
      "learning_rate": 9.529988772018699e-05,
      "loss": 1.1982,
      "step": 1620
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.859375,
      "learning_rate": 9.526492006601337e-05,
      "loss": 0.9936,
      "step": 1625
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.2890625,
      "learning_rate": 9.52298292821912e-05,
      "loss": 1.1356,
      "step": 1630
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.75390625,
      "learning_rate": 9.519461546417441e-05,
      "loss": 0.9811,
      "step": 1635
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6875,
      "learning_rate": 9.515927870775165e-05,
      "loss": 1.1542,
      "step": 1640
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.7265625,
      "learning_rate": 9.5123819109046e-05,
      "loss": 1.0644,
      "step": 1645
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.7109375,
      "learning_rate": 9.508823676451462e-05,
      "loss": 1.1402,
      "step": 1650
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.7109375,
      "learning_rate": 9.505253177094864e-05,
      "loss": 1.076,
      "step": 1655
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.1640625,
      "learning_rate": 9.50167042254728e-05,
      "loss": 1.1421,
      "step": 1660
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.1953125,
      "learning_rate": 9.49807542255452e-05,
      "loss": 1.1005,
      "step": 1665
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.609375,
      "learning_rate": 9.494468186895701e-05,
      "loss": 1.1378,
      "step": 1670
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.875,
      "learning_rate": 9.49084872538323e-05,
      "loss": 1.0166,
      "step": 1675
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.71875,
      "learning_rate": 9.487217047862767e-05,
      "loss": 1.068,
      "step": 1680
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.734375,
      "learning_rate": 9.4835731642132e-05,
      "loss": 1.0219,
      "step": 1685
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.81640625,
      "learning_rate": 9.479917084346623e-05,
      "loss": 1.2396,
      "step": 1690
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.0859375,
      "learning_rate": 9.476248818208308e-05,
      "loss": 1.0547,
      "step": 1695
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.93359375,
      "learning_rate": 9.472568375776669e-05,
      "loss": 1.2844,
      "step": 1700
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.765625,
      "learning_rate": 9.468875767063252e-05,
      "loss": 1.0235,
      "step": 1705
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.84765625,
      "learning_rate": 9.465171002112688e-05,
      "loss": 1.0473,
      "step": 1710
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.78125,
      "learning_rate": 9.461454091002681e-05,
      "loss": 1.091,
      "step": 1715
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.81640625,
      "learning_rate": 9.457725043843975e-05,
      "loss": 0.9108,
      "step": 1720
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.828125,
      "learning_rate": 9.453983870780322e-05,
      "loss": 0.9136,
      "step": 1725
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.8046875,
      "learning_rate": 9.450230581988466e-05,
      "loss": 1.0787,
      "step": 1730
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.75390625,
      "learning_rate": 9.446465187678103e-05,
      "loss": 1.2171,
      "step": 1735
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6875,
      "learning_rate": 9.442687698091857e-05,
      "loss": 1.1358,
      "step": 1740
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.73046875,
      "learning_rate": 9.43889812350526e-05,
      "loss": 1.0922,
      "step": 1745
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.8359375,
      "learning_rate": 9.43509647422671e-05,
      "loss": 1.1919,
      "step": 1750
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.79296875,
      "learning_rate": 9.431282760597458e-05,
      "loss": 1.1724,
      "step": 1755
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.921875,
      "learning_rate": 9.427456992991568e-05,
      "loss": 1.0825,
      "step": 1760
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.578125,
      "learning_rate": 9.423619181815893e-05,
      "loss": 0.9497,
      "step": 1765
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.69140625,
      "learning_rate": 9.419769337510048e-05,
      "loss": 1.0628,
      "step": 1770
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.765625,
      "learning_rate": 9.415907470546382e-05,
      "loss": 0.9418,
      "step": 1775
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.71484375,
      "learning_rate": 9.412033591429947e-05,
      "loss": 0.9248,
      "step": 1780
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.85546875,
      "learning_rate": 9.408147710698467e-05,
      "loss": 1.201,
      "step": 1785
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.73046875,
      "learning_rate": 9.40424983892232e-05,
      "loss": 1.0609,
      "step": 1790
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.4609375,
      "learning_rate": 9.400339986704494e-05,
      "loss": 1.1425,
      "step": 1795
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.71484375,
      "learning_rate": 9.396418164680572e-05,
      "loss": 1.1577,
      "step": 1800
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.76171875,
      "learning_rate": 9.392484383518696e-05,
      "loss": 0.8667,
      "step": 1805
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.8671875,
      "learning_rate": 9.388538653919539e-05,
      "loss": 1.2117,
      "step": 1810
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.88671875,
      "learning_rate": 9.384580986616273e-05,
      "loss": 1.155,
      "step": 1815
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.84375,
      "learning_rate": 9.380611392374548e-05,
      "loss": 1.2536,
      "step": 1820
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.7109375,
      "learning_rate": 9.376629881992454e-05,
      "loss": 1.075,
      "step": 1825
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.7578125,
      "learning_rate": 9.3726364663005e-05,
      "loss": 1.1615,
      "step": 1830
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.734375,
      "learning_rate": 9.368631156161571e-05,
      "loss": 0.97,
      "step": 1835
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.76171875,
      "learning_rate": 9.364613962470918e-05,
      "loss": 0.9049,
      "step": 1840
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.70703125,
      "learning_rate": 9.360584896156111e-05,
      "loss": 1.0722,
      "step": 1845
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.69140625,
      "learning_rate": 9.356543968177015e-05,
      "loss": 1.0683,
      "step": 1850
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.734375,
      "learning_rate": 9.352491189525766e-05,
      "loss": 1.0876,
      "step": 1855
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.83984375,
      "learning_rate": 9.348426571226732e-05,
      "loss": 1.1982,
      "step": 1860
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.6171875,
      "learning_rate": 9.344350124336491e-05,
      "loss": 1.0043,
      "step": 1865
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.75390625,
      "learning_rate": 9.340261859943793e-05,
      "loss": 1.1613,
      "step": 1870
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.875,
      "learning_rate": 9.336161789169536e-05,
      "loss": 1.155,
      "step": 1875
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.8984375,
      "learning_rate": 9.332049923166735e-05,
      "loss": 1.3299,
      "step": 1880
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.62109375,
      "learning_rate": 9.327926273120487e-05,
      "loss": 0.9898,
      "step": 1885
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.0859375,
      "learning_rate": 9.323790850247949e-05,
      "loss": 1.0147,
      "step": 1890
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.89453125,
      "learning_rate": 9.319643665798297e-05,
      "loss": 1.123,
      "step": 1895
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.734375,
      "learning_rate": 9.315484731052706e-05,
      "loss": 1.0347,
      "step": 1900
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.7109375,
      "learning_rate": 9.311314057324307e-05,
      "loss": 1.0325,
      "step": 1905
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.87890625,
      "learning_rate": 9.307131655958173e-05,
      "loss": 1.0965,
      "step": 1910
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.7265625,
      "learning_rate": 9.302937538331271e-05,
      "loss": 0.9937,
      "step": 1915
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.80078125,
      "learning_rate": 9.29873171585244e-05,
      "loss": 1.0982,
      "step": 1920
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.67578125,
      "learning_rate": 9.294514199962359e-05,
      "loss": 0.9308,
      "step": 1925
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.71875,
      "learning_rate": 9.29028500213352e-05,
      "loss": 1.0217,
      "step": 1930
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.6484375,
      "learning_rate": 9.286044133870185e-05,
      "loss": 1.0372,
      "step": 1935
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.76171875,
      "learning_rate": 9.281791606708366e-05,
      "loss": 1.0679,
      "step": 1940
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.83984375,
      "learning_rate": 9.277527432215786e-05,
      "loss": 1.054,
      "step": 1945
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.98828125,
      "learning_rate": 9.273251621991858e-05,
      "loss": 1.1235,
      "step": 1950
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.75,
      "learning_rate": 9.26896418766764e-05,
      "loss": 1.1428,
      "step": 1955
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.73828125,
      "learning_rate": 9.264665140905813e-05,
      "loss": 1.0621,
      "step": 1960
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.84375,
      "learning_rate": 9.260354493400643e-05,
      "loss": 0.9373,
      "step": 1965
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.9765625,
      "learning_rate": 9.256032256877956e-05,
      "loss": 1.0627,
      "step": 1970
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.734375,
      "learning_rate": 9.251698443095102e-05,
      "loss": 0.9658,
      "step": 1975
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.81640625,
      "learning_rate": 9.24735306384092e-05,
      "loss": 1.0197,
      "step": 1980
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.6875,
      "learning_rate": 9.242996130935716e-05,
      "loss": 1.0304,
      "step": 1985
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.7109375,
      "learning_rate": 9.238627656231215e-05,
      "loss": 1.0051,
      "step": 1990
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.9296875,
      "learning_rate": 9.234247651610547e-05,
      "loss": 1.1458,
      "step": 1995
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.73828125,
      "learning_rate": 9.229856128988201e-05,
      "loss": 1.111,
      "step": 2000
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.73828125,
      "learning_rate": 9.225453100309996e-05,
      "loss": 1.1457,
      "step": 2005
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.83984375,
      "learning_rate": 9.221038577553055e-05,
      "loss": 0.9936,
      "step": 2010
    },
    {
      "epoch": 0.21,
      "grad_norm": 11.125,
      "learning_rate": 9.216612572725762e-05,
      "loss": 1.0155,
      "step": 2015
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.74609375,
      "learning_rate": 9.212175097867738e-05,
      "loss": 1.0828,
      "step": 2020
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.65625,
      "learning_rate": 9.207726165049801e-05,
      "loss": 1.0193,
      "step": 2025
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.78515625,
      "learning_rate": 9.203265786373942e-05,
      "loss": 1.0699,
      "step": 2030
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.875,
      "learning_rate": 9.198793973973285e-05,
      "loss": 1.0606,
      "step": 2035
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.62109375,
      "learning_rate": 9.19431074001205e-05,
      "loss": 0.9172,
      "step": 2040
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.7578125,
      "learning_rate": 9.18981609668554e-05,
      "loss": 1.0414,
      "step": 2045
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.73046875,
      "learning_rate": 9.185310056220075e-05,
      "loss": 1.044,
      "step": 2050
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.796875,
      "learning_rate": 9.180792630872997e-05,
      "loss": 1.0334,
      "step": 2055
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.78125,
      "learning_rate": 9.1762638329326e-05,
      "loss": 1.1207,
      "step": 2060
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.734375,
      "learning_rate": 9.171723674718127e-05,
      "loss": 1.0033,
      "step": 2065
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.7578125,
      "learning_rate": 9.167172168579714e-05,
      "loss": 1.0764,
      "step": 2070
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.65625,
      "learning_rate": 9.16260932689837e-05,
      "loss": 1.177,
      "step": 2075
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.78125,
      "learning_rate": 9.158035162085941e-05,
      "loss": 1.0129,
      "step": 2080
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.77734375,
      "learning_rate": 9.153449686585068e-05,
      "loss": 1.0506,
      "step": 2085
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.80859375,
      "learning_rate": 9.148852912869166e-05,
      "loss": 1.0257,
      "step": 2090
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.80078125,
      "learning_rate": 9.144244853442376e-05,
      "loss": 1.1325,
      "step": 2095
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.8359375,
      "learning_rate": 9.139625520839548e-05,
      "loss": 1.1309,
      "step": 2100
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.89453125,
      "learning_rate": 9.134994927626187e-05,
      "loss": 1.1229,
      "step": 2105
    },
    {
      "epoch": 0.21,
      "grad_norm": 2.5,
      "learning_rate": 9.130353086398435e-05,
      "loss": 0.9799,
      "step": 2110
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.76953125,
      "learning_rate": 9.125700009783034e-05,
      "loss": 1.0075,
      "step": 2115
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.703125,
      "learning_rate": 9.121035710437278e-05,
      "loss": 1.0522,
      "step": 2120
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.80859375,
      "learning_rate": 9.116360201049e-05,
      "loss": 1.1396,
      "step": 2125
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.76171875,
      "learning_rate": 9.111673494336519e-05,
      "loss": 1.0829,
      "step": 2130
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.95703125,
      "learning_rate": 9.106975603048619e-05,
      "loss": 1.238,
      "step": 2135
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.4609375,
      "learning_rate": 9.102266539964503e-05,
      "loss": 0.8204,
      "step": 2140
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.6640625,
      "learning_rate": 9.097546317893766e-05,
      "loss": 0.7842,
      "step": 2145
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.71875,
      "learning_rate": 9.092814949676358e-05,
      "loss": 1.0007,
      "step": 2150
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.81640625,
      "learning_rate": 9.08807244818255e-05,
      "loss": 1.1469,
      "step": 2155
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.7578125,
      "learning_rate": 9.083318826312893e-05,
      "loss": 1.0594,
      "step": 2160
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.62890625,
      "learning_rate": 9.078554096998188e-05,
      "loss": 0.9549,
      "step": 2165
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.765625,
      "learning_rate": 9.07377827319946e-05,
      "loss": 1.0158,
      "step": 2170
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.71484375,
      "learning_rate": 9.068991367907902e-05,
      "loss": 1.0672,
      "step": 2175
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.078125,
      "learning_rate": 9.064193394144857e-05,
      "loss": 0.9645,
      "step": 2180
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.69921875,
      "learning_rate": 9.059384364961771e-05,
      "loss": 0.989,
      "step": 2185
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.69921875,
      "learning_rate": 9.054564293440172e-05,
      "loss": 1.003,
      "step": 2190
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.6875,
      "learning_rate": 9.049733192691617e-05,
      "loss": 0.8656,
      "step": 2195
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.7734375,
      "learning_rate": 9.044891075857667e-05,
      "loss": 1.1505,
      "step": 2200
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.7734375,
      "learning_rate": 9.04003795610985e-05,
      "loss": 1.0423,
      "step": 2205
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.70703125,
      "learning_rate": 9.035173846649622e-05,
      "loss": 1.2939,
      "step": 2210
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.875,
      "learning_rate": 9.03029876070834e-05,
      "loss": 0.975,
      "step": 2215
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.84375,
      "learning_rate": 9.025412711547211e-05,
      "loss": 1.0369,
      "step": 2220
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.74609375,
      "learning_rate": 9.020515712457267e-05,
      "loss": 0.9975,
      "step": 2225
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.80078125,
      "learning_rate": 9.015607776759328e-05,
      "loss": 0.9609,
      "step": 2230
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.76953125,
      "learning_rate": 9.010688917803958e-05,
      "loss": 0.979,
      "step": 2235
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.6640625,
      "learning_rate": 9.005759148971445e-05,
      "loss": 0.9703,
      "step": 2240
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.703125,
      "learning_rate": 9.000818483671741e-05,
      "loss": 0.9845,
      "step": 2245
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.6484375,
      "learning_rate": 8.99586693534445e-05,
      "loss": 1.0463,
      "step": 2250
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.8671875,
      "learning_rate": 8.990904517458774e-05,
      "loss": 1.0171,
      "step": 2255
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.76953125,
      "learning_rate": 8.98593124351348e-05,
      "loss": 1.0348,
      "step": 2260
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.7109375,
      "learning_rate": 8.980947127036876e-05,
      "loss": 0.9866,
      "step": 2265
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.7421875,
      "learning_rate": 8.975952181586751e-05,
      "loss": 1.0997,
      "step": 2270
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.74609375,
      "learning_rate": 8.970946420750362e-05,
      "loss": 1.0921,
      "step": 2275
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.828125,
      "learning_rate": 8.965929858144375e-05,
      "loss": 1.1113,
      "step": 2280
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.671875,
      "learning_rate": 8.96090250741485e-05,
      "loss": 0.9056,
      "step": 2285
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.70703125,
      "learning_rate": 8.955864382237187e-05,
      "loss": 1.0231,
      "step": 2290
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.71484375,
      "learning_rate": 8.950815496316094e-05,
      "loss": 1.0442,
      "step": 2295
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.765625,
      "learning_rate": 8.945755863385552e-05,
      "loss": 1.0367,
      "step": 2300
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.7109375,
      "learning_rate": 8.940685497208779e-05,
      "loss": 0.9599,
      "step": 2305
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.89453125,
      "learning_rate": 8.935604411578182e-05,
      "loss": 1.0799,
      "step": 2310
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.7109375,
      "learning_rate": 8.930512620315337e-05,
      "loss": 1.2683,
      "step": 2315
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.66796875,
      "learning_rate": 8.92541013727093e-05,
      "loss": 0.9654,
      "step": 2320
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.75390625,
      "learning_rate": 8.92029697632474e-05,
      "loss": 1.0312,
      "step": 2325
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.59375,
      "learning_rate": 8.915173151385589e-05,
      "loss": 1.0048,
      "step": 2330
    },
    {
      "epoch": 0.24,
      "grad_norm": 2.65625,
      "learning_rate": 8.910038676391308e-05,
      "loss": 1.0796,
      "step": 2335
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.7109375,
      "learning_rate": 8.904893565308696e-05,
      "loss": 0.9861,
      "step": 2340
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.71875,
      "learning_rate": 8.899737832133486e-05,
      "loss": 0.9987,
      "step": 2345
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.6953125,
      "learning_rate": 8.894571490890303e-05,
      "loss": 0.9757,
      "step": 2350
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.0703125,
      "learning_rate": 8.889394555632633e-05,
      "loss": 1.1203,
      "step": 2355
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.984375,
      "learning_rate": 8.884207040442773e-05,
      "loss": 1.0323,
      "step": 2360
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.70703125,
      "learning_rate": 8.87900895943181e-05,
      "loss": 1.03,
      "step": 2365
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.6484375,
      "learning_rate": 8.873800326739558e-05,
      "loss": 0.9085,
      "step": 2370
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.77734375,
      "learning_rate": 8.868581156534544e-05,
      "loss": 1.204,
      "step": 2375
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.94921875,
      "learning_rate": 8.863351463013959e-05,
      "loss": 1.2113,
      "step": 2380
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.6328125,
      "learning_rate": 8.85811126040361e-05,
      "loss": 1.1689,
      "step": 2385
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.6328125,
      "learning_rate": 8.852860562957906e-05,
      "loss": 0.9359,
      "step": 2390
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.8828125,
      "learning_rate": 8.847599384959791e-05,
      "loss": 1.2104,
      "step": 2395
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.76953125,
      "learning_rate": 8.842327740720722e-05,
      "loss": 1.2473,
      "step": 2400
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.62109375,
      "learning_rate": 8.83704564458063e-05,
      "loss": 1.1854,
      "step": 2405
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.69140625,
      "learning_rate": 8.831753110907873e-05,
      "loss": 1.0163,
      "step": 2410
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.78125,
      "learning_rate": 8.826450154099203e-05,
      "loss": 1.0576,
      "step": 2415
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.87109375,
      "learning_rate": 8.821136788579725e-05,
      "loss": 1.2287,
      "step": 2420
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.6953125,
      "learning_rate": 8.815813028802855e-05,
      "loss": 1.0083,
      "step": 2425
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.75,
      "learning_rate": 8.810478889250289e-05,
      "loss": 0.9742,
      "step": 2430
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.72265625,
      "learning_rate": 8.805134384431954e-05,
      "loss": 1.1751,
      "step": 2435
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.82421875,
      "learning_rate": 8.799779528885973e-05,
      "loss": 1.0532,
      "step": 2440
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.68359375,
      "learning_rate": 8.794414337178626e-05,
      "loss": 0.9992,
      "step": 2445
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.71875,
      "learning_rate": 8.789038823904307e-05,
      "loss": 1.0536,
      "step": 2450
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.85546875,
      "learning_rate": 8.783653003685493e-05,
      "loss": 1.0741,
      "step": 2455
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.72265625,
      "learning_rate": 8.778256891172691e-05,
      "loss": 1.1178,
      "step": 2460
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.76953125,
      "learning_rate": 8.772850501044408e-05,
      "loss": 1.2603,
      "step": 2465
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.7109375,
      "learning_rate": 8.767433848007107e-05,
      "loss": 1.0352,
      "step": 2470
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.74609375,
      "learning_rate": 8.762006946795171e-05,
      "loss": 1.0574,
      "step": 2475
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.8671875,
      "learning_rate": 8.756569812170859e-05,
      "loss": 1.1272,
      "step": 2480
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.7265625,
      "learning_rate": 8.751122458924263e-05,
      "loss": 1.2053,
      "step": 2485
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.69921875,
      "learning_rate": 8.745664901873276e-05,
      "loss": 0.9419,
      "step": 2490
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.70703125,
      "learning_rate": 8.740197155863548e-05,
      "loss": 1.0645,
      "step": 2495
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.72265625,
      "learning_rate": 8.734719235768442e-05,
      "loss": 1.1206,
      "step": 2500
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.92578125,
      "learning_rate": 8.729231156488997e-05,
      "loss": 1.0041,
      "step": 2505
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.78515625,
      "learning_rate": 8.723732932953889e-05,
      "loss": 1.0053,
      "step": 2510
    },
    {
      "epoch": 0.26,
      "grad_norm": 2.953125,
      "learning_rate": 8.718224580119387e-05,
      "loss": 1.0583,
      "step": 2515
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.734375,
      "learning_rate": 8.712706112969314e-05,
      "loss": 1.008,
      "step": 2520
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.69140625,
      "learning_rate": 8.707177546515007e-05,
      "loss": 1.1302,
      "step": 2525
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.7734375,
      "learning_rate": 8.701638895795271e-05,
      "loss": 1.006,
      "step": 2530
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.78125,
      "learning_rate": 8.696090175876348e-05,
      "loss": 1.1817,
      "step": 2535
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.6796875,
      "learning_rate": 8.690531401851866e-05,
      "loss": 1.1655,
      "step": 2540
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.75,
      "learning_rate": 8.684962588842805e-05,
      "loss": 1.0119,
      "step": 2545
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.7890625,
      "learning_rate": 8.679383751997452e-05,
      "loss": 1.0204,
      "step": 2550
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.76171875,
      "learning_rate": 8.673794906491358e-05,
      "loss": 1.1571,
      "step": 2555
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.890625,
      "learning_rate": 8.668196067527306e-05,
      "loss": 1.3189,
      "step": 2560
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.84375,
      "learning_rate": 8.662587250335256e-05,
      "loss": 1.0633,
      "step": 2565
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.7265625,
      "learning_rate": 8.656968470172315e-05,
      "loss": 1.0841,
      "step": 2570
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.6953125,
      "learning_rate": 8.65133974232269e-05,
      "loss": 1.1659,
      "step": 2575
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.92578125,
      "learning_rate": 8.645701082097651e-05,
      "loss": 1.0921,
      "step": 2580
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.6484375,
      "learning_rate": 8.64005250483548e-05,
      "loss": 0.9961,
      "step": 2585
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.63671875,
      "learning_rate": 8.63439402590144e-05,
      "loss": 1.0388,
      "step": 2590
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.71875,
      "learning_rate": 8.628725660687726e-05,
      "loss": 1.0681,
      "step": 2595
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.8203125,
      "learning_rate": 8.623047424613427e-05,
      "loss": 1.2429,
      "step": 2600
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.88671875,
      "learning_rate": 8.617359333124484e-05,
      "loss": 1.1379,
      "step": 2605
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.6640625,
      "learning_rate": 8.611661401693643e-05,
      "loss": 1.0119,
      "step": 2610
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.94921875,
      "learning_rate": 8.605953645820419e-05,
      "loss": 1.1663,
      "step": 2615
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.82421875,
      "learning_rate": 8.600236081031052e-05,
      "loss": 1.3278,
      "step": 2620
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.73046875,
      "learning_rate": 8.594508722878462e-05,
      "loss": 1.0438,
      "step": 2625
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.8203125,
      "learning_rate": 8.588771586942213e-05,
      "loss": 1.1375,
      "step": 2630
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.74609375,
      "learning_rate": 8.583024688828462e-05,
      "loss": 1.0629,
      "step": 2635
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.70703125,
      "learning_rate": 8.577268044169922e-05,
      "loss": 1.0521,
      "step": 2640
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.7734375,
      "learning_rate": 8.571501668625821e-05,
      "loss": 1.116,
      "step": 2645
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.6953125,
      "learning_rate": 8.565725577881856e-05,
      "loss": 0.9471,
      "step": 2650
    },
    {
      "epoch": 0.27,
      "grad_norm": 9.0,
      "learning_rate": 8.559939787650149e-05,
      "loss": 1.0076,
      "step": 2655
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.72265625,
      "learning_rate": 8.554144313669207e-05,
      "loss": 1.2238,
      "step": 2660
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.71484375,
      "learning_rate": 8.548339171703883e-05,
      "loss": 1.1868,
      "step": 2665
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.7421875,
      "learning_rate": 8.542524377545325e-05,
      "loss": 1.1218,
      "step": 2670
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.0078125,
      "learning_rate": 8.536699947010937e-05,
      "loss": 1.0396,
      "step": 2675
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.63671875,
      "learning_rate": 8.530865895944337e-05,
      "loss": 0.9681,
      "step": 2680
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.703125,
      "learning_rate": 8.525022240215314e-05,
      "loss": 1.0103,
      "step": 2685
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.74609375,
      "learning_rate": 8.51916899571978e-05,
      "loss": 1.109,
      "step": 2690
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.0625,
      "learning_rate": 8.513306178379732e-05,
      "loss": 1.0134,
      "step": 2695
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.80859375,
      "learning_rate": 8.507433804143208e-05,
      "loss": 1.0635,
      "step": 2700
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.8359375,
      "learning_rate": 8.501551888984241e-05,
      "loss": 1.0984,
      "step": 2705
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.71484375,
      "learning_rate": 8.495660448902823e-05,
      "loss": 1.0672,
      "step": 2710
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.78515625,
      "learning_rate": 8.489759499924845e-05,
      "loss": 1.0909,
      "step": 2715
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.69921875,
      "learning_rate": 8.483849058102073e-05,
      "loss": 1.0751,
      "step": 2720
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.7265625,
      "learning_rate": 8.477929139512093e-05,
      "loss": 0.9671,
      "step": 2725
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.71484375,
      "learning_rate": 8.471999760258268e-05,
      "loss": 1.1997,
      "step": 2730
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.67578125,
      "learning_rate": 8.466060936469697e-05,
      "loss": 1.013,
      "step": 2735
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.7421875,
      "learning_rate": 8.460112684301172e-05,
      "loss": 1.0004,
      "step": 2740
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.8046875,
      "learning_rate": 8.454155019933127e-05,
      "loss": 1.3402,
      "step": 2745
    },
    {
      "epoch": 0.28,
      "grad_norm": 2.203125,
      "learning_rate": 8.448187959571607e-05,
      "loss": 1.198,
      "step": 2750
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.76171875,
      "learning_rate": 8.442211519448209e-05,
      "loss": 1.1905,
      "step": 2755
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.80859375,
      "learning_rate": 8.436225715820046e-05,
      "loss": 1.0614,
      "step": 2760
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.8203125,
      "learning_rate": 8.430230564969702e-05,
      "loss": 1.0983,
      "step": 2765
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.6171875,
      "learning_rate": 8.424226083205192e-05,
      "loss": 1.0469,
      "step": 2770
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.8203125,
      "learning_rate": 8.418212286859904e-05,
      "loss": 1.1048,
      "step": 2775
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.76953125,
      "learning_rate": 8.412189192292572e-05,
      "loss": 0.9553,
      "step": 2780
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.8515625,
      "learning_rate": 8.406156815887221e-05,
      "loss": 1.0917,
      "step": 2785
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.8125,
      "learning_rate": 8.400115174053119e-05,
      "loss": 0.9901,
      "step": 2790
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.6953125,
      "learning_rate": 8.394064283224743e-05,
      "loss": 1.1655,
      "step": 2795
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.7890625,
      "learning_rate": 8.388004159861731e-05,
      "loss": 0.9846,
      "step": 2800
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.8515625,
      "learning_rate": 8.381934820448829e-05,
      "loss": 1.1505,
      "step": 2805
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.609375,
      "learning_rate": 8.375856281495856e-05,
      "loss": 0.8496,
      "step": 2810
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.34375,
      "learning_rate": 8.369768559537657e-05,
      "loss": 1.0266,
      "step": 2815
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.74609375,
      "learning_rate": 8.363671671134054e-05,
      "loss": 1.0651,
      "step": 2820
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.6015625,
      "learning_rate": 8.357565632869804e-05,
      "loss": 1.0375,
      "step": 2825
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.234375,
      "learning_rate": 8.351450461354556e-05,
      "loss": 1.0987,
      "step": 2830
    },
    {
      "epoch": 0.29,
      "grad_norm": 3.21875,
      "learning_rate": 8.345326173222799e-05,
      "loss": 1.0204,
      "step": 2835
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.76171875,
      "learning_rate": 8.339192785133824e-05,
      "loss": 0.8794,
      "step": 2840
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.75390625,
      "learning_rate": 8.333050313771677e-05,
      "loss": 1.0895,
      "step": 2845
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.64453125,
      "learning_rate": 8.326898775845108e-05,
      "loss": 1.0476,
      "step": 2850
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.8359375,
      "learning_rate": 8.320738188087533e-05,
      "loss": 1.1581,
      "step": 2855
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.734375,
      "learning_rate": 8.314568567256986e-05,
      "loss": 1.0111,
      "step": 2860
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.7890625,
      "learning_rate": 8.30838993013607e-05,
      "loss": 0.9779,
      "step": 2865
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.7421875,
      "learning_rate": 8.302202293531915e-05,
      "loss": 0.9491,
      "step": 2870
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.76171875,
      "learning_rate": 8.296005674276133e-05,
      "loss": 1.1357,
      "step": 2875
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.6796875,
      "learning_rate": 8.289800089224768e-05,
      "loss": 0.9299,
      "step": 2880
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.71875,
      "learning_rate": 8.283585555258256e-05,
      "loss": 1.0777,
      "step": 2885
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.7109375,
      "learning_rate": 8.277362089281371e-05,
      "loss": 1.0231,
      "step": 2890
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.71875,
      "learning_rate": 8.27112970822319e-05,
      "loss": 1.1078,
      "step": 2895
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.78125,
      "learning_rate": 8.264888429037039e-05,
      "loss": 1.106,
      "step": 2900
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.87890625,
      "learning_rate": 8.258638268700443e-05,
      "loss": 1.1014,
      "step": 2905
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.73046875,
      "learning_rate": 8.252379244215094e-05,
      "loss": 1.0533,
      "step": 2910
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.6953125,
      "learning_rate": 8.246111372606789e-05,
      "loss": 0.9655,
      "step": 2915
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.71484375,
      "learning_rate": 8.239834670925396e-05,
      "loss": 1.0225,
      "step": 2920
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.72265625,
      "learning_rate": 8.2335491562448e-05,
      "loss": 1.0847,
      "step": 2925
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.78515625,
      "learning_rate": 8.227254845662861e-05,
      "loss": 1.1701,
      "step": 2930
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.74609375,
      "learning_rate": 8.220951756301364e-05,
      "loss": 1.147,
      "step": 2935
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.71484375,
      "learning_rate": 8.214639905305974e-05,
      "loss": 1.1362,
      "step": 2940
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.76171875,
      "learning_rate": 8.208319309846188e-05,
      "loss": 1.1103,
      "step": 2945
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.87109375,
      "learning_rate": 8.201989987115296e-05,
      "loss": 0.927,
      "step": 2950
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.62890625,
      "learning_rate": 8.19565195433032e-05,
      "loss": 0.9902,
      "step": 2955
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.6875,
      "learning_rate": 8.18930522873198e-05,
      "loss": 0.9363,
      "step": 2960
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.78515625,
      "learning_rate": 8.182949827584641e-05,
      "loss": 1.1759,
      "step": 2965
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.76171875,
      "learning_rate": 8.176585768176266e-05,
      "loss": 1.0668,
      "step": 2970
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.75390625,
      "learning_rate": 8.170213067818371e-05,
      "loss": 1.035,
      "step": 2975
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.7578125,
      "learning_rate": 8.16383174384598e-05,
      "loss": 0.9191,
      "step": 2980
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.796875,
      "learning_rate": 8.157441813617568e-05,
      "loss": 1.146,
      "step": 2985
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.7265625,
      "learning_rate": 8.15104329451503e-05,
      "loss": 0.9065,
      "step": 2990
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.76953125,
      "learning_rate": 8.144636203943616e-05,
      "loss": 1.0684,
      "step": 2995
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.953125,
      "learning_rate": 8.138220559331895e-05,
      "loss": 1.1734,
      "step": 3000
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.81640625,
      "learning_rate": 8.131796378131709e-05,
      "loss": 1.0927,
      "step": 3005
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.6328125,
      "learning_rate": 8.125363677818114e-05,
      "loss": 0.9621,
      "step": 3010
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.6953125,
      "learning_rate": 8.118922475889346e-05,
      "loss": 1.1715,
      "step": 3015
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.78125,
      "learning_rate": 8.112472789866763e-05,
      "loss": 1.1307,
      "step": 3020
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.78515625,
      "learning_rate": 8.106014637294801e-05,
      "loss": 1.1674,
      "step": 3025
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.80078125,
      "learning_rate": 8.099548035740932e-05,
      "loss": 0.9488,
      "step": 3030
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.76953125,
      "learning_rate": 8.093073002795605e-05,
      "loss": 1.1273,
      "step": 3035
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.66015625,
      "learning_rate": 8.086589556072208e-05,
      "loss": 1.0848,
      "step": 3040
    },
    {
      "epoch": 0.31,
      "grad_norm": 3.421875,
      "learning_rate": 8.080097713207011e-05,
      "loss": 1.0033,
      "step": 3045
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.828125,
      "learning_rate": 8.073597491859132e-05,
      "loss": 1.0488,
      "step": 3050
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.73046875,
      "learning_rate": 8.067088909710473e-05,
      "loss": 0.97,
      "step": 3055
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.73046875,
      "learning_rate": 8.060571984465679e-05,
      "loss": 1.0538,
      "step": 3060
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.67578125,
      "learning_rate": 8.054046733852095e-05,
      "loss": 1.1165,
      "step": 3065
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.74609375,
      "learning_rate": 8.047513175619708e-05,
      "loss": 1.0492,
      "step": 3070
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.87890625,
      "learning_rate": 8.040971327541105e-05,
      "loss": 1.0117,
      "step": 3075
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.796875,
      "learning_rate": 8.034421207411423e-05,
      "loss": 1.119,
      "step": 3080
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.80859375,
      "learning_rate": 8.0278628330483e-05,
      "loss": 1.1995,
      "step": 3085
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.77734375,
      "learning_rate": 8.021296222291827e-05,
      "loss": 1.023,
      "step": 3090
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.796875,
      "learning_rate": 8.014721393004506e-05,
      "loss": 0.9752,
      "step": 3095
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.78515625,
      "learning_rate": 8.008138363071184e-05,
      "loss": 1.1086,
      "step": 3100
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.7734375,
      "learning_rate": 8.001547150399023e-05,
      "loss": 1.1296,
      "step": 3105
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.9453125,
      "learning_rate": 7.99494777291744e-05,
      "loss": 1.0221,
      "step": 3110
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.83984375,
      "learning_rate": 7.988340248578066e-05,
      "loss": 0.8845,
      "step": 3115
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.97265625,
      "learning_rate": 7.981724595354687e-05,
      "loss": 1.1321,
      "step": 3120
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.84375,
      "learning_rate": 7.97510083124321e-05,
      "loss": 1.0893,
      "step": 3125
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.75390625,
      "learning_rate": 7.968468974261596e-05,
      "loss": 1.1259,
      "step": 3130
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.80859375,
      "learning_rate": 7.961829042449825e-05,
      "loss": 1.0753,
      "step": 3135
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.8203125,
      "learning_rate": 7.955181053869841e-05,
      "loss": 1.0884,
      "step": 3140
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.66796875,
      "learning_rate": 7.948525026605506e-05,
      "loss": 1.0389,
      "step": 3145
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.640625,
      "learning_rate": 7.941860978762548e-05,
      "loss": 0.9641,
      "step": 3150
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.6875,
      "learning_rate": 7.935188928468508e-05,
      "loss": 1.0516,
      "step": 3155
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.734375,
      "learning_rate": 7.928508893872701e-05,
      "loss": 1.0447,
      "step": 3160
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.82421875,
      "learning_rate": 7.921820893146162e-05,
      "loss": 0.9953,
      "step": 3165
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.67578125,
      "learning_rate": 7.915124944481589e-05,
      "loss": 1.0326,
      "step": 3170
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.73046875,
      "learning_rate": 7.908421066093305e-05,
      "loss": 1.026,
      "step": 3175
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.93359375,
      "learning_rate": 7.9017092762172e-05,
      "loss": 1.1395,
      "step": 3180
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.0546875,
      "learning_rate": 7.894989593110688e-05,
      "loss": 1.0432,
      "step": 3185
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.59375,
      "learning_rate": 7.888262035052656e-05,
      "loss": 1.1498,
      "step": 3190
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.80078125,
      "learning_rate": 7.881526620343405e-05,
      "loss": 1.1971,
      "step": 3195
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.73828125,
      "learning_rate": 7.874783367304612e-05,
      "loss": 1.0781,
      "step": 3200
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.8515625,
      "learning_rate": 7.868032294279279e-05,
      "loss": 1.1643,
      "step": 3205
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.8125,
      "learning_rate": 7.861273419631674e-05,
      "loss": 1.0102,
      "step": 3210
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.78125,
      "learning_rate": 7.854506761747291e-05,
      "loss": 0.9955,
      "step": 3215
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.75390625,
      "learning_rate": 7.847732339032796e-05,
      "loss": 1.0998,
      "step": 3220
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.80078125,
      "learning_rate": 7.840950169915973e-05,
      "loss": 0.9991,
      "step": 3225
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.80078125,
      "learning_rate": 7.834160272845681e-05,
      "loss": 1.0961,
      "step": 3230
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.68359375,
      "learning_rate": 7.827362666291802e-05,
      "loss": 1.0051,
      "step": 3235
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.85546875,
      "learning_rate": 7.820557368745185e-05,
      "loss": 1.1509,
      "step": 3240
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.69921875,
      "learning_rate": 7.813744398717603e-05,
      "loss": 1.0741,
      "step": 3245
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.7421875,
      "learning_rate": 7.8069237747417e-05,
      "loss": 0.913,
      "step": 3250
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.7734375,
      "learning_rate": 7.800095515370938e-05,
      "loss": 1.0218,
      "step": 3255
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.83203125,
      "learning_rate": 7.793259639179551e-05,
      "loss": 1.1609,
      "step": 3260
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.7265625,
      "learning_rate": 7.786416164762492e-05,
      "loss": 1.0585,
      "step": 3265
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.87890625,
      "learning_rate": 7.779565110735378e-05,
      "loss": 1.2063,
      "step": 3270
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.7421875,
      "learning_rate": 7.772706495734454e-05,
      "loss": 1.0923,
      "step": 3275
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.85546875,
      "learning_rate": 7.765840338416523e-05,
      "loss": 1.1602,
      "step": 3280
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.75,
      "learning_rate": 7.758966657458908e-05,
      "loss": 1.1422,
      "step": 3285
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.6796875,
      "learning_rate": 7.752085471559399e-05,
      "loss": 1.0115,
      "step": 3290
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.86328125,
      "learning_rate": 7.7451967994362e-05,
      "loss": 1.027,
      "step": 3295
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.83203125,
      "learning_rate": 7.738300659827877e-05,
      "loss": 1.0176,
      "step": 3300
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.2265625,
      "learning_rate": 7.731397071493315e-05,
      "loss": 1.1252,
      "step": 3305
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.72265625,
      "learning_rate": 7.724486053211652e-05,
      "loss": 0.9394,
      "step": 3310
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.72265625,
      "learning_rate": 7.717567623782246e-05,
      "loss": 1.0385,
      "step": 3315
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.8515625,
      "learning_rate": 7.710641802024608e-05,
      "loss": 0.9274,
      "step": 3320
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.69921875,
      "learning_rate": 7.703708606778361e-05,
      "loss": 1.0951,
      "step": 3325
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.875,
      "learning_rate": 7.696768056903186e-05,
      "loss": 1.1682,
      "step": 3330
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.7890625,
      "learning_rate": 7.689820171278769e-05,
      "loss": 1.0186,
      "step": 3335
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.8125,
      "learning_rate": 7.682864968804748e-05,
      "loss": 1.1767,
      "step": 3340
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.859375,
      "learning_rate": 7.675902468400668e-05,
      "loss": 1.2458,
      "step": 3345
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.8359375,
      "learning_rate": 7.668932689005925e-05,
      "loss": 1.1877,
      "step": 3350
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.84375,
      "learning_rate": 7.661955649579712e-05,
      "loss": 1.0722,
      "step": 3355
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.7109375,
      "learning_rate": 7.65497136910098e-05,
      "loss": 1.0606,
      "step": 3360
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.765625,
      "learning_rate": 7.647979866568365e-05,
      "loss": 0.9516,
      "step": 3365
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.96875,
      "learning_rate": 7.640981161000157e-05,
      "loss": 1.2059,
      "step": 3370
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.62109375,
      "learning_rate": 7.633975271434236e-05,
      "loss": 1.0378,
      "step": 3375
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.7890625,
      "learning_rate": 7.626962216928025e-05,
      "loss": 0.9668,
      "step": 3380
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.7265625,
      "learning_rate": 7.619942016558434e-05,
      "loss": 1.0575,
      "step": 3385
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.3671875,
      "learning_rate": 7.612914689421821e-05,
      "loss": 1.1209,
      "step": 3390
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.6484375,
      "learning_rate": 7.605880254633917e-05,
      "loss": 1.1557,
      "step": 3395
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.8515625,
      "learning_rate": 7.598838731329796e-05,
      "loss": 1.0978,
      "step": 3400
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.78125,
      "learning_rate": 7.591790138663813e-05,
      "loss": 1.0088,
      "step": 3405
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.7421875,
      "learning_rate": 7.584734495809549e-05,
      "loss": 1.2496,
      "step": 3410
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.8125,
      "learning_rate": 7.577671821959766e-05,
      "loss": 1.1008,
      "step": 3415
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.7109375,
      "learning_rate": 7.570602136326352e-05,
      "loss": 1.0393,
      "step": 3420
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.78515625,
      "learning_rate": 7.563525458140269e-05,
      "loss": 1.0792,
      "step": 3425
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.66796875,
      "learning_rate": 7.5564418066515e-05,
      "loss": 1.0352,
      "step": 3430
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.6875,
      "learning_rate": 7.549351201128991e-05,
      "loss": 0.9773,
      "step": 3435
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.875,
      "learning_rate": 7.542253660860614e-05,
      "loss": 1.0221,
      "step": 3440
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.81640625,
      "learning_rate": 7.535149205153098e-05,
      "loss": 1.1252,
      "step": 3445
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.7734375,
      "learning_rate": 7.528037853331987e-05,
      "loss": 1.0447,
      "step": 3450
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.890625,
      "learning_rate": 7.520919624741578e-05,
      "loss": 1.0,
      "step": 3455
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.890625,
      "learning_rate": 7.513794538744885e-05,
      "loss": 1.1898,
      "step": 3460
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.83203125,
      "learning_rate": 7.506662614723563e-05,
      "loss": 1.0415,
      "step": 3465
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.7265625,
      "learning_rate": 7.499523872077878e-05,
      "loss": 0.9453,
      "step": 3470
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.76171875,
      "learning_rate": 7.492378330226637e-05,
      "loss": 1.0141,
      "step": 3475
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.7421875,
      "learning_rate": 7.485226008607146e-05,
      "loss": 1.072,
      "step": 3480
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.76953125,
      "learning_rate": 7.478066926675153e-05,
      "loss": 0.995,
      "step": 3485
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.69921875,
      "learning_rate": 7.470901103904794e-05,
      "loss": 1.0155,
      "step": 3490
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.7109375,
      "learning_rate": 7.463728559788541e-05,
      "loss": 1.1286,
      "step": 3495
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.77734375,
      "learning_rate": 7.456549313837153e-05,
      "loss": 0.864,
      "step": 3500
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.8515625,
      "learning_rate": 7.449363385579616e-05,
      "loss": 1.173,
      "step": 3505
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.64453125,
      "learning_rate": 7.442170794563094e-05,
      "loss": 1.1935,
      "step": 3510
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.89453125,
      "learning_rate": 7.434971560352873e-05,
      "loss": 1.2068,
      "step": 3515
    },
    {
      "epoch": 0.36,
      "grad_norm": 6.46875,
      "learning_rate": 7.427765702532315e-05,
      "loss": 1.1444,
      "step": 3520
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.78515625,
      "learning_rate": 7.420553240702798e-05,
      "loss": 1.1498,
      "step": 3525
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.671875,
      "learning_rate": 7.413334194483657e-05,
      "loss": 0.9728,
      "step": 3530
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.796875,
      "learning_rate": 7.406108583512148e-05,
      "loss": 0.9861,
      "step": 3535
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.80859375,
      "learning_rate": 7.398876427443379e-05,
      "loss": 0.9792,
      "step": 3540
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.8125,
      "learning_rate": 7.391637745950262e-05,
      "loss": 0.9392,
      "step": 3545
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.0546875,
      "learning_rate": 7.384392558723461e-05,
      "loss": 1.2004,
      "step": 3550
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.71875,
      "learning_rate": 7.377140885471339e-05,
      "loss": 1.0013,
      "step": 3555
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.8046875,
      "learning_rate": 7.369882745919896e-05,
      "loss": 1.02,
      "step": 3560
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.80078125,
      "learning_rate": 7.362618159812726e-05,
      "loss": 1.0134,
      "step": 3565
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.671875,
      "learning_rate": 7.355347146910961e-05,
      "loss": 0.9055,
      "step": 3570
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.6796875,
      "learning_rate": 7.348069726993208e-05,
      "loss": 0.9618,
      "step": 3575
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.75390625,
      "learning_rate": 7.34078591985551e-05,
      "loss": 0.9983,
      "step": 3580
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.80859375,
      "learning_rate": 7.333495745311279e-05,
      "loss": 1.1512,
      "step": 3585
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.84375,
      "learning_rate": 7.326199223191253e-05,
      "loss": 1.076,
      "step": 3590
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.73046875,
      "learning_rate": 7.318896373343432e-05,
      "loss": 1.0545,
      "step": 3595
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.79296875,
      "learning_rate": 7.311587215633029e-05,
      "loss": 0.974,
      "step": 3600
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.80078125,
      "learning_rate": 7.304271769942417e-05,
      "loss": 0.9138,
      "step": 3605
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.94140625,
      "learning_rate": 7.296950056171076e-05,
      "loss": 1.0022,
      "step": 3610
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.77734375,
      "learning_rate": 7.289622094235531e-05,
      "loss": 1.1084,
      "step": 3615
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.609375,
      "learning_rate": 7.282287904069308e-05,
      "loss": 1.2029,
      "step": 3620
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.7421875,
      "learning_rate": 7.274947505622875e-05,
      "loss": 1.0783,
      "step": 3625
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.84765625,
      "learning_rate": 7.26760091886358e-05,
      "loss": 1.0601,
      "step": 3630
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.7734375,
      "learning_rate": 7.260248163775616e-05,
      "loss": 1.2118,
      "step": 3635
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.74609375,
      "learning_rate": 7.252889260359947e-05,
      "loss": 1.0463,
      "step": 3640
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.765625,
      "learning_rate": 7.245524228634264e-05,
      "loss": 1.072,
      "step": 3645
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.828125,
      "learning_rate": 7.23815308863293e-05,
      "loss": 1.1773,
      "step": 3650
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.6796875,
      "learning_rate": 7.230775860406922e-05,
      "loss": 1.0873,
      "step": 3655
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.9609375,
      "learning_rate": 7.223392564023776e-05,
      "loss": 1.1419,
      "step": 3660
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.80078125,
      "learning_rate": 7.216003219567539e-05,
      "loss": 1.1429,
      "step": 3665
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.6875,
      "learning_rate": 7.208607847138709e-05,
      "loss": 1.0969,
      "step": 3670
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.75,
      "learning_rate": 7.201206466854176e-05,
      "loss": 1.0345,
      "step": 3675
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.7578125,
      "learning_rate": 7.193799098847181e-05,
      "loss": 1.0727,
      "step": 3680
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.8359375,
      "learning_rate": 7.186385763267247e-05,
      "loss": 1.0523,
      "step": 3685
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.7578125,
      "learning_rate": 7.178966480280131e-05,
      "loss": 1.0141,
      "step": 3690
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.72265625,
      "learning_rate": 7.171541270067771e-05,
      "loss": 0.9112,
      "step": 3695
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.875,
      "learning_rate": 7.164110152828223e-05,
      "loss": 1.0412,
      "step": 3700
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.8359375,
      "learning_rate": 7.156673148775615e-05,
      "loss": 1.066,
      "step": 3705
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.9765625,
      "learning_rate": 7.149230278140089e-05,
      "loss": 1.1692,
      "step": 3710
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.80078125,
      "learning_rate": 7.141781561167742e-05,
      "loss": 1.0214,
      "step": 3715
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.83984375,
      "learning_rate": 7.134327018120578e-05,
      "loss": 1.0949,
      "step": 3720
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.66796875,
      "learning_rate": 7.126866669276447e-05,
      "loss": 1.0195,
      "step": 3725
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.78515625,
      "learning_rate": 7.119400534928988e-05,
      "loss": 1.0325,
      "step": 3730
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.92578125,
      "learning_rate": 7.111928635387588e-05,
      "loss": 1.0255,
      "step": 3735
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.8984375,
      "learning_rate": 7.104450990977306e-05,
      "loss": 1.0661,
      "step": 3740
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.796875,
      "learning_rate": 7.096967622038834e-05,
      "loss": 1.0796,
      "step": 3745
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.4296875,
      "learning_rate": 7.089478548928434e-05,
      "loss": 1.0412,
      "step": 3750
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.796875,
      "learning_rate": 7.081983792017885e-05,
      "loss": 1.1041,
      "step": 3755
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.8515625,
      "learning_rate": 7.074483371694426e-05,
      "loss": 0.9852,
      "step": 3760
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.76171875,
      "learning_rate": 7.066977308360704e-05,
      "loss": 1.0722,
      "step": 3765
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.2265625,
      "learning_rate": 7.059465622434713e-05,
      "loss": 0.9913,
      "step": 3770
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.8125,
      "learning_rate": 7.051948334349746e-05,
      "loss": 1.0868,
      "step": 3775
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.77734375,
      "learning_rate": 7.044425464554329e-05,
      "loss": 0.9402,
      "step": 3780
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.79296875,
      "learning_rate": 7.036897033512177e-05,
      "loss": 1.0482,
      "step": 3785
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.75,
      "learning_rate": 7.029363061702129e-05,
      "loss": 0.9428,
      "step": 3790
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.734375,
      "learning_rate": 7.021823569618097e-05,
      "loss": 1.0352,
      "step": 3795
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.734375,
      "learning_rate": 7.01427857776901e-05,
      "loss": 1.2236,
      "step": 3800
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.8359375,
      "learning_rate": 7.006728106678757e-05,
      "loss": 1.2762,
      "step": 3805
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.78125,
      "learning_rate": 6.999172176886133e-05,
      "loss": 0.993,
      "step": 3810
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.8515625,
      "learning_rate": 6.991610808944778e-05,
      "loss": 0.941,
      "step": 3815
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.703125,
      "learning_rate": 6.984044023423128e-05,
      "loss": 1.104,
      "step": 3820
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.2734375,
      "learning_rate": 6.976471840904355e-05,
      "loss": 1.0966,
      "step": 3825
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.7578125,
      "learning_rate": 6.968894281986313e-05,
      "loss": 1.0537,
      "step": 3830
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.76171875,
      "learning_rate": 6.96131136728148e-05,
      "loss": 0.8949,
      "step": 3835
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.765625,
      "learning_rate": 6.953723117416901e-05,
      "loss": 1.0956,
      "step": 3840
    },
    {
      "epoch": 0.39,
      "grad_norm": 3.515625,
      "learning_rate": 6.94612955303414e-05,
      "loss": 1.1337,
      "step": 3845
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.7578125,
      "learning_rate": 6.938530694789206e-05,
      "loss": 0.9737,
      "step": 3850
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.74609375,
      "learning_rate": 6.930926563352521e-05,
      "loss": 1.0535,
      "step": 3855
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.78515625,
      "learning_rate": 6.923317179408844e-05,
      "loss": 1.0223,
      "step": 3860
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.6796875,
      "learning_rate": 6.91570256365722e-05,
      "loss": 1.0118,
      "step": 3865
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.80078125,
      "learning_rate": 6.908082736810935e-05,
      "loss": 1.0233,
      "step": 3870
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.765625,
      "learning_rate": 6.90045771959744e-05,
      "loss": 0.9254,
      "step": 3875
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.6875,
      "learning_rate": 6.892827532758311e-05,
      "loss": 1.0422,
      "step": 3880
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.890625,
      "learning_rate": 6.885192197049182e-05,
      "loss": 1.0743,
      "step": 3885
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.86328125,
      "learning_rate": 6.877551733239699e-05,
      "loss": 1.0699,
      "step": 3890
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.84765625,
      "learning_rate": 6.869906162113449e-05,
      "loss": 1.1193,
      "step": 3895
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.671875,
      "learning_rate": 6.862255504467924e-05,
      "loss": 0.9723,
      "step": 3900
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.640625,
      "learning_rate": 6.854599781114437e-05,
      "loss": 1.0641,
      "step": 3905
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.71484375,
      "learning_rate": 6.846939012878094e-05,
      "loss": 1.1135,
      "step": 3910
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.703125,
      "learning_rate": 6.839273220597717e-05,
      "loss": 1.1871,
      "step": 3915
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.703125,
      "learning_rate": 6.831602425125796e-05,
      "loss": 1.0599,
      "step": 3920
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.70703125,
      "learning_rate": 6.823926647328434e-05,
      "loss": 1.0594,
      "step": 3925
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.82421875,
      "learning_rate": 6.81624590808528e-05,
      "loss": 1.1122,
      "step": 3930
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.75,
      "learning_rate": 6.808560228289487e-05,
      "loss": 0.9967,
      "step": 3935
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.71875,
      "learning_rate": 6.80086962884764e-05,
      "loss": 1.0192,
      "step": 3940
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.68359375,
      "learning_rate": 6.793174130679711e-05,
      "loss": 1.0385,
      "step": 3945
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.75,
      "learning_rate": 6.785473754718997e-05,
      "loss": 0.9171,
      "step": 3950
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.8046875,
      "learning_rate": 6.777768521912062e-05,
      "loss": 0.9,
      "step": 3955
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.7890625,
      "learning_rate": 6.770058453218683e-05,
      "loss": 1.0755,
      "step": 3960
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.75390625,
      "learning_rate": 6.762343569611794e-05,
      "loss": 1.0492,
      "step": 3965
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.671875,
      "learning_rate": 6.754623892077418e-05,
      "loss": 1.0071,
      "step": 3970
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.74609375,
      "learning_rate": 6.746899441614624e-05,
      "loss": 0.9688,
      "step": 3975
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.6875,
      "learning_rate": 6.739170239235471e-05,
      "loss": 0.8293,
      "step": 3980
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.859375,
      "learning_rate": 6.731436305964933e-05,
      "loss": 1.0995,
      "step": 3985
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.57421875,
      "learning_rate": 6.723697662840857e-05,
      "loss": 0.9918,
      "step": 3990
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.69140625,
      "learning_rate": 6.715954330913902e-05,
      "loss": 0.9673,
      "step": 3995
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.80078125,
      "learning_rate": 6.708206331247487e-05,
      "loss": 1.1082,
      "step": 4000
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.73828125,
      "learning_rate": 6.700453684917716e-05,
      "loss": 1.0181,
      "step": 4005
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.74609375,
      "learning_rate": 6.692696413013344e-05,
      "loss": 0.9705,
      "step": 4010
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.66796875,
      "learning_rate": 6.684934536635702e-05,
      "loss": 1.0555,
      "step": 4015
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.83984375,
      "learning_rate": 6.67716807689865e-05,
      "loss": 1.0467,
      "step": 4020
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.7890625,
      "learning_rate": 6.669397054928514e-05,
      "loss": 1.1325,
      "step": 4025
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.046875,
      "learning_rate": 6.661621491864029e-05,
      "loss": 1.0789,
      "step": 4030
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.71875,
      "learning_rate": 6.653841408856284e-05,
      "loss": 1.1021,
      "step": 4035
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.765625,
      "learning_rate": 6.646056827068664e-05,
      "loss": 1.1895,
      "step": 4040
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.62109375,
      "learning_rate": 6.638267767676791e-05,
      "loss": 1.1359,
      "step": 4045
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.7890625,
      "learning_rate": 6.630474251868462e-05,
      "loss": 1.0581,
      "step": 4050
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.7578125,
      "learning_rate": 6.622676300843608e-05,
      "loss": 1.065,
      "step": 4055
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.6875,
      "learning_rate": 6.614873935814214e-05,
      "loss": 1.1056,
      "step": 4060
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.72265625,
      "learning_rate": 6.607067178004275e-05,
      "loss": 1.069,
      "step": 4065
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.73828125,
      "learning_rate": 6.599256048649737e-05,
      "loss": 1.007,
      "step": 4070
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.765625,
      "learning_rate": 6.591440568998435e-05,
      "loss": 1.0672,
      "step": 4075
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.76953125,
      "learning_rate": 6.583620760310041e-05,
      "loss": 0.9907,
      "step": 4080
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.82421875,
      "learning_rate": 6.575796643856e-05,
      "loss": 1.067,
      "step": 4085
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.78125,
      "learning_rate": 6.567968240919475e-05,
      "loss": 1.2485,
      "step": 4090
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.70703125,
      "learning_rate": 6.560135572795292e-05,
      "loss": 1.0458,
      "step": 4095
    },
    {
      "epoch": 0.42,
      "grad_norm": 2.421875,
      "learning_rate": 6.552298660789875e-05,
      "loss": 1.0591,
      "step": 4100
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.0234375,
      "learning_rate": 6.544457526221192e-05,
      "loss": 1.1354,
      "step": 4105
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.7578125,
      "learning_rate": 6.536612190418707e-05,
      "loss": 0.9666,
      "step": 4110
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.71875,
      "learning_rate": 6.528762674723298e-05,
      "loss": 1.1244,
      "step": 4115
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.67578125,
      "learning_rate": 6.520909000487223e-05,
      "loss": 1.0685,
      "step": 4120
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.0859375,
      "learning_rate": 6.513051189074047e-05,
      "loss": 1.171,
      "step": 4125
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.7109375,
      "learning_rate": 6.505189261858591e-05,
      "loss": 1.042,
      "step": 4130
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.73046875,
      "learning_rate": 6.497323240226874e-05,
      "loss": 1.0154,
      "step": 4135
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.734375,
      "learning_rate": 6.489453145576048e-05,
      "loss": 1.0528,
      "step": 4140
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.84765625,
      "learning_rate": 6.481578999314347e-05,
      "loss": 1.2128,
      "step": 4145
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.6484375,
      "learning_rate": 6.473700822861027e-05,
      "loss": 0.8893,
      "step": 4150
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.7265625,
      "learning_rate": 6.465818637646305e-05,
      "loss": 1.0033,
      "step": 4155
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.69921875,
      "learning_rate": 6.457932465111303e-05,
      "loss": 1.0655,
      "step": 4160
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.8046875,
      "learning_rate": 6.450042326707992e-05,
      "loss": 1.0127,
      "step": 4165
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.73046875,
      "learning_rate": 6.442148243899126e-05,
      "loss": 1.1003,
      "step": 4170
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.875,
      "learning_rate": 6.434250238158193e-05,
      "loss": 1.0191,
      "step": 4175
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.80859375,
      "learning_rate": 6.426348330969353e-05,
      "loss": 1.2239,
      "step": 4180
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.83203125,
      "learning_rate": 6.41844254382737e-05,
      "loss": 1.0521,
      "step": 4185
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.94140625,
      "learning_rate": 6.410532898237572e-05,
      "loss": 1.2184,
      "step": 4190
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.4140625,
      "learning_rate": 6.40261941571578e-05,
      "loss": 1.0664,
      "step": 4195
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.67578125,
      "learning_rate": 6.394702117788249e-05,
      "loss": 0.8452,
      "step": 4200
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.80859375,
      "learning_rate": 6.386781025991617e-05,
      "loss": 1.0528,
      "step": 4205
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.80078125,
      "learning_rate": 6.37885616187284e-05,
      "loss": 1.0768,
      "step": 4210
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.76171875,
      "learning_rate": 6.370927546989136e-05,
      "loss": 1.105,
      "step": 4215
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.84375,
      "learning_rate": 6.362995202907924e-05,
      "loss": 1.136,
      "step": 4220
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.703125,
      "learning_rate": 6.35505915120677e-05,
      "loss": 1.1266,
      "step": 4225
    },
    {
      "epoch": 0.43,
      "grad_norm": 4.53125,
      "learning_rate": 6.347119413473323e-05,
      "loss": 0.9888,
      "step": 4230
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.69921875,
      "learning_rate": 6.339176011305262e-05,
      "loss": 1.0379,
      "step": 4235
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.171875,
      "learning_rate": 6.331228966310228e-05,
      "loss": 1.226,
      "step": 4240
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.671875,
      "learning_rate": 6.323278300105778e-05,
      "loss": 0.9981,
      "step": 4245
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.78515625,
      "learning_rate": 6.315324034319317e-05,
      "loss": 1.2093,
      "step": 4250
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.65234375,
      "learning_rate": 6.30736619058804e-05,
      "loss": 0.9749,
      "step": 4255
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.6796875,
      "learning_rate": 6.299404790558874e-05,
      "loss": 1.0927,
      "step": 4260
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.6953125,
      "learning_rate": 6.291439855888423e-05,
      "loss": 1.0008,
      "step": 4265
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.76171875,
      "learning_rate": 6.283471408242907e-05,
      "loss": 1.2255,
      "step": 4270
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.7734375,
      "learning_rate": 6.275499469298097e-05,
      "loss": 1.0456,
      "step": 4275
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.73828125,
      "learning_rate": 6.267524060739264e-05,
      "loss": 1.04,
      "step": 4280
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.7734375,
      "learning_rate": 6.259545204261117e-05,
      "loss": 1.0479,
      "step": 4285
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.56640625,
      "learning_rate": 6.251562921567744e-05,
      "loss": 0.708,
      "step": 4290
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.69921875,
      "learning_rate": 6.243577234372551e-05,
      "loss": 1.2084,
      "step": 4295
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.78515625,
      "learning_rate": 6.235588164398209e-05,
      "loss": 1.1408,
      "step": 4300
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.69921875,
      "learning_rate": 6.227595733376587e-05,
      "loss": 0.9636,
      "step": 4305
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.921875,
      "learning_rate": 6.219599963048697e-05,
      "loss": 1.2366,
      "step": 4310
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.9296875,
      "learning_rate": 6.211600875164637e-05,
      "loss": 1.0225,
      "step": 4315
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.76953125,
      "learning_rate": 6.20359849148353e-05,
      "loss": 1.2267,
      "step": 4320
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.73828125,
      "learning_rate": 6.19559283377346e-05,
      "loss": 1.0573,
      "step": 4325
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.75,
      "learning_rate": 6.18758392381142e-05,
      "loss": 1.1054,
      "step": 4330
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.7265625,
      "learning_rate": 6.17957178338325e-05,
      "loss": 1.1948,
      "step": 4335
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.71875,
      "learning_rate": 6.171556434283574e-05,
      "loss": 1.191,
      "step": 4340
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.79296875,
      "learning_rate": 6.163537898315752e-05,
      "loss": 1.0744,
      "step": 4345
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.75390625,
      "learning_rate": 6.155516197291802e-05,
      "loss": 1.0738,
      "step": 4350
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.73828125,
      "learning_rate": 6.147491353032361e-05,
      "loss": 1.012,
      "step": 4355
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.76953125,
      "learning_rate": 6.139463387366612e-05,
      "loss": 0.9968,
      "step": 4360
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.7265625,
      "learning_rate": 6.131432322132228e-05,
      "loss": 1.0121,
      "step": 4365
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.73046875,
      "learning_rate": 6.123398179175318e-05,
      "loss": 1.0607,
      "step": 4370
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.734375,
      "learning_rate": 6.115360980350358e-05,
      "loss": 0.8946,
      "step": 4375
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.7734375,
      "learning_rate": 6.107320747520138e-05,
      "loss": 1.0477,
      "step": 4380
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.80859375,
      "learning_rate": 6.0992775025557025e-05,
      "loss": 0.9904,
      "step": 4385
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.7265625,
      "learning_rate": 6.0912312673362905e-05,
      "loss": 0.9623,
      "step": 4390
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.6796875,
      "learning_rate": 6.083182063749269e-05,
      "loss": 1.1777,
      "step": 4395
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.83203125,
      "learning_rate": 6.0751299136900886e-05,
      "loss": 0.9931,
      "step": 4400
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.53125,
      "learning_rate": 6.067074839062207e-05,
      "loss": 1.1213,
      "step": 4405
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.7578125,
      "learning_rate": 6.0590168617770416e-05,
      "loss": 1.1679,
      "step": 4410
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.70703125,
      "learning_rate": 6.050956003753905e-05,
      "loss": 0.8842,
      "step": 4415
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.82421875,
      "learning_rate": 6.0428922869199434e-05,
      "loss": 1.1319,
      "step": 4420
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.171875,
      "learning_rate": 6.034825733210086e-05,
      "loss": 1.06,
      "step": 4425
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.74609375,
      "learning_rate": 6.02675636456697e-05,
      "loss": 1.1352,
      "step": 4430
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.77734375,
      "learning_rate": 6.018684202940896e-05,
      "loss": 1.0763,
      "step": 4435
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.73046875,
      "learning_rate": 6.010609270289761e-05,
      "loss": 1.0631,
      "step": 4440
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.72265625,
      "learning_rate": 6.002531588579e-05,
      "loss": 0.9551,
      "step": 4445
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.71484375,
      "learning_rate": 5.9944511797815215e-05,
      "loss": 1.0763,
      "step": 4450
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.74609375,
      "learning_rate": 5.986368065877659e-05,
      "loss": 1.145,
      "step": 4455
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.796875,
      "learning_rate": 5.9782822688551e-05,
      "loss": 1.0967,
      "step": 4460
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.83984375,
      "learning_rate": 5.970193810708833e-05,
      "loss": 1.0671,
      "step": 4465
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.9296875,
      "learning_rate": 5.962102713441083e-05,
      "loss": 1.1535,
      "step": 4470
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.7734375,
      "learning_rate": 5.9540089990612546e-05,
      "loss": 0.9838,
      "step": 4475
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.83984375,
      "learning_rate": 5.9459126895858744e-05,
      "loss": 0.8705,
      "step": 4480
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.75,
      "learning_rate": 5.937813807038524e-05,
      "loss": 1.0007,
      "step": 4485
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.7265625,
      "learning_rate": 5.929712373449785e-05,
      "loss": 1.2598,
      "step": 4490
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.8125,
      "learning_rate": 5.92160841085718e-05,
      "loss": 1.1775,
      "step": 4495
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.80078125,
      "learning_rate": 5.913501941305114e-05,
      "loss": 1.1022,
      "step": 4500
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.7734375,
      "learning_rate": 5.905392986844802e-05,
      "loss": 1.2405,
      "step": 4505
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.671875,
      "learning_rate": 5.897281569534229e-05,
      "loss": 0.9214,
      "step": 4510
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.78125,
      "learning_rate": 5.8891677114380715e-05,
      "loss": 1.0663,
      "step": 4515
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.79296875,
      "learning_rate": 5.88105143462765e-05,
      "loss": 1.0043,
      "step": 4520
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.65625,
      "learning_rate": 5.8729327611808646e-05,
      "loss": 1.12,
      "step": 4525
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.86328125,
      "learning_rate": 5.864811713182129e-05,
      "loss": 1.0425,
      "step": 4530
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.70703125,
      "learning_rate": 5.856688312722324e-05,
      "loss": 1.0047,
      "step": 4535
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.6875,
      "learning_rate": 5.848562581898724e-05,
      "loss": 0.9652,
      "step": 4540
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.6953125,
      "learning_rate": 5.840434542814947e-05,
      "loss": 1.1617,
      "step": 4545
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.796875,
      "learning_rate": 5.8323042175808815e-05,
      "loss": 1.1901,
      "step": 4550
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.7578125,
      "learning_rate": 5.824171628312648e-05,
      "loss": 0.9598,
      "step": 4555
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.671875,
      "learning_rate": 5.8160367971325115e-05,
      "loss": 0.9963,
      "step": 4560
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.671875,
      "learning_rate": 5.807899746168847e-05,
      "loss": 0.9893,
      "step": 4565
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.87109375,
      "learning_rate": 5.7997604975560594e-05,
      "loss": 1.0825,
      "step": 4570
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.76171875,
      "learning_rate": 5.791619073434536e-05,
      "loss": 0.9896,
      "step": 4575
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.0,
      "learning_rate": 5.7834754959505836e-05,
      "loss": 1.074,
      "step": 4580
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.8125,
      "learning_rate": 5.775329787256362e-05,
      "loss": 1.1366,
      "step": 4585
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.8671875,
      "learning_rate": 5.7671819695098294e-05,
      "loss": 1.0909,
      "step": 4590
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.73046875,
      "learning_rate": 5.759032064874683e-05,
      "loss": 1.1507,
      "step": 4595
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.7421875,
      "learning_rate": 5.750880095520296e-05,
      "loss": 1.006,
      "step": 4600
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.73828125,
      "learning_rate": 5.7427260836216556e-05,
      "loss": 1.0907,
      "step": 4605
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.76953125,
      "learning_rate": 5.734570051359312e-05,
      "loss": 1.0693,
      "step": 4610
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.76953125,
      "learning_rate": 5.7264120209193015e-05,
      "loss": 1.1219,
      "step": 4615
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.953125,
      "learning_rate": 5.718252014493104e-05,
      "loss": 0.9902,
      "step": 4620
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.765625,
      "learning_rate": 5.7100900542775705e-05,
      "loss": 0.9542,
      "step": 4625
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.6484375,
      "learning_rate": 5.7019261624748664e-05,
      "loss": 0.993,
      "step": 4630
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.8515625,
      "learning_rate": 5.693760361292414e-05,
      "loss": 1.0169,
      "step": 4635
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.76171875,
      "learning_rate": 5.6855926729428274e-05,
      "loss": 1.0525,
      "step": 4640
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.6328125,
      "learning_rate": 5.677423119643856e-05,
      "loss": 0.8732,
      "step": 4645
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.1953125,
      "learning_rate": 5.66925172361832e-05,
      "loss": 1.0501,
      "step": 4650
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.74609375,
      "learning_rate": 5.661078507094052e-05,
      "loss": 0.9633,
      "step": 4655
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.6796875,
      "learning_rate": 5.652903492303838e-05,
      "loss": 1.0973,
      "step": 4660
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.69921875,
      "learning_rate": 5.644726701485358e-05,
      "loss": 1.1023,
      "step": 4665
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.86328125,
      "learning_rate": 5.6365481568811195e-05,
      "loss": 1.0983,
      "step": 4670
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.71875,
      "learning_rate": 5.628367880738401e-05,
      "loss": 1.1221,
      "step": 4675
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.7578125,
      "learning_rate": 5.620185895309195e-05,
      "loss": 1.0069,
      "step": 4680
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.86328125,
      "learning_rate": 5.6120022228501346e-05,
      "loss": 1.0823,
      "step": 4685
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.73828125,
      "learning_rate": 5.603816885622455e-05,
      "loss": 1.0401,
      "step": 4690
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.90625,
      "learning_rate": 5.595629905891908e-05,
      "loss": 1.1202,
      "step": 4695
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.71484375,
      "learning_rate": 5.5874413059287214e-05,
      "loss": 1.1821,
      "step": 4700
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.7265625,
      "learning_rate": 5.579251108007523e-05,
      "loss": 0.902,
      "step": 4705
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.828125,
      "learning_rate": 5.571059334407297e-05,
      "loss": 1.2474,
      "step": 4710
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.80859375,
      "learning_rate": 5.5628660074113034e-05,
      "loss": 1.0546,
      "step": 4715
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.74609375,
      "learning_rate": 5.554671149307036e-05,
      "loss": 1.1297,
      "step": 4720
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.83203125,
      "learning_rate": 5.5464747823861486e-05,
      "loss": 1.2551,
      "step": 4725
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.7265625,
      "learning_rate": 5.5382769289444e-05,
      "loss": 1.0594,
      "step": 4730
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.875,
      "learning_rate": 5.530077611281598e-05,
      "loss": 1.2795,
      "step": 4735
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.71875,
      "learning_rate": 5.5218768517015216e-05,
      "loss": 1.0416,
      "step": 4740
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.8203125,
      "learning_rate": 5.513674672511885e-05,
      "loss": 1.0523,
      "step": 4745
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.6796875,
      "learning_rate": 5.5054710960242574e-05,
      "loss": 1.1531,
      "step": 4750
    },
    {
      "epoch": 0.48,
      "grad_norm": 5.40625,
      "learning_rate": 5.497266144554007e-05,
      "loss": 1.2132,
      "step": 4755
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.63671875,
      "learning_rate": 5.489059840420249e-05,
      "loss": 1.0111,
      "step": 4760
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.8125,
      "learning_rate": 5.4808522059457724e-05,
      "loss": 1.0953,
      "step": 4765
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.6953125,
      "learning_rate": 5.472643263456987e-05,
      "loss": 1.1535,
      "step": 4770
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.7109375,
      "learning_rate": 5.4644330352838605e-05,
      "loss": 0.8999,
      "step": 4775
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.7421875,
      "learning_rate": 5.456221543759857e-05,
      "loss": 1.0552,
      "step": 4780
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.7265625,
      "learning_rate": 5.4480088112218795e-05,
      "loss": 1.0709,
      "step": 4785
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.64453125,
      "learning_rate": 5.439794860010207e-05,
      "loss": 1.0218,
      "step": 4790
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.78125,
      "learning_rate": 5.431579712468428e-05,
      "loss": 1.0769,
      "step": 4795
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.671875,
      "learning_rate": 5.423363390943391e-05,
      "loss": 0.945,
      "step": 4800
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.70703125,
      "learning_rate": 5.415145917785137e-05,
      "loss": 1.0062,
      "step": 4805
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.890625,
      "learning_rate": 5.406927315346839e-05,
      "loss": 1.2441,
      "step": 4810
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.7734375,
      "learning_rate": 5.398707605984739e-05,
      "loss": 0.9563,
      "step": 4815
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.734375,
      "learning_rate": 5.390486812058095e-05,
      "loss": 1.0443,
      "step": 4820
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.71875,
      "learning_rate": 5.382264955929114e-05,
      "loss": 0.9431,
      "step": 4825
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.703125,
      "learning_rate": 5.374042059962888e-05,
      "loss": 0.9862,
      "step": 4830
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.74609375,
      "learning_rate": 5.365818146527346e-05,
      "loss": 1.0034,
      "step": 4835
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.734375,
      "learning_rate": 5.357593237993174e-05,
      "loss": 0.9622,
      "step": 4840
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.6640625,
      "learning_rate": 5.3493673567337754e-05,
      "loss": 1.0207,
      "step": 4845
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.6484375,
      "learning_rate": 5.341140525125191e-05,
      "loss": 1.137,
      "step": 4850
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.78515625,
      "learning_rate": 5.3329127655460533e-05,
      "loss": 1.0355,
      "step": 4855
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.6875,
      "learning_rate": 5.3246841003775136e-05,
      "loss": 1.2037,
      "step": 4860
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.71484375,
      "learning_rate": 5.3164545520031926e-05,
      "loss": 0.9678,
      "step": 4865
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.66015625,
      "learning_rate": 5.3082241428091065e-05,
      "loss": 1.051,
      "step": 4870
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.7890625,
      "learning_rate": 5.2999928951836195e-05,
      "loss": 1.101,
      "step": 4875
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.88671875,
      "learning_rate": 5.291760831517373e-05,
      "loss": 1.1423,
      "step": 4880
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.81640625,
      "learning_rate": 5.283527974203227e-05,
      "loss": 1.2217,
      "step": 4885
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.72265625,
      "learning_rate": 5.275294345636208e-05,
      "loss": 1.0533,
      "step": 4890
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.71484375,
      "learning_rate": 5.267059968213428e-05,
      "loss": 0.9988,
      "step": 4895
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.75,
      "learning_rate": 5.258824864334047e-05,
      "loss": 0.9561,
      "step": 4900
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.7578125,
      "learning_rate": 5.250589056399194e-05,
      "loss": 1.1255,
      "step": 4905
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.96484375,
      "learning_rate": 5.242352566811919e-05,
      "loss": 1.1792,
      "step": 4910
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.8046875,
      "learning_rate": 5.2341154179771224e-05,
      "loss": 1.1269,
      "step": 4915
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.734375,
      "learning_rate": 5.225877632301497e-05,
      "loss": 1.238,
      "step": 4920
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.75,
      "learning_rate": 5.217639232193473e-05,
      "loss": 1.1802,
      "step": 4925
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.69921875,
      "learning_rate": 5.209400240063147e-05,
      "loss": 1.0109,
      "step": 4930
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.8203125,
      "learning_rate": 5.201160678322227e-05,
      "loss": 1.0919,
      "step": 4935
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.8125,
      "learning_rate": 5.192920569383971e-05,
      "loss": 1.0738,
      "step": 4940
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.69921875,
      "learning_rate": 5.184679935663131e-05,
      "loss": 1.0073,
      "step": 4945
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.66796875,
      "learning_rate": 5.1764387995758726e-05,
      "loss": 1.0793,
      "step": 4950
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.7734375,
      "learning_rate": 5.168197183539742e-05,
      "loss": 1.077,
      "step": 4955
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.71484375,
      "learning_rate": 5.159955109973585e-05,
      "loss": 1.1918,
      "step": 4960
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.77734375,
      "learning_rate": 5.151712601297491e-05,
      "loss": 1.1241,
      "step": 4965
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.74609375,
      "learning_rate": 5.1434696799327374e-05,
      "loss": 0.996,
      "step": 4970
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.7421875,
      "learning_rate": 5.1352263683017165e-05,
      "loss": 1.1889,
      "step": 4975
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.73828125,
      "learning_rate": 5.126982688827892e-05,
      "loss": 0.9517,
      "step": 4980
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.80859375,
      "learning_rate": 5.118738663935719e-05,
      "loss": 1.0765,
      "step": 4985
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.734375,
      "learning_rate": 5.1104943160506005e-05,
      "loss": 1.2389,
      "step": 4990
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.6484375,
      "learning_rate": 5.1022496675988085e-05,
      "loss": 0.9763,
      "step": 4995
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.74609375,
      "learning_rate": 5.0940047410074446e-05,
      "loss": 1.0224,
      "step": 5000
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.7890625,
      "learning_rate": 5.085759558704355e-05,
      "loss": 1.0476,
      "step": 5005
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.8359375,
      "learning_rate": 5.077514143118091e-05,
      "loss": 1.0783,
      "step": 5010
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.78515625,
      "learning_rate": 5.069268516677831e-05,
      "loss": 0.9991,
      "step": 5015
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.70703125,
      "learning_rate": 5.0610227018133314e-05,
      "loss": 0.9524,
      "step": 5020
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.68359375,
      "learning_rate": 5.052776720954862e-05,
      "loss": 1.0971,
      "step": 5025
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.75,
      "learning_rate": 5.0445305965331415e-05,
      "loss": 1.1505,
      "step": 5030
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.78125,
      "learning_rate": 5.03628435097928e-05,
      "loss": 0.9684,
      "step": 5035
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.75390625,
      "learning_rate": 5.028038006724716e-05,
      "loss": 1.0793,
      "step": 5040
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.70703125,
      "learning_rate": 5.019791586201157e-05,
      "loss": 0.927,
      "step": 5045
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.734375,
      "learning_rate": 5.01154511184052e-05,
      "loss": 1.0578,
      "step": 5050
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.6875,
      "learning_rate": 5.0032986060748676e-05,
      "loss": 1.016,
      "step": 5055
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.65234375,
      "learning_rate": 4.9950520913363444e-05,
      "loss": 0.8672,
      "step": 5060
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.671875,
      "learning_rate": 4.9868055900571245e-05,
      "loss": 0.9281,
      "step": 5065
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.73828125,
      "learning_rate": 4.9785591246693415e-05,
      "loss": 1.1488,
      "step": 5070
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.66796875,
      "learning_rate": 4.9703127176050346e-05,
      "loss": 1.0526,
      "step": 5075
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.78125,
      "learning_rate": 4.962066391296083e-05,
      "loss": 1.0239,
      "step": 5080
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.98828125,
      "learning_rate": 4.9538201681741416e-05,
      "loss": 1.098,
      "step": 5085
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.7109375,
      "learning_rate": 4.945574070670595e-05,
      "loss": 1.0512,
      "step": 5090
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.75,
      "learning_rate": 4.9373281212164764e-05,
      "loss": 0.9719,
      "step": 5095
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.68359375,
      "learning_rate": 4.929082342242419e-05,
      "loss": 0.8981,
      "step": 5100
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.81640625,
      "learning_rate": 4.9208367561785953e-05,
      "loss": 1.0382,
      "step": 5105
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.734375,
      "learning_rate": 4.912591385454651e-05,
      "loss": 1.052,
      "step": 5110
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.59765625,
      "learning_rate": 4.904346252499644e-05,
      "loss": 0.9151,
      "step": 5115
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.74609375,
      "learning_rate": 4.89610137974199e-05,
      "loss": 1.0986,
      "step": 5120
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.7265625,
      "learning_rate": 4.8878567896093915e-05,
      "loss": 1.1447,
      "step": 5125
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.7890625,
      "learning_rate": 4.8796125045287864e-05,
      "loss": 1.1617,
      "step": 5130
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.75,
      "learning_rate": 4.87136854692628e-05,
      "loss": 1.045,
      "step": 5135
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.75,
      "learning_rate": 4.86312493922709e-05,
      "loss": 1.0253,
      "step": 5140
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.73046875,
      "learning_rate": 4.854881703855481e-05,
      "loss": 0.9743,
      "step": 5145
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.76171875,
      "learning_rate": 4.8466388632347e-05,
      "loss": 1.0227,
      "step": 5150
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.77734375,
      "learning_rate": 4.83839643978693e-05,
      "loss": 1.1735,
      "step": 5155
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.78515625,
      "learning_rate": 4.83015445593321e-05,
      "loss": 1.1489,
      "step": 5160
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.84375,
      "learning_rate": 4.8219129340933865e-05,
      "loss": 1.0653,
      "step": 5165
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.66015625,
      "learning_rate": 4.813671896686052e-05,
      "loss": 1.016,
      "step": 5170
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.71875,
      "learning_rate": 4.805431366128478e-05,
      "loss": 0.8193,
      "step": 5175
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.8359375,
      "learning_rate": 4.797191364836557e-05,
      "loss": 0.9678,
      "step": 5180
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.73046875,
      "learning_rate": 4.7889519152247445e-05,
      "loss": 1.0295,
      "step": 5185
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.734375,
      "learning_rate": 4.780713039705993e-05,
      "loss": 1.0114,
      "step": 5190
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.78125,
      "learning_rate": 4.772474760691692e-05,
      "loss": 1.1057,
      "step": 5195
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.8046875,
      "learning_rate": 4.764237100591614e-05,
      "loss": 1.1135,
      "step": 5200
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.8359375,
      "learning_rate": 4.756000081813843e-05,
      "loss": 1.0649,
      "step": 5205
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.73046875,
      "learning_rate": 4.7477637267647174e-05,
      "loss": 1.078,
      "step": 5210
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.6796875,
      "learning_rate": 4.739528057848777e-05,
      "loss": 0.9255,
      "step": 5215
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.73828125,
      "learning_rate": 4.731293097468688e-05,
      "loss": 1.1124,
      "step": 5220
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.6640625,
      "learning_rate": 4.7230588680251904e-05,
      "loss": 1.1173,
      "step": 5225
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.66015625,
      "learning_rate": 4.7148253919170415e-05,
      "loss": 1.1483,
      "step": 5230
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.58203125,
      "learning_rate": 4.706592691540944e-05,
      "loss": 1.0426,
      "step": 5235
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.6953125,
      "learning_rate": 4.698360789291489e-05,
      "loss": 1.2384,
      "step": 5240
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.8125,
      "learning_rate": 4.690129707561104e-05,
      "loss": 1.0778,
      "step": 5245
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.703125,
      "learning_rate": 4.6818994687399767e-05,
      "loss": 0.9883,
      "step": 5250
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.796875,
      "learning_rate": 4.673670095216006e-05,
      "loss": 0.9607,
      "step": 5255
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.80859375,
      "learning_rate": 4.665441609374735e-05,
      "loss": 1.0143,
      "step": 5260
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.82421875,
      "learning_rate": 4.6572140335992956e-05,
      "loss": 1.1046,
      "step": 5265
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.734375,
      "learning_rate": 4.648987390270341e-05,
      "loss": 0.9611,
      "step": 5270
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.6953125,
      "learning_rate": 4.6407617017659864e-05,
      "loss": 1.042,
      "step": 5275
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.6875,
      "learning_rate": 4.632536990461756e-05,
      "loss": 0.9466,
      "step": 5280
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.78515625,
      "learning_rate": 4.624313278730508e-05,
      "loss": 1.1763,
      "step": 5285
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.75,
      "learning_rate": 4.6160905889423864e-05,
      "loss": 1.0396,
      "step": 5290
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.79296875,
      "learning_rate": 4.607868943464757e-05,
      "loss": 1.0803,
      "step": 5295
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.2890625,
      "learning_rate": 4.599648364662141e-05,
      "loss": 1.2114,
      "step": 5300
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.703125,
      "learning_rate": 4.591428874896155e-05,
      "loss": 0.8715,
      "step": 5305
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.71875,
      "learning_rate": 4.583210496525464e-05,
      "loss": 1.002,
      "step": 5310
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.7421875,
      "learning_rate": 4.5749932519056994e-05,
      "loss": 1.0924,
      "step": 5315
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.67578125,
      "learning_rate": 4.566777163389411e-05,
      "loss": 0.9119,
      "step": 5320
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.90234375,
      "learning_rate": 4.5585622533260095e-05,
      "loss": 1.1491,
      "step": 5325
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.703125,
      "learning_rate": 4.5503485440616925e-05,
      "loss": 0.9818,
      "step": 5330
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.72265625,
      "learning_rate": 4.5421360579393944e-05,
      "loss": 0.8872,
      "step": 5335
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.68359375,
      "learning_rate": 4.533924817298724e-05,
      "loss": 0.956,
      "step": 5340
    },
    {
      "epoch": 0.54,
      "grad_norm": 2.578125,
      "learning_rate": 4.525714844475901e-05,
      "loss": 0.952,
      "step": 5345
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.6875,
      "learning_rate": 4.5175061618036906e-05,
      "loss": 0.9298,
      "step": 5350
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.7109375,
      "learning_rate": 4.509298791611361e-05,
      "loss": 1.0396,
      "step": 5355
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.76171875,
      "learning_rate": 4.501092756224598e-05,
      "loss": 1.0364,
      "step": 5360
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.7578125,
      "learning_rate": 4.492888077965462e-05,
      "loss": 1.1525,
      "step": 5365
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.69921875,
      "learning_rate": 4.484684779152324e-05,
      "loss": 1.0566,
      "step": 5370
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.57421875,
      "learning_rate": 4.4764828820997965e-05,
      "loss": 0.9149,
      "step": 5375
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.828125,
      "learning_rate": 4.4682824091186855e-05,
      "loss": 1.0611,
      "step": 5380
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.80859375,
      "learning_rate": 4.460083382515914e-05,
      "loss": 1.2073,
      "step": 5385
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.71875,
      "learning_rate": 4.4518858245944836e-05,
      "loss": 0.949,
      "step": 5390
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.71875,
      "learning_rate": 4.4436897576533904e-05,
      "loss": 0.9548,
      "step": 5395
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.921875,
      "learning_rate": 4.435495203987576e-05,
      "loss": 0.9157,
      "step": 5400
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.75390625,
      "learning_rate": 4.427302185887872e-05,
      "loss": 1.0325,
      "step": 5405
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.83203125,
      "learning_rate": 4.4191107256409264e-05,
      "loss": 0.92,
      "step": 5410
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.78515625,
      "learning_rate": 4.410920845529151e-05,
      "loss": 1.0232,
      "step": 5415
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.95703125,
      "learning_rate": 4.4027325678306606e-05,
      "loss": 1.1333,
      "step": 5420
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.64453125,
      "learning_rate": 4.394545914819213e-05,
      "loss": 1.0389,
      "step": 5425
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.73828125,
      "learning_rate": 4.38636090876414e-05,
      "loss": 1.0456,
      "step": 5430
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.72265625,
      "learning_rate": 4.3781775719302994e-05,
      "loss": 1.1344,
      "step": 5435
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.7421875,
      "learning_rate": 4.3699959265780066e-05,
      "loss": 1.0087,
      "step": 5440
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.6953125,
      "learning_rate": 4.361815994962974e-05,
      "loss": 1.0076,
      "step": 5445
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.69140625,
      "learning_rate": 4.353637799336257e-05,
      "loss": 0.954,
      "step": 5450
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.6796875,
      "learning_rate": 4.345461361944184e-05,
      "loss": 1.1507,
      "step": 5455
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.046875,
      "learning_rate": 4.3372867050283e-05,
      "loss": 1.0085,
      "step": 5460
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.765625,
      "learning_rate": 4.329113850825314e-05,
      "loss": 0.8788,
      "step": 5465
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.83984375,
      "learning_rate": 4.320942821567023e-05,
      "loss": 1.0546,
      "step": 5470
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.69140625,
      "learning_rate": 4.312773639480263e-05,
      "loss": 1.0464,
      "step": 5475
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.7578125,
      "learning_rate": 4.304606326786847e-05,
      "loss": 0.9899,
      "step": 5480
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.0390625,
      "learning_rate": 4.296440905703501e-05,
      "loss": 0.9215,
      "step": 5485
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.71875,
      "learning_rate": 4.288277398441804e-05,
      "loss": 0.9959,
      "step": 5490
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.21875,
      "learning_rate": 4.280115827208134e-05,
      "loss": 1.0586,
      "step": 5495
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.77734375,
      "learning_rate": 4.271956214203598e-05,
      "loss": 1.0914,
      "step": 5500
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.671875,
      "learning_rate": 4.263798581623976e-05,
      "loss": 0.9507,
      "step": 5505
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.796875,
      "learning_rate": 4.255642951659664e-05,
      "loss": 1.158,
      "step": 5510
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.78125,
      "learning_rate": 4.2474893464956086e-05,
      "loss": 0.9905,
      "step": 5515
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.80859375,
      "learning_rate": 4.2393377883112495e-05,
      "loss": 1.1065,
      "step": 5520
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.734375,
      "learning_rate": 4.231188299280456e-05,
      "loss": 1.0051,
      "step": 5525
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.77734375,
      "learning_rate": 4.2230409015714724e-05,
      "loss": 0.9964,
      "step": 5530
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.671875,
      "learning_rate": 4.214895617346853e-05,
      "loss": 1.0241,
      "step": 5535
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.65625,
      "learning_rate": 4.206752468763397e-05,
      "loss": 1.0364,
      "step": 5540
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.70703125,
      "learning_rate": 4.198611477972108e-05,
      "loss": 1.0132,
      "step": 5545
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.875,
      "learning_rate": 4.190472667118105e-05,
      "loss": 1.0339,
      "step": 5550
    },
    {
      "epoch": 0.57,
      "grad_norm": 2.921875,
      "learning_rate": 4.182336058340585e-05,
      "loss": 0.9761,
      "step": 5555
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.78515625,
      "learning_rate": 4.174201673772754e-05,
      "loss": 1.1701,
      "step": 5560
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.69921875,
      "learning_rate": 4.166069535541768e-05,
      "loss": 0.917,
      "step": 5565
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.69140625,
      "learning_rate": 4.1579396657686693e-05,
      "loss": 1.0072,
      "step": 5570
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.79296875,
      "learning_rate": 4.1498120865683355e-05,
      "loss": 1.1712,
      "step": 5575
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.87109375,
      "learning_rate": 4.141686820049409e-05,
      "loss": 1.11,
      "step": 5580
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.75390625,
      "learning_rate": 4.1335638883142384e-05,
      "loss": 1.1543,
      "step": 5585
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.75390625,
      "learning_rate": 4.12544331345883e-05,
      "loss": 1.0131,
      "step": 5590
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.76953125,
      "learning_rate": 4.11732511757277e-05,
      "loss": 1.1151,
      "step": 5595
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.6875,
      "learning_rate": 4.1092093227391795e-05,
      "loss": 1.0872,
      "step": 5600
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.6015625,
      "learning_rate": 4.101095951034645e-05,
      "loss": 0.8416,
      "step": 5605
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.8203125,
      "learning_rate": 4.092985024529164e-05,
      "loss": 0.9728,
      "step": 5610
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.7734375,
      "learning_rate": 4.0848765652860766e-05,
      "loss": 0.9884,
      "step": 5615
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.71875,
      "learning_rate": 4.0767705953620226e-05,
      "loss": 0.9671,
      "step": 5620
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.828125,
      "learning_rate": 4.06866713680686e-05,
      "loss": 1.1356,
      "step": 5625
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.75390625,
      "learning_rate": 4.0605662116636185e-05,
      "loss": 1.0336,
      "step": 5630
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.671875,
      "learning_rate": 4.052467841968437e-05,
      "loss": 1.014,
      "step": 5635
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.7265625,
      "learning_rate": 4.0443720497505054e-05,
      "loss": 1.1874,
      "step": 5640
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.71484375,
      "learning_rate": 4.0362788570319995e-05,
      "loss": 1.1396,
      "step": 5645
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.75,
      "learning_rate": 4.0281882858280205e-05,
      "loss": 1.0163,
      "step": 5650
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.7890625,
      "learning_rate": 4.0201003581465484e-05,
      "loss": 1.0544,
      "step": 5655
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.7890625,
      "learning_rate": 4.012015095988363e-05,
      "loss": 1.0106,
      "step": 5660
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.78515625,
      "learning_rate": 4.003932521346996e-05,
      "loss": 1.0598,
      "step": 5665
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.69140625,
      "learning_rate": 3.995852656208672e-05,
      "loss": 1.0287,
      "step": 5670
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.81640625,
      "learning_rate": 3.9877755225522403e-05,
      "loss": 1.2901,
      "step": 5675
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.69921875,
      "learning_rate": 3.979701142349123e-05,
      "loss": 0.9457,
      "step": 5680
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.71484375,
      "learning_rate": 3.971629537563252e-05,
      "loss": 1.174,
      "step": 5685
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.8359375,
      "learning_rate": 3.9635607301510095e-05,
      "loss": 0.9278,
      "step": 5690
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.74609375,
      "learning_rate": 3.955494742061163e-05,
      "loss": 0.8117,
      "step": 5695
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.91015625,
      "learning_rate": 3.947431595234823e-05,
      "loss": 1.1099,
      "step": 5700
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.75390625,
      "learning_rate": 3.939371311605358e-05,
      "loss": 0.9595,
      "step": 5705
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.62109375,
      "learning_rate": 3.931313913098356e-05,
      "loss": 1.0319,
      "step": 5710
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.75390625,
      "learning_rate": 3.923259421631555e-05,
      "loss": 1.0311,
      "step": 5715
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.71875,
      "learning_rate": 3.915207859114785e-05,
      "loss": 1.0689,
      "step": 5720
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.8359375,
      "learning_rate": 3.907159247449907e-05,
      "loss": 1.1307,
      "step": 5725
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.7890625,
      "learning_rate": 3.899113608530759e-05,
      "loss": 1.1061,
      "step": 5730
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.875,
      "learning_rate": 3.891070964243091e-05,
      "loss": 1.1362,
      "step": 5735
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.86328125,
      "learning_rate": 3.883031336464502e-05,
      "loss": 1.1102,
      "step": 5740
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.7734375,
      "learning_rate": 3.874994747064394e-05,
      "loss": 1.1218,
      "step": 5745
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.75,
      "learning_rate": 3.866961217903897e-05,
      "loss": 1.0652,
      "step": 5750
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.81640625,
      "learning_rate": 3.8589307708358217e-05,
      "loss": 0.9907,
      "step": 5755
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.76953125,
      "learning_rate": 3.850903427704591e-05,
      "loss": 1.025,
      "step": 5760
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.90625,
      "learning_rate": 3.8428792103461875e-05,
      "loss": 0.9414,
      "step": 5765
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.8359375,
      "learning_rate": 3.834858140588087e-05,
      "loss": 1.0071,
      "step": 5770
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.79296875,
      "learning_rate": 3.826840240249207e-05,
      "loss": 0.9806,
      "step": 5775
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.73828125,
      "learning_rate": 3.8188255311398434e-05,
      "loss": 0.9352,
      "step": 5780
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.73828125,
      "learning_rate": 3.8108140350616086e-05,
      "loss": 0.9285,
      "step": 5785
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.66796875,
      "learning_rate": 3.802805773807377e-05,
      "loss": 0.9716,
      "step": 5790
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.77734375,
      "learning_rate": 3.794800769161225e-05,
      "loss": 1.197,
      "step": 5795
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.81640625,
      "learning_rate": 3.7867990428983675e-05,
      "loss": 1.2058,
      "step": 5800
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.7421875,
      "learning_rate": 3.7788006167850984e-05,
      "loss": 0.9454,
      "step": 5805
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.6875,
      "learning_rate": 3.770805512578746e-05,
      "loss": 0.97,
      "step": 5810
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.67578125,
      "learning_rate": 3.76281375202759e-05,
      "loss": 1.0012,
      "step": 5815
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.8359375,
      "learning_rate": 3.7548253568708206e-05,
      "loss": 1.0222,
      "step": 5820
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.72265625,
      "learning_rate": 3.746840348838474e-05,
      "loss": 1.135,
      "step": 5825
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.66796875,
      "learning_rate": 3.738858749651371e-05,
      "loss": 0.9011,
      "step": 5830
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.72265625,
      "learning_rate": 3.730880581021058e-05,
      "loss": 1.1707,
      "step": 5835
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.76953125,
      "learning_rate": 3.722905864649754e-05,
      "loss": 1.1671,
      "step": 5840
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.75,
      "learning_rate": 3.7149346222302854e-05,
      "loss": 1.0802,
      "step": 5845
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.6953125,
      "learning_rate": 3.7069668754460236e-05,
      "loss": 0.9061,
      "step": 5850
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.671875,
      "learning_rate": 3.699002645970842e-05,
      "loss": 1.0918,
      "step": 5855
    },
    {
      "epoch": 0.6,
      "grad_norm": 4.8125,
      "learning_rate": 3.6910419554690344e-05,
      "loss": 1.0834,
      "step": 5860
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.78125,
      "learning_rate": 3.6830848255952755e-05,
      "loss": 1.0767,
      "step": 5865
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.75,
      "learning_rate": 3.675131277994553e-05,
      "loss": 1.0405,
      "step": 5870
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.78515625,
      "learning_rate": 3.667181334302109e-05,
      "loss": 1.1731,
      "step": 5875
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.69921875,
      "learning_rate": 3.659235016143383e-05,
      "loss": 1.1153,
      "step": 5880
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.7421875,
      "learning_rate": 3.6512923451339483e-05,
      "loss": 0.972,
      "step": 5885
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.79296875,
      "learning_rate": 3.6433533428794674e-05,
      "loss": 0.9887,
      "step": 5890
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.75,
      "learning_rate": 3.635418030975612e-05,
      "loss": 1.0424,
      "step": 5895
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.734375,
      "learning_rate": 3.627486431008019e-05,
      "loss": 1.0604,
      "step": 5900
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.71484375,
      "learning_rate": 3.619558564552232e-05,
      "loss": 1.2235,
      "step": 5905
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.140625,
      "learning_rate": 3.611634453173634e-05,
      "loss": 1.0018,
      "step": 5910
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.703125,
      "learning_rate": 3.6037141184273955e-05,
      "loss": 1.0535,
      "step": 5915
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.80859375,
      "learning_rate": 3.5957975818584135e-05,
      "loss": 1.0967,
      "step": 5920
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.7734375,
      "learning_rate": 3.587884865001254e-05,
      "loss": 1.1435,
      "step": 5925
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.76953125,
      "learning_rate": 3.579975989380088e-05,
      "loss": 0.9315,
      "step": 5930
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.7421875,
      "learning_rate": 3.572070976508645e-05,
      "loss": 1.1153,
      "step": 5935
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.8203125,
      "learning_rate": 3.564169847890141e-05,
      "loss": 1.0063,
      "step": 5940
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.70703125,
      "learning_rate": 3.5562726250172294e-05,
      "loss": 0.9381,
      "step": 5945
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.81640625,
      "learning_rate": 3.548379329371939e-05,
      "loss": 0.9847,
      "step": 5950
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.98828125,
      "learning_rate": 3.5404899824256144e-05,
      "loss": 1.0556,
      "step": 5955
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.76953125,
      "learning_rate": 3.532604605638856e-05,
      "loss": 1.0517,
      "step": 5960
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.71875,
      "learning_rate": 3.5247232204614745e-05,
      "loss": 0.8894,
      "step": 5965
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.73046875,
      "learning_rate": 3.516845848332411e-05,
      "loss": 0.9472,
      "step": 5970
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.7421875,
      "learning_rate": 3.5089725106796955e-05,
      "loss": 1.1139,
      "step": 5975
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.65625,
      "learning_rate": 3.5011032289203863e-05,
      "loss": 0.9535,
      "step": 5980
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.7734375,
      "learning_rate": 3.4932380244605046e-05,
      "loss": 1.0412,
      "step": 5985
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.7734375,
      "learning_rate": 3.485376918694981e-05,
      "loss": 1.0978,
      "step": 5990
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.671875,
      "learning_rate": 3.477519933007599e-05,
      "loss": 1.022,
      "step": 5995
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.8046875,
      "learning_rate": 3.469667088770934e-05,
      "loss": 1.1041,
      "step": 6000
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.8203125,
      "learning_rate": 3.461818407346292e-05,
      "loss": 0.9351,
      "step": 6005
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.66796875,
      "learning_rate": 3.453973910083659e-05,
      "loss": 0.9167,
      "step": 6010
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.828125,
      "learning_rate": 3.446133618321642e-05,
      "loss": 1.0682,
      "step": 6015
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.671875,
      "learning_rate": 3.438297553387402e-05,
      "loss": 1.0559,
      "step": 6020
    },
    {
      "epoch": 0.61,
      "grad_norm": 4.8125,
      "learning_rate": 3.430465736596605e-05,
      "loss": 1.1181,
      "step": 6025
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.8046875,
      "learning_rate": 3.422638189253364e-05,
      "loss": 1.0976,
      "step": 6030
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.75390625,
      "learning_rate": 3.4148149326501745e-05,
      "loss": 0.9938,
      "step": 6035
    },
    {
      "epoch": 0.62,
      "grad_norm": 2.515625,
      "learning_rate": 3.406995988067858e-05,
      "loss": 1.0343,
      "step": 6040
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.6640625,
      "learning_rate": 3.399181376775514e-05,
      "loss": 1.1531,
      "step": 6045
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.74609375,
      "learning_rate": 3.391371120030449e-05,
      "loss": 0.9443,
      "step": 6050
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.828125,
      "learning_rate": 3.3835652390781246e-05,
      "loss": 1.0006,
      "step": 6055
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.78515625,
      "learning_rate": 3.375763755152101e-05,
      "loss": 1.0689,
      "step": 6060
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.8515625,
      "learning_rate": 3.367966689473979e-05,
      "loss": 1.0221,
      "step": 6065
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.6953125,
      "learning_rate": 3.360174063253335e-05,
      "loss": 0.9852,
      "step": 6070
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.68359375,
      "learning_rate": 3.3523858976876774e-05,
      "loss": 1.1333,
      "step": 6075
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.7578125,
      "learning_rate": 3.3446022139623744e-05,
      "loss": 1.0534,
      "step": 6080
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.63671875,
      "learning_rate": 3.336823033250602e-05,
      "loss": 0.961,
      "step": 6085
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.65625,
      "learning_rate": 3.329048376713294e-05,
      "loss": 0.9828,
      "step": 6090
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.84765625,
      "learning_rate": 3.321278265499072e-05,
      "loss": 0.9971,
      "step": 6095
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.70703125,
      "learning_rate": 3.3135127207441934e-05,
      "loss": 1.0486,
      "step": 6100
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.7421875,
      "learning_rate": 3.305751763572497e-05,
      "loss": 0.9744,
      "step": 6105
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.8046875,
      "learning_rate": 3.297995415095342e-05,
      "loss": 1.1336,
      "step": 6110
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.7890625,
      "learning_rate": 3.2902436964115435e-05,
      "loss": 0.9069,
      "step": 6115
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.73828125,
      "learning_rate": 3.282496628607337e-05,
      "loss": 0.8783,
      "step": 6120
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.75,
      "learning_rate": 3.274754232756294e-05,
      "loss": 0.9949,
      "step": 6125
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.765625,
      "learning_rate": 3.267016529919282e-05,
      "loss": 1.1357,
      "step": 6130
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.84375,
      "learning_rate": 3.259283541144403e-05,
      "loss": 1.1777,
      "step": 6135
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.74609375,
      "learning_rate": 3.251555287466936e-05,
      "loss": 1.0011,
      "step": 6140
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.9296875,
      "learning_rate": 3.2438317899092796e-05,
      "loss": 1.0628,
      "step": 6145
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.671875,
      "learning_rate": 3.2361130694808895e-05,
      "loss": 1.042,
      "step": 6150
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.828125,
      "learning_rate": 3.228399147178238e-05,
      "loss": 1.3185,
      "step": 6155
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.66015625,
      "learning_rate": 3.220690043984734e-05,
      "loss": 1.18,
      "step": 6160
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.83984375,
      "learning_rate": 3.212985780870683e-05,
      "loss": 1.1598,
      "step": 6165
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.72265625,
      "learning_rate": 3.2052863787932245e-05,
      "loss": 1.1073,
      "step": 6170
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.80859375,
      "learning_rate": 3.197591858696275e-05,
      "loss": 0.9834,
      "step": 6175
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.83203125,
      "learning_rate": 3.189902241510467e-05,
      "loss": 0.9688,
      "step": 6180
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.796875,
      "learning_rate": 3.1822175481531034e-05,
      "loss": 1.2171,
      "step": 6185
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.88671875,
      "learning_rate": 3.174537799528089e-05,
      "loss": 1.1525,
      "step": 6190
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.66796875,
      "learning_rate": 3.1668630165258736e-05,
      "loss": 0.9984,
      "step": 6195
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.81640625,
      "learning_rate": 3.1591932200234095e-05,
      "loss": 1.2287,
      "step": 6200
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.765625,
      "learning_rate": 3.151528430884078e-05,
      "loss": 1.0306,
      "step": 6205
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.71875,
      "learning_rate": 3.143868669957637e-05,
      "loss": 0.9547,
      "step": 6210
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.81640625,
      "learning_rate": 3.136213958080175e-05,
      "loss": 1.0548,
      "step": 6215
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.921875,
      "learning_rate": 3.1285643160740394e-05,
      "loss": 1.1733,
      "step": 6220
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.9375,
      "learning_rate": 3.1209197647477885e-05,
      "loss": 1.092,
      "step": 6225
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.8984375,
      "learning_rate": 3.113280324896134e-05,
      "loss": 1.2049,
      "step": 6230
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.79296875,
      "learning_rate": 3.105646017299882e-05,
      "loss": 1.0478,
      "step": 6235
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.8515625,
      "learning_rate": 3.0980168627258775e-05,
      "loss": 1.0877,
      "step": 6240
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.95703125,
      "learning_rate": 3.09039288192695e-05,
      "loss": 1.0203,
      "step": 6245
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.703125,
      "learning_rate": 3.082774095641853e-05,
      "loss": 0.9786,
      "step": 6250
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.7890625,
      "learning_rate": 3.075160524595212e-05,
      "loss": 1.0277,
      "step": 6255
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.765625,
      "learning_rate": 3.0675521894974645e-05,
      "loss": 1.0803,
      "step": 6260
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.67578125,
      "learning_rate": 3.059949111044809e-05,
      "loss": 0.9742,
      "step": 6265
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.046875,
      "learning_rate": 3.052351309919136e-05,
      "loss": 1.092,
      "step": 6270
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.7109375,
      "learning_rate": 3.04475880678799e-05,
      "loss": 0.9507,
      "step": 6275
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.71875,
      "learning_rate": 3.0371716223044998e-05,
      "loss": 1.0087,
      "step": 6280
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.796875,
      "learning_rate": 3.0295897771073266e-05,
      "loss": 1.2292,
      "step": 6285
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.3359375,
      "learning_rate": 3.0220132918206073e-05,
      "loss": 1.06,
      "step": 6290
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.76171875,
      "learning_rate": 3.014442187053901e-05,
      "loss": 1.0489,
      "step": 6295
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.7734375,
      "learning_rate": 3.006876483402128e-05,
      "loss": 1.0905,
      "step": 6300
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.74609375,
      "learning_rate": 2.9993162014455145e-05,
      "loss": 0.9829,
      "step": 6305
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.7578125,
      "learning_rate": 2.9917613617495484e-05,
      "loss": 1.1404,
      "step": 6310
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.75,
      "learning_rate": 2.984211984864902e-05,
      "loss": 0.8786,
      "step": 6315
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.76953125,
      "learning_rate": 2.9766680913273915e-05,
      "loss": 1.0549,
      "step": 6320
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.7265625,
      "learning_rate": 2.9691297016579223e-05,
      "loss": 1.004,
      "step": 6325
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.8125,
      "learning_rate": 2.9615968363624212e-05,
      "loss": 1.1635,
      "step": 6330
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.71484375,
      "learning_rate": 2.9540695159317915e-05,
      "loss": 0.9876,
      "step": 6335
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.76953125,
      "learning_rate": 2.946547760841853e-05,
      "loss": 0.9598,
      "step": 6340
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.5859375,
      "learning_rate": 2.939031591553286e-05,
      "loss": 1.0414,
      "step": 6345
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.73828125,
      "learning_rate": 2.9315210285115736e-05,
      "loss": 1.0624,
      "step": 6350
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.62109375,
      "learning_rate": 2.924016092146956e-05,
      "loss": 1.074,
      "step": 6355
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.76953125,
      "learning_rate": 2.9165168028743605e-05,
      "loss": 1.1979,
      "step": 6360
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.765625,
      "learning_rate": 2.9090231810933538e-05,
      "loss": 1.1097,
      "step": 6365
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.46875,
      "learning_rate": 2.9015352471880953e-05,
      "loss": 1.2338,
      "step": 6370
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.734375,
      "learning_rate": 2.8940530215272577e-05,
      "loss": 1.0206,
      "step": 6375
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.77734375,
      "learning_rate": 2.8865765244639963e-05,
      "loss": 1.0175,
      "step": 6380
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.69140625,
      "learning_rate": 2.879105776335878e-05,
      "loss": 1.2262,
      "step": 6385
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.77734375,
      "learning_rate": 2.8716407974648385e-05,
      "loss": 1.0278,
      "step": 6390
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.109375,
      "learning_rate": 2.8641816081571148e-05,
      "loss": 1.1596,
      "step": 6395
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.8828125,
      "learning_rate": 2.8567282287031904e-05,
      "loss": 1.007,
      "step": 6400
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.68359375,
      "learning_rate": 2.8492806793777566e-05,
      "loss": 0.9987,
      "step": 6405
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.671875,
      "learning_rate": 2.841838980439636e-05,
      "loss": 0.9187,
      "step": 6410
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.66015625,
      "learning_rate": 2.8344031521317415e-05,
      "loss": 0.9507,
      "step": 6415
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.7265625,
      "learning_rate": 2.8269732146810145e-05,
      "loss": 1.0025,
      "step": 6420
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.79296875,
      "learning_rate": 2.8195491882983726e-05,
      "loss": 1.0142,
      "step": 6425
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.75390625,
      "learning_rate": 2.8121310931786537e-05,
      "loss": 1.0992,
      "step": 6430
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.75,
      "learning_rate": 2.8047189495005656e-05,
      "loss": 0.9699,
      "step": 6435
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.70703125,
      "learning_rate": 2.797312777426624e-05,
      "loss": 1.0096,
      "step": 6440
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.69921875,
      "learning_rate": 2.7899125971030938e-05,
      "loss": 0.9341,
      "step": 6445
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.765625,
      "learning_rate": 2.7825184286599543e-05,
      "loss": 0.9735,
      "step": 6450
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.65625,
      "learning_rate": 2.775130292210822e-05,
      "loss": 0.9685,
      "step": 6455
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.78125,
      "learning_rate": 2.7677482078529083e-05,
      "loss": 1.1064,
      "step": 6460
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.75390625,
      "learning_rate": 2.7603721956669604e-05,
      "loss": 1.1283,
      "step": 6465
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.7734375,
      "learning_rate": 2.7530022757172092e-05,
      "loss": 0.987,
      "step": 6470
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.109375,
      "learning_rate": 2.7456384680513114e-05,
      "loss": 1.0236,
      "step": 6475
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.0859375,
      "learning_rate": 2.738280792700303e-05,
      "loss": 0.9482,
      "step": 6480
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.79296875,
      "learning_rate": 2.730929269678534e-05,
      "loss": 1.1663,
      "step": 6485
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.78125,
      "learning_rate": 2.7235839189836127e-05,
      "loss": 0.9718,
      "step": 6490
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.7109375,
      "learning_rate": 2.7162447605963716e-05,
      "loss": 1.1799,
      "step": 6495
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.703125,
      "learning_rate": 2.7089118144807886e-05,
      "loss": 0.9379,
      "step": 6500
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.80078125,
      "learning_rate": 2.7015851005839465e-05,
      "loss": 1.0892,
      "step": 6505
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.70703125,
      "learning_rate": 2.694264638835974e-05,
      "loss": 0.979,
      "step": 6510
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.90625,
      "learning_rate": 2.686950449149993e-05,
      "loss": 1.1582,
      "step": 6515
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.72265625,
      "learning_rate": 2.6796425514220646e-05,
      "loss": 0.9576,
      "step": 6520
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.828125,
      "learning_rate": 2.672340965531135e-05,
      "loss": 1.1961,
      "step": 6525
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.71875,
      "learning_rate": 2.6650457113389794e-05,
      "loss": 1.0329,
      "step": 6530
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.71875,
      "learning_rate": 2.657756808690151e-05,
      "loss": 0.9499,
      "step": 6535
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.80859375,
      "learning_rate": 2.6504742774119225e-05,
      "loss": 1.1772,
      "step": 6540
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.71875,
      "learning_rate": 2.6431981373142413e-05,
      "loss": 1.0316,
      "step": 6545
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.828125,
      "learning_rate": 2.635928408189666e-05,
      "loss": 0.9518,
      "step": 6550
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.79296875,
      "learning_rate": 2.6286651098133107e-05,
      "loss": 1.063,
      "step": 6555
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.74609375,
      "learning_rate": 2.621408261942806e-05,
      "loss": 1.0582,
      "step": 6560
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.7734375,
      "learning_rate": 2.6141578843182312e-05,
      "loss": 0.9849,
      "step": 6565
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.73828125,
      "learning_rate": 2.6069139966620648e-05,
      "loss": 0.96,
      "step": 6570
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.7734375,
      "learning_rate": 2.5996766186791323e-05,
      "loss": 1.1181,
      "step": 6575
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.74609375,
      "learning_rate": 2.592445770056551e-05,
      "loss": 0.9954,
      "step": 6580
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.7421875,
      "learning_rate": 2.5852214704636756e-05,
      "loss": 0.9518,
      "step": 6585
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.7421875,
      "learning_rate": 2.578003739552053e-05,
      "loss": 0.9734,
      "step": 6590
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.7265625,
      "learning_rate": 2.570792596955355e-05,
      "loss": 0.9734,
      "step": 6595
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.74609375,
      "learning_rate": 2.5635880622893282e-05,
      "loss": 0.9918,
      "step": 6600
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.67578125,
      "learning_rate": 2.5563901551517576e-05,
      "loss": 1.0667,
      "step": 6605
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.6796875,
      "learning_rate": 2.5491988951223878e-05,
      "loss": 0.9904,
      "step": 6610
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.76171875,
      "learning_rate": 2.5420143017628885e-05,
      "loss": 1.134,
      "step": 6615
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.78515625,
      "learning_rate": 2.5348363946167913e-05,
      "loss": 0.9767,
      "step": 6620
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.80078125,
      "learning_rate": 2.527665193209442e-05,
      "loss": 1.0921,
      "step": 6625
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.1171875,
      "learning_rate": 2.5205007170479444e-05,
      "loss": 0.9865,
      "step": 6630
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.7578125,
      "learning_rate": 2.5133429856211066e-05,
      "loss": 1.035,
      "step": 6635
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.6953125,
      "learning_rate": 2.5061920183993977e-05,
      "loss": 1.0116,
      "step": 6640
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.9453125,
      "learning_rate": 2.4990478348348744e-05,
      "loss": 0.8975,
      "step": 6645
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.796875,
      "learning_rate": 2.4919104543611476e-05,
      "loss": 1.1114,
      "step": 6650
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.796875,
      "learning_rate": 2.4847798963933256e-05,
      "loss": 1.0895,
      "step": 6655
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.171875,
      "learning_rate": 2.4776561803279524e-05,
      "loss": 1.2028,
      "step": 6660
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.796875,
      "learning_rate": 2.4705393255429616e-05,
      "loss": 1.1115,
      "step": 6665
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.79296875,
      "learning_rate": 2.463429351397624e-05,
      "loss": 1.1289,
      "step": 6670
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.80078125,
      "learning_rate": 2.456326277232494e-05,
      "loss": 1.049,
      "step": 6675
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.7421875,
      "learning_rate": 2.4492301223693553e-05,
      "loss": 1.039,
      "step": 6680
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.65234375,
      "learning_rate": 2.4421409061111704e-05,
      "loss": 1.0935,
      "step": 6685
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.72265625,
      "learning_rate": 2.435058647742028e-05,
      "loss": 1.0483,
      "step": 6690
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.8046875,
      "learning_rate": 2.4279833665270862e-05,
      "loss": 1.0762,
      "step": 6695
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.9453125,
      "learning_rate": 2.4209150817125327e-05,
      "loss": 1.1581,
      "step": 6700
    },
    {
      "epoch": 0.68,
      "grad_norm": 2.234375,
      "learning_rate": 2.413853812525516e-05,
      "loss": 1.0193,
      "step": 6705
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.55859375,
      "learning_rate": 2.4067995781740975e-05,
      "loss": 1.0372,
      "step": 6710
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.703125,
      "learning_rate": 2.399752397847214e-05,
      "loss": 1.0451,
      "step": 6715
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.76953125,
      "learning_rate": 2.392712290714605e-05,
      "loss": 1.0981,
      "step": 6720
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.75390625,
      "learning_rate": 2.3856792759267716e-05,
      "loss": 1.2807,
      "step": 6725
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.76171875,
      "learning_rate": 2.378653372614923e-05,
      "loss": 1.1395,
      "step": 6730
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.80859375,
      "learning_rate": 2.371634599890923e-05,
      "loss": 0.9596,
      "step": 6735
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.7734375,
      "learning_rate": 2.3646229768472376e-05,
      "loss": 1.1523,
      "step": 6740
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.75,
      "learning_rate": 2.3576185225568904e-05,
      "loss": 1.1069,
      "step": 6745
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.6640625,
      "learning_rate": 2.3506212560733986e-05,
      "loss": 1.2153,
      "step": 6750
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.6328125,
      "learning_rate": 2.3436311964307252e-05,
      "loss": 0.8835,
      "step": 6755
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.640625,
      "learning_rate": 2.3366483626432328e-05,
      "loss": 0.9273,
      "step": 6760
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.6953125,
      "learning_rate": 2.329672773705631e-05,
      "loss": 0.9277,
      "step": 6765
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.7265625,
      "learning_rate": 2.3227044485929185e-05,
      "loss": 0.8623,
      "step": 6770
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.78125,
      "learning_rate": 2.3157434062603327e-05,
      "loss": 1.0534,
      "step": 6775
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.83984375,
      "learning_rate": 2.3087896656433057e-05,
      "loss": 0.9354,
      "step": 6780
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.83203125,
      "learning_rate": 2.3018432456574025e-05,
      "loss": 0.9725,
      "step": 6785
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.69921875,
      "learning_rate": 2.2949041651982763e-05,
      "loss": 1.0576,
      "step": 6790
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.71875,
      "learning_rate": 2.28797244314162e-05,
      "loss": 1.006,
      "step": 6795
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.69140625,
      "learning_rate": 2.2810480983431022e-05,
      "loss": 0.9865,
      "step": 6800
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.83984375,
      "learning_rate": 2.2741311496383254e-05,
      "loss": 1.3458,
      "step": 6805
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.796875,
      "learning_rate": 2.2672216158427816e-05,
      "loss": 0.9439,
      "step": 6810
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.81640625,
      "learning_rate": 2.2603195157517825e-05,
      "loss": 1.0935,
      "step": 6815
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.65234375,
      "learning_rate": 2.2534248681404247e-05,
      "loss": 1.0236,
      "step": 6820
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.71484375,
      "learning_rate": 2.246537691763529e-05,
      "loss": 1.0428,
      "step": 6825
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.71484375,
      "learning_rate": 2.2396580053555956e-05,
      "loss": 1.1085,
      "step": 6830
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.64453125,
      "learning_rate": 2.232785827630749e-05,
      "loss": 1.0015,
      "step": 6835
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.796875,
      "learning_rate": 2.2259211772826886e-05,
      "loss": 1.0249,
      "step": 6840
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.76171875,
      "learning_rate": 2.2190640729846396e-05,
      "loss": 1.1303,
      "step": 6845
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.74609375,
      "learning_rate": 2.2122145333892964e-05,
      "loss": 1.0429,
      "step": 6850
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.7265625,
      "learning_rate": 2.2053725771287816e-05,
      "loss": 1.0856,
      "step": 6855
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.80078125,
      "learning_rate": 2.1985382228145873e-05,
      "loss": 1.0638,
      "step": 6860
    },
    {
      "epoch": 0.7,
      "grad_norm": 4.1875,
      "learning_rate": 2.1917114890375205e-05,
      "loss": 1.0854,
      "step": 6865
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.7265625,
      "learning_rate": 2.1848923943676703e-05,
      "loss": 1.0279,
      "step": 6870
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.8515625,
      "learning_rate": 2.1780809573543377e-05,
      "loss": 0.9827,
      "step": 6875
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.7265625,
      "learning_rate": 2.1712771965259953e-05,
      "loss": 0.8368,
      "step": 6880
    },
    {
      "epoch": 0.7,
      "grad_norm": 2.71875,
      "learning_rate": 2.1644811303902357e-05,
      "loss": 1.0574,
      "step": 6885
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.8828125,
      "learning_rate": 2.1576927774337196e-05,
      "loss": 1.0771,
      "step": 6890
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.7578125,
      "learning_rate": 2.1509121561221274e-05,
      "loss": 0.9592,
      "step": 6895
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.796875,
      "learning_rate": 2.1441392849001048e-05,
      "loss": 1.1832,
      "step": 6900
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.7734375,
      "learning_rate": 2.1373741821912234e-05,
      "loss": 1.0718,
      "step": 6905
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.73046875,
      "learning_rate": 2.130616866397912e-05,
      "loss": 0.9542,
      "step": 6910
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.69921875,
      "learning_rate": 2.1238673559014237e-05,
      "loss": 1.0957,
      "step": 6915
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.765625,
      "learning_rate": 2.117125669061782e-05,
      "loss": 1.1128,
      "step": 6920
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.77734375,
      "learning_rate": 2.110391824217723e-05,
      "loss": 0.9538,
      "step": 6925
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.86328125,
      "learning_rate": 2.1036658396866543e-05,
      "loss": 1.2052,
      "step": 6930
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.796875,
      "learning_rate": 2.0969477337646005e-05,
      "loss": 0.9397,
      "step": 6935
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.76171875,
      "learning_rate": 2.090237524726156e-05,
      "loss": 1.038,
      "step": 6940
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.67578125,
      "learning_rate": 2.0835352308244298e-05,
      "loss": 1.0399,
      "step": 6945
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.83984375,
      "learning_rate": 2.076840870291011e-05,
      "loss": 1.2341,
      "step": 6950
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.2890625,
      "learning_rate": 2.070154461335894e-05,
      "loss": 0.9457,
      "step": 6955
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.70703125,
      "learning_rate": 2.0634760221474526e-05,
      "loss": 1.1267,
      "step": 6960
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.78515625,
      "learning_rate": 2.05680557089238e-05,
      "loss": 0.9812,
      "step": 6965
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.79296875,
      "learning_rate": 2.0501431257156395e-05,
      "loss": 1.202,
      "step": 6970
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.6875,
      "learning_rate": 2.0434887047404163e-05,
      "loss": 0.9614,
      "step": 6975
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.77734375,
      "learning_rate": 2.0368423260680674e-05,
      "loss": 0.9797,
      "step": 6980
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.7578125,
      "learning_rate": 2.0302040077780744e-05,
      "loss": 1.0879,
      "step": 6985
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.78515625,
      "learning_rate": 2.023573767927993e-05,
      "loss": 0.984,
      "step": 6990
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.6328125,
      "learning_rate": 2.0169516245534032e-05,
      "loss": 0.8387,
      "step": 6995
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.796875,
      "learning_rate": 2.010337595667861e-05,
      "loss": 0.9795,
      "step": 7000
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.7109375,
      "learning_rate": 2.00373169926285e-05,
      "loss": 1.1734,
      "step": 7005
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.75,
      "learning_rate": 1.99713395330773e-05,
      "loss": 1.006,
      "step": 7010
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.82421875,
      "learning_rate": 1.9905443757496966e-05,
      "loss": 1.1849,
      "step": 7015
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.734375,
      "learning_rate": 1.9839629845137147e-05,
      "loss": 0.9772,
      "step": 7020
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.79296875,
      "learning_rate": 1.9773897975024873e-05,
      "loss": 1.0551,
      "step": 7025
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.83984375,
      "learning_rate": 1.9708248325964028e-05,
      "loss": 1.107,
      "step": 7030
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.65234375,
      "learning_rate": 1.9642681076534792e-05,
      "loss": 0.9541,
      "step": 7035
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.765625,
      "learning_rate": 1.957719640509321e-05,
      "loss": 1.0616,
      "step": 7040
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.89453125,
      "learning_rate": 1.9511794489770712e-05,
      "loss": 1.0631,
      "step": 7045
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.62109375,
      "learning_rate": 1.94464755084736e-05,
      "loss": 1.0395,
      "step": 7050
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.7265625,
      "learning_rate": 1.938123963888257e-05,
      "loss": 0.9842,
      "step": 7055
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.7734375,
      "learning_rate": 1.9316087058452304e-05,
      "loss": 1.0299,
      "step": 7060
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.8125,
      "learning_rate": 1.9251017944410822e-05,
      "loss": 1.0622,
      "step": 7065
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.8984375,
      "learning_rate": 1.9186032473759147e-05,
      "loss": 1.134,
      "step": 7070
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.796875,
      "learning_rate": 1.91211308232708e-05,
      "loss": 0.9231,
      "step": 7075
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.828125,
      "learning_rate": 1.9056313169491264e-05,
      "loss": 0.928,
      "step": 7080
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.64453125,
      "learning_rate": 1.8991579688737538e-05,
      "loss": 0.8661,
      "step": 7085
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.77734375,
      "learning_rate": 1.8926930557097644e-05,
      "loss": 1.0144,
      "step": 7090
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.796875,
      "learning_rate": 1.8862365950430176e-05,
      "loss": 1.1465,
      "step": 7095
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.6484375,
      "learning_rate": 1.8797886044363788e-05,
      "loss": 0.9353,
      "step": 7100
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.703125,
      "learning_rate": 1.8733491014296772e-05,
      "loss": 1.0119,
      "step": 7105
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.78515625,
      "learning_rate": 1.8669181035396465e-05,
      "loss": 0.9608,
      "step": 7110
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.64453125,
      "learning_rate": 1.8604956282598894e-05,
      "loss": 1.2149,
      "step": 7115
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.921875,
      "learning_rate": 1.854081693060828e-05,
      "loss": 0.9177,
      "step": 7120
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.7109375,
      "learning_rate": 1.8476763153896486e-05,
      "loss": 1.0206,
      "step": 7125
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.80078125,
      "learning_rate": 1.8412795126702638e-05,
      "loss": 1.1597,
      "step": 7130
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.72265625,
      "learning_rate": 1.8348913023032522e-05,
      "loss": 1.0858,
      "step": 7135
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.76171875,
      "learning_rate": 1.8285117016658314e-05,
      "loss": 1.1216,
      "step": 7140
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.93359375,
      "learning_rate": 1.8221407281117913e-05,
      "loss": 1.029,
      "step": 7145
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.796875,
      "learning_rate": 1.815778398971455e-05,
      "loss": 1.0972,
      "step": 7150
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.79296875,
      "learning_rate": 1.8094247315516333e-05,
      "loss": 0.899,
      "step": 7155
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.7734375,
      "learning_rate": 1.803079743135573e-05,
      "loss": 0.8719,
      "step": 7160
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.68359375,
      "learning_rate": 1.7967434509829124e-05,
      "loss": 1.0492,
      "step": 7165
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.79296875,
      "learning_rate": 1.7904158723296398e-05,
      "loss": 1.0277,
      "step": 7170
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.65234375,
      "learning_rate": 1.7840970243880317e-05,
      "loss": 1.0728,
      "step": 7175
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.58203125,
      "learning_rate": 1.7777869243466207e-05,
      "loss": 1.0019,
      "step": 7180
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.828125,
      "learning_rate": 1.771485589370145e-05,
      "loss": 1.0241,
      "step": 7185
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.625,
      "learning_rate": 1.7651930365994966e-05,
      "loss": 1.016,
      "step": 7190
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.765625,
      "learning_rate": 1.75890928315168e-05,
      "loss": 0.8753,
      "step": 7195
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.75390625,
      "learning_rate": 1.7526343461197618e-05,
      "loss": 0.936,
      "step": 7200
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.73046875,
      "learning_rate": 1.746368242572829e-05,
      "loss": 1.093,
      "step": 7205
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.3671875,
      "learning_rate": 1.7401109895559353e-05,
      "loss": 1.2597,
      "step": 7210
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.7421875,
      "learning_rate": 1.7338626040900678e-05,
      "loss": 1.0184,
      "step": 7215
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.75,
      "learning_rate": 1.727623103172082e-05,
      "loss": 1.0955,
      "step": 7220
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.7421875,
      "learning_rate": 1.7213925037746688e-05,
      "loss": 1.0579,
      "step": 7225
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.8671875,
      "learning_rate": 1.7151708228463104e-05,
      "loss": 1.0101,
      "step": 7230
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.8125,
      "learning_rate": 1.7089580773112235e-05,
      "loss": 1.0389,
      "step": 7235
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.79296875,
      "learning_rate": 1.7027542840693205e-05,
      "loss": 0.9624,
      "step": 7240
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.72265625,
      "learning_rate": 1.6965594599961608e-05,
      "loss": 1.0722,
      "step": 7245
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.765625,
      "learning_rate": 1.690373621942908e-05,
      "loss": 1.0638,
      "step": 7250
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.69921875,
      "learning_rate": 1.6841967867362784e-05,
      "loss": 0.9502,
      "step": 7255
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.75,
      "learning_rate": 1.678028971178503e-05,
      "loss": 0.9155,
      "step": 7260
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.7421875,
      "learning_rate": 1.6718701920472746e-05,
      "loss": 1.0022,
      "step": 7265
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.765625,
      "learning_rate": 1.665720466095706e-05,
      "loss": 1.0691,
      "step": 7270
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.77734375,
      "learning_rate": 1.6595798100522813e-05,
      "loss": 1.0905,
      "step": 7275
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.66015625,
      "learning_rate": 1.6534482406208192e-05,
      "loss": 1.0457,
      "step": 7280
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.5859375,
      "learning_rate": 1.6473257744804154e-05,
      "loss": 0.9816,
      "step": 7285
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.69140625,
      "learning_rate": 1.6412124282854002e-05,
      "loss": 1.0526,
      "step": 7290
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.79296875,
      "learning_rate": 1.6351082186653045e-05,
      "loss": 0.9644,
      "step": 7295
    },
    {
      "epoch": 0.74,
      "grad_norm": 2.015625,
      "learning_rate": 1.629013162224799e-05,
      "loss": 1.0989,
      "step": 7300
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.82421875,
      "learning_rate": 1.6229272755436574e-05,
      "loss": 0.982,
      "step": 7305
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.84375,
      "learning_rate": 1.616850575176712e-05,
      "loss": 1.0199,
      "step": 7310
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.74609375,
      "learning_rate": 1.6107830776538034e-05,
      "loss": 0.9466,
      "step": 7315
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.75390625,
      "learning_rate": 1.6047247994797397e-05,
      "loss": 0.9556,
      "step": 7320
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.765625,
      "learning_rate": 1.5986757571342547e-05,
      "loss": 1.0774,
      "step": 7325
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.71875,
      "learning_rate": 1.592635967071951e-05,
      "loss": 0.9787,
      "step": 7330
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.21875,
      "learning_rate": 1.5866054457222667e-05,
      "loss": 1.0819,
      "step": 7335
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.9765625,
      "learning_rate": 1.5805842094894313e-05,
      "loss": 1.0096,
      "step": 7340
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.7421875,
      "learning_rate": 1.5745722747524117e-05,
      "loss": 1.0094,
      "step": 7345
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.8125,
      "learning_rate": 1.5685696578648746e-05,
      "loss": 1.2053,
      "step": 7350
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.0625,
      "learning_rate": 1.5625763751551396e-05,
      "loss": 1.1395,
      "step": 7355
    },
    {
      "epoch": 0.75,
      "grad_norm": 3.984375,
      "learning_rate": 1.5565924429261368e-05,
      "loss": 0.8549,
      "step": 7360
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.6953125,
      "learning_rate": 1.5506178774553585e-05,
      "loss": 0.9433,
      "step": 7365
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.921875,
      "learning_rate": 1.5446526949948243e-05,
      "loss": 1.127,
      "step": 7370
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.73828125,
      "learning_rate": 1.53869691177102e-05,
      "loss": 1.0448,
      "step": 7375
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.64453125,
      "learning_rate": 1.5327505439848704e-05,
      "loss": 1.0138,
      "step": 7380
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.76953125,
      "learning_rate": 1.5268136078116864e-05,
      "loss": 1.0345,
      "step": 7385
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.0234375,
      "learning_rate": 1.5208861194011254e-05,
      "loss": 1.0045,
      "step": 7390
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.7890625,
      "learning_rate": 1.5149680948771439e-05,
      "loss": 1.1314,
      "step": 7395
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.6875,
      "learning_rate": 1.5090595503379484e-05,
      "loss": 0.9383,
      "step": 7400
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.8828125,
      "learning_rate": 1.5031605018559685e-05,
      "loss": 1.0632,
      "step": 7405
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.83203125,
      "learning_rate": 1.4972709654777967e-05,
      "loss": 1.1253,
      "step": 7410
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.703125,
      "learning_rate": 1.4913909572241503e-05,
      "loss": 1.2378,
      "step": 7415
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.65234375,
      "learning_rate": 1.4855204930898303e-05,
      "loss": 0.8846,
      "step": 7420
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.703125,
      "learning_rate": 1.4796595890436748e-05,
      "loss": 1.0447,
      "step": 7425
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.73046875,
      "learning_rate": 1.4738082610285153e-05,
      "loss": 1.1328,
      "step": 7430
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.671875,
      "learning_rate": 1.4679665249611386e-05,
      "loss": 1.0008,
      "step": 7435
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.71484375,
      "learning_rate": 1.462134396732237e-05,
      "loss": 1.0664,
      "step": 7440
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.79296875,
      "learning_rate": 1.4563118922063623e-05,
      "loss": 1.1858,
      "step": 7445
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.7890625,
      "learning_rate": 1.4504990272218988e-05,
      "loss": 1.0722,
      "step": 7450
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.73828125,
      "learning_rate": 1.4446958175910008e-05,
      "loss": 1.1279,
      "step": 7455
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.9453125,
      "learning_rate": 1.438902279099561e-05,
      "loss": 1.0442,
      "step": 7460
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.80859375,
      "learning_rate": 1.4331184275071651e-05,
      "loss": 1.051,
      "step": 7465
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.76171875,
      "learning_rate": 1.4273442785470475e-05,
      "loss": 1.0136,
      "step": 7470
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.7421875,
      "learning_rate": 1.4215798479260484e-05,
      "loss": 1.1228,
      "step": 7475
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.75390625,
      "learning_rate": 1.4158251513245774e-05,
      "loss": 0.8455,
      "step": 7480
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.81640625,
      "learning_rate": 1.4100802043965588e-05,
      "loss": 1.1439,
      "step": 7485
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.8046875,
      "learning_rate": 1.4043450227693978e-05,
      "loss": 1.0626,
      "step": 7490
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.73046875,
      "learning_rate": 1.3986196220439402e-05,
      "loss": 0.9875,
      "step": 7495
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.78125,
      "learning_rate": 1.3929040177944213e-05,
      "loss": 1.2451,
      "step": 7500
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.69140625,
      "learning_rate": 1.3871982255684313e-05,
      "loss": 1.0016,
      "step": 7505
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.8203125,
      "learning_rate": 1.3815022608868634e-05,
      "loss": 1.2081,
      "step": 7510
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.0078125,
      "learning_rate": 1.3758161392438862e-05,
      "loss": 0.9329,
      "step": 7515
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.890625,
      "learning_rate": 1.3701398761068878e-05,
      "loss": 1.0586,
      "step": 7520
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.8984375,
      "learning_rate": 1.3644734869164422e-05,
      "loss": 1.2397,
      "step": 7525
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.76171875,
      "learning_rate": 1.358816987086261e-05,
      "loss": 0.9554,
      "step": 7530
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.6875,
      "learning_rate": 1.3531703920031586e-05,
      "loss": 1.0263,
      "step": 7535
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.74609375,
      "learning_rate": 1.3475337170270013e-05,
      "loss": 0.9861,
      "step": 7540
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.70703125,
      "learning_rate": 1.3419069774906784e-05,
      "loss": 0.9712,
      "step": 7545
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.671875,
      "learning_rate": 1.3362901887000473e-05,
      "loss": 1.1714,
      "step": 7550
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.70703125,
      "learning_rate": 1.3306833659338952e-05,
      "loss": 1.0479,
      "step": 7555
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.7421875,
      "learning_rate": 1.325086524443906e-05,
      "loss": 0.9922,
      "step": 7560
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.7265625,
      "learning_rate": 1.3194996794546094e-05,
      "loss": 1.1754,
      "step": 7565
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.80078125,
      "learning_rate": 1.3139228461633418e-05,
      "loss": 0.9961,
      "step": 7570
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.78125,
      "learning_rate": 1.3083560397402072e-05,
      "loss": 1.1122,
      "step": 7575
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.67578125,
      "learning_rate": 1.3027992753280338e-05,
      "loss": 1.0002,
      "step": 7580
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.75390625,
      "learning_rate": 1.2972525680423325e-05,
      "loss": 1.0337,
      "step": 7585
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.8203125,
      "learning_rate": 1.2917159329712608e-05,
      "loss": 1.1011,
      "step": 7590
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.8046875,
      "learning_rate": 1.2861893851755753e-05,
      "loss": 1.0683,
      "step": 7595
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.7265625,
      "learning_rate": 1.280672939688588e-05,
      "loss": 1.0678,
      "step": 7600
    },
    {
      "epoch": 0.77,
      "grad_norm": 2.046875,
      "learning_rate": 1.2751666115161409e-05,
      "loss": 1.0422,
      "step": 7605
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.7421875,
      "learning_rate": 1.2696704156365463e-05,
      "loss": 1.017,
      "step": 7610
    },
    {
      "epoch": 0.78,
      "grad_norm": 10.875,
      "learning_rate": 1.2641843670005593e-05,
      "loss": 0.9963,
      "step": 7615
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.703125,
      "learning_rate": 1.258708480531331e-05,
      "loss": 0.8597,
      "step": 7620
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.66796875,
      "learning_rate": 1.2532427711243682e-05,
      "loss": 1.1199,
      "step": 7625
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.6796875,
      "learning_rate": 1.2477872536474961e-05,
      "loss": 0.9385,
      "step": 7630
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.64453125,
      "learning_rate": 1.2423419429408146e-05,
      "loss": 0.9348,
      "step": 7635
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.6171875,
      "learning_rate": 1.236906853816659e-05,
      "loss": 0.9485,
      "step": 7640
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.703125,
      "learning_rate": 1.2314820010595612e-05,
      "loss": 0.8493,
      "step": 7645
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.76171875,
      "learning_rate": 1.2260673994262056e-05,
      "loss": 1.0026,
      "step": 7650
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.671875,
      "learning_rate": 1.2206630636453953e-05,
      "loss": 1.18,
      "step": 7655
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.76171875,
      "learning_rate": 1.2152690084180068e-05,
      "loss": 1.0291,
      "step": 7660
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.82421875,
      "learning_rate": 1.209885248416946e-05,
      "loss": 1.2367,
      "step": 7665
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.0859375,
      "learning_rate": 1.2045117982871229e-05,
      "loss": 1.0427,
      "step": 7670
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.7265625,
      "learning_rate": 1.1991486726453965e-05,
      "loss": 1.0732,
      "step": 7675
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.85546875,
      "learning_rate": 1.1937958860805427e-05,
      "loss": 1.0942,
      "step": 7680
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.84375,
      "learning_rate": 1.1884534531532132e-05,
      "loss": 1.1168,
      "step": 7685
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.66796875,
      "learning_rate": 1.1831213883958957e-05,
      "loss": 0.9727,
      "step": 7690
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.76171875,
      "learning_rate": 1.177799706312872e-05,
      "loss": 1.0578,
      "step": 7695
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.75,
      "learning_rate": 1.1724884213801873e-05,
      "loss": 1.0348,
      "step": 7700
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.8046875,
      "learning_rate": 1.1671875480455996e-05,
      "loss": 1.1907,
      "step": 7705
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.75,
      "learning_rate": 1.1618971007285423e-05,
      "loss": 1.0709,
      "step": 7710
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.6875,
      "learning_rate": 1.1566170938200954e-05,
      "loss": 1.1659,
      "step": 7715
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.76953125,
      "learning_rate": 1.1513475416829344e-05,
      "loss": 1.1204,
      "step": 7720
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.86328125,
      "learning_rate": 1.146088458651296e-05,
      "loss": 1.0912,
      "step": 7725
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.6953125,
      "learning_rate": 1.14083985903094e-05,
      "loss": 1.1206,
      "step": 7730
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.1640625,
      "learning_rate": 1.1356017570991079e-05,
      "loss": 1.1407,
      "step": 7735
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.6953125,
      "learning_rate": 1.1303741671044848e-05,
      "loss": 1.0298,
      "step": 7740
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.015625,
      "learning_rate": 1.125157103267166e-05,
      "loss": 1.1292,
      "step": 7745
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.73828125,
      "learning_rate": 1.1199505797786091e-05,
      "loss": 1.1263,
      "step": 7750
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.8515625,
      "learning_rate": 1.1147546108015983e-05,
      "loss": 1.0643,
      "step": 7755
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.796875,
      "learning_rate": 1.1095692104702104e-05,
      "loss": 1.1432,
      "step": 7760
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.62109375,
      "learning_rate": 1.1043943928897754e-05,
      "loss": 0.9644,
      "step": 7765
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.7734375,
      "learning_rate": 1.0992301721368319e-05,
      "loss": 1.1377,
      "step": 7770
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.73046875,
      "learning_rate": 1.0940765622590949e-05,
      "loss": 1.1033,
      "step": 7775
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.77734375,
      "learning_rate": 1.088933577275415e-05,
      "loss": 1.1233,
      "step": 7780
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.74609375,
      "learning_rate": 1.0838012311757423e-05,
      "loss": 1.111,
      "step": 7785
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.0078125,
      "learning_rate": 1.0786795379210856e-05,
      "loss": 1.0881,
      "step": 7790
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.76953125,
      "learning_rate": 1.0735685114434763e-05,
      "loss": 1.0869,
      "step": 7795
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.8203125,
      "learning_rate": 1.0684681656459305e-05,
      "loss": 1.1783,
      "step": 7800
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.63671875,
      "learning_rate": 1.0633785144024084e-05,
      "loss": 1.0289,
      "step": 7805
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.7734375,
      "learning_rate": 1.0582995715577843e-05,
      "loss": 1.0741,
      "step": 7810
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.71875,
      "learning_rate": 1.0532313509277992e-05,
      "loss": 1.0633,
      "step": 7815
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.6953125,
      "learning_rate": 1.0481738662990242e-05,
      "loss": 1.0032,
      "step": 7820
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.74609375,
      "learning_rate": 1.0431271314288348e-05,
      "loss": 0.908,
      "step": 7825
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.82421875,
      "learning_rate": 1.0380911600453597e-05,
      "loss": 1.191,
      "step": 7830
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.7578125,
      "learning_rate": 1.033065965847449e-05,
      "loss": 0.8685,
      "step": 7835
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.765625,
      "learning_rate": 1.0280515625046378e-05,
      "loss": 0.92,
      "step": 7840
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.8046875,
      "learning_rate": 1.0230479636571078e-05,
      "loss": 1.0653,
      "step": 7845
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.76953125,
      "learning_rate": 1.0180551829156476e-05,
      "loss": 1.1654,
      "step": 7850
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.66015625,
      "learning_rate": 1.0130732338616244e-05,
      "loss": 1.0986,
      "step": 7855
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.79296875,
      "learning_rate": 1.008102130046938e-05,
      "loss": 1.0657,
      "step": 7860
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.73046875,
      "learning_rate": 1.0031418849939806e-05,
      "loss": 0.9849,
      "step": 7865
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.1328125,
      "learning_rate": 9.981925121956177e-06,
      "loss": 0.9419,
      "step": 7870
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.9296875,
      "learning_rate": 9.932540251151328e-06,
      "loss": 1.2451,
      "step": 7875
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.79296875,
      "learning_rate": 9.883264371862e-06,
      "loss": 1.0926,
      "step": 7880
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.69140625,
      "learning_rate": 9.834097618128458e-06,
      "loss": 0.99,
      "step": 7885
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.8671875,
      "learning_rate": 9.785040123694107e-06,
      "loss": 1.0726,
      "step": 7890
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.7265625,
      "learning_rate": 9.736092022005172e-06,
      "loss": 1.2484,
      "step": 7895
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.73046875,
      "learning_rate": 9.687253446210271e-06,
      "loss": 0.9666,
      "step": 7900
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.76171875,
      "learning_rate": 9.638524529160147e-06,
      "loss": 0.9516,
      "step": 7905
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.75390625,
      "learning_rate": 9.58990540340719e-06,
      "loss": 1.0633,
      "step": 7910
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.703125,
      "learning_rate": 9.541396201205149e-06,
      "loss": 0.9372,
      "step": 7915
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.66015625,
      "learning_rate": 9.492997054508795e-06,
      "loss": 0.9173,
      "step": 7920
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.703125,
      "learning_rate": 9.444708094973498e-06,
      "loss": 1.0616,
      "step": 7925
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.78125,
      "learning_rate": 9.396529453954889e-06,
      "loss": 1.2095,
      "step": 7930
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.73046875,
      "learning_rate": 9.348461262508523e-06,
      "loss": 1.1098,
      "step": 7935
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.70703125,
      "learning_rate": 9.300503651389514e-06,
      "loss": 0.9954,
      "step": 7940
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.0546875,
      "learning_rate": 9.252656751052163e-06,
      "loss": 1.09,
      "step": 7945
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.6953125,
      "learning_rate": 9.20492069164962e-06,
      "loss": 0.9648,
      "step": 7950
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.79296875,
      "learning_rate": 9.157295603033539e-06,
      "loss": 1.1242,
      "step": 7955
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.71484375,
      "learning_rate": 9.109781614753682e-06,
      "loss": 0.971,
      "step": 7960
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.73828125,
      "learning_rate": 9.062378856057629e-06,
      "loss": 0.9582,
      "step": 7965
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.7109375,
      "learning_rate": 9.01508745589038e-06,
      "loss": 1.0426,
      "step": 7970
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.765625,
      "learning_rate": 8.967907542894006e-06,
      "loss": 1.0823,
      "step": 7975
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.75390625,
      "learning_rate": 8.92083924540732e-06,
      "loss": 1.071,
      "step": 7980
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.76953125,
      "learning_rate": 8.873882691465523e-06,
      "loss": 1.0752,
      "step": 7985
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.0546875,
      "learning_rate": 8.827038008799843e-06,
      "loss": 0.9109,
      "step": 7990
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.73046875,
      "learning_rate": 8.780305324837201e-06,
      "loss": 0.9449,
      "step": 7995
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.7109375,
      "learning_rate": 8.733684766699862e-06,
      "loss": 1.0107,
      "step": 8000
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.74609375,
      "learning_rate": 8.68717646120507e-06,
      "loss": 1.0883,
      "step": 8005
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.73046875,
      "learning_rate": 8.640780534864729e-06,
      "loss": 1.058,
      "step": 8010
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.78125,
      "learning_rate": 8.594497113885081e-06,
      "loss": 1.0544,
      "step": 8015
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.8203125,
      "learning_rate": 8.548326324166267e-06,
      "loss": 1.1859,
      "step": 8020
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.90625,
      "learning_rate": 8.502268291302085e-06,
      "loss": 1.0732,
      "step": 8025
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.6484375,
      "learning_rate": 8.456323140579631e-06,
      "loss": 0.9538,
      "step": 8030
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.9453125,
      "learning_rate": 8.410490996978908e-06,
      "loss": 0.9876,
      "step": 8035
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.71875,
      "learning_rate": 8.36477198517252e-06,
      "loss": 0.9445,
      "step": 8040
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.7265625,
      "learning_rate": 8.319166229525344e-06,
      "loss": 0.9873,
      "step": 8045
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.75,
      "learning_rate": 8.27367385409416e-06,
      "loss": 1.0131,
      "step": 8050
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.73828125,
      "learning_rate": 8.228294982627337e-06,
      "loss": 1.0492,
      "step": 8055
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.66796875,
      "learning_rate": 8.18302973856453e-06,
      "loss": 0.9502,
      "step": 8060
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.93359375,
      "learning_rate": 8.137878245036229e-06,
      "loss": 1.1096,
      "step": 8065
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.58984375,
      "learning_rate": 8.092840624863535e-06,
      "loss": 0.993,
      "step": 8070
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.765625,
      "learning_rate": 8.047917000557832e-06,
      "loss": 0.9962,
      "step": 8075
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.98046875,
      "learning_rate": 8.003107494320356e-06,
      "loss": 1.0157,
      "step": 8080
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.70703125,
      "learning_rate": 7.95841222804194e-06,
      "loss": 0.9748,
      "step": 8085
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.84375,
      "learning_rate": 7.913831323302656e-06,
      "loss": 0.9751,
      "step": 8090
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.4140625,
      "learning_rate": 7.869364901371501e-06,
      "loss": 1.1196,
      "step": 8095
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.85546875,
      "learning_rate": 7.82501308320603e-06,
      "loss": 1.0587,
      "step": 8100
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.7890625,
      "learning_rate": 7.780775989452117e-06,
      "loss": 0.9451,
      "step": 8105
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.7890625,
      "learning_rate": 7.736653740443466e-06,
      "loss": 1.1029,
      "step": 8110
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.92578125,
      "learning_rate": 7.692646456201453e-06,
      "loss": 0.8981,
      "step": 8115
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.72265625,
      "learning_rate": 7.64875425643471e-06,
      "loss": 1.119,
      "step": 8120
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.78515625,
      "learning_rate": 7.604977260538804e-06,
      "loss": 1.1263,
      "step": 8125
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.73046875,
      "learning_rate": 7.561315587595946e-06,
      "loss": 1.0309,
      "step": 8130
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.61328125,
      "learning_rate": 7.517769356374599e-06,
      "loss": 0.9766,
      "step": 8135
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.66796875,
      "learning_rate": 7.474338685329263e-06,
      "loss": 0.9325,
      "step": 8140
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.734375,
      "learning_rate": 7.431023692600059e-06,
      "loss": 1.0075,
      "step": 8145
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.8046875,
      "learning_rate": 7.38782449601243e-06,
      "loss": 1.1524,
      "step": 8150
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.7578125,
      "learning_rate": 7.34474121307685e-06,
      "loss": 1.214,
      "step": 8155
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.61328125,
      "learning_rate": 7.301773960988479e-06,
      "loss": 0.9925,
      "step": 8160
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.71484375,
      "learning_rate": 7.258922856626837e-06,
      "loss": 1.1018,
      "step": 8165
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.79296875,
      "learning_rate": 7.216188016555542e-06,
      "loss": 0.9746,
      "step": 8170
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.66015625,
      "learning_rate": 7.173569557021881e-06,
      "loss": 0.8846,
      "step": 8175
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.73828125,
      "learning_rate": 7.131067593956608e-06,
      "loss": 1.0179,
      "step": 8180
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.8046875,
      "learning_rate": 7.0886822429735875e-06,
      "loss": 1.0651,
      "step": 8185
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.7734375,
      "learning_rate": 7.046413619369463e-06,
      "loss": 0.9921,
      "step": 8190
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.71875,
      "learning_rate": 7.004261838123355e-06,
      "loss": 1.1201,
      "step": 8195
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.734375,
      "learning_rate": 6.962227013896544e-06,
      "loss": 1.1401,
      "step": 8200
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.8828125,
      "learning_rate": 6.92030926103217e-06,
      "loss": 1.0509,
      "step": 8205
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.74609375,
      "learning_rate": 6.87850869355492e-06,
      "loss": 1.1444,
      "step": 8210
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.79296875,
      "learning_rate": 6.8368254251707234e-06,
      "loss": 1.1066,
      "step": 8215
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.7265625,
      "learning_rate": 6.795259569266399e-06,
      "loss": 0.9891,
      "step": 8220
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.7890625,
      "learning_rate": 6.7538112389093965e-06,
      "loss": 1.1345,
      "step": 8225
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.76171875,
      "learning_rate": 6.7124805468475e-06,
      "loss": 0.9894,
      "step": 8230
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.7109375,
      "learning_rate": 6.671267605508452e-06,
      "loss": 0.9833,
      "step": 8235
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.7109375,
      "learning_rate": 6.630172526999712e-06,
      "loss": 0.9544,
      "step": 8240
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.8515625,
      "learning_rate": 6.589195423108124e-06,
      "loss": 1.203,
      "step": 8245
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.6953125,
      "learning_rate": 6.548336405299621e-06,
      "loss": 0.9743,
      "step": 8250
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.8203125,
      "learning_rate": 6.507595584718912e-06,
      "loss": 1.1241,
      "step": 8255
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.78125,
      "learning_rate": 6.466973072189186e-06,
      "loss": 1.0756,
      "step": 8260
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.77734375,
      "learning_rate": 6.426468978211819e-06,
      "loss": 1.0216,
      "step": 8265
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.7890625,
      "learning_rate": 6.386083412966054e-06,
      "loss": 1.114,
      "step": 8270
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.765625,
      "learning_rate": 6.345816486308709e-06,
      "loss": 0.9978,
      "step": 8275
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.85546875,
      "learning_rate": 6.305668307773915e-06,
      "loss": 1.1581,
      "step": 8280
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.7265625,
      "learning_rate": 6.265638986572747e-06,
      "loss": 0.9827,
      "step": 8285
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.65234375,
      "learning_rate": 6.225728631592958e-06,
      "loss": 0.9278,
      "step": 8290
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.67578125,
      "learning_rate": 6.18593735139873e-06,
      "loss": 0.9989,
      "step": 8295
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.73828125,
      "learning_rate": 6.146265254230304e-06,
      "loss": 1.0704,
      "step": 8300
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.63671875,
      "learning_rate": 6.106712448003738e-06,
      "loss": 0.9833,
      "step": 8305
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.7890625,
      "learning_rate": 6.067279040310575e-06,
      "loss": 1.0683,
      "step": 8310
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.67578125,
      "learning_rate": 6.027965138417591e-06,
      "loss": 1.0245,
      "step": 8315
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.7421875,
      "learning_rate": 5.988770849266462e-06,
      "loss": 1.0429,
      "step": 8320
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.0234375,
      "learning_rate": 5.949696279473533e-06,
      "loss": 1.0661,
      "step": 8325
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.734375,
      "learning_rate": 5.9107415353294285e-06,
      "loss": 1.1553,
      "step": 8330
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.7265625,
      "learning_rate": 5.871906722798859e-06,
      "loss": 1.0681,
      "step": 8335
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.9765625,
      "learning_rate": 5.833191947520311e-06,
      "loss": 1.0529,
      "step": 8340
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.69921875,
      "learning_rate": 5.794597314805711e-06,
      "loss": 1.0056,
      "step": 8345
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.7578125,
      "learning_rate": 5.7561229296402005e-06,
      "loss": 0.9506,
      "step": 8350
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.8046875,
      "learning_rate": 5.717768896681808e-06,
      "loss": 1.0626,
      "step": 8355
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.75,
      "learning_rate": 5.679535320261187e-06,
      "loss": 0.9996,
      "step": 8360
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.9765625,
      "learning_rate": 5.641422304381305e-06,
      "loss": 0.9982,
      "step": 8365
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.79296875,
      "learning_rate": 5.603429952717238e-06,
      "loss": 1.1246,
      "step": 8370
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.69140625,
      "learning_rate": 5.5655583686157585e-06,
      "loss": 1.1155,
      "step": 8375
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.7578125,
      "learning_rate": 5.527807655095174e-06,
      "loss": 1.0422,
      "step": 8380
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.6640625,
      "learning_rate": 5.490177914844968e-06,
      "loss": 1.0936,
      "step": 8385
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.7890625,
      "learning_rate": 5.452669250225601e-06,
      "loss": 1.1358,
      "step": 8390
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.66796875,
      "learning_rate": 5.415281763268143e-06,
      "loss": 0.98,
      "step": 8395
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.0078125,
      "learning_rate": 5.378015555674026e-06,
      "loss": 1.0793,
      "step": 8400
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.72265625,
      "learning_rate": 5.340870728814823e-06,
      "loss": 1.0334,
      "step": 8405
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.8203125,
      "learning_rate": 5.303847383731891e-06,
      "loss": 1.2618,
      "step": 8410
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.671875,
      "learning_rate": 5.266945621136149e-06,
      "loss": 1.0013,
      "step": 8415
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.60546875,
      "learning_rate": 5.2301655414077834e-06,
      "loss": 0.9661,
      "step": 8420
    },
    {
      "epoch": 0.86,
      "grad_norm": 2.234375,
      "learning_rate": 5.193507244595969e-06,
      "loss": 0.9258,
      "step": 8425
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.80859375,
      "learning_rate": 5.1569708304186174e-06,
      "loss": 1.1952,
      "step": 8430
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.73828125,
      "learning_rate": 5.120556398262099e-06,
      "loss": 1.0236,
      "step": 8435
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.8515625,
      "learning_rate": 5.084264047180976e-06,
      "loss": 0.8937,
      "step": 8440
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.7890625,
      "learning_rate": 5.048093875897674e-06,
      "loss": 1.1052,
      "step": 8445
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.7734375,
      "learning_rate": 5.012045982802332e-06,
      "loss": 1.1172,
      "step": 8450
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.80859375,
      "learning_rate": 4.976120465952438e-06,
      "loss": 0.9773,
      "step": 8455
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.7421875,
      "learning_rate": 4.940317423072588e-06,
      "loss": 1.0267,
      "step": 8460
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.88671875,
      "learning_rate": 4.904636951554226e-06,
      "loss": 1.0397,
      "step": 8465
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.8203125,
      "learning_rate": 4.869079148455385e-06,
      "loss": 1.067,
      "step": 8470
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.83984375,
      "learning_rate": 4.8336441105004016e-06,
      "loss": 1.2063,
      "step": 8475
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.82421875,
      "learning_rate": 4.798331934079709e-06,
      "loss": 1.2085,
      "step": 8480
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.7734375,
      "learning_rate": 4.763142715249458e-06,
      "loss": 0.8973,
      "step": 8485
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.7421875,
      "learning_rate": 4.728076549731386e-06,
      "loss": 1.0066,
      "step": 8490
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.73828125,
      "learning_rate": 4.693133532912497e-06,
      "loss": 1.1551,
      "step": 8495
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.76171875,
      "learning_rate": 4.65831375984479e-06,
      "loss": 0.9847,
      "step": 8500
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.63671875,
      "learning_rate": 4.623617325245027e-06,
      "loss": 0.9894,
      "step": 8505
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.83203125,
      "learning_rate": 4.589044323494435e-06,
      "loss": 1.1926,
      "step": 8510
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.65234375,
      "learning_rate": 4.554594848638533e-06,
      "loss": 1.0448,
      "step": 8515
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.78515625,
      "learning_rate": 4.520268994386778e-06,
      "loss": 0.9993,
      "step": 8520
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.99609375,
      "learning_rate": 4.4860668541123695e-06,
      "loss": 0.9977,
      "step": 8525
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.70703125,
      "learning_rate": 4.451988520851985e-06,
      "loss": 0.9981,
      "step": 8530
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.79296875,
      "learning_rate": 4.418034087305506e-06,
      "loss": 1.0088,
      "step": 8535
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.84765625,
      "learning_rate": 4.3842036458357935e-06,
      "loss": 1.1097,
      "step": 8540
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.296875,
      "learning_rate": 4.350497288468436e-06,
      "loss": 0.9324,
      "step": 8545
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.80078125,
      "learning_rate": 4.316915106891484e-06,
      "loss": 0.9506,
      "step": 8550
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.82421875,
      "learning_rate": 4.283457192455159e-06,
      "loss": 1.1521,
      "step": 8555
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.7421875,
      "learning_rate": 4.2501236361717266e-06,
      "loss": 1.2179,
      "step": 8560
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.83203125,
      "learning_rate": 4.2169145287151215e-06,
      "loss": 1.1016,
      "step": 8565
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.68359375,
      "learning_rate": 4.183829960420765e-06,
      "loss": 0.9785,
      "step": 8570
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.6640625,
      "learning_rate": 4.150870021285319e-06,
      "loss": 0.9116,
      "step": 8575
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.76953125,
      "learning_rate": 4.118034800966408e-06,
      "loss": 1.0457,
      "step": 8580
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.6640625,
      "learning_rate": 4.085324388782407e-06,
      "loss": 0.9744,
      "step": 8585
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.8359375,
      "learning_rate": 4.052738873712208e-06,
      "loss": 1.0935,
      "step": 8590
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.7890625,
      "learning_rate": 4.020278344394929e-06,
      "loss": 1.0172,
      "step": 8595
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.73828125,
      "learning_rate": 3.987942889129692e-06,
      "loss": 0.9883,
      "step": 8600
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.7265625,
      "learning_rate": 3.955732595875439e-06,
      "loss": 0.8696,
      "step": 8605
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.72265625,
      "learning_rate": 3.923647552250604e-06,
      "loss": 1.0511,
      "step": 8610
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.83203125,
      "learning_rate": 3.891687845532932e-06,
      "loss": 1.0887,
      "step": 8615
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.86328125,
      "learning_rate": 3.859853562659232e-06,
      "loss": 1.1004,
      "step": 8620
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.921875,
      "learning_rate": 3.82814479022513e-06,
      "loss": 0.8922,
      "step": 8625
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.69921875,
      "learning_rate": 3.7965616144848337e-06,
      "loss": 1.1908,
      "step": 8630
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.82421875,
      "learning_rate": 3.7651041213509043e-06,
      "loss": 0.9542,
      "step": 8635
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.8671875,
      "learning_rate": 3.733772396394031e-06,
      "loss": 1.2244,
      "step": 8640
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.69921875,
      "learning_rate": 3.7025665248427766e-06,
      "loss": 0.9537,
      "step": 8645
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.78515625,
      "learning_rate": 3.6714865915833563e-06,
      "loss": 1.1107,
      "step": 8650
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.8046875,
      "learning_rate": 3.6405326811594256e-06,
      "loss": 1.0089,
      "step": 8655
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.75390625,
      "learning_rate": 3.6097048777718246e-06,
      "loss": 1.3491,
      "step": 8660
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.7578125,
      "learning_rate": 3.5790032652783247e-06,
      "loss": 1.143,
      "step": 8665
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.8046875,
      "learning_rate": 3.5484279271934873e-06,
      "loss": 1.1867,
      "step": 8670
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.1015625,
      "learning_rate": 3.5179789466883495e-06,
      "loss": 1.1987,
      "step": 8675
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.7734375,
      "learning_rate": 3.487656406590234e-06,
      "loss": 0.9052,
      "step": 8680
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.015625,
      "learning_rate": 3.4574603893825274e-06,
      "loss": 1.0106,
      "step": 8685
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.7109375,
      "learning_rate": 3.427390977204442e-06,
      "loss": 1.1506,
      "step": 8690
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.71484375,
      "learning_rate": 3.397448251850788e-06,
      "loss": 0.9958,
      "step": 8695
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.796875,
      "learning_rate": 3.3676322947717997e-06,
      "loss": 1.2572,
      "step": 8700
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.94140625,
      "learning_rate": 3.337943187072845e-06,
      "loss": 0.8616,
      "step": 8705
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.77734375,
      "learning_rate": 3.3083810095142278e-06,
      "loss": 1.0765,
      "step": 8710
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.7890625,
      "learning_rate": 3.2789458425110054e-06,
      "loss": 1.2654,
      "step": 8715
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.7578125,
      "learning_rate": 3.2496377661327294e-06,
      "loss": 1.0817,
      "step": 8720
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.65234375,
      "learning_rate": 3.2204568601032327e-06,
      "loss": 0.9803,
      "step": 8725
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.67578125,
      "learning_rate": 3.191403203800425e-06,
      "loss": 1.1562,
      "step": 8730
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.76171875,
      "learning_rate": 3.162476876256071e-06,
      "loss": 1.1149,
      "step": 8735
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.71484375,
      "learning_rate": 3.133677956155567e-06,
      "loss": 1.3157,
      "step": 8740
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.73046875,
      "learning_rate": 3.1050065218377556e-06,
      "loss": 0.9424,
      "step": 8745
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.73046875,
      "learning_rate": 3.076462651294687e-06,
      "loss": 0.9379,
      "step": 8750
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.734375,
      "learning_rate": 3.048046422171391e-06,
      "loss": 0.8838,
      "step": 8755
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.7578125,
      "learning_rate": 3.0197579117656972e-06,
      "loss": 1.1311,
      "step": 8760
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.77734375,
      "learning_rate": 2.9915971970280344e-06,
      "loss": 1.0632,
      "step": 8765
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.80859375,
      "learning_rate": 2.9635643545611826e-06,
      "loss": 1.2597,
      "step": 8770
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.80078125,
      "learning_rate": 2.9356594606200883e-06,
      "loss": 0.9993,
      "step": 8775
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.84765625,
      "learning_rate": 2.9078825911116493e-06,
      "loss": 1.0114,
      "step": 8780
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.7109375,
      "learning_rate": 2.880233821594519e-06,
      "loss": 1.1114,
      "step": 8785
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.80078125,
      "learning_rate": 2.8527132272788806e-06,
      "loss": 1.2655,
      "step": 8790
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.8125,
      "learning_rate": 2.825320883026267e-06,
      "loss": 1.0745,
      "step": 8795
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.74609375,
      "learning_rate": 2.7980568633493463e-06,
      "loss": 1.1723,
      "step": 8800
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.65625,
      "learning_rate": 2.7709212424116946e-06,
      "loss": 0.9589,
      "step": 8805
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.703125,
      "learning_rate": 2.7439140940276554e-06,
      "loss": 1.1502,
      "step": 8810
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.73046875,
      "learning_rate": 2.7170354916620842e-06,
      "loss": 1.0675,
      "step": 8815
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.72265625,
      "learning_rate": 2.690285508430135e-06,
      "loss": 0.977,
      "step": 8820
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.70703125,
      "learning_rate": 2.6636642170971506e-06,
      "loss": 1.1135,
      "step": 8825
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.69921875,
      "learning_rate": 2.6371716900783606e-06,
      "loss": 0.966,
      "step": 8830
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.86328125,
      "learning_rate": 2.610807999438747e-06,
      "loss": 1.0174,
      "step": 8835
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.69140625,
      "learning_rate": 2.584573216892827e-06,
      "loss": 1.0872,
      "step": 8840
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.734375,
      "learning_rate": 2.5584674138044606e-06,
      "loss": 1.0395,
      "step": 8845
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.8203125,
      "learning_rate": 2.532490661186665e-06,
      "loss": 1.1921,
      "step": 8850
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.8203125,
      "learning_rate": 2.5066430297014165e-06,
      "loss": 1.1705,
      "step": 8855
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.7265625,
      "learning_rate": 2.4809245896594448e-06,
      "loss": 1.1343,
      "step": 8860
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.73046875,
      "learning_rate": 2.4553354110200487e-06,
      "loss": 0.9917,
      "step": 8865
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.7890625,
      "learning_rate": 2.429875563390932e-06,
      "loss": 1.0901,
      "step": 8870
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.6640625,
      "learning_rate": 2.4045451160279785e-06,
      "loss": 1.1292,
      "step": 8875
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.6953125,
      "learning_rate": 2.3793441378350713e-06,
      "loss": 1.1216,
      "step": 8880
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.79296875,
      "learning_rate": 2.3542726973639184e-06,
      "loss": 1.0914,
      "step": 8885
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.75,
      "learning_rate": 2.3293308628138668e-06,
      "loss": 0.9495,
      "step": 8890
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.70703125,
      "learning_rate": 2.3045187020316994e-06,
      "loss": 0.9418,
      "step": 8895
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.71484375,
      "learning_rate": 2.2798362825114494e-06,
      "loss": 1.035,
      "step": 8900
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.80859375,
      "learning_rate": 2.2552836713942705e-06,
      "loss": 1.0294,
      "step": 8905
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.734375,
      "learning_rate": 2.2308609354681597e-06,
      "loss": 1.1058,
      "step": 8910
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.76953125,
      "learning_rate": 2.206568141167853e-06,
      "loss": 0.9427,
      "step": 8915
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.68359375,
      "learning_rate": 2.1824053545746236e-06,
      "loss": 1.0727,
      "step": 8920
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.7890625,
      "learning_rate": 2.1583726414160843e-06,
      "loss": 0.9588,
      "step": 8925
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.5859375,
      "learning_rate": 2.13447006706603e-06,
      "loss": 0.9136,
      "step": 8930
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.76171875,
      "learning_rate": 2.110697696544256e-06,
      "loss": 1.1282,
      "step": 8935
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.953125,
      "learning_rate": 2.0870555945163627e-06,
      "loss": 1.2367,
      "step": 8940
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.734375,
      "learning_rate": 2.0635438252936124e-06,
      "loss": 0.9674,
      "step": 8945
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.68359375,
      "learning_rate": 2.0401624528327114e-06,
      "loss": 0.9966,
      "step": 8950
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.81640625,
      "learning_rate": 2.0169115407356943e-06,
      "loss": 1.1108,
      "step": 8955
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.81640625,
      "learning_rate": 1.99379115224968e-06,
      "loss": 1.0323,
      "step": 8960
    },
    {
      "epoch": 0.91,
      "grad_norm": 1.53125,
      "learning_rate": 1.970801350266771e-06,
      "loss": 1.1575,
      "step": 8965
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.94921875,
      "learning_rate": 1.9479421973238322e-06,
      "loss": 1.132,
      "step": 8970
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.6953125,
      "learning_rate": 1.9252137556023233e-06,
      "loss": 1.1065,
      "step": 8975
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.7890625,
      "learning_rate": 1.9026160869281773e-06,
      "loss": 1.279,
      "step": 8980
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.66015625,
      "learning_rate": 1.880149252771568e-06,
      "loss": 0.9565,
      "step": 8985
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.7109375,
      "learning_rate": 1.8578133142467812e-06,
      "loss": 1.0308,
      "step": 8990
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.77734375,
      "learning_rate": 1.8356083321120542e-06,
      "loss": 0.8321,
      "step": 8995
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.7578125,
      "learning_rate": 1.8135343667693816e-06,
      "loss": 1.1632,
      "step": 9000
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.7734375,
      "learning_rate": 1.791591478264365e-06,
      "loss": 1.0824,
      "step": 9005
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.76953125,
      "learning_rate": 1.7697797262860527e-06,
      "loss": 1.0118,
      "step": 9010
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.6484375,
      "learning_rate": 1.7480991701668003e-06,
      "loss": 0.8796,
      "step": 9015
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.75390625,
      "learning_rate": 1.726549868882038e-06,
      "loss": 1.0019,
      "step": 9020
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.83984375,
      "learning_rate": 1.705131881050187e-06,
      "loss": 1.0832,
      "step": 9025
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.6328125,
      "learning_rate": 1.6838452649324765e-06,
      "loss": 0.9772,
      "step": 9030
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.8203125,
      "learning_rate": 1.6626900784327603e-06,
      "loss": 1.0851,
      "step": 9035
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.6796875,
      "learning_rate": 1.641666379097373e-06,
      "loss": 0.9743,
      "step": 9040
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.63671875,
      "learning_rate": 1.6207742241150016e-06,
      "loss": 1.0845,
      "step": 9045
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.75,
      "learning_rate": 1.6000136703164803e-06,
      "loss": 1.0218,
      "step": 9050
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.71484375,
      "learning_rate": 1.5793847741746748e-06,
      "loss": 1.1603,
      "step": 9055
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.6328125,
      "learning_rate": 1.5588875918043255e-06,
      "loss": 1.028,
      "step": 9060
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.8046875,
      "learning_rate": 1.5385221789618654e-06,
      "loss": 1.0683,
      "step": 9065
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.75,
      "learning_rate": 1.5182885910452914e-06,
      "loss": 0.958,
      "step": 9070
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.84375,
      "learning_rate": 1.4981868830940216e-06,
      "loss": 1.0407,
      "step": 9075
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.69921875,
      "learning_rate": 1.478217109788732e-06,
      "loss": 1.0951,
      "step": 9080
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.8203125,
      "learning_rate": 1.4583793254511979e-06,
      "loss": 1.2087,
      "step": 9085
    },
    {
      "epoch": 0.93,
      "grad_norm": 1.2265625,
      "learning_rate": 1.43867358404417e-06,
      "loss": 1.0511,
      "step": 9090
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.8671875,
      "learning_rate": 1.4190999391712146e-06,
      "loss": 1.2468,
      "step": 9095
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.75,
      "learning_rate": 1.399658444076557e-06,
      "loss": 1.1579,
      "step": 9100
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.70703125,
      "learning_rate": 1.3803491516449608e-06,
      "loss": 0.9643,
      "step": 9105
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.796875,
      "learning_rate": 1.3611721144015765e-06,
      "loss": 1.0911,
      "step": 9110
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.71875,
      "learning_rate": 1.342127384511771e-06,
      "loss": 0.9783,
      "step": 9115
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.65625,
      "learning_rate": 1.3232150137810428e-06,
      "loss": 0.9269,
      "step": 9120
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.74609375,
      "learning_rate": 1.3044350536548178e-06,
      "loss": 1.1495,
      "step": 9125
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.84375,
      "learning_rate": 1.2857875552183485e-06,
      "loss": 1.0372,
      "step": 9130
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.6640625,
      "learning_rate": 1.2672725691965648e-06,
      "loss": 1.0084,
      "step": 9135
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.68359375,
      "learning_rate": 1.2488901459539403e-06,
      "loss": 1.1416,
      "step": 9140
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.71484375,
      "learning_rate": 1.230640335494354e-06,
      "loss": 1.0988,
      "step": 9145
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.67578125,
      "learning_rate": 1.2125231874609288e-06,
      "loss": 1.0674,
      "step": 9150
    },
    {
      "epoch": 0.93,
      "grad_norm": 1.1015625,
      "learning_rate": 1.194538751135954e-06,
      "loss": 1.042,
      "step": 9155
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.765625,
      "learning_rate": 1.1766870754406855e-06,
      "loss": 1.0034,
      "step": 9160
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.7109375,
      "learning_rate": 1.158968208935257e-06,
      "loss": 1.2456,
      "step": 9165
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.921875,
      "learning_rate": 1.1413821998185526e-06,
      "loss": 1.0782,
      "step": 9170
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.78515625,
      "learning_rate": 1.1239290959280168e-06,
      "loss": 1.1104,
      "step": 9175
    },
    {
      "epoch": 0.93,
      "grad_norm": 2.59375,
      "learning_rate": 1.10660894473959e-06,
      "loss": 0.9997,
      "step": 9180
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.94140625,
      "learning_rate": 1.0894217933675677e-06,
      "loss": 1.1067,
      "step": 9185
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.8984375,
      "learning_rate": 1.0723676885644295e-06,
      "loss": 1.1516,
      "step": 9190
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.7265625,
      "learning_rate": 1.0554466767207672e-06,
      "loss": 1.1205,
      "step": 9195
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.82421875,
      "learning_rate": 1.0386588038651167e-06,
      "loss": 1.0301,
      "step": 9200
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.72265625,
      "learning_rate": 1.022004115663855e-06,
      "loss": 1.1526,
      "step": 9205
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.765625,
      "learning_rate": 1.0054826574210584e-06,
      "loss": 1.1847,
      "step": 9210
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.828125,
      "learning_rate": 9.89094474078428e-07,
      "loss": 1.1413,
      "step": 9215
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.81640625,
      "learning_rate": 9.728396102150872e-07,
      "loss": 1.2559,
      "step": 9220
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.67578125,
      "learning_rate": 9.567181100475164e-07,
      "loss": 0.9214,
      "step": 9225
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.73046875,
      "learning_rate": 9.407300174294365e-07,
      "loss": 1.1769,
      "step": 9230
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.703125,
      "learning_rate": 9.248753758516581e-07,
      "loss": 0.8906,
      "step": 9235
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.94140625,
      "learning_rate": 9.09154228441983e-07,
      "loss": 1.096,
      "step": 9240
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.7109375,
      "learning_rate": 8.935666179650804e-07,
      "loss": 0.9857,
      "step": 9245
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.79296875,
      "learning_rate": 8.781125868223716e-07,
      "loss": 1.0866,
      "step": 9250
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.82421875,
      "learning_rate": 8.627921770519187e-07,
      "loss": 1.1719,
      "step": 9255
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.6328125,
      "learning_rate": 8.476054303282966e-07,
      "loss": 0.99,
      "step": 9260
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.77734375,
      "learning_rate": 8.325523879625096e-07,
      "loss": 1.1741,
      "step": 9265
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.82421875,
      "learning_rate": 8.176330909018426e-07,
      "loss": 1.1881,
      "step": 9270
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.765625,
      "learning_rate": 8.028475797297652e-07,
      "loss": 0.8986,
      "step": 9275
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.57421875,
      "learning_rate": 7.881958946658496e-07,
      "loss": 0.8785,
      "step": 9280
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.76171875,
      "learning_rate": 7.736780755655871e-07,
      "loss": 1.0865,
      "step": 9285
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.859375,
      "learning_rate": 7.592941619203708e-07,
      "loss": 1.0307,
      "step": 9290
    },
    {
      "epoch": 0.95,
      "grad_norm": 1.125,
      "learning_rate": 7.450441928573193e-07,
      "loss": 1.0542,
      "step": 9295
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.84375,
      "learning_rate": 7.309282071392087e-07,
      "loss": 1.0614,
      "step": 9300
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.875,
      "learning_rate": 7.169462431643459e-07,
      "loss": 0.9932,
      "step": 9305
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.69921875,
      "learning_rate": 7.030983389664681e-07,
      "loss": 0.9256,
      "step": 9310
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.66015625,
      "learning_rate": 6.893845322146542e-07,
      "loss": 1.0094,
      "step": 9315
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.82421875,
      "learning_rate": 6.758048602131972e-07,
      "loss": 1.1286,
      "step": 9320
    },
    {
      "epoch": 0.95,
      "grad_norm": 1.265625,
      "learning_rate": 6.623593599015432e-07,
      "loss": 1.1295,
      "step": 9325
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.71484375,
      "learning_rate": 6.490480678541244e-07,
      "loss": 1.1013,
      "step": 9330
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.6953125,
      "learning_rate": 6.358710202803264e-07,
      "loss": 0.9872,
      "step": 9335
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.71875,
      "learning_rate": 6.228282530243656e-07,
      "loss": 1.083,
      "step": 9340
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.703125,
      "learning_rate": 6.099198015651786e-07,
      "loss": 0.9032,
      "step": 9345
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.6796875,
      "learning_rate": 5.971457010163329e-07,
      "loss": 1.0754,
      "step": 9350
    },
    {
      "epoch": 0.95,
      "grad_norm": 1.6796875,
      "learning_rate": 5.845059861259494e-07,
      "loss": 1.1483,
      "step": 9355
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.64453125,
      "learning_rate": 5.720006912765752e-07,
      "loss": 1.0902,
      "step": 9360
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.88671875,
      "learning_rate": 5.596298504851216e-07,
      "loss": 1.118,
      "step": 9365
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.9453125,
      "learning_rate": 5.473934974027595e-07,
      "loss": 1.0928,
      "step": 9370
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.88671875,
      "learning_rate": 5.352916653148077e-07,
      "loss": 1.0034,
      "step": 9375
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.7734375,
      "learning_rate": 5.233243871406779e-07,
      "loss": 1.2228,
      "step": 9380
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.6484375,
      "learning_rate": 5.11491695433769e-07,
      "loss": 0.8809,
      "step": 9385
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.765625,
      "learning_rate": 4.997936223813671e-07,
      "loss": 1.0031,
      "step": 9390
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.59765625,
      "learning_rate": 4.88230199804579e-07,
      "loss": 0.9258,
      "step": 9395
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.6796875,
      "learning_rate": 4.768014591582215e-07,
      "loss": 1.0305,
      "step": 9400
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.734375,
      "learning_rate": 4.65507431530765e-07,
      "loss": 0.9031,
      "step": 9405
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.82421875,
      "learning_rate": 4.543481476442235e-07,
      "loss": 1.2693,
      "step": 9410
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.76953125,
      "learning_rate": 4.4332363785408173e-07,
      "loss": 1.1679,
      "step": 9415
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.78125,
      "learning_rate": 4.324339321492232e-07,
      "loss": 1.0913,
      "step": 9420
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.73046875,
      "learning_rate": 4.21679060151825e-07,
      "loss": 0.9403,
      "step": 9425
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.80859375,
      "learning_rate": 4.1105905111728513e-07,
      "loss": 1.0614,
      "step": 9430
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.69921875,
      "learning_rate": 4.0057393393416185e-07,
      "loss": 1.1454,
      "step": 9435
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.68359375,
      "learning_rate": 3.902237371240791e-07,
      "loss": 1.0216,
      "step": 9440
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.78125,
      "learning_rate": 3.800084888416322e-07,
      "loss": 1.3414,
      "step": 9445
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.76171875,
      "learning_rate": 3.6992821687435454e-07,
      "loss": 1.0328,
      "step": 9450
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.75,
      "learning_rate": 3.5998294864260093e-07,
      "loss": 1.1775,
      "step": 9455
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.83984375,
      "learning_rate": 3.5017271119949236e-07,
      "loss": 1.1039,
      "step": 9460
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.69140625,
      "learning_rate": 3.404975312308378e-07,
      "loss": 0.9896,
      "step": 9465
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.83984375,
      "learning_rate": 3.309574350550626e-07,
      "loss": 0.9686,
      "step": 9470
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.64453125,
      "learning_rate": 3.2155244862314695e-07,
      "loss": 1.0276,
      "step": 9475
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.75390625,
      "learning_rate": 3.1228259751853174e-07,
      "loss": 1.1181,
      "step": 9480
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.85546875,
      "learning_rate": 3.0314790695706865e-07,
      "loss": 1.0357,
      "step": 9485
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.640625,
      "learning_rate": 2.941484017869478e-07,
      "loss": 0.9547,
      "step": 9490
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.77734375,
      "learning_rate": 2.852841064886258e-07,
      "loss": 1.0971,
      "step": 9495
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.82421875,
      "learning_rate": 2.765550451747645e-07,
      "loss": 1.2219,
      "step": 9500
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.8046875,
      "learning_rate": 2.679612415901478e-07,
      "loss": 1.0231,
      "step": 9505
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.71484375,
      "learning_rate": 2.595027191116539e-07,
      "loss": 1.0979,
      "step": 9510
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.6328125,
      "learning_rate": 2.5117950074815545e-07,
      "loss": 1.0184,
      "step": 9515
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.78125,
      "learning_rate": 2.4299160914046934e-07,
      "loss": 0.884,
      "step": 9520
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.81640625,
      "learning_rate": 2.3493906656130161e-07,
      "loss": 1.2306,
      "step": 9525
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.72265625,
      "learning_rate": 2.2702189491518587e-07,
      "loss": 0.9829,
      "step": 9530
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.73046875,
      "learning_rate": 2.1924011573841163e-07,
      "loss": 0.9877,
      "step": 9535
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.74609375,
      "learning_rate": 2.1159375019897397e-07,
      "loss": 1.1163,
      "step": 9540
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.75390625,
      "learning_rate": 2.0408281909652382e-07,
      "loss": 1.0729,
      "step": 9545
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.6796875,
      "learning_rate": 1.9670734286229565e-07,
      "loss": 1.1664,
      "step": 9550
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.875,
      "learning_rate": 1.8946734155905755e-07,
      "loss": 1.1987,
      "step": 9555
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.77734375,
      "learning_rate": 1.8236283488107243e-07,
      "loss": 1.0129,
      "step": 9560
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.6796875,
      "learning_rate": 1.7539384215401468e-07,
      "loss": 0.9329,
      "step": 9565
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.77734375,
      "learning_rate": 1.6856038233494243e-07,
      "loss": 1.3408,
      "step": 9570
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.734375,
      "learning_rate": 1.6186247401223653e-07,
      "loss": 1.0876,
      "step": 9575
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.65234375,
      "learning_rate": 1.55300135405545e-07,
      "loss": 1.138,
      "step": 9580
    },
    {
      "epoch": 0.98,
      "grad_norm": 1.3515625,
      "learning_rate": 1.4887338436574417e-07,
      "loss": 1.0621,
      "step": 9585
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.76171875,
      "learning_rate": 1.425822383748887e-07,
      "loss": 0.9534,
      "step": 9590
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.7578125,
      "learning_rate": 1.3642671454615065e-07,
      "loss": 1.0583,
      "step": 9595
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.71875,
      "learning_rate": 1.3040682962379148e-07,
      "loss": 0.8461,
      "step": 9600
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.6953125,
      "learning_rate": 1.2452259998310123e-07,
      "loss": 0.9082,
      "step": 9605
    },
    {
      "epoch": 0.98,
      "grad_norm": 1.3125,
      "learning_rate": 1.1877404163035954e-07,
      "loss": 1.1428,
      "step": 9610
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.59765625,
      "learning_rate": 1.1316117020280792e-07,
      "loss": 0.8789,
      "step": 9615
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.89453125,
      "learning_rate": 1.0768400096856645e-07,
      "loss": 1.034,
      "step": 9620
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.8125,
      "learning_rate": 1.0234254882664496e-07,
      "loss": 1.0811,
      "step": 9625
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.796875,
      "learning_rate": 9.713682830685411e-08,
      "loss": 1.0992,
      "step": 9630
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.84375,
      "learning_rate": 9.206685356980549e-08,
      "loss": 1.1375,
      "step": 9635
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.74609375,
      "learning_rate": 8.713263840683939e-08,
      "loss": 1.0984,
      "step": 9640
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.73046875,
      "learning_rate": 8.233419624000816e-08,
      "loss": 1.1631,
      "step": 9645
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.70703125,
      "learning_rate": 7.767154012203736e-08,
      "loss": 1.0627,
      "step": 9650
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.74609375,
      "learning_rate": 7.314468273628695e-08,
      "loss": 0.9164,
      "step": 9655
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.79296875,
      "learning_rate": 6.875363639671229e-08,
      "loss": 1.016,
      "step": 9660
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.74609375,
      "learning_rate": 6.44984130478421e-08,
      "loss": 0.9653,
      "step": 9665
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.63671875,
      "learning_rate": 6.037902426473952e-08,
      "loss": 0.9356,
      "step": 9670
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.7109375,
      "learning_rate": 5.639548125295768e-08,
      "loss": 1.1079,
      "step": 9675
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.72265625,
      "learning_rate": 5.2547794848539775e-08,
      "loss": 0.9644,
      "step": 9680
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.76171875,
      "learning_rate": 4.883597551795793e-08,
      "loss": 1.0018,
      "step": 9685
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.74609375,
      "learning_rate": 4.5260033358107686e-08,
      "loss": 1.0907,
      "step": 9690
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.703125,
      "learning_rate": 4.181997809626914e-08,
      "loss": 1.1787,
      "step": 9695
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.78125,
      "learning_rate": 3.8515819090073627e-08,
      "loss": 0.9285,
      "step": 9700
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.77734375,
      "learning_rate": 3.534756532750927e-08,
      "loss": 0.981,
      "step": 9705
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.625,
      "learning_rate": 3.2315225426843286e-08,
      "loss": 1.1156,
      "step": 9710
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.78515625,
      "learning_rate": 2.941880763666083e-08,
      "loss": 1.1838,
      "step": 9715
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.8125,
      "learning_rate": 2.665831983579836e-08,
      "loss": 1.1042,
      "step": 9720
    },
    {
      "epoch": 0.99,
      "grad_norm": 1.0,
      "learning_rate": 2.4033769533327034e-08,
      "loss": 1.0008,
      "step": 9725
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.78515625,
      "learning_rate": 2.1545163868552654e-08,
      "loss": 1.1072,
      "step": 9730
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.62109375,
      "learning_rate": 1.919250961098795e-08,
      "loss": 0.8663,
      "step": 9735
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.8671875,
      "learning_rate": 1.6975813160313713e-08,
      "loss": 0.9007,
      "step": 9740
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.89453125,
      "learning_rate": 1.4895080546395435e-08,
      "loss": 0.823,
      "step": 9745
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.69921875,
      "learning_rate": 1.2950317429238911e-08,
      "loss": 0.9001,
      "step": 9750
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.80859375,
      "learning_rate": 1.1141529098990244e-08,
      "loss": 1.1884,
      "step": 9755
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.7265625,
      "learning_rate": 9.468720475913628e-09,
      "loss": 0.9437,
      "step": 9760
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.72265625,
      "learning_rate": 7.931896110391357e-09,
      "loss": 1.0478,
      "step": 9765
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.87890625,
      "learning_rate": 6.531060182884962e-09,
      "loss": 1.1108,
      "step": 9770
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.70703125,
      "learning_rate": 5.2662165039518704e-09,
      "loss": 0.9931,
      "step": 9775
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.6953125,
      "learning_rate": 4.137368514217643e-09,
      "loss": 1.0868,
      "step": 9780
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.79296875,
      "learning_rate": 3.144519284381531e-09,
      "loss": 1.1434,
      "step": 9785
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.76953125,
      "learning_rate": 2.287671515183165e-09,
      "loss": 1.0663,
      "step": 9790
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.63671875,
      "learning_rate": 1.5668275374303153e-09,
      "loss": 0.9415,
      "step": 9795
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.91015625,
      "learning_rate": 9.819893119544788e-10,
      "loss": 0.9083,
      "step": 9800
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.796875,
      "learning_rate": 5.33158429638636e-10,
      "loss": 1.12,
      "step": 9805
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.59375,
      "learning_rate": 2.203361113894964e-10,
      "loss": 0.9311,
      "step": 9810
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.671875,
      "learning_rate": 4.352320814859923e-11,
      "loss": 0.9898,
      "step": 9815
    },
    {
      "epoch": 1.0,
      "step": 9819,
      "total_flos": 1.2142702597546967e+18,
      "train_loss": 1.0705042276791343,
      "train_runtime": 37728.5727,
      "train_samples_per_second": 1.041,
      "train_steps_per_second": 0.26
    }
  ],
  "logging_steps": 5,
  "max_steps": 9819,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "total_flos": 1.2142702597546967e+18,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}