{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9989651604001379,
  "eval_steps": 500,
  "global_step": 724,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0013797861331493618,
      "grad_norm": 0.37928757071495056,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 10.3875,
      "step": 1
    },
    {
      "epoch": 0.0027595722662987236,
      "grad_norm": 0.4158506393432617,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 10.3804,
      "step": 2
    },
    {
      "epoch": 0.004139358399448086,
      "grad_norm": 0.4176561236381531,
      "learning_rate": 3e-06,
      "loss": 10.3729,
      "step": 3
    },
    {
      "epoch": 0.005519144532597447,
      "grad_norm": 0.4464573562145233,
      "learning_rate": 4.000000000000001e-06,
      "loss": 10.3778,
      "step": 4
    },
    {
      "epoch": 0.006898930665746809,
      "grad_norm": 0.4155403673648834,
      "learning_rate": 5e-06,
      "loss": 10.3817,
      "step": 5
    },
    {
      "epoch": 0.008278716798896172,
      "grad_norm": 0.4856280982494354,
      "learning_rate": 6e-06,
      "loss": 10.3782,
      "step": 6
    },
    {
      "epoch": 0.009658502932045532,
      "grad_norm": 0.4716717600822449,
      "learning_rate": 7.000000000000001e-06,
      "loss": 10.3727,
      "step": 7
    },
    {
      "epoch": 0.011038289065194894,
      "grad_norm": 0.4803764820098877,
      "learning_rate": 8.000000000000001e-06,
      "loss": 10.3701,
      "step": 8
    },
    {
      "epoch": 0.012418075198344257,
      "grad_norm": 0.4964543282985687,
      "learning_rate": 9e-06,
      "loss": 10.3719,
      "step": 9
    },
    {
      "epoch": 0.013797861331493619,
      "grad_norm": 0.5420424938201904,
      "learning_rate": 1e-05,
      "loss": 10.3879,
      "step": 10
    },
    {
      "epoch": 0.015177647464642981,
      "grad_norm": 0.5523043274879456,
      "learning_rate": 1.1000000000000001e-05,
      "loss": 10.3767,
      "step": 11
    },
    {
      "epoch": 0.016557433597792343,
      "grad_norm": 0.5524217486381531,
      "learning_rate": 1.2e-05,
      "loss": 10.3849,
      "step": 12
    },
    {
      "epoch": 0.017937219730941704,
      "grad_norm": 0.47942158579826355,
      "learning_rate": 1.3000000000000001e-05,
      "loss": 10.3652,
      "step": 13
    },
    {
      "epoch": 0.019317005864091064,
      "grad_norm": 0.3707212805747986,
      "learning_rate": 1.4000000000000001e-05,
      "loss": 10.3772,
      "step": 14
    },
    {
      "epoch": 0.020696791997240428,
      "grad_norm": 0.43476638197898865,
      "learning_rate": 1.5e-05,
      "loss": 10.3805,
      "step": 15
    },
    {
      "epoch": 0.02207657813038979,
      "grad_norm": 0.44521450996398926,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 10.3757,
      "step": 16
    },
    {
      "epoch": 0.023456364263539153,
      "grad_norm": 0.5112989544868469,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 10.3853,
      "step": 17
    },
    {
      "epoch": 0.024836150396688513,
      "grad_norm": 0.48292139172554016,
      "learning_rate": 1.8e-05,
      "loss": 10.3791,
      "step": 18
    },
    {
      "epoch": 0.026215936529837874,
      "grad_norm": 0.46689826250076294,
      "learning_rate": 1.9e-05,
      "loss": 10.3874,
      "step": 19
    },
    {
      "epoch": 0.027595722662987238,
      "grad_norm": 0.4777892231941223,
      "learning_rate": 2e-05,
      "loss": 10.3751,
      "step": 20
    },
    {
      "epoch": 0.028975508796136598,
      "grad_norm": 0.5070079565048218,
      "learning_rate": 2.1e-05,
      "loss": 10.3844,
      "step": 21
    },
    {
      "epoch": 0.030355294929285962,
      "grad_norm": 0.5363792181015015,
      "learning_rate": 2.2000000000000003e-05,
      "loss": 10.3806,
      "step": 22
    },
    {
      "epoch": 0.03173508106243532,
      "grad_norm": 0.5464498400688171,
      "learning_rate": 2.3000000000000003e-05,
      "loss": 10.3758,
      "step": 23
    },
    {
      "epoch": 0.03311486719558469,
      "grad_norm": 0.6037337779998779,
      "learning_rate": 2.4e-05,
      "loss": 10.3716,
      "step": 24
    },
    {
      "epoch": 0.03449465332873405,
      "grad_norm": 0.6836243867874146,
      "learning_rate": 2.5e-05,
      "loss": 10.3811,
      "step": 25
    },
    {
      "epoch": 0.03587443946188341,
      "grad_norm": 0.3856373727321625,
      "learning_rate": 2.6000000000000002e-05,
      "loss": 10.3743,
      "step": 26
    },
    {
      "epoch": 0.03725422559503277,
      "grad_norm": 0.40399476885795593,
      "learning_rate": 2.7000000000000002e-05,
      "loss": 10.3781,
      "step": 27
    },
    {
      "epoch": 0.03863401172818213,
      "grad_norm": 0.45152705907821655,
      "learning_rate": 2.8000000000000003e-05,
      "loss": 10.3788,
      "step": 28
    },
    {
      "epoch": 0.040013797861331496,
      "grad_norm": 0.4663512110710144,
      "learning_rate": 2.9e-05,
      "loss": 10.3869,
      "step": 29
    },
    {
      "epoch": 0.041393583994480856,
      "grad_norm": 0.494383305311203,
      "learning_rate": 3e-05,
      "loss": 10.3756,
      "step": 30
    },
    {
      "epoch": 0.04277337012763022,
      "grad_norm": 0.48948773741722107,
      "learning_rate": 3.1e-05,
      "loss": 10.3787,
      "step": 31
    },
    {
      "epoch": 0.04415315626077958,
      "grad_norm": 0.45743992924690247,
      "learning_rate": 3.2000000000000005e-05,
      "loss": 10.3724,
      "step": 32
    },
    {
      "epoch": 0.04553294239392894,
      "grad_norm": 0.4397124946117401,
      "learning_rate": 3.3e-05,
      "loss": 10.3667,
      "step": 33
    },
    {
      "epoch": 0.046912728527078305,
      "grad_norm": 0.5066558122634888,
      "learning_rate": 3.4000000000000007e-05,
      "loss": 10.3686,
      "step": 34
    },
    {
      "epoch": 0.048292514660227666,
      "grad_norm": 0.5163968801498413,
      "learning_rate": 3.5e-05,
      "loss": 10.3699,
      "step": 35
    },
    {
      "epoch": 0.049672300793377026,
      "grad_norm": 0.6014847159385681,
      "learning_rate": 3.6e-05,
      "loss": 10.373,
      "step": 36
    },
    {
      "epoch": 0.05105208692652639,
      "grad_norm": 0.5788203477859497,
      "learning_rate": 3.7e-05,
      "loss": 10.3686,
      "step": 37
    },
    {
      "epoch": 0.05243187305967575,
      "grad_norm": 0.4628985524177551,
      "learning_rate": 3.8e-05,
      "loss": 10.3772,
      "step": 38
    },
    {
      "epoch": 0.053811659192825115,
      "grad_norm": 0.4443260133266449,
      "learning_rate": 3.9000000000000006e-05,
      "loss": 10.3745,
      "step": 39
    },
    {
      "epoch": 0.055191445325974475,
      "grad_norm": 0.42313191294670105,
      "learning_rate": 4e-05,
      "loss": 10.3679,
      "step": 40
    },
    {
      "epoch": 0.056571231459123836,
      "grad_norm": 0.4354301691055298,
      "learning_rate": 4.1e-05,
      "loss": 10.3646,
      "step": 41
    },
    {
      "epoch": 0.057951017592273196,
      "grad_norm": 0.45086437463760376,
      "learning_rate": 4.2e-05,
      "loss": 10.377,
      "step": 42
    },
    {
      "epoch": 0.05933080372542256,
      "grad_norm": 0.4813697040081024,
      "learning_rate": 4.3e-05,
      "loss": 10.3761,
      "step": 43
    },
    {
      "epoch": 0.060710589858571924,
      "grad_norm": 0.46686020493507385,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 10.3674,
      "step": 44
    },
    {
      "epoch": 0.062090375991721285,
      "grad_norm": 0.5052739381790161,
      "learning_rate": 4.5e-05,
      "loss": 10.3637,
      "step": 45
    },
    {
      "epoch": 0.06347016212487064,
      "grad_norm": 0.5201326012611389,
      "learning_rate": 4.600000000000001e-05,
      "loss": 10.3658,
      "step": 46
    },
    {
      "epoch": 0.06484994825802001,
      "grad_norm": 0.536177933216095,
      "learning_rate": 4.7e-05,
      "loss": 10.3699,
      "step": 47
    },
    {
      "epoch": 0.06622973439116937,
      "grad_norm": 0.5888517498970032,
      "learning_rate": 4.8e-05,
      "loss": 10.3689,
      "step": 48
    },
    {
      "epoch": 0.06760952052431873,
      "grad_norm": 0.5909680724143982,
      "learning_rate": 4.9e-05,
      "loss": 10.3739,
      "step": 49
    },
    {
      "epoch": 0.0689893066574681,
      "grad_norm": 0.7720197439193726,
      "learning_rate": 5e-05,
      "loss": 10.3599,
      "step": 50
    },
    {
      "epoch": 0.07036909279061745,
      "grad_norm": 0.4170343279838562,
      "learning_rate": 5.1000000000000006e-05,
      "loss": 10.37,
      "step": 51
    },
    {
      "epoch": 0.07174887892376682,
      "grad_norm": 0.45865100622177124,
      "learning_rate": 5.2000000000000004e-05,
      "loss": 10.3616,
      "step": 52
    },
    {
      "epoch": 0.07312866505691618,
      "grad_norm": 0.4740890562534332,
      "learning_rate": 5.300000000000001e-05,
      "loss": 10.3721,
      "step": 53
    },
    {
      "epoch": 0.07450845119006554,
      "grad_norm": 0.44731295108795166,
      "learning_rate": 5.4000000000000005e-05,
      "loss": 10.3611,
      "step": 54
    },
    {
      "epoch": 0.0758882373232149,
      "grad_norm": 0.5211865305900574,
      "learning_rate": 5.500000000000001e-05,
      "loss": 10.359,
      "step": 55
    },
    {
      "epoch": 0.07726802345636426,
      "grad_norm": 0.49909907579421997,
      "learning_rate": 5.6000000000000006e-05,
      "loss": 10.3668,
      "step": 56
    },
    {
      "epoch": 0.07864780958951363,
      "grad_norm": 0.5443109273910522,
      "learning_rate": 5.6999999999999996e-05,
      "loss": 10.3489,
      "step": 57
    },
    {
      "epoch": 0.08002759572266299,
      "grad_norm": 0.554502546787262,
      "learning_rate": 5.8e-05,
      "loss": 10.3692,
      "step": 58
    },
    {
      "epoch": 0.08140738185581235,
      "grad_norm": 0.6308146119117737,
      "learning_rate": 5.9e-05,
      "loss": 10.3573,
      "step": 59
    },
    {
      "epoch": 0.08278716798896171,
      "grad_norm": 0.6720134019851685,
      "learning_rate": 6e-05,
      "loss": 10.3514,
      "step": 60
    },
    {
      "epoch": 0.08416695412211107,
      "grad_norm": 0.748083770275116,
      "learning_rate": 6.1e-05,
      "loss": 10.3545,
      "step": 61
    },
    {
      "epoch": 0.08554674025526043,
      "grad_norm": 0.8078204989433289,
      "learning_rate": 6.2e-05,
      "loss": 10.3506,
      "step": 62
    },
    {
      "epoch": 0.0869265263884098,
      "grad_norm": 0.5997583270072937,
      "learning_rate": 6.3e-05,
      "loss": 10.3588,
      "step": 63
    },
    {
      "epoch": 0.08830631252155915,
      "grad_norm": 0.5249990820884705,
      "learning_rate": 6.400000000000001e-05,
      "loss": 10.3635,
      "step": 64
    },
    {
      "epoch": 0.08968609865470852,
      "grad_norm": 0.5661075115203857,
      "learning_rate": 6.500000000000001e-05,
      "loss": 10.3594,
      "step": 65
    },
    {
      "epoch": 0.09106588478785788,
      "grad_norm": 0.5764796137809753,
      "learning_rate": 6.6e-05,
      "loss": 10.351,
      "step": 66
    },
    {
      "epoch": 0.09244567092100725,
      "grad_norm": 0.6323254108428955,
      "learning_rate": 6.7e-05,
      "loss": 10.357,
      "step": 67
    },
    {
      "epoch": 0.09382545705415661,
      "grad_norm": 0.6489927172660828,
      "learning_rate": 6.800000000000001e-05,
      "loss": 10.3467,
      "step": 68
    },
    {
      "epoch": 0.09520524318730597,
      "grad_norm": 0.7028964757919312,
      "learning_rate": 6.9e-05,
      "loss": 10.3459,
      "step": 69
    },
    {
      "epoch": 0.09658502932045533,
      "grad_norm": 0.7597895264625549,
      "learning_rate": 7e-05,
      "loss": 10.3482,
      "step": 70
    },
    {
      "epoch": 0.09796481545360469,
      "grad_norm": 0.8057466149330139,
      "learning_rate": 7.1e-05,
      "loss": 10.3374,
      "step": 71
    },
    {
      "epoch": 0.09934460158675405,
      "grad_norm": 0.8530853390693665,
      "learning_rate": 7.2e-05,
      "loss": 10.3444,
      "step": 72
    },
    {
      "epoch": 0.10072438771990341,
      "grad_norm": 0.9716468453407288,
      "learning_rate": 7.3e-05,
      "loss": 10.3242,
      "step": 73
    },
    {
      "epoch": 0.10210417385305277,
      "grad_norm": 1.105352759361267,
      "learning_rate": 7.4e-05,
      "loss": 10.3172,
      "step": 74
    },
    {
      "epoch": 0.10348395998620213,
      "grad_norm": 1.2654082775115967,
      "learning_rate": 7.500000000000001e-05,
      "loss": 10.3131,
      "step": 75
    },
    {
      "epoch": 0.1048637461193515,
      "grad_norm": 0.7005993723869324,
      "learning_rate": 7.6e-05,
      "loss": 10.3454,
      "step": 76
    },
    {
      "epoch": 0.10624353225250087,
      "grad_norm": 0.8157169222831726,
      "learning_rate": 7.7e-05,
      "loss": 10.328,
      "step": 77
    },
    {
      "epoch": 0.10762331838565023,
      "grad_norm": 0.8883787989616394,
      "learning_rate": 7.800000000000001e-05,
      "loss": 10.3259,
      "step": 78
    },
    {
      "epoch": 0.10900310451879959,
      "grad_norm": 0.9492635130882263,
      "learning_rate": 7.900000000000001e-05,
      "loss": 10.3214,
      "step": 79
    },
    {
      "epoch": 0.11038289065194895,
      "grad_norm": 0.8827070593833923,
      "learning_rate": 8e-05,
      "loss": 10.3172,
      "step": 80
    },
    {
      "epoch": 0.11176267678509831,
      "grad_norm": 1.0507211685180664,
      "learning_rate": 8.1e-05,
      "loss": 10.3165,
      "step": 81
    },
    {
      "epoch": 0.11314246291824767,
      "grad_norm": 1.1131185293197632,
      "learning_rate": 8.2e-05,
      "loss": 10.3091,
      "step": 82
    },
    {
      "epoch": 0.11452224905139703,
      "grad_norm": 1.1873068809509277,
      "learning_rate": 8.3e-05,
      "loss": 10.2996,
      "step": 83
    },
    {
      "epoch": 0.11590203518454639,
      "grad_norm": 1.1595804691314697,
      "learning_rate": 8.4e-05,
      "loss": 10.2908,
      "step": 84
    },
    {
      "epoch": 0.11728182131769575,
      "grad_norm": 1.2971093654632568,
      "learning_rate": 8.5e-05,
      "loss": 10.2807,
      "step": 85
    },
    {
      "epoch": 0.11866160745084511,
      "grad_norm": 1.3693597316741943,
      "learning_rate": 8.6e-05,
      "loss": 10.2609,
      "step": 86
    },
    {
      "epoch": 0.12004139358399447,
      "grad_norm": 1.4227279424667358,
      "learning_rate": 8.7e-05,
      "loss": 10.237,
      "step": 87
    },
    {
      "epoch": 0.12142117971714385,
      "grad_norm": 1.0355181694030762,
      "learning_rate": 8.800000000000001e-05,
      "loss": 10.2805,
      "step": 88
    },
    {
      "epoch": 0.12280096585029321,
      "grad_norm": 1.0092841386795044,
      "learning_rate": 8.900000000000001e-05,
      "loss": 10.2912,
      "step": 89
    },
    {
      "epoch": 0.12418075198344257,
      "grad_norm": 1.089247465133667,
      "learning_rate": 9e-05,
      "loss": 10.2767,
      "step": 90
    },
    {
      "epoch": 0.12556053811659193,
      "grad_norm": 1.1349172592163086,
      "learning_rate": 9.1e-05,
      "loss": 10.2677,
      "step": 91
    },
    {
      "epoch": 0.12694032424974128,
      "grad_norm": 1.1714922189712524,
      "learning_rate": 9.200000000000001e-05,
      "loss": 10.2609,
      "step": 92
    },
    {
      "epoch": 0.12832011038289065,
      "grad_norm": 1.191255807876587,
      "learning_rate": 9.300000000000001e-05,
      "loss": 10.2532,
      "step": 93
    },
    {
      "epoch": 0.12969989651604003,
      "grad_norm": 1.232477068901062,
      "learning_rate": 9.4e-05,
      "loss": 10.2439,
      "step": 94
    },
    {
      "epoch": 0.13107968264918937,
      "grad_norm": 1.2855937480926514,
      "learning_rate": 9.5e-05,
      "loss": 10.2264,
      "step": 95
    },
    {
      "epoch": 0.13245946878233875,
      "grad_norm": 1.232479453086853,
      "learning_rate": 9.6e-05,
      "loss": 10.2317,
      "step": 96
    },
    {
      "epoch": 0.1338392549154881,
      "grad_norm": 1.3270705938339233,
      "learning_rate": 9.7e-05,
      "loss": 10.2043,
      "step": 97
    },
    {
      "epoch": 0.13521904104863747,
      "grad_norm": 1.3535412549972534,
      "learning_rate": 9.8e-05,
      "loss": 10.1987,
      "step": 98
    },
    {
      "epoch": 0.1365988271817868,
      "grad_norm": 1.473716378211975,
      "learning_rate": 9.900000000000001e-05,
      "loss": 10.183,
      "step": 99
    },
    {
      "epoch": 0.1379786133149362,
      "grad_norm": 1.5545284748077393,
      "learning_rate": 0.0001,
      "loss": 10.1764,
      "step": 100
    },
    {
      "epoch": 0.13935839944808553,
      "grad_norm": 0.9037075638771057,
      "learning_rate": 9.999936632050037e-05,
      "loss": 10.2408,
      "step": 101
    },
    {
      "epoch": 0.1407381855812349,
      "grad_norm": 1.025711178779602,
      "learning_rate": 9.999746529806349e-05,
      "loss": 10.2258,
      "step": 102
    },
    {
      "epoch": 0.14211797171438428,
      "grad_norm": 1.0868299007415771,
      "learning_rate": 9.99942969808749e-05,
      "loss": 10.219,
      "step": 103
    },
    {
      "epoch": 0.14349775784753363,
      "grad_norm": 1.1319340467453003,
      "learning_rate": 9.998986144924251e-05,
      "loss": 10.2066,
      "step": 104
    },
    {
      "epoch": 0.144877543980683,
      "grad_norm": 1.1912342309951782,
      "learning_rate": 9.998415881559454e-05,
      "loss": 10.1997,
      "step": 105
    },
    {
      "epoch": 0.14625733011383235,
      "grad_norm": 1.1592702865600586,
      "learning_rate": 9.997718922447667e-05,
      "loss": 10.1898,
      "step": 106
    },
    {
      "epoch": 0.14763711624698173,
      "grad_norm": 1.2085384130477905,
      "learning_rate": 9.99689528525484e-05,
      "loss": 10.1825,
      "step": 107
    },
    {
      "epoch": 0.14901690238013107,
      "grad_norm": 1.211334466934204,
      "learning_rate": 9.995944990857849e-05,
      "loss": 10.1712,
      "step": 108
    },
    {
      "epoch": 0.15039668851328045,
      "grad_norm": 1.3283801078796387,
      "learning_rate": 9.994868063343981e-05,
      "loss": 10.1671,
      "step": 109
    },
    {
      "epoch": 0.1517764746464298,
      "grad_norm": 1.3618557453155518,
      "learning_rate": 9.993664530010308e-05,
      "loss": 10.1597,
      "step": 110
    },
    {
      "epoch": 0.15315626077957917,
      "grad_norm": 1.3570090532302856,
      "learning_rate": 9.99233442136301e-05,
      "loss": 10.1517,
      "step": 111
    },
    {
      "epoch": 0.15453604691272851,
      "grad_norm": 1.4205886125564575,
      "learning_rate": 9.990877771116589e-05,
      "loss": 10.1415,
      "step": 112
    },
    {
      "epoch": 0.1559158330458779,
      "grad_norm": 1.2085868120193481,
      "learning_rate": 9.989294616193017e-05,
      "loss": 10.1706,
      "step": 113
    },
    {
      "epoch": 0.15729561917902726,
      "grad_norm": 1.0051339864730835,
      "learning_rate": 9.987584996720814e-05,
      "loss": 10.1806,
      "step": 114
    },
    {
      "epoch": 0.1586754053121766,
      "grad_norm": 1.044624924659729,
      "learning_rate": 9.985748956034006e-05,
      "loss": 10.1836,
      "step": 115
    },
    {
      "epoch": 0.16005519144532598,
      "grad_norm": 1.1088407039642334,
      "learning_rate": 9.983786540671051e-05,
      "loss": 10.1793,
      "step": 116
    },
    {
      "epoch": 0.16143497757847533,
      "grad_norm": 1.156091332435608,
      "learning_rate": 9.981697800373642e-05,
      "loss": 10.1542,
      "step": 117
    },
    {
      "epoch": 0.1628147637116247,
      "grad_norm": 1.1839971542358398,
      "learning_rate": 9.979482788085454e-05,
      "loss": 10.1521,
      "step": 118
    },
    {
      "epoch": 0.16419454984477405,
      "grad_norm": 1.257676362991333,
      "learning_rate": 9.977141559950807e-05,
      "loss": 10.146,
      "step": 119
    },
    {
      "epoch": 0.16557433597792343,
      "grad_norm": 1.2359771728515625,
      "learning_rate": 9.974674175313228e-05,
      "loss": 10.1465,
      "step": 120
    },
    {
      "epoch": 0.16695412211107277,
      "grad_norm": 1.3480055332183838,
      "learning_rate": 9.972080696713961e-05,
      "loss": 10.1324,
      "step": 121
    },
    {
      "epoch": 0.16833390824422215,
      "grad_norm": 1.3127009868621826,
      "learning_rate": 9.969361189890373e-05,
      "loss": 10.1225,
      "step": 122
    },
    {
      "epoch": 0.16971369437737152,
      "grad_norm": 1.369981050491333,
      "learning_rate": 9.966515723774297e-05,
      "loss": 10.108,
      "step": 123
    },
    {
      "epoch": 0.17109348051052087,
      "grad_norm": 1.384888768196106,
      "learning_rate": 9.96354437049027e-05,
      "loss": 10.1046,
      "step": 124
    },
    {
      "epoch": 0.17247326664367024,
      "grad_norm": 1.5264196395874023,
      "learning_rate": 9.960447205353723e-05,
      "loss": 10.0868,
      "step": 125
    },
    {
      "epoch": 0.1738530527768196,
      "grad_norm": 0.9188018441200256,
      "learning_rate": 9.957224306869053e-05,
      "loss": 10.1521,
      "step": 126
    },
    {
      "epoch": 0.17523283890996896,
      "grad_norm": 1.0055698156356812,
      "learning_rate": 9.953875756727651e-05,
      "loss": 10.1447,
      "step": 127
    },
    {
      "epoch": 0.1766126250431183,
      "grad_norm": 1.115513563156128,
      "learning_rate": 9.95040163980582e-05,
      "loss": 10.1354,
      "step": 128
    },
    {
      "epoch": 0.17799241117626768,
      "grad_norm": 1.0765373706817627,
      "learning_rate": 9.946802044162629e-05,
      "loss": 10.1282,
      "step": 129
    },
    {
      "epoch": 0.17937219730941703,
      "grad_norm": 1.133651852607727,
      "learning_rate": 9.943077061037671e-05,
      "loss": 10.1215,
      "step": 130
    },
    {
      "epoch": 0.1807519834425664,
      "grad_norm": 1.1984049081802368,
      "learning_rate": 9.939226784848769e-05,
      "loss": 10.1199,
      "step": 131
    },
    {
      "epoch": 0.18213176957571575,
      "grad_norm": 1.1945085525512695,
      "learning_rate": 9.935251313189564e-05,
      "loss": 10.1057,
      "step": 132
    },
    {
      "epoch": 0.18351155570886513,
      "grad_norm": 1.28304123878479,
      "learning_rate": 9.931150746827054e-05,
      "loss": 10.1038,
      "step": 133
    },
    {
      "epoch": 0.1848913418420145,
      "grad_norm": 1.3560444116592407,
      "learning_rate": 9.92692518969903e-05,
      "loss": 10.0773,
      "step": 134
    },
    {
      "epoch": 0.18627112797516385,
      "grad_norm": 1.2892886400222778,
      "learning_rate": 9.922574748911448e-05,
      "loss": 10.0791,
      "step": 135
    },
    {
      "epoch": 0.18765091410831322,
      "grad_norm": 1.3138704299926758,
      "learning_rate": 9.918099534735718e-05,
      "loss": 10.0841,
      "step": 136
    },
    {
      "epoch": 0.18903070024146257,
      "grad_norm": 1.466147541999817,
      "learning_rate": 9.913499660605897e-05,
      "loss": 10.0606,
      "step": 137
    },
    {
      "epoch": 0.19041048637461194,
      "grad_norm": 1.1251062154769897,
      "learning_rate": 9.908775243115821e-05,
      "loss": 10.0973,
      "step": 138
    },
    {
      "epoch": 0.1917902725077613,
      "grad_norm": 1.0003620386123657,
      "learning_rate": 9.903926402016153e-05,
      "loss": 10.1169,
      "step": 139
    },
    {
      "epoch": 0.19317005864091066,
      "grad_norm": 1.0686968564987183,
      "learning_rate": 9.898953260211338e-05,
      "loss": 10.116,
      "step": 140
    },
    {
      "epoch": 0.19454984477406,
      "grad_norm": 1.1628526449203491,
      "learning_rate": 9.8938559437565e-05,
      "loss": 10.0882,
      "step": 141
    },
    {
      "epoch": 0.19592963090720938,
      "grad_norm": 1.116454005241394,
      "learning_rate": 9.888634581854234e-05,
      "loss": 10.0891,
      "step": 142
    },
    {
      "epoch": 0.19730941704035873,
      "grad_norm": 1.2207541465759277,
      "learning_rate": 9.883289306851342e-05,
      "loss": 10.075,
      "step": 143
    },
    {
      "epoch": 0.1986892031735081,
      "grad_norm": 1.2331547737121582,
      "learning_rate": 9.877820254235471e-05,
      "loss": 10.0643,
      "step": 144
    },
    {
      "epoch": 0.20006898930665748,
      "grad_norm": 1.2393301725387573,
      "learning_rate": 9.87222756263168e-05,
      "loss": 10.0757,
      "step": 145
    },
    {
      "epoch": 0.20144877543980683,
      "grad_norm": 1.288958191871643,
      "learning_rate": 9.86651137379893e-05,
      "loss": 10.0672,
      "step": 146
    },
    {
      "epoch": 0.2028285615729562,
      "grad_norm": 1.3286681175231934,
      "learning_rate": 9.86067183262649e-05,
      "loss": 10.0466,
      "step": 147
    },
    {
      "epoch": 0.20420834770610555,
      "grad_norm": 1.352874994277954,
      "learning_rate": 9.85470908713026e-05,
      "loss": 10.0367,
      "step": 148
    },
    {
      "epoch": 0.20558813383925492,
      "grad_norm": 1.4209654331207275,
      "learning_rate": 9.848623288449025e-05,
      "loss": 10.0265,
      "step": 149
    },
    {
      "epoch": 0.20696791997240427,
      "grad_norm": 1.5714178085327148,
      "learning_rate": 9.842414590840617e-05,
      "loss": 9.9945,
      "step": 150
    },
    {
      "epoch": 0.20834770610555364,
      "grad_norm": 0.9352539777755737,
      "learning_rate": 9.836083151678014e-05,
      "loss": 10.0962,
      "step": 151
    },
    {
      "epoch": 0.209727492238703,
      "grad_norm": 1.0083916187286377,
      "learning_rate": 9.829629131445342e-05,
      "loss": 10.0812,
      "step": 152
    },
    {
      "epoch": 0.21110727837185236,
      "grad_norm": 1.155738353729248,
      "learning_rate": 9.823052693733814e-05,
      "loss": 10.0547,
      "step": 153
    },
    {
      "epoch": 0.21248706450500174,
      "grad_norm": 1.1547437906265259,
      "learning_rate": 9.816354005237583e-05,
      "loss": 10.0552,
      "step": 154
    },
    {
      "epoch": 0.21386685063815108,
      "grad_norm": 1.1641536951065063,
      "learning_rate": 9.809533235749507e-05,
      "loss": 10.061,
      "step": 155
    },
    {
      "epoch": 0.21524663677130046,
      "grad_norm": 1.2230818271636963,
      "learning_rate": 9.802590558156862e-05,
      "loss": 10.0426,
      "step": 156
    },
    {
      "epoch": 0.2166264229044498,
      "grad_norm": 1.2705639600753784,
      "learning_rate": 9.795526148436944e-05,
      "loss": 10.0289,
      "step": 157
    },
    {
      "epoch": 0.21800620903759918,
      "grad_norm": 1.2639011144638062,
      "learning_rate": 9.78834018565262e-05,
      "loss": 10.0266,
      "step": 158
    },
    {
      "epoch": 0.21938599517074853,
      "grad_norm": 1.295678973197937,
      "learning_rate": 9.78103285194778e-05,
      "loss": 10.0191,
      "step": 159
    },
    {
      "epoch": 0.2207657813038979,
      "grad_norm": 1.3106251955032349,
      "learning_rate": 9.773604332542729e-05,
      "loss": 10.0146,
      "step": 160
    },
    {
      "epoch": 0.22214556743704725,
      "grad_norm": 1.3326599597930908,
      "learning_rate": 9.766054815729482e-05,
      "loss": 10.0172,
      "step": 161
    },
    {
      "epoch": 0.22352535357019662,
      "grad_norm": 1.4855268001556396,
      "learning_rate": 9.758384492867003e-05,
      "loss": 9.9864,
      "step": 162
    },
    {
      "epoch": 0.22490513970334597,
      "grad_norm": 1.2385770082473755,
      "learning_rate": 9.750593558376347e-05,
      "loss": 10.0079,
      "step": 163
    },
    {
      "epoch": 0.22628492583649534,
      "grad_norm": 1.0151700973510742,
      "learning_rate": 9.742682209735727e-05,
      "loss": 10.0535,
      "step": 164
    },
    {
      "epoch": 0.22766471196964472,
      "grad_norm": 1.0624631643295288,
      "learning_rate": 9.73465064747553e-05,
      "loss": 10.0475,
      "step": 165
    },
    {
      "epoch": 0.22904449810279406,
      "grad_norm": 1.0995757579803467,
      "learning_rate": 9.726499075173201e-05,
      "loss": 10.0349,
      "step": 166
    },
    {
      "epoch": 0.23042428423594344,
      "grad_norm": 1.2015033960342407,
      "learning_rate": 9.718227699448116e-05,
      "loss": 10.0152,
      "step": 167
    },
    {
      "epoch": 0.23180407036909279,
      "grad_norm": 1.1979053020477295,
      "learning_rate": 9.709836729956325e-05,
      "loss": 10.0255,
      "step": 168
    },
    {
      "epoch": 0.23318385650224216,
      "grad_norm": 1.2255140542984009,
      "learning_rate": 9.701326379385238e-05,
      "loss": 10.0096,
      "step": 169
    },
    {
      "epoch": 0.2345636426353915,
      "grad_norm": 1.2472561597824097,
      "learning_rate": 9.692696863448245e-05,
      "loss": 10.0073,
      "step": 170
    },
    {
      "epoch": 0.23594342876854088,
      "grad_norm": 1.328076958656311,
      "learning_rate": 9.68394840087924e-05,
      "loss": 9.983,
      "step": 171
    },
    {
      "epoch": 0.23732321490169023,
      "grad_norm": 1.318142294883728,
      "learning_rate": 9.675081213427076e-05,
      "loss": 9.9967,
      "step": 172
    },
    {
      "epoch": 0.2387030010348396,
      "grad_norm": 1.432610034942627,
      "learning_rate": 9.66609552584995e-05,
      "loss": 9.9658,
      "step": 173
    },
    {
      "epoch": 0.24008278716798895,
      "grad_norm": 1.4061625003814697,
      "learning_rate": 9.656991565909704e-05,
      "loss": 9.9669,
      "step": 174
    },
    {
      "epoch": 0.24146257330113832,
      "grad_norm": 1.5450146198272705,
      "learning_rate": 9.647769564366046e-05,
      "loss": 9.9462,
      "step": 175
    },
    {
      "epoch": 0.2428423594342877,
      "grad_norm": 0.9310745000839233,
      "learning_rate": 9.638429754970715e-05,
      "loss": 10.0483,
      "step": 176
    },
    {
      "epoch": 0.24422214556743704,
      "grad_norm": 1.0601836442947388,
      "learning_rate": 9.628972374461536e-05,
      "loss": 10.0273,
      "step": 177
    },
    {
      "epoch": 0.24560193170058642,
      "grad_norm": 1.1165393590927124,
      "learning_rate": 9.619397662556435e-05,
      "loss": 10.0129,
      "step": 178
    },
    {
      "epoch": 0.24698171783373576,
      "grad_norm": 1.0721098184585571,
      "learning_rate": 9.609705861947359e-05,
      "loss": 10.0034,
      "step": 179
    },
    {
      "epoch": 0.24836150396688514,
      "grad_norm": 1.1826632022857666,
      "learning_rate": 9.599897218294122e-05,
      "loss": 9.9898,
      "step": 180
    },
    {
      "epoch": 0.24974129010003449,
      "grad_norm": 1.1779521703720093,
      "learning_rate": 9.589971980218179e-05,
      "loss": 9.9933,
      "step": 181
    },
    {
      "epoch": 0.25112107623318386,
      "grad_norm": 1.261836051940918,
      "learning_rate": 9.579930399296331e-05,
      "loss": 9.9738,
      "step": 182
    },
    {
      "epoch": 0.25250086236633323,
      "grad_norm": 1.2774072885513306,
      "learning_rate": 9.569772730054329e-05,
      "loss": 9.9662,
      "step": 183
    },
    {
      "epoch": 0.25388064849948255,
      "grad_norm": 1.3094655275344849,
      "learning_rate": 9.559499229960451e-05,
      "loss": 9.962,
      "step": 184
    },
    {
      "epoch": 0.2552604346326319,
      "grad_norm": 1.3480101823806763,
      "learning_rate": 9.549110159418948e-05,
      "loss": 9.9655,
      "step": 185
    },
    {
      "epoch": 0.2566402207657813,
      "grad_norm": 1.3947056531906128,
      "learning_rate": 9.538605781763463e-05,
      "loss": 9.9447,
      "step": 186
    },
    {
      "epoch": 0.2580200068989307,
      "grad_norm": 1.5284061431884766,
      "learning_rate": 9.527986363250348e-05,
      "loss": 9.9349,
      "step": 187
    },
    {
      "epoch": 0.25939979303208005,
      "grad_norm": 1.1224453449249268,
      "learning_rate": 9.517252173051911e-05,
      "loss": 9.9821,
      "step": 188
    },
    {
      "epoch": 0.26077957916522937,
      "grad_norm": 0.9561300873756409,
      "learning_rate": 9.506403483249608e-05,
      "loss": 10.0054,
      "step": 189
    },
    {
      "epoch": 0.26215936529837874,
      "grad_norm": 1.0607901811599731,
      "learning_rate": 9.495440568827129e-05,
      "loss": 10.0007,
      "step": 190
    },
    {
      "epoch": 0.2635391514315281,
      "grad_norm": 1.1778087615966797,
      "learning_rate": 9.484363707663442e-05,
      "loss": 9.9694,
      "step": 191
    },
    {
      "epoch": 0.2649189375646775,
      "grad_norm": 1.1621663570404053,
      "learning_rate": 9.473173180525737e-05,
      "loss": 9.9717,
      "step": 192
    },
    {
      "epoch": 0.2662987236978268,
      "grad_norm": 1.1913270950317383,
      "learning_rate": 9.461869271062322e-05,
      "loss": 9.9754,
      "step": 193
    },
    {
      "epoch": 0.2676785098309762,
      "grad_norm": 1.207746982574463,
      "learning_rate": 9.450452265795423e-05,
      "loss": 9.9772,
      "step": 194
    },
    {
      "epoch": 0.26905829596412556,
      "grad_norm": 1.3012892007827759,
      "learning_rate": 9.438922454113928e-05,
      "loss": 9.9452,
      "step": 195
    },
    {
      "epoch": 0.27043808209727493,
      "grad_norm": 1.3070969581604004,
      "learning_rate": 9.42728012826605e-05,
      "loss": 9.9479,
      "step": 196
    },
    {
      "epoch": 0.2718178682304243,
      "grad_norm": 1.2681845426559448,
      "learning_rate": 9.415525583351916e-05,
      "loss": 9.9496,
      "step": 197
    },
    {
      "epoch": 0.2731976543635736,
      "grad_norm": 1.410770297050476,
      "learning_rate": 9.403659117316093e-05,
      "loss": 9.9055,
      "step": 198
    },
    {
      "epoch": 0.274577440496723,
      "grad_norm": 1.4336166381835938,
      "learning_rate": 9.391681030940031e-05,
      "loss": 9.9137,
      "step": 199
    },
    {
      "epoch": 0.2759572266298724,
      "grad_norm": 1.6375997066497803,
      "learning_rate": 9.37959162783444e-05,
      "loss": 9.8763,
      "step": 200
    },
    {
      "epoch": 0.27733701276302175,
      "grad_norm": 0.9288918972015381,
      "learning_rate": 9.367391214431598e-05,
      "loss": 9.9892,
      "step": 201
    },
    {
      "epoch": 0.27871679889617107,
      "grad_norm": 1.0222079753875732,
      "learning_rate": 9.355080099977578e-05,
      "loss": 9.9729,
      "step": 202
    },
    {
      "epoch": 0.28009658502932044,
      "grad_norm": 1.1373820304870605,
      "learning_rate": 9.342658596524416e-05,
      "loss": 9.953,
      "step": 203
    },
    {
      "epoch": 0.2814763711624698,
      "grad_norm": 1.1755473613739014,
      "learning_rate": 9.330127018922194e-05,
      "loss": 9.9483,
      "step": 204
    },
    {
      "epoch": 0.2828561572956192,
      "grad_norm": 1.1603400707244873,
      "learning_rate": 9.317485684811065e-05,
      "loss": 9.9461,
      "step": 205
    },
    {
      "epoch": 0.28423594342876857,
      "grad_norm": 1.2315527200698853,
      "learning_rate": 9.3047349146132e-05,
      "loss": 9.9464,
      "step": 206
    },
    {
      "epoch": 0.2856157295619179,
      "grad_norm": 1.2375974655151367,
      "learning_rate": 9.291875031524668e-05,
      "loss": 9.9391,
      "step": 207
    },
    {
      "epoch": 0.28699551569506726,
      "grad_norm": 1.3307669162750244,
      "learning_rate": 9.278906361507238e-05,
      "loss": 9.9194,
      "step": 208
    },
    {
      "epoch": 0.28837530182821663,
      "grad_norm": 1.3934468030929565,
      "learning_rate": 9.265829233280123e-05,
      "loss": 9.9021,
      "step": 209
    },
    {
      "epoch": 0.289755087961366,
      "grad_norm": 1.3561367988586426,
      "learning_rate": 9.252643978311649e-05,
      "loss": 9.9037,
      "step": 210
    },
    {
      "epoch": 0.29113487409451533,
      "grad_norm": 1.3786364793777466,
      "learning_rate": 9.239350930810843e-05,
      "loss": 9.8974,
      "step": 211
    },
    {
      "epoch": 0.2925146602276647,
      "grad_norm": 1.5267988443374634,
      "learning_rate": 9.225950427718975e-05,
      "loss": 9.8763,
      "step": 212
    },
    {
      "epoch": 0.2938944463608141,
      "grad_norm": 1.1748323440551758,
      "learning_rate": 9.212442808701007e-05,
      "loss": 9.9215,
      "step": 213
    },
    {
      "epoch": 0.29527423249396345,
      "grad_norm": 0.9960755705833435,
      "learning_rate": 9.19882841613699e-05,
      "loss": 9.9739,
      "step": 214
    },
    {
      "epoch": 0.29665401862711277,
      "grad_norm": 1.1767911911010742,
      "learning_rate": 9.185107595113384e-05,
      "loss": 9.9264,
      "step": 215
    },
    {
      "epoch": 0.29803380476026214,
      "grad_norm": 1.1542165279388428,
      "learning_rate": 9.171280693414307e-05,
      "loss": 9.9306,
      "step": 216
    },
    {
      "epoch": 0.2994135908934115,
      "grad_norm": 1.1749402284622192,
      "learning_rate": 9.157348061512727e-05,
      "loss": 9.9255,
      "step": 217
    },
    {
      "epoch": 0.3007933770265609,
      "grad_norm": 1.1707139015197754,
      "learning_rate": 9.143310052561571e-05,
      "loss": 9.9199,
      "step": 218
    },
    {
      "epoch": 0.30217316315971027,
      "grad_norm": 1.2597498893737793,
      "learning_rate": 9.129167022384779e-05,
      "loss": 9.9216,
      "step": 219
    },
    {
      "epoch": 0.3035529492928596,
      "grad_norm": 1.3159873485565186,
      "learning_rate": 9.114919329468282e-05,
      "loss": 9.9231,
      "step": 220
    },
    {
      "epoch": 0.30493273542600896,
      "grad_norm": 1.36250638961792,
      "learning_rate": 9.100567334950918e-05,
      "loss": 9.9024,
      "step": 221
    },
    {
      "epoch": 0.30631252155915834,
      "grad_norm": 1.3537706136703491,
      "learning_rate": 9.086111402615273e-05,
      "loss": 9.8925,
      "step": 222
    },
    {
      "epoch": 0.3076923076923077,
      "grad_norm": 1.3712196350097656,
      "learning_rate": 9.07155189887847e-05,
      "loss": 9.8801,
      "step": 223
    },
    {
      "epoch": 0.30907209382545703,
      "grad_norm": 1.4884405136108398,
      "learning_rate": 9.056889192782866e-05,
      "loss": 9.8664,
      "step": 224
    },
    {
      "epoch": 0.3104518799586064,
      "grad_norm": 1.5910539627075195,
      "learning_rate": 9.042123655986713e-05,
      "loss": 9.8475,
      "step": 225
    },
    {
      "epoch": 0.3118316660917558,
      "grad_norm": 0.9714315533638,
      "learning_rate": 9.02725566275473e-05,
      "loss": 9.9451,
      "step": 226
    },
    {
      "epoch": 0.31321145222490515,
      "grad_norm": 1.1220028400421143,
      "learning_rate": 9.012285589948617e-05,
      "loss": 9.9377,
      "step": 227
    },
    {
      "epoch": 0.3145912383580545,
      "grad_norm": 1.099237084388733,
      "learning_rate": 8.997213817017507e-05,
      "loss": 9.9214,
      "step": 228
    },
    {
      "epoch": 0.31597102449120384,
      "grad_norm": 1.145241141319275,
      "learning_rate": 8.982040725988336e-05,
      "loss": 9.9099,
      "step": 229
    },
    {
      "epoch": 0.3173508106243532,
      "grad_norm": 1.2392131090164185,
      "learning_rate": 8.966766701456177e-05,
      "loss": 9.8892,
      "step": 230
    },
    {
      "epoch": 0.3187305967575026,
      "grad_norm": 1.1928961277008057,
      "learning_rate": 8.951392130574477e-05,
      "loss": 9.9052,
      "step": 231
    },
    {
      "epoch": 0.32011038289065197,
      "grad_norm": 1.2984650135040283,
      "learning_rate": 8.935917403045251e-05,
      "loss": 9.8974,
      "step": 232
    },
    {
      "epoch": 0.3214901690238013,
      "grad_norm": 1.3149031400680542,
      "learning_rate": 8.920342911109206e-05,
      "loss": 9.8815,
      "step": 233
    },
    {
      "epoch": 0.32286995515695066,
      "grad_norm": 1.269370436668396,
      "learning_rate": 8.904669049535789e-05,
      "loss": 9.8841,
      "step": 234
    },
    {
      "epoch": 0.32424974129010004,
      "grad_norm": 1.29131019115448,
      "learning_rate": 8.888896215613191e-05,
      "loss": 9.8848,
      "step": 235
    },
    {
      "epoch": 0.3256295274232494,
      "grad_norm": 1.4416961669921875,
      "learning_rate": 8.873024809138272e-05,
      "loss": 9.8468,
      "step": 236
    },
    {
      "epoch": 0.3270093135563988,
      "grad_norm": 1.495208978652954,
      "learning_rate": 8.857055232406432e-05,
      "loss": 9.8455,
      "step": 237
    },
    {
      "epoch": 0.3283890996895481,
      "grad_norm": 1.190208911895752,
      "learning_rate": 8.840987890201403e-05,
      "loss": 9.8895,
      "step": 238
    },
    {
      "epoch": 0.3297688858226975,
      "grad_norm": 1.0209683179855347,
      "learning_rate": 8.824823189785e-05,
      "loss": 9.9142,
      "step": 239
    },
    {
      "epoch": 0.33114867195584685,
      "grad_norm": 1.0553463697433472,
      "learning_rate": 8.808561540886796e-05,
      "loss": 9.9172,
      "step": 240
    },
    {
      "epoch": 0.3325284580889962,
      "grad_norm": 1.152671217918396,
      "learning_rate": 8.792203355693732e-05,
      "loss": 9.9054,
      "step": 241
    },
    {
      "epoch": 0.33390824422214554,
      "grad_norm": 1.2217587232589722,
      "learning_rate": 8.775749048839669e-05,
      "loss": 9.8803,
      "step": 242
    },
    {
      "epoch": 0.3352880303552949,
      "grad_norm": 1.1504801511764526,
      "learning_rate": 8.759199037394887e-05,
      "loss": 9.8892,
      "step": 243
    },
    {
      "epoch": 0.3366678164884443,
      "grad_norm": 1.2767583131790161,
      "learning_rate": 8.742553740855506e-05,
      "loss": 9.8781,
      "step": 244
    },
    {
      "epoch": 0.33804760262159367,
      "grad_norm": 1.2522735595703125,
      "learning_rate": 8.725813581132852e-05,
      "loss": 9.8718,
      "step": 245
    },
    {
      "epoch": 0.33942738875474304,
      "grad_norm": 1.3533169031143188,
      "learning_rate": 8.708978982542765e-05,
      "loss": 9.8469,
      "step": 246
    },
    {
      "epoch": 0.34080717488789236,
      "grad_norm": 1.2770718336105347,
      "learning_rate": 8.692050371794848e-05,
      "loss": 9.8582,
      "step": 247
    },
    {
      "epoch": 0.34218696102104174,
      "grad_norm": 1.4352041482925415,
      "learning_rate": 8.675028177981643e-05,
      "loss": 9.8288,
      "step": 248
    },
    {
      "epoch": 0.3435667471541911,
      "grad_norm": 1.4807460308074951,
      "learning_rate": 8.657912832567761e-05,
      "loss": 9.8256,
      "step": 249
    },
    {
      "epoch": 0.3449465332873405,
      "grad_norm": 1.586320400238037,
      "learning_rate": 8.640704769378942e-05,
      "loss": 9.7933,
      "step": 250
    },
    {
      "epoch": 0.3463263194204898,
      "grad_norm": 0.9499781131744385,
      "learning_rate": 8.623404424591062e-05,
      "loss": 9.9,
      "step": 251
    },
    {
      "epoch": 0.3477061055536392,
      "grad_norm": 1.0685551166534424,
      "learning_rate": 8.606012236719073e-05,
      "loss": 9.8789,
      "step": 252
    },
    {
      "epoch": 0.34908589168678855,
      "grad_norm": 1.059103012084961,
      "learning_rate": 8.588528646605892e-05,
      "loss": 9.8886,
      "step": 253
    },
    {
      "epoch": 0.3504656778199379,
      "grad_norm": 1.1596115827560425,
      "learning_rate": 8.570954097411223e-05,
      "loss": 9.8802,
      "step": 254
    },
    {
      "epoch": 0.35184546395308725,
      "grad_norm": 1.1892733573913574,
      "learning_rate": 8.55328903460033e-05,
      "loss": 9.8839,
      "step": 255
    },
    {
      "epoch": 0.3532252500862366,
      "grad_norm": 1.239803433418274,
      "learning_rate": 8.535533905932738e-05,
      "loss": 9.8661,
      "step": 256
    },
    {
      "epoch": 0.354605036219386,
      "grad_norm": 1.2063547372817993,
      "learning_rate": 8.51768916145089e-05,
      "loss": 9.8716,
      "step": 257
    },
    {
      "epoch": 0.35598482235253537,
      "grad_norm": 1.2528349161148071,
      "learning_rate": 8.499755253468733e-05,
      "loss": 9.8567,
      "step": 258
    },
    {
      "epoch": 0.35736460848568474,
      "grad_norm": 1.2830719947814941,
      "learning_rate": 8.481732636560265e-05,
      "loss": 9.8407,
      "step": 259
    },
    {
      "epoch": 0.35874439461883406,
      "grad_norm": 1.275932788848877,
      "learning_rate": 8.463621767547998e-05,
      "loss": 9.847,
      "step": 260
    },
    {
      "epoch": 0.36012418075198344,
      "grad_norm": 1.4095110893249512,
      "learning_rate": 8.445423105491388e-05,
      "loss": 9.8133,
      "step": 261
    },
    {
      "epoch": 0.3615039668851328,
      "grad_norm": 1.4888107776641846,
      "learning_rate": 8.427137111675199e-05,
      "loss": 9.7952,
      "step": 262
    },
    {
      "epoch": 0.3628837530182822,
      "grad_norm": 1.1439027786254883,
      "learning_rate": 8.408764249597807e-05,
      "loss": 9.8783,
      "step": 263
    },
    {
      "epoch": 0.3642635391514315,
      "grad_norm": 0.9842033386230469,
      "learning_rate": 8.390304984959454e-05,
      "loss": 9.911,
      "step": 264
    },
    {
      "epoch": 0.3656433252845809,
      "grad_norm": 1.0976605415344238,
      "learning_rate": 8.371759785650444e-05,
      "loss": 9.8691,
      "step": 265
    },
    {
      "epoch": 0.36702311141773025,
      "grad_norm": 1.135542869567871,
      "learning_rate": 8.353129121739281e-05,
      "loss": 9.8513,
      "step": 266
    },
    {
      "epoch": 0.3684028975508796,
      "grad_norm": 1.1516274213790894,
      "learning_rate": 8.334413465460755e-05,
      "loss": 9.8695,
      "step": 267
    },
    {
      "epoch": 0.369782683684029,
      "grad_norm": 1.2673522233963013,
      "learning_rate": 8.315613291203976e-05,
      "loss": 9.8429,
      "step": 268
    },
    {
      "epoch": 0.3711624698171783,
      "grad_norm": 1.2768874168395996,
      "learning_rate": 8.296729075500344e-05,
      "loss": 9.8504,
      "step": 269
    },
    {
      "epoch": 0.3725422559503277,
      "grad_norm": 1.191030502319336,
      "learning_rate": 8.277761297011475e-05,
      "loss": 9.851,
      "step": 270
    },
    {
      "epoch": 0.37392204208347707,
      "grad_norm": 1.3136379718780518,
      "learning_rate": 8.258710436517061e-05,
      "loss": 9.8302,
      "step": 271
    },
    {
      "epoch": 0.37530182821662644,
      "grad_norm": 1.3082833290100098,
      "learning_rate": 8.239576976902695e-05,
      "loss": 9.8142,
      "step": 272
    },
    {
      "epoch": 0.37668161434977576,
      "grad_norm": 1.3219044208526611,
      "learning_rate": 8.220361403147619e-05,
      "loss": 9.8299,
      "step": 273
    },
    {
      "epoch": 0.37806140048292514,
      "grad_norm": 1.3385120630264282,
      "learning_rate": 8.201064202312441e-05,
      "loss": 9.8181,
      "step": 274
    },
    {
      "epoch": 0.3794411866160745,
      "grad_norm": 1.527003288269043,
      "learning_rate": 8.181685863526783e-05,
      "loss": 9.7977,
      "step": 275
    },
    {
      "epoch": 0.3808209727492239,
      "grad_norm": 0.9443302750587463,
      "learning_rate": 8.162226877976887e-05,
      "loss": 9.8974,
      "step": 276
    },
    {
      "epoch": 0.38220075888237326,
      "grad_norm": 1.0438883304595947,
      "learning_rate": 8.142687738893161e-05,
      "loss": 9.871,
      "step": 277
    },
    {
      "epoch": 0.3835805450155226,
      "grad_norm": 1.0688831806182861,
      "learning_rate": 8.123068941537682e-05,
      "loss": 9.8699,
      "step": 278
    },
    {
      "epoch": 0.38496033114867195,
      "grad_norm": 1.1512502431869507,
      "learning_rate": 8.103370983191634e-05,
      "loss": 9.8404,
      "step": 279
    },
    {
      "epoch": 0.3863401172818213,
      "grad_norm": 1.1607447862625122,
      "learning_rate": 8.083594363142717e-05,
      "loss": 9.8431,
      "step": 280
    },
    {
      "epoch": 0.3877199034149707,
      "grad_norm": 1.190542459487915,
      "learning_rate": 8.063739582672474e-05,
      "loss": 9.8434,
      "step": 281
    },
    {
      "epoch": 0.38909968954812,
      "grad_norm": 1.3068957328796387,
      "learning_rate": 8.043807145043604e-05,
      "loss": 9.8091,
      "step": 282
    },
    {
      "epoch": 0.3904794756812694,
      "grad_norm": 1.3318898677825928,
      "learning_rate": 8.023797555487188e-05,
      "loss": 9.8226,
      "step": 283
    },
    {
      "epoch": 0.39185926181441877,
      "grad_norm": 1.283746361732483,
      "learning_rate": 8.003711321189895e-05,
      "loss": 9.822,
      "step": 284
    },
    {
      "epoch": 0.39323904794756814,
      "grad_norm": 1.3395658731460571,
      "learning_rate": 7.983548951281121e-05,
      "loss": 9.8164,
      "step": 285
    },
    {
      "epoch": 0.39461883408071746,
      "grad_norm": 1.4497466087341309,
      "learning_rate": 7.963310956820085e-05,
      "loss": 9.792,
      "step": 286
    },
    {
      "epoch": 0.39599862021386684,
      "grad_norm": 1.5034538507461548,
      "learning_rate": 7.942997850782873e-05,
      "loss": 9.7734,
      "step": 287
    },
    {
      "epoch": 0.3973784063470162,
      "grad_norm": 1.080247163772583,
      "learning_rate": 7.922610148049445e-05,
      "loss": 9.8361,
      "step": 288
    },
    {
      "epoch": 0.3987581924801656,
      "grad_norm": 1.011601209640503,
      "learning_rate": 7.902148365390568e-05,
      "loss": 9.8534,
      "step": 289
    },
    {
      "epoch": 0.40013797861331496,
      "grad_norm": 1.0181865692138672,
      "learning_rate": 7.881613021454727e-05,
      "loss": 9.8572,
      "step": 290
    },
    {
      "epoch": 0.4015177647464643,
      "grad_norm": 1.1343557834625244,
      "learning_rate": 7.861004636754986e-05,
      "loss": 9.8366,
      "step": 291
    },
    {
      "epoch": 0.40289755087961365,
      "grad_norm": 1.18169105052948,
      "learning_rate": 7.840323733655778e-05,
      "loss": 9.8386,
      "step": 292
    },
    {
      "epoch": 0.404277337012763,
      "grad_norm": 1.1739155054092407,
      "learning_rate": 7.81957083635968e-05,
      "loss": 9.8157,
      "step": 293
    },
    {
      "epoch": 0.4056571231459124,
      "grad_norm": 1.198239803314209,
      "learning_rate": 7.798746470894111e-05,
      "loss": 9.8257,
      "step": 294
    },
    {
      "epoch": 0.4070369092790617,
      "grad_norm": 1.2070894241333008,
      "learning_rate": 7.777851165098012e-05,
      "loss": 9.8203,
      "step": 295
    },
    {
      "epoch": 0.4084166954122111,
      "grad_norm": 1.220963478088379,
      "learning_rate": 7.756885448608459e-05,
      "loss": 9.8194,
      "step": 296
    },
    {
      "epoch": 0.40979648154536047,
      "grad_norm": 1.353140950202942,
      "learning_rate": 7.735849852847244e-05,
      "loss": 9.7964,
      "step": 297
    },
    {
      "epoch": 0.41117626767850984,
      "grad_norm": 1.3784527778625488,
      "learning_rate": 7.714744911007394e-05,
      "loss": 9.7895,
      "step": 298
    },
    {
      "epoch": 0.4125560538116592,
      "grad_norm": 1.4098964929580688,
      "learning_rate": 7.693571158039673e-05,
      "loss": 9.7867,
      "step": 299
    },
    {
      "epoch": 0.41393583994480854,
      "grad_norm": 1.496354579925537,
      "learning_rate": 7.672329130639005e-05,
      "loss": 9.758,
      "step": 300
    },
    {
      "epoch": 0.4153156260779579,
      "grad_norm": 0.9038072228431702,
      "learning_rate": 7.651019367230885e-05,
      "loss": 9.8562,
      "step": 301
    },
    {
      "epoch": 0.4166954122111073,
      "grad_norm": 1.0591639280319214,
      "learning_rate": 7.62964240795772e-05,
      "loss": 9.8421,
      "step": 302
    },
    {
      "epoch": 0.41807519834425666,
      "grad_norm": 1.0653377771377563,
      "learning_rate": 7.608198794665144e-05,
      "loss": 9.8323,
      "step": 303
    },
    {
      "epoch": 0.419454984477406,
      "grad_norm": 1.136434555053711,
      "learning_rate": 7.586689070888284e-05,
      "loss": 9.8485,
      "step": 304
    },
    {
      "epoch": 0.42083477061055535,
      "grad_norm": 1.1839878559112549,
      "learning_rate": 7.565113781837978e-05,
      "loss": 9.814,
      "step": 305
    },
    {
      "epoch": 0.4222145567437047,
      "grad_norm": 1.2694019079208374,
      "learning_rate": 7.543473474386961e-05,
      "loss": 9.8085,
      "step": 306
    },
    {
      "epoch": 0.4235943428768541,
      "grad_norm": 1.2437082529067993,
      "learning_rate": 7.521768697056004e-05,
      "loss": 9.7813,
      "step": 307
    },
    {
      "epoch": 0.4249741290100035,
      "grad_norm": 1.259202480316162,
      "learning_rate": 7.500000000000001e-05,
      "loss": 9.7942,
      "step": 308
    },
    {
      "epoch": 0.4263539151431528,
      "grad_norm": 1.314386248588562,
      "learning_rate": 7.478167934994034e-05,
      "loss": 9.7937,
      "step": 309
    },
    {
      "epoch": 0.42773370127630217,
      "grad_norm": 1.3308764696121216,
      "learning_rate": 7.456273055419388e-05,
      "loss": 9.7809,
      "step": 310
    },
    {
      "epoch": 0.42911348740945154,
      "grad_norm": 1.3930072784423828,
      "learning_rate": 7.434315916249513e-05,
      "loss": 9.787,
      "step": 311
    },
    {
      "epoch": 0.4304932735426009,
      "grad_norm": 1.4639109373092651,
      "learning_rate": 7.412297074035967e-05,
      "loss": 9.758,
      "step": 312
    },
    {
      "epoch": 0.43187305967575024,
      "grad_norm": 1.1825257539749146,
      "learning_rate": 7.390217086894308e-05,
      "loss": 9.7985,
      "step": 313
    },
    {
      "epoch": 0.4332528458088996,
      "grad_norm": 1.009463906288147,
      "learning_rate": 7.368076514489947e-05,
      "loss": 9.8391,
      "step": 314
    },
    {
      "epoch": 0.434632631942049,
      "grad_norm": 1.0567901134490967,
      "learning_rate": 7.345875918023953e-05,
      "loss": 9.8389,
      "step": 315
    },
    {
      "epoch": 0.43601241807519836,
      "grad_norm": 1.1119906902313232,
      "learning_rate": 7.323615860218843e-05,
      "loss": 9.8111,
      "step": 316
    },
    {
      "epoch": 0.4373922042083477,
      "grad_norm": 1.1427531242370605,
      "learning_rate": 7.301296905304309e-05,
      "loss": 9.8093,
      "step": 317
    },
    {
      "epoch": 0.43877199034149705,
      "grad_norm": 1.2292324304580688,
      "learning_rate": 7.278919619002916e-05,
      "loss": 9.796,
      "step": 318
    },
    {
      "epoch": 0.44015177647464643,
      "grad_norm": 1.2442058324813843,
      "learning_rate": 7.256484568515769e-05,
      "loss": 9.793,
      "step": 319
    },
    {
      "epoch": 0.4415315626077958,
      "grad_norm": 1.3148218393325806,
      "learning_rate": 7.233992322508129e-05,
      "loss": 9.7753,
      "step": 320
    },
    {
      "epoch": 0.4429113487409452,
      "grad_norm": 1.2520376443862915,
      "learning_rate": 7.211443451095007e-05,
      "loss": 9.8126,
      "step": 321
    },
    {
      "epoch": 0.4442911348740945,
      "grad_norm": 1.2993510961532593,
      "learning_rate": 7.188838525826702e-05,
      "loss": 9.7952,
      "step": 322
    },
    {
      "epoch": 0.44567092100724387,
      "grad_norm": 1.3563902378082275,
      "learning_rate": 7.166178119674327e-05,
      "loss": 9.7627,
      "step": 323
    },
    {
      "epoch": 0.44705070714039324,
      "grad_norm": 1.4324527978897095,
      "learning_rate": 7.143462807015271e-05,
      "loss": 9.7579,
      "step": 324
    },
    {
      "epoch": 0.4484304932735426,
      "grad_norm": 1.5741610527038574,
      "learning_rate": 7.120693163618655e-05,
      "loss": 9.7154,
      "step": 325
    },
    {
      "epoch": 0.44981027940669194,
      "grad_norm": 0.9442249536514282,
      "learning_rate": 7.097869766630729e-05,
      "loss": 9.84,
      "step": 326
    },
    {
      "epoch": 0.4511900655398413,
      "grad_norm": 1.0636165142059326,
      "learning_rate": 7.074993194560242e-05,
      "loss": 9.8089,
      "step": 327
    },
    {
      "epoch": 0.4525698516729907,
      "grad_norm": 1.1189924478530884,
      "learning_rate": 7.052064027263786e-05,
      "loss": 9.806,
      "step": 328
    },
    {
      "epoch": 0.45394963780614006,
      "grad_norm": 1.1921933889389038,
      "learning_rate": 7.029082845931087e-05,
      "loss": 9.8117,
      "step": 329
    },
    {
      "epoch": 0.45532942393928943,
      "grad_norm": 1.0726195573806763,
      "learning_rate": 7.006050233070289e-05,
      "loss": 9.8132,
      "step": 330
    },
    {
      "epoch": 0.45670921007243875,
      "grad_norm": 1.2054996490478516,
      "learning_rate": 6.982966772493176e-05,
      "loss": 9.7919,
      "step": 331
    },
    {
      "epoch": 0.45808899620558813,
      "grad_norm": 1.2369368076324463,
      "learning_rate": 6.959833049300377e-05,
      "loss": 9.7817,
      "step": 332
    },
    {
      "epoch": 0.4594687823387375,
      "grad_norm": 1.3058596849441528,
      "learning_rate": 6.936649649866533e-05,
      "loss": 9.763,
      "step": 333
    },
    {
      "epoch": 0.4608485684718869,
      "grad_norm": 1.3070313930511475,
      "learning_rate": 6.91341716182545e-05,
      "loss": 9.7857,
      "step": 334
    },
    {
      "epoch": 0.4622283546050362,
      "grad_norm": 1.2690885066986084,
      "learning_rate": 6.890136174055178e-05,
      "loss": 9.7814,
      "step": 335
    },
    {
      "epoch": 0.46360814073818557,
      "grad_norm": 1.4390944242477417,
      "learning_rate": 6.866807276663106e-05,
      "loss": 9.7484,
      "step": 336
    },
    {
      "epoch": 0.46498792687133494,
      "grad_norm": 1.5518074035644531,
      "learning_rate": 6.843431060970995e-05,
      "loss": 9.7171,
      "step": 337
    },
    {
      "epoch": 0.4663677130044843,
      "grad_norm": 1.1752872467041016,
      "learning_rate": 6.820008119499992e-05,
      "loss": 9.7991,
      "step": 338
    },
    {
      "epoch": 0.4677474991376337,
      "grad_norm": 1.0036665201187134,
      "learning_rate": 6.796539045955609e-05,
      "loss": 9.8279,
      "step": 339
    },
    {
      "epoch": 0.469127285270783,
      "grad_norm": 1.0736103057861328,
      "learning_rate": 6.773024435212678e-05,
      "loss": 9.7977,
      "step": 340
    },
    {
      "epoch": 0.4705070714039324,
      "grad_norm": 1.157914400100708,
      "learning_rate": 6.749464883300272e-05,
      "loss": 9.7976,
      "step": 341
    },
    {
      "epoch": 0.47188685753708176,
      "grad_norm": 1.2276322841644287,
      "learning_rate": 6.72586098738659e-05,
      "loss": 9.7955,
      "step": 342
    },
    {
      "epoch": 0.47326664367023114,
      "grad_norm": 1.2520910501480103,
      "learning_rate": 6.70221334576383e-05,
      "loss": 9.794,
      "step": 343
    },
    {
      "epoch": 0.47464642980338045,
      "grad_norm": 1.2590833902359009,
      "learning_rate": 6.678522557833024e-05,
      "loss": 9.7907,
      "step": 344
    },
    {
      "epoch": 0.47602621593652983,
      "grad_norm": 1.2129725217819214,
      "learning_rate": 6.654789224088836e-05,
      "loss": 9.7921,
      "step": 345
    },
    {
      "epoch": 0.4774060020696792,
      "grad_norm": 1.2681193351745605,
      "learning_rate": 6.631013946104347e-05,
      "loss": 9.7804,
      "step": 346
    },
    {
      "epoch": 0.4787857882028286,
      "grad_norm": 1.3535261154174805,
      "learning_rate": 6.607197326515808e-05,
      "loss": 9.7609,
      "step": 347
    },
    {
      "epoch": 0.4801655743359779,
      "grad_norm": 1.4084678888320923,
      "learning_rate": 6.583339969007363e-05,
      "loss": 9.7531,
      "step": 348
    },
    {
      "epoch": 0.48154536046912727,
      "grad_norm": 1.4405699968338013,
      "learning_rate": 6.559442478295745e-05,
      "loss": 9.7356,
      "step": 349
    },
    {
      "epoch": 0.48292514660227664,
      "grad_norm": 1.6622633934020996,
      "learning_rate": 6.535505460114954e-05,
      "loss": 9.7062,
      "step": 350
    },
    {
      "epoch": 0.484304932735426,
      "grad_norm": 0.928680956363678,
      "learning_rate": 6.511529521200899e-05,
      "loss": 9.8214,
      "step": 351
    },
    {
      "epoch": 0.4856847188685754,
      "grad_norm": 1.0722042322158813,
      "learning_rate": 6.487515269276016e-05,
      "loss": 9.8076,
      "step": 352
    },
    {
      "epoch": 0.4870645050017247,
      "grad_norm": 1.0820056200027466,
      "learning_rate": 6.463463313033872e-05,
      "loss": 9.8085,
      "step": 353
    },
    {
      "epoch": 0.4884442911348741,
      "grad_norm": 1.1904188394546509,
      "learning_rate": 6.439374262123731e-05,
      "loss": 9.8048,
      "step": 354
    },
    {
      "epoch": 0.48982407726802346,
      "grad_norm": 1.1891876459121704,
      "learning_rate": 6.415248727135103e-05,
      "loss": 9.7957,
      "step": 355
    },
    {
      "epoch": 0.49120386340117284,
      "grad_norm": 1.1970020532608032,
      "learning_rate": 6.391087319582264e-05,
      "loss": 9.7968,
      "step": 356
    },
    {
      "epoch": 0.49258364953432215,
      "grad_norm": 1.1900156736373901,
      "learning_rate": 6.366890651888761e-05,
      "loss": 9.7892,
      "step": 357
    },
    {
      "epoch": 0.49396343566747153,
      "grad_norm": 1.2589484453201294,
      "learning_rate": 6.342659337371885e-05,
      "loss": 9.7747,
      "step": 358
    },
    {
      "epoch": 0.4953432218006209,
      "grad_norm": 1.1971404552459717,
      "learning_rate": 6.318393990227126e-05,
      "loss": 9.7989,
      "step": 359
    },
    {
      "epoch": 0.4967230079337703,
      "grad_norm": 1.3852773904800415,
      "learning_rate": 6.294095225512603e-05,
      "loss": 9.7408,
      "step": 360
    },
    {
      "epoch": 0.49810279406691965,
      "grad_norm": 1.4242140054702759,
      "learning_rate": 6.269763659133485e-05,
      "loss": 9.7427,
      "step": 361
    },
    {
      "epoch": 0.49948258020006897,
      "grad_norm": 1.488462209701538,
      "learning_rate": 6.24539990782636e-05,
      "loss": 9.7151,
      "step": 362
    },
    {
      "epoch": 0.5008623663332183,
      "grad_norm": 1.1975644826889038,
      "learning_rate": 6.221004589143615e-05,
      "loss": 9.7742,
      "step": 363
    },
    {
      "epoch": 0.5022421524663677,
      "grad_norm": 1.0266568660736084,
      "learning_rate": 6.19657832143779e-05,
      "loss": 9.8229,
      "step": 364
    },
    {
      "epoch": 0.5036219385995171,
      "grad_norm": 1.0892388820648193,
      "learning_rate": 6.172121723845883e-05,
      "loss": 9.7969,
      "step": 365
    },
    {
      "epoch": 0.5050017247326665,
      "grad_norm": 1.09567129611969,
      "learning_rate": 6.147635416273678e-05,
      "loss": 9.7972,
      "step": 366
    },
    {
      "epoch": 0.5063815108658158,
      "grad_norm": 1.1262130737304688,
      "learning_rate": 6.123120019380021e-05,
      "loss": 9.7919,
      "step": 367
    },
    {
      "epoch": 0.5077612969989651,
      "grad_norm": 1.2206709384918213,
      "learning_rate": 6.098576154561087e-05,
      "loss": 9.7791,
      "step": 368
    },
    {
      "epoch": 0.5091410831321145,
      "grad_norm": 1.2213554382324219,
      "learning_rate": 6.0740044439346345e-05,
      "loss": 9.7859,
      "step": 369
    },
    {
      "epoch": 0.5105208692652639,
      "grad_norm": 1.307381510734558,
      "learning_rate": 6.049405510324238e-05,
      "loss": 9.7559,
      "step": 370
    },
    {
      "epoch": 0.5119006553984132,
      "grad_norm": 1.25954270362854,
      "learning_rate": 6.0247799772434945e-05,
      "loss": 9.7746,
      "step": 371
    },
    {
      "epoch": 0.5132804415315626,
      "grad_norm": 1.3374980688095093,
      "learning_rate": 6.0001284688802226e-05,
      "loss": 9.7513,
      "step": 372
    },
    {
      "epoch": 0.514660227664712,
      "grad_norm": 1.456072449684143,
      "learning_rate": 5.9754516100806423e-05,
      "loss": 9.7068,
      "step": 373
    },
    {
      "epoch": 0.5160400137978614,
      "grad_norm": 1.369471788406372,
      "learning_rate": 5.950750026333534e-05,
      "loss": 9.7335,
      "step": 374
    },
    {
      "epoch": 0.5174197999310107,
      "grad_norm": 1.5743073225021362,
      "learning_rate": 5.926024343754387e-05,
      "loss": 9.7155,
      "step": 375
    },
    {
      "epoch": 0.5187995860641601,
      "grad_norm": 0.9387170076370239,
      "learning_rate": 5.90127518906953e-05,
      "loss": 9.8177,
      "step": 376
    },
    {
      "epoch": 0.5201793721973094,
      "grad_norm": 1.0029942989349365,
      "learning_rate": 5.8765031896002384e-05,
      "loss": 9.8126,
      "step": 377
    },
    {
      "epoch": 0.5215591583304587,
      "grad_norm": 1.0483084917068481,
      "learning_rate": 5.851708973246841e-05,
      "loss": 9.8039,
      "step": 378
    },
    {
      "epoch": 0.5229389444636081,
      "grad_norm": 1.1270887851715088,
      "learning_rate": 5.826893168472807e-05,
      "loss": 9.7818,
      "step": 379
    },
    {
      "epoch": 0.5243187305967575,
      "grad_norm": 1.1814236640930176,
      "learning_rate": 5.8020564042888015e-05,
      "loss": 9.7942,
      "step": 380
    },
    {
      "epoch": 0.5256985167299069,
      "grad_norm": 1.2680063247680664,
      "learning_rate": 5.777199310236758e-05,
      "loss": 9.773,
      "step": 381
    },
    {
      "epoch": 0.5270783028630562,
      "grad_norm": 1.2922178506851196,
      "learning_rate": 5.752322516373916e-05,
      "loss": 9.7523,
      "step": 382
    },
    {
      "epoch": 0.5284580889962056,
      "grad_norm": 1.2834199666976929,
      "learning_rate": 5.727426653256843e-05,
      "loss": 9.7698,
      "step": 383
    },
    {
      "epoch": 0.529837875129355,
      "grad_norm": 1.3297151327133179,
      "learning_rate": 5.702512351925464e-05,
      "loss": 9.7516,
      "step": 384
    },
    {
      "epoch": 0.5312176612625044,
      "grad_norm": 1.330753207206726,
      "learning_rate": 5.67758024388706e-05,
      "loss": 9.7552,
      "step": 385
    },
    {
      "epoch": 0.5325974473956536,
      "grad_norm": 1.3623725175857544,
      "learning_rate": 5.6526309611002594e-05,
      "loss": 9.7394,
      "step": 386
    },
    {
      "epoch": 0.533977233528803,
      "grad_norm": 1.4568856954574585,
      "learning_rate": 5.6276651359590235e-05,
      "loss": 9.7095,
      "step": 387
    },
    {
      "epoch": 0.5353570196619524,
      "grad_norm": 1.1733113527297974,
      "learning_rate": 5.602683401276615e-05,
      "loss": 9.7655,
      "step": 388
    },
    {
      "epoch": 0.5367368057951017,
      "grad_norm": 1.0663663148880005,
      "learning_rate": 5.577686390269562e-05,
      "loss": 9.7791,
      "step": 389
    },
    {
      "epoch": 0.5381165919282511,
      "grad_norm": 1.112555742263794,
      "learning_rate": 5.5526747365416e-05,
      "loss": 9.7898,
      "step": 390
    },
    {
      "epoch": 0.5394963780614005,
      "grad_norm": 1.11070716381073,
      "learning_rate": 5.527649074067618e-05,
      "loss": 9.7931,
      "step": 391
    },
    {
      "epoch": 0.5408761641945499,
      "grad_norm": 1.204613447189331,
      "learning_rate": 5.502610037177586e-05,
      "loss": 9.7795,
      "step": 392
    },
    {
      "epoch": 0.5422559503276992,
      "grad_norm": 1.2208102941513062,
      "learning_rate": 5.4775582605404793e-05,
      "loss": 9.7933,
      "step": 393
    },
    {
      "epoch": 0.5436357364608486,
      "grad_norm": 1.1350970268249512,
      "learning_rate": 5.45249437914819e-05,
      "loss": 9.7878,
      "step": 394
    },
    {
      "epoch": 0.5450155225939979,
      "grad_norm": 1.3300670385360718,
      "learning_rate": 5.42741902829943e-05,
      "loss": 9.7566,
      "step": 395
    },
    {
      "epoch": 0.5463953087271473,
      "grad_norm": 1.3456814289093018,
      "learning_rate": 5.402332843583631e-05,
      "loss": 9.7377,
      "step": 396
    },
    {
      "epoch": 0.5477750948602966,
      "grad_norm": 1.2675468921661377,
      "learning_rate": 5.3772364608648306e-05,
      "loss": 9.7507,
      "step": 397
    },
    {
      "epoch": 0.549154880993446,
      "grad_norm": 1.3765349388122559,
      "learning_rate": 5.35213051626556e-05,
      "loss": 9.7327,
      "step": 398
    },
    {
      "epoch": 0.5505346671265954,
      "grad_norm": 1.4263334274291992,
      "learning_rate": 5.327015646150716e-05,
      "loss": 9.7307,
      "step": 399
    },
    {
      "epoch": 0.5519144532597448,
      "grad_norm": 1.5264482498168945,
      "learning_rate": 5.3018924871114305e-05,
      "loss": 9.7105,
      "step": 400
    },
    {
      "epoch": 0.5532942393928941,
      "grad_norm": 0.9876317977905273,
      "learning_rate": 5.276761675948939e-05,
      "loss": 9.802,
      "step": 401
    },
    {
      "epoch": 0.5546740255260435,
      "grad_norm": 1.0349249839782715,
      "learning_rate": 5.2516238496584335e-05,
      "loss": 9.8119,
      "step": 402
    },
    {
      "epoch": 0.5560538116591929,
      "grad_norm": 1.1278293132781982,
      "learning_rate": 5.226479645412923e-05,
      "loss": 9.783,
      "step": 403
    },
    {
      "epoch": 0.5574335977923421,
      "grad_norm": 1.1814225912094116,
      "learning_rate": 5.201329700547076e-05,
      "loss": 9.7796,
      "step": 404
    },
    {
      "epoch": 0.5588133839254915,
      "grad_norm": 1.190556287765503,
      "learning_rate": 5.176174652541074e-05,
      "loss": 9.787,
      "step": 405
    },
    {
      "epoch": 0.5601931700586409,
      "grad_norm": 1.188617467880249,
      "learning_rate": 5.151015139004445e-05,
      "loss": 9.7872,
      "step": 406
    },
    {
      "epoch": 0.5615729561917903,
      "grad_norm": 1.2366420030593872,
      "learning_rate": 5.1258517976599075e-05,
      "loss": 9.7721,
      "step": 407
    },
    {
      "epoch": 0.5629527423249396,
      "grad_norm": 1.3134856224060059,
      "learning_rate": 5.100685266327202e-05,
      "loss": 9.7565,
      "step": 408
    },
    {
      "epoch": 0.564332528458089,
      "grad_norm": 1.3218770027160645,
      "learning_rate": 5.0755161829069294e-05,
      "loss": 9.7603,
      "step": 409
    },
    {
      "epoch": 0.5657123145912384,
      "grad_norm": 1.3142060041427612,
      "learning_rate": 5.0503451853643776e-05,
      "loss": 9.7497,
      "step": 410
    },
    {
      "epoch": 0.5670921007243878,
      "grad_norm": 1.4091325998306274,
      "learning_rate": 5.0251729117133514e-05,
      "loss": 9.7189,
      "step": 411
    },
    {
      "epoch": 0.5684718868575371,
      "grad_norm": 1.5563430786132812,
      "learning_rate": 5e-05,
      "loss": 9.7153,
      "step": 412
    },
    {
      "epoch": 0.5698516729906864,
      "grad_norm": 1.032839059829712,
      "learning_rate": 4.97482708828665e-05,
      "loss": 9.8046,
      "step": 413
    },
    {
      "epoch": 0.5712314591238358,
      "grad_norm": 0.999853789806366,
      "learning_rate": 4.949654814635623e-05,
      "loss": 9.8077,
      "step": 414
    },
    {
      "epoch": 0.5726112452569851,
      "grad_norm": 1.0494047403335571,
      "learning_rate": 4.924483817093071e-05,
      "loss": 9.7826,
      "step": 415
    },
    {
      "epoch": 0.5739910313901345,
      "grad_norm": 1.1431876420974731,
      "learning_rate": 4.899314733672799e-05,
      "loss": 9.7938,
      "step": 416
    },
    {
      "epoch": 0.5753708175232839,
      "grad_norm": 1.16597580909729,
      "learning_rate": 4.874148202340095e-05,
      "loss": 9.778,
      "step": 417
    },
    {
      "epoch": 0.5767506036564333,
      "grad_norm": 1.2290675640106201,
      "learning_rate": 4.848984860995557e-05,
      "loss": 9.7661,
      "step": 418
    },
    {
      "epoch": 0.5781303897895826,
      "grad_norm": 1.2121100425720215,
      "learning_rate": 4.823825347458928e-05,
      "loss": 9.7712,
      "step": 419
    },
    {
      "epoch": 0.579510175922732,
      "grad_norm": 1.2210222482681274,
      "learning_rate": 4.798670299452926e-05,
      "loss": 9.7731,
      "step": 420
    },
    {
      "epoch": 0.5808899620558814,
      "grad_norm": 1.3147337436676025,
      "learning_rate": 4.7735203545870794e-05,
      "loss": 9.7444,
      "step": 421
    },
    {
      "epoch": 0.5822697481890307,
      "grad_norm": 1.3694804906845093,
      "learning_rate": 4.748376150341566e-05,
      "loss": 9.7403,
      "step": 422
    },
    {
      "epoch": 0.58364953432218,
      "grad_norm": 1.3305609226226807,
      "learning_rate": 4.723238324051061e-05,
      "loss": 9.7368,
      "step": 423
    },
    {
      "epoch": 0.5850293204553294,
      "grad_norm": 1.6528981924057007,
      "learning_rate": 4.6981075128885693e-05,
      "loss": 9.6905,
      "step": 424
    },
    {
      "epoch": 0.5864091065884788,
      "grad_norm": 1.5188294649124146,
      "learning_rate": 4.6729843538492847e-05,
      "loss": 9.6925,
      "step": 425
    },
    {
      "epoch": 0.5877888927216282,
      "grad_norm": 0.9169396162033081,
      "learning_rate": 4.6478694837344404e-05,
      "loss": 9.8025,
      "step": 426
    },
    {
      "epoch": 0.5891686788547775,
      "grad_norm": 1.0337716341018677,
      "learning_rate": 4.6227635391351705e-05,
      "loss": 9.7932,
      "step": 427
    },
    {
      "epoch": 0.5905484649879269,
      "grad_norm": 1.0950019359588623,
      "learning_rate": 4.597667156416371e-05,
      "loss": 9.7856,
      "step": 428
    },
    {
      "epoch": 0.5919282511210763,
      "grad_norm": 1.0964305400848389,
      "learning_rate": 4.572580971700571e-05,
      "loss": 9.7829,
      "step": 429
    },
    {
      "epoch": 0.5933080372542255,
      "grad_norm": 1.2117314338684082,
      "learning_rate": 4.547505620851811e-05,
      "loss": 9.7921,
      "step": 430
    },
    {
      "epoch": 0.5946878233873749,
      "grad_norm": 1.150464653968811,
      "learning_rate": 4.522441739459521e-05,
      "loss": 9.7754,
      "step": 431
    },
    {
      "epoch": 0.5960676095205243,
      "grad_norm": 1.2205294370651245,
      "learning_rate": 4.4973899628224154e-05,
      "loss": 9.7841,
      "step": 432
    },
    {
      "epoch": 0.5974473956536737,
      "grad_norm": 1.2993420362472534,
      "learning_rate": 4.472350925932384e-05,
      "loss": 9.7478,
      "step": 433
    },
    {
      "epoch": 0.598827181786823,
      "grad_norm": 1.2837934494018555,
      "learning_rate": 4.4473252634584015e-05,
      "loss": 9.7516,
      "step": 434
    },
    {
      "epoch": 0.6002069679199724,
      "grad_norm": 1.3853527307510376,
      "learning_rate": 4.422313609730439e-05,
      "loss": 9.7443,
      "step": 435
    },
    {
      "epoch": 0.6015867540531218,
      "grad_norm": 1.4042359590530396,
      "learning_rate": 4.397316598723385e-05,
      "loss": 9.7504,
      "step": 436
    },
    {
      "epoch": 0.6029665401862712,
      "grad_norm": 1.4547556638717651,
      "learning_rate": 4.372334864040979e-05,
      "loss": 9.7129,
      "step": 437
    },
    {
      "epoch": 0.6043463263194205,
      "grad_norm": 1.2045806646347046,
      "learning_rate": 4.347369038899744e-05,
      "loss": 9.7568,
      "step": 438
    },
    {
      "epoch": 0.6057261124525698,
      "grad_norm": 0.9907721877098083,
      "learning_rate": 4.322419756112943e-05,
      "loss": 9.8035,
      "step": 439
    },
    {
      "epoch": 0.6071058985857192,
      "grad_norm": 1.065327763557434,
      "learning_rate": 4.297487648074538e-05,
      "loss": 9.7859,
      "step": 440
    },
    {
      "epoch": 0.6084856847188685,
      "grad_norm": 1.110983967781067,
      "learning_rate": 4.2725733467431595e-05,
      "loss": 9.7891,
      "step": 441
    },
    {
      "epoch": 0.6098654708520179,
      "grad_norm": 1.1270619630813599,
      "learning_rate": 4.2476774836260845e-05,
      "loss": 9.7671,
      "step": 442
    },
    {
      "epoch": 0.6112452569851673,
      "grad_norm": 1.204433560371399,
      "learning_rate": 4.222800689763241e-05,
      "loss": 9.7617,
      "step": 443
    },
    {
      "epoch": 0.6126250431183167,
      "grad_norm": 1.243435263633728,
      "learning_rate": 4.197943595711198e-05,
      "loss": 9.7765,
      "step": 444
    },
    {
      "epoch": 0.614004829251466,
      "grad_norm": 1.2665642499923706,
      "learning_rate": 4.1731068315271934e-05,
      "loss": 9.7423,
      "step": 445
    },
    {
      "epoch": 0.6153846153846154,
      "grad_norm": 1.2910243272781372,
      "learning_rate": 4.1482910267531585e-05,
      "loss": 9.762,
      "step": 446
    },
    {
      "epoch": 0.6167644015177648,
      "grad_norm": 1.3219455480575562,
      "learning_rate": 4.123496810399763e-05,
      "loss": 9.752,
      "step": 447
    },
    {
      "epoch": 0.6181441876509141,
      "grad_norm": 1.3940999507904053,
      "learning_rate": 4.0987248109304714e-05,
      "loss": 9.7159,
      "step": 448
    },
    {
      "epoch": 0.6195239737840634,
      "grad_norm": 1.3925243616104126,
      "learning_rate": 4.0739756562456134e-05,
      "loss": 9.7123,
      "step": 449
    },
    {
      "epoch": 0.6209037599172128,
      "grad_norm": 1.5753551721572876,
      "learning_rate": 4.049249973666467e-05,
      "loss": 9.6993,
      "step": 450
    },
    {
      "epoch": 0.6222835460503622,
      "grad_norm": 0.9300054907798767,
      "learning_rate": 4.0245483899193595e-05,
      "loss": 9.8026,
      "step": 451
    },
    {
      "epoch": 0.6236633321835116,
      "grad_norm": 1.050060749053955,
      "learning_rate": 3.9998715311197785e-05,
      "loss": 9.7811,
      "step": 452
    },
    {
      "epoch": 0.6250431183166609,
      "grad_norm": 1.0960569381713867,
      "learning_rate": 3.9752200227565066e-05,
      "loss": 9.7914,
      "step": 453
    },
    {
      "epoch": 0.6264229044498103,
      "grad_norm": 1.1422584056854248,
      "learning_rate": 3.950594489675763e-05,
      "loss": 9.7838,
      "step": 454
    },
    {
      "epoch": 0.6278026905829597,
      "grad_norm": 1.1648497581481934,
      "learning_rate": 3.925995556065367e-05,
      "loss": 9.7672,
      "step": 455
    },
    {
      "epoch": 0.629182476716109,
      "grad_norm": 1.2434571981430054,
      "learning_rate": 3.901423845438916e-05,
      "loss": 9.7519,
      "step": 456
    },
    {
      "epoch": 0.6305622628492583,
      "grad_norm": 1.1653773784637451,
      "learning_rate": 3.876879980619982e-05,
      "loss": 9.7658,
      "step": 457
    },
    {
      "epoch": 0.6319420489824077,
      "grad_norm": 1.3217308521270752,
      "learning_rate": 3.852364583726324e-05,
      "loss": 9.7428,
      "step": 458
    },
    {
      "epoch": 0.6333218351155571,
      "grad_norm": 1.3371676206588745,
      "learning_rate": 3.827878276154119e-05,
      "loss": 9.7647,
      "step": 459
    },
    {
      "epoch": 0.6347016212487064,
      "grad_norm": 1.3296242952346802,
      "learning_rate": 3.803421678562213e-05,
      "loss": 9.7292,
      "step": 460
    },
    {
      "epoch": 0.6360814073818558,
      "grad_norm": 1.344732642173767,
      "learning_rate": 3.7789954108563843e-05,
      "loss": 9.7548,
      "step": 461
    },
    {
      "epoch": 0.6374611935150052,
      "grad_norm": 1.5095053911209106,
      "learning_rate": 3.754600092173641e-05,
      "loss": 9.7073,
      "step": 462
    },
    {
      "epoch": 0.6388409796481546,
      "grad_norm": 1.2287043333053589,
      "learning_rate": 3.7302363408665156e-05,
      "loss": 9.7332,
      "step": 463
    },
    {
      "epoch": 0.6402207657813039,
      "grad_norm": 1.0486985445022583,
      "learning_rate": 3.705904774487396e-05,
      "loss": 9.7796,
      "step": 464
    },
    {
      "epoch": 0.6416005519144533,
      "grad_norm": 1.0872530937194824,
      "learning_rate": 3.6816060097728756e-05,
      "loss": 9.7794,
      "step": 465
    },
    {
      "epoch": 0.6429803380476026,
      "grad_norm": 1.1105153560638428,
      "learning_rate": 3.657340662628116e-05,
      "loss": 9.7643,
      "step": 466
    },
    {
      "epoch": 0.644360124180752,
      "grad_norm": 1.229119896888733,
      "learning_rate": 3.63310934811124e-05,
      "loss": 9.7515,
      "step": 467
    },
    {
      "epoch": 0.6457399103139013,
      "grad_norm": 1.204601526260376,
      "learning_rate": 3.608912680417737e-05,
      "loss": 9.7704,
      "step": 468
    },
    {
      "epoch": 0.6471196964470507,
      "grad_norm": 1.1653997898101807,
      "learning_rate": 3.5847512728648985e-05,
      "loss": 9.7585,
      "step": 469
    },
    {
      "epoch": 0.6484994825802001,
      "grad_norm": 1.305881381034851,
      "learning_rate": 3.5606257378762696e-05,
      "loss": 9.7481,
      "step": 470
    },
    {
      "epoch": 0.6498792687133494,
      "grad_norm": 1.3310476541519165,
      "learning_rate": 3.5365366869661296e-05,
      "loss": 9.7463,
      "step": 471
    },
    {
      "epoch": 0.6512590548464988,
      "grad_norm": 1.2370867729187012,
      "learning_rate": 3.512484730723986e-05,
      "loss": 9.75,
      "step": 472
    },
    {
      "epoch": 0.6526388409796482,
      "grad_norm": 1.4093278646469116,
      "learning_rate": 3.488470478799103e-05,
      "loss": 9.7231,
      "step": 473
    },
    {
      "epoch": 0.6540186271127976,
      "grad_norm": 1.4781489372253418,
      "learning_rate": 3.464494539885047e-05,
      "loss": 9.695,
      "step": 474
    },
    {
      "epoch": 0.6553984132459468,
      "grad_norm": 1.5731643438339233,
      "learning_rate": 3.4405575217042557e-05,
      "loss": 9.6821,
      "step": 475
    },
    {
      "epoch": 0.6567781993790962,
      "grad_norm": 0.9464230537414551,
      "learning_rate": 3.4166600309926387e-05,
      "loss": 9.8025,
      "step": 476
    },
    {
      "epoch": 0.6581579855122456,
      "grad_norm": 1.003790259361267,
      "learning_rate": 3.392802673484193e-05,
      "loss": 9.78,
      "step": 477
    },
    {
      "epoch": 0.659537771645395,
      "grad_norm": 1.1202031373977661,
      "learning_rate": 3.368986053895655e-05,
      "loss": 9.7648,
      "step": 478
    },
    {
      "epoch": 0.6609175577785443,
      "grad_norm": 1.1496732234954834,
      "learning_rate": 3.345210775911166e-05,
      "loss": 9.7602,
      "step": 479
    },
    {
      "epoch": 0.6622973439116937,
      "grad_norm": 1.1272746324539185,
      "learning_rate": 3.3214774421669774e-05,
      "loss": 9.7809,
      "step": 480
    },
    {
      "epoch": 0.6636771300448431,
      "grad_norm": 1.2115309238433838,
      "learning_rate": 3.297786654236169e-05,
      "loss": 9.7554,
      "step": 481
    },
    {
      "epoch": 0.6650569161779925,
      "grad_norm": 1.1891155242919922,
      "learning_rate": 3.2741390126134106e-05,
      "loss": 9.766,
      "step": 482
    },
    {
      "epoch": 0.6664367023111418,
      "grad_norm": 1.2283670902252197,
      "learning_rate": 3.250535116699729e-05,
      "loss": 9.7619,
      "step": 483
    },
    {
      "epoch": 0.6678164884442911,
      "grad_norm": 1.259109377861023,
      "learning_rate": 3.226975564787322e-05,
      "loss": 9.7414,
      "step": 484
    },
    {
      "epoch": 0.6691962745774405,
      "grad_norm": 1.322930932044983,
      "learning_rate": 3.2034609540443915e-05,
      "loss": 9.7465,
      "step": 485
    },
    {
      "epoch": 0.6705760607105898,
      "grad_norm": 1.3455944061279297,
      "learning_rate": 3.179991880500009e-05,
      "loss": 9.7221,
      "step": 486
    },
    {
      "epoch": 0.6719558468437392,
      "grad_norm": 1.5582841634750366,
      "learning_rate": 3.1565689390290066e-05,
      "loss": 9.6821,
      "step": 487
    },
    {
      "epoch": 0.6733356329768886,
      "grad_norm": 1.1807150840759277,
      "learning_rate": 3.133192723336895e-05,
      "loss": 9.7562,
      "step": 488
    },
    {
      "epoch": 0.674715419110038,
      "grad_norm": 0.9975462555885315,
      "learning_rate": 3.109863825944823e-05,
      "loss": 9.7919,
      "step": 489
    },
    {
      "epoch": 0.6760952052431873,
      "grad_norm": 1.0711045265197754,
      "learning_rate": 3.086582838174551e-05,
      "loss": 9.7813,
      "step": 490
    },
    {
      "epoch": 0.6774749913763367,
      "grad_norm": 1.1586861610412598,
      "learning_rate": 3.063350350133467e-05,
      "loss": 9.7722,
      "step": 491
    },
    {
      "epoch": 0.6788547775094861,
      "grad_norm": 1.228026270866394,
      "learning_rate": 3.0401669506996256e-05,
      "loss": 9.7511,
      "step": 492
    },
    {
      "epoch": 0.6802345636426353,
      "grad_norm": 1.205870270729065,
      "learning_rate": 3.0170332275068246e-05,
      "loss": 9.7468,
      "step": 493
    },
    {
      "epoch": 0.6816143497757847,
      "grad_norm": 1.283031702041626,
      "learning_rate": 2.9939497669297112e-05,
      "loss": 9.7567,
      "step": 494
    },
    {
      "epoch": 0.6829941359089341,
      "grad_norm": 1.2276543378829956,
      "learning_rate": 2.970917154068914e-05,
      "loss": 9.7648,
      "step": 495
    },
    {
      "epoch": 0.6843739220420835,
      "grad_norm": 1.2356555461883545,
      "learning_rate": 2.9479359727362173e-05,
      "loss": 9.7521,
      "step": 496
    },
    {
      "epoch": 0.6857537081752328,
      "grad_norm": 1.301045298576355,
      "learning_rate": 2.92500680543976e-05,
      "loss": 9.7518,
      "step": 497
    },
    {
      "epoch": 0.6871334943083822,
      "grad_norm": 1.424596905708313,
      "learning_rate": 2.9021302333692734e-05,
      "loss": 9.7215,
      "step": 498
    },
    {
      "epoch": 0.6885132804415316,
      "grad_norm": 1.409056305885315,
      "learning_rate": 2.879306836381345e-05,
      "loss": 9.7073,
      "step": 499
    },
    {
      "epoch": 0.689893066574681,
      "grad_norm": 1.6761122941970825,
      "learning_rate": 2.8565371929847284e-05,
      "loss": 9.6887,
      "step": 500
    },
    {
      "epoch": 0.6912728527078302,
      "grad_norm": 0.9244851469993591,
      "learning_rate": 2.8338218803256745e-05,
      "loss": 9.8076,
      "step": 501
    },
    {
      "epoch": 0.6926526388409796,
      "grad_norm": 1.1031217575073242,
      "learning_rate": 2.811161474173297e-05,
      "loss": 9.7844,
      "step": 502
    },
    {
      "epoch": 0.694032424974129,
      "grad_norm": 1.1576576232910156,
      "learning_rate": 2.7885565489049946e-05,
      "loss": 9.7647,
      "step": 503
    },
    {
      "epoch": 0.6954122111072784,
      "grad_norm": 1.1267688274383545,
      "learning_rate": 2.7660076774918708e-05,
      "loss": 9.786,
      "step": 504
    },
    {
      "epoch": 0.6967919972404277,
      "grad_norm": 1.165797472000122,
      "learning_rate": 2.7435154314842338e-05,
      "loss": 9.7736,
      "step": 505
    },
    {
      "epoch": 0.6981717833735771,
      "grad_norm": 1.1887449026107788,
      "learning_rate": 2.7210803809970853e-05,
      "loss": 9.7515,
      "step": 506
    },
    {
      "epoch": 0.6995515695067265,
      "grad_norm": 1.2904964685440063,
      "learning_rate": 2.6987030946956936e-05,
      "loss": 9.7602,
      "step": 507
    },
    {
      "epoch": 0.7009313556398759,
      "grad_norm": 1.2434381246566772,
      "learning_rate": 2.6763841397811573e-05,
      "loss": 9.7476,
      "step": 508
    },
    {
      "epoch": 0.7023111417730252,
      "grad_norm": 1.282375454902649,
      "learning_rate": 2.654124081976048e-05,
      "loss": 9.7474,
      "step": 509
    },
    {
      "epoch": 0.7036909279061745,
      "grad_norm": 1.3851263523101807,
      "learning_rate": 2.631923485510054e-05,
      "loss": 9.7208,
      "step": 510
    },
    {
      "epoch": 0.7050707140393239,
      "grad_norm": 1.4085386991500854,
      "learning_rate": 2.6097829131056907e-05,
      "loss": 9.7317,
      "step": 511
    },
    {
      "epoch": 0.7064505001724732,
      "grad_norm": 1.4174658060073853,
      "learning_rate": 2.587702925964034e-05,
      "loss": 9.7054,
      "step": 512
    },
    {
      "epoch": 0.7078302863056226,
      "grad_norm": 1.1571439504623413,
      "learning_rate": 2.565684083750488e-05,
      "loss": 9.7617,
      "step": 513
    },
    {
      "epoch": 0.709210072438772,
      "grad_norm": 1.0106278657913208,
      "learning_rate": 2.5437269445806145e-05,
      "loss": 9.7957,
      "step": 514
    },
    {
      "epoch": 0.7105898585719214,
      "grad_norm": 1.1029208898544312,
      "learning_rate": 2.5218320650059657e-05,
      "loss": 9.7729,
      "step": 515
    },
    {
      "epoch": 0.7119696447050707,
      "grad_norm": 1.134572148323059,
      "learning_rate": 2.500000000000001e-05,
      "loss": 9.7702,
      "step": 516
    },
    {
      "epoch": 0.7133494308382201,
      "grad_norm": 1.1734071969985962,
      "learning_rate": 2.478231302943997e-05,
      "loss": 9.782,
      "step": 517
    },
    {
      "epoch": 0.7147292169713695,
      "grad_norm": 1.220053791999817,
      "learning_rate": 2.4565265256130394e-05,
      "loss": 9.7545,
      "step": 518
    },
    {
      "epoch": 0.7161090031045187,
      "grad_norm": 1.2671356201171875,
      "learning_rate": 2.4348862181620224e-05,
      "loss": 9.7482,
      "step": 519
    },
    {
      "epoch": 0.7174887892376681,
      "grad_norm": 1.2364708185195923,
      "learning_rate": 2.4133109291117156e-05,
      "loss": 9.7619,
      "step": 520
    },
    {
      "epoch": 0.7188685753708175,
      "grad_norm": 1.251885175704956,
      "learning_rate": 2.391801205334856e-05,
      "loss": 9.7535,
      "step": 521
    },
    {
      "epoch": 0.7202483615039669,
      "grad_norm": 1.345314621925354,
      "learning_rate": 2.3703575920422795e-05,
      "loss": 9.7491,
      "step": 522
    },
    {
      "epoch": 0.7216281476371162,
      "grad_norm": 1.4312492609024048,
      "learning_rate": 2.3489806327691155e-05,
      "loss": 9.7244,
      "step": 523
    },
    {
      "epoch": 0.7230079337702656,
      "grad_norm": 1.4245972633361816,
      "learning_rate": 2.3276708693609943e-05,
      "loss": 9.7038,
      "step": 524
    },
    {
      "epoch": 0.724387719903415,
      "grad_norm": 1.5328452587127686,
      "learning_rate": 2.3064288419603292e-05,
      "loss": 9.6869,
      "step": 525
    },
    {
      "epoch": 0.7257675060365644,
      "grad_norm": 0.9019702672958374,
      "learning_rate": 2.2852550889926067e-05,
      "loss": 9.8176,
      "step": 526
    },
    {
      "epoch": 0.7271472921697137,
      "grad_norm": 1.0406774282455444,
      "learning_rate": 2.2641501471527595e-05,
      "loss": 9.7709,
      "step": 527
    },
    {
      "epoch": 0.728527078302863,
      "grad_norm": 1.0880190134048462,
      "learning_rate": 2.243114551391542e-05,
      "loss": 9.7687,
      "step": 528
    },
    {
      "epoch": 0.7299068644360124,
      "grad_norm": 1.1651344299316406,
      "learning_rate": 2.2221488349019903e-05,
      "loss": 9.7504,
      "step": 529
    },
    {
      "epoch": 0.7312866505691618,
      "grad_norm": 1.1736459732055664,
      "learning_rate": 2.20125352910589e-05,
      "loss": 9.7507,
      "step": 530
    },
    {
      "epoch": 0.7326664367023111,
      "grad_norm": 1.196784257888794,
      "learning_rate": 2.18042916364032e-05,
      "loss": 9.7769,
      "step": 531
    },
    {
      "epoch": 0.7340462228354605,
      "grad_norm": 1.2587947845458984,
      "learning_rate": 2.1596762663442218e-05,
      "loss": 9.7598,
      "step": 532
    },
    {
      "epoch": 0.7354260089686099,
      "grad_norm": 1.1886099576950073,
      "learning_rate": 2.1389953632450145e-05,
      "loss": 9.7561,
      "step": 533
    },
    {
      "epoch": 0.7368057951017593,
      "grad_norm": 1.3150951862335205,
      "learning_rate": 2.118386978545274e-05,
      "loss": 9.7332,
      "step": 534
    },
    {
      "epoch": 0.7381855812349086,
      "grad_norm": 1.3374465703964233,
      "learning_rate": 2.097851634609434e-05,
      "loss": 9.7546,
      "step": 535
    },
    {
      "epoch": 0.739565367368058,
      "grad_norm": 1.3926987648010254,
      "learning_rate": 2.077389851950557e-05,
      "loss": 9.7265,
      "step": 536
    },
    {
      "epoch": 0.7409451535012073,
      "grad_norm": 1.4018597602844238,
      "learning_rate": 2.0570021492171264e-05,
      "loss": 9.7042,
      "step": 537
    },
    {
      "epoch": 0.7423249396343566,
      "grad_norm": 1.165271520614624,
      "learning_rate": 2.0366890431799167e-05,
      "loss": 9.7577,
      "step": 538
    },
    {
      "epoch": 0.743704725767506,
      "grad_norm": 1.0279182195663452,
      "learning_rate": 2.0164510487188798e-05,
      "loss": 9.7801,
      "step": 539
    },
    {
      "epoch": 0.7450845119006554,
      "grad_norm": 1.0791254043579102,
      "learning_rate": 1.996288678810105e-05,
      "loss": 9.7826,
      "step": 540
    },
    {
      "epoch": 0.7464642980338048,
      "grad_norm": 1.1344528198242188,
      "learning_rate": 1.976202444512813e-05,
      "loss": 9.7708,
      "step": 541
    },
    {
      "epoch": 0.7478440841669541,
      "grad_norm": 1.1656285524368286,
      "learning_rate": 1.9561928549563968e-05,
      "loss": 9.7673,
      "step": 542
    },
    {
      "epoch": 0.7492238703001035,
      "grad_norm": 1.1736102104187012,
      "learning_rate": 1.9362604173275268e-05,
      "loss": 9.7814,
      "step": 543
    },
    {
      "epoch": 0.7506036564332529,
      "grad_norm": 1.2199715375900269,
      "learning_rate": 1.9164056368572846e-05,
      "loss": 9.7803,
      "step": 544
    },
    {
      "epoch": 0.7519834425664023,
      "grad_norm": 1.196496844291687,
      "learning_rate": 1.8966290168083677e-05,
      "loss": 9.7406,
      "step": 545
    },
    {
      "epoch": 0.7533632286995515,
      "grad_norm": 1.2514899969100952,
      "learning_rate": 1.87693105846232e-05,
      "loss": 9.7567,
      "step": 546
    },
    {
      "epoch": 0.7547430148327009,
      "grad_norm": 1.3064500093460083,
      "learning_rate": 1.8573122611068406e-05,
      "loss": 9.7289,
      "step": 547
    },
    {
      "epoch": 0.7561228009658503,
      "grad_norm": 1.400439739227295,
      "learning_rate": 1.837773122023114e-05,
      "loss": 9.7237,
      "step": 548
    },
    {
      "epoch": 0.7575025870989996,
      "grad_norm": 1.3692952394485474,
      "learning_rate": 1.818314136473217e-05,
      "loss": 9.7226,
      "step": 549
    },
    {
      "epoch": 0.758882373232149,
      "grad_norm": 1.5424178838729858,
      "learning_rate": 1.7989357976875603e-05,
      "loss": 9.6992,
      "step": 550
    },
    {
      "epoch": 0.7602621593652984,
      "grad_norm": 0.8980839848518372,
      "learning_rate": 1.7796385968523815e-05,
      "loss": 9.811,
      "step": 551
    },
    {
      "epoch": 0.7616419454984478,
      "grad_norm": 0.9639766216278076,
      "learning_rate": 1.760423023097307e-05,
      "loss": 9.7805,
      "step": 552
    },
    {
      "epoch": 0.7630217316315971,
      "grad_norm": 1.0791096687316895,
      "learning_rate": 1.741289563482939e-05,
      "loss": 9.7779,
      "step": 553
    },
    {
      "epoch": 0.7644015177647465,
      "grad_norm": 1.1886143684387207,
      "learning_rate": 1.7222387029885268e-05,
      "loss": 9.7533,
      "step": 554
    },
    {
      "epoch": 0.7657813038978958,
      "grad_norm": 1.2121399641036987,
      "learning_rate": 1.703270924499656e-05,
      "loss": 9.756,
      "step": 555
    },
    {
      "epoch": 0.7671610900310452,
      "grad_norm": 1.2668524980545044,
      "learning_rate": 1.684386708796025e-05,
      "loss": 9.7412,
      "step": 556
    },
    {
      "epoch": 0.7685408761641945,
      "grad_norm": 1.212928056716919,
      "learning_rate": 1.665586534539246e-05,
      "loss": 9.7641,
      "step": 557
    },
    {
      "epoch": 0.7699206622973439,
      "grad_norm": 1.2839542627334595,
      "learning_rate": 1.646870878260721e-05,
      "loss": 9.7659,
      "step": 558
    },
    {
      "epoch": 0.7713004484304933,
      "grad_norm": 1.2510404586791992,
      "learning_rate": 1.6282402143495568e-05,
      "loss": 9.7529,
      "step": 559
    },
    {
      "epoch": 0.7726802345636427,
      "grad_norm": 1.3847991228103638,
      "learning_rate": 1.6096950150405454e-05,
      "loss": 9.7254,
      "step": 560
    },
    {
      "epoch": 0.774060020696792,
      "grad_norm": 1.3537993431091309,
      "learning_rate": 1.5912357504021937e-05,
      "loss": 9.7168,
      "step": 561
    },
    {
      "epoch": 0.7754398068299414,
      "grad_norm": 1.5678364038467407,
      "learning_rate": 1.5728628883248007e-05,
      "loss": 9.6773,
      "step": 562
    },
    {
      "epoch": 0.7768195929630907,
      "grad_norm": 1.173730731010437,
      "learning_rate": 1.554576894508613e-05,
      "loss": 9.7439,
      "step": 563
    },
    {
      "epoch": 0.77819937909624,
      "grad_norm": 1.0499186515808105,
      "learning_rate": 1.536378232452003e-05,
      "loss": 9.7698,
      "step": 564
    },
    {
      "epoch": 0.7795791652293894,
      "grad_norm": 1.048542857170105,
      "learning_rate": 1.5182673634397365e-05,
      "loss": 9.7965,
      "step": 565
    },
    {
      "epoch": 0.7809589513625388,
      "grad_norm": 1.149909496307373,
      "learning_rate": 1.5002447465312675e-05,
      "loss": 9.7774,
      "step": 566
    },
    {
      "epoch": 0.7823387374956882,
      "grad_norm": 1.1968557834625244,
      "learning_rate": 1.4823108385491124e-05,
      "loss": 9.7471,
      "step": 567
    },
    {
      "epoch": 0.7837185236288375,
      "grad_norm": 1.135711908340454,
      "learning_rate": 1.4644660940672627e-05,
      "loss": 9.7686,
      "step": 568
    },
    {
      "epoch": 0.7850983097619869,
      "grad_norm": 1.3147321939468384,
      "learning_rate": 1.4467109653996697e-05,
      "loss": 9.7424,
      "step": 569
    },
    {
      "epoch": 0.7864780958951363,
      "grad_norm": 1.1964170932769775,
      "learning_rate": 1.429045902588777e-05,
      "loss": 9.7611,
      "step": 570
    },
    {
      "epoch": 0.7878578820282857,
      "grad_norm": 1.376891016960144,
      "learning_rate": 1.4114713533941082e-05,
      "loss": 9.7149,
      "step": 571
    },
    {
      "epoch": 0.7892376681614349,
      "grad_norm": 1.3214612007141113,
      "learning_rate": 1.3939877632809278e-05,
      "loss": 9.7467,
      "step": 572
    },
    {
      "epoch": 0.7906174542945843,
      "grad_norm": 1.346970796585083,
      "learning_rate": 1.3765955754089383e-05,
      "loss": 9.7122,
      "step": 573
    },
    {
      "epoch": 0.7919972404277337,
      "grad_norm": 1.4399006366729736,
      "learning_rate": 1.3592952306210588e-05,
      "loss": 9.7232,
      "step": 574
    },
    {
      "epoch": 0.793377026560883,
      "grad_norm": 1.645202875137329,
      "learning_rate": 1.3420871674322394e-05,
      "loss": 9.6723,
      "step": 575
    },
    {
      "epoch": 0.7947568126940324,
      "grad_norm": 0.9395657181739807,
      "learning_rate": 1.3249718220183583e-05,
      "loss": 9.8069,
      "step": 576
    },
    {
      "epoch": 0.7961365988271818,
      "grad_norm": 1.003074049949646,
      "learning_rate": 1.3079496282051529e-05,
      "loss": 9.7872,
      "step": 577
    },
    {
      "epoch": 0.7975163849603312,
      "grad_norm": 1.1187260150909424,
      "learning_rate": 1.2910210174572346e-05,
      "loss": 9.7854,
      "step": 578
    },
    {
      "epoch": 0.7988961710934805,
      "grad_norm": 1.1820377111434937,
      "learning_rate": 1.2741864188671492e-05,
      "loss": 9.7445,
      "step": 579
    },
    {
      "epoch": 0.8002759572266299,
      "grad_norm": 1.135115146636963,
      "learning_rate": 1.257446259144494e-05,
      "loss": 9.743,
      "step": 580
    },
    {
      "epoch": 0.8016557433597792,
      "grad_norm": 1.2748380899429321,
      "learning_rate": 1.2408009626051137e-05,
      "loss": 9.7491,
      "step": 581
    },
    {
      "epoch": 0.8030355294929286,
      "grad_norm": 1.2434227466583252,
      "learning_rate": 1.2242509511603317e-05,
      "loss": 9.7696,
      "step": 582
    },
    {
      "epoch": 0.8044153156260779,
      "grad_norm": 1.251225233078003,
      "learning_rate": 1.2077966443062705e-05,
      "loss": 9.7418,
      "step": 583
    },
    {
      "epoch": 0.8057951017592273,
      "grad_norm": 1.2516130208969116,
      "learning_rate": 1.1914384591132044e-05,
      "loss": 9.7266,
      "step": 584
    },
    {
      "epoch": 0.8071748878923767,
      "grad_norm": 1.392945647239685,
      "learning_rate": 1.1751768102150007e-05,
      "loss": 9.7269,
      "step": 585
    },
    {
      "epoch": 0.808554674025526,
      "grad_norm": 1.510158896446228,
      "learning_rate": 1.159012109798598e-05,
      "loss": 9.7113,
      "step": 586
    },
    {
      "epoch": 0.8099344601586754,
      "grad_norm": 1.5267751216888428,
      "learning_rate": 1.1429447675935695e-05,
      "loss": 9.704,
      "step": 587
    },
    {
      "epoch": 0.8113142462918248,
      "grad_norm": 1.1726975440979004,
      "learning_rate": 1.1269751908617277e-05,
      "loss": 9.7488,
      "step": 588
    },
    {
      "epoch": 0.8126940324249742,
      "grad_norm": 0.9718952775001526,
      "learning_rate": 1.1111037843868094e-05,
      "loss": 9.7921,
      "step": 589
    },
    {
      "epoch": 0.8140738185581234,
      "grad_norm": 1.0413875579833984,
      "learning_rate": 1.0953309504642128e-05,
      "loss": 9.7741,
      "step": 590
    },
    {
      "epoch": 0.8154536046912728,
      "grad_norm": 1.1728605031967163,
      "learning_rate": 1.0796570888907953e-05,
      "loss": 9.7641,
      "step": 591
    },
    {
      "epoch": 0.8168333908244222,
      "grad_norm": 1.157538652420044,
      "learning_rate": 1.0640825969547496e-05,
      "loss": 9.7696,
      "step": 592
    },
    {
      "epoch": 0.8182131769575716,
      "grad_norm": 1.2358477115631104,
      "learning_rate": 1.0486078694255242e-05,
      "loss": 9.7469,
      "step": 593
    },
    {
      "epoch": 0.8195929630907209,
      "grad_norm": 1.2600133419036865,
      "learning_rate": 1.0332332985438248e-05,
      "loss": 9.747,
      "step": 594
    },
    {
      "epoch": 0.8209727492238703,
      "grad_norm": 1.2197083234786987,
      "learning_rate": 1.017959274011665e-05,
      "loss": 9.7532,
      "step": 595
    },
    {
      "epoch": 0.8223525353570197,
      "grad_norm": 1.3774051666259766,
      "learning_rate": 1.0027861829824952e-05,
      "loss": 9.7325,
      "step": 596
    },
    {
      "epoch": 0.8237323214901691,
      "grad_norm": 1.38478422164917,
      "learning_rate": 9.87714410051383e-06,
      "loss": 9.7306,
      "step": 597
    },
    {
      "epoch": 0.8251121076233184,
      "grad_norm": 1.432771921157837,
      "learning_rate": 9.7274433724527e-06,
      "loss": 9.7276,
      "step": 598
    },
    {
      "epoch": 0.8264918937564677,
      "grad_norm": 1.5427794456481934,
      "learning_rate": 9.578763440132883e-06,
      "loss": 9.7113,
      "step": 599
    },
    {
      "epoch": 0.8278716798896171,
      "grad_norm": 1.5743299722671509,
      "learning_rate": 9.431108072171346e-06,
      "loss": 9.6803,
      "step": 600
    },
    {
      "epoch": 0.8292514660227664,
      "grad_norm": 0.921704113483429,
      "learning_rate": 9.284481011215318e-06,
      "loss": 9.7834,
      "step": 601
    },
    {
      "epoch": 0.8306312521559158,
      "grad_norm": 1.0345510244369507,
      "learning_rate": 9.138885973847261e-06,
      "loss": 9.79,
      "step": 602
    },
    {
      "epoch": 0.8320110382890652,
      "grad_norm": 1.182210087776184,
      "learning_rate": 8.994326650490831e-06,
      "loss": 9.7632,
      "step": 603
    },
    {
      "epoch": 0.8333908244222146,
      "grad_norm": 1.1576429605484009,
      "learning_rate": 8.850806705317183e-06,
      "loss": 9.7689,
      "step": 604
    },
    {
      "epoch": 0.834770610555364,
      "grad_norm": 1.2053581476211548,
      "learning_rate": 8.708329776152224e-06,
      "loss": 9.7558,
      "step": 605
    },
    {
      "epoch": 0.8361503966885133,
      "grad_norm": 1.2201980352401733,
      "learning_rate": 8.566899474384299e-06,
      "loss": 9.7571,
      "step": 606
    },
    {
      "epoch": 0.8375301828216627,
      "grad_norm": 1.1964430809020996,
      "learning_rate": 8.426519384872733e-06,
      "loss": 9.7458,
      "step": 607
    },
    {
      "epoch": 0.838909968954812,
      "grad_norm": 1.321816086769104,
      "learning_rate": 8.287193065856935e-06,
      "loss": 9.7307,
      "step": 608
    },
    {
      "epoch": 0.8402897550879613,
      "grad_norm": 1.290749430656433,
      "learning_rate": 8.148924048866169e-06,
      "loss": 9.7321,
      "step": 609
    },
    {
      "epoch": 0.8416695412211107,
      "grad_norm": 1.353271722793579,
      "learning_rate": 8.011715838630107e-06,
      "loss": 9.7231,
      "step": 610
    },
    {
      "epoch": 0.8430493273542601,
      "grad_norm": 1.4000368118286133,
      "learning_rate": 7.875571912989938e-06,
      "loss": 9.7083,
      "step": 611
    },
    {
      "epoch": 0.8444291134874095,
      "grad_norm": 1.4400883913040161,
      "learning_rate": 7.740495722810271e-06,
      "loss": 9.7085,
      "step": 612
    },
    {
      "epoch": 0.8458088996205588,
      "grad_norm": 1.134522795677185,
      "learning_rate": 7.606490691891577e-06,
      "loss": 9.7551,
      "step": 613
    },
    {
      "epoch": 0.8471886857537082,
      "grad_norm": 1.0418236255645752,
      "learning_rate": 7.4735602168835236e-06,
      "loss": 9.7693,
      "step": 614
    },
    {
      "epoch": 0.8485684718868576,
      "grad_norm": 1.080651879310608,
      "learning_rate": 7.341707667198766e-06,
      "loss": 9.7787,
      "step": 615
    },
    {
      "epoch": 0.849948258020007,
      "grad_norm": 1.1420639753341675,
      "learning_rate": 7.21093638492763e-06,
      "loss": 9.7674,
      "step": 616
    },
    {
      "epoch": 0.8513280441531562,
      "grad_norm": 1.221085786819458,
      "learning_rate": 7.081249684753327e-06,
      "loss": 9.7556,
      "step": 617
    },
    {
      "epoch": 0.8527078302863056,
      "grad_norm": 1.2512654066085815,
      "learning_rate": 6.952650853867993e-06,
      "loss": 9.7305,
      "step": 618
    },
    {
      "epoch": 0.854087616419455,
      "grad_norm": 1.2443575859069824,
      "learning_rate": 6.825143151889357e-06,
      "loss": 9.7577,
      "step": 619
    },
    {
      "epoch": 0.8554674025526043,
      "grad_norm": 1.2522335052490234,
      "learning_rate": 6.698729810778065e-06,
      "loss": 9.7628,
      "step": 620
    },
    {
      "epoch": 0.8568471886857537,
      "grad_norm": 1.360615611076355,
      "learning_rate": 6.5734140347558536e-06,
      "loss": 9.7367,
      "step": 621
    },
    {
      "epoch": 0.8582269748189031,
      "grad_norm": 1.3139582872390747,
      "learning_rate": 6.449199000224221e-06,
      "loss": 9.732,
      "step": 622
    },
    {
      "epoch": 0.8596067609520525,
      "grad_norm": 1.3527382612228394,
      "learning_rate": 6.3260878556840375e-06,
      "loss": 9.7361,
      "step": 623
    },
    {
      "epoch": 0.8609865470852018,
      "grad_norm": 1.4311867952346802,
      "learning_rate": 6.204083721655607e-06,
      "loss": 9.706,
      "step": 624
    },
    {
      "epoch": 0.8623663332183512,
      "grad_norm": 1.6295082569122314,
      "learning_rate": 6.083189690599711e-06,
      "loss": 9.6803,
      "step": 625
    },
    {
      "epoch": 0.8637461193515005,
      "grad_norm": 0.9568197131156921,
      "learning_rate": 5.9634088268390784e-06,
      "loss": 9.7839,
      "step": 626
    },
    {
      "epoch": 0.8651259054846498,
      "grad_norm": 1.080190896987915,
      "learning_rate": 5.8447441664808424e-06,
      "loss": 9.7847,
      "step": 627
    },
    {
      "epoch": 0.8665056916177992,
      "grad_norm": 1.0810461044311523,
      "learning_rate": 5.727198717339511e-06,
      "loss": 9.7586,
      "step": 628
    },
    {
      "epoch": 0.8678854777509486,
      "grad_norm": 1.1114495992660522,
      "learning_rate": 5.610775458860718e-06,
      "loss": 9.7832,
      "step": 629
    },
    {
      "epoch": 0.869265263884098,
      "grad_norm": 1.2362908124923706,
      "learning_rate": 5.495477342045779e-06,
      "loss": 9.7565,
      "step": 630
    },
    {
      "epoch": 0.8706450500172473,
      "grad_norm": 1.2435539960861206,
      "learning_rate": 5.381307289376786e-06,
      "loss": 9.7561,
      "step": 631
    },
    {
      "epoch": 0.8720248361503967,
      "grad_norm": 1.3071742057800293,
      "learning_rate": 5.2682681947426375e-06,
      "loss": 9.7482,
      "step": 632
    },
    {
      "epoch": 0.8734046222835461,
      "grad_norm": 1.2518724203109741,
      "learning_rate": 5.156362923365588e-06,
      "loss": 9.7569,
      "step": 633
    },
    {
      "epoch": 0.8747844084166954,
      "grad_norm": 1.3225029706954956,
      "learning_rate": 5.045594311728707e-06,
      "loss": 9.7413,
      "step": 634
    },
    {
      "epoch": 0.8761641945498447,
      "grad_norm": 1.3300327062606812,
      "learning_rate": 4.93596516750392e-06,
      "loss": 9.7283,
      "step": 635
    },
    {
      "epoch": 0.8775439806829941,
      "grad_norm": 1.3841952085494995,
      "learning_rate": 4.827478269480895e-06,
      "loss": 9.7273,
      "step": 636
    },
    {
      "epoch": 0.8789237668161435,
      "grad_norm": 1.5817680358886719,
      "learning_rate": 4.720136367496536e-06,
      "loss": 9.671,
      "step": 637
    },
    {
      "epoch": 0.8803035529492929,
      "grad_norm": 1.1424212455749512,
      "learning_rate": 4.613942182365372e-06,
      "loss": 9.7741,
      "step": 638
    },
    {
      "epoch": 0.8816833390824422,
      "grad_norm": 1.0098145008087158,
      "learning_rate": 4.508898405810524e-06,
      "loss": 9.7891,
      "step": 639
    },
    {
      "epoch": 0.8830631252155916,
      "grad_norm": 1.0495284795761108,
      "learning_rate": 4.405007700395497e-06,
      "loss": 9.7944,
      "step": 640
    },
    {
      "epoch": 0.884442911348741,
      "grad_norm": 1.165906310081482,
      "learning_rate": 4.3022726994567105e-06,
      "loss": 9.773,
      "step": 641
    },
    {
      "epoch": 0.8858226974818904,
      "grad_norm": 1.1741012334823608,
      "learning_rate": 4.200696007036703e-06,
      "loss": 9.767,
      "step": 642
    },
    {
      "epoch": 0.8872024836150396,
      "grad_norm": 1.2205920219421387,
      "learning_rate": 4.100280197818207e-06,
      "loss": 9.7574,
      "step": 643
    },
    {
      "epoch": 0.888582269748189,
      "grad_norm": 1.2672699689865112,
      "learning_rate": 4.001027817058789e-06,
      "loss": 9.7688,
      "step": 644
    },
    {
      "epoch": 0.8899620558813384,
      "grad_norm": 1.2361886501312256,
      "learning_rate": 3.902941380526426e-06,
      "loss": 9.7636,
      "step": 645
    },
    {
      "epoch": 0.8913418420144877,
      "grad_norm": 1.322023868560791,
      "learning_rate": 3.8060233744356633e-06,
      "loss": 9.7413,
      "step": 646
    },
    {
      "epoch": 0.8927216281476371,
      "grad_norm": 1.3382331132888794,
      "learning_rate": 3.7102762553846536e-06,
      "loss": 9.7279,
      "step": 647
    },
    {
      "epoch": 0.8941014142807865,
      "grad_norm": 1.345215082168579,
      "learning_rate": 3.615702450292857e-06,
      "loss": 9.745,
      "step": 648
    },
    {
      "epoch": 0.8954812004139359,
      "grad_norm": 1.3454034328460693,
      "learning_rate": 3.5223043563395288e-06,
      "loss": 9.721,
      "step": 649
    },
    {
      "epoch": 0.8968609865470852,
      "grad_norm": 1.4947688579559326,
      "learning_rate": 3.4300843409029726e-06,
      "loss": 9.7119,
      "step": 650
    },
    {
      "epoch": 0.8982407726802346,
      "grad_norm": 0.9093929529190063,
      "learning_rate": 3.3390447415005077e-06,
      "loss": 9.8062,
      "step": 651
    },
    {
      "epoch": 0.8996205588133839,
      "grad_norm": 1.040328860282898,
      "learning_rate": 3.249187865729264e-06,
      "loss": 9.7664,
      "step": 652
    },
    {
      "epoch": 0.9010003449465332,
      "grad_norm": 1.1193190813064575,
      "learning_rate": 3.160515991207619e-06,
      "loss": 9.7838,
      "step": 653
    },
    {
      "epoch": 0.9023801310796826,
      "grad_norm": 1.14997136592865,
      "learning_rate": 3.0730313655175645e-06,
      "loss": 9.768,
      "step": 654
    },
    {
      "epoch": 0.903759917212832,
      "grad_norm": 1.1968692541122437,
      "learning_rate": 2.9867362061476276e-06,
      "loss": 9.7622,
      "step": 655
    },
    {
      "epoch": 0.9051397033459814,
      "grad_norm": 1.1504422426223755,
      "learning_rate": 2.901632700436757e-06,
      "loss": 9.7521,
      "step": 656
    },
    {
      "epoch": 0.9065194894791307,
      "grad_norm": 1.2824842929840088,
      "learning_rate": 2.8177230055188353e-06,
      "loss": 9.7633,
      "step": 657
    },
    {
      "epoch": 0.9078992756122801,
      "grad_norm": 1.2600338459014893,
      "learning_rate": 2.7350092482679836e-06,
      "loss": 9.7494,
      "step": 658
    },
    {
      "epoch": 0.9092790617454295,
      "grad_norm": 1.2912122011184692,
      "learning_rate": 2.653493525244721e-06,
      "loss": 9.7263,
      "step": 659
    },
    {
      "epoch": 0.9106588478785789,
      "grad_norm": 1.415069580078125,
      "learning_rate": 2.573177902642726e-06,
      "loss": 9.7187,
      "step": 660
    },
    {
      "epoch": 0.9120386340117281,
      "grad_norm": 1.3458380699157715,
      "learning_rate": 2.4940644162365522e-06,
      "loss": 9.7266,
      "step": 661
    },
    {
      "epoch": 0.9134184201448775,
      "grad_norm": 1.5353480577468872,
      "learning_rate": 2.416155071329973e-06,
      "loss": 9.6964,
      "step": 662
    },
    {
      "epoch": 0.9147982062780269,
      "grad_norm": 1.0634307861328125,
      "learning_rate": 2.339451842705187e-06,
      "loss": 9.7851,
      "step": 663
    },
    {
      "epoch": 0.9161779924111763,
      "grad_norm": 1.0400745868682861,
      "learning_rate": 2.2639566745727205e-06,
      "loss": 9.7713,
      "step": 664
    },
    {
      "epoch": 0.9175577785443256,
      "grad_norm": 1.134487509727478,
      "learning_rate": 2.189671480522204e-06,
      "loss": 9.7789,
      "step": 665
    },
    {
      "epoch": 0.918937564677475,
      "grad_norm": 1.1036391258239746,
      "learning_rate": 2.1165981434738026e-06,
      "loss": 9.7722,
      "step": 666
    },
    {
      "epoch": 0.9203173508106244,
      "grad_norm": 1.1182793378829956,
      "learning_rate": 2.0447385156305565e-06,
      "loss": 9.7933,
      "step": 667
    },
    {
      "epoch": 0.9216971369437738,
      "grad_norm": 1.2758280038833618,
      "learning_rate": 1.974094418431388e-06,
      "loss": 9.7561,
      "step": 668
    },
    {
      "epoch": 0.9230769230769231,
      "grad_norm": 1.2668101787567139,
      "learning_rate": 1.9046676425049315e-06,
      "loss": 9.7434,
      "step": 669
    },
    {
      "epoch": 0.9244567092100724,
      "grad_norm": 1.2824689149856567,
      "learning_rate": 1.8364599476241862e-06,
      "loss": 9.7397,
      "step": 670
    },
    {
      "epoch": 0.9258364953432218,
      "grad_norm": 1.2909704446792603,
      "learning_rate": 1.7694730626618583e-06,
      "loss": 9.7306,
      "step": 671
    },
    {
      "epoch": 0.9272162814763711,
      "grad_norm": 1.3211469650268555,
      "learning_rate": 1.70370868554659e-06,
      "loss": 9.7482,
      "step": 672
    },
    {
      "epoch": 0.9285960676095205,
      "grad_norm": 1.3531687259674072,
      "learning_rate": 1.6391684832198718e-06,
      "loss": 9.7382,
      "step": 673
    },
    {
      "epoch": 0.9299758537426699,
      "grad_norm": 1.4724624156951904,
      "learning_rate": 1.5758540915938368e-06,
      "loss": 9.7051,
      "step": 674
    },
    {
      "epoch": 0.9313556398758193,
      "grad_norm": 1.6450470685958862,
      "learning_rate": 1.5137671155097565e-06,
      "loss": 9.6571,
      "step": 675
    },
    {
      "epoch": 0.9327354260089686,
      "grad_norm": 0.8748622536659241,
      "learning_rate": 1.4529091286973995e-06,
      "loss": 9.8211,
      "step": 676
    },
    {
      "epoch": 0.934115212142118,
      "grad_norm": 1.065389633178711,
      "learning_rate": 1.3932816737351085e-06,
      "loss": 9.7792,
      "step": 677
    },
    {
      "epoch": 0.9354949982752674,
      "grad_norm": 1.1271086931228638,
      "learning_rate": 1.3348862620107038e-06,
      "loss": 9.7635,
      "step": 678
    },
    {
      "epoch": 0.9368747844084166,
      "grad_norm": 1.126579999923706,
      "learning_rate": 1.27772437368322e-06,
      "loss": 9.7736,
      "step": 679
    },
    {
      "epoch": 0.938254570541566,
      "grad_norm": 1.1667218208312988,
      "learning_rate": 1.2217974576453073e-06,
      "loss": 9.7564,
      "step": 680
    },
    {
      "epoch": 0.9396343566747154,
      "grad_norm": 1.243585228919983,
      "learning_rate": 1.1671069314865834e-06,
      "loss": 9.7614,
      "step": 681
    },
    {
      "epoch": 0.9410141428078648,
      "grad_norm": 1.1980286836624146,
      "learning_rate": 1.1136541814576573e-06,
      "loss": 9.7782,
      "step": 682
    },
    {
      "epoch": 0.9423939289410141,
      "grad_norm": 1.2429455518722534,
      "learning_rate": 1.061440562435001e-06,
      "loss": 9.7618,
      "step": 683
    },
    {
      "epoch": 0.9437737150741635,
      "grad_norm": 1.306204080581665,
      "learning_rate": 1.0104673978866164e-06,
      "loss": 9.7464,
      "step": 684
    },
    {
      "epoch": 0.9451535012073129,
      "grad_norm": 1.3777480125427246,
      "learning_rate": 9.607359798384785e-07,
      "loss": 9.7229,
      "step": 685
    },
    {
      "epoch": 0.9465332873404623,
      "grad_norm": 1.3626405000686646,
      "learning_rate": 9.122475688417953e-07,
      "loss": 9.712,
      "step": 686
    },
    {
      "epoch": 0.9479130734736116,
      "grad_norm": 1.4798493385314941,
      "learning_rate": 8.650033939410385e-07,
      "loss": 9.7036,
      "step": 687
    },
    {
      "epoch": 0.9492928596067609,
      "grad_norm": 1.1905264854431152,
      "learning_rate": 8.190046526428242e-07,
      "loss": 9.7456,
      "step": 688
    },
    {
      "epoch": 0.9506726457399103,
      "grad_norm": 0.9911084175109863,
      "learning_rate": 7.742525108855148e-07,
      "loss": 9.7855,
      "step": 689
    },
    {
      "epoch": 0.9520524318730597,
      "grad_norm": 1.0882097482681274,
      "learning_rate": 7.307481030097152e-07,
      "loss": 9.7704,
      "step": 690
    },
    {
      "epoch": 0.953432218006209,
      "grad_norm": 1.1580933332443237,
      "learning_rate": 6.884925317294677e-07,
      "loss": 9.7733,
      "step": 691
    },
    {
      "epoch": 0.9548120041393584,
      "grad_norm": 1.1494512557983398,
      "learning_rate": 6.474868681043578e-07,
      "loss": 9.7834,
      "step": 692
    },
    {
      "epoch": 0.9561917902725078,
      "grad_norm": 1.1650831699371338,
      "learning_rate": 6.077321515123136e-07,
      "loss": 9.7708,
      "step": 693
    },
    {
      "epoch": 0.9575715764056572,
      "grad_norm": 1.2284369468688965,
      "learning_rate": 5.692293896232936e-07,
      "loss": 9.7564,
      "step": 694
    },
    {
      "epoch": 0.9589513625388065,
      "grad_norm": 1.2123266458511353,
      "learning_rate": 5.319795583737242e-07,
      "loss": 9.7493,
      "step": 695
    },
    {
      "epoch": 0.9603311486719558,
      "grad_norm": 1.2671525478363037,
      "learning_rate": 4.959836019417963e-07,
      "loss": 9.7491,
      "step": 696
    },
    {
      "epoch": 0.9617109348051052,
      "grad_norm": 1.2901546955108643,
      "learning_rate": 4.6124243272349653e-07,
      "loss": 9.745,
      "step": 697
    },
    {
      "epoch": 0.9630907209382545,
      "grad_norm": 1.3228528499603271,
      "learning_rate": 4.277569313094809e-07,
      "loss": 9.7245,
      "step": 698
    },
    {
      "epoch": 0.9644705070714039,
      "grad_norm": 1.4477964639663696,
      "learning_rate": 3.9552794646279255e-07,
      "loss": 9.7099,
      "step": 699
    },
    {
      "epoch": 0.9658502932045533,
      "grad_norm": 1.4955841302871704,
      "learning_rate": 3.6455629509730136e-07,
      "loss": 9.6912,
      "step": 700
    },
    {
      "epoch": 0.9672300793377027,
      "grad_norm": 0.8777285218238831,
      "learning_rate": 3.3484276225703714e-07,
      "loss": 9.8028,
      "step": 701
    },
    {
      "epoch": 0.968609865470852,
      "grad_norm": 1.04829740524292,
      "learning_rate": 3.0638810109626103e-07,
      "loss": 9.7746,
      "step": 702
    },
    {
      "epoch": 0.9699896516040014,
      "grad_norm": 1.18967604637146,
      "learning_rate": 2.79193032860392e-07,
      "loss": 9.7497,
      "step": 703
    },
    {
      "epoch": 0.9713694377371508,
      "grad_norm": 1.118967890739441,
      "learning_rate": 2.532582468677214e-07,
      "loss": 9.7728,
      "step": 704
    },
    {
      "epoch": 0.9727492238703,
      "grad_norm": 1.229331612586975,
      "learning_rate": 2.285844004919324e-07,
      "loss": 9.7464,
      "step": 705
    },
    {
      "epoch": 0.9741290100034494,
      "grad_norm": 1.1736466884613037,
      "learning_rate": 2.0517211914545254e-07,
      "loss": 9.7577,
      "step": 706
    },
    {
      "epoch": 0.9755087961365988,
      "grad_norm": 1.2047674655914307,
      "learning_rate": 1.8302199626358818e-07,
      "loss": 9.752,
      "step": 707
    },
    {
      "epoch": 0.9768885822697482,
      "grad_norm": 1.251903772354126,
      "learning_rate": 1.6213459328950352e-07,
      "loss": 9.7461,
      "step": 708
    },
    {
      "epoch": 0.9782683684028975,
      "grad_norm": 1.369354486465454,
      "learning_rate": 1.4251043965994304e-07,
      "loss": 9.7395,
      "step": 709
    },
    {
      "epoch": 0.9796481545360469,
      "grad_norm": 1.3757730722427368,
      "learning_rate": 1.2415003279186987e-07,
      "loss": 9.7186,
      "step": 710
    },
    {
      "epoch": 0.9810279406691963,
      "grad_norm": 1.3921722173690796,
      "learning_rate": 1.0705383806982606e-07,
      "loss": 9.7241,
      "step": 711
    },
    {
      "epoch": 0.9824077268023457,
      "grad_norm": 1.4798989295959473,
      "learning_rate": 9.12222888341252e-08,
      "loss": 9.7088,
      "step": 712
    },
    {
      "epoch": 0.983787512935495,
      "grad_norm": 1.149962067604065,
      "learning_rate": 7.665578636990579e-08,
      "loss": 9.7613,
      "step": 713
    },
    {
      "epoch": 0.9851672990686443,
      "grad_norm": 0.952008843421936,
      "learning_rate": 6.335469989692256e-08,
      "loss": 9.7976,
      "step": 714
    },
    {
      "epoch": 0.9865470852017937,
      "grad_norm": 1.064375400543213,
      "learning_rate": 5.131936656020409e-08,
      "loss": 9.7691,
      "step": 715
    },
    {
      "epoch": 0.9879268713349431,
      "grad_norm": 1.1418269872665405,
      "learning_rate": 4.055009142152067e-08,
      "loss": 9.7665,
      "step": 716
    },
    {
      "epoch": 0.9893066574680924,
      "grad_norm": 1.1035135984420776,
      "learning_rate": 3.10471474516183e-08,
      "loss": 9.782,
      "step": 717
    },
    {
      "epoch": 0.9906864436012418,
      "grad_norm": 1.1444123983383179,
      "learning_rate": 2.2810775523329773e-08,
      "loss": 9.7736,
      "step": 718
    },
    {
      "epoch": 0.9920662297343912,
      "grad_norm": 1.2123403549194336,
      "learning_rate": 1.5841184405462895e-08,
      "loss": 9.7574,
      "step": 719
    },
    {
      "epoch": 0.9934460158675406,
      "grad_norm": 1.313451886177063,
      "learning_rate": 1.0138550757493592e-08,
      "loss": 9.757,
      "step": 720
    },
    {
      "epoch": 0.9948258020006899,
      "grad_norm": 1.2594588994979858,
      "learning_rate": 5.703019125102849e-09,
      "loss": 9.744,
      "step": 721
    },
    {
      "epoch": 0.9962055881338393,
      "grad_norm": 1.3528109788894653,
      "learning_rate": 2.534701936512951e-09,
      "loss": 9.7332,
      "step": 722
    },
    {
      "epoch": 0.9975853742669886,
      "grad_norm": 1.4873722791671753,
      "learning_rate": 6.336794996231188e-10,
      "loss": 9.7142,
      "step": 723
    },
    {
      "epoch": 0.9989651604001379,
      "grad_norm": 1.4233864545822144,
      "learning_rate": 0.0,
      "loss": 9.7126,
      "step": 724
    }
  ],
  "logging_steps": 1,
  "max_steps": 724,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 239,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 601161656696832.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}