{
  "best_metric": 1.2914531230926514,
  "best_model_checkpoint": "runs/deepseek_lora_20240422-141601/checkpoint-30000",
  "epoch": 0.75,
  "eval_steps": 2500,
  "global_step": 30000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 2.851395606994629,
      "learning_rate": 4.0000000000000003e-07,
      "loss": 2.2277,
      "step": 10
    },
    {
      "epoch": 0.0,
      "grad_norm": 5.176904201507568,
      "learning_rate": 8.000000000000001e-07,
      "loss": 2.0736,
      "step": 20
    },
    {
      "epoch": 0.0,
      "grad_norm": 3.9631340503692627,
      "learning_rate": 1.2000000000000002e-06,
      "loss": 2.2476,
      "step": 30
    },
    {
      "epoch": 0.0,
      "grad_norm": 6.812822341918945,
      "learning_rate": 1.6000000000000001e-06,
      "loss": 2.3009,
      "step": 40
    },
    {
      "epoch": 0.0,
      "grad_norm": 9.3452730178833,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 2.1243,
      "step": 50
    },
    {
      "epoch": 0.0,
      "grad_norm": 4.052621841430664,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 2.1649,
      "step": 60
    },
    {
      "epoch": 0.0,
      "grad_norm": 3.975917339324951,
      "learning_rate": 2.8000000000000003e-06,
      "loss": 2.1861,
      "step": 70
    },
    {
      "epoch": 0.0,
      "grad_norm": 3.4422502517700195,
      "learning_rate": 3.2000000000000003e-06,
      "loss": 2.1845,
      "step": 80
    },
    {
      "epoch": 0.0,
      "grad_norm": 5.118084907531738,
      "learning_rate": 3.6000000000000003e-06,
      "loss": 2.1471,
      "step": 90
    },
    {
      "epoch": 0.0,
      "grad_norm": 4.582140922546387,
      "learning_rate": 4.000000000000001e-06,
      "loss": 1.9872,
      "step": 100
    },
    {
      "epoch": 0.0,
      "grad_norm": 5.58018159866333,
      "learning_rate": 4.4e-06,
      "loss": 1.8719,
      "step": 110
    },
    {
      "epoch": 0.0,
      "grad_norm": 3.4979188442230225,
      "learning_rate": 4.800000000000001e-06,
      "loss": 1.9481,
      "step": 120
    },
    {
      "epoch": 0.0,
      "grad_norm": 6.800895690917969,
      "learning_rate": 5.2e-06,
      "loss": 1.8681,
      "step": 130
    },
    {
      "epoch": 0.0,
      "grad_norm": 3.454385757446289,
      "learning_rate": 5.600000000000001e-06,
      "loss": 2.026,
      "step": 140
    },
    {
      "epoch": 0.0,
      "grad_norm": 7.374110698699951,
      "learning_rate": 6e-06,
      "loss": 1.8847,
      "step": 150
    },
    {
      "epoch": 0.0,
      "grad_norm": 3.2150352001190186,
      "learning_rate": 6.4000000000000006e-06,
      "loss": 1.8317,
      "step": 160
    },
    {
      "epoch": 0.0,
      "grad_norm": 4.306232452392578,
      "learning_rate": 6.800000000000001e-06,
      "loss": 1.6211,
      "step": 170
    },
    {
      "epoch": 0.0,
      "grad_norm": 8.484739303588867,
      "learning_rate": 7.2000000000000005e-06,
      "loss": 1.7555,
      "step": 180
    },
    {
      "epoch": 0.0,
      "grad_norm": 3.582071542739868,
      "learning_rate": 7.600000000000001e-06,
      "loss": 1.8707,
      "step": 190
    },
    {
      "epoch": 0.01,
      "grad_norm": 8.12852954864502,
      "learning_rate": 8.000000000000001e-06,
      "loss": 1.7256,
      "step": 200
    },
    {
      "epoch": 0.01,
      "grad_norm": 8.658639907836914,
      "learning_rate": 8.400000000000001e-06,
      "loss": 1.5517,
      "step": 210
    },
    {
      "epoch": 0.01,
      "grad_norm": 5.2935709953308105,
      "learning_rate": 8.8e-06,
      "loss": 1.6941,
      "step": 220
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.7728030681610107,
      "learning_rate": 9.200000000000002e-06,
      "loss": 1.9548,
      "step": 230
    },
    {
      "epoch": 0.01,
      "grad_norm": 4.062624931335449,
      "learning_rate": 9.600000000000001e-06,
      "loss": 1.5029,
      "step": 240
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.485808849334717,
      "learning_rate": 1e-05,
      "loss": 1.7361,
      "step": 250
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.051875591278076,
      "learning_rate": 1.04e-05,
      "loss": 1.8557,
      "step": 260
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.840186834335327,
      "learning_rate": 1.0800000000000002e-05,
      "loss": 1.6347,
      "step": 270
    },
    {
      "epoch": 0.01,
      "grad_norm": 4.363358974456787,
      "learning_rate": 1.1200000000000001e-05,
      "loss": 1.6483,
      "step": 280
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.1710777282714844,
      "learning_rate": 1.16e-05,
      "loss": 1.8262,
      "step": 290
    },
    {
      "epoch": 0.01,
      "grad_norm": 4.845742225646973,
      "learning_rate": 1.2e-05,
      "loss": 1.6229,
      "step": 300
    },
    {
      "epoch": 0.01,
      "grad_norm": 6.983277320861816,
      "learning_rate": 1.2400000000000002e-05,
      "loss": 1.7023,
      "step": 310
    },
    {
      "epoch": 0.01,
      "grad_norm": 19.834449768066406,
      "learning_rate": 1.2800000000000001e-05,
      "loss": 1.6652,
      "step": 320
    },
    {
      "epoch": 0.01,
      "grad_norm": 7.883560657501221,
      "learning_rate": 1.3200000000000002e-05,
      "loss": 1.8418,
      "step": 330
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.3526132106781006,
      "learning_rate": 1.3600000000000002e-05,
      "loss": 1.7664,
      "step": 340
    },
    {
      "epoch": 0.01,
      "grad_norm": 4.037511348724365,
      "learning_rate": 1.4e-05,
      "loss": 1.6846,
      "step": 350
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.0859925746917725,
      "learning_rate": 1.4400000000000001e-05,
      "loss": 1.6175,
      "step": 360
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.3323440551757812,
      "learning_rate": 1.48e-05,
      "loss": 1.61,
      "step": 370
    },
    {
      "epoch": 0.01,
      "grad_norm": 6.743717193603516,
      "learning_rate": 1.5200000000000002e-05,
      "loss": 1.7058,
      "step": 380
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.9727015495300293,
      "learning_rate": 1.5600000000000003e-05,
      "loss": 1.6392,
      "step": 390
    },
    {
      "epoch": 0.01,
      "grad_norm": 15.913630485534668,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 1.6514,
      "step": 400
    },
    {
      "epoch": 0.01,
      "grad_norm": 6.065343856811523,
      "learning_rate": 1.64e-05,
      "loss": 1.6155,
      "step": 410
    },
    {
      "epoch": 0.01,
      "grad_norm": 7.500320911407471,
      "learning_rate": 1.6800000000000002e-05,
      "loss": 1.5588,
      "step": 420
    },
    {
      "epoch": 0.01,
      "grad_norm": 24.69054412841797,
      "learning_rate": 1.72e-05,
      "loss": 1.7351,
      "step": 430
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.404951572418213,
      "learning_rate": 1.76e-05,
      "loss": 1.4672,
      "step": 440
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.41680645942688,
      "learning_rate": 1.8e-05,
      "loss": 1.5949,
      "step": 450
    },
    {
      "epoch": 0.01,
      "grad_norm": 5.663394927978516,
      "learning_rate": 1.8400000000000003e-05,
      "loss": 1.6688,
      "step": 460
    },
    {
      "epoch": 0.01,
      "grad_norm": 6.914289474487305,
      "learning_rate": 1.88e-05,
      "loss": 1.731,
      "step": 470
    },
    {
      "epoch": 0.01,
      "grad_norm": 5.951695919036865,
      "learning_rate": 1.9200000000000003e-05,
      "loss": 1.6382,
      "step": 480
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.513606309890747,
      "learning_rate": 1.9600000000000002e-05,
      "loss": 1.5352,
      "step": 490
    },
    {
      "epoch": 0.01,
      "grad_norm": 10.61838436126709,
      "learning_rate": 2e-05,
      "loss": 1.7062,
      "step": 500
    },
    {
      "epoch": 0.01,
      "grad_norm": 4.863495349884033,
      "learning_rate": 1.999322033898305e-05,
      "loss": 1.6271,
      "step": 510
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.60945725440979,
      "learning_rate": 1.9986440677966104e-05,
      "loss": 1.5564,
      "step": 520
    },
    {
      "epoch": 0.01,
      "grad_norm": 5.073423385620117,
      "learning_rate": 1.9979661016949154e-05,
      "loss": 1.61,
      "step": 530
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.496952772140503,
      "learning_rate": 1.9972881355932204e-05,
      "loss": 1.6975,
      "step": 540
    },
    {
      "epoch": 0.01,
      "grad_norm": 6.6176958084106445,
      "learning_rate": 1.9966101694915257e-05,
      "loss": 1.6037,
      "step": 550
    },
    {
      "epoch": 0.01,
      "grad_norm": 4.7368364334106445,
      "learning_rate": 1.9959322033898307e-05,
      "loss": 1.6465,
      "step": 560
    },
    {
      "epoch": 0.01,
      "grad_norm": 4.264565944671631,
      "learning_rate": 1.995254237288136e-05,
      "loss": 1.4169,
      "step": 570
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.9882798194885254,
      "learning_rate": 1.994576271186441e-05,
      "loss": 1.5441,
      "step": 580
    },
    {
      "epoch": 0.01,
      "grad_norm": 8.446619987487793,
      "learning_rate": 1.993898305084746e-05,
      "loss": 1.8324,
      "step": 590
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.198861598968506,
      "learning_rate": 1.9932203389830512e-05,
      "loss": 1.6048,
      "step": 600
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.164740800857544,
      "learning_rate": 1.992542372881356e-05,
      "loss": 1.6014,
      "step": 610
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.619685649871826,
      "learning_rate": 1.991864406779661e-05,
      "loss": 1.7197,
      "step": 620
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.9726085662841797,
      "learning_rate": 1.991186440677966e-05,
      "loss": 1.6246,
      "step": 630
    },
    {
      "epoch": 0.02,
      "grad_norm": 4.476212024688721,
      "learning_rate": 1.990508474576271e-05,
      "loss": 1.6854,
      "step": 640
    },
    {
      "epoch": 0.02,
      "grad_norm": 3.6580896377563477,
      "learning_rate": 1.9898305084745764e-05,
      "loss": 1.6336,
      "step": 650
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.021207571029663,
      "learning_rate": 1.9891525423728814e-05,
      "loss": 1.8665,
      "step": 660
    },
    {
      "epoch": 0.02,
      "grad_norm": 3.3528759479522705,
      "learning_rate": 1.9884745762711867e-05,
      "loss": 1.5498,
      "step": 670
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.3135390281677246,
      "learning_rate": 1.9877966101694917e-05,
      "loss": 1.5024,
      "step": 680
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.831437110900879,
      "learning_rate": 1.9871186440677966e-05,
      "loss": 1.6879,
      "step": 690
    },
    {
      "epoch": 0.02,
      "grad_norm": 4.128018379211426,
      "learning_rate": 1.986440677966102e-05,
      "loss": 1.4852,
      "step": 700
    },
    {
      "epoch": 0.02,
      "grad_norm": 3.344834327697754,
      "learning_rate": 1.985762711864407e-05,
      "loss": 1.637,
      "step": 710
    },
    {
      "epoch": 0.02,
      "grad_norm": 3.829740285873413,
      "learning_rate": 1.985084745762712e-05,
      "loss": 1.5666,
      "step": 720
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.7154682874679565,
      "learning_rate": 1.9844067796610172e-05,
      "loss": 1.6689,
      "step": 730
    },
    {
      "epoch": 0.02,
      "grad_norm": 3.3414573669433594,
      "learning_rate": 1.9837288135593222e-05,
      "loss": 1.5791,
      "step": 740
    },
    {
      "epoch": 0.02,
      "grad_norm": 3.9933433532714844,
      "learning_rate": 1.9830508474576275e-05,
      "loss": 1.692,
      "step": 750
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.381664991378784,
      "learning_rate": 1.9823728813559324e-05,
      "loss": 1.5871,
      "step": 760
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.823326349258423,
      "learning_rate": 1.9816949152542374e-05,
      "loss": 1.6241,
      "step": 770
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.7634427547454834,
      "learning_rate": 1.9810169491525427e-05,
      "loss": 1.5557,
      "step": 780
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.1946895122528076,
      "learning_rate": 1.9803389830508477e-05,
      "loss": 1.427,
      "step": 790
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.7877142429351807,
      "learning_rate": 1.9796610169491527e-05,
      "loss": 1.6299,
      "step": 800
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.4223766326904297,
      "learning_rate": 1.978983050847458e-05,
      "loss": 1.5892,
      "step": 810
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.841574192047119,
      "learning_rate": 1.9783050847457626e-05,
      "loss": 1.5237,
      "step": 820
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.028319835662842,
      "learning_rate": 1.977627118644068e-05,
      "loss": 1.6722,
      "step": 830
    },
    {
      "epoch": 0.02,
      "grad_norm": 9.368087768554688,
      "learning_rate": 1.976949152542373e-05,
      "loss": 1.6136,
      "step": 840
    },
    {
      "epoch": 0.02,
      "grad_norm": 4.162208557128906,
      "learning_rate": 1.9762711864406782e-05,
      "loss": 1.6591,
      "step": 850
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.6572774648666382,
      "learning_rate": 1.9755932203389832e-05,
      "loss": 1.7035,
      "step": 860
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.6563576459884644,
      "learning_rate": 1.974915254237288e-05,
      "loss": 1.6079,
      "step": 870
    },
    {
      "epoch": 0.02,
      "grad_norm": 4.1615142822265625,
      "learning_rate": 1.9742372881355935e-05,
      "loss": 1.5603,
      "step": 880
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.3207154273986816,
      "learning_rate": 1.9735593220338984e-05,
      "loss": 1.6503,
      "step": 890
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.771116256713867,
      "learning_rate": 1.9728813559322034e-05,
      "loss": 1.6087,
      "step": 900
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.7232967615127563,
      "learning_rate": 1.9722033898305087e-05,
      "loss": 1.6093,
      "step": 910
    },
    {
      "epoch": 0.02,
      "grad_norm": 3.521620035171509,
      "learning_rate": 1.9715254237288137e-05,
      "loss": 1.6089,
      "step": 920
    },
    {
      "epoch": 0.02,
      "grad_norm": 4.505629539489746,
      "learning_rate": 1.970847457627119e-05,
      "loss": 1.5096,
      "step": 930
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.2191290855407715,
      "learning_rate": 1.970169491525424e-05,
      "loss": 1.6299,
      "step": 940
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.3791440725326538,
      "learning_rate": 1.969491525423729e-05,
      "loss": 1.6391,
      "step": 950
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.8153390884399414,
      "learning_rate": 1.9688135593220342e-05,
      "loss": 1.5753,
      "step": 960
    },
    {
      "epoch": 0.02,
      "grad_norm": 4.593536853790283,
      "learning_rate": 1.9681355932203392e-05,
      "loss": 1.6299,
      "step": 970
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.1231645345687866,
      "learning_rate": 1.9674576271186442e-05,
      "loss": 1.5701,
      "step": 980
    },
    {
      "epoch": 0.02,
      "grad_norm": 3.257448434829712,
      "learning_rate": 1.9667796610169495e-05,
      "loss": 1.5542,
      "step": 990
    },
    {
      "epoch": 0.03,
      "grad_norm": 3.4285597801208496,
      "learning_rate": 1.9661016949152545e-05,
      "loss": 1.6853,
      "step": 1000
    },
    {
      "epoch": 0.03,
      "grad_norm": 4.221822738647461,
      "learning_rate": 1.9654237288135594e-05,
      "loss": 1.688,
      "step": 1010
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.7846754789352417,
      "learning_rate": 1.9647457627118644e-05,
      "loss": 1.6791,
      "step": 1020
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.9467215538024902,
      "learning_rate": 1.9640677966101697e-05,
      "loss": 1.6426,
      "step": 1030
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.0467166900634766,
      "learning_rate": 1.9633898305084747e-05,
      "loss": 1.6687,
      "step": 1040
    },
    {
      "epoch": 0.03,
      "grad_norm": 4.110260963439941,
      "learning_rate": 1.9627118644067796e-05,
      "loss": 1.5767,
      "step": 1050
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.858794927597046,
      "learning_rate": 1.962033898305085e-05,
      "loss": 1.7274,
      "step": 1060
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.4475393295288086,
      "learning_rate": 1.96135593220339e-05,
      "loss": 1.44,
      "step": 1070
    },
    {
      "epoch": 0.03,
      "grad_norm": 8.013075828552246,
      "learning_rate": 1.960677966101695e-05,
      "loss": 1.7872,
      "step": 1080
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.881619453430176,
      "learning_rate": 1.9600000000000002e-05,
      "loss": 1.616,
      "step": 1090
    },
    {
      "epoch": 0.03,
      "grad_norm": 3.5345535278320312,
      "learning_rate": 1.9593220338983052e-05,
      "loss": 1.7476,
      "step": 1100
    },
    {
      "epoch": 0.03,
      "grad_norm": 4.891156196594238,
      "learning_rate": 1.95864406779661e-05,
      "loss": 1.7048,
      "step": 1110
    },
    {
      "epoch": 0.03,
      "grad_norm": 3.030557870864868,
      "learning_rate": 1.9579661016949155e-05,
      "loss": 1.5614,
      "step": 1120
    },
    {
      "epoch": 0.03,
      "grad_norm": 3.2785491943359375,
      "learning_rate": 1.9572881355932204e-05,
      "loss": 1.5347,
      "step": 1130
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.5175490379333496,
      "learning_rate": 1.9566101694915257e-05,
      "loss": 1.5167,
      "step": 1140
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.8572545051574707,
      "learning_rate": 1.9559322033898307e-05,
      "loss": 1.6463,
      "step": 1150
    },
    {
      "epoch": 0.03,
      "grad_norm": 3.903278350830078,
      "learning_rate": 1.9552542372881357e-05,
      "loss": 1.5609,
      "step": 1160
    },
    {
      "epoch": 0.03,
      "grad_norm": 5.839925289154053,
      "learning_rate": 1.954576271186441e-05,
      "loss": 1.5802,
      "step": 1170
    },
    {
      "epoch": 0.03,
      "grad_norm": 7.085844039916992,
      "learning_rate": 1.953898305084746e-05,
      "loss": 1.4595,
      "step": 1180
    },
    {
      "epoch": 0.03,
      "grad_norm": 4.327020645141602,
      "learning_rate": 1.953220338983051e-05,
      "loss": 1.622,
      "step": 1190
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.1976447105407715,
      "learning_rate": 1.9525423728813562e-05,
      "loss": 1.5931,
      "step": 1200
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.9214314222335815,
      "learning_rate": 1.9518644067796612e-05,
      "loss": 1.5809,
      "step": 1210
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.8724608421325684,
      "learning_rate": 1.9511864406779665e-05,
      "loss": 1.5122,
      "step": 1220
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.5664455890655518,
      "learning_rate": 1.950508474576271e-05,
      "loss": 1.5742,
      "step": 1230
    },
    {
      "epoch": 0.03,
      "grad_norm": 4.146188259124756,
      "learning_rate": 1.9498305084745765e-05,
      "loss": 1.7236,
      "step": 1240
    },
    {
      "epoch": 0.03,
      "grad_norm": 4.019920349121094,
      "learning_rate": 1.9491525423728814e-05,
      "loss": 1.7064,
      "step": 1250
    },
    {
      "epoch": 0.03,
      "grad_norm": 7.946535587310791,
      "learning_rate": 1.9484745762711864e-05,
      "loss": 1.4223,
      "step": 1260
    },
    {
      "epoch": 0.03,
      "grad_norm": 3.4898784160614014,
      "learning_rate": 1.9477966101694917e-05,
      "loss": 1.5758,
      "step": 1270
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.5326132774353027,
      "learning_rate": 1.9471186440677967e-05,
      "loss": 1.4112,
      "step": 1280
    },
    {
      "epoch": 0.03,
      "grad_norm": 4.736443519592285,
      "learning_rate": 1.9464406779661017e-05,
      "loss": 1.7355,
      "step": 1290
    },
    {
      "epoch": 0.03,
      "grad_norm": 5.109076499938965,
      "learning_rate": 1.945762711864407e-05,
      "loss": 1.4924,
      "step": 1300
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.268367052078247,
      "learning_rate": 1.945084745762712e-05,
      "loss": 1.5755,
      "step": 1310
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.643416404724121,
      "learning_rate": 1.9444067796610172e-05,
      "loss": 1.5164,
      "step": 1320
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.7688724994659424,
      "learning_rate": 1.9437288135593222e-05,
      "loss": 1.4642,
      "step": 1330
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.5019948482513428,
      "learning_rate": 1.9430508474576272e-05,
      "loss": 1.7102,
      "step": 1340
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.474804401397705,
      "learning_rate": 1.9423728813559325e-05,
      "loss": 1.3884,
      "step": 1350
    },
    {
      "epoch": 0.03,
      "grad_norm": 4.1147780418396,
      "learning_rate": 1.9416949152542375e-05,
      "loss": 1.5273,
      "step": 1360
    },
    {
      "epoch": 0.03,
      "grad_norm": 3.650137424468994,
      "learning_rate": 1.9410169491525424e-05,
      "loss": 1.5234,
      "step": 1370
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.3730149269104004,
      "learning_rate": 1.9403389830508477e-05,
      "loss": 1.5617,
      "step": 1380
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.4931895732879639,
      "learning_rate": 1.9396610169491527e-05,
      "loss": 1.432,
      "step": 1390
    },
    {
      "epoch": 0.04,
      "grad_norm": 12.553142547607422,
      "learning_rate": 1.938983050847458e-05,
      "loss": 1.7025,
      "step": 1400
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.4376004934310913,
      "learning_rate": 1.938305084745763e-05,
      "loss": 1.5554,
      "step": 1410
    },
    {
      "epoch": 0.04,
      "grad_norm": 4.463530540466309,
      "learning_rate": 1.937627118644068e-05,
      "loss": 1.7339,
      "step": 1420
    },
    {
      "epoch": 0.04,
      "grad_norm": 4.199666500091553,
      "learning_rate": 1.9369491525423733e-05,
      "loss": 1.4764,
      "step": 1430
    },
    {
      "epoch": 0.04,
      "grad_norm": 5.547006130218506,
      "learning_rate": 1.936271186440678e-05,
      "loss": 1.5275,
      "step": 1440
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.4305906295776367,
      "learning_rate": 1.9355932203389832e-05,
      "loss": 1.4891,
      "step": 1450
    },
    {
      "epoch": 0.04,
      "grad_norm": 6.539031982421875,
      "learning_rate": 1.9349152542372882e-05,
      "loss": 1.5861,
      "step": 1460
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.8748772144317627,
      "learning_rate": 1.934237288135593e-05,
      "loss": 1.6962,
      "step": 1470
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.4520819187164307,
      "learning_rate": 1.9335593220338985e-05,
      "loss": 1.528,
      "step": 1480
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.8652116060256958,
      "learning_rate": 1.9328813559322034e-05,
      "loss": 1.4667,
      "step": 1490
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.033538341522217,
      "learning_rate": 1.9322033898305087e-05,
      "loss": 1.541,
      "step": 1500
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.597931146621704,
      "learning_rate": 1.9315254237288137e-05,
      "loss": 1.5174,
      "step": 1510
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.4136037826538086,
      "learning_rate": 1.9308474576271187e-05,
      "loss": 1.6064,
      "step": 1520
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.753058671951294,
      "learning_rate": 1.930169491525424e-05,
      "loss": 1.5042,
      "step": 1530
    },
    {
      "epoch": 0.04,
      "grad_norm": 4.677073001861572,
      "learning_rate": 1.929491525423729e-05,
      "loss": 1.5494,
      "step": 1540
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.8957511186599731,
      "learning_rate": 1.928813559322034e-05,
      "loss": 1.6484,
      "step": 1550
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.838806390762329,
      "learning_rate": 1.9281355932203392e-05,
      "loss": 1.497,
      "step": 1560
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.078474283218384,
      "learning_rate": 1.9274576271186442e-05,
      "loss": 1.5427,
      "step": 1570
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.899158239364624,
      "learning_rate": 1.9267796610169492e-05,
      "loss": 1.4632,
      "step": 1580
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.7859959602355957,
      "learning_rate": 1.9261016949152545e-05,
      "loss": 1.5454,
      "step": 1590
    },
    {
      "epoch": 0.04,
      "grad_norm": 6.8981781005859375,
      "learning_rate": 1.9254237288135595e-05,
      "loss": 1.6573,
      "step": 1600
    },
    {
      "epoch": 0.04,
      "grad_norm": 9.905054092407227,
      "learning_rate": 1.9247457627118648e-05,
      "loss": 1.537,
      "step": 1610
    },
    {
      "epoch": 0.04,
      "grad_norm": 5.012923717498779,
      "learning_rate": 1.9240677966101698e-05,
      "loss": 1.4633,
      "step": 1620
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.548527240753174,
      "learning_rate": 1.9233898305084747e-05,
      "loss": 1.5818,
      "step": 1630
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.3274333477020264,
      "learning_rate": 1.92271186440678e-05,
      "loss": 1.6789,
      "step": 1640
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.535564422607422,
      "learning_rate": 1.9220338983050847e-05,
      "loss": 1.6132,
      "step": 1650
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.165702223777771,
      "learning_rate": 1.92135593220339e-05,
      "loss": 1.4798,
      "step": 1660
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.9253801107406616,
      "learning_rate": 1.920677966101695e-05,
      "loss": 1.5539,
      "step": 1670
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.7573797702789307,
      "learning_rate": 1.9200000000000003e-05,
      "loss": 1.6876,
      "step": 1680
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.5003459453582764,
      "learning_rate": 1.9193220338983052e-05,
      "loss": 1.6733,
      "step": 1690
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.403451681137085,
      "learning_rate": 1.9186440677966102e-05,
      "loss": 1.5298,
      "step": 1700
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.6052262783050537,
      "learning_rate": 1.9179661016949155e-05,
      "loss": 1.602,
      "step": 1710
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.5912582874298096,
      "learning_rate": 1.9172881355932205e-05,
      "loss": 1.6402,
      "step": 1720
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.5052791833877563,
      "learning_rate": 1.9166101694915254e-05,
      "loss": 1.3943,
      "step": 1730
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.4026551246643066,
      "learning_rate": 1.9159322033898308e-05,
      "loss": 1.4108,
      "step": 1740
    },
    {
      "epoch": 0.04,
      "grad_norm": 4.763739109039307,
      "learning_rate": 1.9152542372881357e-05,
      "loss": 1.6907,
      "step": 1750
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.252786874771118,
      "learning_rate": 1.9145762711864407e-05,
      "loss": 1.7057,
      "step": 1760
    },
    {
      "epoch": 0.04,
      "grad_norm": 4.1158576011657715,
      "learning_rate": 1.913898305084746e-05,
      "loss": 1.6021,
      "step": 1770
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.2215704917907715,
      "learning_rate": 1.913220338983051e-05,
      "loss": 1.7631,
      "step": 1780
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.5683887004852295,
      "learning_rate": 1.9125423728813563e-05,
      "loss": 1.513,
      "step": 1790
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.8528956174850464,
      "learning_rate": 1.9118644067796613e-05,
      "loss": 1.5257,
      "step": 1800
    },
    {
      "epoch": 0.05,
      "grad_norm": 4.791789531707764,
      "learning_rate": 1.9111864406779662e-05,
      "loss": 1.6476,
      "step": 1810
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.6491265296936035,
      "learning_rate": 1.9105084745762715e-05,
      "loss": 1.4177,
      "step": 1820
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.918846607208252,
      "learning_rate": 1.9098305084745765e-05,
      "loss": 1.5514,
      "step": 1830
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.498677968978882,
      "learning_rate": 1.9091525423728815e-05,
      "loss": 1.5839,
      "step": 1840
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.9992542266845703,
      "learning_rate": 1.9084745762711868e-05,
      "loss": 1.5403,
      "step": 1850
    },
    {
      "epoch": 0.05,
      "grad_norm": 4.272899150848389,
      "learning_rate": 1.9077966101694914e-05,
      "loss": 1.5413,
      "step": 1860
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.8770604133605957,
      "learning_rate": 1.9071186440677967e-05,
      "loss": 1.5309,
      "step": 1870
    },
    {
      "epoch": 0.05,
      "grad_norm": 7.052611827850342,
      "learning_rate": 1.9064406779661017e-05,
      "loss": 1.4491,
      "step": 1880
    },
    {
      "epoch": 0.05,
      "grad_norm": 3.6247971057891846,
      "learning_rate": 1.905762711864407e-05,
      "loss": 1.5832,
      "step": 1890
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.8061251640319824,
      "learning_rate": 1.905084745762712e-05,
      "loss": 1.4693,
      "step": 1900
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.4889063835144043,
      "learning_rate": 1.904406779661017e-05,
      "loss": 1.506,
      "step": 1910
    },
    {
      "epoch": 0.05,
      "grad_norm": 4.2278218269348145,
      "learning_rate": 1.9037288135593223e-05,
      "loss": 1.5761,
      "step": 1920
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.5221449136734009,
      "learning_rate": 1.9030508474576272e-05,
      "loss": 1.5925,
      "step": 1930
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.404806613922119,
      "learning_rate": 1.9023728813559322e-05,
      "loss": 1.5151,
      "step": 1940
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.544687271118164,
      "learning_rate": 1.9016949152542375e-05,
      "loss": 1.4932,
      "step": 1950
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.3219454288482666,
      "learning_rate": 1.9010169491525425e-05,
      "loss": 1.394,
      "step": 1960
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.491605758666992,
      "learning_rate": 1.9003389830508478e-05,
      "loss": 1.3943,
      "step": 1970
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.306079149246216,
      "learning_rate": 1.8996610169491528e-05,
      "loss": 1.5792,
      "step": 1980
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.6108038425445557,
      "learning_rate": 1.8989830508474577e-05,
      "loss": 1.6123,
      "step": 1990
    },
    {
      "epoch": 0.05,
      "grad_norm": 6.570626735687256,
      "learning_rate": 1.898305084745763e-05,
      "loss": 1.5036,
      "step": 2000
    },
    {
      "epoch": 0.05,
      "grad_norm": 6.583322525024414,
      "learning_rate": 1.897627118644068e-05,
      "loss": 1.6835,
      "step": 2010
    },
    {
      "epoch": 0.05,
      "grad_norm": 6.5043745040893555,
      "learning_rate": 1.896949152542373e-05,
      "loss": 1.6837,
      "step": 2020
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.6753697395324707,
      "learning_rate": 1.8962711864406783e-05,
      "loss": 1.5402,
      "step": 2030
    },
    {
      "epoch": 0.05,
      "grad_norm": 3.444967746734619,
      "learning_rate": 1.8955932203389833e-05,
      "loss": 1.5361,
      "step": 2040
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.729161024093628,
      "learning_rate": 1.8949152542372882e-05,
      "loss": 1.5347,
      "step": 2050
    },
    {
      "epoch": 0.05,
      "grad_norm": 3.8078458309173584,
      "learning_rate": 1.8942372881355932e-05,
      "loss": 1.6621,
      "step": 2060
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.9905569553375244,
      "learning_rate": 1.8935593220338985e-05,
      "loss": 1.5948,
      "step": 2070
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.4287803173065186,
      "learning_rate": 1.8928813559322035e-05,
      "loss": 1.549,
      "step": 2080
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.8749382495880127,
      "learning_rate": 1.8922033898305085e-05,
      "loss": 1.5382,
      "step": 2090
    },
    {
      "epoch": 0.05,
      "grad_norm": 3.3191275596618652,
      "learning_rate": 1.8915254237288138e-05,
      "loss": 1.4513,
      "step": 2100
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.9423046112060547,
      "learning_rate": 1.8908474576271187e-05,
      "loss": 1.6089,
      "step": 2110
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.974191188812256,
      "learning_rate": 1.8901694915254237e-05,
      "loss": 1.4364,
      "step": 2120
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.6282615661621094,
      "learning_rate": 1.889491525423729e-05,
      "loss": 1.564,
      "step": 2130
    },
    {
      "epoch": 0.05,
      "grad_norm": 3.031837224960327,
      "learning_rate": 1.888813559322034e-05,
      "loss": 1.5629,
      "step": 2140
    },
    {
      "epoch": 0.05,
      "grad_norm": 4.550152778625488,
      "learning_rate": 1.8881355932203393e-05,
      "loss": 1.4063,
      "step": 2150
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.0126075744628906,
      "learning_rate": 1.8874576271186443e-05,
      "loss": 1.6668,
      "step": 2160
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.3012149333953857,
      "learning_rate": 1.8867796610169492e-05,
      "loss": 1.5176,
      "step": 2170
    },
    {
      "epoch": 0.05,
      "grad_norm": 5.637124538421631,
      "learning_rate": 1.8861016949152545e-05,
      "loss": 1.6126,
      "step": 2180
    },
    {
      "epoch": 0.05,
      "grad_norm": 4.235476493835449,
      "learning_rate": 1.8854237288135595e-05,
      "loss": 1.6699,
      "step": 2190
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.229443311691284,
      "learning_rate": 1.8847457627118645e-05,
      "loss": 1.6809,
      "step": 2200
    },
    {
      "epoch": 0.06,
      "grad_norm": 3.459526538848877,
      "learning_rate": 1.8840677966101698e-05,
      "loss": 1.5235,
      "step": 2210
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.7822821140289307,
      "learning_rate": 1.8833898305084748e-05,
      "loss": 1.5493,
      "step": 2220
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.9952603578567505,
      "learning_rate": 1.8827118644067797e-05,
      "loss": 1.641,
      "step": 2230
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.978151798248291,
      "learning_rate": 1.882033898305085e-05,
      "loss": 1.4723,
      "step": 2240
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.273080348968506,
      "learning_rate": 1.88135593220339e-05,
      "loss": 1.5592,
      "step": 2250
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.7028123140335083,
      "learning_rate": 1.8806779661016953e-05,
      "loss": 1.5641,
      "step": 2260
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.5020806789398193,
      "learning_rate": 1.88e-05,
      "loss": 1.5493,
      "step": 2270
    },
    {
      "epoch": 0.06,
      "grad_norm": 3.8906872272491455,
      "learning_rate": 1.8793220338983053e-05,
      "loss": 1.4362,
      "step": 2280
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.847545623779297,
      "learning_rate": 1.8786440677966102e-05,
      "loss": 1.4759,
      "step": 2290
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.5495567321777344,
      "learning_rate": 1.8779661016949152e-05,
      "loss": 1.5692,
      "step": 2300
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.9365819692611694,
      "learning_rate": 1.8772881355932205e-05,
      "loss": 1.4154,
      "step": 2310
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.2686619758605957,
      "learning_rate": 1.8766101694915255e-05,
      "loss": 1.5987,
      "step": 2320
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.376857280731201,
      "learning_rate": 1.8759322033898305e-05,
      "loss": 1.4986,
      "step": 2330
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.3899834156036377,
      "learning_rate": 1.8752542372881358e-05,
      "loss": 1.4986,
      "step": 2340
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.5679335594177246,
      "learning_rate": 1.8745762711864407e-05,
      "loss": 1.7439,
      "step": 2350
    },
    {
      "epoch": 0.06,
      "grad_norm": 3.375790596008301,
      "learning_rate": 1.873898305084746e-05,
      "loss": 1.4942,
      "step": 2360
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.9815618395805359,
      "learning_rate": 1.873220338983051e-05,
      "loss": 1.7029,
      "step": 2370
    },
    {
      "epoch": 0.06,
      "grad_norm": 3.3285017013549805,
      "learning_rate": 1.872542372881356e-05,
      "loss": 1.4137,
      "step": 2380
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.6229037046432495,
      "learning_rate": 1.8718644067796613e-05,
      "loss": 1.4346,
      "step": 2390
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.9402897357940674,
      "learning_rate": 1.8711864406779663e-05,
      "loss": 1.5237,
      "step": 2400
    },
    {
      "epoch": 0.06,
      "grad_norm": 4.087438106536865,
      "learning_rate": 1.8705084745762712e-05,
      "loss": 1.5599,
      "step": 2410
    },
    {
      "epoch": 0.06,
      "grad_norm": 4.989218235015869,
      "learning_rate": 1.8698305084745765e-05,
      "loss": 1.2149,
      "step": 2420
    },
    {
      "epoch": 0.06,
      "grad_norm": 5.38515043258667,
      "learning_rate": 1.8691525423728815e-05,
      "loss": 1.4826,
      "step": 2430
    },
    {
      "epoch": 0.06,
      "grad_norm": 4.267625331878662,
      "learning_rate": 1.8684745762711868e-05,
      "loss": 1.4021,
      "step": 2440
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.202053427696228,
      "learning_rate": 1.8677966101694918e-05,
      "loss": 1.5868,
      "step": 2450
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.121800184249878,
      "learning_rate": 1.8671186440677968e-05,
      "loss": 1.5964,
      "step": 2460
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.9189884662628174,
      "learning_rate": 1.866440677966102e-05,
      "loss": 1.6586,
      "step": 2470
    },
    {
      "epoch": 0.06,
      "grad_norm": 3.1505911350250244,
      "learning_rate": 1.8657627118644067e-05,
      "loss": 1.4105,
      "step": 2480
    },
    {
      "epoch": 0.06,
      "grad_norm": 3.378533363342285,
      "learning_rate": 1.865084745762712e-05,
      "loss": 1.4924,
      "step": 2490
    },
    {
      "epoch": 0.06,
      "grad_norm": 3.24145770072937,
      "learning_rate": 1.864406779661017e-05,
      "loss": 1.5821,
      "step": 2500
    },
    {
      "epoch": 0.06,
      "eval_loss": 1.5105478763580322,
      "eval_runtime": 122.1138,
      "eval_samples_per_second": 8.189,
      "eval_steps_per_second": 8.189,
      "step": 2500
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.311624765396118,
      "learning_rate": 1.863728813559322e-05,
      "loss": 1.5278,
      "step": 2510
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.0834193229675293,
      "learning_rate": 1.8630508474576273e-05,
      "loss": 1.5296,
      "step": 2520
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.0099519491195679,
      "learning_rate": 1.8623728813559322e-05,
      "loss": 1.4238,
      "step": 2530
    },
    {
      "epoch": 0.06,
      "grad_norm": 4.44098424911499,
      "learning_rate": 1.8616949152542376e-05,
      "loss": 1.6053,
      "step": 2540
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.1189584732055664,
      "learning_rate": 1.8610169491525425e-05,
      "loss": 1.6757,
      "step": 2550
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.7386863231658936,
      "learning_rate": 1.8603389830508475e-05,
      "loss": 1.632,
      "step": 2560
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.454979181289673,
      "learning_rate": 1.8596610169491528e-05,
      "loss": 1.6346,
      "step": 2570
    },
    {
      "epoch": 0.06,
      "grad_norm": 13.50500774383545,
      "learning_rate": 1.8589830508474578e-05,
      "loss": 1.5434,
      "step": 2580
    },
    {
      "epoch": 0.06,
      "grad_norm": 3.2502903938293457,
      "learning_rate": 1.8583050847457627e-05,
      "loss": 1.5045,
      "step": 2590
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.067229747772217,
      "learning_rate": 1.857627118644068e-05,
      "loss": 1.5083,
      "step": 2600
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.341421127319336,
      "learning_rate": 1.856949152542373e-05,
      "loss": 1.3129,
      "step": 2610
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.1210529804229736,
      "learning_rate": 1.856271186440678e-05,
      "loss": 1.5939,
      "step": 2620
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.8607630729675293,
      "learning_rate": 1.8555932203389833e-05,
      "loss": 1.5752,
      "step": 2630
    },
    {
      "epoch": 0.07,
      "grad_norm": 3.6864287853240967,
      "learning_rate": 1.8549152542372883e-05,
      "loss": 1.5115,
      "step": 2640
    },
    {
      "epoch": 0.07,
      "grad_norm": 3.188952922821045,
      "learning_rate": 1.8542372881355936e-05,
      "loss": 1.5831,
      "step": 2650
    },
    {
      "epoch": 0.07,
      "grad_norm": 3.022368907928467,
      "learning_rate": 1.8535593220338986e-05,
      "loss": 1.4768,
      "step": 2660
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.8109090328216553,
      "learning_rate": 1.8528813559322035e-05,
      "loss": 1.5777,
      "step": 2670
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.9495521783828735,
      "learning_rate": 1.852203389830509e-05,
      "loss": 1.5209,
      "step": 2680
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.639312505722046,
      "learning_rate": 1.8515254237288135e-05,
      "loss": 1.4686,
      "step": 2690
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.118788719177246,
      "learning_rate": 1.8508474576271188e-05,
      "loss": 1.593,
      "step": 2700
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.9422922134399414,
      "learning_rate": 1.8501694915254237e-05,
      "loss": 1.5504,
      "step": 2710
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.6859073638916016,
      "learning_rate": 1.849491525423729e-05,
      "loss": 1.503,
      "step": 2720
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.1287014484405518,
      "learning_rate": 1.848813559322034e-05,
      "loss": 1.4379,
      "step": 2730
    },
    {
      "epoch": 0.07,
      "grad_norm": 5.31379508972168,
      "learning_rate": 1.848135593220339e-05,
      "loss": 1.4124,
      "step": 2740
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.2716705799102783,
      "learning_rate": 1.8474576271186443e-05,
      "loss": 1.3028,
      "step": 2750
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.1232097148895264,
      "learning_rate": 1.8467796610169493e-05,
      "loss": 1.4397,
      "step": 2760
    },
    {
      "epoch": 0.07,
      "grad_norm": 8.15327262878418,
      "learning_rate": 1.8461016949152542e-05,
      "loss": 1.4602,
      "step": 2770
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.911167025566101,
      "learning_rate": 1.8454237288135596e-05,
      "loss": 1.4358,
      "step": 2780
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.435959815979004,
      "learning_rate": 1.8447457627118645e-05,
      "loss": 1.5363,
      "step": 2790
    },
    {
      "epoch": 0.07,
      "grad_norm": 3.965622901916504,
      "learning_rate": 1.8440677966101695e-05,
      "loss": 1.4572,
      "step": 2800
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.4959452152252197,
      "learning_rate": 1.8433898305084748e-05,
      "loss": 1.462,
      "step": 2810
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.4225144386291504,
      "learning_rate": 1.8427118644067798e-05,
      "loss": 1.6232,
      "step": 2820
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.0637855529785156,
      "learning_rate": 1.842033898305085e-05,
      "loss": 1.3907,
      "step": 2830
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.5297822952270508,
      "learning_rate": 1.84135593220339e-05,
      "loss": 1.4691,
      "step": 2840
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.9830141067504883,
      "learning_rate": 1.840677966101695e-05,
      "loss": 1.4954,
      "step": 2850
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.0993255376815796,
      "learning_rate": 1.8400000000000003e-05,
      "loss": 1.5532,
      "step": 2860
    },
    {
      "epoch": 0.07,
      "grad_norm": 3.029752016067505,
      "learning_rate": 1.8393220338983053e-05,
      "loss": 1.3707,
      "step": 2870
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.2476751804351807,
      "learning_rate": 1.8386440677966103e-05,
      "loss": 1.5881,
      "step": 2880
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.877641201019287,
      "learning_rate": 1.8379661016949153e-05,
      "loss": 1.7004,
      "step": 2890
    },
    {
      "epoch": 0.07,
      "grad_norm": 7.551489353179932,
      "learning_rate": 1.8372881355932202e-05,
      "loss": 1.4955,
      "step": 2900
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.323641777038574,
      "learning_rate": 1.8366101694915255e-05,
      "loss": 1.4646,
      "step": 2910
    },
    {
      "epoch": 0.07,
      "grad_norm": 7.538690567016602,
      "learning_rate": 1.8359322033898305e-05,
      "loss": 1.4349,
      "step": 2920
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.6787673234939575,
      "learning_rate": 1.8352542372881358e-05,
      "loss": 1.4466,
      "step": 2930
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.8904561400413513,
      "learning_rate": 1.8345762711864408e-05,
      "loss": 1.4905,
      "step": 2940
    },
    {
      "epoch": 0.07,
      "grad_norm": 3.4237475395202637,
      "learning_rate": 1.8338983050847458e-05,
      "loss": 1.7526,
      "step": 2950
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.0563924312591553,
      "learning_rate": 1.833220338983051e-05,
      "loss": 1.5989,
      "step": 2960
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.3719455003738403,
      "learning_rate": 1.832542372881356e-05,
      "loss": 1.3786,
      "step": 2970
    },
    {
      "epoch": 0.07,
      "grad_norm": 3.358279228210449,
      "learning_rate": 1.831864406779661e-05,
      "loss": 1.4405,
      "step": 2980
    },
    {
      "epoch": 0.07,
      "grad_norm": 3.859485149383545,
      "learning_rate": 1.8311864406779663e-05,
      "loss": 1.4132,
      "step": 2990
    },
    {
      "epoch": 0.07,
      "grad_norm": 3.4341726303100586,
      "learning_rate": 1.8305084745762713e-05,
      "loss": 1.5347,
      "step": 3000
    },
    {
      "epoch": 0.08,
      "grad_norm": 3.9264137744903564,
      "learning_rate": 1.8298305084745766e-05,
      "loss": 1.2906,
      "step": 3010
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.0573365688323975,
      "learning_rate": 1.8291525423728816e-05,
      "loss": 1.3471,
      "step": 3020
    },
    {
      "epoch": 0.08,
      "grad_norm": 3.907181978225708,
      "learning_rate": 1.8284745762711865e-05,
      "loss": 1.4731,
      "step": 3030
    },
    {
      "epoch": 0.08,
      "grad_norm": 3.8897969722747803,
      "learning_rate": 1.827796610169492e-05,
      "loss": 1.4531,
      "step": 3040
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.1601696014404297,
      "learning_rate": 1.8271186440677968e-05,
      "loss": 1.3284,
      "step": 3050
    },
    {
      "epoch": 0.08,
      "grad_norm": 4.193431854248047,
      "learning_rate": 1.8264406779661018e-05,
      "loss": 1.6109,
      "step": 3060
    },
    {
      "epoch": 0.08,
      "grad_norm": 4.300390720367432,
      "learning_rate": 1.825762711864407e-05,
      "loss": 1.5134,
      "step": 3070
    },
    {
      "epoch": 0.08,
      "grad_norm": 3.635544538497925,
      "learning_rate": 1.825084745762712e-05,
      "loss": 1.578,
      "step": 3080
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.216878890991211,
      "learning_rate": 1.824406779661017e-05,
      "loss": 1.4583,
      "step": 3090
    },
    {
      "epoch": 0.08,
      "grad_norm": 4.205069065093994,
      "learning_rate": 1.823728813559322e-05,
      "loss": 1.675,
      "step": 3100
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.846705913543701,
      "learning_rate": 1.8230508474576273e-05,
      "loss": 1.6475,
      "step": 3110
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.2543482780456543,
      "learning_rate": 1.8223728813559323e-05,
      "loss": 1.5776,
      "step": 3120
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.5159026384353638,
      "learning_rate": 1.8216949152542373e-05,
      "loss": 1.5459,
      "step": 3130
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.0293331146240234,
      "learning_rate": 1.8210169491525426e-05,
      "loss": 1.425,
      "step": 3140
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.2342400550842285,
      "learning_rate": 1.8203389830508475e-05,
      "loss": 1.5727,
      "step": 3150
    },
    {
      "epoch": 0.08,
      "grad_norm": 3.4542758464813232,
      "learning_rate": 1.8196610169491525e-05,
      "loss": 1.4062,
      "step": 3160
    },
    {
      "epoch": 0.08,
      "grad_norm": 3.661266565322876,
      "learning_rate": 1.8189830508474578e-05,
      "loss": 1.5048,
      "step": 3170
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.260097622871399,
      "learning_rate": 1.8183050847457628e-05,
      "loss": 1.5345,
      "step": 3180
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.0968689918518066,
      "learning_rate": 1.817627118644068e-05,
      "loss": 1.5731,
      "step": 3190
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.311238169670105,
      "learning_rate": 1.816949152542373e-05,
      "loss": 1.4738,
      "step": 3200
    },
    {
      "epoch": 0.08,
      "grad_norm": 5.0531511306762695,
      "learning_rate": 1.816271186440678e-05,
      "loss": 1.6322,
      "step": 3210
    },
    {
      "epoch": 0.08,
      "grad_norm": 9.276644706726074,
      "learning_rate": 1.8155932203389833e-05,
      "loss": 1.6929,
      "step": 3220
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.991140127182007,
      "learning_rate": 1.8149152542372883e-05,
      "loss": 1.5816,
      "step": 3230
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.175997018814087,
      "learning_rate": 1.8142372881355933e-05,
      "loss": 1.5072,
      "step": 3240
    },
    {
      "epoch": 0.08,
      "grad_norm": 3.3827784061431885,
      "learning_rate": 1.8135593220338986e-05,
      "loss": 1.5224,
      "step": 3250
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.6336770057678223,
      "learning_rate": 1.8128813559322036e-05,
      "loss": 1.39,
      "step": 3260
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.821932554244995,
      "learning_rate": 1.8122033898305085e-05,
      "loss": 1.3934,
      "step": 3270
    },
    {
      "epoch": 0.08,
      "grad_norm": 4.799144268035889,
      "learning_rate": 1.811525423728814e-05,
      "loss": 1.6669,
      "step": 3280
    },
    {
      "epoch": 0.08,
      "grad_norm": 4.200830936431885,
      "learning_rate": 1.8108474576271188e-05,
      "loss": 1.5016,
      "step": 3290
    },
    {
      "epoch": 0.08,
      "grad_norm": 3.38613224029541,
      "learning_rate": 1.810169491525424e-05,
      "loss": 1.3987,
      "step": 3300
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.1738901138305664,
      "learning_rate": 1.8094915254237288e-05,
      "loss": 1.6103,
      "step": 3310
    },
    {
      "epoch": 0.08,
      "grad_norm": 3.5386528968811035,
      "learning_rate": 1.808813559322034e-05,
      "loss": 1.5953,
      "step": 3320
    },
    {
      "epoch": 0.08,
      "grad_norm": 3.7797794342041016,
      "learning_rate": 1.808135593220339e-05,
      "loss": 1.488,
      "step": 3330
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.5344908237457275,
      "learning_rate": 1.807457627118644e-05,
      "loss": 1.4928,
      "step": 3340
    },
    {
      "epoch": 0.08,
      "grad_norm": 3.421039342880249,
      "learning_rate": 1.8067796610169493e-05,
      "loss": 1.6482,
      "step": 3350
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.393103837966919,
      "learning_rate": 1.8061016949152543e-05,
      "loss": 1.6305,
      "step": 3360
    },
    {
      "epoch": 0.08,
      "grad_norm": 3.145881175994873,
      "learning_rate": 1.8054237288135593e-05,
      "loss": 1.4474,
      "step": 3370
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.6975643634796143,
      "learning_rate": 1.8047457627118646e-05,
      "loss": 1.381,
      "step": 3380
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.3912140130996704,
      "learning_rate": 1.8040677966101695e-05,
      "loss": 1.4391,
      "step": 3390
    },
    {
      "epoch": 0.09,
      "grad_norm": 3.034607172012329,
      "learning_rate": 1.803389830508475e-05,
      "loss": 1.5088,
      "step": 3400
    },
    {
      "epoch": 0.09,
      "grad_norm": 4.336439609527588,
      "learning_rate": 1.8027118644067798e-05,
      "loss": 1.4072,
      "step": 3410
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.483102798461914,
      "learning_rate": 1.8020338983050848e-05,
      "loss": 1.6016,
      "step": 3420
    },
    {
      "epoch": 0.09,
      "grad_norm": 3.4058072566986084,
      "learning_rate": 1.80135593220339e-05,
      "loss": 1.4985,
      "step": 3430
    },
    {
      "epoch": 0.09,
      "grad_norm": 3.850084066390991,
      "learning_rate": 1.800677966101695e-05,
      "loss": 1.4373,
      "step": 3440
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.5974807739257812,
      "learning_rate": 1.8e-05,
      "loss": 1.4761,
      "step": 3450
    },
    {
      "epoch": 0.09,
      "grad_norm": 3.012868642807007,
      "learning_rate": 1.7993220338983054e-05,
      "loss": 1.4755,
      "step": 3460
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.389517307281494,
      "learning_rate": 1.7986440677966103e-05,
      "loss": 1.4965,
      "step": 3470
    },
    {
      "epoch": 0.09,
      "grad_norm": 4.1623125076293945,
      "learning_rate": 1.7979661016949156e-05,
      "loss": 1.5227,
      "step": 3480
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.7761058807373047,
      "learning_rate": 1.7972881355932206e-05,
      "loss": 1.435,
      "step": 3490
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.5522000789642334,
      "learning_rate": 1.7966101694915256e-05,
      "loss": 1.5159,
      "step": 3500
    },
    {
      "epoch": 0.09,
      "grad_norm": 5.255176067352295,
      "learning_rate": 1.795932203389831e-05,
      "loss": 1.5027,
      "step": 3510
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.806614875793457,
      "learning_rate": 1.7952542372881355e-05,
      "loss": 1.5081,
      "step": 3520
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.0580685138702393,
      "learning_rate": 1.7945762711864408e-05,
      "loss": 1.3191,
      "step": 3530
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.131335496902466,
      "learning_rate": 1.7938983050847458e-05,
      "loss": 1.2758,
      "step": 3540
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.671358346939087,
      "learning_rate": 1.7932203389830508e-05,
      "loss": 1.6377,
      "step": 3550
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.681123971939087,
      "learning_rate": 1.792542372881356e-05,
      "loss": 1.5285,
      "step": 3560
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.8902277946472168,
      "learning_rate": 1.791864406779661e-05,
      "loss": 1.4772,
      "step": 3570
    },
    {
      "epoch": 0.09,
      "grad_norm": 3.0269978046417236,
      "learning_rate": 1.7911864406779664e-05,
      "loss": 1.5333,
      "step": 3580
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.9900496006011963,
      "learning_rate": 1.7905084745762713e-05,
      "loss": 1.36,
      "step": 3590
    },
    {
      "epoch": 0.09,
      "grad_norm": 5.621304988861084,
      "learning_rate": 1.7898305084745763e-05,
      "loss": 1.588,
      "step": 3600
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.4060802459716797,
      "learning_rate": 1.7891525423728816e-05,
      "loss": 1.4178,
      "step": 3610
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.8500341176986694,
      "learning_rate": 1.7884745762711866e-05,
      "loss": 1.4456,
      "step": 3620
    },
    {
      "epoch": 0.09,
      "grad_norm": 5.456387996673584,
      "learning_rate": 1.7877966101694916e-05,
      "loss": 1.5612,
      "step": 3630
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.2822142839431763,
      "learning_rate": 1.787118644067797e-05,
      "loss": 1.5491,
      "step": 3640
    },
    {
      "epoch": 0.09,
      "grad_norm": 3.71099591255188,
      "learning_rate": 1.7864406779661018e-05,
      "loss": 1.5208,
      "step": 3650
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.7524642944335938,
      "learning_rate": 1.785762711864407e-05,
      "loss": 1.445,
      "step": 3660
    },
    {
      "epoch": 0.09,
      "grad_norm": 4.494740009307861,
      "learning_rate": 1.785084745762712e-05,
      "loss": 1.3007,
      "step": 3670
    },
    {
      "epoch": 0.09,
      "grad_norm": 3.742347002029419,
      "learning_rate": 1.784406779661017e-05,
      "loss": 1.3394,
      "step": 3680
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.215082883834839,
      "learning_rate": 1.7837288135593224e-05,
      "loss": 1.4787,
      "step": 3690
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.095433473587036,
      "learning_rate": 1.7830508474576274e-05,
      "loss": 1.5177,
      "step": 3700
    },
    {
      "epoch": 0.09,
      "grad_norm": 9.498424530029297,
      "learning_rate": 1.7823728813559323e-05,
      "loss": 1.81,
      "step": 3710
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.3474382162094116,
      "learning_rate": 1.7816949152542376e-05,
      "loss": 1.4691,
      "step": 3720
    },
    {
      "epoch": 0.09,
      "grad_norm": 4.452912330627441,
      "learning_rate": 1.7810169491525423e-05,
      "loss": 1.4073,
      "step": 3730
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.7433382272720337,
      "learning_rate": 1.7803389830508476e-05,
      "loss": 1.4741,
      "step": 3740
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.274219036102295,
      "learning_rate": 1.7796610169491526e-05,
      "loss": 1.5086,
      "step": 3750
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.883714437484741,
      "learning_rate": 1.778983050847458e-05,
      "loss": 1.4372,
      "step": 3760
    },
    {
      "epoch": 0.09,
      "grad_norm": 3.5310864448547363,
      "learning_rate": 1.778305084745763e-05,
      "loss": 1.4292,
      "step": 3770
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.4181811809539795,
      "learning_rate": 1.7776271186440678e-05,
      "loss": 1.4853,
      "step": 3780
    },
    {
      "epoch": 0.09,
      "grad_norm": 9.955339431762695,
      "learning_rate": 1.776949152542373e-05,
      "loss": 1.6139,
      "step": 3790
    },
    {
      "epoch": 0.1,
      "grad_norm": 3.45771861076355,
      "learning_rate": 1.776271186440678e-05,
      "loss": 1.5383,
      "step": 3800
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.7813302278518677,
      "learning_rate": 1.775593220338983e-05,
      "loss": 1.5597,
      "step": 3810
    },
    {
      "epoch": 0.1,
      "grad_norm": 6.3414387702941895,
      "learning_rate": 1.7749152542372884e-05,
      "loss": 1.6369,
      "step": 3820
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.190211772918701,
      "learning_rate": 1.7742372881355933e-05,
      "loss": 1.5379,
      "step": 3830
    },
    {
      "epoch": 0.1,
      "grad_norm": 7.890295505523682,
      "learning_rate": 1.7735593220338983e-05,
      "loss": 1.445,
      "step": 3840
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.8373866081237793,
      "learning_rate": 1.7728813559322036e-05,
      "loss": 1.4636,
      "step": 3850
    },
    {
      "epoch": 0.1,
      "grad_norm": 3.3169384002685547,
      "learning_rate": 1.7722033898305086e-05,
      "loss": 1.3491,
      "step": 3860
    },
    {
      "epoch": 0.1,
      "grad_norm": 3.449866771697998,
      "learning_rate": 1.771525423728814e-05,
      "loss": 1.2964,
      "step": 3870
    },
    {
      "epoch": 0.1,
      "grad_norm": 4.36476993560791,
      "learning_rate": 1.770847457627119e-05,
      "loss": 1.5852,
      "step": 3880
    },
    {
      "epoch": 0.1,
      "grad_norm": 4.784100532531738,
      "learning_rate": 1.770169491525424e-05,
      "loss": 1.4905,
      "step": 3890
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.7355469465255737,
      "learning_rate": 1.769491525423729e-05,
      "loss": 1.5827,
      "step": 3900
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.377225399017334,
      "learning_rate": 1.768813559322034e-05,
      "loss": 1.5132,
      "step": 3910
    },
    {
      "epoch": 0.1,
      "grad_norm": 3.315483808517456,
      "learning_rate": 1.768135593220339e-05,
      "loss": 1.5375,
      "step": 3920
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.3933663368225098,
      "learning_rate": 1.767457627118644e-05,
      "loss": 1.6134,
      "step": 3930
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.9260916709899902,
      "learning_rate": 1.7667796610169494e-05,
      "loss": 1.5169,
      "step": 3940
    },
    {
      "epoch": 0.1,
      "grad_norm": 3.126014471054077,
      "learning_rate": 1.7661016949152543e-05,
      "loss": 1.3977,
      "step": 3950
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.4948694705963135,
      "learning_rate": 1.7654237288135593e-05,
      "loss": 1.5331,
      "step": 3960
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.0737502574920654,
      "learning_rate": 1.7647457627118646e-05,
      "loss": 1.6188,
      "step": 3970
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.267319917678833,
      "learning_rate": 1.7640677966101696e-05,
      "loss": 1.4179,
      "step": 3980
    },
    {
      "epoch": 0.1,
      "grad_norm": 4.286787033081055,
      "learning_rate": 1.7633898305084746e-05,
      "loss": 1.498,
      "step": 3990
    },
    {
      "epoch": 0.1,
      "grad_norm": 4.745983123779297,
      "learning_rate": 1.76271186440678e-05,
      "loss": 1.4002,
      "step": 4000
    },
    {
      "epoch": 0.1,
      "grad_norm": 5.1735148429870605,
      "learning_rate": 1.762033898305085e-05,
      "loss": 1.6062,
      "step": 4010
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.88619327545166,
      "learning_rate": 1.7613559322033898e-05,
      "loss": 1.395,
      "step": 4020
    },
    {
      "epoch": 0.1,
      "grad_norm": 5.35923957824707,
      "learning_rate": 1.760677966101695e-05,
      "loss": 1.3457,
      "step": 4030
    },
    {
      "epoch": 0.1,
      "grad_norm": 5.583395481109619,
      "learning_rate": 1.76e-05,
      "loss": 1.6342,
      "step": 4040
    },
    {
      "epoch": 0.1,
      "grad_norm": 10.570828437805176,
      "learning_rate": 1.7593220338983054e-05,
      "loss": 1.5859,
      "step": 4050
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.0806658267974854,
      "learning_rate": 1.7586440677966104e-05,
      "loss": 1.4284,
      "step": 4060
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.9427568912506104,
      "learning_rate": 1.7579661016949153e-05,
      "loss": 1.6711,
      "step": 4070
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.499094009399414,
      "learning_rate": 1.7572881355932206e-05,
      "loss": 1.6458,
      "step": 4080
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.1996042728424072,
      "learning_rate": 1.7566101694915256e-05,
      "loss": 1.3328,
      "step": 4090
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.4075920581817627,
      "learning_rate": 1.7559322033898306e-05,
      "loss": 1.4951,
      "step": 4100
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.1332108974456787,
      "learning_rate": 1.755254237288136e-05,
      "loss": 1.2751,
      "step": 4110
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.0534520149230957,
      "learning_rate": 1.754576271186441e-05,
      "loss": 1.6714,
      "step": 4120
    },
    {
      "epoch": 0.1,
      "grad_norm": 3.056049108505249,
      "learning_rate": 1.753898305084746e-05,
      "loss": 1.5304,
      "step": 4130
    },
    {
      "epoch": 0.1,
      "grad_norm": 3.6233339309692383,
      "learning_rate": 1.7532203389830508e-05,
      "loss": 1.574,
      "step": 4140
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.0271669626235962,
      "learning_rate": 1.752542372881356e-05,
      "loss": 1.5791,
      "step": 4150
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.7830803394317627,
      "learning_rate": 1.751864406779661e-05,
      "loss": 1.6381,
      "step": 4160
    },
    {
      "epoch": 0.1,
      "grad_norm": 5.869181156158447,
      "learning_rate": 1.751186440677966e-05,
      "loss": 1.5053,
      "step": 4170
    },
    {
      "epoch": 0.1,
      "grad_norm": 4.090358734130859,
      "learning_rate": 1.7505084745762714e-05,
      "loss": 1.3603,
      "step": 4180
    },
    {
      "epoch": 0.1,
      "grad_norm": 3.8125693798065186,
      "learning_rate": 1.7498305084745763e-05,
      "loss": 1.424,
      "step": 4190
    },
    {
      "epoch": 0.1,
      "grad_norm": 4.374596118927002,
      "learning_rate": 1.7491525423728813e-05,
      "loss": 1.5618,
      "step": 4200
    },
    {
      "epoch": 0.11,
      "grad_norm": 2.011169195175171,
      "learning_rate": 1.7484745762711866e-05,
      "loss": 1.5105,
      "step": 4210
    },
    {
      "epoch": 0.11,
      "grad_norm": 4.436703681945801,
      "learning_rate": 1.7477966101694916e-05,
      "loss": 1.4749,
      "step": 4220
    },
    {
      "epoch": 0.11,
      "grad_norm": 4.226778984069824,
      "learning_rate": 1.747118644067797e-05,
      "loss": 1.3816,
      "step": 4230
    },
    {
      "epoch": 0.11,
      "grad_norm": 3.130694627761841,
      "learning_rate": 1.746440677966102e-05,
      "loss": 1.341,
      "step": 4240
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.9768272638320923,
      "learning_rate": 1.745762711864407e-05,
      "loss": 1.6542,
      "step": 4250
    },
    {
      "epoch": 0.11,
      "grad_norm": 5.091753005981445,
      "learning_rate": 1.745084745762712e-05,
      "loss": 1.4983,
      "step": 4260
    },
    {
      "epoch": 0.11,
      "grad_norm": 2.771695613861084,
      "learning_rate": 1.744406779661017e-05,
      "loss": 1.5171,
      "step": 4270
    },
    {
      "epoch": 0.11,
      "grad_norm": 2.413747787475586,
      "learning_rate": 1.743728813559322e-05,
      "loss": 1.5318,
      "step": 4280
    },
    {
      "epoch": 0.11,
      "grad_norm": 4.8205766677856445,
      "learning_rate": 1.7430508474576274e-05,
      "loss": 1.5135,
      "step": 4290
    },
    {
      "epoch": 0.11,
      "grad_norm": 4.749906539916992,
      "learning_rate": 1.7423728813559324e-05,
      "loss": 1.4479,
      "step": 4300
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.5895742177963257,
      "learning_rate": 1.7416949152542373e-05,
      "loss": 1.5604,
      "step": 4310
    },
    {
      "epoch": 0.11,
      "grad_norm": 3.7786502838134766,
      "learning_rate": 1.7410169491525427e-05,
      "loss": 1.5736,
      "step": 4320
    },
    {
      "epoch": 0.11,
      "grad_norm": 2.570530891418457,
      "learning_rate": 1.7403389830508476e-05,
      "loss": 1.4447,
      "step": 4330
    },
    {
      "epoch": 0.11,
      "grad_norm": 2.9958982467651367,
      "learning_rate": 1.739661016949153e-05,
      "loss": 1.4472,
      "step": 4340
    },
    {
      "epoch": 0.11,
      "grad_norm": 5.140355110168457,
      "learning_rate": 1.7389830508474576e-05,
      "loss": 1.3618,
      "step": 4350
    },
    {
      "epoch": 0.11,
      "grad_norm": 2.1568965911865234,
      "learning_rate": 1.738305084745763e-05,
      "loss": 1.583,
      "step": 4360
    },
    {
      "epoch": 0.11,
      "grad_norm": 3.0182366371154785,
      "learning_rate": 1.737627118644068e-05,
      "loss": 1.6505,
      "step": 4370
    },
    {
      "epoch": 0.11,
      "grad_norm": 3.633270502090454,
      "learning_rate": 1.7369491525423728e-05,
      "loss": 1.4935,
      "step": 4380
    },
    {
      "epoch": 0.11,
      "grad_norm": 5.959003448486328,
      "learning_rate": 1.736271186440678e-05,
      "loss": 1.6776,
      "step": 4390
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.8102015256881714,
      "learning_rate": 1.735593220338983e-05,
      "loss": 1.5289,
      "step": 4400
    },
    {
      "epoch": 0.11,
      "grad_norm": 3.290053367614746,
      "learning_rate": 1.734915254237288e-05,
      "loss": 1.5028,
      "step": 4410
    },
    {
      "epoch": 0.11,
      "grad_norm": 5.087380886077881,
      "learning_rate": 1.7342372881355934e-05,
      "loss": 1.5375,
      "step": 4420
    },
    {
      "epoch": 0.11,
      "grad_norm": 7.137427806854248,
      "learning_rate": 1.7335593220338983e-05,
      "loss": 1.5013,
      "step": 4430
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.8010659217834473,
      "learning_rate": 1.7328813559322037e-05,
      "loss": 1.5357,
      "step": 4440
    },
    {
      "epoch": 0.11,
      "grad_norm": 2.408257246017456,
      "learning_rate": 1.7322033898305086e-05,
      "loss": 1.5419,
      "step": 4450
    },
    {
      "epoch": 0.11,
      "grad_norm": 2.6391518115997314,
      "learning_rate": 1.7315254237288136e-05,
      "loss": 1.5335,
      "step": 4460
    },
    {
      "epoch": 0.11,
      "grad_norm": 2.6913039684295654,
      "learning_rate": 1.730847457627119e-05,
      "loss": 1.421,
      "step": 4470
    },
    {
      "epoch": 0.11,
      "grad_norm": 2.694584608078003,
      "learning_rate": 1.730169491525424e-05,
      "loss": 1.4711,
      "step": 4480
    },
    {
      "epoch": 0.11,
      "grad_norm": 3.1425373554229736,
      "learning_rate": 1.729491525423729e-05,
      "loss": 1.5384,
      "step": 4490
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.6462829113006592,
      "learning_rate": 1.728813559322034e-05,
      "loss": 1.5503,
      "step": 4500
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.7344395518302917,
      "learning_rate": 1.728135593220339e-05,
      "loss": 1.597,
      "step": 4510
    },
    {
      "epoch": 0.11,
      "grad_norm": 4.82660436630249,
      "learning_rate": 1.7274576271186444e-05,
      "loss": 1.6279,
      "step": 4520
    },
    {
      "epoch": 0.11,
      "grad_norm": 3.49877667427063,
      "learning_rate": 1.7267796610169494e-05,
      "loss": 1.527,
      "step": 4530
    },
    {
      "epoch": 0.11,
      "grad_norm": 4.440892219543457,
      "learning_rate": 1.7261016949152544e-05,
      "loss": 1.5338,
      "step": 4540
    },
    {
      "epoch": 0.11,
      "grad_norm": 4.456193447113037,
      "learning_rate": 1.7254237288135597e-05,
      "loss": 1.5171,
      "step": 4550
    },
    {
      "epoch": 0.11,
      "grad_norm": 4.547098636627197,
      "learning_rate": 1.7247457627118643e-05,
      "loss": 1.691,
      "step": 4560
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.4237561225891113,
      "learning_rate": 1.7240677966101696e-05,
      "loss": 1.3744,
      "step": 4570
    },
    {
      "epoch": 0.11,
      "grad_norm": 3.008183717727661,
      "learning_rate": 1.7233898305084746e-05,
      "loss": 1.4857,
      "step": 4580
    },
    {
      "epoch": 0.11,
      "grad_norm": 5.0977582931518555,
      "learning_rate": 1.7227118644067796e-05,
      "loss": 1.6282,
      "step": 4590
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.9749881625175476,
      "learning_rate": 1.722033898305085e-05,
      "loss": 1.5292,
      "step": 4600
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.287003993988037,
      "learning_rate": 1.72135593220339e-05,
      "loss": 1.5087,
      "step": 4610
    },
    {
      "epoch": 0.12,
      "grad_norm": 2.0710980892181396,
      "learning_rate": 1.720677966101695e-05,
      "loss": 1.4257,
      "step": 4620
    },
    {
      "epoch": 0.12,
      "grad_norm": 3.2857139110565186,
      "learning_rate": 1.72e-05,
      "loss": 1.515,
      "step": 4630
    },
    {
      "epoch": 0.12,
      "grad_norm": 3.0228002071380615,
      "learning_rate": 1.719322033898305e-05,
      "loss": 1.2487,
      "step": 4640
    },
    {
      "epoch": 0.12,
      "grad_norm": 5.693451404571533,
      "learning_rate": 1.7186440677966104e-05,
      "loss": 1.5388,
      "step": 4650
    },
    {
      "epoch": 0.12,
      "grad_norm": 2.892563581466675,
      "learning_rate": 1.7179661016949154e-05,
      "loss": 1.4028,
      "step": 4660
    },
    {
      "epoch": 0.12,
      "grad_norm": 6.046909332275391,
      "learning_rate": 1.7172881355932204e-05,
      "loss": 1.4841,
      "step": 4670
    },
    {
      "epoch": 0.12,
      "grad_norm": 3.291569709777832,
      "learning_rate": 1.7166101694915257e-05,
      "loss": 1.5581,
      "step": 4680
    },
    {
      "epoch": 0.12,
      "grad_norm": 4.125076770782471,
      "learning_rate": 1.7159322033898306e-05,
      "loss": 1.4597,
      "step": 4690
    },
    {
      "epoch": 0.12,
      "grad_norm": 5.526379108428955,
      "learning_rate": 1.715254237288136e-05,
      "loss": 1.412,
      "step": 4700
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.2300543785095215,
      "learning_rate": 1.714576271186441e-05,
      "loss": 1.6105,
      "step": 4710
    },
    {
      "epoch": 0.12,
      "grad_norm": 3.772667646408081,
      "learning_rate": 1.713898305084746e-05,
      "loss": 1.598,
      "step": 4720
    },
    {
      "epoch": 0.12,
      "grad_norm": 5.655450820922852,
      "learning_rate": 1.7132203389830512e-05,
      "loss": 1.7285,
      "step": 4730
    },
    {
      "epoch": 0.12,
      "grad_norm": 4.088421821594238,
      "learning_rate": 1.712542372881356e-05,
      "loss": 1.364,
      "step": 4740
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.3764408826828003,
      "learning_rate": 1.711864406779661e-05,
      "loss": 1.543,
      "step": 4750
    },
    {
      "epoch": 0.12,
      "grad_norm": 3.439382314682007,
      "learning_rate": 1.711186440677966e-05,
      "loss": 1.639,
      "step": 4760
    },
    {
      "epoch": 0.12,
      "grad_norm": 3.9168601036071777,
      "learning_rate": 1.710508474576271e-05,
      "loss": 1.6387,
      "step": 4770
    },
    {
      "epoch": 0.12,
      "grad_norm": 3.363762617111206,
      "learning_rate": 1.7098305084745764e-05,
      "loss": 1.4295,
      "step": 4780
    },
    {
      "epoch": 0.12,
      "grad_norm": 6.131990909576416,
      "learning_rate": 1.7091525423728814e-05,
      "loss": 1.4635,
      "step": 4790
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.546452283859253,
      "learning_rate": 1.7084745762711867e-05,
      "loss": 1.435,
      "step": 4800
    },
    {
      "epoch": 0.12,
      "grad_norm": 2.090942144393921,
      "learning_rate": 1.7077966101694916e-05,
      "loss": 1.4256,
      "step": 4810
    },
    {
      "epoch": 0.12,
      "grad_norm": 3.1422488689422607,
      "learning_rate": 1.7071186440677966e-05,
      "loss": 1.467,
      "step": 4820
    },
    {
      "epoch": 0.12,
      "grad_norm": 2.450874090194702,
      "learning_rate": 1.706440677966102e-05,
      "loss": 1.5574,
      "step": 4830
    },
    {
      "epoch": 0.12,
      "grad_norm": 2.1604795455932617,
      "learning_rate": 1.705762711864407e-05,
      "loss": 1.5203,
      "step": 4840
    },
    {
      "epoch": 0.12,
      "grad_norm": 2.723970890045166,
      "learning_rate": 1.705084745762712e-05,
      "loss": 1.3351,
      "step": 4850
    },
    {
      "epoch": 0.12,
      "grad_norm": 4.656442642211914,
      "learning_rate": 1.704406779661017e-05,
      "loss": 1.519,
      "step": 4860
    },
    {
      "epoch": 0.12,
      "grad_norm": 2.0349249839782715,
      "learning_rate": 1.703728813559322e-05,
      "loss": 1.5239,
      "step": 4870
    },
    {
      "epoch": 0.12,
      "grad_norm": 4.223779678344727,
      "learning_rate": 1.703050847457627e-05,
      "loss": 1.5797,
      "step": 4880
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.9316459894180298,
      "learning_rate": 1.7023728813559324e-05,
      "loss": 1.514,
      "step": 4890
    },
    {
      "epoch": 0.12,
      "grad_norm": 2.5943896770477295,
      "learning_rate": 1.7016949152542374e-05,
      "loss": 1.6564,
      "step": 4900
    },
    {
      "epoch": 0.12,
      "grad_norm": 4.332151889801025,
      "learning_rate": 1.7010169491525427e-05,
      "loss": 1.458,
      "step": 4910
    },
    {
      "epoch": 0.12,
      "grad_norm": 4.935928821563721,
      "learning_rate": 1.7003389830508477e-05,
      "loss": 1.5565,
      "step": 4920
    },
    {
      "epoch": 0.12,
      "grad_norm": 3.392824411392212,
      "learning_rate": 1.6996610169491526e-05,
      "loss": 1.5043,
      "step": 4930
    },
    {
      "epoch": 0.12,
      "grad_norm": 2.030325412750244,
      "learning_rate": 1.698983050847458e-05,
      "loss": 1.4464,
      "step": 4940
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.1359357833862305,
      "learning_rate": 1.698305084745763e-05,
      "loss": 1.4193,
      "step": 4950
    },
    {
      "epoch": 0.12,
      "grad_norm": 3.806464910507202,
      "learning_rate": 1.697627118644068e-05,
      "loss": 1.5843,
      "step": 4960
    },
    {
      "epoch": 0.12,
      "grad_norm": 4.528138160705566,
      "learning_rate": 1.696949152542373e-05,
      "loss": 1.5183,
      "step": 4970
    },
    {
      "epoch": 0.12,
      "grad_norm": 2.763155698776245,
      "learning_rate": 1.6962711864406782e-05,
      "loss": 1.4913,
      "step": 4980
    },
    {
      "epoch": 0.12,
      "grad_norm": 6.776627063751221,
      "learning_rate": 1.695593220338983e-05,
      "loss": 1.5184,
      "step": 4990
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.0963356494903564,
      "learning_rate": 1.694915254237288e-05,
      "loss": 1.4424,
      "step": 5000
    },
    {
      "epoch": 0.12,
      "eval_loss": 1.4299169778823853,
      "eval_runtime": 122.3677,
      "eval_samples_per_second": 8.172,
      "eval_steps_per_second": 8.172,
      "step": 5000
    },
    {
      "epoch": 0.13,
      "grad_norm": 4.315555095672607,
      "learning_rate": 1.6942372881355934e-05,
      "loss": 1.6895,
      "step": 5010
    },
    {
      "epoch": 0.13,
      "grad_norm": 4.546899318695068,
      "learning_rate": 1.6935593220338984e-05,
      "loss": 1.4389,
      "step": 5020
    },
    {
      "epoch": 0.13,
      "grad_norm": 2.4407567977905273,
      "learning_rate": 1.6928813559322034e-05,
      "loss": 1.5297,
      "step": 5030
    },
    {
      "epoch": 0.13,
      "grad_norm": 4.0706706047058105,
      "learning_rate": 1.6922033898305087e-05,
      "loss": 1.5126,
      "step": 5040
    },
    {
      "epoch": 0.13,
      "grad_norm": 2.0037994384765625,
      "learning_rate": 1.6915254237288136e-05,
      "loss": 1.5133,
      "step": 5050
    },
    {
      "epoch": 0.13,
      "grad_norm": 2.827160358428955,
      "learning_rate": 1.6908474576271186e-05,
      "loss": 1.3906,
      "step": 5060
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.2849539518356323,
      "learning_rate": 1.690169491525424e-05,
      "loss": 1.4447,
      "step": 5070
    },
    {
      "epoch": 0.13,
      "grad_norm": 4.018255233764648,
      "learning_rate": 1.689491525423729e-05,
      "loss": 1.4943,
      "step": 5080
    },
    {
      "epoch": 0.13,
      "grad_norm": 2.2751851081848145,
      "learning_rate": 1.6888135593220342e-05,
      "loss": 1.5349,
      "step": 5090
    },
    {
      "epoch": 0.13,
      "grad_norm": 2.5244412422180176,
      "learning_rate": 1.6881355932203392e-05,
      "loss": 1.4901,
      "step": 5100
    },
    {
      "epoch": 0.13,
      "grad_norm": 3.3271806240081787,
      "learning_rate": 1.687457627118644e-05,
      "loss": 1.6362,
      "step": 5110
    },
    {
      "epoch": 0.13,
      "grad_norm": 2.7805681228637695,
      "learning_rate": 1.6867796610169495e-05,
      "loss": 1.4829,
      "step": 5120
    },
    {
      "epoch": 0.13,
      "grad_norm": 9.150943756103516,
      "learning_rate": 1.6861016949152544e-05,
      "loss": 1.6398,
      "step": 5130
    },
    {
      "epoch": 0.13,
      "grad_norm": 2.734816074371338,
      "learning_rate": 1.6854237288135594e-05,
      "loss": 1.4215,
      "step": 5140
    },
    {
      "epoch": 0.13,
      "grad_norm": 2.646103858947754,
      "learning_rate": 1.6847457627118647e-05,
      "loss": 1.4505,
      "step": 5150
    },
    {
      "epoch": 0.13,
      "grad_norm": 3.8651649951934814,
      "learning_rate": 1.6840677966101697e-05,
      "loss": 1.5377,
      "step": 5160
    },
    {
      "epoch": 0.13,
      "grad_norm": 2.219148874282837,
      "learning_rate": 1.683389830508475e-05,
      "loss": 1.4299,
      "step": 5170
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.7768529653549194,
      "learning_rate": 1.6827118644067796e-05,
      "loss": 1.5656,
      "step": 5180
    },
    {
      "epoch": 0.13,
      "grad_norm": 7.3894829750061035,
      "learning_rate": 1.682033898305085e-05,
      "loss": 1.4995,
      "step": 5190
    },
    {
      "epoch": 0.13,
      "grad_norm": 5.121682643890381,
      "learning_rate": 1.68135593220339e-05,
      "loss": 1.4559,
      "step": 5200
    },
    {
      "epoch": 0.13,
      "grad_norm": 5.169993877410889,
      "learning_rate": 1.680677966101695e-05,
      "loss": 1.4945,
      "step": 5210
    },
    {
      "epoch": 0.13,
      "grad_norm": 4.57939338684082,
      "learning_rate": 1.6800000000000002e-05,
      "loss": 1.3416,
      "step": 5220
    },
    {
      "epoch": 0.13,
      "grad_norm": 5.06122350692749,
      "learning_rate": 1.679322033898305e-05,
      "loss": 1.6018,
      "step": 5230
    },
    {
      "epoch": 0.13,
      "grad_norm": 3.2775559425354004,
      "learning_rate": 1.67864406779661e-05,
      "loss": 1.4793,
      "step": 5240
    },
    {
      "epoch": 0.13,
      "grad_norm": 4.888992786407471,
      "learning_rate": 1.6779661016949154e-05,
      "loss": 1.577,
      "step": 5250
    },
    {
      "epoch": 0.13,
      "grad_norm": 4.167829990386963,
      "learning_rate": 1.6772881355932204e-05,
      "loss": 1.5335,
      "step": 5260
    },
    {
      "epoch": 0.13,
      "grad_norm": 6.000946521759033,
      "learning_rate": 1.6766101694915257e-05,
      "loss": 1.382,
      "step": 5270
    },
    {
      "epoch": 0.13,
      "grad_norm": 4.0460524559021,
      "learning_rate": 1.6759322033898307e-05,
      "loss": 1.4355,
      "step": 5280
    },
    {
      "epoch": 0.13,
      "grad_norm": 3.21203351020813,
      "learning_rate": 1.6752542372881357e-05,
      "loss": 1.3427,
      "step": 5290
    },
    {
      "epoch": 0.13,
      "grad_norm": 6.206517219543457,
      "learning_rate": 1.674576271186441e-05,
      "loss": 1.3813,
      "step": 5300
    },
    {
      "epoch": 0.13,
      "grad_norm": 7.900985240936279,
      "learning_rate": 1.673898305084746e-05,
      "loss": 1.4964,
      "step": 5310
    },
    {
      "epoch": 0.13,
      "grad_norm": 2.3290903568267822,
      "learning_rate": 1.673220338983051e-05,
      "loss": 1.5764,
      "step": 5320
    },
    {
      "epoch": 0.13,
      "grad_norm": 2.4964051246643066,
      "learning_rate": 1.6725423728813562e-05,
      "loss": 1.4856,
      "step": 5330
    },
    {
      "epoch": 0.13,
      "grad_norm": 2.75761079788208,
      "learning_rate": 1.6718644067796612e-05,
      "loss": 1.5954,
      "step": 5340
    },
    {
      "epoch": 0.13,
      "grad_norm": 3.5791542530059814,
      "learning_rate": 1.671186440677966e-05,
      "loss": 1.5569,
      "step": 5350
    },
    {
      "epoch": 0.13,
      "grad_norm": 3.3503384590148926,
      "learning_rate": 1.6705084745762715e-05,
      "loss": 1.4643,
      "step": 5360
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.691370964050293,
      "learning_rate": 1.6698305084745764e-05,
      "loss": 1.451,
      "step": 5370
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.7851871252059937,
      "learning_rate": 1.6691525423728817e-05,
      "loss": 1.4943,
      "step": 5380
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.7007778882980347,
      "learning_rate": 1.6684745762711864e-05,
      "loss": 1.4363,
      "step": 5390
    },
    {
      "epoch": 0.14,
      "grad_norm": 3.409649133682251,
      "learning_rate": 1.6677966101694917e-05,
      "loss": 1.4374,
      "step": 5400
    },
    {
      "epoch": 0.14,
      "grad_norm": 3.221937894821167,
      "learning_rate": 1.6671186440677967e-05,
      "loss": 1.3033,
      "step": 5410
    },
    {
      "epoch": 0.14,
      "grad_norm": 2.0081026554107666,
      "learning_rate": 1.6664406779661016e-05,
      "loss": 1.6267,
      "step": 5420
    },
    {
      "epoch": 0.14,
      "grad_norm": 3.105792760848999,
      "learning_rate": 1.665762711864407e-05,
      "loss": 1.3978,
      "step": 5430
    },
    {
      "epoch": 0.14,
      "grad_norm": 2.11507248878479,
      "learning_rate": 1.665084745762712e-05,
      "loss": 1.4233,
      "step": 5440
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.7791744470596313,
      "learning_rate": 1.6644067796610172e-05,
      "loss": 1.4315,
      "step": 5450
    },
    {
      "epoch": 0.14,
      "grad_norm": 2.0917255878448486,
      "learning_rate": 1.6637288135593222e-05,
      "loss": 1.3977,
      "step": 5460
    },
    {
      "epoch": 0.14,
      "grad_norm": 4.012594699859619,
      "learning_rate": 1.663050847457627e-05,
      "loss": 1.5531,
      "step": 5470
    },
    {
      "epoch": 0.14,
      "grad_norm": 3.4779622554779053,
      "learning_rate": 1.6623728813559325e-05,
      "loss": 1.3889,
      "step": 5480
    },
    {
      "epoch": 0.14,
      "grad_norm": 2.1180970668792725,
      "learning_rate": 1.6616949152542374e-05,
      "loss": 1.4765,
      "step": 5490
    },
    {
      "epoch": 0.14,
      "grad_norm": 3.9273595809936523,
      "learning_rate": 1.6610169491525424e-05,
      "loss": 1.5436,
      "step": 5500
    },
    {
      "epoch": 0.14,
      "grad_norm": 2.0900495052337646,
      "learning_rate": 1.6603389830508477e-05,
      "loss": 1.2453,
      "step": 5510
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.6529499292373657,
      "learning_rate": 1.6596610169491527e-05,
      "loss": 1.6197,
      "step": 5520
    },
    {
      "epoch": 0.14,
      "grad_norm": 2.8194429874420166,
      "learning_rate": 1.6589830508474577e-05,
      "loss": 1.4349,
      "step": 5530
    },
    {
      "epoch": 0.14,
      "grad_norm": 2.102729320526123,
      "learning_rate": 1.658305084745763e-05,
      "loss": 1.3745,
      "step": 5540
    },
    {
      "epoch": 0.14,
      "grad_norm": 2.953857421875,
      "learning_rate": 1.657627118644068e-05,
      "loss": 1.4064,
      "step": 5550
    },
    {
      "epoch": 0.14,
      "grad_norm": 2.9009859561920166,
      "learning_rate": 1.6569491525423732e-05,
      "loss": 1.5503,
      "step": 5560
    },
    {
      "epoch": 0.14,
      "grad_norm": 3.959721565246582,
      "learning_rate": 1.6562711864406782e-05,
      "loss": 1.4913,
      "step": 5570
    },
    {
      "epoch": 0.14,
      "grad_norm": 2.8297066688537598,
      "learning_rate": 1.6555932203389832e-05,
      "loss": 1.3178,
      "step": 5580
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.7589664459228516,
      "learning_rate": 1.654915254237288e-05,
      "loss": 1.434,
      "step": 5590
    },
    {
      "epoch": 0.14,
      "grad_norm": 2.0058228969573975,
      "learning_rate": 1.654237288135593e-05,
      "loss": 1.3941,
      "step": 5600
    },
    {
      "epoch": 0.14,
      "grad_norm": 3.0510635375976562,
      "learning_rate": 1.6535593220338984e-05,
      "loss": 1.3438,
      "step": 5610
    },
    {
      "epoch": 0.14,
      "grad_norm": 2.7488441467285156,
      "learning_rate": 1.6528813559322034e-05,
      "loss": 1.4706,
      "step": 5620
    },
    {
      "epoch": 0.14,
      "grad_norm": 2.8862850666046143,
      "learning_rate": 1.6522033898305084e-05,
      "loss": 1.4544,
      "step": 5630
    },
    {
      "epoch": 0.14,
      "grad_norm": 4.727209568023682,
      "learning_rate": 1.6515254237288137e-05,
      "loss": 1.6279,
      "step": 5640
    },
    {
      "epoch": 0.14,
      "grad_norm": 2.8114616870880127,
      "learning_rate": 1.6508474576271187e-05,
      "loss": 1.4337,
      "step": 5650
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.8815864324569702,
      "learning_rate": 1.650169491525424e-05,
      "loss": 1.498,
      "step": 5660
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.9600774049758911,
      "learning_rate": 1.649491525423729e-05,
      "loss": 1.4432,
      "step": 5670
    },
    {
      "epoch": 0.14,
      "grad_norm": 9.273777961730957,
      "learning_rate": 1.648813559322034e-05,
      "loss": 1.5237,
      "step": 5680
    },
    {
      "epoch": 0.14,
      "grad_norm": 2.540921688079834,
      "learning_rate": 1.6481355932203392e-05,
      "loss": 1.5756,
      "step": 5690
    },
    {
      "epoch": 0.14,
      "grad_norm": 4.218064785003662,
      "learning_rate": 1.6474576271186442e-05,
      "loss": 1.5389,
      "step": 5700
    },
    {
      "epoch": 0.14,
      "grad_norm": 9.966960906982422,
      "learning_rate": 1.646779661016949e-05,
      "loss": 1.4562,
      "step": 5710
    },
    {
      "epoch": 0.14,
      "grad_norm": 5.023383140563965,
      "learning_rate": 1.6461016949152545e-05,
      "loss": 1.5459,
      "step": 5720
    },
    {
      "epoch": 0.14,
      "grad_norm": 2.2492129802703857,
      "learning_rate": 1.6454237288135594e-05,
      "loss": 1.3522,
      "step": 5730
    },
    {
      "epoch": 0.14,
      "grad_norm": 4.066563606262207,
      "learning_rate": 1.6447457627118648e-05,
      "loss": 1.5359,
      "step": 5740
    },
    {
      "epoch": 0.14,
      "grad_norm": 3.71022629737854,
      "learning_rate": 1.6440677966101697e-05,
      "loss": 1.5531,
      "step": 5750
    },
    {
      "epoch": 0.14,
      "grad_norm": 4.129106044769287,
      "learning_rate": 1.6433898305084747e-05,
      "loss": 1.4467,
      "step": 5760
    },
    {
      "epoch": 0.14,
      "grad_norm": 3.7006003856658936,
      "learning_rate": 1.64271186440678e-05,
      "loss": 1.6005,
      "step": 5770
    },
    {
      "epoch": 0.14,
      "grad_norm": 3.2597343921661377,
      "learning_rate": 1.642033898305085e-05,
      "loss": 1.3539,
      "step": 5780
    },
    {
      "epoch": 0.14,
      "grad_norm": 2.197850227355957,
      "learning_rate": 1.64135593220339e-05,
      "loss": 1.3549,
      "step": 5790
    },
    {
      "epoch": 0.14,
      "grad_norm": 2.6192123889923096,
      "learning_rate": 1.640677966101695e-05,
      "loss": 1.5776,
      "step": 5800
    },
    {
      "epoch": 0.15,
      "grad_norm": 2.787466049194336,
      "learning_rate": 1.64e-05,
      "loss": 1.5177,
      "step": 5810
    },
    {
      "epoch": 0.15,
      "grad_norm": 5.628580093383789,
      "learning_rate": 1.6393220338983052e-05,
      "loss": 1.6389,
      "step": 5820
    },
    {
      "epoch": 0.15,
      "grad_norm": 6.1815619468688965,
      "learning_rate": 1.63864406779661e-05,
      "loss": 1.467,
      "step": 5830
    },
    {
      "epoch": 0.15,
      "grad_norm": 2.9296717643737793,
      "learning_rate": 1.6379661016949155e-05,
      "loss": 1.5134,
      "step": 5840
    },
    {
      "epoch": 0.15,
      "grad_norm": 2.6761577129364014,
      "learning_rate": 1.6372881355932204e-05,
      "loss": 1.5576,
      "step": 5850
    },
    {
      "epoch": 0.15,
      "grad_norm": 3.5270586013793945,
      "learning_rate": 1.6366101694915254e-05,
      "loss": 1.4483,
      "step": 5860
    },
    {
      "epoch": 0.15,
      "grad_norm": 2.8878560066223145,
      "learning_rate": 1.6359322033898307e-05,
      "loss": 1.4017,
      "step": 5870
    },
    {
      "epoch": 0.15,
      "grad_norm": 3.653306245803833,
      "learning_rate": 1.6352542372881357e-05,
      "loss": 1.3913,
      "step": 5880
    },
    {
      "epoch": 0.15,
      "grad_norm": 3.2331554889678955,
      "learning_rate": 1.6345762711864407e-05,
      "loss": 1.6493,
      "step": 5890
    },
    {
      "epoch": 0.15,
      "grad_norm": 4.781271934509277,
      "learning_rate": 1.633898305084746e-05,
      "loss": 1.6687,
      "step": 5900
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.7907171249389648,
      "learning_rate": 1.633220338983051e-05,
      "loss": 1.4535,
      "step": 5910
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.6242311000823975,
      "learning_rate": 1.6325423728813563e-05,
      "loss": 1.6414,
      "step": 5920
    },
    {
      "epoch": 0.15,
      "grad_norm": 4.622259616851807,
      "learning_rate": 1.6318644067796612e-05,
      "loss": 1.4432,
      "step": 5930
    },
    {
      "epoch": 0.15,
      "grad_norm": 2.9257688522338867,
      "learning_rate": 1.6311864406779662e-05,
      "loss": 1.4025,
      "step": 5940
    },
    {
      "epoch": 0.15,
      "grad_norm": 3.823228359222412,
      "learning_rate": 1.6305084745762715e-05,
      "loss": 1.5054,
      "step": 5950
    },
    {
      "epoch": 0.15,
      "grad_norm": 2.4507181644439697,
      "learning_rate": 1.6298305084745765e-05,
      "loss": 1.3786,
      "step": 5960
    },
    {
      "epoch": 0.15,
      "grad_norm": 2.069478988647461,
      "learning_rate": 1.6291525423728814e-05,
      "loss": 1.319,
      "step": 5970
    },
    {
      "epoch": 0.15,
      "grad_norm": 4.48889684677124,
      "learning_rate": 1.6284745762711868e-05,
      "loss": 1.5197,
      "step": 5980
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.3289880752563477,
      "learning_rate": 1.6277966101694917e-05,
      "loss": 1.4523,
      "step": 5990
    },
    {
      "epoch": 0.15,
      "grad_norm": 3.0325958728790283,
      "learning_rate": 1.6271186440677967e-05,
      "loss": 1.3743,
      "step": 6000
    },
    {
      "epoch": 0.15,
      "grad_norm": 3.5299935340881348,
      "learning_rate": 1.6264406779661017e-05,
      "loss": 1.48,
      "step": 6010
    },
    {
      "epoch": 0.15,
      "grad_norm": 2.4003710746765137,
      "learning_rate": 1.625762711864407e-05,
      "loss": 1.5789,
      "step": 6020
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.6165212392807007,
      "learning_rate": 1.625084745762712e-05,
      "loss": 1.5405,
      "step": 6030
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.7408668994903564,
      "learning_rate": 1.624406779661017e-05,
      "loss": 1.3981,
      "step": 6040
    },
    {
      "epoch": 0.15,
      "grad_norm": 2.9926040172576904,
      "learning_rate": 1.6237288135593222e-05,
      "loss": 1.385,
      "step": 6050
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.5138695240020752,
      "learning_rate": 1.6230508474576272e-05,
      "loss": 1.4577,
      "step": 6060
    },
    {
      "epoch": 0.15,
      "grad_norm": 5.120387077331543,
      "learning_rate": 1.6223728813559322e-05,
      "loss": 1.3119,
      "step": 6070
    },
    {
      "epoch": 0.15,
      "grad_norm": 2.9037554264068604,
      "learning_rate": 1.6216949152542375e-05,
      "loss": 1.6238,
      "step": 6080
    },
    {
      "epoch": 0.15,
      "grad_norm": 2.26347279548645,
      "learning_rate": 1.6210169491525424e-05,
      "loss": 1.4437,
      "step": 6090
    },
    {
      "epoch": 0.15,
      "grad_norm": 5.754712104797363,
      "learning_rate": 1.6203389830508474e-05,
      "loss": 1.396,
      "step": 6100
    },
    {
      "epoch": 0.15,
      "grad_norm": 3.5630972385406494,
      "learning_rate": 1.6196610169491527e-05,
      "loss": 1.548,
      "step": 6110
    },
    {
      "epoch": 0.15,
      "grad_norm": 2.743373155593872,
      "learning_rate": 1.6189830508474577e-05,
      "loss": 1.5371,
      "step": 6120
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.8159428834915161,
      "learning_rate": 1.618305084745763e-05,
      "loss": 1.5093,
      "step": 6130
    },
    {
      "epoch": 0.15,
      "grad_norm": 3.4671523571014404,
      "learning_rate": 1.617627118644068e-05,
      "loss": 1.4732,
      "step": 6140
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.327411413192749,
      "learning_rate": 1.616949152542373e-05,
      "loss": 1.4216,
      "step": 6150
    },
    {
      "epoch": 0.15,
      "grad_norm": 3.7786147594451904,
      "learning_rate": 1.6162711864406783e-05,
      "loss": 1.4521,
      "step": 6160
    },
    {
      "epoch": 0.15,
      "grad_norm": 2.91334867477417,
      "learning_rate": 1.6155932203389832e-05,
      "loss": 1.4175,
      "step": 6170
    },
    {
      "epoch": 0.15,
      "grad_norm": 2.8965654373168945,
      "learning_rate": 1.6149152542372882e-05,
      "loss": 1.4037,
      "step": 6180
    },
    {
      "epoch": 0.15,
      "grad_norm": 2.145796775817871,
      "learning_rate": 1.6142372881355935e-05,
      "loss": 1.271,
      "step": 6190
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.8319767713546753,
      "learning_rate": 1.6135593220338985e-05,
      "loss": 1.5431,
      "step": 6200
    },
    {
      "epoch": 0.16,
      "grad_norm": 4.483182430267334,
      "learning_rate": 1.6128813559322038e-05,
      "loss": 1.5226,
      "step": 6210
    },
    {
      "epoch": 0.16,
      "grad_norm": 5.25124454498291,
      "learning_rate": 1.6122033898305084e-05,
      "loss": 1.5631,
      "step": 6220
    },
    {
      "epoch": 0.16,
      "grad_norm": 2.759519577026367,
      "learning_rate": 1.6115254237288137e-05,
      "loss": 1.4629,
      "step": 6230
    },
    {
      "epoch": 0.16,
      "grad_norm": 3.544701099395752,
      "learning_rate": 1.6108474576271187e-05,
      "loss": 1.563,
      "step": 6240
    },
    {
      "epoch": 0.16,
      "grad_norm": 2.6991095542907715,
      "learning_rate": 1.6101694915254237e-05,
      "loss": 1.346,
      "step": 6250
    },
    {
      "epoch": 0.16,
      "grad_norm": 3.9894187450408936,
      "learning_rate": 1.609491525423729e-05,
      "loss": 1.5077,
      "step": 6260
    },
    {
      "epoch": 0.16,
      "grad_norm": 5.335521221160889,
      "learning_rate": 1.608813559322034e-05,
      "loss": 1.4778,
      "step": 6270
    },
    {
      "epoch": 0.16,
      "grad_norm": 8.659845352172852,
      "learning_rate": 1.608135593220339e-05,
      "loss": 1.4799,
      "step": 6280
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.5903798341751099,
      "learning_rate": 1.6074576271186442e-05,
      "loss": 1.5992,
      "step": 6290
    },
    {
      "epoch": 0.16,
      "grad_norm": 3.451918840408325,
      "learning_rate": 1.6067796610169492e-05,
      "loss": 1.4438,
      "step": 6300
    },
    {
      "epoch": 0.16,
      "grad_norm": 4.562588214874268,
      "learning_rate": 1.6061016949152545e-05,
      "loss": 1.5701,
      "step": 6310
    },
    {
      "epoch": 0.16,
      "grad_norm": 5.422329902648926,
      "learning_rate": 1.6054237288135595e-05,
      "loss": 1.4549,
      "step": 6320
    },
    {
      "epoch": 0.16,
      "grad_norm": 2.3192145824432373,
      "learning_rate": 1.6047457627118645e-05,
      "loss": 1.5323,
      "step": 6330
    },
    {
      "epoch": 0.16,
      "grad_norm": 3.22367525100708,
      "learning_rate": 1.6040677966101698e-05,
      "loss": 1.7072,
      "step": 6340
    },
    {
      "epoch": 0.16,
      "grad_norm": 3.406710147857666,
      "learning_rate": 1.6033898305084747e-05,
      "loss": 1.542,
      "step": 6350
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.2928985357284546,
      "learning_rate": 1.6027118644067797e-05,
      "loss": 1.6213,
      "step": 6360
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.5963776111602783,
      "learning_rate": 1.602033898305085e-05,
      "loss": 1.4635,
      "step": 6370
    },
    {
      "epoch": 0.16,
      "grad_norm": 2.3411450386047363,
      "learning_rate": 1.60135593220339e-05,
      "loss": 1.5172,
      "step": 6380
    },
    {
      "epoch": 0.16,
      "grad_norm": 3.735161066055298,
      "learning_rate": 1.600677966101695e-05,
      "loss": 1.3673,
      "step": 6390
    },
    {
      "epoch": 0.16,
      "grad_norm": 2.5860648155212402,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 1.6392,
      "step": 6400
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.479199767112732,
      "learning_rate": 1.5993220338983052e-05,
      "loss": 1.3837,
      "step": 6410
    },
    {
      "epoch": 0.16,
      "grad_norm": 3.9291670322418213,
      "learning_rate": 1.5986440677966105e-05,
      "loss": 1.5389,
      "step": 6420
    },
    {
      "epoch": 0.16,
      "grad_norm": 3.7710626125335693,
      "learning_rate": 1.5979661016949152e-05,
      "loss": 1.4428,
      "step": 6430
    },
    {
      "epoch": 0.16,
      "grad_norm": 3.6627652645111084,
      "learning_rate": 1.5972881355932205e-05,
      "loss": 1.4286,
      "step": 6440
    },
    {
      "epoch": 0.16,
      "grad_norm": 2.351447582244873,
      "learning_rate": 1.5966101694915255e-05,
      "loss": 1.5782,
      "step": 6450
    },
    {
      "epoch": 0.16,
      "grad_norm": 3.5684151649475098,
      "learning_rate": 1.5959322033898304e-05,
      "loss": 1.3914,
      "step": 6460
    },
    {
      "epoch": 0.16,
      "grad_norm": 2.841437816619873,
      "learning_rate": 1.5952542372881357e-05,
      "loss": 1.4759,
      "step": 6470
    },
    {
      "epoch": 0.16,
      "grad_norm": 6.185434818267822,
      "learning_rate": 1.5945762711864407e-05,
      "loss": 1.5259,
      "step": 6480
    },
    {
      "epoch": 0.16,
      "grad_norm": 2.650010347366333,
      "learning_rate": 1.593898305084746e-05,
      "loss": 1.4768,
      "step": 6490
    },
    {
      "epoch": 0.16,
      "grad_norm": 3.858945846557617,
      "learning_rate": 1.593220338983051e-05,
      "loss": 1.3447,
      "step": 6500
    },
    {
      "epoch": 0.16,
      "grad_norm": 2.4795072078704834,
      "learning_rate": 1.592542372881356e-05,
      "loss": 1.5608,
      "step": 6510
    },
    {
      "epoch": 0.16,
      "grad_norm": 8.102066993713379,
      "learning_rate": 1.5918644067796613e-05,
      "loss": 1.6269,
      "step": 6520
    },
    {
      "epoch": 0.16,
      "grad_norm": 3.7153491973876953,
      "learning_rate": 1.5911864406779662e-05,
      "loss": 1.5927,
      "step": 6530
    },
    {
      "epoch": 0.16,
      "grad_norm": 3.4244496822357178,
      "learning_rate": 1.5905084745762712e-05,
      "loss": 1.2983,
      "step": 6540
    },
    {
      "epoch": 0.16,
      "grad_norm": 2.9444518089294434,
      "learning_rate": 1.5898305084745765e-05,
      "loss": 1.6322,
      "step": 6550
    },
    {
      "epoch": 0.16,
      "grad_norm": 6.136363983154297,
      "learning_rate": 1.5891525423728815e-05,
      "loss": 1.2941,
      "step": 6560
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.8231093883514404,
      "learning_rate": 1.5884745762711865e-05,
      "loss": 1.4843,
      "step": 6570
    },
    {
      "epoch": 0.16,
      "grad_norm": 9.1654052734375,
      "learning_rate": 1.5877966101694918e-05,
      "loss": 1.6293,
      "step": 6580
    },
    {
      "epoch": 0.16,
      "grad_norm": 2.462616205215454,
      "learning_rate": 1.5871186440677967e-05,
      "loss": 1.3849,
      "step": 6590
    },
    {
      "epoch": 0.17,
      "grad_norm": 4.495391368865967,
      "learning_rate": 1.586440677966102e-05,
      "loss": 1.6263,
      "step": 6600
    },
    {
      "epoch": 0.17,
      "grad_norm": 4.207485198974609,
      "learning_rate": 1.585762711864407e-05,
      "loss": 1.394,
      "step": 6610
    },
    {
      "epoch": 0.17,
      "grad_norm": 5.1345014572143555,
      "learning_rate": 1.585084745762712e-05,
      "loss": 1.6417,
      "step": 6620
    },
    {
      "epoch": 0.17,
      "grad_norm": 3.3824915885925293,
      "learning_rate": 1.584406779661017e-05,
      "loss": 1.4342,
      "step": 6630
    },
    {
      "epoch": 0.17,
      "grad_norm": 2.7321133613586426,
      "learning_rate": 1.583728813559322e-05,
      "loss": 1.5275,
      "step": 6640
    },
    {
      "epoch": 0.17,
      "grad_norm": 2.0641205310821533,
      "learning_rate": 1.5830508474576272e-05,
      "loss": 1.4921,
      "step": 6650
    },
    {
      "epoch": 0.17,
      "grad_norm": 2.6981923580169678,
      "learning_rate": 1.5823728813559322e-05,
      "loss": 1.4881,
      "step": 6660
    },
    {
      "epoch": 0.17,
      "grad_norm": 7.957277297973633,
      "learning_rate": 1.5816949152542372e-05,
      "loss": 1.6394,
      "step": 6670
    },
    {
      "epoch": 0.17,
      "grad_norm": 3.696110486984253,
      "learning_rate": 1.5810169491525425e-05,
      "loss": 1.5631,
      "step": 6680
    },
    {
      "epoch": 0.17,
      "grad_norm": 2.163233995437622,
      "learning_rate": 1.5803389830508475e-05,
      "loss": 1.6774,
      "step": 6690
    },
    {
      "epoch": 0.17,
      "grad_norm": 7.901493072509766,
      "learning_rate": 1.5796610169491528e-05,
      "loss": 1.5779,
      "step": 6700
    },
    {
      "epoch": 0.17,
      "grad_norm": 4.805121898651123,
      "learning_rate": 1.5789830508474577e-05,
      "loss": 1.4051,
      "step": 6710
    },
    {
      "epoch": 0.17,
      "grad_norm": 3.0310895442962646,
      "learning_rate": 1.5783050847457627e-05,
      "loss": 1.4071,
      "step": 6720
    },
    {
      "epoch": 0.17,
      "grad_norm": 5.031834602355957,
      "learning_rate": 1.577627118644068e-05,
      "loss": 1.4699,
      "step": 6730
    },
    {
      "epoch": 0.17,
      "grad_norm": 2.9163765907287598,
      "learning_rate": 1.576949152542373e-05,
      "loss": 1.5672,
      "step": 6740
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.5567353963851929,
      "learning_rate": 1.576271186440678e-05,
      "loss": 1.3926,
      "step": 6750
    },
    {
      "epoch": 0.17,
      "grad_norm": 5.556354522705078,
      "learning_rate": 1.5755932203389833e-05,
      "loss": 1.4875,
      "step": 6760
    },
    {
      "epoch": 0.17,
      "grad_norm": 2.687858819961548,
      "learning_rate": 1.5749152542372882e-05,
      "loss": 1.3951,
      "step": 6770
    },
    {
      "epoch": 0.17,
      "grad_norm": 2.278468608856201,
      "learning_rate": 1.5742372881355936e-05,
      "loss": 1.5664,
      "step": 6780
    },
    {
      "epoch": 0.17,
      "grad_norm": 4.735833168029785,
      "learning_rate": 1.5735593220338985e-05,
      "loss": 1.4692,
      "step": 6790
    },
    {
      "epoch": 0.17,
      "grad_norm": 2.8482844829559326,
      "learning_rate": 1.5728813559322035e-05,
      "loss": 1.5033,
      "step": 6800
    },
    {
      "epoch": 0.17,
      "grad_norm": 3.8273816108703613,
      "learning_rate": 1.5722033898305088e-05,
      "loss": 1.3153,
      "step": 6810
    },
    {
      "epoch": 0.17,
      "grad_norm": 5.1907958984375,
      "learning_rate": 1.5715254237288138e-05,
      "loss": 1.5143,
      "step": 6820
    },
    {
      "epoch": 0.17,
      "grad_norm": 2.969869375228882,
      "learning_rate": 1.5708474576271187e-05,
      "loss": 1.5629,
      "step": 6830
    },
    {
      "epoch": 0.17,
      "grad_norm": 2.6443393230438232,
      "learning_rate": 1.5701694915254237e-05,
      "loss": 1.5282,
      "step": 6840
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.8287936449050903,
      "learning_rate": 1.5694915254237287e-05,
      "loss": 1.5981,
      "step": 6850
    },
    {
      "epoch": 0.17,
      "grad_norm": 2.707810401916504,
      "learning_rate": 1.568813559322034e-05,
      "loss": 1.5634,
      "step": 6860
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.9375061988830566,
      "learning_rate": 1.568135593220339e-05,
      "loss": 1.4435,
      "step": 6870
    },
    {
      "epoch": 0.17,
      "grad_norm": 2.3451223373413086,
      "learning_rate": 1.5674576271186443e-05,
      "loss": 1.3045,
      "step": 6880
    },
    {
      "epoch": 0.17,
      "grad_norm": 3.8591105937957764,
      "learning_rate": 1.5667796610169492e-05,
      "loss": 1.4744,
      "step": 6890
    },
    {
      "epoch": 0.17,
      "grad_norm": 5.085460662841797,
      "learning_rate": 1.5661016949152542e-05,
      "loss": 1.4577,
      "step": 6900
    },
    {
      "epoch": 0.17,
      "grad_norm": 3.881556272506714,
      "learning_rate": 1.5654237288135595e-05,
      "loss": 1.5804,
      "step": 6910
    },
    {
      "epoch": 0.17,
      "grad_norm": 2.529904365539551,
      "learning_rate": 1.5647457627118645e-05,
      "loss": 1.3308,
      "step": 6920
    },
    {
      "epoch": 0.17,
      "grad_norm": 3.365994453430176,
      "learning_rate": 1.5640677966101695e-05,
      "loss": 1.6405,
      "step": 6930
    },
    {
      "epoch": 0.17,
      "grad_norm": 5.989173889160156,
      "learning_rate": 1.5633898305084748e-05,
      "loss": 1.2849,
      "step": 6940
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.963987946510315,
      "learning_rate": 1.5627118644067798e-05,
      "loss": 1.4939,
      "step": 6950
    },
    {
      "epoch": 0.17,
      "grad_norm": 2.617427110671997,
      "learning_rate": 1.562033898305085e-05,
      "loss": 1.3725,
      "step": 6960
    },
    {
      "epoch": 0.17,
      "grad_norm": 4.356164932250977,
      "learning_rate": 1.56135593220339e-05,
      "loss": 1.5418,
      "step": 6970
    },
    {
      "epoch": 0.17,
      "grad_norm": 3.704021692276001,
      "learning_rate": 1.560677966101695e-05,
      "loss": 1.4649,
      "step": 6980
    },
    {
      "epoch": 0.17,
      "grad_norm": 3.9556939601898193,
      "learning_rate": 1.5600000000000003e-05,
      "loss": 1.4474,
      "step": 6990
    },
    {
      "epoch": 0.17,
      "grad_norm": 3.5361435413360596,
      "learning_rate": 1.5593220338983053e-05,
      "loss": 1.4319,
      "step": 7000
    },
    {
      "epoch": 0.18,
      "grad_norm": 5.4693803787231445,
      "learning_rate": 1.5586440677966103e-05,
      "loss": 1.2996,
      "step": 7010
    },
    {
      "epoch": 0.18,
      "grad_norm": 4.020200729370117,
      "learning_rate": 1.5579661016949156e-05,
      "loss": 1.4498,
      "step": 7020
    },
    {
      "epoch": 0.18,
      "grad_norm": 2.733577013015747,
      "learning_rate": 1.5572881355932205e-05,
      "loss": 1.5082,
      "step": 7030
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.3573766946792603,
      "learning_rate": 1.5566101694915255e-05,
      "loss": 1.615,
      "step": 7040
    },
    {
      "epoch": 0.18,
      "grad_norm": 2.2789108753204346,
      "learning_rate": 1.5559322033898305e-05,
      "loss": 1.4904,
      "step": 7050
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.6954723596572876,
      "learning_rate": 1.5552542372881358e-05,
      "loss": 1.3797,
      "step": 7060
    },
    {
      "epoch": 0.18,
      "grad_norm": 5.36355447769165,
      "learning_rate": 1.5545762711864408e-05,
      "loss": 1.3624,
      "step": 7070
    },
    {
      "epoch": 0.18,
      "grad_norm": 2.569018602371216,
      "learning_rate": 1.5538983050847457e-05,
      "loss": 1.4751,
      "step": 7080
    },
    {
      "epoch": 0.18,
      "grad_norm": 6.11487340927124,
      "learning_rate": 1.553220338983051e-05,
      "loss": 1.3237,
      "step": 7090
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.2943029403686523,
      "learning_rate": 1.552542372881356e-05,
      "loss": 1.3676,
      "step": 7100
    },
    {
      "epoch": 0.18,
      "grad_norm": 3.214148998260498,
      "learning_rate": 1.551864406779661e-05,
      "loss": 1.4629,
      "step": 7110
    },
    {
      "epoch": 0.18,
      "grad_norm": 3.229365587234497,
      "learning_rate": 1.5511864406779663e-05,
      "loss": 1.4889,
      "step": 7120
    },
    {
      "epoch": 0.18,
      "grad_norm": 4.385135173797607,
      "learning_rate": 1.5505084745762713e-05,
      "loss": 1.4697,
      "step": 7130
    },
    {
      "epoch": 0.18,
      "grad_norm": 4.022393226623535,
      "learning_rate": 1.5498305084745762e-05,
      "loss": 1.5645,
      "step": 7140
    },
    {
      "epoch": 0.18,
      "grad_norm": 6.193070411682129,
      "learning_rate": 1.5491525423728815e-05,
      "loss": 1.462,
      "step": 7150
    },
    {
      "epoch": 0.18,
      "grad_norm": 5.172171592712402,
      "learning_rate": 1.5484745762711865e-05,
      "loss": 1.3287,
      "step": 7160
    },
    {
      "epoch": 0.18,
      "grad_norm": 3.0748867988586426,
      "learning_rate": 1.5477966101694918e-05,
      "loss": 1.3064,
      "step": 7170
    },
    {
      "epoch": 0.18,
      "grad_norm": 2.094850540161133,
      "learning_rate": 1.5471186440677968e-05,
      "loss": 1.3475,
      "step": 7180
    },
    {
      "epoch": 0.18,
      "grad_norm": 3.5158042907714844,
      "learning_rate": 1.5464406779661018e-05,
      "loss": 1.3971,
      "step": 7190
    },
    {
      "epoch": 0.18,
      "grad_norm": 2.3553078174591064,
      "learning_rate": 1.545762711864407e-05,
      "loss": 1.5563,
      "step": 7200
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.0270248651504517,
      "learning_rate": 1.545084745762712e-05,
      "loss": 1.5208,
      "step": 7210
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.056679368019104,
      "learning_rate": 1.544406779661017e-05,
      "loss": 1.4852,
      "step": 7220
    },
    {
      "epoch": 0.18,
      "grad_norm": 7.35222053527832,
      "learning_rate": 1.5437288135593223e-05,
      "loss": 1.5029,
      "step": 7230
    },
    {
      "epoch": 0.18,
      "grad_norm": 3.2837350368499756,
      "learning_rate": 1.5430508474576273e-05,
      "loss": 1.4196,
      "step": 7240
    },
    {
      "epoch": 0.18,
      "grad_norm": 5.04086446762085,
      "learning_rate": 1.5423728813559326e-05,
      "loss": 1.3399,
      "step": 7250
    },
    {
      "epoch": 0.18,
      "grad_norm": 5.913218021392822,
      "learning_rate": 1.5416949152542372e-05,
      "loss": 1.5924,
      "step": 7260
    },
    {
      "epoch": 0.18,
      "grad_norm": 2.7377588748931885,
      "learning_rate": 1.5410169491525425e-05,
      "loss": 1.5574,
      "step": 7270
    },
    {
      "epoch": 0.18,
      "grad_norm": 2.2820863723754883,
      "learning_rate": 1.5403389830508475e-05,
      "loss": 1.3394,
      "step": 7280
    },
    {
      "epoch": 0.18,
      "grad_norm": 2.7436327934265137,
      "learning_rate": 1.5396610169491525e-05,
      "loss": 1.3762,
      "step": 7290
    },
    {
      "epoch": 0.18,
      "grad_norm": 5.309854030609131,
      "learning_rate": 1.5389830508474578e-05,
      "loss": 1.5029,
      "step": 7300
    },
    {
      "epoch": 0.18,
      "grad_norm": 6.098525047302246,
      "learning_rate": 1.5383050847457628e-05,
      "loss": 1.4066,
      "step": 7310
    },
    {
      "epoch": 0.18,
      "grad_norm": 4.842358112335205,
      "learning_rate": 1.5376271186440677e-05,
      "loss": 1.5155,
      "step": 7320
    },
    {
      "epoch": 0.18,
      "grad_norm": 2.510955333709717,
      "learning_rate": 1.536949152542373e-05,
      "loss": 1.3376,
      "step": 7330
    },
    {
      "epoch": 0.18,
      "grad_norm": 3.17315673828125,
      "learning_rate": 1.536271186440678e-05,
      "loss": 1.2744,
      "step": 7340
    },
    {
      "epoch": 0.18,
      "grad_norm": 4.362697124481201,
      "learning_rate": 1.5355932203389833e-05,
      "loss": 1.4229,
      "step": 7350
    },
    {
      "epoch": 0.18,
      "grad_norm": 2.723053216934204,
      "learning_rate": 1.5349152542372883e-05,
      "loss": 1.4073,
      "step": 7360
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.8270694017410278,
      "learning_rate": 1.5342372881355933e-05,
      "loss": 1.4346,
      "step": 7370
    },
    {
      "epoch": 0.18,
      "grad_norm": 3.8290789127349854,
      "learning_rate": 1.5335593220338986e-05,
      "loss": 1.4864,
      "step": 7380
    },
    {
      "epoch": 0.18,
      "grad_norm": 2.812735080718994,
      "learning_rate": 1.5328813559322035e-05,
      "loss": 1.41,
      "step": 7390
    },
    {
      "epoch": 0.18,
      "grad_norm": 2.151588201522827,
      "learning_rate": 1.5322033898305085e-05,
      "loss": 1.4377,
      "step": 7400
    },
    {
      "epoch": 0.19,
      "grad_norm": 2.5662014484405518,
      "learning_rate": 1.5315254237288138e-05,
      "loss": 1.3758,
      "step": 7410
    },
    {
      "epoch": 0.19,
      "grad_norm": 3.8229243755340576,
      "learning_rate": 1.5308474576271188e-05,
      "loss": 1.5029,
      "step": 7420
    },
    {
      "epoch": 0.19,
      "grad_norm": 2.436012029647827,
      "learning_rate": 1.530169491525424e-05,
      "loss": 1.6117,
      "step": 7430
    },
    {
      "epoch": 0.19,
      "grad_norm": 5.254758358001709,
      "learning_rate": 1.529491525423729e-05,
      "loss": 1.6131,
      "step": 7440
    },
    {
      "epoch": 0.19,
      "grad_norm": 3.4107015132904053,
      "learning_rate": 1.528813559322034e-05,
      "loss": 1.4869,
      "step": 7450
    },
    {
      "epoch": 0.19,
      "grad_norm": 3.7725350856781006,
      "learning_rate": 1.528135593220339e-05,
      "loss": 1.4612,
      "step": 7460
    },
    {
      "epoch": 0.19,
      "grad_norm": 4.748162746429443,
      "learning_rate": 1.527457627118644e-05,
      "loss": 1.5007,
      "step": 7470
    },
    {
      "epoch": 0.19,
      "grad_norm": 4.55190372467041,
      "learning_rate": 1.5267796610169493e-05,
      "loss": 1.4602,
      "step": 7480
    },
    {
      "epoch": 0.19,
      "grad_norm": 3.2815232276916504,
      "learning_rate": 1.5261016949152543e-05,
      "loss": 1.5087,
      "step": 7490
    },
    {
      "epoch": 0.19,
      "grad_norm": 3.3302783966064453,
      "learning_rate": 1.5254237288135594e-05,
      "loss": 1.4179,
      "step": 7500
    },
    {
      "epoch": 0.19,
      "eval_loss": 1.429421305656433,
      "eval_runtime": 122.4826,
      "eval_samples_per_second": 8.164,
      "eval_steps_per_second": 8.164,
      "step": 7500
    },
    {
      "epoch": 0.19,
      "grad_norm": 4.227777004241943,
      "learning_rate": 1.5247457627118645e-05,
      "loss": 1.4771,
      "step": 7510
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.8159816265106201,
      "learning_rate": 1.5240677966101695e-05,
      "loss": 1.4061,
      "step": 7520
    },
    {
      "epoch": 0.19,
      "grad_norm": 2.051342248916626,
      "learning_rate": 1.5233898305084747e-05,
      "loss": 1.5657,
      "step": 7530
    },
    {
      "epoch": 0.19,
      "grad_norm": 2.7177517414093018,
      "learning_rate": 1.5227118644067798e-05,
      "loss": 1.405,
      "step": 7540
    },
    {
      "epoch": 0.19,
      "grad_norm": 2.609527587890625,
      "learning_rate": 1.522033898305085e-05,
      "loss": 1.2861,
      "step": 7550
    },
    {
      "epoch": 0.19,
      "grad_norm": 3.252938747406006,
      "learning_rate": 1.5213559322033899e-05,
      "loss": 1.5148,
      "step": 7560
    },
    {
      "epoch": 0.19,
      "grad_norm": 2.3884074687957764,
      "learning_rate": 1.520677966101695e-05,
      "loss": 1.4052,
      "step": 7570
    },
    {
      "epoch": 0.19,
      "grad_norm": 3.9711310863494873,
      "learning_rate": 1.5200000000000002e-05,
      "loss": 1.6173,
      "step": 7580
    },
    {
      "epoch": 0.19,
      "grad_norm": 8.52052116394043,
      "learning_rate": 1.5193220338983052e-05,
      "loss": 1.4508,
      "step": 7590
    },
    {
      "epoch": 0.19,
      "grad_norm": 4.199940204620361,
      "learning_rate": 1.5186440677966103e-05,
      "loss": 1.3637,
      "step": 7600
    },
    {
      "epoch": 0.19,
      "grad_norm": 4.584707260131836,
      "learning_rate": 1.5179661016949154e-05,
      "loss": 1.5719,
      "step": 7610
    },
    {
      "epoch": 0.19,
      "grad_norm": 3.160292387008667,
      "learning_rate": 1.5172881355932206e-05,
      "loss": 1.3207,
      "step": 7620
    },
    {
      "epoch": 0.19,
      "grad_norm": 3.286118268966675,
      "learning_rate": 1.5166101694915255e-05,
      "loss": 1.4304,
      "step": 7630
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.5094937086105347,
      "learning_rate": 1.5159322033898307e-05,
      "loss": 1.4086,
      "step": 7640
    },
    {
      "epoch": 0.19,
      "grad_norm": 2.407759189605713,
      "learning_rate": 1.5152542372881358e-05,
      "loss": 1.5933,
      "step": 7650
    },
    {
      "epoch": 0.19,
      "grad_norm": 3.6175222396850586,
      "learning_rate": 1.514576271186441e-05,
      "loss": 1.358,
      "step": 7660
    },
    {
      "epoch": 0.19,
      "grad_norm": 4.4565839767456055,
      "learning_rate": 1.5138983050847458e-05,
      "loss": 1.6017,
      "step": 7670
    },
    {
      "epoch": 0.19,
      "grad_norm": 5.479851722717285,
      "learning_rate": 1.5132203389830509e-05,
      "loss": 1.5845,
      "step": 7680
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.8264964818954468,
      "learning_rate": 1.512542372881356e-05,
      "loss": 1.5275,
      "step": 7690
    },
    {
      "epoch": 0.19,
      "grad_norm": 3.3226988315582275,
      "learning_rate": 1.511864406779661e-05,
      "loss": 1.2622,
      "step": 7700
    },
    {
      "epoch": 0.19,
      "grad_norm": 5.333388328552246,
      "learning_rate": 1.5111864406779662e-05,
      "loss": 1.2151,
      "step": 7710
    },
    {
      "epoch": 0.19,
      "grad_norm": 4.269395351409912,
      "learning_rate": 1.5105084745762713e-05,
      "loss": 1.5154,
      "step": 7720
    },
    {
      "epoch": 0.19,
      "grad_norm": 2.333357572555542,
      "learning_rate": 1.5098305084745763e-05,
      "loss": 1.4878,
      "step": 7730
    },
    {
      "epoch": 0.19,
      "grad_norm": 5.043832778930664,
      "learning_rate": 1.5091525423728814e-05,
      "loss": 1.5049,
      "step": 7740
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.9587914943695068,
      "learning_rate": 1.5084745762711865e-05,
      "loss": 1.4318,
      "step": 7750
    },
    {
      "epoch": 0.19,
      "grad_norm": 2.670542001724243,
      "learning_rate": 1.5077966101694917e-05,
      "loss": 1.5539,
      "step": 7760
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.336397647857666,
      "learning_rate": 1.5071186440677967e-05,
      "loss": 1.4664,
      "step": 7770
    },
    {
      "epoch": 0.19,
      "grad_norm": 2.9761111736297607,
      "learning_rate": 1.5064406779661018e-05,
      "loss": 1.3044,
      "step": 7780
    },
    {
      "epoch": 0.19,
      "grad_norm": 3.650763988494873,
      "learning_rate": 1.505762711864407e-05,
      "loss": 1.3202,
      "step": 7790
    },
    {
      "epoch": 0.2,
      "grad_norm": 3.2990357875823975,
      "learning_rate": 1.505084745762712e-05,
      "loss": 1.4289,
      "step": 7800
    },
    {
      "epoch": 0.2,
      "grad_norm": 4.1509108543396,
      "learning_rate": 1.504406779661017e-05,
      "loss": 1.2721,
      "step": 7810
    },
    {
      "epoch": 0.2,
      "grad_norm": 2.029076337814331,
      "learning_rate": 1.5037288135593222e-05,
      "loss": 1.4522,
      "step": 7820
    },
    {
      "epoch": 0.2,
      "grad_norm": 3.3788859844207764,
      "learning_rate": 1.5030508474576273e-05,
      "loss": 1.5012,
      "step": 7830
    },
    {
      "epoch": 0.2,
      "grad_norm": 2.077439546585083,
      "learning_rate": 1.5023728813559325e-05,
      "loss": 1.468,
      "step": 7840
    },
    {
      "epoch": 0.2,
      "grad_norm": 5.250193119049072,
      "learning_rate": 1.5016949152542374e-05,
      "loss": 1.3041,
      "step": 7850
    },
    {
      "epoch": 0.2,
      "grad_norm": 3.2319061756134033,
      "learning_rate": 1.5010169491525426e-05,
      "loss": 1.5433,
      "step": 7860
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.9006788730621338,
      "learning_rate": 1.5003389830508477e-05,
      "loss": 1.5216,
      "step": 7870
    },
    {
      "epoch": 0.2,
      "grad_norm": 2.611494302749634,
      "learning_rate": 1.4996610169491525e-05,
      "loss": 1.4739,
      "step": 7880
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.7585150003433228,
      "learning_rate": 1.4989830508474577e-05,
      "loss": 1.3797,
      "step": 7890
    },
    {
      "epoch": 0.2,
      "grad_norm": 2.0867931842803955,
      "learning_rate": 1.4983050847457628e-05,
      "loss": 1.4749,
      "step": 7900
    },
    {
      "epoch": 0.2,
      "grad_norm": 3.2771809101104736,
      "learning_rate": 1.4976271186440678e-05,
      "loss": 1.4982,
      "step": 7910
    },
    {
      "epoch": 0.2,
      "grad_norm": 3.2540512084960938,
      "learning_rate": 1.4969491525423729e-05,
      "loss": 1.1921,
      "step": 7920
    },
    {
      "epoch": 0.2,
      "grad_norm": 2.752269744873047,
      "learning_rate": 1.496271186440678e-05,
      "loss": 1.4736,
      "step": 7930
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.9583200216293335,
      "learning_rate": 1.4955932203389832e-05,
      "loss": 1.5394,
      "step": 7940
    },
    {
      "epoch": 0.2,
      "grad_norm": 2.7196719646453857,
      "learning_rate": 1.4949152542372882e-05,
      "loss": 1.4254,
      "step": 7950
    },
    {
      "epoch": 0.2,
      "grad_norm": 5.240522384643555,
      "learning_rate": 1.4942372881355933e-05,
      "loss": 1.3019,
      "step": 7960
    },
    {
      "epoch": 0.2,
      "grad_norm": 2.6920173168182373,
      "learning_rate": 1.4935593220338984e-05,
      "loss": 1.495,
      "step": 7970
    },
    {
      "epoch": 0.2,
      "grad_norm": 2.3263721466064453,
      "learning_rate": 1.4928813559322036e-05,
      "loss": 1.3855,
      "step": 7980
    },
    {
      "epoch": 0.2,
      "grad_norm": 4.875345230102539,
      "learning_rate": 1.4922033898305086e-05,
      "loss": 1.5045,
      "step": 7990
    },
    {
      "epoch": 0.2,
      "grad_norm": 7.168892860412598,
      "learning_rate": 1.4915254237288137e-05,
      "loss": 1.6184,
      "step": 8000
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.6748573780059814,
      "learning_rate": 1.4908474576271188e-05,
      "loss": 1.4415,
      "step": 8010
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.8683879375457764,
      "learning_rate": 1.490169491525424e-05,
      "loss": 1.3246,
      "step": 8020
    },
    {
      "epoch": 0.2,
      "grad_norm": 5.742013931274414,
      "learning_rate": 1.489491525423729e-05,
      "loss": 1.3101,
      "step": 8030
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.9626543521881104,
      "learning_rate": 1.4888135593220341e-05,
      "loss": 1.4654,
      "step": 8040
    },
    {
      "epoch": 0.2,
      "grad_norm": 2.5889735221862793,
      "learning_rate": 1.4881355932203392e-05,
      "loss": 1.4796,
      "step": 8050
    },
    {
      "epoch": 0.2,
      "grad_norm": 2.22261381149292,
      "learning_rate": 1.4874576271186442e-05,
      "loss": 1.4254,
      "step": 8060
    },
    {
      "epoch": 0.2,
      "grad_norm": 5.62781286239624,
      "learning_rate": 1.4867796610169493e-05,
      "loss": 1.1633,
      "step": 8070
    },
    {
      "epoch": 0.2,
      "grad_norm": 3.9774527549743652,
      "learning_rate": 1.4861016949152545e-05,
      "loss": 1.5795,
      "step": 8080
    },
    {
      "epoch": 0.2,
      "grad_norm": 4.619643688201904,
      "learning_rate": 1.4854237288135593e-05,
      "loss": 1.3261,
      "step": 8090
    },
    {
      "epoch": 0.2,
      "grad_norm": 11.18410873413086,
      "learning_rate": 1.4847457627118644e-05,
      "loss": 1.5561,
      "step": 8100
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.7388226985931396,
      "learning_rate": 1.4840677966101696e-05,
      "loss": 1.5926,
      "step": 8110
    },
    {
      "epoch": 0.2,
      "grad_norm": 5.982414722442627,
      "learning_rate": 1.4833898305084747e-05,
      "loss": 1.6116,
      "step": 8120
    },
    {
      "epoch": 0.2,
      "grad_norm": 4.2974700927734375,
      "learning_rate": 1.4827118644067797e-05,
      "loss": 1.476,
      "step": 8130
    },
    {
      "epoch": 0.2,
      "grad_norm": 2.618072986602783,
      "learning_rate": 1.4820338983050848e-05,
      "loss": 1.3992,
      "step": 8140
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.971746802330017,
      "learning_rate": 1.48135593220339e-05,
      "loss": 1.5074,
      "step": 8150
    },
    {
      "epoch": 0.2,
      "grad_norm": 4.787137985229492,
      "learning_rate": 1.4806779661016951e-05,
      "loss": 1.4629,
      "step": 8160
    },
    {
      "epoch": 0.2,
      "grad_norm": 2.4519615173339844,
      "learning_rate": 1.48e-05,
      "loss": 1.395,
      "step": 8170
    },
    {
      "epoch": 0.2,
      "grad_norm": 6.289714336395264,
      "learning_rate": 1.4793220338983052e-05,
      "loss": 1.4422,
      "step": 8180
    },
    {
      "epoch": 0.2,
      "grad_norm": 3.205418109893799,
      "learning_rate": 1.4786440677966103e-05,
      "loss": 1.389,
      "step": 8190
    },
    {
      "epoch": 0.2,
      "grad_norm": 4.528128147125244,
      "learning_rate": 1.4779661016949153e-05,
      "loss": 1.4964,
      "step": 8200
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.8594542741775513,
      "learning_rate": 1.4772881355932205e-05,
      "loss": 1.4583,
      "step": 8210
    },
    {
      "epoch": 0.21,
      "grad_norm": 3.368403196334839,
      "learning_rate": 1.4766101694915256e-05,
      "loss": 1.3053,
      "step": 8220
    },
    {
      "epoch": 0.21,
      "grad_norm": 4.075615406036377,
      "learning_rate": 1.4759322033898307e-05,
      "loss": 1.3606,
      "step": 8230
    },
    {
      "epoch": 0.21,
      "grad_norm": 4.870560646057129,
      "learning_rate": 1.4752542372881357e-05,
      "loss": 1.4153,
      "step": 8240
    },
    {
      "epoch": 0.21,
      "grad_norm": 2.1522116661071777,
      "learning_rate": 1.4745762711864408e-05,
      "loss": 1.4056,
      "step": 8250
    },
    {
      "epoch": 0.21,
      "grad_norm": 4.34472131729126,
      "learning_rate": 1.473898305084746e-05,
      "loss": 1.4331,
      "step": 8260
    },
    {
      "epoch": 0.21,
      "grad_norm": 3.235410451889038,
      "learning_rate": 1.4732203389830511e-05,
      "loss": 1.354,
      "step": 8270
    },
    {
      "epoch": 0.21,
      "grad_norm": 6.468446254730225,
      "learning_rate": 1.4725423728813561e-05,
      "loss": 1.5236,
      "step": 8280
    },
    {
      "epoch": 0.21,
      "grad_norm": 2.415300130844116,
      "learning_rate": 1.4718644067796612e-05,
      "loss": 1.4708,
      "step": 8290
    },
    {
      "epoch": 0.21,
      "grad_norm": 2.976445198059082,
      "learning_rate": 1.4711864406779662e-05,
      "loss": 1.3956,
      "step": 8300
    },
    {
      "epoch": 0.21,
      "grad_norm": 3.6008412837982178,
      "learning_rate": 1.4705084745762712e-05,
      "loss": 1.4145,
      "step": 8310
    },
    {
      "epoch": 0.21,
      "grad_norm": 5.5210065841674805,
      "learning_rate": 1.4698305084745763e-05,
      "loss": 1.252,
      "step": 8320
    },
    {
      "epoch": 0.21,
      "grad_norm": 2.950364589691162,
      "learning_rate": 1.4691525423728815e-05,
      "loss": 1.1882,
      "step": 8330
    },
    {
      "epoch": 0.21,
      "grad_norm": 2.646656036376953,
      "learning_rate": 1.4684745762711864e-05,
      "loss": 1.3265,
      "step": 8340
    },
    {
      "epoch": 0.21,
      "grad_norm": 5.378515243530273,
      "learning_rate": 1.4677966101694916e-05,
      "loss": 1.4286,
      "step": 8350
    },
    {
      "epoch": 0.21,
      "grad_norm": 6.115920066833496,
      "learning_rate": 1.4671186440677967e-05,
      "loss": 1.446,
      "step": 8360
    },
    {
      "epoch": 0.21,
      "grad_norm": 6.6454596519470215,
      "learning_rate": 1.4664406779661018e-05,
      "loss": 1.434,
      "step": 8370
    },
    {
      "epoch": 0.21,
      "grad_norm": 4.864197254180908,
      "learning_rate": 1.4657627118644068e-05,
      "loss": 1.2945,
      "step": 8380
    },
    {
      "epoch": 0.21,
      "grad_norm": 5.1658549308776855,
      "learning_rate": 1.465084745762712e-05,
      "loss": 1.303,
      "step": 8390
    },
    {
      "epoch": 0.21,
      "grad_norm": 2.6847732067108154,
      "learning_rate": 1.4644067796610171e-05,
      "loss": 1.5418,
      "step": 8400
    },
    {
      "epoch": 0.21,
      "grad_norm": 4.768235683441162,
      "learning_rate": 1.4637288135593222e-05,
      "loss": 1.3809,
      "step": 8410
    },
    {
      "epoch": 0.21,
      "grad_norm": 3.4850351810455322,
      "learning_rate": 1.4630508474576272e-05,
      "loss": 1.4256,
      "step": 8420
    },
    {
      "epoch": 0.21,
      "grad_norm": 3.0509157180786133,
      "learning_rate": 1.4623728813559323e-05,
      "loss": 1.4593,
      "step": 8430
    },
    {
      "epoch": 0.21,
      "grad_norm": 5.4713969230651855,
      "learning_rate": 1.4616949152542375e-05,
      "loss": 1.5795,
      "step": 8440
    },
    {
      "epoch": 0.21,
      "grad_norm": 8.919637680053711,
      "learning_rate": 1.4610169491525426e-05,
      "loss": 1.4627,
      "step": 8450
    },
    {
      "epoch": 0.21,
      "grad_norm": 4.516945838928223,
      "learning_rate": 1.4603389830508476e-05,
      "loss": 1.3197,
      "step": 8460
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.3506308794021606,
      "learning_rate": 1.4596610169491527e-05,
      "loss": 1.3147,
      "step": 8470
    },
    {
      "epoch": 0.21,
      "grad_norm": 3.706346273422241,
      "learning_rate": 1.4589830508474579e-05,
      "loss": 1.4724,
      "step": 8480
    },
    {
      "epoch": 0.21,
      "grad_norm": 3.0023765563964844,
      "learning_rate": 1.458305084745763e-05,
      "loss": 1.4043,
      "step": 8490
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.961065411567688,
      "learning_rate": 1.4576271186440678e-05,
      "loss": 1.5058,
      "step": 8500
    },
    {
      "epoch": 0.21,
      "grad_norm": 2.3559446334838867,
      "learning_rate": 1.456949152542373e-05,
      "loss": 1.4364,
      "step": 8510
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.442697525024414,
      "learning_rate": 1.456271186440678e-05,
      "loss": 1.3921,
      "step": 8520
    },
    {
      "epoch": 0.21,
      "grad_norm": 5.632051467895508,
      "learning_rate": 1.455593220338983e-05,
      "loss": 1.5338,
      "step": 8530
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.7581976652145386,
      "learning_rate": 1.4549152542372882e-05,
      "loss": 1.4413,
      "step": 8540
    },
    {
      "epoch": 0.21,
      "grad_norm": 2.5327064990997314,
      "learning_rate": 1.4542372881355933e-05,
      "loss": 1.3584,
      "step": 8550
    },
    {
      "epoch": 0.21,
      "grad_norm": 2.1675007343292236,
      "learning_rate": 1.4535593220338983e-05,
      "loss": 1.4535,
      "step": 8560
    },
    {
      "epoch": 0.21,
      "grad_norm": 2.6816961765289307,
      "learning_rate": 1.4528813559322035e-05,
      "loss": 1.4044,
      "step": 8570
    },
    {
      "epoch": 0.21,
      "grad_norm": 4.016593933105469,
      "learning_rate": 1.4522033898305086e-05,
      "loss": 1.4604,
      "step": 8580
    },
    {
      "epoch": 0.21,
      "grad_norm": 3.1831982135772705,
      "learning_rate": 1.4515254237288137e-05,
      "loss": 1.5224,
      "step": 8590
    },
    {
      "epoch": 0.21,
      "grad_norm": 2.7489213943481445,
      "learning_rate": 1.4508474576271187e-05,
      "loss": 1.5482,
      "step": 8600
    },
    {
      "epoch": 0.22,
      "grad_norm": 3.789846658706665,
      "learning_rate": 1.4501694915254239e-05,
      "loss": 1.3677,
      "step": 8610
    },
    {
      "epoch": 0.22,
      "grad_norm": 11.019224166870117,
      "learning_rate": 1.449491525423729e-05,
      "loss": 1.2979,
      "step": 8620
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.5070538520812988,
      "learning_rate": 1.4488135593220341e-05,
      "loss": 1.536,
      "step": 8630
    },
    {
      "epoch": 0.22,
      "grad_norm": 3.577901601791382,
      "learning_rate": 1.4481355932203391e-05,
      "loss": 1.594,
      "step": 8640
    },
    {
      "epoch": 0.22,
      "grad_norm": 2.9272878170013428,
      "learning_rate": 1.4474576271186442e-05,
      "loss": 1.4688,
      "step": 8650
    },
    {
      "epoch": 0.22,
      "grad_norm": 3.842909097671509,
      "learning_rate": 1.4467796610169494e-05,
      "loss": 1.4701,
      "step": 8660
    },
    {
      "epoch": 0.22,
      "grad_norm": 4.70853853225708,
      "learning_rate": 1.4461016949152544e-05,
      "loss": 1.4818,
      "step": 8670
    },
    {
      "epoch": 0.22,
      "grad_norm": 3.0963170528411865,
      "learning_rate": 1.4454237288135595e-05,
      "loss": 1.554,
      "step": 8680
    },
    {
      "epoch": 0.22,
      "grad_norm": 2.4648854732513428,
      "learning_rate": 1.4447457627118646e-05,
      "loss": 1.5977,
      "step": 8690
    },
    {
      "epoch": 0.22,
      "grad_norm": 2.417259693145752,
      "learning_rate": 1.4440677966101698e-05,
      "loss": 1.4642,
      "step": 8700
    },
    {
      "epoch": 0.22,
      "grad_norm": 6.033206939697266,
      "learning_rate": 1.4433898305084746e-05,
      "loss": 1.4625,
      "step": 8710
    },
    {
      "epoch": 0.22,
      "grad_norm": 8.898385047912598,
      "learning_rate": 1.4427118644067797e-05,
      "loss": 1.3722,
      "step": 8720
    },
    {
      "epoch": 0.22,
      "grad_norm": 4.355011463165283,
      "learning_rate": 1.4420338983050849e-05,
      "loss": 1.4332,
      "step": 8730
    },
    {
      "epoch": 0.22,
      "grad_norm": 7.822190761566162,
      "learning_rate": 1.4413559322033898e-05,
      "loss": 1.4841,
      "step": 8740
    },
    {
      "epoch": 0.22,
      "grad_norm": 4.317648887634277,
      "learning_rate": 1.440677966101695e-05,
      "loss": 1.5808,
      "step": 8750
    },
    {
      "epoch": 0.22,
      "grad_norm": 3.815460681915283,
      "learning_rate": 1.4400000000000001e-05,
      "loss": 1.5992,
      "step": 8760
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.0499264001846313,
      "learning_rate": 1.4393220338983052e-05,
      "loss": 1.5435,
      "step": 8770
    },
    {
      "epoch": 0.22,
      "grad_norm": 2.490985155105591,
      "learning_rate": 1.4386440677966102e-05,
      "loss": 1.4187,
      "step": 8780
    },
    {
      "epoch": 0.22,
      "grad_norm": 11.051963806152344,
      "learning_rate": 1.4379661016949154e-05,
      "loss": 1.4923,
      "step": 8790
    },
    {
      "epoch": 0.22,
      "grad_norm": 3.150280475616455,
      "learning_rate": 1.4372881355932205e-05,
      "loss": 1.5572,
      "step": 8800
    },
    {
      "epoch": 0.22,
      "grad_norm": 6.757431983947754,
      "learning_rate": 1.4366101694915255e-05,
      "loss": 1.4393,
      "step": 8810
    },
    {
      "epoch": 0.22,
      "grad_norm": 2.0655736923217773,
      "learning_rate": 1.4359322033898306e-05,
      "loss": 1.5541,
      "step": 8820
    },
    {
      "epoch": 0.22,
      "grad_norm": 3.7942028045654297,
      "learning_rate": 1.4352542372881357e-05,
      "loss": 1.2834,
      "step": 8830
    },
    {
      "epoch": 0.22,
      "grad_norm": 5.594410419464111,
      "learning_rate": 1.4345762711864409e-05,
      "loss": 1.3622,
      "step": 8840
    },
    {
      "epoch": 0.22,
      "grad_norm": 3.7694172859191895,
      "learning_rate": 1.4338983050847459e-05,
      "loss": 1.5003,
      "step": 8850
    },
    {
      "epoch": 0.22,
      "grad_norm": 3.6056268215179443,
      "learning_rate": 1.433220338983051e-05,
      "loss": 1.5703,
      "step": 8860
    },
    {
      "epoch": 0.22,
      "grad_norm": 5.224764347076416,
      "learning_rate": 1.4325423728813561e-05,
      "loss": 1.5649,
      "step": 8870
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.6841343641281128,
      "learning_rate": 1.4318644067796613e-05,
      "loss": 1.3928,
      "step": 8880
    },
    {
      "epoch": 0.22,
      "grad_norm": 2.2923529148101807,
      "learning_rate": 1.4311864406779662e-05,
      "loss": 1.4374,
      "step": 8890
    },
    {
      "epoch": 0.22,
      "grad_norm": 3.346454381942749,
      "learning_rate": 1.4305084745762714e-05,
      "loss": 1.4881,
      "step": 8900
    },
    {
      "epoch": 0.22,
      "grad_norm": 2.629026412963867,
      "learning_rate": 1.4298305084745765e-05,
      "loss": 1.5866,
      "step": 8910
    },
    {
      "epoch": 0.22,
      "grad_norm": 3.6858720779418945,
      "learning_rate": 1.4291525423728813e-05,
      "loss": 1.3559,
      "step": 8920
    },
    {
      "epoch": 0.22,
      "grad_norm": 2.0351126194000244,
      "learning_rate": 1.4284745762711865e-05,
      "loss": 1.3109,
      "step": 8930
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.4689570665359497,
      "learning_rate": 1.4277966101694916e-05,
      "loss": 1.3227,
      "step": 8940
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.7875679731369019,
      "learning_rate": 1.4271186440677966e-05,
      "loss": 1.6738,
      "step": 8950
    },
    {
      "epoch": 0.22,
      "grad_norm": 7.477941989898682,
      "learning_rate": 1.4264406779661017e-05,
      "loss": 1.4238,
      "step": 8960
    },
    {
      "epoch": 0.22,
      "grad_norm": 2.491065502166748,
      "learning_rate": 1.4257627118644069e-05,
      "loss": 1.5972,
      "step": 8970
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.9341115951538086,
      "learning_rate": 1.425084745762712e-05,
      "loss": 1.4388,
      "step": 8980
    },
    {
      "epoch": 0.22,
      "grad_norm": 2.093651056289673,
      "learning_rate": 1.424406779661017e-05,
      "loss": 1.3772,
      "step": 8990
    },
    {
      "epoch": 0.23,
      "grad_norm": 4.210513591766357,
      "learning_rate": 1.4237288135593221e-05,
      "loss": 1.3585,
      "step": 9000
    },
    {
      "epoch": 0.23,
      "grad_norm": 4.9974470138549805,
      "learning_rate": 1.4230508474576273e-05,
      "loss": 1.4506,
      "step": 9010
    },
    {
      "epoch": 0.23,
      "grad_norm": 3.516511917114258,
      "learning_rate": 1.4223728813559324e-05,
      "loss": 1.3557,
      "step": 9020
    },
    {
      "epoch": 0.23,
      "grad_norm": 2.2725930213928223,
      "learning_rate": 1.4216949152542374e-05,
      "loss": 1.6309,
      "step": 9030
    },
    {
      "epoch": 0.23,
      "grad_norm": 2.875182867050171,
      "learning_rate": 1.4210169491525425e-05,
      "loss": 1.4972,
      "step": 9040
    },
    {
      "epoch": 0.23,
      "grad_norm": 3.496995210647583,
      "learning_rate": 1.4203389830508476e-05,
      "loss": 1.4366,
      "step": 9050
    },
    {
      "epoch": 0.23,
      "grad_norm": 4.719113349914551,
      "learning_rate": 1.4196610169491528e-05,
      "loss": 1.4736,
      "step": 9060
    },
    {
      "epoch": 0.23,
      "grad_norm": 3.0286519527435303,
      "learning_rate": 1.4189830508474578e-05,
      "loss": 1.1393,
      "step": 9070
    },
    {
      "epoch": 0.23,
      "grad_norm": 6.375524520874023,
      "learning_rate": 1.4183050847457629e-05,
      "loss": 1.4686,
      "step": 9080
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.3006030321121216,
      "learning_rate": 1.417627118644068e-05,
      "loss": 1.5033,
      "step": 9090
    },
    {
      "epoch": 0.23,
      "grad_norm": 11.15798568725586,
      "learning_rate": 1.416949152542373e-05,
      "loss": 1.5374,
      "step": 9100
    },
    {
      "epoch": 0.23,
      "grad_norm": 4.659414291381836,
      "learning_rate": 1.4162711864406781e-05,
      "loss": 1.4185,
      "step": 9110
    },
    {
      "epoch": 0.23,
      "grad_norm": 4.72214937210083,
      "learning_rate": 1.4155932203389833e-05,
      "loss": 1.4763,
      "step": 9120
    },
    {
      "epoch": 0.23,
      "grad_norm": 3.757380723953247,
      "learning_rate": 1.414915254237288e-05,
      "loss": 1.4475,
      "step": 9130
    },
    {
      "epoch": 0.23,
      "grad_norm": 3.793137788772583,
      "learning_rate": 1.4142372881355932e-05,
      "loss": 1.4503,
      "step": 9140
    },
    {
      "epoch": 0.23,
      "grad_norm": 2.3396735191345215,
      "learning_rate": 1.4135593220338984e-05,
      "loss": 1.3816,
      "step": 9150
    },
    {
      "epoch": 0.23,
      "grad_norm": 2.8242173194885254,
      "learning_rate": 1.4128813559322035e-05,
      "loss": 1.4597,
      "step": 9160
    },
    {
      "epoch": 0.23,
      "grad_norm": 3.4906699657440186,
      "learning_rate": 1.4122033898305085e-05,
      "loss": 1.2626,
      "step": 9170
    },
    {
      "epoch": 0.23,
      "grad_norm": 6.765458106994629,
      "learning_rate": 1.4115254237288136e-05,
      "loss": 1.3799,
      "step": 9180
    },
    {
      "epoch": 0.23,
      "grad_norm": 5.313246250152588,
      "learning_rate": 1.4108474576271188e-05,
      "loss": 1.4815,
      "step": 9190
    },
    {
      "epoch": 0.23,
      "grad_norm": 3.532703161239624,
      "learning_rate": 1.4101694915254239e-05,
      "loss": 1.4935,
      "step": 9200
    },
    {
      "epoch": 0.23,
      "grad_norm": 2.790750026702881,
      "learning_rate": 1.4094915254237289e-05,
      "loss": 1.3319,
      "step": 9210
    },
    {
      "epoch": 0.23,
      "grad_norm": 7.300121307373047,
      "learning_rate": 1.408813559322034e-05,
      "loss": 1.5353,
      "step": 9220
    },
    {
      "epoch": 0.23,
      "grad_norm": 5.028524398803711,
      "learning_rate": 1.4081355932203391e-05,
      "loss": 1.4873,
      "step": 9230
    },
    {
      "epoch": 0.23,
      "grad_norm": 4.20883846282959,
      "learning_rate": 1.4074576271186441e-05,
      "loss": 1.4465,
      "step": 9240
    },
    {
      "epoch": 0.23,
      "grad_norm": 2.486734390258789,
      "learning_rate": 1.4067796610169493e-05,
      "loss": 1.3537,
      "step": 9250
    },
    {
      "epoch": 0.23,
      "grad_norm": 2.898331642150879,
      "learning_rate": 1.4061016949152544e-05,
      "loss": 1.4738,
      "step": 9260
    },
    {
      "epoch": 0.23,
      "grad_norm": 2.013429641723633,
      "learning_rate": 1.4054237288135595e-05,
      "loss": 1.296,
      "step": 9270
    },
    {
      "epoch": 0.23,
      "grad_norm": 3.667020559310913,
      "learning_rate": 1.4047457627118645e-05,
      "loss": 1.4978,
      "step": 9280
    },
    {
      "epoch": 0.23,
      "grad_norm": 2.7311887741088867,
      "learning_rate": 1.4040677966101696e-05,
      "loss": 1.3249,
      "step": 9290
    },
    {
      "epoch": 0.23,
      "grad_norm": 2.2948977947235107,
      "learning_rate": 1.4033898305084748e-05,
      "loss": 1.4484,
      "step": 9300
    },
    {
      "epoch": 0.23,
      "grad_norm": 3.781008243560791,
      "learning_rate": 1.40271186440678e-05,
      "loss": 1.3611,
      "step": 9310
    },
    {
      "epoch": 0.23,
      "grad_norm": 5.073706150054932,
      "learning_rate": 1.4020338983050849e-05,
      "loss": 1.4997,
      "step": 9320
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.5371689796447754,
      "learning_rate": 1.4013559322033899e-05,
      "loss": 1.3964,
      "step": 9330
    },
    {
      "epoch": 0.23,
      "grad_norm": 4.89487886428833,
      "learning_rate": 1.400677966101695e-05,
      "loss": 1.4582,
      "step": 9340
    },
    {
      "epoch": 0.23,
      "grad_norm": 3.8304333686828613,
      "learning_rate": 1.4e-05,
      "loss": 1.3845,
      "step": 9350
    },
    {
      "epoch": 0.23,
      "grad_norm": 4.8575286865234375,
      "learning_rate": 1.3993220338983051e-05,
      "loss": 1.5418,
      "step": 9360
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.648848295211792,
      "learning_rate": 1.3986440677966103e-05,
      "loss": 1.3597,
      "step": 9370
    },
    {
      "epoch": 0.23,
      "grad_norm": 2.135981559753418,
      "learning_rate": 1.3979661016949152e-05,
      "loss": 1.6127,
      "step": 9380
    },
    {
      "epoch": 0.23,
      "grad_norm": 2.0944900512695312,
      "learning_rate": 1.3972881355932204e-05,
      "loss": 1.4883,
      "step": 9390
    },
    {
      "epoch": 0.23,
      "grad_norm": 3.724740743637085,
      "learning_rate": 1.3966101694915255e-05,
      "loss": 1.341,
      "step": 9400
    },
    {
      "epoch": 0.24,
      "grad_norm": 2.3570704460144043,
      "learning_rate": 1.3959322033898306e-05,
      "loss": 1.35,
      "step": 9410
    },
    {
      "epoch": 0.24,
      "grad_norm": 2.7155792713165283,
      "learning_rate": 1.3952542372881356e-05,
      "loss": 1.5027,
      "step": 9420
    },
    {
      "epoch": 0.24,
      "grad_norm": 9.99693489074707,
      "learning_rate": 1.3945762711864408e-05,
      "loss": 1.4363,
      "step": 9430
    },
    {
      "epoch": 0.24,
      "grad_norm": 5.962501525878906,
      "learning_rate": 1.3938983050847459e-05,
      "loss": 1.3491,
      "step": 9440
    },
    {
      "epoch": 0.24,
      "grad_norm": 4.423463344573975,
      "learning_rate": 1.393220338983051e-05,
      "loss": 1.6445,
      "step": 9450
    },
    {
      "epoch": 0.24,
      "grad_norm": 5.279736518859863,
      "learning_rate": 1.392542372881356e-05,
      "loss": 1.4616,
      "step": 9460
    },
    {
      "epoch": 0.24,
      "grad_norm": 2.550114870071411,
      "learning_rate": 1.3918644067796612e-05,
      "loss": 1.4488,
      "step": 9470
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.9499937295913696,
      "learning_rate": 1.3911864406779663e-05,
      "loss": 1.5656,
      "step": 9480
    },
    {
      "epoch": 0.24,
      "grad_norm": 6.516053199768066,
      "learning_rate": 1.3905084745762714e-05,
      "loss": 1.3519,
      "step": 9490
    },
    {
      "epoch": 0.24,
      "grad_norm": 3.5697858333587646,
      "learning_rate": 1.3898305084745764e-05,
      "loss": 1.3955,
      "step": 9500
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.6801804304122925,
      "learning_rate": 1.3891525423728815e-05,
      "loss": 1.6211,
      "step": 9510
    },
    {
      "epoch": 0.24,
      "grad_norm": 5.539935111999512,
      "learning_rate": 1.3884745762711867e-05,
      "loss": 1.4392,
      "step": 9520
    },
    {
      "epoch": 0.24,
      "grad_norm": 4.8439178466796875,
      "learning_rate": 1.3877966101694918e-05,
      "loss": 1.4856,
      "step": 9530
    },
    {
      "epoch": 0.24,
      "grad_norm": 4.0676798820495605,
      "learning_rate": 1.3871186440677966e-05,
      "loss": 1.5147,
      "step": 9540
    },
    {
      "epoch": 0.24,
      "grad_norm": 5.304898738861084,
      "learning_rate": 1.3864406779661018e-05,
      "loss": 1.4325,
      "step": 9550
    },
    {
      "epoch": 0.24,
      "grad_norm": 2.493945837020874,
      "learning_rate": 1.3857627118644067e-05,
      "loss": 1.4402,
      "step": 9560
    },
    {
      "epoch": 0.24,
      "grad_norm": 10.46382999420166,
      "learning_rate": 1.3850847457627119e-05,
      "loss": 1.3506,
      "step": 9570
    },
    {
      "epoch": 0.24,
      "grad_norm": 2.717838764190674,
      "learning_rate": 1.384406779661017e-05,
      "loss": 1.317,
      "step": 9580
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.8320386409759521,
      "learning_rate": 1.3837288135593222e-05,
      "loss": 1.5172,
      "step": 9590
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.4577361345291138,
      "learning_rate": 1.3830508474576271e-05,
      "loss": 1.3049,
      "step": 9600
    },
    {
      "epoch": 0.24,
      "grad_norm": 3.4513816833496094,
      "learning_rate": 1.3823728813559323e-05,
      "loss": 1.5363,
      "step": 9610
    },
    {
      "epoch": 0.24,
      "grad_norm": 5.6405134201049805,
      "learning_rate": 1.3816949152542374e-05,
      "loss": 1.5302,
      "step": 9620
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.441107988357544,
      "learning_rate": 1.3810169491525425e-05,
      "loss": 1.2896,
      "step": 9630
    },
    {
      "epoch": 0.24,
      "grad_norm": 8.282943725585938,
      "learning_rate": 1.3803389830508475e-05,
      "loss": 1.3448,
      "step": 9640
    },
    {
      "epoch": 0.24,
      "grad_norm": 3.812814474105835,
      "learning_rate": 1.3796610169491527e-05,
      "loss": 1.3764,
      "step": 9650
    },
    {
      "epoch": 0.24,
      "grad_norm": 7.053981304168701,
      "learning_rate": 1.3789830508474578e-05,
      "loss": 1.2607,
      "step": 9660
    },
    {
      "epoch": 0.24,
      "grad_norm": 7.5300517082214355,
      "learning_rate": 1.378305084745763e-05,
      "loss": 1.509,
      "step": 9670
    },
    {
      "epoch": 0.24,
      "grad_norm": 3.457367420196533,
      "learning_rate": 1.3776271186440679e-05,
      "loss": 1.2809,
      "step": 9680
    },
    {
      "epoch": 0.24,
      "grad_norm": 3.1301324367523193,
      "learning_rate": 1.376949152542373e-05,
      "loss": 1.4504,
      "step": 9690
    },
    {
      "epoch": 0.24,
      "grad_norm": 7.733901023864746,
      "learning_rate": 1.3762711864406782e-05,
      "loss": 1.4695,
      "step": 9700
    },
    {
      "epoch": 0.24,
      "grad_norm": 3.885404348373413,
      "learning_rate": 1.3755932203389832e-05,
      "loss": 1.487,
      "step": 9710
    },
    {
      "epoch": 0.24,
      "grad_norm": 7.525615215301514,
      "learning_rate": 1.3749152542372883e-05,
      "loss": 1.4926,
      "step": 9720
    },
    {
      "epoch": 0.24,
      "grad_norm": 4.785201072692871,
      "learning_rate": 1.3742372881355934e-05,
      "loss": 1.4202,
      "step": 9730
    },
    {
      "epoch": 0.24,
      "grad_norm": 2.442051649093628,
      "learning_rate": 1.3735593220338986e-05,
      "loss": 1.4976,
      "step": 9740
    },
    {
      "epoch": 0.24,
      "grad_norm": 3.9468209743499756,
      "learning_rate": 1.3728813559322034e-05,
      "loss": 1.3574,
      "step": 9750
    },
    {
      "epoch": 0.24,
      "grad_norm": 2.7709028720855713,
      "learning_rate": 1.3722033898305085e-05,
      "loss": 1.4645,
      "step": 9760
    },
    {
      "epoch": 0.24,
      "grad_norm": 6.794541358947754,
      "learning_rate": 1.3715254237288137e-05,
      "loss": 1.5522,
      "step": 9770
    },
    {
      "epoch": 0.24,
      "grad_norm": 4.260660648345947,
      "learning_rate": 1.3708474576271186e-05,
      "loss": 1.4093,
      "step": 9780
    },
    {
      "epoch": 0.24,
      "grad_norm": 3.1224398612976074,
      "learning_rate": 1.3701694915254238e-05,
      "loss": 1.3535,
      "step": 9790
    },
    {
      "epoch": 0.24,
      "grad_norm": 4.128883361816406,
      "learning_rate": 1.3694915254237289e-05,
      "loss": 1.5056,
      "step": 9800
    },
    {
      "epoch": 0.25,
      "grad_norm": 2.559934139251709,
      "learning_rate": 1.368813559322034e-05,
      "loss": 1.4926,
      "step": 9810
    },
    {
      "epoch": 0.25,
      "grad_norm": 4.7702202796936035,
      "learning_rate": 1.368135593220339e-05,
      "loss": 1.4751,
      "step": 9820
    },
    {
      "epoch": 0.25,
      "grad_norm": 4.250860214233398,
      "learning_rate": 1.3674576271186442e-05,
      "loss": 1.3841,
      "step": 9830
    },
    {
      "epoch": 0.25,
      "grad_norm": 3.4019124507904053,
      "learning_rate": 1.3667796610169493e-05,
      "loss": 1.304,
      "step": 9840
    },
    {
      "epoch": 0.25,
      "grad_norm": 3.3869032859802246,
      "learning_rate": 1.3661016949152543e-05,
      "loss": 1.4193,
      "step": 9850
    },
    {
      "epoch": 0.25,
      "grad_norm": 7.182471752166748,
      "learning_rate": 1.3654237288135594e-05,
      "loss": 1.5191,
      "step": 9860
    },
    {
      "epoch": 0.25,
      "grad_norm": 3.5116724967956543,
      "learning_rate": 1.3647457627118646e-05,
      "loss": 1.3531,
      "step": 9870
    },
    {
      "epoch": 0.25,
      "grad_norm": 3.9179270267486572,
      "learning_rate": 1.3640677966101697e-05,
      "loss": 1.5527,
      "step": 9880
    },
    {
      "epoch": 0.25,
      "grad_norm": 7.844541072845459,
      "learning_rate": 1.3633898305084747e-05,
      "loss": 1.3132,
      "step": 9890
    },
    {
      "epoch": 0.25,
      "grad_norm": 2.2254388332366943,
      "learning_rate": 1.3627118644067798e-05,
      "loss": 1.3693,
      "step": 9900
    },
    {
      "epoch": 0.25,
      "grad_norm": 6.887940406799316,
      "learning_rate": 1.362033898305085e-05,
      "loss": 1.3659,
      "step": 9910
    },
    {
      "epoch": 0.25,
      "grad_norm": 6.743874549865723,
      "learning_rate": 1.36135593220339e-05,
      "loss": 1.3034,
      "step": 9920
    },
    {
      "epoch": 0.25,
      "grad_norm": 6.084103107452393,
      "learning_rate": 1.360677966101695e-05,
      "loss": 1.4221,
      "step": 9930
    },
    {
      "epoch": 0.25,
      "grad_norm": 3.6407341957092285,
      "learning_rate": 1.3600000000000002e-05,
      "loss": 1.3658,
      "step": 9940
    },
    {
      "epoch": 0.25,
      "grad_norm": 3.9673774242401123,
      "learning_rate": 1.3593220338983053e-05,
      "loss": 1.4771,
      "step": 9950
    },
    {
      "epoch": 0.25,
      "grad_norm": 4.6172919273376465,
      "learning_rate": 1.3586440677966101e-05,
      "loss": 1.4567,
      "step": 9960
    },
    {
      "epoch": 0.25,
      "grad_norm": 3.7408809661865234,
      "learning_rate": 1.3579661016949153e-05,
      "loss": 1.4905,
      "step": 9970
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.6898508071899414,
      "learning_rate": 1.3572881355932204e-05,
      "loss": 1.5118,
      "step": 9980
    },
    {
      "epoch": 0.25,
      "grad_norm": 3.319342613220215,
      "learning_rate": 1.3566101694915254e-05,
      "loss": 1.4178,
      "step": 9990
    },
    {
      "epoch": 0.25,
      "grad_norm": 4.550518989562988,
      "learning_rate": 1.3559322033898305e-05,
      "loss": 1.5203,
      "step": 10000
    },
    {
      "epoch": 0.25,
      "eval_loss": 1.4036680459976196,
      "eval_runtime": 122.1773,
      "eval_samples_per_second": 8.185,
      "eval_steps_per_second": 8.185,
      "step": 10000
    },
    {
      "epoch": 0.25,
      "grad_norm": 9.537081718444824,
      "learning_rate": 1.3552542372881357e-05,
      "loss": 1.3961,
      "step": 10010
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.5685465335845947,
      "learning_rate": 1.3545762711864408e-05,
      "loss": 1.3974,
      "step": 10020
    },
    {
      "epoch": 0.25,
      "grad_norm": 2.5580856800079346,
      "learning_rate": 1.3538983050847458e-05,
      "loss": 1.5613,
      "step": 10030
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.2093429565429688,
      "learning_rate": 1.353220338983051e-05,
      "loss": 1.5301,
      "step": 10040
    },
    {
      "epoch": 0.25,
      "grad_norm": 4.904521942138672,
      "learning_rate": 1.352542372881356e-05,
      "loss": 1.4335,
      "step": 10050
    },
    {
      "epoch": 0.25,
      "grad_norm": 3.853315830230713,
      "learning_rate": 1.3518644067796612e-05,
      "loss": 1.314,
      "step": 10060
    },
    {
      "epoch": 0.25,
      "grad_norm": 3.8313851356506348,
      "learning_rate": 1.3511864406779662e-05,
      "loss": 1.4974,
      "step": 10070
    },
    {
      "epoch": 0.25,
      "grad_norm": 3.672273635864258,
      "learning_rate": 1.3505084745762713e-05,
      "loss": 1.3091,
      "step": 10080
    },
    {
      "epoch": 0.25,
      "grad_norm": 3.773167371749878,
      "learning_rate": 1.3498305084745764e-05,
      "loss": 1.5238,
      "step": 10090
    },
    {
      "epoch": 0.25,
      "grad_norm": 7.905582427978516,
      "learning_rate": 1.3491525423728816e-05,
      "loss": 1.4102,
      "step": 10100
    },
    {
      "epoch": 0.25,
      "grad_norm": 5.653487682342529,
      "learning_rate": 1.3484745762711866e-05,
      "loss": 1.4226,
      "step": 10110
    },
    {
      "epoch": 0.25,
      "grad_norm": 2.5647854804992676,
      "learning_rate": 1.3477966101694917e-05,
      "loss": 1.4584,
      "step": 10120
    },
    {
      "epoch": 0.25,
      "grad_norm": 3.5463757514953613,
      "learning_rate": 1.3471186440677968e-05,
      "loss": 1.437,
      "step": 10130
    },
    {
      "epoch": 0.25,
      "grad_norm": 3.1035516262054443,
      "learning_rate": 1.346440677966102e-05,
      "loss": 1.4322,
      "step": 10140
    },
    {
      "epoch": 0.25,
      "grad_norm": 4.035073757171631,
      "learning_rate": 1.345762711864407e-05,
      "loss": 1.4553,
      "step": 10150
    },
    {
      "epoch": 0.25,
      "grad_norm": 5.994688987731934,
      "learning_rate": 1.3450847457627121e-05,
      "loss": 1.409,
      "step": 10160
    },
    {
      "epoch": 0.25,
      "grad_norm": 5.073398113250732,
      "learning_rate": 1.3444067796610169e-05,
      "loss": 1.4535,
      "step": 10170
    },
    {
      "epoch": 0.25,
      "grad_norm": 2.7315621376037598,
      "learning_rate": 1.343728813559322e-05,
      "loss": 1.4017,
      "step": 10180
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.8778635263442993,
      "learning_rate": 1.3430508474576272e-05,
      "loss": 1.3789,
      "step": 10190
    },
    {
      "epoch": 0.26,
      "grad_norm": 4.361875057220459,
      "learning_rate": 1.3423728813559323e-05,
      "loss": 1.5064,
      "step": 10200
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.1118437051773071,
      "learning_rate": 1.3416949152542373e-05,
      "loss": 1.4267,
      "step": 10210
    },
    {
      "epoch": 0.26,
      "grad_norm": 5.846581935882568,
      "learning_rate": 1.3410169491525424e-05,
      "loss": 1.5792,
      "step": 10220
    },
    {
      "epoch": 0.26,
      "grad_norm": 2.2367305755615234,
      "learning_rate": 1.3403389830508476e-05,
      "loss": 1.5034,
      "step": 10230
    },
    {
      "epoch": 0.26,
      "grad_norm": 7.08678674697876,
      "learning_rate": 1.3396610169491527e-05,
      "loss": 1.4434,
      "step": 10240
    },
    {
      "epoch": 0.26,
      "grad_norm": 4.037341117858887,
      "learning_rate": 1.3389830508474577e-05,
      "loss": 1.418,
      "step": 10250
    },
    {
      "epoch": 0.26,
      "grad_norm": 2.9176080226898193,
      "learning_rate": 1.3383050847457628e-05,
      "loss": 1.3746,
      "step": 10260
    },
    {
      "epoch": 0.26,
      "grad_norm": 6.231027603149414,
      "learning_rate": 1.337627118644068e-05,
      "loss": 1.4908,
      "step": 10270
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.6426746845245361,
      "learning_rate": 1.3369491525423731e-05,
      "loss": 1.4773,
      "step": 10280
    },
    {
      "epoch": 0.26,
      "grad_norm": 5.620302200317383,
      "learning_rate": 1.336271186440678e-05,
      "loss": 1.3068,
      "step": 10290
    },
    {
      "epoch": 0.26,
      "grad_norm": 2.1782431602478027,
      "learning_rate": 1.3355932203389832e-05,
      "loss": 1.3158,
      "step": 10300
    },
    {
      "epoch": 0.26,
      "grad_norm": 2.4901323318481445,
      "learning_rate": 1.3349152542372883e-05,
      "loss": 1.4511,
      "step": 10310
    },
    {
      "epoch": 0.26,
      "grad_norm": 3.7503662109375,
      "learning_rate": 1.3342372881355933e-05,
      "loss": 1.5034,
      "step": 10320
    },
    {
      "epoch": 0.26,
      "grad_norm": 2.144787311553955,
      "learning_rate": 1.3335593220338985e-05,
      "loss": 1.3971,
      "step": 10330
    },
    {
      "epoch": 0.26,
      "grad_norm": 5.365964889526367,
      "learning_rate": 1.3328813559322036e-05,
      "loss": 1.3059,
      "step": 10340
    },
    {
      "epoch": 0.26,
      "grad_norm": 2.337282419204712,
      "learning_rate": 1.3322033898305087e-05,
      "loss": 1.4279,
      "step": 10350
    },
    {
      "epoch": 0.26,
      "grad_norm": 13.3304443359375,
      "learning_rate": 1.3315254237288137e-05,
      "loss": 1.5418,
      "step": 10360
    },
    {
      "epoch": 0.26,
      "grad_norm": 6.966452598571777,
      "learning_rate": 1.3308474576271187e-05,
      "loss": 1.5011,
      "step": 10370
    },
    {
      "epoch": 0.26,
      "grad_norm": 2.3808274269104004,
      "learning_rate": 1.3301694915254238e-05,
      "loss": 1.5145,
      "step": 10380
    },
    {
      "epoch": 0.26,
      "grad_norm": 6.16119909286499,
      "learning_rate": 1.3294915254237288e-05,
      "loss": 1.298,
      "step": 10390
    },
    {
      "epoch": 0.26,
      "grad_norm": 4.000382423400879,
      "learning_rate": 1.328813559322034e-05,
      "loss": 1.4955,
      "step": 10400
    },
    {
      "epoch": 0.26,
      "grad_norm": 3.535064697265625,
      "learning_rate": 1.328135593220339e-05,
      "loss": 1.3485,
      "step": 10410
    },
    {
      "epoch": 0.26,
      "grad_norm": 9.773163795471191,
      "learning_rate": 1.3274576271186442e-05,
      "loss": 1.4454,
      "step": 10420
    },
    {
      "epoch": 0.26,
      "grad_norm": 2.1642231941223145,
      "learning_rate": 1.3267796610169492e-05,
      "loss": 1.5313,
      "step": 10430
    },
    {
      "epoch": 0.26,
      "grad_norm": 4.243091106414795,
      "learning_rate": 1.3261016949152543e-05,
      "loss": 1.3294,
      "step": 10440
    },
    {
      "epoch": 0.26,
      "grad_norm": 5.057590961456299,
      "learning_rate": 1.3254237288135595e-05,
      "loss": 1.537,
      "step": 10450
    },
    {
      "epoch": 0.26,
      "grad_norm": 2.3524413108825684,
      "learning_rate": 1.3247457627118644e-05,
      "loss": 1.4868,
      "step": 10460
    },
    {
      "epoch": 0.26,
      "grad_norm": 8.08791732788086,
      "learning_rate": 1.3240677966101696e-05,
      "loss": 1.3867,
      "step": 10470
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.56927490234375,
      "learning_rate": 1.3233898305084747e-05,
      "loss": 1.2215,
      "step": 10480
    },
    {
      "epoch": 0.26,
      "grad_norm": 2.067445993423462,
      "learning_rate": 1.3227118644067798e-05,
      "loss": 1.4979,
      "step": 10490
    },
    {
      "epoch": 0.26,
      "grad_norm": 2.5484492778778076,
      "learning_rate": 1.3220338983050848e-05,
      "loss": 1.4181,
      "step": 10500
    },
    {
      "epoch": 0.26,
      "grad_norm": 3.527362585067749,
      "learning_rate": 1.32135593220339e-05,
      "loss": 1.6717,
      "step": 10510
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.6202934980392456,
      "learning_rate": 1.3206779661016951e-05,
      "loss": 1.4794,
      "step": 10520
    },
    {
      "epoch": 0.26,
      "grad_norm": 2.4861209392547607,
      "learning_rate": 1.3200000000000002e-05,
      "loss": 1.5406,
      "step": 10530
    },
    {
      "epoch": 0.26,
      "grad_norm": 6.074316501617432,
      "learning_rate": 1.3193220338983052e-05,
      "loss": 1.4283,
      "step": 10540
    },
    {
      "epoch": 0.26,
      "grad_norm": 4.089766025543213,
      "learning_rate": 1.3186440677966103e-05,
      "loss": 1.454,
      "step": 10550
    },
    {
      "epoch": 0.26,
      "grad_norm": 3.9060821533203125,
      "learning_rate": 1.3179661016949155e-05,
      "loss": 1.5692,
      "step": 10560
    },
    {
      "epoch": 0.26,
      "grad_norm": 2.4621622562408447,
      "learning_rate": 1.3172881355932206e-05,
      "loss": 1.418,
      "step": 10570
    },
    {
      "epoch": 0.26,
      "grad_norm": 2.693464994430542,
      "learning_rate": 1.3166101694915254e-05,
      "loss": 1.4534,
      "step": 10580
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.2386515140533447,
      "learning_rate": 1.3159322033898306e-05,
      "loss": 1.5029,
      "step": 10590
    },
    {
      "epoch": 0.27,
      "grad_norm": 5.441234111785889,
      "learning_rate": 1.3152542372881355e-05,
      "loss": 1.4803,
      "step": 10600
    },
    {
      "epoch": 0.27,
      "grad_norm": 5.770966053009033,
      "learning_rate": 1.3145762711864407e-05,
      "loss": 1.348,
      "step": 10610
    },
    {
      "epoch": 0.27,
      "grad_norm": 3.701178550720215,
      "learning_rate": 1.3138983050847458e-05,
      "loss": 1.3741,
      "step": 10620
    },
    {
      "epoch": 0.27,
      "grad_norm": 4.069987773895264,
      "learning_rate": 1.313220338983051e-05,
      "loss": 1.3451,
      "step": 10630
    },
    {
      "epoch": 0.27,
      "grad_norm": 3.717097759246826,
      "learning_rate": 1.312542372881356e-05,
      "loss": 1.5033,
      "step": 10640
    },
    {
      "epoch": 0.27,
      "grad_norm": 5.649170398712158,
      "learning_rate": 1.311864406779661e-05,
      "loss": 1.1766,
      "step": 10650
    },
    {
      "epoch": 0.27,
      "grad_norm": 6.903465747833252,
      "learning_rate": 1.3111864406779662e-05,
      "loss": 1.4381,
      "step": 10660
    },
    {
      "epoch": 0.27,
      "grad_norm": 10.294591903686523,
      "learning_rate": 1.3105084745762714e-05,
      "loss": 1.5075,
      "step": 10670
    },
    {
      "epoch": 0.27,
      "grad_norm": 4.394117832183838,
      "learning_rate": 1.3098305084745763e-05,
      "loss": 1.4169,
      "step": 10680
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.8759058713912964,
      "learning_rate": 1.3091525423728815e-05,
      "loss": 1.3812,
      "step": 10690
    },
    {
      "epoch": 0.27,
      "grad_norm": 6.578223705291748,
      "learning_rate": 1.3084745762711866e-05,
      "loss": 1.4703,
      "step": 10700
    },
    {
      "epoch": 0.27,
      "grad_norm": 2.173283100128174,
      "learning_rate": 1.3077966101694917e-05,
      "loss": 1.473,
      "step": 10710
    },
    {
      "epoch": 0.27,
      "grad_norm": 3.296926736831665,
      "learning_rate": 1.3071186440677967e-05,
      "loss": 1.3823,
      "step": 10720
    },
    {
      "epoch": 0.27,
      "grad_norm": 2.94889497756958,
      "learning_rate": 1.3064406779661019e-05,
      "loss": 1.4051,
      "step": 10730
    },
    {
      "epoch": 0.27,
      "grad_norm": 4.6259284019470215,
      "learning_rate": 1.305762711864407e-05,
      "loss": 1.3028,
      "step": 10740
    },
    {
      "epoch": 0.27,
      "grad_norm": 3.5650174617767334,
      "learning_rate": 1.305084745762712e-05,
      "loss": 1.3598,
      "step": 10750
    },
    {
      "epoch": 0.27,
      "grad_norm": 4.751513481140137,
      "learning_rate": 1.3044067796610171e-05,
      "loss": 1.3922,
      "step": 10760
    },
    {
      "epoch": 0.27,
      "grad_norm": 3.40189528465271,
      "learning_rate": 1.3037288135593222e-05,
      "loss": 1.4305,
      "step": 10770
    },
    {
      "epoch": 0.27,
      "grad_norm": 6.658262252807617,
      "learning_rate": 1.3030508474576274e-05,
      "loss": 1.391,
      "step": 10780
    },
    {
      "epoch": 0.27,
      "grad_norm": 4.6862359046936035,
      "learning_rate": 1.3023728813559322e-05,
      "loss": 1.4643,
      "step": 10790
    },
    {
      "epoch": 0.27,
      "grad_norm": 3.3371682167053223,
      "learning_rate": 1.3016949152542373e-05,
      "loss": 1.4081,
      "step": 10800
    },
    {
      "epoch": 0.27,
      "grad_norm": 5.789614677429199,
      "learning_rate": 1.3010169491525425e-05,
      "loss": 1.532,
      "step": 10810
    },
    {
      "epoch": 0.27,
      "grad_norm": 3.1568942070007324,
      "learning_rate": 1.3003389830508474e-05,
      "loss": 1.6236,
      "step": 10820
    },
    {
      "epoch": 0.27,
      "grad_norm": 4.688432693481445,
      "learning_rate": 1.2996610169491526e-05,
      "loss": 1.3767,
      "step": 10830
    },
    {
      "epoch": 0.27,
      "grad_norm": 4.97307825088501,
      "learning_rate": 1.2989830508474577e-05,
      "loss": 1.4694,
      "step": 10840
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.5381091833114624,
      "learning_rate": 1.2983050847457629e-05,
      "loss": 1.3772,
      "step": 10850
    },
    {
      "epoch": 0.27,
      "grad_norm": 5.210800647735596,
      "learning_rate": 1.2976271186440678e-05,
      "loss": 1.4061,
      "step": 10860
    },
    {
      "epoch": 0.27,
      "grad_norm": 6.574726104736328,
      "learning_rate": 1.296949152542373e-05,
      "loss": 1.348,
      "step": 10870
    },
    {
      "epoch": 0.27,
      "grad_norm": 3.936906576156616,
      "learning_rate": 1.2962711864406781e-05,
      "loss": 1.4018,
      "step": 10880
    },
    {
      "epoch": 0.27,
      "grad_norm": 4.410968780517578,
      "learning_rate": 1.295593220338983e-05,
      "loss": 1.5595,
      "step": 10890
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.9505410194396973,
      "learning_rate": 1.2949152542372882e-05,
      "loss": 1.4414,
      "step": 10900
    },
    {
      "epoch": 0.27,
      "grad_norm": 3.7938876152038574,
      "learning_rate": 1.2942372881355934e-05,
      "loss": 1.3939,
      "step": 10910
    },
    {
      "epoch": 0.27,
      "grad_norm": 2.310593366622925,
      "learning_rate": 1.2935593220338985e-05,
      "loss": 1.4798,
      "step": 10920
    },
    {
      "epoch": 0.27,
      "grad_norm": 2.3763818740844727,
      "learning_rate": 1.2928813559322035e-05,
      "loss": 1.3079,
      "step": 10930
    },
    {
      "epoch": 0.27,
      "grad_norm": 3.2649786472320557,
      "learning_rate": 1.2922033898305086e-05,
      "loss": 1.4095,
      "step": 10940
    },
    {
      "epoch": 0.27,
      "grad_norm": 4.654965400695801,
      "learning_rate": 1.2915254237288137e-05,
      "loss": 1.399,
      "step": 10950
    },
    {
      "epoch": 0.27,
      "grad_norm": 6.813911437988281,
      "learning_rate": 1.2908474576271189e-05,
      "loss": 1.5587,
      "step": 10960
    },
    {
      "epoch": 0.27,
      "grad_norm": 3.309264659881592,
      "learning_rate": 1.2901694915254239e-05,
      "loss": 1.2123,
      "step": 10970
    },
    {
      "epoch": 0.27,
      "grad_norm": 3.3811533451080322,
      "learning_rate": 1.289491525423729e-05,
      "loss": 1.4726,
      "step": 10980
    },
    {
      "epoch": 0.27,
      "grad_norm": 4.549508571624756,
      "learning_rate": 1.2888135593220341e-05,
      "loss": 1.4868,
      "step": 10990
    },
    {
      "epoch": 0.28,
      "grad_norm": 3.4169394969940186,
      "learning_rate": 1.288135593220339e-05,
      "loss": 1.6193,
      "step": 11000
    },
    {
      "epoch": 0.28,
      "grad_norm": 10.524312973022461,
      "learning_rate": 1.287457627118644e-05,
      "loss": 1.5691,
      "step": 11010
    },
    {
      "epoch": 0.28,
      "grad_norm": 6.23699951171875,
      "learning_rate": 1.2867796610169492e-05,
      "loss": 1.479,
      "step": 11020
    },
    {
      "epoch": 0.28,
      "grad_norm": 2.1807405948638916,
      "learning_rate": 1.2861016949152542e-05,
      "loss": 1.4947,
      "step": 11030
    },
    {
      "epoch": 0.28,
      "grad_norm": 5.321918487548828,
      "learning_rate": 1.2854237288135593e-05,
      "loss": 1.6062,
      "step": 11040
    },
    {
      "epoch": 0.28,
      "grad_norm": 6.338499546051025,
      "learning_rate": 1.2847457627118645e-05,
      "loss": 1.4442,
      "step": 11050
    },
    {
      "epoch": 0.28,
      "grad_norm": 2.1677632331848145,
      "learning_rate": 1.2840677966101696e-05,
      "loss": 1.4006,
      "step": 11060
    },
    {
      "epoch": 0.28,
      "grad_norm": 4.229844093322754,
      "learning_rate": 1.2833898305084746e-05,
      "loss": 1.3553,
      "step": 11070
    },
    {
      "epoch": 0.28,
      "grad_norm": 2.8787291049957275,
      "learning_rate": 1.2827118644067797e-05,
      "loss": 1.6257,
      "step": 11080
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.544262170791626,
      "learning_rate": 1.2820338983050849e-05,
      "loss": 1.4121,
      "step": 11090
    },
    {
      "epoch": 0.28,
      "grad_norm": 2.342297315597534,
      "learning_rate": 1.28135593220339e-05,
      "loss": 1.6108,
      "step": 11100
    },
    {
      "epoch": 0.28,
      "grad_norm": 2.0289993286132812,
      "learning_rate": 1.280677966101695e-05,
      "loss": 1.3715,
      "step": 11110
    },
    {
      "epoch": 0.28,
      "grad_norm": 6.402794361114502,
      "learning_rate": 1.2800000000000001e-05,
      "loss": 1.3021,
      "step": 11120
    },
    {
      "epoch": 0.28,
      "grad_norm": 7.015219211578369,
      "learning_rate": 1.2793220338983053e-05,
      "loss": 1.5143,
      "step": 11130
    },
    {
      "epoch": 0.28,
      "grad_norm": 5.17816162109375,
      "learning_rate": 1.2786440677966104e-05,
      "loss": 1.2957,
      "step": 11140
    },
    {
      "epoch": 0.28,
      "grad_norm": 3.089566230773926,
      "learning_rate": 1.2779661016949154e-05,
      "loss": 1.339,
      "step": 11150
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.9675359725952148,
      "learning_rate": 1.2772881355932205e-05,
      "loss": 1.293,
      "step": 11160
    },
    {
      "epoch": 0.28,
      "grad_norm": 4.397621154785156,
      "learning_rate": 1.2766101694915256e-05,
      "loss": 1.2851,
      "step": 11170
    },
    {
      "epoch": 0.28,
      "grad_norm": 8.32104778289795,
      "learning_rate": 1.2759322033898308e-05,
      "loss": 1.555,
      "step": 11180
    },
    {
      "epoch": 0.28,
      "grad_norm": 2.426379442214966,
      "learning_rate": 1.2752542372881358e-05,
      "loss": 1.2762,
      "step": 11190
    },
    {
      "epoch": 0.28,
      "grad_norm": 5.169369220733643,
      "learning_rate": 1.2745762711864407e-05,
      "loss": 1.4836,
      "step": 11200
    },
    {
      "epoch": 0.28,
      "grad_norm": 2.9541831016540527,
      "learning_rate": 1.2738983050847457e-05,
      "loss": 1.4324,
      "step": 11210
    },
    {
      "epoch": 0.28,
      "grad_norm": 3.389583110809326,
      "learning_rate": 1.2732203389830508e-05,
      "loss": 1.3144,
      "step": 11220
    },
    {
      "epoch": 0.28,
      "grad_norm": 3.6163461208343506,
      "learning_rate": 1.272542372881356e-05,
      "loss": 1.5519,
      "step": 11230
    },
    {
      "epoch": 0.28,
      "grad_norm": 5.6586012840271,
      "learning_rate": 1.2718644067796611e-05,
      "loss": 1.5043,
      "step": 11240
    },
    {
      "epoch": 0.28,
      "grad_norm": 5.7041144371032715,
      "learning_rate": 1.2711864406779661e-05,
      "loss": 1.3471,
      "step": 11250
    },
    {
      "epoch": 0.28,
      "grad_norm": 2.085888624191284,
      "learning_rate": 1.2705084745762712e-05,
      "loss": 1.4054,
      "step": 11260
    },
    {
      "epoch": 0.28,
      "grad_norm": 5.836471080780029,
      "learning_rate": 1.2698305084745764e-05,
      "loss": 1.6426,
      "step": 11270
    },
    {
      "epoch": 0.28,
      "grad_norm": 6.930868625640869,
      "learning_rate": 1.2691525423728815e-05,
      "loss": 1.3756,
      "step": 11280
    },
    {
      "epoch": 0.28,
      "grad_norm": 4.569734573364258,
      "learning_rate": 1.2684745762711865e-05,
      "loss": 1.3019,
      "step": 11290
    },
    {
      "epoch": 0.28,
      "grad_norm": 2.0089516639709473,
      "learning_rate": 1.2677966101694916e-05,
      "loss": 1.3678,
      "step": 11300
    },
    {
      "epoch": 0.28,
      "grad_norm": 2.9268321990966797,
      "learning_rate": 1.2671186440677968e-05,
      "loss": 1.3469,
      "step": 11310
    },
    {
      "epoch": 0.28,
      "grad_norm": 2.9947781562805176,
      "learning_rate": 1.2664406779661019e-05,
      "loss": 1.3904,
      "step": 11320
    },
    {
      "epoch": 0.28,
      "grad_norm": 4.66842794418335,
      "learning_rate": 1.2657627118644069e-05,
      "loss": 1.4373,
      "step": 11330
    },
    {
      "epoch": 0.28,
      "grad_norm": 3.0476136207580566,
      "learning_rate": 1.265084745762712e-05,
      "loss": 1.4361,
      "step": 11340
    },
    {
      "epoch": 0.28,
      "grad_norm": 4.6397175788879395,
      "learning_rate": 1.2644067796610171e-05,
      "loss": 1.3457,
      "step": 11350
    },
    {
      "epoch": 0.28,
      "grad_norm": 3.0197949409484863,
      "learning_rate": 1.2637288135593221e-05,
      "loss": 1.3823,
      "step": 11360
    },
    {
      "epoch": 0.28,
      "grad_norm": 6.175907135009766,
      "learning_rate": 1.2630508474576273e-05,
      "loss": 1.4754,
      "step": 11370
    },
    {
      "epoch": 0.28,
      "grad_norm": 5.634079933166504,
      "learning_rate": 1.2623728813559324e-05,
      "loss": 1.4998,
      "step": 11380
    },
    {
      "epoch": 0.28,
      "grad_norm": 6.4973039627075195,
      "learning_rate": 1.2616949152542375e-05,
      "loss": 1.3785,
      "step": 11390
    },
    {
      "epoch": 0.28,
      "grad_norm": 3.8919670581817627,
      "learning_rate": 1.2610169491525425e-05,
      "loss": 1.2079,
      "step": 11400
    },
    {
      "epoch": 0.29,
      "grad_norm": 4.296518325805664,
      "learning_rate": 1.2603389830508475e-05,
      "loss": 1.5827,
      "step": 11410
    },
    {
      "epoch": 0.29,
      "grad_norm": 7.156797409057617,
      "learning_rate": 1.2596610169491526e-05,
      "loss": 1.4112,
      "step": 11420
    },
    {
      "epoch": 0.29,
      "grad_norm": 7.084222793579102,
      "learning_rate": 1.2589830508474576e-05,
      "loss": 1.4342,
      "step": 11430
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.8836684226989746,
      "learning_rate": 1.2583050847457627e-05,
      "loss": 1.3912,
      "step": 11440
    },
    {
      "epoch": 0.29,
      "grad_norm": 5.910703659057617,
      "learning_rate": 1.2576271186440679e-05,
      "loss": 1.3993,
      "step": 11450
    },
    {
      "epoch": 0.29,
      "grad_norm": 2.0015511512756348,
      "learning_rate": 1.256949152542373e-05,
      "loss": 1.304,
      "step": 11460
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.927198052406311,
      "learning_rate": 1.256271186440678e-05,
      "loss": 1.4272,
      "step": 11470
    },
    {
      "epoch": 0.29,
      "grad_norm": 4.085020542144775,
      "learning_rate": 1.2555932203389831e-05,
      "loss": 1.421,
      "step": 11480
    },
    {
      "epoch": 0.29,
      "grad_norm": 4.327631950378418,
      "learning_rate": 1.2549152542372883e-05,
      "loss": 1.3904,
      "step": 11490
    },
    {
      "epoch": 0.29,
      "grad_norm": 7.12744665145874,
      "learning_rate": 1.2542372881355932e-05,
      "loss": 1.4889,
      "step": 11500
    },
    {
      "epoch": 0.29,
      "grad_norm": 3.7715964317321777,
      "learning_rate": 1.2535593220338984e-05,
      "loss": 1.4941,
      "step": 11510
    },
    {
      "epoch": 0.29,
      "grad_norm": 3.0724148750305176,
      "learning_rate": 1.2528813559322035e-05,
      "loss": 1.3117,
      "step": 11520
    },
    {
      "epoch": 0.29,
      "grad_norm": 2.8994596004486084,
      "learning_rate": 1.2522033898305087e-05,
      "loss": 1.4414,
      "step": 11530
    },
    {
      "epoch": 0.29,
      "grad_norm": 4.751483917236328,
      "learning_rate": 1.2515254237288136e-05,
      "loss": 1.3516,
      "step": 11540
    },
    {
      "epoch": 0.29,
      "grad_norm": 5.237547874450684,
      "learning_rate": 1.2508474576271188e-05,
      "loss": 1.2714,
      "step": 11550
    },
    {
      "epoch": 0.29,
      "grad_norm": 2.0692265033721924,
      "learning_rate": 1.2501694915254239e-05,
      "loss": 1.5128,
      "step": 11560
    },
    {
      "epoch": 0.29,
      "grad_norm": 3.6956844329833984,
      "learning_rate": 1.249491525423729e-05,
      "loss": 1.5017,
      "step": 11570
    },
    {
      "epoch": 0.29,
      "grad_norm": 3.947849750518799,
      "learning_rate": 1.248813559322034e-05,
      "loss": 1.3873,
      "step": 11580
    },
    {
      "epoch": 0.29,
      "grad_norm": 3.6663479804992676,
      "learning_rate": 1.2481355932203392e-05,
      "loss": 1.3116,
      "step": 11590
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.384424090385437,
      "learning_rate": 1.2474576271186443e-05,
      "loss": 1.3077,
      "step": 11600
    },
    {
      "epoch": 0.29,
      "grad_norm": 2.7401161193847656,
      "learning_rate": 1.2467796610169494e-05,
      "loss": 1.3776,
      "step": 11610
    },
    {
      "epoch": 0.29,
      "grad_norm": 2.758903980255127,
      "learning_rate": 1.2461016949152542e-05,
      "loss": 1.3805,
      "step": 11620
    },
    {
      "epoch": 0.29,
      "grad_norm": 3.3219079971313477,
      "learning_rate": 1.2454237288135594e-05,
      "loss": 1.4713,
      "step": 11630
    },
    {
      "epoch": 0.29,
      "grad_norm": 4.310435771942139,
      "learning_rate": 1.2447457627118643e-05,
      "loss": 1.3415,
      "step": 11640
    },
    {
      "epoch": 0.29,
      "grad_norm": 7.007094860076904,
      "learning_rate": 1.2440677966101695e-05,
      "loss": 1.2731,
      "step": 11650
    },
    {
      "epoch": 0.29,
      "grad_norm": 3.9648146629333496,
      "learning_rate": 1.2433898305084746e-05,
      "loss": 1.4948,
      "step": 11660
    },
    {
      "epoch": 0.29,
      "grad_norm": 3.944697380065918,
      "learning_rate": 1.2427118644067798e-05,
      "loss": 1.4037,
      "step": 11670
    },
    {
      "epoch": 0.29,
      "grad_norm": 3.199932098388672,
      "learning_rate": 1.2420338983050847e-05,
      "loss": 1.3689,
      "step": 11680
    },
    {
      "epoch": 0.29,
      "grad_norm": 2.170687675476074,
      "learning_rate": 1.2413559322033899e-05,
      "loss": 1.4211,
      "step": 11690
    },
    {
      "epoch": 0.29,
      "grad_norm": 2.944251775741577,
      "learning_rate": 1.240677966101695e-05,
      "loss": 1.4267,
      "step": 11700
    },
    {
      "epoch": 0.29,
      "grad_norm": 3.6329870223999023,
      "learning_rate": 1.2400000000000002e-05,
      "loss": 1.41,
      "step": 11710
    },
    {
      "epoch": 0.29,
      "grad_norm": 3.745835304260254,
      "learning_rate": 1.2393220338983051e-05,
      "loss": 1.5216,
      "step": 11720
    },
    {
      "epoch": 0.29,
      "grad_norm": 3.793372869491577,
      "learning_rate": 1.2386440677966103e-05,
      "loss": 1.2819,
      "step": 11730
    },
    {
      "epoch": 0.29,
      "grad_norm": 4.974847316741943,
      "learning_rate": 1.2379661016949154e-05,
      "loss": 1.3499,
      "step": 11740
    },
    {
      "epoch": 0.29,
      "grad_norm": 6.320126533508301,
      "learning_rate": 1.2372881355932205e-05,
      "loss": 1.4321,
      "step": 11750
    },
    {
      "epoch": 0.29,
      "grad_norm": 3.2078051567077637,
      "learning_rate": 1.2366101694915255e-05,
      "loss": 1.5682,
      "step": 11760
    },
    {
      "epoch": 0.29,
      "grad_norm": 8.024663925170898,
      "learning_rate": 1.2359322033898307e-05,
      "loss": 1.6118,
      "step": 11770
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.9363720417022705,
      "learning_rate": 1.2352542372881358e-05,
      "loss": 1.5759,
      "step": 11780
    },
    {
      "epoch": 0.29,
      "grad_norm": 3.957169771194458,
      "learning_rate": 1.234576271186441e-05,
      "loss": 1.482,
      "step": 11790
    },
    {
      "epoch": 0.29,
      "grad_norm": 2.8198018074035645,
      "learning_rate": 1.2338983050847459e-05,
      "loss": 1.2723,
      "step": 11800
    },
    {
      "epoch": 0.3,
      "grad_norm": 3.2750072479248047,
      "learning_rate": 1.233220338983051e-05,
      "loss": 1.3456,
      "step": 11810
    },
    {
      "epoch": 0.3,
      "grad_norm": 2.286834478378296,
      "learning_rate": 1.2325423728813562e-05,
      "loss": 1.402,
      "step": 11820
    },
    {
      "epoch": 0.3,
      "grad_norm": 2.499225616455078,
      "learning_rate": 1.231864406779661e-05,
      "loss": 1.6833,
      "step": 11830
    },
    {
      "epoch": 0.3,
      "grad_norm": 2.649601697921753,
      "learning_rate": 1.2311864406779661e-05,
      "loss": 1.4015,
      "step": 11840
    },
    {
      "epoch": 0.3,
      "grad_norm": 4.886670112609863,
      "learning_rate": 1.2305084745762713e-05,
      "loss": 1.4322,
      "step": 11850
    },
    {
      "epoch": 0.3,
      "grad_norm": 2.639349937438965,
      "learning_rate": 1.2298305084745762e-05,
      "loss": 1.3,
      "step": 11860
    },
    {
      "epoch": 0.3,
      "grad_norm": 3.6703052520751953,
      "learning_rate": 1.2291525423728814e-05,
      "loss": 1.581,
      "step": 11870
    },
    {
      "epoch": 0.3,
      "grad_norm": 2.8534090518951416,
      "learning_rate": 1.2284745762711865e-05,
      "loss": 1.2386,
      "step": 11880
    },
    {
      "epoch": 0.3,
      "grad_norm": 5.143392086029053,
      "learning_rate": 1.2277966101694917e-05,
      "loss": 1.3647,
      "step": 11890
    },
    {
      "epoch": 0.3,
      "grad_norm": 7.419843673706055,
      "learning_rate": 1.2271186440677966e-05,
      "loss": 1.438,
      "step": 11900
    },
    {
      "epoch": 0.3,
      "grad_norm": 5.128147125244141,
      "learning_rate": 1.2264406779661018e-05,
      "loss": 1.6595,
      "step": 11910
    },
    {
      "epoch": 0.3,
      "grad_norm": 8.509882926940918,
      "learning_rate": 1.2257627118644069e-05,
      "loss": 1.269,
      "step": 11920
    },
    {
      "epoch": 0.3,
      "grad_norm": 2.279961585998535,
      "learning_rate": 1.225084745762712e-05,
      "loss": 1.5161,
      "step": 11930
    },
    {
      "epoch": 0.3,
      "grad_norm": 7.543710708618164,
      "learning_rate": 1.224406779661017e-05,
      "loss": 1.532,
      "step": 11940
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.636099100112915,
      "learning_rate": 1.2237288135593222e-05,
      "loss": 1.5582,
      "step": 11950
    },
    {
      "epoch": 0.3,
      "grad_norm": 2.6481144428253174,
      "learning_rate": 1.2230508474576273e-05,
      "loss": 1.3488,
      "step": 11960
    },
    {
      "epoch": 0.3,
      "grad_norm": 10.184081077575684,
      "learning_rate": 1.2223728813559323e-05,
      "loss": 1.4207,
      "step": 11970
    },
    {
      "epoch": 0.3,
      "grad_norm": 2.9163167476654053,
      "learning_rate": 1.2216949152542374e-05,
      "loss": 1.34,
      "step": 11980
    },
    {
      "epoch": 0.3,
      "grad_norm": 3.474348306655884,
      "learning_rate": 1.2210169491525426e-05,
      "loss": 1.2853,
      "step": 11990
    },
    {
      "epoch": 0.3,
      "grad_norm": 3.3969898223876953,
      "learning_rate": 1.2203389830508477e-05,
      "loss": 1.531,
      "step": 12000
    },
    {
      "epoch": 0.3,
      "grad_norm": 4.142295837402344,
      "learning_rate": 1.2196610169491527e-05,
      "loss": 1.4381,
      "step": 12010
    },
    {
      "epoch": 0.3,
      "grad_norm": 2.671096086502075,
      "learning_rate": 1.2189830508474578e-05,
      "loss": 1.3529,
      "step": 12020
    },
    {
      "epoch": 0.3,
      "grad_norm": 4.420811176300049,
      "learning_rate": 1.2183050847457628e-05,
      "loss": 1.3238,
      "step": 12030
    },
    {
      "epoch": 0.3,
      "grad_norm": 4.227188587188721,
      "learning_rate": 1.2176271186440677e-05,
      "loss": 1.4057,
      "step": 12040
    },
    {
      "epoch": 0.3,
      "grad_norm": 7.683736801147461,
      "learning_rate": 1.2169491525423729e-05,
      "loss": 1.3622,
      "step": 12050
    },
    {
      "epoch": 0.3,
      "grad_norm": 7.022848129272461,
      "learning_rate": 1.216271186440678e-05,
      "loss": 1.3424,
      "step": 12060
    },
    {
      "epoch": 0.3,
      "grad_norm": 5.885775089263916,
      "learning_rate": 1.2155932203389832e-05,
      "loss": 1.4362,
      "step": 12070
    },
    {
      "epoch": 0.3,
      "grad_norm": 7.27533483505249,
      "learning_rate": 1.2149152542372881e-05,
      "loss": 1.4683,
      "step": 12080
    },
    {
      "epoch": 0.3,
      "grad_norm": 2.617831230163574,
      "learning_rate": 1.2142372881355933e-05,
      "loss": 1.4669,
      "step": 12090
    },
    {
      "epoch": 0.3,
      "grad_norm": 5.099781036376953,
      "learning_rate": 1.2135593220338984e-05,
      "loss": 1.3095,
      "step": 12100
    },
    {
      "epoch": 0.3,
      "grad_norm": 8.695642471313477,
      "learning_rate": 1.2128813559322034e-05,
      "loss": 1.3228,
      "step": 12110
    },
    {
      "epoch": 0.3,
      "grad_norm": 3.8995704650878906,
      "learning_rate": 1.2122033898305085e-05,
      "loss": 1.539,
      "step": 12120
    },
    {
      "epoch": 0.3,
      "grad_norm": 3.012927770614624,
      "learning_rate": 1.2115254237288137e-05,
      "loss": 1.3339,
      "step": 12130
    },
    {
      "epoch": 0.3,
      "grad_norm": 2.5025315284729004,
      "learning_rate": 1.2108474576271188e-05,
      "loss": 1.3498,
      "step": 12140
    },
    {
      "epoch": 0.3,
      "grad_norm": 8.83749771118164,
      "learning_rate": 1.2101694915254238e-05,
      "loss": 1.2714,
      "step": 12150
    },
    {
      "epoch": 0.3,
      "grad_norm": 8.859834671020508,
      "learning_rate": 1.209491525423729e-05,
      "loss": 1.4884,
      "step": 12160
    },
    {
      "epoch": 0.3,
      "grad_norm": 4.210524559020996,
      "learning_rate": 1.208813559322034e-05,
      "loss": 1.4433,
      "step": 12170
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.851257562637329,
      "learning_rate": 1.2081355932203392e-05,
      "loss": 1.2866,
      "step": 12180
    },
    {
      "epoch": 0.3,
      "grad_norm": 8.631868362426758,
      "learning_rate": 1.2074576271186442e-05,
      "loss": 1.2828,
      "step": 12190
    },
    {
      "epoch": 0.3,
      "grad_norm": 3.778810739517212,
      "learning_rate": 1.2067796610169493e-05,
      "loss": 1.3999,
      "step": 12200
    },
    {
      "epoch": 0.31,
      "grad_norm": 4.637775421142578,
      "learning_rate": 1.2061016949152544e-05,
      "loss": 1.6437,
      "step": 12210
    },
    {
      "epoch": 0.31,
      "grad_norm": 5.261911392211914,
      "learning_rate": 1.2054237288135596e-05,
      "loss": 1.4595,
      "step": 12220
    },
    {
      "epoch": 0.31,
      "grad_norm": 2.9827585220336914,
      "learning_rate": 1.2047457627118646e-05,
      "loss": 1.2828,
      "step": 12230
    },
    {
      "epoch": 0.31,
      "grad_norm": 3.5697154998779297,
      "learning_rate": 1.2040677966101695e-05,
      "loss": 1.3189,
      "step": 12240
    },
    {
      "epoch": 0.31,
      "grad_norm": 2.249147891998291,
      "learning_rate": 1.2033898305084745e-05,
      "loss": 1.4488,
      "step": 12250
    },
    {
      "epoch": 0.31,
      "grad_norm": 14.768622398376465,
      "learning_rate": 1.2027118644067796e-05,
      "loss": 1.3953,
      "step": 12260
    },
    {
      "epoch": 0.31,
      "grad_norm": 8.961536407470703,
      "learning_rate": 1.2020338983050848e-05,
      "loss": 1.4439,
      "step": 12270
    },
    {
      "epoch": 0.31,
      "grad_norm": 2.7521257400512695,
      "learning_rate": 1.20135593220339e-05,
      "loss": 1.4273,
      "step": 12280
    },
    {
      "epoch": 0.31,
      "grad_norm": 2.267268419265747,
      "learning_rate": 1.2006779661016949e-05,
      "loss": 1.4253,
      "step": 12290
    },
    {
      "epoch": 0.31,
      "grad_norm": 2.8530664443969727,
      "learning_rate": 1.2e-05,
      "loss": 1.46,
      "step": 12300
    },
    {
      "epoch": 0.31,
      "grad_norm": 6.780263423919678,
      "learning_rate": 1.1993220338983052e-05,
      "loss": 1.4492,
      "step": 12310
    },
    {
      "epoch": 0.31,
      "grad_norm": 3.0507030487060547,
      "learning_rate": 1.1986440677966103e-05,
      "loss": 1.3701,
      "step": 12320
    },
    {
      "epoch": 0.31,
      "grad_norm": 9.925793647766113,
      "learning_rate": 1.1979661016949153e-05,
      "loss": 1.4798,
      "step": 12330
    },
    {
      "epoch": 0.31,
      "grad_norm": 3.48189377784729,
      "learning_rate": 1.1972881355932204e-05,
      "loss": 1.4494,
      "step": 12340
    },
    {
      "epoch": 0.31,
      "grad_norm": 4.056149005889893,
      "learning_rate": 1.1966101694915256e-05,
      "loss": 1.5467,
      "step": 12350
    },
    {
      "epoch": 0.31,
      "grad_norm": 4.1265692710876465,
      "learning_rate": 1.1959322033898307e-05,
      "loss": 1.3824,
      "step": 12360
    },
    {
      "epoch": 0.31,
      "grad_norm": 2.876166582107544,
      "learning_rate": 1.1952542372881357e-05,
      "loss": 1.7261,
      "step": 12370
    },
    {
      "epoch": 0.31,
      "grad_norm": 3.7193408012390137,
      "learning_rate": 1.1945762711864408e-05,
      "loss": 1.5968,
      "step": 12380
    },
    {
      "epoch": 0.31,
      "grad_norm": 7.140708923339844,
      "learning_rate": 1.193898305084746e-05,
      "loss": 1.3799,
      "step": 12390
    },
    {
      "epoch": 0.31,
      "grad_norm": 4.3301920890808105,
      "learning_rate": 1.1932203389830511e-05,
      "loss": 1.4771,
      "step": 12400
    },
    {
      "epoch": 0.31,
      "grad_norm": 5.242763996124268,
      "learning_rate": 1.192542372881356e-05,
      "loss": 1.4309,
      "step": 12410
    },
    {
      "epoch": 0.31,
      "grad_norm": 5.091261386871338,
      "learning_rate": 1.1918644067796612e-05,
      "loss": 1.4398,
      "step": 12420
    },
    {
      "epoch": 0.31,
      "grad_norm": 3.501070737838745,
      "learning_rate": 1.1911864406779663e-05,
      "loss": 1.2981,
      "step": 12430
    },
    {
      "epoch": 0.31,
      "grad_norm": 3.7128777503967285,
      "learning_rate": 1.1905084745762713e-05,
      "loss": 1.3998,
      "step": 12440
    },
    {
      "epoch": 0.31,
      "grad_norm": 3.6176857948303223,
      "learning_rate": 1.1898305084745763e-05,
      "loss": 1.5625,
      "step": 12450
    },
    {
      "epoch": 0.31,
      "grad_norm": 9.238510131835938,
      "learning_rate": 1.1891525423728814e-05,
      "loss": 1.4737,
      "step": 12460
    },
    {
      "epoch": 0.31,
      "grad_norm": 7.006262302398682,
      "learning_rate": 1.1884745762711864e-05,
      "loss": 1.3525,
      "step": 12470
    },
    {
      "epoch": 0.31,
      "grad_norm": 3.2028143405914307,
      "learning_rate": 1.1877966101694915e-05,
      "loss": 1.3041,
      "step": 12480
    },
    {
      "epoch": 0.31,
      "grad_norm": 6.272420883178711,
      "learning_rate": 1.1871186440677967e-05,
      "loss": 1.3837,
      "step": 12490
    },
    {
      "epoch": 0.31,
      "grad_norm": 2.0181572437286377,
      "learning_rate": 1.1864406779661018e-05,
      "loss": 1.2586,
      "step": 12500
    },
    {
      "epoch": 0.31,
      "eval_loss": 1.41282320022583,
      "eval_runtime": 122.1863,
      "eval_samples_per_second": 8.184,
      "eval_steps_per_second": 8.184,
      "step": 12500
    },
    {
      "epoch": 0.31,
      "grad_norm": 3.7258734703063965,
      "learning_rate": 1.1857627118644068e-05,
      "loss": 1.6044,
      "step": 12510
    },
    {
      "epoch": 0.31,
      "grad_norm": 2.6150224208831787,
      "learning_rate": 1.185084745762712e-05,
      "loss": 1.3926,
      "step": 12520
    },
    {
      "epoch": 0.31,
      "grad_norm": 4.713764667510986,
      "learning_rate": 1.184406779661017e-05,
      "loss": 1.4248,
      "step": 12530
    },
    {
      "epoch": 0.31,
      "grad_norm": 4.03065824508667,
      "learning_rate": 1.183728813559322e-05,
      "loss": 1.2203,
      "step": 12540
    },
    {
      "epoch": 0.31,
      "grad_norm": 7.120413303375244,
      "learning_rate": 1.1830508474576272e-05,
      "loss": 1.2521,
      "step": 12550
    },
    {
      "epoch": 0.31,
      "grad_norm": 7.475151062011719,
      "learning_rate": 1.1823728813559323e-05,
      "loss": 1.4073,
      "step": 12560
    },
    {
      "epoch": 0.31,
      "grad_norm": 7.4751715660095215,
      "learning_rate": 1.1816949152542375e-05,
      "loss": 1.4283,
      "step": 12570
    },
    {
      "epoch": 0.31,
      "grad_norm": 2.401461601257324,
      "learning_rate": 1.1810169491525424e-05,
      "loss": 1.3334,
      "step": 12580
    },
    {
      "epoch": 0.31,
      "grad_norm": 2.8532662391662598,
      "learning_rate": 1.1803389830508476e-05,
      "loss": 1.1893,
      "step": 12590
    },
    {
      "epoch": 0.32,
      "grad_norm": 6.916831970214844,
      "learning_rate": 1.1796610169491527e-05,
      "loss": 1.3783,
      "step": 12600
    },
    {
      "epoch": 0.32,
      "grad_norm": 3.3440134525299072,
      "learning_rate": 1.1789830508474578e-05,
      "loss": 1.4821,
      "step": 12610
    },
    {
      "epoch": 0.32,
      "grad_norm": 5.114617347717285,
      "learning_rate": 1.1783050847457628e-05,
      "loss": 1.3289,
      "step": 12620
    },
    {
      "epoch": 0.32,
      "grad_norm": 5.534918308258057,
      "learning_rate": 1.177627118644068e-05,
      "loss": 1.4112,
      "step": 12630
    },
    {
      "epoch": 0.32,
      "grad_norm": 6.668290615081787,
      "learning_rate": 1.1769491525423731e-05,
      "loss": 1.6591,
      "step": 12640
    },
    {
      "epoch": 0.32,
      "grad_norm": 5.515822410583496,
      "learning_rate": 1.1762711864406782e-05,
      "loss": 1.3174,
      "step": 12650
    },
    {
      "epoch": 0.32,
      "grad_norm": 4.382727146148682,
      "learning_rate": 1.175593220338983e-05,
      "loss": 1.4316,
      "step": 12660
    },
    {
      "epoch": 0.32,
      "grad_norm": 2.1675164699554443,
      "learning_rate": 1.1749152542372882e-05,
      "loss": 1.3716,
      "step": 12670
    },
    {
      "epoch": 0.32,
      "grad_norm": 3.56939697265625,
      "learning_rate": 1.1742372881355931e-05,
      "loss": 1.2994,
      "step": 12680
    },
    {
      "epoch": 0.32,
      "grad_norm": 2.873751401901245,
      "learning_rate": 1.1735593220338983e-05,
      "loss": 1.5063,
      "step": 12690
    },
    {
      "epoch": 0.32,
      "grad_norm": 5.886043071746826,
      "learning_rate": 1.1728813559322034e-05,
      "loss": 1.4374,
      "step": 12700
    },
    {
      "epoch": 0.32,
      "grad_norm": 4.226538181304932,
      "learning_rate": 1.1722033898305086e-05,
      "loss": 1.3556,
      "step": 12710
    },
    {
      "epoch": 0.32,
      "grad_norm": 8.882972717285156,
      "learning_rate": 1.1715254237288135e-05,
      "loss": 1.3683,
      "step": 12720
    },
    {
      "epoch": 0.32,
      "grad_norm": 9.191460609436035,
      "learning_rate": 1.1708474576271187e-05,
      "loss": 1.361,
      "step": 12730
    },
    {
      "epoch": 0.32,
      "grad_norm": 3.978660821914673,
      "learning_rate": 1.1701694915254238e-05,
      "loss": 1.4778,
      "step": 12740
    },
    {
      "epoch": 0.32,
      "grad_norm": 3.6160736083984375,
      "learning_rate": 1.169491525423729e-05,
      "loss": 1.4695,
      "step": 12750
    },
    {
      "epoch": 0.32,
      "grad_norm": 3.862130880355835,
      "learning_rate": 1.168813559322034e-05,
      "loss": 1.4542,
      "step": 12760
    },
    {
      "epoch": 0.32,
      "grad_norm": 4.235650062561035,
      "learning_rate": 1.168135593220339e-05,
      "loss": 1.3565,
      "step": 12770
    },
    {
      "epoch": 0.32,
      "grad_norm": 4.322425842285156,
      "learning_rate": 1.1674576271186442e-05,
      "loss": 1.5157,
      "step": 12780
    },
    {
      "epoch": 0.32,
      "grad_norm": 2.8006017208099365,
      "learning_rate": 1.1667796610169494e-05,
      "loss": 1.3883,
      "step": 12790
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.3275543451309204,
      "learning_rate": 1.1661016949152543e-05,
      "loss": 1.5993,
      "step": 12800
    },
    {
      "epoch": 0.32,
      "grad_norm": 8.304267883300781,
      "learning_rate": 1.1654237288135595e-05,
      "loss": 1.3,
      "step": 12810
    },
    {
      "epoch": 0.32,
      "grad_norm": 4.160703182220459,
      "learning_rate": 1.1647457627118646e-05,
      "loss": 1.2988,
      "step": 12820
    },
    {
      "epoch": 0.32,
      "grad_norm": 6.857654571533203,
      "learning_rate": 1.1640677966101697e-05,
      "loss": 1.3433,
      "step": 12830
    },
    {
      "epoch": 0.32,
      "grad_norm": 3.394209861755371,
      "learning_rate": 1.1633898305084747e-05,
      "loss": 1.4598,
      "step": 12840
    },
    {
      "epoch": 0.32,
      "grad_norm": 4.151514530181885,
      "learning_rate": 1.1627118644067799e-05,
      "loss": 1.386,
      "step": 12850
    },
    {
      "epoch": 0.32,
      "grad_norm": 3.8545305728912354,
      "learning_rate": 1.162033898305085e-05,
      "loss": 1.4558,
      "step": 12860
    },
    {
      "epoch": 0.32,
      "grad_norm": 4.819375514984131,
      "learning_rate": 1.1613559322033898e-05,
      "loss": 1.2603,
      "step": 12870
    },
    {
      "epoch": 0.32,
      "grad_norm": 4.475069999694824,
      "learning_rate": 1.160677966101695e-05,
      "loss": 1.433,
      "step": 12880
    },
    {
      "epoch": 0.32,
      "grad_norm": 11.210762977600098,
      "learning_rate": 1.16e-05,
      "loss": 1.4697,
      "step": 12890
    },
    {
      "epoch": 0.32,
      "grad_norm": 3.463872194290161,
      "learning_rate": 1.159322033898305e-05,
      "loss": 1.3284,
      "step": 12900
    },
    {
      "epoch": 0.32,
      "grad_norm": 2.506638765335083,
      "learning_rate": 1.1586440677966102e-05,
      "loss": 1.3739,
      "step": 12910
    },
    {
      "epoch": 0.32,
      "grad_norm": 4.121403694152832,
      "learning_rate": 1.1579661016949153e-05,
      "loss": 1.3458,
      "step": 12920
    },
    {
      "epoch": 0.32,
      "grad_norm": 5.18405818939209,
      "learning_rate": 1.1572881355932205e-05,
      "loss": 1.3417,
      "step": 12930
    },
    {
      "epoch": 0.32,
      "grad_norm": 2.7213258743286133,
      "learning_rate": 1.1566101694915254e-05,
      "loss": 1.1582,
      "step": 12940
    },
    {
      "epoch": 0.32,
      "grad_norm": 2.769882917404175,
      "learning_rate": 1.1559322033898306e-05,
      "loss": 1.5011,
      "step": 12950
    },
    {
      "epoch": 0.32,
      "grad_norm": 5.9535651206970215,
      "learning_rate": 1.1552542372881357e-05,
      "loss": 1.4084,
      "step": 12960
    },
    {
      "epoch": 0.32,
      "grad_norm": 5.238331317901611,
      "learning_rate": 1.1545762711864409e-05,
      "loss": 1.2475,
      "step": 12970
    },
    {
      "epoch": 0.32,
      "grad_norm": 2.8670809268951416,
      "learning_rate": 1.1538983050847458e-05,
      "loss": 1.3614,
      "step": 12980
    },
    {
      "epoch": 0.32,
      "grad_norm": 5.805054664611816,
      "learning_rate": 1.153220338983051e-05,
      "loss": 1.4508,
      "step": 12990
    },
    {
      "epoch": 0.33,
      "grad_norm": 3.9865105152130127,
      "learning_rate": 1.1525423728813561e-05,
      "loss": 1.408,
      "step": 13000
    },
    {
      "epoch": 0.33,
      "grad_norm": 5.843343734741211,
      "learning_rate": 1.151864406779661e-05,
      "loss": 1.6078,
      "step": 13010
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.3116165399551392,
      "learning_rate": 1.1511864406779662e-05,
      "loss": 1.4039,
      "step": 13020
    },
    {
      "epoch": 0.33,
      "grad_norm": 4.26644229888916,
      "learning_rate": 1.1505084745762714e-05,
      "loss": 1.431,
      "step": 13030
    },
    {
      "epoch": 0.33,
      "grad_norm": 5.766468048095703,
      "learning_rate": 1.1498305084745765e-05,
      "loss": 1.3031,
      "step": 13040
    },
    {
      "epoch": 0.33,
      "grad_norm": 2.0089643001556396,
      "learning_rate": 1.1491525423728815e-05,
      "loss": 1.3925,
      "step": 13050
    },
    {
      "epoch": 0.33,
      "grad_norm": 2.539196729660034,
      "learning_rate": 1.1484745762711866e-05,
      "loss": 1.476,
      "step": 13060
    },
    {
      "epoch": 0.33,
      "grad_norm": 2.942608118057251,
      "learning_rate": 1.1477966101694916e-05,
      "loss": 1.3719,
      "step": 13070
    },
    {
      "epoch": 0.33,
      "grad_norm": 7.528717517852783,
      "learning_rate": 1.1471186440677965e-05,
      "loss": 1.4201,
      "step": 13080
    },
    {
      "epoch": 0.33,
      "grad_norm": 6.850097179412842,
      "learning_rate": 1.1464406779661017e-05,
      "loss": 1.3329,
      "step": 13090
    },
    {
      "epoch": 0.33,
      "grad_norm": 5.788763523101807,
      "learning_rate": 1.1457627118644068e-05,
      "loss": 1.3512,
      "step": 13100
    },
    {
      "epoch": 0.33,
      "grad_norm": 3.462327003479004,
      "learning_rate": 1.145084745762712e-05,
      "loss": 1.4148,
      "step": 13110
    },
    {
      "epoch": 0.33,
      "grad_norm": 4.319177150726318,
      "learning_rate": 1.144406779661017e-05,
      "loss": 1.393,
      "step": 13120
    },
    {
      "epoch": 0.33,
      "grad_norm": 3.8986408710479736,
      "learning_rate": 1.143728813559322e-05,
      "loss": 1.4604,
      "step": 13130
    },
    {
      "epoch": 0.33,
      "grad_norm": 7.318924427032471,
      "learning_rate": 1.1430508474576272e-05,
      "loss": 1.6399,
      "step": 13140
    },
    {
      "epoch": 0.33,
      "grad_norm": 5.47678804397583,
      "learning_rate": 1.1423728813559322e-05,
      "loss": 1.3703,
      "step": 13150
    },
    {
      "epoch": 0.33,
      "grad_norm": 3.4975647926330566,
      "learning_rate": 1.1416949152542373e-05,
      "loss": 1.3331,
      "step": 13160
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.8587815761566162,
      "learning_rate": 1.1410169491525425e-05,
      "loss": 1.5117,
      "step": 13170
    },
    {
      "epoch": 0.33,
      "grad_norm": 11.561251640319824,
      "learning_rate": 1.1403389830508476e-05,
      "loss": 1.4811,
      "step": 13180
    },
    {
      "epoch": 0.33,
      "grad_norm": 4.3838348388671875,
      "learning_rate": 1.1396610169491526e-05,
      "loss": 1.4262,
      "step": 13190
    },
    {
      "epoch": 0.33,
      "grad_norm": 3.973987340927124,
      "learning_rate": 1.1389830508474577e-05,
      "loss": 1.3881,
      "step": 13200
    },
    {
      "epoch": 0.33,
      "grad_norm": 3.3548011779785156,
      "learning_rate": 1.1383050847457629e-05,
      "loss": 1.5162,
      "step": 13210
    },
    {
      "epoch": 0.33,
      "grad_norm": 10.37120246887207,
      "learning_rate": 1.137627118644068e-05,
      "loss": 1.453,
      "step": 13220
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.9512081146240234,
      "learning_rate": 1.136949152542373e-05,
      "loss": 1.2927,
      "step": 13230
    },
    {
      "epoch": 0.33,
      "grad_norm": 3.3641445636749268,
      "learning_rate": 1.1362711864406781e-05,
      "loss": 1.3378,
      "step": 13240
    },
    {
      "epoch": 0.33,
      "grad_norm": 5.190877437591553,
      "learning_rate": 1.1355932203389833e-05,
      "loss": 1.4249,
      "step": 13250
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.8611725568771362,
      "learning_rate": 1.1349152542372884e-05,
      "loss": 1.4188,
      "step": 13260
    },
    {
      "epoch": 0.33,
      "grad_norm": 3.0206570625305176,
      "learning_rate": 1.1342372881355934e-05,
      "loss": 1.3686,
      "step": 13270
    },
    {
      "epoch": 0.33,
      "grad_norm": 5.976974010467529,
      "learning_rate": 1.1335593220338983e-05,
      "loss": 1.566,
      "step": 13280
    },
    {
      "epoch": 0.33,
      "grad_norm": 7.103373050689697,
      "learning_rate": 1.1328813559322033e-05,
      "loss": 1.5635,
      "step": 13290
    },
    {
      "epoch": 0.33,
      "grad_norm": 5.803537845611572,
      "learning_rate": 1.1322033898305084e-05,
      "loss": 1.2443,
      "step": 13300
    },
    {
      "epoch": 0.33,
      "grad_norm": 4.929863929748535,
      "learning_rate": 1.1315254237288136e-05,
      "loss": 1.4459,
      "step": 13310
    },
    {
      "epoch": 0.33,
      "grad_norm": 2.3878164291381836,
      "learning_rate": 1.1308474576271187e-05,
      "loss": 1.362,
      "step": 13320
    },
    {
      "epoch": 0.33,
      "grad_norm": 3.999992609024048,
      "learning_rate": 1.1301694915254237e-05,
      "loss": 1.3284,
      "step": 13330
    },
    {
      "epoch": 0.33,
      "grad_norm": 3.5180516242980957,
      "learning_rate": 1.1294915254237288e-05,
      "loss": 1.4911,
      "step": 13340
    },
    {
      "epoch": 0.33,
      "grad_norm": 7.478114128112793,
      "learning_rate": 1.128813559322034e-05,
      "loss": 1.5124,
      "step": 13350
    },
    {
      "epoch": 0.33,
      "grad_norm": 2.4897103309631348,
      "learning_rate": 1.1281355932203391e-05,
      "loss": 1.3379,
      "step": 13360
    },
    {
      "epoch": 0.33,
      "grad_norm": 6.019432544708252,
      "learning_rate": 1.1274576271186441e-05,
      "loss": 1.4773,
      "step": 13370
    },
    {
      "epoch": 0.33,
      "grad_norm": 2.173727512359619,
      "learning_rate": 1.1267796610169492e-05,
      "loss": 1.4248,
      "step": 13380
    },
    {
      "epoch": 0.33,
      "grad_norm": 7.0194854736328125,
      "learning_rate": 1.1261016949152544e-05,
      "loss": 1.4419,
      "step": 13390
    },
    {
      "epoch": 0.34,
      "grad_norm": 3.5669920444488525,
      "learning_rate": 1.1254237288135595e-05,
      "loss": 1.518,
      "step": 13400
    },
    {
      "epoch": 0.34,
      "grad_norm": 5.0139689445495605,
      "learning_rate": 1.1247457627118645e-05,
      "loss": 1.3365,
      "step": 13410
    },
    {
      "epoch": 0.34,
      "grad_norm": 2.6222822666168213,
      "learning_rate": 1.1240677966101696e-05,
      "loss": 1.3748,
      "step": 13420
    },
    {
      "epoch": 0.34,
      "grad_norm": 4.982034683227539,
      "learning_rate": 1.1233898305084748e-05,
      "loss": 1.4444,
      "step": 13430
    },
    {
      "epoch": 0.34,
      "grad_norm": 5.214868068695068,
      "learning_rate": 1.1227118644067799e-05,
      "loss": 1.3249,
      "step": 13440
    },
    {
      "epoch": 0.34,
      "grad_norm": 3.7210988998413086,
      "learning_rate": 1.1220338983050849e-05,
      "loss": 1.1682,
      "step": 13450
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.919372797012329,
      "learning_rate": 1.12135593220339e-05,
      "loss": 1.4065,
      "step": 13460
    },
    {
      "epoch": 0.34,
      "grad_norm": 3.2649168968200684,
      "learning_rate": 1.1206779661016951e-05,
      "loss": 1.378,
      "step": 13470
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.508327603340149,
      "learning_rate": 1.1200000000000001e-05,
      "loss": 1.4103,
      "step": 13480
    },
    {
      "epoch": 0.34,
      "grad_norm": 3.5368258953094482,
      "learning_rate": 1.1193220338983051e-05,
      "loss": 1.409,
      "step": 13490
    },
    {
      "epoch": 0.34,
      "grad_norm": 6.36169958114624,
      "learning_rate": 1.1186440677966102e-05,
      "loss": 1.2165,
      "step": 13500
    },
    {
      "epoch": 0.34,
      "grad_norm": 4.591517925262451,
      "learning_rate": 1.1179661016949152e-05,
      "loss": 1.3166,
      "step": 13510
    },
    {
      "epoch": 0.34,
      "grad_norm": 3.9070346355438232,
      "learning_rate": 1.1172881355932203e-05,
      "loss": 1.5214,
      "step": 13520
    },
    {
      "epoch": 0.34,
      "grad_norm": 11.359418869018555,
      "learning_rate": 1.1166101694915255e-05,
      "loss": 1.3789,
      "step": 13530
    },
    {
      "epoch": 0.34,
      "grad_norm": 2.793360710144043,
      "learning_rate": 1.1159322033898306e-05,
      "loss": 1.3596,
      "step": 13540
    },
    {
      "epoch": 0.34,
      "grad_norm": 5.872134685516357,
      "learning_rate": 1.1152542372881356e-05,
      "loss": 1.3774,
      "step": 13550
    },
    {
      "epoch": 0.34,
      "grad_norm": 9.555739402770996,
      "learning_rate": 1.1145762711864407e-05,
      "loss": 1.4067,
      "step": 13560
    },
    {
      "epoch": 0.34,
      "grad_norm": 6.3838725090026855,
      "learning_rate": 1.1138983050847459e-05,
      "loss": 1.405,
      "step": 13570
    },
    {
      "epoch": 0.34,
      "grad_norm": 7.605794429779053,
      "learning_rate": 1.113220338983051e-05,
      "loss": 1.3688,
      "step": 13580
    },
    {
      "epoch": 0.34,
      "grad_norm": 4.44448184967041,
      "learning_rate": 1.112542372881356e-05,
      "loss": 1.5755,
      "step": 13590
    },
    {
      "epoch": 0.34,
      "grad_norm": 10.151824951171875,
      "learning_rate": 1.1118644067796611e-05,
      "loss": 1.2428,
      "step": 13600
    },
    {
      "epoch": 0.34,
      "grad_norm": 6.918173789978027,
      "learning_rate": 1.1111864406779663e-05,
      "loss": 1.1972,
      "step": 13610
    },
    {
      "epoch": 0.34,
      "grad_norm": 2.1551120281219482,
      "learning_rate": 1.1105084745762712e-05,
      "loss": 1.3142,
      "step": 13620
    },
    {
      "epoch": 0.34,
      "grad_norm": 10.687677383422852,
      "learning_rate": 1.1098305084745764e-05,
      "loss": 1.4632,
      "step": 13630
    },
    {
      "epoch": 0.34,
      "grad_norm": 4.853692531585693,
      "learning_rate": 1.1091525423728815e-05,
      "loss": 1.2948,
      "step": 13640
    },
    {
      "epoch": 0.34,
      "grad_norm": 4.251141548156738,
      "learning_rate": 1.1084745762711867e-05,
      "loss": 1.3907,
      "step": 13650
    },
    {
      "epoch": 0.34,
      "grad_norm": 5.989134788513184,
      "learning_rate": 1.1077966101694916e-05,
      "loss": 1.5161,
      "step": 13660
    },
    {
      "epoch": 0.34,
      "grad_norm": 4.6931471824646,
      "learning_rate": 1.1071186440677968e-05,
      "loss": 1.2562,
      "step": 13670
    },
    {
      "epoch": 0.34,
      "grad_norm": 7.449806213378906,
      "learning_rate": 1.1064406779661019e-05,
      "loss": 1.3159,
      "step": 13680
    },
    {
      "epoch": 0.34,
      "grad_norm": 6.833204746246338,
      "learning_rate": 1.105762711864407e-05,
      "loss": 1.4215,
      "step": 13690
    },
    {
      "epoch": 0.34,
      "grad_norm": 2.45497465133667,
      "learning_rate": 1.1050847457627118e-05,
      "loss": 1.5511,
      "step": 13700
    },
    {
      "epoch": 0.34,
      "grad_norm": 4.091574192047119,
      "learning_rate": 1.104406779661017e-05,
      "loss": 1.5469,
      "step": 13710
    },
    {
      "epoch": 0.34,
      "grad_norm": 2.3454174995422363,
      "learning_rate": 1.1037288135593221e-05,
      "loss": 1.4261,
      "step": 13720
    },
    {
      "epoch": 0.34,
      "grad_norm": 8.664597511291504,
      "learning_rate": 1.1030508474576271e-05,
      "loss": 1.5009,
      "step": 13730
    },
    {
      "epoch": 0.34,
      "grad_norm": 3.5096991062164307,
      "learning_rate": 1.1023728813559322e-05,
      "loss": 1.3663,
      "step": 13740
    },
    {
      "epoch": 0.34,
      "grad_norm": 11.86909294128418,
      "learning_rate": 1.1016949152542374e-05,
      "loss": 1.3403,
      "step": 13750
    },
    {
      "epoch": 0.34,
      "grad_norm": 4.446667671203613,
      "learning_rate": 1.1010169491525423e-05,
      "loss": 1.4076,
      "step": 13760
    },
    {
      "epoch": 0.34,
      "grad_norm": 4.6884636878967285,
      "learning_rate": 1.1003389830508475e-05,
      "loss": 1.5521,
      "step": 13770
    },
    {
      "epoch": 0.34,
      "grad_norm": 6.024641513824463,
      "learning_rate": 1.0996610169491526e-05,
      "loss": 1.451,
      "step": 13780
    },
    {
      "epoch": 0.34,
      "grad_norm": 7.03015661239624,
      "learning_rate": 1.0989830508474578e-05,
      "loss": 1.3397,
      "step": 13790
    },
    {
      "epoch": 0.34,
      "grad_norm": 2.7456510066986084,
      "learning_rate": 1.0983050847457627e-05,
      "loss": 1.1161,
      "step": 13800
    },
    {
      "epoch": 0.35,
      "grad_norm": 2.9939138889312744,
      "learning_rate": 1.0976271186440679e-05,
      "loss": 1.5048,
      "step": 13810
    },
    {
      "epoch": 0.35,
      "grad_norm": 5.112384796142578,
      "learning_rate": 1.096949152542373e-05,
      "loss": 1.3451,
      "step": 13820
    },
    {
      "epoch": 0.35,
      "grad_norm": 3.0827462673187256,
      "learning_rate": 1.0962711864406782e-05,
      "loss": 1.4395,
      "step": 13830
    },
    {
      "epoch": 0.35,
      "grad_norm": 2.1749086380004883,
      "learning_rate": 1.0955932203389831e-05,
      "loss": 1.5113,
      "step": 13840
    },
    {
      "epoch": 0.35,
      "grad_norm": 2.877530097961426,
      "learning_rate": 1.0949152542372883e-05,
      "loss": 1.3075,
      "step": 13850
    },
    {
      "epoch": 0.35,
      "grad_norm": 4.4052042961120605,
      "learning_rate": 1.0942372881355934e-05,
      "loss": 1.348,
      "step": 13860
    },
    {
      "epoch": 0.35,
      "grad_norm": 8.872933387756348,
      "learning_rate": 1.0935593220338985e-05,
      "loss": 1.459,
      "step": 13870
    },
    {
      "epoch": 0.35,
      "grad_norm": 3.286579132080078,
      "learning_rate": 1.0928813559322035e-05,
      "loss": 1.3108,
      "step": 13880
    },
    {
      "epoch": 0.35,
      "grad_norm": 11.492996215820312,
      "learning_rate": 1.0922033898305087e-05,
      "loss": 1.4302,
      "step": 13890
    },
    {
      "epoch": 0.35,
      "grad_norm": 3.9962315559387207,
      "learning_rate": 1.0915254237288135e-05,
      "loss": 1.246,
      "step": 13900
    },
    {
      "epoch": 0.35,
      "grad_norm": 3.57492995262146,
      "learning_rate": 1.0908474576271186e-05,
      "loss": 1.2576,
      "step": 13910
    },
    {
      "epoch": 0.35,
      "grad_norm": 4.833920478820801,
      "learning_rate": 1.0901694915254237e-05,
      "loss": 1.5454,
      "step": 13920
    },
    {
      "epoch": 0.35,
      "grad_norm": 2.4575819969177246,
      "learning_rate": 1.0894915254237289e-05,
      "loss": 1.1803,
      "step": 13930
    },
    {
      "epoch": 0.35,
      "grad_norm": 6.771847724914551,
      "learning_rate": 1.0888135593220339e-05,
      "loss": 1.381,
      "step": 13940
    },
    {
      "epoch": 0.35,
      "grad_norm": 6.094201564788818,
      "learning_rate": 1.088135593220339e-05,
      "loss": 1.3127,
      "step": 13950
    },
    {
      "epoch": 0.35,
      "grad_norm": 2.1938679218292236,
      "learning_rate": 1.0874576271186441e-05,
      "loss": 1.4033,
      "step": 13960
    },
    {
      "epoch": 0.35,
      "grad_norm": 7.476833343505859,
      "learning_rate": 1.0867796610169493e-05,
      "loss": 1.4817,
      "step": 13970
    },
    {
      "epoch": 0.35,
      "grad_norm": 5.379706382751465,
      "learning_rate": 1.0861016949152542e-05,
      "loss": 1.4558,
      "step": 13980
    },
    {
      "epoch": 0.35,
      "grad_norm": 4.60951042175293,
      "learning_rate": 1.0854237288135594e-05,
      "loss": 1.3056,
      "step": 13990
    },
    {
      "epoch": 0.35,
      "grad_norm": 3.469287395477295,
      "learning_rate": 1.0847457627118645e-05,
      "loss": 1.4666,
      "step": 14000
    },
    {
      "epoch": 0.35,
      "grad_norm": 6.565362930297852,
      "learning_rate": 1.0840677966101697e-05,
      "loss": 1.2819,
      "step": 14010
    },
    {
      "epoch": 0.35,
      "grad_norm": 4.5218281745910645,
      "learning_rate": 1.0833898305084746e-05,
      "loss": 1.4682,
      "step": 14020
    },
    {
      "epoch": 0.35,
      "grad_norm": 3.104768991470337,
      "learning_rate": 1.0827118644067798e-05,
      "loss": 1.521,
      "step": 14030
    },
    {
      "epoch": 0.35,
      "grad_norm": 4.510459899902344,
      "learning_rate": 1.0820338983050849e-05,
      "loss": 1.3285,
      "step": 14040
    },
    {
      "epoch": 0.35,
      "grad_norm": 3.237248659133911,
      "learning_rate": 1.08135593220339e-05,
      "loss": 1.3539,
      "step": 14050
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.2550790309906006,
      "learning_rate": 1.080677966101695e-05,
      "loss": 1.4968,
      "step": 14060
    },
    {
      "epoch": 0.35,
      "grad_norm": 4.4144487380981445,
      "learning_rate": 1.0800000000000002e-05,
      "loss": 1.3554,
      "step": 14070
    },
    {
      "epoch": 0.35,
      "grad_norm": 8.557780265808105,
      "learning_rate": 1.0793220338983053e-05,
      "loss": 1.2483,
      "step": 14080
    },
    {
      "epoch": 0.35,
      "grad_norm": 6.656067371368408,
      "learning_rate": 1.0786440677966103e-05,
      "loss": 1.3265,
      "step": 14090
    },
    {
      "epoch": 0.35,
      "grad_norm": 2.8041324615478516,
      "learning_rate": 1.0779661016949154e-05,
      "loss": 1.4114,
      "step": 14100
    },
    {
      "epoch": 0.35,
      "grad_norm": 4.193390369415283,
      "learning_rate": 1.0772881355932204e-05,
      "loss": 1.4974,
      "step": 14110
    },
    {
      "epoch": 0.35,
      "grad_norm": 3.4639577865600586,
      "learning_rate": 1.0766101694915254e-05,
      "loss": 1.4462,
      "step": 14120
    },
    {
      "epoch": 0.35,
      "grad_norm": 5.817545413970947,
      "learning_rate": 1.0759322033898305e-05,
      "loss": 1.3408,
      "step": 14130
    },
    {
      "epoch": 0.35,
      "grad_norm": 3.1863584518432617,
      "learning_rate": 1.0752542372881356e-05,
      "loss": 1.2525,
      "step": 14140
    },
    {
      "epoch": 0.35,
      "grad_norm": 4.928628444671631,
      "learning_rate": 1.0745762711864408e-05,
      "loss": 1.3993,
      "step": 14150
    },
    {
      "epoch": 0.35,
      "grad_norm": 8.653220176696777,
      "learning_rate": 1.0738983050847457e-05,
      "loss": 1.3923,
      "step": 14160
    },
    {
      "epoch": 0.35,
      "grad_norm": 2.3015573024749756,
      "learning_rate": 1.0732203389830509e-05,
      "loss": 1.4113,
      "step": 14170
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.8576148748397827,
      "learning_rate": 1.072542372881356e-05,
      "loss": 1.4169,
      "step": 14180
    },
    {
      "epoch": 0.35,
      "grad_norm": 3.376601219177246,
      "learning_rate": 1.0718644067796612e-05,
      "loss": 1.2976,
      "step": 14190
    },
    {
      "epoch": 0.35,
      "grad_norm": 5.912999153137207,
      "learning_rate": 1.0711864406779661e-05,
      "loss": 1.3656,
      "step": 14200
    },
    {
      "epoch": 0.36,
      "grad_norm": 5.299818992614746,
      "learning_rate": 1.0705084745762713e-05,
      "loss": 1.3591,
      "step": 14210
    },
    {
      "epoch": 0.36,
      "grad_norm": 4.026183605194092,
      "learning_rate": 1.0698305084745764e-05,
      "loss": 1.3621,
      "step": 14220
    },
    {
      "epoch": 0.36,
      "grad_norm": 7.459228992462158,
      "learning_rate": 1.0691525423728814e-05,
      "loss": 1.4314,
      "step": 14230
    },
    {
      "epoch": 0.36,
      "grad_norm": 3.479918956756592,
      "learning_rate": 1.0684745762711865e-05,
      "loss": 1.23,
      "step": 14240
    },
    {
      "epoch": 0.36,
      "grad_norm": 6.066028118133545,
      "learning_rate": 1.0677966101694917e-05,
      "loss": 1.3377,
      "step": 14250
    },
    {
      "epoch": 0.36,
      "grad_norm": 6.887609004974365,
      "learning_rate": 1.0671186440677968e-05,
      "loss": 1.2611,
      "step": 14260
    },
    {
      "epoch": 0.36,
      "grad_norm": 3.8576271533966064,
      "learning_rate": 1.0664406779661018e-05,
      "loss": 1.3682,
      "step": 14270
    },
    {
      "epoch": 0.36,
      "grad_norm": 3.8831918239593506,
      "learning_rate": 1.065762711864407e-05,
      "loss": 1.4227,
      "step": 14280
    },
    {
      "epoch": 0.36,
      "grad_norm": 3.9178073406219482,
      "learning_rate": 1.065084745762712e-05,
      "loss": 1.4681,
      "step": 14290
    },
    {
      "epoch": 0.36,
      "grad_norm": 4.398986339569092,
      "learning_rate": 1.0644067796610172e-05,
      "loss": 1.3255,
      "step": 14300
    },
    {
      "epoch": 0.36,
      "grad_norm": 9.885825157165527,
      "learning_rate": 1.0637288135593222e-05,
      "loss": 1.3654,
      "step": 14310
    },
    {
      "epoch": 0.36,
      "grad_norm": 7.351196765899658,
      "learning_rate": 1.0630508474576271e-05,
      "loss": 1.5729,
      "step": 14320
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.9780303239822388,
      "learning_rate": 1.0623728813559323e-05,
      "loss": 1.4228,
      "step": 14330
    },
    {
      "epoch": 0.36,
      "grad_norm": 10.410249710083008,
      "learning_rate": 1.0616949152542373e-05,
      "loss": 1.5135,
      "step": 14340
    },
    {
      "epoch": 0.36,
      "grad_norm": 2.9553821086883545,
      "learning_rate": 1.0610169491525424e-05,
      "loss": 1.5216,
      "step": 14350
    },
    {
      "epoch": 0.36,
      "grad_norm": 6.547674179077148,
      "learning_rate": 1.0603389830508475e-05,
      "loss": 1.24,
      "step": 14360
    },
    {
      "epoch": 0.36,
      "grad_norm": 2.1439294815063477,
      "learning_rate": 1.0596610169491525e-05,
      "loss": 1.3668,
      "step": 14370
    },
    {
      "epoch": 0.36,
      "grad_norm": 5.144618988037109,
      "learning_rate": 1.0589830508474576e-05,
      "loss": 1.5374,
      "step": 14380
    },
    {
      "epoch": 0.36,
      "grad_norm": 2.8388068675994873,
      "learning_rate": 1.0583050847457628e-05,
      "loss": 1.5478,
      "step": 14390
    },
    {
      "epoch": 0.36,
      "grad_norm": 4.369128227233887,
      "learning_rate": 1.057627118644068e-05,
      "loss": 1.4203,
      "step": 14400
    },
    {
      "epoch": 0.36,
      "grad_norm": 7.508152484893799,
      "learning_rate": 1.0569491525423729e-05,
      "loss": 1.3867,
      "step": 14410
    },
    {
      "epoch": 0.36,
      "grad_norm": 5.237878799438477,
      "learning_rate": 1.056271186440678e-05,
      "loss": 1.4668,
      "step": 14420
    },
    {
      "epoch": 0.36,
      "grad_norm": 6.717671871185303,
      "learning_rate": 1.0555932203389832e-05,
      "loss": 1.4708,
      "step": 14430
    },
    {
      "epoch": 0.36,
      "grad_norm": 5.358588218688965,
      "learning_rate": 1.0549152542372883e-05,
      "loss": 1.6207,
      "step": 14440
    },
    {
      "epoch": 0.36,
      "grad_norm": 4.034215450286865,
      "learning_rate": 1.0542372881355933e-05,
      "loss": 1.4627,
      "step": 14450
    },
    {
      "epoch": 0.36,
      "grad_norm": 2.0506346225738525,
      "learning_rate": 1.0535593220338984e-05,
      "loss": 1.367,
      "step": 14460
    },
    {
      "epoch": 0.36,
      "grad_norm": 16.490341186523438,
      "learning_rate": 1.0528813559322036e-05,
      "loss": 1.4104,
      "step": 14470
    },
    {
      "epoch": 0.36,
      "grad_norm": 3.040726900100708,
      "learning_rate": 1.0522033898305087e-05,
      "loss": 1.349,
      "step": 14480
    },
    {
      "epoch": 0.36,
      "grad_norm": 2.0654332637786865,
      "learning_rate": 1.0515254237288137e-05,
      "loss": 1.1943,
      "step": 14490
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.6009739637374878,
      "learning_rate": 1.0508474576271188e-05,
      "loss": 1.5093,
      "step": 14500
    },
    {
      "epoch": 0.36,
      "grad_norm": 6.735856533050537,
      "learning_rate": 1.050169491525424e-05,
      "loss": 1.4568,
      "step": 14510
    },
    {
      "epoch": 0.36,
      "grad_norm": 2.8814077377319336,
      "learning_rate": 1.049491525423729e-05,
      "loss": 1.4789,
      "step": 14520
    },
    {
      "epoch": 0.36,
      "grad_norm": 2.9417665004730225,
      "learning_rate": 1.0488135593220339e-05,
      "loss": 1.342,
      "step": 14530
    },
    {
      "epoch": 0.36,
      "grad_norm": 2.7162630558013916,
      "learning_rate": 1.048135593220339e-05,
      "loss": 1.3405,
      "step": 14540
    },
    {
      "epoch": 0.36,
      "grad_norm": 16.989566802978516,
      "learning_rate": 1.047457627118644e-05,
      "loss": 1.2926,
      "step": 14550
    },
    {
      "epoch": 0.36,
      "grad_norm": 5.988175392150879,
      "learning_rate": 1.0467796610169491e-05,
      "loss": 1.312,
      "step": 14560
    },
    {
      "epoch": 0.36,
      "grad_norm": 5.802346706390381,
      "learning_rate": 1.0461016949152543e-05,
      "loss": 1.308,
      "step": 14570
    },
    {
      "epoch": 0.36,
      "grad_norm": 9.407906532287598,
      "learning_rate": 1.0454237288135594e-05,
      "loss": 1.4781,
      "step": 14580
    },
    {
      "epoch": 0.36,
      "grad_norm": 2.769667863845825,
      "learning_rate": 1.0447457627118644e-05,
      "loss": 1.5887,
      "step": 14590
    },
    {
      "epoch": 0.36,
      "grad_norm": 5.349637985229492,
      "learning_rate": 1.0440677966101695e-05,
      "loss": 1.4238,
      "step": 14600
    },
    {
      "epoch": 0.37,
      "grad_norm": 8.598146438598633,
      "learning_rate": 1.0433898305084747e-05,
      "loss": 1.3711,
      "step": 14610
    },
    {
      "epoch": 0.37,
      "grad_norm": 11.67712116241455,
      "learning_rate": 1.0427118644067798e-05,
      "loss": 1.5724,
      "step": 14620
    },
    {
      "epoch": 0.37,
      "grad_norm": 6.301398754119873,
      "learning_rate": 1.0420338983050848e-05,
      "loss": 1.3599,
      "step": 14630
    },
    {
      "epoch": 0.37,
      "grad_norm": 5.270076274871826,
      "learning_rate": 1.04135593220339e-05,
      "loss": 1.4483,
      "step": 14640
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.520828366279602,
      "learning_rate": 1.040677966101695e-05,
      "loss": 1.5316,
      "step": 14650
    },
    {
      "epoch": 0.37,
      "grad_norm": 4.574060440063477,
      "learning_rate": 1.04e-05,
      "loss": 1.3414,
      "step": 14660
    },
    {
      "epoch": 0.37,
      "grad_norm": 9.551538467407227,
      "learning_rate": 1.0393220338983052e-05,
      "loss": 1.2514,
      "step": 14670
    },
    {
      "epoch": 0.37,
      "grad_norm": 2.1911561489105225,
      "learning_rate": 1.0386440677966103e-05,
      "loss": 1.4909,
      "step": 14680
    },
    {
      "epoch": 0.37,
      "grad_norm": 7.988354206085205,
      "learning_rate": 1.0379661016949155e-05,
      "loss": 1.4171,
      "step": 14690
    },
    {
      "epoch": 0.37,
      "grad_norm": 3.835221529006958,
      "learning_rate": 1.0372881355932204e-05,
      "loss": 1.3332,
      "step": 14700
    },
    {
      "epoch": 0.37,
      "grad_norm": 4.325536251068115,
      "learning_rate": 1.0366101694915256e-05,
      "loss": 1.4785,
      "step": 14710
    },
    {
      "epoch": 0.37,
      "grad_norm": 4.421005725860596,
      "learning_rate": 1.0359322033898307e-05,
      "loss": 1.3731,
      "step": 14720
    },
    {
      "epoch": 0.37,
      "grad_norm": 7.538575172424316,
      "learning_rate": 1.0352542372881358e-05,
      "loss": 1.454,
      "step": 14730
    },
    {
      "epoch": 0.37,
      "grad_norm": 7.058475494384766,
      "learning_rate": 1.0345762711864406e-05,
      "loss": 1.3607,
      "step": 14740
    },
    {
      "epoch": 0.37,
      "grad_norm": 3.7093636989593506,
      "learning_rate": 1.0338983050847458e-05,
      "loss": 1.3594,
      "step": 14750
    },
    {
      "epoch": 0.37,
      "grad_norm": 3.534818649291992,
      "learning_rate": 1.033220338983051e-05,
      "loss": 1.4324,
      "step": 14760
    },
    {
      "epoch": 0.37,
      "grad_norm": 2.282867908477783,
      "learning_rate": 1.0325423728813559e-05,
      "loss": 1.4294,
      "step": 14770
    },
    {
      "epoch": 0.37,
      "grad_norm": 6.613044261932373,
      "learning_rate": 1.031864406779661e-05,
      "loss": 1.4023,
      "step": 14780
    },
    {
      "epoch": 0.37,
      "grad_norm": 2.0816056728363037,
      "learning_rate": 1.0311864406779662e-05,
      "loss": 1.4054,
      "step": 14790
    },
    {
      "epoch": 0.37,
      "grad_norm": 4.677524089813232,
      "learning_rate": 1.0305084745762712e-05,
      "loss": 1.3661,
      "step": 14800
    },
    {
      "epoch": 0.37,
      "grad_norm": 3.064871072769165,
      "learning_rate": 1.0298305084745763e-05,
      "loss": 1.4508,
      "step": 14810
    },
    {
      "epoch": 0.37,
      "grad_norm": 3.247441530227661,
      "learning_rate": 1.0291525423728814e-05,
      "loss": 1.4769,
      "step": 14820
    },
    {
      "epoch": 0.37,
      "grad_norm": 3.2026891708374023,
      "learning_rate": 1.0284745762711866e-05,
      "loss": 1.5051,
      "step": 14830
    },
    {
      "epoch": 0.37,
      "grad_norm": 9.62253189086914,
      "learning_rate": 1.0277966101694915e-05,
      "loss": 1.4169,
      "step": 14840
    },
    {
      "epoch": 0.37,
      "grad_norm": 9.10492992401123,
      "learning_rate": 1.0271186440677967e-05,
      "loss": 1.3729,
      "step": 14850
    },
    {
      "epoch": 0.37,
      "grad_norm": 3.3863179683685303,
      "learning_rate": 1.0264406779661018e-05,
      "loss": 1.3174,
      "step": 14860
    },
    {
      "epoch": 0.37,
      "grad_norm": 3.950000286102295,
      "learning_rate": 1.025762711864407e-05,
      "loss": 1.2499,
      "step": 14870
    },
    {
      "epoch": 0.37,
      "grad_norm": 3.975641965866089,
      "learning_rate": 1.025084745762712e-05,
      "loss": 1.174,
      "step": 14880
    },
    {
      "epoch": 0.37,
      "grad_norm": 2.310656785964966,
      "learning_rate": 1.024406779661017e-05,
      "loss": 1.3562,
      "step": 14890
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.7508426904678345,
      "learning_rate": 1.0237288135593222e-05,
      "loss": 1.416,
      "step": 14900
    },
    {
      "epoch": 0.37,
      "grad_norm": 10.431480407714844,
      "learning_rate": 1.0230508474576274e-05,
      "loss": 1.346,
      "step": 14910
    },
    {
      "epoch": 0.37,
      "grad_norm": 3.3649802207946777,
      "learning_rate": 1.0223728813559323e-05,
      "loss": 1.3324,
      "step": 14920
    },
    {
      "epoch": 0.37,
      "grad_norm": 8.90680980682373,
      "learning_rate": 1.0216949152542375e-05,
      "loss": 1.4269,
      "step": 14930
    },
    {
      "epoch": 0.37,
      "grad_norm": 3.7382590770721436,
      "learning_rate": 1.0210169491525423e-05,
      "loss": 1.4904,
      "step": 14940
    },
    {
      "epoch": 0.37,
      "grad_norm": 2.7689971923828125,
      "learning_rate": 1.0203389830508474e-05,
      "loss": 1.2617,
      "step": 14950
    },
    {
      "epoch": 0.37,
      "grad_norm": 4.021928787231445,
      "learning_rate": 1.0196610169491525e-05,
      "loss": 1.4237,
      "step": 14960
    },
    {
      "epoch": 0.37,
      "grad_norm": 4.612102031707764,
      "learning_rate": 1.0189830508474577e-05,
      "loss": 1.3606,
      "step": 14970
    },
    {
      "epoch": 0.37,
      "grad_norm": 4.465333938598633,
      "learning_rate": 1.0183050847457627e-05,
      "loss": 1.4363,
      "step": 14980
    },
    {
      "epoch": 0.37,
      "grad_norm": 3.054535150527954,
      "learning_rate": 1.0176271186440678e-05,
      "loss": 1.4046,
      "step": 14990
    },
    {
      "epoch": 0.38,
      "grad_norm": 9.842041969299316,
      "learning_rate": 1.016949152542373e-05,
      "loss": 1.3313,
      "step": 15000
    },
    {
      "epoch": 0.38,
      "eval_loss": 1.3888890743255615,
      "eval_runtime": 122.3446,
      "eval_samples_per_second": 8.174,
      "eval_steps_per_second": 8.174,
      "step": 15000
    },
    {
      "epoch": 0.38,
      "grad_norm": 3.3704476356506348,
      "learning_rate": 1.016271186440678e-05,
      "loss": 1.5013,
      "step": 15010
    },
    {
      "epoch": 0.38,
      "grad_norm": 4.385085105895996,
      "learning_rate": 1.015593220338983e-05,
      "loss": 1.2783,
      "step": 15020
    },
    {
      "epoch": 0.38,
      "grad_norm": 2.657679557800293,
      "learning_rate": 1.0149152542372882e-05,
      "loss": 1.2901,
      "step": 15030
    },
    {
      "epoch": 0.38,
      "grad_norm": 6.710740089416504,
      "learning_rate": 1.0142372881355933e-05,
      "loss": 1.3512,
      "step": 15040
    },
    {
      "epoch": 0.38,
      "grad_norm": 7.799945831298828,
      "learning_rate": 1.0135593220338985e-05,
      "loss": 1.3848,
      "step": 15050
    },
    {
      "epoch": 0.38,
      "grad_norm": 10.693659782409668,
      "learning_rate": 1.0128813559322034e-05,
      "loss": 1.2004,
      "step": 15060
    },
    {
      "epoch": 0.38,
      "grad_norm": 3.980273723602295,
      "learning_rate": 1.0122033898305086e-05,
      "loss": 1.5739,
      "step": 15070
    },
    {
      "epoch": 0.38,
      "grad_norm": 4.042763710021973,
      "learning_rate": 1.0115254237288137e-05,
      "loss": 1.2657,
      "step": 15080
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.3417140245437622,
      "learning_rate": 1.0108474576271189e-05,
      "loss": 1.206,
      "step": 15090
    },
    {
      "epoch": 0.38,
      "grad_norm": 5.994216442108154,
      "learning_rate": 1.0101694915254238e-05,
      "loss": 1.3248,
      "step": 15100
    },
    {
      "epoch": 0.38,
      "grad_norm": 2.938751697540283,
      "learning_rate": 1.009491525423729e-05,
      "loss": 1.4234,
      "step": 15110
    },
    {
      "epoch": 0.38,
      "grad_norm": 5.216578006744385,
      "learning_rate": 1.0088135593220341e-05,
      "loss": 1.3464,
      "step": 15120
    },
    {
      "epoch": 0.38,
      "grad_norm": 11.384958267211914,
      "learning_rate": 1.008135593220339e-05,
      "loss": 1.4197,
      "step": 15130
    },
    {
      "epoch": 0.38,
      "grad_norm": 5.775814056396484,
      "learning_rate": 1.0074576271186442e-05,
      "loss": 1.4297,
      "step": 15140
    },
    {
      "epoch": 0.38,
      "grad_norm": 6.48391580581665,
      "learning_rate": 1.0067796610169492e-05,
      "loss": 1.3123,
      "step": 15150
    },
    {
      "epoch": 0.38,
      "grad_norm": 4.773738861083984,
      "learning_rate": 1.0061016949152542e-05,
      "loss": 1.3538,
      "step": 15160
    },
    {
      "epoch": 0.38,
      "grad_norm": 5.0771942138671875,
      "learning_rate": 1.0054237288135593e-05,
      "loss": 1.349,
      "step": 15170
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.8072324991226196,
      "learning_rate": 1.0047457627118644e-05,
      "loss": 1.3801,
      "step": 15180
    },
    {
      "epoch": 0.38,
      "grad_norm": 4.061614513397217,
      "learning_rate": 1.0040677966101696e-05,
      "loss": 1.3398,
      "step": 15190
    },
    {
      "epoch": 0.38,
      "grad_norm": 5.61275053024292,
      "learning_rate": 1.0033898305084746e-05,
      "loss": 1.4425,
      "step": 15200
    },
    {
      "epoch": 0.38,
      "grad_norm": 16.69525718688965,
      "learning_rate": 1.0027118644067797e-05,
      "loss": 1.4937,
      "step": 15210
    },
    {
      "epoch": 0.38,
      "grad_norm": 4.031248092651367,
      "learning_rate": 1.0020338983050848e-05,
      "loss": 1.3415,
      "step": 15220
    },
    {
      "epoch": 0.38,
      "grad_norm": 2.307645559310913,
      "learning_rate": 1.00135593220339e-05,
      "loss": 1.5268,
      "step": 15230
    },
    {
      "epoch": 0.38,
      "grad_norm": 3.7098755836486816,
      "learning_rate": 1.000677966101695e-05,
      "loss": 1.3768,
      "step": 15240
    },
    {
      "epoch": 0.38,
      "grad_norm": 6.115889072418213,
      "learning_rate": 1e-05,
      "loss": 1.4285,
      "step": 15250
    },
    {
      "epoch": 0.38,
      "grad_norm": 7.988094329833984,
      "learning_rate": 9.993220338983052e-06,
      "loss": 1.3858,
      "step": 15260
    },
    {
      "epoch": 0.38,
      "grad_norm": 10.41092300415039,
      "learning_rate": 9.986440677966102e-06,
      "loss": 1.4315,
      "step": 15270
    },
    {
      "epoch": 0.38,
      "grad_norm": 3.125478744506836,
      "learning_rate": 9.979661016949153e-06,
      "loss": 1.3608,
      "step": 15280
    },
    {
      "epoch": 0.38,
      "grad_norm": 3.9737863540649414,
      "learning_rate": 9.972881355932205e-06,
      "loss": 1.4055,
      "step": 15290
    },
    {
      "epoch": 0.38,
      "grad_norm": 11.59518814086914,
      "learning_rate": 9.966101694915256e-06,
      "loss": 1.5088,
      "step": 15300
    },
    {
      "epoch": 0.38,
      "grad_norm": 6.9823079109191895,
      "learning_rate": 9.959322033898306e-06,
      "loss": 1.4821,
      "step": 15310
    },
    {
      "epoch": 0.38,
      "grad_norm": 3.4408843517303467,
      "learning_rate": 9.952542372881356e-06,
      "loss": 1.4964,
      "step": 15320
    },
    {
      "epoch": 0.38,
      "grad_norm": 2.332047700881958,
      "learning_rate": 9.945762711864407e-06,
      "loss": 1.4629,
      "step": 15330
    },
    {
      "epoch": 0.38,
      "grad_norm": 2.7520997524261475,
      "learning_rate": 9.938983050847458e-06,
      "loss": 1.3057,
      "step": 15340
    },
    {
      "epoch": 0.38,
      "grad_norm": 4.296947002410889,
      "learning_rate": 9.93220338983051e-06,
      "loss": 1.3537,
      "step": 15350
    },
    {
      "epoch": 0.38,
      "grad_norm": 6.33931303024292,
      "learning_rate": 9.92542372881356e-06,
      "loss": 1.2661,
      "step": 15360
    },
    {
      "epoch": 0.38,
      "grad_norm": 7.359936237335205,
      "learning_rate": 9.918644067796611e-06,
      "loss": 1.4041,
      "step": 15370
    },
    {
      "epoch": 0.38,
      "grad_norm": 4.600747108459473,
      "learning_rate": 9.911864406779662e-06,
      "loss": 1.5198,
      "step": 15380
    },
    {
      "epoch": 0.38,
      "grad_norm": 3.2126359939575195,
      "learning_rate": 9.905084745762714e-06,
      "loss": 1.2859,
      "step": 15390
    },
    {
      "epoch": 0.39,
      "grad_norm": 3.7875165939331055,
      "learning_rate": 9.898305084745763e-06,
      "loss": 1.4244,
      "step": 15400
    },
    {
      "epoch": 0.39,
      "grad_norm": 5.999313831329346,
      "learning_rate": 9.891525423728813e-06,
      "loss": 1.2918,
      "step": 15410
    },
    {
      "epoch": 0.39,
      "grad_norm": 2.040942668914795,
      "learning_rate": 9.884745762711864e-06,
      "loss": 1.3725,
      "step": 15420
    },
    {
      "epoch": 0.39,
      "grad_norm": 6.45195198059082,
      "learning_rate": 9.877966101694916e-06,
      "loss": 1.3233,
      "step": 15430
    },
    {
      "epoch": 0.39,
      "grad_norm": 3.630338668823242,
      "learning_rate": 9.871186440677967e-06,
      "loss": 1.3428,
      "step": 15440
    },
    {
      "epoch": 0.39,
      "grad_norm": 4.121795177459717,
      "learning_rate": 9.864406779661017e-06,
      "loss": 1.3865,
      "step": 15450
    },
    {
      "epoch": 0.39,
      "grad_norm": 4.676584243774414,
      "learning_rate": 9.857627118644068e-06,
      "loss": 1.4434,
      "step": 15460
    },
    {
      "epoch": 0.39,
      "grad_norm": 8.199736595153809,
      "learning_rate": 9.85084745762712e-06,
      "loss": 1.5816,
      "step": 15470
    },
    {
      "epoch": 0.39,
      "grad_norm": 2.099637031555176,
      "learning_rate": 9.844067796610171e-06,
      "loss": 1.421,
      "step": 15480
    },
    {
      "epoch": 0.39,
      "grad_norm": 4.684576034545898,
      "learning_rate": 9.837288135593221e-06,
      "loss": 1.3991,
      "step": 15490
    },
    {
      "epoch": 0.39,
      "grad_norm": 5.442004203796387,
      "learning_rate": 9.830508474576272e-06,
      "loss": 1.3609,
      "step": 15500
    },
    {
      "epoch": 0.39,
      "grad_norm": 6.5514984130859375,
      "learning_rate": 9.823728813559322e-06,
      "loss": 1.4227,
      "step": 15510
    },
    {
      "epoch": 0.39,
      "grad_norm": 3.1997768878936768,
      "learning_rate": 9.816949152542373e-06,
      "loss": 1.3313,
      "step": 15520
    },
    {
      "epoch": 0.39,
      "grad_norm": 6.121874809265137,
      "learning_rate": 9.810169491525425e-06,
      "loss": 1.4085,
      "step": 15530
    },
    {
      "epoch": 0.39,
      "grad_norm": 12.801511764526367,
      "learning_rate": 9.803389830508474e-06,
      "loss": 1.4496,
      "step": 15540
    },
    {
      "epoch": 0.39,
      "grad_norm": 2.6913018226623535,
      "learning_rate": 9.796610169491526e-06,
      "loss": 1.2745,
      "step": 15550
    },
    {
      "epoch": 0.39,
      "grad_norm": 2.5916895866394043,
      "learning_rate": 9.789830508474577e-06,
      "loss": 1.2973,
      "step": 15560
    },
    {
      "epoch": 0.39,
      "grad_norm": 3.151700258255005,
      "learning_rate": 9.783050847457629e-06,
      "loss": 1.4092,
      "step": 15570
    },
    {
      "epoch": 0.39,
      "grad_norm": 2.532691717147827,
      "learning_rate": 9.776271186440678e-06,
      "loss": 1.2579,
      "step": 15580
    },
    {
      "epoch": 0.39,
      "grad_norm": 2.8922550678253174,
      "learning_rate": 9.76949152542373e-06,
      "loss": 1.1696,
      "step": 15590
    },
    {
      "epoch": 0.39,
      "grad_norm": 11.087928771972656,
      "learning_rate": 9.762711864406781e-06,
      "loss": 1.3841,
      "step": 15600
    },
    {
      "epoch": 0.39,
      "grad_norm": 6.364797115325928,
      "learning_rate": 9.755932203389833e-06,
      "loss": 1.3911,
      "step": 15610
    },
    {
      "epoch": 0.39,
      "grad_norm": 3.893240451812744,
      "learning_rate": 9.749152542372882e-06,
      "loss": 1.3766,
      "step": 15620
    },
    {
      "epoch": 0.39,
      "grad_norm": 3.214137554168701,
      "learning_rate": 9.742372881355932e-06,
      "loss": 1.3021,
      "step": 15630
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.6587047576904297,
      "learning_rate": 9.735593220338983e-06,
      "loss": 1.3929,
      "step": 15640
    },
    {
      "epoch": 0.39,
      "grad_norm": 6.39413595199585,
      "learning_rate": 9.728813559322035e-06,
      "loss": 1.5443,
      "step": 15650
    },
    {
      "epoch": 0.39,
      "grad_norm": 5.886016845703125,
      "learning_rate": 9.722033898305086e-06,
      "loss": 1.343,
      "step": 15660
    },
    {
      "epoch": 0.39,
      "grad_norm": 6.696247577667236,
      "learning_rate": 9.715254237288136e-06,
      "loss": 1.4697,
      "step": 15670
    },
    {
      "epoch": 0.39,
      "grad_norm": 16.11737823486328,
      "learning_rate": 9.708474576271187e-06,
      "loss": 1.1966,
      "step": 15680
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.0612120628356934,
      "learning_rate": 9.701694915254239e-06,
      "loss": 1.5481,
      "step": 15690
    },
    {
      "epoch": 0.39,
      "grad_norm": 4.875497817993164,
      "learning_rate": 9.69491525423729e-06,
      "loss": 1.4698,
      "step": 15700
    },
    {
      "epoch": 0.39,
      "grad_norm": 3.7041900157928467,
      "learning_rate": 9.68813559322034e-06,
      "loss": 1.4084,
      "step": 15710
    },
    {
      "epoch": 0.39,
      "grad_norm": 3.641465187072754,
      "learning_rate": 9.68135593220339e-06,
      "loss": 1.1715,
      "step": 15720
    },
    {
      "epoch": 0.39,
      "grad_norm": 8.047656059265137,
      "learning_rate": 9.674576271186441e-06,
      "loss": 1.2772,
      "step": 15730
    },
    {
      "epoch": 0.39,
      "grad_norm": 10.9264554977417,
      "learning_rate": 9.667796610169492e-06,
      "loss": 1.4736,
      "step": 15740
    },
    {
      "epoch": 0.39,
      "grad_norm": 6.976169109344482,
      "learning_rate": 9.661016949152544e-06,
      "loss": 1.4825,
      "step": 15750
    },
    {
      "epoch": 0.39,
      "grad_norm": 4.377677917480469,
      "learning_rate": 9.654237288135593e-06,
      "loss": 1.4606,
      "step": 15760
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.9935078620910645,
      "learning_rate": 9.647457627118645e-06,
      "loss": 1.3048,
      "step": 15770
    },
    {
      "epoch": 0.39,
      "grad_norm": 2.633688449859619,
      "learning_rate": 9.640677966101696e-06,
      "loss": 1.312,
      "step": 15780
    },
    {
      "epoch": 0.39,
      "grad_norm": 2.4499807357788086,
      "learning_rate": 9.633898305084746e-06,
      "loss": 1.3615,
      "step": 15790
    },
    {
      "epoch": 0.4,
      "grad_norm": 2.0773448944091797,
      "learning_rate": 9.627118644067797e-06,
      "loss": 1.289,
      "step": 15800
    },
    {
      "epoch": 0.4,
      "grad_norm": 4.064075469970703,
      "learning_rate": 9.620338983050849e-06,
      "loss": 1.5547,
      "step": 15810
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.4866830110549927,
      "learning_rate": 9.6135593220339e-06,
      "loss": 1.4588,
      "step": 15820
    },
    {
      "epoch": 0.4,
      "grad_norm": 5.019138813018799,
      "learning_rate": 9.60677966101695e-06,
      "loss": 1.4697,
      "step": 15830
    },
    {
      "epoch": 0.4,
      "grad_norm": 10.953125,
      "learning_rate": 9.600000000000001e-06,
      "loss": 1.3294,
      "step": 15840
    },
    {
      "epoch": 0.4,
      "grad_norm": 6.1235761642456055,
      "learning_rate": 9.593220338983051e-06,
      "loss": 1.3492,
      "step": 15850
    },
    {
      "epoch": 0.4,
      "grad_norm": 6.580984592437744,
      "learning_rate": 9.586440677966102e-06,
      "loss": 1.2905,
      "step": 15860
    },
    {
      "epoch": 0.4,
      "grad_norm": 2.7498395442962646,
      "learning_rate": 9.579661016949154e-06,
      "loss": 1.3735,
      "step": 15870
    },
    {
      "epoch": 0.4,
      "grad_norm": 4.033597469329834,
      "learning_rate": 9.572881355932203e-06,
      "loss": 1.0994,
      "step": 15880
    },
    {
      "epoch": 0.4,
      "grad_norm": 8.705718994140625,
      "learning_rate": 9.566101694915255e-06,
      "loss": 1.4352,
      "step": 15890
    },
    {
      "epoch": 0.4,
      "grad_norm": 4.520132541656494,
      "learning_rate": 9.559322033898306e-06,
      "loss": 1.509,
      "step": 15900
    },
    {
      "epoch": 0.4,
      "grad_norm": 4.481790065765381,
      "learning_rate": 9.552542372881358e-06,
      "loss": 1.304,
      "step": 15910
    },
    {
      "epoch": 0.4,
      "grad_norm": 4.652289867401123,
      "learning_rate": 9.545762711864407e-06,
      "loss": 1.4473,
      "step": 15920
    },
    {
      "epoch": 0.4,
      "grad_norm": 4.510603904724121,
      "learning_rate": 9.538983050847457e-06,
      "loss": 1.4935,
      "step": 15930
    },
    {
      "epoch": 0.4,
      "grad_norm": 7.690815448760986,
      "learning_rate": 9.532203389830508e-06,
      "loss": 1.534,
      "step": 15940
    },
    {
      "epoch": 0.4,
      "grad_norm": 6.279505729675293,
      "learning_rate": 9.52542372881356e-06,
      "loss": 1.3972,
      "step": 15950
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.9571785926818848,
      "learning_rate": 9.518644067796611e-06,
      "loss": 1.465,
      "step": 15960
    },
    {
      "epoch": 0.4,
      "grad_norm": 4.379458904266357,
      "learning_rate": 9.511864406779661e-06,
      "loss": 1.3987,
      "step": 15970
    },
    {
      "epoch": 0.4,
      "grad_norm": 4.896313190460205,
      "learning_rate": 9.505084745762712e-06,
      "loss": 1.3982,
      "step": 15980
    },
    {
      "epoch": 0.4,
      "grad_norm": 3.3697798252105713,
      "learning_rate": 9.498305084745764e-06,
      "loss": 1.3195,
      "step": 15990
    },
    {
      "epoch": 0.4,
      "grad_norm": 7.6551361083984375,
      "learning_rate": 9.491525423728815e-06,
      "loss": 1.5165,
      "step": 16000
    },
    {
      "epoch": 0.4,
      "grad_norm": 6.0134053230285645,
      "learning_rate": 9.484745762711865e-06,
      "loss": 1.4543,
      "step": 16010
    },
    {
      "epoch": 0.4,
      "grad_norm": 2.629333019256592,
      "learning_rate": 9.477966101694916e-06,
      "loss": 1.3365,
      "step": 16020
    },
    {
      "epoch": 0.4,
      "grad_norm": 3.2119452953338623,
      "learning_rate": 9.471186440677966e-06,
      "loss": 1.1254,
      "step": 16030
    },
    {
      "epoch": 0.4,
      "grad_norm": 3.6267120838165283,
      "learning_rate": 9.464406779661017e-06,
      "loss": 1.388,
      "step": 16040
    },
    {
      "epoch": 0.4,
      "grad_norm": 3.0831100940704346,
      "learning_rate": 9.457627118644069e-06,
      "loss": 1.3232,
      "step": 16050
    },
    {
      "epoch": 0.4,
      "grad_norm": 4.732052803039551,
      "learning_rate": 9.450847457627119e-06,
      "loss": 1.3444,
      "step": 16060
    },
    {
      "epoch": 0.4,
      "grad_norm": 2.5624067783355713,
      "learning_rate": 9.44406779661017e-06,
      "loss": 1.3221,
      "step": 16070
    },
    {
      "epoch": 0.4,
      "grad_norm": 4.598475933074951,
      "learning_rate": 9.437288135593221e-06,
      "loss": 1.2786,
      "step": 16080
    },
    {
      "epoch": 0.4,
      "grad_norm": 2.739576816558838,
      "learning_rate": 9.430508474576273e-06,
      "loss": 1.1858,
      "step": 16090
    },
    {
      "epoch": 0.4,
      "grad_norm": 11.181398391723633,
      "learning_rate": 9.423728813559322e-06,
      "loss": 1.3946,
      "step": 16100
    },
    {
      "epoch": 0.4,
      "grad_norm": 5.001202583312988,
      "learning_rate": 9.416949152542374e-06,
      "loss": 1.4787,
      "step": 16110
    },
    {
      "epoch": 0.4,
      "grad_norm": 6.388422966003418,
      "learning_rate": 9.410169491525425e-06,
      "loss": 1.2745,
      "step": 16120
    },
    {
      "epoch": 0.4,
      "grad_norm": 5.090276718139648,
      "learning_rate": 9.403389830508477e-06,
      "loss": 1.2736,
      "step": 16130
    },
    {
      "epoch": 0.4,
      "grad_norm": 4.6817779541015625,
      "learning_rate": 9.396610169491526e-06,
      "loss": 1.3128,
      "step": 16140
    },
    {
      "epoch": 0.4,
      "grad_norm": 3.3270206451416016,
      "learning_rate": 9.389830508474576e-06,
      "loss": 1.3573,
      "step": 16150
    },
    {
      "epoch": 0.4,
      "grad_norm": 2.4766674041748047,
      "learning_rate": 9.383050847457627e-06,
      "loss": 1.5016,
      "step": 16160
    },
    {
      "epoch": 0.4,
      "grad_norm": 5.795687198638916,
      "learning_rate": 9.376271186440679e-06,
      "loss": 1.4935,
      "step": 16170
    },
    {
      "epoch": 0.4,
      "grad_norm": 8.751638412475586,
      "learning_rate": 9.36949152542373e-06,
      "loss": 1.5272,
      "step": 16180
    },
    {
      "epoch": 0.4,
      "grad_norm": 5.794180870056152,
      "learning_rate": 9.36271186440678e-06,
      "loss": 1.2899,
      "step": 16190
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.3093836307525635,
      "learning_rate": 9.355932203389831e-06,
      "loss": 1.3653,
      "step": 16200
    },
    {
      "epoch": 0.41,
      "grad_norm": 2.4276533126831055,
      "learning_rate": 9.349152542372883e-06,
      "loss": 1.4175,
      "step": 16210
    },
    {
      "epoch": 0.41,
      "grad_norm": 3.2700552940368652,
      "learning_rate": 9.342372881355934e-06,
      "loss": 1.271,
      "step": 16220
    },
    {
      "epoch": 0.41,
      "grad_norm": 2.873892307281494,
      "learning_rate": 9.335593220338984e-06,
      "loss": 1.3345,
      "step": 16230
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.3676258325576782,
      "learning_rate": 9.328813559322034e-06,
      "loss": 1.4357,
      "step": 16240
    },
    {
      "epoch": 0.41,
      "grad_norm": 10.726542472839355,
      "learning_rate": 9.322033898305085e-06,
      "loss": 1.5011,
      "step": 16250
    },
    {
      "epoch": 0.41,
      "grad_norm": 3.9172821044921875,
      "learning_rate": 9.315254237288136e-06,
      "loss": 1.2187,
      "step": 16260
    },
    {
      "epoch": 0.41,
      "grad_norm": 3.4887006282806396,
      "learning_rate": 9.308474576271188e-06,
      "loss": 1.291,
      "step": 16270
    },
    {
      "epoch": 0.41,
      "grad_norm": 2.8565285205841064,
      "learning_rate": 9.301694915254237e-06,
      "loss": 1.3272,
      "step": 16280
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.6669663190841675,
      "learning_rate": 9.294915254237289e-06,
      "loss": 1.4696,
      "step": 16290
    },
    {
      "epoch": 0.41,
      "grad_norm": 3.704602003097534,
      "learning_rate": 9.28813559322034e-06,
      "loss": 1.2163,
      "step": 16300
    },
    {
      "epoch": 0.41,
      "grad_norm": 2.4821040630340576,
      "learning_rate": 9.28135593220339e-06,
      "loss": 1.4636,
      "step": 16310
    },
    {
      "epoch": 0.41,
      "grad_norm": 5.994017124176025,
      "learning_rate": 9.274576271186441e-06,
      "loss": 1.2135,
      "step": 16320
    },
    {
      "epoch": 0.41,
      "grad_norm": 4.454836845397949,
      "learning_rate": 9.267796610169493e-06,
      "loss": 1.3951,
      "step": 16330
    },
    {
      "epoch": 0.41,
      "grad_norm": 3.0423309803009033,
      "learning_rate": 9.261016949152544e-06,
      "loss": 1.3479,
      "step": 16340
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.7580335140228271,
      "learning_rate": 9.254237288135594e-06,
      "loss": 1.4471,
      "step": 16350
    },
    {
      "epoch": 0.41,
      "grad_norm": 4.719851016998291,
      "learning_rate": 9.247457627118645e-06,
      "loss": 1.2366,
      "step": 16360
    },
    {
      "epoch": 0.41,
      "grad_norm": 7.321564197540283,
      "learning_rate": 9.240677966101695e-06,
      "loss": 1.2232,
      "step": 16370
    },
    {
      "epoch": 0.41,
      "grad_norm": 4.101144790649414,
      "learning_rate": 9.233898305084746e-06,
      "loss": 1.5191,
      "step": 16380
    },
    {
      "epoch": 0.41,
      "grad_norm": 7.778228282928467,
      "learning_rate": 9.227118644067798e-06,
      "loss": 1.3983,
      "step": 16390
    },
    {
      "epoch": 0.41,
      "grad_norm": 2.54646635055542,
      "learning_rate": 9.220338983050847e-06,
      "loss": 1.4351,
      "step": 16400
    },
    {
      "epoch": 0.41,
      "grad_norm": 2.159886598587036,
      "learning_rate": 9.213559322033899e-06,
      "loss": 1.4239,
      "step": 16410
    },
    {
      "epoch": 0.41,
      "grad_norm": 3.386319160461426,
      "learning_rate": 9.20677966101695e-06,
      "loss": 1.4871,
      "step": 16420
    },
    {
      "epoch": 0.41,
      "grad_norm": 7.528735160827637,
      "learning_rate": 9.200000000000002e-06,
      "loss": 1.1861,
      "step": 16430
    },
    {
      "epoch": 0.41,
      "grad_norm": 2.3670830726623535,
      "learning_rate": 9.193220338983051e-06,
      "loss": 1.3077,
      "step": 16440
    },
    {
      "epoch": 0.41,
      "grad_norm": 9.332890510559082,
      "learning_rate": 9.186440677966101e-06,
      "loss": 1.5842,
      "step": 16450
    },
    {
      "epoch": 0.41,
      "grad_norm": 8.82120418548584,
      "learning_rate": 9.179661016949153e-06,
      "loss": 1.5407,
      "step": 16460
    },
    {
      "epoch": 0.41,
      "grad_norm": 7.420884609222412,
      "learning_rate": 9.172881355932204e-06,
      "loss": 1.3197,
      "step": 16470
    },
    {
      "epoch": 0.41,
      "grad_norm": 5.81050443649292,
      "learning_rate": 9.166101694915255e-06,
      "loss": 1.4797,
      "step": 16480
    },
    {
      "epoch": 0.41,
      "grad_norm": 4.399102210998535,
      "learning_rate": 9.159322033898305e-06,
      "loss": 1.2532,
      "step": 16490
    },
    {
      "epoch": 0.41,
      "grad_norm": 4.553720951080322,
      "learning_rate": 9.152542372881356e-06,
      "loss": 1.3293,
      "step": 16500
    },
    {
      "epoch": 0.41,
      "grad_norm": 6.8112053871154785,
      "learning_rate": 9.145762711864408e-06,
      "loss": 1.6102,
      "step": 16510
    },
    {
      "epoch": 0.41,
      "grad_norm": 3.031045913696289,
      "learning_rate": 9.13898305084746e-06,
      "loss": 1.3466,
      "step": 16520
    },
    {
      "epoch": 0.41,
      "grad_norm": 3.8192317485809326,
      "learning_rate": 9.132203389830509e-06,
      "loss": 1.4567,
      "step": 16530
    },
    {
      "epoch": 0.41,
      "grad_norm": 3.7835776805877686,
      "learning_rate": 9.12542372881356e-06,
      "loss": 1.2723,
      "step": 16540
    },
    {
      "epoch": 0.41,
      "grad_norm": 5.986169338226318,
      "learning_rate": 9.11864406779661e-06,
      "loss": 1.3469,
      "step": 16550
    },
    {
      "epoch": 0.41,
      "grad_norm": 5.293610095977783,
      "learning_rate": 9.111864406779661e-06,
      "loss": 1.4987,
      "step": 16560
    },
    {
      "epoch": 0.41,
      "grad_norm": 3.9783549308776855,
      "learning_rate": 9.105084745762713e-06,
      "loss": 1.3997,
      "step": 16570
    },
    {
      "epoch": 0.41,
      "grad_norm": 6.57890510559082,
      "learning_rate": 9.098305084745763e-06,
      "loss": 1.4637,
      "step": 16580
    },
    {
      "epoch": 0.41,
      "grad_norm": 4.2959394454956055,
      "learning_rate": 9.091525423728814e-06,
      "loss": 1.4215,
      "step": 16590
    },
    {
      "epoch": 0.41,
      "grad_norm": 2.5835142135620117,
      "learning_rate": 9.084745762711865e-06,
      "loss": 1.2732,
      "step": 16600
    },
    {
      "epoch": 0.42,
      "grad_norm": 2.248013734817505,
      "learning_rate": 9.077966101694917e-06,
      "loss": 1.3547,
      "step": 16610
    },
    {
      "epoch": 0.42,
      "grad_norm": 4.238186359405518,
      "learning_rate": 9.071186440677966e-06,
      "loss": 1.4386,
      "step": 16620
    },
    {
      "epoch": 0.42,
      "grad_norm": 2.6734702587127686,
      "learning_rate": 9.064406779661018e-06,
      "loss": 1.5402,
      "step": 16630
    },
    {
      "epoch": 0.42,
      "grad_norm": 7.255619049072266,
      "learning_rate": 9.05762711864407e-06,
      "loss": 1.3871,
      "step": 16640
    },
    {
      "epoch": 0.42,
      "grad_norm": 2.775770425796509,
      "learning_rate": 9.05084745762712e-06,
      "loss": 1.3824,
      "step": 16650
    },
    {
      "epoch": 0.42,
      "grad_norm": 7.425033092498779,
      "learning_rate": 9.04406779661017e-06,
      "loss": 1.2615,
      "step": 16660
    },
    {
      "epoch": 0.42,
      "grad_norm": 5.410216808319092,
      "learning_rate": 9.03728813559322e-06,
      "loss": 1.3343,
      "step": 16670
    },
    {
      "epoch": 0.42,
      "grad_norm": 3.405527353286743,
      "learning_rate": 9.030508474576271e-06,
      "loss": 1.4312,
      "step": 16680
    },
    {
      "epoch": 0.42,
      "grad_norm": 4.094501495361328,
      "learning_rate": 9.023728813559323e-06,
      "loss": 1.2288,
      "step": 16690
    },
    {
      "epoch": 0.42,
      "grad_norm": 2.783144950866699,
      "learning_rate": 9.016949152542374e-06,
      "loss": 1.4211,
      "step": 16700
    },
    {
      "epoch": 0.42,
      "grad_norm": 6.911566734313965,
      "learning_rate": 9.010169491525424e-06,
      "loss": 1.3325,
      "step": 16710
    },
    {
      "epoch": 0.42,
      "grad_norm": 4.0294108390808105,
      "learning_rate": 9.003389830508475e-06,
      "loss": 1.3507,
      "step": 16720
    },
    {
      "epoch": 0.42,
      "grad_norm": 4.844551086425781,
      "learning_rate": 8.996610169491527e-06,
      "loss": 1.3769,
      "step": 16730
    },
    {
      "epoch": 0.42,
      "grad_norm": 3.546492338180542,
      "learning_rate": 8.989830508474578e-06,
      "loss": 1.4002,
      "step": 16740
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.3420721292495728,
      "learning_rate": 8.983050847457628e-06,
      "loss": 1.2954,
      "step": 16750
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.8581109046936035,
      "learning_rate": 8.976271186440678e-06,
      "loss": 1.5132,
      "step": 16760
    },
    {
      "epoch": 0.42,
      "grad_norm": 4.649177551269531,
      "learning_rate": 8.969491525423729e-06,
      "loss": 1.4765,
      "step": 16770
    },
    {
      "epoch": 0.42,
      "grad_norm": 6.321155548095703,
      "learning_rate": 8.96271186440678e-06,
      "loss": 1.5169,
      "step": 16780
    },
    {
      "epoch": 0.42,
      "grad_norm": 6.5299530029296875,
      "learning_rate": 8.955932203389832e-06,
      "loss": 1.157,
      "step": 16790
    },
    {
      "epoch": 0.42,
      "grad_norm": 6.718413829803467,
      "learning_rate": 8.949152542372881e-06,
      "loss": 1.346,
      "step": 16800
    },
    {
      "epoch": 0.42,
      "grad_norm": 3.934128761291504,
      "learning_rate": 8.942372881355933e-06,
      "loss": 1.4273,
      "step": 16810
    },
    {
      "epoch": 0.42,
      "grad_norm": 2.2934927940368652,
      "learning_rate": 8.935593220338984e-06,
      "loss": 1.3726,
      "step": 16820
    },
    {
      "epoch": 0.42,
      "grad_norm": 5.354536533355713,
      "learning_rate": 8.928813559322036e-06,
      "loss": 1.4685,
      "step": 16830
    },
    {
      "epoch": 0.42,
      "grad_norm": 3.372624158859253,
      "learning_rate": 8.922033898305085e-06,
      "loss": 1.4789,
      "step": 16840
    },
    {
      "epoch": 0.42,
      "grad_norm": 4.341814994812012,
      "learning_rate": 8.915254237288137e-06,
      "loss": 1.4677,
      "step": 16850
    },
    {
      "epoch": 0.42,
      "grad_norm": 9.57802677154541,
      "learning_rate": 8.908474576271188e-06,
      "loss": 1.2754,
      "step": 16860
    },
    {
      "epoch": 0.42,
      "grad_norm": 3.5526912212371826,
      "learning_rate": 8.901694915254238e-06,
      "loss": 1.4577,
      "step": 16870
    },
    {
      "epoch": 0.42,
      "grad_norm": 5.3464789390563965,
      "learning_rate": 8.89491525423729e-06,
      "loss": 1.3144,
      "step": 16880
    },
    {
      "epoch": 0.42,
      "grad_norm": 4.8993377685546875,
      "learning_rate": 8.888135593220339e-06,
      "loss": 1.3245,
      "step": 16890
    },
    {
      "epoch": 0.42,
      "grad_norm": 9.462966918945312,
      "learning_rate": 8.88135593220339e-06,
      "loss": 1.5916,
      "step": 16900
    },
    {
      "epoch": 0.42,
      "grad_norm": 5.578088760375977,
      "learning_rate": 8.874576271186442e-06,
      "loss": 1.3036,
      "step": 16910
    },
    {
      "epoch": 0.42,
      "grad_norm": 7.304584503173828,
      "learning_rate": 8.867796610169492e-06,
      "loss": 1.267,
      "step": 16920
    },
    {
      "epoch": 0.42,
      "grad_norm": 8.829882621765137,
      "learning_rate": 8.861016949152543e-06,
      "loss": 1.4429,
      "step": 16930
    },
    {
      "epoch": 0.42,
      "grad_norm": 2.3376832008361816,
      "learning_rate": 8.854237288135594e-06,
      "loss": 1.3813,
      "step": 16940
    },
    {
      "epoch": 0.42,
      "grad_norm": 2.7193713188171387,
      "learning_rate": 8.847457627118646e-06,
      "loss": 1.3494,
      "step": 16950
    },
    {
      "epoch": 0.42,
      "grad_norm": 6.182406902313232,
      "learning_rate": 8.840677966101695e-06,
      "loss": 1.4056,
      "step": 16960
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.9391649961471558,
      "learning_rate": 8.833898305084747e-06,
      "loss": 1.5418,
      "step": 16970
    },
    {
      "epoch": 0.42,
      "grad_norm": 5.3956522941589355,
      "learning_rate": 8.827118644067797e-06,
      "loss": 1.4257,
      "step": 16980
    },
    {
      "epoch": 0.42,
      "grad_norm": 3.2380099296569824,
      "learning_rate": 8.820338983050848e-06,
      "loss": 1.4544,
      "step": 16990
    },
    {
      "epoch": 0.42,
      "grad_norm": 3.9518375396728516,
      "learning_rate": 8.8135593220339e-06,
      "loss": 1.1458,
      "step": 17000
    },
    {
      "epoch": 0.43,
      "grad_norm": 4.615858554840088,
      "learning_rate": 8.806779661016949e-06,
      "loss": 1.2322,
      "step": 17010
    },
    {
      "epoch": 0.43,
      "grad_norm": 3.6182825565338135,
      "learning_rate": 8.8e-06,
      "loss": 1.3042,
      "step": 17020
    },
    {
      "epoch": 0.43,
      "grad_norm": 5.142463207244873,
      "learning_rate": 8.793220338983052e-06,
      "loss": 1.3827,
      "step": 17030
    },
    {
      "epoch": 0.43,
      "grad_norm": 3.647782802581787,
      "learning_rate": 8.786440677966103e-06,
      "loss": 1.181,
      "step": 17040
    },
    {
      "epoch": 0.43,
      "grad_norm": 3.557016372680664,
      "learning_rate": 8.779661016949153e-06,
      "loss": 1.462,
      "step": 17050
    },
    {
      "epoch": 0.43,
      "grad_norm": 8.3936128616333,
      "learning_rate": 8.772881355932204e-06,
      "loss": 1.1388,
      "step": 17060
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.7268701791763306,
      "learning_rate": 8.766101694915254e-06,
      "loss": 1.5298,
      "step": 17070
    },
    {
      "epoch": 0.43,
      "grad_norm": 5.313370227813721,
      "learning_rate": 8.759322033898305e-06,
      "loss": 1.2205,
      "step": 17080
    },
    {
      "epoch": 0.43,
      "grad_norm": 5.083520889282227,
      "learning_rate": 8.752542372881357e-06,
      "loss": 1.3117,
      "step": 17090
    },
    {
      "epoch": 0.43,
      "grad_norm": 5.750375747680664,
      "learning_rate": 8.745762711864407e-06,
      "loss": 1.3815,
      "step": 17100
    },
    {
      "epoch": 0.43,
      "grad_norm": 6.129281520843506,
      "learning_rate": 8.738983050847458e-06,
      "loss": 1.2509,
      "step": 17110
    },
    {
      "epoch": 0.43,
      "grad_norm": 8.429251670837402,
      "learning_rate": 8.73220338983051e-06,
      "loss": 1.2116,
      "step": 17120
    },
    {
      "epoch": 0.43,
      "grad_norm": 4.266468048095703,
      "learning_rate": 8.72542372881356e-06,
      "loss": 1.4818,
      "step": 17130
    },
    {
      "epoch": 0.43,
      "grad_norm": 3.4918172359466553,
      "learning_rate": 8.71864406779661e-06,
      "loss": 1.2947,
      "step": 17140
    },
    {
      "epoch": 0.43,
      "grad_norm": 10.345577239990234,
      "learning_rate": 8.711864406779662e-06,
      "loss": 1.4044,
      "step": 17150
    },
    {
      "epoch": 0.43,
      "grad_norm": 6.321402549743652,
      "learning_rate": 8.705084745762713e-06,
      "loss": 1.2271,
      "step": 17160
    },
    {
      "epoch": 0.43,
      "grad_norm": 4.5922369956970215,
      "learning_rate": 8.698305084745765e-06,
      "loss": 1.2652,
      "step": 17170
    },
    {
      "epoch": 0.43,
      "grad_norm": 3.2404696941375732,
      "learning_rate": 8.691525423728814e-06,
      "loss": 1.2832,
      "step": 17180
    },
    {
      "epoch": 0.43,
      "grad_norm": 6.5807271003723145,
      "learning_rate": 8.684745762711864e-06,
      "loss": 1.3336,
      "step": 17190
    },
    {
      "epoch": 0.43,
      "grad_norm": 7.412044525146484,
      "learning_rate": 8.677966101694915e-06,
      "loss": 1.3573,
      "step": 17200
    },
    {
      "epoch": 0.43,
      "grad_norm": 11.085704803466797,
      "learning_rate": 8.671186440677967e-06,
      "loss": 1.4329,
      "step": 17210
    },
    {
      "epoch": 0.43,
      "grad_norm": 5.468708515167236,
      "learning_rate": 8.664406779661018e-06,
      "loss": 1.1951,
      "step": 17220
    },
    {
      "epoch": 0.43,
      "grad_norm": 9.419631004333496,
      "learning_rate": 8.657627118644068e-06,
      "loss": 1.3981,
      "step": 17230
    },
    {
      "epoch": 0.43,
      "grad_norm": 4.822430610656738,
      "learning_rate": 8.65084745762712e-06,
      "loss": 1.3057,
      "step": 17240
    },
    {
      "epoch": 0.43,
      "grad_norm": 5.3872971534729,
      "learning_rate": 8.64406779661017e-06,
      "loss": 1.3604,
      "step": 17250
    },
    {
      "epoch": 0.43,
      "grad_norm": 5.837887287139893,
      "learning_rate": 8.637288135593222e-06,
      "loss": 1.4484,
      "step": 17260
    },
    {
      "epoch": 0.43,
      "grad_norm": 3.392738103866577,
      "learning_rate": 8.630508474576272e-06,
      "loss": 1.314,
      "step": 17270
    },
    {
      "epoch": 0.43,
      "grad_norm": 3.168947219848633,
      "learning_rate": 8.623728813559322e-06,
      "loss": 1.3126,
      "step": 17280
    },
    {
      "epoch": 0.43,
      "grad_norm": 5.4609832763671875,
      "learning_rate": 8.616949152542373e-06,
      "loss": 1.5409,
      "step": 17290
    },
    {
      "epoch": 0.43,
      "grad_norm": 2.6080710887908936,
      "learning_rate": 8.610169491525424e-06,
      "loss": 1.4408,
      "step": 17300
    },
    {
      "epoch": 0.43,
      "grad_norm": 4.541211128234863,
      "learning_rate": 8.603389830508476e-06,
      "loss": 1.2828,
      "step": 17310
    },
    {
      "epoch": 0.43,
      "grad_norm": 6.359476089477539,
      "learning_rate": 8.596610169491526e-06,
      "loss": 1.3239,
      "step": 17320
    },
    {
      "epoch": 0.43,
      "grad_norm": 2.774082660675049,
      "learning_rate": 8.589830508474577e-06,
      "loss": 1.2819,
      "step": 17330
    },
    {
      "epoch": 0.43,
      "grad_norm": 5.138007640838623,
      "learning_rate": 8.583050847457628e-06,
      "loss": 1.3396,
      "step": 17340
    },
    {
      "epoch": 0.43,
      "grad_norm": 6.311397552490234,
      "learning_rate": 8.57627118644068e-06,
      "loss": 1.3651,
      "step": 17350
    },
    {
      "epoch": 0.43,
      "grad_norm": 4.323927402496338,
      "learning_rate": 8.56949152542373e-06,
      "loss": 1.3118,
      "step": 17360
    },
    {
      "epoch": 0.43,
      "grad_norm": 2.3549041748046875,
      "learning_rate": 8.56271186440678e-06,
      "loss": 1.3429,
      "step": 17370
    },
    {
      "epoch": 0.43,
      "grad_norm": 9.139942169189453,
      "learning_rate": 8.55593220338983e-06,
      "loss": 1.5345,
      "step": 17380
    },
    {
      "epoch": 0.43,
      "grad_norm": 7.705909252166748,
      "learning_rate": 8.549152542372882e-06,
      "loss": 1.4523,
      "step": 17390
    },
    {
      "epoch": 0.43,
      "grad_norm": 4.607798099517822,
      "learning_rate": 8.542372881355933e-06,
      "loss": 1.5757,
      "step": 17400
    },
    {
      "epoch": 0.44,
      "grad_norm": 2.9559946060180664,
      "learning_rate": 8.535593220338983e-06,
      "loss": 1.2822,
      "step": 17410
    },
    {
      "epoch": 0.44,
      "grad_norm": 3.2365543842315674,
      "learning_rate": 8.528813559322034e-06,
      "loss": 1.3692,
      "step": 17420
    },
    {
      "epoch": 0.44,
      "grad_norm": 3.374927043914795,
      "learning_rate": 8.522033898305086e-06,
      "loss": 1.3532,
      "step": 17430
    },
    {
      "epoch": 0.44,
      "grad_norm": 3.463963747024536,
      "learning_rate": 8.515254237288136e-06,
      "loss": 1.4345,
      "step": 17440
    },
    {
      "epoch": 0.44,
      "grad_norm": 4.043869495391846,
      "learning_rate": 8.508474576271187e-06,
      "loss": 1.2886,
      "step": 17450
    },
    {
      "epoch": 0.44,
      "grad_norm": 4.573731899261475,
      "learning_rate": 8.501694915254238e-06,
      "loss": 1.1644,
      "step": 17460
    },
    {
      "epoch": 0.44,
      "grad_norm": 5.096779823303223,
      "learning_rate": 8.49491525423729e-06,
      "loss": 1.1855,
      "step": 17470
    },
    {
      "epoch": 0.44,
      "grad_norm": 12.950697898864746,
      "learning_rate": 8.48813559322034e-06,
      "loss": 1.5736,
      "step": 17480
    },
    {
      "epoch": 0.44,
      "grad_norm": 4.568536281585693,
      "learning_rate": 8.481355932203391e-06,
      "loss": 1.4281,
      "step": 17490
    },
    {
      "epoch": 0.44,
      "grad_norm": 3.575422763824463,
      "learning_rate": 8.47457627118644e-06,
      "loss": 1.4935,
      "step": 17500
    },
    {
      "epoch": 0.44,
      "eval_loss": 1.3699909448623657,
      "eval_runtime": 122.3848,
      "eval_samples_per_second": 8.171,
      "eval_steps_per_second": 8.171,
      "step": 17500
    },
    {
      "epoch": 0.44,
      "grad_norm": 4.047749996185303,
      "learning_rate": 8.467796610169492e-06,
      "loss": 1.3602,
      "step": 17510
    },
    {
      "epoch": 0.44,
      "grad_norm": 4.81995964050293,
      "learning_rate": 8.461016949152543e-06,
      "loss": 1.4881,
      "step": 17520
    },
    {
      "epoch": 0.44,
      "grad_norm": 2.7308261394500732,
      "learning_rate": 8.454237288135593e-06,
      "loss": 1.3149,
      "step": 17530
    },
    {
      "epoch": 0.44,
      "grad_norm": 2.6615195274353027,
      "learning_rate": 8.447457627118644e-06,
      "loss": 1.2734,
      "step": 17540
    },
    {
      "epoch": 0.44,
      "grad_norm": 3.65482759475708,
      "learning_rate": 8.440677966101696e-06,
      "loss": 1.335,
      "step": 17550
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.8164212703704834,
      "learning_rate": 8.433898305084747e-06,
      "loss": 1.3527,
      "step": 17560
    },
    {
      "epoch": 0.44,
      "grad_norm": 4.036699295043945,
      "learning_rate": 8.427118644067797e-06,
      "loss": 1.3178,
      "step": 17570
    },
    {
      "epoch": 0.44,
      "grad_norm": 5.2271623611450195,
      "learning_rate": 8.420338983050848e-06,
      "loss": 1.3275,
      "step": 17580
    },
    {
      "epoch": 0.44,
      "grad_norm": 6.8988423347473145,
      "learning_rate": 8.413559322033898e-06,
      "loss": 1.3007,
      "step": 17590
    },
    {
      "epoch": 0.44,
      "grad_norm": 9.273791313171387,
      "learning_rate": 8.40677966101695e-06,
      "loss": 1.3349,
      "step": 17600
    },
    {
      "epoch": 0.44,
      "grad_norm": 7.584418296813965,
      "learning_rate": 8.400000000000001e-06,
      "loss": 1.2669,
      "step": 17610
    },
    {
      "epoch": 0.44,
      "grad_norm": 6.503374099731445,
      "learning_rate": 8.39322033898305e-06,
      "loss": 1.4744,
      "step": 17620
    },
    {
      "epoch": 0.44,
      "grad_norm": 5.538229465484619,
      "learning_rate": 8.386440677966102e-06,
      "loss": 1.3492,
      "step": 17630
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.4738035202026367,
      "learning_rate": 8.379661016949153e-06,
      "loss": 1.3839,
      "step": 17640
    },
    {
      "epoch": 0.44,
      "grad_norm": 2.364210367202759,
      "learning_rate": 8.372881355932205e-06,
      "loss": 1.4777,
      "step": 17650
    },
    {
      "epoch": 0.44,
      "grad_norm": 5.341204643249512,
      "learning_rate": 8.366101694915255e-06,
      "loss": 1.4599,
      "step": 17660
    },
    {
      "epoch": 0.44,
      "grad_norm": 8.549229621887207,
      "learning_rate": 8.359322033898306e-06,
      "loss": 1.477,
      "step": 17670
    },
    {
      "epoch": 0.44,
      "grad_norm": 7.134267807006836,
      "learning_rate": 8.352542372881357e-06,
      "loss": 1.377,
      "step": 17680
    },
    {
      "epoch": 0.44,
      "grad_norm": 4.102703094482422,
      "learning_rate": 8.345762711864409e-06,
      "loss": 1.079,
      "step": 17690
    },
    {
      "epoch": 0.44,
      "grad_norm": 5.486017227172852,
      "learning_rate": 8.338983050847458e-06,
      "loss": 1.2945,
      "step": 17700
    },
    {
      "epoch": 0.44,
      "grad_norm": 2.1329519748687744,
      "learning_rate": 8.332203389830508e-06,
      "loss": 1.2709,
      "step": 17710
    },
    {
      "epoch": 0.44,
      "grad_norm": 3.166658401489258,
      "learning_rate": 8.32542372881356e-06,
      "loss": 1.4188,
      "step": 17720
    },
    {
      "epoch": 0.44,
      "grad_norm": 3.9792320728302,
      "learning_rate": 8.318644067796611e-06,
      "loss": 1.2297,
      "step": 17730
    },
    {
      "epoch": 0.44,
      "grad_norm": 5.598804473876953,
      "learning_rate": 8.311864406779662e-06,
      "loss": 1.352,
      "step": 17740
    },
    {
      "epoch": 0.44,
      "grad_norm": 4.075491428375244,
      "learning_rate": 8.305084745762712e-06,
      "loss": 1.4048,
      "step": 17750
    },
    {
      "epoch": 0.44,
      "grad_norm": 4.653989315032959,
      "learning_rate": 8.298305084745763e-06,
      "loss": 1.2515,
      "step": 17760
    },
    {
      "epoch": 0.44,
      "grad_norm": 2.3320040702819824,
      "learning_rate": 8.291525423728815e-06,
      "loss": 1.4444,
      "step": 17770
    },
    {
      "epoch": 0.44,
      "grad_norm": 4.202153205871582,
      "learning_rate": 8.284745762711866e-06,
      "loss": 1.3681,
      "step": 17780
    },
    {
      "epoch": 0.44,
      "grad_norm": 3.0815560817718506,
      "learning_rate": 8.277966101694916e-06,
      "loss": 1.2924,
      "step": 17790
    },
    {
      "epoch": 0.45,
      "grad_norm": 5.612250804901123,
      "learning_rate": 8.271186440677966e-06,
      "loss": 1.394,
      "step": 17800
    },
    {
      "epoch": 0.45,
      "grad_norm": 4.899747848510742,
      "learning_rate": 8.264406779661017e-06,
      "loss": 1.3257,
      "step": 17810
    },
    {
      "epoch": 0.45,
      "grad_norm": 3.0536115169525146,
      "learning_rate": 8.257627118644068e-06,
      "loss": 1.2796,
      "step": 17820
    },
    {
      "epoch": 0.45,
      "grad_norm": 7.068782329559326,
      "learning_rate": 8.25084745762712e-06,
      "loss": 1.3163,
      "step": 17830
    },
    {
      "epoch": 0.45,
      "grad_norm": 4.943035125732422,
      "learning_rate": 8.24406779661017e-06,
      "loss": 1.3712,
      "step": 17840
    },
    {
      "epoch": 0.45,
      "grad_norm": 7.196053504943848,
      "learning_rate": 8.237288135593221e-06,
      "loss": 1.4155,
      "step": 17850
    },
    {
      "epoch": 0.45,
      "grad_norm": 5.740854263305664,
      "learning_rate": 8.230508474576272e-06,
      "loss": 1.3311,
      "step": 17860
    },
    {
      "epoch": 0.45,
      "grad_norm": 4.434732913970947,
      "learning_rate": 8.223728813559324e-06,
      "loss": 1.1361,
      "step": 17870
    },
    {
      "epoch": 0.45,
      "grad_norm": 9.674403190612793,
      "learning_rate": 8.216949152542373e-06,
      "loss": 1.2832,
      "step": 17880
    },
    {
      "epoch": 0.45,
      "grad_norm": 6.101053714752197,
      "learning_rate": 8.210169491525425e-06,
      "loss": 1.4947,
      "step": 17890
    },
    {
      "epoch": 0.45,
      "grad_norm": 5.026559352874756,
      "learning_rate": 8.203389830508475e-06,
      "loss": 1.4749,
      "step": 17900
    },
    {
      "epoch": 0.45,
      "grad_norm": 3.4650232791900635,
      "learning_rate": 8.196610169491526e-06,
      "loss": 1.3734,
      "step": 17910
    },
    {
      "epoch": 0.45,
      "grad_norm": 5.595499515533447,
      "learning_rate": 8.189830508474577e-06,
      "loss": 1.3842,
      "step": 17920
    },
    {
      "epoch": 0.45,
      "grad_norm": 9.26806354522705,
      "learning_rate": 8.183050847457627e-06,
      "loss": 1.4243,
      "step": 17930
    },
    {
      "epoch": 0.45,
      "grad_norm": 5.305275917053223,
      "learning_rate": 8.176271186440678e-06,
      "loss": 1.359,
      "step": 17940
    },
    {
      "epoch": 0.45,
      "grad_norm": 13.687361717224121,
      "learning_rate": 8.16949152542373e-06,
      "loss": 1.3508,
      "step": 17950
    },
    {
      "epoch": 0.45,
      "grad_norm": 2.9817168712615967,
      "learning_rate": 8.162711864406781e-06,
      "loss": 1.4238,
      "step": 17960
    },
    {
      "epoch": 0.45,
      "grad_norm": 55.80167007446289,
      "learning_rate": 8.155932203389831e-06,
      "loss": 1.6129,
      "step": 17970
    },
    {
      "epoch": 0.45,
      "grad_norm": 2.6112301349639893,
      "learning_rate": 8.149152542372882e-06,
      "loss": 1.2966,
      "step": 17980
    },
    {
      "epoch": 0.45,
      "grad_norm": 7.075587749481201,
      "learning_rate": 8.142372881355934e-06,
      "loss": 1.3541,
      "step": 17990
    },
    {
      "epoch": 0.45,
      "grad_norm": 5.8252129554748535,
      "learning_rate": 8.135593220338983e-06,
      "loss": 1.3036,
      "step": 18000
    },
    {
      "epoch": 0.45,
      "grad_norm": 2.099008083343506,
      "learning_rate": 8.128813559322035e-06,
      "loss": 1.3123,
      "step": 18010
    },
    {
      "epoch": 0.45,
      "grad_norm": 18.689348220825195,
      "learning_rate": 8.122033898305085e-06,
      "loss": 1.4644,
      "step": 18020
    },
    {
      "epoch": 0.45,
      "grad_norm": 4.318515300750732,
      "learning_rate": 8.115254237288136e-06,
      "loss": 1.5108,
      "step": 18030
    },
    {
      "epoch": 0.45,
      "grad_norm": 5.690342426300049,
      "learning_rate": 8.108474576271187e-06,
      "loss": 1.3619,
      "step": 18040
    },
    {
      "epoch": 0.45,
      "grad_norm": 5.543004512786865,
      "learning_rate": 8.101694915254237e-06,
      "loss": 1.5159,
      "step": 18050
    },
    {
      "epoch": 0.45,
      "grad_norm": 8.042171478271484,
      "learning_rate": 8.094915254237289e-06,
      "loss": 1.3401,
      "step": 18060
    },
    {
      "epoch": 0.45,
      "grad_norm": 4.132726669311523,
      "learning_rate": 8.08813559322034e-06,
      "loss": 1.2238,
      "step": 18070
    },
    {
      "epoch": 0.45,
      "grad_norm": 2.9036052227020264,
      "learning_rate": 8.081355932203391e-06,
      "loss": 1.335,
      "step": 18080
    },
    {
      "epoch": 0.45,
      "grad_norm": 10.872087478637695,
      "learning_rate": 8.074576271186441e-06,
      "loss": 1.4685,
      "step": 18090
    },
    {
      "epoch": 0.45,
      "grad_norm": 5.2267022132873535,
      "learning_rate": 8.067796610169492e-06,
      "loss": 1.3548,
      "step": 18100
    },
    {
      "epoch": 0.45,
      "grad_norm": 4.5072922706604,
      "learning_rate": 8.061016949152542e-06,
      "loss": 1.2696,
      "step": 18110
    },
    {
      "epoch": 0.45,
      "grad_norm": 2.018592357635498,
      "learning_rate": 8.054237288135594e-06,
      "loss": 1.2825,
      "step": 18120
    },
    {
      "epoch": 0.45,
      "grad_norm": 6.42473030090332,
      "learning_rate": 8.047457627118645e-06,
      "loss": 1.2975,
      "step": 18130
    },
    {
      "epoch": 0.45,
      "grad_norm": 9.191697120666504,
      "learning_rate": 8.040677966101695e-06,
      "loss": 1.3941,
      "step": 18140
    },
    {
      "epoch": 0.45,
      "grad_norm": 5.8174591064453125,
      "learning_rate": 8.033898305084746e-06,
      "loss": 1.2577,
      "step": 18150
    },
    {
      "epoch": 0.45,
      "grad_norm": 3.370523452758789,
      "learning_rate": 8.027118644067797e-06,
      "loss": 1.3591,
      "step": 18160
    },
    {
      "epoch": 0.45,
      "grad_norm": 3.2100830078125,
      "learning_rate": 8.020338983050849e-06,
      "loss": 1.0972,
      "step": 18170
    },
    {
      "epoch": 0.45,
      "grad_norm": 3.962881326675415,
      "learning_rate": 8.013559322033899e-06,
      "loss": 1.4304,
      "step": 18180
    },
    {
      "epoch": 0.45,
      "grad_norm": 3.239504337310791,
      "learning_rate": 8.00677966101695e-06,
      "loss": 1.4001,
      "step": 18190
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.865474820137024,
      "learning_rate": 8.000000000000001e-06,
      "loss": 1.4586,
      "step": 18200
    },
    {
      "epoch": 0.46,
      "grad_norm": 5.256229877471924,
      "learning_rate": 7.993220338983053e-06,
      "loss": 1.352,
      "step": 18210
    },
    {
      "epoch": 0.46,
      "grad_norm": 3.299957752227783,
      "learning_rate": 7.986440677966102e-06,
      "loss": 1.2885,
      "step": 18220
    },
    {
      "epoch": 0.46,
      "grad_norm": 3.1765966415405273,
      "learning_rate": 7.979661016949152e-06,
      "loss": 1.3131,
      "step": 18230
    },
    {
      "epoch": 0.46,
      "grad_norm": 9.922146797180176,
      "learning_rate": 7.972881355932204e-06,
      "loss": 1.435,
      "step": 18240
    },
    {
      "epoch": 0.46,
      "grad_norm": 12.427643775939941,
      "learning_rate": 7.966101694915255e-06,
      "loss": 1.5732,
      "step": 18250
    },
    {
      "epoch": 0.46,
      "grad_norm": 11.859210014343262,
      "learning_rate": 7.959322033898306e-06,
      "loss": 1.3325,
      "step": 18260
    },
    {
      "epoch": 0.46,
      "grad_norm": 5.892291069030762,
      "learning_rate": 7.952542372881356e-06,
      "loss": 1.3961,
      "step": 18270
    },
    {
      "epoch": 0.46,
      "grad_norm": 7.953782081604004,
      "learning_rate": 7.945762711864407e-06,
      "loss": 1.3579,
      "step": 18280
    },
    {
      "epoch": 0.46,
      "grad_norm": 8.380058288574219,
      "learning_rate": 7.938983050847459e-06,
      "loss": 1.2266,
      "step": 18290
    },
    {
      "epoch": 0.46,
      "grad_norm": 4.986574172973633,
      "learning_rate": 7.93220338983051e-06,
      "loss": 1.2345,
      "step": 18300
    },
    {
      "epoch": 0.46,
      "grad_norm": 6.323670864105225,
      "learning_rate": 7.92542372881356e-06,
      "loss": 1.3459,
      "step": 18310
    },
    {
      "epoch": 0.46,
      "grad_norm": 3.625586986541748,
      "learning_rate": 7.91864406779661e-06,
      "loss": 1.4266,
      "step": 18320
    },
    {
      "epoch": 0.46,
      "grad_norm": 11.654215812683105,
      "learning_rate": 7.911864406779661e-06,
      "loss": 1.2831,
      "step": 18330
    },
    {
      "epoch": 0.46,
      "grad_norm": 6.866698265075684,
      "learning_rate": 7.905084745762712e-06,
      "loss": 1.3215,
      "step": 18340
    },
    {
      "epoch": 0.46,
      "grad_norm": 9.294219017028809,
      "learning_rate": 7.898305084745764e-06,
      "loss": 1.2964,
      "step": 18350
    },
    {
      "epoch": 0.46,
      "grad_norm": 5.652096271514893,
      "learning_rate": 7.891525423728814e-06,
      "loss": 1.3691,
      "step": 18360
    },
    {
      "epoch": 0.46,
      "grad_norm": 3.31866192817688,
      "learning_rate": 7.884745762711865e-06,
      "loss": 1.5261,
      "step": 18370
    },
    {
      "epoch": 0.46,
      "grad_norm": 3.0881481170654297,
      "learning_rate": 7.877966101694916e-06,
      "loss": 1.3618,
      "step": 18380
    },
    {
      "epoch": 0.46,
      "grad_norm": 4.305927276611328,
      "learning_rate": 7.871186440677968e-06,
      "loss": 1.5422,
      "step": 18390
    },
    {
      "epoch": 0.46,
      "grad_norm": 11.496697425842285,
      "learning_rate": 7.864406779661017e-06,
      "loss": 1.2874,
      "step": 18400
    },
    {
      "epoch": 0.46,
      "grad_norm": 6.3191070556640625,
      "learning_rate": 7.857627118644069e-06,
      "loss": 1.2414,
      "step": 18410
    },
    {
      "epoch": 0.46,
      "grad_norm": 2.2075703144073486,
      "learning_rate": 7.850847457627119e-06,
      "loss": 1.5212,
      "step": 18420
    },
    {
      "epoch": 0.46,
      "grad_norm": 6.782857418060303,
      "learning_rate": 7.84406779661017e-06,
      "loss": 1.4201,
      "step": 18430
    },
    {
      "epoch": 0.46,
      "grad_norm": 6.062523365020752,
      "learning_rate": 7.837288135593221e-06,
      "loss": 1.4158,
      "step": 18440
    },
    {
      "epoch": 0.46,
      "grad_norm": 3.9009385108947754,
      "learning_rate": 7.830508474576271e-06,
      "loss": 1.2588,
      "step": 18450
    },
    {
      "epoch": 0.46,
      "grad_norm": 5.026491165161133,
      "learning_rate": 7.823728813559322e-06,
      "loss": 1.2953,
      "step": 18460
    },
    {
      "epoch": 0.46,
      "grad_norm": 4.660696029663086,
      "learning_rate": 7.816949152542374e-06,
      "loss": 1.3458,
      "step": 18470
    },
    {
      "epoch": 0.46,
      "grad_norm": 6.407046794891357,
      "learning_rate": 7.810169491525425e-06,
      "loss": 1.5457,
      "step": 18480
    },
    {
      "epoch": 0.46,
      "grad_norm": 2.1768410205841064,
      "learning_rate": 7.803389830508475e-06,
      "loss": 1.4588,
      "step": 18490
    },
    {
      "epoch": 0.46,
      "grad_norm": 6.523290157318115,
      "learning_rate": 7.796610169491526e-06,
      "loss": 1.3751,
      "step": 18500
    },
    {
      "epoch": 0.46,
      "grad_norm": 4.551753520965576,
      "learning_rate": 7.789830508474578e-06,
      "loss": 1.3863,
      "step": 18510
    },
    {
      "epoch": 0.46,
      "grad_norm": 3.3955557346343994,
      "learning_rate": 7.783050847457628e-06,
      "loss": 1.3863,
      "step": 18520
    },
    {
      "epoch": 0.46,
      "grad_norm": 3.465635299682617,
      "learning_rate": 7.776271186440679e-06,
      "loss": 1.5153,
      "step": 18530
    },
    {
      "epoch": 0.46,
      "grad_norm": 5.9394073486328125,
      "learning_rate": 7.769491525423729e-06,
      "loss": 1.434,
      "step": 18540
    },
    {
      "epoch": 0.46,
      "grad_norm": 3.1613309383392334,
      "learning_rate": 7.76271186440678e-06,
      "loss": 1.1107,
      "step": 18550
    },
    {
      "epoch": 0.46,
      "grad_norm": 3.1706387996673584,
      "learning_rate": 7.755932203389831e-06,
      "loss": 1.2724,
      "step": 18560
    },
    {
      "epoch": 0.46,
      "grad_norm": 7.352231025695801,
      "learning_rate": 7.749152542372881e-06,
      "loss": 1.3335,
      "step": 18570
    },
    {
      "epoch": 0.46,
      "grad_norm": 7.294050216674805,
      "learning_rate": 7.742372881355933e-06,
      "loss": 1.2153,
      "step": 18580
    },
    {
      "epoch": 0.46,
      "grad_norm": 10.764616966247559,
      "learning_rate": 7.735593220338984e-06,
      "loss": 1.4353,
      "step": 18590
    },
    {
      "epoch": 0.47,
      "grad_norm": 5.407416343688965,
      "learning_rate": 7.728813559322035e-06,
      "loss": 1.3868,
      "step": 18600
    },
    {
      "epoch": 0.47,
      "grad_norm": 8.108033180236816,
      "learning_rate": 7.722033898305085e-06,
      "loss": 1.0827,
      "step": 18610
    },
    {
      "epoch": 0.47,
      "grad_norm": 7.3071608543396,
      "learning_rate": 7.715254237288136e-06,
      "loss": 1.3612,
      "step": 18620
    },
    {
      "epoch": 0.47,
      "grad_norm": 3.4926538467407227,
      "learning_rate": 7.708474576271186e-06,
      "loss": 1.3849,
      "step": 18630
    },
    {
      "epoch": 0.47,
      "grad_norm": 3.5705881118774414,
      "learning_rate": 7.701694915254238e-06,
      "loss": 1.4653,
      "step": 18640
    },
    {
      "epoch": 0.47,
      "grad_norm": 5.107929229736328,
      "learning_rate": 7.694915254237289e-06,
      "loss": 1.4117,
      "step": 18650
    },
    {
      "epoch": 0.47,
      "grad_norm": 5.992781639099121,
      "learning_rate": 7.688135593220339e-06,
      "loss": 1.3545,
      "step": 18660
    },
    {
      "epoch": 0.47,
      "grad_norm": 3.1811888217926025,
      "learning_rate": 7.68135593220339e-06,
      "loss": 1.3933,
      "step": 18670
    },
    {
      "epoch": 0.47,
      "grad_norm": 2.376875162124634,
      "learning_rate": 7.674576271186441e-06,
      "loss": 1.3486,
      "step": 18680
    },
    {
      "epoch": 0.47,
      "grad_norm": 5.675966739654541,
      "learning_rate": 7.667796610169493e-06,
      "loss": 1.2512,
      "step": 18690
    },
    {
      "epoch": 0.47,
      "grad_norm": 2.2476232051849365,
      "learning_rate": 7.661016949152543e-06,
      "loss": 1.3607,
      "step": 18700
    },
    {
      "epoch": 0.47,
      "grad_norm": 10.595720291137695,
      "learning_rate": 7.654237288135594e-06,
      "loss": 1.2379,
      "step": 18710
    },
    {
      "epoch": 0.47,
      "grad_norm": 3.0006022453308105,
      "learning_rate": 7.647457627118645e-06,
      "loss": 1.4312,
      "step": 18720
    },
    {
      "epoch": 0.47,
      "grad_norm": 7.10250186920166,
      "learning_rate": 7.640677966101695e-06,
      "loss": 1.4699,
      "step": 18730
    },
    {
      "epoch": 0.47,
      "grad_norm": 4.450944900512695,
      "learning_rate": 7.633898305084746e-06,
      "loss": 1.4344,
      "step": 18740
    },
    {
      "epoch": 0.47,
      "grad_norm": 4.79910945892334,
      "learning_rate": 7.627118644067797e-06,
      "loss": 1.3741,
      "step": 18750
    },
    {
      "epoch": 0.47,
      "grad_norm": 14.28135871887207,
      "learning_rate": 7.6203389830508476e-06,
      "loss": 1.4289,
      "step": 18760
    },
    {
      "epoch": 0.47,
      "grad_norm": 9.035604476928711,
      "learning_rate": 7.613559322033899e-06,
      "loss": 1.2811,
      "step": 18770
    },
    {
      "epoch": 0.47,
      "grad_norm": 9.290374755859375,
      "learning_rate": 7.6067796610169495e-06,
      "loss": 1.4093,
      "step": 18780
    },
    {
      "epoch": 0.47,
      "grad_norm": 5.805250644683838,
      "learning_rate": 7.600000000000001e-06,
      "loss": 1.3041,
      "step": 18790
    },
    {
      "epoch": 0.47,
      "grad_norm": 8.446578025817871,
      "learning_rate": 7.5932203389830515e-06,
      "loss": 1.3443,
      "step": 18800
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.7226201295852661,
      "learning_rate": 7.586440677966103e-06,
      "loss": 1.3828,
      "step": 18810
    },
    {
      "epoch": 0.47,
      "grad_norm": 2.5866918563842773,
      "learning_rate": 7.5796610169491534e-06,
      "loss": 1.2293,
      "step": 18820
    },
    {
      "epoch": 0.47,
      "grad_norm": 3.274022340774536,
      "learning_rate": 7.572881355932205e-06,
      "loss": 1.3091,
      "step": 18830
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.421252727508545,
      "learning_rate": 7.5661016949152545e-06,
      "loss": 1.2415,
      "step": 18840
    },
    {
      "epoch": 0.47,
      "grad_norm": 2.9152557849884033,
      "learning_rate": 7.559322033898305e-06,
      "loss": 1.5037,
      "step": 18850
    },
    {
      "epoch": 0.47,
      "grad_norm": 7.876719951629639,
      "learning_rate": 7.5525423728813565e-06,
      "loss": 1.075,
      "step": 18860
    },
    {
      "epoch": 0.47,
      "grad_norm": 7.329843997955322,
      "learning_rate": 7.545762711864407e-06,
      "loss": 1.4631,
      "step": 18870
    },
    {
      "epoch": 0.47,
      "grad_norm": 12.64434814453125,
      "learning_rate": 7.5389830508474584e-06,
      "loss": 1.374,
      "step": 18880
    },
    {
      "epoch": 0.47,
      "grad_norm": 6.557058334350586,
      "learning_rate": 7.532203389830509e-06,
      "loss": 1.3634,
      "step": 18890
    },
    {
      "epoch": 0.47,
      "grad_norm": 5.131541728973389,
      "learning_rate": 7.52542372881356e-06,
      "loss": 1.4829,
      "step": 18900
    },
    {
      "epoch": 0.47,
      "grad_norm": 5.675949573516846,
      "learning_rate": 7.518644067796611e-06,
      "loss": 1.278,
      "step": 18910
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.365838885307312,
      "learning_rate": 7.511864406779662e-06,
      "loss": 1.3541,
      "step": 18920
    },
    {
      "epoch": 0.47,
      "grad_norm": 9.97677230834961,
      "learning_rate": 7.505084745762713e-06,
      "loss": 1.3591,
      "step": 18930
    },
    {
      "epoch": 0.47,
      "grad_norm": 4.036107540130615,
      "learning_rate": 7.498305084745763e-06,
      "loss": 1.2616,
      "step": 18940
    },
    {
      "epoch": 0.47,
      "grad_norm": 4.973388195037842,
      "learning_rate": 7.491525423728814e-06,
      "loss": 1.2,
      "step": 18950
    },
    {
      "epoch": 0.47,
      "grad_norm": 7.6102752685546875,
      "learning_rate": 7.4847457627118646e-06,
      "loss": 1.2785,
      "step": 18960
    },
    {
      "epoch": 0.47,
      "grad_norm": 7.575082778930664,
      "learning_rate": 7.477966101694916e-06,
      "loss": 1.336,
      "step": 18970
    },
    {
      "epoch": 0.47,
      "grad_norm": 4.849128246307373,
      "learning_rate": 7.4711864406779665e-06,
      "loss": 1.3286,
      "step": 18980
    },
    {
      "epoch": 0.47,
      "grad_norm": 4.723095417022705,
      "learning_rate": 7.464406779661018e-06,
      "loss": 1.409,
      "step": 18990
    },
    {
      "epoch": 0.47,
      "grad_norm": 8.499979972839355,
      "learning_rate": 7.4576271186440685e-06,
      "loss": 1.4162,
      "step": 19000
    },
    {
      "epoch": 0.48,
      "grad_norm": 4.321350574493408,
      "learning_rate": 7.45084745762712e-06,
      "loss": 1.271,
      "step": 19010
    },
    {
      "epoch": 0.48,
      "grad_norm": 3.6212425231933594,
      "learning_rate": 7.4440677966101704e-06,
      "loss": 1.3802,
      "step": 19020
    },
    {
      "epoch": 0.48,
      "grad_norm": 4.067981243133545,
      "learning_rate": 7.437288135593221e-06,
      "loss": 1.4292,
      "step": 19030
    },
    {
      "epoch": 0.48,
      "grad_norm": 12.705096244812012,
      "learning_rate": 7.430508474576272e-06,
      "loss": 1.2203,
      "step": 19040
    },
    {
      "epoch": 0.48,
      "grad_norm": 5.270137310028076,
      "learning_rate": 7.423728813559322e-06,
      "loss": 1.3766,
      "step": 19050
    },
    {
      "epoch": 0.48,
      "grad_norm": 6.715325355529785,
      "learning_rate": 7.4169491525423735e-06,
      "loss": 1.362,
      "step": 19060
    },
    {
      "epoch": 0.48,
      "grad_norm": 6.584071636199951,
      "learning_rate": 7.410169491525424e-06,
      "loss": 1.4266,
      "step": 19070
    },
    {
      "epoch": 0.48,
      "grad_norm": 2.778653860092163,
      "learning_rate": 7.4033898305084754e-06,
      "loss": 1.4062,
      "step": 19080
    },
    {
      "epoch": 0.48,
      "grad_norm": 2.6927425861358643,
      "learning_rate": 7.396610169491526e-06,
      "loss": 1.4085,
      "step": 19090
    },
    {
      "epoch": 0.48,
      "grad_norm": 6.936356544494629,
      "learning_rate": 7.3898305084745766e-06,
      "loss": 1.3418,
      "step": 19100
    },
    {
      "epoch": 0.48,
      "grad_norm": 11.453264236450195,
      "learning_rate": 7.383050847457628e-06,
      "loss": 1.341,
      "step": 19110
    },
    {
      "epoch": 0.48,
      "grad_norm": 4.236923694610596,
      "learning_rate": 7.3762711864406785e-06,
      "loss": 1.4257,
      "step": 19120
    },
    {
      "epoch": 0.48,
      "grad_norm": 9.215226173400879,
      "learning_rate": 7.36949152542373e-06,
      "loss": 1.2916,
      "step": 19130
    },
    {
      "epoch": 0.48,
      "grad_norm": 8.170127868652344,
      "learning_rate": 7.3627118644067805e-06,
      "loss": 1.2401,
      "step": 19140
    },
    {
      "epoch": 0.48,
      "grad_norm": 3.121269702911377,
      "learning_rate": 7.355932203389831e-06,
      "loss": 1.353,
      "step": 19150
    },
    {
      "epoch": 0.48,
      "grad_norm": 5.2461724281311035,
      "learning_rate": 7.3491525423728816e-06,
      "loss": 1.2722,
      "step": 19160
    },
    {
      "epoch": 0.48,
      "grad_norm": 7.478322982788086,
      "learning_rate": 7.342372881355932e-06,
      "loss": 1.3965,
      "step": 19170
    },
    {
      "epoch": 0.48,
      "grad_norm": 4.277684211730957,
      "learning_rate": 7.3355932203389835e-06,
      "loss": 1.4419,
      "step": 19180
    },
    {
      "epoch": 0.48,
      "grad_norm": 17.70556640625,
      "learning_rate": 7.328813559322034e-06,
      "loss": 1.3851,
      "step": 19190
    },
    {
      "epoch": 0.48,
      "grad_norm": 9.973118782043457,
      "learning_rate": 7.3220338983050855e-06,
      "loss": 1.2506,
      "step": 19200
    },
    {
      "epoch": 0.48,
      "grad_norm": 2.630134105682373,
      "learning_rate": 7.315254237288136e-06,
      "loss": 1.3592,
      "step": 19210
    },
    {
      "epoch": 0.48,
      "grad_norm": 7.502743721008301,
      "learning_rate": 7.3084745762711874e-06,
      "loss": 1.3197,
      "step": 19220
    },
    {
      "epoch": 0.48,
      "grad_norm": 5.648642539978027,
      "learning_rate": 7.301694915254238e-06,
      "loss": 1.2157,
      "step": 19230
    },
    {
      "epoch": 0.48,
      "grad_norm": 8.529255867004395,
      "learning_rate": 7.294915254237289e-06,
      "loss": 1.3086,
      "step": 19240
    },
    {
      "epoch": 0.48,
      "grad_norm": 5.723903179168701,
      "learning_rate": 7.288135593220339e-06,
      "loss": 1.4993,
      "step": 19250
    },
    {
      "epoch": 0.48,
      "grad_norm": 2.296700954437256,
      "learning_rate": 7.28135593220339e-06,
      "loss": 1.4155,
      "step": 19260
    },
    {
      "epoch": 0.48,
      "grad_norm": 7.792112350463867,
      "learning_rate": 7.274576271186441e-06,
      "loss": 1.3285,
      "step": 19270
    },
    {
      "epoch": 0.48,
      "grad_norm": 4.722753047943115,
      "learning_rate": 7.267796610169492e-06,
      "loss": 1.3531,
      "step": 19280
    },
    {
      "epoch": 0.48,
      "grad_norm": 10.849143028259277,
      "learning_rate": 7.261016949152543e-06,
      "loss": 1.3917,
      "step": 19290
    },
    {
      "epoch": 0.48,
      "grad_norm": 6.434514999389648,
      "learning_rate": 7.2542372881355936e-06,
      "loss": 1.5461,
      "step": 19300
    },
    {
      "epoch": 0.48,
      "grad_norm": 5.793635845184326,
      "learning_rate": 7.247457627118645e-06,
      "loss": 1.1435,
      "step": 19310
    },
    {
      "epoch": 0.48,
      "grad_norm": 5.777005195617676,
      "learning_rate": 7.2406779661016955e-06,
      "loss": 1.2511,
      "step": 19320
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.983711838722229,
      "learning_rate": 7.233898305084747e-06,
      "loss": 1.2053,
      "step": 19330
    },
    {
      "epoch": 0.48,
      "grad_norm": 3.716196298599243,
      "learning_rate": 7.2271186440677975e-06,
      "loss": 1.2022,
      "step": 19340
    },
    {
      "epoch": 0.48,
      "grad_norm": 2.878067970275879,
      "learning_rate": 7.220338983050849e-06,
      "loss": 1.498,
      "step": 19350
    },
    {
      "epoch": 0.48,
      "grad_norm": 3.5594191551208496,
      "learning_rate": 7.2135593220338986e-06,
      "loss": 1.4644,
      "step": 19360
    },
    {
      "epoch": 0.48,
      "grad_norm": 3.587618827819824,
      "learning_rate": 7.206779661016949e-06,
      "loss": 1.119,
      "step": 19370
    },
    {
      "epoch": 0.48,
      "grad_norm": 6.6539716720581055,
      "learning_rate": 7.2000000000000005e-06,
      "loss": 1.4482,
      "step": 19380
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.7582392692565918,
      "learning_rate": 7.193220338983051e-06,
      "loss": 1.4473,
      "step": 19390
    },
    {
      "epoch": 0.48,
      "grad_norm": 5.808141231536865,
      "learning_rate": 7.1864406779661025e-06,
      "loss": 1.3603,
      "step": 19400
    },
    {
      "epoch": 0.49,
      "grad_norm": 5.359313488006592,
      "learning_rate": 7.179661016949153e-06,
      "loss": 1.3978,
      "step": 19410
    },
    {
      "epoch": 0.49,
      "grad_norm": 3.27229380607605,
      "learning_rate": 7.1728813559322044e-06,
      "loss": 1.2838,
      "step": 19420
    },
    {
      "epoch": 0.49,
      "grad_norm": 6.635318756103516,
      "learning_rate": 7.166101694915255e-06,
      "loss": 1.5034,
      "step": 19430
    },
    {
      "epoch": 0.49,
      "grad_norm": 6.988661766052246,
      "learning_rate": 7.159322033898306e-06,
      "loss": 1.4469,
      "step": 19440
    },
    {
      "epoch": 0.49,
      "grad_norm": 11.993741989135742,
      "learning_rate": 7.152542372881357e-06,
      "loss": 1.285,
      "step": 19450
    },
    {
      "epoch": 0.49,
      "grad_norm": 2.6057329177856445,
      "learning_rate": 7.145762711864407e-06,
      "loss": 1.394,
      "step": 19460
    },
    {
      "epoch": 0.49,
      "grad_norm": 3.688753128051758,
      "learning_rate": 7.138983050847458e-06,
      "loss": 1.3991,
      "step": 19470
    },
    {
      "epoch": 0.49,
      "grad_norm": 5.323999404907227,
      "learning_rate": 7.132203389830509e-06,
      "loss": 1.2589,
      "step": 19480
    },
    {
      "epoch": 0.49,
      "grad_norm": 3.189335823059082,
      "learning_rate": 7.12542372881356e-06,
      "loss": 1.2682,
      "step": 19490
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.8446924686431885,
      "learning_rate": 7.1186440677966106e-06,
      "loss": 1.2333,
      "step": 19500
    },
    {
      "epoch": 0.49,
      "grad_norm": 2.597740411758423,
      "learning_rate": 7.111864406779662e-06,
      "loss": 1.3978,
      "step": 19510
    },
    {
      "epoch": 0.49,
      "grad_norm": 6.010899543762207,
      "learning_rate": 7.1050847457627125e-06,
      "loss": 1.4077,
      "step": 19520
    },
    {
      "epoch": 0.49,
      "grad_norm": 4.375614643096924,
      "learning_rate": 7.098305084745764e-06,
      "loss": 1.2913,
      "step": 19530
    },
    {
      "epoch": 0.49,
      "grad_norm": 3.3142616748809814,
      "learning_rate": 7.0915254237288145e-06,
      "loss": 1.3794,
      "step": 19540
    },
    {
      "epoch": 0.49,
      "grad_norm": 3.074887275695801,
      "learning_rate": 7.084745762711865e-06,
      "loss": 1.3545,
      "step": 19550
    },
    {
      "epoch": 0.49,
      "grad_norm": 2.700955390930176,
      "learning_rate": 7.077966101694916e-06,
      "loss": 1.2601,
      "step": 19560
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.4296528100967407,
      "learning_rate": 7.071186440677966e-06,
      "loss": 1.2389,
      "step": 19570
    },
    {
      "epoch": 0.49,
      "grad_norm": 7.203572750091553,
      "learning_rate": 7.0644067796610175e-06,
      "loss": 1.2697,
      "step": 19580
    },
    {
      "epoch": 0.49,
      "grad_norm": 4.294157028198242,
      "learning_rate": 7.057627118644068e-06,
      "loss": 1.2931,
      "step": 19590
    },
    {
      "epoch": 0.49,
      "grad_norm": 10.261987686157227,
      "learning_rate": 7.0508474576271195e-06,
      "loss": 1.295,
      "step": 19600
    },
    {
      "epoch": 0.49,
      "grad_norm": 4.492446422576904,
      "learning_rate": 7.04406779661017e-06,
      "loss": 1.229,
      "step": 19610
    },
    {
      "epoch": 0.49,
      "grad_norm": 2.3853440284729004,
      "learning_rate": 7.037288135593221e-06,
      "loss": 1.3616,
      "step": 19620
    },
    {
      "epoch": 0.49,
      "grad_norm": 6.414808750152588,
      "learning_rate": 7.030508474576272e-06,
      "loss": 1.2847,
      "step": 19630
    },
    {
      "epoch": 0.49,
      "grad_norm": 7.093498706817627,
      "learning_rate": 7.0237288135593225e-06,
      "loss": 1.3191,
      "step": 19640
    },
    {
      "epoch": 0.49,
      "grad_norm": 2.8542966842651367,
      "learning_rate": 7.016949152542374e-06,
      "loss": 1.1424,
      "step": 19650
    },
    {
      "epoch": 0.49,
      "grad_norm": 3.1772007942199707,
      "learning_rate": 7.0101694915254245e-06,
      "loss": 1.2958,
      "step": 19660
    },
    {
      "epoch": 0.49,
      "grad_norm": 8.018784523010254,
      "learning_rate": 7.003389830508475e-06,
      "loss": 1.2579,
      "step": 19670
    },
    {
      "epoch": 0.49,
      "grad_norm": 13.49968147277832,
      "learning_rate": 6.996610169491526e-06,
      "loss": 1.3161,
      "step": 19680
    },
    {
      "epoch": 0.49,
      "grad_norm": 4.020516395568848,
      "learning_rate": 6.989830508474576e-06,
      "loss": 1.3075,
      "step": 19690
    },
    {
      "epoch": 0.49,
      "grad_norm": 6.960037708282471,
      "learning_rate": 6.9830508474576275e-06,
      "loss": 1.3718,
      "step": 19700
    },
    {
      "epoch": 0.49,
      "grad_norm": 6.919734954833984,
      "learning_rate": 6.976271186440678e-06,
      "loss": 1.5833,
      "step": 19710
    },
    {
      "epoch": 0.49,
      "grad_norm": 4.464432716369629,
      "learning_rate": 6.9694915254237295e-06,
      "loss": 1.3472,
      "step": 19720
    },
    {
      "epoch": 0.49,
      "grad_norm": 9.147382736206055,
      "learning_rate": 6.96271186440678e-06,
      "loss": 1.3751,
      "step": 19730
    },
    {
      "epoch": 0.49,
      "grad_norm": 3.2477524280548096,
      "learning_rate": 6.9559322033898315e-06,
      "loss": 1.3522,
      "step": 19740
    },
    {
      "epoch": 0.49,
      "grad_norm": 12.42885684967041,
      "learning_rate": 6.949152542372882e-06,
      "loss": 1.4472,
      "step": 19750
    },
    {
      "epoch": 0.49,
      "grad_norm": 3.450083017349243,
      "learning_rate": 6.942372881355933e-06,
      "loss": 1.3504,
      "step": 19760
    },
    {
      "epoch": 0.49,
      "grad_norm": 8.696680068969727,
      "learning_rate": 6.935593220338983e-06,
      "loss": 1.4917,
      "step": 19770
    },
    {
      "epoch": 0.49,
      "grad_norm": 4.729881763458252,
      "learning_rate": 6.928813559322034e-06,
      "loss": 1.3966,
      "step": 19780
    },
    {
      "epoch": 0.49,
      "grad_norm": 12.553901672363281,
      "learning_rate": 6.922033898305085e-06,
      "loss": 1.1048,
      "step": 19790
    },
    {
      "epoch": 0.49,
      "grad_norm": 10.372455596923828,
      "learning_rate": 6.915254237288136e-06,
      "loss": 1.5142,
      "step": 19800
    },
    {
      "epoch": 0.5,
      "grad_norm": 5.72446346282959,
      "learning_rate": 6.908474576271187e-06,
      "loss": 1.3617,
      "step": 19810
    },
    {
      "epoch": 0.5,
      "grad_norm": 7.227321147918701,
      "learning_rate": 6.901694915254238e-06,
      "loss": 1.2474,
      "step": 19820
    },
    {
      "epoch": 0.5,
      "grad_norm": 4.264225006103516,
      "learning_rate": 6.894915254237289e-06,
      "loss": 1.1393,
      "step": 19830
    },
    {
      "epoch": 0.5,
      "grad_norm": 2.9691781997680664,
      "learning_rate": 6.8881355932203395e-06,
      "loss": 1.209,
      "step": 19840
    },
    {
      "epoch": 0.5,
      "grad_norm": 4.267544269561768,
      "learning_rate": 6.881355932203391e-06,
      "loss": 1.3269,
      "step": 19850
    },
    {
      "epoch": 0.5,
      "grad_norm": 9.143953323364258,
      "learning_rate": 6.8745762711864415e-06,
      "loss": 1.1848,
      "step": 19860
    },
    {
      "epoch": 0.5,
      "grad_norm": 3.567173719406128,
      "learning_rate": 6.867796610169493e-06,
      "loss": 1.3936,
      "step": 19870
    },
    {
      "epoch": 0.5,
      "grad_norm": 7.46988582611084,
      "learning_rate": 6.861016949152543e-06,
      "loss": 1.2892,
      "step": 19880
    },
    {
      "epoch": 0.5,
      "grad_norm": 3.349632978439331,
      "learning_rate": 6.854237288135593e-06,
      "loss": 1.4199,
      "step": 19890
    },
    {
      "epoch": 0.5,
      "grad_norm": 7.5824713706970215,
      "learning_rate": 6.8474576271186445e-06,
      "loss": 1.4164,
      "step": 19900
    },
    {
      "epoch": 0.5,
      "grad_norm": 9.003057479858398,
      "learning_rate": 6.840677966101695e-06,
      "loss": 1.3074,
      "step": 19910
    },
    {
      "epoch": 0.5,
      "grad_norm": 3.6586782932281494,
      "learning_rate": 6.8338983050847465e-06,
      "loss": 1.3554,
      "step": 19920
    },
    {
      "epoch": 0.5,
      "grad_norm": 5.403897285461426,
      "learning_rate": 6.827118644067797e-06,
      "loss": 1.4533,
      "step": 19930
    },
    {
      "epoch": 0.5,
      "grad_norm": 3.7970988750457764,
      "learning_rate": 6.8203389830508485e-06,
      "loss": 1.3033,
      "step": 19940
    },
    {
      "epoch": 0.5,
      "grad_norm": 5.425084590911865,
      "learning_rate": 6.813559322033899e-06,
      "loss": 1.2651,
      "step": 19950
    },
    {
      "epoch": 0.5,
      "grad_norm": 4.225399017333984,
      "learning_rate": 6.80677966101695e-06,
      "loss": 1.4934,
      "step": 19960
    },
    {
      "epoch": 0.5,
      "grad_norm": 6.519874572753906,
      "learning_rate": 6.800000000000001e-06,
      "loss": 1.434,
      "step": 19970
    },
    {
      "epoch": 0.5,
      "grad_norm": 28.235916137695312,
      "learning_rate": 6.793220338983051e-06,
      "loss": 1.4958,
      "step": 19980
    },
    {
      "epoch": 0.5,
      "grad_norm": 9.091018676757812,
      "learning_rate": 6.786440677966102e-06,
      "loss": 1.4321,
      "step": 19990
    },
    {
      "epoch": 0.5,
      "grad_norm": 4.438204288482666,
      "learning_rate": 6.779661016949153e-06,
      "loss": 1.4244,
      "step": 20000
    },
    {
      "epoch": 0.5,
      "eval_loss": 1.3505966663360596,
      "eval_runtime": 122.3846,
      "eval_samples_per_second": 8.171,
      "eval_steps_per_second": 8.171,
      "step": 20000
    },
    {
      "epoch": 0.5,
      "grad_norm": 3.0722553730010986,
      "learning_rate": 6.772881355932204e-06,
      "loss": 1.4289,
      "step": 20010
    },
    {
      "epoch": 0.5,
      "grad_norm": 9.793889045715332,
      "learning_rate": 6.766101694915255e-06,
      "loss": 1.4726,
      "step": 20020
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.4246736764907837,
      "learning_rate": 6.759322033898306e-06,
      "loss": 1.3273,
      "step": 20030
    },
    {
      "epoch": 0.5,
      "grad_norm": 6.397233009338379,
      "learning_rate": 6.7525423728813565e-06,
      "loss": 1.316,
      "step": 20040
    },
    {
      "epoch": 0.5,
      "grad_norm": 11.257098197937012,
      "learning_rate": 6.745762711864408e-06,
      "loss": 1.5277,
      "step": 20050
    },
    {
      "epoch": 0.5,
      "grad_norm": 7.300049781799316,
      "learning_rate": 6.7389830508474585e-06,
      "loss": 1.2847,
      "step": 20060
    },
    {
      "epoch": 0.5,
      "grad_norm": 6.414507865905762,
      "learning_rate": 6.73220338983051e-06,
      "loss": 1.3158,
      "step": 20070
    },
    {
      "epoch": 0.5,
      "grad_norm": 5.005466461181641,
      "learning_rate": 6.7254237288135604e-06,
      "loss": 1.4015,
      "step": 20080
    },
    {
      "epoch": 0.5,
      "grad_norm": 3.0672664642333984,
      "learning_rate": 6.71864406779661e-06,
      "loss": 1.368,
      "step": 20090
    },
    {
      "epoch": 0.5,
      "grad_norm": 2.401742935180664,
      "learning_rate": 6.7118644067796615e-06,
      "loss": 1.5319,
      "step": 20100
    },
    {
      "epoch": 0.5,
      "grad_norm": 5.489325523376465,
      "learning_rate": 6.705084745762712e-06,
      "loss": 1.4877,
      "step": 20110
    },
    {
      "epoch": 0.5,
      "grad_norm": 9.200358390808105,
      "learning_rate": 6.6983050847457635e-06,
      "loss": 1.3385,
      "step": 20120
    },
    {
      "epoch": 0.5,
      "grad_norm": 4.872713565826416,
      "learning_rate": 6.691525423728814e-06,
      "loss": 1.3536,
      "step": 20130
    },
    {
      "epoch": 0.5,
      "grad_norm": 3.130598545074463,
      "learning_rate": 6.6847457627118655e-06,
      "loss": 1.4881,
      "step": 20140
    },
    {
      "epoch": 0.5,
      "grad_norm": 3.361140251159668,
      "learning_rate": 6.677966101694916e-06,
      "loss": 1.4141,
      "step": 20150
    },
    {
      "epoch": 0.5,
      "grad_norm": 11.690096855163574,
      "learning_rate": 6.6711864406779666e-06,
      "loss": 1.3506,
      "step": 20160
    },
    {
      "epoch": 0.5,
      "grad_norm": 10.660921096801758,
      "learning_rate": 6.664406779661018e-06,
      "loss": 1.365,
      "step": 20170
    },
    {
      "epoch": 0.5,
      "grad_norm": 11.015972137451172,
      "learning_rate": 6.6576271186440685e-06,
      "loss": 1.4309,
      "step": 20180
    },
    {
      "epoch": 0.5,
      "grad_norm": 11.133783340454102,
      "learning_rate": 6.650847457627119e-06,
      "loss": 1.3618,
      "step": 20190
    },
    {
      "epoch": 0.51,
      "grad_norm": 4.559747695922852,
      "learning_rate": 6.64406779661017e-06,
      "loss": 1.3693,
      "step": 20200
    },
    {
      "epoch": 0.51,
      "grad_norm": 4.533148288726807,
      "learning_rate": 6.637288135593221e-06,
      "loss": 1.4986,
      "step": 20210
    },
    {
      "epoch": 0.51,
      "grad_norm": 5.510108470916748,
      "learning_rate": 6.6305084745762716e-06,
      "loss": 1.3255,
      "step": 20220
    },
    {
      "epoch": 0.51,
      "grad_norm": 5.735023021697998,
      "learning_rate": 6.623728813559322e-06,
      "loss": 1.4119,
      "step": 20230
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.9268147945404053,
      "learning_rate": 6.6169491525423735e-06,
      "loss": 1.4887,
      "step": 20240
    },
    {
      "epoch": 0.51,
      "grad_norm": 8.361649513244629,
      "learning_rate": 6.610169491525424e-06,
      "loss": 1.4613,
      "step": 20250
    },
    {
      "epoch": 0.51,
      "grad_norm": 4.189188480377197,
      "learning_rate": 6.6033898305084755e-06,
      "loss": 1.3955,
      "step": 20260
    },
    {
      "epoch": 0.51,
      "grad_norm": 24.131383895874023,
      "learning_rate": 6.596610169491526e-06,
      "loss": 1.309,
      "step": 20270
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.8833867311477661,
      "learning_rate": 6.5898305084745774e-06,
      "loss": 1.4364,
      "step": 20280
    },
    {
      "epoch": 0.51,
      "grad_norm": 9.247352600097656,
      "learning_rate": 6.583050847457627e-06,
      "loss": 1.2739,
      "step": 20290
    },
    {
      "epoch": 0.51,
      "grad_norm": 12.675292015075684,
      "learning_rate": 6.576271186440678e-06,
      "loss": 1.3688,
      "step": 20300
    },
    {
      "epoch": 0.51,
      "grad_norm": 3.1554551124572754,
      "learning_rate": 6.569491525423729e-06,
      "loss": 1.4123,
      "step": 20310
    },
    {
      "epoch": 0.51,
      "grad_norm": 8.197525024414062,
      "learning_rate": 6.56271186440678e-06,
      "loss": 1.4314,
      "step": 20320
    },
    {
      "epoch": 0.51,
      "grad_norm": 3.4080681800842285,
      "learning_rate": 6.555932203389831e-06,
      "loss": 1.4029,
      "step": 20330
    },
    {
      "epoch": 0.51,
      "grad_norm": 5.85288143157959,
      "learning_rate": 6.549152542372882e-06,
      "loss": 1.3295,
      "step": 20340
    },
    {
      "epoch": 0.51,
      "grad_norm": 4.488935947418213,
      "learning_rate": 6.542372881355933e-06,
      "loss": 1.3361,
      "step": 20350
    },
    {
      "epoch": 0.51,
      "grad_norm": 3.925982713699341,
      "learning_rate": 6.5355932203389836e-06,
      "loss": 1.355,
      "step": 20360
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.382690191268921,
      "learning_rate": 6.528813559322035e-06,
      "loss": 1.3186,
      "step": 20370
    },
    {
      "epoch": 0.51,
      "grad_norm": 4.629077434539795,
      "learning_rate": 6.5220338983050855e-06,
      "loss": 1.3272,
      "step": 20380
    },
    {
      "epoch": 0.51,
      "grad_norm": 4.436100482940674,
      "learning_rate": 6.515254237288137e-06,
      "loss": 1.3382,
      "step": 20390
    },
    {
      "epoch": 0.51,
      "grad_norm": 3.2458138465881348,
      "learning_rate": 6.508474576271187e-06,
      "loss": 1.2103,
      "step": 20400
    },
    {
      "epoch": 0.51,
      "grad_norm": 3.328113555908203,
      "learning_rate": 6.501694915254237e-06,
      "loss": 1.33,
      "step": 20410
    },
    {
      "epoch": 0.51,
      "grad_norm": 4.0282135009765625,
      "learning_rate": 6.4949152542372886e-06,
      "loss": 1.4258,
      "step": 20420
    },
    {
      "epoch": 0.51,
      "grad_norm": 6.514222621917725,
      "learning_rate": 6.488135593220339e-06,
      "loss": 1.353,
      "step": 20430
    },
    {
      "epoch": 0.51,
      "grad_norm": 8.407407760620117,
      "learning_rate": 6.4813559322033905e-06,
      "loss": 1.4448,
      "step": 20440
    },
    {
      "epoch": 0.51,
      "grad_norm": 8.151182174682617,
      "learning_rate": 6.474576271186441e-06,
      "loss": 1.4544,
      "step": 20450
    },
    {
      "epoch": 0.51,
      "grad_norm": 8.823492050170898,
      "learning_rate": 6.4677966101694925e-06,
      "loss": 1.2728,
      "step": 20460
    },
    {
      "epoch": 0.51,
      "grad_norm": 4.913995742797852,
      "learning_rate": 6.461016949152543e-06,
      "loss": 1.4342,
      "step": 20470
    },
    {
      "epoch": 0.51,
      "grad_norm": 5.866121292114258,
      "learning_rate": 6.4542372881355944e-06,
      "loss": 1.3472,
      "step": 20480
    },
    {
      "epoch": 0.51,
      "grad_norm": 2.4927186965942383,
      "learning_rate": 6.447457627118645e-06,
      "loss": 1.4063,
      "step": 20490
    },
    {
      "epoch": 0.51,
      "grad_norm": 4.486921787261963,
      "learning_rate": 6.440677966101695e-06,
      "loss": 1.4572,
      "step": 20500
    },
    {
      "epoch": 0.51,
      "grad_norm": 21.873701095581055,
      "learning_rate": 6.433898305084746e-06,
      "loss": 1.3167,
      "step": 20510
    },
    {
      "epoch": 0.51,
      "grad_norm": 8.287447929382324,
      "learning_rate": 6.427118644067797e-06,
      "loss": 1.3568,
      "step": 20520
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.3230948448181152,
      "learning_rate": 6.420338983050848e-06,
      "loss": 1.3569,
      "step": 20530
    },
    {
      "epoch": 0.51,
      "grad_norm": 6.885702610015869,
      "learning_rate": 6.413559322033899e-06,
      "loss": 1.2216,
      "step": 20540
    },
    {
      "epoch": 0.51,
      "grad_norm": 13.705862998962402,
      "learning_rate": 6.40677966101695e-06,
      "loss": 1.191,
      "step": 20550
    },
    {
      "epoch": 0.51,
      "grad_norm": 2.434264659881592,
      "learning_rate": 6.4000000000000006e-06,
      "loss": 1.2421,
      "step": 20560
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.7716752290725708,
      "learning_rate": 6.393220338983052e-06,
      "loss": 1.3514,
      "step": 20570
    },
    {
      "epoch": 0.51,
      "grad_norm": 7.752338409423828,
      "learning_rate": 6.3864406779661025e-06,
      "loss": 1.2612,
      "step": 20580
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.8300265073776245,
      "learning_rate": 6.379661016949154e-06,
      "loss": 1.2573,
      "step": 20590
    },
    {
      "epoch": 0.52,
      "grad_norm": 5.662349700927734,
      "learning_rate": 6.372881355932204e-06,
      "loss": 1.1762,
      "step": 20600
    },
    {
      "epoch": 0.52,
      "grad_norm": 9.122199058532715,
      "learning_rate": 6.366101694915254e-06,
      "loss": 1.3679,
      "step": 20610
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.3133958578109741,
      "learning_rate": 6.3593220338983056e-06,
      "loss": 1.274,
      "step": 20620
    },
    {
      "epoch": 0.52,
      "grad_norm": 7.175881385803223,
      "learning_rate": 6.352542372881356e-06,
      "loss": 1.5124,
      "step": 20630
    },
    {
      "epoch": 0.52,
      "grad_norm": 11.994564056396484,
      "learning_rate": 6.3457627118644075e-06,
      "loss": 1.251,
      "step": 20640
    },
    {
      "epoch": 0.52,
      "grad_norm": 2.729207992553711,
      "learning_rate": 6.338983050847458e-06,
      "loss": 1.4331,
      "step": 20650
    },
    {
      "epoch": 0.52,
      "grad_norm": 2.5460658073425293,
      "learning_rate": 6.3322033898305095e-06,
      "loss": 1.3789,
      "step": 20660
    },
    {
      "epoch": 0.52,
      "grad_norm": 6.080118656158447,
      "learning_rate": 6.32542372881356e-06,
      "loss": 1.2998,
      "step": 20670
    },
    {
      "epoch": 0.52,
      "grad_norm": 21.435016632080078,
      "learning_rate": 6.318644067796611e-06,
      "loss": 1.4741,
      "step": 20680
    },
    {
      "epoch": 0.52,
      "grad_norm": 3.296051025390625,
      "learning_rate": 6.311864406779662e-06,
      "loss": 1.329,
      "step": 20690
    },
    {
      "epoch": 0.52,
      "grad_norm": 7.729842662811279,
      "learning_rate": 6.3050847457627125e-06,
      "loss": 1.4129,
      "step": 20700
    },
    {
      "epoch": 0.52,
      "grad_norm": 6.633853912353516,
      "learning_rate": 6.298305084745763e-06,
      "loss": 1.4395,
      "step": 20710
    },
    {
      "epoch": 0.52,
      "grad_norm": 11.433235168457031,
      "learning_rate": 6.291525423728814e-06,
      "loss": 1.347,
      "step": 20720
    },
    {
      "epoch": 0.52,
      "grad_norm": 10.14087963104248,
      "learning_rate": 6.284745762711865e-06,
      "loss": 1.3332,
      "step": 20730
    },
    {
      "epoch": 0.52,
      "grad_norm": 4.155704975128174,
      "learning_rate": 6.277966101694916e-06,
      "loss": 1.5413,
      "step": 20740
    },
    {
      "epoch": 0.52,
      "grad_norm": 9.64389705657959,
      "learning_rate": 6.271186440677966e-06,
      "loss": 1.0952,
      "step": 20750
    },
    {
      "epoch": 0.52,
      "grad_norm": 6.665346145629883,
      "learning_rate": 6.2644067796610176e-06,
      "loss": 1.2514,
      "step": 20760
    },
    {
      "epoch": 0.52,
      "grad_norm": 3.899184465408325,
      "learning_rate": 6.257627118644068e-06,
      "loss": 1.3202,
      "step": 20770
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.938454508781433,
      "learning_rate": 6.2508474576271195e-06,
      "loss": 1.3795,
      "step": 20780
    },
    {
      "epoch": 0.52,
      "grad_norm": 3.743913412094116,
      "learning_rate": 6.24406779661017e-06,
      "loss": 1.3904,
      "step": 20790
    },
    {
      "epoch": 0.52,
      "grad_norm": 2.653268814086914,
      "learning_rate": 6.2372881355932215e-06,
      "loss": 1.457,
      "step": 20800
    },
    {
      "epoch": 0.52,
      "grad_norm": 9.194693565368652,
      "learning_rate": 6.230508474576271e-06,
      "loss": 1.337,
      "step": 20810
    },
    {
      "epoch": 0.52,
      "grad_norm": 3.662729024887085,
      "learning_rate": 6.223728813559322e-06,
      "loss": 1.0587,
      "step": 20820
    },
    {
      "epoch": 0.52,
      "grad_norm": 4.549154758453369,
      "learning_rate": 6.216949152542373e-06,
      "loss": 1.3009,
      "step": 20830
    },
    {
      "epoch": 0.52,
      "grad_norm": 8.112665176391602,
      "learning_rate": 6.210169491525424e-06,
      "loss": 1.3788,
      "step": 20840
    },
    {
      "epoch": 0.52,
      "grad_norm": 11.637565612792969,
      "learning_rate": 6.203389830508475e-06,
      "loss": 1.5249,
      "step": 20850
    },
    {
      "epoch": 0.52,
      "grad_norm": 2.3856558799743652,
      "learning_rate": 6.196610169491526e-06,
      "loss": 1.3971,
      "step": 20860
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.4000920057296753,
      "learning_rate": 6.189830508474577e-06,
      "loss": 1.2676,
      "step": 20870
    },
    {
      "epoch": 0.52,
      "grad_norm": 6.194108486175537,
      "learning_rate": 6.183050847457628e-06,
      "loss": 1.4893,
      "step": 20880
    },
    {
      "epoch": 0.52,
      "grad_norm": 2.3196473121643066,
      "learning_rate": 6.176271186440679e-06,
      "loss": 1.443,
      "step": 20890
    },
    {
      "epoch": 0.52,
      "grad_norm": 3.8083395957946777,
      "learning_rate": 6.1694915254237295e-06,
      "loss": 1.3592,
      "step": 20900
    },
    {
      "epoch": 0.52,
      "grad_norm": 8.59618854522705,
      "learning_rate": 6.162711864406781e-06,
      "loss": 1.3369,
      "step": 20910
    },
    {
      "epoch": 0.52,
      "grad_norm": 5.316215991973877,
      "learning_rate": 6.155932203389831e-06,
      "loss": 1.3355,
      "step": 20920
    },
    {
      "epoch": 0.52,
      "grad_norm": 9.462787628173828,
      "learning_rate": 6.149152542372881e-06,
      "loss": 1.4838,
      "step": 20930
    },
    {
      "epoch": 0.52,
      "grad_norm": 4.59434700012207,
      "learning_rate": 6.142372881355933e-06,
      "loss": 1.5266,
      "step": 20940
    },
    {
      "epoch": 0.52,
      "grad_norm": 4.569182872772217,
      "learning_rate": 6.135593220338983e-06,
      "loss": 1.2482,
      "step": 20950
    },
    {
      "epoch": 0.52,
      "grad_norm": 2.1293554306030273,
      "learning_rate": 6.1288135593220346e-06,
      "loss": 1.2922,
      "step": 20960
    },
    {
      "epoch": 0.52,
      "grad_norm": 8.99962043762207,
      "learning_rate": 6.122033898305085e-06,
      "loss": 1.4384,
      "step": 20970
    },
    {
      "epoch": 0.52,
      "grad_norm": 16.70253562927246,
      "learning_rate": 6.1152542372881365e-06,
      "loss": 1.2083,
      "step": 20980
    },
    {
      "epoch": 0.52,
      "grad_norm": 2.752915859222412,
      "learning_rate": 6.108474576271187e-06,
      "loss": 1.4159,
      "step": 20990
    },
    {
      "epoch": 0.53,
      "grad_norm": 5.210763454437256,
      "learning_rate": 6.1016949152542385e-06,
      "loss": 1.3307,
      "step": 21000
    },
    {
      "epoch": 0.53,
      "grad_norm": 7.698249816894531,
      "learning_rate": 6.094915254237289e-06,
      "loss": 1.3053,
      "step": 21010
    },
    {
      "epoch": 0.53,
      "grad_norm": 10.784698486328125,
      "learning_rate": 6.088135593220339e-06,
      "loss": 1.3872,
      "step": 21020
    },
    {
      "epoch": 0.53,
      "grad_norm": 6.999063014984131,
      "learning_rate": 6.08135593220339e-06,
      "loss": 1.3749,
      "step": 21030
    },
    {
      "epoch": 0.53,
      "grad_norm": 4.021378993988037,
      "learning_rate": 6.074576271186441e-06,
      "loss": 1.4373,
      "step": 21040
    },
    {
      "epoch": 0.53,
      "grad_norm": 7.960516929626465,
      "learning_rate": 6.067796610169492e-06,
      "loss": 1.3342,
      "step": 21050
    },
    {
      "epoch": 0.53,
      "grad_norm": 4.862764358520508,
      "learning_rate": 6.061016949152543e-06,
      "loss": 1.369,
      "step": 21060
    },
    {
      "epoch": 0.53,
      "grad_norm": 2.728264570236206,
      "learning_rate": 6.054237288135594e-06,
      "loss": 1.3669,
      "step": 21070
    },
    {
      "epoch": 0.53,
      "grad_norm": 4.100936412811279,
      "learning_rate": 6.047457627118645e-06,
      "loss": 1.3545,
      "step": 21080
    },
    {
      "epoch": 0.53,
      "grad_norm": 10.657527923583984,
      "learning_rate": 6.040677966101696e-06,
      "loss": 1.4269,
      "step": 21090
    },
    {
      "epoch": 0.53,
      "grad_norm": 9.700201988220215,
      "learning_rate": 6.0338983050847465e-06,
      "loss": 1.4317,
      "step": 21100
    },
    {
      "epoch": 0.53,
      "grad_norm": 8.100811958312988,
      "learning_rate": 6.027118644067798e-06,
      "loss": 1.337,
      "step": 21110
    },
    {
      "epoch": 0.53,
      "grad_norm": 3.148442268371582,
      "learning_rate": 6.020338983050848e-06,
      "loss": 1.2295,
      "step": 21120
    },
    {
      "epoch": 0.53,
      "grad_norm": 4.263021469116211,
      "learning_rate": 6.013559322033898e-06,
      "loss": 1.4917,
      "step": 21130
    },
    {
      "epoch": 0.53,
      "grad_norm": 8.611481666564941,
      "learning_rate": 6.00677966101695e-06,
      "loss": 1.4328,
      "step": 21140
    },
    {
      "epoch": 0.53,
      "grad_norm": 4.446702480316162,
      "learning_rate": 6e-06,
      "loss": 1.2638,
      "step": 21150
    },
    {
      "epoch": 0.53,
      "grad_norm": 9.882699966430664,
      "learning_rate": 5.9932203389830516e-06,
      "loss": 1.2,
      "step": 21160
    },
    {
      "epoch": 0.53,
      "grad_norm": 2.4835803508758545,
      "learning_rate": 5.986440677966102e-06,
      "loss": 1.4001,
      "step": 21170
    },
    {
      "epoch": 0.53,
      "grad_norm": 7.206079006195068,
      "learning_rate": 5.9796610169491535e-06,
      "loss": 1.3011,
      "step": 21180
    },
    {
      "epoch": 0.53,
      "grad_norm": 2.2151267528533936,
      "learning_rate": 5.972881355932204e-06,
      "loss": 1.4225,
      "step": 21190
    },
    {
      "epoch": 0.53,
      "grad_norm": 4.7154221534729,
      "learning_rate": 5.9661016949152555e-06,
      "loss": 1.3216,
      "step": 21200
    },
    {
      "epoch": 0.53,
      "grad_norm": 2.9739983081817627,
      "learning_rate": 5.959322033898306e-06,
      "loss": 1.2715,
      "step": 21210
    },
    {
      "epoch": 0.53,
      "grad_norm": 3.9512410163879395,
      "learning_rate": 5.9525423728813566e-06,
      "loss": 1.3083,
      "step": 21220
    },
    {
      "epoch": 0.53,
      "grad_norm": 3.705937385559082,
      "learning_rate": 5.945762711864407e-06,
      "loss": 1.3187,
      "step": 21230
    },
    {
      "epoch": 0.53,
      "grad_norm": 5.944228172302246,
      "learning_rate": 5.938983050847458e-06,
      "loss": 1.4633,
      "step": 21240
    },
    {
      "epoch": 0.53,
      "grad_norm": 2.7101802825927734,
      "learning_rate": 5.932203389830509e-06,
      "loss": 1.4035,
      "step": 21250
    },
    {
      "epoch": 0.53,
      "grad_norm": 12.047493934631348,
      "learning_rate": 5.92542372881356e-06,
      "loss": 1.2315,
      "step": 21260
    },
    {
      "epoch": 0.53,
      "grad_norm": 6.789846897125244,
      "learning_rate": 5.91864406779661e-06,
      "loss": 1.3253,
      "step": 21270
    },
    {
      "epoch": 0.53,
      "grad_norm": 7.4159369468688965,
      "learning_rate": 5.911864406779662e-06,
      "loss": 1.4368,
      "step": 21280
    },
    {
      "epoch": 0.53,
      "grad_norm": 5.49143648147583,
      "learning_rate": 5.905084745762712e-06,
      "loss": 1.5592,
      "step": 21290
    },
    {
      "epoch": 0.53,
      "grad_norm": 4.616772174835205,
      "learning_rate": 5.8983050847457635e-06,
      "loss": 1.4645,
      "step": 21300
    },
    {
      "epoch": 0.53,
      "grad_norm": 4.112008094787598,
      "learning_rate": 5.891525423728814e-06,
      "loss": 1.3504,
      "step": 21310
    },
    {
      "epoch": 0.53,
      "grad_norm": 14.327009201049805,
      "learning_rate": 5.8847457627118655e-06,
      "loss": 1.2757,
      "step": 21320
    },
    {
      "epoch": 0.53,
      "grad_norm": 3.6715610027313232,
      "learning_rate": 5.877966101694915e-06,
      "loss": 1.3965,
      "step": 21330
    },
    {
      "epoch": 0.53,
      "grad_norm": 2.6030914783477783,
      "learning_rate": 5.871186440677966e-06,
      "loss": 1.503,
      "step": 21340
    },
    {
      "epoch": 0.53,
      "grad_norm": 6.329497337341309,
      "learning_rate": 5.864406779661017e-06,
      "loss": 1.226,
      "step": 21350
    },
    {
      "epoch": 0.53,
      "grad_norm": 7.985304355621338,
      "learning_rate": 5.857627118644068e-06,
      "loss": 1.3999,
      "step": 21360
    },
    {
      "epoch": 0.53,
      "grad_norm": 2.7633464336395264,
      "learning_rate": 5.850847457627119e-06,
      "loss": 1.4605,
      "step": 21370
    },
    {
      "epoch": 0.53,
      "grad_norm": 5.338439464569092,
      "learning_rate": 5.84406779661017e-06,
      "loss": 1.2479,
      "step": 21380
    },
    {
      "epoch": 0.53,
      "grad_norm": 5.68145751953125,
      "learning_rate": 5.837288135593221e-06,
      "loss": 1.4275,
      "step": 21390
    },
    {
      "epoch": 0.54,
      "grad_norm": 2.4100029468536377,
      "learning_rate": 5.830508474576272e-06,
      "loss": 1.1839,
      "step": 21400
    },
    {
      "epoch": 0.54,
      "grad_norm": 9.0816650390625,
      "learning_rate": 5.823728813559323e-06,
      "loss": 1.4315,
      "step": 21410
    },
    {
      "epoch": 0.54,
      "grad_norm": 6.4322190284729,
      "learning_rate": 5.8169491525423736e-06,
      "loss": 1.4943,
      "step": 21420
    },
    {
      "epoch": 0.54,
      "grad_norm": 4.693780422210693,
      "learning_rate": 5.810169491525425e-06,
      "loss": 1.2531,
      "step": 21430
    },
    {
      "epoch": 0.54,
      "grad_norm": 10.678814888000488,
      "learning_rate": 5.803389830508475e-06,
      "loss": 1.3508,
      "step": 21440
    },
    {
      "epoch": 0.54,
      "grad_norm": 6.265130043029785,
      "learning_rate": 5.796610169491525e-06,
      "loss": 1.2625,
      "step": 21450
    },
    {
      "epoch": 0.54,
      "grad_norm": 3.7817535400390625,
      "learning_rate": 5.789830508474577e-06,
      "loss": 1.3654,
      "step": 21460
    },
    {
      "epoch": 0.54,
      "grad_norm": 16.397951126098633,
      "learning_rate": 5.783050847457627e-06,
      "loss": 1.3014,
      "step": 21470
    },
    {
      "epoch": 0.54,
      "grad_norm": 4.194436073303223,
      "learning_rate": 5.776271186440679e-06,
      "loss": 1.1604,
      "step": 21480
    },
    {
      "epoch": 0.54,
      "grad_norm": 8.044507026672363,
      "learning_rate": 5.769491525423729e-06,
      "loss": 1.3226,
      "step": 21490
    },
    {
      "epoch": 0.54,
      "grad_norm": 13.019102096557617,
      "learning_rate": 5.7627118644067805e-06,
      "loss": 1.4915,
      "step": 21500
    },
    {
      "epoch": 0.54,
      "grad_norm": 7.105216026306152,
      "learning_rate": 5.755932203389831e-06,
      "loss": 1.231,
      "step": 21510
    },
    {
      "epoch": 0.54,
      "grad_norm": 4.86849308013916,
      "learning_rate": 5.7491525423728825e-06,
      "loss": 1.2058,
      "step": 21520
    },
    {
      "epoch": 0.54,
      "grad_norm": 5.364449977874756,
      "learning_rate": 5.742372881355933e-06,
      "loss": 1.1875,
      "step": 21530
    },
    {
      "epoch": 0.54,
      "grad_norm": 3.9380345344543457,
      "learning_rate": 5.735593220338983e-06,
      "loss": 1.153,
      "step": 21540
    },
    {
      "epoch": 0.54,
      "grad_norm": 10.023425102233887,
      "learning_rate": 5.728813559322034e-06,
      "loss": 1.188,
      "step": 21550
    },
    {
      "epoch": 0.54,
      "grad_norm": 13.787605285644531,
      "learning_rate": 5.722033898305085e-06,
      "loss": 1.3615,
      "step": 21560
    },
    {
      "epoch": 0.54,
      "grad_norm": 10.42817211151123,
      "learning_rate": 5.715254237288136e-06,
      "loss": 1.3286,
      "step": 21570
    },
    {
      "epoch": 0.54,
      "grad_norm": 6.196288585662842,
      "learning_rate": 5.708474576271187e-06,
      "loss": 1.2818,
      "step": 21580
    },
    {
      "epoch": 0.54,
      "grad_norm": 6.792708873748779,
      "learning_rate": 5.701694915254238e-06,
      "loss": 1.3055,
      "step": 21590
    },
    {
      "epoch": 0.54,
      "grad_norm": 9.340911865234375,
      "learning_rate": 5.694915254237289e-06,
      "loss": 1.4033,
      "step": 21600
    },
    {
      "epoch": 0.54,
      "grad_norm": 3.1234211921691895,
      "learning_rate": 5.68813559322034e-06,
      "loss": 1.2611,
      "step": 21610
    },
    {
      "epoch": 0.54,
      "grad_norm": 3.1399428844451904,
      "learning_rate": 5.6813559322033906e-06,
      "loss": 1.1944,
      "step": 21620
    },
    {
      "epoch": 0.54,
      "grad_norm": 9.483480453491211,
      "learning_rate": 5.674576271186442e-06,
      "loss": 1.5517,
      "step": 21630
    },
    {
      "epoch": 0.54,
      "grad_norm": 3.204786539077759,
      "learning_rate": 5.667796610169492e-06,
      "loss": 1.5206,
      "step": 21640
    },
    {
      "epoch": 0.54,
      "grad_norm": 3.714053153991699,
      "learning_rate": 5.661016949152542e-06,
      "loss": 1.5099,
      "step": 21650
    },
    {
      "epoch": 0.54,
      "grad_norm": 5.080292701721191,
      "learning_rate": 5.654237288135594e-06,
      "loss": 1.4777,
      "step": 21660
    },
    {
      "epoch": 0.54,
      "grad_norm": 4.1797943115234375,
      "learning_rate": 5.647457627118644e-06,
      "loss": 1.3334,
      "step": 21670
    },
    {
      "epoch": 0.54,
      "grad_norm": 2.6116831302642822,
      "learning_rate": 5.640677966101696e-06,
      "loss": 1.4182,
      "step": 21680
    },
    {
      "epoch": 0.54,
      "grad_norm": 3.7096669673919678,
      "learning_rate": 5.633898305084746e-06,
      "loss": 1.2448,
      "step": 21690
    },
    {
      "epoch": 0.54,
      "grad_norm": 5.467226982116699,
      "learning_rate": 5.6271186440677975e-06,
      "loss": 1.2938,
      "step": 21700
    },
    {
      "epoch": 0.54,
      "grad_norm": 5.12249231338501,
      "learning_rate": 5.620338983050848e-06,
      "loss": 1.4177,
      "step": 21710
    },
    {
      "epoch": 0.54,
      "grad_norm": 3.1156530380249023,
      "learning_rate": 5.6135593220338995e-06,
      "loss": 1.4467,
      "step": 21720
    },
    {
      "epoch": 0.54,
      "grad_norm": 3.0598835945129395,
      "learning_rate": 5.60677966101695e-06,
      "loss": 1.2954,
      "step": 21730
    },
    {
      "epoch": 0.54,
      "grad_norm": 3.4932334423065186,
      "learning_rate": 5.600000000000001e-06,
      "loss": 1.3108,
      "step": 21740
    },
    {
      "epoch": 0.54,
      "grad_norm": 9.232585906982422,
      "learning_rate": 5.593220338983051e-06,
      "loss": 1.3295,
      "step": 21750
    },
    {
      "epoch": 0.54,
      "grad_norm": 3.7384567260742188,
      "learning_rate": 5.586440677966102e-06,
      "loss": 1.2195,
      "step": 21760
    },
    {
      "epoch": 0.54,
      "grad_norm": 10.520264625549316,
      "learning_rate": 5.579661016949153e-06,
      "loss": 1.3308,
      "step": 21770
    },
    {
      "epoch": 0.54,
      "grad_norm": 8.967552185058594,
      "learning_rate": 5.572881355932204e-06,
      "loss": 1.2524,
      "step": 21780
    },
    {
      "epoch": 0.54,
      "grad_norm": 4.718537330627441,
      "learning_rate": 5.566101694915255e-06,
      "loss": 1.6225,
      "step": 21790
    },
    {
      "epoch": 0.55,
      "grad_norm": 7.520408630371094,
      "learning_rate": 5.559322033898306e-06,
      "loss": 1.4489,
      "step": 21800
    },
    {
      "epoch": 0.55,
      "grad_norm": 2.3625175952911377,
      "learning_rate": 5.552542372881356e-06,
      "loss": 1.2048,
      "step": 21810
    },
    {
      "epoch": 0.55,
      "grad_norm": 11.236878395080566,
      "learning_rate": 5.5457627118644076e-06,
      "loss": 1.2144,
      "step": 21820
    },
    {
      "epoch": 0.55,
      "grad_norm": 2.8770816326141357,
      "learning_rate": 5.538983050847458e-06,
      "loss": 1.3672,
      "step": 21830
    },
    {
      "epoch": 0.55,
      "grad_norm": 4.691359519958496,
      "learning_rate": 5.5322033898305095e-06,
      "loss": 1.3181,
      "step": 21840
    },
    {
      "epoch": 0.55,
      "grad_norm": 11.01724624633789,
      "learning_rate": 5.525423728813559e-06,
      "loss": 1.1618,
      "step": 21850
    },
    {
      "epoch": 0.55,
      "grad_norm": 3.48175311088562,
      "learning_rate": 5.518644067796611e-06,
      "loss": 1.4492,
      "step": 21860
    },
    {
      "epoch": 0.55,
      "grad_norm": 4.6883544921875,
      "learning_rate": 5.511864406779661e-06,
      "loss": 1.2777,
      "step": 21870
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.7653623819351196,
      "learning_rate": 5.505084745762712e-06,
      "loss": 1.3169,
      "step": 21880
    },
    {
      "epoch": 0.55,
      "grad_norm": 6.6760029792785645,
      "learning_rate": 5.498305084745763e-06,
      "loss": 1.3427,
      "step": 21890
    },
    {
      "epoch": 0.55,
      "grad_norm": 8.014688491821289,
      "learning_rate": 5.491525423728814e-06,
      "loss": 1.4466,
      "step": 21900
    },
    {
      "epoch": 0.55,
      "grad_norm": 4.489105224609375,
      "learning_rate": 5.484745762711865e-06,
      "loss": 1.2728,
      "step": 21910
    },
    {
      "epoch": 0.55,
      "grad_norm": 15.908321380615234,
      "learning_rate": 5.477966101694916e-06,
      "loss": 1.5261,
      "step": 21920
    },
    {
      "epoch": 0.55,
      "grad_norm": 5.994565010070801,
      "learning_rate": 5.471186440677967e-06,
      "loss": 1.3377,
      "step": 21930
    },
    {
      "epoch": 0.55,
      "grad_norm": 5.997050762176514,
      "learning_rate": 5.464406779661018e-06,
      "loss": 1.2874,
      "step": 21940
    },
    {
      "epoch": 0.55,
      "grad_norm": 6.5358099937438965,
      "learning_rate": 5.457627118644067e-06,
      "loss": 1.4312,
      "step": 21950
    },
    {
      "epoch": 0.55,
      "grad_norm": 11.344245910644531,
      "learning_rate": 5.450847457627119e-06,
      "loss": 1.5362,
      "step": 21960
    },
    {
      "epoch": 0.55,
      "grad_norm": 6.406681060791016,
      "learning_rate": 5.444067796610169e-06,
      "loss": 1.3455,
      "step": 21970
    },
    {
      "epoch": 0.55,
      "grad_norm": 9.506949424743652,
      "learning_rate": 5.437288135593221e-06,
      "loss": 1.5093,
      "step": 21980
    },
    {
      "epoch": 0.55,
      "grad_norm": 2.633558988571167,
      "learning_rate": 5.430508474576271e-06,
      "loss": 1.4614,
      "step": 21990
    },
    {
      "epoch": 0.55,
      "grad_norm": 5.4223713874816895,
      "learning_rate": 5.423728813559323e-06,
      "loss": 1.217,
      "step": 22000
    },
    {
      "epoch": 0.55,
      "grad_norm": 9.218213081359863,
      "learning_rate": 5.416949152542373e-06,
      "loss": 1.1343,
      "step": 22010
    },
    {
      "epoch": 0.55,
      "grad_norm": 7.161883354187012,
      "learning_rate": 5.4101694915254246e-06,
      "loss": 1.3996,
      "step": 22020
    },
    {
      "epoch": 0.55,
      "grad_norm": 4.1903157234191895,
      "learning_rate": 5.403389830508475e-06,
      "loss": 1.3594,
      "step": 22030
    },
    {
      "epoch": 0.55,
      "grad_norm": 2.753154754638672,
      "learning_rate": 5.3966101694915265e-06,
      "loss": 1.0266,
      "step": 22040
    },
    {
      "epoch": 0.55,
      "grad_norm": 6.781494140625,
      "learning_rate": 5.389830508474577e-06,
      "loss": 1.3484,
      "step": 22050
    },
    {
      "epoch": 0.55,
      "grad_norm": 4.839456558227539,
      "learning_rate": 5.383050847457627e-06,
      "loss": 1.1969,
      "step": 22060
    },
    {
      "epoch": 0.55,
      "grad_norm": 2.400866746902466,
      "learning_rate": 5.376271186440678e-06,
      "loss": 1.3895,
      "step": 22070
    },
    {
      "epoch": 0.55,
      "grad_norm": 11.034051895141602,
      "learning_rate": 5.369491525423729e-06,
      "loss": 1.2501,
      "step": 22080
    },
    {
      "epoch": 0.55,
      "grad_norm": 8.162158012390137,
      "learning_rate": 5.36271186440678e-06,
      "loss": 1.1053,
      "step": 22090
    },
    {
      "epoch": 0.55,
      "grad_norm": 6.1298747062683105,
      "learning_rate": 5.355932203389831e-06,
      "loss": 1.2716,
      "step": 22100
    },
    {
      "epoch": 0.55,
      "grad_norm": 3.0399742126464844,
      "learning_rate": 5.349152542372882e-06,
      "loss": 1.1162,
      "step": 22110
    },
    {
      "epoch": 0.55,
      "grad_norm": 11.782633781433105,
      "learning_rate": 5.342372881355933e-06,
      "loss": 1.298,
      "step": 22120
    },
    {
      "epoch": 0.55,
      "grad_norm": 5.440155982971191,
      "learning_rate": 5.335593220338984e-06,
      "loss": 1.5204,
      "step": 22130
    },
    {
      "epoch": 0.55,
      "grad_norm": 4.491126537322998,
      "learning_rate": 5.328813559322035e-06,
      "loss": 1.2911,
      "step": 22140
    },
    {
      "epoch": 0.55,
      "grad_norm": 9.803009033203125,
      "learning_rate": 5.322033898305086e-06,
      "loss": 1.2635,
      "step": 22150
    },
    {
      "epoch": 0.55,
      "grad_norm": 10.113853454589844,
      "learning_rate": 5.315254237288136e-06,
      "loss": 1.4671,
      "step": 22160
    },
    {
      "epoch": 0.55,
      "grad_norm": 12.36583423614502,
      "learning_rate": 5.308474576271186e-06,
      "loss": 1.3344,
      "step": 22170
    },
    {
      "epoch": 0.55,
      "grad_norm": 11.591397285461426,
      "learning_rate": 5.301694915254238e-06,
      "loss": 1.3677,
      "step": 22180
    },
    {
      "epoch": 0.55,
      "grad_norm": 3.309582233428955,
      "learning_rate": 5.294915254237288e-06,
      "loss": 1.314,
      "step": 22190
    },
    {
      "epoch": 0.56,
      "grad_norm": 8.943580627441406,
      "learning_rate": 5.28813559322034e-06,
      "loss": 1.3398,
      "step": 22200
    },
    {
      "epoch": 0.56,
      "grad_norm": 4.133789539337158,
      "learning_rate": 5.28135593220339e-06,
      "loss": 1.4291,
      "step": 22210
    },
    {
      "epoch": 0.56,
      "grad_norm": 10.557209968566895,
      "learning_rate": 5.2745762711864416e-06,
      "loss": 1.2711,
      "step": 22220
    },
    {
      "epoch": 0.56,
      "grad_norm": 8.362191200256348,
      "learning_rate": 5.267796610169492e-06,
      "loss": 1.25,
      "step": 22230
    },
    {
      "epoch": 0.56,
      "grad_norm": 8.049225807189941,
      "learning_rate": 5.2610169491525435e-06,
      "loss": 1.3188,
      "step": 22240
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.0453487634658813,
      "learning_rate": 5.254237288135594e-06,
      "loss": 1.305,
      "step": 22250
    },
    {
      "epoch": 0.56,
      "grad_norm": 6.4410834312438965,
      "learning_rate": 5.247457627118645e-06,
      "loss": 1.3933,
      "step": 22260
    },
    {
      "epoch": 0.56,
      "grad_norm": 6.164778709411621,
      "learning_rate": 5.240677966101695e-06,
      "loss": 1.1667,
      "step": 22270
    },
    {
      "epoch": 0.56,
      "grad_norm": 8.114805221557617,
      "learning_rate": 5.233898305084746e-06,
      "loss": 1.3263,
      "step": 22280
    },
    {
      "epoch": 0.56,
      "grad_norm": 3.8220670223236084,
      "learning_rate": 5.227118644067797e-06,
      "loss": 1.3337,
      "step": 22290
    },
    {
      "epoch": 0.56,
      "grad_norm": 8.982809066772461,
      "learning_rate": 5.220338983050848e-06,
      "loss": 1.4235,
      "step": 22300
    },
    {
      "epoch": 0.56,
      "grad_norm": 5.933160781860352,
      "learning_rate": 5.213559322033899e-06,
      "loss": 1.2963,
      "step": 22310
    },
    {
      "epoch": 0.56,
      "grad_norm": 5.841404914855957,
      "learning_rate": 5.20677966101695e-06,
      "loss": 1.3441,
      "step": 22320
    },
    {
      "epoch": 0.56,
      "grad_norm": 9.481009483337402,
      "learning_rate": 5.2e-06,
      "loss": 1.5076,
      "step": 22330
    },
    {
      "epoch": 0.56,
      "grad_norm": 4.357295036315918,
      "learning_rate": 5.193220338983052e-06,
      "loss": 1.3962,
      "step": 22340
    },
    {
      "epoch": 0.56,
      "grad_norm": 4.579665184020996,
      "learning_rate": 5.186440677966102e-06,
      "loss": 1.2782,
      "step": 22350
    },
    {
      "epoch": 0.56,
      "grad_norm": 6.721803188323975,
      "learning_rate": 5.1796610169491535e-06,
      "loss": 1.1839,
      "step": 22360
    },
    {
      "epoch": 0.56,
      "grad_norm": 6.047186851501465,
      "learning_rate": 5.172881355932203e-06,
      "loss": 1.242,
      "step": 22370
    },
    {
      "epoch": 0.56,
      "grad_norm": 5.540093898773193,
      "learning_rate": 5.166101694915255e-06,
      "loss": 1.5321,
      "step": 22380
    },
    {
      "epoch": 0.56,
      "grad_norm": 10.6552734375,
      "learning_rate": 5.159322033898305e-06,
      "loss": 1.4318,
      "step": 22390
    },
    {
      "epoch": 0.56,
      "grad_norm": 2.812580108642578,
      "learning_rate": 5.152542372881356e-06,
      "loss": 1.3183,
      "step": 22400
    },
    {
      "epoch": 0.56,
      "grad_norm": 11.109842300415039,
      "learning_rate": 5.145762711864407e-06,
      "loss": 1.3034,
      "step": 22410
    },
    {
      "epoch": 0.56,
      "grad_norm": 6.639523506164551,
      "learning_rate": 5.138983050847458e-06,
      "loss": 1.3265,
      "step": 22420
    },
    {
      "epoch": 0.56,
      "grad_norm": 3.581305503845215,
      "learning_rate": 5.132203389830509e-06,
      "loss": 1.3837,
      "step": 22430
    },
    {
      "epoch": 0.56,
      "grad_norm": 6.955338954925537,
      "learning_rate": 5.12542372881356e-06,
      "loss": 1.2203,
      "step": 22440
    },
    {
      "epoch": 0.56,
      "grad_norm": 6.101033687591553,
      "learning_rate": 5.118644067796611e-06,
      "loss": 1.2535,
      "step": 22450
    },
    {
      "epoch": 0.56,
      "grad_norm": 5.1015801429748535,
      "learning_rate": 5.111864406779662e-06,
      "loss": 1.2731,
      "step": 22460
    },
    {
      "epoch": 0.56,
      "grad_norm": 14.434350967407227,
      "learning_rate": 5.105084745762711e-06,
      "loss": 1.2948,
      "step": 22470
    },
    {
      "epoch": 0.56,
      "grad_norm": 3.6449780464172363,
      "learning_rate": 5.098305084745763e-06,
      "loss": 1.2662,
      "step": 22480
    },
    {
      "epoch": 0.56,
      "grad_norm": 2.7186925411224365,
      "learning_rate": 5.091525423728813e-06,
      "loss": 1.3551,
      "step": 22490
    },
    {
      "epoch": 0.56,
      "grad_norm": 4.75358772277832,
      "learning_rate": 5.084745762711865e-06,
      "loss": 1.2254,
      "step": 22500
    },
    {
      "epoch": 0.56,
      "eval_loss": 1.3262169361114502,
      "eval_runtime": 122.373,
      "eval_samples_per_second": 8.172,
      "eval_steps_per_second": 8.172,
      "step": 22500
    },
    {
      "epoch": 0.56,
      "grad_norm": 3.8966474533081055,
      "learning_rate": 5.077966101694915e-06,
      "loss": 1.3701,
      "step": 22510
    },
    {
      "epoch": 0.56,
      "grad_norm": 7.942519664764404,
      "learning_rate": 5.071186440677967e-06,
      "loss": 1.3176,
      "step": 22520
    },
    {
      "epoch": 0.56,
      "grad_norm": 11.244969367980957,
      "learning_rate": 5.064406779661017e-06,
      "loss": 1.2549,
      "step": 22530
    },
    {
      "epoch": 0.56,
      "grad_norm": 6.425158500671387,
      "learning_rate": 5.057627118644069e-06,
      "loss": 1.3243,
      "step": 22540
    },
    {
      "epoch": 0.56,
      "grad_norm": 3.9619619846343994,
      "learning_rate": 5.050847457627119e-06,
      "loss": 1.4702,
      "step": 22550
    },
    {
      "epoch": 0.56,
      "grad_norm": 6.954550266265869,
      "learning_rate": 5.0440677966101705e-06,
      "loss": 1.3348,
      "step": 22560
    },
    {
      "epoch": 0.56,
      "grad_norm": 8.0310697555542,
      "learning_rate": 5.037288135593221e-06,
      "loss": 1.3938,
      "step": 22570
    },
    {
      "epoch": 0.56,
      "grad_norm": 4.217766284942627,
      "learning_rate": 5.030508474576271e-06,
      "loss": 1.322,
      "step": 22580
    },
    {
      "epoch": 0.56,
      "grad_norm": 7.419365406036377,
      "learning_rate": 5.023728813559322e-06,
      "loss": 1.4899,
      "step": 22590
    },
    {
      "epoch": 0.56,
      "grad_norm": 8.665319442749023,
      "learning_rate": 5.016949152542373e-06,
      "loss": 1.3262,
      "step": 22600
    },
    {
      "epoch": 0.57,
      "grad_norm": 4.781732559204102,
      "learning_rate": 5.010169491525424e-06,
      "loss": 1.3748,
      "step": 22610
    },
    {
      "epoch": 0.57,
      "grad_norm": 2.8004636764526367,
      "learning_rate": 5.003389830508475e-06,
      "loss": 1.3445,
      "step": 22620
    },
    {
      "epoch": 0.57,
      "grad_norm": 12.941588401794434,
      "learning_rate": 4.996610169491526e-06,
      "loss": 1.2112,
      "step": 22630
    },
    {
      "epoch": 0.57,
      "grad_norm": 9.608789443969727,
      "learning_rate": 4.989830508474577e-06,
      "loss": 1.3931,
      "step": 22640
    },
    {
      "epoch": 0.57,
      "grad_norm": 2.8279383182525635,
      "learning_rate": 4.983050847457628e-06,
      "loss": 1.2481,
      "step": 22650
    },
    {
      "epoch": 0.57,
      "grad_norm": 7.210662364959717,
      "learning_rate": 4.976271186440678e-06,
      "loss": 1.3074,
      "step": 22660
    },
    {
      "epoch": 0.57,
      "grad_norm": 3.413175582885742,
      "learning_rate": 4.969491525423729e-06,
      "loss": 1.3209,
      "step": 22670
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.654166579246521,
      "learning_rate": 4.96271186440678e-06,
      "loss": 1.3002,
      "step": 22680
    },
    {
      "epoch": 0.57,
      "grad_norm": 10.318363189697266,
      "learning_rate": 4.955932203389831e-06,
      "loss": 1.1209,
      "step": 22690
    },
    {
      "epoch": 0.57,
      "grad_norm": 4.4343791007995605,
      "learning_rate": 4.949152542372882e-06,
      "loss": 1.4163,
      "step": 22700
    },
    {
      "epoch": 0.57,
      "grad_norm": 9.116500854492188,
      "learning_rate": 4.942372881355932e-06,
      "loss": 1.3966,
      "step": 22710
    },
    {
      "epoch": 0.57,
      "grad_norm": 5.384235858917236,
      "learning_rate": 4.935593220338984e-06,
      "loss": 1.3546,
      "step": 22720
    },
    {
      "epoch": 0.57,
      "grad_norm": 4.688304424285889,
      "learning_rate": 4.928813559322034e-06,
      "loss": 1.3766,
      "step": 22730
    },
    {
      "epoch": 0.57,
      "grad_norm": 4.299511909484863,
      "learning_rate": 4.922033898305086e-06,
      "loss": 1.3755,
      "step": 22740
    },
    {
      "epoch": 0.57,
      "grad_norm": 6.685275077819824,
      "learning_rate": 4.915254237288136e-06,
      "loss": 1.3884,
      "step": 22750
    },
    {
      "epoch": 0.57,
      "grad_norm": 7.280564308166504,
      "learning_rate": 4.908474576271187e-06,
      "loss": 1.2153,
      "step": 22760
    },
    {
      "epoch": 0.57,
      "grad_norm": 2.573958396911621,
      "learning_rate": 4.901694915254237e-06,
      "loss": 1.3893,
      "step": 22770
    },
    {
      "epoch": 0.57,
      "grad_norm": 8.996614456176758,
      "learning_rate": 4.894915254237289e-06,
      "loss": 1.3945,
      "step": 22780
    },
    {
      "epoch": 0.57,
      "grad_norm": 6.839596271514893,
      "learning_rate": 4.888135593220339e-06,
      "loss": 1.4355,
      "step": 22790
    },
    {
      "epoch": 0.57,
      "grad_norm": 3.997774839401245,
      "learning_rate": 4.881355932203391e-06,
      "loss": 1.1918,
      "step": 22800
    },
    {
      "epoch": 0.57,
      "grad_norm": 9.9241943359375,
      "learning_rate": 4.874576271186441e-06,
      "loss": 1.311,
      "step": 22810
    },
    {
      "epoch": 0.57,
      "grad_norm": 7.339332103729248,
      "learning_rate": 4.867796610169492e-06,
      "loss": 1.2219,
      "step": 22820
    },
    {
      "epoch": 0.57,
      "grad_norm": 5.831264972686768,
      "learning_rate": 4.861016949152543e-06,
      "loss": 1.4528,
      "step": 22830
    },
    {
      "epoch": 0.57,
      "grad_norm": 13.206710815429688,
      "learning_rate": 4.854237288135594e-06,
      "loss": 1.5454,
      "step": 22840
    },
    {
      "epoch": 0.57,
      "grad_norm": 11.570894241333008,
      "learning_rate": 4.847457627118645e-06,
      "loss": 1.4312,
      "step": 22850
    },
    {
      "epoch": 0.57,
      "grad_norm": 3.9631407260894775,
      "learning_rate": 4.840677966101695e-06,
      "loss": 1.2286,
      "step": 22860
    },
    {
      "epoch": 0.57,
      "grad_norm": 2.9281585216522217,
      "learning_rate": 4.833898305084746e-06,
      "loss": 1.2974,
      "step": 22870
    },
    {
      "epoch": 0.57,
      "grad_norm": 3.692300796508789,
      "learning_rate": 4.827118644067797e-06,
      "loss": 1.2091,
      "step": 22880
    },
    {
      "epoch": 0.57,
      "grad_norm": 5.737148761749268,
      "learning_rate": 4.820338983050848e-06,
      "loss": 1.3564,
      "step": 22890
    },
    {
      "epoch": 0.57,
      "grad_norm": 11.296788215637207,
      "learning_rate": 4.813559322033899e-06,
      "loss": 1.3316,
      "step": 22900
    },
    {
      "epoch": 0.57,
      "grad_norm": 2.6350669860839844,
      "learning_rate": 4.80677966101695e-06,
      "loss": 1.2617,
      "step": 22910
    },
    {
      "epoch": 0.57,
      "grad_norm": 4.150533676147461,
      "learning_rate": 4.800000000000001e-06,
      "loss": 1.3223,
      "step": 22920
    },
    {
      "epoch": 0.57,
      "grad_norm": 11.198905944824219,
      "learning_rate": 4.793220338983051e-06,
      "loss": 1.3737,
      "step": 22930
    },
    {
      "epoch": 0.57,
      "grad_norm": 10.89981460571289,
      "learning_rate": 4.786440677966102e-06,
      "loss": 1.4232,
      "step": 22940
    },
    {
      "epoch": 0.57,
      "grad_norm": 6.690379619598389,
      "learning_rate": 4.779661016949153e-06,
      "loss": 1.3398,
      "step": 22950
    },
    {
      "epoch": 0.57,
      "grad_norm": 4.480560779571533,
      "learning_rate": 4.772881355932204e-06,
      "loss": 1.2081,
      "step": 22960
    },
    {
      "epoch": 0.57,
      "grad_norm": 6.472361087799072,
      "learning_rate": 4.766101694915254e-06,
      "loss": 1.3287,
      "step": 22970
    },
    {
      "epoch": 0.57,
      "grad_norm": 9.797123908996582,
      "learning_rate": 4.759322033898306e-06,
      "loss": 1.1217,
      "step": 22980
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.5400909185409546,
      "learning_rate": 4.752542372881356e-06,
      "loss": 1.1864,
      "step": 22990
    },
    {
      "epoch": 0.57,
      "grad_norm": 6.199005126953125,
      "learning_rate": 4.745762711864408e-06,
      "loss": 1.4267,
      "step": 23000
    },
    {
      "epoch": 0.58,
      "grad_norm": 8.349159240722656,
      "learning_rate": 4.738983050847458e-06,
      "loss": 1.436,
      "step": 23010
    },
    {
      "epoch": 0.58,
      "grad_norm": 17.42652702331543,
      "learning_rate": 4.732203389830509e-06,
      "loss": 1.3219,
      "step": 23020
    },
    {
      "epoch": 0.58,
      "grad_norm": 14.882533073425293,
      "learning_rate": 4.725423728813559e-06,
      "loss": 1.4409,
      "step": 23030
    },
    {
      "epoch": 0.58,
      "grad_norm": 8.824116706848145,
      "learning_rate": 4.718644067796611e-06,
      "loss": 1.3715,
      "step": 23040
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.6905938386917114,
      "learning_rate": 4.711864406779661e-06,
      "loss": 1.2218,
      "step": 23050
    },
    {
      "epoch": 0.58,
      "grad_norm": 6.688169002532959,
      "learning_rate": 4.705084745762713e-06,
      "loss": 1.4205,
      "step": 23060
    },
    {
      "epoch": 0.58,
      "grad_norm": 4.521963596343994,
      "learning_rate": 4.698305084745763e-06,
      "loss": 1.262,
      "step": 23070
    },
    {
      "epoch": 0.58,
      "grad_norm": 2.7517573833465576,
      "learning_rate": 4.691525423728814e-06,
      "loss": 1.4178,
      "step": 23080
    },
    {
      "epoch": 0.58,
      "grad_norm": 8.5391263961792,
      "learning_rate": 4.684745762711865e-06,
      "loss": 1.344,
      "step": 23090
    },
    {
      "epoch": 0.58,
      "grad_norm": 4.232736110687256,
      "learning_rate": 4.677966101694916e-06,
      "loss": 1.4065,
      "step": 23100
    },
    {
      "epoch": 0.58,
      "grad_norm": 12.701478958129883,
      "learning_rate": 4.671186440677967e-06,
      "loss": 1.4527,
      "step": 23110
    },
    {
      "epoch": 0.58,
      "grad_norm": 12.91329288482666,
      "learning_rate": 4.664406779661017e-06,
      "loss": 1.3096,
      "step": 23120
    },
    {
      "epoch": 0.58,
      "grad_norm": 3.6490163803100586,
      "learning_rate": 4.657627118644068e-06,
      "loss": 1.4439,
      "step": 23130
    },
    {
      "epoch": 0.58,
      "grad_norm": 5.199461936950684,
      "learning_rate": 4.650847457627119e-06,
      "loss": 1.456,
      "step": 23140
    },
    {
      "epoch": 0.58,
      "grad_norm": 3.675081253051758,
      "learning_rate": 4.64406779661017e-06,
      "loss": 1.2643,
      "step": 23150
    },
    {
      "epoch": 0.58,
      "grad_norm": 5.199882507324219,
      "learning_rate": 4.637288135593221e-06,
      "loss": 1.2117,
      "step": 23160
    },
    {
      "epoch": 0.58,
      "grad_norm": 8.903018951416016,
      "learning_rate": 4.630508474576272e-06,
      "loss": 1.3633,
      "step": 23170
    },
    {
      "epoch": 0.58,
      "grad_norm": 6.658726692199707,
      "learning_rate": 4.623728813559323e-06,
      "loss": 1.2576,
      "step": 23180
    },
    {
      "epoch": 0.58,
      "grad_norm": 7.379878997802734,
      "learning_rate": 4.616949152542373e-06,
      "loss": 1.4597,
      "step": 23190
    },
    {
      "epoch": 0.58,
      "grad_norm": 6.420439720153809,
      "learning_rate": 4.610169491525424e-06,
      "loss": 1.1719,
      "step": 23200
    },
    {
      "epoch": 0.58,
      "grad_norm": 10.046388626098633,
      "learning_rate": 4.603389830508475e-06,
      "loss": 1.4573,
      "step": 23210
    },
    {
      "epoch": 0.58,
      "grad_norm": 6.4325761795043945,
      "learning_rate": 4.596610169491526e-06,
      "loss": 1.3136,
      "step": 23220
    },
    {
      "epoch": 0.58,
      "grad_norm": 5.889609336853027,
      "learning_rate": 4.589830508474576e-06,
      "loss": 1.2557,
      "step": 23230
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.747876524925232,
      "learning_rate": 4.583050847457628e-06,
      "loss": 1.1637,
      "step": 23240
    },
    {
      "epoch": 0.58,
      "grad_norm": 7.574699878692627,
      "learning_rate": 4.576271186440678e-06,
      "loss": 1.2097,
      "step": 23250
    },
    {
      "epoch": 0.58,
      "grad_norm": 8.996368408203125,
      "learning_rate": 4.56949152542373e-06,
      "loss": 1.3375,
      "step": 23260
    },
    {
      "epoch": 0.58,
      "grad_norm": 11.68962574005127,
      "learning_rate": 4.56271186440678e-06,
      "loss": 1.2507,
      "step": 23270
    },
    {
      "epoch": 0.58,
      "grad_norm": 10.069473266601562,
      "learning_rate": 4.555932203389831e-06,
      "loss": 1.3118,
      "step": 23280
    },
    {
      "epoch": 0.58,
      "grad_norm": 2.3804116249084473,
      "learning_rate": 4.549152542372881e-06,
      "loss": 1.3264,
      "step": 23290
    },
    {
      "epoch": 0.58,
      "grad_norm": 4.807265758514404,
      "learning_rate": 4.542372881355933e-06,
      "loss": 1.1123,
      "step": 23300
    },
    {
      "epoch": 0.58,
      "grad_norm": 8.738534927368164,
      "learning_rate": 4.535593220338983e-06,
      "loss": 1.3119,
      "step": 23310
    },
    {
      "epoch": 0.58,
      "grad_norm": 4.734846591949463,
      "learning_rate": 4.528813559322035e-06,
      "loss": 1.2911,
      "step": 23320
    },
    {
      "epoch": 0.58,
      "grad_norm": 5.872142314910889,
      "learning_rate": 4.522033898305085e-06,
      "loss": 1.27,
      "step": 23330
    },
    {
      "epoch": 0.58,
      "grad_norm": 9.100564956665039,
      "learning_rate": 4.515254237288136e-06,
      "loss": 1.2584,
      "step": 23340
    },
    {
      "epoch": 0.58,
      "grad_norm": 8.202118873596191,
      "learning_rate": 4.508474576271187e-06,
      "loss": 1.2007,
      "step": 23350
    },
    {
      "epoch": 0.58,
      "grad_norm": 6.78044319152832,
      "learning_rate": 4.501694915254238e-06,
      "loss": 1.396,
      "step": 23360
    },
    {
      "epoch": 0.58,
      "grad_norm": 7.169397830963135,
      "learning_rate": 4.494915254237289e-06,
      "loss": 1.2857,
      "step": 23370
    },
    {
      "epoch": 0.58,
      "grad_norm": 8.966153144836426,
      "learning_rate": 4.488135593220339e-06,
      "loss": 1.3533,
      "step": 23380
    },
    {
      "epoch": 0.58,
      "grad_norm": 17.64482879638672,
      "learning_rate": 4.48135593220339e-06,
      "loss": 1.3687,
      "step": 23390
    },
    {
      "epoch": 0.58,
      "grad_norm": 13.492774963378906,
      "learning_rate": 4.474576271186441e-06,
      "loss": 1.4947,
      "step": 23400
    },
    {
      "epoch": 0.59,
      "grad_norm": 5.877525806427002,
      "learning_rate": 4.467796610169492e-06,
      "loss": 1.2806,
      "step": 23410
    },
    {
      "epoch": 0.59,
      "grad_norm": 5.113012790679932,
      "learning_rate": 4.461016949152543e-06,
      "loss": 1.3645,
      "step": 23420
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.907209038734436,
      "learning_rate": 4.454237288135594e-06,
      "loss": 1.4367,
      "step": 23430
    },
    {
      "epoch": 0.59,
      "grad_norm": 2.958543539047241,
      "learning_rate": 4.447457627118645e-06,
      "loss": 1.2299,
      "step": 23440
    },
    {
      "epoch": 0.59,
      "grad_norm": 9.819035530090332,
      "learning_rate": 4.440677966101695e-06,
      "loss": 1.3009,
      "step": 23450
    },
    {
      "epoch": 0.59,
      "grad_norm": 9.349372863769531,
      "learning_rate": 4.433898305084746e-06,
      "loss": 1.1882,
      "step": 23460
    },
    {
      "epoch": 0.59,
      "grad_norm": 5.309628963470459,
      "learning_rate": 4.427118644067797e-06,
      "loss": 1.4121,
      "step": 23470
    },
    {
      "epoch": 0.59,
      "grad_norm": 7.17758846282959,
      "learning_rate": 4.420338983050848e-06,
      "loss": 1.3117,
      "step": 23480
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.7299296855926514,
      "learning_rate": 4.413559322033898e-06,
      "loss": 1.4104,
      "step": 23490
    },
    {
      "epoch": 0.59,
      "grad_norm": 10.265968322753906,
      "learning_rate": 4.40677966101695e-06,
      "loss": 1.1101,
      "step": 23500
    },
    {
      "epoch": 0.59,
      "grad_norm": 10.072466850280762,
      "learning_rate": 4.4e-06,
      "loss": 1.1136,
      "step": 23510
    },
    {
      "epoch": 0.59,
      "grad_norm": 7.807421684265137,
      "learning_rate": 4.393220338983052e-06,
      "loss": 1.2777,
      "step": 23520
    },
    {
      "epoch": 0.59,
      "grad_norm": 5.091457843780518,
      "learning_rate": 4.386440677966102e-06,
      "loss": 1.3011,
      "step": 23530
    },
    {
      "epoch": 0.59,
      "grad_norm": 12.106415748596191,
      "learning_rate": 4.379661016949153e-06,
      "loss": 1.2999,
      "step": 23540
    },
    {
      "epoch": 0.59,
      "grad_norm": 3.699613094329834,
      "learning_rate": 4.372881355932203e-06,
      "loss": 1.5223,
      "step": 23550
    },
    {
      "epoch": 0.59,
      "grad_norm": 23.751169204711914,
      "learning_rate": 4.366101694915255e-06,
      "loss": 1.2738,
      "step": 23560
    },
    {
      "epoch": 0.59,
      "grad_norm": 7.125929832458496,
      "learning_rate": 4.359322033898305e-06,
      "loss": 1.184,
      "step": 23570
    },
    {
      "epoch": 0.59,
      "grad_norm": 4.1947760581970215,
      "learning_rate": 4.352542372881357e-06,
      "loss": 1.5137,
      "step": 23580
    },
    {
      "epoch": 0.59,
      "grad_norm": 7.801207065582275,
      "learning_rate": 4.345762711864407e-06,
      "loss": 1.3148,
      "step": 23590
    },
    {
      "epoch": 0.59,
      "grad_norm": 8.398948669433594,
      "learning_rate": 4.338983050847458e-06,
      "loss": 1.1618,
      "step": 23600
    },
    {
      "epoch": 0.59,
      "grad_norm": 6.113057613372803,
      "learning_rate": 4.332203389830509e-06,
      "loss": 1.2272,
      "step": 23610
    },
    {
      "epoch": 0.59,
      "grad_norm": 5.134265899658203,
      "learning_rate": 4.32542372881356e-06,
      "loss": 1.3766,
      "step": 23620
    },
    {
      "epoch": 0.59,
      "grad_norm": 7.657702922821045,
      "learning_rate": 4.318644067796611e-06,
      "loss": 1.4195,
      "step": 23630
    },
    {
      "epoch": 0.59,
      "grad_norm": 5.876861572265625,
      "learning_rate": 4.311864406779661e-06,
      "loss": 1.3812,
      "step": 23640
    },
    {
      "epoch": 0.59,
      "grad_norm": 5.893710136413574,
      "learning_rate": 4.305084745762712e-06,
      "loss": 1.2947,
      "step": 23650
    },
    {
      "epoch": 0.59,
      "grad_norm": 6.3075032234191895,
      "learning_rate": 4.298305084745763e-06,
      "loss": 1.2734,
      "step": 23660
    },
    {
      "epoch": 0.59,
      "grad_norm": 3.498833417892456,
      "learning_rate": 4.291525423728814e-06,
      "loss": 1.2876,
      "step": 23670
    },
    {
      "epoch": 0.59,
      "grad_norm": 4.680170059204102,
      "learning_rate": 4.284745762711865e-06,
      "loss": 1.3012,
      "step": 23680
    },
    {
      "epoch": 0.59,
      "grad_norm": 10.015076637268066,
      "learning_rate": 4.277966101694915e-06,
      "loss": 1.2618,
      "step": 23690
    },
    {
      "epoch": 0.59,
      "grad_norm": 4.386192798614502,
      "learning_rate": 4.271186440677967e-06,
      "loss": 1.4271,
      "step": 23700
    },
    {
      "epoch": 0.59,
      "grad_norm": 3.4567041397094727,
      "learning_rate": 4.264406779661017e-06,
      "loss": 1.4776,
      "step": 23710
    },
    {
      "epoch": 0.59,
      "grad_norm": 17.128454208374023,
      "learning_rate": 4.257627118644068e-06,
      "loss": 1.5807,
      "step": 23720
    },
    {
      "epoch": 0.59,
      "grad_norm": 5.857600688934326,
      "learning_rate": 4.250847457627119e-06,
      "loss": 1.213,
      "step": 23730
    },
    {
      "epoch": 0.59,
      "grad_norm": 12.647801399230957,
      "learning_rate": 4.24406779661017e-06,
      "loss": 1.4099,
      "step": 23740
    },
    {
      "epoch": 0.59,
      "grad_norm": 8.02133846282959,
      "learning_rate": 4.23728813559322e-06,
      "loss": 1.2556,
      "step": 23750
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.8410090208053589,
      "learning_rate": 4.230508474576272e-06,
      "loss": 1.3158,
      "step": 23760
    },
    {
      "epoch": 0.59,
      "grad_norm": 6.88528299331665,
      "learning_rate": 4.223728813559322e-06,
      "loss": 1.1572,
      "step": 23770
    },
    {
      "epoch": 0.59,
      "grad_norm": 12.544876098632812,
      "learning_rate": 4.216949152542374e-06,
      "loss": 1.2918,
      "step": 23780
    },
    {
      "epoch": 0.59,
      "grad_norm": 4.425920486450195,
      "learning_rate": 4.210169491525424e-06,
      "loss": 1.2342,
      "step": 23790
    },
    {
      "epoch": 0.59,
      "grad_norm": 6.000210285186768,
      "learning_rate": 4.203389830508475e-06,
      "loss": 1.4569,
      "step": 23800
    },
    {
      "epoch": 0.6,
      "grad_norm": 6.639040946960449,
      "learning_rate": 4.196610169491525e-06,
      "loss": 1.3527,
      "step": 23810
    },
    {
      "epoch": 0.6,
      "grad_norm": 24.1419620513916,
      "learning_rate": 4.189830508474577e-06,
      "loss": 1.2617,
      "step": 23820
    },
    {
      "epoch": 0.6,
      "grad_norm": 19.9375,
      "learning_rate": 4.183050847457627e-06,
      "loss": 1.5564,
      "step": 23830
    },
    {
      "epoch": 0.6,
      "grad_norm": 4.195712566375732,
      "learning_rate": 4.176271186440679e-06,
      "loss": 1.4757,
      "step": 23840
    },
    {
      "epoch": 0.6,
      "grad_norm": 5.61302375793457,
      "learning_rate": 4.169491525423729e-06,
      "loss": 1.3569,
      "step": 23850
    },
    {
      "epoch": 0.6,
      "grad_norm": 2.8037874698638916,
      "learning_rate": 4.16271186440678e-06,
      "loss": 1.451,
      "step": 23860
    },
    {
      "epoch": 0.6,
      "grad_norm": 10.095121383666992,
      "learning_rate": 4.155932203389831e-06,
      "loss": 1.1986,
      "step": 23870
    },
    {
      "epoch": 0.6,
      "grad_norm": 6.2578864097595215,
      "learning_rate": 4.149152542372882e-06,
      "loss": 1.3903,
      "step": 23880
    },
    {
      "epoch": 0.6,
      "grad_norm": 12.088221549987793,
      "learning_rate": 4.142372881355933e-06,
      "loss": 1.4658,
      "step": 23890
    },
    {
      "epoch": 0.6,
      "grad_norm": 9.102261543273926,
      "learning_rate": 4.135593220338983e-06,
      "loss": 1.3583,
      "step": 23900
    },
    {
      "epoch": 0.6,
      "grad_norm": 7.043952941894531,
      "learning_rate": 4.128813559322034e-06,
      "loss": 1.3913,
      "step": 23910
    },
    {
      "epoch": 0.6,
      "grad_norm": 2.0804154872894287,
      "learning_rate": 4.122033898305085e-06,
      "loss": 1.2291,
      "step": 23920
    },
    {
      "epoch": 0.6,
      "grad_norm": 7.723446846008301,
      "learning_rate": 4.115254237288136e-06,
      "loss": 1.375,
      "step": 23930
    },
    {
      "epoch": 0.6,
      "grad_norm": 10.418320655822754,
      "learning_rate": 4.108474576271187e-06,
      "loss": 1.3592,
      "step": 23940
    },
    {
      "epoch": 0.6,
      "grad_norm": 16.461467742919922,
      "learning_rate": 4.101694915254237e-06,
      "loss": 1.3827,
      "step": 23950
    },
    {
      "epoch": 0.6,
      "grad_norm": 13.934067726135254,
      "learning_rate": 4.094915254237289e-06,
      "loss": 1.1478,
      "step": 23960
    },
    {
      "epoch": 0.6,
      "grad_norm": 7.638941287994385,
      "learning_rate": 4.088135593220339e-06,
      "loss": 1.237,
      "step": 23970
    },
    {
      "epoch": 0.6,
      "grad_norm": 2.192106008529663,
      "learning_rate": 4.081355932203391e-06,
      "loss": 1.2372,
      "step": 23980
    },
    {
      "epoch": 0.6,
      "grad_norm": 8.600611686706543,
      "learning_rate": 4.074576271186441e-06,
      "loss": 1.2415,
      "step": 23990
    },
    {
      "epoch": 0.6,
      "grad_norm": 6.099268913269043,
      "learning_rate": 4.067796610169492e-06,
      "loss": 1.2564,
      "step": 24000
    },
    {
      "epoch": 0.6,
      "grad_norm": 6.219475269317627,
      "learning_rate": 4.061016949152542e-06,
      "loss": 1.1678,
      "step": 24010
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.8029667735099792,
      "learning_rate": 4.054237288135594e-06,
      "loss": 1.1462,
      "step": 24020
    },
    {
      "epoch": 0.6,
      "grad_norm": 12.5972261428833,
      "learning_rate": 4.047457627118644e-06,
      "loss": 1.2604,
      "step": 24030
    },
    {
      "epoch": 0.6,
      "grad_norm": 4.760711669921875,
      "learning_rate": 4.040677966101696e-06,
      "loss": 1.2439,
      "step": 24040
    },
    {
      "epoch": 0.6,
      "grad_norm": 7.038741588592529,
      "learning_rate": 4.033898305084746e-06,
      "loss": 1.205,
      "step": 24050
    },
    {
      "epoch": 0.6,
      "grad_norm": 8.526154518127441,
      "learning_rate": 4.027118644067797e-06,
      "loss": 1.4461,
      "step": 24060
    },
    {
      "epoch": 0.6,
      "grad_norm": 13.55893325805664,
      "learning_rate": 4.020338983050847e-06,
      "loss": 1.4073,
      "step": 24070
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.2988280057907104,
      "learning_rate": 4.013559322033899e-06,
      "loss": 1.3205,
      "step": 24080
    },
    {
      "epoch": 0.6,
      "grad_norm": 2.728271961212158,
      "learning_rate": 4.006779661016949e-06,
      "loss": 1.1551,
      "step": 24090
    },
    {
      "epoch": 0.6,
      "grad_norm": 5.839344024658203,
      "learning_rate": 4.000000000000001e-06,
      "loss": 1.3477,
      "step": 24100
    },
    {
      "epoch": 0.6,
      "grad_norm": 6.8664445877075195,
      "learning_rate": 3.993220338983051e-06,
      "loss": 1.1994,
      "step": 24110
    },
    {
      "epoch": 0.6,
      "grad_norm": 8.018701553344727,
      "learning_rate": 3.986440677966102e-06,
      "loss": 1.4816,
      "step": 24120
    },
    {
      "epoch": 0.6,
      "grad_norm": 6.4340925216674805,
      "learning_rate": 3.979661016949153e-06,
      "loss": 1.3655,
      "step": 24130
    },
    {
      "epoch": 0.6,
      "grad_norm": 2.2245099544525146,
      "learning_rate": 3.972881355932204e-06,
      "loss": 1.1541,
      "step": 24140
    },
    {
      "epoch": 0.6,
      "grad_norm": 9.884227752685547,
      "learning_rate": 3.966101694915255e-06,
      "loss": 1.3908,
      "step": 24150
    },
    {
      "epoch": 0.6,
      "grad_norm": 7.214573383331299,
      "learning_rate": 3.959322033898305e-06,
      "loss": 1.4161,
      "step": 24160
    },
    {
      "epoch": 0.6,
      "grad_norm": 9.771342277526855,
      "learning_rate": 3.952542372881356e-06,
      "loss": 1.4587,
      "step": 24170
    },
    {
      "epoch": 0.6,
      "grad_norm": 4.8412604331970215,
      "learning_rate": 3.945762711864407e-06,
      "loss": 1.2941,
      "step": 24180
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.9710540771484375,
      "learning_rate": 3.938983050847458e-06,
      "loss": 1.3964,
      "step": 24190
    },
    {
      "epoch": 0.6,
      "grad_norm": 3.252842903137207,
      "learning_rate": 3.932203389830509e-06,
      "loss": 1.4568,
      "step": 24200
    },
    {
      "epoch": 0.61,
      "grad_norm": 3.3610448837280273,
      "learning_rate": 3.925423728813559e-06,
      "loss": 1.3459,
      "step": 24210
    },
    {
      "epoch": 0.61,
      "grad_norm": 10.054492950439453,
      "learning_rate": 3.918644067796611e-06,
      "loss": 1.3659,
      "step": 24220
    },
    {
      "epoch": 0.61,
      "grad_norm": 8.003775596618652,
      "learning_rate": 3.911864406779661e-06,
      "loss": 1.486,
      "step": 24230
    },
    {
      "epoch": 0.61,
      "grad_norm": 7.817092418670654,
      "learning_rate": 3.905084745762713e-06,
      "loss": 1.3845,
      "step": 24240
    },
    {
      "epoch": 0.61,
      "grad_norm": 7.940910339355469,
      "learning_rate": 3.898305084745763e-06,
      "loss": 1.2693,
      "step": 24250
    },
    {
      "epoch": 0.61,
      "grad_norm": 7.764880180358887,
      "learning_rate": 3.891525423728814e-06,
      "loss": 1.2101,
      "step": 24260
    },
    {
      "epoch": 0.61,
      "grad_norm": 3.04939866065979,
      "learning_rate": 3.884745762711864e-06,
      "loss": 1.3573,
      "step": 24270
    },
    {
      "epoch": 0.61,
      "grad_norm": 8.402636528015137,
      "learning_rate": 3.877966101694916e-06,
      "loss": 1.2909,
      "step": 24280
    },
    {
      "epoch": 0.61,
      "grad_norm": 5.9454345703125,
      "learning_rate": 3.871186440677966e-06,
      "loss": 1.3091,
      "step": 24290
    },
    {
      "epoch": 0.61,
      "grad_norm": 7.828911781311035,
      "learning_rate": 3.864406779661018e-06,
      "loss": 1.3207,
      "step": 24300
    },
    {
      "epoch": 0.61,
      "grad_norm": 6.573239326477051,
      "learning_rate": 3.857627118644068e-06,
      "loss": 1.32,
      "step": 24310
    },
    {
      "epoch": 0.61,
      "grad_norm": 9.533820152282715,
      "learning_rate": 3.850847457627119e-06,
      "loss": 1.4469,
      "step": 24320
    },
    {
      "epoch": 0.61,
      "grad_norm": 10.16154670715332,
      "learning_rate": 3.844067796610169e-06,
      "loss": 1.2962,
      "step": 24330
    },
    {
      "epoch": 0.61,
      "grad_norm": 8.2395658493042,
      "learning_rate": 3.837288135593221e-06,
      "loss": 1.3448,
      "step": 24340
    },
    {
      "epoch": 0.61,
      "grad_norm": 18.230348587036133,
      "learning_rate": 3.830508474576271e-06,
      "loss": 1.3504,
      "step": 24350
    },
    {
      "epoch": 0.61,
      "grad_norm": 7.205873012542725,
      "learning_rate": 3.823728813559323e-06,
      "loss": 1.252,
      "step": 24360
    },
    {
      "epoch": 0.61,
      "grad_norm": 6.645435810089111,
      "learning_rate": 3.816949152542373e-06,
      "loss": 1.3114,
      "step": 24370
    },
    {
      "epoch": 0.61,
      "grad_norm": 4.2720842361450195,
      "learning_rate": 3.8101694915254238e-06,
      "loss": 1.3314,
      "step": 24380
    },
    {
      "epoch": 0.61,
      "grad_norm": 6.6986775398254395,
      "learning_rate": 3.8033898305084748e-06,
      "loss": 1.2547,
      "step": 24390
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.2966694831848145,
      "learning_rate": 3.7966101694915257e-06,
      "loss": 1.1595,
      "step": 24400
    },
    {
      "epoch": 0.61,
      "grad_norm": 4.688185214996338,
      "learning_rate": 3.7898305084745767e-06,
      "loss": 1.3229,
      "step": 24410
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.245527744293213,
      "learning_rate": 3.7830508474576273e-06,
      "loss": 1.4375,
      "step": 24420
    },
    {
      "epoch": 0.61,
      "grad_norm": 9.676859855651855,
      "learning_rate": 3.7762711864406782e-06,
      "loss": 1.2591,
      "step": 24430
    },
    {
      "epoch": 0.61,
      "grad_norm": 6.991977691650391,
      "learning_rate": 3.7694915254237292e-06,
      "loss": 1.1408,
      "step": 24440
    },
    {
      "epoch": 0.61,
      "grad_norm": 4.704678058624268,
      "learning_rate": 3.76271186440678e-06,
      "loss": 1.1997,
      "step": 24450
    },
    {
      "epoch": 0.61,
      "grad_norm": 4.673946380615234,
      "learning_rate": 3.755932203389831e-06,
      "loss": 1.3852,
      "step": 24460
    },
    {
      "epoch": 0.61,
      "grad_norm": 2.77323579788208,
      "learning_rate": 3.7491525423728813e-06,
      "loss": 1.2162,
      "step": 24470
    },
    {
      "epoch": 0.61,
      "grad_norm": 5.128834247589111,
      "learning_rate": 3.7423728813559323e-06,
      "loss": 1.3063,
      "step": 24480
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.9332771301269531,
      "learning_rate": 3.7355932203389833e-06,
      "loss": 1.4053,
      "step": 24490
    },
    {
      "epoch": 0.61,
      "grad_norm": 4.355793476104736,
      "learning_rate": 3.7288135593220342e-06,
      "loss": 1.4419,
      "step": 24500
    },
    {
      "epoch": 0.61,
      "grad_norm": 8.115254402160645,
      "learning_rate": 3.7220338983050852e-06,
      "loss": 1.2131,
      "step": 24510
    },
    {
      "epoch": 0.61,
      "grad_norm": 4.985610485076904,
      "learning_rate": 3.715254237288136e-06,
      "loss": 1.2573,
      "step": 24520
    },
    {
      "epoch": 0.61,
      "grad_norm": 6.433516502380371,
      "learning_rate": 3.7084745762711867e-06,
      "loss": 1.228,
      "step": 24530
    },
    {
      "epoch": 0.61,
      "grad_norm": 6.564544200897217,
      "learning_rate": 3.7016949152542377e-06,
      "loss": 1.2727,
      "step": 24540
    },
    {
      "epoch": 0.61,
      "grad_norm": 9.307107925415039,
      "learning_rate": 3.6949152542372883e-06,
      "loss": 1.4235,
      "step": 24550
    },
    {
      "epoch": 0.61,
      "grad_norm": 3.663797616958618,
      "learning_rate": 3.6881355932203393e-06,
      "loss": 1.3806,
      "step": 24560
    },
    {
      "epoch": 0.61,
      "grad_norm": 5.164207458496094,
      "learning_rate": 3.6813559322033902e-06,
      "loss": 1.2492,
      "step": 24570
    },
    {
      "epoch": 0.61,
      "grad_norm": 8.45071029663086,
      "learning_rate": 3.6745762711864408e-06,
      "loss": 1.333,
      "step": 24580
    },
    {
      "epoch": 0.61,
      "grad_norm": 14.722311973571777,
      "learning_rate": 3.6677966101694918e-06,
      "loss": 1.325,
      "step": 24590
    },
    {
      "epoch": 0.61,
      "grad_norm": 9.298062324523926,
      "learning_rate": 3.6610169491525427e-06,
      "loss": 1.4456,
      "step": 24600
    },
    {
      "epoch": 0.62,
      "grad_norm": 7.502054214477539,
      "learning_rate": 3.6542372881355937e-06,
      "loss": 1.2475,
      "step": 24610
    },
    {
      "epoch": 0.62,
      "grad_norm": 12.15906047821045,
      "learning_rate": 3.6474576271186447e-06,
      "loss": 1.2091,
      "step": 24620
    },
    {
      "epoch": 0.62,
      "grad_norm": 2.4973747730255127,
      "learning_rate": 3.640677966101695e-06,
      "loss": 1.2963,
      "step": 24630
    },
    {
      "epoch": 0.62,
      "grad_norm": 3.5087554454803467,
      "learning_rate": 3.633898305084746e-06,
      "loss": 1.2813,
      "step": 24640
    },
    {
      "epoch": 0.62,
      "grad_norm": 8.784186363220215,
      "learning_rate": 3.6271186440677968e-06,
      "loss": 1.466,
      "step": 24650
    },
    {
      "epoch": 0.62,
      "grad_norm": 6.409433841705322,
      "learning_rate": 3.6203389830508478e-06,
      "loss": 1.3055,
      "step": 24660
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.883354663848877,
      "learning_rate": 3.6135593220338987e-06,
      "loss": 1.407,
      "step": 24670
    },
    {
      "epoch": 0.62,
      "grad_norm": 2.1597604751586914,
      "learning_rate": 3.6067796610169493e-06,
      "loss": 1.2795,
      "step": 24680
    },
    {
      "epoch": 0.62,
      "grad_norm": 9.558140754699707,
      "learning_rate": 3.6000000000000003e-06,
      "loss": 1.3779,
      "step": 24690
    },
    {
      "epoch": 0.62,
      "grad_norm": 16.530136108398438,
      "learning_rate": 3.5932203389830512e-06,
      "loss": 1.2392,
      "step": 24700
    },
    {
      "epoch": 0.62,
      "grad_norm": 6.770777702331543,
      "learning_rate": 3.5864406779661022e-06,
      "loss": 1.2862,
      "step": 24710
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.7149447202682495,
      "learning_rate": 3.579661016949153e-06,
      "loss": 1.3564,
      "step": 24720
    },
    {
      "epoch": 0.62,
      "grad_norm": 8.497549057006836,
      "learning_rate": 3.5728813559322033e-06,
      "loss": 1.3284,
      "step": 24730
    },
    {
      "epoch": 0.62,
      "grad_norm": 6.891111373901367,
      "learning_rate": 3.5661016949152543e-06,
      "loss": 1.2681,
      "step": 24740
    },
    {
      "epoch": 0.62,
      "grad_norm": 10.200005531311035,
      "learning_rate": 3.5593220338983053e-06,
      "loss": 1.3538,
      "step": 24750
    },
    {
      "epoch": 0.62,
      "grad_norm": 5.764871597290039,
      "learning_rate": 3.5525423728813563e-06,
      "loss": 1.0213,
      "step": 24760
    },
    {
      "epoch": 0.62,
      "grad_norm": 8.506179809570312,
      "learning_rate": 3.5457627118644072e-06,
      "loss": 1.3816,
      "step": 24770
    },
    {
      "epoch": 0.62,
      "grad_norm": 7.045341968536377,
      "learning_rate": 3.538983050847458e-06,
      "loss": 1.4771,
      "step": 24780
    },
    {
      "epoch": 0.62,
      "grad_norm": 5.495665550231934,
      "learning_rate": 3.5322033898305088e-06,
      "loss": 1.3835,
      "step": 24790
    },
    {
      "epoch": 0.62,
      "grad_norm": 12.816548347473145,
      "learning_rate": 3.5254237288135597e-06,
      "loss": 1.2952,
      "step": 24800
    },
    {
      "epoch": 0.62,
      "grad_norm": 4.715461254119873,
      "learning_rate": 3.5186440677966103e-06,
      "loss": 1.3881,
      "step": 24810
    },
    {
      "epoch": 0.62,
      "grad_norm": 6.540507793426514,
      "learning_rate": 3.5118644067796613e-06,
      "loss": 1.454,
      "step": 24820
    },
    {
      "epoch": 0.62,
      "grad_norm": 2.6611897945404053,
      "learning_rate": 3.5050847457627122e-06,
      "loss": 1.2097,
      "step": 24830
    },
    {
      "epoch": 0.62,
      "grad_norm": 8.780037879943848,
      "learning_rate": 3.498305084745763e-06,
      "loss": 1.3843,
      "step": 24840
    },
    {
      "epoch": 0.62,
      "grad_norm": 7.406546592712402,
      "learning_rate": 3.4915254237288138e-06,
      "loss": 1.2006,
      "step": 24850
    },
    {
      "epoch": 0.62,
      "grad_norm": 2.9691720008850098,
      "learning_rate": 3.4847457627118648e-06,
      "loss": 1.3474,
      "step": 24860
    },
    {
      "epoch": 0.62,
      "grad_norm": 5.585392475128174,
      "learning_rate": 3.4779661016949157e-06,
      "loss": 1.2814,
      "step": 24870
    },
    {
      "epoch": 0.62,
      "grad_norm": 5.655838489532471,
      "learning_rate": 3.4711864406779667e-06,
      "loss": 1.4713,
      "step": 24880
    },
    {
      "epoch": 0.62,
      "grad_norm": 5.474581718444824,
      "learning_rate": 3.464406779661017e-06,
      "loss": 1.3336,
      "step": 24890
    },
    {
      "epoch": 0.62,
      "grad_norm": 6.543440818786621,
      "learning_rate": 3.457627118644068e-06,
      "loss": 1.3815,
      "step": 24900
    },
    {
      "epoch": 0.62,
      "grad_norm": 4.729587078094482,
      "learning_rate": 3.450847457627119e-06,
      "loss": 1.1675,
      "step": 24910
    },
    {
      "epoch": 0.62,
      "grad_norm": 13.120586395263672,
      "learning_rate": 3.4440677966101698e-06,
      "loss": 1.3373,
      "step": 24920
    },
    {
      "epoch": 0.62,
      "grad_norm": 2.1029560565948486,
      "learning_rate": 3.4372881355932207e-06,
      "loss": 1.4426,
      "step": 24930
    },
    {
      "epoch": 0.62,
      "grad_norm": 4.8036322593688965,
      "learning_rate": 3.4305084745762713e-06,
      "loss": 1.4996,
      "step": 24940
    },
    {
      "epoch": 0.62,
      "grad_norm": 5.592566967010498,
      "learning_rate": 3.4237288135593223e-06,
      "loss": 1.2189,
      "step": 24950
    },
    {
      "epoch": 0.62,
      "grad_norm": 2.1203982830047607,
      "learning_rate": 3.4169491525423733e-06,
      "loss": 1.3607,
      "step": 24960
    },
    {
      "epoch": 0.62,
      "grad_norm": 4.516960144042969,
      "learning_rate": 3.4101694915254242e-06,
      "loss": 1.3102,
      "step": 24970
    },
    {
      "epoch": 0.62,
      "grad_norm": 6.249980449676514,
      "learning_rate": 3.403389830508475e-06,
      "loss": 1.136,
      "step": 24980
    },
    {
      "epoch": 0.62,
      "grad_norm": 13.181388854980469,
      "learning_rate": 3.3966101694915253e-06,
      "loss": 1.4147,
      "step": 24990
    },
    {
      "epoch": 0.62,
      "grad_norm": 2.8158318996429443,
      "learning_rate": 3.3898305084745763e-06,
      "loss": 1.3012,
      "step": 25000
    },
    {
      "epoch": 0.62,
      "eval_loss": 1.2961933612823486,
      "eval_runtime": 122.3461,
      "eval_samples_per_second": 8.174,
      "eval_steps_per_second": 8.174,
      "step": 25000
    },
    {
      "epoch": 0.63,
      "grad_norm": 7.878478527069092,
      "learning_rate": 3.3830508474576273e-06,
      "loss": 1.5645,
      "step": 25010
    },
    {
      "epoch": 0.63,
      "grad_norm": 4.464993000030518,
      "learning_rate": 3.3762711864406783e-06,
      "loss": 1.2522,
      "step": 25020
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.824704885482788,
      "learning_rate": 3.3694915254237292e-06,
      "loss": 1.1667,
      "step": 25030
    },
    {
      "epoch": 0.63,
      "grad_norm": 4.979220390319824,
      "learning_rate": 3.3627118644067802e-06,
      "loss": 1.4592,
      "step": 25040
    },
    {
      "epoch": 0.63,
      "grad_norm": 7.964636325836182,
      "learning_rate": 3.3559322033898308e-06,
      "loss": 1.3505,
      "step": 25050
    },
    {
      "epoch": 0.63,
      "grad_norm": 3.848740816116333,
      "learning_rate": 3.3491525423728817e-06,
      "loss": 1.3424,
      "step": 25060
    },
    {
      "epoch": 0.63,
      "grad_norm": 5.446021556854248,
      "learning_rate": 3.3423728813559327e-06,
      "loss": 1.3514,
      "step": 25070
    },
    {
      "epoch": 0.63,
      "grad_norm": 4.195797920227051,
      "learning_rate": 3.3355932203389833e-06,
      "loss": 1.4442,
      "step": 25080
    },
    {
      "epoch": 0.63,
      "grad_norm": 5.259161949157715,
      "learning_rate": 3.3288135593220343e-06,
      "loss": 1.4244,
      "step": 25090
    },
    {
      "epoch": 0.63,
      "grad_norm": 4.59972620010376,
      "learning_rate": 3.322033898305085e-06,
      "loss": 1.3104,
      "step": 25100
    },
    {
      "epoch": 0.63,
      "grad_norm": 4.761903762817383,
      "learning_rate": 3.3152542372881358e-06,
      "loss": 1.3031,
      "step": 25110
    },
    {
      "epoch": 0.63,
      "grad_norm": 10.678994178771973,
      "learning_rate": 3.3084745762711868e-06,
      "loss": 1.372,
      "step": 25120
    },
    {
      "epoch": 0.63,
      "grad_norm": 5.033021926879883,
      "learning_rate": 3.3016949152542377e-06,
      "loss": 1.3018,
      "step": 25130
    },
    {
      "epoch": 0.63,
      "grad_norm": 14.870203971862793,
      "learning_rate": 3.2949152542372887e-06,
      "loss": 1.3284,
      "step": 25140
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.7636396884918213,
      "learning_rate": 3.288135593220339e-06,
      "loss": 1.3044,
      "step": 25150
    },
    {
      "epoch": 0.63,
      "grad_norm": 2.6163928508758545,
      "learning_rate": 3.28135593220339e-06,
      "loss": 1.1875,
      "step": 25160
    },
    {
      "epoch": 0.63,
      "grad_norm": 4.163614273071289,
      "learning_rate": 3.274576271186441e-06,
      "loss": 1.3458,
      "step": 25170
    },
    {
      "epoch": 0.63,
      "grad_norm": 9.529121398925781,
      "learning_rate": 3.2677966101694918e-06,
      "loss": 1.2659,
      "step": 25180
    },
    {
      "epoch": 0.63,
      "grad_norm": 10.017566680908203,
      "learning_rate": 3.2610169491525428e-06,
      "loss": 1.4239,
      "step": 25190
    },
    {
      "epoch": 0.63,
      "grad_norm": 4.03933048248291,
      "learning_rate": 3.2542372881355933e-06,
      "loss": 1.252,
      "step": 25200
    },
    {
      "epoch": 0.63,
      "grad_norm": 6.725574970245361,
      "learning_rate": 3.2474576271186443e-06,
      "loss": 1.1699,
      "step": 25210
    },
    {
      "epoch": 0.63,
      "grad_norm": 8.446992874145508,
      "learning_rate": 3.2406779661016953e-06,
      "loss": 1.2996,
      "step": 25220
    },
    {
      "epoch": 0.63,
      "grad_norm": 15.130813598632812,
      "learning_rate": 3.2338983050847462e-06,
      "loss": 1.1601,
      "step": 25230
    },
    {
      "epoch": 0.63,
      "grad_norm": 8.915867805480957,
      "learning_rate": 3.2271186440677972e-06,
      "loss": 1.4512,
      "step": 25240
    },
    {
      "epoch": 0.63,
      "grad_norm": 3.813676595687866,
      "learning_rate": 3.2203389830508473e-06,
      "loss": 1.4311,
      "step": 25250
    },
    {
      "epoch": 0.63,
      "grad_norm": 8.306417465209961,
      "learning_rate": 3.2135593220338983e-06,
      "loss": 1.2554,
      "step": 25260
    },
    {
      "epoch": 0.63,
      "grad_norm": 4.557586193084717,
      "learning_rate": 3.2067796610169493e-06,
      "loss": 1.4084,
      "step": 25270
    },
    {
      "epoch": 0.63,
      "grad_norm": 3.206493616104126,
      "learning_rate": 3.2000000000000003e-06,
      "loss": 1.3265,
      "step": 25280
    },
    {
      "epoch": 0.63,
      "grad_norm": 4.9644622802734375,
      "learning_rate": 3.1932203389830513e-06,
      "loss": 1.3932,
      "step": 25290
    },
    {
      "epoch": 0.63,
      "grad_norm": 3.4366214275360107,
      "learning_rate": 3.186440677966102e-06,
      "loss": 1.4233,
      "step": 25300
    },
    {
      "epoch": 0.63,
      "grad_norm": 6.7854461669921875,
      "learning_rate": 3.1796610169491528e-06,
      "loss": 1.2838,
      "step": 25310
    },
    {
      "epoch": 0.63,
      "grad_norm": 6.945521831512451,
      "learning_rate": 3.1728813559322038e-06,
      "loss": 1.2936,
      "step": 25320
    },
    {
      "epoch": 0.63,
      "grad_norm": 4.550187587738037,
      "learning_rate": 3.1661016949152547e-06,
      "loss": 1.3477,
      "step": 25330
    },
    {
      "epoch": 0.63,
      "grad_norm": 5.256103515625,
      "learning_rate": 3.1593220338983053e-06,
      "loss": 1.4062,
      "step": 25340
    },
    {
      "epoch": 0.63,
      "grad_norm": 11.691208839416504,
      "learning_rate": 3.1525423728813563e-06,
      "loss": 1.1473,
      "step": 25350
    },
    {
      "epoch": 0.63,
      "grad_norm": 5.051747798919678,
      "learning_rate": 3.145762711864407e-06,
      "loss": 1.4299,
      "step": 25360
    },
    {
      "epoch": 0.63,
      "grad_norm": 2.562920093536377,
      "learning_rate": 3.138983050847458e-06,
      "loss": 1.2428,
      "step": 25370
    },
    {
      "epoch": 0.63,
      "grad_norm": 7.520709037780762,
      "learning_rate": 3.1322033898305088e-06,
      "loss": 1.1736,
      "step": 25380
    },
    {
      "epoch": 0.63,
      "grad_norm": 5.788994789123535,
      "learning_rate": 3.1254237288135598e-06,
      "loss": 1.3141,
      "step": 25390
    },
    {
      "epoch": 0.64,
      "grad_norm": 17.664766311645508,
      "learning_rate": 3.1186440677966107e-06,
      "loss": 1.357,
      "step": 25400
    },
    {
      "epoch": 0.64,
      "grad_norm": 4.366672992706299,
      "learning_rate": 3.111864406779661e-06,
      "loss": 1.3631,
      "step": 25410
    },
    {
      "epoch": 0.64,
      "grad_norm": 15.109098434448242,
      "learning_rate": 3.105084745762712e-06,
      "loss": 1.5056,
      "step": 25420
    },
    {
      "epoch": 0.64,
      "grad_norm": 12.550411224365234,
      "learning_rate": 3.098305084745763e-06,
      "loss": 1.2255,
      "step": 25430
    },
    {
      "epoch": 0.64,
      "grad_norm": 5.603880882263184,
      "learning_rate": 3.091525423728814e-06,
      "loss": 1.4206,
      "step": 25440
    },
    {
      "epoch": 0.64,
      "grad_norm": 5.365425109863281,
      "learning_rate": 3.0847457627118648e-06,
      "loss": 1.1442,
      "step": 25450
    },
    {
      "epoch": 0.64,
      "grad_norm": 6.900291442871094,
      "learning_rate": 3.0779661016949153e-06,
      "loss": 1.3042,
      "step": 25460
    },
    {
      "epoch": 0.64,
      "grad_norm": 6.29402494430542,
      "learning_rate": 3.0711864406779663e-06,
      "loss": 1.2615,
      "step": 25470
    },
    {
      "epoch": 0.64,
      "grad_norm": 11.43036937713623,
      "learning_rate": 3.0644067796610173e-06,
      "loss": 1.1831,
      "step": 25480
    },
    {
      "epoch": 0.64,
      "grad_norm": 8.487439155578613,
      "learning_rate": 3.0576271186440683e-06,
      "loss": 1.2347,
      "step": 25490
    },
    {
      "epoch": 0.64,
      "grad_norm": 4.873865127563477,
      "learning_rate": 3.0508474576271192e-06,
      "loss": 1.0908,
      "step": 25500
    },
    {
      "epoch": 0.64,
      "grad_norm": 2.734248161315918,
      "learning_rate": 3.0440677966101694e-06,
      "loss": 1.0498,
      "step": 25510
    },
    {
      "epoch": 0.64,
      "grad_norm": 12.489617347717285,
      "learning_rate": 3.0372881355932203e-06,
      "loss": 1.2103,
      "step": 25520
    },
    {
      "epoch": 0.64,
      "grad_norm": 11.316421508789062,
      "learning_rate": 3.0305084745762713e-06,
      "loss": 1.5457,
      "step": 25530
    },
    {
      "epoch": 0.64,
      "grad_norm": 7.098801612854004,
      "learning_rate": 3.0237288135593223e-06,
      "loss": 1.4746,
      "step": 25540
    },
    {
      "epoch": 0.64,
      "grad_norm": 5.496173858642578,
      "learning_rate": 3.0169491525423733e-06,
      "loss": 1.296,
      "step": 25550
    },
    {
      "epoch": 0.64,
      "grad_norm": 3.706704616546631,
      "learning_rate": 3.010169491525424e-06,
      "loss": 1.4458,
      "step": 25560
    },
    {
      "epoch": 0.64,
      "grad_norm": 10.640968322753906,
      "learning_rate": 3.003389830508475e-06,
      "loss": 1.2303,
      "step": 25570
    },
    {
      "epoch": 0.64,
      "grad_norm": 9.76960563659668,
      "learning_rate": 2.9966101694915258e-06,
      "loss": 1.3352,
      "step": 25580
    },
    {
      "epoch": 0.64,
      "grad_norm": 6.274062633514404,
      "learning_rate": 2.9898305084745768e-06,
      "loss": 1.4795,
      "step": 25590
    },
    {
      "epoch": 0.64,
      "grad_norm": 2.7021098136901855,
      "learning_rate": 2.9830508474576277e-06,
      "loss": 1.3396,
      "step": 25600
    },
    {
      "epoch": 0.64,
      "grad_norm": 4.872988224029541,
      "learning_rate": 2.9762711864406783e-06,
      "loss": 1.4767,
      "step": 25610
    },
    {
      "epoch": 0.64,
      "grad_norm": 8.984478950500488,
      "learning_rate": 2.969491525423729e-06,
      "loss": 1.3939,
      "step": 25620
    },
    {
      "epoch": 0.64,
      "grad_norm": 18.204336166381836,
      "learning_rate": 2.96271186440678e-06,
      "loss": 1.3875,
      "step": 25630
    },
    {
      "epoch": 0.64,
      "grad_norm": 6.397688388824463,
      "learning_rate": 2.955932203389831e-06,
      "loss": 1.1992,
      "step": 25640
    },
    {
      "epoch": 0.64,
      "grad_norm": 3.8524389266967773,
      "learning_rate": 2.9491525423728818e-06,
      "loss": 1.5454,
      "step": 25650
    },
    {
      "epoch": 0.64,
      "grad_norm": 6.4902191162109375,
      "learning_rate": 2.9423728813559327e-06,
      "loss": 1.3023,
      "step": 25660
    },
    {
      "epoch": 0.64,
      "grad_norm": 2.5433766841888428,
      "learning_rate": 2.935593220338983e-06,
      "loss": 1.3424,
      "step": 25670
    },
    {
      "epoch": 0.64,
      "grad_norm": 2.7309176921844482,
      "learning_rate": 2.928813559322034e-06,
      "loss": 1.2601,
      "step": 25680
    },
    {
      "epoch": 0.64,
      "grad_norm": 5.759544849395752,
      "learning_rate": 2.922033898305085e-06,
      "loss": 1.4213,
      "step": 25690
    },
    {
      "epoch": 0.64,
      "grad_norm": 8.862116813659668,
      "learning_rate": 2.915254237288136e-06,
      "loss": 1.2039,
      "step": 25700
    },
    {
      "epoch": 0.64,
      "grad_norm": 4.360088348388672,
      "learning_rate": 2.9084745762711868e-06,
      "loss": 1.2741,
      "step": 25710
    },
    {
      "epoch": 0.64,
      "grad_norm": 6.417861461639404,
      "learning_rate": 2.9016949152542373e-06,
      "loss": 1.173,
      "step": 25720
    },
    {
      "epoch": 0.64,
      "grad_norm": 4.8081254959106445,
      "learning_rate": 2.8949152542372883e-06,
      "loss": 1.3977,
      "step": 25730
    },
    {
      "epoch": 0.64,
      "grad_norm": 4.581600189208984,
      "learning_rate": 2.8881355932203393e-06,
      "loss": 1.2773,
      "step": 25740
    },
    {
      "epoch": 0.64,
      "grad_norm": 7.737278938293457,
      "learning_rate": 2.8813559322033903e-06,
      "loss": 1.3239,
      "step": 25750
    },
    {
      "epoch": 0.64,
      "grad_norm": 8.583956718444824,
      "learning_rate": 2.8745762711864412e-06,
      "loss": 1.3715,
      "step": 25760
    },
    {
      "epoch": 0.64,
      "grad_norm": 6.567660331726074,
      "learning_rate": 2.8677966101694914e-06,
      "loss": 1.0883,
      "step": 25770
    },
    {
      "epoch": 0.64,
      "grad_norm": 8.99410343170166,
      "learning_rate": 2.8610169491525424e-06,
      "loss": 1.213,
      "step": 25780
    },
    {
      "epoch": 0.64,
      "grad_norm": 7.688558101654053,
      "learning_rate": 2.8542372881355933e-06,
      "loss": 1.147,
      "step": 25790
    },
    {
      "epoch": 0.65,
      "grad_norm": 9.962355613708496,
      "learning_rate": 2.8474576271186443e-06,
      "loss": 1.243,
      "step": 25800
    },
    {
      "epoch": 0.65,
      "grad_norm": 2.2275478839874268,
      "learning_rate": 2.8406779661016953e-06,
      "loss": 1.2328,
      "step": 25810
    },
    {
      "epoch": 0.65,
      "grad_norm": 3.4127414226531982,
      "learning_rate": 2.833898305084746e-06,
      "loss": 1.4244,
      "step": 25820
    },
    {
      "epoch": 0.65,
      "grad_norm": 12.15282917022705,
      "learning_rate": 2.827118644067797e-06,
      "loss": 1.401,
      "step": 25830
    },
    {
      "epoch": 0.65,
      "grad_norm": 8.007610321044922,
      "learning_rate": 2.820338983050848e-06,
      "loss": 1.3701,
      "step": 25840
    },
    {
      "epoch": 0.65,
      "grad_norm": 9.589988708496094,
      "learning_rate": 2.8135593220338988e-06,
      "loss": 1.2886,
      "step": 25850
    },
    {
      "epoch": 0.65,
      "grad_norm": 4.063002109527588,
      "learning_rate": 2.8067796610169497e-06,
      "loss": 1.3383,
      "step": 25860
    },
    {
      "epoch": 0.65,
      "grad_norm": 2.1042330265045166,
      "learning_rate": 2.8000000000000003e-06,
      "loss": 1.2753,
      "step": 25870
    },
    {
      "epoch": 0.65,
      "grad_norm": 23.256053924560547,
      "learning_rate": 2.793220338983051e-06,
      "loss": 1.3617,
      "step": 25880
    },
    {
      "epoch": 0.65,
      "grad_norm": 7.6475911140441895,
      "learning_rate": 2.786440677966102e-06,
      "loss": 1.2862,
      "step": 25890
    },
    {
      "epoch": 0.65,
      "grad_norm": 16.81471824645996,
      "learning_rate": 2.779661016949153e-06,
      "loss": 1.5188,
      "step": 25900
    },
    {
      "epoch": 0.65,
      "grad_norm": 10.20080852508545,
      "learning_rate": 2.7728813559322038e-06,
      "loss": 1.5414,
      "step": 25910
    },
    {
      "epoch": 0.65,
      "grad_norm": 9.043648719787598,
      "learning_rate": 2.7661016949152548e-06,
      "loss": 1.3299,
      "step": 25920
    },
    {
      "epoch": 0.65,
      "grad_norm": 4.330277442932129,
      "learning_rate": 2.7593220338983053e-06,
      "loss": 1.1826,
      "step": 25930
    },
    {
      "epoch": 0.65,
      "grad_norm": 7.962332725524902,
      "learning_rate": 2.752542372881356e-06,
      "loss": 1.3566,
      "step": 25940
    },
    {
      "epoch": 0.65,
      "grad_norm": 8.763781547546387,
      "learning_rate": 2.745762711864407e-06,
      "loss": 1.3485,
      "step": 25950
    },
    {
      "epoch": 0.65,
      "grad_norm": 6.160745620727539,
      "learning_rate": 2.738983050847458e-06,
      "loss": 1.303,
      "step": 25960
    },
    {
      "epoch": 0.65,
      "grad_norm": 2.6451609134674072,
      "learning_rate": 2.732203389830509e-06,
      "loss": 1.4117,
      "step": 25970
    },
    {
      "epoch": 0.65,
      "grad_norm": 10.857940673828125,
      "learning_rate": 2.7254237288135593e-06,
      "loss": 1.1566,
      "step": 25980
    },
    {
      "epoch": 0.65,
      "grad_norm": 5.1549601554870605,
      "learning_rate": 2.7186440677966103e-06,
      "loss": 1.2738,
      "step": 25990
    },
    {
      "epoch": 0.65,
      "grad_norm": 7.437528610229492,
      "learning_rate": 2.7118644067796613e-06,
      "loss": 1.4507,
      "step": 26000
    },
    {
      "epoch": 0.65,
      "grad_norm": 6.80765962600708,
      "learning_rate": 2.7050847457627123e-06,
      "loss": 1.1497,
      "step": 26010
    },
    {
      "epoch": 0.65,
      "grad_norm": 6.173390865325928,
      "learning_rate": 2.6983050847457633e-06,
      "loss": 1.3426,
      "step": 26020
    },
    {
      "epoch": 0.65,
      "grad_norm": 6.7725911140441895,
      "learning_rate": 2.6915254237288134e-06,
      "loss": 1.3951,
      "step": 26030
    },
    {
      "epoch": 0.65,
      "grad_norm": 6.6503777503967285,
      "learning_rate": 2.6847457627118644e-06,
      "loss": 1.4507,
      "step": 26040
    },
    {
      "epoch": 0.65,
      "grad_norm": 5.210537433624268,
      "learning_rate": 2.6779661016949153e-06,
      "loss": 1.3876,
      "step": 26050
    },
    {
      "epoch": 0.65,
      "grad_norm": 3.615936756134033,
      "learning_rate": 2.6711864406779663e-06,
      "loss": 1.4476,
      "step": 26060
    },
    {
      "epoch": 0.65,
      "grad_norm": 9.065774917602539,
      "learning_rate": 2.6644067796610173e-06,
      "loss": 1.3124,
      "step": 26070
    },
    {
      "epoch": 0.65,
      "grad_norm": 13.187819480895996,
      "learning_rate": 2.657627118644068e-06,
      "loss": 1.2038,
      "step": 26080
    },
    {
      "epoch": 0.65,
      "grad_norm": 13.807534217834473,
      "learning_rate": 2.650847457627119e-06,
      "loss": 1.2118,
      "step": 26090
    },
    {
      "epoch": 0.65,
      "grad_norm": 16.19401741027832,
      "learning_rate": 2.64406779661017e-06,
      "loss": 1.2734,
      "step": 26100
    },
    {
      "epoch": 0.65,
      "grad_norm": 7.493824005126953,
      "learning_rate": 2.6372881355932208e-06,
      "loss": 1.3692,
      "step": 26110
    },
    {
      "epoch": 0.65,
      "grad_norm": 16.29339599609375,
      "learning_rate": 2.6305084745762718e-06,
      "loss": 1.3132,
      "step": 26120
    },
    {
      "epoch": 0.65,
      "grad_norm": 6.331333160400391,
      "learning_rate": 2.6237288135593223e-06,
      "loss": 1.3203,
      "step": 26130
    },
    {
      "epoch": 0.65,
      "grad_norm": 6.448307991027832,
      "learning_rate": 2.616949152542373e-06,
      "loss": 1.2753,
      "step": 26140
    },
    {
      "epoch": 0.65,
      "grad_norm": 6.896134376525879,
      "learning_rate": 2.610169491525424e-06,
      "loss": 1.349,
      "step": 26150
    },
    {
      "epoch": 0.65,
      "grad_norm": 5.189770698547363,
      "learning_rate": 2.603389830508475e-06,
      "loss": 1.289,
      "step": 26160
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.7721081972122192,
      "learning_rate": 2.596610169491526e-06,
      "loss": 1.4432,
      "step": 26170
    },
    {
      "epoch": 0.65,
      "grad_norm": 3.9717156887054443,
      "learning_rate": 2.5898305084745768e-06,
      "loss": 1.2982,
      "step": 26180
    },
    {
      "epoch": 0.65,
      "grad_norm": 19.084896087646484,
      "learning_rate": 2.5830508474576273e-06,
      "loss": 1.3897,
      "step": 26190
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.6735248565673828,
      "learning_rate": 2.576271186440678e-06,
      "loss": 1.4137,
      "step": 26200
    },
    {
      "epoch": 0.66,
      "grad_norm": 8.025382041931152,
      "learning_rate": 2.569491525423729e-06,
      "loss": 1.2629,
      "step": 26210
    },
    {
      "epoch": 0.66,
      "grad_norm": 4.695014476776123,
      "learning_rate": 2.56271186440678e-06,
      "loss": 1.5102,
      "step": 26220
    },
    {
      "epoch": 0.66,
      "grad_norm": 4.135346412658691,
      "learning_rate": 2.555932203389831e-06,
      "loss": 1.1628,
      "step": 26230
    },
    {
      "epoch": 0.66,
      "grad_norm": 6.607401371002197,
      "learning_rate": 2.5491525423728814e-06,
      "loss": 1.305,
      "step": 26240
    },
    {
      "epoch": 0.66,
      "grad_norm": 17.407390594482422,
      "learning_rate": 2.5423728813559323e-06,
      "loss": 1.3841,
      "step": 26250
    },
    {
      "epoch": 0.66,
      "grad_norm": 13.363433837890625,
      "learning_rate": 2.5355932203389833e-06,
      "loss": 1.4593,
      "step": 26260
    },
    {
      "epoch": 0.66,
      "grad_norm": 4.77979040145874,
      "learning_rate": 2.5288135593220343e-06,
      "loss": 1.2445,
      "step": 26270
    },
    {
      "epoch": 0.66,
      "grad_norm": 10.652926445007324,
      "learning_rate": 2.5220338983050853e-06,
      "loss": 1.3092,
      "step": 26280
    },
    {
      "epoch": 0.66,
      "grad_norm": 5.278314113616943,
      "learning_rate": 2.5152542372881354e-06,
      "loss": 1.447,
      "step": 26290
    },
    {
      "epoch": 0.66,
      "grad_norm": 6.439229488372803,
      "learning_rate": 2.5084745762711864e-06,
      "loss": 1.3421,
      "step": 26300
    },
    {
      "epoch": 0.66,
      "grad_norm": 4.738833904266357,
      "learning_rate": 2.5016949152542374e-06,
      "loss": 1.3454,
      "step": 26310
    },
    {
      "epoch": 0.66,
      "grad_norm": 4.070488929748535,
      "learning_rate": 2.4949152542372883e-06,
      "loss": 1.377,
      "step": 26320
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.9005275964736938,
      "learning_rate": 2.488135593220339e-06,
      "loss": 1.4501,
      "step": 26330
    },
    {
      "epoch": 0.66,
      "grad_norm": 9.970990180969238,
      "learning_rate": 2.48135593220339e-06,
      "loss": 1.1534,
      "step": 26340
    },
    {
      "epoch": 0.66,
      "grad_norm": 2.7662065029144287,
      "learning_rate": 2.474576271186441e-06,
      "loss": 1.4422,
      "step": 26350
    },
    {
      "epoch": 0.66,
      "grad_norm": 11.093968391418457,
      "learning_rate": 2.467796610169492e-06,
      "loss": 1.198,
      "step": 26360
    },
    {
      "epoch": 0.66,
      "grad_norm": 6.7317280769348145,
      "learning_rate": 2.461016949152543e-06,
      "loss": 1.244,
      "step": 26370
    },
    {
      "epoch": 0.66,
      "grad_norm": 8.76866340637207,
      "learning_rate": 2.4542372881355933e-06,
      "loss": 1.3653,
      "step": 26380
    },
    {
      "epoch": 0.66,
      "grad_norm": 11.940791130065918,
      "learning_rate": 2.4474576271186443e-06,
      "loss": 1.304,
      "step": 26390
    },
    {
      "epoch": 0.66,
      "grad_norm": 6.687407970428467,
      "learning_rate": 2.4406779661016953e-06,
      "loss": 1.1965,
      "step": 26400
    },
    {
      "epoch": 0.66,
      "grad_norm": 5.42927885055542,
      "learning_rate": 2.433898305084746e-06,
      "loss": 1.3501,
      "step": 26410
    },
    {
      "epoch": 0.66,
      "grad_norm": 2.731924057006836,
      "learning_rate": 2.427118644067797e-06,
      "loss": 1.4572,
      "step": 26420
    },
    {
      "epoch": 0.66,
      "grad_norm": 5.305939197540283,
      "learning_rate": 2.4203389830508474e-06,
      "loss": 1.3428,
      "step": 26430
    },
    {
      "epoch": 0.66,
      "grad_norm": 10.32532787322998,
      "learning_rate": 2.4135593220338984e-06,
      "loss": 1.4827,
      "step": 26440
    },
    {
      "epoch": 0.66,
      "grad_norm": 7.55979585647583,
      "learning_rate": 2.4067796610169493e-06,
      "loss": 1.2877,
      "step": 26450
    },
    {
      "epoch": 0.66,
      "grad_norm": 9.092228889465332,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 1.3516,
      "step": 26460
    },
    {
      "epoch": 0.66,
      "grad_norm": 4.732894420623779,
      "learning_rate": 2.393220338983051e-06,
      "loss": 1.3218,
      "step": 26470
    },
    {
      "epoch": 0.66,
      "grad_norm": 8.649917602539062,
      "learning_rate": 2.386440677966102e-06,
      "loss": 1.2708,
      "step": 26480
    },
    {
      "epoch": 0.66,
      "grad_norm": 4.569608211517334,
      "learning_rate": 2.379661016949153e-06,
      "loss": 1.3422,
      "step": 26490
    },
    {
      "epoch": 0.66,
      "grad_norm": 3.702059030532837,
      "learning_rate": 2.372881355932204e-06,
      "loss": 1.4754,
      "step": 26500
    },
    {
      "epoch": 0.66,
      "grad_norm": 3.9777114391326904,
      "learning_rate": 2.3661016949152544e-06,
      "loss": 1.3133,
      "step": 26510
    },
    {
      "epoch": 0.66,
      "grad_norm": 9.692605018615723,
      "learning_rate": 2.3593220338983053e-06,
      "loss": 1.2866,
      "step": 26520
    },
    {
      "epoch": 0.66,
      "grad_norm": 2.1870622634887695,
      "learning_rate": 2.3525423728813563e-06,
      "loss": 1.4271,
      "step": 26530
    },
    {
      "epoch": 0.66,
      "grad_norm": 6.799996852874756,
      "learning_rate": 2.345762711864407e-06,
      "loss": 1.3768,
      "step": 26540
    },
    {
      "epoch": 0.66,
      "grad_norm": 2.3258345127105713,
      "learning_rate": 2.338983050847458e-06,
      "loss": 1.4023,
      "step": 26550
    },
    {
      "epoch": 0.66,
      "grad_norm": 3.950892925262451,
      "learning_rate": 2.3322033898305084e-06,
      "loss": 1.2986,
      "step": 26560
    },
    {
      "epoch": 0.66,
      "grad_norm": 11.114343643188477,
      "learning_rate": 2.3254237288135594e-06,
      "loss": 1.3538,
      "step": 26570
    },
    {
      "epoch": 0.66,
      "grad_norm": 8.47208023071289,
      "learning_rate": 2.3186440677966103e-06,
      "loss": 1.3557,
      "step": 26580
    },
    {
      "epoch": 0.66,
      "grad_norm": 15.794944763183594,
      "learning_rate": 2.3118644067796613e-06,
      "loss": 1.3295,
      "step": 26590
    },
    {
      "epoch": 0.67,
      "grad_norm": 5.596043586730957,
      "learning_rate": 2.305084745762712e-06,
      "loss": 1.2669,
      "step": 26600
    },
    {
      "epoch": 0.67,
      "grad_norm": 8.727288246154785,
      "learning_rate": 2.298305084745763e-06,
      "loss": 1.339,
      "step": 26610
    },
    {
      "epoch": 0.67,
      "grad_norm": 10.842510223388672,
      "learning_rate": 2.291525423728814e-06,
      "loss": 1.2154,
      "step": 26620
    },
    {
      "epoch": 0.67,
      "grad_norm": 11.826702117919922,
      "learning_rate": 2.284745762711865e-06,
      "loss": 1.1599,
      "step": 26630
    },
    {
      "epoch": 0.67,
      "grad_norm": 16.47806167602539,
      "learning_rate": 2.2779661016949154e-06,
      "loss": 1.3005,
      "step": 26640
    },
    {
      "epoch": 0.67,
      "grad_norm": 4.797351837158203,
      "learning_rate": 2.2711864406779663e-06,
      "loss": 1.2779,
      "step": 26650
    },
    {
      "epoch": 0.67,
      "grad_norm": 6.353465557098389,
      "learning_rate": 2.2644067796610173e-06,
      "loss": 1.2419,
      "step": 26660
    },
    {
      "epoch": 0.67,
      "grad_norm": 10.895663261413574,
      "learning_rate": 2.257627118644068e-06,
      "loss": 1.2078,
      "step": 26670
    },
    {
      "epoch": 0.67,
      "grad_norm": 7.583923816680908,
      "learning_rate": 2.250847457627119e-06,
      "loss": 1.3381,
      "step": 26680
    },
    {
      "epoch": 0.67,
      "grad_norm": 10.84875774383545,
      "learning_rate": 2.2440677966101694e-06,
      "loss": 1.4887,
      "step": 26690
    },
    {
      "epoch": 0.67,
      "grad_norm": 5.171149253845215,
      "learning_rate": 2.2372881355932204e-06,
      "loss": 1.3469,
      "step": 26700
    },
    {
      "epoch": 0.67,
      "grad_norm": 6.136636734008789,
      "learning_rate": 2.2305084745762714e-06,
      "loss": 1.1804,
      "step": 26710
    },
    {
      "epoch": 0.67,
      "grad_norm": 12.95764446258545,
      "learning_rate": 2.2237288135593223e-06,
      "loss": 1.2808,
      "step": 26720
    },
    {
      "epoch": 0.67,
      "grad_norm": 4.0281453132629395,
      "learning_rate": 2.216949152542373e-06,
      "loss": 1.4454,
      "step": 26730
    },
    {
      "epoch": 0.67,
      "grad_norm": 5.7566609382629395,
      "learning_rate": 2.210169491525424e-06,
      "loss": 1.2968,
      "step": 26740
    },
    {
      "epoch": 0.67,
      "grad_norm": 4.710749626159668,
      "learning_rate": 2.203389830508475e-06,
      "loss": 1.355,
      "step": 26750
    },
    {
      "epoch": 0.67,
      "grad_norm": 3.0553205013275146,
      "learning_rate": 2.196610169491526e-06,
      "loss": 1.3319,
      "step": 26760
    },
    {
      "epoch": 0.67,
      "grad_norm": 15.849903106689453,
      "learning_rate": 2.1898305084745764e-06,
      "loss": 1.3233,
      "step": 26770
    },
    {
      "epoch": 0.67,
      "grad_norm": 31.49736785888672,
      "learning_rate": 2.1830508474576273e-06,
      "loss": 1.3991,
      "step": 26780
    },
    {
      "epoch": 0.67,
      "grad_norm": 11.734864234924316,
      "learning_rate": 2.1762711864406783e-06,
      "loss": 1.5305,
      "step": 26790
    },
    {
      "epoch": 0.67,
      "grad_norm": 6.124046325683594,
      "learning_rate": 2.169491525423729e-06,
      "loss": 1.249,
      "step": 26800
    },
    {
      "epoch": 0.67,
      "grad_norm": 11.438417434692383,
      "learning_rate": 2.16271186440678e-06,
      "loss": 1.2922,
      "step": 26810
    },
    {
      "epoch": 0.67,
      "grad_norm": 12.979373931884766,
      "learning_rate": 2.1559322033898304e-06,
      "loss": 1.3554,
      "step": 26820
    },
    {
      "epoch": 0.67,
      "grad_norm": 3.8955001831054688,
      "learning_rate": 2.1491525423728814e-06,
      "loss": 1.3303,
      "step": 26830
    },
    {
      "epoch": 0.67,
      "grad_norm": 9.349483489990234,
      "learning_rate": 2.1423728813559324e-06,
      "loss": 1.2548,
      "step": 26840
    },
    {
      "epoch": 0.67,
      "grad_norm": 2.8842084407806396,
      "learning_rate": 2.1355932203389833e-06,
      "loss": 1.3517,
      "step": 26850
    },
    {
      "epoch": 0.67,
      "grad_norm": 3.986353635787964,
      "learning_rate": 2.128813559322034e-06,
      "loss": 1.5566,
      "step": 26860
    },
    {
      "epoch": 0.67,
      "grad_norm": 14.33786392211914,
      "learning_rate": 2.122033898305085e-06,
      "loss": 1.2168,
      "step": 26870
    },
    {
      "epoch": 0.67,
      "grad_norm": 4.677867889404297,
      "learning_rate": 2.115254237288136e-06,
      "loss": 1.2758,
      "step": 26880
    },
    {
      "epoch": 0.67,
      "grad_norm": 3.638185977935791,
      "learning_rate": 2.108474576271187e-06,
      "loss": 1.2348,
      "step": 26890
    },
    {
      "epoch": 0.67,
      "grad_norm": 2.7823917865753174,
      "learning_rate": 2.1016949152542374e-06,
      "loss": 1.4506,
      "step": 26900
    },
    {
      "epoch": 0.67,
      "grad_norm": 14.349405288696289,
      "learning_rate": 2.0949152542372883e-06,
      "loss": 1.2189,
      "step": 26910
    },
    {
      "epoch": 0.67,
      "grad_norm": 5.958116054534912,
      "learning_rate": 2.0881355932203393e-06,
      "loss": 1.0776,
      "step": 26920
    },
    {
      "epoch": 0.67,
      "grad_norm": 5.689637184143066,
      "learning_rate": 2.08135593220339e-06,
      "loss": 1.5885,
      "step": 26930
    },
    {
      "epoch": 0.67,
      "grad_norm": 17.451379776000977,
      "learning_rate": 2.074576271186441e-06,
      "loss": 1.4116,
      "step": 26940
    },
    {
      "epoch": 0.67,
      "grad_norm": 6.859378814697266,
      "learning_rate": 2.0677966101694914e-06,
      "loss": 1.1968,
      "step": 26950
    },
    {
      "epoch": 0.67,
      "grad_norm": 5.8354082107543945,
      "learning_rate": 2.0610169491525424e-06,
      "loss": 1.3615,
      "step": 26960
    },
    {
      "epoch": 0.67,
      "grad_norm": 4.631415367126465,
      "learning_rate": 2.0542372881355934e-06,
      "loss": 1.3517,
      "step": 26970
    },
    {
      "epoch": 0.67,
      "grad_norm": 9.62684154510498,
      "learning_rate": 2.0474576271186443e-06,
      "loss": 1.3364,
      "step": 26980
    },
    {
      "epoch": 0.67,
      "grad_norm": 4.8851447105407715,
      "learning_rate": 2.0406779661016953e-06,
      "loss": 1.2893,
      "step": 26990
    },
    {
      "epoch": 0.68,
      "grad_norm": 5.123071670532227,
      "learning_rate": 2.033898305084746e-06,
      "loss": 1.4069,
      "step": 27000
    },
    {
      "epoch": 0.68,
      "grad_norm": 3.8006324768066406,
      "learning_rate": 2.027118644067797e-06,
      "loss": 1.3013,
      "step": 27010
    },
    {
      "epoch": 0.68,
      "grad_norm": 3.052011728286743,
      "learning_rate": 2.020338983050848e-06,
      "loss": 1.3073,
      "step": 27020
    },
    {
      "epoch": 0.68,
      "grad_norm": 6.314701080322266,
      "learning_rate": 2.0135593220338984e-06,
      "loss": 1.3644,
      "step": 27030
    },
    {
      "epoch": 0.68,
      "grad_norm": 2.868659257888794,
      "learning_rate": 2.0067796610169494e-06,
      "loss": 1.4159,
      "step": 27040
    },
    {
      "epoch": 0.68,
      "grad_norm": 3.1452548503875732,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 1.2184,
      "step": 27050
    },
    {
      "epoch": 0.68,
      "grad_norm": 7.137606620788574,
      "learning_rate": 1.993220338983051e-06,
      "loss": 1.1827,
      "step": 27060
    },
    {
      "epoch": 0.68,
      "grad_norm": 4.114950180053711,
      "learning_rate": 1.986440677966102e-06,
      "loss": 1.3316,
      "step": 27070
    },
    {
      "epoch": 0.68,
      "grad_norm": 4.815858364105225,
      "learning_rate": 1.9796610169491524e-06,
      "loss": 1.286,
      "step": 27080
    },
    {
      "epoch": 0.68,
      "grad_norm": 2.6551191806793213,
      "learning_rate": 1.9728813559322034e-06,
      "loss": 1.2562,
      "step": 27090
    },
    {
      "epoch": 0.68,
      "grad_norm": 10.009329795837402,
      "learning_rate": 1.9661016949152544e-06,
      "loss": 1.3984,
      "step": 27100
    },
    {
      "epoch": 0.68,
      "grad_norm": 2.3229589462280273,
      "learning_rate": 1.9593220338983053e-06,
      "loss": 1.3478,
      "step": 27110
    },
    {
      "epoch": 0.68,
      "grad_norm": 5.636902332305908,
      "learning_rate": 1.9525423728813563e-06,
      "loss": 1.3959,
      "step": 27120
    },
    {
      "epoch": 0.68,
      "grad_norm": 15.958221435546875,
      "learning_rate": 1.945762711864407e-06,
      "loss": 1.4275,
      "step": 27130
    },
    {
      "epoch": 0.68,
      "grad_norm": 5.097884654998779,
      "learning_rate": 1.938983050847458e-06,
      "loss": 1.3826,
      "step": 27140
    },
    {
      "epoch": 0.68,
      "grad_norm": 6.786471366882324,
      "learning_rate": 1.932203389830509e-06,
      "loss": 1.3073,
      "step": 27150
    },
    {
      "epoch": 0.68,
      "grad_norm": 7.529135704040527,
      "learning_rate": 1.9254237288135594e-06,
      "loss": 1.2605,
      "step": 27160
    },
    {
      "epoch": 0.68,
      "grad_norm": 10.065816879272461,
      "learning_rate": 1.9186440677966104e-06,
      "loss": 1.2377,
      "step": 27170
    },
    {
      "epoch": 0.68,
      "grad_norm": 5.812075614929199,
      "learning_rate": 1.9118644067796613e-06,
      "loss": 1.3309,
      "step": 27180
    },
    {
      "epoch": 0.68,
      "grad_norm": 4.214847564697266,
      "learning_rate": 1.9050847457627119e-06,
      "loss": 1.4063,
      "step": 27190
    },
    {
      "epoch": 0.68,
      "grad_norm": 4.021416187286377,
      "learning_rate": 1.8983050847457629e-06,
      "loss": 1.2773,
      "step": 27200
    },
    {
      "epoch": 0.68,
      "grad_norm": 7.9816083908081055,
      "learning_rate": 1.8915254237288136e-06,
      "loss": 1.2151,
      "step": 27210
    },
    {
      "epoch": 0.68,
      "grad_norm": 9.82458209991455,
      "learning_rate": 1.8847457627118646e-06,
      "loss": 1.393,
      "step": 27220
    },
    {
      "epoch": 0.68,
      "grad_norm": 16.87822914123535,
      "learning_rate": 1.8779661016949156e-06,
      "loss": 1.269,
      "step": 27230
    },
    {
      "epoch": 0.68,
      "grad_norm": 6.815838813781738,
      "learning_rate": 1.8711864406779661e-06,
      "loss": 1.4237,
      "step": 27240
    },
    {
      "epoch": 0.68,
      "grad_norm": 3.9835774898529053,
      "learning_rate": 1.8644067796610171e-06,
      "loss": 1.2872,
      "step": 27250
    },
    {
      "epoch": 0.68,
      "grad_norm": 4.0309953689575195,
      "learning_rate": 1.857627118644068e-06,
      "loss": 1.2629,
      "step": 27260
    },
    {
      "epoch": 0.68,
      "grad_norm": 12.035406112670898,
      "learning_rate": 1.8508474576271189e-06,
      "loss": 1.3755,
      "step": 27270
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.8347936868667603,
      "learning_rate": 1.8440677966101696e-06,
      "loss": 1.3831,
      "step": 27280
    },
    {
      "epoch": 0.68,
      "grad_norm": 8.658760070800781,
      "learning_rate": 1.8372881355932204e-06,
      "loss": 1.3428,
      "step": 27290
    },
    {
      "epoch": 0.68,
      "grad_norm": 3.145319938659668,
      "learning_rate": 1.8305084745762714e-06,
      "loss": 1.4272,
      "step": 27300
    },
    {
      "epoch": 0.68,
      "grad_norm": 9.265095710754395,
      "learning_rate": 1.8237288135593223e-06,
      "loss": 1.4133,
      "step": 27310
    },
    {
      "epoch": 0.68,
      "grad_norm": 7.101969242095947,
      "learning_rate": 1.816949152542373e-06,
      "loss": 1.1041,
      "step": 27320
    },
    {
      "epoch": 0.68,
      "grad_norm": 16.614511489868164,
      "learning_rate": 1.8101694915254239e-06,
      "loss": 1.3473,
      "step": 27330
    },
    {
      "epoch": 0.68,
      "grad_norm": 2.301051378250122,
      "learning_rate": 1.8033898305084746e-06,
      "loss": 1.2432,
      "step": 27340
    },
    {
      "epoch": 0.68,
      "grad_norm": 5.643409729003906,
      "learning_rate": 1.7966101694915256e-06,
      "loss": 1.2781,
      "step": 27350
    },
    {
      "epoch": 0.68,
      "grad_norm": 7.286752223968506,
      "learning_rate": 1.7898305084745766e-06,
      "loss": 1.328,
      "step": 27360
    },
    {
      "epoch": 0.68,
      "grad_norm": 3.3953471183776855,
      "learning_rate": 1.7830508474576271e-06,
      "loss": 1.3025,
      "step": 27370
    },
    {
      "epoch": 0.68,
      "grad_norm": 8.240042686462402,
      "learning_rate": 1.7762711864406781e-06,
      "loss": 1.3847,
      "step": 27380
    },
    {
      "epoch": 0.68,
      "grad_norm": 9.705995559692383,
      "learning_rate": 1.769491525423729e-06,
      "loss": 1.427,
      "step": 27390
    },
    {
      "epoch": 0.69,
      "grad_norm": 3.306814193725586,
      "learning_rate": 1.7627118644067799e-06,
      "loss": 1.4487,
      "step": 27400
    },
    {
      "epoch": 0.69,
      "grad_norm": 15.25204086303711,
      "learning_rate": 1.7559322033898306e-06,
      "loss": 1.1448,
      "step": 27410
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.065521240234375,
      "learning_rate": 1.7491525423728814e-06,
      "loss": 1.4139,
      "step": 27420
    },
    {
      "epoch": 0.69,
      "grad_norm": 7.884547233581543,
      "learning_rate": 1.7423728813559324e-06,
      "loss": 1.3313,
      "step": 27430
    },
    {
      "epoch": 0.69,
      "grad_norm": 8.109780311584473,
      "learning_rate": 1.7355932203389834e-06,
      "loss": 1.2523,
      "step": 27440
    },
    {
      "epoch": 0.69,
      "grad_norm": 23.829362869262695,
      "learning_rate": 1.728813559322034e-06,
      "loss": 1.2634,
      "step": 27450
    },
    {
      "epoch": 0.69,
      "grad_norm": 8.787532806396484,
      "learning_rate": 1.7220338983050849e-06,
      "loss": 1.3913,
      "step": 27460
    },
    {
      "epoch": 0.69,
      "grad_norm": 17.018415451049805,
      "learning_rate": 1.7152542372881356e-06,
      "loss": 1.2223,
      "step": 27470
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.56651782989502,
      "learning_rate": 1.7084745762711866e-06,
      "loss": 1.3977,
      "step": 27480
    },
    {
      "epoch": 0.69,
      "grad_norm": 4.520813941955566,
      "learning_rate": 1.7016949152542376e-06,
      "loss": 1.3356,
      "step": 27490
    },
    {
      "epoch": 0.69,
      "grad_norm": 5.794963359832764,
      "learning_rate": 1.6949152542372882e-06,
      "loss": 1.2999,
      "step": 27500
    },
    {
      "epoch": 0.69,
      "eval_loss": 1.2965120077133179,
      "eval_runtime": 122.3839,
      "eval_samples_per_second": 8.171,
      "eval_steps_per_second": 8.171,
      "step": 27500
    },
    {
      "epoch": 0.69,
      "grad_norm": 4.339414119720459,
      "learning_rate": 1.6881355932203391e-06,
      "loss": 1.3429,
      "step": 27510
    },
    {
      "epoch": 0.69,
      "grad_norm": 2.123643159866333,
      "learning_rate": 1.6813559322033901e-06,
      "loss": 1.2949,
      "step": 27520
    },
    {
      "epoch": 0.69,
      "grad_norm": 6.727291584014893,
      "learning_rate": 1.6745762711864409e-06,
      "loss": 1.4269,
      "step": 27530
    },
    {
      "epoch": 0.69,
      "grad_norm": 5.464277744293213,
      "learning_rate": 1.6677966101694916e-06,
      "loss": 1.3967,
      "step": 27540
    },
    {
      "epoch": 0.69,
      "grad_norm": 3.3927230834960938,
      "learning_rate": 1.6610169491525424e-06,
      "loss": 1.2701,
      "step": 27550
    },
    {
      "epoch": 0.69,
      "grad_norm": 6.092131614685059,
      "learning_rate": 1.6542372881355934e-06,
      "loss": 1.1989,
      "step": 27560
    },
    {
      "epoch": 0.69,
      "grad_norm": 4.336033821105957,
      "learning_rate": 1.6474576271186444e-06,
      "loss": 1.2679,
      "step": 27570
    },
    {
      "epoch": 0.69,
      "grad_norm": 3.4040894508361816,
      "learning_rate": 1.640677966101695e-06,
      "loss": 1.4791,
      "step": 27580
    },
    {
      "epoch": 0.69,
      "grad_norm": 5.1817626953125,
      "learning_rate": 1.6338983050847459e-06,
      "loss": 1.3513,
      "step": 27590
    },
    {
      "epoch": 0.69,
      "grad_norm": 14.48134708404541,
      "learning_rate": 1.6271186440677967e-06,
      "loss": 1.3058,
      "step": 27600
    },
    {
      "epoch": 0.69,
      "grad_norm": 6.538887977600098,
      "learning_rate": 1.6203389830508476e-06,
      "loss": 1.3431,
      "step": 27610
    },
    {
      "epoch": 0.69,
      "grad_norm": 7.871098518371582,
      "learning_rate": 1.6135593220338986e-06,
      "loss": 1.3213,
      "step": 27620
    },
    {
      "epoch": 0.69,
      "grad_norm": 6.496534824371338,
      "learning_rate": 1.6067796610169492e-06,
      "loss": 1.4389,
      "step": 27630
    },
    {
      "epoch": 0.69,
      "grad_norm": 6.736118793487549,
      "learning_rate": 1.6000000000000001e-06,
      "loss": 1.3323,
      "step": 27640
    },
    {
      "epoch": 0.69,
      "grad_norm": 13.042952537536621,
      "learning_rate": 1.593220338983051e-06,
      "loss": 1.1813,
      "step": 27650
    },
    {
      "epoch": 0.69,
      "grad_norm": 2.258842706680298,
      "learning_rate": 1.5864406779661019e-06,
      "loss": 1.3224,
      "step": 27660
    },
    {
      "epoch": 0.69,
      "grad_norm": 6.339553356170654,
      "learning_rate": 1.5796610169491526e-06,
      "loss": 1.3433,
      "step": 27670
    },
    {
      "epoch": 0.69,
      "grad_norm": 7.741724491119385,
      "learning_rate": 1.5728813559322034e-06,
      "loss": 1.3168,
      "step": 27680
    },
    {
      "epoch": 0.69,
      "grad_norm": 16.320072174072266,
      "learning_rate": 1.5661016949152544e-06,
      "loss": 1.2347,
      "step": 27690
    },
    {
      "epoch": 0.69,
      "grad_norm": 5.457180023193359,
      "learning_rate": 1.5593220338983054e-06,
      "loss": 1.2835,
      "step": 27700
    },
    {
      "epoch": 0.69,
      "grad_norm": 5.750472068786621,
      "learning_rate": 1.552542372881356e-06,
      "loss": 1.2346,
      "step": 27710
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.067400574684143,
      "learning_rate": 1.545762711864407e-06,
      "loss": 1.26,
      "step": 27720
    },
    {
      "epoch": 0.69,
      "grad_norm": 14.453791618347168,
      "learning_rate": 1.5389830508474577e-06,
      "loss": 1.1674,
      "step": 27730
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.273608207702637,
      "learning_rate": 1.5322033898305086e-06,
      "loss": 1.2715,
      "step": 27740
    },
    {
      "epoch": 0.69,
      "grad_norm": 2.4199931621551514,
      "learning_rate": 1.5254237288135596e-06,
      "loss": 1.3738,
      "step": 27750
    },
    {
      "epoch": 0.69,
      "grad_norm": 11.302939414978027,
      "learning_rate": 1.5186440677966102e-06,
      "loss": 1.2809,
      "step": 27760
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.895012855529785,
      "learning_rate": 1.5118644067796611e-06,
      "loss": 1.2577,
      "step": 27770
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.7351655960083,
      "learning_rate": 1.505084745762712e-06,
      "loss": 1.2743,
      "step": 27780
    },
    {
      "epoch": 0.69,
      "grad_norm": 3.2647464275360107,
      "learning_rate": 1.4983050847457629e-06,
      "loss": 1.32,
      "step": 27790
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.726398468017578,
      "learning_rate": 1.4915254237288139e-06,
      "loss": 1.3836,
      "step": 27800
    },
    {
      "epoch": 0.7,
      "grad_norm": 7.257278919219971,
      "learning_rate": 1.4847457627118644e-06,
      "loss": 1.4125,
      "step": 27810
    },
    {
      "epoch": 0.7,
      "grad_norm": 10.737898826599121,
      "learning_rate": 1.4779661016949154e-06,
      "loss": 1.3774,
      "step": 27820
    },
    {
      "epoch": 0.7,
      "grad_norm": 10.912455558776855,
      "learning_rate": 1.4711864406779664e-06,
      "loss": 1.302,
      "step": 27830
    },
    {
      "epoch": 0.7,
      "grad_norm": 4.399783611297607,
      "learning_rate": 1.464406779661017e-06,
      "loss": 1.2622,
      "step": 27840
    },
    {
      "epoch": 0.7,
      "grad_norm": 14.4951171875,
      "learning_rate": 1.457627118644068e-06,
      "loss": 1.415,
      "step": 27850
    },
    {
      "epoch": 0.7,
      "grad_norm": 5.326625347137451,
      "learning_rate": 1.4508474576271187e-06,
      "loss": 1.2882,
      "step": 27860
    },
    {
      "epoch": 0.7,
      "grad_norm": 6.698885917663574,
      "learning_rate": 1.4440677966101696e-06,
      "loss": 1.3461,
      "step": 27870
    },
    {
      "epoch": 0.7,
      "grad_norm": 2.7243850231170654,
      "learning_rate": 1.4372881355932206e-06,
      "loss": 1.4233,
      "step": 27880
    },
    {
      "epoch": 0.7,
      "grad_norm": 2.273730516433716,
      "learning_rate": 1.4305084745762712e-06,
      "loss": 1.2245,
      "step": 27890
    },
    {
      "epoch": 0.7,
      "grad_norm": 3.1223866939544678,
      "learning_rate": 1.4237288135593222e-06,
      "loss": 1.3181,
      "step": 27900
    },
    {
      "epoch": 0.7,
      "grad_norm": 8.419058799743652,
      "learning_rate": 1.416949152542373e-06,
      "loss": 1.2782,
      "step": 27910
    },
    {
      "epoch": 0.7,
      "grad_norm": 4.810690879821777,
      "learning_rate": 1.410169491525424e-06,
      "loss": 1.2527,
      "step": 27920
    },
    {
      "epoch": 0.7,
      "grad_norm": 2.7679851055145264,
      "learning_rate": 1.4033898305084749e-06,
      "loss": 1.1995,
      "step": 27930
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.7833703756332397,
      "learning_rate": 1.3966101694915254e-06,
      "loss": 1.3988,
      "step": 27940
    },
    {
      "epoch": 0.7,
      "grad_norm": 8.348339080810547,
      "learning_rate": 1.3898305084745764e-06,
      "loss": 1.2983,
      "step": 27950
    },
    {
      "epoch": 0.7,
      "grad_norm": 8.304106712341309,
      "learning_rate": 1.3830508474576274e-06,
      "loss": 1.2926,
      "step": 27960
    },
    {
      "epoch": 0.7,
      "grad_norm": 5.610420227050781,
      "learning_rate": 1.376271186440678e-06,
      "loss": 1.2064,
      "step": 27970
    },
    {
      "epoch": 0.7,
      "grad_norm": 13.0822114944458,
      "learning_rate": 1.369491525423729e-06,
      "loss": 1.3171,
      "step": 27980
    },
    {
      "epoch": 0.7,
      "grad_norm": 3.2502501010894775,
      "learning_rate": 1.3627118644067797e-06,
      "loss": 1.2354,
      "step": 27990
    },
    {
      "epoch": 0.7,
      "grad_norm": 19.23220443725586,
      "learning_rate": 1.3559322033898307e-06,
      "loss": 1.3304,
      "step": 28000
    },
    {
      "epoch": 0.7,
      "grad_norm": 5.642111778259277,
      "learning_rate": 1.3491525423728816e-06,
      "loss": 1.2085,
      "step": 28010
    },
    {
      "epoch": 0.7,
      "grad_norm": 5.464583396911621,
      "learning_rate": 1.3423728813559322e-06,
      "loss": 1.3588,
      "step": 28020
    },
    {
      "epoch": 0.7,
      "grad_norm": 2.51157808303833,
      "learning_rate": 1.3355932203389832e-06,
      "loss": 1.1803,
      "step": 28030
    },
    {
      "epoch": 0.7,
      "grad_norm": 10.629277229309082,
      "learning_rate": 1.328813559322034e-06,
      "loss": 1.3943,
      "step": 28040
    },
    {
      "epoch": 0.7,
      "grad_norm": 10.221532821655273,
      "learning_rate": 1.322033898305085e-06,
      "loss": 1.4424,
      "step": 28050
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.1695493459701538,
      "learning_rate": 1.3152542372881359e-06,
      "loss": 1.3071,
      "step": 28060
    },
    {
      "epoch": 0.7,
      "grad_norm": 4.1398539543151855,
      "learning_rate": 1.3084745762711864e-06,
      "loss": 1.2239,
      "step": 28070
    },
    {
      "epoch": 0.7,
      "grad_norm": 20.721601486206055,
      "learning_rate": 1.3016949152542374e-06,
      "loss": 1.3761,
      "step": 28080
    },
    {
      "epoch": 0.7,
      "grad_norm": 8.006892204284668,
      "learning_rate": 1.2949152542372884e-06,
      "loss": 1.2867,
      "step": 28090
    },
    {
      "epoch": 0.7,
      "grad_norm": 4.007428169250488,
      "learning_rate": 1.288135593220339e-06,
      "loss": 1.2511,
      "step": 28100
    },
    {
      "epoch": 0.7,
      "grad_norm": 8.041426658630371,
      "learning_rate": 1.28135593220339e-06,
      "loss": 1.3597,
      "step": 28110
    },
    {
      "epoch": 0.7,
      "grad_norm": 4.553910732269287,
      "learning_rate": 1.2745762711864407e-06,
      "loss": 1.4066,
      "step": 28120
    },
    {
      "epoch": 0.7,
      "grad_norm": 8.02763843536377,
      "learning_rate": 1.2677966101694917e-06,
      "loss": 1.1639,
      "step": 28130
    },
    {
      "epoch": 0.7,
      "grad_norm": 9.010587692260742,
      "learning_rate": 1.2610169491525426e-06,
      "loss": 1.5038,
      "step": 28140
    },
    {
      "epoch": 0.7,
      "grad_norm": 4.593667984008789,
      "learning_rate": 1.2542372881355932e-06,
      "loss": 1.4871,
      "step": 28150
    },
    {
      "epoch": 0.7,
      "grad_norm": 6.7055158615112305,
      "learning_rate": 1.2474576271186442e-06,
      "loss": 1.2533,
      "step": 28160
    },
    {
      "epoch": 0.7,
      "grad_norm": 13.491644859313965,
      "learning_rate": 1.240677966101695e-06,
      "loss": 1.2957,
      "step": 28170
    },
    {
      "epoch": 0.7,
      "grad_norm": 5.513734340667725,
      "learning_rate": 1.233898305084746e-06,
      "loss": 1.2898,
      "step": 28180
    },
    {
      "epoch": 0.7,
      "grad_norm": 11.136821746826172,
      "learning_rate": 1.2271186440677967e-06,
      "loss": 1.4257,
      "step": 28190
    },
    {
      "epoch": 0.7,
      "grad_norm": 4.229417324066162,
      "learning_rate": 1.2203389830508477e-06,
      "loss": 1.322,
      "step": 28200
    },
    {
      "epoch": 0.71,
      "grad_norm": 4.852339267730713,
      "learning_rate": 1.2135593220338984e-06,
      "loss": 1.3371,
      "step": 28210
    },
    {
      "epoch": 0.71,
      "grad_norm": 9.924147605895996,
      "learning_rate": 1.2067796610169492e-06,
      "loss": 1.1373,
      "step": 28220
    },
    {
      "epoch": 0.71,
      "grad_norm": 3.660102128982544,
      "learning_rate": 1.2000000000000002e-06,
      "loss": 1.253,
      "step": 28230
    },
    {
      "epoch": 0.71,
      "grad_norm": 5.641490936279297,
      "learning_rate": 1.193220338983051e-06,
      "loss": 1.3156,
      "step": 28240
    },
    {
      "epoch": 0.71,
      "grad_norm": 7.371779441833496,
      "learning_rate": 1.186440677966102e-06,
      "loss": 1.3273,
      "step": 28250
    },
    {
      "epoch": 0.71,
      "grad_norm": 11.060622215270996,
      "learning_rate": 1.1796610169491527e-06,
      "loss": 1.2729,
      "step": 28260
    },
    {
      "epoch": 0.71,
      "grad_norm": 2.1023409366607666,
      "learning_rate": 1.1728813559322034e-06,
      "loss": 1.4814,
      "step": 28270
    },
    {
      "epoch": 0.71,
      "grad_norm": 2.1587941646575928,
      "learning_rate": 1.1661016949152542e-06,
      "loss": 1.4983,
      "step": 28280
    },
    {
      "epoch": 0.71,
      "grad_norm": 10.272698402404785,
      "learning_rate": 1.1593220338983052e-06,
      "loss": 1.2093,
      "step": 28290
    },
    {
      "epoch": 0.71,
      "grad_norm": 3.677959442138672,
      "learning_rate": 1.152542372881356e-06,
      "loss": 1.1745,
      "step": 28300
    },
    {
      "epoch": 0.71,
      "grad_norm": 2.187302350997925,
      "learning_rate": 1.145762711864407e-06,
      "loss": 1.3257,
      "step": 28310
    },
    {
      "epoch": 0.71,
      "grad_norm": 5.065845489501953,
      "learning_rate": 1.1389830508474577e-06,
      "loss": 1.3038,
      "step": 28320
    },
    {
      "epoch": 0.71,
      "grad_norm": 6.725832462310791,
      "learning_rate": 1.1322033898305087e-06,
      "loss": 1.2525,
      "step": 28330
    },
    {
      "epoch": 0.71,
      "grad_norm": 13.034323692321777,
      "learning_rate": 1.1254237288135594e-06,
      "loss": 1.3096,
      "step": 28340
    },
    {
      "epoch": 0.71,
      "grad_norm": 12.125313758850098,
      "learning_rate": 1.1186440677966102e-06,
      "loss": 1.1823,
      "step": 28350
    },
    {
      "epoch": 0.71,
      "grad_norm": 12.07535457611084,
      "learning_rate": 1.1118644067796612e-06,
      "loss": 1.2526,
      "step": 28360
    },
    {
      "epoch": 0.71,
      "grad_norm": 6.179876804351807,
      "learning_rate": 1.105084745762712e-06,
      "loss": 1.3062,
      "step": 28370
    },
    {
      "epoch": 0.71,
      "grad_norm": 2.814276933670044,
      "learning_rate": 1.098305084745763e-06,
      "loss": 1.2911,
      "step": 28380
    },
    {
      "epoch": 0.71,
      "grad_norm": 14.710918426513672,
      "learning_rate": 1.0915254237288137e-06,
      "loss": 1.1594,
      "step": 28390
    },
    {
      "epoch": 0.71,
      "grad_norm": 12.056600570678711,
      "learning_rate": 1.0847457627118644e-06,
      "loss": 1.2203,
      "step": 28400
    },
    {
      "epoch": 0.71,
      "grad_norm": 9.935985565185547,
      "learning_rate": 1.0779661016949152e-06,
      "loss": 1.4101,
      "step": 28410
    },
    {
      "epoch": 0.71,
      "grad_norm": 15.330471992492676,
      "learning_rate": 1.0711864406779662e-06,
      "loss": 1.1429,
      "step": 28420
    },
    {
      "epoch": 0.71,
      "grad_norm": 2.6031837463378906,
      "learning_rate": 1.064406779661017e-06,
      "loss": 1.336,
      "step": 28430
    },
    {
      "epoch": 0.71,
      "grad_norm": 10.921528816223145,
      "learning_rate": 1.057627118644068e-06,
      "loss": 1.3067,
      "step": 28440
    },
    {
      "epoch": 0.71,
      "grad_norm": 10.396923065185547,
      "learning_rate": 1.0508474576271187e-06,
      "loss": 1.3811,
      "step": 28450
    },
    {
      "epoch": 0.71,
      "grad_norm": 10.552454948425293,
      "learning_rate": 1.0440677966101697e-06,
      "loss": 1.3656,
      "step": 28460
    },
    {
      "epoch": 0.71,
      "grad_norm": 11.751737594604492,
      "learning_rate": 1.0372881355932204e-06,
      "loss": 1.2437,
      "step": 28470
    },
    {
      "epoch": 0.71,
      "grad_norm": 3.7091989517211914,
      "learning_rate": 1.0305084745762712e-06,
      "loss": 1.3765,
      "step": 28480
    },
    {
      "epoch": 0.71,
      "grad_norm": 5.946925163269043,
      "learning_rate": 1.0237288135593222e-06,
      "loss": 1.4501,
      "step": 28490
    },
    {
      "epoch": 0.71,
      "grad_norm": 10.043877601623535,
      "learning_rate": 1.016949152542373e-06,
      "loss": 1.3166,
      "step": 28500
    },
    {
      "epoch": 0.71,
      "grad_norm": 12.921855926513672,
      "learning_rate": 1.010169491525424e-06,
      "loss": 1.2181,
      "step": 28510
    },
    {
      "epoch": 0.71,
      "grad_norm": 8.31167984008789,
      "learning_rate": 1.0033898305084747e-06,
      "loss": 1.1729,
      "step": 28520
    },
    {
      "epoch": 0.71,
      "grad_norm": 13.391186714172363,
      "learning_rate": 9.966101694915254e-07,
      "loss": 1.4122,
      "step": 28530
    },
    {
      "epoch": 0.71,
      "grad_norm": 3.428011178970337,
      "learning_rate": 9.898305084745762e-07,
      "loss": 1.4873,
      "step": 28540
    },
    {
      "epoch": 0.71,
      "grad_norm": 5.871237754821777,
      "learning_rate": 9.830508474576272e-07,
      "loss": 1.5205,
      "step": 28550
    },
    {
      "epoch": 0.71,
      "grad_norm": 11.823165893554688,
      "learning_rate": 9.762711864406782e-07,
      "loss": 1.1785,
      "step": 28560
    },
    {
      "epoch": 0.71,
      "grad_norm": 13.24584674835205,
      "learning_rate": 9.69491525423729e-07,
      "loss": 1.3375,
      "step": 28570
    },
    {
      "epoch": 0.71,
      "grad_norm": 4.6865386962890625,
      "learning_rate": 9.627118644067797e-07,
      "loss": 1.3289,
      "step": 28580
    },
    {
      "epoch": 0.71,
      "grad_norm": 4.408345699310303,
      "learning_rate": 9.559322033898307e-07,
      "loss": 1.3748,
      "step": 28590
    },
    {
      "epoch": 0.71,
      "grad_norm": 6.063961982727051,
      "learning_rate": 9.491525423728814e-07,
      "loss": 1.2393,
      "step": 28600
    },
    {
      "epoch": 0.72,
      "grad_norm": 12.433189392089844,
      "learning_rate": 9.423728813559323e-07,
      "loss": 1.295,
      "step": 28610
    },
    {
      "epoch": 0.72,
      "grad_norm": 4.160829067230225,
      "learning_rate": 9.355932203389831e-07,
      "loss": 1.1346,
      "step": 28620
    },
    {
      "epoch": 0.72,
      "grad_norm": 4.698163032531738,
      "learning_rate": 9.28813559322034e-07,
      "loss": 1.2795,
      "step": 28630
    },
    {
      "epoch": 0.72,
      "grad_norm": 9.586724281311035,
      "learning_rate": 9.220338983050848e-07,
      "loss": 1.3546,
      "step": 28640
    },
    {
      "epoch": 0.72,
      "grad_norm": 5.39924955368042,
      "learning_rate": 9.152542372881357e-07,
      "loss": 1.3835,
      "step": 28650
    },
    {
      "epoch": 0.72,
      "grad_norm": 9.498052597045898,
      "learning_rate": 9.084745762711864e-07,
      "loss": 1.3516,
      "step": 28660
    },
    {
      "epoch": 0.72,
      "grad_norm": 7.17230749130249,
      "learning_rate": 9.016949152542373e-07,
      "loss": 1.2309,
      "step": 28670
    },
    {
      "epoch": 0.72,
      "grad_norm": 5.45039176940918,
      "learning_rate": 8.949152542372883e-07,
      "loss": 1.2632,
      "step": 28680
    },
    {
      "epoch": 0.72,
      "grad_norm": 5.3587260246276855,
      "learning_rate": 8.881355932203391e-07,
      "loss": 1.4843,
      "step": 28690
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.8675847053527832,
      "learning_rate": 8.813559322033899e-07,
      "loss": 1.4973,
      "step": 28700
    },
    {
      "epoch": 0.72,
      "grad_norm": 3.713160753250122,
      "learning_rate": 8.745762711864407e-07,
      "loss": 1.309,
      "step": 28710
    },
    {
      "epoch": 0.72,
      "grad_norm": 10.668930053710938,
      "learning_rate": 8.677966101694917e-07,
      "loss": 1.3676,
      "step": 28720
    },
    {
      "epoch": 0.72,
      "grad_norm": 2.1564583778381348,
      "learning_rate": 8.610169491525424e-07,
      "loss": 1.2518,
      "step": 28730
    },
    {
      "epoch": 0.72,
      "grad_norm": 4.37905216217041,
      "learning_rate": 8.542372881355933e-07,
      "loss": 1.3369,
      "step": 28740
    },
    {
      "epoch": 0.72,
      "grad_norm": 15.10600471496582,
      "learning_rate": 8.474576271186441e-07,
      "loss": 1.1867,
      "step": 28750
    },
    {
      "epoch": 0.72,
      "grad_norm": 6.618597030639648,
      "learning_rate": 8.406779661016951e-07,
      "loss": 1.2391,
      "step": 28760
    },
    {
      "epoch": 0.72,
      "grad_norm": 16.292184829711914,
      "learning_rate": 8.338983050847458e-07,
      "loss": 1.4165,
      "step": 28770
    },
    {
      "epoch": 0.72,
      "grad_norm": 8.334046363830566,
      "learning_rate": 8.271186440677967e-07,
      "loss": 1.344,
      "step": 28780
    },
    {
      "epoch": 0.72,
      "grad_norm": 10.543550491333008,
      "learning_rate": 8.203389830508475e-07,
      "loss": 1.3328,
      "step": 28790
    },
    {
      "epoch": 0.72,
      "grad_norm": 2.6791608333587646,
      "learning_rate": 8.135593220338983e-07,
      "loss": 1.4221,
      "step": 28800
    },
    {
      "epoch": 0.72,
      "grad_norm": 3.6508536338806152,
      "learning_rate": 8.067796610169493e-07,
      "loss": 1.3269,
      "step": 28810
    },
    {
      "epoch": 0.72,
      "grad_norm": 7.089112758636475,
      "learning_rate": 8.000000000000001e-07,
      "loss": 1.3038,
      "step": 28820
    },
    {
      "epoch": 0.72,
      "grad_norm": 6.554429531097412,
      "learning_rate": 7.932203389830509e-07,
      "loss": 1.2598,
      "step": 28830
    },
    {
      "epoch": 0.72,
      "grad_norm": 5.9939799308776855,
      "learning_rate": 7.864406779661017e-07,
      "loss": 1.419,
      "step": 28840
    },
    {
      "epoch": 0.72,
      "grad_norm": 10.088254928588867,
      "learning_rate": 7.796610169491527e-07,
      "loss": 1.2858,
      "step": 28850
    },
    {
      "epoch": 0.72,
      "grad_norm": 16.476806640625,
      "learning_rate": 7.728813559322034e-07,
      "loss": 1.3498,
      "step": 28860
    },
    {
      "epoch": 0.72,
      "grad_norm": 11.453723907470703,
      "learning_rate": 7.661016949152543e-07,
      "loss": 1.2346,
      "step": 28870
    },
    {
      "epoch": 0.72,
      "grad_norm": 4.832727909088135,
      "learning_rate": 7.593220338983051e-07,
      "loss": 1.252,
      "step": 28880
    },
    {
      "epoch": 0.72,
      "grad_norm": 8.281596183776855,
      "learning_rate": 7.52542372881356e-07,
      "loss": 1.3162,
      "step": 28890
    },
    {
      "epoch": 0.72,
      "grad_norm": 8.322393417358398,
      "learning_rate": 7.457627118644069e-07,
      "loss": 1.1593,
      "step": 28900
    },
    {
      "epoch": 0.72,
      "grad_norm": 10.499345779418945,
      "learning_rate": 7.389830508474577e-07,
      "loss": 1.243,
      "step": 28910
    },
    {
      "epoch": 0.72,
      "grad_norm": 14.577847480773926,
      "learning_rate": 7.322033898305085e-07,
      "loss": 1.3396,
      "step": 28920
    },
    {
      "epoch": 0.72,
      "grad_norm": 3.6276347637176514,
      "learning_rate": 7.254237288135593e-07,
      "loss": 1.4325,
      "step": 28930
    },
    {
      "epoch": 0.72,
      "grad_norm": 10.75643539428711,
      "learning_rate": 7.186440677966103e-07,
      "loss": 1.3319,
      "step": 28940
    },
    {
      "epoch": 0.72,
      "grad_norm": 8.437240600585938,
      "learning_rate": 7.118644067796611e-07,
      "loss": 1.2917,
      "step": 28950
    },
    {
      "epoch": 0.72,
      "grad_norm": 7.0036139488220215,
      "learning_rate": 7.05084745762712e-07,
      "loss": 1.306,
      "step": 28960
    },
    {
      "epoch": 0.72,
      "grad_norm": 6.939202308654785,
      "learning_rate": 6.983050847457627e-07,
      "loss": 1.1919,
      "step": 28970
    },
    {
      "epoch": 0.72,
      "grad_norm": 5.897158145904541,
      "learning_rate": 6.915254237288137e-07,
      "loss": 1.3183,
      "step": 28980
    },
    {
      "epoch": 0.72,
      "grad_norm": 8.489838600158691,
      "learning_rate": 6.847457627118645e-07,
      "loss": 1.3525,
      "step": 28990
    },
    {
      "epoch": 0.72,
      "grad_norm": 11.85141658782959,
      "learning_rate": 6.779661016949153e-07,
      "loss": 1.3708,
      "step": 29000
    },
    {
      "epoch": 0.73,
      "grad_norm": 12.649805068969727,
      "learning_rate": 6.711864406779661e-07,
      "loss": 1.2521,
      "step": 29010
    },
    {
      "epoch": 0.73,
      "grad_norm": 4.55898380279541,
      "learning_rate": 6.64406779661017e-07,
      "loss": 1.3484,
      "step": 29020
    },
    {
      "epoch": 0.73,
      "grad_norm": 7.323060989379883,
      "learning_rate": 6.576271186440679e-07,
      "loss": 1.4211,
      "step": 29030
    },
    {
      "epoch": 0.73,
      "grad_norm": 4.497279167175293,
      "learning_rate": 6.508474576271187e-07,
      "loss": 1.3893,
      "step": 29040
    },
    {
      "epoch": 0.73,
      "grad_norm": 10.126681327819824,
      "learning_rate": 6.440677966101695e-07,
      "loss": 1.2969,
      "step": 29050
    },
    {
      "epoch": 0.73,
      "grad_norm": 4.058706760406494,
      "learning_rate": 6.372881355932203e-07,
      "loss": 1.297,
      "step": 29060
    },
    {
      "epoch": 0.73,
      "grad_norm": 3.2129762172698975,
      "learning_rate": 6.305084745762713e-07,
      "loss": 1.336,
      "step": 29070
    },
    {
      "epoch": 0.73,
      "grad_norm": 2.8219425678253174,
      "learning_rate": 6.237288135593221e-07,
      "loss": 1.2462,
      "step": 29080
    },
    {
      "epoch": 0.73,
      "grad_norm": 3.939793109893799,
      "learning_rate": 6.16949152542373e-07,
      "loss": 1.417,
      "step": 29090
    },
    {
      "epoch": 0.73,
      "grad_norm": 5.102091312408447,
      "learning_rate": 6.101694915254238e-07,
      "loss": 1.467,
      "step": 29100
    },
    {
      "epoch": 0.73,
      "grad_norm": 24.394020080566406,
      "learning_rate": 6.033898305084746e-07,
      "loss": 1.0779,
      "step": 29110
    },
    {
      "epoch": 0.73,
      "grad_norm": 9.672082901000977,
      "learning_rate": 5.966101694915255e-07,
      "loss": 1.2263,
      "step": 29120
    },
    {
      "epoch": 0.73,
      "grad_norm": 3.5495758056640625,
      "learning_rate": 5.898305084745763e-07,
      "loss": 1.3195,
      "step": 29130
    },
    {
      "epoch": 0.73,
      "grad_norm": 3.160689115524292,
      "learning_rate": 5.830508474576271e-07,
      "loss": 1.4946,
      "step": 29140
    },
    {
      "epoch": 0.73,
      "grad_norm": 7.806612014770508,
      "learning_rate": 5.76271186440678e-07,
      "loss": 1.2732,
      "step": 29150
    },
    {
      "epoch": 0.73,
      "grad_norm": 16.695268630981445,
      "learning_rate": 5.694915254237288e-07,
      "loss": 1.4132,
      "step": 29160
    },
    {
      "epoch": 0.73,
      "grad_norm": 9.606965065002441,
      "learning_rate": 5.627118644067797e-07,
      "loss": 1.2185,
      "step": 29170
    },
    {
      "epoch": 0.73,
      "grad_norm": 3.9440929889678955,
      "learning_rate": 5.559322033898306e-07,
      "loss": 1.3204,
      "step": 29180
    },
    {
      "epoch": 0.73,
      "grad_norm": 11.45638370513916,
      "learning_rate": 5.491525423728815e-07,
      "loss": 1.3867,
      "step": 29190
    },
    {
      "epoch": 0.73,
      "grad_norm": 5.335674285888672,
      "learning_rate": 5.423728813559322e-07,
      "loss": 1.2608,
      "step": 29200
    },
    {
      "epoch": 0.73,
      "grad_norm": 5.219457149505615,
      "learning_rate": 5.355932203389831e-07,
      "loss": 1.3998,
      "step": 29210
    },
    {
      "epoch": 0.73,
      "grad_norm": 8.143280029296875,
      "learning_rate": 5.28813559322034e-07,
      "loss": 1.2506,
      "step": 29220
    },
    {
      "epoch": 0.73,
      "grad_norm": 21.4418888092041,
      "learning_rate": 5.220338983050848e-07,
      "loss": 1.3894,
      "step": 29230
    },
    {
      "epoch": 0.73,
      "grad_norm": 9.489405632019043,
      "learning_rate": 5.152542372881356e-07,
      "loss": 1.3329,
      "step": 29240
    },
    {
      "epoch": 0.73,
      "grad_norm": 4.9300031661987305,
      "learning_rate": 5.084745762711865e-07,
      "loss": 1.2334,
      "step": 29250
    },
    {
      "epoch": 0.73,
      "grad_norm": 4.382762908935547,
      "learning_rate": 5.016949152542373e-07,
      "loss": 1.2921,
      "step": 29260
    },
    {
      "epoch": 0.73,
      "grad_norm": 9.448204040527344,
      "learning_rate": 4.949152542372881e-07,
      "loss": 1.2535,
      "step": 29270
    },
    {
      "epoch": 0.73,
      "grad_norm": 12.2984619140625,
      "learning_rate": 4.881355932203391e-07,
      "loss": 1.1073,
      "step": 29280
    },
    {
      "epoch": 0.73,
      "grad_norm": 8.206221580505371,
      "learning_rate": 4.813559322033898e-07,
      "loss": 1.0739,
      "step": 29290
    },
    {
      "epoch": 0.73,
      "grad_norm": 16.299449920654297,
      "learning_rate": 4.745762711864407e-07,
      "loss": 1.2505,
      "step": 29300
    },
    {
      "epoch": 0.73,
      "grad_norm": 5.297379493713379,
      "learning_rate": 4.6779661016949154e-07,
      "loss": 1.2957,
      "step": 29310
    },
    {
      "epoch": 0.73,
      "grad_norm": 11.336212158203125,
      "learning_rate": 4.610169491525424e-07,
      "loss": 1.2837,
      "step": 29320
    },
    {
      "epoch": 0.73,
      "grad_norm": 8.407357215881348,
      "learning_rate": 4.542372881355932e-07,
      "loss": 1.2719,
      "step": 29330
    },
    {
      "epoch": 0.73,
      "grad_norm": 4.997983932495117,
      "learning_rate": 4.4745762711864415e-07,
      "loss": 1.3578,
      "step": 29340
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.8061422109603882,
      "learning_rate": 4.4067796610169497e-07,
      "loss": 1.2524,
      "step": 29350
    },
    {
      "epoch": 0.73,
      "grad_norm": 7.534559726715088,
      "learning_rate": 4.3389830508474584e-07,
      "loss": 1.2727,
      "step": 29360
    },
    {
      "epoch": 0.73,
      "grad_norm": 2.897261619567871,
      "learning_rate": 4.2711864406779666e-07,
      "loss": 1.35,
      "step": 29370
    },
    {
      "epoch": 0.73,
      "grad_norm": 11.500360488891602,
      "learning_rate": 4.2033898305084753e-07,
      "loss": 1.3056,
      "step": 29380
    },
    {
      "epoch": 0.73,
      "grad_norm": 11.723714828491211,
      "learning_rate": 4.1355932203389835e-07,
      "loss": 1.3156,
      "step": 29390
    },
    {
      "epoch": 0.73,
      "grad_norm": 3.963747978210449,
      "learning_rate": 4.0677966101694916e-07,
      "loss": 1.3509,
      "step": 29400
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.9269698858261108,
      "learning_rate": 4.0000000000000003e-07,
      "loss": 1.3019,
      "step": 29410
    },
    {
      "epoch": 0.74,
      "grad_norm": 6.010022163391113,
      "learning_rate": 3.9322033898305085e-07,
      "loss": 1.2579,
      "step": 29420
    },
    {
      "epoch": 0.74,
      "grad_norm": 11.123760223388672,
      "learning_rate": 3.864406779661017e-07,
      "loss": 1.3221,
      "step": 29430
    },
    {
      "epoch": 0.74,
      "grad_norm": 11.927657127380371,
      "learning_rate": 3.7966101694915254e-07,
      "loss": 1.2219,
      "step": 29440
    },
    {
      "epoch": 0.74,
      "grad_norm": 8.796635627746582,
      "learning_rate": 3.7288135593220347e-07,
      "loss": 1.2437,
      "step": 29450
    },
    {
      "epoch": 0.74,
      "grad_norm": 15.227128028869629,
      "learning_rate": 3.6610169491525423e-07,
      "loss": 1.2653,
      "step": 29460
    },
    {
      "epoch": 0.74,
      "grad_norm": 3.172653913497925,
      "learning_rate": 3.5932203389830516e-07,
      "loss": 1.4635,
      "step": 29470
    },
    {
      "epoch": 0.74,
      "grad_norm": 5.841287612915039,
      "learning_rate": 3.52542372881356e-07,
      "loss": 1.2766,
      "step": 29480
    },
    {
      "epoch": 0.74,
      "grad_norm": 9.996382713317871,
      "learning_rate": 3.4576271186440684e-07,
      "loss": 1.3705,
      "step": 29490
    },
    {
      "epoch": 0.74,
      "grad_norm": 9.059891700744629,
      "learning_rate": 3.3898305084745766e-07,
      "loss": 1.282,
      "step": 29500
    },
    {
      "epoch": 0.74,
      "grad_norm": 6.326866626739502,
      "learning_rate": 3.322033898305085e-07,
      "loss": 1.1478,
      "step": 29510
    },
    {
      "epoch": 0.74,
      "grad_norm": 4.546393394470215,
      "learning_rate": 3.2542372881355935e-07,
      "loss": 1.3345,
      "step": 29520
    },
    {
      "epoch": 0.74,
      "grad_norm": 3.654453992843628,
      "learning_rate": 3.1864406779661017e-07,
      "loss": 1.5465,
      "step": 29530
    },
    {
      "epoch": 0.74,
      "grad_norm": 8.529802322387695,
      "learning_rate": 3.1186440677966104e-07,
      "loss": 1.5145,
      "step": 29540
    },
    {
      "epoch": 0.74,
      "grad_norm": 4.871727466583252,
      "learning_rate": 3.050847457627119e-07,
      "loss": 1.3838,
      "step": 29550
    },
    {
      "epoch": 0.74,
      "grad_norm": 6.172389507293701,
      "learning_rate": 2.9830508474576273e-07,
      "loss": 1.3202,
      "step": 29560
    },
    {
      "epoch": 0.74,
      "grad_norm": 9.824962615966797,
      "learning_rate": 2.9152542372881355e-07,
      "loss": 1.2167,
      "step": 29570
    },
    {
      "epoch": 0.74,
      "grad_norm": 10.725250244140625,
      "learning_rate": 2.847457627118644e-07,
      "loss": 1.2796,
      "step": 29580
    },
    {
      "epoch": 0.74,
      "grad_norm": 14.172257423400879,
      "learning_rate": 2.779661016949153e-07,
      "loss": 1.1175,
      "step": 29590
    },
    {
      "epoch": 0.74,
      "grad_norm": 6.030600547790527,
      "learning_rate": 2.711864406779661e-07,
      "loss": 1.0479,
      "step": 29600
    },
    {
      "epoch": 0.74,
      "grad_norm": 3.860551357269287,
      "learning_rate": 2.64406779661017e-07,
      "loss": 1.1551,
      "step": 29610
    },
    {
      "epoch": 0.74,
      "grad_norm": 8.990274429321289,
      "learning_rate": 2.576271186440678e-07,
      "loss": 1.2577,
      "step": 29620
    },
    {
      "epoch": 0.74,
      "grad_norm": 9.873722076416016,
      "learning_rate": 2.5084745762711867e-07,
      "loss": 1.333,
      "step": 29630
    },
    {
      "epoch": 0.74,
      "grad_norm": 10.069340705871582,
      "learning_rate": 2.4406779661016954e-07,
      "loss": 1.4405,
      "step": 29640
    },
    {
      "epoch": 0.74,
      "grad_norm": 4.778994083404541,
      "learning_rate": 2.3728813559322036e-07,
      "loss": 1.5492,
      "step": 29650
    },
    {
      "epoch": 0.74,
      "grad_norm": 14.848403930664062,
      "learning_rate": 2.305084745762712e-07,
      "loss": 1.136,
      "step": 29660
    },
    {
      "epoch": 0.74,
      "grad_norm": 17.621976852416992,
      "learning_rate": 2.2372881355932207e-07,
      "loss": 1.3884,
      "step": 29670
    },
    {
      "epoch": 0.74,
      "grad_norm": 5.0682597160339355,
      "learning_rate": 2.1694915254237292e-07,
      "loss": 1.2395,
      "step": 29680
    },
    {
      "epoch": 0.74,
      "grad_norm": 10.465557098388672,
      "learning_rate": 2.1016949152542376e-07,
      "loss": 1.3657,
      "step": 29690
    },
    {
      "epoch": 0.74,
      "grad_norm": 16.993274688720703,
      "learning_rate": 2.0338983050847458e-07,
      "loss": 1.2805,
      "step": 29700
    },
    {
      "epoch": 0.74,
      "grad_norm": 8.59652042388916,
      "learning_rate": 1.9661016949152543e-07,
      "loss": 1.0588,
      "step": 29710
    },
    {
      "epoch": 0.74,
      "grad_norm": 3.580928087234497,
      "learning_rate": 1.8983050847457627e-07,
      "loss": 1.3986,
      "step": 29720
    },
    {
      "epoch": 0.74,
      "grad_norm": 5.2861552238464355,
      "learning_rate": 1.8305084745762712e-07,
      "loss": 1.1782,
      "step": 29730
    },
    {
      "epoch": 0.74,
      "grad_norm": 16.308183670043945,
      "learning_rate": 1.76271186440678e-07,
      "loss": 1.153,
      "step": 29740
    },
    {
      "epoch": 0.74,
      "grad_norm": 4.2665863037109375,
      "learning_rate": 1.6949152542372883e-07,
      "loss": 1.3667,
      "step": 29750
    },
    {
      "epoch": 0.74,
      "grad_norm": 7.868191719055176,
      "learning_rate": 1.6271186440677968e-07,
      "loss": 1.2511,
      "step": 29760
    },
    {
      "epoch": 0.74,
      "grad_norm": 6.3579936027526855,
      "learning_rate": 1.5593220338983052e-07,
      "loss": 1.4062,
      "step": 29770
    },
    {
      "epoch": 0.74,
      "grad_norm": 9.477784156799316,
      "learning_rate": 1.4915254237288137e-07,
      "loss": 1.2801,
      "step": 29780
    },
    {
      "epoch": 0.74,
      "grad_norm": 5.881575584411621,
      "learning_rate": 1.423728813559322e-07,
      "loss": 1.343,
      "step": 29790
    },
    {
      "epoch": 0.74,
      "grad_norm": 6.489737510681152,
      "learning_rate": 1.3559322033898305e-07,
      "loss": 1.302,
      "step": 29800
    },
    {
      "epoch": 0.75,
      "grad_norm": 14.81570816040039,
      "learning_rate": 1.288135593220339e-07,
      "loss": 1.251,
      "step": 29810
    },
    {
      "epoch": 0.75,
      "grad_norm": 15.79887866973877,
      "learning_rate": 1.2203389830508477e-07,
      "loss": 1.3992,
      "step": 29820
    },
    {
      "epoch": 0.75,
      "grad_norm": 5.4701828956604,
      "learning_rate": 1.152542372881356e-07,
      "loss": 1.3728,
      "step": 29830
    },
    {
      "epoch": 0.75,
      "grad_norm": 5.326042175292969,
      "learning_rate": 1.0847457627118646e-07,
      "loss": 1.3645,
      "step": 29840
    },
    {
      "epoch": 0.75,
      "grad_norm": 3.9920434951782227,
      "learning_rate": 1.0169491525423729e-07,
      "loss": 1.4768,
      "step": 29850
    },
    {
      "epoch": 0.75,
      "grad_norm": 10.122292518615723,
      "learning_rate": 9.491525423728814e-08,
      "loss": 1.4596,
      "step": 29860
    },
    {
      "epoch": 0.75,
      "grad_norm": 6.636433124542236,
      "learning_rate": 8.8135593220339e-08,
      "loss": 1.3858,
      "step": 29870
    },
    {
      "epoch": 0.75,
      "grad_norm": 6.18862247467041,
      "learning_rate": 8.135593220338984e-08,
      "loss": 1.3912,
      "step": 29880
    },
    {
      "epoch": 0.75,
      "grad_norm": 10.676321983337402,
      "learning_rate": 7.457627118644068e-08,
      "loss": 1.3445,
      "step": 29890
    },
    {
      "epoch": 0.75,
      "grad_norm": 14.703264236450195,
      "learning_rate": 6.779661016949153e-08,
      "loss": 1.3391,
      "step": 29900
    },
    {
      "epoch": 0.75,
      "grad_norm": 11.101775169372559,
      "learning_rate": 6.101694915254239e-08,
      "loss": 1.2044,
      "step": 29910
    },
    {
      "epoch": 0.75,
      "grad_norm": 5.793562889099121,
      "learning_rate": 5.423728813559323e-08,
      "loss": 1.3205,
      "step": 29920
    },
    {
      "epoch": 0.75,
      "grad_norm": 4.186032295227051,
      "learning_rate": 4.745762711864407e-08,
      "loss": 1.3155,
      "step": 29930
    },
    {
      "epoch": 0.75,
      "grad_norm": 5.155704021453857,
      "learning_rate": 4.067796610169492e-08,
      "loss": 1.3661,
      "step": 29940
    },
    {
      "epoch": 0.75,
      "grad_norm": 10.625030517578125,
      "learning_rate": 3.3898305084745764e-08,
      "loss": 1.2094,
      "step": 29950
    },
    {
      "epoch": 0.75,
      "grad_norm": 9.991485595703125,
      "learning_rate": 2.7118644067796615e-08,
      "loss": 1.2198,
      "step": 29960
    },
    {
      "epoch": 0.75,
      "grad_norm": 3.559561014175415,
      "learning_rate": 2.033898305084746e-08,
      "loss": 1.33,
      "step": 29970
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.8645702600479126,
      "learning_rate": 1.3559322033898307e-08,
      "loss": 1.3138,
      "step": 29980
    },
    {
      "epoch": 0.75,
      "grad_norm": 4.156149387359619,
      "learning_rate": 6.779661016949154e-09,
      "loss": 1.2674,
      "step": 29990
    },
    {
      "epoch": 0.75,
      "grad_norm": 4.104331016540527,
      "learning_rate": 0.0,
      "loss": 1.2401,
      "step": 30000
    },
    {
      "epoch": 0.75,
      "eval_loss": 1.2914531230926514,
      "eval_runtime": 122.439,
      "eval_samples_per_second": 8.167,
      "eval_steps_per_second": 8.167,
      "step": 30000
    }
  ],
  "logging_steps": 10,
  "max_steps": 30000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 2500,
  "total_flos": 4.8306377981952e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}