{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0588137976673009,
  "eval_steps": 1600,
  "global_step": 16000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 1.6154260635375977,
      "learning_rate": 6.250000000000001e-08,
      "loss": 0.6581,
      "step": 10
    },
    {
      "epoch": 0.0,
      "grad_norm": 1.434083104133606,
      "learning_rate": 1.2500000000000002e-07,
      "loss": 0.65,
      "step": 20
    },
    {
      "epoch": 0.0,
      "grad_norm": 1.770402193069458,
      "learning_rate": 1.875e-07,
      "loss": 0.6538,
      "step": 30
    },
    {
      "epoch": 0.0,
      "grad_norm": 1.5259037017822266,
      "learning_rate": 2.5000000000000004e-07,
      "loss": 0.6661,
      "step": 40
    },
    {
      "epoch": 0.0,
      "grad_norm": 1.3413686752319336,
      "learning_rate": 3.125e-07,
      "loss": 0.6629,
      "step": 50
    },
    {
      "epoch": 0.0,
      "grad_norm": 1.8524203300476074,
      "learning_rate": 3.75e-07,
      "loss": 0.652,
      "step": 60
    },
    {
      "epoch": 0.0,
      "grad_norm": 1.6129018068313599,
      "learning_rate": 4.375e-07,
      "loss": 0.6503,
      "step": 70
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.3771299123764038,
      "learning_rate": 5.000000000000001e-07,
      "loss": 0.6568,
      "step": 80
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.4569002389907837,
      "learning_rate": 5.625e-07,
      "loss": 0.6592,
      "step": 90
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.3211089372634888,
      "learning_rate": 6.25e-07,
      "loss": 0.6609,
      "step": 100
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.3682845830917358,
      "learning_rate": 6.875000000000001e-07,
      "loss": 0.6655,
      "step": 110
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.0182406902313232,
      "learning_rate": 7.5e-07,
      "loss": 0.6581,
      "step": 120
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.5314549207687378,
      "learning_rate": 8.125000000000001e-07,
      "loss": 0.6554,
      "step": 130
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.4846184253692627,
      "learning_rate": 8.75e-07,
      "loss": 0.6555,
      "step": 140
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.568630576133728,
      "learning_rate": 9.375000000000001e-07,
      "loss": 0.6567,
      "step": 150
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.1758819818496704,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 0.6448,
      "step": 160
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.6889770030975342,
      "learning_rate": 1.0625e-06,
      "loss": 0.6352,
      "step": 170
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.4104373455047607,
      "learning_rate": 1.125e-06,
      "loss": 0.64,
      "step": 180
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.5935137271881104,
      "learning_rate": 1.1875e-06,
      "loss": 0.6647,
      "step": 190
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.5472041368484497,
      "learning_rate": 1.25e-06,
      "loss": 0.637,
      "step": 200
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.2572060823440552,
      "learning_rate": 1.3125000000000001e-06,
      "loss": 0.6642,
      "step": 210
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.3120354413986206,
      "learning_rate": 1.3750000000000002e-06,
      "loss": 0.6713,
      "step": 220
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.480773687362671,
      "learning_rate": 1.4375e-06,
      "loss": 0.6765,
      "step": 230
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.6741840839385986,
      "learning_rate": 1.5e-06,
      "loss": 0.662,
      "step": 240
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.2385046482086182,
      "learning_rate": 1.5625e-06,
      "loss": 0.6563,
      "step": 250
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.4410344362258911,
      "learning_rate": 1.6250000000000001e-06,
      "loss": 0.6524,
      "step": 260
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.2795886993408203,
      "learning_rate": 1.6875000000000001e-06,
      "loss": 0.655,
      "step": 270
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.6680865287780762,
      "learning_rate": 1.75e-06,
      "loss": 0.6577,
      "step": 280
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.456658959388733,
      "learning_rate": 1.8125e-06,
      "loss": 0.6687,
      "step": 290
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.4947397708892822,
      "learning_rate": 1.8750000000000003e-06,
      "loss": 0.6685,
      "step": 300
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.6426098346710205,
      "learning_rate": 1.9375e-06,
      "loss": 0.653,
      "step": 310
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.3753637075424194,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.6543,
      "step": 320
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.4655407667160034,
      "learning_rate": 2.0625e-06,
      "loss": 0.654,
      "step": 330
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.6397826671600342,
      "learning_rate": 2.125e-06,
      "loss": 0.6604,
      "step": 340
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.4749575853347778,
      "learning_rate": 2.1875000000000002e-06,
      "loss": 0.6587,
      "step": 350
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.650220274925232,
      "learning_rate": 2.25e-06,
      "loss": 0.6578,
      "step": 360
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.290277123451233,
      "learning_rate": 2.3125000000000003e-06,
      "loss": 0.6607,
      "step": 370
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.5285893678665161,
      "learning_rate": 2.375e-06,
      "loss": 0.6594,
      "step": 380
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.8043378591537476,
      "learning_rate": 2.4375e-06,
      "loss": 0.6539,
      "step": 390
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.5191413164138794,
      "learning_rate": 2.5e-06,
      "loss": 0.6716,
      "step": 400
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.7285609245300293,
      "learning_rate": 2.5625e-06,
      "loss": 0.6462,
      "step": 410
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.2920969724655151,
      "learning_rate": 2.6250000000000003e-06,
      "loss": 0.6601,
      "step": 420
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.4754414558410645,
      "learning_rate": 2.6875e-06,
      "loss": 0.6574,
      "step": 430
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.6130346059799194,
      "learning_rate": 2.7500000000000004e-06,
      "loss": 0.6625,
      "step": 440
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.4590799808502197,
      "learning_rate": 2.8125e-06,
      "loss": 0.6793,
      "step": 450
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.7691385746002197,
      "learning_rate": 2.875e-06,
      "loss": 0.6616,
      "step": 460
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.7910367250442505,
      "learning_rate": 2.9375000000000003e-06,
      "loss": 0.6534,
      "step": 470
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.179307460784912,
      "learning_rate": 3e-06,
      "loss": 0.6698,
      "step": 480
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.0902700424194336,
      "learning_rate": 3.0625000000000003e-06,
      "loss": 0.6707,
      "step": 490
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.3620158433914185,
      "learning_rate": 3.125e-06,
      "loss": 0.6648,
      "step": 500
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.5206565856933594,
      "learning_rate": 3.1875e-06,
      "loss": 0.6733,
      "step": 510
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.7129814624786377,
      "learning_rate": 3.2500000000000002e-06,
      "loss": 0.6648,
      "step": 520
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.7053027153015137,
      "learning_rate": 3.3125e-06,
      "loss": 0.6711,
      "step": 530
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.8278532028198242,
      "learning_rate": 3.3750000000000003e-06,
      "loss": 0.6582,
      "step": 540
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.8324699401855469,
      "learning_rate": 3.4375e-06,
      "loss": 0.6746,
      "step": 550
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.597606897354126,
      "learning_rate": 3.5e-06,
      "loss": 0.6647,
      "step": 560
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.7364826202392578,
      "learning_rate": 3.5625e-06,
      "loss": 0.6577,
      "step": 570
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.0563762187957764,
      "learning_rate": 3.625e-06,
      "loss": 0.6711,
      "step": 580
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.8743737936019897,
      "learning_rate": 3.6875000000000007e-06,
      "loss": 0.6623,
      "step": 590
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.7393046617507935,
      "learning_rate": 3.7500000000000005e-06,
      "loss": 0.6453,
      "step": 600
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.464277744293213,
      "learning_rate": 3.8125e-06,
      "loss": 0.6528,
      "step": 610
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.5163260698318481,
      "learning_rate": 3.875e-06,
      "loss": 0.6682,
      "step": 620
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.5761061906814575,
      "learning_rate": 3.9375e-06,
      "loss": 0.6647,
      "step": 630
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.468314528465271,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.6572,
      "step": 640
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.7023948431015015,
      "learning_rate": 4.0625000000000005e-06,
      "loss": 0.6527,
      "step": 650
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.502794861793518,
      "learning_rate": 4.125e-06,
      "loss": 0.6444,
      "step": 660
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.8254650831222534,
      "learning_rate": 4.1875e-06,
      "loss": 0.6524,
      "step": 670
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.577621340751648,
      "learning_rate": 4.25e-06,
      "loss": 0.644,
      "step": 680
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.2948737144470215,
      "learning_rate": 4.312500000000001e-06,
      "loss": 0.6617,
      "step": 690
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.8099135160446167,
      "learning_rate": 4.3750000000000005e-06,
      "loss": 0.6569,
      "step": 700
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.51678466796875,
      "learning_rate": 4.4375e-06,
      "loss": 0.6637,
      "step": 710
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.8125680685043335,
      "learning_rate": 4.5e-06,
      "loss": 0.6525,
      "step": 720
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.2971110343933105,
      "learning_rate": 4.5625e-06,
      "loss": 0.6658,
      "step": 730
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.7269909381866455,
      "learning_rate": 4.625000000000001e-06,
      "loss": 0.6461,
      "step": 740
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.6699669361114502,
      "learning_rate": 4.6875000000000004e-06,
      "loss": 0.6569,
      "step": 750
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.5642304420471191,
      "learning_rate": 4.75e-06,
      "loss": 0.6394,
      "step": 760
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.1117033958435059,
      "learning_rate": 4.8125e-06,
      "loss": 0.6547,
      "step": 770
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.453066349029541,
      "learning_rate": 4.875e-06,
      "loss": 0.6565,
      "step": 780
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.6391645669937134,
      "learning_rate": 4.937500000000001e-06,
      "loss": 0.6586,
      "step": 790
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.4576597213745117,
      "learning_rate": 5e-06,
      "loss": 0.6569,
      "step": 800
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.4210525751113892,
      "learning_rate": 5.0625e-06,
      "loss": 0.6439,
      "step": 810
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.8609915971755981,
      "learning_rate": 5.125e-06,
      "loss": 0.6543,
      "step": 820
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.8662998676300049,
      "learning_rate": 5.187500000000001e-06,
      "loss": 0.6596,
      "step": 830
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.3369812965393066,
      "learning_rate": 5.2500000000000006e-06,
      "loss": 0.6538,
      "step": 840
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.643251657485962,
      "learning_rate": 5.3125e-06,
      "loss": 0.6422,
      "step": 850
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.5795202255249023,
      "learning_rate": 5.375e-06,
      "loss": 0.6567,
      "step": 860
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.4530309438705444,
      "learning_rate": 5.4375e-06,
      "loss": 0.6613,
      "step": 870
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.514719843864441,
      "learning_rate": 5.500000000000001e-06,
      "loss": 0.6644,
      "step": 880
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.7065812349319458,
      "learning_rate": 5.5625000000000005e-06,
      "loss": 0.6512,
      "step": 890
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.272714376449585,
      "learning_rate": 5.625e-06,
      "loss": 0.6693,
      "step": 900
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.5543745756149292,
      "learning_rate": 5.6875e-06,
      "loss": 0.6469,
      "step": 910
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.5970263481140137,
      "learning_rate": 5.75e-06,
      "loss": 0.6492,
      "step": 920
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.426156759262085,
      "learning_rate": 5.812500000000001e-06,
      "loss": 0.6232,
      "step": 930
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.5355912446975708,
      "learning_rate": 5.8750000000000005e-06,
      "loss": 0.6603,
      "step": 940
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.58833909034729,
      "learning_rate": 5.9375e-06,
      "loss": 0.6448,
      "step": 950
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.565136194229126,
      "learning_rate": 6e-06,
      "loss": 0.6434,
      "step": 960
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.7123429775238037,
      "learning_rate": 6.0625e-06,
      "loss": 0.6724,
      "step": 970
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.8091198205947876,
      "learning_rate": 6.125000000000001e-06,
      "loss": 0.659,
      "step": 980
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.606202244758606,
      "learning_rate": 6.1875000000000005e-06,
      "loss": 0.6351,
      "step": 990
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.8575665950775146,
      "learning_rate": 6.25e-06,
      "loss": 0.6766,
      "step": 1000
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.4026097059249878,
      "learning_rate": 6.3125e-06,
      "loss": 0.6549,
      "step": 1010
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.772989273071289,
      "learning_rate": 6.375e-06,
      "loss": 0.6579,
      "step": 1020
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.7279027700424194,
      "learning_rate": 6.437500000000001e-06,
      "loss": 0.6604,
      "step": 1030
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.0833723545074463,
      "learning_rate": 6.5000000000000004e-06,
      "loss": 0.6434,
      "step": 1040
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.8391436338424683,
      "learning_rate": 6.5625e-06,
      "loss": 0.6495,
      "step": 1050
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.104515790939331,
      "learning_rate": 6.625e-06,
      "loss": 0.6573,
      "step": 1060
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.012073516845703,
      "learning_rate": 6.6875e-06,
      "loss": 0.6601,
      "step": 1070
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.5166912078857422,
      "learning_rate": 6.750000000000001e-06,
      "loss": 0.687,
      "step": 1080
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.0484936237335205,
      "learning_rate": 6.8125e-06,
      "loss": 0.6362,
      "step": 1090
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.7620980739593506,
      "learning_rate": 6.875e-06,
      "loss": 0.6295,
      "step": 1100
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.7010966539382935,
      "learning_rate": 6.9375e-06,
      "loss": 0.662,
      "step": 1110
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.4270750284194946,
      "learning_rate": 7e-06,
      "loss": 0.6696,
      "step": 1120
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.3873461484909058,
      "learning_rate": 7.062500000000001e-06,
      "loss": 0.6568,
      "step": 1130
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.8650726079940796,
      "learning_rate": 7.125e-06,
      "loss": 0.6574,
      "step": 1140
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.807304859161377,
      "learning_rate": 7.1875e-06,
      "loss": 0.6714,
      "step": 1150
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.5310503244400024,
      "learning_rate": 7.25e-06,
      "loss": 0.6476,
      "step": 1160
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.4306132793426514,
      "learning_rate": 7.3125e-06,
      "loss": 0.6508,
      "step": 1170
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.687558650970459,
      "learning_rate": 7.375000000000001e-06,
      "loss": 0.6534,
      "step": 1180
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.5304369926452637,
      "learning_rate": 7.437500000000001e-06,
      "loss": 0.6651,
      "step": 1190
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.6289629936218262,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.6712,
      "step": 1200
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.8669979572296143,
      "learning_rate": 7.5625e-06,
      "loss": 0.6637,
      "step": 1210
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.877084732055664,
      "learning_rate": 7.625e-06,
      "loss": 0.6521,
      "step": 1220
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.906940221786499,
      "learning_rate": 7.6875e-06,
      "loss": 0.6381,
      "step": 1230
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.4054601192474365,
      "learning_rate": 7.75e-06,
      "loss": 0.6528,
      "step": 1240
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.1831490993499756,
      "learning_rate": 7.8125e-06,
      "loss": 0.6424,
      "step": 1250
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.7532954216003418,
      "learning_rate": 7.875e-06,
      "loss": 0.6591,
      "step": 1260
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.4256421327590942,
      "learning_rate": 7.9375e-06,
      "loss": 0.644,
      "step": 1270
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.645342230796814,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.6786,
      "step": 1280
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.749700665473938,
      "learning_rate": 8.062500000000001e-06,
      "loss": 0.6583,
      "step": 1290
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.5855311155319214,
      "learning_rate": 8.125000000000001e-06,
      "loss": 0.6388,
      "step": 1300
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.7551072835922241,
      "learning_rate": 8.1875e-06,
      "loss": 0.6638,
      "step": 1310
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.656398892402649,
      "learning_rate": 8.25e-06,
      "loss": 0.6486,
      "step": 1320
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.3058027029037476,
      "learning_rate": 8.3125e-06,
      "loss": 0.6678,
      "step": 1330
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.9038392305374146,
      "learning_rate": 8.375e-06,
      "loss": 0.6576,
      "step": 1340
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.7255905866622925,
      "learning_rate": 8.4375e-06,
      "loss": 0.6432,
      "step": 1350
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.0054244995117188,
      "learning_rate": 8.5e-06,
      "loss": 0.6642,
      "step": 1360
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.624997615814209,
      "learning_rate": 8.5625e-06,
      "loss": 0.6659,
      "step": 1370
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.5902334451675415,
      "learning_rate": 8.625000000000001e-06,
      "loss": 0.6707,
      "step": 1380
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.2210817337036133,
      "learning_rate": 8.687500000000001e-06,
      "loss": 0.6713,
      "step": 1390
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.6660022735595703,
      "learning_rate": 8.750000000000001e-06,
      "loss": 0.6637,
      "step": 1400
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.5590497255325317,
      "learning_rate": 8.8125e-06,
      "loss": 0.6643,
      "step": 1410
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.6343599557876587,
      "learning_rate": 8.875e-06,
      "loss": 0.6496,
      "step": 1420
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.755144476890564,
      "learning_rate": 8.9375e-06,
      "loss": 0.654,
      "step": 1430
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.1182198524475098,
      "learning_rate": 9e-06,
      "loss": 0.6567,
      "step": 1440
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.6853889226913452,
      "learning_rate": 9.0625e-06,
      "loss": 0.6651,
      "step": 1450
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.7039555311203003,
      "learning_rate": 9.125e-06,
      "loss": 0.6376,
      "step": 1460
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.2666831016540527,
      "learning_rate": 9.1875e-06,
      "loss": 0.638,
      "step": 1470
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.2994011640548706,
      "learning_rate": 9.250000000000001e-06,
      "loss": 0.665,
      "step": 1480
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.4676392078399658,
      "learning_rate": 9.312500000000001e-06,
      "loss": 0.6592,
      "step": 1490
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.501640558242798,
      "learning_rate": 9.375000000000001e-06,
      "loss": 0.6532,
      "step": 1500
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.9528775215148926,
      "learning_rate": 9.4375e-06,
      "loss": 0.6697,
      "step": 1510
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.0028085708618164,
      "learning_rate": 9.5e-06,
      "loss": 0.6705,
      "step": 1520
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.638511061668396,
      "learning_rate": 9.562500000000002e-06,
      "loss": 0.6712,
      "step": 1530
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.7564353942871094,
      "learning_rate": 9.625e-06,
      "loss": 0.6543,
      "step": 1540
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.4182738065719604,
      "learning_rate": 9.6875e-06,
      "loss": 0.6614,
      "step": 1550
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.9909614324569702,
      "learning_rate": 9.75e-06,
      "loss": 0.6676,
      "step": 1560
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.5370928049087524,
      "learning_rate": 9.8125e-06,
      "loss": 0.6585,
      "step": 1570
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.7557249069213867,
      "learning_rate": 9.875000000000001e-06,
      "loss": 0.6462,
      "step": 1580
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.671527624130249,
      "learning_rate": 9.937500000000001e-06,
      "loss": 0.6862,
      "step": 1590
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.6722038984298706,
      "learning_rate": 1e-05,
      "loss": 0.6518,
      "step": 1600
    },
    {
      "epoch": 0.11,
      "eval_loss": 0.7220360636711121,
      "eval_runtime": 134.694,
      "eval_samples_per_second": 81.667,
      "eval_steps_per_second": 10.208,
      "step": 1600
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.6332907676696777,
      "learning_rate": 9.99998810088676e-06,
      "loss": 0.6458,
      "step": 1610
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.5682039260864258,
      "learning_rate": 9.999952403603674e-06,
      "loss": 0.6709,
      "step": 1620
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.3811382055282593,
      "learning_rate": 9.999892908320647e-06,
      "loss": 0.6468,
      "step": 1630
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.5780935287475586,
      "learning_rate": 9.999809615320857e-06,
      "loss": 0.6345,
      "step": 1640
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.7821502685546875,
      "learning_rate": 9.99970252500075e-06,
      "loss": 0.6425,
      "step": 1650
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.6573759317398071,
      "learning_rate": 9.999571637870035e-06,
      "loss": 0.6448,
      "step": 1660
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.8963578939437866,
      "learning_rate": 9.999416954551693e-06,
      "loss": 0.6699,
      "step": 1670
    },
    {
      "epoch": 0.11,
      "grad_norm": 2.0272181034088135,
      "learning_rate": 9.999238475781957e-06,
      "loss": 0.6593,
      "step": 1680
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.6698834896087646,
      "learning_rate": 9.999036202410324e-06,
      "loss": 0.6589,
      "step": 1690
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.7105271816253662,
      "learning_rate": 9.998810135399545e-06,
      "loss": 0.645,
      "step": 1700
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.5127520561218262,
      "learning_rate": 9.99856027582562e-06,
      "loss": 0.6435,
      "step": 1710
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.930246114730835,
      "learning_rate": 9.998286624877786e-06,
      "loss": 0.658,
      "step": 1720
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.6141360998153687,
      "learning_rate": 9.997989183858531e-06,
      "loss": 0.6503,
      "step": 1730
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.8381562232971191,
      "learning_rate": 9.997667954183566e-06,
      "loss": 0.6619,
      "step": 1740
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.8966178894042969,
      "learning_rate": 9.997322937381829e-06,
      "loss": 0.644,
      "step": 1750
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.9953837394714355,
      "learning_rate": 9.99695413509548e-06,
      "loss": 0.6716,
      "step": 1760
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.9095193147659302,
      "learning_rate": 9.996561549079886e-06,
      "loss": 0.6741,
      "step": 1770
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.4451298713684082,
      "learning_rate": 9.996145181203616e-06,
      "loss": 0.6781,
      "step": 1780
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.8144304752349854,
      "learning_rate": 9.995705033448435e-06,
      "loss": 0.6751,
      "step": 1790
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.5360352993011475,
      "learning_rate": 9.99524110790929e-06,
      "loss": 0.6658,
      "step": 1800
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.858272910118103,
      "learning_rate": 9.994753406794303e-06,
      "loss": 0.6563,
      "step": 1810
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.853493571281433,
      "learning_rate": 9.994241932424755e-06,
      "loss": 0.6695,
      "step": 1820
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.5680876970291138,
      "learning_rate": 9.993706687235085e-06,
      "loss": 0.6632,
      "step": 1830
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.5058163404464722,
      "learning_rate": 9.993147673772869e-06,
      "loss": 0.6767,
      "step": 1840
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.5722607374191284,
      "learning_rate": 9.992564894698816e-06,
      "loss": 0.6661,
      "step": 1850
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.6947115659713745,
      "learning_rate": 9.991958352786744e-06,
      "loss": 0.6316,
      "step": 1860
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.8827723264694214,
      "learning_rate": 9.99132805092358e-06,
      "loss": 0.6517,
      "step": 1870
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.6634174585342407,
      "learning_rate": 9.990673992109335e-06,
      "loss": 0.6425,
      "step": 1880
    },
    {
      "epoch": 0.13,
      "grad_norm": 2.319075107574463,
      "learning_rate": 9.9899961794571e-06,
      "loss": 0.6636,
      "step": 1890
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.733626365661621,
      "learning_rate": 9.989294616193018e-06,
      "loss": 0.6459,
      "step": 1900
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.7393757104873657,
      "learning_rate": 9.988569305656286e-06,
      "loss": 0.6623,
      "step": 1910
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.7465229034423828,
      "learning_rate": 9.987820251299121e-06,
      "loss": 0.6554,
      "step": 1920
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.940171718597412,
      "learning_rate": 9.98704745668676e-06,
      "loss": 0.641,
      "step": 1930
    },
    {
      "epoch": 0.13,
      "grad_norm": 2.001396656036377,
      "learning_rate": 9.986250925497429e-06,
      "loss": 0.6558,
      "step": 1940
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.8247671127319336,
      "learning_rate": 9.985430661522333e-06,
      "loss": 0.6703,
      "step": 1950
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.653157114982605,
      "learning_rate": 9.984586668665641e-06,
      "loss": 0.6426,
      "step": 1960
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.9459385871887207,
      "learning_rate": 9.983718950944457e-06,
      "loss": 0.656,
      "step": 1970
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.8265575170516968,
      "learning_rate": 9.982827512488809e-06,
      "loss": 0.6623,
      "step": 1980
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.6862218379974365,
      "learning_rate": 9.981912357541628e-06,
      "loss": 0.6457,
      "step": 1990
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.6471173763275146,
      "learning_rate": 9.980973490458728e-06,
      "loss": 0.6565,
      "step": 2000
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.7987889051437378,
      "learning_rate": 9.980010915708785e-06,
      "loss": 0.6526,
      "step": 2010
    },
    {
      "epoch": 0.13,
      "grad_norm": 2.383647918701172,
      "learning_rate": 9.979024637873309e-06,
      "loss": 0.6532,
      "step": 2020
    },
    {
      "epoch": 0.13,
      "grad_norm": 2.297119379043579,
      "learning_rate": 9.978014661646637e-06,
      "loss": 0.6488,
      "step": 2030
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.8215795755386353,
      "learning_rate": 9.976980991835896e-06,
      "loss": 0.6537,
      "step": 2040
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.737383484840393,
      "learning_rate": 9.975923633360985e-06,
      "loss": 0.6375,
      "step": 2050
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.5321128368377686,
      "learning_rate": 9.974842591254559e-06,
      "loss": 0.6346,
      "step": 2060
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.6683794260025024,
      "learning_rate": 9.973737870661995e-06,
      "loss": 0.6484,
      "step": 2070
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.3653274774551392,
      "learning_rate": 9.972609476841368e-06,
      "loss": 0.6352,
      "step": 2080
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.6334515810012817,
      "learning_rate": 9.971457415163435e-06,
      "loss": 0.6514,
      "step": 2090
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.4465519189834595,
      "learning_rate": 9.970281691111598e-06,
      "loss": 0.6532,
      "step": 2100
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.7188860177993774,
      "learning_rate": 9.96908231028189e-06,
      "loss": 0.6427,
      "step": 2110
    },
    {
      "epoch": 0.14,
      "grad_norm": 2.0040030479431152,
      "learning_rate": 9.967859278382939e-06,
      "loss": 0.6457,
      "step": 2120
    },
    {
      "epoch": 0.14,
      "grad_norm": 2.0880534648895264,
      "learning_rate": 9.96661260123594e-06,
      "loss": 0.6407,
      "step": 2130
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.6864984035491943,
      "learning_rate": 9.965342284774633e-06,
      "loss": 0.6569,
      "step": 2140
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.4719489812850952,
      "learning_rate": 9.964048335045276e-06,
      "loss": 0.6439,
      "step": 2150
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.5283831357955933,
      "learning_rate": 9.962730758206612e-06,
      "loss": 0.6606,
      "step": 2160
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.6295084953308105,
      "learning_rate": 9.961389560529835e-06,
      "loss": 0.6726,
      "step": 2170
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.9193845987319946,
      "learning_rate": 9.960024748398576e-06,
      "loss": 0.6464,
      "step": 2180
    },
    {
      "epoch": 0.14,
      "grad_norm": 2.2357025146484375,
      "learning_rate": 9.958636328308852e-06,
      "loss": 0.6371,
      "step": 2190
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.66888427734375,
      "learning_rate": 9.957224306869053e-06,
      "loss": 0.6607,
      "step": 2200
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.6502246856689453,
      "learning_rate": 9.9557886907999e-06,
      "loss": 0.6436,
      "step": 2210
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.6644810438156128,
      "learning_rate": 9.954329486934411e-06,
      "loss": 0.6552,
      "step": 2220
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.6509498357772827,
      "learning_rate": 9.952846702217886e-06,
      "loss": 0.6501,
      "step": 2230
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.7857375144958496,
      "learning_rate": 9.951340343707852e-06,
      "loss": 0.6603,
      "step": 2240
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.4443598985671997,
      "learning_rate": 9.94981041857404e-06,
      "loss": 0.6351,
      "step": 2250
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.636020302772522,
      "learning_rate": 9.948256934098353e-06,
      "loss": 0.6548,
      "step": 2260
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.6805977821350098,
      "learning_rate": 9.946679897674823e-06,
      "loss": 0.6643,
      "step": 2270
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.5527812242507935,
      "learning_rate": 9.945079316809585e-06,
      "loss": 0.6417,
      "step": 2280
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.5074429512023926,
      "learning_rate": 9.943455199120836e-06,
      "loss": 0.662,
      "step": 2290
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.5941535234451294,
      "learning_rate": 9.941807552338805e-06,
      "loss": 0.6613,
      "step": 2300
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.7628566026687622,
      "learning_rate": 9.940136384305699e-06,
      "loss": 0.6624,
      "step": 2310
    },
    {
      "epoch": 0.15,
      "grad_norm": 2.036191940307617,
      "learning_rate": 9.938441702975689e-06,
      "loss": 0.6615,
      "step": 2320
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.7616183757781982,
      "learning_rate": 9.936723516414857e-06,
      "loss": 0.6636,
      "step": 2330
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.732585072517395,
      "learning_rate": 9.934981832801161e-06,
      "loss": 0.6687,
      "step": 2340
    },
    {
      "epoch": 0.16,
      "grad_norm": 2.074768543243408,
      "learning_rate": 9.933216660424396e-06,
      "loss": 0.6649,
      "step": 2350
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.9834880828857422,
      "learning_rate": 9.931428007686158e-06,
      "loss": 0.645,
      "step": 2360
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.7717509269714355,
      "learning_rate": 9.9296158830998e-06,
      "loss": 0.6571,
      "step": 2370
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.8364372253417969,
      "learning_rate": 9.92778029529039e-06,
      "loss": 0.6428,
      "step": 2380
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.92979896068573,
      "learning_rate": 9.925921252994677e-06,
      "loss": 0.6629,
      "step": 2390
    },
    {
      "epoch": 0.16,
      "grad_norm": 2.358120918273926,
      "learning_rate": 9.924038765061042e-06,
      "loss": 0.6375,
      "step": 2400
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.8555611371994019,
      "learning_rate": 9.922132840449459e-06,
      "loss": 0.6455,
      "step": 2410
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.9320988655090332,
      "learning_rate": 9.920203488231455e-06,
      "loss": 0.6426,
      "step": 2420
    },
    {
      "epoch": 0.16,
      "grad_norm": 2.835221767425537,
      "learning_rate": 9.918250717590061e-06,
      "loss": 0.6584,
      "step": 2430
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.8801295757293701,
      "learning_rate": 9.916274537819774e-06,
      "loss": 0.6462,
      "step": 2440
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.548391580581665,
      "learning_rate": 9.914274958326507e-06,
      "loss": 0.6579,
      "step": 2450
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.7065672874450684,
      "learning_rate": 9.91225198862755e-06,
      "loss": 0.6617,
      "step": 2460
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.8889693021774292,
      "learning_rate": 9.91020563835152e-06,
      "loss": 0.6748,
      "step": 2470
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.5888632535934448,
      "learning_rate": 9.908135917238321e-06,
      "loss": 0.6484,
      "step": 2480
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.8980170488357544,
      "learning_rate": 9.90604283513909e-06,
      "loss": 0.6489,
      "step": 2490
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.585455298423767,
      "learning_rate": 9.903926402016153e-06,
      "loss": 0.667,
      "step": 2500
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.7540241479873657,
      "learning_rate": 9.901786627942984e-06,
      "loss": 0.6521,
      "step": 2510
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.97389554977417,
      "learning_rate": 9.899623523104149e-06,
      "loss": 0.6614,
      "step": 2520
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.4442778825759888,
      "learning_rate": 9.897437097795257e-06,
      "loss": 0.6573,
      "step": 2530
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.446113109588623,
      "learning_rate": 9.89522736242292e-06,
      "loss": 0.6522,
      "step": 2540
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.8758691549301147,
      "learning_rate": 9.892994327504693e-06,
      "loss": 0.6511,
      "step": 2550
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.9266222715377808,
      "learning_rate": 9.890738003669029e-06,
      "loss": 0.646,
      "step": 2560
    },
    {
      "epoch": 0.17,
      "grad_norm": 2.077249050140381,
      "learning_rate": 9.888458401655231e-06,
      "loss": 0.6555,
      "step": 2570
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.661763310432434,
      "learning_rate": 9.886155532313396e-06,
      "loss": 0.6391,
      "step": 2580
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.98686683177948,
      "learning_rate": 9.883829406604363e-06,
      "loss": 0.6679,
      "step": 2590
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.9729160070419312,
      "learning_rate": 9.881480035599667e-06,
      "loss": 0.6719,
      "step": 2600
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.4754536151885986,
      "learning_rate": 9.879107430481482e-06,
      "loss": 0.6443,
      "step": 2610
    },
    {
      "epoch": 0.17,
      "grad_norm": 2.3644585609436035,
      "learning_rate": 9.876711602542564e-06,
      "loss": 0.655,
      "step": 2620
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.5546903610229492,
      "learning_rate": 9.874292563186206e-06,
      "loss": 0.6488,
      "step": 2630
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.5309034585952759,
      "learning_rate": 9.871850323926178e-06,
      "loss": 0.6597,
      "step": 2640
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.701562762260437,
      "learning_rate": 9.869384896386669e-06,
      "loss": 0.6409,
      "step": 2650
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.8511505126953125,
      "learning_rate": 9.866896292302243e-06,
      "loss": 0.6562,
      "step": 2660
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.7161003351211548,
      "learning_rate": 9.86438452351777e-06,
      "loss": 0.6519,
      "step": 2670
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.749366283416748,
      "learning_rate": 9.861849601988384e-06,
      "loss": 0.6697,
      "step": 2680
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.707053780555725,
      "learning_rate": 9.859291539779407e-06,
      "loss": 0.664,
      "step": 2690
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.8223627805709839,
      "learning_rate": 9.856710349066307e-06,
      "loss": 0.6626,
      "step": 2700
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.9889775514602661,
      "learning_rate": 9.854106042134642e-06,
      "loss": 0.6317,
      "step": 2710
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.9672682285308838,
      "learning_rate": 9.851478631379982e-06,
      "loss": 0.6414,
      "step": 2720
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.818117618560791,
      "learning_rate": 9.848828129307876e-06,
      "loss": 0.6687,
      "step": 2730
    },
    {
      "epoch": 0.18,
      "grad_norm": 2.1144392490386963,
      "learning_rate": 9.846154548533773e-06,
      "loss": 0.6473,
      "step": 2740
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.5458323955535889,
      "learning_rate": 9.843457901782967e-06,
      "loss": 0.6568,
      "step": 2750
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.7970044612884521,
      "learning_rate": 9.84073820189054e-06,
      "loss": 0.6609,
      "step": 2760
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.4563696384429932,
      "learning_rate": 9.8379954618013e-06,
      "loss": 0.6626,
      "step": 2770
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.5996843576431274,
      "learning_rate": 9.835229694569717e-06,
      "loss": 0.6556,
      "step": 2780
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.6190502643585205,
      "learning_rate": 9.83244091335986e-06,
      "loss": 0.6492,
      "step": 2790
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.9234681129455566,
      "learning_rate": 9.829629131445342e-06,
      "loss": 0.6645,
      "step": 2800
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.7610886096954346,
      "learning_rate": 9.826794362209246e-06,
      "loss": 0.6548,
      "step": 2810
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.6596527099609375,
      "learning_rate": 9.823936619144065e-06,
      "loss": 0.6493,
      "step": 2820
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.7359377145767212,
      "learning_rate": 9.821055915851647e-06,
      "loss": 0.658,
      "step": 2830
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.7483757734298706,
      "learning_rate": 9.818152266043115e-06,
      "loss": 0.64,
      "step": 2840
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.4559893608093262,
      "learning_rate": 9.815225683538814e-06,
      "loss": 0.6546,
      "step": 2850
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.773544192314148,
      "learning_rate": 9.812276182268236e-06,
      "loss": 0.6565,
      "step": 2860
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.798349142074585,
      "learning_rate": 9.809303776269964e-06,
      "loss": 0.6534,
      "step": 2870
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.7769262790679932,
      "learning_rate": 9.806308479691595e-06,
      "loss": 0.6556,
      "step": 2880
    },
    {
      "epoch": 0.19,
      "grad_norm": 2.2573776245117188,
      "learning_rate": 9.803290306789676e-06,
      "loss": 0.6344,
      "step": 2890
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.619582176208496,
      "learning_rate": 9.800249271929645e-06,
      "loss": 0.6554,
      "step": 2900
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.765291690826416,
      "learning_rate": 9.797185389585742e-06,
      "loss": 0.6565,
      "step": 2910
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.5608848333358765,
      "learning_rate": 9.794098674340966e-06,
      "loss": 0.6485,
      "step": 2920
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.6000419855117798,
      "learning_rate": 9.790989140886983e-06,
      "loss": 0.6413,
      "step": 2930
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.367274522781372,
      "learning_rate": 9.787856804024073e-06,
      "loss": 0.6414,
      "step": 2940
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.7870447635650635,
      "learning_rate": 9.784701678661045e-06,
      "loss": 0.6904,
      "step": 2950
    },
    {
      "epoch": 0.2,
      "grad_norm": 2.1525564193725586,
      "learning_rate": 9.781523779815178e-06,
      "loss": 0.658,
      "step": 2960
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.6607723236083984,
      "learning_rate": 9.778323122612143e-06,
      "loss": 0.6476,
      "step": 2970
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.6392195224761963,
      "learning_rate": 9.775099722285934e-06,
      "loss": 0.6591,
      "step": 2980
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.522789478302002,
      "learning_rate": 9.771853594178791e-06,
      "loss": 0.6598,
      "step": 2990
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.791340708732605,
      "learning_rate": 9.768584753741134e-06,
      "loss": 0.6698,
      "step": 3000
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.3436764478683472,
      "learning_rate": 9.765293216531486e-06,
      "loss": 0.6819,
      "step": 3010
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.5194346904754639,
      "learning_rate": 9.761978998216392e-06,
      "loss": 0.671,
      "step": 3020
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.87282133102417,
      "learning_rate": 9.758642114570359e-06,
      "loss": 0.6684,
      "step": 3030
    },
    {
      "epoch": 0.2,
      "grad_norm": 2.3428215980529785,
      "learning_rate": 9.755282581475769e-06,
      "loss": 0.6635,
      "step": 3040
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.9552640914916992,
      "learning_rate": 9.751900414922807e-06,
      "loss": 0.6384,
      "step": 3050
    },
    {
      "epoch": 0.2,
      "grad_norm": 2.020320415496826,
      "learning_rate": 9.748495631009386e-06,
      "loss": 0.6606,
      "step": 3060
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.3121072053909302,
      "learning_rate": 9.745068245941071e-06,
      "loss": 0.671,
      "step": 3070
    },
    {
      "epoch": 0.2,
      "grad_norm": 2.007598400115967,
      "learning_rate": 9.741618276030998e-06,
      "loss": 0.6521,
      "step": 3080
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.9580893516540527,
      "learning_rate": 9.7381457376998e-06,
      "loss": 0.6292,
      "step": 3090
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.6906355619430542,
      "learning_rate": 9.73465064747553e-06,
      "loss": 0.6485,
      "step": 3100
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.9698902368545532,
      "learning_rate": 9.731133021993574e-06,
      "loss": 0.6404,
      "step": 3110
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.9455699920654297,
      "learning_rate": 9.727592877996585e-06,
      "loss": 0.6355,
      "step": 3120
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.5921604633331299,
      "learning_rate": 9.72403023233439e-06,
      "loss": 0.656,
      "step": 3130
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.6062285900115967,
      "learning_rate": 9.720445101963923e-06,
      "loss": 0.6658,
      "step": 3140
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.8297983407974243,
      "learning_rate": 9.716837503949128e-06,
      "loss": 0.6573,
      "step": 3150
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.9983500242233276,
      "learning_rate": 9.713207455460893e-06,
      "loss": 0.6729,
      "step": 3160
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.587420105934143,
      "learning_rate": 9.709554973776962e-06,
      "loss": 0.669,
      "step": 3170
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.468428611755371,
      "learning_rate": 9.705880076281854e-06,
      "loss": 0.6473,
      "step": 3180
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.710042119026184,
      "learning_rate": 9.702182780466775e-06,
      "loss": 0.684,
      "step": 3190
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.8687783479690552,
      "learning_rate": 9.698463103929542e-06,
      "loss": 0.6583,
      "step": 3200
    },
    {
      "epoch": 0.21,
      "eval_loss": 0.7434597015380859,
      "eval_runtime": 134.6713,
      "eval_samples_per_second": 81.68,
      "eval_steps_per_second": 10.21,
      "step": 3200
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.8008785247802734,
      "learning_rate": 9.694721064374497e-06,
      "loss": 0.6572,
      "step": 3210
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.7104676961898804,
      "learning_rate": 9.690956679612422e-06,
      "loss": 0.6357,
      "step": 3220
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.9379181861877441,
      "learning_rate": 9.68716996756045e-06,
      "loss": 0.6741,
      "step": 3230
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.993609070777893,
      "learning_rate": 9.683360946241988e-06,
      "loss": 0.658,
      "step": 3240
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.755682349205017,
      "learning_rate": 9.67952963378663e-06,
      "loss": 0.6803,
      "step": 3250
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.6151280403137207,
      "learning_rate": 9.67567604843006e-06,
      "loss": 0.6693,
      "step": 3260
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.4773739576339722,
      "learning_rate": 9.671800208513978e-06,
      "loss": 0.636,
      "step": 3270
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.6857538223266602,
      "learning_rate": 9.667902132486009e-06,
      "loss": 0.6762,
      "step": 3280
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.871492624282837,
      "learning_rate": 9.663981838899612e-06,
      "loss": 0.6595,
      "step": 3290
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.8173590898513794,
      "learning_rate": 9.660039346413994e-06,
      "loss": 0.6605,
      "step": 3300
    },
    {
      "epoch": 0.22,
      "grad_norm": 2.4486677646636963,
      "learning_rate": 9.656074673794018e-06,
      "loss": 0.6604,
      "step": 3310
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.5115715265274048,
      "learning_rate": 9.652087839910123e-06,
      "loss": 0.6732,
      "step": 3320
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.5585874319076538,
      "learning_rate": 9.648078863738224e-06,
      "loss": 0.6651,
      "step": 3330
    },
    {
      "epoch": 0.22,
      "grad_norm": 2.0782039165496826,
      "learning_rate": 9.644047764359623e-06,
      "loss": 0.6467,
      "step": 3340
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.5483053922653198,
      "learning_rate": 9.639994560960923e-06,
      "loss": 0.6625,
      "step": 3350
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.6379517316818237,
      "learning_rate": 9.635919272833938e-06,
      "loss": 0.668,
      "step": 3360
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.7890597581863403,
      "learning_rate": 9.63182191937559e-06,
      "loss": 0.6528,
      "step": 3370
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.5837633609771729,
      "learning_rate": 9.627702520087833e-06,
      "loss": 0.656,
      "step": 3380
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.770031213760376,
      "learning_rate": 9.623561094577541e-06,
      "loss": 0.6479,
      "step": 3390
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.7771351337432861,
      "learning_rate": 9.619397662556434e-06,
      "loss": 0.6554,
      "step": 3400
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.8997645378112793,
      "learning_rate": 9.615212243840972e-06,
      "loss": 0.6503,
      "step": 3410
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.9709876775741577,
      "learning_rate": 9.61100485835226e-06,
      "loss": 0.6737,
      "step": 3420
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.9905493259429932,
      "learning_rate": 9.606775526115963e-06,
      "loss": 0.6574,
      "step": 3430
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.7140249013900757,
      "learning_rate": 9.602524267262202e-06,
      "loss": 0.6502,
      "step": 3440
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.849632978439331,
      "learning_rate": 9.598251102025463e-06,
      "loss": 0.6641,
      "step": 3450
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.5897934436798096,
      "learning_rate": 9.593956050744493e-06,
      "loss": 0.6373,
      "step": 3460
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.7657532691955566,
      "learning_rate": 9.589639133862214e-06,
      "loss": 0.6568,
      "step": 3470
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.7160342931747437,
      "learning_rate": 9.58530037192562e-06,
      "loss": 0.6671,
      "step": 3480
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.909600019454956,
      "learning_rate": 9.58093978558568e-06,
      "loss": 0.6514,
      "step": 3490
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.8497976064682007,
      "learning_rate": 9.576557395597237e-06,
      "loss": 0.6471,
      "step": 3500
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.4226924180984497,
      "learning_rate": 9.572153222818911e-06,
      "loss": 0.6686,
      "step": 3510
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.7128634452819824,
      "learning_rate": 9.567727288213005e-06,
      "loss": 0.6838,
      "step": 3520
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.6018590927124023,
      "learning_rate": 9.563279612845398e-06,
      "loss": 0.6336,
      "step": 3530
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.5354275703430176,
      "learning_rate": 9.558810217885444e-06,
      "loss": 0.6544,
      "step": 3540
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.7476013898849487,
      "learning_rate": 9.55431912460588e-06,
      "loss": 0.6717,
      "step": 3550
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.68599534034729,
      "learning_rate": 9.549806354382716e-06,
      "loss": 0.6623,
      "step": 3560
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.7562646865844727,
      "learning_rate": 9.54527192869514e-06,
      "loss": 0.6655,
      "step": 3570
    },
    {
      "epoch": 0.24,
      "grad_norm": 2.3129031658172607,
      "learning_rate": 9.540715869125407e-06,
      "loss": 0.6798,
      "step": 3580
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.9537733793258667,
      "learning_rate": 9.536138197358747e-06,
      "loss": 0.638,
      "step": 3590
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.7580291032791138,
      "learning_rate": 9.531538935183252e-06,
      "loss": 0.6663,
      "step": 3600
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.713660717010498,
      "learning_rate": 9.526918104489777e-06,
      "loss": 0.653,
      "step": 3610
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.7625364065170288,
      "learning_rate": 9.522275727271842e-06,
      "loss": 0.6597,
      "step": 3620
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.5197455883026123,
      "learning_rate": 9.51761182562551e-06,
      "loss": 0.6719,
      "step": 3630
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.7578450441360474,
      "learning_rate": 9.512926421749305e-06,
      "loss": 0.649,
      "step": 3640
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.4946767091751099,
      "learning_rate": 9.50821953794408e-06,
      "loss": 0.6493,
      "step": 3650
    },
    {
      "epoch": 0.24,
      "grad_norm": 2.112987518310547,
      "learning_rate": 9.503491196612939e-06,
      "loss": 0.6544,
      "step": 3660
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.4539880752563477,
      "learning_rate": 9.498741420261109e-06,
      "loss": 0.6452,
      "step": 3670
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.7577303647994995,
      "learning_rate": 9.493970231495836e-06,
      "loss": 0.6705,
      "step": 3680
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.4653581380844116,
      "learning_rate": 9.48917765302629e-06,
      "loss": 0.6759,
      "step": 3690
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.4176025390625,
      "learning_rate": 9.484363707663443e-06,
      "loss": 0.6525,
      "step": 3700
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.6279628276824951,
      "learning_rate": 9.479528418319968e-06,
      "loss": 0.6493,
      "step": 3710
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.9044269323349,
      "learning_rate": 9.474671808010126e-06,
      "loss": 0.6398,
      "step": 3720
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.878974199295044,
      "learning_rate": 9.469793899849663e-06,
      "loss": 0.6533,
      "step": 3730
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.6365597248077393,
      "learning_rate": 9.464894717055686e-06,
      "loss": 0.6601,
      "step": 3740
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.9638291597366333,
      "learning_rate": 9.459974282946572e-06,
      "loss": 0.6496,
      "step": 3750
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.6757596731185913,
      "learning_rate": 9.45503262094184e-06,
      "loss": 0.6566,
      "step": 3760
    },
    {
      "epoch": 0.25,
      "grad_norm": 2.156317949295044,
      "learning_rate": 9.45006975456205e-06,
      "loss": 0.6722,
      "step": 3770
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.5976033210754395,
      "learning_rate": 9.445085707428683e-06,
      "loss": 0.6596,
      "step": 3780
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.6568489074707031,
      "learning_rate": 9.440080503264038e-06,
      "loss": 0.6412,
      "step": 3790
    },
    {
      "epoch": 0.25,
      "grad_norm": 2.008394479751587,
      "learning_rate": 9.43505416589111e-06,
      "loss": 0.6643,
      "step": 3800
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.5924795866012573,
      "learning_rate": 9.430006719233483e-06,
      "loss": 0.6374,
      "step": 3810
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.8586095571517944,
      "learning_rate": 9.42493818731521e-06,
      "loss": 0.6545,
      "step": 3820
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.717989444732666,
      "learning_rate": 9.419848594260708e-06,
      "loss": 0.6411,
      "step": 3830
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.6271024942398071,
      "learning_rate": 9.414737964294636e-06,
      "loss": 0.6737,
      "step": 3840
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.7664674520492554,
      "learning_rate": 9.409606321741776e-06,
      "loss": 0.6635,
      "step": 3850
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.7622829675674438,
      "learning_rate": 9.40445369102693e-06,
      "loss": 0.673,
      "step": 3860
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.5368982553482056,
      "learning_rate": 9.399280096674788e-06,
      "loss": 0.6383,
      "step": 3870
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.9596257209777832,
      "learning_rate": 9.394085563309827e-06,
      "loss": 0.6818,
      "step": 3880
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.8420953750610352,
      "learning_rate": 9.388870115656185e-06,
      "loss": 0.6641,
      "step": 3890
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.407446026802063,
      "learning_rate": 9.38363377853754e-06,
      "loss": 0.6591,
      "step": 3900
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.393359661102295,
      "learning_rate": 9.378376576876999e-06,
      "loss": 0.6496,
      "step": 3910
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.487106442451477,
      "learning_rate": 9.37309853569698e-06,
      "loss": 0.6501,
      "step": 3920
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.5764119625091553,
      "learning_rate": 9.367799680119085e-06,
      "loss": 0.6609,
      "step": 3930
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.7985124588012695,
      "learning_rate": 9.362480035363987e-06,
      "loss": 0.6597,
      "step": 3940
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.5843743085861206,
      "learning_rate": 9.357139626751308e-06,
      "loss": 0.662,
      "step": 3950
    },
    {
      "epoch": 0.26,
      "grad_norm": 2.2033445835113525,
      "learning_rate": 9.351778479699499e-06,
      "loss": 0.6505,
      "step": 3960
    },
    {
      "epoch": 0.26,
      "grad_norm": 2.089184284210205,
      "learning_rate": 9.34639661972572e-06,
      "loss": 0.6703,
      "step": 3970
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.5206127166748047,
      "learning_rate": 9.340994072445713e-06,
      "loss": 0.6676,
      "step": 3980
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.3185709714889526,
      "learning_rate": 9.335570863573687e-06,
      "loss": 0.6517,
      "step": 3990
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.6669631004333496,
      "learning_rate": 9.330127018922195e-06,
      "loss": 0.6785,
      "step": 4000
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.6120712757110596,
      "learning_rate": 9.324662564402004e-06,
      "loss": 0.667,
      "step": 4010
    },
    {
      "epoch": 0.27,
      "grad_norm": 2.04732608795166,
      "learning_rate": 9.31917752602198e-06,
      "loss": 0.6623,
      "step": 4020
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.6198598146438599,
      "learning_rate": 9.31367192988896e-06,
      "loss": 0.6431,
      "step": 4030
    },
    {
      "epoch": 0.27,
      "grad_norm": 2.13211727142334,
      "learning_rate": 9.30814580220763e-06,
      "loss": 0.6589,
      "step": 4040
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.837581753730774,
      "learning_rate": 9.302599169280395e-06,
      "loss": 0.6542,
      "step": 4050
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.8382261991500854,
      "learning_rate": 9.297032057507264e-06,
      "loss": 0.6527,
      "step": 4060
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.9884930849075317,
      "learning_rate": 9.291444493385712e-06,
      "loss": 0.6681,
      "step": 4070
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.6666454076766968,
      "learning_rate": 9.285836503510562e-06,
      "loss": 0.6809,
      "step": 4080
    },
    {
      "epoch": 0.27,
      "grad_norm": 2.0016543865203857,
      "learning_rate": 9.280208114573859e-06,
      "loss": 0.6808,
      "step": 4090
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.6018117666244507,
      "learning_rate": 9.274559353364734e-06,
      "loss": 0.6751,
      "step": 4100
    },
    {
      "epoch": 0.27,
      "grad_norm": 2.8057656288146973,
      "learning_rate": 9.268890246769288e-06,
      "loss": 0.6618,
      "step": 4110
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.394476294517517,
      "learning_rate": 9.263200821770462e-06,
      "loss": 0.6596,
      "step": 4120
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.8551517724990845,
      "learning_rate": 9.257491105447895e-06,
      "loss": 0.6557,
      "step": 4130
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.7693849802017212,
      "learning_rate": 9.251761124977816e-06,
      "loss": 0.6389,
      "step": 4140
    },
    {
      "epoch": 0.27,
      "grad_norm": 2.032569169998169,
      "learning_rate": 9.246010907632894e-06,
      "loss": 0.653,
      "step": 4150
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.4892268180847168,
      "learning_rate": 9.24024048078213e-06,
      "loss": 0.657,
      "step": 4160
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.8956682682037354,
      "learning_rate": 9.234449871890708e-06,
      "loss": 0.6585,
      "step": 4170
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.9287517070770264,
      "learning_rate": 9.228639108519867e-06,
      "loss": 0.6556,
      "step": 4180
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.6487663984298706,
      "learning_rate": 9.222808218326784e-06,
      "loss": 0.6572,
      "step": 4190
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.7571004629135132,
      "learning_rate": 9.21695722906443e-06,
      "loss": 0.6681,
      "step": 4200
    },
    {
      "epoch": 0.28,
      "grad_norm": 2.26200532913208,
      "learning_rate": 9.211086168581433e-06,
      "loss": 0.6369,
      "step": 4210
    },
    {
      "epoch": 0.28,
      "grad_norm": 2.210916042327881,
      "learning_rate": 9.205195064821964e-06,
      "loss": 0.6748,
      "step": 4220
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.523188829421997,
      "learning_rate": 9.199283945825582e-06,
      "loss": 0.6595,
      "step": 4230
    },
    {
      "epoch": 0.28,
      "grad_norm": 2.346522092819214,
      "learning_rate": 9.193352839727122e-06,
      "loss": 0.6908,
      "step": 4240
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.611079454421997,
      "learning_rate": 9.18740177475654e-06,
      "loss": 0.6476,
      "step": 4250
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.8673412799835205,
      "learning_rate": 9.181430779238799e-06,
      "loss": 0.6544,
      "step": 4260
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.6414918899536133,
      "learning_rate": 9.175439881593716e-06,
      "loss": 0.6579,
      "step": 4270
    },
    {
      "epoch": 0.28,
      "grad_norm": 2.150317668914795,
      "learning_rate": 9.169429110335842e-06,
      "loss": 0.6687,
      "step": 4280
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.9575879573822021,
      "learning_rate": 9.163398494074314e-06,
      "loss": 0.6423,
      "step": 4290
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.7684667110443115,
      "learning_rate": 9.157348061512728e-06,
      "loss": 0.6457,
      "step": 4300
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.60320246219635,
      "learning_rate": 9.151277841448993e-06,
      "loss": 0.632,
      "step": 4310
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.604935884475708,
      "learning_rate": 9.145187862775208e-06,
      "loss": 0.6633,
      "step": 4320
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.5652973651885986,
      "learning_rate": 9.139078154477512e-06,
      "loss": 0.6496,
      "step": 4330
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.8579405546188354,
      "learning_rate": 9.132948745635943e-06,
      "loss": 0.652,
      "step": 4340
    },
    {
      "epoch": 0.29,
      "grad_norm": 2.2190017700195312,
      "learning_rate": 9.126799665424319e-06,
      "loss": 0.6789,
      "step": 4350
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.978559970855713,
      "learning_rate": 9.120630943110078e-06,
      "loss": 0.6849,
      "step": 4360
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.7971248626708984,
      "learning_rate": 9.114442608054153e-06,
      "loss": 0.6637,
      "step": 4370
    },
    {
      "epoch": 0.29,
      "grad_norm": 2.0373167991638184,
      "learning_rate": 9.10823468971082e-06,
      "loss": 0.6718,
      "step": 4380
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.748665690422058,
      "learning_rate": 9.102007217627568e-06,
      "loss": 0.6666,
      "step": 4390
    },
    {
      "epoch": 0.29,
      "grad_norm": 2.1095681190490723,
      "learning_rate": 9.09576022144496e-06,
      "loss": 0.6676,
      "step": 4400
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.4815317392349243,
      "learning_rate": 9.089493730896478e-06,
      "loss": 0.6414,
      "step": 4410
    },
    {
      "epoch": 0.29,
      "grad_norm": 10.00068187713623,
      "learning_rate": 9.083207775808395e-06,
      "loss": 0.6533,
      "step": 4420
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.62544584274292,
      "learning_rate": 9.076902386099628e-06,
      "loss": 0.6549,
      "step": 4430
    },
    {
      "epoch": 0.29,
      "grad_norm": 2.158773183822632,
      "learning_rate": 9.070577591781598e-06,
      "loss": 0.6627,
      "step": 4440
    },
    {
      "epoch": 0.29,
      "grad_norm": 2.012834072113037,
      "learning_rate": 9.064233422958078e-06,
      "loss": 0.6542,
      "step": 4450
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.6546082496643066,
      "learning_rate": 9.057869909825062e-06,
      "loss": 0.6541,
      "step": 4460
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.8351281881332397,
      "learning_rate": 9.051487082670618e-06,
      "loss": 0.6714,
      "step": 4470
    },
    {
      "epoch": 0.3,
      "grad_norm": 2.1694746017456055,
      "learning_rate": 9.045084971874738e-06,
      "loss": 0.6666,
      "step": 4480
    },
    {
      "epoch": 0.3,
      "grad_norm": 2.0474767684936523,
      "learning_rate": 9.038663607909198e-06,
      "loss": 0.6649,
      "step": 4490
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.6256054639816284,
      "learning_rate": 9.032223021337415e-06,
      "loss": 0.6588,
      "step": 4500
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.5998213291168213,
      "learning_rate": 9.025763242814291e-06,
      "loss": 0.6857,
      "step": 4510
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.970272421836853,
      "learning_rate": 9.019284303086086e-06,
      "loss": 0.6657,
      "step": 4520
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.555245041847229,
      "learning_rate": 9.012786232990256e-06,
      "loss": 0.651,
      "step": 4530
    },
    {
      "epoch": 0.3,
      "grad_norm": 2.0598926544189453,
      "learning_rate": 9.006269063455305e-06,
      "loss": 0.6689,
      "step": 4540
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.8089969158172607,
      "learning_rate": 8.999732825500649e-06,
      "loss": 0.6528,
      "step": 4550
    },
    {
      "epoch": 0.3,
      "grad_norm": 2.0560412406921387,
      "learning_rate": 8.993177550236464e-06,
      "loss": 0.6599,
      "step": 4560
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.9101848602294922,
      "learning_rate": 8.986603268863536e-06,
      "loss": 0.6429,
      "step": 4570
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.8303155899047852,
      "learning_rate": 8.98001001267311e-06,
      "loss": 0.6315,
      "step": 4580
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.8667141199111938,
      "learning_rate": 8.97339781304675e-06,
      "loss": 0.6405,
      "step": 4590
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.8446400165557861,
      "learning_rate": 8.966766701456177e-06,
      "loss": 0.6378,
      "step": 4600
    },
    {
      "epoch": 0.31,
      "grad_norm": 2.6912033557891846,
      "learning_rate": 8.960116709463131e-06,
      "loss": 0.6727,
      "step": 4610
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.9695560932159424,
      "learning_rate": 8.953447868719218e-06,
      "loss": 0.6554,
      "step": 4620
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.7779700756072998,
      "learning_rate": 8.94676021096575e-06,
      "loss": 0.674,
      "step": 4630
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.5401631593704224,
      "learning_rate": 8.94005376803361e-06,
      "loss": 0.6557,
      "step": 4640
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.6267428398132324,
      "learning_rate": 8.933328571843086e-06,
      "loss": 0.65,
      "step": 4650
    },
    {
      "epoch": 0.31,
      "grad_norm": 2.1260504722595215,
      "learning_rate": 8.926584654403725e-06,
      "loss": 0.6648,
      "step": 4660
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.7050879001617432,
      "learning_rate": 8.919822047814184e-06,
      "loss": 0.6471,
      "step": 4670
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.5733059644699097,
      "learning_rate": 8.91304078426207e-06,
      "loss": 0.6494,
      "step": 4680
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.788745403289795,
      "learning_rate": 8.906240896023794e-06,
      "loss": 0.6619,
      "step": 4690
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.6663099527359009,
      "learning_rate": 8.899422415464409e-06,
      "loss": 0.656,
      "step": 4700
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.934832215309143,
      "learning_rate": 8.892585375037469e-06,
      "loss": 0.6489,
      "step": 4710
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.9886648654937744,
      "learning_rate": 8.885729807284855e-06,
      "loss": 0.6524,
      "step": 4720
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.718221664428711,
      "learning_rate": 8.878855744836643e-06,
      "loss": 0.657,
      "step": 4730
    },
    {
      "epoch": 0.31,
      "grad_norm": 2.0010931491851807,
      "learning_rate": 8.871963220410929e-06,
      "loss": 0.6532,
      "step": 4740
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.7940354347229004,
      "learning_rate": 8.865052266813686e-06,
      "loss": 0.6639,
      "step": 4750
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.7344415187835693,
      "learning_rate": 8.858122916938601e-06,
      "loss": 0.6588,
      "step": 4760
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.5673871040344238,
      "learning_rate": 8.851175203766922e-06,
      "loss": 0.6526,
      "step": 4770
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.609823226928711,
      "learning_rate": 8.844209160367298e-06,
      "loss": 0.6836,
      "step": 4780
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.8896527290344238,
      "learning_rate": 8.837224819895627e-06,
      "loss": 0.6739,
      "step": 4790
    },
    {
      "epoch": 0.32,
      "grad_norm": 2.3576793670654297,
      "learning_rate": 8.83022221559489e-06,
      "loss": 0.6755,
      "step": 4800
    },
    {
      "epoch": 0.32,
      "eval_loss": 0.7543163895606995,
      "eval_runtime": 134.2076,
      "eval_samples_per_second": 81.963,
      "eval_steps_per_second": 10.245,
      "step": 4800
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.6949530839920044,
      "learning_rate": 8.823201380795003e-06,
      "loss": 0.6621,
      "step": 4810
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.8176947832107544,
      "learning_rate": 8.816162348912644e-06,
      "loss": 0.6603,
      "step": 4820
    },
    {
      "epoch": 0.32,
      "grad_norm": 2.184191942214966,
      "learning_rate": 8.809105153451113e-06,
      "loss": 0.6721,
      "step": 4830
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.877064824104309,
      "learning_rate": 8.802029828000157e-06,
      "loss": 0.6696,
      "step": 4840
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.3812588453292847,
      "learning_rate": 8.79493640623581e-06,
      "loss": 0.6698,
      "step": 4850
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.5951484441757202,
      "learning_rate": 8.78782492192025e-06,
      "loss": 0.6656,
      "step": 4860
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.7040786743164062,
      "learning_rate": 8.780695408901613e-06,
      "loss": 0.6648,
      "step": 4870
    },
    {
      "epoch": 0.32,
      "grad_norm": 2.1103320121765137,
      "learning_rate": 8.773547901113862e-06,
      "loss": 0.6809,
      "step": 4880
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.5602185726165771,
      "learning_rate": 8.766382432576589e-06,
      "loss": 0.6597,
      "step": 4890
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.905117392539978,
      "learning_rate": 8.759199037394888e-06,
      "loss": 0.6619,
      "step": 4900
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.5752969980239868,
      "learning_rate": 8.75199774975917e-06,
      "loss": 0.6549,
      "step": 4910
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.9230555295944214,
      "learning_rate": 8.744778603945013e-06,
      "loss": 0.6364,
      "step": 4920
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.3653078079223633,
      "learning_rate": 8.737541634312985e-06,
      "loss": 0.6222,
      "step": 4930
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.6335278749465942,
      "learning_rate": 8.730286875308498e-06,
      "loss": 0.665,
      "step": 4940
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.7848576307296753,
      "learning_rate": 8.723014361461633e-06,
      "loss": 0.6575,
      "step": 4950
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.5796133279800415,
      "learning_rate": 8.715724127386971e-06,
      "loss": 0.6556,
      "step": 4960
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.7556233406066895,
      "learning_rate": 8.708416207783447e-06,
      "loss": 0.6569,
      "step": 4970
    },
    {
      "epoch": 0.33,
      "grad_norm": 2.42543888092041,
      "learning_rate": 8.701090637434161e-06,
      "loss": 0.6654,
      "step": 4980
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.9830849170684814,
      "learning_rate": 8.693747451206231e-06,
      "loss": 0.6418,
      "step": 4990
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.6270414590835571,
      "learning_rate": 8.68638668405062e-06,
      "loss": 0.6666,
      "step": 5000
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.8726800680160522,
      "learning_rate": 8.679008371001969e-06,
      "loss": 0.6598,
      "step": 5010
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.2677507400512695,
      "learning_rate": 8.671612547178428e-06,
      "loss": 0.6625,
      "step": 5020
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.7537394762039185,
      "learning_rate": 8.664199247781497e-06,
      "loss": 0.6797,
      "step": 5030
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.82877779006958,
      "learning_rate": 8.656768508095853e-06,
      "loss": 0.6734,
      "step": 5040
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.8775991201400757,
      "learning_rate": 8.649320363489178e-06,
      "loss": 0.6722,
      "step": 5050
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.7584172487258911,
      "learning_rate": 8.641854849412002e-06,
      "loss": 0.6744,
      "step": 5060
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.6870256662368774,
      "learning_rate": 8.634372001397521e-06,
      "loss": 0.6596,
      "step": 5070
    },
    {
      "epoch": 0.34,
      "grad_norm": 2.059516191482544,
      "learning_rate": 8.626871855061438e-06,
      "loss": 0.6584,
      "step": 5080
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.5890916585922241,
      "learning_rate": 8.61935444610179e-06,
      "loss": 0.6461,
      "step": 5090
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.5515724420547485,
      "learning_rate": 8.611819810298778e-06,
      "loss": 0.6651,
      "step": 5100
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.4988518953323364,
      "learning_rate": 8.604267983514595e-06,
      "loss": 0.6616,
      "step": 5110
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.6072263717651367,
      "learning_rate": 8.596699001693257e-06,
      "loss": 0.6694,
      "step": 5120
    },
    {
      "epoch": 0.34,
      "grad_norm": 2.2334463596343994,
      "learning_rate": 8.589112900860432e-06,
      "loss": 0.6647,
      "step": 5130
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.476638674736023,
      "learning_rate": 8.581509717123272e-06,
      "loss": 0.6542,
      "step": 5140
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.5825482606887817,
      "learning_rate": 8.573889486670233e-06,
      "loss": 0.6628,
      "step": 5150
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.387061357498169,
      "learning_rate": 8.56625224577091e-06,
      "loss": 0.646,
      "step": 5160
    },
    {
      "epoch": 0.34,
      "grad_norm": 2.2527670860290527,
      "learning_rate": 8.558598030775857e-06,
      "loss": 0.6675,
      "step": 5170
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.8437511920928955,
      "learning_rate": 8.550926878116428e-06,
      "loss": 0.646,
      "step": 5180
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.7316181659698486,
      "learning_rate": 8.543238824304585e-06,
      "loss": 0.6828,
      "step": 5190
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.5785106420516968,
      "learning_rate": 8.535533905932739e-06,
      "loss": 0.6675,
      "step": 5200
    },
    {
      "epoch": 0.34,
      "grad_norm": 2.5922186374664307,
      "learning_rate": 8.527812159673567e-06,
      "loss": 0.675,
      "step": 5210
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.8569538593292236,
      "learning_rate": 8.520073622279844e-06,
      "loss": 0.6397,
      "step": 5220
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.9984098672866821,
      "learning_rate": 8.51231833058426e-06,
      "loss": 0.6806,
      "step": 5230
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.8538262844085693,
      "learning_rate": 8.504546321499255e-06,
      "loss": 0.654,
      "step": 5240
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.672841191291809,
      "learning_rate": 8.496757632016836e-06,
      "loss": 0.6429,
      "step": 5250
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.5475361347198486,
      "learning_rate": 8.488952299208402e-06,
      "loss": 0.6574,
      "step": 5260
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.6212105751037598,
      "learning_rate": 8.481130360224567e-06,
      "loss": 0.6504,
      "step": 5270
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.7608546018600464,
      "learning_rate": 8.473291852294986e-06,
      "loss": 0.6709,
      "step": 5280
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.720782995223999,
      "learning_rate": 8.465436812728181e-06,
      "loss": 0.6558,
      "step": 5290
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.6840670108795166,
      "learning_rate": 8.457565278911349e-06,
      "loss": 0.6542,
      "step": 5300
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.7344157695770264,
      "learning_rate": 8.449677288310198e-06,
      "loss": 0.6694,
      "step": 5310
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.6819674968719482,
      "learning_rate": 8.44177287846877e-06,
      "loss": 0.6453,
      "step": 5320
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.9622447490692139,
      "learning_rate": 8.433852087009251e-06,
      "loss": 0.6861,
      "step": 5330
    },
    {
      "epoch": 0.35,
      "grad_norm": 2.322389602661133,
      "learning_rate": 8.425914951631796e-06,
      "loss": 0.6838,
      "step": 5340
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.4469890594482422,
      "learning_rate": 8.417961510114357e-06,
      "loss": 0.6752,
      "step": 5350
    },
    {
      "epoch": 0.35,
      "grad_norm": 2.1829028129577637,
      "learning_rate": 8.409991800312493e-06,
      "loss": 0.6567,
      "step": 5360
    },
    {
      "epoch": 0.36,
      "grad_norm": 2.2404589653015137,
      "learning_rate": 8.402005860159197e-06,
      "loss": 0.6509,
      "step": 5370
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.9056888818740845,
      "learning_rate": 8.39400372766471e-06,
      "loss": 0.6588,
      "step": 5380
    },
    {
      "epoch": 0.36,
      "grad_norm": 3.867516040802002,
      "learning_rate": 8.385985440916344e-06,
      "loss": 0.6476,
      "step": 5390
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.9217302799224854,
      "learning_rate": 8.377951038078303e-06,
      "loss": 0.6568,
      "step": 5400
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.806114912033081,
      "learning_rate": 8.36990055739149e-06,
      "loss": 0.6729,
      "step": 5410
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.9048370122909546,
      "learning_rate": 8.36183403717334e-06,
      "loss": 0.6783,
      "step": 5420
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.9079678058624268,
      "learning_rate": 8.353751515817629e-06,
      "loss": 0.6659,
      "step": 5430
    },
    {
      "epoch": 0.36,
      "grad_norm": 2.0406830310821533,
      "learning_rate": 8.345653031794292e-06,
      "loss": 0.6696,
      "step": 5440
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.7874845266342163,
      "learning_rate": 8.337538623649237e-06,
      "loss": 0.6506,
      "step": 5450
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.866437554359436,
      "learning_rate": 8.329408330004172e-06,
      "loss": 0.6631,
      "step": 5460
    },
    {
      "epoch": 0.36,
      "grad_norm": 2.0198395252227783,
      "learning_rate": 8.32126218955641e-06,
      "loss": 0.656,
      "step": 5470
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.6846990585327148,
      "learning_rate": 8.313100241078689e-06,
      "loss": 0.6639,
      "step": 5480
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.445802927017212,
      "learning_rate": 8.304922523418988e-06,
      "loss": 0.6772,
      "step": 5490
    },
    {
      "epoch": 0.36,
      "grad_norm": 2.063717842102051,
      "learning_rate": 8.296729075500345e-06,
      "loss": 0.6666,
      "step": 5500
    },
    {
      "epoch": 0.36,
      "grad_norm": 2.563687324523926,
      "learning_rate": 8.288519936320664e-06,
      "loss": 0.6528,
      "step": 5510
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.9258893728256226,
      "learning_rate": 8.280295144952537e-06,
      "loss": 0.6656,
      "step": 5520
    },
    {
      "epoch": 0.37,
      "grad_norm": 2.109924077987671,
      "learning_rate": 8.272054740543053e-06,
      "loss": 0.6509,
      "step": 5530
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.950067400932312,
      "learning_rate": 8.263798762313613e-06,
      "loss": 0.6524,
      "step": 5540
    },
    {
      "epoch": 0.37,
      "grad_norm": 2.1017370223999023,
      "learning_rate": 8.255527249559747e-06,
      "loss": 0.662,
      "step": 5550
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.7278681993484497,
      "learning_rate": 8.247240241650918e-06,
      "loss": 0.6602,
      "step": 5560
    },
    {
      "epoch": 0.37,
      "grad_norm": 3.6589279174804688,
      "learning_rate": 8.23893777803035e-06,
      "loss": 0.6454,
      "step": 5570
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.8462167978286743,
      "learning_rate": 8.23061989821482e-06,
      "loss": 0.6468,
      "step": 5580
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.9224498271942139,
      "learning_rate": 8.222286641794488e-06,
      "loss": 0.6608,
      "step": 5590
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.796586513519287,
      "learning_rate": 8.213938048432697e-06,
      "loss": 0.6582,
      "step": 5600
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.8218111991882324,
      "learning_rate": 8.205574157865791e-06,
      "loss": 0.665,
      "step": 5610
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.6219843626022339,
      "learning_rate": 8.197195009902924e-06,
      "loss": 0.643,
      "step": 5620
    },
    {
      "epoch": 0.37,
      "grad_norm": 2.4031009674072266,
      "learning_rate": 8.188800644425867e-06,
      "loss": 0.6543,
      "step": 5630
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.7068158388137817,
      "learning_rate": 8.18039110138882e-06,
      "loss": 0.6671,
      "step": 5640
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.9731248617172241,
      "learning_rate": 8.171966420818227e-06,
      "loss": 0.6717,
      "step": 5650
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.647336721420288,
      "learning_rate": 8.163526642812582e-06,
      "loss": 0.6717,
      "step": 5660
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.3182628154754639,
      "learning_rate": 8.15507180754223e-06,
      "loss": 0.6597,
      "step": 5670
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.8658872842788696,
      "learning_rate": 8.146601955249187e-06,
      "loss": 0.6751,
      "step": 5680
    },
    {
      "epoch": 0.38,
      "grad_norm": 2.511075019836426,
      "learning_rate": 8.138117126246951e-06,
      "loss": 0.6608,
      "step": 5690
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.5444680452346802,
      "learning_rate": 8.129617360920297e-06,
      "loss": 0.6621,
      "step": 5700
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.8167874813079834,
      "learning_rate": 8.12110269972509e-06,
      "loss": 0.6596,
      "step": 5710
    },
    {
      "epoch": 0.38,
      "grad_norm": 2.235473155975342,
      "learning_rate": 8.112573183188099e-06,
      "loss": 0.6696,
      "step": 5720
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.5982747077941895,
      "learning_rate": 8.104028851906797e-06,
      "loss": 0.6697,
      "step": 5730
    },
    {
      "epoch": 0.38,
      "grad_norm": 2.0763542652130127,
      "learning_rate": 8.095469746549172e-06,
      "loss": 0.6757,
      "step": 5740
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.66940176486969,
      "learning_rate": 8.086895907853526e-06,
      "loss": 0.676,
      "step": 5750
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.759155035018921,
      "learning_rate": 8.078307376628292e-06,
      "loss": 0.6539,
      "step": 5760
    },
    {
      "epoch": 0.38,
      "grad_norm": 2.2886273860931396,
      "learning_rate": 8.069704193751834e-06,
      "loss": 0.6575,
      "step": 5770
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.7834458351135254,
      "learning_rate": 8.061086400172247e-06,
      "loss": 0.6827,
      "step": 5780
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.9700424671173096,
      "learning_rate": 8.052454036907174e-06,
      "loss": 0.6629,
      "step": 5790
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.9198191165924072,
      "learning_rate": 8.043807145043604e-06,
      "loss": 0.6695,
      "step": 5800
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.5736877918243408,
      "learning_rate": 8.035145765737671e-06,
      "loss": 0.6676,
      "step": 5810
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.728734016418457,
      "learning_rate": 8.026469940214471e-06,
      "loss": 0.6684,
      "step": 5820
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.5845164060592651,
      "learning_rate": 8.017779709767857e-06,
      "loss": 0.6597,
      "step": 5830
    },
    {
      "epoch": 0.39,
      "grad_norm": 2.1340720653533936,
      "learning_rate": 8.009075115760243e-06,
      "loss": 0.6703,
      "step": 5840
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.7079795598983765,
      "learning_rate": 8.000356199622406e-06,
      "loss": 0.6652,
      "step": 5850
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.5601955652236938,
      "learning_rate": 7.991623002853296e-06,
      "loss": 0.6583,
      "step": 5860
    },
    {
      "epoch": 0.39,
      "grad_norm": 2.0047013759613037,
      "learning_rate": 7.982875567019833e-06,
      "loss": 0.6733,
      "step": 5870
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.8969409465789795,
      "learning_rate": 7.974113933756708e-06,
      "loss": 0.678,
      "step": 5880
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.5502653121948242,
      "learning_rate": 7.965338144766186e-06,
      "loss": 0.6474,
      "step": 5890
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.8099123239517212,
      "learning_rate": 7.956548241817914e-06,
      "loss": 0.6564,
      "step": 5900
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.6473009586334229,
      "learning_rate": 7.947744266748707e-06,
      "loss": 0.6785,
      "step": 5910
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.9546464681625366,
      "learning_rate": 7.938926261462366e-06,
      "loss": 0.6852,
      "step": 5920
    },
    {
      "epoch": 0.39,
      "grad_norm": 2.04304838180542,
      "learning_rate": 7.93009426792947e-06,
      "loss": 0.6521,
      "step": 5930
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.7775088548660278,
      "learning_rate": 7.921248328187174e-06,
      "loss": 0.6755,
      "step": 5940
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.7072761058807373,
      "learning_rate": 7.912388484339012e-06,
      "loss": 0.6681,
      "step": 5950
    },
    {
      "epoch": 0.39,
      "grad_norm": 2.046266555786133,
      "learning_rate": 7.903514778554699e-06,
      "loss": 0.6622,
      "step": 5960
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.9110488891601562,
      "learning_rate": 7.89462725306993e-06,
      "loss": 0.6676,
      "step": 5970
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.9375412464141846,
      "learning_rate": 7.88572595018617e-06,
      "loss": 0.6635,
      "step": 5980
    },
    {
      "epoch": 0.4,
      "grad_norm": 2.0740785598754883,
      "learning_rate": 7.876810912270462e-06,
      "loss": 0.6592,
      "step": 5990
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.6654475927352905,
      "learning_rate": 7.86788218175523e-06,
      "loss": 0.6598,
      "step": 6000
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.9562811851501465,
      "learning_rate": 7.858939801138061e-06,
      "loss": 0.6831,
      "step": 6010
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.9873039722442627,
      "learning_rate": 7.849983812981516e-06,
      "loss": 0.6548,
      "step": 6020
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.7364007234573364,
      "learning_rate": 7.84101425991292e-06,
      "loss": 0.6452,
      "step": 6030
    },
    {
      "epoch": 0.4,
      "grad_norm": 2.040637731552124,
      "learning_rate": 7.832031184624165e-06,
      "loss": 0.6533,
      "step": 6040
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.9943041801452637,
      "learning_rate": 7.823034629871503e-06,
      "loss": 0.6601,
      "step": 6050
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.8325495719909668,
      "learning_rate": 7.814024638475344e-06,
      "loss": 0.6666,
      "step": 6060
    },
    {
      "epoch": 0.4,
      "grad_norm": 2.264528274536133,
      "learning_rate": 7.80500125332005e-06,
      "loss": 0.645,
      "step": 6070
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.8350125551223755,
      "learning_rate": 7.795964517353734e-06,
      "loss": 0.681,
      "step": 6080
    },
    {
      "epoch": 0.4,
      "grad_norm": 5.209187030792236,
      "learning_rate": 7.786914473588057e-06,
      "loss": 0.6385,
      "step": 6090
    },
    {
      "epoch": 0.4,
      "grad_norm": 2.010577917098999,
      "learning_rate": 7.777851165098012e-06,
      "loss": 0.6375,
      "step": 6100
    },
    {
      "epoch": 0.4,
      "grad_norm": 2.004896402359009,
      "learning_rate": 7.768774635021737e-06,
      "loss": 0.6817,
      "step": 6110
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.7268619537353516,
      "learning_rate": 7.759684926560292e-06,
      "loss": 0.665,
      "step": 6120
    },
    {
      "epoch": 0.41,
      "grad_norm": 2.00075101852417,
      "learning_rate": 7.750582082977468e-06,
      "loss": 0.6602,
      "step": 6130
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.9209833145141602,
      "learning_rate": 7.74146614759957e-06,
      "loss": 0.6795,
      "step": 6140
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.7252877950668335,
      "learning_rate": 7.732337163815218e-06,
      "loss": 0.6608,
      "step": 6150
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.6505799293518066,
      "learning_rate": 7.723195175075136e-06,
      "loss": 0.6519,
      "step": 6160
    },
    {
      "epoch": 0.41,
      "grad_norm": 2.175870656967163,
      "learning_rate": 7.714040224891949e-06,
      "loss": 0.6877,
      "step": 6170
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.762647271156311,
      "learning_rate": 7.704872356839971e-06,
      "loss": 0.6582,
      "step": 6180
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.860929250717163,
      "learning_rate": 7.695691614555002e-06,
      "loss": 0.6719,
      "step": 6190
    },
    {
      "epoch": 0.41,
      "grad_norm": 6.883901596069336,
      "learning_rate": 7.686498041734121e-06,
      "loss": 0.6604,
      "step": 6200
    },
    {
      "epoch": 0.41,
      "grad_norm": 2.181346893310547,
      "learning_rate": 7.67729168213547e-06,
      "loss": 0.672,
      "step": 6210
    },
    {
      "epoch": 0.41,
      "grad_norm": 2.26829195022583,
      "learning_rate": 7.66807257957806e-06,
      "loss": 0.671,
      "step": 6220
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.7630722522735596,
      "learning_rate": 7.658840777941543e-06,
      "loss": 0.6623,
      "step": 6230
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.9417176246643066,
      "learning_rate": 7.649596321166024e-06,
      "loss": 0.6476,
      "step": 6240
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.660746693611145,
      "learning_rate": 7.64033925325184e-06,
      "loss": 0.6778,
      "step": 6250
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.5244559049606323,
      "learning_rate": 7.631069618259347e-06,
      "loss": 0.6471,
      "step": 6260
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.8127204179763794,
      "learning_rate": 7.621787460308723e-06,
      "loss": 0.6633,
      "step": 6270
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.7429029941558838,
      "learning_rate": 7.612492823579744e-06,
      "loss": 0.674,
      "step": 6280
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.903221845626831,
      "learning_rate": 7.603185752311587e-06,
      "loss": 0.6593,
      "step": 6290
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.6729519367218018,
      "learning_rate": 7.593866290802608e-06,
      "loss": 0.664,
      "step": 6300
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.7710927724838257,
      "learning_rate": 7.584534483410137e-06,
      "loss": 0.6639,
      "step": 6310
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.5772103071212769,
      "learning_rate": 7.575190374550272e-06,
      "loss": 0.6746,
      "step": 6320
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.7170156240463257,
      "learning_rate": 7.565834008697652e-06,
      "loss": 0.6721,
      "step": 6330
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.7289538383483887,
      "learning_rate": 7.55646543038526e-06,
      "loss": 0.6507,
      "step": 6340
    },
    {
      "epoch": 0.42,
      "grad_norm": 2.6091270446777344,
      "learning_rate": 7.54708468420421e-06,
      "loss": 0.6584,
      "step": 6350
    },
    {
      "epoch": 0.42,
      "grad_norm": 2.401001214981079,
      "learning_rate": 7.537691814803522e-06,
      "loss": 0.6503,
      "step": 6360
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.2781227827072144,
      "learning_rate": 7.528286866889924e-06,
      "loss": 0.6545,
      "step": 6370
    },
    {
      "epoch": 0.42,
      "grad_norm": 2.448362112045288,
      "learning_rate": 7.518869885227632e-06,
      "loss": 0.6722,
      "step": 6380
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.9541414976119995,
      "learning_rate": 7.50944091463814e-06,
      "loss": 0.6553,
      "step": 6390
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.3395729064941406,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.6601,
      "step": 6400
    },
    {
      "epoch": 0.42,
      "eval_loss": 0.7648976445198059,
      "eval_runtime": 134.0186,
      "eval_samples_per_second": 82.078,
      "eval_steps_per_second": 10.26,
      "step": 6400
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.6933283805847168,
      "learning_rate": 7.4905471862486215e-06,
      "loss": 0.6627,
      "step": 6410
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.5258747339248657,
      "learning_rate": 7.4810825183760425e-06,
      "loss": 0.6685,
      "step": 6420
    },
    {
      "epoch": 0.43,
      "grad_norm": 2.085606336593628,
      "learning_rate": 7.471606041430724e-06,
      "loss": 0.6646,
      "step": 6430
    },
    {
      "epoch": 0.43,
      "grad_norm": 2.31076979637146,
      "learning_rate": 7.462117800517337e-06,
      "loss": 0.6467,
      "step": 6440
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.5522764921188354,
      "learning_rate": 7.4526178407965396e-06,
      "loss": 0.6559,
      "step": 6450
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.5559918880462646,
      "learning_rate": 7.443106207484776e-06,
      "loss": 0.6486,
      "step": 6460
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.428256630897522,
      "learning_rate": 7.433582945854041e-06,
      "loss": 0.6696,
      "step": 6470
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.6516914367675781,
      "learning_rate": 7.424048101231687e-06,
      "loss": 0.6571,
      "step": 6480
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.6921088695526123,
      "learning_rate": 7.414501719000187e-06,
      "loss": 0.6595,
      "step": 6490
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.556532621383667,
      "learning_rate": 7.404943844596939e-06,
      "loss": 0.657,
      "step": 6500
    },
    {
      "epoch": 0.43,
      "grad_norm": 2.2812130451202393,
      "learning_rate": 7.3953745235140325e-06,
      "loss": 0.677,
      "step": 6510
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.9262441396713257,
      "learning_rate": 7.3857938012980425e-06,
      "loss": 0.6451,
      "step": 6520
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.781899094581604,
      "learning_rate": 7.3762017235498084e-06,
      "loss": 0.6774,
      "step": 6530
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.732818841934204,
      "learning_rate": 7.3665983359242175e-06,
      "loss": 0.6724,
      "step": 6540
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.8356454372406006,
      "learning_rate": 7.3569836841299905e-06,
      "loss": 0.682,
      "step": 6550
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.787606954574585,
      "learning_rate": 7.347357813929455e-06,
      "loss": 0.6719,
      "step": 6560
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.9841190576553345,
      "learning_rate": 7.337720771138343e-06,
      "loss": 0.6547,
      "step": 6570
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.8264681100845337,
      "learning_rate": 7.328072601625558e-06,
      "loss": 0.6545,
      "step": 6580
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.4462283849716187,
      "learning_rate": 7.318413351312965e-06,
      "loss": 0.6714,
      "step": 6590
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.7890430688858032,
      "learning_rate": 7.308743066175172e-06,
      "loss": 0.6563,
      "step": 6600
    },
    {
      "epoch": 0.44,
      "grad_norm": 2.616171360015869,
      "learning_rate": 7.2990617922393e-06,
      "loss": 0.657,
      "step": 6610
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.8816206455230713,
      "learning_rate": 7.289369575584783e-06,
      "loss": 0.6662,
      "step": 6620
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.7342751026153564,
      "learning_rate": 7.279666462343138e-06,
      "loss": 0.674,
      "step": 6630
    },
    {
      "epoch": 0.44,
      "grad_norm": 2.6490211486816406,
      "learning_rate": 7.269952498697734e-06,
      "loss": 0.6869,
      "step": 6640
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.4362220764160156,
      "learning_rate": 7.2602277308836e-06,
      "loss": 0.6465,
      "step": 6650
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.6366841793060303,
      "learning_rate": 7.250492205187176e-06,
      "loss": 0.6526,
      "step": 6660
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.906667709350586,
      "learning_rate": 7.240745967946113e-06,
      "loss": 0.6552,
      "step": 6670
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.9176589250564575,
      "learning_rate": 7.2309890655490446e-06,
      "loss": 0.6732,
      "step": 6680
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.6638537645339966,
      "learning_rate": 7.221221544435364e-06,
      "loss": 0.6552,
      "step": 6690
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.9146836996078491,
      "learning_rate": 7.211443451095007e-06,
      "loss": 0.6516,
      "step": 6700
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.5971161127090454,
      "learning_rate": 7.20165483206823e-06,
      "loss": 0.6429,
      "step": 6710
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.771937370300293,
      "learning_rate": 7.191855733945388e-06,
      "loss": 0.6407,
      "step": 6720
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.5045241117477417,
      "learning_rate": 7.18204620336671e-06,
      "loss": 0.6663,
      "step": 6730
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.8442574739456177,
      "learning_rate": 7.172226287022086e-06,
      "loss": 0.6637,
      "step": 6740
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.5092518329620361,
      "learning_rate": 7.162396031650831e-06,
      "loss": 0.6665,
      "step": 6750
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.6536009311676025,
      "learning_rate": 7.1525554840414765e-06,
      "loss": 0.6544,
      "step": 6760
    },
    {
      "epoch": 0.45,
      "grad_norm": 2.160461902618408,
      "learning_rate": 7.142704691031537e-06,
      "loss": 0.6886,
      "step": 6770
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.5910495519638062,
      "learning_rate": 7.132843699507292e-06,
      "loss": 0.6634,
      "step": 6780
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.7818145751953125,
      "learning_rate": 7.1229725564035665e-06,
      "loss": 0.6761,
      "step": 6790
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.350142478942871,
      "learning_rate": 7.113091308703498e-06,
      "loss": 0.6484,
      "step": 6800
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.6214317083358765,
      "learning_rate": 7.103200003438322e-06,
      "loss": 0.6699,
      "step": 6810
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.9810750484466553,
      "learning_rate": 7.093298687687141e-06,
      "loss": 0.6579,
      "step": 6820
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.8099583387374878,
      "learning_rate": 7.08338740857671e-06,
      "loss": 0.6747,
      "step": 6830
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.903903841972351,
      "learning_rate": 7.073466213281196e-06,
      "loss": 0.6625,
      "step": 6840
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.7253398895263672,
      "learning_rate": 7.063535149021974e-06,
      "loss": 0.6655,
      "step": 6850
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.8018195629119873,
      "learning_rate": 7.053594263067387e-06,
      "loss": 0.6587,
      "step": 6860
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.7250744104385376,
      "learning_rate": 7.043643602732525e-06,
      "loss": 0.6414,
      "step": 6870
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.4878661632537842,
      "learning_rate": 7.033683215379002e-06,
      "loss": 0.6509,
      "step": 6880
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.621942400932312,
      "learning_rate": 7.023713148414728e-06,
      "loss": 0.6672,
      "step": 6890
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.8879963159561157,
      "learning_rate": 7.0137334492936875e-06,
      "loss": 0.6588,
      "step": 6900
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.4919434785842896,
      "learning_rate": 7.0037441655157045e-06,
      "loss": 0.6529,
      "step": 6910
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.991425633430481,
      "learning_rate": 6.993745344626232e-06,
      "loss": 0.6455,
      "step": 6920
    },
    {
      "epoch": 0.46,
      "grad_norm": 3.1762428283691406,
      "learning_rate": 6.983737034216106e-06,
      "loss": 0.6531,
      "step": 6930
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.9187873601913452,
      "learning_rate": 6.973719281921336e-06,
      "loss": 0.6828,
      "step": 6940
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.8147242069244385,
      "learning_rate": 6.963692135422872e-06,
      "loss": 0.6663,
      "step": 6950
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.9920296669006348,
      "learning_rate": 6.953655642446368e-06,
      "loss": 0.6571,
      "step": 6960
    },
    {
      "epoch": 0.46,
      "grad_norm": 2.697572946548462,
      "learning_rate": 6.943609850761979e-06,
      "loss": 0.6564,
      "step": 6970
    },
    {
      "epoch": 0.46,
      "grad_norm": 2.1568384170532227,
      "learning_rate": 6.933554808184104e-06,
      "loss": 0.6625,
      "step": 6980
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.5813344717025757,
      "learning_rate": 6.9234905625711816e-06,
      "loss": 0.6872,
      "step": 6990
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.5985502004623413,
      "learning_rate": 6.913417161825449e-06,
      "loss": 0.6762,
      "step": 7000
    },
    {
      "epoch": 0.46,
      "grad_norm": 2.154099941253662,
      "learning_rate": 6.9033346538927235e-06,
      "loss": 0.6392,
      "step": 7010
    },
    {
      "epoch": 0.46,
      "grad_norm": 2.5765979290008545,
      "learning_rate": 6.8932430867621655e-06,
      "loss": 0.6707,
      "step": 7020
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.7588094472885132,
      "learning_rate": 6.883142508466054e-06,
      "loss": 0.6718,
      "step": 7030
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.5742931365966797,
      "learning_rate": 6.873032967079562e-06,
      "loss": 0.6586,
      "step": 7040
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.7835898399353027,
      "learning_rate": 6.862914510720515e-06,
      "loss": 0.6703,
      "step": 7050
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.8514904975891113,
      "learning_rate": 6.852787187549182e-06,
      "loss": 0.6803,
      "step": 7060
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.6340848207473755,
      "learning_rate": 6.842651045768026e-06,
      "loss": 0.6881,
      "step": 7070
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.552843451499939,
      "learning_rate": 6.832506133621487e-06,
      "loss": 0.6789,
      "step": 7080
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.7644469738006592,
      "learning_rate": 6.822352499395751e-06,
      "loss": 0.6853,
      "step": 7090
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.6095765829086304,
      "learning_rate": 6.812190191418508e-06,
      "loss": 0.6491,
      "step": 7100
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.5338112115859985,
      "learning_rate": 6.80201925805875e-06,
      "loss": 0.6573,
      "step": 7110
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.9101743698120117,
      "learning_rate": 6.7918397477265e-06,
      "loss": 0.6727,
      "step": 7120
    },
    {
      "epoch": 0.47,
      "grad_norm": 2.3680825233459473,
      "learning_rate": 6.781651708872629e-06,
      "loss": 0.6728,
      "step": 7130
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.8265564441680908,
      "learning_rate": 6.771455189988579e-06,
      "loss": 0.6851,
      "step": 7140
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.4049782752990723,
      "learning_rate": 6.7612502396061685e-06,
      "loss": 0.67,
      "step": 7150
    },
    {
      "epoch": 0.47,
      "grad_norm": 2.181297779083252,
      "learning_rate": 6.751036906297338e-06,
      "loss": 0.6373,
      "step": 7160
    },
    {
      "epoch": 0.47,
      "grad_norm": 2.1673145294189453,
      "learning_rate": 6.740815238673932e-06,
      "loss": 0.6639,
      "step": 7170
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.8220609426498413,
      "learning_rate": 6.730585285387465e-06,
      "loss": 0.6712,
      "step": 7180
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.3301570415496826,
      "learning_rate": 6.720347095128884e-06,
      "loss": 0.6792,
      "step": 7190
    },
    {
      "epoch": 0.48,
      "grad_norm": 2.4812264442443848,
      "learning_rate": 6.710100716628345e-06,
      "loss": 0.6863,
      "step": 7200
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.278617024421692,
      "learning_rate": 6.6998461986549715e-06,
      "loss": 0.677,
      "step": 7210
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.8752747774124146,
      "learning_rate": 6.689583590016636e-06,
      "loss": 0.666,
      "step": 7220
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.41091787815094,
      "learning_rate": 6.679312939559712e-06,
      "loss": 0.6868,
      "step": 7230
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.9605522155761719,
      "learning_rate": 6.669034296168855e-06,
      "loss": 0.6784,
      "step": 7240
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.8834712505340576,
      "learning_rate": 6.6587477087667615e-06,
      "loss": 0.6703,
      "step": 7250
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.7570191621780396,
      "learning_rate": 6.648453226313937e-06,
      "loss": 0.6847,
      "step": 7260
    },
    {
      "epoch": 0.48,
      "grad_norm": 2.3550283908843994,
      "learning_rate": 6.638150897808469e-06,
      "loss": 0.6582,
      "step": 7270
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.2605594396591187,
      "learning_rate": 6.627840772285784e-06,
      "loss": 0.6684,
      "step": 7280
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.6544970273971558,
      "learning_rate": 6.617522898818426e-06,
      "loss": 0.6698,
      "step": 7290
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.7877496480941772,
      "learning_rate": 6.607197326515808e-06,
      "loss": 0.6608,
      "step": 7300
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.8587825298309326,
      "learning_rate": 6.596864104523996e-06,
      "loss": 0.6637,
      "step": 7310
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.8580244779586792,
      "learning_rate": 6.586523282025462e-06,
      "loss": 0.6681,
      "step": 7320
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.5482275485992432,
      "learning_rate": 6.57617490823885e-06,
      "loss": 0.6583,
      "step": 7330
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.6002042293548584,
      "learning_rate": 6.565819032418748e-06,
      "loss": 0.682,
      "step": 7340
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.9164557456970215,
      "learning_rate": 6.555455703855454e-06,
      "loss": 0.667,
      "step": 7350
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.4982038736343384,
      "learning_rate": 6.545084971874738e-06,
      "loss": 0.6675,
      "step": 7360
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.3651514053344727,
      "learning_rate": 6.534706885837601e-06,
      "loss": 0.659,
      "step": 7370
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.6981933116912842,
      "learning_rate": 6.5243214951400545e-06,
      "loss": 0.6691,
      "step": 7380
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.5667850971221924,
      "learning_rate": 6.513928849212874e-06,
      "loss": 0.6888,
      "step": 7390
    },
    {
      "epoch": 0.49,
      "grad_norm": 2.9156270027160645,
      "learning_rate": 6.503528997521365e-06,
      "loss": 0.665,
      "step": 7400
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.765031099319458,
      "learning_rate": 6.4931219895651384e-06,
      "loss": 0.6694,
      "step": 7410
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.2956502437591553,
      "learning_rate": 6.482707874877855e-06,
      "loss": 0.6606,
      "step": 7420
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.3494573831558228,
      "learning_rate": 6.472286703027011e-06,
      "loss": 0.6469,
      "step": 7430
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.6514666080474854,
      "learning_rate": 6.461858523613684e-06,
      "loss": 0.6688,
      "step": 7440
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.7511714696884155,
      "learning_rate": 6.451423386272312e-06,
      "loss": 0.655,
      "step": 7450
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.893713355064392,
      "learning_rate": 6.440981340670447e-06,
      "loss": 0.6512,
      "step": 7460
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.3743759393692017,
      "learning_rate": 6.430532436508522e-06,
      "loss": 0.6542,
      "step": 7470
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.3243224620819092,
      "learning_rate": 6.420076723519615e-06,
      "loss": 0.6565,
      "step": 7480
    },
    {
      "epoch": 0.5,
      "grad_norm": 2.08793568611145,
      "learning_rate": 6.4096142514692085e-06,
      "loss": 0.6679,
      "step": 7490
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.69426691532135,
      "learning_rate": 6.399145070154962e-06,
      "loss": 0.6747,
      "step": 7500
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.9869012832641602,
      "learning_rate": 6.388669229406462e-06,
      "loss": 0.6755,
      "step": 7510
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.9419549703598022,
      "learning_rate": 6.378186779084996e-06,
      "loss": 0.6928,
      "step": 7520
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.703644871711731,
      "learning_rate": 6.367697769083312e-06,
      "loss": 0.6848,
      "step": 7530
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.7564888000488281,
      "learning_rate": 6.3572022493253715e-06,
      "loss": 0.6626,
      "step": 7540
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.867537260055542,
      "learning_rate": 6.346700269766132e-06,
      "loss": 0.6472,
      "step": 7550
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.6203207969665527,
      "learning_rate": 6.336191880391285e-06,
      "loss": 0.6706,
      "step": 7560
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.374058485031128,
      "learning_rate": 6.325677131217041e-06,
      "loss": 0.6603,
      "step": 7570
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.7284061908721924,
      "learning_rate": 6.315156072289874e-06,
      "loss": 0.6659,
      "step": 7580
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.3828376531600952,
      "learning_rate": 6.304628753686295e-06,
      "loss": 0.666,
      "step": 7590
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.9280568361282349,
      "learning_rate": 6.294095225512604e-06,
      "loss": 0.6584,
      "step": 7600
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.9138083457946777,
      "learning_rate": 6.283555537904659e-06,
      "loss": 0.6575,
      "step": 7610
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.4988782405853271,
      "learning_rate": 6.273009741027638e-06,
      "loss": 0.676,
      "step": 7620
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.5602288246154785,
      "learning_rate": 6.26245788507579e-06,
      "loss": 0.6582,
      "step": 7630
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.3849523067474365,
      "learning_rate": 6.251900020272208e-06,
      "loss": 0.6835,
      "step": 7640
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.7585147619247437,
      "learning_rate": 6.241336196868582e-06,
      "loss": 0.673,
      "step": 7650
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.720861792564392,
      "learning_rate": 6.230766465144966e-06,
      "loss": 0.6857,
      "step": 7660
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.6793100833892822,
      "learning_rate": 6.220190875409533e-06,
      "loss": 0.6667,
      "step": 7670
    },
    {
      "epoch": 0.51,
      "grad_norm": 2.7875709533691406,
      "learning_rate": 6.209609477998339e-06,
      "loss": 0.6624,
      "step": 7680
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.640952706336975,
      "learning_rate": 6.199022323275083e-06,
      "loss": 0.6723,
      "step": 7690
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.5932248830795288,
      "learning_rate": 6.188429461630866e-06,
      "loss": 0.6525,
      "step": 7700
    },
    {
      "epoch": 0.51,
      "grad_norm": 2.364243984222412,
      "learning_rate": 6.177830943483952e-06,
      "loss": 0.6771,
      "step": 7710
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.9060790538787842,
      "learning_rate": 6.1672268192795285e-06,
      "loss": 0.6932,
      "step": 7720
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.844760537147522,
      "learning_rate": 6.156617139489465e-06,
      "loss": 0.6684,
      "step": 7730
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.5574620962142944,
      "learning_rate": 6.146001954612072e-06,
      "loss": 0.6814,
      "step": 7740
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.8839787244796753,
      "learning_rate": 6.135381315171867e-06,
      "loss": 0.6783,
      "step": 7750
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.6583342552185059,
      "learning_rate": 6.124755271719326e-06,
      "loss": 0.6655,
      "step": 7760
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.7036808729171753,
      "learning_rate": 6.114123874830647e-06,
      "loss": 0.6814,
      "step": 7770
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.3549695014953613,
      "learning_rate": 6.103487175107508e-06,
      "loss": 0.6738,
      "step": 7780
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.3941162824630737,
      "learning_rate": 6.092845223176823e-06,
      "loss": 0.6722,
      "step": 7790
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.9601216316223145,
      "learning_rate": 6.0821980696905145e-06,
      "loss": 0.6645,
      "step": 7800
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.757865309715271,
      "learning_rate": 6.071545765325254e-06,
      "loss": 0.6782,
      "step": 7810
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.4168590307235718,
      "learning_rate": 6.060888360782232e-06,
      "loss": 0.6648,
      "step": 7820
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.7266621589660645,
      "learning_rate": 6.050225906786913e-06,
      "loss": 0.6729,
      "step": 7830
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.7137846946716309,
      "learning_rate": 6.039558454088796e-06,
      "loss": 0.6863,
      "step": 7840
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.9252254962921143,
      "learning_rate": 6.028886053461175e-06,
      "loss": 0.6348,
      "step": 7850
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.7567081451416016,
      "learning_rate": 6.0182087557008875e-06,
      "loss": 0.6672,
      "step": 7860
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.5034328699111938,
      "learning_rate": 6.0075266116280865e-06,
      "loss": 0.6714,
      "step": 7870
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.4937865734100342,
      "learning_rate": 5.996839672085986e-06,
      "loss": 0.6555,
      "step": 7880
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.4067682027816772,
      "learning_rate": 5.986147987940632e-06,
      "loss": 0.6473,
      "step": 7890
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.5189076662063599,
      "learning_rate": 5.975451610080643e-06,
      "loss": 0.6583,
      "step": 7900
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.8422696590423584,
      "learning_rate": 5.964750589416985e-06,
      "loss": 0.6659,
      "step": 7910
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.4895750284194946,
      "learning_rate": 5.954044976882725e-06,
      "loss": 0.6653,
      "step": 7920
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.71246337890625,
      "learning_rate": 5.943334823432777e-06,
      "loss": 0.6742,
      "step": 7930
    },
    {
      "epoch": 0.53,
      "grad_norm": 2.6499202251434326,
      "learning_rate": 5.932620180043674e-06,
      "loss": 0.6748,
      "step": 7940
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.4720969200134277,
      "learning_rate": 5.921901097713317e-06,
      "loss": 0.6714,
      "step": 7950
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.7344847917556763,
      "learning_rate": 5.911177627460739e-06,
      "loss": 0.6691,
      "step": 7960
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.5558526515960693,
      "learning_rate": 5.9004498203258495e-06,
      "loss": 0.7011,
      "step": 7970
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.4514857530593872,
      "learning_rate": 5.88971772736921e-06,
      "loss": 0.6828,
      "step": 7980
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.6652244329452515,
      "learning_rate": 5.878981399671774e-06,
      "loss": 0.6727,
      "step": 7990
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.811303734779358,
      "learning_rate": 5.8682408883346535e-06,
      "loss": 0.6786,
      "step": 8000
    },
    {
      "epoch": 0.53,
      "eval_loss": 0.7684772610664368,
      "eval_runtime": 134.4971,
      "eval_samples_per_second": 81.786,
      "eval_steps_per_second": 10.223,
      "step": 8000
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.600232481956482,
      "learning_rate": 5.8574962444788715e-06,
      "loss": 0.6763,
      "step": 8010
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.4658622741699219,
      "learning_rate": 5.846747519245123e-06,
      "loss": 0.6784,
      "step": 8020
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.8189136981964111,
      "learning_rate": 5.835994763793529e-06,
      "loss": 0.6676,
      "step": 8030
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.8251430988311768,
      "learning_rate": 5.825238029303388e-06,
      "loss": 0.6829,
      "step": 8040
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.769020676612854,
      "learning_rate": 5.814477366972945e-06,
      "loss": 0.6711,
      "step": 8050
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.7342435121536255,
      "learning_rate": 5.8037128280191315e-06,
      "loss": 0.6458,
      "step": 8060
    },
    {
      "epoch": 0.53,
      "grad_norm": 2.0114922523498535,
      "learning_rate": 5.792944463677336e-06,
      "loss": 0.6412,
      "step": 8070
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.960257649421692,
      "learning_rate": 5.782172325201155e-06,
      "loss": 0.6765,
      "step": 8080
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.7859301567077637,
      "learning_rate": 5.771396463862145e-06,
      "loss": 0.6751,
      "step": 8090
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.7072672843933105,
      "learning_rate": 5.760616930949584e-06,
      "loss": 0.6676,
      "step": 8100
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.50044584274292,
      "learning_rate": 5.749833777770225e-06,
      "loss": 0.6729,
      "step": 8110
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.5489919185638428,
      "learning_rate": 5.7390470556480545e-06,
      "loss": 0.6773,
      "step": 8120
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.5351699590682983,
      "learning_rate": 5.7282568159240405e-06,
      "loss": 0.6531,
      "step": 8130
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.8078864812850952,
      "learning_rate": 5.717463109955896e-06,
      "loss": 0.6546,
      "step": 8140
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.7729016542434692,
      "learning_rate": 5.7066659891178385e-06,
      "loss": 0.6725,
      "step": 8150
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.5250924825668335,
      "learning_rate": 5.695865504800328e-06,
      "loss": 0.6748,
      "step": 8160
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.7680206298828125,
      "learning_rate": 5.6850617084098416e-06,
      "loss": 0.6629,
      "step": 8170
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.5990138053894043,
      "learning_rate": 5.674254651368616e-06,
      "loss": 0.6725,
      "step": 8180
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.87677001953125,
      "learning_rate": 5.6634443851144115e-06,
      "loss": 0.6628,
      "step": 8190
    },
    {
      "epoch": 0.54,
      "grad_norm": 6.757842540740967,
      "learning_rate": 5.65263096110026e-06,
      "loss": 0.6609,
      "step": 8200
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.3478131294250488,
      "learning_rate": 5.641814430794222e-06,
      "loss": 0.6476,
      "step": 8210
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.5324090719223022,
      "learning_rate": 5.63099484567915e-06,
      "loss": 0.648,
      "step": 8220
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.4732797145843506,
      "learning_rate": 5.620172257252427e-06,
      "loss": 0.6745,
      "step": 8230
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.4768403768539429,
      "learning_rate": 5.609346717025738e-06,
      "loss": 0.6811,
      "step": 8240
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.409543752670288,
      "learning_rate": 5.598518276524813e-06,
      "loss": 0.6795,
      "step": 8250
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.5199010372161865,
      "learning_rate": 5.587686987289189e-06,
      "loss": 0.6879,
      "step": 8260
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.5905330181121826,
      "learning_rate": 5.57685290087196e-06,
      "loss": 0.6585,
      "step": 8270
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.3945244550704956,
      "learning_rate": 5.566016068839535e-06,
      "loss": 0.6677,
      "step": 8280
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.805010437965393,
      "learning_rate": 5.555176542771389e-06,
      "loss": 0.682,
      "step": 8290
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.46513032913208,
      "learning_rate": 5.544334374259823e-06,
      "loss": 0.654,
      "step": 8300
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.642730474472046,
      "learning_rate": 5.533489614909714e-06,
      "loss": 0.6548,
      "step": 8310
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.5549567937850952,
      "learning_rate": 5.522642316338268e-06,
      "loss": 0.66,
      "step": 8320
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.639410376548767,
      "learning_rate": 5.51179253017478e-06,
      "loss": 0.6582,
      "step": 8330
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.937220811843872,
      "learning_rate": 5.500940308060382e-06,
      "loss": 0.6769,
      "step": 8340
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.5718787908554077,
      "learning_rate": 5.490085701647805e-06,
      "loss": 0.6638,
      "step": 8350
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.5767078399658203,
      "learning_rate": 5.4792287626011206e-06,
      "loss": 0.6472,
      "step": 8360
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.781279444694519,
      "learning_rate": 5.468369542595512e-06,
      "loss": 0.6778,
      "step": 8370
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.7731683254241943,
      "learning_rate": 5.457508093317013e-06,
      "loss": 0.6707,
      "step": 8380
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.4716190099716187,
      "learning_rate": 5.446644466462269e-06,
      "loss": 0.6691,
      "step": 8390
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.4966431856155396,
      "learning_rate": 5.435778713738292e-06,
      "loss": 0.6587,
      "step": 8400
    },
    {
      "epoch": 0.56,
      "grad_norm": 2.055002212524414,
      "learning_rate": 5.4249108868622095e-06,
      "loss": 0.6513,
      "step": 8410
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.5098994970321655,
      "learning_rate": 5.414041037561022e-06,
      "loss": 0.6579,
      "step": 8420
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.787880301475525,
      "learning_rate": 5.403169217571359e-06,
      "loss": 0.6577,
      "step": 8430
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.8359959125518799,
      "learning_rate": 5.392295478639226e-06,
      "loss": 0.6628,
      "step": 8440
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.8961970806121826,
      "learning_rate": 5.381419872519763e-06,
      "loss": 0.6444,
      "step": 8450
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.7812308073043823,
      "learning_rate": 5.3705424509769976e-06,
      "loss": 0.643,
      "step": 8460
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.8749055862426758,
      "learning_rate": 5.3596632657835975e-06,
      "loss": 0.694,
      "step": 8470
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.7846895456314087,
      "learning_rate": 5.348782368720627e-06,
      "loss": 0.6745,
      "step": 8480
    },
    {
      "epoch": 0.56,
      "grad_norm": 2.077484607696533,
      "learning_rate": 5.337899811577297e-06,
      "loss": 0.6886,
      "step": 8490
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.7645025253295898,
      "learning_rate": 5.327015646150716e-06,
      "loss": 0.6339,
      "step": 8500
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.7799251079559326,
      "learning_rate": 5.3161299242456535e-06,
      "loss": 0.6564,
      "step": 8510
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.8336862325668335,
      "learning_rate": 5.3052426976742855e-06,
      "loss": 0.6761,
      "step": 8520
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.5300918817520142,
      "learning_rate": 5.294354018255945e-06,
      "loss": 0.6634,
      "step": 8530
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.747183084487915,
      "learning_rate": 5.283463937816888e-06,
      "loss": 0.6465,
      "step": 8540
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.5539292097091675,
      "learning_rate": 5.272572508190033e-06,
      "loss": 0.6844,
      "step": 8550
    },
    {
      "epoch": 0.57,
      "grad_norm": 2.0540542602539062,
      "learning_rate": 5.2616797812147205e-06,
      "loss": 0.6938,
      "step": 8560
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.6536123752593994,
      "learning_rate": 5.250785808736467e-06,
      "loss": 0.6707,
      "step": 8570
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.6305638551712036,
      "learning_rate": 5.23989064260672e-06,
      "loss": 0.6685,
      "step": 8580
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.6948388814926147,
      "learning_rate": 5.228994334682605e-06,
      "loss": 0.6685,
      "step": 8590
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.391671895980835,
      "learning_rate": 5.218096936826681e-06,
      "loss": 0.6609,
      "step": 8600
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.8620556592941284,
      "learning_rate": 5.207198500906698e-06,
      "loss": 0.6701,
      "step": 8610
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.3646891117095947,
      "learning_rate": 5.1962990787953436e-06,
      "loss": 0.6421,
      "step": 8620
    },
    {
      "epoch": 0.57,
      "grad_norm": 2.0592386722564697,
      "learning_rate": 5.185398722370004e-06,
      "loss": 0.6551,
      "step": 8630
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.6097017526626587,
      "learning_rate": 5.174497483512506e-06,
      "loss": 0.6714,
      "step": 8640
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.8083919286727905,
      "learning_rate": 5.1635954141088815e-06,
      "loss": 0.6529,
      "step": 8650
    },
    {
      "epoch": 0.57,
      "grad_norm": 2.0661661624908447,
      "learning_rate": 5.1526925660491145e-06,
      "loss": 0.6454,
      "step": 8660
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.97969388961792,
      "learning_rate": 5.141788991226892e-06,
      "loss": 0.6633,
      "step": 8670
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.696077823638916,
      "learning_rate": 5.130884741539367e-06,
      "loss": 0.6604,
      "step": 8680
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.7397289276123047,
      "learning_rate": 5.1199798688868955e-06,
      "loss": 0.6747,
      "step": 8690
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.664778232574463,
      "learning_rate": 5.109074425172806e-06,
      "loss": 0.6572,
      "step": 8700
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.5999622344970703,
      "learning_rate": 5.098168462303141e-06,
      "loss": 0.6849,
      "step": 8710
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.667110562324524,
      "learning_rate": 5.087262032186418e-06,
      "loss": 0.6791,
      "step": 8720
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.889437198638916,
      "learning_rate": 5.076355186733373e-06,
      "loss": 0.6677,
      "step": 8730
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.882400631904602,
      "learning_rate": 5.065447977856723e-06,
      "loss": 0.638,
      "step": 8740
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.7043653726577759,
      "learning_rate": 5.054540457470912e-06,
      "loss": 0.6703,
      "step": 8750
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.6736963987350464,
      "learning_rate": 5.04363267749187e-06,
      "loss": 0.656,
      "step": 8760
    },
    {
      "epoch": 0.58,
      "grad_norm": 2.609456777572632,
      "learning_rate": 5.0327246898367595e-06,
      "loss": 0.6665,
      "step": 8770
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.9416409730911255,
      "learning_rate": 5.021816546423734e-06,
      "loss": 0.6706,
      "step": 8780
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.6183526515960693,
      "learning_rate": 5.010908299171685e-06,
      "loss": 0.6759,
      "step": 8790
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.77249014377594,
      "learning_rate": 5e-06,
      "loss": 0.6524,
      "step": 8800
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.4985134601593018,
      "learning_rate": 4.989091700828316e-06,
      "loss": 0.6703,
      "step": 8810
    },
    {
      "epoch": 0.58,
      "grad_norm": 2.138824462890625,
      "learning_rate": 4.978183453576268e-06,
      "loss": 0.6905,
      "step": 8820
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.2800077199935913,
      "learning_rate": 4.967275310163241e-06,
      "loss": 0.6383,
      "step": 8830
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.545478105545044,
      "learning_rate": 4.956367322508131e-06,
      "loss": 0.6501,
      "step": 8840
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.9019412994384766,
      "learning_rate": 4.945459542529089e-06,
      "loss": 0.6876,
      "step": 8850
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.6927233934402466,
      "learning_rate": 4.934552022143279e-06,
      "loss": 0.6647,
      "step": 8860
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.4633018970489502,
      "learning_rate": 4.923644813266629e-06,
      "loss": 0.6555,
      "step": 8870
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.59763765335083,
      "learning_rate": 4.9127379678135825e-06,
      "loss": 0.6411,
      "step": 8880
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.7652654647827148,
      "learning_rate": 4.90183153769686e-06,
      "loss": 0.6775,
      "step": 8890
    },
    {
      "epoch": 0.59,
      "grad_norm": 3.181339740753174,
      "learning_rate": 4.890925574827195e-06,
      "loss": 0.659,
      "step": 8900
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.6873351335525513,
      "learning_rate": 4.880020131113107e-06,
      "loss": 0.6394,
      "step": 8910
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.513166904449463,
      "learning_rate": 4.869115258460636e-06,
      "loss": 0.6576,
      "step": 8920
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.5530864000320435,
      "learning_rate": 4.8582110087731095e-06,
      "loss": 0.6396,
      "step": 8930
    },
    {
      "epoch": 0.59,
      "grad_norm": 2.308940887451172,
      "learning_rate": 4.847307433950888e-06,
      "loss": 0.6985,
      "step": 8940
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.679738998413086,
      "learning_rate": 4.83640458589112e-06,
      "loss": 0.6722,
      "step": 8950
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.9519011974334717,
      "learning_rate": 4.825502516487497e-06,
      "loss": 0.676,
      "step": 8960
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.7706700563430786,
      "learning_rate": 4.8146012776299984e-06,
      "loss": 0.6791,
      "step": 8970
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.718878984451294,
      "learning_rate": 4.803700921204659e-06,
      "loss": 0.66,
      "step": 8980
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.953933835029602,
      "learning_rate": 4.792801499093305e-06,
      "loss": 0.6872,
      "step": 8990
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.6954289674758911,
      "learning_rate": 4.781903063173321e-06,
      "loss": 0.6759,
      "step": 9000
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.7253254652023315,
      "learning_rate": 4.771005665317398e-06,
      "loss": 0.6621,
      "step": 9010
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.4226269721984863,
      "learning_rate": 4.760109357393282e-06,
      "loss": 0.643,
      "step": 9020
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.6937118768692017,
      "learning_rate": 4.749214191263533e-06,
      "loss": 0.6726,
      "step": 9030
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.4342418909072876,
      "learning_rate": 4.738320218785281e-06,
      "loss": 0.681,
      "step": 9040
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.6379367113113403,
      "learning_rate": 4.727427491809968e-06,
      "loss": 0.6942,
      "step": 9050
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.7488796710968018,
      "learning_rate": 4.716536062183112e-06,
      "loss": 0.6669,
      "step": 9060
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.671850323677063,
      "learning_rate": 4.705645981744055e-06,
      "loss": 0.6713,
      "step": 9070
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.6623774766921997,
      "learning_rate": 4.694757302325715e-06,
      "loss": 0.668,
      "step": 9080
    },
    {
      "epoch": 0.6,
      "grad_norm": 2.3808863162994385,
      "learning_rate": 4.683870075754347e-06,
      "loss": 0.688,
      "step": 9090
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.9445726871490479,
      "learning_rate": 4.672984353849285e-06,
      "loss": 0.669,
      "step": 9100
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.3454090356826782,
      "learning_rate": 4.662100188422705e-06,
      "loss": 0.6616,
      "step": 9110
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.649337649345398,
      "learning_rate": 4.651217631279374e-06,
      "loss": 0.6416,
      "step": 9120
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.760851502418518,
      "learning_rate": 4.640336734216403e-06,
      "loss": 0.678,
      "step": 9130
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.77865469455719,
      "learning_rate": 4.629457549023004e-06,
      "loss": 0.6744,
      "step": 9140
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.9781075716018677,
      "learning_rate": 4.618580127480239e-06,
      "loss": 0.6548,
      "step": 9150
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.4706860780715942,
      "learning_rate": 4.6077045213607765e-06,
      "loss": 0.6828,
      "step": 9160
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.6444385051727295,
      "learning_rate": 4.596830782428642e-06,
      "loss": 0.6629,
      "step": 9170
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.650578498840332,
      "learning_rate": 4.58595896243898e-06,
      "loss": 0.6529,
      "step": 9180
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.7814977169036865,
      "learning_rate": 4.575089113137792e-06,
      "loss": 0.6633,
      "step": 9190
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.5504720211029053,
      "learning_rate": 4.564221286261709e-06,
      "loss": 0.6463,
      "step": 9200
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.6057378053665161,
      "learning_rate": 4.553355533537732e-06,
      "loss": 0.669,
      "step": 9210
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.5203750133514404,
      "learning_rate": 4.542491906682988e-06,
      "loss": 0.6646,
      "step": 9220
    },
    {
      "epoch": 0.61,
      "grad_norm": 2.8241419792175293,
      "learning_rate": 4.53163045740449e-06,
      "loss": 0.6652,
      "step": 9230
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.680282711982727,
      "learning_rate": 4.52077123739888e-06,
      "loss": 0.6757,
      "step": 9240
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.5383038520812988,
      "learning_rate": 4.509914298352197e-06,
      "loss": 0.6775,
      "step": 9250
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.3444230556488037,
      "learning_rate": 4.49905969193962e-06,
      "loss": 0.6407,
      "step": 9260
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.9666281938552856,
      "learning_rate": 4.488207469825221e-06,
      "loss": 0.6731,
      "step": 9270
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.1215275526046753,
      "learning_rate": 4.477357683661734e-06,
      "loss": 0.6692,
      "step": 9280
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.7040753364562988,
      "learning_rate": 4.466510385090287e-06,
      "loss": 0.6571,
      "step": 9290
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.5992120504379272,
      "learning_rate": 4.4556656257401786e-06,
      "loss": 0.6658,
      "step": 9300
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.7472400665283203,
      "learning_rate": 4.4448234572286126e-06,
      "loss": 0.6786,
      "step": 9310
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.5891025066375732,
      "learning_rate": 4.4339839311604675e-06,
      "loss": 0.6595,
      "step": 9320
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.8652875423431396,
      "learning_rate": 4.4231470991280425e-06,
      "loss": 0.6568,
      "step": 9330
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.7360942363739014,
      "learning_rate": 4.4123130127108125e-06,
      "loss": 0.6626,
      "step": 9340
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.7511422634124756,
      "learning_rate": 4.401481723475189e-06,
      "loss": 0.6569,
      "step": 9350
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.7602146863937378,
      "learning_rate": 4.390653282974264e-06,
      "loss": 0.6431,
      "step": 9360
    },
    {
      "epoch": 0.62,
      "grad_norm": 2.336517095565796,
      "learning_rate": 4.379827742747575e-06,
      "loss": 0.6788,
      "step": 9370
    },
    {
      "epoch": 0.62,
      "grad_norm": 2.613835096359253,
      "learning_rate": 4.3690051543208526e-06,
      "loss": 0.6776,
      "step": 9380
    },
    {
      "epoch": 0.62,
      "grad_norm": 2.495633840560913,
      "learning_rate": 4.358185569205779e-06,
      "loss": 0.6867,
      "step": 9390
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.8492158651351929,
      "learning_rate": 4.347369038899744e-06,
      "loss": 0.6553,
      "step": 9400
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.860273838043213,
      "learning_rate": 4.336555614885591e-06,
      "loss": 0.6648,
      "step": 9410
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.934490442276001,
      "learning_rate": 4.3257453486313865e-06,
      "loss": 0.641,
      "step": 9420
    },
    {
      "epoch": 0.62,
      "grad_norm": 2.0558342933654785,
      "learning_rate": 4.314938291590161e-06,
      "loss": 0.6659,
      "step": 9430
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.751317024230957,
      "learning_rate": 4.304134495199675e-06,
      "loss": 0.6718,
      "step": 9440
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.4357839822769165,
      "learning_rate": 4.293334010882164e-06,
      "loss": 0.674,
      "step": 9450
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.4586100578308105,
      "learning_rate": 4.282536890044105e-06,
      "loss": 0.6694,
      "step": 9460
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.403393030166626,
      "learning_rate": 4.271743184075963e-06,
      "loss": 0.6625,
      "step": 9470
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.9740585088729858,
      "learning_rate": 4.260952944351947e-06,
      "loss": 0.65,
      "step": 9480
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.788071632385254,
      "learning_rate": 4.250166222229775e-06,
      "loss": 0.661,
      "step": 9490
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.5685087442398071,
      "learning_rate": 4.239383069050417e-06,
      "loss": 0.6735,
      "step": 9500
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.4203977584838867,
      "learning_rate": 4.228603536137856e-06,
      "loss": 0.6416,
      "step": 9510
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.93641197681427,
      "learning_rate": 4.217827674798845e-06,
      "loss": 0.6681,
      "step": 9520
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.8250054121017456,
      "learning_rate": 4.207055536322665e-06,
      "loss": 0.6954,
      "step": 9530
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.5296562910079956,
      "learning_rate": 4.196287171980869e-06,
      "loss": 0.6809,
      "step": 9540
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.7069146633148193,
      "learning_rate": 4.185522633027057e-06,
      "loss": 0.648,
      "step": 9550
    },
    {
      "epoch": 0.63,
      "grad_norm": 2.05324125289917,
      "learning_rate": 4.174761970696612e-06,
      "loss": 0.6818,
      "step": 9560
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.6765975952148438,
      "learning_rate": 4.164005236206471e-06,
      "loss": 0.674,
      "step": 9570
    },
    {
      "epoch": 0.63,
      "grad_norm": 2.0910682678222656,
      "learning_rate": 4.1532524807548776e-06,
      "loss": 0.6609,
      "step": 9580
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.6869258880615234,
      "learning_rate": 4.142503755521129e-06,
      "loss": 0.6618,
      "step": 9590
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.5661019086837769,
      "learning_rate": 4.131759111665349e-06,
      "loss": 0.6749,
      "step": 9600
    },
    {
      "epoch": 0.64,
      "eval_loss": 0.7704323530197144,
      "eval_runtime": 134.2109,
      "eval_samples_per_second": 81.961,
      "eval_steps_per_second": 10.245,
      "step": 9600
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.7838884592056274,
      "learning_rate": 4.1210186003282275e-06,
      "loss": 0.6592,
      "step": 9610
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.9045015573501587,
      "learning_rate": 4.1102822726307925e-06,
      "loss": 0.66,
      "step": 9620
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.6812734603881836,
      "learning_rate": 4.099550179674151e-06,
      "loss": 0.6734,
      "step": 9630
    },
    {
      "epoch": 0.64,
      "grad_norm": 2.4321088790893555,
      "learning_rate": 4.088822372539263e-06,
      "loss": 0.6621,
      "step": 9640
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.5010099411010742,
      "learning_rate": 4.078098902286684e-06,
      "loss": 0.674,
      "step": 9650
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.6574801206588745,
      "learning_rate": 4.067379819956327e-06,
      "loss": 0.6388,
      "step": 9660
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.5412184000015259,
      "learning_rate": 4.056665176567225e-06,
      "loss": 0.6437,
      "step": 9670
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.5895485877990723,
      "learning_rate": 4.045955023117276e-06,
      "loss": 0.6687,
      "step": 9680
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.414804458618164,
      "learning_rate": 4.0352494105830155e-06,
      "loss": 0.6747,
      "step": 9690
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.5721184015274048,
      "learning_rate": 4.02454838991936e-06,
      "loss": 0.6679,
      "step": 9700
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.5480763912200928,
      "learning_rate": 4.013852012059371e-06,
      "loss": 0.6684,
      "step": 9710
    },
    {
      "epoch": 0.64,
      "grad_norm": 2.122615337371826,
      "learning_rate": 4.003160327914015e-06,
      "loss": 0.6544,
      "step": 9720
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.6763825416564941,
      "learning_rate": 3.992473388371914e-06,
      "loss": 0.6747,
      "step": 9730
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.469360589981079,
      "learning_rate": 3.981791244299113e-06,
      "loss": 0.659,
      "step": 9740
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.746013879776001,
      "learning_rate": 3.971113946538826e-06,
      "loss": 0.6904,
      "step": 9750
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.5463660955429077,
      "learning_rate": 3.960441545911205e-06,
      "loss": 0.6837,
      "step": 9760
    },
    {
      "epoch": 0.65,
      "grad_norm": 2.29986310005188,
      "learning_rate": 3.949774093213089e-06,
      "loss": 0.6809,
      "step": 9770
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.6813066005706787,
      "learning_rate": 3.939111639217769e-06,
      "loss": 0.6804,
      "step": 9780
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.5592458248138428,
      "learning_rate": 3.928454234674748e-06,
      "loss": 0.6664,
      "step": 9790
    },
    {
      "epoch": 0.65,
      "grad_norm": 2.1687581539154053,
      "learning_rate": 3.917801930309486e-06,
      "loss": 0.6589,
      "step": 9800
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.744254469871521,
      "learning_rate": 3.907154776823179e-06,
      "loss": 0.6624,
      "step": 9810
    },
    {
      "epoch": 0.65,
      "grad_norm": 2.446599245071411,
      "learning_rate": 3.8965128248924956e-06,
      "loss": 0.661,
      "step": 9820
    },
    {
      "epoch": 0.65,
      "grad_norm": 2.0234007835388184,
      "learning_rate": 3.885876125169356e-06,
      "loss": 0.6654,
      "step": 9830
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.8161579370498657,
      "learning_rate": 3.875244728280676e-06,
      "loss": 0.6779,
      "step": 9840
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.6402438879013062,
      "learning_rate": 3.864618684828135e-06,
      "loss": 0.6477,
      "step": 9850
    },
    {
      "epoch": 0.65,
      "grad_norm": 2.011474370956421,
      "learning_rate": 3.85399804538793e-06,
      "loss": 0.649,
      "step": 9860
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.7880947589874268,
      "learning_rate": 3.8433828605105385e-06,
      "loss": 0.6751,
      "step": 9870
    },
    {
      "epoch": 0.65,
      "grad_norm": 2.0285022258758545,
      "learning_rate": 3.832773180720475e-06,
      "loss": 0.662,
      "step": 9880
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.7321562767028809,
      "learning_rate": 3.822169056516051e-06,
      "loss": 0.6698,
      "step": 9890
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.807761788368225,
      "learning_rate": 3.8115705383691354e-06,
      "loss": 0.658,
      "step": 9900
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.7578504085540771,
      "learning_rate": 3.800977676724919e-06,
      "loss": 0.6343,
      "step": 9910
    },
    {
      "epoch": 0.66,
      "grad_norm": 2.058124542236328,
      "learning_rate": 3.790390522001662e-06,
      "loss": 0.6572,
      "step": 9920
    },
    {
      "epoch": 0.66,
      "grad_norm": 2.71362566947937,
      "learning_rate": 3.7798091245904674e-06,
      "loss": 0.6728,
      "step": 9930
    },
    {
      "epoch": 0.66,
      "grad_norm": 2.2041659355163574,
      "learning_rate": 3.769233534855035e-06,
      "loss": 0.6564,
      "step": 9940
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.7433922290802002,
      "learning_rate": 3.7586638031314182e-06,
      "loss": 0.6592,
      "step": 9950
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.8850117921829224,
      "learning_rate": 3.748099979727792e-06,
      "loss": 0.6677,
      "step": 9960
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.5975645780563354,
      "learning_rate": 3.7375421149242102e-06,
      "loss": 0.6874,
      "step": 9970
    },
    {
      "epoch": 0.66,
      "grad_norm": 2.081080198287964,
      "learning_rate": 3.7269902589723617e-06,
      "loss": 0.6508,
      "step": 9980
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.5484403371810913,
      "learning_rate": 3.7164444620953397e-06,
      "loss": 0.6681,
      "step": 9990
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.7869611978530884,
      "learning_rate": 3.705904774487396e-06,
      "loss": 0.6762,
      "step": 10000
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.8547977209091187,
      "learning_rate": 3.6953712463137057e-06,
      "loss": 0.6572,
      "step": 10010
    },
    {
      "epoch": 0.66,
      "grad_norm": 2.011695146560669,
      "learning_rate": 3.6848439277101262e-06,
      "loss": 0.6574,
      "step": 10020
    },
    {
      "epoch": 0.66,
      "grad_norm": 2.0033137798309326,
      "learning_rate": 3.6743228687829596e-06,
      "loss": 0.6723,
      "step": 10030
    },
    {
      "epoch": 0.66,
      "grad_norm": 2.6469576358795166,
      "learning_rate": 3.663808119608716e-06,
      "loss": 0.6693,
      "step": 10040
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.7878594398498535,
      "learning_rate": 3.6532997302338704e-06,
      "loss": 0.6653,
      "step": 10050
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.6353760957717896,
      "learning_rate": 3.6427977506746293e-06,
      "loss": 0.6429,
      "step": 10060
    },
    {
      "epoch": 0.67,
      "grad_norm": 2.0853710174560547,
      "learning_rate": 3.63230223091669e-06,
      "loss": 0.6467,
      "step": 10070
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.7960524559020996,
      "learning_rate": 3.6218132209150047e-06,
      "loss": 0.6812,
      "step": 10080
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.5491673946380615,
      "learning_rate": 3.6113307705935398e-06,
      "loss": 0.6718,
      "step": 10090
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.6183891296386719,
      "learning_rate": 3.6008549298450403e-06,
      "loss": 0.6671,
      "step": 10100
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.508072018623352,
      "learning_rate": 3.5903857485307936e-06,
      "loss": 0.6755,
      "step": 10110
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.8748259544372559,
      "learning_rate": 3.579923276480387e-06,
      "loss": 0.6664,
      "step": 10120
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.6908599138259888,
      "learning_rate": 3.56946756349148e-06,
      "loss": 0.6732,
      "step": 10130
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.944483995437622,
      "learning_rate": 3.559018659329554e-06,
      "loss": 0.6482,
      "step": 10140
    },
    {
      "epoch": 0.67,
      "grad_norm": 2.147538900375366,
      "learning_rate": 3.5485766137276894e-06,
      "loss": 0.6454,
      "step": 10150
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.6188044548034668,
      "learning_rate": 3.538141476386317e-06,
      "loss": 0.6458,
      "step": 10160
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.9587546586990356,
      "learning_rate": 3.527713296972991e-06,
      "loss": 0.662,
      "step": 10170
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.46161949634552,
      "learning_rate": 3.517292125122146e-06,
      "loss": 0.6857,
      "step": 10180
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.7037651538848877,
      "learning_rate": 3.5068780104348632e-06,
      "loss": 0.6547,
      "step": 10190
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.446110486984253,
      "learning_rate": 3.4964710024786354e-06,
      "loss": 0.6737,
      "step": 10200
    },
    {
      "epoch": 0.68,
      "grad_norm": 2.2284021377563477,
      "learning_rate": 3.486071150787128e-06,
      "loss": 0.6568,
      "step": 10210
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.4104644060134888,
      "learning_rate": 3.4756785048599464e-06,
      "loss": 0.6572,
      "step": 10220
    },
    {
      "epoch": 0.68,
      "grad_norm": 2.4139363765716553,
      "learning_rate": 3.4652931141624002e-06,
      "loss": 0.6568,
      "step": 10230
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.8519604206085205,
      "learning_rate": 3.4549150281252635e-06,
      "loss": 0.6626,
      "step": 10240
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.8511260747909546,
      "learning_rate": 3.444544296144546e-06,
      "loss": 0.688,
      "step": 10250
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.454230785369873,
      "learning_rate": 3.4341809675812532e-06,
      "loss": 0.6693,
      "step": 10260
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.7652877569198608,
      "learning_rate": 3.4238250917611533e-06,
      "loss": 0.6523,
      "step": 10270
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.9757181406021118,
      "learning_rate": 3.4134767179745404e-06,
      "loss": 0.6672,
      "step": 10280
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.6871960163116455,
      "learning_rate": 3.403135895476004e-06,
      "loss": 0.6604,
      "step": 10290
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.3390458822250366,
      "learning_rate": 3.3928026734841935e-06,
      "loss": 0.6668,
      "step": 10300
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.4733284711837769,
      "learning_rate": 3.3824771011815772e-06,
      "loss": 0.6843,
      "step": 10310
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.77756667137146,
      "learning_rate": 3.372159227714218e-06,
      "loss": 0.6809,
      "step": 10320
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.5924615859985352,
      "learning_rate": 3.3618491021915334e-06,
      "loss": 0.6658,
      "step": 10330
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.7609484195709229,
      "learning_rate": 3.351546773686065e-06,
      "loss": 0.6797,
      "step": 10340
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.5677520036697388,
      "learning_rate": 3.341252291233241e-06,
      "loss": 0.668,
      "step": 10350
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.59241783618927,
      "learning_rate": 3.330965703831146e-06,
      "loss": 0.6561,
      "step": 10360
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.4724187850952148,
      "learning_rate": 3.3206870604402897e-06,
      "loss": 0.6326,
      "step": 10370
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.5326093435287476,
      "learning_rate": 3.3104164099833652e-06,
      "loss": 0.6764,
      "step": 10380
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.5835292339324951,
      "learning_rate": 3.3001538013450285e-06,
      "loss": 0.6678,
      "step": 10390
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.7297430038452148,
      "learning_rate": 3.289899283371657e-06,
      "loss": 0.6459,
      "step": 10400
    },
    {
      "epoch": 0.69,
      "grad_norm": 2.1451315879821777,
      "learning_rate": 3.2796529048711158e-06,
      "loss": 0.65,
      "step": 10410
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.720035433769226,
      "learning_rate": 3.269414714612534e-06,
      "loss": 0.687,
      "step": 10420
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.6080423593521118,
      "learning_rate": 3.259184761326068e-06,
      "loss": 0.677,
      "step": 10430
    },
    {
      "epoch": 0.69,
      "grad_norm": 2.1213626861572266,
      "learning_rate": 3.248963093702663e-06,
      "loss": 0.6687,
      "step": 10440
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.6701629161834717,
      "learning_rate": 3.2387497603938327e-06,
      "loss": 0.6816,
      "step": 10450
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.812422513961792,
      "learning_rate": 3.2285448100114208e-06,
      "loss": 0.6695,
      "step": 10460
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.7101892232894897,
      "learning_rate": 3.218348291127371e-06,
      "loss": 0.6676,
      "step": 10470
    },
    {
      "epoch": 0.69,
      "grad_norm": 2.012482166290283,
      "learning_rate": 3.2081602522734987e-06,
      "loss": 0.6616,
      "step": 10480
    },
    {
      "epoch": 0.69,
      "grad_norm": 2.179459571838379,
      "learning_rate": 3.1979807419412523e-06,
      "loss": 0.6681,
      "step": 10490
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.7281560897827148,
      "learning_rate": 3.1878098085814926e-06,
      "loss": 0.6808,
      "step": 10500
    },
    {
      "epoch": 0.7,
      "grad_norm": 2.1194772720336914,
      "learning_rate": 3.177647500604252e-06,
      "loss": 0.6662,
      "step": 10510
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.462747573852539,
      "learning_rate": 3.167493866378514e-06,
      "loss": 0.6606,
      "step": 10520
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.694826602935791,
      "learning_rate": 3.1573489542319754e-06,
      "loss": 0.6513,
      "step": 10530
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.4207332134246826,
      "learning_rate": 3.147212812450819e-06,
      "loss": 0.6592,
      "step": 10540
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.5110808610916138,
      "learning_rate": 3.1370854892794855e-06,
      "loss": 0.6504,
      "step": 10550
    },
    {
      "epoch": 0.7,
      "grad_norm": 3.3117215633392334,
      "learning_rate": 3.12696703292044e-06,
      "loss": 0.6538,
      "step": 10560
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.655723214149475,
      "learning_rate": 3.1168574915339465e-06,
      "loss": 0.6742,
      "step": 10570
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.747663974761963,
      "learning_rate": 3.1067569132378358e-06,
      "loss": 0.6694,
      "step": 10580
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.6091103553771973,
      "learning_rate": 3.0966653461072778e-06,
      "loss": 0.6666,
      "step": 10590
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.784797191619873,
      "learning_rate": 3.0865828381745515e-06,
      "loss": 0.6683,
      "step": 10600
    },
    {
      "epoch": 0.7,
      "grad_norm": 2.061067819595337,
      "learning_rate": 3.0765094374288197e-06,
      "loss": 0.666,
      "step": 10610
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.6201362609863281,
      "learning_rate": 3.066445191815898e-06,
      "loss": 0.6738,
      "step": 10620
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.7196377515792847,
      "learning_rate": 3.056390149238022e-06,
      "loss": 0.6561,
      "step": 10630
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.8749741315841675,
      "learning_rate": 3.0463443575536324e-06,
      "loss": 0.6817,
      "step": 10640
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.473533034324646,
      "learning_rate": 3.0363078645771303e-06,
      "loss": 0.6449,
      "step": 10650
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.9876571893692017,
      "learning_rate": 3.0262807180786647e-06,
      "loss": 0.6657,
      "step": 10660
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.949334979057312,
      "learning_rate": 3.0162629657838947e-06,
      "loss": 0.6462,
      "step": 10670
    },
    {
      "epoch": 0.71,
      "grad_norm": 3.1843159198760986,
      "learning_rate": 3.0062546553737692e-06,
      "loss": 0.6448,
      "step": 10680
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.9066513776779175,
      "learning_rate": 2.9962558344842963e-06,
      "loss": 0.6639,
      "step": 10690
    },
    {
      "epoch": 0.71,
      "grad_norm": 3.2927820682525635,
      "learning_rate": 2.986266550706315e-06,
      "loss": 0.6532,
      "step": 10700
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.4722814559936523,
      "learning_rate": 2.976286851585274e-06,
      "loss": 0.6665,
      "step": 10710
    },
    {
      "epoch": 0.71,
      "grad_norm": 2.8656790256500244,
      "learning_rate": 2.966316784621e-06,
      "loss": 0.6719,
      "step": 10720
    },
    {
      "epoch": 0.71,
      "grad_norm": 2.7842459678649902,
      "learning_rate": 2.956356397267477e-06,
      "loss": 0.6726,
      "step": 10730
    },
    {
      "epoch": 0.71,
      "grad_norm": 2.024540662765503,
      "learning_rate": 2.946405736932615e-06,
      "loss": 0.6587,
      "step": 10740
    },
    {
      "epoch": 0.71,
      "grad_norm": 2.1006271839141846,
      "learning_rate": 2.936464850978027e-06,
      "loss": 0.6585,
      "step": 10750
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.8146822452545166,
      "learning_rate": 2.926533786718806e-06,
      "loss": 0.6355,
      "step": 10760
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.8565179109573364,
      "learning_rate": 2.9166125914232935e-06,
      "loss": 0.6917,
      "step": 10770
    },
    {
      "epoch": 0.71,
      "grad_norm": 2.3497860431671143,
      "learning_rate": 2.906701312312861e-06,
      "loss": 0.6522,
      "step": 10780
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.8022676706314087,
      "learning_rate": 2.8967999965616815e-06,
      "loss": 0.6548,
      "step": 10790
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.768591284751892,
      "learning_rate": 2.886908691296504e-06,
      "loss": 0.6607,
      "step": 10800
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.492566466331482,
      "learning_rate": 2.8770274435964356e-06,
      "loss": 0.6817,
      "step": 10810
    },
    {
      "epoch": 0.72,
      "grad_norm": 2.3266677856445312,
      "learning_rate": 2.8671563004927107e-06,
      "loss": 0.6659,
      "step": 10820
    },
    {
      "epoch": 0.72,
      "grad_norm": 2.3747525215148926,
      "learning_rate": 2.8572953089684654e-06,
      "loss": 0.6794,
      "step": 10830
    },
    {
      "epoch": 0.72,
      "grad_norm": 2.2825326919555664,
      "learning_rate": 2.8474445159585235e-06,
      "loss": 0.6911,
      "step": 10840
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.5767141580581665,
      "learning_rate": 2.8376039683491683e-06,
      "loss": 0.6322,
      "step": 10850
    },
    {
      "epoch": 0.72,
      "grad_norm": 2.076864242553711,
      "learning_rate": 2.827773712977915e-06,
      "loss": 0.6793,
      "step": 10860
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.8241063356399536,
      "learning_rate": 2.817953796633289e-06,
      "loss": 0.6607,
      "step": 10870
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.8080527782440186,
      "learning_rate": 2.8081442660546126e-06,
      "loss": 0.6799,
      "step": 10880
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.869896650314331,
      "learning_rate": 2.798345167931771e-06,
      "loss": 0.6643,
      "step": 10890
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.5855712890625,
      "learning_rate": 2.7885565489049948e-06,
      "loss": 0.6549,
      "step": 10900
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.8657243251800537,
      "learning_rate": 2.7787784555646363e-06,
      "loss": 0.6615,
      "step": 10910
    },
    {
      "epoch": 0.72,
      "grad_norm": 2.6397323608398438,
      "learning_rate": 2.7690109344509563e-06,
      "loss": 0.671,
      "step": 10920
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.839142084121704,
      "learning_rate": 2.759254032053888e-06,
      "loss": 0.6648,
      "step": 10930
    },
    {
      "epoch": 0.72,
      "grad_norm": 2.0795602798461914,
      "learning_rate": 2.7495077948128245e-06,
      "loss": 0.6503,
      "step": 10940
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.419287919998169,
      "learning_rate": 2.739772269116402e-06,
      "loss": 0.6647,
      "step": 10950
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.8919901847839355,
      "learning_rate": 2.7300475013022666e-06,
      "loss": 0.6565,
      "step": 10960
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.7516885995864868,
      "learning_rate": 2.720333537656865e-06,
      "loss": 0.6556,
      "step": 10970
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.7074434757232666,
      "learning_rate": 2.710630424415216e-06,
      "loss": 0.6839,
      "step": 10980
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.6884958744049072,
      "learning_rate": 2.700938207760701e-06,
      "loss": 0.6645,
      "step": 10990
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.643620252609253,
      "learning_rate": 2.6912569338248317e-06,
      "loss": 0.6514,
      "step": 11000
    },
    {
      "epoch": 0.73,
      "grad_norm": 2.157878875732422,
      "learning_rate": 2.681586648687035e-06,
      "loss": 0.6604,
      "step": 11010
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.6927180290222168,
      "learning_rate": 2.671927398374443e-06,
      "loss": 0.6996,
      "step": 11020
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.435312032699585,
      "learning_rate": 2.6622792288616595e-06,
      "loss": 0.6723,
      "step": 11030
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.5182280540466309,
      "learning_rate": 2.6526421860705474e-06,
      "loss": 0.6723,
      "step": 11040
    },
    {
      "epoch": 0.73,
      "grad_norm": 2.380584478378296,
      "learning_rate": 2.6430163158700116e-06,
      "loss": 0.666,
      "step": 11050
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.787930965423584,
      "learning_rate": 2.6334016640757838e-06,
      "loss": 0.6569,
      "step": 11060
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.6916040182113647,
      "learning_rate": 2.6237982764501936e-06,
      "loss": 0.6599,
      "step": 11070
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.7497210502624512,
      "learning_rate": 2.614206198701958e-06,
      "loss": 0.6667,
      "step": 11080
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.8590047359466553,
      "learning_rate": 2.6046254764859687e-06,
      "loss": 0.6574,
      "step": 11090
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.6491410732269287,
      "learning_rate": 2.595056155403063e-06,
      "loss": 0.685,
      "step": 11100
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.8310904502868652,
      "learning_rate": 2.5854982809998154e-06,
      "loss": 0.6726,
      "step": 11110
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.8105839490890503,
      "learning_rate": 2.5759518987683154e-06,
      "loss": 0.6754,
      "step": 11120
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.6606043577194214,
      "learning_rate": 2.56641705414596e-06,
      "loss": 0.6668,
      "step": 11130
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.6423386335372925,
      "learning_rate": 2.5568937925152272e-06,
      "loss": 0.6726,
      "step": 11140
    },
    {
      "epoch": 0.74,
      "grad_norm": 2.008108377456665,
      "learning_rate": 2.5473821592034604e-06,
      "loss": 0.6687,
      "step": 11150
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.9500596523284912,
      "learning_rate": 2.5378821994826654e-06,
      "loss": 0.6748,
      "step": 11160
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.762308120727539,
      "learning_rate": 2.5283939585692787e-06,
      "loss": 0.664,
      "step": 11170
    },
    {
      "epoch": 0.74,
      "grad_norm": 2.1707937717437744,
      "learning_rate": 2.518917481623961e-06,
      "loss": 0.6662,
      "step": 11180
    },
    {
      "epoch": 0.74,
      "grad_norm": 2.0620369911193848,
      "learning_rate": 2.5094528137513797e-06,
      "loss": 0.6487,
      "step": 11190
    },
    {
      "epoch": 0.74,
      "grad_norm": 2.011185884475708,
      "learning_rate": 2.5000000000000015e-06,
      "loss": 0.6789,
      "step": 11200
    },
    {
      "epoch": 0.74,
      "eval_loss": 0.7809577584266663,
      "eval_runtime": 134.3045,
      "eval_samples_per_second": 81.903,
      "eval_steps_per_second": 10.238,
      "step": 11200
    },
    {
      "epoch": 0.74,
      "grad_norm": 2.186656951904297,
      "learning_rate": 2.490559085361863e-06,
      "loss": 0.6671,
      "step": 11210
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.421526312828064,
      "learning_rate": 2.481130114772369e-06,
      "loss": 0.6711,
      "step": 11220
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.6306933164596558,
      "learning_rate": 2.471713133110078e-06,
      "loss": 0.6659,
      "step": 11230
    },
    {
      "epoch": 0.74,
      "grad_norm": 3.0042200088500977,
      "learning_rate": 2.462308185196481e-06,
      "loss": 0.6408,
      "step": 11240
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.5799850225448608,
      "learning_rate": 2.4529153157957913e-06,
      "loss": 0.658,
      "step": 11250
    },
    {
      "epoch": 0.75,
      "grad_norm": 2.0872106552124023,
      "learning_rate": 2.4435345696147404e-06,
      "loss": 0.6609,
      "step": 11260
    },
    {
      "epoch": 0.75,
      "grad_norm": 2.4066195487976074,
      "learning_rate": 2.43416599130235e-06,
      "loss": 0.6471,
      "step": 11270
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.88292396068573,
      "learning_rate": 2.424809625449729e-06,
      "loss": 0.6777,
      "step": 11280
    },
    {
      "epoch": 0.75,
      "grad_norm": 2.071256160736084,
      "learning_rate": 2.4154655165898626e-06,
      "loss": 0.6567,
      "step": 11290
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.9656749963760376,
      "learning_rate": 2.406133709197392e-06,
      "loss": 0.6701,
      "step": 11300
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.7675248384475708,
      "learning_rate": 2.396814247688413e-06,
      "loss": 0.6567,
      "step": 11310
    },
    {
      "epoch": 0.75,
      "grad_norm": 2.7106385231018066,
      "learning_rate": 2.387507176420256e-06,
      "loss": 0.6772,
      "step": 11320
    },
    {
      "epoch": 0.75,
      "grad_norm": 2.049154281616211,
      "learning_rate": 2.3782125396912765e-06,
      "loss": 0.6517,
      "step": 11330
    },
    {
      "epoch": 0.75,
      "grad_norm": 2.511610984802246,
      "learning_rate": 2.3689303817406523e-06,
      "loss": 0.6672,
      "step": 11340
    },
    {
      "epoch": 0.75,
      "grad_norm": 2.223442554473877,
      "learning_rate": 2.3596607467481602e-06,
      "loss": 0.6525,
      "step": 11350
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.762532114982605,
      "learning_rate": 2.3504036788339763e-06,
      "loss": 0.6863,
      "step": 11360
    },
    {
      "epoch": 0.75,
      "grad_norm": 2.122451066970825,
      "learning_rate": 2.3411592220584574e-06,
      "loss": 0.6608,
      "step": 11370
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.7999773025512695,
      "learning_rate": 2.3319274204219427e-06,
      "loss": 0.6354,
      "step": 11380
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.5697572231292725,
      "learning_rate": 2.3227083178645316e-06,
      "loss": 0.6651,
      "step": 11390
    },
    {
      "epoch": 0.75,
      "grad_norm": 2.309673309326172,
      "learning_rate": 2.3135019582658803e-06,
      "loss": 0.6694,
      "step": 11400
    },
    {
      "epoch": 0.76,
      "grad_norm": 2.324636220932007,
      "learning_rate": 2.304308385444999e-06,
      "loss": 0.6593,
      "step": 11410
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.683274745941162,
      "learning_rate": 2.295127643160031e-06,
      "loss": 0.6747,
      "step": 11420
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.913460373878479,
      "learning_rate": 2.2859597751080536e-06,
      "loss": 0.6553,
      "step": 11430
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.3622136116027832,
      "learning_rate": 2.2768048249248648e-06,
      "loss": 0.6639,
      "step": 11440
    },
    {
      "epoch": 0.76,
      "grad_norm": 2.03890323638916,
      "learning_rate": 2.2676628361847834e-06,
      "loss": 0.663,
      "step": 11450
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.982285737991333,
      "learning_rate": 2.258533852400432e-06,
      "loss": 0.663,
      "step": 11460
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.8850750923156738,
      "learning_rate": 2.2494179170225333e-06,
      "loss": 0.6667,
      "step": 11470
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.936200737953186,
      "learning_rate": 2.2403150734397095e-06,
      "loss": 0.6626,
      "step": 11480
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.9150683879852295,
      "learning_rate": 2.2312253649782655e-06,
      "loss": 0.663,
      "step": 11490
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.6289910078048706,
      "learning_rate": 2.2221488349019903e-06,
      "loss": 0.6719,
      "step": 11500
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.7412772178649902,
      "learning_rate": 2.213085526411945e-06,
      "loss": 0.6666,
      "step": 11510
    },
    {
      "epoch": 0.76,
      "grad_norm": 2.1215097904205322,
      "learning_rate": 2.204035482646267e-06,
      "loss": 0.6705,
      "step": 11520
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.5671436786651611,
      "learning_rate": 2.1949987466799524e-06,
      "loss": 0.6633,
      "step": 11530
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.7265719175338745,
      "learning_rate": 2.185975361524657e-06,
      "loss": 0.6555,
      "step": 11540
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.697858214378357,
      "learning_rate": 2.1769653701284983e-06,
      "loss": 0.6563,
      "step": 11550
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.9649640321731567,
      "learning_rate": 2.1679688153758373e-06,
      "loss": 0.6479,
      "step": 11560
    },
    {
      "epoch": 0.77,
      "grad_norm": 2.096909523010254,
      "learning_rate": 2.1589857400870804e-06,
      "loss": 0.642,
      "step": 11570
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.91984224319458,
      "learning_rate": 2.150016187018485e-06,
      "loss": 0.656,
      "step": 11580
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.696636438369751,
      "learning_rate": 2.1410601988619394e-06,
      "loss": 0.6838,
      "step": 11590
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.7546021938323975,
      "learning_rate": 2.132117818244771e-06,
      "loss": 0.6651,
      "step": 11600
    },
    {
      "epoch": 0.77,
      "grad_norm": 2.2934517860412598,
      "learning_rate": 2.1231890877295374e-06,
      "loss": 0.6734,
      "step": 11610
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.7087199687957764,
      "learning_rate": 2.1142740498138327e-06,
      "loss": 0.6764,
      "step": 11620
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.9262723922729492,
      "learning_rate": 2.105372746930073e-06,
      "loss": 0.6512,
      "step": 11630
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.9107701778411865,
      "learning_rate": 2.096485221445301e-06,
      "loss": 0.6493,
      "step": 11640
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.5462172031402588,
      "learning_rate": 2.08761151566099e-06,
      "loss": 0.6609,
      "step": 11650
    },
    {
      "epoch": 0.77,
      "grad_norm": 2.213968515396118,
      "learning_rate": 2.0787516718128294e-06,
      "loss": 0.6764,
      "step": 11660
    },
    {
      "epoch": 0.77,
      "grad_norm": 2.0844807624816895,
      "learning_rate": 2.0699057320705328e-06,
      "loss": 0.6553,
      "step": 11670
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.6982314586639404,
      "learning_rate": 2.061073738537635e-06,
      "loss": 0.6562,
      "step": 11680
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.8195136785507202,
      "learning_rate": 2.0522557332512953e-06,
      "loss": 0.6799,
      "step": 11690
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.861391305923462,
      "learning_rate": 2.0434517581820893e-06,
      "loss": 0.6695,
      "step": 11700
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.5561097860336304,
      "learning_rate": 2.034661855233815e-06,
      "loss": 0.6468,
      "step": 11710
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.8640862703323364,
      "learning_rate": 2.0258860662432946e-06,
      "loss": 0.6619,
      "step": 11720
    },
    {
      "epoch": 0.78,
      "grad_norm": 2.2300779819488525,
      "learning_rate": 2.0171244329801677e-06,
      "loss": 0.6681,
      "step": 11730
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.967496633529663,
      "learning_rate": 2.008376997146705e-06,
      "loss": 0.6359,
      "step": 11740
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.6537777185440063,
      "learning_rate": 1.999643800377596e-06,
      "loss": 0.6803,
      "step": 11750
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.5698853731155396,
      "learning_rate": 1.990924884239758e-06,
      "loss": 0.6403,
      "step": 11760
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.752765417098999,
      "learning_rate": 1.982220290232143e-06,
      "loss": 0.6745,
      "step": 11770
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.6937237977981567,
      "learning_rate": 1.9735300597855287e-06,
      "loss": 0.6675,
      "step": 11780
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.685810923576355,
      "learning_rate": 1.9648542342623276e-06,
      "loss": 0.6527,
      "step": 11790
    },
    {
      "epoch": 0.78,
      "grad_norm": 2.0320212841033936,
      "learning_rate": 1.956192854956397e-06,
      "loss": 0.6843,
      "step": 11800
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.7817845344543457,
      "learning_rate": 1.9475459630928263e-06,
      "loss": 0.667,
      "step": 11810
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.8626519441604614,
      "learning_rate": 1.938913599827753e-06,
      "loss": 0.6823,
      "step": 11820
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.6484193801879883,
      "learning_rate": 1.9302958062481673e-06,
      "loss": 0.6713,
      "step": 11830
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.940313696861267,
      "learning_rate": 1.9216926233717087e-06,
      "loss": 0.6425,
      "step": 11840
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.8807305097579956,
      "learning_rate": 1.913104092146476e-06,
      "loss": 0.6591,
      "step": 11850
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.7303801774978638,
      "learning_rate": 1.9045302534508298e-06,
      "loss": 0.6569,
      "step": 11860
    },
    {
      "epoch": 0.79,
      "grad_norm": 2.3046600818634033,
      "learning_rate": 1.8959711480932042e-06,
      "loss": 0.6785,
      "step": 11870
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.4674731492996216,
      "learning_rate": 1.887426816811903e-06,
      "loss": 0.6738,
      "step": 11880
    },
    {
      "epoch": 0.79,
      "grad_norm": 2.161259651184082,
      "learning_rate": 1.8788973002749112e-06,
      "loss": 0.6712,
      "step": 11890
    },
    {
      "epoch": 0.79,
      "grad_norm": 2.3141844272613525,
      "learning_rate": 1.8703826390797047e-06,
      "loss": 0.6787,
      "step": 11900
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.51625657081604,
      "learning_rate": 1.8618828737530497e-06,
      "loss": 0.6646,
      "step": 11910
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.8078784942626953,
      "learning_rate": 1.8533980447508138e-06,
      "loss": 0.6605,
      "step": 11920
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.5935428142547607,
      "learning_rate": 1.8449281924577716e-06,
      "loss": 0.681,
      "step": 11930
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.6151143312454224,
      "learning_rate": 1.83647335718742e-06,
      "loss": 0.6732,
      "step": 11940
    },
    {
      "epoch": 0.79,
      "grad_norm": 2.1067352294921875,
      "learning_rate": 1.8280335791817733e-06,
      "loss": 0.6642,
      "step": 11950
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.8711411952972412,
      "learning_rate": 1.8196088986111798e-06,
      "loss": 0.6602,
      "step": 11960
    },
    {
      "epoch": 0.79,
      "grad_norm": 2.235433578491211,
      "learning_rate": 1.8111993555741342e-06,
      "loss": 0.662,
      "step": 11970
    },
    {
      "epoch": 0.79,
      "grad_norm": 2.0249688625335693,
      "learning_rate": 1.8028049900970768e-06,
      "loss": 0.6615,
      "step": 11980
    },
    {
      "epoch": 0.79,
      "grad_norm": 2.3812663555145264,
      "learning_rate": 1.7944258421342097e-06,
      "loss": 0.6833,
      "step": 11990
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.5996013879776,
      "learning_rate": 1.7860619515673034e-06,
      "loss": 0.6463,
      "step": 12000
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.5530682802200317,
      "learning_rate": 1.777713358205514e-06,
      "loss": 0.6812,
      "step": 12010
    },
    {
      "epoch": 0.8,
      "grad_norm": 2.069619655609131,
      "learning_rate": 1.7693801017851818e-06,
      "loss": 0.6601,
      "step": 12020
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.6531022787094116,
      "learning_rate": 1.761062221969651e-06,
      "loss": 0.65,
      "step": 12030
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.5818767547607422,
      "learning_rate": 1.7527597583490825e-06,
      "loss": 0.6664,
      "step": 12040
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.9506007432937622,
      "learning_rate": 1.7444727504402554e-06,
      "loss": 0.6709,
      "step": 12050
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.5742695331573486,
      "learning_rate": 1.736201237686389e-06,
      "loss": 0.6492,
      "step": 12060
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.3699088096618652,
      "learning_rate": 1.7279452594569484e-06,
      "loss": 0.6909,
      "step": 12070
    },
    {
      "epoch": 0.8,
      "grad_norm": 2.039612293243408,
      "learning_rate": 1.7197048550474643e-06,
      "loss": 0.6643,
      "step": 12080
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.7419607639312744,
      "learning_rate": 1.7114800636793378e-06,
      "loss": 0.6795,
      "step": 12090
    },
    {
      "epoch": 0.8,
      "grad_norm": 3.1645009517669678,
      "learning_rate": 1.7032709244996559e-06,
      "loss": 0.6452,
      "step": 12100
    },
    {
      "epoch": 0.8,
      "grad_norm": 2.338763952255249,
      "learning_rate": 1.695077476581013e-06,
      "loss": 0.6519,
      "step": 12110
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.672167420387268,
      "learning_rate": 1.6868997589213138e-06,
      "loss": 0.6373,
      "step": 12120
    },
    {
      "epoch": 0.8,
      "grad_norm": 2.539010763168335,
      "learning_rate": 1.6787378104435931e-06,
      "loss": 0.6561,
      "step": 12130
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.6293374300003052,
      "learning_rate": 1.6705916699958292e-06,
      "loss": 0.6846,
      "step": 12140
    },
    {
      "epoch": 0.8,
      "grad_norm": 2.007550001144409,
      "learning_rate": 1.662461376350764e-06,
      "loss": 0.6662,
      "step": 12150
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.6836156845092773,
      "learning_rate": 1.6543469682057105e-06,
      "loss": 0.6571,
      "step": 12160
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.7401918172836304,
      "learning_rate": 1.6462484841823712e-06,
      "loss": 0.6443,
      "step": 12170
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.6568111181259155,
      "learning_rate": 1.6381659628266589e-06,
      "loss": 0.6764,
      "step": 12180
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.7030593156814575,
      "learning_rate": 1.6300994426085103e-06,
      "loss": 0.6521,
      "step": 12190
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.7264918088912964,
      "learning_rate": 1.6220489619216988e-06,
      "loss": 0.6512,
      "step": 12200
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.9158384799957275,
      "learning_rate": 1.6140145590836554e-06,
      "loss": 0.6806,
      "step": 12210
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.5750938653945923,
      "learning_rate": 1.6059962723352912e-06,
      "loss": 0.6463,
      "step": 12220
    },
    {
      "epoch": 0.81,
      "grad_norm": 2.226882219314575,
      "learning_rate": 1.5979941398408045e-06,
      "loss": 0.6607,
      "step": 12230
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.5629756450653076,
      "learning_rate": 1.5900081996875083e-06,
      "loss": 0.6355,
      "step": 12240
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.8792222738265991,
      "learning_rate": 1.5820384898856433e-06,
      "loss": 0.6825,
      "step": 12250
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.9516578912734985,
      "learning_rate": 1.574085048368204e-06,
      "loss": 0.639,
      "step": 12260
    },
    {
      "epoch": 0.81,
      "grad_norm": 2.400202512741089,
      "learning_rate": 1.5661479129907508e-06,
      "loss": 0.6697,
      "step": 12270
    },
    {
      "epoch": 0.81,
      "grad_norm": 2.255854606628418,
      "learning_rate": 1.5582271215312294e-06,
      "loss": 0.6621,
      "step": 12280
    },
    {
      "epoch": 0.81,
      "grad_norm": 2.195754289627075,
      "learning_rate": 1.5503227116898017e-06,
      "loss": 0.6412,
      "step": 12290
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.8883105516433716,
      "learning_rate": 1.5424347210886538e-06,
      "loss": 0.6692,
      "step": 12300
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.464462161064148,
      "learning_rate": 1.5345631872718214e-06,
      "loss": 0.6669,
      "step": 12310
    },
    {
      "epoch": 0.82,
      "grad_norm": 2.4194180965423584,
      "learning_rate": 1.5267081477050132e-06,
      "loss": 0.6795,
      "step": 12320
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.6516296863555908,
      "learning_rate": 1.5188696397754344e-06,
      "loss": 0.6681,
      "step": 12330
    },
    {
      "epoch": 0.82,
      "grad_norm": 2.128793716430664,
      "learning_rate": 1.5110477007916002e-06,
      "loss": 0.6767,
      "step": 12340
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.7647573947906494,
      "learning_rate": 1.5032423679831642e-06,
      "loss": 0.6705,
      "step": 12350
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.5265520811080933,
      "learning_rate": 1.4954536785007456e-06,
      "loss": 0.662,
      "step": 12360
    },
    {
      "epoch": 0.82,
      "grad_norm": 2.2167117595672607,
      "learning_rate": 1.487681669415742e-06,
      "loss": 0.6628,
      "step": 12370
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.4803221225738525,
      "learning_rate": 1.4799263777201594e-06,
      "loss": 0.6688,
      "step": 12380
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.5486646890640259,
      "learning_rate": 1.4721878403264344e-06,
      "loss": 0.66,
      "step": 12390
    },
    {
      "epoch": 0.82,
      "grad_norm": 2.1329715251922607,
      "learning_rate": 1.4644660940672628e-06,
      "loss": 0.6564,
      "step": 12400
    },
    {
      "epoch": 0.82,
      "grad_norm": 2.2044312953948975,
      "learning_rate": 1.456761175695417e-06,
      "loss": 0.6582,
      "step": 12410
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.7977498769760132,
      "learning_rate": 1.449073121883573e-06,
      "loss": 0.6713,
      "step": 12420
    },
    {
      "epoch": 0.82,
      "grad_norm": 2.147284507751465,
      "learning_rate": 1.4414019692241437e-06,
      "loss": 0.6577,
      "step": 12430
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.5342192649841309,
      "learning_rate": 1.433747754229093e-06,
      "loss": 0.6673,
      "step": 12440
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.5152603387832642,
      "learning_rate": 1.4261105133297693e-06,
      "loss": 0.64,
      "step": 12450
    },
    {
      "epoch": 0.82,
      "grad_norm": 2.131248950958252,
      "learning_rate": 1.4184902828767288e-06,
      "loss": 0.6667,
      "step": 12460
    },
    {
      "epoch": 0.83,
      "grad_norm": 2.3400142192840576,
      "learning_rate": 1.410887099139569e-06,
      "loss": 0.6758,
      "step": 12470
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.8582547903060913,
      "learning_rate": 1.4033009983067454e-06,
      "loss": 0.6707,
      "step": 12480
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.4715774059295654,
      "learning_rate": 1.395732016485406e-06,
      "loss": 0.6738,
      "step": 12490
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.6715984344482422,
      "learning_rate": 1.3881801897012225e-06,
      "loss": 0.6513,
      "step": 12500
    },
    {
      "epoch": 0.83,
      "grad_norm": 2.1860718727111816,
      "learning_rate": 1.3806455538982106e-06,
      "loss": 0.6655,
      "step": 12510
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.513540506362915,
      "learning_rate": 1.373128144938563e-06,
      "loss": 0.6441,
      "step": 12520
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.9079211950302124,
      "learning_rate": 1.3656279986024802e-06,
      "loss": 0.6635,
      "step": 12530
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.9643515348434448,
      "learning_rate": 1.3581451505879995e-06,
      "loss": 0.6565,
      "step": 12540
    },
    {
      "epoch": 0.83,
      "grad_norm": 2.0323612689971924,
      "learning_rate": 1.3506796365108232e-06,
      "loss": 0.7029,
      "step": 12550
    },
    {
      "epoch": 0.83,
      "grad_norm": 2.054985761642456,
      "learning_rate": 1.3432314919041478e-06,
      "loss": 0.6354,
      "step": 12560
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.635546326637268,
      "learning_rate": 1.3358007522185035e-06,
      "loss": 0.6554,
      "step": 12570
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.7966934442520142,
      "learning_rate": 1.3283874528215735e-06,
      "loss": 0.653,
      "step": 12580
    },
    {
      "epoch": 0.83,
      "grad_norm": 2.113840341567993,
      "learning_rate": 1.3209916289980336e-06,
      "loss": 0.6667,
      "step": 12590
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.962321400642395,
      "learning_rate": 1.3136133159493803e-06,
      "loss": 0.6496,
      "step": 12600
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.854382038116455,
      "learning_rate": 1.30625254879377e-06,
      "loss": 0.6693,
      "step": 12610
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.7110834121704102,
      "learning_rate": 1.2989093625658411e-06,
      "loss": 0.684,
      "step": 12620
    },
    {
      "epoch": 0.84,
      "grad_norm": 2.151606798171997,
      "learning_rate": 1.2915837922165547e-06,
      "loss": 0.6638,
      "step": 12630
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.8074573278427124,
      "learning_rate": 1.2842758726130283e-06,
      "loss": 0.6783,
      "step": 12640
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.9045308828353882,
      "learning_rate": 1.2769856385383689e-06,
      "loss": 0.6646,
      "step": 12650
    },
    {
      "epoch": 0.84,
      "grad_norm": 2.033857583999634,
      "learning_rate": 1.269713124691503e-06,
      "loss": 0.6633,
      "step": 12660
    },
    {
      "epoch": 0.84,
      "grad_norm": 2.591675281524658,
      "learning_rate": 1.2624583656870153e-06,
      "loss": 0.6656,
      "step": 12670
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.8951250314712524,
      "learning_rate": 1.2552213960549891e-06,
      "loss": 0.6642,
      "step": 12680
    },
    {
      "epoch": 0.84,
      "grad_norm": 2.0589277744293213,
      "learning_rate": 1.2480022502408306e-06,
      "loss": 0.6691,
      "step": 12690
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.7472984790802002,
      "learning_rate": 1.2408009626051137e-06,
      "loss": 0.6548,
      "step": 12700
    },
    {
      "epoch": 0.84,
      "grad_norm": 2.1019725799560547,
      "learning_rate": 1.2336175674234112e-06,
      "loss": 0.6853,
      "step": 12710
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.7122173309326172,
      "learning_rate": 1.22645209888614e-06,
      "loss": 0.6643,
      "step": 12720
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.4647725820541382,
      "learning_rate": 1.2193045910983864e-06,
      "loss": 0.6637,
      "step": 12730
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.424286127090454,
      "learning_rate": 1.2121750780797514e-06,
      "loss": 0.6656,
      "step": 12740
    },
    {
      "epoch": 0.84,
      "grad_norm": 2.3147671222686768,
      "learning_rate": 1.2050635937641909e-06,
      "loss": 0.6865,
      "step": 12750
    },
    {
      "epoch": 0.84,
      "grad_norm": 2.1018872261047363,
      "learning_rate": 1.1979701719998454e-06,
      "loss": 0.6708,
      "step": 12760
    },
    {
      "epoch": 0.85,
      "grad_norm": 2.553410291671753,
      "learning_rate": 1.1908948465488878e-06,
      "loss": 0.6679,
      "step": 12770
    },
    {
      "epoch": 0.85,
      "grad_norm": 2.1423566341400146,
      "learning_rate": 1.1838376510873557e-06,
      "loss": 0.6652,
      "step": 12780
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.8816630840301514,
      "learning_rate": 1.1767986192049986e-06,
      "loss": 0.6632,
      "step": 12790
    },
    {
      "epoch": 0.85,
      "grad_norm": 2.626129627227783,
      "learning_rate": 1.1697777844051105e-06,
      "loss": 0.6678,
      "step": 12800
    },
    {
      "epoch": 0.85,
      "eval_loss": 0.7841401100158691,
      "eval_runtime": 134.3092,
      "eval_samples_per_second": 81.901,
      "eval_steps_per_second": 10.238,
      "step": 12800
    },
    {
      "epoch": 0.85,
      "grad_norm": 2.2205734252929688,
      "learning_rate": 1.1627751801043736e-06,
      "loss": 0.6638,
      "step": 12810
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.9543955326080322,
      "learning_rate": 1.1557908396327028e-06,
      "loss": 0.6792,
      "step": 12820
    },
    {
      "epoch": 0.85,
      "grad_norm": 2.1611430644989014,
      "learning_rate": 1.14882479623308e-06,
      "loss": 0.667,
      "step": 12830
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.858528971672058,
      "learning_rate": 1.1418770830614012e-06,
      "loss": 0.6918,
      "step": 12840
    },
    {
      "epoch": 0.85,
      "grad_norm": 2.0636982917785645,
      "learning_rate": 1.134947733186315e-06,
      "loss": 0.664,
      "step": 12850
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.9815447330474854,
      "learning_rate": 1.1280367795890724e-06,
      "loss": 0.6562,
      "step": 12860
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.684668779373169,
      "learning_rate": 1.1211442551633595e-06,
      "loss": 0.6825,
      "step": 12870
    },
    {
      "epoch": 0.85,
      "grad_norm": 2.0503435134887695,
      "learning_rate": 1.1142701927151456e-06,
      "loss": 0.6696,
      "step": 12880
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.9878673553466797,
      "learning_rate": 1.1074146249625334e-06,
      "loss": 0.6593,
      "step": 12890
    },
    {
      "epoch": 0.85,
      "grad_norm": 2.233013391494751,
      "learning_rate": 1.100577584535592e-06,
      "loss": 0.6665,
      "step": 12900
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.5695899724960327,
      "learning_rate": 1.0937591039762086e-06,
      "loss": 0.6692,
      "step": 12910
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.6886929273605347,
      "learning_rate": 1.0869592157379305e-06,
      "loss": 0.6749,
      "step": 12920
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.81573486328125,
      "learning_rate": 1.0801779521858175e-06,
      "loss": 0.642,
      "step": 12930
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.753726840019226,
      "learning_rate": 1.0734153455962765e-06,
      "loss": 0.6506,
      "step": 12940
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.8571149110794067,
      "learning_rate": 1.0666714281569152e-06,
      "loss": 0.6661,
      "step": 12950
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.9926731586456299,
      "learning_rate": 1.0599462319663906e-06,
      "loss": 0.6618,
      "step": 12960
    },
    {
      "epoch": 0.86,
      "grad_norm": 2.0180611610412598,
      "learning_rate": 1.0532397890342506e-06,
      "loss": 0.6381,
      "step": 12970
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.8695250749588013,
      "learning_rate": 1.0465521312807846e-06,
      "loss": 0.6609,
      "step": 12980
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.7631117105484009,
      "learning_rate": 1.0398832905368693e-06,
      "loss": 0.6786,
      "step": 12990
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.9388374090194702,
      "learning_rate": 1.0332332985438248e-06,
      "loss": 0.6765,
      "step": 13000
    },
    {
      "epoch": 0.86,
      "grad_norm": 2.051642417907715,
      "learning_rate": 1.0266021869532527e-06,
      "loss": 0.6743,
      "step": 13010
    },
    {
      "epoch": 0.86,
      "grad_norm": 2.41388201713562,
      "learning_rate": 1.0199899873268903e-06,
      "loss": 0.6845,
      "step": 13020
    },
    {
      "epoch": 0.86,
      "grad_norm": 2.17926025390625,
      "learning_rate": 1.013396731136465e-06,
      "loss": 0.6765,
      "step": 13030
    },
    {
      "epoch": 0.86,
      "grad_norm": 2.158048391342163,
      "learning_rate": 1.006822449763537e-06,
      "loss": 0.66,
      "step": 13040
    },
    {
      "epoch": 0.86,
      "grad_norm": 2.1093921661376953,
      "learning_rate": 1.0002671744993519e-06,
      "loss": 0.6716,
      "step": 13050
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.8924907445907593,
      "learning_rate": 9.937309365446973e-07,
      "loss": 0.6559,
      "step": 13060
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.9550564289093018,
      "learning_rate": 9.872137670097465e-07,
      "loss": 0.6692,
      "step": 13070
    },
    {
      "epoch": 0.87,
      "grad_norm": 2.1951191425323486,
      "learning_rate": 9.807156969139136e-07,
      "loss": 0.6782,
      "step": 13080
    },
    {
      "epoch": 0.87,
      "grad_norm": 2.0226147174835205,
      "learning_rate": 9.742367571857092e-07,
      "loss": 0.6669,
      "step": 13090
    },
    {
      "epoch": 0.87,
      "grad_norm": 2.125762701034546,
      "learning_rate": 9.677769786625869e-07,
      "loss": 0.6674,
      "step": 13100
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.7338881492614746,
      "learning_rate": 9.613363920908025e-07,
      "loss": 0.6481,
      "step": 13110
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.5108554363250732,
      "learning_rate": 9.549150281252633e-07,
      "loss": 0.6305,
      "step": 13120
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.688029408454895,
      "learning_rate": 9.485129173293823e-07,
      "loss": 0.6535,
      "step": 13130
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.406068205833435,
      "learning_rate": 9.421300901749386e-07,
      "loss": 0.6413,
      "step": 13140
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.9008201360702515,
      "learning_rate": 9.357665770419244e-07,
      "loss": 0.6724,
      "step": 13150
    },
    {
      "epoch": 0.87,
      "grad_norm": 2.259673833847046,
      "learning_rate": 9.294224082184045e-07,
      "loss": 0.6548,
      "step": 13160
    },
    {
      "epoch": 0.87,
      "grad_norm": 2.195957660675049,
      "learning_rate": 9.230976139003717e-07,
      "loss": 0.6643,
      "step": 13170
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.6432466506958008,
      "learning_rate": 9.167922241916055e-07,
      "loss": 0.6644,
      "step": 13180
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.8441758155822754,
      "learning_rate": 9.105062691035233e-07,
      "loss": 0.6508,
      "step": 13190
    },
    {
      "epoch": 0.87,
      "grad_norm": 2.337584972381592,
      "learning_rate": 9.042397785550405e-07,
      "loss": 0.6821,
      "step": 13200
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.7856107950210571,
      "learning_rate": 8.979927823724321e-07,
      "loss": 0.6543,
      "step": 13210
    },
    {
      "epoch": 0.87,
      "grad_norm": 2.0911343097686768,
      "learning_rate": 8.917653102891822e-07,
      "loss": 0.6605,
      "step": 13220
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.6215004920959473,
      "learning_rate": 8.855573919458494e-07,
      "loss": 0.6685,
      "step": 13230
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.819148302078247,
      "learning_rate": 8.793690568899216e-07,
      "loss": 0.6603,
      "step": 13240
    },
    {
      "epoch": 0.88,
      "grad_norm": 2.3011019229888916,
      "learning_rate": 8.732003345756812e-07,
      "loss": 0.6738,
      "step": 13250
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.663140892982483,
      "learning_rate": 8.670512543640574e-07,
      "loss": 0.6618,
      "step": 13260
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.843697190284729,
      "learning_rate": 8.609218455224893e-07,
      "loss": 0.6638,
      "step": 13270
    },
    {
      "epoch": 0.88,
      "grad_norm": 2.5996809005737305,
      "learning_rate": 8.54812137224792e-07,
      "loss": 0.6497,
      "step": 13280
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.8779548406600952,
      "learning_rate": 8.487221585510075e-07,
      "loss": 0.6613,
      "step": 13290
    },
    {
      "epoch": 0.88,
      "grad_norm": 2.365200996398926,
      "learning_rate": 8.426519384872733e-07,
      "loss": 0.665,
      "step": 13300
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.9242346286773682,
      "learning_rate": 8.366015059256871e-07,
      "loss": 0.6502,
      "step": 13310
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.5760527849197388,
      "learning_rate": 8.305708896641596e-07,
      "loss": 0.6626,
      "step": 13320
    },
    {
      "epoch": 0.88,
      "grad_norm": 2.014904022216797,
      "learning_rate": 8.245601184062851e-07,
      "loss": 0.6584,
      "step": 13330
    },
    {
      "epoch": 0.88,
      "grad_norm": 2.12705135345459,
      "learning_rate": 8.185692207612023e-07,
      "loss": 0.6799,
      "step": 13340
    },
    {
      "epoch": 0.88,
      "grad_norm": 2.026439905166626,
      "learning_rate": 8.125982252434611e-07,
      "loss": 0.6819,
      "step": 13350
    },
    {
      "epoch": 0.88,
      "grad_norm": 2.0270392894744873,
      "learning_rate": 8.066471602728804e-07,
      "loss": 0.6701,
      "step": 13360
    },
    {
      "epoch": 0.88,
      "grad_norm": 2.5948376655578613,
      "learning_rate": 8.007160541744186e-07,
      "loss": 0.657,
      "step": 13370
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.7466539144515991,
      "learning_rate": 7.94804935178038e-07,
      "loss": 0.665,
      "step": 13380
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.7160353660583496,
      "learning_rate": 7.88913831418568e-07,
      "loss": 0.6628,
      "step": 13390
    },
    {
      "epoch": 0.89,
      "grad_norm": 2.2262911796569824,
      "learning_rate": 7.830427709355726e-07,
      "loss": 0.6555,
      "step": 13400
    },
    {
      "epoch": 0.89,
      "grad_norm": 2.0489237308502197,
      "learning_rate": 7.771917816732161e-07,
      "loss": 0.6678,
      "step": 13410
    },
    {
      "epoch": 0.89,
      "grad_norm": 2.559640645980835,
      "learning_rate": 7.71360891480134e-07,
      "loss": 0.6431,
      "step": 13420
    },
    {
      "epoch": 0.89,
      "grad_norm": 2.476576089859009,
      "learning_rate": 7.655501281092953e-07,
      "loss": 0.6667,
      "step": 13430
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.941450595855713,
      "learning_rate": 7.597595192178702e-07,
      "loss": 0.6734,
      "step": 13440
    },
    {
      "epoch": 0.89,
      "grad_norm": 2.5670228004455566,
      "learning_rate": 7.539890923671061e-07,
      "loss": 0.6781,
      "step": 13450
    },
    {
      "epoch": 0.89,
      "grad_norm": 2.5730628967285156,
      "learning_rate": 7.482388750221864e-07,
      "loss": 0.6738,
      "step": 13460
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.977264165878296,
      "learning_rate": 7.425088945521064e-07,
      "loss": 0.6633,
      "step": 13470
    },
    {
      "epoch": 0.89,
      "grad_norm": 2.168454647064209,
      "learning_rate": 7.367991782295392e-07,
      "loss": 0.6604,
      "step": 13480
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.8109989166259766,
      "learning_rate": 7.311097532307121e-07,
      "loss": 0.6542,
      "step": 13490
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.8640021085739136,
      "learning_rate": 7.254406466352682e-07,
      "loss": 0.6736,
      "step": 13500
    },
    {
      "epoch": 0.89,
      "grad_norm": 2.002460241317749,
      "learning_rate": 7.197918854261432e-07,
      "loss": 0.6951,
      "step": 13510
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.855389952659607,
      "learning_rate": 7.141634964894389e-07,
      "loss": 0.6749,
      "step": 13520
    },
    {
      "epoch": 0.9,
      "grad_norm": 2.7581233978271484,
      "learning_rate": 7.085555066142886e-07,
      "loss": 0.6567,
      "step": 13530
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.7739157676696777,
      "learning_rate": 7.029679424927366e-07,
      "loss": 0.6697,
      "step": 13540
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.4929020404815674,
      "learning_rate": 6.974008307196057e-07,
      "loss": 0.6861,
      "step": 13550
    },
    {
      "epoch": 0.9,
      "grad_norm": 2.034844398498535,
      "learning_rate": 6.918541977923709e-07,
      "loss": 0.6607,
      "step": 13560
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.5017189979553223,
      "learning_rate": 6.863280701110409e-07,
      "loss": 0.6726,
      "step": 13570
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.9159743785858154,
      "learning_rate": 6.808224739780217e-07,
      "loss": 0.6781,
      "step": 13580
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.910868763923645,
      "learning_rate": 6.753374355979975e-07,
      "loss": 0.6363,
      "step": 13590
    },
    {
      "epoch": 0.9,
      "grad_norm": 2.351539373397827,
      "learning_rate": 6.698729810778065e-07,
      "loss": 0.6571,
      "step": 13600
    },
    {
      "epoch": 0.9,
      "grad_norm": 3.010953664779663,
      "learning_rate": 6.644291364263139e-07,
      "loss": 0.6714,
      "step": 13610
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.7596651315689087,
      "learning_rate": 6.590059275542882e-07,
      "loss": 0.6746,
      "step": 13620
    },
    {
      "epoch": 0.9,
      "grad_norm": 2.1174886226654053,
      "learning_rate": 6.536033802742814e-07,
      "loss": 0.6606,
      "step": 13630
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.700536847114563,
      "learning_rate": 6.482215203005016e-07,
      "loss": 0.6512,
      "step": 13640
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.9607415199279785,
      "learning_rate": 6.428603732486938e-07,
      "loss": 0.6769,
      "step": 13650
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.9792729616165161,
      "learning_rate": 6.375199646360142e-07,
      "loss": 0.6615,
      "step": 13660
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.2958414554595947,
      "learning_rate": 6.322003198809162e-07,
      "loss": 0.6879,
      "step": 13670
    },
    {
      "epoch": 0.91,
      "grad_norm": 2.095305919647217,
      "learning_rate": 6.269014643030214e-07,
      "loss": 0.6753,
      "step": 13680
    },
    {
      "epoch": 0.91,
      "grad_norm": 2.2841269969940186,
      "learning_rate": 6.216234231230012e-07,
      "loss": 0.6827,
      "step": 13690
    },
    {
      "epoch": 0.91,
      "grad_norm": 1.9332698583602905,
      "learning_rate": 6.163662214624616e-07,
      "loss": 0.6721,
      "step": 13700
    },
    {
      "epoch": 0.91,
      "grad_norm": 2.1489346027374268,
      "learning_rate": 6.111298843438169e-07,
      "loss": 0.6442,
      "step": 13710
    },
    {
      "epoch": 0.91,
      "grad_norm": 1.6940524578094482,
      "learning_rate": 6.059144366901737e-07,
      "loss": 0.6637,
      "step": 13720
    },
    {
      "epoch": 0.91,
      "grad_norm": 1.6115096807479858,
      "learning_rate": 6.007199033252131e-07,
      "loss": 0.669,
      "step": 13730
    },
    {
      "epoch": 0.91,
      "grad_norm": 2.0993459224700928,
      "learning_rate": 5.955463089730723e-07,
      "loss": 0.6644,
      "step": 13740
    },
    {
      "epoch": 0.91,
      "grad_norm": 1.9472441673278809,
      "learning_rate": 5.903936782582253e-07,
      "loss": 0.6715,
      "step": 13750
    },
    {
      "epoch": 0.91,
      "grad_norm": 2.4469716548919678,
      "learning_rate": 5.852620357053651e-07,
      "loss": 0.6633,
      "step": 13760
    },
    {
      "epoch": 0.91,
      "grad_norm": 2.319772481918335,
      "learning_rate": 5.80151405739292e-07,
      "loss": 0.689,
      "step": 13770
    },
    {
      "epoch": 0.91,
      "grad_norm": 1.8905223608016968,
      "learning_rate": 5.750618126847912e-07,
      "loss": 0.6469,
      "step": 13780
    },
    {
      "epoch": 0.91,
      "grad_norm": 1.8720320463180542,
      "learning_rate": 5.699932807665198e-07,
      "loss": 0.6647,
      "step": 13790
    },
    {
      "epoch": 0.91,
      "grad_norm": 1.8318226337432861,
      "learning_rate": 5.649458341088915e-07,
      "loss": 0.6804,
      "step": 13800
    },
    {
      "epoch": 0.91,
      "grad_norm": 2.6785199642181396,
      "learning_rate": 5.599194967359639e-07,
      "loss": 0.6481,
      "step": 13810
    },
    {
      "epoch": 0.91,
      "grad_norm": 2.7276508808135986,
      "learning_rate": 5.549142925713186e-07,
      "loss": 0.6818,
      "step": 13820
    },
    {
      "epoch": 0.92,
      "grad_norm": 2.1352598667144775,
      "learning_rate": 5.499302454379512e-07,
      "loss": 0.6594,
      "step": 13830
    },
    {
      "epoch": 0.92,
      "grad_norm": 1.9602961540222168,
      "learning_rate": 5.449673790581611e-07,
      "loss": 0.6741,
      "step": 13840
    },
    {
      "epoch": 0.92,
      "grad_norm": 2.205834150314331,
      "learning_rate": 5.400257170534296e-07,
      "loss": 0.6669,
      "step": 13850
    },
    {
      "epoch": 0.92,
      "grad_norm": 2.2802295684814453,
      "learning_rate": 5.351052829443159e-07,
      "loss": 0.6576,
      "step": 13860
    },
    {
      "epoch": 0.92,
      "grad_norm": 1.6849833726882935,
      "learning_rate": 5.302061001503395e-07,
      "loss": 0.6648,
      "step": 13870
    },
    {
      "epoch": 0.92,
      "grad_norm": 2.004652976989746,
      "learning_rate": 5.253281919898751e-07,
      "loss": 0.6302,
      "step": 13880
    },
    {
      "epoch": 0.92,
      "grad_norm": 1.8887594938278198,
      "learning_rate": 5.204715816800343e-07,
      "loss": 0.6642,
      "step": 13890
    },
    {
      "epoch": 0.92,
      "grad_norm": 2.5696282386779785,
      "learning_rate": 5.156362923365587e-07,
      "loss": 0.6419,
      "step": 13900
    },
    {
      "epoch": 0.92,
      "grad_norm": 1.7427834272384644,
      "learning_rate": 5.108223469737117e-07,
      "loss": 0.6784,
      "step": 13910
    },
    {
      "epoch": 0.92,
      "grad_norm": 2.1054744720458984,
      "learning_rate": 5.06029768504166e-07,
      "loss": 0.6545,
      "step": 13920
    },
    {
      "epoch": 0.92,
      "grad_norm": 2.091437339782715,
      "learning_rate": 5.012585797388936e-07,
      "loss": 0.655,
      "step": 13930
    },
    {
      "epoch": 0.92,
      "grad_norm": 2.2512803077697754,
      "learning_rate": 4.965088033870608e-07,
      "loss": 0.6354,
      "step": 13940
    },
    {
      "epoch": 0.92,
      "grad_norm": 2.2140800952911377,
      "learning_rate": 4.917804620559202e-07,
      "loss": 0.6559,
      "step": 13950
    },
    {
      "epoch": 0.92,
      "grad_norm": 1.8559973239898682,
      "learning_rate": 4.87073578250698e-07,
      "loss": 0.6559,
      "step": 13960
    },
    {
      "epoch": 0.92,
      "grad_norm": 3.173664093017578,
      "learning_rate": 4.823881743744907e-07,
      "loss": 0.6733,
      "step": 13970
    },
    {
      "epoch": 0.93,
      "grad_norm": 2.235106945037842,
      "learning_rate": 4.777242727281594e-07,
      "loss": 0.6404,
      "step": 13980
    },
    {
      "epoch": 0.93,
      "grad_norm": 1.6829653978347778,
      "learning_rate": 4.730818955102234e-07,
      "loss": 0.6621,
      "step": 13990
    },
    {
      "epoch": 0.93,
      "grad_norm": 2.1085221767425537,
      "learning_rate": 4.6846106481675035e-07,
      "loss": 0.6712,
      "step": 14000
    },
    {
      "epoch": 0.93,
      "grad_norm": 2.800367593765259,
      "learning_rate": 4.638618026412539e-07,
      "loss": 0.6713,
      "step": 14010
    },
    {
      "epoch": 0.93,
      "grad_norm": 1.9451186656951904,
      "learning_rate": 4.5928413087459325e-07,
      "loss": 0.664,
      "step": 14020
    },
    {
      "epoch": 0.93,
      "grad_norm": 1.687585711479187,
      "learning_rate": 4.5472807130486075e-07,
      "loss": 0.6699,
      "step": 14030
    },
    {
      "epoch": 0.93,
      "grad_norm": 2.3365931510925293,
      "learning_rate": 4.501936456172845e-07,
      "loss": 0.6572,
      "step": 14040
    },
    {
      "epoch": 0.93,
      "grad_norm": 1.7187267541885376,
      "learning_rate": 4.456808753941205e-07,
      "loss": 0.6675,
      "step": 14050
    },
    {
      "epoch": 0.93,
      "grad_norm": 2.139224052429199,
      "learning_rate": 4.4118978211455723e-07,
      "loss": 0.6786,
      "step": 14060
    },
    {
      "epoch": 0.93,
      "grad_norm": 2.3092565536499023,
      "learning_rate": 4.367203871546039e-07,
      "loss": 0.6549,
      "step": 14070
    },
    {
      "epoch": 0.93,
      "grad_norm": 1.6337780952453613,
      "learning_rate": 4.322727117869951e-07,
      "loss": 0.6408,
      "step": 14080
    },
    {
      "epoch": 0.93,
      "grad_norm": 1.9629242420196533,
      "learning_rate": 4.278467771810896e-07,
      "loss": 0.6744,
      "step": 14090
    },
    {
      "epoch": 0.93,
      "grad_norm": 2.3631951808929443,
      "learning_rate": 4.2344260440276455e-07,
      "loss": 0.6699,
      "step": 14100
    },
    {
      "epoch": 0.93,
      "grad_norm": 1.8263096809387207,
      "learning_rate": 4.1906021441432074e-07,
      "loss": 0.6651,
      "step": 14110
    },
    {
      "epoch": 0.93,
      "grad_norm": 2.167978286743164,
      "learning_rate": 4.146996280743798e-07,
      "loss": 0.6738,
      "step": 14120
    },
    {
      "epoch": 0.94,
      "grad_norm": 1.9163544178009033,
      "learning_rate": 4.103608661377867e-07,
      "loss": 0.6654,
      "step": 14130
    },
    {
      "epoch": 0.94,
      "grad_norm": 3.2101268768310547,
      "learning_rate": 4.0604394925550906e-07,
      "loss": 0.6709,
      "step": 14140
    },
    {
      "epoch": 0.94,
      "grad_norm": 1.9373053312301636,
      "learning_rate": 4.0174889797453875e-07,
      "loss": 0.6675,
      "step": 14150
    },
    {
      "epoch": 0.94,
      "grad_norm": 2.323124647140503,
      "learning_rate": 3.9747573273779816e-07,
      "loss": 0.6869,
      "step": 14160
    },
    {
      "epoch": 0.94,
      "grad_norm": 2.504706621170044,
      "learning_rate": 3.9322447388403796e-07,
      "loss": 0.6576,
      "step": 14170
    },
    {
      "epoch": 0.94,
      "grad_norm": 2.201514959335327,
      "learning_rate": 3.8899514164774166e-07,
      "loss": 0.6594,
      "step": 14180
    },
    {
      "epoch": 0.94,
      "grad_norm": 2.347700834274292,
      "learning_rate": 3.8478775615902965e-07,
      "loss": 0.6843,
      "step": 14190
    },
    {
      "epoch": 0.94,
      "grad_norm": 2.5682637691497803,
      "learning_rate": 3.8060233744356634e-07,
      "loss": 0.6636,
      "step": 14200
    },
    {
      "epoch": 0.94,
      "grad_norm": 2.0093626976013184,
      "learning_rate": 3.7643890542245985e-07,
      "loss": 0.6518,
      "step": 14210
    },
    {
      "epoch": 0.94,
      "grad_norm": 2.4383957386016846,
      "learning_rate": 3.722974799121681e-07,
      "loss": 0.6349,
      "step": 14220
    },
    {
      "epoch": 0.94,
      "grad_norm": 3.2455692291259766,
      "learning_rate": 3.6817808062440953e-07,
      "loss": 0.6564,
      "step": 14230
    },
    {
      "epoch": 0.94,
      "grad_norm": 1.918816328048706,
      "learning_rate": 3.6408072716606346e-07,
      "loss": 0.6615,
      "step": 14240
    },
    {
      "epoch": 0.94,
      "grad_norm": 2.3905696868896484,
      "learning_rate": 3.600054390390778e-07,
      "loss": 0.6458,
      "step": 14250
    },
    {
      "epoch": 0.94,
      "grad_norm": 2.004607677459717,
      "learning_rate": 3.5595223564037884e-07,
      "loss": 0.679,
      "step": 14260
    },
    {
      "epoch": 0.94,
      "grad_norm": 1.7747329473495483,
      "learning_rate": 3.5192113626177806e-07,
      "loss": 0.6708,
      "step": 14270
    },
    {
      "epoch": 0.94,
      "grad_norm": 2.097956657409668,
      "learning_rate": 3.479121600898777e-07,
      "loss": 0.652,
      "step": 14280
    },
    {
      "epoch": 0.95,
      "grad_norm": 2.358074903488159,
      "learning_rate": 3.439253262059822e-07,
      "loss": 0.6811,
      "step": 14290
    },
    {
      "epoch": 0.95,
      "grad_norm": 2.267530679702759,
      "learning_rate": 3.399606535860078e-07,
      "loss": 0.6504,
      "step": 14300
    },
    {
      "epoch": 0.95,
      "grad_norm": 2.4041032791137695,
      "learning_rate": 3.360181611003893e-07,
      "loss": 0.6658,
      "step": 14310
    },
    {
      "epoch": 0.95,
      "grad_norm": 1.7667807340621948,
      "learning_rate": 3.320978675139919e-07,
      "loss": 0.6827,
      "step": 14320
    },
    {
      "epoch": 0.95,
      "grad_norm": 2.0427334308624268,
      "learning_rate": 3.2819979148602245e-07,
      "loss": 0.6617,
      "step": 14330
    },
    {
      "epoch": 0.95,
      "grad_norm": 2.4956207275390625,
      "learning_rate": 3.24323951569942e-07,
      "loss": 0.6386,
      "step": 14340
    },
    {
      "epoch": 0.95,
      "grad_norm": 2.769589900970459,
      "learning_rate": 3.204703662133724e-07,
      "loss": 0.6644,
      "step": 14350
    },
    {
      "epoch": 0.95,
      "grad_norm": 2.495105743408203,
      "learning_rate": 3.166390537580122e-07,
      "loss": 0.6637,
      "step": 14360
    },
    {
      "epoch": 0.95,
      "grad_norm": 2.220623731613159,
      "learning_rate": 3.128300324395517e-07,
      "loss": 0.6637,
      "step": 14370
    },
    {
      "epoch": 0.95,
      "grad_norm": 1.9806053638458252,
      "learning_rate": 3.0904332038757977e-07,
      "loss": 0.652,
      "step": 14380
    },
    {
      "epoch": 0.95,
      "grad_norm": 1.9179275035858154,
      "learning_rate": 3.052789356255037e-07,
      "loss": 0.6586,
      "step": 14390
    },
    {
      "epoch": 0.95,
      "grad_norm": 2.090574026107788,
      "learning_rate": 3.015368960704584e-07,
      "loss": 0.6507,
      "step": 14400
    },
    {
      "epoch": 0.95,
      "eval_loss": 0.7874770760536194,
      "eval_runtime": 133.7124,
      "eval_samples_per_second": 82.266,
      "eval_steps_per_second": 10.283,
      "step": 14400
    },
    {
      "epoch": 0.95,
      "grad_norm": 5.8708295822143555,
      "learning_rate": 2.9781721953322627e-07,
      "loss": 0.6516,
      "step": 14410
    },
    {
      "epoch": 0.95,
      "grad_norm": 1.9250444173812866,
      "learning_rate": 2.9411992371814744e-07,
      "loss": 0.6604,
      "step": 14420
    },
    {
      "epoch": 0.95,
      "grad_norm": 1.5833460092544556,
      "learning_rate": 2.904450262230385e-07,
      "loss": 0.6766,
      "step": 14430
    },
    {
      "epoch": 0.96,
      "grad_norm": 2.0383031368255615,
      "learning_rate": 2.867925445391079e-07,
      "loss": 0.6542,
      "step": 14440
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.9491095542907715,
      "learning_rate": 2.8316249605087386e-07,
      "loss": 0.6496,
      "step": 14450
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.9134398698806763,
      "learning_rate": 2.7955489803607907e-07,
      "loss": 0.6666,
      "step": 14460
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.8715418577194214,
      "learning_rate": 2.7596976766560977e-07,
      "loss": 0.6934,
      "step": 14470
    },
    {
      "epoch": 0.96,
      "grad_norm": 2.399040699005127,
      "learning_rate": 2.724071220034158e-07,
      "loss": 0.6518,
      "step": 14480
    },
    {
      "epoch": 0.96,
      "grad_norm": 2.1483490467071533,
      "learning_rate": 2.688669780064268e-07,
      "loss": 0.6539,
      "step": 14490
    },
    {
      "epoch": 0.96,
      "grad_norm": 2.5796451568603516,
      "learning_rate": 2.653493525244721e-07,
      "loss": 0.6554,
      "step": 14500
    },
    {
      "epoch": 0.96,
      "grad_norm": 2.631964683532715,
      "learning_rate": 2.6185426230020074e-07,
      "loss": 0.6617,
      "step": 14510
    },
    {
      "epoch": 0.96,
      "grad_norm": 2.705181837081909,
      "learning_rate": 2.583817239690034e-07,
      "loss": 0.6644,
      "step": 14520
    },
    {
      "epoch": 0.96,
      "grad_norm": 2.420186758041382,
      "learning_rate": 2.5493175405893076e-07,
      "loss": 0.6518,
      "step": 14530
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.8278932571411133,
      "learning_rate": 2.5150436899061494e-07,
      "loss": 0.6563,
      "step": 14540
    },
    {
      "epoch": 0.96,
      "grad_norm": 2.477051019668579,
      "learning_rate": 2.4809958507719444e-07,
      "loss": 0.6658,
      "step": 14550
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.8384480476379395,
      "learning_rate": 2.447174185242324e-07,
      "loss": 0.6573,
      "step": 14560
    },
    {
      "epoch": 0.96,
      "grad_norm": 2.3078768253326416,
      "learning_rate": 2.413578854296417e-07,
      "loss": 0.6715,
      "step": 14570
    },
    {
      "epoch": 0.96,
      "grad_norm": 2.314739227294922,
      "learning_rate": 2.3802100178360822e-07,
      "loss": 0.6414,
      "step": 14580
    },
    {
      "epoch": 0.97,
      "grad_norm": 1.8191229104995728,
      "learning_rate": 2.3470678346851517e-07,
      "loss": 0.6532,
      "step": 14590
    },
    {
      "epoch": 0.97,
      "grad_norm": 2.201284170150757,
      "learning_rate": 2.314152462588659e-07,
      "loss": 0.6845,
      "step": 14600
    },
    {
      "epoch": 0.97,
      "grad_norm": 1.9875273704528809,
      "learning_rate": 2.2814640582120905e-07,
      "loss": 0.6565,
      "step": 14610
    },
    {
      "epoch": 0.97,
      "grad_norm": 2.196948289871216,
      "learning_rate": 2.2490027771406686e-07,
      "loss": 0.6689,
      "step": 14620
    },
    {
      "epoch": 0.97,
      "grad_norm": 2.2184298038482666,
      "learning_rate": 2.2167687738785748e-07,
      "loss": 0.6884,
      "step": 14630
    },
    {
      "epoch": 0.97,
      "grad_norm": 2.3369548320770264,
      "learning_rate": 2.1847622018482283e-07,
      "loss": 0.6681,
      "step": 14640
    },
    {
      "epoch": 0.97,
      "grad_norm": 2.206387996673584,
      "learning_rate": 2.152983213389559e-07,
      "loss": 0.6722,
      "step": 14650
    },
    {
      "epoch": 0.97,
      "grad_norm": 1.7726385593414307,
      "learning_rate": 2.1214319597592792e-07,
      "loss": 0.6777,
      "step": 14660
    },
    {
      "epoch": 0.97,
      "grad_norm": 2.626243829727173,
      "learning_rate": 2.090108591130169e-07,
      "loss": 0.6818,
      "step": 14670
    },
    {
      "epoch": 0.97,
      "grad_norm": 2.6065804958343506,
      "learning_rate": 2.0590132565903475e-07,
      "loss": 0.6521,
      "step": 14680
    },
    {
      "epoch": 0.97,
      "grad_norm": 3.029127359390259,
      "learning_rate": 2.028146104142581e-07,
      "loss": 0.6711,
      "step": 14690
    },
    {
      "epoch": 0.97,
      "grad_norm": 2.4063072204589844,
      "learning_rate": 1.99750728070357e-07,
      "loss": 0.6603,
      "step": 14700
    },
    {
      "epoch": 0.97,
      "grad_norm": 2.1884777545928955,
      "learning_rate": 1.9670969321032407e-07,
      "loss": 0.6469,
      "step": 14710
    },
    {
      "epoch": 0.97,
      "grad_norm": 1.968603253364563,
      "learning_rate": 1.9369152030840553e-07,
      "loss": 0.6654,
      "step": 14720
    },
    {
      "epoch": 0.97,
      "grad_norm": 2.769115686416626,
      "learning_rate": 1.9069622373003638e-07,
      "loss": 0.6621,
      "step": 14730
    },
    {
      "epoch": 0.98,
      "grad_norm": 2.50546932220459,
      "learning_rate": 1.8772381773176417e-07,
      "loss": 0.6727,
      "step": 14740
    },
    {
      "epoch": 0.98,
      "grad_norm": 1.992620587348938,
      "learning_rate": 1.8477431646118648e-07,
      "loss": 0.6462,
      "step": 14750
    },
    {
      "epoch": 0.98,
      "grad_norm": 1.9770318269729614,
      "learning_rate": 1.8184773395688527e-07,
      "loss": 0.6522,
      "step": 14760
    },
    {
      "epoch": 0.98,
      "grad_norm": 2.0310850143432617,
      "learning_rate": 1.7894408414835362e-07,
      "loss": 0.6826,
      "step": 14770
    },
    {
      "epoch": 0.98,
      "grad_norm": 2.093679904937744,
      "learning_rate": 1.7606338085593532e-07,
      "loss": 0.675,
      "step": 14780
    },
    {
      "epoch": 0.98,
      "grad_norm": 1.967199444770813,
      "learning_rate": 1.7320563779075595e-07,
      "loss": 0.6918,
      "step": 14790
    },
    {
      "epoch": 0.98,
      "grad_norm": 1.9936221837997437,
      "learning_rate": 1.7037086855465902e-07,
      "loss": 0.6603,
      "step": 14800
    },
    {
      "epoch": 0.98,
      "grad_norm": 2.7689085006713867,
      "learning_rate": 1.6755908664014054e-07,
      "loss": 0.659,
      "step": 14810
    },
    {
      "epoch": 0.98,
      "grad_norm": 2.3982479572296143,
      "learning_rate": 1.6477030543028462e-07,
      "loss": 0.6478,
      "step": 14820
    },
    {
      "epoch": 0.98,
      "grad_norm": 2.1643288135528564,
      "learning_rate": 1.6200453819870122e-07,
      "loss": 0.6382,
      "step": 14830
    },
    {
      "epoch": 0.98,
      "grad_norm": 1.9189598560333252,
      "learning_rate": 1.5926179810946185e-07,
      "loss": 0.6403,
      "step": 14840
    },
    {
      "epoch": 0.98,
      "grad_norm": 1.7351678609848022,
      "learning_rate": 1.5654209821703458e-07,
      "loss": 0.6642,
      "step": 14850
    },
    {
      "epoch": 0.98,
      "grad_norm": 1.879950761795044,
      "learning_rate": 1.5384545146622854e-07,
      "loss": 0.6572,
      "step": 14860
    },
    {
      "epoch": 0.98,
      "grad_norm": 2.0498573780059814,
      "learning_rate": 1.511718706921239e-07,
      "loss": 0.6626,
      "step": 14870
    },
    {
      "epoch": 0.98,
      "grad_norm": 2.135221004486084,
      "learning_rate": 1.4852136862001766e-07,
      "loss": 0.662,
      "step": 14880
    },
    {
      "epoch": 0.99,
      "grad_norm": 2.5622806549072266,
      "learning_rate": 1.4589395786535954e-07,
      "loss": 0.6911,
      "step": 14890
    },
    {
      "epoch": 0.99,
      "grad_norm": 3.4293711185455322,
      "learning_rate": 1.4328965093369284e-07,
      "loss": 0.6699,
      "step": 14900
    },
    {
      "epoch": 0.99,
      "grad_norm": 2.0343523025512695,
      "learning_rate": 1.4070846022059437e-07,
      "loss": 0.6785,
      "step": 14910
    },
    {
      "epoch": 0.99,
      "grad_norm": 2.4741287231445312,
      "learning_rate": 1.3815039801161723e-07,
      "loss": 0.6639,
      "step": 14920
    },
    {
      "epoch": 0.99,
      "grad_norm": 1.8715765476226807,
      "learning_rate": 1.3561547648222871e-07,
      "loss": 0.6485,
      "step": 14930
    },
    {
      "epoch": 0.99,
      "grad_norm": 2.449462890625,
      "learning_rate": 1.331037076977576e-07,
      "loss": 0.664,
      "step": 14940
    },
    {
      "epoch": 0.99,
      "grad_norm": 2.398143768310547,
      "learning_rate": 1.3061510361333186e-07,
      "loss": 0.6662,
      "step": 14950
    },
    {
      "epoch": 0.99,
      "grad_norm": 2.0606908798217773,
      "learning_rate": 1.2814967607382433e-07,
      "loss": 0.6564,
      "step": 14960
    },
    {
      "epoch": 0.99,
      "grad_norm": 2.3036439418792725,
      "learning_rate": 1.257074368137945e-07,
      "loss": 0.6524,
      "step": 14970
    },
    {
      "epoch": 0.99,
      "grad_norm": 2.2730531692504883,
      "learning_rate": 1.232883974574367e-07,
      "loss": 0.6413,
      "step": 14980
    },
    {
      "epoch": 0.99,
      "grad_norm": 2.5283987522125244,
      "learning_rate": 1.2089256951851923e-07,
      "loss": 0.6682,
      "step": 14990
    },
    {
      "epoch": 0.99,
      "grad_norm": 2.1954846382141113,
      "learning_rate": 1.185199644003332e-07,
      "loss": 0.6377,
      "step": 15000
    },
    {
      "epoch": 0.99,
      "grad_norm": 1.8074020147323608,
      "learning_rate": 1.1617059339563807e-07,
      "loss": 0.6575,
      "step": 15010
    },
    {
      "epoch": 0.99,
      "grad_norm": 1.7706438302993774,
      "learning_rate": 1.1384446768660572e-07,
      "loss": 0.666,
      "step": 15020
    },
    {
      "epoch": 0.99,
      "grad_norm": 2.4781429767608643,
      "learning_rate": 1.115415983447704e-07,
      "loss": 0.6667,
      "step": 15030
    },
    {
      "epoch": 1.0,
      "grad_norm": 2.2024221420288086,
      "learning_rate": 1.0926199633097156e-07,
      "loss": 0.662,
      "step": 15040
    },
    {
      "epoch": 1.0,
      "grad_norm": 2.2697882652282715,
      "learning_rate": 1.0700567249530835e-07,
      "loss": 0.6578,
      "step": 15050
    },
    {
      "epoch": 1.0,
      "grad_norm": 2.4836816787719727,
      "learning_rate": 1.0477263757708078e-07,
      "loss": 0.6539,
      "step": 15060
    },
    {
      "epoch": 1.0,
      "grad_norm": 2.1601076126098633,
      "learning_rate": 1.0256290220474308e-07,
      "loss": 0.6474,
      "step": 15070
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.9255223274230957,
      "learning_rate": 1.0037647689585207e-07,
      "loss": 0.6431,
      "step": 15080
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.622870922088623,
      "learning_rate": 9.821337205701664e-08,
      "loss": 0.6397,
      "step": 15090
    },
    {
      "epoch": 1.0,
      "grad_norm": 3.9312548637390137,
      "learning_rate": 9.607359798384785e-08,
      "loss": 0.6416,
      "step": 15100
    },
    {
      "epoch": 1.0,
      "grad_norm": 2.5819900035858154,
      "learning_rate": 9.395716486091222e-08,
      "loss": 0.6512,
      "step": 15110
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.9174243211746216,
      "learning_rate": 9.186408276168012e-08,
      "loss": 0.6391,
      "step": 15120
    },
    {
      "epoch": 1.0,
      "grad_norm": 2.371643304824829,
      "learning_rate": 8.979436164848088e-08,
      "loss": 0.6606,
      "step": 15130
    },
    {
      "epoch": 1.0,
      "grad_norm": 2.4533822536468506,
      "learning_rate": 8.77480113724516e-08,
      "loss": 0.6642,
      "step": 15140
    },
    {
      "epoch": 1.0,
      "grad_norm": 2.1988747119903564,
      "learning_rate": 8.572504167349449e-08,
      "loss": 0.637,
      "step": 15150
    },
    {
      "epoch": 1.0,
      "grad_norm": 2.2295937538146973,
      "learning_rate": 8.372546218022747e-08,
      "loss": 0.6612,
      "step": 15160
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.8617255687713623,
      "learning_rate": 8.174928240993917e-08,
      "loss": 0.6626,
      "step": 15170
    },
    {
      "epoch": 1.0,
      "grad_norm": 2.5846171379089355,
      "learning_rate": 7.979651176854564e-08,
      "loss": 0.6715,
      "step": 15180
    },
    {
      "epoch": 1.01,
      "grad_norm": 1.8019553422927856,
      "learning_rate": 7.786715955054202e-08,
      "loss": 0.6487,
      "step": 15190
    },
    {
      "epoch": 1.01,
      "grad_norm": 2.963513135910034,
      "learning_rate": 7.59612349389599e-08,
      "loss": 0.6428,
      "step": 15200
    },
    {
      "epoch": 1.01,
      "grad_norm": 2.1513736248016357,
      "learning_rate": 7.407874700532447e-08,
      "loss": 0.6655,
      "step": 15210
    },
    {
      "epoch": 1.01,
      "grad_norm": 1.9389243125915527,
      "learning_rate": 7.221970470961125e-08,
      "loss": 0.6522,
      "step": 15220
    },
    {
      "epoch": 1.01,
      "grad_norm": 2.251108407974243,
      "learning_rate": 7.03841169002023e-08,
      "loss": 0.6634,
      "step": 15230
    },
    {
      "epoch": 1.01,
      "grad_norm": 2.1588563919067383,
      "learning_rate": 6.857199231384282e-08,
      "loss": 0.6442,
      "step": 15240
    },
    {
      "epoch": 1.01,
      "grad_norm": 1.868743896484375,
      "learning_rate": 6.678333957560513e-08,
      "loss": 0.6628,
      "step": 15250
    },
    {
      "epoch": 1.01,
      "grad_norm": 1.8450775146484375,
      "learning_rate": 6.501816719884091e-08,
      "loss": 0.6362,
      "step": 15260
    },
    {
      "epoch": 1.01,
      "grad_norm": 1.971536636352539,
      "learning_rate": 6.327648358514404e-08,
      "loss": 0.655,
      "step": 15270
    },
    {
      "epoch": 1.01,
      "grad_norm": 2.585538625717163,
      "learning_rate": 6.15582970243117e-08,
      "loss": 0.6618,
      "step": 15280
    },
    {
      "epoch": 1.01,
      "grad_norm": 2.5381572246551514,
      "learning_rate": 5.986361569430166e-08,
      "loss": 0.6577,
      "step": 15290
    },
    {
      "epoch": 1.01,
      "grad_norm": 2.4174349308013916,
      "learning_rate": 5.8192447661196694e-08,
      "loss": 0.6678,
      "step": 15300
    },
    {
      "epoch": 1.01,
      "grad_norm": 1.9419946670532227,
      "learning_rate": 5.654480087916303e-08,
      "loss": 0.6855,
      "step": 15310
    },
    {
      "epoch": 1.01,
      "grad_norm": 2.323552131652832,
      "learning_rate": 5.492068319041588e-08,
      "loss": 0.6705,
      "step": 15320
    },
    {
      "epoch": 1.01,
      "grad_norm": 2.2662553787231445,
      "learning_rate": 5.332010232517892e-08,
      "loss": 0.6549,
      "step": 15330
    },
    {
      "epoch": 1.02,
      "grad_norm": 1.6378850936889648,
      "learning_rate": 5.174306590164879e-08,
      "loss": 0.6512,
      "step": 15340
    },
    {
      "epoch": 1.02,
      "grad_norm": 2.1355671882629395,
      "learning_rate": 5.0189581425960644e-08,
      "loss": 0.6373,
      "step": 15350
    },
    {
      "epoch": 1.02,
      "grad_norm": 2.738846778869629,
      "learning_rate": 4.865965629214819e-08,
      "loss": 0.6454,
      "step": 15360
    },
    {
      "epoch": 1.02,
      "grad_norm": 2.610119342803955,
      "learning_rate": 4.715329778211375e-08,
      "loss": 0.6471,
      "step": 15370
    },
    {
      "epoch": 1.02,
      "grad_norm": 2.306705951690674,
      "learning_rate": 4.5670513065588785e-08,
      "loss": 0.6486,
      "step": 15380
    },
    {
      "epoch": 1.02,
      "grad_norm": 2.4253878593444824,
      "learning_rate": 4.42113092001023e-08,
      "loss": 0.6577,
      "step": 15390
    },
    {
      "epoch": 1.02,
      "grad_norm": 2.209932565689087,
      "learning_rate": 4.2775693130948094e-08,
      "loss": 0.6449,
      "step": 15400
    },
    {
      "epoch": 1.02,
      "grad_norm": 2.779481887817383,
      "learning_rate": 4.1363671691148633e-08,
      "loss": 0.6569,
      "step": 15410
    },
    {
      "epoch": 1.02,
      "grad_norm": 2.4713690280914307,
      "learning_rate": 3.9975251601425126e-08,
      "loss": 0.6672,
      "step": 15420
    },
    {
      "epoch": 1.02,
      "grad_norm": 2.153374195098877,
      "learning_rate": 3.861043947016474e-08,
      "loss": 0.6839,
      "step": 15430
    },
    {
      "epoch": 1.02,
      "grad_norm": 1.948896884918213,
      "learning_rate": 3.726924179339009e-08,
      "loss": 0.6471,
      "step": 15440
    },
    {
      "epoch": 1.02,
      "grad_norm": 3.0682265758514404,
      "learning_rate": 3.59516649547248e-08,
      "loss": 0.6591,
      "step": 15450
    },
    {
      "epoch": 1.02,
      "grad_norm": 2.3854193687438965,
      "learning_rate": 3.465771522536854e-08,
      "loss": 0.6489,
      "step": 15460
    },
    {
      "epoch": 1.02,
      "grad_norm": 1.635633111000061,
      "learning_rate": 3.3387398764062603e-08,
      "loss": 0.6699,
      "step": 15470
    },
    {
      "epoch": 1.02,
      "grad_norm": 1.99565589427948,
      "learning_rate": 3.214072161706272e-08,
      "loss": 0.6636,
      "step": 15480
    },
    {
      "epoch": 1.03,
      "grad_norm": 1.8887382745742798,
      "learning_rate": 3.09176897181096e-08,
      "loss": 0.651,
      "step": 15490
    },
    {
      "epoch": 1.03,
      "grad_norm": 2.1598098278045654,
      "learning_rate": 2.971830888840177e-08,
      "loss": 0.6575,
      "step": 15500
    },
    {
      "epoch": 1.03,
      "grad_norm": 2.4078586101531982,
      "learning_rate": 2.854258483656669e-08,
      "loss": 0.6666,
      "step": 15510
    },
    {
      "epoch": 1.03,
      "grad_norm": 2.4672226905822754,
      "learning_rate": 2.7390523158633552e-08,
      "loss": 0.631,
      "step": 15520
    },
    {
      "epoch": 1.03,
      "grad_norm": 2.4110639095306396,
      "learning_rate": 2.6262129338006647e-08,
      "loss": 0.6673,
      "step": 15530
    },
    {
      "epoch": 1.03,
      "grad_norm": 1.6732667684555054,
      "learning_rate": 2.515740874544148e-08,
      "loss": 0.6444,
      "step": 15540
    },
    {
      "epoch": 1.03,
      "grad_norm": 2.4348301887512207,
      "learning_rate": 2.4076366639015914e-08,
      "loss": 0.6613,
      "step": 15550
    },
    {
      "epoch": 1.03,
      "grad_norm": 2.3126699924468994,
      "learning_rate": 2.301900816410574e-08,
      "loss": 0.6694,
      "step": 15560
    },
    {
      "epoch": 1.03,
      "grad_norm": 2.0329172611236572,
      "learning_rate": 2.198533835336414e-08,
      "loss": 0.6501,
      "step": 15570
    },
    {
      "epoch": 1.03,
      "grad_norm": 2.472811460494995,
      "learning_rate": 2.097536212669171e-08,
      "loss": 0.6594,
      "step": 15580
    },
    {
      "epoch": 1.03,
      "grad_norm": 2.291701555252075,
      "learning_rate": 1.9989084291216487e-08,
      "loss": 0.658,
      "step": 15590
    },
    {
      "epoch": 1.03,
      "grad_norm": 1.8039275407791138,
      "learning_rate": 1.9026509541272276e-08,
      "loss": 0.6577,
      "step": 15600
    },
    {
      "epoch": 1.03,
      "grad_norm": 2.6124210357666016,
      "learning_rate": 1.8087642458373135e-08,
      "loss": 0.6733,
      "step": 15610
    },
    {
      "epoch": 1.03,
      "grad_norm": 2.3919432163238525,
      "learning_rate": 1.7172487511192827e-08,
      "loss": 0.6597,
      "step": 15620
    },
    {
      "epoch": 1.03,
      "grad_norm": 1.9119877815246582,
      "learning_rate": 1.628104905554484e-08,
      "loss": 0.6733,
      "step": 15630
    },
    {
      "epoch": 1.03,
      "grad_norm": 2.052133083343506,
      "learning_rate": 1.541333133436018e-08,
      "loss": 0.6405,
      "step": 15640
    },
    {
      "epoch": 1.04,
      "grad_norm": 2.703890323638916,
      "learning_rate": 1.4569338477666838e-08,
      "loss": 0.6659,
      "step": 15650
    },
    {
      "epoch": 1.04,
      "grad_norm": 2.1235244274139404,
      "learning_rate": 1.3749074502572012e-08,
      "loss": 0.6451,
      "step": 15660
    },
    {
      "epoch": 1.04,
      "grad_norm": 3.593223810195923,
      "learning_rate": 1.2952543313240474e-08,
      "loss": 0.6553,
      "step": 15670
    },
    {
      "epoch": 1.04,
      "grad_norm": 1.7909396886825562,
      "learning_rate": 1.2179748700879013e-08,
      "loss": 0.6514,
      "step": 15680
    },
    {
      "epoch": 1.04,
      "grad_norm": 2.645784854888916,
      "learning_rate": 1.1430694343715354e-08,
      "loss": 0.6431,
      "step": 15690
    },
    {
      "epoch": 1.04,
      "grad_norm": 2.3556981086730957,
      "learning_rate": 1.0705383806982606e-08,
      "loss": 0.6731,
      "step": 15700
    },
    {
      "epoch": 1.04,
      "grad_norm": 2.335996389389038,
      "learning_rate": 1.00038205429015e-08,
      "loss": 0.6665,
      "step": 15710
    },
    {
      "epoch": 1.04,
      "grad_norm": 2.2005608081817627,
      "learning_rate": 9.3260078906654e-09,
      "loss": 0.6463,
      "step": 15720
    },
    {
      "epoch": 1.04,
      "grad_norm": 2.0543875694274902,
      "learning_rate": 8.671949076420883e-09,
      "loss": 0.6582,
      "step": 15730
    },
    {
      "epoch": 1.04,
      "grad_norm": 1.907952070236206,
      "learning_rate": 8.041647213256066e-09,
      "loss": 0.6501,
      "step": 15740
    },
    {
      "epoch": 1.04,
      "grad_norm": 3.233694076538086,
      "learning_rate": 7.43510530118452e-09,
      "loss": 0.6657,
      "step": 15750
    },
    {
      "epoch": 1.04,
      "grad_norm": 2.057260036468506,
      "learning_rate": 6.852326227130835e-09,
      "loss": 0.6804,
      "step": 15760
    },
    {
      "epoch": 1.04,
      "grad_norm": 2.1894454956054688,
      "learning_rate": 6.293312764916182e-09,
      "loss": 0.6562,
      "step": 15770
    },
    {
      "epoch": 1.04,
      "grad_norm": 2.1258082389831543,
      "learning_rate": 5.758067575246662e-09,
      "loss": 0.6551,
      "step": 15780
    },
    {
      "epoch": 1.04,
      "grad_norm": 1.4697284698486328,
      "learning_rate": 5.246593205699424e-09,
      "loss": 0.6638,
      "step": 15790
    },
    {
      "epoch": 1.05,
      "grad_norm": 2.231762409210205,
      "learning_rate": 4.758892090711009e-09,
      "loss": 0.6746,
      "step": 15800
    },
    {
      "epoch": 1.05,
      "grad_norm": 2.007432222366333,
      "learning_rate": 4.294966551566249e-09,
      "loss": 0.6382,
      "step": 15810
    },
    {
      "epoch": 1.05,
      "grad_norm": 2.1840968132019043,
      "learning_rate": 3.854818796385495e-09,
      "loss": 0.6444,
      "step": 15820
    },
    {
      "epoch": 1.05,
      "grad_norm": 2.2293922901153564,
      "learning_rate": 3.4384509201157435e-09,
      "loss": 0.6611,
      "step": 15830
    },
    {
      "epoch": 1.05,
      "grad_norm": 1.7496803998947144,
      "learning_rate": 3.0458649045211897e-09,
      "loss": 0.6644,
      "step": 15840
    },
    {
      "epoch": 1.05,
      "grad_norm": 2.6910362243652344,
      "learning_rate": 2.6770626181715776e-09,
      "loss": 0.6383,
      "step": 15850
    },
    {
      "epoch": 1.05,
      "grad_norm": 2.11971378326416,
      "learning_rate": 2.3320458164355352e-09,
      "loss": 0.6382,
      "step": 15860
    },
    {
      "epoch": 1.05,
      "grad_norm": 1.7193676233291626,
      "learning_rate": 2.0108161414694736e-09,
      "loss": 0.6607,
      "step": 15870
    },
    {
      "epoch": 1.05,
      "grad_norm": 2.7598321437835693,
      "learning_rate": 1.7133751222137007e-09,
      "loss": 0.6503,
      "step": 15880
    },
    {
      "epoch": 1.05,
      "grad_norm": 2.2920305728912354,
      "learning_rate": 1.4397241743813185e-09,
      "loss": 0.6791,
      "step": 15890
    },
    {
      "epoch": 1.05,
      "grad_norm": 2.376326560974121,
      "learning_rate": 1.189864600454338e-09,
      "loss": 0.666,
      "step": 15900
    },
    {
      "epoch": 1.05,
      "grad_norm": 2.610793352127075,
      "learning_rate": 9.637975896759077e-10,
      "loss": 0.6424,
      "step": 15910
    },
    {
      "epoch": 1.05,
      "grad_norm": 2.694457530975342,
      "learning_rate": 7.615242180436521e-10,
      "loss": 0.6644,
      "step": 15920
    },
    {
      "epoch": 1.05,
      "grad_norm": 2.2358672618865967,
      "learning_rate": 5.830454483085612e-10,
      "loss": 0.6916,
      "step": 15930
    },
    {
      "epoch": 1.05,
      "grad_norm": 1.7769392728805542,
      "learning_rate": 4.283621299649987e-10,
      "loss": 0.6549,
      "step": 15940
    },
    {
      "epoch": 1.06,
      "grad_norm": 2.700413227081299,
      "learning_rate": 2.974749992512571e-10,
      "loss": 0.6605,
      "step": 15950
    },
    {
      "epoch": 1.06,
      "grad_norm": 2.442838191986084,
      "learning_rate": 1.903846791434516e-10,
      "loss": 0.6544,
      "step": 15960
    },
    {
      "epoch": 1.06,
      "grad_norm": 1.9941034317016602,
      "learning_rate": 1.0709167935385456e-10,
      "loss": 0.6723,
      "step": 15970
    },
    {
      "epoch": 1.06,
      "grad_norm": 1.7016099691390991,
      "learning_rate": 4.7596396327564966e-11,
      "loss": 0.6498,
      "step": 15980
    },
    {
      "epoch": 1.06,
      "grad_norm": 2.3243091106414795,
      "learning_rate": 1.189911324084303e-11,
      "loss": 0.6521,
      "step": 15990
    },
    {
      "epoch": 1.06,
      "grad_norm": 2.1339619159698486,
      "learning_rate": 0.0,
      "loss": 0.6481,
      "step": 16000
    },
    {
      "epoch": 1.06,
      "eval_loss": 0.7880622148513794,
      "eval_runtime": 133.9187,
      "eval_samples_per_second": 82.139,
      "eval_steps_per_second": 10.267,
      "step": 16000
    }
  ],
  "logging_steps": 10,
  "max_steps": 16000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 1,
  "total_flos": 3.248711155315009e+19,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}