ds-filtered-3 / trainer_state.json
vitaleantonio's picture
Upload ds-filtered-3
9921579 verified
Raw
History Blame Contribute Delete
411 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"eval_steps": 500,
"global_step": 2535,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0011837821840781297,
"grad_norm": 100.0,
"learning_rate": 4.739336492890996e-08,
"loss": 6.1011,
"step": 1
},
{
"epoch": 0.0023675643681562593,
"grad_norm": 87.5,
"learning_rate": 9.478672985781992e-08,
"loss": 5.2723,
"step": 2
},
{
"epoch": 0.003551346552234389,
"grad_norm": 91.0,
"learning_rate": 1.4218009478672986e-07,
"loss": 6.0872,
"step": 3
},
{
"epoch": 0.004735128736312519,
"grad_norm": 98.5,
"learning_rate": 1.8957345971563984e-07,
"loss": 6.3863,
"step": 4
},
{
"epoch": 0.005918910920390648,
"grad_norm": 83.5,
"learning_rate": 2.3696682464454978e-07,
"loss": 5.4804,
"step": 5
},
{
"epoch": 0.007102693104468778,
"grad_norm": 89.0,
"learning_rate": 2.843601895734597e-07,
"loss": 5.7869,
"step": 6
},
{
"epoch": 0.008286475288546907,
"grad_norm": 73.0,
"learning_rate": 3.317535545023697e-07,
"loss": 5.6744,
"step": 7
},
{
"epoch": 0.009470257472625037,
"grad_norm": 71.5,
"learning_rate": 3.791469194312797e-07,
"loss": 4.7048,
"step": 8
},
{
"epoch": 0.010654039656703166,
"grad_norm": 71.5,
"learning_rate": 4.265402843601896e-07,
"loss": 5.3761,
"step": 9
},
{
"epoch": 0.011837821840781295,
"grad_norm": 86.0,
"learning_rate": 4.7393364928909956e-07,
"loss": 5.4346,
"step": 10
},
{
"epoch": 0.013021604024859426,
"grad_norm": 93.5,
"learning_rate": 5.213270142180095e-07,
"loss": 5.9582,
"step": 11
},
{
"epoch": 0.014205386208937555,
"grad_norm": 74.0,
"learning_rate": 5.687203791469194e-07,
"loss": 5.335,
"step": 12
},
{
"epoch": 0.015389168393015686,
"grad_norm": 71.5,
"learning_rate": 6.161137440758294e-07,
"loss": 4.9707,
"step": 13
},
{
"epoch": 0.016572950577093813,
"grad_norm": 72.0,
"learning_rate": 6.635071090047394e-07,
"loss": 5.0601,
"step": 14
},
{
"epoch": 0.017756732761171946,
"grad_norm": 89.0,
"learning_rate": 7.109004739336493e-07,
"loss": 6.2194,
"step": 15
},
{
"epoch": 0.018940514945250075,
"grad_norm": 79.0,
"learning_rate": 7.582938388625594e-07,
"loss": 5.2448,
"step": 16
},
{
"epoch": 0.020124297129328204,
"grad_norm": 84.0,
"learning_rate": 8.056872037914692e-07,
"loss": 5.6702,
"step": 17
},
{
"epoch": 0.021308079313406333,
"grad_norm": 92.0,
"learning_rate": 8.530805687203792e-07,
"loss": 5.7547,
"step": 18
},
{
"epoch": 0.022491861497484462,
"grad_norm": 87.5,
"learning_rate": 9.004739336492892e-07,
"loss": 5.3102,
"step": 19
},
{
"epoch": 0.02367564368156259,
"grad_norm": 61.75,
"learning_rate": 9.478672985781991e-07,
"loss": 4.7304,
"step": 20
},
{
"epoch": 0.024859425865640723,
"grad_norm": 87.5,
"learning_rate": 9.95260663507109e-07,
"loss": 5.9793,
"step": 21
},
{
"epoch": 0.026043208049718852,
"grad_norm": 78.5,
"learning_rate": 1.042654028436019e-06,
"loss": 5.4827,
"step": 22
},
{
"epoch": 0.02722699023379698,
"grad_norm": 75.0,
"learning_rate": 1.090047393364929e-06,
"loss": 4.9969,
"step": 23
},
{
"epoch": 0.02841077241787511,
"grad_norm": 74.5,
"learning_rate": 1.1374407582938388e-06,
"loss": 5.3642,
"step": 24
},
{
"epoch": 0.02959455460195324,
"grad_norm": 101.5,
"learning_rate": 1.184834123222749e-06,
"loss": 6.2325,
"step": 25
},
{
"epoch": 0.030778336786031372,
"grad_norm": 74.0,
"learning_rate": 1.2322274881516587e-06,
"loss": 5.6023,
"step": 26
},
{
"epoch": 0.0319621189701095,
"grad_norm": 76.0,
"learning_rate": 1.2796208530805687e-06,
"loss": 5.0527,
"step": 27
},
{
"epoch": 0.033145901154187626,
"grad_norm": 101.0,
"learning_rate": 1.3270142180094788e-06,
"loss": 5.3848,
"step": 28
},
{
"epoch": 0.03432968333826576,
"grad_norm": 83.0,
"learning_rate": 1.3744075829383887e-06,
"loss": 5.2706,
"step": 29
},
{
"epoch": 0.03551346552234389,
"grad_norm": 91.5,
"learning_rate": 1.4218009478672987e-06,
"loss": 5.5121,
"step": 30
},
{
"epoch": 0.03669724770642202,
"grad_norm": 76.0,
"learning_rate": 1.4691943127962086e-06,
"loss": 5.3106,
"step": 31
},
{
"epoch": 0.03788102989050015,
"grad_norm": 94.5,
"learning_rate": 1.5165876777251187e-06,
"loss": 5.6794,
"step": 32
},
{
"epoch": 0.03906481207457828,
"grad_norm": 84.5,
"learning_rate": 1.5639810426540287e-06,
"loss": 5.2462,
"step": 33
},
{
"epoch": 0.04024859425865641,
"grad_norm": 86.0,
"learning_rate": 1.6113744075829384e-06,
"loss": 5.404,
"step": 34
},
{
"epoch": 0.041432376442734536,
"grad_norm": 71.0,
"learning_rate": 1.6587677725118483e-06,
"loss": 5.0895,
"step": 35
},
{
"epoch": 0.042616158626812665,
"grad_norm": 77.5,
"learning_rate": 1.7061611374407585e-06,
"loss": 5.2412,
"step": 36
},
{
"epoch": 0.043799940810890794,
"grad_norm": 66.5,
"learning_rate": 1.7535545023696684e-06,
"loss": 3.6924,
"step": 37
},
{
"epoch": 0.044983722994968924,
"grad_norm": 72.5,
"learning_rate": 1.8009478672985784e-06,
"loss": 4.7869,
"step": 38
},
{
"epoch": 0.04616750517904705,
"grad_norm": 82.5,
"learning_rate": 1.8483412322274883e-06,
"loss": 4.9911,
"step": 39
},
{
"epoch": 0.04735128736312518,
"grad_norm": 95.0,
"learning_rate": 1.8957345971563982e-06,
"loss": 5.2916,
"step": 40
},
{
"epoch": 0.04853506954720332,
"grad_norm": 75.5,
"learning_rate": 1.943127962085308e-06,
"loss": 4.324,
"step": 41
},
{
"epoch": 0.049718851731281447,
"grad_norm": 81.5,
"learning_rate": 1.990521327014218e-06,
"loss": 4.7464,
"step": 42
},
{
"epoch": 0.050902633915359576,
"grad_norm": 92.0,
"learning_rate": 2.037914691943128e-06,
"loss": 5.2764,
"step": 43
},
{
"epoch": 0.052086416099437705,
"grad_norm": 92.5,
"learning_rate": 2.085308056872038e-06,
"loss": 4.978,
"step": 44
},
{
"epoch": 0.053270198283515834,
"grad_norm": 88.0,
"learning_rate": 2.1327014218009483e-06,
"loss": 4.5148,
"step": 45
},
{
"epoch": 0.05445398046759396,
"grad_norm": 84.0,
"learning_rate": 2.180094786729858e-06,
"loss": 4.2065,
"step": 46
},
{
"epoch": 0.05563776265167209,
"grad_norm": 89.0,
"learning_rate": 2.2274881516587678e-06,
"loss": 4.2973,
"step": 47
},
{
"epoch": 0.05682154483575022,
"grad_norm": 79.5,
"learning_rate": 2.2748815165876777e-06,
"loss": 4.5567,
"step": 48
},
{
"epoch": 0.05800532701982835,
"grad_norm": 92.0,
"learning_rate": 2.322274881516588e-06,
"loss": 4.3378,
"step": 49
},
{
"epoch": 0.05918910920390648,
"grad_norm": 91.5,
"learning_rate": 2.369668246445498e-06,
"loss": 3.7227,
"step": 50
},
{
"epoch": 0.06037289138798461,
"grad_norm": 87.5,
"learning_rate": 2.417061611374408e-06,
"loss": 4.0515,
"step": 51
},
{
"epoch": 0.061556673572062744,
"grad_norm": 96.0,
"learning_rate": 2.4644549763033174e-06,
"loss": 4.048,
"step": 52
},
{
"epoch": 0.06274045575614087,
"grad_norm": 78.0,
"learning_rate": 2.5118483412322274e-06,
"loss": 4.4733,
"step": 53
},
{
"epoch": 0.063924237940219,
"grad_norm": 69.5,
"learning_rate": 2.5592417061611373e-06,
"loss": 3.7112,
"step": 54
},
{
"epoch": 0.06510802012429713,
"grad_norm": 66.0,
"learning_rate": 2.606635071090048e-06,
"loss": 4.039,
"step": 55
},
{
"epoch": 0.06629180230837525,
"grad_norm": 75.0,
"learning_rate": 2.6540284360189576e-06,
"loss": 3.8792,
"step": 56
},
{
"epoch": 0.06747558449245339,
"grad_norm": 71.0,
"learning_rate": 2.7014218009478675e-06,
"loss": 3.0576,
"step": 57
},
{
"epoch": 0.06865936667653152,
"grad_norm": 66.0,
"learning_rate": 2.7488151658767775e-06,
"loss": 3.5826,
"step": 58
},
{
"epoch": 0.06984314886060965,
"grad_norm": 69.0,
"learning_rate": 2.7962085308056874e-06,
"loss": 3.4863,
"step": 59
},
{
"epoch": 0.07102693104468778,
"grad_norm": 57.5,
"learning_rate": 2.8436018957345973e-06,
"loss": 3.7599,
"step": 60
},
{
"epoch": 0.0722107132287659,
"grad_norm": 53.25,
"learning_rate": 2.8909952606635073e-06,
"loss": 3.7882,
"step": 61
},
{
"epoch": 0.07339449541284404,
"grad_norm": 47.25,
"learning_rate": 2.938388625592417e-06,
"loss": 3.4144,
"step": 62
},
{
"epoch": 0.07457827759692216,
"grad_norm": 49.25,
"learning_rate": 2.985781990521327e-06,
"loss": 3.0334,
"step": 63
},
{
"epoch": 0.0757620597810003,
"grad_norm": 42.25,
"learning_rate": 3.0331753554502375e-06,
"loss": 3.0094,
"step": 64
},
{
"epoch": 0.07694584196507842,
"grad_norm": 41.75,
"learning_rate": 3.0805687203791474e-06,
"loss": 3.0529,
"step": 65
},
{
"epoch": 0.07812962414915656,
"grad_norm": 34.5,
"learning_rate": 3.1279620853080574e-06,
"loss": 2.3809,
"step": 66
},
{
"epoch": 0.07931340633323468,
"grad_norm": 37.0,
"learning_rate": 3.1753554502369673e-06,
"loss": 2.5117,
"step": 67
},
{
"epoch": 0.08049718851731281,
"grad_norm": 31.25,
"learning_rate": 3.222748815165877e-06,
"loss": 2.8075,
"step": 68
},
{
"epoch": 0.08168097070139095,
"grad_norm": 29.75,
"learning_rate": 3.2701421800947867e-06,
"loss": 2.6948,
"step": 69
},
{
"epoch": 0.08286475288546907,
"grad_norm": 27.375,
"learning_rate": 3.3175355450236967e-06,
"loss": 1.9687,
"step": 70
},
{
"epoch": 0.08404853506954721,
"grad_norm": 27.625,
"learning_rate": 3.3649289099526066e-06,
"loss": 2.2163,
"step": 71
},
{
"epoch": 0.08523231725362533,
"grad_norm": 25.25,
"learning_rate": 3.412322274881517e-06,
"loss": 2.5418,
"step": 72
},
{
"epoch": 0.08641609943770347,
"grad_norm": 24.25,
"learning_rate": 3.459715639810427e-06,
"loss": 2.2988,
"step": 73
},
{
"epoch": 0.08759988162178159,
"grad_norm": 25.625,
"learning_rate": 3.507109004739337e-06,
"loss": 2.3544,
"step": 74
},
{
"epoch": 0.08878366380585972,
"grad_norm": 20.625,
"learning_rate": 3.5545023696682468e-06,
"loss": 2.76,
"step": 75
},
{
"epoch": 0.08996744598993785,
"grad_norm": 22.25,
"learning_rate": 3.6018957345971567e-06,
"loss": 2.4018,
"step": 76
},
{
"epoch": 0.09115122817401598,
"grad_norm": 20.5,
"learning_rate": 3.6492890995260666e-06,
"loss": 2.1251,
"step": 77
},
{
"epoch": 0.0923350103580941,
"grad_norm": 15.1875,
"learning_rate": 3.6966824644549766e-06,
"loss": 1.8724,
"step": 78
},
{
"epoch": 0.09351879254217224,
"grad_norm": 19.25,
"learning_rate": 3.7440758293838865e-06,
"loss": 2.7706,
"step": 79
},
{
"epoch": 0.09470257472625036,
"grad_norm": 16.375,
"learning_rate": 3.7914691943127964e-06,
"loss": 1.7835,
"step": 80
},
{
"epoch": 0.0958863569103285,
"grad_norm": 14.8125,
"learning_rate": 3.838862559241707e-06,
"loss": 1.8068,
"step": 81
},
{
"epoch": 0.09707013909440664,
"grad_norm": 16.25,
"learning_rate": 3.886255924170616e-06,
"loss": 2.3608,
"step": 82
},
{
"epoch": 0.09825392127848476,
"grad_norm": 14.1875,
"learning_rate": 3.933649289099527e-06,
"loss": 1.7893,
"step": 83
},
{
"epoch": 0.09943770346256289,
"grad_norm": 13.0625,
"learning_rate": 3.981042654028436e-06,
"loss": 2.1508,
"step": 84
},
{
"epoch": 0.10062148564664102,
"grad_norm": 13.6875,
"learning_rate": 4.0284360189573465e-06,
"loss": 1.9684,
"step": 85
},
{
"epoch": 0.10180526783071915,
"grad_norm": 15.125,
"learning_rate": 4.075829383886256e-06,
"loss": 2.6762,
"step": 86
},
{
"epoch": 0.10298905001479727,
"grad_norm": 14.1875,
"learning_rate": 4.123222748815166e-06,
"loss": 2.1105,
"step": 87
},
{
"epoch": 0.10417283219887541,
"grad_norm": 12.375,
"learning_rate": 4.170616113744076e-06,
"loss": 2.1101,
"step": 88
},
{
"epoch": 0.10535661438295353,
"grad_norm": 16.125,
"learning_rate": 4.218009478672986e-06,
"loss": 2.371,
"step": 89
},
{
"epoch": 0.10654039656703167,
"grad_norm": 11.875,
"learning_rate": 4.265402843601897e-06,
"loss": 1.5986,
"step": 90
},
{
"epoch": 0.10772417875110979,
"grad_norm": 11.5,
"learning_rate": 4.312796208530806e-06,
"loss": 1.645,
"step": 91
},
{
"epoch": 0.10890796093518793,
"grad_norm": 13.25,
"learning_rate": 4.360189573459716e-06,
"loss": 2.0038,
"step": 92
},
{
"epoch": 0.11009174311926606,
"grad_norm": 12.9375,
"learning_rate": 4.407582938388626e-06,
"loss": 1.7194,
"step": 93
},
{
"epoch": 0.11127552530334418,
"grad_norm": 15.3125,
"learning_rate": 4.4549763033175355e-06,
"loss": 1.7307,
"step": 94
},
{
"epoch": 0.11245930748742232,
"grad_norm": 13.8125,
"learning_rate": 4.502369668246446e-06,
"loss": 2.1517,
"step": 95
},
{
"epoch": 0.11364308967150044,
"grad_norm": 11.1875,
"learning_rate": 4.549763033175355e-06,
"loss": 1.8286,
"step": 96
},
{
"epoch": 0.11482687185557858,
"grad_norm": 12.75,
"learning_rate": 4.597156398104266e-06,
"loss": 2.0655,
"step": 97
},
{
"epoch": 0.1160106540396567,
"grad_norm": 11.8125,
"learning_rate": 4.644549763033176e-06,
"loss": 1.7114,
"step": 98
},
{
"epoch": 0.11719443622373484,
"grad_norm": 10.8125,
"learning_rate": 4.691943127962086e-06,
"loss": 1.3346,
"step": 99
},
{
"epoch": 0.11837821840781296,
"grad_norm": 12.0,
"learning_rate": 4.739336492890996e-06,
"loss": 2.0403,
"step": 100
},
{
"epoch": 0.1195620005918911,
"grad_norm": 10.9375,
"learning_rate": 4.7867298578199055e-06,
"loss": 1.5846,
"step": 101
},
{
"epoch": 0.12074578277596922,
"grad_norm": 9.8125,
"learning_rate": 4.834123222748816e-06,
"loss": 1.8092,
"step": 102
},
{
"epoch": 0.12192956496004735,
"grad_norm": 10.5625,
"learning_rate": 4.881516587677725e-06,
"loss": 1.7133,
"step": 103
},
{
"epoch": 0.12311334714412549,
"grad_norm": 12.1875,
"learning_rate": 4.928909952606635e-06,
"loss": 2.3588,
"step": 104
},
{
"epoch": 0.12429712932820361,
"grad_norm": 11.0625,
"learning_rate": 4.976303317535545e-06,
"loss": 1.9257,
"step": 105
},
{
"epoch": 0.12548091151228175,
"grad_norm": 12.3125,
"learning_rate": 5.023696682464455e-06,
"loss": 2.0995,
"step": 106
},
{
"epoch": 0.12666469369635988,
"grad_norm": 9.3125,
"learning_rate": 5.071090047393366e-06,
"loss": 1.5075,
"step": 107
},
{
"epoch": 0.127848475880438,
"grad_norm": 10.6875,
"learning_rate": 5.118483412322275e-06,
"loss": 2.0765,
"step": 108
},
{
"epoch": 0.12903225806451613,
"grad_norm": 9.3125,
"learning_rate": 5.165876777251185e-06,
"loss": 1.516,
"step": 109
},
{
"epoch": 0.13021604024859426,
"grad_norm": 11.375,
"learning_rate": 5.213270142180096e-06,
"loss": 2.2197,
"step": 110
},
{
"epoch": 0.1313998224326724,
"grad_norm": 9.5,
"learning_rate": 5.260663507109005e-06,
"loss": 1.4677,
"step": 111
},
{
"epoch": 0.1325836046167505,
"grad_norm": 9.9375,
"learning_rate": 5.308056872037915e-06,
"loss": 2.0798,
"step": 112
},
{
"epoch": 0.13376738680082864,
"grad_norm": 12.1875,
"learning_rate": 5.355450236966825e-06,
"loss": 2.4164,
"step": 113
},
{
"epoch": 0.13495116898490678,
"grad_norm": 9.625,
"learning_rate": 5.402843601895735e-06,
"loss": 1.7392,
"step": 114
},
{
"epoch": 0.1361349511689849,
"grad_norm": 9.8125,
"learning_rate": 5.4502369668246446e-06,
"loss": 1.742,
"step": 115
},
{
"epoch": 0.13731873335306305,
"grad_norm": 8.3125,
"learning_rate": 5.497630331753555e-06,
"loss": 1.7232,
"step": 116
},
{
"epoch": 0.13850251553714116,
"grad_norm": 8.0,
"learning_rate": 5.5450236966824644e-06,
"loss": 1.4722,
"step": 117
},
{
"epoch": 0.1396862977212193,
"grad_norm": 8.9375,
"learning_rate": 5.592417061611375e-06,
"loss": 1.7305,
"step": 118
},
{
"epoch": 0.14087007990529743,
"grad_norm": 10.6875,
"learning_rate": 5.639810426540285e-06,
"loss": 2.0884,
"step": 119
},
{
"epoch": 0.14205386208937557,
"grad_norm": 8.875,
"learning_rate": 5.687203791469195e-06,
"loss": 1.741,
"step": 120
},
{
"epoch": 0.14323764427345367,
"grad_norm": 10.6875,
"learning_rate": 5.734597156398105e-06,
"loss": 2.0642,
"step": 121
},
{
"epoch": 0.1444214264575318,
"grad_norm": 9.625,
"learning_rate": 5.7819905213270145e-06,
"loss": 1.9281,
"step": 122
},
{
"epoch": 0.14560520864160995,
"grad_norm": 10.0,
"learning_rate": 5.829383886255925e-06,
"loss": 1.8116,
"step": 123
},
{
"epoch": 0.14678899082568808,
"grad_norm": 8.375,
"learning_rate": 5.876777251184834e-06,
"loss": 1.8983,
"step": 124
},
{
"epoch": 0.1479727730097662,
"grad_norm": 7.15625,
"learning_rate": 5.924170616113745e-06,
"loss": 1.4895,
"step": 125
},
{
"epoch": 0.14915655519384433,
"grad_norm": 10.5625,
"learning_rate": 5.971563981042654e-06,
"loss": 1.9107,
"step": 126
},
{
"epoch": 0.15034033737792246,
"grad_norm": 8.5,
"learning_rate": 6.018957345971565e-06,
"loss": 1.6736,
"step": 127
},
{
"epoch": 0.1515241195620006,
"grad_norm": 8.5,
"learning_rate": 6.066350710900475e-06,
"loss": 1.5571,
"step": 128
},
{
"epoch": 0.15270790174607873,
"grad_norm": 9.375,
"learning_rate": 6.1137440758293845e-06,
"loss": 1.8141,
"step": 129
},
{
"epoch": 0.15389168393015684,
"grad_norm": 8.625,
"learning_rate": 6.161137440758295e-06,
"loss": 1.9877,
"step": 130
},
{
"epoch": 0.15507546611423498,
"grad_norm": 6.78125,
"learning_rate": 6.208530805687204e-06,
"loss": 1.4212,
"step": 131
},
{
"epoch": 0.1562592482983131,
"grad_norm": 10.5,
"learning_rate": 6.255924170616115e-06,
"loss": 1.7437,
"step": 132
},
{
"epoch": 0.15744303048239125,
"grad_norm": 8.9375,
"learning_rate": 6.303317535545023e-06,
"loss": 1.9611,
"step": 133
},
{
"epoch": 0.15862681266646936,
"grad_norm": 9.375,
"learning_rate": 6.350710900473935e-06,
"loss": 1.3534,
"step": 134
},
{
"epoch": 0.1598105948505475,
"grad_norm": 7.34375,
"learning_rate": 6.398104265402843e-06,
"loss": 1.2178,
"step": 135
},
{
"epoch": 0.16099437703462563,
"grad_norm": 8.25,
"learning_rate": 6.445497630331754e-06,
"loss": 1.2171,
"step": 136
},
{
"epoch": 0.16217815921870377,
"grad_norm": 9.25,
"learning_rate": 6.492890995260665e-06,
"loss": 2.0541,
"step": 137
},
{
"epoch": 0.1633619414027819,
"grad_norm": 9.0,
"learning_rate": 6.5402843601895735e-06,
"loss": 1.3016,
"step": 138
},
{
"epoch": 0.16454572358686,
"grad_norm": 10.0,
"learning_rate": 6.587677725118484e-06,
"loss": 1.8663,
"step": 139
},
{
"epoch": 0.16572950577093815,
"grad_norm": 8.125,
"learning_rate": 6.635071090047393e-06,
"loss": 1.6733,
"step": 140
},
{
"epoch": 0.16691328795501628,
"grad_norm": 9.5625,
"learning_rate": 6.682464454976304e-06,
"loss": 1.8208,
"step": 141
},
{
"epoch": 0.16809707013909442,
"grad_norm": 9.75,
"learning_rate": 6.729857819905213e-06,
"loss": 1.9689,
"step": 142
},
{
"epoch": 0.16928085232317253,
"grad_norm": 9.8125,
"learning_rate": 6.777251184834124e-06,
"loss": 1.8051,
"step": 143
},
{
"epoch": 0.17046463450725066,
"grad_norm": 7.59375,
"learning_rate": 6.824644549763034e-06,
"loss": 1.5444,
"step": 144
},
{
"epoch": 0.1716484166913288,
"grad_norm": 8.125,
"learning_rate": 6.8720379146919435e-06,
"loss": 1.9915,
"step": 145
},
{
"epoch": 0.17283219887540693,
"grad_norm": 9.25,
"learning_rate": 6.919431279620854e-06,
"loss": 2.1294,
"step": 146
},
{
"epoch": 0.17401598105948504,
"grad_norm": 10.3125,
"learning_rate": 6.966824644549763e-06,
"loss": 1.9772,
"step": 147
},
{
"epoch": 0.17519976324356318,
"grad_norm": 8.25,
"learning_rate": 7.014218009478674e-06,
"loss": 1.5921,
"step": 148
},
{
"epoch": 0.17638354542764131,
"grad_norm": 7.21875,
"learning_rate": 7.061611374407583e-06,
"loss": 1.6138,
"step": 149
},
{
"epoch": 0.17756732761171945,
"grad_norm": 8.5,
"learning_rate": 7.1090047393364935e-06,
"loss": 1.5423,
"step": 150
},
{
"epoch": 0.17875110979579759,
"grad_norm": 8.125,
"learning_rate": 7.156398104265403e-06,
"loss": 1.2566,
"step": 151
},
{
"epoch": 0.1799348919798757,
"grad_norm": 9.375,
"learning_rate": 7.203791469194313e-06,
"loss": 2.0622,
"step": 152
},
{
"epoch": 0.18111867416395383,
"grad_norm": 7.4375,
"learning_rate": 7.251184834123224e-06,
"loss": 1.0066,
"step": 153
},
{
"epoch": 0.18230245634803197,
"grad_norm": 8.875,
"learning_rate": 7.298578199052133e-06,
"loss": 2.0575,
"step": 154
},
{
"epoch": 0.1834862385321101,
"grad_norm": 8.625,
"learning_rate": 7.345971563981044e-06,
"loss": 1.7389,
"step": 155
},
{
"epoch": 0.1846700207161882,
"grad_norm": 7.34375,
"learning_rate": 7.393364928909953e-06,
"loss": 1.3202,
"step": 156
},
{
"epoch": 0.18585380290026635,
"grad_norm": 8.1875,
"learning_rate": 7.4407582938388635e-06,
"loss": 1.4784,
"step": 157
},
{
"epoch": 0.18703758508434448,
"grad_norm": 8.8125,
"learning_rate": 7.488151658767773e-06,
"loss": 1.8262,
"step": 158
},
{
"epoch": 0.18822136726842262,
"grad_norm": 10.625,
"learning_rate": 7.535545023696683e-06,
"loss": 1.4627,
"step": 159
},
{
"epoch": 0.18940514945250073,
"grad_norm": 9.0625,
"learning_rate": 7.582938388625593e-06,
"loss": 1.9267,
"step": 160
},
{
"epoch": 0.19058893163657886,
"grad_norm": 8.875,
"learning_rate": 7.630331753554503e-06,
"loss": 2.0932,
"step": 161
},
{
"epoch": 0.191772713820657,
"grad_norm": 6.59375,
"learning_rate": 7.677725118483414e-06,
"loss": 1.5374,
"step": 162
},
{
"epoch": 0.19295649600473513,
"grad_norm": 8.75,
"learning_rate": 7.725118483412322e-06,
"loss": 1.4268,
"step": 163
},
{
"epoch": 0.19414027818881327,
"grad_norm": 8.75,
"learning_rate": 7.772511848341233e-06,
"loss": 2.0425,
"step": 164
},
{
"epoch": 0.19532406037289138,
"grad_norm": 6.375,
"learning_rate": 7.819905213270143e-06,
"loss": 1.3523,
"step": 165
},
{
"epoch": 0.19650784255696951,
"grad_norm": 9.0625,
"learning_rate": 7.867298578199053e-06,
"loss": 1.4911,
"step": 166
},
{
"epoch": 0.19769162474104765,
"grad_norm": 8.6875,
"learning_rate": 7.914691943127962e-06,
"loss": 1.8161,
"step": 167
},
{
"epoch": 0.19887540692512579,
"grad_norm": 7.78125,
"learning_rate": 7.962085308056872e-06,
"loss": 1.7052,
"step": 168
},
{
"epoch": 0.2000591891092039,
"grad_norm": 8.1875,
"learning_rate": 8.009478672985783e-06,
"loss": 1.6209,
"step": 169
},
{
"epoch": 0.20124297129328203,
"grad_norm": 8.125,
"learning_rate": 8.056872037914693e-06,
"loss": 1.7739,
"step": 170
},
{
"epoch": 0.20242675347736017,
"grad_norm": 7.1875,
"learning_rate": 8.104265402843603e-06,
"loss": 1.7383,
"step": 171
},
{
"epoch": 0.2036105356614383,
"grad_norm": 7.65625,
"learning_rate": 8.151658767772512e-06,
"loss": 2.0716,
"step": 172
},
{
"epoch": 0.20479431784551644,
"grad_norm": 6.375,
"learning_rate": 8.199052132701422e-06,
"loss": 1.3346,
"step": 173
},
{
"epoch": 0.20597810002959455,
"grad_norm": 8.4375,
"learning_rate": 8.246445497630333e-06,
"loss": 1.9454,
"step": 174
},
{
"epoch": 0.20716188221367268,
"grad_norm": 8.3125,
"learning_rate": 8.293838862559243e-06,
"loss": 1.7686,
"step": 175
},
{
"epoch": 0.20834566439775082,
"grad_norm": 6.78125,
"learning_rate": 8.341232227488152e-06,
"loss": 1.3993,
"step": 176
},
{
"epoch": 0.20952944658182895,
"grad_norm": 8.875,
"learning_rate": 8.388625592417062e-06,
"loss": 2.0951,
"step": 177
},
{
"epoch": 0.21071322876590706,
"grad_norm": 7.34375,
"learning_rate": 8.436018957345973e-06,
"loss": 1.4131,
"step": 178
},
{
"epoch": 0.2118970109499852,
"grad_norm": 7.0625,
"learning_rate": 8.483412322274883e-06,
"loss": 1.9326,
"step": 179
},
{
"epoch": 0.21308079313406333,
"grad_norm": 7.78125,
"learning_rate": 8.530805687203793e-06,
"loss": 2.0047,
"step": 180
},
{
"epoch": 0.21426457531814147,
"grad_norm": 7.84375,
"learning_rate": 8.578199052132702e-06,
"loss": 1.3454,
"step": 181
},
{
"epoch": 0.21544835750221958,
"grad_norm": 8.125,
"learning_rate": 8.625592417061612e-06,
"loss": 1.6605,
"step": 182
},
{
"epoch": 0.21663213968629771,
"grad_norm": 7.375,
"learning_rate": 8.672985781990521e-06,
"loss": 1.8343,
"step": 183
},
{
"epoch": 0.21781592187037585,
"grad_norm": 7.375,
"learning_rate": 8.720379146919431e-06,
"loss": 1.6831,
"step": 184
},
{
"epoch": 0.218999704054454,
"grad_norm": 6.1875,
"learning_rate": 8.767772511848342e-06,
"loss": 1.077,
"step": 185
},
{
"epoch": 0.22018348623853212,
"grad_norm": 7.875,
"learning_rate": 8.815165876777252e-06,
"loss": 1.7032,
"step": 186
},
{
"epoch": 0.22136726842261023,
"grad_norm": 6.78125,
"learning_rate": 8.862559241706162e-06,
"loss": 1.1127,
"step": 187
},
{
"epoch": 0.22255105060668837,
"grad_norm": 6.90625,
"learning_rate": 8.909952606635071e-06,
"loss": 1.6881,
"step": 188
},
{
"epoch": 0.2237348327907665,
"grad_norm": 7.625,
"learning_rate": 8.957345971563981e-06,
"loss": 1.8561,
"step": 189
},
{
"epoch": 0.22491861497484464,
"grad_norm": 7.0,
"learning_rate": 9.004739336492892e-06,
"loss": 1.6105,
"step": 190
},
{
"epoch": 0.22610239715892275,
"grad_norm": 5.90625,
"learning_rate": 9.052132701421802e-06,
"loss": 1.1169,
"step": 191
},
{
"epoch": 0.22728617934300088,
"grad_norm": 7.59375,
"learning_rate": 9.09952606635071e-06,
"loss": 1.5174,
"step": 192
},
{
"epoch": 0.22846996152707902,
"grad_norm": 8.875,
"learning_rate": 9.146919431279621e-06,
"loss": 1.4527,
"step": 193
},
{
"epoch": 0.22965374371115715,
"grad_norm": 7.75,
"learning_rate": 9.194312796208532e-06,
"loss": 1.6945,
"step": 194
},
{
"epoch": 0.2308375258952353,
"grad_norm": 7.75,
"learning_rate": 9.241706161137442e-06,
"loss": 1.4465,
"step": 195
},
{
"epoch": 0.2320213080793134,
"grad_norm": 8.1875,
"learning_rate": 9.289099526066352e-06,
"loss": 1.9824,
"step": 196
},
{
"epoch": 0.23320509026339153,
"grad_norm": 6.40625,
"learning_rate": 9.336492890995261e-06,
"loss": 1.4604,
"step": 197
},
{
"epoch": 0.23438887244746967,
"grad_norm": 7.09375,
"learning_rate": 9.383886255924171e-06,
"loss": 1.5298,
"step": 198
},
{
"epoch": 0.2355726546315478,
"grad_norm": 6.5,
"learning_rate": 9.431279620853082e-06,
"loss": 1.4588,
"step": 199
},
{
"epoch": 0.23675643681562591,
"grad_norm": 7.53125,
"learning_rate": 9.478672985781992e-06,
"loss": 1.5856,
"step": 200
},
{
"epoch": 0.23794021899970405,
"grad_norm": 7.5625,
"learning_rate": 9.5260663507109e-06,
"loss": 1.5782,
"step": 201
},
{
"epoch": 0.2391240011837822,
"grad_norm": 7.28125,
"learning_rate": 9.573459715639811e-06,
"loss": 1.9513,
"step": 202
},
{
"epoch": 0.24030778336786032,
"grad_norm": 7.5625,
"learning_rate": 9.620853080568721e-06,
"loss": 1.6181,
"step": 203
},
{
"epoch": 0.24149156555193843,
"grad_norm": 8.4375,
"learning_rate": 9.668246445497632e-06,
"loss": 1.6805,
"step": 204
},
{
"epoch": 0.24267534773601657,
"grad_norm": 8.125,
"learning_rate": 9.715639810426542e-06,
"loss": 1.4319,
"step": 205
},
{
"epoch": 0.2438591299200947,
"grad_norm": 7.71875,
"learning_rate": 9.76303317535545e-06,
"loss": 1.4366,
"step": 206
},
{
"epoch": 0.24504291210417284,
"grad_norm": 6.0625,
"learning_rate": 9.810426540284361e-06,
"loss": 1.2241,
"step": 207
},
{
"epoch": 0.24622669428825097,
"grad_norm": 5.71875,
"learning_rate": 9.85781990521327e-06,
"loss": 1.1994,
"step": 208
},
{
"epoch": 0.24741047647232908,
"grad_norm": 7.09375,
"learning_rate": 9.905213270142182e-06,
"loss": 1.4188,
"step": 209
},
{
"epoch": 0.24859425865640722,
"grad_norm": 7.03125,
"learning_rate": 9.95260663507109e-06,
"loss": 1.8346,
"step": 210
},
{
"epoch": 0.24977804084048535,
"grad_norm": 7.96875,
"learning_rate": 1e-05,
"loss": 1.8444,
"step": 211
},
{
"epoch": 0.2509618230245635,
"grad_norm": 6.625,
"learning_rate": 1.004739336492891e-05,
"loss": 1.5463,
"step": 212
},
{
"epoch": 0.2521456052086416,
"grad_norm": 7.0625,
"learning_rate": 1.0094786729857822e-05,
"loss": 1.4663,
"step": 213
},
{
"epoch": 0.25332938739271976,
"grad_norm": 5.1875,
"learning_rate": 1.0142180094786732e-05,
"loss": 1.0079,
"step": 214
},
{
"epoch": 0.25451316957679787,
"grad_norm": 6.46875,
"learning_rate": 1.018957345971564e-05,
"loss": 1.6467,
"step": 215
},
{
"epoch": 0.255696951760876,
"grad_norm": 6.5625,
"learning_rate": 1.023696682464455e-05,
"loss": 1.5228,
"step": 216
},
{
"epoch": 0.25688073394495414,
"grad_norm": 6.46875,
"learning_rate": 1.0284360189573461e-05,
"loss": 1.2899,
"step": 217
},
{
"epoch": 0.25806451612903225,
"grad_norm": 7.46875,
"learning_rate": 1.033175355450237e-05,
"loss": 1.6715,
"step": 218
},
{
"epoch": 0.2592482983131104,
"grad_norm": 8.0,
"learning_rate": 1.037914691943128e-05,
"loss": 1.7818,
"step": 219
},
{
"epoch": 0.2604320804971885,
"grad_norm": 8.125,
"learning_rate": 1.0426540284360192e-05,
"loss": 2.1323,
"step": 220
},
{
"epoch": 0.26161586268126663,
"grad_norm": 6.9375,
"learning_rate": 1.0473933649289101e-05,
"loss": 1.7469,
"step": 221
},
{
"epoch": 0.2627996448653448,
"grad_norm": 4.875,
"learning_rate": 1.052132701421801e-05,
"loss": 0.8507,
"step": 222
},
{
"epoch": 0.2639834270494229,
"grad_norm": 6.625,
"learning_rate": 1.056872037914692e-05,
"loss": 1.4815,
"step": 223
},
{
"epoch": 0.265167209233501,
"grad_norm": 6.6875,
"learning_rate": 1.061611374407583e-05,
"loss": 1.5332,
"step": 224
},
{
"epoch": 0.2663509914175792,
"grad_norm": 7.15625,
"learning_rate": 1.066350710900474e-05,
"loss": 1.5742,
"step": 225
},
{
"epoch": 0.2675347736016573,
"grad_norm": 6.875,
"learning_rate": 1.071090047393365e-05,
"loss": 1.6495,
"step": 226
},
{
"epoch": 0.26871855578573545,
"grad_norm": 7.4375,
"learning_rate": 1.075829383886256e-05,
"loss": 1.4193,
"step": 227
},
{
"epoch": 0.26990233796981355,
"grad_norm": 6.40625,
"learning_rate": 1.080568720379147e-05,
"loss": 1.4858,
"step": 228
},
{
"epoch": 0.27108612015389166,
"grad_norm": 7.28125,
"learning_rate": 1.085308056872038e-05,
"loss": 1.5865,
"step": 229
},
{
"epoch": 0.2722699023379698,
"grad_norm": 7.75,
"learning_rate": 1.0900473933649289e-05,
"loss": 1.9832,
"step": 230
},
{
"epoch": 0.27345368452204794,
"grad_norm": 6.5625,
"learning_rate": 1.0947867298578201e-05,
"loss": 1.6942,
"step": 231
},
{
"epoch": 0.2746374667061261,
"grad_norm": 7.375,
"learning_rate": 1.099526066350711e-05,
"loss": 1.9318,
"step": 232
},
{
"epoch": 0.2758212488902042,
"grad_norm": 7.25,
"learning_rate": 1.104265402843602e-05,
"loss": 1.8121,
"step": 233
},
{
"epoch": 0.2770050310742823,
"grad_norm": 8.125,
"learning_rate": 1.1090047393364929e-05,
"loss": 1.4727,
"step": 234
},
{
"epoch": 0.2781888132583605,
"grad_norm": 8.3125,
"learning_rate": 1.1137440758293841e-05,
"loss": 2.1618,
"step": 235
},
{
"epoch": 0.2793725954424386,
"grad_norm": 6.6875,
"learning_rate": 1.118483412322275e-05,
"loss": 1.4174,
"step": 236
},
{
"epoch": 0.2805563776265167,
"grad_norm": 5.5625,
"learning_rate": 1.1232227488151658e-05,
"loss": 1.3391,
"step": 237
},
{
"epoch": 0.28174015981059486,
"grad_norm": 5.6875,
"learning_rate": 1.127962085308057e-05,
"loss": 1.2655,
"step": 238
},
{
"epoch": 0.28292394199467297,
"grad_norm": 6.84375,
"learning_rate": 1.132701421800948e-05,
"loss": 1.4712,
"step": 239
},
{
"epoch": 0.28410772417875113,
"grad_norm": 8.0625,
"learning_rate": 1.137440758293839e-05,
"loss": 1.7757,
"step": 240
},
{
"epoch": 0.28529150636282924,
"grad_norm": 6.40625,
"learning_rate": 1.1421800947867298e-05,
"loss": 1.5906,
"step": 241
},
{
"epoch": 0.28647528854690735,
"grad_norm": 7.25,
"learning_rate": 1.146919431279621e-05,
"loss": 1.8919,
"step": 242
},
{
"epoch": 0.2876590707309855,
"grad_norm": 6.78125,
"learning_rate": 1.1516587677725119e-05,
"loss": 1.5768,
"step": 243
},
{
"epoch": 0.2888428529150636,
"grad_norm": 7.25,
"learning_rate": 1.1563981042654029e-05,
"loss": 1.6047,
"step": 244
},
{
"epoch": 0.2900266350991418,
"grad_norm": 7.0,
"learning_rate": 1.1611374407582941e-05,
"loss": 1.4856,
"step": 245
},
{
"epoch": 0.2912104172832199,
"grad_norm": 6.28125,
"learning_rate": 1.165876777251185e-05,
"loss": 1.4449,
"step": 246
},
{
"epoch": 0.292394199467298,
"grad_norm": 6.84375,
"learning_rate": 1.1706161137440758e-05,
"loss": 1.5361,
"step": 247
},
{
"epoch": 0.29357798165137616,
"grad_norm": 6.96875,
"learning_rate": 1.1753554502369669e-05,
"loss": 1.7108,
"step": 248
},
{
"epoch": 0.29476176383545427,
"grad_norm": 6.3125,
"learning_rate": 1.180094786729858e-05,
"loss": 1.7784,
"step": 249
},
{
"epoch": 0.2959455460195324,
"grad_norm": 7.78125,
"learning_rate": 1.184834123222749e-05,
"loss": 1.2305,
"step": 250
},
{
"epoch": 0.29712932820361054,
"grad_norm": 5.8125,
"learning_rate": 1.1895734597156398e-05,
"loss": 1.3055,
"step": 251
},
{
"epoch": 0.29831311038768865,
"grad_norm": 5.53125,
"learning_rate": 1.1943127962085309e-05,
"loss": 0.9123,
"step": 252
},
{
"epoch": 0.2994968925717668,
"grad_norm": 6.21875,
"learning_rate": 1.1990521327014219e-05,
"loss": 1.8654,
"step": 253
},
{
"epoch": 0.3006806747558449,
"grad_norm": 7.375,
"learning_rate": 1.203791469194313e-05,
"loss": 1.5772,
"step": 254
},
{
"epoch": 0.30186445693992303,
"grad_norm": 6.5,
"learning_rate": 1.2085308056872038e-05,
"loss": 1.6384,
"step": 255
},
{
"epoch": 0.3030482391240012,
"grad_norm": 7.3125,
"learning_rate": 1.213270142180095e-05,
"loss": 1.3352,
"step": 256
},
{
"epoch": 0.3042320213080793,
"grad_norm": 6.96875,
"learning_rate": 1.2180094786729859e-05,
"loss": 1.6957,
"step": 257
},
{
"epoch": 0.30541580349215747,
"grad_norm": 8.1875,
"learning_rate": 1.2227488151658769e-05,
"loss": 1.8857,
"step": 258
},
{
"epoch": 0.3065995856762356,
"grad_norm": 7.53125,
"learning_rate": 1.2274881516587678e-05,
"loss": 1.8975,
"step": 259
},
{
"epoch": 0.3077833678603137,
"grad_norm": 6.46875,
"learning_rate": 1.232227488151659e-05,
"loss": 1.4811,
"step": 260
},
{
"epoch": 0.30896715004439185,
"grad_norm": 6.3125,
"learning_rate": 1.2369668246445498e-05,
"loss": 1.5917,
"step": 261
},
{
"epoch": 0.31015093222846996,
"grad_norm": 5.125,
"learning_rate": 1.2417061611374409e-05,
"loss": 1.089,
"step": 262
},
{
"epoch": 0.31133471441254806,
"grad_norm": 6.875,
"learning_rate": 1.2464454976303319e-05,
"loss": 1.1836,
"step": 263
},
{
"epoch": 0.3125184965966262,
"grad_norm": 7.0,
"learning_rate": 1.251184834123223e-05,
"loss": 1.6104,
"step": 264
},
{
"epoch": 0.31370227878070434,
"grad_norm": 7.5625,
"learning_rate": 1.2559241706161138e-05,
"loss": 2.0875,
"step": 265
},
{
"epoch": 0.3148860609647825,
"grad_norm": 6.1875,
"learning_rate": 1.2606635071090047e-05,
"loss": 1.2757,
"step": 266
},
{
"epoch": 0.3160698431488606,
"grad_norm": 6.75,
"learning_rate": 1.2654028436018959e-05,
"loss": 1.3903,
"step": 267
},
{
"epoch": 0.3172536253329387,
"grad_norm": 7.65625,
"learning_rate": 1.270142180094787e-05,
"loss": 1.8969,
"step": 268
},
{
"epoch": 0.3184374075170169,
"grad_norm": 6.53125,
"learning_rate": 1.2748815165876778e-05,
"loss": 1.426,
"step": 269
},
{
"epoch": 0.319621189701095,
"grad_norm": 6.03125,
"learning_rate": 1.2796208530805687e-05,
"loss": 1.127,
"step": 270
},
{
"epoch": 0.32080497188517315,
"grad_norm": 7.84375,
"learning_rate": 1.2843601895734599e-05,
"loss": 1.6832,
"step": 271
},
{
"epoch": 0.32198875406925126,
"grad_norm": 8.375,
"learning_rate": 1.2890995260663507e-05,
"loss": 1.4658,
"step": 272
},
{
"epoch": 0.32317253625332937,
"grad_norm": 7.84375,
"learning_rate": 1.2938388625592418e-05,
"loss": 1.7787,
"step": 273
},
{
"epoch": 0.32435631843740753,
"grad_norm": 7.0625,
"learning_rate": 1.298578199052133e-05,
"loss": 1.5036,
"step": 274
},
{
"epoch": 0.32554010062148564,
"grad_norm": 7.4375,
"learning_rate": 1.3033175355450238e-05,
"loss": 1.7151,
"step": 275
},
{
"epoch": 0.3267238828055638,
"grad_norm": 6.78125,
"learning_rate": 1.3080568720379147e-05,
"loss": 1.586,
"step": 276
},
{
"epoch": 0.3279076649896419,
"grad_norm": 6.15625,
"learning_rate": 1.3127962085308057e-05,
"loss": 1.5204,
"step": 277
},
{
"epoch": 0.32909144717372,
"grad_norm": 7.625,
"learning_rate": 1.3175355450236968e-05,
"loss": 1.6707,
"step": 278
},
{
"epoch": 0.3302752293577982,
"grad_norm": 6.9375,
"learning_rate": 1.3222748815165878e-05,
"loss": 1.8481,
"step": 279
},
{
"epoch": 0.3314590115418763,
"grad_norm": 5.15625,
"learning_rate": 1.3270142180094787e-05,
"loss": 1.0195,
"step": 280
},
{
"epoch": 0.3326427937259544,
"grad_norm": 6.9375,
"learning_rate": 1.3317535545023699e-05,
"loss": 2.0383,
"step": 281
},
{
"epoch": 0.33382657591003256,
"grad_norm": 6.46875,
"learning_rate": 1.3364928909952607e-05,
"loss": 1.083,
"step": 282
},
{
"epoch": 0.33501035809411067,
"grad_norm": 7.78125,
"learning_rate": 1.3412322274881518e-05,
"loss": 2.0276,
"step": 283
},
{
"epoch": 0.33619414027818884,
"grad_norm": 6.8125,
"learning_rate": 1.3459715639810426e-05,
"loss": 1.3273,
"step": 284
},
{
"epoch": 0.33737792246226694,
"grad_norm": 5.625,
"learning_rate": 1.3507109004739339e-05,
"loss": 1.1611,
"step": 285
},
{
"epoch": 0.33856170464634505,
"grad_norm": 6.90625,
"learning_rate": 1.3554502369668247e-05,
"loss": 1.4576,
"step": 286
},
{
"epoch": 0.3397454868304232,
"grad_norm": 6.6875,
"learning_rate": 1.3601895734597158e-05,
"loss": 1.5571,
"step": 287
},
{
"epoch": 0.3409292690145013,
"grad_norm": 5.71875,
"learning_rate": 1.3649289099526068e-05,
"loss": 0.9773,
"step": 288
},
{
"epoch": 0.3421130511985795,
"grad_norm": 7.09375,
"learning_rate": 1.3696682464454978e-05,
"loss": 1.5571,
"step": 289
},
{
"epoch": 0.3432968333826576,
"grad_norm": 5.21875,
"learning_rate": 1.3744075829383887e-05,
"loss": 1.0917,
"step": 290
},
{
"epoch": 0.3444806155667357,
"grad_norm": 6.96875,
"learning_rate": 1.3791469194312797e-05,
"loss": 2.0931,
"step": 291
},
{
"epoch": 0.34566439775081387,
"grad_norm": 5.9375,
"learning_rate": 1.3838862559241708e-05,
"loss": 1.3784,
"step": 292
},
{
"epoch": 0.346848179934892,
"grad_norm": 6.53125,
"learning_rate": 1.3886255924170618e-05,
"loss": 1.5724,
"step": 293
},
{
"epoch": 0.3480319621189701,
"grad_norm": 5.75,
"learning_rate": 1.3933649289099527e-05,
"loss": 1.4233,
"step": 294
},
{
"epoch": 0.34921574430304825,
"grad_norm": 7.375,
"learning_rate": 1.3981042654028435e-05,
"loss": 1.5996,
"step": 295
},
{
"epoch": 0.35039952648712636,
"grad_norm": 6.375,
"learning_rate": 1.4028436018957347e-05,
"loss": 1.2861,
"step": 296
},
{
"epoch": 0.3515833086712045,
"grad_norm": 8.125,
"learning_rate": 1.4075829383886258e-05,
"loss": 1.6614,
"step": 297
},
{
"epoch": 0.35276709085528263,
"grad_norm": 6.375,
"learning_rate": 1.4123222748815166e-05,
"loss": 1.5181,
"step": 298
},
{
"epoch": 0.35395087303936074,
"grad_norm": 7.625,
"learning_rate": 1.4170616113744078e-05,
"loss": 1.7997,
"step": 299
},
{
"epoch": 0.3551346552234389,
"grad_norm": 6.9375,
"learning_rate": 1.4218009478672987e-05,
"loss": 1.0847,
"step": 300
},
{
"epoch": 0.356318437407517,
"grad_norm": 8.1875,
"learning_rate": 1.4265402843601896e-05,
"loss": 2.3314,
"step": 301
},
{
"epoch": 0.35750221959159517,
"grad_norm": 6.53125,
"learning_rate": 1.4312796208530806e-05,
"loss": 1.3107,
"step": 302
},
{
"epoch": 0.3586860017756733,
"grad_norm": 7.5625,
"learning_rate": 1.4360189573459718e-05,
"loss": 2.4327,
"step": 303
},
{
"epoch": 0.3598697839597514,
"grad_norm": 5.90625,
"learning_rate": 1.4407582938388627e-05,
"loss": 1.092,
"step": 304
},
{
"epoch": 0.36105356614382955,
"grad_norm": 6.40625,
"learning_rate": 1.4454976303317535e-05,
"loss": 1.4988,
"step": 305
},
{
"epoch": 0.36223734832790766,
"grad_norm": 10.25,
"learning_rate": 1.4502369668246448e-05,
"loss": 1.6112,
"step": 306
},
{
"epoch": 0.36342113051198577,
"grad_norm": 5.90625,
"learning_rate": 1.4549763033175356e-05,
"loss": 1.7611,
"step": 307
},
{
"epoch": 0.36460491269606393,
"grad_norm": 5.71875,
"learning_rate": 1.4597156398104267e-05,
"loss": 1.314,
"step": 308
},
{
"epoch": 0.36578869488014204,
"grad_norm": 5.34375,
"learning_rate": 1.4644549763033175e-05,
"loss": 1.2535,
"step": 309
},
{
"epoch": 0.3669724770642202,
"grad_norm": 7.59375,
"learning_rate": 1.4691943127962087e-05,
"loss": 1.4578,
"step": 310
},
{
"epoch": 0.3681562592482983,
"grad_norm": 6.03125,
"learning_rate": 1.4739336492890996e-05,
"loss": 1.3504,
"step": 311
},
{
"epoch": 0.3693400414323764,
"grad_norm": 6.6875,
"learning_rate": 1.4786729857819906e-05,
"loss": 1.6423,
"step": 312
},
{
"epoch": 0.3705238236164546,
"grad_norm": 6.46875,
"learning_rate": 1.4834123222748817e-05,
"loss": 1.5561,
"step": 313
},
{
"epoch": 0.3717076058005327,
"grad_norm": 4.40625,
"learning_rate": 1.4881516587677727e-05,
"loss": 0.7511,
"step": 314
},
{
"epoch": 0.37289138798461086,
"grad_norm": 6.65625,
"learning_rate": 1.4928909952606636e-05,
"loss": 1.3844,
"step": 315
},
{
"epoch": 0.37407517016868896,
"grad_norm": 5.46875,
"learning_rate": 1.4976303317535546e-05,
"loss": 1.4113,
"step": 316
},
{
"epoch": 0.3752589523527671,
"grad_norm": 5.46875,
"learning_rate": 1.5023696682464456e-05,
"loss": 1.3451,
"step": 317
},
{
"epoch": 0.37644273453684524,
"grad_norm": 7.96875,
"learning_rate": 1.5071090047393367e-05,
"loss": 2.1936,
"step": 318
},
{
"epoch": 0.37762651672092334,
"grad_norm": 6.5,
"learning_rate": 1.5118483412322275e-05,
"loss": 1.3217,
"step": 319
},
{
"epoch": 0.37881029890500145,
"grad_norm": 6.09375,
"learning_rate": 1.5165876777251186e-05,
"loss": 1.6756,
"step": 320
},
{
"epoch": 0.3799940810890796,
"grad_norm": 5.4375,
"learning_rate": 1.5213270142180096e-05,
"loss": 1.2053,
"step": 321
},
{
"epoch": 0.3811778632731577,
"grad_norm": 6.46875,
"learning_rate": 1.5260663507109007e-05,
"loss": 1.5314,
"step": 322
},
{
"epoch": 0.3823616454572359,
"grad_norm": 5.40625,
"learning_rate": 1.5308056872037915e-05,
"loss": 1.1509,
"step": 323
},
{
"epoch": 0.383545427641314,
"grad_norm": 7.1875,
"learning_rate": 1.5355450236966827e-05,
"loss": 1.8336,
"step": 324
},
{
"epoch": 0.3847292098253921,
"grad_norm": 5.71875,
"learning_rate": 1.5402843601895736e-05,
"loss": 1.4806,
"step": 325
},
{
"epoch": 0.38591299200947027,
"grad_norm": 6.28125,
"learning_rate": 1.5450236966824645e-05,
"loss": 1.4553,
"step": 326
},
{
"epoch": 0.3870967741935484,
"grad_norm": 6.71875,
"learning_rate": 1.5497630331753553e-05,
"loss": 1.3168,
"step": 327
},
{
"epoch": 0.38828055637762654,
"grad_norm": 7.8125,
"learning_rate": 1.5545023696682465e-05,
"loss": 1.7397,
"step": 328
},
{
"epoch": 0.38946433856170465,
"grad_norm": 6.71875,
"learning_rate": 1.5592417061611377e-05,
"loss": 1.2998,
"step": 329
},
{
"epoch": 0.39064812074578276,
"grad_norm": 6.625,
"learning_rate": 1.5639810426540286e-05,
"loss": 2.1414,
"step": 330
},
{
"epoch": 0.3918319029298609,
"grad_norm": 6.1875,
"learning_rate": 1.5687203791469198e-05,
"loss": 1.716,
"step": 331
},
{
"epoch": 0.39301568511393903,
"grad_norm": 6.1875,
"learning_rate": 1.5734597156398107e-05,
"loss": 1.6864,
"step": 332
},
{
"epoch": 0.3941994672980172,
"grad_norm": 5.71875,
"learning_rate": 1.5781990521327015e-05,
"loss": 1.5223,
"step": 333
},
{
"epoch": 0.3953832494820953,
"grad_norm": 6.03125,
"learning_rate": 1.5829383886255924e-05,
"loss": 1.2502,
"step": 334
},
{
"epoch": 0.3965670316661734,
"grad_norm": 6.9375,
"learning_rate": 1.5876777251184836e-05,
"loss": 1.4129,
"step": 335
},
{
"epoch": 0.39775081385025157,
"grad_norm": 6.625,
"learning_rate": 1.5924170616113745e-05,
"loss": 1.5021,
"step": 336
},
{
"epoch": 0.3989345960343297,
"grad_norm": 4.34375,
"learning_rate": 1.5971563981042653e-05,
"loss": 0.7062,
"step": 337
},
{
"epoch": 0.4001183782184078,
"grad_norm": 6.25,
"learning_rate": 1.6018957345971565e-05,
"loss": 1.5001,
"step": 338
},
{
"epoch": 0.40130216040248595,
"grad_norm": 5.875,
"learning_rate": 1.6066350710900474e-05,
"loss": 1.2073,
"step": 339
},
{
"epoch": 0.40248594258656406,
"grad_norm": 5.59375,
"learning_rate": 1.6113744075829386e-05,
"loss": 1.2846,
"step": 340
},
{
"epoch": 0.4036697247706422,
"grad_norm": 6.875,
"learning_rate": 1.6161137440758295e-05,
"loss": 1.4911,
"step": 341
},
{
"epoch": 0.40485350695472033,
"grad_norm": 6.84375,
"learning_rate": 1.6208530805687207e-05,
"loss": 1.332,
"step": 342
},
{
"epoch": 0.40603728913879844,
"grad_norm": 5.28125,
"learning_rate": 1.6255924170616116e-05,
"loss": 0.9023,
"step": 343
},
{
"epoch": 0.4072210713228766,
"grad_norm": 7.34375,
"learning_rate": 1.6303317535545024e-05,
"loss": 1.7327,
"step": 344
},
{
"epoch": 0.4084048535069547,
"grad_norm": 7.0,
"learning_rate": 1.6350710900473933e-05,
"loss": 1.9256,
"step": 345
},
{
"epoch": 0.4095886356910329,
"grad_norm": 6.40625,
"learning_rate": 1.6398104265402845e-05,
"loss": 1.5407,
"step": 346
},
{
"epoch": 0.410772417875111,
"grad_norm": 7.4375,
"learning_rate": 1.6445497630331754e-05,
"loss": 1.6706,
"step": 347
},
{
"epoch": 0.4119562000591891,
"grad_norm": 5.96875,
"learning_rate": 1.6492890995260666e-05,
"loss": 1.4983,
"step": 348
},
{
"epoch": 0.41313998224326726,
"grad_norm": 6.09375,
"learning_rate": 1.6540284360189574e-05,
"loss": 1.4087,
"step": 349
},
{
"epoch": 0.41432376442734536,
"grad_norm": 6.65625,
"learning_rate": 1.6587677725118486e-05,
"loss": 1.7009,
"step": 350
},
{
"epoch": 0.4155075466114235,
"grad_norm": 5.59375,
"learning_rate": 1.6635071090047395e-05,
"loss": 1.4062,
"step": 351
},
{
"epoch": 0.41669132879550164,
"grad_norm": 6.25,
"learning_rate": 1.6682464454976304e-05,
"loss": 1.2782,
"step": 352
},
{
"epoch": 0.41787511097957974,
"grad_norm": 6.59375,
"learning_rate": 1.6729857819905216e-05,
"loss": 1.7223,
"step": 353
},
{
"epoch": 0.4190588931636579,
"grad_norm": 5.46875,
"learning_rate": 1.6777251184834124e-05,
"loss": 1.1698,
"step": 354
},
{
"epoch": 0.420242675347736,
"grad_norm": 6.34375,
"learning_rate": 1.6824644549763033e-05,
"loss": 1.7743,
"step": 355
},
{
"epoch": 0.4214264575318141,
"grad_norm": 6.40625,
"learning_rate": 1.6872037914691945e-05,
"loss": 1.6943,
"step": 356
},
{
"epoch": 0.4226102397158923,
"grad_norm": 5.84375,
"learning_rate": 1.6919431279620854e-05,
"loss": 1.483,
"step": 357
},
{
"epoch": 0.4237940218999704,
"grad_norm": 5.875,
"learning_rate": 1.6966824644549766e-05,
"loss": 1.003,
"step": 358
},
{
"epoch": 0.42497780408404856,
"grad_norm": 5.59375,
"learning_rate": 1.7014218009478674e-05,
"loss": 0.9547,
"step": 359
},
{
"epoch": 0.42616158626812667,
"grad_norm": 6.40625,
"learning_rate": 1.7061611374407587e-05,
"loss": 1.6017,
"step": 360
},
{
"epoch": 0.4273453684522048,
"grad_norm": 6.6875,
"learning_rate": 1.7109004739336495e-05,
"loss": 1.4875,
"step": 361
},
{
"epoch": 0.42852915063628294,
"grad_norm": 6.21875,
"learning_rate": 1.7156398104265404e-05,
"loss": 1.535,
"step": 362
},
{
"epoch": 0.42971293282036105,
"grad_norm": 5.65625,
"learning_rate": 1.7203791469194316e-05,
"loss": 1.0059,
"step": 363
},
{
"epoch": 0.43089671500443916,
"grad_norm": 6.71875,
"learning_rate": 1.7251184834123225e-05,
"loss": 1.1937,
"step": 364
},
{
"epoch": 0.4320804971885173,
"grad_norm": 5.875,
"learning_rate": 1.7298578199052133e-05,
"loss": 1.0674,
"step": 365
},
{
"epoch": 0.43326427937259543,
"grad_norm": 6.125,
"learning_rate": 1.7345971563981042e-05,
"loss": 1.7837,
"step": 366
},
{
"epoch": 0.4344480615566736,
"grad_norm": 6.875,
"learning_rate": 1.7393364928909954e-05,
"loss": 1.6398,
"step": 367
},
{
"epoch": 0.4356318437407517,
"grad_norm": 6.28125,
"learning_rate": 1.7440758293838863e-05,
"loss": 1.5272,
"step": 368
},
{
"epoch": 0.4368156259248298,
"grad_norm": 6.5625,
"learning_rate": 1.7488151658767775e-05,
"loss": 1.7976,
"step": 369
},
{
"epoch": 0.437999408108908,
"grad_norm": 5.78125,
"learning_rate": 1.7535545023696683e-05,
"loss": 1.3614,
"step": 370
},
{
"epoch": 0.4391831902929861,
"grad_norm": 5.96875,
"learning_rate": 1.7582938388625595e-05,
"loss": 0.9627,
"step": 371
},
{
"epoch": 0.44036697247706424,
"grad_norm": 5.5625,
"learning_rate": 1.7630331753554504e-05,
"loss": 1.3153,
"step": 372
},
{
"epoch": 0.44155075466114235,
"grad_norm": 5.625,
"learning_rate": 1.7677725118483413e-05,
"loss": 0.8717,
"step": 373
},
{
"epoch": 0.44273453684522046,
"grad_norm": 6.46875,
"learning_rate": 1.7725118483412325e-05,
"loss": 1.3221,
"step": 374
},
{
"epoch": 0.4439183190292986,
"grad_norm": 6.1875,
"learning_rate": 1.7772511848341233e-05,
"loss": 1.6051,
"step": 375
},
{
"epoch": 0.44510210121337673,
"grad_norm": 6.03125,
"learning_rate": 1.7819905213270142e-05,
"loss": 1.6669,
"step": 376
},
{
"epoch": 0.44628588339745484,
"grad_norm": 5.0,
"learning_rate": 1.7867298578199054e-05,
"loss": 1.0366,
"step": 377
},
{
"epoch": 0.447469665581533,
"grad_norm": 6.90625,
"learning_rate": 1.7914691943127963e-05,
"loss": 2.0126,
"step": 378
},
{
"epoch": 0.4486534477656111,
"grad_norm": 6.0,
"learning_rate": 1.7962085308056875e-05,
"loss": 1.461,
"step": 379
},
{
"epoch": 0.4498372299496893,
"grad_norm": 6.53125,
"learning_rate": 1.8009478672985784e-05,
"loss": 1.4245,
"step": 380
},
{
"epoch": 0.4510210121337674,
"grad_norm": 6.125,
"learning_rate": 1.8056872037914696e-05,
"loss": 1.9704,
"step": 381
},
{
"epoch": 0.4522047943178455,
"grad_norm": 6.59375,
"learning_rate": 1.8104265402843604e-05,
"loss": 1.2223,
"step": 382
},
{
"epoch": 0.45338857650192366,
"grad_norm": 5.34375,
"learning_rate": 1.8151658767772513e-05,
"loss": 1.2987,
"step": 383
},
{
"epoch": 0.45457235868600177,
"grad_norm": 5.5625,
"learning_rate": 1.819905213270142e-05,
"loss": 1.5421,
"step": 384
},
{
"epoch": 0.45575614087007993,
"grad_norm": 7.40625,
"learning_rate": 1.8246445497630334e-05,
"loss": 1.4026,
"step": 385
},
{
"epoch": 0.45693992305415804,
"grad_norm": 5.34375,
"learning_rate": 1.8293838862559242e-05,
"loss": 1.2173,
"step": 386
},
{
"epoch": 0.45812370523823615,
"grad_norm": 6.46875,
"learning_rate": 1.8341232227488154e-05,
"loss": 1.7088,
"step": 387
},
{
"epoch": 0.4593074874223143,
"grad_norm": 7.0,
"learning_rate": 1.8388625592417063e-05,
"loss": 1.6153,
"step": 388
},
{
"epoch": 0.4604912696063924,
"grad_norm": 6.53125,
"learning_rate": 1.8436018957345975e-05,
"loss": 1.9964,
"step": 389
},
{
"epoch": 0.4616750517904706,
"grad_norm": 6.9375,
"learning_rate": 1.8483412322274884e-05,
"loss": 1.847,
"step": 390
},
{
"epoch": 0.4628588339745487,
"grad_norm": 4.96875,
"learning_rate": 1.8530805687203792e-05,
"loss": 0.9636,
"step": 391
},
{
"epoch": 0.4640426161586268,
"grad_norm": 5.34375,
"learning_rate": 1.8578199052132704e-05,
"loss": 0.8995,
"step": 392
},
{
"epoch": 0.46522639834270496,
"grad_norm": 6.25,
"learning_rate": 1.8625592417061613e-05,
"loss": 1.5158,
"step": 393
},
{
"epoch": 0.46641018052678307,
"grad_norm": 6.21875,
"learning_rate": 1.8672985781990522e-05,
"loss": 1.494,
"step": 394
},
{
"epoch": 0.4675939627108612,
"grad_norm": 5.6875,
"learning_rate": 1.872037914691943e-05,
"loss": 1.3589,
"step": 395
},
{
"epoch": 0.46877774489493934,
"grad_norm": 6.5625,
"learning_rate": 1.8767772511848342e-05,
"loss": 1.2567,
"step": 396
},
{
"epoch": 0.46996152707901745,
"grad_norm": 6.71875,
"learning_rate": 1.881516587677725e-05,
"loss": 1.6705,
"step": 397
},
{
"epoch": 0.4711453092630956,
"grad_norm": 5.375,
"learning_rate": 1.8862559241706163e-05,
"loss": 1.2286,
"step": 398
},
{
"epoch": 0.4723290914471737,
"grad_norm": 6.6875,
"learning_rate": 1.8909952606635075e-05,
"loss": 1.7102,
"step": 399
},
{
"epoch": 0.47351287363125183,
"grad_norm": 6.9375,
"learning_rate": 1.8957345971563984e-05,
"loss": 1.6101,
"step": 400
},
{
"epoch": 0.47469665581533,
"grad_norm": 6.5625,
"learning_rate": 1.9004739336492893e-05,
"loss": 1.7861,
"step": 401
},
{
"epoch": 0.4758804379994081,
"grad_norm": 5.71875,
"learning_rate": 1.90521327014218e-05,
"loss": 1.3258,
"step": 402
},
{
"epoch": 0.47706422018348627,
"grad_norm": 5.3125,
"learning_rate": 1.9099526066350713e-05,
"loss": 1.1771,
"step": 403
},
{
"epoch": 0.4782480023675644,
"grad_norm": 5.09375,
"learning_rate": 1.9146919431279622e-05,
"loss": 1.0935,
"step": 404
},
{
"epoch": 0.4794317845516425,
"grad_norm": 6.1875,
"learning_rate": 1.919431279620853e-05,
"loss": 1.5818,
"step": 405
},
{
"epoch": 0.48061556673572065,
"grad_norm": 5.90625,
"learning_rate": 1.9241706161137443e-05,
"loss": 1.4282,
"step": 406
},
{
"epoch": 0.48179934891979875,
"grad_norm": 5.875,
"learning_rate": 1.928909952606635e-05,
"loss": 1.4932,
"step": 407
},
{
"epoch": 0.48298313110387686,
"grad_norm": 6.53125,
"learning_rate": 1.9336492890995263e-05,
"loss": 1.2801,
"step": 408
},
{
"epoch": 0.484166913287955,
"grad_norm": 6.25,
"learning_rate": 1.9383886255924172e-05,
"loss": 1.6586,
"step": 409
},
{
"epoch": 0.48535069547203313,
"grad_norm": 6.65625,
"learning_rate": 1.9431279620853084e-05,
"loss": 1.2842,
"step": 410
},
{
"epoch": 0.4865344776561113,
"grad_norm": 6.15625,
"learning_rate": 1.9478672985781993e-05,
"loss": 1.4286,
"step": 411
},
{
"epoch": 0.4877182598401894,
"grad_norm": 5.65625,
"learning_rate": 1.95260663507109e-05,
"loss": 1.3082,
"step": 412
},
{
"epoch": 0.4889020420242675,
"grad_norm": 6.5625,
"learning_rate": 1.957345971563981e-05,
"loss": 1.717,
"step": 413
},
{
"epoch": 0.4900858242083457,
"grad_norm": 6.0,
"learning_rate": 1.9620853080568722e-05,
"loss": 1.355,
"step": 414
},
{
"epoch": 0.4912696063924238,
"grad_norm": 6.5,
"learning_rate": 1.966824644549763e-05,
"loss": 2.0023,
"step": 415
},
{
"epoch": 0.49245338857650195,
"grad_norm": 7.0,
"learning_rate": 1.971563981042654e-05,
"loss": 1.7158,
"step": 416
},
{
"epoch": 0.49363717076058006,
"grad_norm": 6.5625,
"learning_rate": 1.976303317535545e-05,
"loss": 1.4434,
"step": 417
},
{
"epoch": 0.49482095294465817,
"grad_norm": 6.5,
"learning_rate": 1.9810426540284364e-05,
"loss": 1.4009,
"step": 418
},
{
"epoch": 0.49600473512873633,
"grad_norm": 6.3125,
"learning_rate": 1.9857819905213272e-05,
"loss": 1.6251,
"step": 419
},
{
"epoch": 0.49718851731281444,
"grad_norm": 6.6875,
"learning_rate": 1.990521327014218e-05,
"loss": 1.4363,
"step": 420
},
{
"epoch": 0.49837229949689255,
"grad_norm": 6.375,
"learning_rate": 1.9952606635071093e-05,
"loss": 1.3963,
"step": 421
},
{
"epoch": 0.4995560816809707,
"grad_norm": 7.5,
"learning_rate": 2e-05,
"loss": 1.4491,
"step": 422
},
{
"epoch": 0.5007398638650489,
"grad_norm": 6.09375,
"learning_rate": 1.9999996578946986e-05,
"loss": 1.3617,
"step": 423
},
{
"epoch": 0.501923646049127,
"grad_norm": 6.84375,
"learning_rate": 1.9999986315790288e-05,
"loss": 1.5132,
"step": 424
},
{
"epoch": 0.5031074282332051,
"grad_norm": 6.625,
"learning_rate": 1.9999969210536924e-05,
"loss": 1.938,
"step": 425
},
{
"epoch": 0.5042912104172832,
"grad_norm": 6.46875,
"learning_rate": 1.99999452631986e-05,
"loss": 1.5161,
"step": 426
},
{
"epoch": 0.5054749926013613,
"grad_norm": 5.78125,
"learning_rate": 1.9999914473791704e-05,
"loss": 0.9499,
"step": 427
},
{
"epoch": 0.5066587747854395,
"grad_norm": 7.34375,
"learning_rate": 1.9999876842337294e-05,
"loss": 1.4799,
"step": 428
},
{
"epoch": 0.5078425569695176,
"grad_norm": 6.75,
"learning_rate": 1.9999832368861126e-05,
"loss": 1.5696,
"step": 429
},
{
"epoch": 0.5090263391535957,
"grad_norm": 5.9375,
"learning_rate": 1.9999781053393626e-05,
"loss": 1.2694,
"step": 430
},
{
"epoch": 0.5102101213376738,
"grad_norm": 3.609375,
"learning_rate": 1.9999722895969904e-05,
"loss": 0.6298,
"step": 431
},
{
"epoch": 0.511393903521752,
"grad_norm": 5.65625,
"learning_rate": 1.9999657896629752e-05,
"loss": 1.3767,
"step": 432
},
{
"epoch": 0.5125776857058302,
"grad_norm": 5.9375,
"learning_rate": 1.9999586055417645e-05,
"loss": 1.7057,
"step": 433
},
{
"epoch": 0.5137614678899083,
"grad_norm": 5.8125,
"learning_rate": 1.9999507372382738e-05,
"loss": 1.4546,
"step": 434
},
{
"epoch": 0.5149452500739864,
"grad_norm": 6.34375,
"learning_rate": 1.9999421847578864e-05,
"loss": 1.1604,
"step": 435
},
{
"epoch": 0.5161290322580645,
"grad_norm": 7.5,
"learning_rate": 1.999932948106454e-05,
"loss": 2.0266,
"step": 436
},
{
"epoch": 0.5173128144421426,
"grad_norm": 7.375,
"learning_rate": 1.999923027290297e-05,
"loss": 1.7186,
"step": 437
},
{
"epoch": 0.5184965966262208,
"grad_norm": 6.34375,
"learning_rate": 1.9999124223162023e-05,
"loss": 1.8008,
"step": 438
},
{
"epoch": 0.5196803788102989,
"grad_norm": 6.28125,
"learning_rate": 1.9999011331914264e-05,
"loss": 1.5211,
"step": 439
},
{
"epoch": 0.520864160994377,
"grad_norm": 6.15625,
"learning_rate": 1.999889159923694e-05,
"loss": 1.5905,
"step": 440
},
{
"epoch": 0.5220479431784552,
"grad_norm": 6.25,
"learning_rate": 1.999876502521197e-05,
"loss": 1.4618,
"step": 441
},
{
"epoch": 0.5232317253625333,
"grad_norm": 4.8125,
"learning_rate": 1.999863160992595e-05,
"loss": 0.9064,
"step": 442
},
{
"epoch": 0.5244155075466114,
"grad_norm": 5.8125,
"learning_rate": 1.9998491353470176e-05,
"loss": 1.509,
"step": 443
},
{
"epoch": 0.5255992897306896,
"grad_norm": 5.53125,
"learning_rate": 1.9998344255940602e-05,
"loss": 1.4633,
"step": 444
},
{
"epoch": 0.5267830719147677,
"grad_norm": 4.34375,
"learning_rate": 1.9998190317437883e-05,
"loss": 0.7265,
"step": 445
},
{
"epoch": 0.5279668540988458,
"grad_norm": 6.09375,
"learning_rate": 1.999802953806734e-05,
"loss": 1.0154,
"step": 446
},
{
"epoch": 0.5291506362829239,
"grad_norm": 6.3125,
"learning_rate": 1.999786191793898e-05,
"loss": 1.6285,
"step": 447
},
{
"epoch": 0.530334418467002,
"grad_norm": 5.59375,
"learning_rate": 1.9997687457167495e-05,
"loss": 1.5863,
"step": 448
},
{
"epoch": 0.5315182006510802,
"grad_norm": 5.59375,
"learning_rate": 1.9997506155872246e-05,
"loss": 1.1283,
"step": 449
},
{
"epoch": 0.5327019828351583,
"grad_norm": 6.09375,
"learning_rate": 1.9997318014177283e-05,
"loss": 1.0686,
"step": 450
},
{
"epoch": 0.5338857650192365,
"grad_norm": 5.84375,
"learning_rate": 1.9997123032211343e-05,
"loss": 1.6487,
"step": 451
},
{
"epoch": 0.5350695472033146,
"grad_norm": 5.78125,
"learning_rate": 1.9996921210107823e-05,
"loss": 1.5499,
"step": 452
},
{
"epoch": 0.5362533293873927,
"grad_norm": 5.9375,
"learning_rate": 1.999671254800482e-05,
"loss": 1.5186,
"step": 453
},
{
"epoch": 0.5374371115714709,
"grad_norm": 6.3125,
"learning_rate": 1.9996497046045096e-05,
"loss": 1.2583,
"step": 454
},
{
"epoch": 0.538620893755549,
"grad_norm": 5.1875,
"learning_rate": 1.9996274704376105e-05,
"loss": 1.3144,
"step": 455
},
{
"epoch": 0.5398046759396271,
"grad_norm": 4.59375,
"learning_rate": 1.9996045523149977e-05,
"loss": 1.0985,
"step": 456
},
{
"epoch": 0.5409884581237052,
"grad_norm": 5.5,
"learning_rate": 1.9995809502523513e-05,
"loss": 1.1678,
"step": 457
},
{
"epoch": 0.5421722403077833,
"grad_norm": 7.125,
"learning_rate": 1.9995566642658208e-05,
"loss": 1.6436,
"step": 458
},
{
"epoch": 0.5433560224918615,
"grad_norm": 7.0625,
"learning_rate": 1.9995316943720225e-05,
"loss": 1.8564,
"step": 459
},
{
"epoch": 0.5445398046759397,
"grad_norm": 5.59375,
"learning_rate": 1.9995060405880412e-05,
"loss": 1.6981,
"step": 460
},
{
"epoch": 0.5457235868600178,
"grad_norm": 5.5625,
"learning_rate": 1.999479702931429e-05,
"loss": 1.335,
"step": 461
},
{
"epoch": 0.5469073690440959,
"grad_norm": 5.3125,
"learning_rate": 1.9994526814202076e-05,
"loss": 1.1632,
"step": 462
},
{
"epoch": 0.548091151228174,
"grad_norm": 7.875,
"learning_rate": 1.9994249760728645e-05,
"loss": 1.3636,
"step": 463
},
{
"epoch": 0.5492749334122522,
"grad_norm": 6.1875,
"learning_rate": 1.9993965869083562e-05,
"loss": 1.427,
"step": 464
},
{
"epoch": 0.5504587155963303,
"grad_norm": 5.0,
"learning_rate": 1.999367513946107e-05,
"loss": 1.2135,
"step": 465
},
{
"epoch": 0.5516424977804084,
"grad_norm": 5.5625,
"learning_rate": 1.9993377572060084e-05,
"loss": 1.3433,
"step": 466
},
{
"epoch": 0.5528262799644865,
"grad_norm": 5.03125,
"learning_rate": 1.999307316708421e-05,
"loss": 1.0626,
"step": 467
},
{
"epoch": 0.5540100621485646,
"grad_norm": 5.875,
"learning_rate": 1.9992761924741718e-05,
"loss": 1.3228,
"step": 468
},
{
"epoch": 0.5551938443326427,
"grad_norm": 4.9375,
"learning_rate": 1.9992443845245572e-05,
"loss": 1.1424,
"step": 469
},
{
"epoch": 0.556377626516721,
"grad_norm": 6.09375,
"learning_rate": 1.99921189288134e-05,
"loss": 1.1614,
"step": 470
},
{
"epoch": 0.5575614087007991,
"grad_norm": 6.65625,
"learning_rate": 1.999178717566751e-05,
"loss": 1.8883,
"step": 471
},
{
"epoch": 0.5587451908848772,
"grad_norm": 6.4375,
"learning_rate": 1.9991448586034897e-05,
"loss": 2.1599,
"step": 472
},
{
"epoch": 0.5599289730689553,
"grad_norm": 6.1875,
"learning_rate": 1.9991103160147226e-05,
"loss": 1.0956,
"step": 473
},
{
"epoch": 0.5611127552530334,
"grad_norm": 6.09375,
"learning_rate": 1.9990750898240837e-05,
"loss": 1.7254,
"step": 474
},
{
"epoch": 0.5622965374371116,
"grad_norm": 5.875,
"learning_rate": 1.9990391800556757e-05,
"loss": 1.6371,
"step": 475
},
{
"epoch": 0.5634803196211897,
"grad_norm": 5.9375,
"learning_rate": 1.9990025867340683e-05,
"loss": 1.2713,
"step": 476
},
{
"epoch": 0.5646641018052678,
"grad_norm": 9.0,
"learning_rate": 1.998965309884299e-05,
"loss": 1.5838,
"step": 477
},
{
"epoch": 0.5658478839893459,
"grad_norm": 4.96875,
"learning_rate": 1.9989273495318726e-05,
"loss": 1.2414,
"step": 478
},
{
"epoch": 0.567031666173424,
"grad_norm": 5.90625,
"learning_rate": 1.9988887057027627e-05,
"loss": 1.1634,
"step": 479
},
{
"epoch": 0.5682154483575023,
"grad_norm": 5.65625,
"learning_rate": 1.998849378423409e-05,
"loss": 1.0765,
"step": 480
},
{
"epoch": 0.5693992305415804,
"grad_norm": 4.71875,
"learning_rate": 1.998809367720721e-05,
"loss": 1.2561,
"step": 481
},
{
"epoch": 0.5705830127256585,
"grad_norm": 7.15625,
"learning_rate": 1.9987686736220726e-05,
"loss": 1.8462,
"step": 482
},
{
"epoch": 0.5717667949097366,
"grad_norm": 6.40625,
"learning_rate": 1.9987272961553087e-05,
"loss": 2.2064,
"step": 483
},
{
"epoch": 0.5729505770938147,
"grad_norm": 5.84375,
"learning_rate": 1.998685235348739e-05,
"loss": 1.561,
"step": 484
},
{
"epoch": 0.5741343592778929,
"grad_norm": 5.84375,
"learning_rate": 1.998642491231143e-05,
"loss": 1.6648,
"step": 485
},
{
"epoch": 0.575318141461971,
"grad_norm": 6.4375,
"learning_rate": 1.9985990638317663e-05,
"loss": 1.5712,
"step": 486
},
{
"epoch": 0.5765019236460491,
"grad_norm": 6.65625,
"learning_rate": 1.998554953180322e-05,
"loss": 1.5301,
"step": 487
},
{
"epoch": 0.5776857058301272,
"grad_norm": 5.8125,
"learning_rate": 1.9985101593069914e-05,
"loss": 1.3492,
"step": 488
},
{
"epoch": 0.5788694880142053,
"grad_norm": 6.5625,
"learning_rate": 1.998464682242423e-05,
"loss": 1.9684,
"step": 489
},
{
"epoch": 0.5800532701982836,
"grad_norm": 5.65625,
"learning_rate": 1.9984185220177327e-05,
"loss": 1.7987,
"step": 490
},
{
"epoch": 0.5812370523823617,
"grad_norm": 5.40625,
"learning_rate": 1.9983716786645035e-05,
"loss": 1.0255,
"step": 491
},
{
"epoch": 0.5824208345664398,
"grad_norm": 6.5625,
"learning_rate": 1.9983241522147864e-05,
"loss": 1.5068,
"step": 492
},
{
"epoch": 0.5836046167505179,
"grad_norm": 7.40625,
"learning_rate": 1.9982759427011e-05,
"loss": 1.7253,
"step": 493
},
{
"epoch": 0.584788398934596,
"grad_norm": 6.0,
"learning_rate": 1.9982270501564286e-05,
"loss": 1.58,
"step": 494
},
{
"epoch": 0.5859721811186742,
"grad_norm": 5.84375,
"learning_rate": 1.9981774746142254e-05,
"loss": 1.3031,
"step": 495
},
{
"epoch": 0.5871559633027523,
"grad_norm": 6.34375,
"learning_rate": 1.9981272161084113e-05,
"loss": 1.4522,
"step": 496
},
{
"epoch": 0.5883397454868304,
"grad_norm": 5.4375,
"learning_rate": 1.998076274673373e-05,
"loss": 1.5209,
"step": 497
},
{
"epoch": 0.5895235276709085,
"grad_norm": 5.78125,
"learning_rate": 1.998024650343965e-05,
"loss": 1.2254,
"step": 498
},
{
"epoch": 0.5907073098549867,
"grad_norm": 5.75,
"learning_rate": 1.9979723431555098e-05,
"loss": 1.3998,
"step": 499
},
{
"epoch": 0.5918910920390648,
"grad_norm": 5.65625,
"learning_rate": 1.9979193531437962e-05,
"loss": 1.4027,
"step": 500
},
{
"epoch": 0.593074874223143,
"grad_norm": 5.15625,
"learning_rate": 1.9978656803450804e-05,
"loss": 1.0404,
"step": 501
},
{
"epoch": 0.5942586564072211,
"grad_norm": 5.375,
"learning_rate": 1.9978113247960862e-05,
"loss": 1.317,
"step": 502
},
{
"epoch": 0.5954424385912992,
"grad_norm": 4.71875,
"learning_rate": 1.997756286534004e-05,
"loss": 1.0851,
"step": 503
},
{
"epoch": 0.5966262207753773,
"grad_norm": 6.125,
"learning_rate": 1.9977005655964916e-05,
"loss": 1.5949,
"step": 504
},
{
"epoch": 0.5978100029594554,
"grad_norm": 6.375,
"learning_rate": 1.997644162021674e-05,
"loss": 1.4574,
"step": 505
},
{
"epoch": 0.5989937851435336,
"grad_norm": 6.40625,
"learning_rate": 1.997587075848143e-05,
"loss": 1.6315,
"step": 506
},
{
"epoch": 0.6001775673276117,
"grad_norm": 6.125,
"learning_rate": 1.9975293071149577e-05,
"loss": 1.3118,
"step": 507
},
{
"epoch": 0.6013613495116898,
"grad_norm": 5.5,
"learning_rate": 1.997470855861644e-05,
"loss": 1.4691,
"step": 508
},
{
"epoch": 0.602545131695768,
"grad_norm": 6.25,
"learning_rate": 1.997411722128195e-05,
"loss": 1.791,
"step": 509
},
{
"epoch": 0.6037289138798461,
"grad_norm": 4.84375,
"learning_rate": 1.99735190595507e-05,
"loss": 1.3449,
"step": 510
},
{
"epoch": 0.6049126960639243,
"grad_norm": 5.9375,
"learning_rate": 1.9972914073831967e-05,
"loss": 1.5904,
"step": 511
},
{
"epoch": 0.6060964782480024,
"grad_norm": 5.71875,
"learning_rate": 1.9972302264539686e-05,
"loss": 1.4242,
"step": 512
},
{
"epoch": 0.6072802604320805,
"grad_norm": 6.15625,
"learning_rate": 1.997168363209246e-05,
"loss": 1.2606,
"step": 513
},
{
"epoch": 0.6084640426161586,
"grad_norm": 6.46875,
"learning_rate": 1.9971058176913566e-05,
"loss": 1.8071,
"step": 514
},
{
"epoch": 0.6096478248002367,
"grad_norm": 5.84375,
"learning_rate": 1.997042589943095e-05,
"loss": 1.5618,
"step": 515
},
{
"epoch": 0.6108316069843149,
"grad_norm": 5.59375,
"learning_rate": 1.9969786800077217e-05,
"loss": 1.2276,
"step": 516
},
{
"epoch": 0.612015389168393,
"grad_norm": 5.28125,
"learning_rate": 1.9969140879289654e-05,
"loss": 1.0109,
"step": 517
},
{
"epoch": 0.6131991713524712,
"grad_norm": 4.78125,
"learning_rate": 1.99684881375102e-05,
"loss": 1.067,
"step": 518
},
{
"epoch": 0.6143829535365493,
"grad_norm": 5.5,
"learning_rate": 1.9967828575185467e-05,
"loss": 1.4382,
"step": 519
},
{
"epoch": 0.6155667357206274,
"grad_norm": 7.03125,
"learning_rate": 1.996716219276674e-05,
"loss": 1.5627,
"step": 520
},
{
"epoch": 0.6167505179047056,
"grad_norm": 5.25,
"learning_rate": 1.996648899070996e-05,
"loss": 1.352,
"step": 521
},
{
"epoch": 0.6179343000887837,
"grad_norm": 6.46875,
"learning_rate": 1.996580896947574e-05,
"loss": 1.3839,
"step": 522
},
{
"epoch": 0.6191180822728618,
"grad_norm": 5.78125,
"learning_rate": 1.9965122129529362e-05,
"loss": 1.3309,
"step": 523
},
{
"epoch": 0.6203018644569399,
"grad_norm": 6.46875,
"learning_rate": 1.9964428471340765e-05,
"loss": 1.785,
"step": 524
},
{
"epoch": 0.621485646641018,
"grad_norm": 5.125,
"learning_rate": 1.9963727995384558e-05,
"loss": 0.947,
"step": 525
},
{
"epoch": 0.6226694288250961,
"grad_norm": 5.75,
"learning_rate": 1.9963020702140016e-05,
"loss": 1.1448,
"step": 526
},
{
"epoch": 0.6238532110091743,
"grad_norm": 5.5625,
"learning_rate": 1.9962306592091073e-05,
"loss": 1.151,
"step": 527
},
{
"epoch": 0.6250369931932525,
"grad_norm": 6.78125,
"learning_rate": 1.9961585665726332e-05,
"loss": 1.4072,
"step": 528
},
{
"epoch": 0.6262207753773306,
"grad_norm": 6.25,
"learning_rate": 1.996085792353906e-05,
"loss": 1.6292,
"step": 529
},
{
"epoch": 0.6274045575614087,
"grad_norm": 6.34375,
"learning_rate": 1.9960123366027187e-05,
"loss": 1.7855,
"step": 530
},
{
"epoch": 0.6285883397454868,
"grad_norm": 5.53125,
"learning_rate": 1.99593819936933e-05,
"loss": 1.4554,
"step": 531
},
{
"epoch": 0.629772121929565,
"grad_norm": 4.875,
"learning_rate": 1.9958633807044657e-05,
"loss": 0.9738,
"step": 532
},
{
"epoch": 0.6309559041136431,
"grad_norm": 7.03125,
"learning_rate": 1.9957878806593176e-05,
"loss": 1.3923,
"step": 533
},
{
"epoch": 0.6321396862977212,
"grad_norm": 5.4375,
"learning_rate": 1.9957116992855435e-05,
"loss": 1.3379,
"step": 534
},
{
"epoch": 0.6333234684817993,
"grad_norm": 5.59375,
"learning_rate": 1.9956348366352675e-05,
"loss": 1.4911,
"step": 535
},
{
"epoch": 0.6345072506658774,
"grad_norm": 6.53125,
"learning_rate": 1.9955572927610798e-05,
"loss": 1.857,
"step": 536
},
{
"epoch": 0.6356910328499557,
"grad_norm": 7.09375,
"learning_rate": 1.9954790677160368e-05,
"loss": 1.9798,
"step": 537
},
{
"epoch": 0.6368748150340338,
"grad_norm": 5.125,
"learning_rate": 1.9954001615536612e-05,
"loss": 0.9763,
"step": 538
},
{
"epoch": 0.6380585972181119,
"grad_norm": 6.75,
"learning_rate": 1.995320574327941e-05,
"loss": 2.5139,
"step": 539
},
{
"epoch": 0.63924237940219,
"grad_norm": 5.34375,
"learning_rate": 1.995240306093331e-05,
"loss": 1.4293,
"step": 540
},
{
"epoch": 0.6404261615862681,
"grad_norm": 6.0,
"learning_rate": 1.9951593569047513e-05,
"loss": 1.398,
"step": 541
},
{
"epoch": 0.6416099437703463,
"grad_norm": 5.53125,
"learning_rate": 1.995077726817588e-05,
"loss": 1.4907,
"step": 542
},
{
"epoch": 0.6427937259544244,
"grad_norm": 5.59375,
"learning_rate": 1.9949954158876938e-05,
"loss": 1.3604,
"step": 543
},
{
"epoch": 0.6439775081385025,
"grad_norm": 5.71875,
"learning_rate": 1.994912424171386e-05,
"loss": 0.9204,
"step": 544
},
{
"epoch": 0.6451612903225806,
"grad_norm": 5.0,
"learning_rate": 1.994828751725449e-05,
"loss": 1.2397,
"step": 545
},
{
"epoch": 0.6463450725066587,
"grad_norm": 5.0,
"learning_rate": 1.9947443986071328e-05,
"loss": 1.1915,
"step": 546
},
{
"epoch": 0.647528854690737,
"grad_norm": 4.84375,
"learning_rate": 1.9946593648741515e-05,
"loss": 1.1682,
"step": 547
},
{
"epoch": 0.6487126368748151,
"grad_norm": 4.3125,
"learning_rate": 1.9945736505846866e-05,
"loss": 0.8412,
"step": 548
},
{
"epoch": 0.6498964190588932,
"grad_norm": 5.9375,
"learning_rate": 1.994487255797385e-05,
"loss": 1.781,
"step": 549
},
{
"epoch": 0.6510802012429713,
"grad_norm": 5.78125,
"learning_rate": 1.994400180571359e-05,
"loss": 1.4809,
"step": 550
},
{
"epoch": 0.6522639834270494,
"grad_norm": 6.8125,
"learning_rate": 1.994312424966186e-05,
"loss": 1.6193,
"step": 551
},
{
"epoch": 0.6534477656111276,
"grad_norm": 5.125,
"learning_rate": 1.9942239890419097e-05,
"loss": 1.3365,
"step": 552
},
{
"epoch": 0.6546315477952057,
"grad_norm": 5.625,
"learning_rate": 1.9941348728590385e-05,
"loss": 1.2196,
"step": 553
},
{
"epoch": 0.6558153299792838,
"grad_norm": 5.34375,
"learning_rate": 1.9940450764785465e-05,
"loss": 0.8375,
"step": 554
},
{
"epoch": 0.6569991121633619,
"grad_norm": 5.0625,
"learning_rate": 1.9939545999618737e-05,
"loss": 1.1208,
"step": 555
},
{
"epoch": 0.65818289434744,
"grad_norm": 5.375,
"learning_rate": 1.9938634433709254e-05,
"loss": 1.2928,
"step": 556
},
{
"epoch": 0.6593666765315181,
"grad_norm": 6.0625,
"learning_rate": 1.9937716067680712e-05,
"loss": 1.4053,
"step": 557
},
{
"epoch": 0.6605504587155964,
"grad_norm": 4.78125,
"learning_rate": 1.993679090216147e-05,
"loss": 1.3616,
"step": 558
},
{
"epoch": 0.6617342408996745,
"grad_norm": 5.8125,
"learning_rate": 1.9935858937784537e-05,
"loss": 1.2772,
"step": 559
},
{
"epoch": 0.6629180230837526,
"grad_norm": 5.40625,
"learning_rate": 1.993492017518757e-05,
"loss": 1.3156,
"step": 560
},
{
"epoch": 0.6641018052678307,
"grad_norm": 6.5,
"learning_rate": 1.9933974615012888e-05,
"loss": 2.2163,
"step": 561
},
{
"epoch": 0.6652855874519088,
"grad_norm": 5.25,
"learning_rate": 1.9933022257907445e-05,
"loss": 1.2261,
"step": 562
},
{
"epoch": 0.666469369635987,
"grad_norm": 4.4375,
"learning_rate": 1.9932063104522857e-05,
"loss": 1.0029,
"step": 563
},
{
"epoch": 0.6676531518200651,
"grad_norm": 5.34375,
"learning_rate": 1.9931097155515383e-05,
"loss": 1.1303,
"step": 564
},
{
"epoch": 0.6688369340041432,
"grad_norm": 5.625,
"learning_rate": 1.9930124411545943e-05,
"loss": 1.6805,
"step": 565
},
{
"epoch": 0.6700207161882213,
"grad_norm": 4.625,
"learning_rate": 1.9929144873280092e-05,
"loss": 1.2781,
"step": 566
},
{
"epoch": 0.6712044983722995,
"grad_norm": 5.5,
"learning_rate": 1.9928158541388046e-05,
"loss": 1.5283,
"step": 567
},
{
"epoch": 0.6723882805563777,
"grad_norm": 5.09375,
"learning_rate": 1.9927165416544656e-05,
"loss": 1.25,
"step": 568
},
{
"epoch": 0.6735720627404558,
"grad_norm": 6.25,
"learning_rate": 1.9926165499429437e-05,
"loss": 1.5649,
"step": 569
},
{
"epoch": 0.6747558449245339,
"grad_norm": 5.0,
"learning_rate": 1.992515879072654e-05,
"loss": 1.1627,
"step": 570
},
{
"epoch": 0.675939627108612,
"grad_norm": 4.65625,
"learning_rate": 1.9924145291124764e-05,
"loss": 1.1447,
"step": 571
},
{
"epoch": 0.6771234092926901,
"grad_norm": 5.0625,
"learning_rate": 1.9923125001317557e-05,
"loss": 1.2525,
"step": 572
},
{
"epoch": 0.6783071914767683,
"grad_norm": 6.09375,
"learning_rate": 1.9922097922003016e-05,
"loss": 0.9096,
"step": 573
},
{
"epoch": 0.6794909736608464,
"grad_norm": 5.875,
"learning_rate": 1.992106405388387e-05,
"loss": 1.6442,
"step": 574
},
{
"epoch": 0.6806747558449245,
"grad_norm": 7.34375,
"learning_rate": 1.992002339766751e-05,
"loss": 1.5446,
"step": 575
},
{
"epoch": 0.6818585380290026,
"grad_norm": 5.21875,
"learning_rate": 1.9918975954065964e-05,
"loss": 1.4214,
"step": 576
},
{
"epoch": 0.6830423202130808,
"grad_norm": 5.25,
"learning_rate": 1.9917921723795903e-05,
"loss": 1.5916,
"step": 577
},
{
"epoch": 0.684226102397159,
"grad_norm": 6.28125,
"learning_rate": 1.9916860707578642e-05,
"loss": 1.8495,
"step": 578
},
{
"epoch": 0.6854098845812371,
"grad_norm": 7.25,
"learning_rate": 1.9915792906140138e-05,
"loss": 1.564,
"step": 579
},
{
"epoch": 0.6865936667653152,
"grad_norm": 6.875,
"learning_rate": 1.9914718320210994e-05,
"loss": 1.622,
"step": 580
},
{
"epoch": 0.6877774489493933,
"grad_norm": 6.53125,
"learning_rate": 1.9913636950526455e-05,
"loss": 1.7955,
"step": 581
},
{
"epoch": 0.6889612311334714,
"grad_norm": 6.03125,
"learning_rate": 1.9912548797826402e-05,
"loss": 1.6859,
"step": 582
},
{
"epoch": 0.6901450133175495,
"grad_norm": 6.84375,
"learning_rate": 1.9911453862855362e-05,
"loss": 1.2656,
"step": 583
},
{
"epoch": 0.6913287955016277,
"grad_norm": 5.1875,
"learning_rate": 1.99103521463625e-05,
"loss": 1.2217,
"step": 584
},
{
"epoch": 0.6925125776857058,
"grad_norm": 5.5625,
"learning_rate": 1.9909243649101625e-05,
"loss": 1.6654,
"step": 585
},
{
"epoch": 0.693696359869784,
"grad_norm": 4.9375,
"learning_rate": 1.9908128371831177e-05,
"loss": 1.1995,
"step": 586
},
{
"epoch": 0.6948801420538621,
"grad_norm": 5.90625,
"learning_rate": 1.990700631531425e-05,
"loss": 1.3534,
"step": 587
},
{
"epoch": 0.6960639242379402,
"grad_norm": 5.78125,
"learning_rate": 1.9905877480318556e-05,
"loss": 1.4224,
"step": 588
},
{
"epoch": 0.6972477064220184,
"grad_norm": 5.53125,
"learning_rate": 1.9904741867616463e-05,
"loss": 1.505,
"step": 589
},
{
"epoch": 0.6984314886060965,
"grad_norm": 6.25,
"learning_rate": 1.990359947798497e-05,
"loss": 1.7375,
"step": 590
},
{
"epoch": 0.6996152707901746,
"grad_norm": 5.125,
"learning_rate": 1.9902450312205707e-05,
"loss": 1.0517,
"step": 591
},
{
"epoch": 0.7007990529742527,
"grad_norm": 6.6875,
"learning_rate": 1.9901294371064947e-05,
"loss": 1.6755,
"step": 592
},
{
"epoch": 0.7019828351583308,
"grad_norm": 6.125,
"learning_rate": 1.9900131655353597e-05,
"loss": 1.5789,
"step": 593
},
{
"epoch": 0.703166617342409,
"grad_norm": 5.0625,
"learning_rate": 1.9898962165867204e-05,
"loss": 0.9411,
"step": 594
},
{
"epoch": 0.7043503995264871,
"grad_norm": 6.3125,
"learning_rate": 1.989778590340594e-05,
"loss": 1.6249,
"step": 595
},
{
"epoch": 0.7055341817105653,
"grad_norm": 6.09375,
"learning_rate": 1.9896602868774617e-05,
"loss": 1.7102,
"step": 596
},
{
"epoch": 0.7067179638946434,
"grad_norm": 6.75,
"learning_rate": 1.9895413062782683e-05,
"loss": 1.7092,
"step": 597
},
{
"epoch": 0.7079017460787215,
"grad_norm": 5.09375,
"learning_rate": 1.9894216486244214e-05,
"loss": 1.1434,
"step": 598
},
{
"epoch": 0.7090855282627997,
"grad_norm": 5.09375,
"learning_rate": 1.9893013139977917e-05,
"loss": 1.0986,
"step": 599
},
{
"epoch": 0.7102693104468778,
"grad_norm": 5.90625,
"learning_rate": 1.989180302480714e-05,
"loss": 1.6445,
"step": 600
},
{
"epoch": 0.7114530926309559,
"grad_norm": 5.15625,
"learning_rate": 1.989058614155985e-05,
"loss": 1.211,
"step": 601
},
{
"epoch": 0.712636874815034,
"grad_norm": 6.28125,
"learning_rate": 1.9889362491068658e-05,
"loss": 1.6315,
"step": 602
},
{
"epoch": 0.7138206569991121,
"grad_norm": 5.71875,
"learning_rate": 1.9888132074170797e-05,
"loss": 1.5468,
"step": 603
},
{
"epoch": 0.7150044391831903,
"grad_norm": 5.28125,
"learning_rate": 1.988689489170813e-05,
"loss": 1.221,
"step": 604
},
{
"epoch": 0.7161882213672685,
"grad_norm": 5.375,
"learning_rate": 1.9885650944527146e-05,
"loss": 1.1613,
"step": 605
},
{
"epoch": 0.7173720035513466,
"grad_norm": 6.625,
"learning_rate": 1.9884400233478976e-05,
"loss": 1.8792,
"step": 606
},
{
"epoch": 0.7185557857354247,
"grad_norm": 5.8125,
"learning_rate": 1.9883142759419364e-05,
"loss": 1.5946,
"step": 607
},
{
"epoch": 0.7197395679195028,
"grad_norm": 5.0,
"learning_rate": 1.9881878523208687e-05,
"loss": 0.8917,
"step": 608
},
{
"epoch": 0.720923350103581,
"grad_norm": 6.21875,
"learning_rate": 1.9880607525711954e-05,
"loss": 1.658,
"step": 609
},
{
"epoch": 0.7221071322876591,
"grad_norm": 5.5,
"learning_rate": 1.987932976779879e-05,
"loss": 1.27,
"step": 610
},
{
"epoch": 0.7232909144717372,
"grad_norm": 6.15625,
"learning_rate": 1.9878045250343445e-05,
"loss": 2.0559,
"step": 611
},
{
"epoch": 0.7244746966558153,
"grad_norm": 6.125,
"learning_rate": 1.987675397422481e-05,
"loss": 1.3066,
"step": 612
},
{
"epoch": 0.7256584788398934,
"grad_norm": 6.0,
"learning_rate": 1.9875455940326386e-05,
"loss": 1.7674,
"step": 613
},
{
"epoch": 0.7268422610239715,
"grad_norm": 5.5,
"learning_rate": 1.9874151149536304e-05,
"loss": 1.5067,
"step": 614
},
{
"epoch": 0.7280260432080498,
"grad_norm": 7.03125,
"learning_rate": 1.9872839602747308e-05,
"loss": 1.7234,
"step": 615
},
{
"epoch": 0.7292098253921279,
"grad_norm": 5.625,
"learning_rate": 1.987152130085678e-05,
"loss": 1.3198,
"step": 616
},
{
"epoch": 0.730393607576206,
"grad_norm": 6.125,
"learning_rate": 1.9870196244766715e-05,
"loss": 1.3736,
"step": 617
},
{
"epoch": 0.7315773897602841,
"grad_norm": 5.34375,
"learning_rate": 1.9868864435383724e-05,
"loss": 1.0221,
"step": 618
},
{
"epoch": 0.7327611719443622,
"grad_norm": 5.09375,
"learning_rate": 1.9867525873619056e-05,
"loss": 1.2022,
"step": 619
},
{
"epoch": 0.7339449541284404,
"grad_norm": 6.03125,
"learning_rate": 1.986618056038856e-05,
"loss": 1.4624,
"step": 620
},
{
"epoch": 0.7351287363125185,
"grad_norm": 5.5625,
"learning_rate": 1.9864828496612717e-05,
"loss": 1.2138,
"step": 621
},
{
"epoch": 0.7363125184965966,
"grad_norm": 6.0,
"learning_rate": 1.9863469683216622e-05,
"loss": 1.5946,
"step": 622
},
{
"epoch": 0.7374963006806747,
"grad_norm": 6.15625,
"learning_rate": 1.9862104121129992e-05,
"loss": 1.4994,
"step": 623
},
{
"epoch": 0.7386800828647528,
"grad_norm": 5.875,
"learning_rate": 1.9860731811287157e-05,
"loss": 1.5608,
"step": 624
},
{
"epoch": 0.7398638650488311,
"grad_norm": 6.59375,
"learning_rate": 1.9859352754627067e-05,
"loss": 1.7837,
"step": 625
},
{
"epoch": 0.7410476472329092,
"grad_norm": 6.59375,
"learning_rate": 1.9857966952093285e-05,
"loss": 1.7659,
"step": 626
},
{
"epoch": 0.7422314294169873,
"grad_norm": 6.09375,
"learning_rate": 1.9856574404633996e-05,
"loss": 1.457,
"step": 627
},
{
"epoch": 0.7434152116010654,
"grad_norm": 5.875,
"learning_rate": 1.9855175113201994e-05,
"loss": 1.6159,
"step": 628
},
{
"epoch": 0.7445989937851435,
"grad_norm": 5.65625,
"learning_rate": 1.9853769078754685e-05,
"loss": 1.1179,
"step": 629
},
{
"epoch": 0.7457827759692217,
"grad_norm": 5.9375,
"learning_rate": 1.9852356302254098e-05,
"loss": 1.764,
"step": 630
},
{
"epoch": 0.7469665581532998,
"grad_norm": 4.9375,
"learning_rate": 1.9850936784666865e-05,
"loss": 0.863,
"step": 631
},
{
"epoch": 0.7481503403373779,
"grad_norm": 4.40625,
"learning_rate": 1.9849510526964243e-05,
"loss": 0.8325,
"step": 632
},
{
"epoch": 0.749334122521456,
"grad_norm": 6.125,
"learning_rate": 1.9848077530122083e-05,
"loss": 1.8687,
"step": 633
},
{
"epoch": 0.7505179047055341,
"grad_norm": 6.09375,
"learning_rate": 1.984663779512086e-05,
"loss": 1.6104,
"step": 634
},
{
"epoch": 0.7517016868896124,
"grad_norm": 5.78125,
"learning_rate": 1.984519132294566e-05,
"loss": 1.2897,
"step": 635
},
{
"epoch": 0.7528854690736905,
"grad_norm": 6.40625,
"learning_rate": 1.984373811458617e-05,
"loss": 1.1805,
"step": 636
},
{
"epoch": 0.7540692512577686,
"grad_norm": 6.0625,
"learning_rate": 1.9842278171036692e-05,
"loss": 1.2362,
"step": 637
},
{
"epoch": 0.7552530334418467,
"grad_norm": 6.75,
"learning_rate": 1.9840811493296134e-05,
"loss": 1.6137,
"step": 638
},
{
"epoch": 0.7564368156259248,
"grad_norm": 6.1875,
"learning_rate": 1.9839338082368017e-05,
"loss": 1.5294,
"step": 639
},
{
"epoch": 0.7576205978100029,
"grad_norm": 5.78125,
"learning_rate": 1.9837857939260457e-05,
"loss": 1.3244,
"step": 640
},
{
"epoch": 0.7588043799940811,
"grad_norm": 6.4375,
"learning_rate": 1.983637106498619e-05,
"loss": 1.7651,
"step": 641
},
{
"epoch": 0.7599881621781592,
"grad_norm": 5.25,
"learning_rate": 1.9834877460562546e-05,
"loss": 1.3837,
"step": 642
},
{
"epoch": 0.7611719443622373,
"grad_norm": 6.75,
"learning_rate": 1.9833377127011468e-05,
"loss": 1.6387,
"step": 643
},
{
"epoch": 0.7623557265463154,
"grad_norm": 5.4375,
"learning_rate": 1.98318700653595e-05,
"loss": 1.1784,
"step": 644
},
{
"epoch": 0.7635395087303936,
"grad_norm": 5.6875,
"learning_rate": 1.9830356276637788e-05,
"loss": 1.5128,
"step": 645
},
{
"epoch": 0.7647232909144718,
"grad_norm": 6.375,
"learning_rate": 1.9828835761882085e-05,
"loss": 1.9819,
"step": 646
},
{
"epoch": 0.7659070730985499,
"grad_norm": 4.28125,
"learning_rate": 1.982730852213274e-05,
"loss": 0.9206,
"step": 647
},
{
"epoch": 0.767090855282628,
"grad_norm": 5.78125,
"learning_rate": 1.9825774558434712e-05,
"loss": 1.6262,
"step": 648
},
{
"epoch": 0.7682746374667061,
"grad_norm": 5.96875,
"learning_rate": 1.9824233871837547e-05,
"loss": 1.7208,
"step": 649
},
{
"epoch": 0.7694584196507842,
"grad_norm": 5.8125,
"learning_rate": 1.9822686463395406e-05,
"loss": 1.5826,
"step": 650
},
{
"epoch": 0.7706422018348624,
"grad_norm": 4.375,
"learning_rate": 1.982113233416704e-05,
"loss": 0.8769,
"step": 651
},
{
"epoch": 0.7718259840189405,
"grad_norm": 6.59375,
"learning_rate": 1.98195714852158e-05,
"loss": 1.7527,
"step": 652
},
{
"epoch": 0.7730097662030186,
"grad_norm": 5.59375,
"learning_rate": 1.9818003917609637e-05,
"loss": 1.2368,
"step": 653
},
{
"epoch": 0.7741935483870968,
"grad_norm": 5.46875,
"learning_rate": 1.9816429632421095e-05,
"loss": 1.1347,
"step": 654
},
{
"epoch": 0.7753773305711749,
"grad_norm": 5.6875,
"learning_rate": 1.9814848630727323e-05,
"loss": 1.3182,
"step": 655
},
{
"epoch": 0.7765611127552531,
"grad_norm": 5.875,
"learning_rate": 1.9813260913610048e-05,
"loss": 1.0803,
"step": 656
},
{
"epoch": 0.7777448949393312,
"grad_norm": 4.625,
"learning_rate": 1.9811666482155608e-05,
"loss": 0.9864,
"step": 657
},
{
"epoch": 0.7789286771234093,
"grad_norm": 5.125,
"learning_rate": 1.9810065337454934e-05,
"loss": 1.6491,
"step": 658
},
{
"epoch": 0.7801124593074874,
"grad_norm": 5.4375,
"learning_rate": 1.9808457480603547e-05,
"loss": 1.2711,
"step": 659
},
{
"epoch": 0.7812962414915655,
"grad_norm": 6.0,
"learning_rate": 1.980684291270155e-05,
"loss": 1.7972,
"step": 660
},
{
"epoch": 0.7824800236756437,
"grad_norm": 5.5625,
"learning_rate": 1.9805221634853656e-05,
"loss": 1.3169,
"step": 661
},
{
"epoch": 0.7836638058597218,
"grad_norm": 5.78125,
"learning_rate": 1.980359364816916e-05,
"loss": 1.4975,
"step": 662
},
{
"epoch": 0.7848475880438,
"grad_norm": 5.0,
"learning_rate": 1.980195895376194e-05,
"loss": 0.9862,
"step": 663
},
{
"epoch": 0.7860313702278781,
"grad_norm": 4.78125,
"learning_rate": 1.980031755275048e-05,
"loss": 0.9706,
"step": 664
},
{
"epoch": 0.7872151524119562,
"grad_norm": 5.75,
"learning_rate": 1.9798669446257844e-05,
"loss": 1.5911,
"step": 665
},
{
"epoch": 0.7883989345960344,
"grad_norm": 5.75,
"learning_rate": 1.9797014635411674e-05,
"loss": 1.5144,
"step": 666
},
{
"epoch": 0.7895827167801125,
"grad_norm": 6.34375,
"learning_rate": 1.979535312134422e-05,
"loss": 1.7656,
"step": 667
},
{
"epoch": 0.7907664989641906,
"grad_norm": 7.03125,
"learning_rate": 1.9793684905192302e-05,
"loss": 2.2322,
"step": 668
},
{
"epoch": 0.7919502811482687,
"grad_norm": 6.125,
"learning_rate": 1.9792009988097334e-05,
"loss": 1.5118,
"step": 669
},
{
"epoch": 0.7931340633323468,
"grad_norm": 7.0625,
"learning_rate": 1.979032837120531e-05,
"loss": 2.2054,
"step": 670
},
{
"epoch": 0.7943178455164249,
"grad_norm": 5.28125,
"learning_rate": 1.978864005566681e-05,
"loss": 1.543,
"step": 671
},
{
"epoch": 0.7955016277005031,
"grad_norm": 5.4375,
"learning_rate": 1.9786945042636998e-05,
"loss": 1.5768,
"step": 672
},
{
"epoch": 0.7966854098845813,
"grad_norm": 5.125,
"learning_rate": 1.9785243333275622e-05,
"loss": 1.4974,
"step": 673
},
{
"epoch": 0.7978691920686594,
"grad_norm": 5.40625,
"learning_rate": 1.9783534928747006e-05,
"loss": 1.1993,
"step": 674
},
{
"epoch": 0.7990529742527375,
"grad_norm": 6.03125,
"learning_rate": 1.9781819830220058e-05,
"loss": 2.1668,
"step": 675
},
{
"epoch": 0.8002367564368156,
"grad_norm": 6.65625,
"learning_rate": 1.978009803886827e-05,
"loss": 1.6233,
"step": 676
},
{
"epoch": 0.8014205386208938,
"grad_norm": 5.71875,
"learning_rate": 1.9778369555869714e-05,
"loss": 1.457,
"step": 677
},
{
"epoch": 0.8026043208049719,
"grad_norm": 6.59375,
"learning_rate": 1.9776634382407027e-05,
"loss": 1.3318,
"step": 678
},
{
"epoch": 0.80378810298905,
"grad_norm": 4.90625,
"learning_rate": 1.9774892519667433e-05,
"loss": 1.5303,
"step": 679
},
{
"epoch": 0.8049718851731281,
"grad_norm": 5.15625,
"learning_rate": 1.9773143968842743e-05,
"loss": 1.4359,
"step": 680
},
{
"epoch": 0.8061556673572062,
"grad_norm": 4.3125,
"learning_rate": 1.9771388731129325e-05,
"loss": 1.0214,
"step": 681
},
{
"epoch": 0.8073394495412844,
"grad_norm": 4.96875,
"learning_rate": 1.9769626807728134e-05,
"loss": 1.2042,
"step": 682
},
{
"epoch": 0.8085232317253626,
"grad_norm": 4.71875,
"learning_rate": 1.9767858199844697e-05,
"loss": 0.9863,
"step": 683
},
{
"epoch": 0.8097070139094407,
"grad_norm": 5.40625,
"learning_rate": 1.9766082908689117e-05,
"loss": 1.2145,
"step": 684
},
{
"epoch": 0.8108907960935188,
"grad_norm": 5.21875,
"learning_rate": 1.9764300935476065e-05,
"loss": 1.2448,
"step": 685
},
{
"epoch": 0.8120745782775969,
"grad_norm": 5.46875,
"learning_rate": 1.9762512281424778e-05,
"loss": 1.207,
"step": 686
},
{
"epoch": 0.8132583604616751,
"grad_norm": 5.21875,
"learning_rate": 1.9760716947759086e-05,
"loss": 1.3856,
"step": 687
},
{
"epoch": 0.8144421426457532,
"grad_norm": 6.53125,
"learning_rate": 1.9758914935707366e-05,
"loss": 1.7468,
"step": 688
},
{
"epoch": 0.8156259248298313,
"grad_norm": 6.09375,
"learning_rate": 1.9757106246502577e-05,
"loss": 1.4507,
"step": 689
},
{
"epoch": 0.8168097070139094,
"grad_norm": 5.46875,
"learning_rate": 1.9755290881382244e-05,
"loss": 1.4116,
"step": 690
},
{
"epoch": 0.8179934891979875,
"grad_norm": 6.8125,
"learning_rate": 1.9753468841588456e-05,
"loss": 1.395,
"step": 691
},
{
"epoch": 0.8191772713820658,
"grad_norm": 5.9375,
"learning_rate": 1.9751640128367872e-05,
"loss": 1.5778,
"step": 692
},
{
"epoch": 0.8203610535661439,
"grad_norm": 6.5625,
"learning_rate": 1.974980474297172e-05,
"loss": 1.757,
"step": 693
},
{
"epoch": 0.821544835750222,
"grad_norm": 5.375,
"learning_rate": 1.974796268665579e-05,
"loss": 1.3214,
"step": 694
},
{
"epoch": 0.8227286179343001,
"grad_norm": 5.5,
"learning_rate": 1.9746113960680435e-05,
"loss": 1.2076,
"step": 695
},
{
"epoch": 0.8239124001183782,
"grad_norm": 5.625,
"learning_rate": 1.9744258566310574e-05,
"loss": 1.8772,
"step": 696
},
{
"epoch": 0.8250961823024563,
"grad_norm": 5.46875,
"learning_rate": 1.9742396504815687e-05,
"loss": 1.1263,
"step": 697
},
{
"epoch": 0.8262799644865345,
"grad_norm": 5.15625,
"learning_rate": 1.974052777746981e-05,
"loss": 1.6026,
"step": 698
},
{
"epoch": 0.8274637466706126,
"grad_norm": 5.59375,
"learning_rate": 1.973865238555156e-05,
"loss": 1.4331,
"step": 699
},
{
"epoch": 0.8286475288546907,
"grad_norm": 4.8125,
"learning_rate": 1.9736770330344084e-05,
"loss": 1.344,
"step": 700
},
{
"epoch": 0.8298313110387688,
"grad_norm": 6.5625,
"learning_rate": 1.973488161313512e-05,
"loss": 1.5337,
"step": 701
},
{
"epoch": 0.831015093222847,
"grad_norm": 4.84375,
"learning_rate": 1.9732986235216937e-05,
"loss": 1.2712,
"step": 702
},
{
"epoch": 0.8321988754069252,
"grad_norm": 6.5625,
"learning_rate": 1.9731084197886376e-05,
"loss": 1.8053,
"step": 703
},
{
"epoch": 0.8333826575910033,
"grad_norm": 5.0,
"learning_rate": 1.9729175502444832e-05,
"loss": 1.2971,
"step": 704
},
{
"epoch": 0.8345664397750814,
"grad_norm": 4.375,
"learning_rate": 1.9727260150198254e-05,
"loss": 0.8897,
"step": 705
},
{
"epoch": 0.8357502219591595,
"grad_norm": 6.6875,
"learning_rate": 1.9725338142457145e-05,
"loss": 1.8121,
"step": 706
},
{
"epoch": 0.8369340041432376,
"grad_norm": 7.625,
"learning_rate": 1.9723409480536566e-05,
"loss": 1.3363,
"step": 707
},
{
"epoch": 0.8381177863273158,
"grad_norm": 4.84375,
"learning_rate": 1.9721474165756124e-05,
"loss": 1.5734,
"step": 708
},
{
"epoch": 0.8393015685113939,
"grad_norm": 6.5,
"learning_rate": 1.9719532199439986e-05,
"loss": 1.5202,
"step": 709
},
{
"epoch": 0.840485350695472,
"grad_norm": 4.90625,
"learning_rate": 1.9717583582916862e-05,
"loss": 0.9867,
"step": 710
},
{
"epoch": 0.8416691328795501,
"grad_norm": 4.8125,
"learning_rate": 1.9715628317520022e-05,
"loss": 0.9933,
"step": 711
},
{
"epoch": 0.8428529150636282,
"grad_norm": 5.375,
"learning_rate": 1.9713666404587274e-05,
"loss": 1.3409,
"step": 712
},
{
"epoch": 0.8440366972477065,
"grad_norm": 5.0625,
"learning_rate": 1.9711697845460978e-05,
"loss": 1.414,
"step": 713
},
{
"epoch": 0.8452204794317846,
"grad_norm": 5.4375,
"learning_rate": 1.970972264148805e-05,
"loss": 1.6005,
"step": 714
},
{
"epoch": 0.8464042616158627,
"grad_norm": 5.9375,
"learning_rate": 1.9707740794019937e-05,
"loss": 1.5261,
"step": 715
},
{
"epoch": 0.8475880437999408,
"grad_norm": 6.0,
"learning_rate": 1.970575230441265e-05,
"loss": 1.3009,
"step": 716
},
{
"epoch": 0.8487718259840189,
"grad_norm": 6.15625,
"learning_rate": 1.9703757174026726e-05,
"loss": 1.6992,
"step": 717
},
{
"epoch": 0.8499556081680971,
"grad_norm": 5.71875,
"learning_rate": 1.970175540422726e-05,
"loss": 1.2445,
"step": 718
},
{
"epoch": 0.8511393903521752,
"grad_norm": 4.375,
"learning_rate": 1.969974699638388e-05,
"loss": 0.9533,
"step": 719
},
{
"epoch": 0.8523231725362533,
"grad_norm": 4.9375,
"learning_rate": 1.969773195187076e-05,
"loss": 1.2177,
"step": 720
},
{
"epoch": 0.8535069547203314,
"grad_norm": 5.09375,
"learning_rate": 1.9695710272066623e-05,
"loss": 1.4562,
"step": 721
},
{
"epoch": 0.8546907369044096,
"grad_norm": 6.4375,
"learning_rate": 1.969368195835471e-05,
"loss": 1.8507,
"step": 722
},
{
"epoch": 0.8558745190884878,
"grad_norm": 6.34375,
"learning_rate": 1.9691647012122825e-05,
"loss": 1.6108,
"step": 723
},
{
"epoch": 0.8570583012725659,
"grad_norm": 5.34375,
"learning_rate": 1.9689605434763298e-05,
"loss": 1.3795,
"step": 724
},
{
"epoch": 0.858242083456644,
"grad_norm": 5.0,
"learning_rate": 1.9687557227672997e-05,
"loss": 1.2847,
"step": 725
},
{
"epoch": 0.8594258656407221,
"grad_norm": 4.125,
"learning_rate": 1.9685502392253326e-05,
"loss": 0.978,
"step": 726
},
{
"epoch": 0.8606096478248002,
"grad_norm": 5.34375,
"learning_rate": 1.9683440929910223e-05,
"loss": 1.0839,
"step": 727
},
{
"epoch": 0.8617934300088783,
"grad_norm": 6.75,
"learning_rate": 1.968137284205417e-05,
"loss": 1.7532,
"step": 728
},
{
"epoch": 0.8629772121929565,
"grad_norm": 5.375,
"learning_rate": 1.9679298130100165e-05,
"loss": 1.2769,
"step": 729
},
{
"epoch": 0.8641609943770346,
"grad_norm": 4.5,
"learning_rate": 1.967721679546776e-05,
"loss": 1.0752,
"step": 730
},
{
"epoch": 0.8653447765611127,
"grad_norm": 7.46875,
"learning_rate": 1.9675128839581013e-05,
"loss": 1.6552,
"step": 731
},
{
"epoch": 0.8665285587451909,
"grad_norm": 4.75,
"learning_rate": 1.9673034263868533e-05,
"loss": 1.0978,
"step": 732
},
{
"epoch": 0.867712340929269,
"grad_norm": 5.96875,
"learning_rate": 1.967093306976345e-05,
"loss": 1.2744,
"step": 733
},
{
"epoch": 0.8688961231133472,
"grad_norm": 5.46875,
"learning_rate": 1.9668825258703424e-05,
"loss": 1.4219,
"step": 734
},
{
"epoch": 0.8700799052974253,
"grad_norm": 5.5,
"learning_rate": 1.966671083213064e-05,
"loss": 1.5471,
"step": 735
},
{
"epoch": 0.8712636874815034,
"grad_norm": 5.59375,
"learning_rate": 1.9664589791491812e-05,
"loss": 1.4807,
"step": 736
},
{
"epoch": 0.8724474696655815,
"grad_norm": 5.15625,
"learning_rate": 1.966246213823818e-05,
"loss": 1.1011,
"step": 737
},
{
"epoch": 0.8736312518496596,
"grad_norm": 5.28125,
"learning_rate": 1.9660327873825503e-05,
"loss": 1.2984,
"step": 738
},
{
"epoch": 0.8748150340337378,
"grad_norm": 5.625,
"learning_rate": 1.9658186999714067e-05,
"loss": 1.1409,
"step": 739
},
{
"epoch": 0.875998816217816,
"grad_norm": 4.5,
"learning_rate": 1.9656039517368686e-05,
"loss": 0.9786,
"step": 740
},
{
"epoch": 0.877182598401894,
"grad_norm": 5.75,
"learning_rate": 1.9653885428258686e-05,
"loss": 1.4138,
"step": 741
},
{
"epoch": 0.8783663805859722,
"grad_norm": 5.5,
"learning_rate": 1.965172473385792e-05,
"loss": 1.1823,
"step": 742
},
{
"epoch": 0.8795501627700503,
"grad_norm": 6.0,
"learning_rate": 1.9649557435644757e-05,
"loss": 2.0114,
"step": 743
},
{
"epoch": 0.8807339449541285,
"grad_norm": 5.875,
"learning_rate": 1.9647383535102085e-05,
"loss": 1.3305,
"step": 744
},
{
"epoch": 0.8819177271382066,
"grad_norm": 5.5625,
"learning_rate": 1.964520303371731e-05,
"loss": 1.4193,
"step": 745
},
{
"epoch": 0.8831015093222847,
"grad_norm": 5.5,
"learning_rate": 1.9643015932982355e-05,
"loss": 0.9923,
"step": 746
},
{
"epoch": 0.8842852915063628,
"grad_norm": 5.25,
"learning_rate": 1.9640822234393654e-05,
"loss": 1.3816,
"step": 747
},
{
"epoch": 0.8854690736904409,
"grad_norm": 5.78125,
"learning_rate": 1.9638621939452166e-05,
"loss": 1.6367,
"step": 748
},
{
"epoch": 0.8866528558745191,
"grad_norm": 6.0,
"learning_rate": 1.963641504966335e-05,
"loss": 1.7965,
"step": 749
},
{
"epoch": 0.8878366380585972,
"grad_norm": 5.0625,
"learning_rate": 1.9634201566537182e-05,
"loss": 1.3075,
"step": 750
},
{
"epoch": 0.8890204202426754,
"grad_norm": 6.84375,
"learning_rate": 1.963198149158816e-05,
"loss": 1.7589,
"step": 751
},
{
"epoch": 0.8902042024267535,
"grad_norm": 5.09375,
"learning_rate": 1.962975482633527e-05,
"loss": 0.9977,
"step": 752
},
{
"epoch": 0.8913879846108316,
"grad_norm": 6.40625,
"learning_rate": 1.9627521572302035e-05,
"loss": 1.5262,
"step": 753
},
{
"epoch": 0.8925717667949097,
"grad_norm": 6.03125,
"learning_rate": 1.9625281731016455e-05,
"loss": 1.513,
"step": 754
},
{
"epoch": 0.8937555489789879,
"grad_norm": 5.40625,
"learning_rate": 1.9623035304011062e-05,
"loss": 1.2644,
"step": 755
},
{
"epoch": 0.894939331163066,
"grad_norm": 6.375,
"learning_rate": 1.962078229282289e-05,
"loss": 1.7802,
"step": 756
},
{
"epoch": 0.8961231133471441,
"grad_norm": 6.59375,
"learning_rate": 1.961852269899346e-05,
"loss": 1.808,
"step": 757
},
{
"epoch": 0.8973068955312222,
"grad_norm": 5.59375,
"learning_rate": 1.9616256524068822e-05,
"loss": 1.4188,
"step": 758
},
{
"epoch": 0.8984906777153003,
"grad_norm": 5.1875,
"learning_rate": 1.961398376959951e-05,
"loss": 1.3978,
"step": 759
},
{
"epoch": 0.8996744598993786,
"grad_norm": 4.65625,
"learning_rate": 1.961170443714057e-05,
"loss": 0.9928,
"step": 760
},
{
"epoch": 0.9008582420834567,
"grad_norm": 5.9375,
"learning_rate": 1.9609418528251543e-05,
"loss": 1.6512,
"step": 761
},
{
"epoch": 0.9020420242675348,
"grad_norm": 6.15625,
"learning_rate": 1.9607126044496476e-05,
"loss": 1.8008,
"step": 762
},
{
"epoch": 0.9032258064516129,
"grad_norm": 5.5625,
"learning_rate": 1.9604826987443905e-05,
"loss": 1.1789,
"step": 763
},
{
"epoch": 0.904409588635691,
"grad_norm": 5.09375,
"learning_rate": 1.960252135866687e-05,
"loss": 1.3402,
"step": 764
},
{
"epoch": 0.9055933708197692,
"grad_norm": 5.9375,
"learning_rate": 1.9600209159742913e-05,
"loss": 1.3972,
"step": 765
},
{
"epoch": 0.9067771530038473,
"grad_norm": 5.625,
"learning_rate": 1.959789039225406e-05,
"loss": 1.4459,
"step": 766
},
{
"epoch": 0.9079609351879254,
"grad_norm": 5.84375,
"learning_rate": 1.9595565057786838e-05,
"loss": 1.6303,
"step": 767
},
{
"epoch": 0.9091447173720035,
"grad_norm": 5.46875,
"learning_rate": 1.9593233157932265e-05,
"loss": 1.2033,
"step": 768
},
{
"epoch": 0.9103284995560816,
"grad_norm": 4.6875,
"learning_rate": 1.9590894694285852e-05,
"loss": 0.9957,
"step": 769
},
{
"epoch": 0.9115122817401599,
"grad_norm": 5.4375,
"learning_rate": 1.9588549668447595e-05,
"loss": 1.4582,
"step": 770
},
{
"epoch": 0.912696063924238,
"grad_norm": 5.53125,
"learning_rate": 1.9586198082021995e-05,
"loss": 1.5325,
"step": 771
},
{
"epoch": 0.9138798461083161,
"grad_norm": 6.0,
"learning_rate": 1.958383993661803e-05,
"loss": 1.1531,
"step": 772
},
{
"epoch": 0.9150636282923942,
"grad_norm": 5.65625,
"learning_rate": 1.9581475233849165e-05,
"loss": 1.8776,
"step": 773
},
{
"epoch": 0.9162474104764723,
"grad_norm": 5.59375,
"learning_rate": 1.957910397533335e-05,
"loss": 1.5848,
"step": 774
},
{
"epoch": 0.9174311926605505,
"grad_norm": 5.125,
"learning_rate": 1.9576726162693036e-05,
"loss": 1.5302,
"step": 775
},
{
"epoch": 0.9186149748446286,
"grad_norm": 5.4375,
"learning_rate": 1.9574341797555143e-05,
"loss": 1.5221,
"step": 776
},
{
"epoch": 0.9197987570287067,
"grad_norm": 5.25,
"learning_rate": 1.9571950881551073e-05,
"loss": 1.3238,
"step": 777
},
{
"epoch": 0.9209825392127848,
"grad_norm": 6.375,
"learning_rate": 1.9569553416316725e-05,
"loss": 1.6467,
"step": 778
},
{
"epoch": 0.9221663213968629,
"grad_norm": 6.03125,
"learning_rate": 1.9567149403492467e-05,
"loss": 1.8354,
"step": 779
},
{
"epoch": 0.9233501035809412,
"grad_norm": 6.5,
"learning_rate": 1.9564738844723153e-05,
"loss": 1.7425,
"step": 780
},
{
"epoch": 0.9245338857650193,
"grad_norm": 4.5625,
"learning_rate": 1.9562321741658106e-05,
"loss": 1.1226,
"step": 781
},
{
"epoch": 0.9257176679490974,
"grad_norm": 5.15625,
"learning_rate": 1.9559898095951137e-05,
"loss": 1.2621,
"step": 782
},
{
"epoch": 0.9269014501331755,
"grad_norm": 5.75,
"learning_rate": 1.955746790926053e-05,
"loss": 1.2224,
"step": 783
},
{
"epoch": 0.9280852323172536,
"grad_norm": 5.28125,
"learning_rate": 1.9555031183249046e-05,
"loss": 1.5868,
"step": 784
},
{
"epoch": 0.9292690145013317,
"grad_norm": 5.75,
"learning_rate": 1.9552587919583915e-05,
"loss": 1.7259,
"step": 785
},
{
"epoch": 0.9304527966854099,
"grad_norm": 5.71875,
"learning_rate": 1.955013811993685e-05,
"loss": 1.4874,
"step": 786
},
{
"epoch": 0.931636578869488,
"grad_norm": 5.6875,
"learning_rate": 1.9547681785984024e-05,
"loss": 1.3721,
"step": 787
},
{
"epoch": 0.9328203610535661,
"grad_norm": 5.65625,
"learning_rate": 1.954521891940609e-05,
"loss": 1.5377,
"step": 788
},
{
"epoch": 0.9340041432376442,
"grad_norm": 5.90625,
"learning_rate": 1.9542749521888165e-05,
"loss": 1.4326,
"step": 789
},
{
"epoch": 0.9351879254217224,
"grad_norm": 5.5,
"learning_rate": 1.9540273595119842e-05,
"loss": 1.3647,
"step": 790
},
{
"epoch": 0.9363717076058006,
"grad_norm": 5.75,
"learning_rate": 1.953779114079517e-05,
"loss": 1.2791,
"step": 791
},
{
"epoch": 0.9375554897898787,
"grad_norm": 4.84375,
"learning_rate": 1.9535302160612674e-05,
"loss": 1.0689,
"step": 792
},
{
"epoch": 0.9387392719739568,
"grad_norm": 5.53125,
"learning_rate": 1.953280665627534e-05,
"loss": 1.0936,
"step": 793
},
{
"epoch": 0.9399230541580349,
"grad_norm": 5.59375,
"learning_rate": 1.953030462949062e-05,
"loss": 1.5988,
"step": 794
},
{
"epoch": 0.941106836342113,
"grad_norm": 6.0,
"learning_rate": 1.9527796081970423e-05,
"loss": 1.5384,
"step": 795
},
{
"epoch": 0.9422906185261912,
"grad_norm": 5.59375,
"learning_rate": 1.9525281015431127e-05,
"loss": 1.4638,
"step": 796
},
{
"epoch": 0.9434744007102693,
"grad_norm": 4.84375,
"learning_rate": 1.9522759431593566e-05,
"loss": 1.0825,
"step": 797
},
{
"epoch": 0.9446581828943474,
"grad_norm": 4.875,
"learning_rate": 1.9520231332183037e-05,
"loss": 0.9433,
"step": 798
},
{
"epoch": 0.9458419650784256,
"grad_norm": 5.65625,
"learning_rate": 1.951769671892929e-05,
"loss": 1.225,
"step": 799
},
{
"epoch": 0.9470257472625037,
"grad_norm": 5.78125,
"learning_rate": 1.9515155593566536e-05,
"loss": 1.7972,
"step": 800
},
{
"epoch": 0.9482095294465819,
"grad_norm": 5.375,
"learning_rate": 1.9512607957833438e-05,
"loss": 1.6504,
"step": 801
},
{
"epoch": 0.94939331163066,
"grad_norm": 5.0,
"learning_rate": 1.9510053813473115e-05,
"loss": 1.197,
"step": 802
},
{
"epoch": 0.9505770938147381,
"grad_norm": 4.84375,
"learning_rate": 1.950749316223314e-05,
"loss": 1.2127,
"step": 803
},
{
"epoch": 0.9517608759988162,
"grad_norm": 5.46875,
"learning_rate": 1.9504926005865542e-05,
"loss": 0.9847,
"step": 804
},
{
"epoch": 0.9529446581828943,
"grad_norm": 5.625,
"learning_rate": 1.950235234612679e-05,
"loss": 1.2296,
"step": 805
},
{
"epoch": 0.9541284403669725,
"grad_norm": 4.4375,
"learning_rate": 1.9499772184777814e-05,
"loss": 0.8717,
"step": 806
},
{
"epoch": 0.9553122225510506,
"grad_norm": 5.75,
"learning_rate": 1.9497185523583986e-05,
"loss": 1.4412,
"step": 807
},
{
"epoch": 0.9564960047351287,
"grad_norm": 5.9375,
"learning_rate": 1.9494592364315125e-05,
"loss": 1.2627,
"step": 808
},
{
"epoch": 0.9576797869192069,
"grad_norm": 5.90625,
"learning_rate": 1.9491992708745502e-05,
"loss": 1.7195,
"step": 809
},
{
"epoch": 0.958863569103285,
"grad_norm": 6.0625,
"learning_rate": 1.948938655865383e-05,
"loss": 1.4862,
"step": 810
},
{
"epoch": 0.9600473512873631,
"grad_norm": 6.0625,
"learning_rate": 1.9486773915823258e-05,
"loss": 1.2426,
"step": 811
},
{
"epoch": 0.9612311334714413,
"grad_norm": 5.5625,
"learning_rate": 1.948415478204139e-05,
"loss": 1.4877,
"step": 812
},
{
"epoch": 0.9624149156555194,
"grad_norm": 5.65625,
"learning_rate": 1.9481529159100258e-05,
"loss": 1.4973,
"step": 813
},
{
"epoch": 0.9635986978395975,
"grad_norm": 5.625,
"learning_rate": 1.947889704879635e-05,
"loss": 1.6016,
"step": 814
},
{
"epoch": 0.9647824800236756,
"grad_norm": 5.90625,
"learning_rate": 1.9476258452930577e-05,
"loss": 1.573,
"step": 815
},
{
"epoch": 0.9659662622077537,
"grad_norm": 6.0,
"learning_rate": 1.94736133733083e-05,
"loss": 1.6285,
"step": 816
},
{
"epoch": 0.9671500443918319,
"grad_norm": 5.28125,
"learning_rate": 1.9470961811739304e-05,
"loss": 1.3072,
"step": 817
},
{
"epoch": 0.96833382657591,
"grad_norm": 6.1875,
"learning_rate": 1.946830377003782e-05,
"loss": 1.6751,
"step": 818
},
{
"epoch": 0.9695176087599882,
"grad_norm": 4.9375,
"learning_rate": 1.946563925002251e-05,
"loss": 1.4002,
"step": 819
},
{
"epoch": 0.9707013909440663,
"grad_norm": 6.03125,
"learning_rate": 1.9462968253516462e-05,
"loss": 1.1341,
"step": 820
},
{
"epoch": 0.9718851731281444,
"grad_norm": 5.21875,
"learning_rate": 1.94602907823472e-05,
"loss": 1.2678,
"step": 821
},
{
"epoch": 0.9730689553122226,
"grad_norm": 5.34375,
"learning_rate": 1.945760683834668e-05,
"loss": 1.7004,
"step": 822
},
{
"epoch": 0.9742527374963007,
"grad_norm": 6.0,
"learning_rate": 1.9454916423351288e-05,
"loss": 2.0058,
"step": 823
},
{
"epoch": 0.9754365196803788,
"grad_norm": 5.59375,
"learning_rate": 1.945221953920183e-05,
"loss": 1.8835,
"step": 824
},
{
"epoch": 0.9766203018644569,
"grad_norm": 5.8125,
"learning_rate": 1.9449516187743546e-05,
"loss": 1.3977,
"step": 825
},
{
"epoch": 0.977804084048535,
"grad_norm": 6.09375,
"learning_rate": 1.94468063708261e-05,
"loss": 1.4607,
"step": 826
},
{
"epoch": 0.9789878662326132,
"grad_norm": 5.875,
"learning_rate": 1.9444090090303567e-05,
"loss": 1.6656,
"step": 827
},
{
"epoch": 0.9801716484166914,
"grad_norm": 4.5625,
"learning_rate": 1.9441367348034464e-05,
"loss": 1.2226,
"step": 828
},
{
"epoch": 0.9813554306007695,
"grad_norm": 6.46875,
"learning_rate": 1.9438638145881718e-05,
"loss": 1.5736,
"step": 829
},
{
"epoch": 0.9825392127848476,
"grad_norm": 5.40625,
"learning_rate": 1.9435902485712677e-05,
"loss": 1.1888,
"step": 830
},
{
"epoch": 0.9837229949689257,
"grad_norm": 5.9375,
"learning_rate": 1.9433160369399108e-05,
"loss": 1.5072,
"step": 831
},
{
"epoch": 0.9849067771530039,
"grad_norm": 5.125,
"learning_rate": 1.94304117988172e-05,
"loss": 1.1913,
"step": 832
},
{
"epoch": 0.986090559337082,
"grad_norm": 5.71875,
"learning_rate": 1.9427656775847553e-05,
"loss": 1.4369,
"step": 833
},
{
"epoch": 0.9872743415211601,
"grad_norm": 5.53125,
"learning_rate": 1.9424895302375177e-05,
"loss": 1.3249,
"step": 834
},
{
"epoch": 0.9884581237052382,
"grad_norm": 5.25,
"learning_rate": 1.9422127380289508e-05,
"loss": 1.3002,
"step": 835
},
{
"epoch": 0.9896419058893163,
"grad_norm": 4.8125,
"learning_rate": 1.941935301148439e-05,
"loss": 1.1471,
"step": 836
},
{
"epoch": 0.9908256880733946,
"grad_norm": 6.4375,
"learning_rate": 1.9416572197858063e-05,
"loss": 1.4162,
"step": 837
},
{
"epoch": 0.9920094702574727,
"grad_norm": 5.15625,
"learning_rate": 1.9413784941313203e-05,
"loss": 1.081,
"step": 838
},
{
"epoch": 0.9931932524415508,
"grad_norm": 6.1875,
"learning_rate": 1.941099124375687e-05,
"loss": 1.3035,
"step": 839
},
{
"epoch": 0.9943770346256289,
"grad_norm": 5.46875,
"learning_rate": 1.9408191107100554e-05,
"loss": 1.2108,
"step": 840
},
{
"epoch": 0.995560816809707,
"grad_norm": 4.78125,
"learning_rate": 1.9405384533260124e-05,
"loss": 1.291,
"step": 841
},
{
"epoch": 0.9967445989937851,
"grad_norm": 5.65625,
"learning_rate": 1.9402571524155877e-05,
"loss": 1.3387,
"step": 842
},
{
"epoch": 0.9979283811778633,
"grad_norm": 7.53125,
"learning_rate": 1.9399752081712498e-05,
"loss": 1.6406,
"step": 843
},
{
"epoch": 0.9991121633619414,
"grad_norm": 5.625,
"learning_rate": 1.9396926207859085e-05,
"loss": 1.4207,
"step": 844
},
{
"epoch": 1.0,
"grad_norm": 4.90625,
"learning_rate": 1.939409390452913e-05,
"loss": 1.0101,
"step": 845
},
{
"epoch": 1.001183782184078,
"grad_norm": 4.875,
"learning_rate": 1.9391255173660516e-05,
"loss": 1.4194,
"step": 846
},
{
"epoch": 1.0023675643681562,
"grad_norm": 5.3125,
"learning_rate": 1.9388410017195544e-05,
"loss": 1.3534,
"step": 847
},
{
"epoch": 1.0035513465522343,
"grad_norm": 5.15625,
"learning_rate": 1.9385558437080898e-05,
"loss": 1.4195,
"step": 848
},
{
"epoch": 1.0047351287363124,
"grad_norm": 5.53125,
"learning_rate": 1.9382700435267653e-05,
"loss": 1.4411,
"step": 849
},
{
"epoch": 1.0059189109203905,
"grad_norm": 5.5625,
"learning_rate": 1.937983601371129e-05,
"loss": 1.1894,
"step": 850
},
{
"epoch": 1.0071026931044689,
"grad_norm": 5.34375,
"learning_rate": 1.9376965174371678e-05,
"loss": 1.4993,
"step": 851
},
{
"epoch": 1.008286475288547,
"grad_norm": 5.25,
"learning_rate": 1.937408791921307e-05,
"loss": 1.1394,
"step": 852
},
{
"epoch": 1.009470257472625,
"grad_norm": 5.90625,
"learning_rate": 1.9371204250204116e-05,
"loss": 1.4995,
"step": 853
},
{
"epoch": 1.0106540396567032,
"grad_norm": 5.1875,
"learning_rate": 1.9368314169317858e-05,
"loss": 0.9036,
"step": 854
},
{
"epoch": 1.0118378218407813,
"grad_norm": 6.0,
"learning_rate": 1.9365417678531713e-05,
"loss": 1.1521,
"step": 855
},
{
"epoch": 1.0130216040248594,
"grad_norm": 6.375,
"learning_rate": 1.9362514779827494e-05,
"loss": 1.1867,
"step": 856
},
{
"epoch": 1.0142053862089375,
"grad_norm": 6.21875,
"learning_rate": 1.9359605475191396e-05,
"loss": 1.9155,
"step": 857
},
{
"epoch": 1.0153891683930156,
"grad_norm": 5.6875,
"learning_rate": 1.9356689766613993e-05,
"loss": 1.3164,
"step": 858
},
{
"epoch": 1.0165729505770937,
"grad_norm": 5.46875,
"learning_rate": 1.935376765609025e-05,
"loss": 1.5309,
"step": 859
},
{
"epoch": 1.0177567327611718,
"grad_norm": 6.09375,
"learning_rate": 1.9350839145619496e-05,
"loss": 1.038,
"step": 860
},
{
"epoch": 1.0189405149452502,
"grad_norm": 4.3125,
"learning_rate": 1.934790423720546e-05,
"loss": 1.1392,
"step": 861
},
{
"epoch": 1.0201242971293283,
"grad_norm": 5.125,
"learning_rate": 1.9344962932856228e-05,
"loss": 1.3084,
"step": 862
},
{
"epoch": 1.0213080793134064,
"grad_norm": 5.375,
"learning_rate": 1.9342015234584277e-05,
"loss": 1.6329,
"step": 863
},
{
"epoch": 1.0224918614974845,
"grad_norm": 6.125,
"learning_rate": 1.9339061144406455e-05,
"loss": 1.6293,
"step": 864
},
{
"epoch": 1.0236756436815626,
"grad_norm": 5.46875,
"learning_rate": 1.9336100664343976e-05,
"loss": 1.1309,
"step": 865
},
{
"epoch": 1.0248594258656407,
"grad_norm": 4.6875,
"learning_rate": 1.9333133796422437e-05,
"loss": 1.0764,
"step": 866
},
{
"epoch": 1.0260432080497188,
"grad_norm": 5.40625,
"learning_rate": 1.9330160542671794e-05,
"loss": 1.2188,
"step": 867
},
{
"epoch": 1.027226990233797,
"grad_norm": 5.0625,
"learning_rate": 1.9327180905126386e-05,
"loss": 1.1279,
"step": 868
},
{
"epoch": 1.028410772417875,
"grad_norm": 5.71875,
"learning_rate": 1.932419488582491e-05,
"loss": 1.4666,
"step": 869
},
{
"epoch": 1.0295945546019531,
"grad_norm": 5.4375,
"learning_rate": 1.9321202486810427e-05,
"loss": 1.3447,
"step": 870
},
{
"epoch": 1.0307783367860315,
"grad_norm": 6.9375,
"learning_rate": 1.931820371013038e-05,
"loss": 1.816,
"step": 871
},
{
"epoch": 1.0319621189701096,
"grad_norm": 5.125,
"learning_rate": 1.9315198557836555e-05,
"loss": 1.3515,
"step": 872
},
{
"epoch": 1.0331459011541877,
"grad_norm": 5.34375,
"learning_rate": 1.9312187031985105e-05,
"loss": 1.3136,
"step": 873
},
{
"epoch": 1.0343296833382658,
"grad_norm": 5.34375,
"learning_rate": 1.9309169134636558e-05,
"loss": 1.4253,
"step": 874
},
{
"epoch": 1.035513465522344,
"grad_norm": 5.5,
"learning_rate": 1.930614486785579e-05,
"loss": 1.1949,
"step": 875
},
{
"epoch": 1.036697247706422,
"grad_norm": 5.375,
"learning_rate": 1.930311423371203e-05,
"loss": 1.2728,
"step": 876
},
{
"epoch": 1.0378810298905001,
"grad_norm": 5.34375,
"learning_rate": 1.930007723427887e-05,
"loss": 1.2248,
"step": 877
},
{
"epoch": 1.0390648120745782,
"grad_norm": 5.5,
"learning_rate": 1.9297033871634266e-05,
"loss": 1.5535,
"step": 878
},
{
"epoch": 1.0402485942586563,
"grad_norm": 4.75,
"learning_rate": 1.929398414786051e-05,
"loss": 1.0748,
"step": 879
},
{
"epoch": 1.0414323764427345,
"grad_norm": 5.375,
"learning_rate": 1.929092806504426e-05,
"loss": 1.3883,
"step": 880
},
{
"epoch": 1.0426161586268126,
"grad_norm": 5.25,
"learning_rate": 1.928786562527652e-05,
"loss": 1.1885,
"step": 881
},
{
"epoch": 1.043799940810891,
"grad_norm": 5.84375,
"learning_rate": 1.9284796830652642e-05,
"loss": 1.6488,
"step": 882
},
{
"epoch": 1.044983722994969,
"grad_norm": 5.0,
"learning_rate": 1.928172168327233e-05,
"loss": 0.9137,
"step": 883
},
{
"epoch": 1.046167505179047,
"grad_norm": 5.84375,
"learning_rate": 1.927864018523963e-05,
"loss": 1.464,
"step": 884
},
{
"epoch": 1.0473512873631252,
"grad_norm": 4.84375,
"learning_rate": 1.927555233866294e-05,
"loss": 1.0184,
"step": 885
},
{
"epoch": 1.0485350695472033,
"grad_norm": 5.3125,
"learning_rate": 1.927245814565499e-05,
"loss": 1.3389,
"step": 886
},
{
"epoch": 1.0497188517312814,
"grad_norm": 6.125,
"learning_rate": 1.9269357608332867e-05,
"loss": 1.5653,
"step": 887
},
{
"epoch": 1.0509026339153595,
"grad_norm": 6.125,
"learning_rate": 1.9266250728817985e-05,
"loss": 1.5631,
"step": 888
},
{
"epoch": 1.0520864160994376,
"grad_norm": 7.03125,
"learning_rate": 1.926313750923611e-05,
"loss": 1.3839,
"step": 889
},
{
"epoch": 1.0532701982835158,
"grad_norm": 6.71875,
"learning_rate": 1.9260017951717334e-05,
"loss": 1.3291,
"step": 890
},
{
"epoch": 1.0544539804675939,
"grad_norm": 5.90625,
"learning_rate": 1.9256892058396098e-05,
"loss": 1.5651,
"step": 891
},
{
"epoch": 1.0556377626516722,
"grad_norm": 5.34375,
"learning_rate": 1.9253759831411166e-05,
"loss": 1.0454,
"step": 892
},
{
"epoch": 1.0568215448357503,
"grad_norm": 5.78125,
"learning_rate": 1.9250621272905643e-05,
"loss": 1.2137,
"step": 893
},
{
"epoch": 1.0580053270198284,
"grad_norm": 4.84375,
"learning_rate": 1.924747638502696e-05,
"loss": 0.9991,
"step": 894
},
{
"epoch": 1.0591891092039065,
"grad_norm": 4.28125,
"learning_rate": 1.924432516992689e-05,
"loss": 0.9066,
"step": 895
},
{
"epoch": 1.0603728913879846,
"grad_norm": 5.34375,
"learning_rate": 1.9241167629761527e-05,
"loss": 1.0054,
"step": 896
},
{
"epoch": 1.0615566735720627,
"grad_norm": 5.25,
"learning_rate": 1.923800376669129e-05,
"loss": 1.351,
"step": 897
},
{
"epoch": 1.0627404557561408,
"grad_norm": 6.53125,
"learning_rate": 1.9234833582880923e-05,
"loss": 1.6873,
"step": 898
},
{
"epoch": 1.063924237940219,
"grad_norm": 5.71875,
"learning_rate": 1.9231657080499507e-05,
"loss": 1.5955,
"step": 899
},
{
"epoch": 1.065108020124297,
"grad_norm": 4.5,
"learning_rate": 1.922847426172044e-05,
"loss": 0.8937,
"step": 900
},
{
"epoch": 1.0662918023083752,
"grad_norm": 6.8125,
"learning_rate": 1.9225285128721435e-05,
"loss": 1.8723,
"step": 901
},
{
"epoch": 1.0674755844924535,
"grad_norm": 5.0625,
"learning_rate": 1.9222089683684527e-05,
"loss": 1.009,
"step": 902
},
{
"epoch": 1.0686593666765316,
"grad_norm": 5.0,
"learning_rate": 1.9218887928796083e-05,
"loss": 1.2936,
"step": 903
},
{
"epoch": 1.0698431488606097,
"grad_norm": 6.40625,
"learning_rate": 1.9215679866246774e-05,
"loss": 1.7399,
"step": 904
},
{
"epoch": 1.0710269310446878,
"grad_norm": 5.875,
"learning_rate": 1.9212465498231587e-05,
"loss": 1.5589,
"step": 905
},
{
"epoch": 1.072210713228766,
"grad_norm": 5.34375,
"learning_rate": 1.9209244826949833e-05,
"loss": 1.4371,
"step": 906
},
{
"epoch": 1.073394495412844,
"grad_norm": 4.40625,
"learning_rate": 1.9206017854605122e-05,
"loss": 0.8713,
"step": 907
},
{
"epoch": 1.0745782775969221,
"grad_norm": 5.375,
"learning_rate": 1.9202784583405386e-05,
"loss": 1.4728,
"step": 908
},
{
"epoch": 1.0757620597810003,
"grad_norm": 5.21875,
"learning_rate": 1.9199545015562866e-05,
"loss": 0.9718,
"step": 909
},
{
"epoch": 1.0769458419650784,
"grad_norm": 4.9375,
"learning_rate": 1.9196299153294107e-05,
"loss": 1.0689,
"step": 910
},
{
"epoch": 1.0781296241491565,
"grad_norm": 6.8125,
"learning_rate": 1.919304699881996e-05,
"loss": 1.3155,
"step": 911
},
{
"epoch": 1.0793134063332346,
"grad_norm": 5.3125,
"learning_rate": 1.9189788554365586e-05,
"loss": 1.1212,
"step": 912
},
{
"epoch": 1.080497188517313,
"grad_norm": 5.4375,
"learning_rate": 1.918652382216045e-05,
"loss": 1.3804,
"step": 913
},
{
"epoch": 1.081680970701391,
"grad_norm": 6.625,
"learning_rate": 1.918325280443831e-05,
"loss": 1.7244,
"step": 914
},
{
"epoch": 1.0828647528854691,
"grad_norm": 4.90625,
"learning_rate": 1.9179975503437235e-05,
"loss": 1.1484,
"step": 915
},
{
"epoch": 1.0840485350695472,
"grad_norm": 5.375,
"learning_rate": 1.9176691921399585e-05,
"loss": 1.3466,
"step": 916
},
{
"epoch": 1.0852323172536253,
"grad_norm": 5.25,
"learning_rate": 1.9173402060572028e-05,
"loss": 1.1894,
"step": 917
},
{
"epoch": 1.0864160994377035,
"grad_norm": 4.875,
"learning_rate": 1.9170105923205517e-05,
"loss": 1.1813,
"step": 918
},
{
"epoch": 1.0875998816217816,
"grad_norm": 5.5625,
"learning_rate": 1.916680351155531e-05,
"loss": 1.6616,
"step": 919
},
{
"epoch": 1.0887836638058597,
"grad_norm": 5.4375,
"learning_rate": 1.9163494827880943e-05,
"loss": 1.26,
"step": 920
},
{
"epoch": 1.0899674459899378,
"grad_norm": 5.375,
"learning_rate": 1.916017987444626e-05,
"loss": 1.474,
"step": 921
},
{
"epoch": 1.0911512281740159,
"grad_norm": 4.9375,
"learning_rate": 1.915685865351938e-05,
"loss": 1.5007,
"step": 922
},
{
"epoch": 1.0923350103580942,
"grad_norm": 5.375,
"learning_rate": 1.9153531167372725e-05,
"loss": 1.2325,
"step": 923
},
{
"epoch": 1.0935187925421723,
"grad_norm": 6.375,
"learning_rate": 1.9150197418282993e-05,
"loss": 1.3736,
"step": 924
},
{
"epoch": 1.0947025747262504,
"grad_norm": 5.21875,
"learning_rate": 1.9146857408531174e-05,
"loss": 1.1506,
"step": 925
},
{
"epoch": 1.0958863569103285,
"grad_norm": 5.71875,
"learning_rate": 1.9143511140402532e-05,
"loss": 1.2378,
"step": 926
},
{
"epoch": 1.0970701390944066,
"grad_norm": 5.90625,
"learning_rate": 1.9140158616186625e-05,
"loss": 1.1685,
"step": 927
},
{
"epoch": 1.0982539212784848,
"grad_norm": 5.625,
"learning_rate": 1.913679983817728e-05,
"loss": 1.3467,
"step": 928
},
{
"epoch": 1.0994377034625629,
"grad_norm": 5.8125,
"learning_rate": 1.913343480867261e-05,
"loss": 1.3123,
"step": 929
},
{
"epoch": 1.100621485646641,
"grad_norm": 6.09375,
"learning_rate": 1.9130063529975004e-05,
"loss": 1.0767,
"step": 930
},
{
"epoch": 1.101805267830719,
"grad_norm": 6.6875,
"learning_rate": 1.912668600439113e-05,
"loss": 1.8134,
"step": 931
},
{
"epoch": 1.1029890500147972,
"grad_norm": 7.0,
"learning_rate": 1.9123302234231924e-05,
"loss": 1.5853,
"step": 932
},
{
"epoch": 1.1041728321988753,
"grad_norm": 6.125,
"learning_rate": 1.91199122218126e-05,
"loss": 1.5315,
"step": 933
},
{
"epoch": 1.1053566143829536,
"grad_norm": 4.84375,
"learning_rate": 1.9116515969452637e-05,
"loss": 1.2047,
"step": 934
},
{
"epoch": 1.1065403965670317,
"grad_norm": 6.40625,
"learning_rate": 1.9113113479475784e-05,
"loss": 1.8077,
"step": 935
},
{
"epoch": 1.1077241787511098,
"grad_norm": 5.0625,
"learning_rate": 1.910970475421007e-05,
"loss": 0.9796,
"step": 936
},
{
"epoch": 1.108907960935188,
"grad_norm": 6.0,
"learning_rate": 1.9106289795987774e-05,
"loss": 1.5812,
"step": 937
},
{
"epoch": 1.110091743119266,
"grad_norm": 6.03125,
"learning_rate": 1.910286860714545e-05,
"loss": 1.7554,
"step": 938
},
{
"epoch": 1.1112755253033442,
"grad_norm": 5.84375,
"learning_rate": 1.909944119002391e-05,
"loss": 1.5331,
"step": 939
},
{
"epoch": 1.1124593074874223,
"grad_norm": 5.71875,
"learning_rate": 1.909600754696823e-05,
"loss": 1.1712,
"step": 940
},
{
"epoch": 1.1136430896715004,
"grad_norm": 5.65625,
"learning_rate": 1.909256768032774e-05,
"loss": 1.4563,
"step": 941
},
{
"epoch": 1.1148268718555785,
"grad_norm": 5.15625,
"learning_rate": 1.9089121592456044e-05,
"loss": 1.1163,
"step": 942
},
{
"epoch": 1.1160106540396566,
"grad_norm": 5.15625,
"learning_rate": 1.908566928571098e-05,
"loss": 1.095,
"step": 943
},
{
"epoch": 1.117194436223735,
"grad_norm": 5.71875,
"learning_rate": 1.908221076245466e-05,
"loss": 1.1738,
"step": 944
},
{
"epoch": 1.118378218407813,
"grad_norm": 7.21875,
"learning_rate": 1.9078746025053442e-05,
"loss": 1.8845,
"step": 945
},
{
"epoch": 1.1195620005918911,
"grad_norm": 5.3125,
"learning_rate": 1.907527507587793e-05,
"loss": 1.3628,
"step": 946
},
{
"epoch": 1.1207457827759693,
"grad_norm": 4.875,
"learning_rate": 1.9071797917302995e-05,
"loss": 0.9133,
"step": 947
},
{
"epoch": 1.1219295649600474,
"grad_norm": 4.84375,
"learning_rate": 1.9068314551707738e-05,
"loss": 1.0523,
"step": 948
},
{
"epoch": 1.1231133471441255,
"grad_norm": 3.90625,
"learning_rate": 1.9064824981475514e-05,
"loss": 1.0088,
"step": 949
},
{
"epoch": 1.1242971293282036,
"grad_norm": 5.75,
"learning_rate": 1.9061329208993927e-05,
"loss": 1.5127,
"step": 950
},
{
"epoch": 1.1254809115122817,
"grad_norm": 6.59375,
"learning_rate": 1.905782723665482e-05,
"loss": 1.6179,
"step": 951
},
{
"epoch": 1.1266646936963598,
"grad_norm": 6.34375,
"learning_rate": 1.9054319066854285e-05,
"loss": 1.9093,
"step": 952
},
{
"epoch": 1.127848475880438,
"grad_norm": 5.5,
"learning_rate": 1.905080470199264e-05,
"loss": 1.1013,
"step": 953
},
{
"epoch": 1.129032258064516,
"grad_norm": 5.6875,
"learning_rate": 1.904728414447446e-05,
"loss": 1.4124,
"step": 954
},
{
"epoch": 1.1302160402485943,
"grad_norm": 5.6875,
"learning_rate": 1.9043757396708536e-05,
"loss": 1.107,
"step": 955
},
{
"epoch": 1.1313998224326725,
"grad_norm": 5.21875,
"learning_rate": 1.9040224461107918e-05,
"loss": 1.1956,
"step": 956
},
{
"epoch": 1.1325836046167506,
"grad_norm": 5.5,
"learning_rate": 1.903668534008987e-05,
"loss": 1.1086,
"step": 957
},
{
"epoch": 1.1337673868008287,
"grad_norm": 5.53125,
"learning_rate": 1.90331400360759e-05,
"loss": 0.9453,
"step": 958
},
{
"epoch": 1.1349511689849068,
"grad_norm": 5.4375,
"learning_rate": 1.9029588551491743e-05,
"loss": 1.2976,
"step": 959
},
{
"epoch": 1.1361349511689849,
"grad_norm": 5.84375,
"learning_rate": 1.9026030888767364e-05,
"loss": 1.2753,
"step": 960
},
{
"epoch": 1.137318733353063,
"grad_norm": 5.8125,
"learning_rate": 1.9022467050336945e-05,
"loss": 1.5271,
"step": 961
},
{
"epoch": 1.138502515537141,
"grad_norm": 5.1875,
"learning_rate": 1.901889703863891e-05,
"loss": 1.0929,
"step": 962
},
{
"epoch": 1.1396862977212192,
"grad_norm": 5.09375,
"learning_rate": 1.90153208561159e-05,
"loss": 1.208,
"step": 963
},
{
"epoch": 1.1408700799052975,
"grad_norm": 5.8125,
"learning_rate": 1.901173850521477e-05,
"loss": 1.005,
"step": 964
},
{
"epoch": 1.1420538620893756,
"grad_norm": 5.53125,
"learning_rate": 1.9008149988386608e-05,
"loss": 1.3335,
"step": 965
},
{
"epoch": 1.1432376442734538,
"grad_norm": 6.4375,
"learning_rate": 1.900455530808671e-05,
"loss": 1.7188,
"step": 966
},
{
"epoch": 1.1444214264575319,
"grad_norm": 5.78125,
"learning_rate": 1.9000954466774603e-05,
"loss": 1.1978,
"step": 967
},
{
"epoch": 1.14560520864161,
"grad_norm": 6.53125,
"learning_rate": 1.8997347466914014e-05,
"loss": 1.5109,
"step": 968
},
{
"epoch": 1.146788990825688,
"grad_norm": 5.1875,
"learning_rate": 1.899373431097289e-05,
"loss": 1.0175,
"step": 969
},
{
"epoch": 1.1479727730097662,
"grad_norm": 5.375,
"learning_rate": 1.8990115001423396e-05,
"loss": 1.4012,
"step": 970
},
{
"epoch": 1.1491565551938443,
"grad_norm": 6.0625,
"learning_rate": 1.8986489540741895e-05,
"loss": 1.2845,
"step": 971
},
{
"epoch": 1.1503403373779224,
"grad_norm": 6.3125,
"learning_rate": 1.8982857931408972e-05,
"loss": 1.4204,
"step": 972
},
{
"epoch": 1.1515241195620005,
"grad_norm": 5.40625,
"learning_rate": 1.8979220175909408e-05,
"loss": 1.3842,
"step": 973
},
{
"epoch": 1.1527079017460786,
"grad_norm": 6.15625,
"learning_rate": 1.8975576276732196e-05,
"loss": 1.5307,
"step": 974
},
{
"epoch": 1.1538916839301567,
"grad_norm": 7.09375,
"learning_rate": 1.8971926236370532e-05,
"loss": 1.3406,
"step": 975
},
{
"epoch": 1.155075466114235,
"grad_norm": 6.96875,
"learning_rate": 1.896827005732181e-05,
"loss": 1.501,
"step": 976
},
{
"epoch": 1.1562592482983132,
"grad_norm": 6.3125,
"learning_rate": 1.8964607742087626e-05,
"loss": 1.1783,
"step": 977
},
{
"epoch": 1.1574430304823913,
"grad_norm": 5.5625,
"learning_rate": 1.8960939293173777e-05,
"loss": 1.0948,
"step": 978
},
{
"epoch": 1.1586268126664694,
"grad_norm": 5.1875,
"learning_rate": 1.8957264713090256e-05,
"loss": 1.0202,
"step": 979
},
{
"epoch": 1.1598105948505475,
"grad_norm": 5.96875,
"learning_rate": 1.8953584004351243e-05,
"loss": 1.3237,
"step": 980
},
{
"epoch": 1.1609943770346256,
"grad_norm": 4.84375,
"learning_rate": 1.8949897169475126e-05,
"loss": 1.0506,
"step": 981
},
{
"epoch": 1.1621781592187037,
"grad_norm": 6.125,
"learning_rate": 1.8946204210984468e-05,
"loss": 1.451,
"step": 982
},
{
"epoch": 1.1633619414027818,
"grad_norm": 5.71875,
"learning_rate": 1.894250513140604e-05,
"loss": 0.9385,
"step": 983
},
{
"epoch": 1.16454572358686,
"grad_norm": 5.15625,
"learning_rate": 1.8938799933270784e-05,
"loss": 1.1655,
"step": 984
},
{
"epoch": 1.1657295057709383,
"grad_norm": 5.4375,
"learning_rate": 1.893508861911384e-05,
"loss": 0.9766,
"step": 985
},
{
"epoch": 1.1669132879550164,
"grad_norm": 5.65625,
"learning_rate": 1.8931371191474526e-05,
"loss": 1.5581,
"step": 986
},
{
"epoch": 1.1680970701390945,
"grad_norm": 5.375,
"learning_rate": 1.8927647652896347e-05,
"loss": 1.4594,
"step": 987
},
{
"epoch": 1.1692808523231726,
"grad_norm": 6.0625,
"learning_rate": 1.8923918005926984e-05,
"loss": 1.1388,
"step": 988
},
{
"epoch": 1.1704646345072507,
"grad_norm": 4.90625,
"learning_rate": 1.892018225311831e-05,
"loss": 1.1581,
"step": 989
},
{
"epoch": 1.1716484166913288,
"grad_norm": 5.75,
"learning_rate": 1.8916440397026355e-05,
"loss": 0.9826,
"step": 990
},
{
"epoch": 1.172832198875407,
"grad_norm": 6.09375,
"learning_rate": 1.8912692440211345e-05,
"loss": 1.383,
"step": 991
},
{
"epoch": 1.174015981059485,
"grad_norm": 4.9375,
"learning_rate": 1.8908938385237667e-05,
"loss": 0.8212,
"step": 992
},
{
"epoch": 1.1751997632435631,
"grad_norm": 4.8125,
"learning_rate": 1.8905178234673886e-05,
"loss": 0.9015,
"step": 993
},
{
"epoch": 1.1763835454276412,
"grad_norm": 7.3125,
"learning_rate": 1.890141199109274e-05,
"loss": 1.7464,
"step": 994
},
{
"epoch": 1.1775673276117193,
"grad_norm": 6.21875,
"learning_rate": 1.8897639657071135e-05,
"loss": 1.4729,
"step": 995
},
{
"epoch": 1.1787511097957977,
"grad_norm": 4.25,
"learning_rate": 1.889386123519013e-05,
"loss": 0.8063,
"step": 996
},
{
"epoch": 1.1799348919798758,
"grad_norm": 6.125,
"learning_rate": 1.8890076728034974e-05,
"loss": 1.2844,
"step": 997
},
{
"epoch": 1.1811186741639539,
"grad_norm": 7.0,
"learning_rate": 1.8886286138195063e-05,
"loss": 1.7225,
"step": 998
},
{
"epoch": 1.182302456348032,
"grad_norm": 6.46875,
"learning_rate": 1.8882489468263954e-05,
"loss": 1.6683,
"step": 999
},
{
"epoch": 1.18348623853211,
"grad_norm": 6.53125,
"learning_rate": 1.8878686720839378e-05,
"loss": 1.8076,
"step": 1000
},
{
"epoch": 1.1846700207161882,
"grad_norm": 5.21875,
"learning_rate": 1.8874877898523204e-05,
"loss": 1.1879,
"step": 1001
},
{
"epoch": 1.1858538029002663,
"grad_norm": 4.34375,
"learning_rate": 1.887106300392148e-05,
"loss": 0.6932,
"step": 1002
},
{
"epoch": 1.1870375850843444,
"grad_norm": 6.21875,
"learning_rate": 1.8867242039644387e-05,
"loss": 1.5336,
"step": 1003
},
{
"epoch": 1.1882213672684225,
"grad_norm": 6.71875,
"learning_rate": 1.8863415008306277e-05,
"loss": 1.4397,
"step": 1004
},
{
"epoch": 1.1894051494525006,
"grad_norm": 5.71875,
"learning_rate": 1.8859581912525636e-05,
"loss": 0.9673,
"step": 1005
},
{
"epoch": 1.190588931636579,
"grad_norm": 7.40625,
"learning_rate": 1.8855742754925122e-05,
"loss": 1.6781,
"step": 1006
},
{
"epoch": 1.191772713820657,
"grad_norm": 5.34375,
"learning_rate": 1.885189753813152e-05,
"loss": 1.2905,
"step": 1007
},
{
"epoch": 1.1929564960047352,
"grad_norm": 6.5625,
"learning_rate": 1.8848046264775766e-05,
"loss": 1.2941,
"step": 1008
},
{
"epoch": 1.1941402781888133,
"grad_norm": 5.25,
"learning_rate": 1.8844188937492948e-05,
"loss": 0.9897,
"step": 1009
},
{
"epoch": 1.1953240603728914,
"grad_norm": 5.875,
"learning_rate": 1.8840325558922285e-05,
"loss": 1.8209,
"step": 1010
},
{
"epoch": 1.1965078425569695,
"grad_norm": 5.53125,
"learning_rate": 1.8836456131707143e-05,
"loss": 1.1735,
"step": 1011
},
{
"epoch": 1.1976916247410476,
"grad_norm": 5.53125,
"learning_rate": 1.8832580658495026e-05,
"loss": 1.5464,
"step": 1012
},
{
"epoch": 1.1988754069251257,
"grad_norm": 5.65625,
"learning_rate": 1.882869914193757e-05,
"loss": 1.3727,
"step": 1013
},
{
"epoch": 1.2000591891092038,
"grad_norm": 6.21875,
"learning_rate": 1.8824811584690557e-05,
"loss": 1.594,
"step": 1014
},
{
"epoch": 1.201242971293282,
"grad_norm": 5.625,
"learning_rate": 1.882091798941389e-05,
"loss": 1.5442,
"step": 1015
},
{
"epoch": 1.20242675347736,
"grad_norm": 6.0,
"learning_rate": 1.8817018358771612e-05,
"loss": 1.2388,
"step": 1016
},
{
"epoch": 1.2036105356614384,
"grad_norm": 6.46875,
"learning_rate": 1.8813112695431885e-05,
"loss": 1.7463,
"step": 1017
},
{
"epoch": 1.2047943178455165,
"grad_norm": 5.875,
"learning_rate": 1.880920100206701e-05,
"loss": 1.3667,
"step": 1018
},
{
"epoch": 1.2059781000295946,
"grad_norm": 5.65625,
"learning_rate": 1.8805283281353413e-05,
"loss": 1.3563,
"step": 1019
},
{
"epoch": 1.2071618822136727,
"grad_norm": 5.9375,
"learning_rate": 1.8801359535971628e-05,
"loss": 1.3621,
"step": 1020
},
{
"epoch": 1.2083456643977508,
"grad_norm": 4.40625,
"learning_rate": 1.8797429768606336e-05,
"loss": 0.957,
"step": 1021
},
{
"epoch": 1.209529446581829,
"grad_norm": 5.71875,
"learning_rate": 1.879349398194632e-05,
"loss": 1.3796,
"step": 1022
},
{
"epoch": 1.210713228765907,
"grad_norm": 5.6875,
"learning_rate": 1.8789552178684483e-05,
"loss": 1.1666,
"step": 1023
},
{
"epoch": 1.2118970109499851,
"grad_norm": 6.5625,
"learning_rate": 1.8785604361517852e-05,
"loss": 1.5825,
"step": 1024
},
{
"epoch": 1.2130807931340633,
"grad_norm": 5.21875,
"learning_rate": 1.8781650533147572e-05,
"loss": 1.1162,
"step": 1025
},
{
"epoch": 1.2142645753181416,
"grad_norm": 6.25,
"learning_rate": 1.877769069627888e-05,
"loss": 1.4441,
"step": 1026
},
{
"epoch": 1.2154483575022197,
"grad_norm": 5.0,
"learning_rate": 1.8773724853621154e-05,
"loss": 1.2393,
"step": 1027
},
{
"epoch": 1.2166321396862978,
"grad_norm": 5.53125,
"learning_rate": 1.8769753007887858e-05,
"loss": 1.2374,
"step": 1028
},
{
"epoch": 1.217815921870376,
"grad_norm": 6.125,
"learning_rate": 1.8765775161796565e-05,
"loss": 1.5663,
"step": 1029
},
{
"epoch": 1.218999704054454,
"grad_norm": 6.21875,
"learning_rate": 1.8761791318068972e-05,
"loss": 1.3138,
"step": 1030
},
{
"epoch": 1.2201834862385321,
"grad_norm": 7.09375,
"learning_rate": 1.8757801479430858e-05,
"loss": 1.8548,
"step": 1031
},
{
"epoch": 1.2213672684226102,
"grad_norm": 5.5625,
"learning_rate": 1.8753805648612115e-05,
"loss": 1.3471,
"step": 1032
},
{
"epoch": 1.2225510506066883,
"grad_norm": 6.09375,
"learning_rate": 1.8749803828346732e-05,
"loss": 1.7264,
"step": 1033
},
{
"epoch": 1.2237348327907664,
"grad_norm": 4.875,
"learning_rate": 1.87457960213728e-05,
"loss": 0.9052,
"step": 1034
},
{
"epoch": 1.2249186149748446,
"grad_norm": 5.875,
"learning_rate": 1.8741782230432503e-05,
"loss": 1.3161,
"step": 1035
},
{
"epoch": 1.2261023971589227,
"grad_norm": 5.90625,
"learning_rate": 1.8737762458272114e-05,
"loss": 1.4371,
"step": 1036
},
{
"epoch": 1.2272861793430008,
"grad_norm": 6.40625,
"learning_rate": 1.8733736707642012e-05,
"loss": 1.5041,
"step": 1037
},
{
"epoch": 1.228469961527079,
"grad_norm": 5.09375,
"learning_rate": 1.8729704981296654e-05,
"loss": 0.8493,
"step": 1038
},
{
"epoch": 1.2296537437111572,
"grad_norm": 5.09375,
"learning_rate": 1.8725667281994584e-05,
"loss": 0.9745,
"step": 1039
},
{
"epoch": 1.2308375258952353,
"grad_norm": 5.1875,
"learning_rate": 1.872162361249845e-05,
"loss": 1.0359,
"step": 1040
},
{
"epoch": 1.2320213080793134,
"grad_norm": 5.46875,
"learning_rate": 1.8717573975574962e-05,
"loss": 0.968,
"step": 1041
},
{
"epoch": 1.2332050902633915,
"grad_norm": 6.53125,
"learning_rate": 1.8713518373994932e-05,
"loss": 1.1988,
"step": 1042
},
{
"epoch": 1.2343888724474696,
"grad_norm": 4.9375,
"learning_rate": 1.8709456810533248e-05,
"loss": 0.8927,
"step": 1043
},
{
"epoch": 1.2355726546315478,
"grad_norm": 5.71875,
"learning_rate": 1.8705389287968866e-05,
"loss": 1.3352,
"step": 1044
},
{
"epoch": 1.2367564368156259,
"grad_norm": 6.25,
"learning_rate": 1.8701315809084835e-05,
"loss": 1.6099,
"step": 1045
},
{
"epoch": 1.237940218999704,
"grad_norm": 5.25,
"learning_rate": 1.869723637666827e-05,
"loss": 0.8058,
"step": 1046
},
{
"epoch": 1.2391240011837823,
"grad_norm": 6.84375,
"learning_rate": 1.869315099351036e-05,
"loss": 1.7114,
"step": 1047
},
{
"epoch": 1.2403077833678604,
"grad_norm": 5.96875,
"learning_rate": 1.8689059662406373e-05,
"loss": 1.3862,
"step": 1048
},
{
"epoch": 1.2414915655519385,
"grad_norm": 6.375,
"learning_rate": 1.8684962386155633e-05,
"loss": 1.5886,
"step": 1049
},
{
"epoch": 1.2426753477360166,
"grad_norm": 5.84375,
"learning_rate": 1.868085916756155e-05,
"loss": 1.2667,
"step": 1050
},
{
"epoch": 1.2438591299200947,
"grad_norm": 6.15625,
"learning_rate": 1.867675000943158e-05,
"loss": 1.3921,
"step": 1051
},
{
"epoch": 1.2450429121041728,
"grad_norm": 5.4375,
"learning_rate": 1.867263491457726e-05,
"loss": 1.0819,
"step": 1052
},
{
"epoch": 1.246226694288251,
"grad_norm": 5.34375,
"learning_rate": 1.8668513885814176e-05,
"loss": 1.0033,
"step": 1053
},
{
"epoch": 1.247410476472329,
"grad_norm": 5.875,
"learning_rate": 1.866438692596198e-05,
"loss": 1.301,
"step": 1054
},
{
"epoch": 1.2485942586564072,
"grad_norm": 5.21875,
"learning_rate": 1.866025403784439e-05,
"loss": 1.1192,
"step": 1055
},
{
"epoch": 1.2497780408404853,
"grad_norm": 4.84375,
"learning_rate": 1.865611522428916e-05,
"loss": 1.1792,
"step": 1056
},
{
"epoch": 1.2509618230245634,
"grad_norm": 5.96875,
"learning_rate": 1.8651970488128115e-05,
"loss": 1.5993,
"step": 1057
},
{
"epoch": 1.2521456052086415,
"grad_norm": 6.5625,
"learning_rate": 1.864781983219713e-05,
"loss": 1.8914,
"step": 1058
},
{
"epoch": 1.2533293873927198,
"grad_norm": 5.96875,
"learning_rate": 1.8643663259336126e-05,
"loss": 1.0465,
"step": 1059
},
{
"epoch": 1.254513169576798,
"grad_norm": 5.5625,
"learning_rate": 1.8639500772389073e-05,
"loss": 1.2585,
"step": 1060
},
{
"epoch": 1.255696951760876,
"grad_norm": 6.59375,
"learning_rate": 1.8635332374203993e-05,
"loss": 1.5641,
"step": 1061
},
{
"epoch": 1.2568807339449541,
"grad_norm": 6.0,
"learning_rate": 1.8631158067632942e-05,
"loss": 1.4758,
"step": 1062
},
{
"epoch": 1.2580645161290323,
"grad_norm": 5.34375,
"learning_rate": 1.862697785553203e-05,
"loss": 1.0488,
"step": 1063
},
{
"epoch": 1.2592482983131104,
"grad_norm": 4.9375,
"learning_rate": 1.8622791740761398e-05,
"loss": 0.9649,
"step": 1064
},
{
"epoch": 1.2604320804971885,
"grad_norm": 6.53125,
"learning_rate": 1.861859972618523e-05,
"loss": 1.4518,
"step": 1065
},
{
"epoch": 1.2616158626812666,
"grad_norm": 5.21875,
"learning_rate": 1.861440181467175e-05,
"loss": 1.1232,
"step": 1066
},
{
"epoch": 1.262799644865345,
"grad_norm": 5.96875,
"learning_rate": 1.861019800909322e-05,
"loss": 1.5442,
"step": 1067
},
{
"epoch": 1.263983427049423,
"grad_norm": 5.6875,
"learning_rate": 1.8605988312325915e-05,
"loss": 1.4329,
"step": 1068
},
{
"epoch": 1.2651672092335011,
"grad_norm": 5.46875,
"learning_rate": 1.8601772727250158e-05,
"loss": 1.3702,
"step": 1069
},
{
"epoch": 1.2663509914175792,
"grad_norm": 5.5,
"learning_rate": 1.85975512567503e-05,
"loss": 1.3982,
"step": 1070
},
{
"epoch": 1.2675347736016573,
"grad_norm": 4.875,
"learning_rate": 1.8593323903714718e-05,
"loss": 0.9871,
"step": 1071
},
{
"epoch": 1.2687185557857354,
"grad_norm": 7.0625,
"learning_rate": 1.858909067103581e-05,
"loss": 1.9463,
"step": 1072
},
{
"epoch": 1.2699023379698136,
"grad_norm": 6.53125,
"learning_rate": 1.8584851561609995e-05,
"loss": 1.5476,
"step": 1073
},
{
"epoch": 1.2710861201538917,
"grad_norm": 5.78125,
"learning_rate": 1.8580606578337718e-05,
"loss": 1.4641,
"step": 1074
},
{
"epoch": 1.2722699023379698,
"grad_norm": 5.53125,
"learning_rate": 1.8576355724123446e-05,
"loss": 1.0996,
"step": 1075
},
{
"epoch": 1.2734536845220479,
"grad_norm": 6.75,
"learning_rate": 1.857209900187565e-05,
"loss": 1.7067,
"step": 1076
},
{
"epoch": 1.274637466706126,
"grad_norm": 5.28125,
"learning_rate": 1.8567836414506835e-05,
"loss": 1.5705,
"step": 1077
},
{
"epoch": 1.275821248890204,
"grad_norm": 5.28125,
"learning_rate": 1.85635679649335e-05,
"loss": 1.1138,
"step": 1078
},
{
"epoch": 1.2770050310742822,
"grad_norm": 7.15625,
"learning_rate": 1.8559293656076167e-05,
"loss": 1.4655,
"step": 1079
},
{
"epoch": 1.2781888132583605,
"grad_norm": 6.78125,
"learning_rate": 1.8555013490859367e-05,
"loss": 1.6313,
"step": 1080
},
{
"epoch": 1.2793725954424386,
"grad_norm": 6.21875,
"learning_rate": 1.8550727472211624e-05,
"loss": 1.3618,
"step": 1081
},
{
"epoch": 1.2805563776265168,
"grad_norm": 6.21875,
"learning_rate": 1.8546435603065488e-05,
"loss": 1.5423,
"step": 1082
},
{
"epoch": 1.2817401598105949,
"grad_norm": 5.6875,
"learning_rate": 1.85421378863575e-05,
"loss": 1.0703,
"step": 1083
},
{
"epoch": 1.282923941994673,
"grad_norm": 5.625,
"learning_rate": 1.8537834325028196e-05,
"loss": 1.1886,
"step": 1084
},
{
"epoch": 1.284107724178751,
"grad_norm": 5.84375,
"learning_rate": 1.8533524922022123e-05,
"loss": 1.1831,
"step": 1085
},
{
"epoch": 1.2852915063628292,
"grad_norm": 5.125,
"learning_rate": 1.852920968028782e-05,
"loss": 1.0677,
"step": 1086
},
{
"epoch": 1.2864752885469073,
"grad_norm": 6.3125,
"learning_rate": 1.852488860277782e-05,
"loss": 1.8992,
"step": 1087
},
{
"epoch": 1.2876590707309856,
"grad_norm": 4.125,
"learning_rate": 1.8520561692448655e-05,
"loss": 0.7481,
"step": 1088
},
{
"epoch": 1.2888428529150637,
"grad_norm": 4.9375,
"learning_rate": 1.851622895226084e-05,
"loss": 0.9621,
"step": 1089
},
{
"epoch": 1.2900266350991418,
"grad_norm": 5.0625,
"learning_rate": 1.851189038517888e-05,
"loss": 0.8731,
"step": 1090
},
{
"epoch": 1.29121041728322,
"grad_norm": 4.9375,
"learning_rate": 1.850754599417127e-05,
"loss": 1.1062,
"step": 1091
},
{
"epoch": 1.292394199467298,
"grad_norm": 6.25,
"learning_rate": 1.8503195782210484e-05,
"loss": 1.7273,
"step": 1092
},
{
"epoch": 1.2935779816513762,
"grad_norm": 5.875,
"learning_rate": 1.849883975227299e-05,
"loss": 1.3938,
"step": 1093
},
{
"epoch": 1.2947617638354543,
"grad_norm": 6.1875,
"learning_rate": 1.8494477907339224e-05,
"loss": 1.2627,
"step": 1094
},
{
"epoch": 1.2959455460195324,
"grad_norm": 6.28125,
"learning_rate": 1.8490110250393617e-05,
"loss": 1.5027,
"step": 1095
},
{
"epoch": 1.2971293282036105,
"grad_norm": 5.375,
"learning_rate": 1.8485736784424554e-05,
"loss": 1.2642,
"step": 1096
},
{
"epoch": 1.2983131103876886,
"grad_norm": 5.65625,
"learning_rate": 1.848135751242441e-05,
"loss": 1.3878,
"step": 1097
},
{
"epoch": 1.2994968925717667,
"grad_norm": 5.15625,
"learning_rate": 1.8476972437389535e-05,
"loss": 0.8602,
"step": 1098
},
{
"epoch": 1.3006806747558448,
"grad_norm": 6.65625,
"learning_rate": 1.8472581562320235e-05,
"loss": 1.3648,
"step": 1099
},
{
"epoch": 1.301864456939923,
"grad_norm": 6.5625,
"learning_rate": 1.8468184890220803e-05,
"loss": 1.3919,
"step": 1100
},
{
"epoch": 1.3030482391240013,
"grad_norm": 7.21875,
"learning_rate": 1.8463782424099478e-05,
"loss": 1.5623,
"step": 1101
},
{
"epoch": 1.3042320213080794,
"grad_norm": 6.40625,
"learning_rate": 1.8459374166968483e-05,
"loss": 1.8396,
"step": 1102
},
{
"epoch": 1.3054158034921575,
"grad_norm": 5.96875,
"learning_rate": 1.845496012184399e-05,
"loss": 1.0512,
"step": 1103
},
{
"epoch": 1.3065995856762356,
"grad_norm": 5.125,
"learning_rate": 1.8450540291746138e-05,
"loss": 1.1027,
"step": 1104
},
{
"epoch": 1.3077833678603137,
"grad_norm": 5.625,
"learning_rate": 1.844611467969902e-05,
"loss": 1.1773,
"step": 1105
},
{
"epoch": 1.3089671500443918,
"grad_norm": 6.09375,
"learning_rate": 1.8441683288730686e-05,
"loss": 1.4142,
"step": 1106
},
{
"epoch": 1.31015093222847,
"grad_norm": 5.28125,
"learning_rate": 1.8437246121873143e-05,
"loss": 1.0883,
"step": 1107
},
{
"epoch": 1.311334714412548,
"grad_norm": 5.75,
"learning_rate": 1.8432803182162346e-05,
"loss": 1.1892,
"step": 1108
},
{
"epoch": 1.3125184965966263,
"grad_norm": 5.71875,
"learning_rate": 1.84283544726382e-05,
"loss": 1.3985,
"step": 1109
},
{
"epoch": 1.3137022787807044,
"grad_norm": 5.46875,
"learning_rate": 1.842389999634456e-05,
"loss": 1.3613,
"step": 1110
},
{
"epoch": 1.3148860609647826,
"grad_norm": 5.78125,
"learning_rate": 1.8419439756329226e-05,
"loss": 1.1586,
"step": 1111
},
{
"epoch": 1.3160698431488607,
"grad_norm": 6.375,
"learning_rate": 1.8414973755643942e-05,
"loss": 1.2295,
"step": 1112
},
{
"epoch": 1.3172536253329388,
"grad_norm": 7.0,
"learning_rate": 1.8410501997344394e-05,
"loss": 1.1786,
"step": 1113
},
{
"epoch": 1.3184374075170169,
"grad_norm": 6.46875,
"learning_rate": 1.8406024484490207e-05,
"loss": 1.8444,
"step": 1114
},
{
"epoch": 1.319621189701095,
"grad_norm": 6.53125,
"learning_rate": 1.840154122014494e-05,
"loss": 1.3313,
"step": 1115
},
{
"epoch": 1.320804971885173,
"grad_norm": 5.0,
"learning_rate": 1.839705220737609e-05,
"loss": 1.1129,
"step": 1116
},
{
"epoch": 1.3219887540692512,
"grad_norm": 5.53125,
"learning_rate": 1.8392557449255094e-05,
"loss": 1.4621,
"step": 1117
},
{
"epoch": 1.3231725362533293,
"grad_norm": 5.21875,
"learning_rate": 1.8388056948857303e-05,
"loss": 1.1684,
"step": 1118
},
{
"epoch": 1.3243563184374074,
"grad_norm": 5.0625,
"learning_rate": 1.838355070926201e-05,
"loss": 1.2198,
"step": 1119
},
{
"epoch": 1.3255401006214855,
"grad_norm": 6.0625,
"learning_rate": 1.8379038733552435e-05,
"loss": 1.3765,
"step": 1120
},
{
"epoch": 1.3267238828055639,
"grad_norm": 5.71875,
"learning_rate": 1.837452102481572e-05,
"loss": 1.1697,
"step": 1121
},
{
"epoch": 1.327907664989642,
"grad_norm": 5.28125,
"learning_rate": 1.836999758614293e-05,
"loss": 1.2712,
"step": 1122
},
{
"epoch": 1.32909144717372,
"grad_norm": 5.71875,
"learning_rate": 1.836546842062904e-05,
"loss": 1.1518,
"step": 1123
},
{
"epoch": 1.3302752293577982,
"grad_norm": 6.21875,
"learning_rate": 1.836093353137297e-05,
"loss": 1.2837,
"step": 1124
},
{
"epoch": 1.3314590115418763,
"grad_norm": 6.71875,
"learning_rate": 1.835639292147752e-05,
"loss": 1.4457,
"step": 1125
},
{
"epoch": 1.3326427937259544,
"grad_norm": 5.34375,
"learning_rate": 1.835184659404944e-05,
"loss": 1.2874,
"step": 1126
},
{
"epoch": 1.3338265759100325,
"grad_norm": 4.8125,
"learning_rate": 1.8347294552199366e-05,
"loss": 1.1977,
"step": 1127
},
{
"epoch": 1.3350103580941106,
"grad_norm": 6.125,
"learning_rate": 1.834273679904185e-05,
"loss": 1.2744,
"step": 1128
},
{
"epoch": 1.336194140278189,
"grad_norm": 5.40625,
"learning_rate": 1.833817333769537e-05,
"loss": 1.2426,
"step": 1129
},
{
"epoch": 1.337377922462267,
"grad_norm": 6.1875,
"learning_rate": 1.833360417128228e-05,
"loss": 1.4472,
"step": 1130
},
{
"epoch": 1.3385617046463452,
"grad_norm": 7.15625,
"learning_rate": 1.8329029302928858e-05,
"loss": 1.4022,
"step": 1131
},
{
"epoch": 1.3397454868304233,
"grad_norm": 5.21875,
"learning_rate": 1.8324448735765278e-05,
"loss": 1.2482,
"step": 1132
},
{
"epoch": 1.3409292690145014,
"grad_norm": 4.8125,
"learning_rate": 1.831986247292561e-05,
"loss": 0.8275,
"step": 1133
},
{
"epoch": 1.3421130511985795,
"grad_norm": 5.78125,
"learning_rate": 1.831527051754783e-05,
"loss": 1.2156,
"step": 1134
},
{
"epoch": 1.3432968333826576,
"grad_norm": 5.46875,
"learning_rate": 1.8310672872773788e-05,
"loss": 1.0232,
"step": 1135
},
{
"epoch": 1.3444806155667357,
"grad_norm": 6.0,
"learning_rate": 1.830606954174926e-05,
"loss": 1.3736,
"step": 1136
},
{
"epoch": 1.3456643977508138,
"grad_norm": 6.53125,
"learning_rate": 1.830146052762388e-05,
"loss": 1.6156,
"step": 1137
},
{
"epoch": 1.346848179934892,
"grad_norm": 5.25,
"learning_rate": 1.8296845833551192e-05,
"loss": 1.123,
"step": 1138
},
{
"epoch": 1.34803196211897,
"grad_norm": 5.5625,
"learning_rate": 1.8292225462688616e-05,
"loss": 1.2332,
"step": 1139
},
{
"epoch": 1.3492157443030481,
"grad_norm": 6.25,
"learning_rate": 1.8287599418197457e-05,
"loss": 1.552,
"step": 1140
},
{
"epoch": 1.3503995264871262,
"grad_norm": 5.375,
"learning_rate": 1.8282967703242905e-05,
"loss": 1.2938,
"step": 1141
},
{
"epoch": 1.3515833086712046,
"grad_norm": 6.5,
"learning_rate": 1.8278330320994035e-05,
"loss": 1.1512,
"step": 1142
},
{
"epoch": 1.3527670908552827,
"grad_norm": 5.84375,
"learning_rate": 1.827368727462378e-05,
"loss": 1.3857,
"step": 1143
},
{
"epoch": 1.3539508730393608,
"grad_norm": 5.59375,
"learning_rate": 1.8269038567308968e-05,
"loss": 0.9229,
"step": 1144
},
{
"epoch": 1.355134655223439,
"grad_norm": 5.625,
"learning_rate": 1.82643842022303e-05,
"loss": 1.1843,
"step": 1145
},
{
"epoch": 1.356318437407517,
"grad_norm": 5.25,
"learning_rate": 1.825972418257233e-05,
"loss": 1.3912,
"step": 1146
},
{
"epoch": 1.3575022195915951,
"grad_norm": 5.625,
"learning_rate": 1.82550585115235e-05,
"loss": 1.1908,
"step": 1147
},
{
"epoch": 1.3586860017756732,
"grad_norm": 6.3125,
"learning_rate": 1.8250387192276115e-05,
"loss": 1.2058,
"step": 1148
},
{
"epoch": 1.3598697839597513,
"grad_norm": 7.15625,
"learning_rate": 1.8245710228026336e-05,
"loss": 1.51,
"step": 1149
},
{
"epoch": 1.3610535661438297,
"grad_norm": 5.15625,
"learning_rate": 1.824102762197419e-05,
"loss": 1.1847,
"step": 1150
},
{
"epoch": 1.3622373483279078,
"grad_norm": 6.46875,
"learning_rate": 1.823633937732357e-05,
"loss": 1.0443,
"step": 1151
},
{
"epoch": 1.3634211305119859,
"grad_norm": 5.3125,
"learning_rate": 1.8231645497282218e-05,
"loss": 1.3024,
"step": 1152
},
{
"epoch": 1.364604912696064,
"grad_norm": 5.4375,
"learning_rate": 1.8226945985061743e-05,
"loss": 0.9063,
"step": 1153
},
{
"epoch": 1.365788694880142,
"grad_norm": 4.96875,
"learning_rate": 1.8222240843877593e-05,
"loss": 1.1351,
"step": 1154
},
{
"epoch": 1.3669724770642202,
"grad_norm": 5.84375,
"learning_rate": 1.8217530076949083e-05,
"loss": 1.3037,
"step": 1155
},
{
"epoch": 1.3681562592482983,
"grad_norm": 6.9375,
"learning_rate": 1.8212813687499365e-05,
"loss": 1.8689,
"step": 1156
},
{
"epoch": 1.3693400414323764,
"grad_norm": 5.84375,
"learning_rate": 1.8208091678755444e-05,
"loss": 1.1192,
"step": 1157
},
{
"epoch": 1.3705238236164545,
"grad_norm": 6.25,
"learning_rate": 1.8203364053948173e-05,
"loss": 1.7293,
"step": 1158
},
{
"epoch": 1.3717076058005326,
"grad_norm": 4.4375,
"learning_rate": 1.8198630816312236e-05,
"loss": 0.6688,
"step": 1159
},
{
"epoch": 1.3728913879846107,
"grad_norm": 5.375,
"learning_rate": 1.8193891969086164e-05,
"loss": 0.94,
"step": 1160
},
{
"epoch": 1.3740751701686889,
"grad_norm": 6.09375,
"learning_rate": 1.818914751551233e-05,
"loss": 1.3082,
"step": 1161
},
{
"epoch": 1.375258952352767,
"grad_norm": 5.15625,
"learning_rate": 1.818439745883694e-05,
"loss": 0.9209,
"step": 1162
},
{
"epoch": 1.3764427345368453,
"grad_norm": 6.25,
"learning_rate": 1.8179641802310035e-05,
"loss": 1.5589,
"step": 1163
},
{
"epoch": 1.3776265167209234,
"grad_norm": 6.59375,
"learning_rate": 1.817488054918548e-05,
"loss": 1.5628,
"step": 1164
},
{
"epoch": 1.3788102989050015,
"grad_norm": 5.40625,
"learning_rate": 1.817011370272098e-05,
"loss": 0.9794,
"step": 1165
},
{
"epoch": 1.3799940810890796,
"grad_norm": 6.125,
"learning_rate": 1.8165341266178056e-05,
"loss": 1.685,
"step": 1166
},
{
"epoch": 1.3811778632731577,
"grad_norm": 5.1875,
"learning_rate": 1.8160563242822068e-05,
"loss": 0.9396,
"step": 1167
},
{
"epoch": 1.3823616454572358,
"grad_norm": 5.625,
"learning_rate": 1.8155779635922178e-05,
"loss": 1.2423,
"step": 1168
},
{
"epoch": 1.383545427641314,
"grad_norm": 5.28125,
"learning_rate": 1.8150990448751393e-05,
"loss": 0.9562,
"step": 1169
},
{
"epoch": 1.384729209825392,
"grad_norm": 5.71875,
"learning_rate": 1.814619568458652e-05,
"loss": 1.5424,
"step": 1170
},
{
"epoch": 1.3859129920094704,
"grad_norm": 6.4375,
"learning_rate": 1.8141395346708192e-05,
"loss": 1.4395,
"step": 1171
},
{
"epoch": 1.3870967741935485,
"grad_norm": 6.3125,
"learning_rate": 1.8136589438400843e-05,
"loss": 1.5371,
"step": 1172
},
{
"epoch": 1.3882805563776266,
"grad_norm": 6.21875,
"learning_rate": 1.8131777962952732e-05,
"loss": 1.4758,
"step": 1173
},
{
"epoch": 1.3894643385617047,
"grad_norm": 5.90625,
"learning_rate": 1.8126960923655915e-05,
"loss": 1.5292,
"step": 1174
},
{
"epoch": 1.3906481207457828,
"grad_norm": 5.6875,
"learning_rate": 1.812213832380627e-05,
"loss": 1.0378,
"step": 1175
},
{
"epoch": 1.391831902929861,
"grad_norm": 7.15625,
"learning_rate": 1.811731016670347e-05,
"loss": 1.6733,
"step": 1176
},
{
"epoch": 1.393015685113939,
"grad_norm": 6.09375,
"learning_rate": 1.8112476455650987e-05,
"loss": 1.6444,
"step": 1177
},
{
"epoch": 1.3941994672980171,
"grad_norm": 5.78125,
"learning_rate": 1.8107637193956102e-05,
"loss": 1.1573,
"step": 1178
},
{
"epoch": 1.3953832494820952,
"grad_norm": 5.90625,
"learning_rate": 1.810279238492988e-05,
"loss": 1.7226,
"step": 1179
},
{
"epoch": 1.3965670316661734,
"grad_norm": 5.53125,
"learning_rate": 1.80979420318872e-05,
"loss": 1.1653,
"step": 1180
},
{
"epoch": 1.3977508138502515,
"grad_norm": 5.3125,
"learning_rate": 1.8093086138146723e-05,
"loss": 1.3516,
"step": 1181
},
{
"epoch": 1.3989345960343296,
"grad_norm": 5.40625,
"learning_rate": 1.8088224707030903e-05,
"loss": 1.1508,
"step": 1182
},
{
"epoch": 1.4001183782184077,
"grad_norm": 4.9375,
"learning_rate": 1.8083357741865976e-05,
"loss": 0.9156,
"step": 1183
},
{
"epoch": 1.401302160402486,
"grad_norm": 7.46875,
"learning_rate": 1.8078485245981982e-05,
"loss": 1.8316,
"step": 1184
},
{
"epoch": 1.4024859425865641,
"grad_norm": 6.0,
"learning_rate": 1.807360722271273e-05,
"loss": 1.2778,
"step": 1185
},
{
"epoch": 1.4036697247706422,
"grad_norm": 7.1875,
"learning_rate": 1.806872367539581e-05,
"loss": 1.4825,
"step": 1186
},
{
"epoch": 1.4048535069547203,
"grad_norm": 7.0625,
"learning_rate": 1.8063834607372603e-05,
"loss": 1.5877,
"step": 1187
},
{
"epoch": 1.4060372891387984,
"grad_norm": 5.75,
"learning_rate": 1.8058940021988257e-05,
"loss": 1.2812,
"step": 1188
},
{
"epoch": 1.4072210713228765,
"grad_norm": 5.75,
"learning_rate": 1.8054039922591705e-05,
"loss": 1.3761,
"step": 1189
},
{
"epoch": 1.4084048535069547,
"grad_norm": 4.53125,
"learning_rate": 1.804913431253564e-05,
"loss": 0.9879,
"step": 1190
},
{
"epoch": 1.409588635691033,
"grad_norm": 6.03125,
"learning_rate": 1.804422319517654e-05,
"loss": 1.1897,
"step": 1191
},
{
"epoch": 1.410772417875111,
"grad_norm": 4.59375,
"learning_rate": 1.803930657387464e-05,
"loss": 0.9774,
"step": 1192
},
{
"epoch": 1.4119562000591892,
"grad_norm": 5.03125,
"learning_rate": 1.803438445199394e-05,
"loss": 1.0439,
"step": 1193
},
{
"epoch": 1.4131399822432673,
"grad_norm": 5.5,
"learning_rate": 1.8029456832902215e-05,
"loss": 1.1948,
"step": 1194
},
{
"epoch": 1.4143237644273454,
"grad_norm": 5.75,
"learning_rate": 1.802452371997099e-05,
"loss": 1.336,
"step": 1195
},
{
"epoch": 1.4155075466114235,
"grad_norm": 5.5,
"learning_rate": 1.8019585116575554e-05,
"loss": 1.406,
"step": 1196
},
{
"epoch": 1.4166913287955016,
"grad_norm": 5.625,
"learning_rate": 1.801464102609495e-05,
"loss": 0.9517,
"step": 1197
},
{
"epoch": 1.4178751109795797,
"grad_norm": 5.84375,
"learning_rate": 1.8009691451911986e-05,
"loss": 1.3934,
"step": 1198
},
{
"epoch": 1.4190588931636579,
"grad_norm": 5.96875,
"learning_rate": 1.8004736397413204e-05,
"loss": 1.4439,
"step": 1199
},
{
"epoch": 1.420242675347736,
"grad_norm": 6.59375,
"learning_rate": 1.7999775865988903e-05,
"loss": 1.3333,
"step": 1200
},
{
"epoch": 1.421426457531814,
"grad_norm": 5.75,
"learning_rate": 1.799480986103314e-05,
"loss": 1.3871,
"step": 1201
},
{
"epoch": 1.4226102397158922,
"grad_norm": 6.59375,
"learning_rate": 1.7989838385943697e-05,
"loss": 1.6928,
"step": 1202
},
{
"epoch": 1.4237940218999703,
"grad_norm": 5.1875,
"learning_rate": 1.7984861444122127e-05,
"loss": 1.1271,
"step": 1203
},
{
"epoch": 1.4249778040840486,
"grad_norm": 5.53125,
"learning_rate": 1.7979879038973686e-05,
"loss": 1.3292,
"step": 1204
},
{
"epoch": 1.4261615862681267,
"grad_norm": 6.03125,
"learning_rate": 1.7974891173907406e-05,
"loss": 1.5227,
"step": 1205
},
{
"epoch": 1.4273453684522048,
"grad_norm": 5.3125,
"learning_rate": 1.7969897852336028e-05,
"loss": 0.9766,
"step": 1206
},
{
"epoch": 1.428529150636283,
"grad_norm": 5.1875,
"learning_rate": 1.7964899077676034e-05,
"loss": 1.1149,
"step": 1207
},
{
"epoch": 1.429712932820361,
"grad_norm": 6.15625,
"learning_rate": 1.795989485334764e-05,
"loss": 1.2373,
"step": 1208
},
{
"epoch": 1.4308967150044392,
"grad_norm": 5.90625,
"learning_rate": 1.7954885182774798e-05,
"loss": 1.1432,
"step": 1209
},
{
"epoch": 1.4320804971885173,
"grad_norm": 5.78125,
"learning_rate": 1.794987006938517e-05,
"loss": 1.5298,
"step": 1210
},
{
"epoch": 1.4332642793725954,
"grad_norm": 4.96875,
"learning_rate": 1.7944849516610147e-05,
"loss": 0.8989,
"step": 1211
},
{
"epoch": 1.4344480615566737,
"grad_norm": 4.75,
"learning_rate": 1.7939823527884845e-05,
"loss": 1.0535,
"step": 1212
},
{
"epoch": 1.4356318437407518,
"grad_norm": 5.0625,
"learning_rate": 1.793479210664811e-05,
"loss": 1.067,
"step": 1213
},
{
"epoch": 1.43681562592483,
"grad_norm": 5.4375,
"learning_rate": 1.792975525634248e-05,
"loss": 1.1348,
"step": 1214
},
{
"epoch": 1.437999408108908,
"grad_norm": 6.1875,
"learning_rate": 1.7924712980414227e-05,
"loss": 1.755,
"step": 1215
},
{
"epoch": 1.4391831902929861,
"grad_norm": 6.25,
"learning_rate": 1.7919665282313335e-05,
"loss": 1.5169,
"step": 1216
},
{
"epoch": 1.4403669724770642,
"grad_norm": 5.71875,
"learning_rate": 1.7914612165493483e-05,
"loss": 1.4729,
"step": 1217
},
{
"epoch": 1.4415507546611424,
"grad_norm": 5.375,
"learning_rate": 1.7909553633412068e-05,
"loss": 1.0991,
"step": 1218
},
{
"epoch": 1.4427345368452205,
"grad_norm": 6.28125,
"learning_rate": 1.79044896895302e-05,
"loss": 1.3705,
"step": 1219
},
{
"epoch": 1.4439183190292986,
"grad_norm": 5.15625,
"learning_rate": 1.7899420337312675e-05,
"loss": 1.2086,
"step": 1220
},
{
"epoch": 1.4451021012133767,
"grad_norm": 6.5625,
"learning_rate": 1.7894345580228e-05,
"loss": 1.3769,
"step": 1221
},
{
"epoch": 1.4462858833974548,
"grad_norm": 5.15625,
"learning_rate": 1.788926542174838e-05,
"loss": 0.9948,
"step": 1222
},
{
"epoch": 1.447469665581533,
"grad_norm": 6.78125,
"learning_rate": 1.7884179865349713e-05,
"loss": 1.9357,
"step": 1223
},
{
"epoch": 1.448653447765611,
"grad_norm": 5.9375,
"learning_rate": 1.7879088914511583e-05,
"loss": 1.3896,
"step": 1224
},
{
"epoch": 1.4498372299496893,
"grad_norm": 7.78125,
"learning_rate": 1.7873992572717282e-05,
"loss": 1.6592,
"step": 1225
},
{
"epoch": 1.4510210121337674,
"grad_norm": 6.03125,
"learning_rate": 1.7868890843453776e-05,
"loss": 1.4291,
"step": 1226
},
{
"epoch": 1.4522047943178455,
"grad_norm": 5.8125,
"learning_rate": 1.7863783730211722e-05,
"loss": 1.3109,
"step": 1227
},
{
"epoch": 1.4533885765019237,
"grad_norm": 5.375,
"learning_rate": 1.7858671236485466e-05,
"loss": 1.5458,
"step": 1228
},
{
"epoch": 1.4545723586860018,
"grad_norm": 6.34375,
"learning_rate": 1.7853553365773027e-05,
"loss": 1.4185,
"step": 1229
},
{
"epoch": 1.4557561408700799,
"grad_norm": 5.59375,
"learning_rate": 1.7848430121576105e-05,
"loss": 1.1607,
"step": 1230
},
{
"epoch": 1.456939923054158,
"grad_norm": 5.71875,
"learning_rate": 1.784330150740008e-05,
"loss": 1.4561,
"step": 1231
},
{
"epoch": 1.458123705238236,
"grad_norm": 5.28125,
"learning_rate": 1.7838167526754002e-05,
"loss": 1.1164,
"step": 1232
},
{
"epoch": 1.4593074874223144,
"grad_norm": 5.65625,
"learning_rate": 1.7833028183150597e-05,
"loss": 1.4132,
"step": 1233
},
{
"epoch": 1.4604912696063925,
"grad_norm": 6.5,
"learning_rate": 1.782788348010626e-05,
"loss": 1.9292,
"step": 1234
},
{
"epoch": 1.4616750517904706,
"grad_norm": 6.3125,
"learning_rate": 1.7822733421141046e-05,
"loss": 1.1536,
"step": 1235
},
{
"epoch": 1.4628588339745487,
"grad_norm": 5.25,
"learning_rate": 1.7817578009778687e-05,
"loss": 1.0918,
"step": 1236
},
{
"epoch": 1.4640426161586269,
"grad_norm": 6.21875,
"learning_rate": 1.7812417249546564e-05,
"loss": 1.6518,
"step": 1237
},
{
"epoch": 1.465226398342705,
"grad_norm": 5.8125,
"learning_rate": 1.7807251143975727e-05,
"loss": 1.2441,
"step": 1238
},
{
"epoch": 1.466410180526783,
"grad_norm": 6.03125,
"learning_rate": 1.780207969660088e-05,
"loss": 1.1598,
"step": 1239
},
{
"epoch": 1.4675939627108612,
"grad_norm": 5.6875,
"learning_rate": 1.779690291096038e-05,
"loss": 1.3816,
"step": 1240
},
{
"epoch": 1.4687777448949393,
"grad_norm": 6.59375,
"learning_rate": 1.7791720790596242e-05,
"loss": 1.7059,
"step": 1241
},
{
"epoch": 1.4699615270790174,
"grad_norm": 6.4375,
"learning_rate": 1.7786533339054126e-05,
"loss": 1.4855,
"step": 1242
},
{
"epoch": 1.4711453092630955,
"grad_norm": 6.6875,
"learning_rate": 1.778134055988334e-05,
"loss": 1.683,
"step": 1243
},
{
"epoch": 1.4723290914471736,
"grad_norm": 5.15625,
"learning_rate": 1.7776142456636843e-05,
"loss": 0.9064,
"step": 1244
},
{
"epoch": 1.4735128736312517,
"grad_norm": 5.625,
"learning_rate": 1.7770939032871227e-05,
"loss": 1.155,
"step": 1245
},
{
"epoch": 1.47469665581533,
"grad_norm": 6.0,
"learning_rate": 1.776573029214673e-05,
"loss": 1.415,
"step": 1246
},
{
"epoch": 1.4758804379994082,
"grad_norm": 5.0,
"learning_rate": 1.7760516238027233e-05,
"loss": 0.932,
"step": 1247
},
{
"epoch": 1.4770642201834863,
"grad_norm": 5.15625,
"learning_rate": 1.7755296874080242e-05,
"loss": 1.2457,
"step": 1248
},
{
"epoch": 1.4782480023675644,
"grad_norm": 3.46875,
"learning_rate": 1.7750072203876905e-05,
"loss": 0.648,
"step": 1249
},
{
"epoch": 1.4794317845516425,
"grad_norm": 5.46875,
"learning_rate": 1.774484223099199e-05,
"loss": 1.185,
"step": 1250
},
{
"epoch": 1.4806155667357206,
"grad_norm": 4.90625,
"learning_rate": 1.773960695900391e-05,
"loss": 0.9739,
"step": 1251
},
{
"epoch": 1.4817993489197987,
"grad_norm": 7.0625,
"learning_rate": 1.7734366391494686e-05,
"loss": 1.4274,
"step": 1252
},
{
"epoch": 1.4829831311038768,
"grad_norm": 5.5625,
"learning_rate": 1.7729120532049967e-05,
"loss": 1.146,
"step": 1253
},
{
"epoch": 1.4841669132879551,
"grad_norm": 5.15625,
"learning_rate": 1.7723869384259038e-05,
"loss": 1.0387,
"step": 1254
},
{
"epoch": 1.4853506954720332,
"grad_norm": 4.625,
"learning_rate": 1.771861295171478e-05,
"loss": 0.9034,
"step": 1255
},
{
"epoch": 1.4865344776561114,
"grad_norm": 4.9375,
"learning_rate": 1.77133512380137e-05,
"loss": 0.8538,
"step": 1256
},
{
"epoch": 1.4877182598401895,
"grad_norm": 5.09375,
"learning_rate": 1.7708084246755924e-05,
"loss": 0.8724,
"step": 1257
},
{
"epoch": 1.4889020420242676,
"grad_norm": 5.1875,
"learning_rate": 1.7702811981545177e-05,
"loss": 0.9999,
"step": 1258
},
{
"epoch": 1.4900858242083457,
"grad_norm": 6.0,
"learning_rate": 1.7697534445988804e-05,
"loss": 1.5709,
"step": 1259
},
{
"epoch": 1.4912696063924238,
"grad_norm": 5.125,
"learning_rate": 1.7692251643697748e-05,
"loss": 1.0102,
"step": 1260
},
{
"epoch": 1.492453388576502,
"grad_norm": 6.125,
"learning_rate": 1.768696357828656e-05,
"loss": 1.0263,
"step": 1261
},
{
"epoch": 1.49363717076058,
"grad_norm": 6.125,
"learning_rate": 1.7681670253373387e-05,
"loss": 1.1874,
"step": 1262
},
{
"epoch": 1.494820952944658,
"grad_norm": 4.1875,
"learning_rate": 1.7676371672579983e-05,
"loss": 0.7004,
"step": 1263
},
{
"epoch": 1.4960047351287362,
"grad_norm": 6.0625,
"learning_rate": 1.767106783953169e-05,
"loss": 1.7209,
"step": 1264
},
{
"epoch": 1.4971885173128143,
"grad_norm": 4.9375,
"learning_rate": 1.7665758757857445e-05,
"loss": 1.0497,
"step": 1265
},
{
"epoch": 1.4983722994968924,
"grad_norm": 4.78125,
"learning_rate": 1.766044443118978e-05,
"loss": 0.8903,
"step": 1266
},
{
"epoch": 1.4995560816809708,
"grad_norm": 6.0625,
"learning_rate": 1.7655124863164817e-05,
"loss": 1.2987,
"step": 1267
},
{
"epoch": 1.5007398638650489,
"grad_norm": 6.6875,
"learning_rate": 1.7649800057422256e-05,
"loss": 1.8082,
"step": 1268
},
{
"epoch": 1.501923646049127,
"grad_norm": 6.25,
"learning_rate": 1.764447001760539e-05,
"loss": 1.0985,
"step": 1269
},
{
"epoch": 1.503107428233205,
"grad_norm": 5.09375,
"learning_rate": 1.7639134747361084e-05,
"loss": 1.051,
"step": 1270
},
{
"epoch": 1.5042912104172832,
"grad_norm": 5.125,
"learning_rate": 1.763379425033979e-05,
"loss": 0.8763,
"step": 1271
},
{
"epoch": 1.5054749926013613,
"grad_norm": 5.90625,
"learning_rate": 1.7628448530195527e-05,
"loss": 1.4842,
"step": 1272
},
{
"epoch": 1.5066587747854396,
"grad_norm": 5.5625,
"learning_rate": 1.7623097590585902e-05,
"loss": 1.2695,
"step": 1273
},
{
"epoch": 1.5078425569695177,
"grad_norm": 5.21875,
"learning_rate": 1.761774143517208e-05,
"loss": 1.1286,
"step": 1274
},
{
"epoch": 1.5090263391535959,
"grad_norm": 4.96875,
"learning_rate": 1.7612380067618796e-05,
"loss": 0.9574,
"step": 1275
},
{
"epoch": 1.510210121337674,
"grad_norm": 5.0625,
"learning_rate": 1.7607013491594363e-05,
"loss": 1.2294,
"step": 1276
},
{
"epoch": 1.511393903521752,
"grad_norm": 6.625,
"learning_rate": 1.760164171077064e-05,
"loss": 1.3758,
"step": 1277
},
{
"epoch": 1.5125776857058302,
"grad_norm": 5.90625,
"learning_rate": 1.7596264728823063e-05,
"loss": 1.031,
"step": 1278
},
{
"epoch": 1.5137614678899083,
"grad_norm": 5.5,
"learning_rate": 1.7590882549430617e-05,
"loss": 1.3964,
"step": 1279
},
{
"epoch": 1.5149452500739864,
"grad_norm": 6.28125,
"learning_rate": 1.758549517627585e-05,
"loss": 1.2461,
"step": 1280
},
{
"epoch": 1.5161290322580645,
"grad_norm": 6.5625,
"learning_rate": 1.7580102613044853e-05,
"loss": 1.8333,
"step": 1281
},
{
"epoch": 1.5173128144421426,
"grad_norm": 6.375,
"learning_rate": 1.7574704863427277e-05,
"loss": 1.3218,
"step": 1282
},
{
"epoch": 1.5184965966262207,
"grad_norm": 6.15625,
"learning_rate": 1.7569301931116326e-05,
"loss": 1.1875,
"step": 1283
},
{
"epoch": 1.5196803788102988,
"grad_norm": 6.6875,
"learning_rate": 1.7563893819808736e-05,
"loss": 1.3062,
"step": 1284
},
{
"epoch": 1.520864160994377,
"grad_norm": 5.875,
"learning_rate": 1.75584805332048e-05,
"loss": 1.467,
"step": 1285
},
{
"epoch": 1.522047943178455,
"grad_norm": 4.78125,
"learning_rate": 1.755306207500834e-05,
"loss": 1.2971,
"step": 1286
},
{
"epoch": 1.5232317253625332,
"grad_norm": 6.46875,
"learning_rate": 1.754763844892673e-05,
"loss": 1.3452,
"step": 1287
},
{
"epoch": 1.5244155075466113,
"grad_norm": 6.34375,
"learning_rate": 1.7542209658670862e-05,
"loss": 1.0722,
"step": 1288
},
{
"epoch": 1.5255992897306896,
"grad_norm": 5.8125,
"learning_rate": 1.7536775707955182e-05,
"loss": 1.2578,
"step": 1289
},
{
"epoch": 1.5267830719147677,
"grad_norm": 5.4375,
"learning_rate": 1.753133660049765e-05,
"loss": 1.2805,
"step": 1290
},
{
"epoch": 1.5279668540988458,
"grad_norm": 6.8125,
"learning_rate": 1.7525892340019766e-05,
"loss": 1.3532,
"step": 1291
},
{
"epoch": 1.529150636282924,
"grad_norm": 6.09375,
"learning_rate": 1.7520442930246546e-05,
"loss": 1.1744,
"step": 1292
},
{
"epoch": 1.530334418467002,
"grad_norm": 6.0625,
"learning_rate": 1.7514988374906535e-05,
"loss": 1.4194,
"step": 1293
},
{
"epoch": 1.5315182006510804,
"grad_norm": 6.125,
"learning_rate": 1.7509528677731806e-05,
"loss": 1.3089,
"step": 1294
},
{
"epoch": 1.5327019828351585,
"grad_norm": 5.71875,
"learning_rate": 1.750406384245793e-05,
"loss": 1.1319,
"step": 1295
},
{
"epoch": 1.5338857650192366,
"grad_norm": 5.53125,
"learning_rate": 1.749859387282401e-05,
"loss": 0.9954,
"step": 1296
},
{
"epoch": 1.5350695472033147,
"grad_norm": 6.21875,
"learning_rate": 1.7493118772572655e-05,
"loss": 1.5434,
"step": 1297
},
{
"epoch": 1.5362533293873928,
"grad_norm": 5.5,
"learning_rate": 1.7487638545449993e-05,
"loss": 1.1891,
"step": 1298
},
{
"epoch": 1.537437111571471,
"grad_norm": 5.1875,
"learning_rate": 1.7482153195205646e-05,
"loss": 1.1625,
"step": 1299
},
{
"epoch": 1.538620893755549,
"grad_norm": 5.96875,
"learning_rate": 1.747666272559275e-05,
"loss": 1.4001,
"step": 1300
},
{
"epoch": 1.539804675939627,
"grad_norm": 4.90625,
"learning_rate": 1.7471167140367948e-05,
"loss": 0.8876,
"step": 1301
},
{
"epoch": 1.5409884581237052,
"grad_norm": 5.46875,
"learning_rate": 1.7465666443291374e-05,
"loss": 1.1814,
"step": 1302
},
{
"epoch": 1.5421722403077833,
"grad_norm": 5.59375,
"learning_rate": 1.7460160638126663e-05,
"loss": 1.06,
"step": 1303
},
{
"epoch": 1.5433560224918614,
"grad_norm": 5.1875,
"learning_rate": 1.7454649728640944e-05,
"loss": 0.8666,
"step": 1304
},
{
"epoch": 1.5445398046759395,
"grad_norm": 5.53125,
"learning_rate": 1.7449133718604845e-05,
"loss": 1.1397,
"step": 1305
},
{
"epoch": 1.5457235868600177,
"grad_norm": 5.53125,
"learning_rate": 1.7443612611792473e-05,
"loss": 0.9162,
"step": 1306
},
{
"epoch": 1.5469073690440958,
"grad_norm": 4.0625,
"learning_rate": 1.743808641198143e-05,
"loss": 0.698,
"step": 1307
},
{
"epoch": 1.5480911512281739,
"grad_norm": 5.21875,
"learning_rate": 1.74325551229528e-05,
"loss": 1.0317,
"step": 1308
},
{
"epoch": 1.5492749334122522,
"grad_norm": 5.46875,
"learning_rate": 1.742701874849115e-05,
"loss": 1.0891,
"step": 1309
},
{
"epoch": 1.5504587155963303,
"grad_norm": 6.375,
"learning_rate": 1.7421477292384524e-05,
"loss": 1.5187,
"step": 1310
},
{
"epoch": 1.5516424977804084,
"grad_norm": 5.90625,
"learning_rate": 1.741593075842445e-05,
"loss": 1.3926,
"step": 1311
},
{
"epoch": 1.5528262799644865,
"grad_norm": 5.15625,
"learning_rate": 1.7410379150405924e-05,
"loss": 1.1211,
"step": 1312
},
{
"epoch": 1.5540100621485646,
"grad_norm": 4.4375,
"learning_rate": 1.7404822472127406e-05,
"loss": 0.7486,
"step": 1313
},
{
"epoch": 1.5551938443326427,
"grad_norm": 5.75,
"learning_rate": 1.7399260727390846e-05,
"loss": 1.3007,
"step": 1314
},
{
"epoch": 1.556377626516721,
"grad_norm": 4.90625,
"learning_rate": 1.7393693920001647e-05,
"loss": 0.9171,
"step": 1315
},
{
"epoch": 1.5575614087007992,
"grad_norm": 4.71875,
"learning_rate": 1.7388122053768673e-05,
"loss": 1.0204,
"step": 1316
},
{
"epoch": 1.5587451908848773,
"grad_norm": 5.25,
"learning_rate": 1.7382545132504255e-05,
"loss": 1.0157,
"step": 1317
},
{
"epoch": 1.5599289730689554,
"grad_norm": 6.46875,
"learning_rate": 1.7376963160024184e-05,
"loss": 1.2015,
"step": 1318
},
{
"epoch": 1.5611127552530335,
"grad_norm": 5.71875,
"learning_rate": 1.7371376140147705e-05,
"loss": 1.1851,
"step": 1319
},
{
"epoch": 1.5622965374371116,
"grad_norm": 5.09375,
"learning_rate": 1.736578407669751e-05,
"loss": 0.8418,
"step": 1320
},
{
"epoch": 1.5634803196211897,
"grad_norm": 6.09375,
"learning_rate": 1.7360186973499752e-05,
"loss": 1.2245,
"step": 1321
},
{
"epoch": 1.5646641018052678,
"grad_norm": 5.125,
"learning_rate": 1.7354584834384036e-05,
"loss": 1.0012,
"step": 1322
},
{
"epoch": 1.565847883989346,
"grad_norm": 6.375,
"learning_rate": 1.7348977663183392e-05,
"loss": 1.3556,
"step": 1323
},
{
"epoch": 1.567031666173424,
"grad_norm": 5.59375,
"learning_rate": 1.7343365463734314e-05,
"loss": 1.35,
"step": 1324
},
{
"epoch": 1.5682154483575022,
"grad_norm": 5.8125,
"learning_rate": 1.733774823987673e-05,
"loss": 1.082,
"step": 1325
},
{
"epoch": 1.5693992305415803,
"grad_norm": 6.6875,
"learning_rate": 1.7332125995453992e-05,
"loss": 1.446,
"step": 1326
},
{
"epoch": 1.5705830127256584,
"grad_norm": 5.875,
"learning_rate": 1.732649873431291e-05,
"loss": 1.1454,
"step": 1327
},
{
"epoch": 1.5717667949097365,
"grad_norm": 5.4375,
"learning_rate": 1.732086646030372e-05,
"loss": 1.1108,
"step": 1328
},
{
"epoch": 1.5729505770938146,
"grad_norm": 5.46875,
"learning_rate": 1.7315229177280075e-05,
"loss": 1.247,
"step": 1329
},
{
"epoch": 1.574134359277893,
"grad_norm": 5.375,
"learning_rate": 1.7309586889099063e-05,
"loss": 1.4245,
"step": 1330
},
{
"epoch": 1.575318141461971,
"grad_norm": 5.78125,
"learning_rate": 1.73039395996212e-05,
"loss": 1.2274,
"step": 1331
},
{
"epoch": 1.5765019236460491,
"grad_norm": 6.21875,
"learning_rate": 1.7298287312710423e-05,
"loss": 1.2951,
"step": 1332
},
{
"epoch": 1.5776857058301272,
"grad_norm": 5.71875,
"learning_rate": 1.7292630032234086e-05,
"loss": 1.3365,
"step": 1333
},
{
"epoch": 1.5788694880142053,
"grad_norm": 5.8125,
"learning_rate": 1.7286967762062958e-05,
"loss": 1.1987,
"step": 1334
},
{
"epoch": 1.5800532701982837,
"grad_norm": 4.96875,
"learning_rate": 1.728130050607123e-05,
"loss": 0.8776,
"step": 1335
},
{
"epoch": 1.5812370523823618,
"grad_norm": 6.03125,
"learning_rate": 1.7275628268136487e-05,
"loss": 1.3505,
"step": 1336
},
{
"epoch": 1.58242083456644,
"grad_norm": 5.71875,
"learning_rate": 1.7269951052139745e-05,
"loss": 1.2599,
"step": 1337
},
{
"epoch": 1.583604616750518,
"grad_norm": 4.9375,
"learning_rate": 1.7264268861965416e-05,
"loss": 1.062,
"step": 1338
},
{
"epoch": 1.584788398934596,
"grad_norm": 4.78125,
"learning_rate": 1.7258581701501305e-05,
"loss": 0.9316,
"step": 1339
},
{
"epoch": 1.5859721811186742,
"grad_norm": 5.3125,
"learning_rate": 1.725288957463864e-05,
"loss": 1.2045,
"step": 1340
},
{
"epoch": 1.5871559633027523,
"grad_norm": 5.28125,
"learning_rate": 1.724719248527202e-05,
"loss": 0.9913,
"step": 1341
},
{
"epoch": 1.5883397454868304,
"grad_norm": 6.15625,
"learning_rate": 1.724149043729947e-05,
"loss": 0.9455,
"step": 1342
},
{
"epoch": 1.5895235276709085,
"grad_norm": 5.59375,
"learning_rate": 1.723578343462238e-05,
"loss": 1.1195,
"step": 1343
},
{
"epoch": 1.5907073098549867,
"grad_norm": 6.09375,
"learning_rate": 1.7230071481145546e-05,
"loss": 1.1511,
"step": 1344
},
{
"epoch": 1.5918910920390648,
"grad_norm": 6.34375,
"learning_rate": 1.7224354580777144e-05,
"loss": 1.5532,
"step": 1345
},
{
"epoch": 1.5930748742231429,
"grad_norm": 6.34375,
"learning_rate": 1.7218632737428743e-05,
"loss": 1.6035,
"step": 1346
},
{
"epoch": 1.594258656407221,
"grad_norm": 6.4375,
"learning_rate": 1.7212905955015285e-05,
"loss": 1.1392,
"step": 1347
},
{
"epoch": 1.595442438591299,
"grad_norm": 5.28125,
"learning_rate": 1.7207174237455095e-05,
"loss": 0.9658,
"step": 1348
},
{
"epoch": 1.5966262207753772,
"grad_norm": 5.84375,
"learning_rate": 1.720143758866988e-05,
"loss": 1.4291,
"step": 1349
},
{
"epoch": 1.5978100029594553,
"grad_norm": 6.03125,
"learning_rate": 1.719569601258471e-05,
"loss": 1.3934,
"step": 1350
},
{
"epoch": 1.5989937851435336,
"grad_norm": 5.375,
"learning_rate": 1.7189949513128033e-05,
"loss": 1.0505,
"step": 1351
},
{
"epoch": 1.6001775673276117,
"grad_norm": 5.0625,
"learning_rate": 1.718419809423167e-05,
"loss": 0.9919,
"step": 1352
},
{
"epoch": 1.6013613495116898,
"grad_norm": 5.46875,
"learning_rate": 1.7178441759830797e-05,
"loss": 1.3151,
"step": 1353
},
{
"epoch": 1.602545131695768,
"grad_norm": 7.40625,
"learning_rate": 1.717268051386396e-05,
"loss": 1.5176,
"step": 1354
},
{
"epoch": 1.603728913879846,
"grad_norm": 6.6875,
"learning_rate": 1.716691436027307e-05,
"loss": 1.3464,
"step": 1355
},
{
"epoch": 1.6049126960639244,
"grad_norm": 6.1875,
"learning_rate": 1.7161143303003382e-05,
"loss": 1.5177,
"step": 1356
},
{
"epoch": 1.6060964782480025,
"grad_norm": 5.34375,
"learning_rate": 1.715536734600352e-05,
"loss": 0.8949,
"step": 1357
},
{
"epoch": 1.6072802604320806,
"grad_norm": 5.65625,
"learning_rate": 1.7149586493225453e-05,
"loss": 0.9173,
"step": 1358
},
{
"epoch": 1.6084640426161587,
"grad_norm": 5.90625,
"learning_rate": 1.7143800748624507e-05,
"loss": 1.372,
"step": 1359
},
{
"epoch": 1.6096478248002368,
"grad_norm": 5.34375,
"learning_rate": 1.7138010116159342e-05,
"loss": 1.1767,
"step": 1360
},
{
"epoch": 1.610831606984315,
"grad_norm": 5.9375,
"learning_rate": 1.7132214599791975e-05,
"loss": 1.4807,
"step": 1361
},
{
"epoch": 1.612015389168393,
"grad_norm": 6.15625,
"learning_rate": 1.7126414203487757e-05,
"loss": 1.6726,
"step": 1362
},
{
"epoch": 1.6131991713524712,
"grad_norm": 6.375,
"learning_rate": 1.712060893121538e-05,
"loss": 1.3641,
"step": 1363
},
{
"epoch": 1.6143829535365493,
"grad_norm": 5.5625,
"learning_rate": 1.7114798786946875e-05,
"loss": 1.0601,
"step": 1364
},
{
"epoch": 1.6155667357206274,
"grad_norm": 6.15625,
"learning_rate": 1.710898377465761e-05,
"loss": 1.2894,
"step": 1365
},
{
"epoch": 1.6167505179047055,
"grad_norm": 5.96875,
"learning_rate": 1.7103163898326265e-05,
"loss": 1.6396,
"step": 1366
},
{
"epoch": 1.6179343000887836,
"grad_norm": 6.53125,
"learning_rate": 1.709733916193487e-05,
"loss": 1.2183,
"step": 1367
},
{
"epoch": 1.6191180822728617,
"grad_norm": 5.28125,
"learning_rate": 1.709150956946877e-05,
"loss": 1.2261,
"step": 1368
},
{
"epoch": 1.6203018644569398,
"grad_norm": 6.4375,
"learning_rate": 1.7085675124916638e-05,
"loss": 1.3444,
"step": 1369
},
{
"epoch": 1.621485646641018,
"grad_norm": 5.28125,
"learning_rate": 1.7079835832270455e-05,
"loss": 1.2922,
"step": 1370
},
{
"epoch": 1.622669428825096,
"grad_norm": 5.4375,
"learning_rate": 1.7073991695525528e-05,
"loss": 1.1105,
"step": 1371
},
{
"epoch": 1.6238532110091743,
"grad_norm": 6.0625,
"learning_rate": 1.706814271868048e-05,
"loss": 1.1449,
"step": 1372
},
{
"epoch": 1.6250369931932525,
"grad_norm": 5.1875,
"learning_rate": 1.7062288905737245e-05,
"loss": 1.0403,
"step": 1373
},
{
"epoch": 1.6262207753773306,
"grad_norm": 5.5,
"learning_rate": 1.705643026070106e-05,
"loss": 1.6333,
"step": 1374
},
{
"epoch": 1.6274045575614087,
"grad_norm": 4.875,
"learning_rate": 1.7050566787580477e-05,
"loss": 0.8834,
"step": 1375
},
{
"epoch": 1.6285883397454868,
"grad_norm": 5.46875,
"learning_rate": 1.704469849038734e-05,
"loss": 1.1659,
"step": 1376
},
{
"epoch": 1.629772121929565,
"grad_norm": 5.96875,
"learning_rate": 1.7038825373136806e-05,
"loss": 1.2523,
"step": 1377
},
{
"epoch": 1.6309559041136432,
"grad_norm": 6.09375,
"learning_rate": 1.7032947439847316e-05,
"loss": 1.323,
"step": 1378
},
{
"epoch": 1.6321396862977213,
"grad_norm": 5.59375,
"learning_rate": 1.7027064694540623e-05,
"loss": 1.264,
"step": 1379
},
{
"epoch": 1.6333234684817994,
"grad_norm": 5.46875,
"learning_rate": 1.7021177141241757e-05,
"loss": 0.9082,
"step": 1380
},
{
"epoch": 1.6345072506658775,
"grad_norm": 6.71875,
"learning_rate": 1.701528478397905e-05,
"loss": 1.4211,
"step": 1381
},
{
"epoch": 1.6356910328499557,
"grad_norm": 5.5,
"learning_rate": 1.7009387626784118e-05,
"loss": 0.8781,
"step": 1382
},
{
"epoch": 1.6368748150340338,
"grad_norm": 5.96875,
"learning_rate": 1.7003485673691845e-05,
"loss": 1.099,
"step": 1383
},
{
"epoch": 1.6380585972181119,
"grad_norm": 4.875,
"learning_rate": 1.6997578928740423e-05,
"loss": 0.8661,
"step": 1384
},
{
"epoch": 1.63924237940219,
"grad_norm": 5.84375,
"learning_rate": 1.6991667395971306e-05,
"loss": 1.081,
"step": 1385
},
{
"epoch": 1.640426161586268,
"grad_norm": 6.15625,
"learning_rate": 1.6985751079429226e-05,
"loss": 1.7194,
"step": 1386
},
{
"epoch": 1.6416099437703462,
"grad_norm": 6.25,
"learning_rate": 1.697982998316219e-05,
"loss": 1.3778,
"step": 1387
},
{
"epoch": 1.6427937259544243,
"grad_norm": 5.65625,
"learning_rate": 1.697390411122147e-05,
"loss": 1.5055,
"step": 1388
},
{
"epoch": 1.6439775081385024,
"grad_norm": 5.5,
"learning_rate": 1.696797346766162e-05,
"loss": 1.2076,
"step": 1389
},
{
"epoch": 1.6451612903225805,
"grad_norm": 5.09375,
"learning_rate": 1.696203805654044e-05,
"loss": 1.2671,
"step": 1390
},
{
"epoch": 1.6463450725066586,
"grad_norm": 5.625,
"learning_rate": 1.6956097881919006e-05,
"loss": 1.331,
"step": 1391
},
{
"epoch": 1.647528854690737,
"grad_norm": 5.875,
"learning_rate": 1.695015294786165e-05,
"loss": 1.2928,
"step": 1392
},
{
"epoch": 1.648712636874815,
"grad_norm": 5.46875,
"learning_rate": 1.6944203258435957e-05,
"loss": 1.0058,
"step": 1393
},
{
"epoch": 1.6498964190588932,
"grad_norm": 5.75,
"learning_rate": 1.6938248817712767e-05,
"loss": 1.0551,
"step": 1394
},
{
"epoch": 1.6510802012429713,
"grad_norm": 5.15625,
"learning_rate": 1.693228962976617e-05,
"loss": 1.2421,
"step": 1395
},
{
"epoch": 1.6522639834270494,
"grad_norm": 4.625,
"learning_rate": 1.6926325698673513e-05,
"loss": 1.026,
"step": 1396
},
{
"epoch": 1.6534477656111277,
"grad_norm": 6.65625,
"learning_rate": 1.6920357028515368e-05,
"loss": 1.9762,
"step": 1397
},
{
"epoch": 1.6546315477952058,
"grad_norm": 7.34375,
"learning_rate": 1.6914383623375577e-05,
"loss": 1.5981,
"step": 1398
},
{
"epoch": 1.655815329979284,
"grad_norm": 5.96875,
"learning_rate": 1.6908405487341195e-05,
"loss": 1.4243,
"step": 1399
},
{
"epoch": 1.656999112163362,
"grad_norm": 6.28125,
"learning_rate": 1.690242262450253e-05,
"loss": 1.3202,
"step": 1400
},
{
"epoch": 1.6581828943474402,
"grad_norm": 6.0,
"learning_rate": 1.6896435038953126e-05,
"loss": 1.3387,
"step": 1401
},
{
"epoch": 1.6593666765315183,
"grad_norm": 6.375,
"learning_rate": 1.6890442734789743e-05,
"loss": 1.2336,
"step": 1402
},
{
"epoch": 1.6605504587155964,
"grad_norm": 5.5625,
"learning_rate": 1.6884445716112388e-05,
"loss": 1.0991,
"step": 1403
},
{
"epoch": 1.6617342408996745,
"grad_norm": 6.125,
"learning_rate": 1.6878443987024276e-05,
"loss": 1.1503,
"step": 1404
},
{
"epoch": 1.6629180230837526,
"grad_norm": 7.125,
"learning_rate": 1.6872437551631863e-05,
"loss": 1.5255,
"step": 1405
},
{
"epoch": 1.6641018052678307,
"grad_norm": 5.6875,
"learning_rate": 1.686642641404481e-05,
"loss": 1.4242,
"step": 1406
},
{
"epoch": 1.6652855874519088,
"grad_norm": 6.34375,
"learning_rate": 1.6860410578375997e-05,
"loss": 1.7046,
"step": 1407
},
{
"epoch": 1.666469369635987,
"grad_norm": 6.96875,
"learning_rate": 1.685439004874153e-05,
"loss": 1.3059,
"step": 1408
},
{
"epoch": 1.667653151820065,
"grad_norm": 5.53125,
"learning_rate": 1.684836482926072e-05,
"loss": 1.2994,
"step": 1409
},
{
"epoch": 1.6688369340041431,
"grad_norm": 4.84375,
"learning_rate": 1.6842334924056082e-05,
"loss": 0.8497,
"step": 1410
},
{
"epoch": 1.6700207161882212,
"grad_norm": 5.59375,
"learning_rate": 1.6836300337253337e-05,
"loss": 1.1749,
"step": 1411
},
{
"epoch": 1.6712044983722993,
"grad_norm": 5.5625,
"learning_rate": 1.6830261072981423e-05,
"loss": 1.2342,
"step": 1412
},
{
"epoch": 1.6723882805563777,
"grad_norm": 5.46875,
"learning_rate": 1.682421713537246e-05,
"loss": 1.0712,
"step": 1413
},
{
"epoch": 1.6735720627404558,
"grad_norm": 7.875,
"learning_rate": 1.6818168528561783e-05,
"loss": 1.5969,
"step": 1414
},
{
"epoch": 1.6747558449245339,
"grad_norm": 6.90625,
"learning_rate": 1.6812115256687906e-05,
"loss": 1.5111,
"step": 1415
},
{
"epoch": 1.675939627108612,
"grad_norm": 6.28125,
"learning_rate": 1.6806057323892542e-05,
"loss": 1.4562,
"step": 1416
},
{
"epoch": 1.67712340929269,
"grad_norm": 5.875,
"learning_rate": 1.6799994734320596e-05,
"loss": 0.8361,
"step": 1417
},
{
"epoch": 1.6783071914767684,
"grad_norm": 7.125,
"learning_rate": 1.679392749212015e-05,
"loss": 1.3219,
"step": 1418
},
{
"epoch": 1.6794909736608465,
"grad_norm": 6.0,
"learning_rate": 1.678785560144249e-05,
"loss": 1.1576,
"step": 1419
},
{
"epoch": 1.6806747558449247,
"grad_norm": 5.75,
"learning_rate": 1.678177906644205e-05,
"loss": 1.0903,
"step": 1420
},
{
"epoch": 1.6818585380290028,
"grad_norm": 5.71875,
"learning_rate": 1.677569789127647e-05,
"loss": 1.2878,
"step": 1421
},
{
"epoch": 1.6830423202130809,
"grad_norm": 5.5,
"learning_rate": 1.6769612080106554e-05,
"loss": 1.1293,
"step": 1422
},
{
"epoch": 1.684226102397159,
"grad_norm": 5.875,
"learning_rate": 1.6763521637096272e-05,
"loss": 1.122,
"step": 1423
},
{
"epoch": 1.685409884581237,
"grad_norm": 7.40625,
"learning_rate": 1.675742656641278e-05,
"loss": 1.4902,
"step": 1424
},
{
"epoch": 1.6865936667653152,
"grad_norm": 5.5625,
"learning_rate": 1.675132687222638e-05,
"loss": 1.0147,
"step": 1425
},
{
"epoch": 1.6877774489493933,
"grad_norm": 5.25,
"learning_rate": 1.6745222558710556e-05,
"loss": 1.0662,
"step": 1426
},
{
"epoch": 1.6889612311334714,
"grad_norm": 5.3125,
"learning_rate": 1.6739113630041937e-05,
"loss": 1.2297,
"step": 1427
},
{
"epoch": 1.6901450133175495,
"grad_norm": 6.65625,
"learning_rate": 1.673300009040032e-05,
"loss": 1.6411,
"step": 1428
},
{
"epoch": 1.6913287955016276,
"grad_norm": 6.6875,
"learning_rate": 1.6726881943968653e-05,
"loss": 1.6398,
"step": 1429
},
{
"epoch": 1.6925125776857057,
"grad_norm": 5.125,
"learning_rate": 1.6720759194933037e-05,
"loss": 1.1018,
"step": 1430
},
{
"epoch": 1.6936963598697838,
"grad_norm": 5.71875,
"learning_rate": 1.6714631847482724e-05,
"loss": 0.8976,
"step": 1431
},
{
"epoch": 1.694880142053862,
"grad_norm": 7.34375,
"learning_rate": 1.6708499905810107e-05,
"loss": 1.8453,
"step": 1432
},
{
"epoch": 1.69606392423794,
"grad_norm": 6.3125,
"learning_rate": 1.6702363374110726e-05,
"loss": 1.1571,
"step": 1433
},
{
"epoch": 1.6972477064220184,
"grad_norm": 6.21875,
"learning_rate": 1.669622225658326e-05,
"loss": 1.5562,
"step": 1434
},
{
"epoch": 1.6984314886060965,
"grad_norm": 6.40625,
"learning_rate": 1.669007655742953e-05,
"loss": 1.2154,
"step": 1435
},
{
"epoch": 1.6996152707901746,
"grad_norm": 5.71875,
"learning_rate": 1.6683926280854484e-05,
"loss": 1.1535,
"step": 1436
},
{
"epoch": 1.7007990529742527,
"grad_norm": 7.90625,
"learning_rate": 1.6677771431066213e-05,
"loss": 1.3267,
"step": 1437
},
{
"epoch": 1.7019828351583308,
"grad_norm": 6.125,
"learning_rate": 1.6671612012275924e-05,
"loss": 1.3567,
"step": 1438
},
{
"epoch": 1.7031666173424092,
"grad_norm": 7.09375,
"learning_rate": 1.666544802869796e-05,
"loss": 1.6159,
"step": 1439
},
{
"epoch": 1.7043503995264873,
"grad_norm": 6.375,
"learning_rate": 1.6659279484549782e-05,
"loss": 1.6842,
"step": 1440
},
{
"epoch": 1.7055341817105654,
"grad_norm": 6.65625,
"learning_rate": 1.6653106384051977e-05,
"loss": 1.6881,
"step": 1441
},
{
"epoch": 1.7067179638946435,
"grad_norm": 5.8125,
"learning_rate": 1.664692873142824e-05,
"loss": 1.0208,
"step": 1442
},
{
"epoch": 1.7079017460787216,
"grad_norm": 5.5,
"learning_rate": 1.6640746530905395e-05,
"loss": 0.9472,
"step": 1443
},
{
"epoch": 1.7090855282627997,
"grad_norm": 5.6875,
"learning_rate": 1.663455978671336e-05,
"loss": 1.1162,
"step": 1444
},
{
"epoch": 1.7102693104468778,
"grad_norm": 4.75,
"learning_rate": 1.6628368503085174e-05,
"loss": 0.9746,
"step": 1445
},
{
"epoch": 1.711453092630956,
"grad_norm": 5.3125,
"learning_rate": 1.6622172684256984e-05,
"loss": 0.9995,
"step": 1446
},
{
"epoch": 1.712636874815034,
"grad_norm": 5.8125,
"learning_rate": 1.6615972334468023e-05,
"loss": 1.192,
"step": 1447
},
{
"epoch": 1.7138206569991121,
"grad_norm": 6.90625,
"learning_rate": 1.660976745796065e-05,
"loss": 1.3989,
"step": 1448
},
{
"epoch": 1.7150044391831902,
"grad_norm": 6.5625,
"learning_rate": 1.6603558058980296e-05,
"loss": 1.5006,
"step": 1449
},
{
"epoch": 1.7161882213672683,
"grad_norm": 6.46875,
"learning_rate": 1.659734414177551e-05,
"loss": 1.7251,
"step": 1450
},
{
"epoch": 1.7173720035513464,
"grad_norm": 5.59375,
"learning_rate": 1.6591125710597908e-05,
"loss": 1.2869,
"step": 1451
},
{
"epoch": 1.7185557857354246,
"grad_norm": 5.96875,
"learning_rate": 1.6584902769702214e-05,
"loss": 1.3166,
"step": 1452
},
{
"epoch": 1.7197395679195027,
"grad_norm": 5.3125,
"learning_rate": 1.6578675323346224e-05,
"loss": 1.1939,
"step": 1453
},
{
"epoch": 1.720923350103581,
"grad_norm": 7.71875,
"learning_rate": 1.6572443375790824e-05,
"loss": 1.7101,
"step": 1454
},
{
"epoch": 1.722107132287659,
"grad_norm": 6.28125,
"learning_rate": 1.6566206931299987e-05,
"loss": 1.2044,
"step": 1455
},
{
"epoch": 1.7232909144717372,
"grad_norm": 6.5,
"learning_rate": 1.655996599414075e-05,
"loss": 1.5694,
"step": 1456
},
{
"epoch": 1.7244746966558153,
"grad_norm": 6.65625,
"learning_rate": 1.655372056858322e-05,
"loss": 1.7294,
"step": 1457
},
{
"epoch": 1.7256584788398934,
"grad_norm": 6.21875,
"learning_rate": 1.6547470658900592e-05,
"loss": 1.1696,
"step": 1458
},
{
"epoch": 1.7268422610239715,
"grad_norm": 5.75,
"learning_rate": 1.654121626936912e-05,
"loss": 1.383,
"step": 1459
},
{
"epoch": 1.7280260432080499,
"grad_norm": 6.03125,
"learning_rate": 1.653495740426812e-05,
"loss": 1.642,
"step": 1460
},
{
"epoch": 1.729209825392128,
"grad_norm": 6.375,
"learning_rate": 1.6528694067879977e-05,
"loss": 1.3902,
"step": 1461
},
{
"epoch": 1.730393607576206,
"grad_norm": 6.59375,
"learning_rate": 1.652242626449013e-05,
"loss": 1.0808,
"step": 1462
},
{
"epoch": 1.7315773897602842,
"grad_norm": 6.34375,
"learning_rate": 1.6516153998387077e-05,
"loss": 1.7502,
"step": 1463
},
{
"epoch": 1.7327611719443623,
"grad_norm": 5.5,
"learning_rate": 1.650987727386237e-05,
"loss": 0.8855,
"step": 1464
},
{
"epoch": 1.7339449541284404,
"grad_norm": 6.6875,
"learning_rate": 1.650359609521061e-05,
"loss": 1.6242,
"step": 1465
},
{
"epoch": 1.7351287363125185,
"grad_norm": 5.15625,
"learning_rate": 1.6497310466729448e-05,
"loss": 0.9131,
"step": 1466
},
{
"epoch": 1.7363125184965966,
"grad_norm": 6.375,
"learning_rate": 1.6491020392719575e-05,
"loss": 1.2403,
"step": 1467
},
{
"epoch": 1.7374963006806747,
"grad_norm": 5.96875,
"learning_rate": 1.648472587748473e-05,
"loss": 1.4171,
"step": 1468
},
{
"epoch": 1.7386800828647528,
"grad_norm": 5.875,
"learning_rate": 1.647842692533168e-05,
"loss": 1.1549,
"step": 1469
},
{
"epoch": 1.739863865048831,
"grad_norm": 5.875,
"learning_rate": 1.647212354057024e-05,
"loss": 1.1008,
"step": 1470
},
{
"epoch": 1.741047647232909,
"grad_norm": 5.125,
"learning_rate": 1.6465815727513253e-05,
"loss": 1.1147,
"step": 1471
},
{
"epoch": 1.7422314294169872,
"grad_norm": 7.34375,
"learning_rate": 1.645950349047659e-05,
"loss": 1.4949,
"step": 1472
},
{
"epoch": 1.7434152116010653,
"grad_norm": 5.6875,
"learning_rate": 1.645318683377915e-05,
"loss": 1.1332,
"step": 1473
},
{
"epoch": 1.7445989937851434,
"grad_norm": 5.875,
"learning_rate": 1.644686576174286e-05,
"loss": 1.2679,
"step": 1474
},
{
"epoch": 1.7457827759692217,
"grad_norm": 5.84375,
"learning_rate": 1.6440540278692656e-05,
"loss": 1.2799,
"step": 1475
},
{
"epoch": 1.7469665581532998,
"grad_norm": 6.28125,
"learning_rate": 1.643421038895651e-05,
"loss": 1.5058,
"step": 1476
},
{
"epoch": 1.748150340337378,
"grad_norm": 5.84375,
"learning_rate": 1.6427876096865394e-05,
"loss": 1.2089,
"step": 1477
},
{
"epoch": 1.749334122521456,
"grad_norm": 6.625,
"learning_rate": 1.6421537406753302e-05,
"loss": 1.5396,
"step": 1478
},
{
"epoch": 1.7505179047055341,
"grad_norm": 6.3125,
"learning_rate": 1.6415194322957232e-05,
"loss": 1.6986,
"step": 1479
},
{
"epoch": 1.7517016868896125,
"grad_norm": 5.8125,
"learning_rate": 1.6408846849817182e-05,
"loss": 0.9811,
"step": 1480
},
{
"epoch": 1.7528854690736906,
"grad_norm": 4.75,
"learning_rate": 1.6402494991676174e-05,
"loss": 1.0357,
"step": 1481
},
{
"epoch": 1.7540692512577687,
"grad_norm": 5.375,
"learning_rate": 1.6396138752880205e-05,
"loss": 0.8781,
"step": 1482
},
{
"epoch": 1.7552530334418468,
"grad_norm": 5.90625,
"learning_rate": 1.6389778137778284e-05,
"loss": 1.3404,
"step": 1483
},
{
"epoch": 1.756436815625925,
"grad_norm": 6.3125,
"learning_rate": 1.6383413150722417e-05,
"loss": 1.3058,
"step": 1484
},
{
"epoch": 1.757620597810003,
"grad_norm": 6.0,
"learning_rate": 1.6377043796067585e-05,
"loss": 1.1635,
"step": 1485
},
{
"epoch": 1.7588043799940811,
"grad_norm": 5.53125,
"learning_rate": 1.637067007817178e-05,
"loss": 1.2193,
"step": 1486
},
{
"epoch": 1.7599881621781592,
"grad_norm": 6.625,
"learning_rate": 1.6364292001395958e-05,
"loss": 1.3951,
"step": 1487
},
{
"epoch": 1.7611719443622373,
"grad_norm": 6.125,
"learning_rate": 1.635790957010407e-05,
"loss": 1.5455,
"step": 1488
},
{
"epoch": 1.7623557265463154,
"grad_norm": 6.625,
"learning_rate": 1.6351522788663044e-05,
"loss": 1.4453,
"step": 1489
},
{
"epoch": 1.7635395087303936,
"grad_norm": 5.59375,
"learning_rate": 1.6345131661442782e-05,
"loss": 0.9829,
"step": 1490
},
{
"epoch": 1.7647232909144717,
"grad_norm": 6.25,
"learning_rate": 1.6338736192816163e-05,
"loss": 1.1098,
"step": 1491
},
{
"epoch": 1.7659070730985498,
"grad_norm": 4.5625,
"learning_rate": 1.6332336387159032e-05,
"loss": 0.7784,
"step": 1492
},
{
"epoch": 1.7670908552826279,
"grad_norm": 5.25,
"learning_rate": 1.6325932248850206e-05,
"loss": 0.9815,
"step": 1493
},
{
"epoch": 1.768274637466706,
"grad_norm": 6.0625,
"learning_rate": 1.6319523782271462e-05,
"loss": 1.2823,
"step": 1494
},
{
"epoch": 1.769458419650784,
"grad_norm": 6.1875,
"learning_rate": 1.6313110991807546e-05,
"loss": 1.5918,
"step": 1495
},
{
"epoch": 1.7706422018348624,
"grad_norm": 6.21875,
"learning_rate": 1.6306693881846152e-05,
"loss": 1.1253,
"step": 1496
},
{
"epoch": 1.7718259840189405,
"grad_norm": 6.78125,
"learning_rate": 1.6300272456777933e-05,
"loss": 1.6937,
"step": 1497
},
{
"epoch": 1.7730097662030186,
"grad_norm": 6.375,
"learning_rate": 1.6293846720996504e-05,
"loss": 1.7121,
"step": 1498
},
{
"epoch": 1.7741935483870968,
"grad_norm": 7.03125,
"learning_rate": 1.6287416678898415e-05,
"loss": 1.1525,
"step": 1499
},
{
"epoch": 1.7753773305711749,
"grad_norm": 5.65625,
"learning_rate": 1.628098233488317e-05,
"loss": 1.167,
"step": 1500
},
{
"epoch": 1.7765611127552532,
"grad_norm": 6.6875,
"learning_rate": 1.6274543693353218e-05,
"loss": 1.2101,
"step": 1501
},
{
"epoch": 1.7777448949393313,
"grad_norm": 6.1875,
"learning_rate": 1.626810075871394e-05,
"loss": 1.0372,
"step": 1502
},
{
"epoch": 1.7789286771234094,
"grad_norm": 5.9375,
"learning_rate": 1.6261653535373668e-05,
"loss": 1.208,
"step": 1503
},
{
"epoch": 1.7801124593074875,
"grad_norm": 5.4375,
"learning_rate": 1.6255202027743658e-05,
"loss": 1.2318,
"step": 1504
},
{
"epoch": 1.7812962414915656,
"grad_norm": 4.5,
"learning_rate": 1.6248746240238095e-05,
"loss": 0.8007,
"step": 1505
},
{
"epoch": 1.7824800236756437,
"grad_norm": 5.6875,
"learning_rate": 1.6242286177274103e-05,
"loss": 1.1835,
"step": 1506
},
{
"epoch": 1.7836638058597218,
"grad_norm": 5.65625,
"learning_rate": 1.6235821843271723e-05,
"loss": 1.1683,
"step": 1507
},
{
"epoch": 1.7848475880438,
"grad_norm": 5.53125,
"learning_rate": 1.6229353242653922e-05,
"loss": 0.9267,
"step": 1508
},
{
"epoch": 1.786031370227878,
"grad_norm": 5.53125,
"learning_rate": 1.6222880379846584e-05,
"loss": 0.9019,
"step": 1509
},
{
"epoch": 1.7872151524119562,
"grad_norm": 4.5,
"learning_rate": 1.6216403259278514e-05,
"loss": 0.8028,
"step": 1510
},
{
"epoch": 1.7883989345960343,
"grad_norm": 5.96875,
"learning_rate": 1.6209921885381418e-05,
"loss": 1.289,
"step": 1511
},
{
"epoch": 1.7895827167801124,
"grad_norm": 5.5,
"learning_rate": 1.6203436262589927e-05,
"loss": 1.3159,
"step": 1512
},
{
"epoch": 1.7907664989641905,
"grad_norm": 5.25,
"learning_rate": 1.6196946395341576e-05,
"loss": 1.0076,
"step": 1513
},
{
"epoch": 1.7919502811482686,
"grad_norm": 4.5,
"learning_rate": 1.6190452288076795e-05,
"loss": 1.1402,
"step": 1514
},
{
"epoch": 1.7931340633323467,
"grad_norm": 6.15625,
"learning_rate": 1.6183953945238924e-05,
"loss": 1.4977,
"step": 1515
},
{
"epoch": 1.7943178455164248,
"grad_norm": 8.5,
"learning_rate": 1.6177451371274197e-05,
"loss": 1.4719,
"step": 1516
},
{
"epoch": 1.7955016277005031,
"grad_norm": 5.9375,
"learning_rate": 1.6170944570631743e-05,
"loss": 1.6843,
"step": 1517
},
{
"epoch": 1.7966854098845813,
"grad_norm": 6.1875,
"learning_rate": 1.6164433547763584e-05,
"loss": 1.4098,
"step": 1518
},
{
"epoch": 1.7978691920686594,
"grad_norm": 6.0,
"learning_rate": 1.6157918307124635e-05,
"loss": 1.2278,
"step": 1519
},
{
"epoch": 1.7990529742527375,
"grad_norm": 5.59375,
"learning_rate": 1.615139885317269e-05,
"loss": 1.363,
"step": 1520
},
{
"epoch": 1.8002367564368156,
"grad_norm": 6.09375,
"learning_rate": 1.6144875190368426e-05,
"loss": 1.4795,
"step": 1521
},
{
"epoch": 1.801420538620894,
"grad_norm": 4.53125,
"learning_rate": 1.6138347323175402e-05,
"loss": 0.7845,
"step": 1522
},
{
"epoch": 1.802604320804972,
"grad_norm": 5.53125,
"learning_rate": 1.6131815256060064e-05,
"loss": 1.3649,
"step": 1523
},
{
"epoch": 1.8037881029890501,
"grad_norm": 5.9375,
"learning_rate": 1.612527899349171e-05,
"loss": 1.4326,
"step": 1524
},
{
"epoch": 1.8049718851731282,
"grad_norm": 6.0625,
"learning_rate": 1.6118738539942524e-05,
"loss": 1.4829,
"step": 1525
},
{
"epoch": 1.8061556673572063,
"grad_norm": 5.25,
"learning_rate": 1.6112193899887555e-05,
"loss": 1.1326,
"step": 1526
},
{
"epoch": 1.8073394495412844,
"grad_norm": 6.375,
"learning_rate": 1.6105645077804713e-05,
"loss": 1.5077,
"step": 1527
},
{
"epoch": 1.8085232317253626,
"grad_norm": 5.84375,
"learning_rate": 1.6099092078174773e-05,
"loss": 1.2742,
"step": 1528
},
{
"epoch": 1.8097070139094407,
"grad_norm": 6.40625,
"learning_rate": 1.6092534905481367e-05,
"loss": 1.7877,
"step": 1529
},
{
"epoch": 1.8108907960935188,
"grad_norm": 5.28125,
"learning_rate": 1.608597356421098e-05,
"loss": 1.1368,
"step": 1530
},
{
"epoch": 1.8120745782775969,
"grad_norm": 7.0625,
"learning_rate": 1.6079408058852956e-05,
"loss": 1.7861,
"step": 1531
},
{
"epoch": 1.813258360461675,
"grad_norm": 5.65625,
"learning_rate": 1.6072838393899476e-05,
"loss": 1.2169,
"step": 1532
},
{
"epoch": 1.814442142645753,
"grad_norm": 6.09375,
"learning_rate": 1.606626457384558e-05,
"loss": 1.4832,
"step": 1533
},
{
"epoch": 1.8156259248298312,
"grad_norm": 6.625,
"learning_rate": 1.6059686603189146e-05,
"loss": 1.647,
"step": 1534
},
{
"epoch": 1.8168097070139093,
"grad_norm": 5.84375,
"learning_rate": 1.6053104486430887e-05,
"loss": 1.3547,
"step": 1535
},
{
"epoch": 1.8179934891979874,
"grad_norm": 7.09375,
"learning_rate": 1.604651822807436e-05,
"loss": 1.4115,
"step": 1536
},
{
"epoch": 1.8191772713820658,
"grad_norm": 5.96875,
"learning_rate": 1.6039927832625953e-05,
"loss": 1.5265,
"step": 1537
},
{
"epoch": 1.8203610535661439,
"grad_norm": 5.4375,
"learning_rate": 1.6033333304594886e-05,
"loss": 1.0446,
"step": 1538
},
{
"epoch": 1.821544835750222,
"grad_norm": 5.03125,
"learning_rate": 1.6026734648493195e-05,
"loss": 1.2012,
"step": 1539
},
{
"epoch": 1.8227286179343,
"grad_norm": 5.34375,
"learning_rate": 1.6020131868835762e-05,
"loss": 0.9209,
"step": 1540
},
{
"epoch": 1.8239124001183782,
"grad_norm": 4.8125,
"learning_rate": 1.6013524970140276e-05,
"loss": 1.0065,
"step": 1541
},
{
"epoch": 1.8250961823024563,
"grad_norm": 4.5,
"learning_rate": 1.6006913956927242e-05,
"loss": 0.8243,
"step": 1542
},
{
"epoch": 1.8262799644865346,
"grad_norm": 4.90625,
"learning_rate": 1.6000298833719996e-05,
"loss": 1.0843,
"step": 1543
},
{
"epoch": 1.8274637466706127,
"grad_norm": 4.75,
"learning_rate": 1.5993679605044662e-05,
"loss": 1.0199,
"step": 1544
},
{
"epoch": 1.8286475288546908,
"grad_norm": 6.0625,
"learning_rate": 1.5987056275430196e-05,
"loss": 1.4643,
"step": 1545
},
{
"epoch": 1.829831311038769,
"grad_norm": 6.1875,
"learning_rate": 1.598042884940835e-05,
"loss": 1.4143,
"step": 1546
},
{
"epoch": 1.831015093222847,
"grad_norm": 6.0625,
"learning_rate": 1.5973797331513674e-05,
"loss": 1.4086,
"step": 1547
},
{
"epoch": 1.8321988754069252,
"grad_norm": 4.875,
"learning_rate": 1.5967161726283527e-05,
"loss": 0.9139,
"step": 1548
},
{
"epoch": 1.8333826575910033,
"grad_norm": 5.25,
"learning_rate": 1.5960522038258057e-05,
"loss": 1.2274,
"step": 1549
},
{
"epoch": 1.8345664397750814,
"grad_norm": 6.0625,
"learning_rate": 1.5953878271980212e-05,
"loss": 1.1662,
"step": 1550
},
{
"epoch": 1.8357502219591595,
"grad_norm": 6.46875,
"learning_rate": 1.5947230431995725e-05,
"loss": 1.2266,
"step": 1551
},
{
"epoch": 1.8369340041432376,
"grad_norm": 5.9375,
"learning_rate": 1.594057852285312e-05,
"loss": 1.1761,
"step": 1552
},
{
"epoch": 1.8381177863273157,
"grad_norm": 6.65625,
"learning_rate": 1.5933922549103708e-05,
"loss": 1.3472,
"step": 1553
},
{
"epoch": 1.8393015685113938,
"grad_norm": 6.0625,
"learning_rate": 1.5927262515301565e-05,
"loss": 1.4622,
"step": 1554
},
{
"epoch": 1.840485350695472,
"grad_norm": 6.625,
"learning_rate": 1.592059842600357e-05,
"loss": 1.438,
"step": 1555
},
{
"epoch": 1.84166913287955,
"grad_norm": 5.78125,
"learning_rate": 1.5913930285769356e-05,
"loss": 1.2777,
"step": 1556
},
{
"epoch": 1.8428529150636281,
"grad_norm": 7.40625,
"learning_rate": 1.5907258099161335e-05,
"loss": 1.9538,
"step": 1557
},
{
"epoch": 1.8440366972477065,
"grad_norm": 6.28125,
"learning_rate": 1.5900581870744692e-05,
"loss": 1.514,
"step": 1558
},
{
"epoch": 1.8452204794317846,
"grad_norm": 6.3125,
"learning_rate": 1.589390160508737e-05,
"loss": 1.0674,
"step": 1559
},
{
"epoch": 1.8464042616158627,
"grad_norm": 6.125,
"learning_rate": 1.5887217306760082e-05,
"loss": 1.21,
"step": 1560
},
{
"epoch": 1.8475880437999408,
"grad_norm": 5.15625,
"learning_rate": 1.588052898033629e-05,
"loss": 1.0875,
"step": 1561
},
{
"epoch": 1.848771825984019,
"grad_norm": 6.25,
"learning_rate": 1.5873836630392218e-05,
"loss": 1.2685,
"step": 1562
},
{
"epoch": 1.8499556081680972,
"grad_norm": 4.9375,
"learning_rate": 1.5867140261506845e-05,
"loss": 0.9204,
"step": 1563
},
{
"epoch": 1.8511393903521753,
"grad_norm": 8.875,
"learning_rate": 1.5860439878261905e-05,
"loss": 1.6964,
"step": 1564
},
{
"epoch": 1.8523231725362534,
"grad_norm": 6.0,
"learning_rate": 1.5853735485241858e-05,
"loss": 1.2348,
"step": 1565
},
{
"epoch": 1.8535069547203316,
"grad_norm": 6.625,
"learning_rate": 1.5847027087033926e-05,
"loss": 1.5894,
"step": 1566
},
{
"epoch": 1.8546907369044097,
"grad_norm": 6.4375,
"learning_rate": 1.584031468822807e-05,
"loss": 1.0699,
"step": 1567
},
{
"epoch": 1.8558745190884878,
"grad_norm": 5.75,
"learning_rate": 1.5833598293416978e-05,
"loss": 1.5189,
"step": 1568
},
{
"epoch": 1.8570583012725659,
"grad_norm": 6.4375,
"learning_rate": 1.5826877907196084e-05,
"loss": 1.726,
"step": 1569
},
{
"epoch": 1.858242083456644,
"grad_norm": 5.625,
"learning_rate": 1.5820153534163543e-05,
"loss": 0.7998,
"step": 1570
},
{
"epoch": 1.859425865640722,
"grad_norm": 4.625,
"learning_rate": 1.581342517892025e-05,
"loss": 0.6937,
"step": 1571
},
{
"epoch": 1.8606096478248002,
"grad_norm": 5.28125,
"learning_rate": 1.5806692846069807e-05,
"loss": 1.0849,
"step": 1572
},
{
"epoch": 1.8617934300088783,
"grad_norm": 5.8125,
"learning_rate": 1.5799956540218553e-05,
"loss": 1.464,
"step": 1573
},
{
"epoch": 1.8629772121929564,
"grad_norm": 6.90625,
"learning_rate": 1.579321626597554e-05,
"loss": 1.7484,
"step": 1574
},
{
"epoch": 1.8641609943770345,
"grad_norm": 4.875,
"learning_rate": 1.5786472027952533e-05,
"loss": 0.842,
"step": 1575
},
{
"epoch": 1.8653447765611126,
"grad_norm": 6.25,
"learning_rate": 1.5779723830764013e-05,
"loss": 1.3273,
"step": 1576
},
{
"epoch": 1.8665285587451907,
"grad_norm": 6.03125,
"learning_rate": 1.5772971679027168e-05,
"loss": 1.1666,
"step": 1577
},
{
"epoch": 1.8677123409292689,
"grad_norm": 5.65625,
"learning_rate": 1.576621557736189e-05,
"loss": 1.2737,
"step": 1578
},
{
"epoch": 1.8688961231133472,
"grad_norm": 6.4375,
"learning_rate": 1.5759455530390782e-05,
"loss": 1.5155,
"step": 1579
},
{
"epoch": 1.8700799052974253,
"grad_norm": 6.28125,
"learning_rate": 1.575269154273913e-05,
"loss": 1.6416,
"step": 1580
},
{
"epoch": 1.8712636874815034,
"grad_norm": 5.71875,
"learning_rate": 1.5745923619034928e-05,
"loss": 1.1113,
"step": 1581
},
{
"epoch": 1.8724474696655815,
"grad_norm": 5.8125,
"learning_rate": 1.5739151763908867e-05,
"loss": 1.5798,
"step": 1582
},
{
"epoch": 1.8736312518496596,
"grad_norm": 5.625,
"learning_rate": 1.573237598199432e-05,
"loss": 1.6145,
"step": 1583
},
{
"epoch": 1.874815034033738,
"grad_norm": 5.90625,
"learning_rate": 1.5725596277927342e-05,
"loss": 1.3546,
"step": 1584
},
{
"epoch": 1.875998816217816,
"grad_norm": 6.71875,
"learning_rate": 1.571881265634669e-05,
"loss": 1.6839,
"step": 1585
},
{
"epoch": 1.8771825984018942,
"grad_norm": 5.78125,
"learning_rate": 1.571202512189378e-05,
"loss": 1.2636,
"step": 1586
},
{
"epoch": 1.8783663805859723,
"grad_norm": 4.96875,
"learning_rate": 1.570523367921272e-05,
"loss": 0.8954,
"step": 1587
},
{
"epoch": 1.8795501627700504,
"grad_norm": 6.5,
"learning_rate": 1.5698438332950287e-05,
"loss": 1.2852,
"step": 1588
},
{
"epoch": 1.8807339449541285,
"grad_norm": 6.78125,
"learning_rate": 1.5691639087755928e-05,
"loss": 1.4471,
"step": 1589
},
{
"epoch": 1.8819177271382066,
"grad_norm": 8.4375,
"learning_rate": 1.568483594828176e-05,
"loss": 1.7659,
"step": 1590
},
{
"epoch": 1.8831015093222847,
"grad_norm": 5.15625,
"learning_rate": 1.5678028919182558e-05,
"loss": 1.106,
"step": 1591
},
{
"epoch": 1.8842852915063628,
"grad_norm": 6.125,
"learning_rate": 1.5671218005115767e-05,
"loss": 1.5331,
"step": 1592
},
{
"epoch": 1.885469073690441,
"grad_norm": 5.59375,
"learning_rate": 1.566440321074149e-05,
"loss": 0.9351,
"step": 1593
},
{
"epoch": 1.886652855874519,
"grad_norm": 6.90625,
"learning_rate": 1.5657584540722476e-05,
"loss": 1.0853,
"step": 1594
},
{
"epoch": 1.8878366380585971,
"grad_norm": 5.59375,
"learning_rate": 1.5650761999724134e-05,
"loss": 1.3752,
"step": 1595
},
{
"epoch": 1.8890204202426752,
"grad_norm": 5.1875,
"learning_rate": 1.5643935592414518e-05,
"loss": 0.7999,
"step": 1596
},
{
"epoch": 1.8902042024267534,
"grad_norm": 5.15625,
"learning_rate": 1.563710532346433e-05,
"loss": 1.2622,
"step": 1597
},
{
"epoch": 1.8913879846108315,
"grad_norm": 6.5,
"learning_rate": 1.563027119754691e-05,
"loss": 1.1718,
"step": 1598
},
{
"epoch": 1.8925717667949096,
"grad_norm": 5.6875,
"learning_rate": 1.5623433219338244e-05,
"loss": 1.147,
"step": 1599
},
{
"epoch": 1.893755548978988,
"grad_norm": 6.375,
"learning_rate": 1.5616591393516946e-05,
"loss": 1.1672,
"step": 1600
},
{
"epoch": 1.894939331163066,
"grad_norm": 6.0,
"learning_rate": 1.5609745724764264e-05,
"loss": 1.3116,
"step": 1601
},
{
"epoch": 1.8961231133471441,
"grad_norm": 6.21875,
"learning_rate": 1.5602896217764075e-05,
"loss": 1.5243,
"step": 1602
},
{
"epoch": 1.8973068955312222,
"grad_norm": 5.71875,
"learning_rate": 1.559604287720289e-05,
"loss": 1.6061,
"step": 1603
},
{
"epoch": 1.8984906777153003,
"grad_norm": 5.6875,
"learning_rate": 1.558918570776984e-05,
"loss": 1.0744,
"step": 1604
},
{
"epoch": 1.8996744598993787,
"grad_norm": 5.75,
"learning_rate": 1.5582324714156663e-05,
"loss": 1.3432,
"step": 1605
},
{
"epoch": 1.9008582420834568,
"grad_norm": 5.1875,
"learning_rate": 1.557545990105773e-05,
"loss": 1.2206,
"step": 1606
},
{
"epoch": 1.9020420242675349,
"grad_norm": 5.65625,
"learning_rate": 1.556859127317002e-05,
"loss": 1.2301,
"step": 1607
},
{
"epoch": 1.903225806451613,
"grad_norm": 6.28125,
"learning_rate": 1.5561718835193117e-05,
"loss": 1.4809,
"step": 1608
},
{
"epoch": 1.904409588635691,
"grad_norm": 6.4375,
"learning_rate": 1.555484259182922e-05,
"loss": 1.4269,
"step": 1609
},
{
"epoch": 1.9055933708197692,
"grad_norm": 6.625,
"learning_rate": 1.5547962547783126e-05,
"loss": 1.5821,
"step": 1610
},
{
"epoch": 1.9067771530038473,
"grad_norm": 5.28125,
"learning_rate": 1.554107870776223e-05,
"loss": 1.316,
"step": 1611
},
{
"epoch": 1.9079609351879254,
"grad_norm": 5.65625,
"learning_rate": 1.553419107647653e-05,
"loss": 1.388,
"step": 1612
},
{
"epoch": 1.9091447173720035,
"grad_norm": 4.96875,
"learning_rate": 1.5527299658638627e-05,
"loss": 1.0944,
"step": 1613
},
{
"epoch": 1.9103284995560816,
"grad_norm": 5.375,
"learning_rate": 1.5520404458963683e-05,
"loss": 1.3085,
"step": 1614
},
{
"epoch": 1.9115122817401597,
"grad_norm": 6.25,
"learning_rate": 1.551350548216948e-05,
"loss": 1.6434,
"step": 1615
},
{
"epoch": 1.9126960639242379,
"grad_norm": 5.78125,
"learning_rate": 1.5506602732976373e-05,
"loss": 1.3168,
"step": 1616
},
{
"epoch": 1.913879846108316,
"grad_norm": 6.09375,
"learning_rate": 1.5499696216107293e-05,
"loss": 1.2865,
"step": 1617
},
{
"epoch": 1.915063628292394,
"grad_norm": 5.5625,
"learning_rate": 1.5492785936287744e-05,
"loss": 1.3117,
"step": 1618
},
{
"epoch": 1.9162474104764722,
"grad_norm": 6.15625,
"learning_rate": 1.5485871898245824e-05,
"loss": 1.1456,
"step": 1619
},
{
"epoch": 1.9174311926605505,
"grad_norm": 5.6875,
"learning_rate": 1.547895410671218e-05,
"loss": 1.348,
"step": 1620
},
{
"epoch": 1.9186149748446286,
"grad_norm": 5.875,
"learning_rate": 1.5472032566420047e-05,
"loss": 1.5114,
"step": 1621
},
{
"epoch": 1.9197987570287067,
"grad_norm": 5.96875,
"learning_rate": 1.546510728210522e-05,
"loss": 1.68,
"step": 1622
},
{
"epoch": 1.9209825392127848,
"grad_norm": 5.71875,
"learning_rate": 1.545817825850604e-05,
"loss": 1.1689,
"step": 1623
},
{
"epoch": 1.922166321396863,
"grad_norm": 5.65625,
"learning_rate": 1.5451245500363424e-05,
"loss": 1.2115,
"step": 1624
},
{
"epoch": 1.9233501035809413,
"grad_norm": 5.53125,
"learning_rate": 1.544430901242084e-05,
"loss": 1.0604,
"step": 1625
},
{
"epoch": 1.9245338857650194,
"grad_norm": 6.65625,
"learning_rate": 1.5437368799424307e-05,
"loss": 1.3335,
"step": 1626
},
{
"epoch": 1.9257176679490975,
"grad_norm": 5.59375,
"learning_rate": 1.5430424866122386e-05,
"loss": 1.1301,
"step": 1627
},
{
"epoch": 1.9269014501331756,
"grad_norm": 4.5,
"learning_rate": 1.54234772172662e-05,
"loss": 0.9096,
"step": 1628
},
{
"epoch": 1.9280852323172537,
"grad_norm": 6.90625,
"learning_rate": 1.5416525857609396e-05,
"loss": 2.0481,
"step": 1629
},
{
"epoch": 1.9292690145013318,
"grad_norm": 5.0625,
"learning_rate": 1.5409570791908172e-05,
"loss": 0.9335,
"step": 1630
},
{
"epoch": 1.93045279668541,
"grad_norm": 6.21875,
"learning_rate": 1.5402612024921254e-05,
"loss": 1.3242,
"step": 1631
},
{
"epoch": 1.931636578869488,
"grad_norm": 5.9375,
"learning_rate": 1.5395649561409903e-05,
"loss": 1.468,
"step": 1632
},
{
"epoch": 1.9328203610535661,
"grad_norm": 6.5,
"learning_rate": 1.5388683406137923e-05,
"loss": 1.4578,
"step": 1633
},
{
"epoch": 1.9340041432376442,
"grad_norm": 5.6875,
"learning_rate": 1.5381713563871616e-05,
"loss": 1.2629,
"step": 1634
},
{
"epoch": 1.9351879254217224,
"grad_norm": 7.0,
"learning_rate": 1.537474003937983e-05,
"loss": 1.6642,
"step": 1635
},
{
"epoch": 1.9363717076058005,
"grad_norm": 6.03125,
"learning_rate": 1.536776283743392e-05,
"loss": 1.246,
"step": 1636
},
{
"epoch": 1.9375554897898786,
"grad_norm": 5.65625,
"learning_rate": 1.536078196280777e-05,
"loss": 1.2676,
"step": 1637
},
{
"epoch": 1.9387392719739567,
"grad_norm": 5.375,
"learning_rate": 1.5353797420277755e-05,
"loss": 1.1858,
"step": 1638
},
{
"epoch": 1.9399230541580348,
"grad_norm": 5.8125,
"learning_rate": 1.5346809214622785e-05,
"loss": 1.048,
"step": 1639
},
{
"epoch": 1.941106836342113,
"grad_norm": 6.25,
"learning_rate": 1.533981735062426e-05,
"loss": 1.1626,
"step": 1640
},
{
"epoch": 1.9422906185261912,
"grad_norm": 6.75,
"learning_rate": 1.5332821833066083e-05,
"loss": 1.429,
"step": 1641
},
{
"epoch": 1.9434744007102693,
"grad_norm": 5.25,
"learning_rate": 1.532582266673467e-05,
"loss": 1.1873,
"step": 1642
},
{
"epoch": 1.9446581828943474,
"grad_norm": 5.65625,
"learning_rate": 1.5318819856418918e-05,
"loss": 1.4501,
"step": 1643
},
{
"epoch": 1.9458419650784256,
"grad_norm": 7.0,
"learning_rate": 1.5311813406910227e-05,
"loss": 1.4162,
"step": 1644
},
{
"epoch": 1.9470257472625037,
"grad_norm": 4.78125,
"learning_rate": 1.5304803323002483e-05,
"loss": 0.9701,
"step": 1645
},
{
"epoch": 1.948209529446582,
"grad_norm": 7.15625,
"learning_rate": 1.5297789609492062e-05,
"loss": 1.3231,
"step": 1646
},
{
"epoch": 1.94939331163066,
"grad_norm": 6.28125,
"learning_rate": 1.5290772271177818e-05,
"loss": 1.4139,
"step": 1647
},
{
"epoch": 1.9505770938147382,
"grad_norm": 6.1875,
"learning_rate": 1.5283751312861093e-05,
"loss": 1.374,
"step": 1648
},
{
"epoch": 1.9517608759988163,
"grad_norm": 5.875,
"learning_rate": 1.527672673934569e-05,
"loss": 1.5372,
"step": 1649
},
{
"epoch": 1.9529446581828944,
"grad_norm": 5.5,
"learning_rate": 1.5269698555437913e-05,
"loss": 1.2469,
"step": 1650
},
{
"epoch": 1.9541284403669725,
"grad_norm": 4.96875,
"learning_rate": 1.526266676594651e-05,
"loss": 0.8888,
"step": 1651
},
{
"epoch": 1.9553122225510506,
"grad_norm": 6.25,
"learning_rate": 1.5255631375682702e-05,
"loss": 1.4135,
"step": 1652
},
{
"epoch": 1.9564960047351287,
"grad_norm": 5.0625,
"learning_rate": 1.5248592389460189e-05,
"loss": 1.1634,
"step": 1653
},
{
"epoch": 1.9576797869192069,
"grad_norm": 6.0,
"learning_rate": 1.5241549812095114e-05,
"loss": 1.2268,
"step": 1654
},
{
"epoch": 1.958863569103285,
"grad_norm": 5.75,
"learning_rate": 1.5234503648406075e-05,
"loss": 1.2457,
"step": 1655
},
{
"epoch": 1.960047351287363,
"grad_norm": 5.75,
"learning_rate": 1.5227453903214147e-05,
"loss": 1.2611,
"step": 1656
},
{
"epoch": 1.9612311334714412,
"grad_norm": 6.03125,
"learning_rate": 1.5220400581342833e-05,
"loss": 0.9867,
"step": 1657
},
{
"epoch": 1.9624149156555193,
"grad_norm": 5.46875,
"learning_rate": 1.521334368761809e-05,
"loss": 1.3796,
"step": 1658
},
{
"epoch": 1.9635986978395974,
"grad_norm": 5.125,
"learning_rate": 1.520628322686832e-05,
"loss": 1.0728,
"step": 1659
},
{
"epoch": 1.9647824800236755,
"grad_norm": 5.6875,
"learning_rate": 1.5199219203924367e-05,
"loss": 1.4536,
"step": 1660
},
{
"epoch": 1.9659662622077536,
"grad_norm": 5.4375,
"learning_rate": 1.5192151623619506e-05,
"loss": 1.1857,
"step": 1661
},
{
"epoch": 1.967150044391832,
"grad_norm": 5.5625,
"learning_rate": 1.5185080490789457e-05,
"loss": 1.4415,
"step": 1662
},
{
"epoch": 1.96833382657591,
"grad_norm": 4.4375,
"learning_rate": 1.517800581027236e-05,
"loss": 0.807,
"step": 1663
},
{
"epoch": 1.9695176087599882,
"grad_norm": 5.0,
"learning_rate": 1.5170927586908787e-05,
"loss": 0.9589,
"step": 1664
},
{
"epoch": 1.9707013909440663,
"grad_norm": 5.28125,
"learning_rate": 1.516384582554173e-05,
"loss": 1.046,
"step": 1665
},
{
"epoch": 1.9718851731281444,
"grad_norm": 5.6875,
"learning_rate": 1.515676053101661e-05,
"loss": 1.1006,
"step": 1666
},
{
"epoch": 1.9730689553122227,
"grad_norm": 6.4375,
"learning_rate": 1.514967170818126e-05,
"loss": 1.538,
"step": 1667
},
{
"epoch": 1.9742527374963008,
"grad_norm": 7.21875,
"learning_rate": 1.5142579361885927e-05,
"loss": 1.2989,
"step": 1668
},
{
"epoch": 1.975436519680379,
"grad_norm": 5.8125,
"learning_rate": 1.5135483496983273e-05,
"loss": 1.2258,
"step": 1669
},
{
"epoch": 1.976620301864457,
"grad_norm": 7.25,
"learning_rate": 1.5128384118328355e-05,
"loss": 1.6718,
"step": 1670
},
{
"epoch": 1.9778040840485351,
"grad_norm": 15.8125,
"learning_rate": 1.512128123077865e-05,
"loss": 1.1139,
"step": 1671
},
{
"epoch": 1.9789878662326132,
"grad_norm": 5.8125,
"learning_rate": 1.5114174839194029e-05,
"loss": 1.4326,
"step": 1672
},
{
"epoch": 1.9801716484166914,
"grad_norm": 6.0,
"learning_rate": 1.5107064948436758e-05,
"loss": 1.402,
"step": 1673
},
{
"epoch": 1.9813554306007695,
"grad_norm": 5.375,
"learning_rate": 1.50999515633715e-05,
"loss": 1.1666,
"step": 1674
},
{
"epoch": 1.9825392127848476,
"grad_norm": 6.65625,
"learning_rate": 1.5092834688865311e-05,
"loss": 1.4487,
"step": 1675
},
{
"epoch": 1.9837229949689257,
"grad_norm": 8.3125,
"learning_rate": 1.5085714329787631e-05,
"loss": 1.7003,
"step": 1676
},
{
"epoch": 1.9849067771530038,
"grad_norm": 6.40625,
"learning_rate": 1.507859049101028e-05,
"loss": 1.6256,
"step": 1677
},
{
"epoch": 1.986090559337082,
"grad_norm": 5.65625,
"learning_rate": 1.5071463177407473e-05,
"loss": 1.4496,
"step": 1678
},
{
"epoch": 1.98727434152116,
"grad_norm": 5.09375,
"learning_rate": 1.5064332393855787e-05,
"loss": 1.0231,
"step": 1679
},
{
"epoch": 1.9884581237052381,
"grad_norm": 6.59375,
"learning_rate": 1.5057198145234182e-05,
"loss": 1.6673,
"step": 1680
},
{
"epoch": 1.9896419058893162,
"grad_norm": 5.5,
"learning_rate": 1.5050060436423986e-05,
"loss": 1.3601,
"step": 1681
},
{
"epoch": 1.9908256880733946,
"grad_norm": 5.5625,
"learning_rate": 1.5042919272308895e-05,
"loss": 1.16,
"step": 1682
},
{
"epoch": 1.9920094702574727,
"grad_norm": 6.65625,
"learning_rate": 1.5035774657774972e-05,
"loss": 1.4967,
"step": 1683
},
{
"epoch": 1.9931932524415508,
"grad_norm": 5.53125,
"learning_rate": 1.5028626597710632e-05,
"loss": 1.2081,
"step": 1684
},
{
"epoch": 1.9943770346256289,
"grad_norm": 5.09375,
"learning_rate": 1.5021475097006657e-05,
"loss": 0.8963,
"step": 1685
},
{
"epoch": 1.995560816809707,
"grad_norm": 6.46875,
"learning_rate": 1.5014320160556182e-05,
"loss": 1.524,
"step": 1686
},
{
"epoch": 1.996744598993785,
"grad_norm": 6.0625,
"learning_rate": 1.5007161793254686e-05,
"loss": 1.0788,
"step": 1687
},
{
"epoch": 1.9979283811778634,
"grad_norm": 6.6875,
"learning_rate": 1.5000000000000002e-05,
"loss": 1.1912,
"step": 1688
},
{
"epoch": 1.9991121633619415,
"grad_norm": 7.375,
"learning_rate": 1.4992834785692303e-05,
"loss": 1.1103,
"step": 1689
},
{
"epoch": 2.0,
"grad_norm": 5.75,
"learning_rate": 1.4985666155234109e-05,
"loss": 1.0301,
"step": 1690
},
{
"epoch": 2.001183782184078,
"grad_norm": 6.25,
"learning_rate": 1.4978494113530268e-05,
"loss": 1.5102,
"step": 1691
},
{
"epoch": 2.002367564368156,
"grad_norm": 5.15625,
"learning_rate": 1.4971318665487974e-05,
"loss": 0.8851,
"step": 1692
},
{
"epoch": 2.0035513465522343,
"grad_norm": 6.28125,
"learning_rate": 1.4964139816016737e-05,
"loss": 1.7302,
"step": 1693
},
{
"epoch": 2.0047351287363124,
"grad_norm": 5.25,
"learning_rate": 1.4956957570028401e-05,
"loss": 1.1702,
"step": 1694
},
{
"epoch": 2.0059189109203905,
"grad_norm": 6.375,
"learning_rate": 1.494977193243714e-05,
"loss": 1.3998,
"step": 1695
},
{
"epoch": 2.0071026931044686,
"grad_norm": 5.15625,
"learning_rate": 1.4942582908159446e-05,
"loss": 0.7946,
"step": 1696
},
{
"epoch": 2.0082864752885468,
"grad_norm": 4.1875,
"learning_rate": 1.4935390502114118e-05,
"loss": 0.8437,
"step": 1697
},
{
"epoch": 2.009470257472625,
"grad_norm": 4.6875,
"learning_rate": 1.492819471922228e-05,
"loss": 0.9858,
"step": 1698
},
{
"epoch": 2.010654039656703,
"grad_norm": 5.46875,
"learning_rate": 1.4920995564407368e-05,
"loss": 1.0343,
"step": 1699
},
{
"epoch": 2.011837821840781,
"grad_norm": 5.625,
"learning_rate": 1.4913793042595109e-05,
"loss": 1.3074,
"step": 1700
},
{
"epoch": 2.0130216040248596,
"grad_norm": 5.21875,
"learning_rate": 1.4906587158713552e-05,
"loss": 1.1474,
"step": 1701
},
{
"epoch": 2.0142053862089377,
"grad_norm": 4.53125,
"learning_rate": 1.4899377917693041e-05,
"loss": 0.8771,
"step": 1702
},
{
"epoch": 2.015389168393016,
"grad_norm": 5.1875,
"learning_rate": 1.489216532446621e-05,
"loss": 0.9176,
"step": 1703
},
{
"epoch": 2.016572950577094,
"grad_norm": 5.09375,
"learning_rate": 1.4884949383967993e-05,
"loss": 1.1296,
"step": 1704
},
{
"epoch": 2.017756732761172,
"grad_norm": 4.65625,
"learning_rate": 1.4877730101135619e-05,
"loss": 1.1406,
"step": 1705
},
{
"epoch": 2.01894051494525,
"grad_norm": 6.59375,
"learning_rate": 1.4870507480908587e-05,
"loss": 1.7316,
"step": 1706
},
{
"epoch": 2.0201242971293283,
"grad_norm": 6.59375,
"learning_rate": 1.4863281528228698e-05,
"loss": 1.2453,
"step": 1707
},
{
"epoch": 2.0213080793134064,
"grad_norm": 5.75,
"learning_rate": 1.4856052248040023e-05,
"loss": 1.3112,
"step": 1708
},
{
"epoch": 2.0224918614974845,
"grad_norm": 5.5,
"learning_rate": 1.4848819645288915e-05,
"loss": 1.2842,
"step": 1709
},
{
"epoch": 2.0236756436815626,
"grad_norm": 6.125,
"learning_rate": 1.4841583724923994e-05,
"loss": 1.4428,
"step": 1710
},
{
"epoch": 2.0248594258656407,
"grad_norm": 4.75,
"learning_rate": 1.4834344491896156e-05,
"loss": 1.0286,
"step": 1711
},
{
"epoch": 2.026043208049719,
"grad_norm": 4.96875,
"learning_rate": 1.4827101951158555e-05,
"loss": 0.8397,
"step": 1712
},
{
"epoch": 2.027226990233797,
"grad_norm": 5.1875,
"learning_rate": 1.4819856107666622e-05,
"loss": 1.1757,
"step": 1713
},
{
"epoch": 2.028410772417875,
"grad_norm": 5.9375,
"learning_rate": 1.4812606966378039e-05,
"loss": 1.4317,
"step": 1714
},
{
"epoch": 2.029594554601953,
"grad_norm": 5.375,
"learning_rate": 1.480535453225274e-05,
"loss": 1.1757,
"step": 1715
},
{
"epoch": 2.0307783367860313,
"grad_norm": 6.8125,
"learning_rate": 1.479809881025292e-05,
"loss": 1.3802,
"step": 1716
},
{
"epoch": 2.0319621189701094,
"grad_norm": 6.4375,
"learning_rate": 1.4790839805343025e-05,
"loss": 1.3051,
"step": 1717
},
{
"epoch": 2.0331459011541875,
"grad_norm": 4.59375,
"learning_rate": 1.4783577522489733e-05,
"loss": 0.5486,
"step": 1718
},
{
"epoch": 2.0343296833382656,
"grad_norm": 5.21875,
"learning_rate": 1.4776311966661987e-05,
"loss": 0.901,
"step": 1719
},
{
"epoch": 2.0355134655223437,
"grad_norm": 5.59375,
"learning_rate": 1.476904314283095e-05,
"loss": 0.9204,
"step": 1720
},
{
"epoch": 2.036697247706422,
"grad_norm": 6.0625,
"learning_rate": 1.4761771055970032e-05,
"loss": 1.1868,
"step": 1721
},
{
"epoch": 2.0378810298905004,
"grad_norm": 5.46875,
"learning_rate": 1.4754495711054866e-05,
"loss": 1.052,
"step": 1722
},
{
"epoch": 2.0390648120745785,
"grad_norm": 6.90625,
"learning_rate": 1.4747217113063328e-05,
"loss": 1.8921,
"step": 1723
},
{
"epoch": 2.0402485942586566,
"grad_norm": 5.75,
"learning_rate": 1.4739935266975503e-05,
"loss": 1.2518,
"step": 1724
},
{
"epoch": 2.0414323764427347,
"grad_norm": 7.03125,
"learning_rate": 1.4732650177773717e-05,
"loss": 1.1072,
"step": 1725
},
{
"epoch": 2.042616158626813,
"grad_norm": 5.59375,
"learning_rate": 1.4725361850442502e-05,
"loss": 0.852,
"step": 1726
},
{
"epoch": 2.043799940810891,
"grad_norm": 5.59375,
"learning_rate": 1.4718070289968602e-05,
"loss": 1.0403,
"step": 1727
},
{
"epoch": 2.044983722994969,
"grad_norm": 7.21875,
"learning_rate": 1.4710775501340988e-05,
"loss": 1.1103,
"step": 1728
},
{
"epoch": 2.046167505179047,
"grad_norm": 6.53125,
"learning_rate": 1.470347748955083e-05,
"loss": 1.2266,
"step": 1729
},
{
"epoch": 2.047351287363125,
"grad_norm": 7.34375,
"learning_rate": 1.4696176259591501e-05,
"loss": 1.8161,
"step": 1730
},
{
"epoch": 2.0485350695472033,
"grad_norm": 6.03125,
"learning_rate": 1.4688871816458585e-05,
"loss": 1.3915,
"step": 1731
},
{
"epoch": 2.0497188517312814,
"grad_norm": 5.0625,
"learning_rate": 1.4681564165149863e-05,
"loss": 0.8888,
"step": 1732
},
{
"epoch": 2.0509026339153595,
"grad_norm": 5.65625,
"learning_rate": 1.4674253310665294e-05,
"loss": 1.0929,
"step": 1733
},
{
"epoch": 2.0520864160994376,
"grad_norm": 5.65625,
"learning_rate": 1.4666939258007054e-05,
"loss": 0.9761,
"step": 1734
},
{
"epoch": 2.0532701982835158,
"grad_norm": 6.03125,
"learning_rate": 1.4659622012179493e-05,
"loss": 1.1665,
"step": 1735
},
{
"epoch": 2.054453980467594,
"grad_norm": 6.40625,
"learning_rate": 1.4652301578189141e-05,
"loss": 1.2553,
"step": 1736
},
{
"epoch": 2.055637762651672,
"grad_norm": 5.625,
"learning_rate": 1.4644977961044725e-05,
"loss": 0.9841,
"step": 1737
},
{
"epoch": 2.05682154483575,
"grad_norm": 6.5625,
"learning_rate": 1.4637651165757143e-05,
"loss": 1.1633,
"step": 1738
},
{
"epoch": 2.058005327019828,
"grad_norm": 5.4375,
"learning_rate": 1.463032119733946e-05,
"loss": 1.3354,
"step": 1739
},
{
"epoch": 2.0591891092039063,
"grad_norm": 5.5625,
"learning_rate": 1.4622988060806917e-05,
"loss": 1.2051,
"step": 1740
},
{
"epoch": 2.0603728913879844,
"grad_norm": 7.375,
"learning_rate": 1.4615651761176928e-05,
"loss": 1.6195,
"step": 1741
},
{
"epoch": 2.061556673572063,
"grad_norm": 4.625,
"learning_rate": 1.4608312303469067e-05,
"loss": 0.7275,
"step": 1742
},
{
"epoch": 2.062740455756141,
"grad_norm": 5.5,
"learning_rate": 1.4600969692705067e-05,
"loss": 1.0974,
"step": 1743
},
{
"epoch": 2.063924237940219,
"grad_norm": 5.5,
"learning_rate": 1.4593623933908822e-05,
"loss": 1.0758,
"step": 1744
},
{
"epoch": 2.0651080201242973,
"grad_norm": 6.34375,
"learning_rate": 1.4586275032106373e-05,
"loss": 1.3741,
"step": 1745
},
{
"epoch": 2.0662918023083754,
"grad_norm": 5.3125,
"learning_rate": 1.4578922992325924e-05,
"loss": 0.9151,
"step": 1746
},
{
"epoch": 2.0674755844924535,
"grad_norm": 5.9375,
"learning_rate": 1.4571567819597812e-05,
"loss": 1.1758,
"step": 1747
},
{
"epoch": 2.0686593666765316,
"grad_norm": 5.53125,
"learning_rate": 1.4564209518954528e-05,
"loss": 1.0833,
"step": 1748
},
{
"epoch": 2.0698431488606097,
"grad_norm": 7.0,
"learning_rate": 1.4556848095430698e-05,
"loss": 1.0948,
"step": 1749
},
{
"epoch": 2.071026931044688,
"grad_norm": 5.90625,
"learning_rate": 1.454948355406309e-05,
"loss": 1.1306,
"step": 1750
},
{
"epoch": 2.072210713228766,
"grad_norm": 7.4375,
"learning_rate": 1.4542115899890594e-05,
"loss": 1.603,
"step": 1751
},
{
"epoch": 2.073394495412844,
"grad_norm": 6.15625,
"learning_rate": 1.453474513795424e-05,
"loss": 1.1448,
"step": 1752
},
{
"epoch": 2.074578277596922,
"grad_norm": 6.125,
"learning_rate": 1.4527371273297184e-05,
"loss": 1.1615,
"step": 1753
},
{
"epoch": 2.0757620597810003,
"grad_norm": 5.0,
"learning_rate": 1.4519994310964697e-05,
"loss": 1.0181,
"step": 1754
},
{
"epoch": 2.0769458419650784,
"grad_norm": 6.46875,
"learning_rate": 1.4512614256004182e-05,
"loss": 1.2854,
"step": 1755
},
{
"epoch": 2.0781296241491565,
"grad_norm": 5.96875,
"learning_rate": 1.4505231113465148e-05,
"loss": 1.1568,
"step": 1756
},
{
"epoch": 2.0793134063332346,
"grad_norm": 4.75,
"learning_rate": 1.4497844888399216e-05,
"loss": 0.9652,
"step": 1757
},
{
"epoch": 2.0804971885173127,
"grad_norm": 6.28125,
"learning_rate": 1.4490455585860124e-05,
"loss": 1.2564,
"step": 1758
},
{
"epoch": 2.081680970701391,
"grad_norm": 5.28125,
"learning_rate": 1.448306321090371e-05,
"loss": 0.9164,
"step": 1759
},
{
"epoch": 2.082864752885469,
"grad_norm": 6.03125,
"learning_rate": 1.447566776858791e-05,
"loss": 1.4022,
"step": 1760
},
{
"epoch": 2.084048535069547,
"grad_norm": 6.03125,
"learning_rate": 1.4468269263972773e-05,
"loss": 1.1979,
"step": 1761
},
{
"epoch": 2.085232317253625,
"grad_norm": 6.0625,
"learning_rate": 1.4460867702120432e-05,
"loss": 1.0738,
"step": 1762
},
{
"epoch": 2.0864160994377037,
"grad_norm": 5.71875,
"learning_rate": 1.4453463088095108e-05,
"loss": 0.7597,
"step": 1763
},
{
"epoch": 2.087599881621782,
"grad_norm": 5.5625,
"learning_rate": 1.4446055426963121e-05,
"loss": 1.3502,
"step": 1764
},
{
"epoch": 2.08878366380586,
"grad_norm": 6.65625,
"learning_rate": 1.443864472379287e-05,
"loss": 1.152,
"step": 1765
},
{
"epoch": 2.089967445989938,
"grad_norm": 4.84375,
"learning_rate": 1.4431230983654839e-05,
"loss": 0.7866,
"step": 1766
},
{
"epoch": 2.091151228174016,
"grad_norm": 5.1875,
"learning_rate": 1.4423814211621586e-05,
"loss": 0.6993,
"step": 1767
},
{
"epoch": 2.092335010358094,
"grad_norm": 5.5625,
"learning_rate": 1.4416394412767747e-05,
"loss": 0.9852,
"step": 1768
},
{
"epoch": 2.0935187925421723,
"grad_norm": 6.75,
"learning_rate": 1.440897159217002e-05,
"loss": 1.4428,
"step": 1769
},
{
"epoch": 2.0947025747262504,
"grad_norm": 5.65625,
"learning_rate": 1.4401545754907187e-05,
"loss": 0.8649,
"step": 1770
},
{
"epoch": 2.0958863569103285,
"grad_norm": 5.40625,
"learning_rate": 1.4394116906060078e-05,
"loss": 1.1372,
"step": 1771
},
{
"epoch": 2.0970701390944066,
"grad_norm": 7.3125,
"learning_rate": 1.4386685050711593e-05,
"loss": 1.5032,
"step": 1772
},
{
"epoch": 2.0982539212784848,
"grad_norm": 5.78125,
"learning_rate": 1.4379250193946685e-05,
"loss": 0.9629,
"step": 1773
},
{
"epoch": 2.099437703462563,
"grad_norm": 6.84375,
"learning_rate": 1.4371812340852366e-05,
"loss": 1.4058,
"step": 1774
},
{
"epoch": 2.100621485646641,
"grad_norm": 5.3125,
"learning_rate": 1.4364371496517688e-05,
"loss": 1.0557,
"step": 1775
},
{
"epoch": 2.101805267830719,
"grad_norm": 6.5,
"learning_rate": 1.435692766603376e-05,
"loss": 1.4222,
"step": 1776
},
{
"epoch": 2.102989050014797,
"grad_norm": 5.5,
"learning_rate": 1.4349480854493724e-05,
"loss": 1.1683,
"step": 1777
},
{
"epoch": 2.1041728321988753,
"grad_norm": 4.9375,
"learning_rate": 1.4342031066992775e-05,
"loss": 0.9764,
"step": 1778
},
{
"epoch": 2.1053566143829534,
"grad_norm": 7.03125,
"learning_rate": 1.4334578308628131e-05,
"loss": 1.4692,
"step": 1779
},
{
"epoch": 2.1065403965670315,
"grad_norm": 6.15625,
"learning_rate": 1.4327122584499052e-05,
"loss": 1.2667,
"step": 1780
},
{
"epoch": 2.1077241787511096,
"grad_norm": 5.09375,
"learning_rate": 1.4319663899706818e-05,
"loss": 0.9703,
"step": 1781
},
{
"epoch": 2.1089079609351877,
"grad_norm": 5.46875,
"learning_rate": 1.4312202259354746e-05,
"loss": 1.0449,
"step": 1782
},
{
"epoch": 2.1100917431192663,
"grad_norm": 6.625,
"learning_rate": 1.4304737668548165e-05,
"loss": 1.287,
"step": 1783
},
{
"epoch": 2.1112755253033444,
"grad_norm": 4.625,
"learning_rate": 1.4297270132394432e-05,
"loss": 0.702,
"step": 1784
},
{
"epoch": 2.1124593074874225,
"grad_norm": 5.75,
"learning_rate": 1.4289799656002912e-05,
"loss": 1.0835,
"step": 1785
},
{
"epoch": 2.1136430896715006,
"grad_norm": 7.03125,
"learning_rate": 1.4282326244484983e-05,
"loss": 1.4025,
"step": 1786
},
{
"epoch": 2.1148268718555787,
"grad_norm": 6.28125,
"learning_rate": 1.4274849902954034e-05,
"loss": 1.3664,
"step": 1787
},
{
"epoch": 2.116010654039657,
"grad_norm": 4.375,
"learning_rate": 1.4267370636525457e-05,
"loss": 0.5946,
"step": 1788
},
{
"epoch": 2.117194436223735,
"grad_norm": 6.65625,
"learning_rate": 1.4259888450316641e-05,
"loss": 1.2476,
"step": 1789
},
{
"epoch": 2.118378218407813,
"grad_norm": 5.59375,
"learning_rate": 1.4252403349446986e-05,
"loss": 1.1951,
"step": 1790
},
{
"epoch": 2.119562000591891,
"grad_norm": 5.78125,
"learning_rate": 1.424491533903787e-05,
"loss": 1.1829,
"step": 1791
},
{
"epoch": 2.1207457827759693,
"grad_norm": 6.34375,
"learning_rate": 1.4237424424212673e-05,
"loss": 1.2083,
"step": 1792
},
{
"epoch": 2.1219295649600474,
"grad_norm": 4.46875,
"learning_rate": 1.4229930610096754e-05,
"loss": 0.8619,
"step": 1793
},
{
"epoch": 2.1231133471441255,
"grad_norm": 6.625,
"learning_rate": 1.4222433901817468e-05,
"loss": 1.1992,
"step": 1794
},
{
"epoch": 2.1242971293282036,
"grad_norm": 6.625,
"learning_rate": 1.4214934304504133e-05,
"loss": 1.3275,
"step": 1795
},
{
"epoch": 2.1254809115122817,
"grad_norm": 6.46875,
"learning_rate": 1.4207431823288058e-05,
"loss": 1.0135,
"step": 1796
},
{
"epoch": 2.12666469369636,
"grad_norm": 6.0,
"learning_rate": 1.419992646330252e-05,
"loss": 0.9161,
"step": 1797
},
{
"epoch": 2.127848475880438,
"grad_norm": 7.0,
"learning_rate": 1.4192418229682768e-05,
"loss": 1.4357,
"step": 1798
},
{
"epoch": 2.129032258064516,
"grad_norm": 5.96875,
"learning_rate": 1.4184907127566006e-05,
"loss": 0.9421,
"step": 1799
},
{
"epoch": 2.130216040248594,
"grad_norm": 5.9375,
"learning_rate": 1.4177393162091421e-05,
"loss": 1.2869,
"step": 1800
},
{
"epoch": 2.1313998224326722,
"grad_norm": 6.375,
"learning_rate": 1.4169876338400141e-05,
"loss": 0.9874,
"step": 1801
},
{
"epoch": 2.1325836046167503,
"grad_norm": 6.21875,
"learning_rate": 1.4162356661635262e-05,
"loss": 1.1604,
"step": 1802
},
{
"epoch": 2.1337673868008284,
"grad_norm": 6.75,
"learning_rate": 1.4154834136941817e-05,
"loss": 1.4548,
"step": 1803
},
{
"epoch": 2.134951168984907,
"grad_norm": 5.625,
"learning_rate": 1.414730876946681e-05,
"loss": 1.0765,
"step": 1804
},
{
"epoch": 2.136134951168985,
"grad_norm": 7.28125,
"learning_rate": 1.4139780564359162e-05,
"loss": 1.9211,
"step": 1805
},
{
"epoch": 2.137318733353063,
"grad_norm": 7.46875,
"learning_rate": 1.4132249526769765e-05,
"loss": 1.0523,
"step": 1806
},
{
"epoch": 2.1385025155371413,
"grad_norm": 5.6875,
"learning_rate": 1.412471566185143e-05,
"loss": 1.2022,
"step": 1807
},
{
"epoch": 2.1396862977212194,
"grad_norm": 5.0,
"learning_rate": 1.4117178974758903e-05,
"loss": 0.9633,
"step": 1808
},
{
"epoch": 2.1408700799052975,
"grad_norm": 6.53125,
"learning_rate": 1.4109639470648868e-05,
"loss": 1.1567,
"step": 1809
},
{
"epoch": 2.1420538620893756,
"grad_norm": 6.28125,
"learning_rate": 1.4102097154679938e-05,
"loss": 1.1429,
"step": 1810
},
{
"epoch": 2.1432376442734538,
"grad_norm": 6.09375,
"learning_rate": 1.4094552032012635e-05,
"loss": 1.027,
"step": 1811
},
{
"epoch": 2.144421426457532,
"grad_norm": 7.84375,
"learning_rate": 1.4087004107809424e-05,
"loss": 1.315,
"step": 1812
},
{
"epoch": 2.14560520864161,
"grad_norm": 6.96875,
"learning_rate": 1.4079453387234667e-05,
"loss": 1.33,
"step": 1813
},
{
"epoch": 2.146788990825688,
"grad_norm": 5.03125,
"learning_rate": 1.407189987545465e-05,
"loss": 0.9134,
"step": 1814
},
{
"epoch": 2.147972773009766,
"grad_norm": 5.6875,
"learning_rate": 1.4064343577637565e-05,
"loss": 1.0441,
"step": 1815
},
{
"epoch": 2.1491565551938443,
"grad_norm": 5.5,
"learning_rate": 1.4056784498953511e-05,
"loss": 1.109,
"step": 1816
},
{
"epoch": 2.1503403373779224,
"grad_norm": 5.5625,
"learning_rate": 1.404922264457449e-05,
"loss": 1.1907,
"step": 1817
},
{
"epoch": 2.1515241195620005,
"grad_norm": 5.9375,
"learning_rate": 1.4041658019674405e-05,
"loss": 1.0564,
"step": 1818
},
{
"epoch": 2.1527079017460786,
"grad_norm": 6.90625,
"learning_rate": 1.4034090629429047e-05,
"loss": 1.5409,
"step": 1819
},
{
"epoch": 2.1538916839301567,
"grad_norm": 5.28125,
"learning_rate": 1.402652047901611e-05,
"loss": 1.0863,
"step": 1820
},
{
"epoch": 2.155075466114235,
"grad_norm": 5.53125,
"learning_rate": 1.401894757361517e-05,
"loss": 0.811,
"step": 1821
},
{
"epoch": 2.156259248298313,
"grad_norm": 7.21875,
"learning_rate": 1.4011371918407686e-05,
"loss": 1.291,
"step": 1822
},
{
"epoch": 2.157443030482391,
"grad_norm": 5.84375,
"learning_rate": 1.4003793518577007e-05,
"loss": 0.9886,
"step": 1823
},
{
"epoch": 2.158626812666469,
"grad_norm": 6.40625,
"learning_rate": 1.3996212379308351e-05,
"loss": 1.4106,
"step": 1824
},
{
"epoch": 2.1598105948505477,
"grad_norm": 7.0,
"learning_rate": 1.3988628505788814e-05,
"loss": 1.7257,
"step": 1825
},
{
"epoch": 2.160994377034626,
"grad_norm": 5.53125,
"learning_rate": 1.3981041903207364e-05,
"loss": 1.0151,
"step": 1826
},
{
"epoch": 2.162178159218704,
"grad_norm": 5.1875,
"learning_rate": 1.3973452576754833e-05,
"loss": 0.8476,
"step": 1827
},
{
"epoch": 2.163361941402782,
"grad_norm": 6.125,
"learning_rate": 1.396586053162392e-05,
"loss": 1.2903,
"step": 1828
},
{
"epoch": 2.16454572358686,
"grad_norm": 6.75,
"learning_rate": 1.3958265773009184e-05,
"loss": 1.1922,
"step": 1829
},
{
"epoch": 2.1657295057709383,
"grad_norm": 5.09375,
"learning_rate": 1.3950668306107035e-05,
"loss": 0.6867,
"step": 1830
},
{
"epoch": 2.1669132879550164,
"grad_norm": 5.53125,
"learning_rate": 1.3943068136115745e-05,
"loss": 1.0349,
"step": 1831
},
{
"epoch": 2.1680970701390945,
"grad_norm": 6.03125,
"learning_rate": 1.3935465268235429e-05,
"loss": 1.0227,
"step": 1832
},
{
"epoch": 2.1692808523231726,
"grad_norm": 5.34375,
"learning_rate": 1.392785970766805e-05,
"loss": 0.9904,
"step": 1833
},
{
"epoch": 2.1704646345072507,
"grad_norm": 6.4375,
"learning_rate": 1.3920251459617413e-05,
"loss": 1.5908,
"step": 1834
},
{
"epoch": 2.171648416691329,
"grad_norm": 5.65625,
"learning_rate": 1.3912640529289163e-05,
"loss": 0.8865,
"step": 1835
},
{
"epoch": 2.172832198875407,
"grad_norm": 5.21875,
"learning_rate": 1.390502692189078e-05,
"loss": 0.9118,
"step": 1836
},
{
"epoch": 2.174015981059485,
"grad_norm": 4.9375,
"learning_rate": 1.389741064263157e-05,
"loss": 0.8927,
"step": 1837
},
{
"epoch": 2.175199763243563,
"grad_norm": 6.40625,
"learning_rate": 1.3889791696722676e-05,
"loss": 1.2936,
"step": 1838
},
{
"epoch": 2.1763835454276412,
"grad_norm": 5.21875,
"learning_rate": 1.388217008937706e-05,
"loss": 0.7724,
"step": 1839
},
{
"epoch": 2.1775673276117193,
"grad_norm": 5.84375,
"learning_rate": 1.3874545825809509e-05,
"loss": 0.9466,
"step": 1840
},
{
"epoch": 2.1787511097957974,
"grad_norm": 5.96875,
"learning_rate": 1.3866918911236624e-05,
"loss": 1.1639,
"step": 1841
},
{
"epoch": 2.1799348919798756,
"grad_norm": 7.1875,
"learning_rate": 1.3859289350876822e-05,
"loss": 1.0583,
"step": 1842
},
{
"epoch": 2.1811186741639537,
"grad_norm": 5.25,
"learning_rate": 1.3851657149950325e-05,
"loss": 0.9877,
"step": 1843
},
{
"epoch": 2.1823024563480318,
"grad_norm": 5.3125,
"learning_rate": 1.3844022313679167e-05,
"loss": 1.0549,
"step": 1844
},
{
"epoch": 2.18348623853211,
"grad_norm": 5.71875,
"learning_rate": 1.3836384847287186e-05,
"loss": 0.8688,
"step": 1845
},
{
"epoch": 2.1846700207161884,
"grad_norm": 7.1875,
"learning_rate": 1.3828744756000015e-05,
"loss": 1.3491,
"step": 1846
},
{
"epoch": 2.1858538029002665,
"grad_norm": 7.15625,
"learning_rate": 1.3821102045045088e-05,
"loss": 0.8907,
"step": 1847
},
{
"epoch": 2.1870375850843446,
"grad_norm": 6.71875,
"learning_rate": 1.381345671965163e-05,
"loss": 1.4325,
"step": 1848
},
{
"epoch": 2.1882213672684228,
"grad_norm": 6.0625,
"learning_rate": 1.380580878505065e-05,
"loss": 0.914,
"step": 1849
},
{
"epoch": 2.189405149452501,
"grad_norm": 6.46875,
"learning_rate": 1.3798158246474946e-05,
"loss": 1.0597,
"step": 1850
},
{
"epoch": 2.190588931636579,
"grad_norm": 6.40625,
"learning_rate": 1.3790505109159101e-05,
"loss": 1.2389,
"step": 1851
},
{
"epoch": 2.191772713820657,
"grad_norm": 5.65625,
"learning_rate": 1.3782849378339469e-05,
"loss": 0.9412,
"step": 1852
},
{
"epoch": 2.192956496004735,
"grad_norm": 7.03125,
"learning_rate": 1.3775191059254185e-05,
"loss": 1.5254,
"step": 1853
},
{
"epoch": 2.1941402781888133,
"grad_norm": 5.8125,
"learning_rate": 1.3767530157143154e-05,
"loss": 1.3729,
"step": 1854
},
{
"epoch": 2.1953240603728914,
"grad_norm": 5.78125,
"learning_rate": 1.3759866677248041e-05,
"loss": 0.6939,
"step": 1855
},
{
"epoch": 2.1965078425569695,
"grad_norm": 6.34375,
"learning_rate": 1.3752200624812282e-05,
"loss": 1.112,
"step": 1856
},
{
"epoch": 2.1976916247410476,
"grad_norm": 6.4375,
"learning_rate": 1.3744532005081072e-05,
"loss": 1.2412,
"step": 1857
},
{
"epoch": 2.1988754069251257,
"grad_norm": 5.5,
"learning_rate": 1.3736860823301364e-05,
"loss": 1.045,
"step": 1858
},
{
"epoch": 2.200059189109204,
"grad_norm": 5.4375,
"learning_rate": 1.3729187084721858e-05,
"loss": 1.1363,
"step": 1859
},
{
"epoch": 2.201242971293282,
"grad_norm": 6.9375,
"learning_rate": 1.3721510794593012e-05,
"loss": 1.3523,
"step": 1860
},
{
"epoch": 2.20242675347736,
"grad_norm": 6.5,
"learning_rate": 1.3713831958167018e-05,
"loss": 1.1485,
"step": 1861
},
{
"epoch": 2.203610535661438,
"grad_norm": 6.03125,
"learning_rate": 1.3706150580697826e-05,
"loss": 1.0084,
"step": 1862
},
{
"epoch": 2.2047943178455163,
"grad_norm": 5.53125,
"learning_rate": 1.369846666744111e-05,
"loss": 0.7215,
"step": 1863
},
{
"epoch": 2.2059781000295944,
"grad_norm": 7.0625,
"learning_rate": 1.3690780223654286e-05,
"loss": 0.9984,
"step": 1864
},
{
"epoch": 2.2071618822136725,
"grad_norm": 6.71875,
"learning_rate": 1.36830912545965e-05,
"loss": 0.9441,
"step": 1865
},
{
"epoch": 2.2083456643977506,
"grad_norm": 5.71875,
"learning_rate": 1.367539976552863e-05,
"loss": 0.9912,
"step": 1866
},
{
"epoch": 2.209529446581829,
"grad_norm": 6.59375,
"learning_rate": 1.366770576171327e-05,
"loss": 0.9525,
"step": 1867
},
{
"epoch": 2.2107132287659073,
"grad_norm": 6.875,
"learning_rate": 1.3660009248414735e-05,
"loss": 1.1646,
"step": 1868
},
{
"epoch": 2.2118970109499854,
"grad_norm": 6.71875,
"learning_rate": 1.3652310230899074e-05,
"loss": 1.0577,
"step": 1869
},
{
"epoch": 2.2130807931340635,
"grad_norm": 6.1875,
"learning_rate": 1.3644608714434027e-05,
"loss": 0.9646,
"step": 1870
},
{
"epoch": 2.2142645753181416,
"grad_norm": 7.78125,
"learning_rate": 1.3636904704289053e-05,
"loss": 1.4812,
"step": 1871
},
{
"epoch": 2.2154483575022197,
"grad_norm": 5.6875,
"learning_rate": 1.362919820573532e-05,
"loss": 0.957,
"step": 1872
},
{
"epoch": 2.216632139686298,
"grad_norm": 5.59375,
"learning_rate": 1.3621489224045696e-05,
"loss": 0.8143,
"step": 1873
},
{
"epoch": 2.217815921870376,
"grad_norm": 5.75,
"learning_rate": 1.3613777764494747e-05,
"loss": 0.8356,
"step": 1874
},
{
"epoch": 2.218999704054454,
"grad_norm": 5.375,
"learning_rate": 1.3606063832358736e-05,
"loss": 1.0012,
"step": 1875
},
{
"epoch": 2.220183486238532,
"grad_norm": 7.4375,
"learning_rate": 1.3598347432915617e-05,
"loss": 1.6375,
"step": 1876
},
{
"epoch": 2.2213672684226102,
"grad_norm": 6.25,
"learning_rate": 1.3590628571445033e-05,
"loss": 1.4594,
"step": 1877
},
{
"epoch": 2.2225510506066883,
"grad_norm": 5.6875,
"learning_rate": 1.358290725322831e-05,
"loss": 0.9049,
"step": 1878
},
{
"epoch": 2.2237348327907664,
"grad_norm": 6.6875,
"learning_rate": 1.3575183483548453e-05,
"loss": 1.47,
"step": 1879
},
{
"epoch": 2.2249186149748446,
"grad_norm": 6.46875,
"learning_rate": 1.3567457267690152e-05,
"loss": 1.1626,
"step": 1880
},
{
"epoch": 2.2261023971589227,
"grad_norm": 6.0625,
"learning_rate": 1.3559728610939767e-05,
"loss": 1.1702,
"step": 1881
},
{
"epoch": 2.2272861793430008,
"grad_norm": 6.65625,
"learning_rate": 1.3551997518585317e-05,
"loss": 1.2687,
"step": 1882
},
{
"epoch": 2.228469961527079,
"grad_norm": 5.78125,
"learning_rate": 1.3544263995916508e-05,
"loss": 0.7509,
"step": 1883
},
{
"epoch": 2.229653743711157,
"grad_norm": 7.40625,
"learning_rate": 1.3536528048224698e-05,
"loss": 1.5871,
"step": 1884
},
{
"epoch": 2.230837525895235,
"grad_norm": 8.375,
"learning_rate": 1.3528789680802896e-05,
"loss": 1.3444,
"step": 1885
},
{
"epoch": 2.232021308079313,
"grad_norm": 6.4375,
"learning_rate": 1.3521048898945778e-05,
"loss": 1.3582,
"step": 1886
},
{
"epoch": 2.2332050902633913,
"grad_norm": 5.28125,
"learning_rate": 1.3513305707949675e-05,
"loss": 0.8996,
"step": 1887
},
{
"epoch": 2.23438887244747,
"grad_norm": 5.09375,
"learning_rate": 1.3505560113112555e-05,
"loss": 0.8061,
"step": 1888
},
{
"epoch": 2.235572654631548,
"grad_norm": 6.25,
"learning_rate": 1.3497812119734037e-05,
"loss": 1.2961,
"step": 1889
},
{
"epoch": 2.236756436815626,
"grad_norm": 5.90625,
"learning_rate": 1.3490061733115383e-05,
"loss": 1.1839,
"step": 1890
},
{
"epoch": 2.237940218999704,
"grad_norm": 6.4375,
"learning_rate": 1.3482308958559483e-05,
"loss": 1.0899,
"step": 1891
},
{
"epoch": 2.2391240011837823,
"grad_norm": 5.59375,
"learning_rate": 1.3474553801370871e-05,
"loss": 1.1478,
"step": 1892
},
{
"epoch": 2.2403077833678604,
"grad_norm": 6.25,
"learning_rate": 1.3466796266855713e-05,
"loss": 1.0528,
"step": 1893
},
{
"epoch": 2.2414915655519385,
"grad_norm": 7.6875,
"learning_rate": 1.345903636032179e-05,
"loss": 1.0466,
"step": 1894
},
{
"epoch": 2.2426753477360166,
"grad_norm": 6.15625,
"learning_rate": 1.3451274087078514e-05,
"loss": 1.1358,
"step": 1895
},
{
"epoch": 2.2438591299200947,
"grad_norm": 5.90625,
"learning_rate": 1.3443509452436917e-05,
"loss": 0.9243,
"step": 1896
},
{
"epoch": 2.245042912104173,
"grad_norm": 7.375,
"learning_rate": 1.3435742461709635e-05,
"loss": 1.4078,
"step": 1897
},
{
"epoch": 2.246226694288251,
"grad_norm": 6.125,
"learning_rate": 1.342797312021094e-05,
"loss": 1.021,
"step": 1898
},
{
"epoch": 2.247410476472329,
"grad_norm": 4.96875,
"learning_rate": 1.342020143325669e-05,
"loss": 0.8379,
"step": 1899
},
{
"epoch": 2.248594258656407,
"grad_norm": 5.84375,
"learning_rate": 1.3412427406164353e-05,
"loss": 1.1004,
"step": 1900
},
{
"epoch": 2.2497780408404853,
"grad_norm": 7.0,
"learning_rate": 1.3404651044253006e-05,
"loss": 1.1769,
"step": 1901
},
{
"epoch": 2.2509618230245634,
"grad_norm": 6.5,
"learning_rate": 1.3396872352843318e-05,
"loss": 1.1685,
"step": 1902
},
{
"epoch": 2.2521456052086415,
"grad_norm": 6.71875,
"learning_rate": 1.3389091337257543e-05,
"loss": 1.0242,
"step": 1903
},
{
"epoch": 2.2533293873927196,
"grad_norm": 7.03125,
"learning_rate": 1.3381308002819546e-05,
"loss": 1.3217,
"step": 1904
},
{
"epoch": 2.2545131695767977,
"grad_norm": 4.625,
"learning_rate": 1.3373522354854765e-05,
"loss": 0.9074,
"step": 1905
},
{
"epoch": 2.255696951760876,
"grad_norm": 6.40625,
"learning_rate": 1.3365734398690217e-05,
"loss": 1.3418,
"step": 1906
},
{
"epoch": 2.2568807339449544,
"grad_norm": 5.4375,
"learning_rate": 1.3357944139654508e-05,
"loss": 1.0252,
"step": 1907
},
{
"epoch": 2.258064516129032,
"grad_norm": 5.84375,
"learning_rate": 1.335015158307782e-05,
"loss": 0.8775,
"step": 1908
},
{
"epoch": 2.2592482983131106,
"grad_norm": 6.71875,
"learning_rate": 1.334235673429189e-05,
"loss": 1.2716,
"step": 1909
},
{
"epoch": 2.2604320804971887,
"grad_norm": 5.5,
"learning_rate": 1.333455959863005e-05,
"loss": 0.6765,
"step": 1910
},
{
"epoch": 2.261615862681267,
"grad_norm": 7.34375,
"learning_rate": 1.332676018142718e-05,
"loss": 1.4399,
"step": 1911
},
{
"epoch": 2.262799644865345,
"grad_norm": 7.59375,
"learning_rate": 1.3318958488019716e-05,
"loss": 1.3465,
"step": 1912
},
{
"epoch": 2.263983427049423,
"grad_norm": 6.3125,
"learning_rate": 1.3311154523745663e-05,
"loss": 1.3535,
"step": 1913
},
{
"epoch": 2.265167209233501,
"grad_norm": 7.53125,
"learning_rate": 1.3303348293944585e-05,
"loss": 1.2832,
"step": 1914
},
{
"epoch": 2.2663509914175792,
"grad_norm": 5.78125,
"learning_rate": 1.3295539803957573e-05,
"loss": 1.1535,
"step": 1915
},
{
"epoch": 2.2675347736016573,
"grad_norm": 7.59375,
"learning_rate": 1.3287729059127288e-05,
"loss": 1.286,
"step": 1916
},
{
"epoch": 2.2687185557857354,
"grad_norm": 6.1875,
"learning_rate": 1.3279916064797924e-05,
"loss": 1.1713,
"step": 1917
},
{
"epoch": 2.2699023379698136,
"grad_norm": 6.21875,
"learning_rate": 1.327210082631521e-05,
"loss": 1.2359,
"step": 1918
},
{
"epoch": 2.2710861201538917,
"grad_norm": 7.0625,
"learning_rate": 1.3264283349026419e-05,
"loss": 1.4523,
"step": 1919
},
{
"epoch": 2.2722699023379698,
"grad_norm": 5.6875,
"learning_rate": 1.325646363828035e-05,
"loss": 1.079,
"step": 1920
},
{
"epoch": 2.273453684522048,
"grad_norm": 6.96875,
"learning_rate": 1.3248641699427332e-05,
"loss": 1.3552,
"step": 1921
},
{
"epoch": 2.274637466706126,
"grad_norm": 5.875,
"learning_rate": 1.324081753781922e-05,
"loss": 0.741,
"step": 1922
},
{
"epoch": 2.275821248890204,
"grad_norm": 6.0,
"learning_rate": 1.323299115880939e-05,
"loss": 1.3015,
"step": 1923
},
{
"epoch": 2.277005031074282,
"grad_norm": 6.71875,
"learning_rate": 1.3225162567752725e-05,
"loss": 0.9838,
"step": 1924
},
{
"epoch": 2.2781888132583603,
"grad_norm": 5.875,
"learning_rate": 1.3217331770005639e-05,
"loss": 0.9462,
"step": 1925
},
{
"epoch": 2.2793725954424384,
"grad_norm": 5.09375,
"learning_rate": 1.3209498770926043e-05,
"loss": 0.8233,
"step": 1926
},
{
"epoch": 2.2805563776265165,
"grad_norm": 6.78125,
"learning_rate": 1.3201663575873359e-05,
"loss": 1.1421,
"step": 1927
},
{
"epoch": 2.281740159810595,
"grad_norm": 6.21875,
"learning_rate": 1.3193826190208507e-05,
"loss": 1.105,
"step": 1928
},
{
"epoch": 2.2829239419946727,
"grad_norm": 5.71875,
"learning_rate": 1.318598661929392e-05,
"loss": 1.0519,
"step": 1929
},
{
"epoch": 2.2841077241787513,
"grad_norm": 5.75,
"learning_rate": 1.3178144868493502e-05,
"loss": 0.979,
"step": 1930
},
{
"epoch": 2.2852915063628294,
"grad_norm": 5.0,
"learning_rate": 1.3170300943172669e-05,
"loss": 0.7673,
"step": 1931
},
{
"epoch": 2.2864752885469075,
"grad_norm": 6.3125,
"learning_rate": 1.3162454848698318e-05,
"loss": 1.347,
"step": 1932
},
{
"epoch": 2.2876590707309856,
"grad_norm": 5.3125,
"learning_rate": 1.3154606590438828e-05,
"loss": 0.9615,
"step": 1933
},
{
"epoch": 2.2888428529150637,
"grad_norm": 5.34375,
"learning_rate": 1.3146756173764061e-05,
"loss": 0.7339,
"step": 1934
},
{
"epoch": 2.290026635099142,
"grad_norm": 6.28125,
"learning_rate": 1.3138903604045358e-05,
"loss": 1.1366,
"step": 1935
},
{
"epoch": 2.29121041728322,
"grad_norm": 6.25,
"learning_rate": 1.3131048886655529e-05,
"loss": 0.8346,
"step": 1936
},
{
"epoch": 2.292394199467298,
"grad_norm": 7.1875,
"learning_rate": 1.3123192026968851e-05,
"loss": 1.0252,
"step": 1937
},
{
"epoch": 2.293577981651376,
"grad_norm": 6.78125,
"learning_rate": 1.3115333030361078e-05,
"loss": 1.4224,
"step": 1938
},
{
"epoch": 2.2947617638354543,
"grad_norm": 6.125,
"learning_rate": 1.310747190220941e-05,
"loss": 1.1148,
"step": 1939
},
{
"epoch": 2.2959455460195324,
"grad_norm": 6.0,
"learning_rate": 1.309960864789252e-05,
"loss": 0.9155,
"step": 1940
},
{
"epoch": 2.2971293282036105,
"grad_norm": 5.90625,
"learning_rate": 1.3091743272790533e-05,
"loss": 0.9712,
"step": 1941
},
{
"epoch": 2.2983131103876886,
"grad_norm": 5.625,
"learning_rate": 1.3083875782285015e-05,
"loss": 0.9281,
"step": 1942
},
{
"epoch": 2.2994968925717667,
"grad_norm": 5.78125,
"learning_rate": 1.3076006181758989e-05,
"loss": 1.0501,
"step": 1943
},
{
"epoch": 2.300680674755845,
"grad_norm": 5.3125,
"learning_rate": 1.306813447659692e-05,
"loss": 0.835,
"step": 1944
},
{
"epoch": 2.301864456939923,
"grad_norm": 6.375,
"learning_rate": 1.3060260672184713e-05,
"loss": 1.3927,
"step": 1945
},
{
"epoch": 2.303048239124001,
"grad_norm": 6.6875,
"learning_rate": 1.3052384773909706e-05,
"loss": 1.312,
"step": 1946
},
{
"epoch": 2.304232021308079,
"grad_norm": 6.53125,
"learning_rate": 1.3044506787160676e-05,
"loss": 1.1177,
"step": 1947
},
{
"epoch": 2.3054158034921572,
"grad_norm": 6.3125,
"learning_rate": 1.3036626717327818e-05,
"loss": 1.1416,
"step": 1948
},
{
"epoch": 2.306599585676236,
"grad_norm": 7.53125,
"learning_rate": 1.3028744569802765e-05,
"loss": 1.4839,
"step": 1949
},
{
"epoch": 2.3077833678603135,
"grad_norm": 6.1875,
"learning_rate": 1.3020860349978564e-05,
"loss": 0.9814,
"step": 1950
},
{
"epoch": 2.308967150044392,
"grad_norm": 4.53125,
"learning_rate": 1.3012974063249684e-05,
"loss": 0.7533,
"step": 1951
},
{
"epoch": 2.31015093222847,
"grad_norm": 9.125,
"learning_rate": 1.3005085715012003e-05,
"loss": 1.5419,
"step": 1952
},
{
"epoch": 2.3113347144125482,
"grad_norm": 6.34375,
"learning_rate": 1.2997195310662816e-05,
"loss": 0.928,
"step": 1953
},
{
"epoch": 2.3125184965966263,
"grad_norm": 7.125,
"learning_rate": 1.2989302855600816e-05,
"loss": 1.2831,
"step": 1954
},
{
"epoch": 2.3137022787807044,
"grad_norm": 7.4375,
"learning_rate": 1.2981408355226109e-05,
"loss": 1.1198,
"step": 1955
},
{
"epoch": 2.3148860609647826,
"grad_norm": 6.5625,
"learning_rate": 1.2973511814940194e-05,
"loss": 1.129,
"step": 1956
},
{
"epoch": 2.3160698431488607,
"grad_norm": 5.65625,
"learning_rate": 1.2965613240145969e-05,
"loss": 1.062,
"step": 1957
},
{
"epoch": 2.3172536253329388,
"grad_norm": 6.8125,
"learning_rate": 1.2957712636247722e-05,
"loss": 1.8541,
"step": 1958
},
{
"epoch": 2.318437407517017,
"grad_norm": 5.65625,
"learning_rate": 1.2949810008651129e-05,
"loss": 0.8959,
"step": 1959
},
{
"epoch": 2.319621189701095,
"grad_norm": 6.625,
"learning_rate": 1.2941905362763252e-05,
"loss": 1.4,
"step": 1960
},
{
"epoch": 2.320804971885173,
"grad_norm": 5.0,
"learning_rate": 1.2933998703992531e-05,
"loss": 0.7484,
"step": 1961
},
{
"epoch": 2.321988754069251,
"grad_norm": 7.375,
"learning_rate": 1.2926090037748793e-05,
"loss": 1.3194,
"step": 1962
},
{
"epoch": 2.3231725362533293,
"grad_norm": 7.34375,
"learning_rate": 1.2918179369443224e-05,
"loss": 1.0944,
"step": 1963
},
{
"epoch": 2.3243563184374074,
"grad_norm": 5.65625,
"learning_rate": 1.291026670448839e-05,
"loss": 0.9337,
"step": 1964
},
{
"epoch": 2.3255401006214855,
"grad_norm": 6.3125,
"learning_rate": 1.2902352048298224e-05,
"loss": 1.2983,
"step": 1965
},
{
"epoch": 2.3267238828055636,
"grad_norm": 5.59375,
"learning_rate": 1.289443540628801e-05,
"loss": 1.2559,
"step": 1966
},
{
"epoch": 2.3279076649896417,
"grad_norm": 6.34375,
"learning_rate": 1.2886516783874403e-05,
"loss": 1.2228,
"step": 1967
},
{
"epoch": 2.32909144717372,
"grad_norm": 7.125,
"learning_rate": 1.2878596186475408e-05,
"loss": 1.0363,
"step": 1968
},
{
"epoch": 2.330275229357798,
"grad_norm": 5.75,
"learning_rate": 1.2870673619510382e-05,
"loss": 0.6718,
"step": 1969
},
{
"epoch": 2.3314590115418765,
"grad_norm": 6.1875,
"learning_rate": 1.2862749088400026e-05,
"loss": 1.0993,
"step": 1970
},
{
"epoch": 2.332642793725954,
"grad_norm": 6.6875,
"learning_rate": 1.2854822598566394e-05,
"loss": 0.9594,
"step": 1971
},
{
"epoch": 2.3338265759100327,
"grad_norm": 7.03125,
"learning_rate": 1.2846894155432868e-05,
"loss": 1.2486,
"step": 1972
},
{
"epoch": 2.335010358094111,
"grad_norm": 8.6875,
"learning_rate": 1.283896376442418e-05,
"loss": 1.6201,
"step": 1973
},
{
"epoch": 2.336194140278189,
"grad_norm": 7.28125,
"learning_rate": 1.283103143096638e-05,
"loss": 1.6344,
"step": 1974
},
{
"epoch": 2.337377922462267,
"grad_norm": 6.4375,
"learning_rate": 1.2823097160486861e-05,
"loss": 1.0768,
"step": 1975
},
{
"epoch": 2.338561704646345,
"grad_norm": 6.90625,
"learning_rate": 1.2815160958414332e-05,
"loss": 1.2733,
"step": 1976
},
{
"epoch": 2.3397454868304233,
"grad_norm": 6.5625,
"learning_rate": 1.280722283017883e-05,
"loss": 1.2414,
"step": 1977
},
{
"epoch": 2.3409292690145014,
"grad_norm": 6.03125,
"learning_rate": 1.2799282781211698e-05,
"loss": 1.2432,
"step": 1978
},
{
"epoch": 2.3421130511985795,
"grad_norm": 5.3125,
"learning_rate": 1.279134081694561e-05,
"loss": 0.8064,
"step": 1979
},
{
"epoch": 2.3432968333826576,
"grad_norm": 5.84375,
"learning_rate": 1.278339694281454e-05,
"loss": 0.9756,
"step": 1980
},
{
"epoch": 2.3444806155667357,
"grad_norm": 5.21875,
"learning_rate": 1.2775451164253767e-05,
"loss": 0.9617,
"step": 1981
},
{
"epoch": 2.345664397750814,
"grad_norm": 6.15625,
"learning_rate": 1.2767503486699884e-05,
"loss": 1.215,
"step": 1982
},
{
"epoch": 2.346848179934892,
"grad_norm": 7.34375,
"learning_rate": 1.275955391559077e-05,
"loss": 1.8853,
"step": 1983
},
{
"epoch": 2.34803196211897,
"grad_norm": 7.03125,
"learning_rate": 1.2751602456365608e-05,
"loss": 1.2337,
"step": 1984
},
{
"epoch": 2.349215744303048,
"grad_norm": 6.90625,
"learning_rate": 1.2743649114464873e-05,
"loss": 1.6591,
"step": 1985
},
{
"epoch": 2.3503995264871262,
"grad_norm": 6.3125,
"learning_rate": 1.2735693895330324e-05,
"loss": 0.8038,
"step": 1986
},
{
"epoch": 2.3515833086712044,
"grad_norm": 6.125,
"learning_rate": 1.2727736804405005e-05,
"loss": 1.1799,
"step": 1987
},
{
"epoch": 2.3527670908552825,
"grad_norm": 5.75,
"learning_rate": 1.2719777847133241e-05,
"loss": 0.9802,
"step": 1988
},
{
"epoch": 2.3539508730393606,
"grad_norm": 5.21875,
"learning_rate": 1.2711817028960642e-05,
"loss": 0.8198,
"step": 1989
},
{
"epoch": 2.3551346552234387,
"grad_norm": 6.375,
"learning_rate": 1.2703854355334073e-05,
"loss": 1.0484,
"step": 1990
},
{
"epoch": 2.3563184374075172,
"grad_norm": 6.4375,
"learning_rate": 1.2695889831701691e-05,
"loss": 1.4473,
"step": 1991
},
{
"epoch": 2.3575022195915953,
"grad_norm": 6.0625,
"learning_rate": 1.26879234635129e-05,
"loss": 1.0174,
"step": 1992
},
{
"epoch": 2.3586860017756734,
"grad_norm": 5.9375,
"learning_rate": 1.2679955256218379e-05,
"loss": 0.835,
"step": 1993
},
{
"epoch": 2.3598697839597516,
"grad_norm": 6.0625,
"learning_rate": 1.2671985215270055e-05,
"loss": 0.9207,
"step": 1994
},
{
"epoch": 2.3610535661438297,
"grad_norm": 6.9375,
"learning_rate": 1.2664013346121115e-05,
"loss": 1.2007,
"step": 1995
},
{
"epoch": 2.3622373483279078,
"grad_norm": 6.3125,
"learning_rate": 1.2656039654226e-05,
"loss": 1.1279,
"step": 1996
},
{
"epoch": 2.363421130511986,
"grad_norm": 6.5,
"learning_rate": 1.2648064145040392e-05,
"loss": 1.1935,
"step": 1997
},
{
"epoch": 2.364604912696064,
"grad_norm": 6.71875,
"learning_rate": 1.264008682402122e-05,
"loss": 1.2537,
"step": 1998
},
{
"epoch": 2.365788694880142,
"grad_norm": 9.3125,
"learning_rate": 1.2632107696626652e-05,
"loss": 1.5378,
"step": 1999
},
{
"epoch": 2.36697247706422,
"grad_norm": 6.78125,
"learning_rate": 1.2624126768316088e-05,
"loss": 1.1336,
"step": 2000
},
{
"epoch": 2.3681562592482983,
"grad_norm": 6.0625,
"learning_rate": 1.2616144044550167e-05,
"loss": 0.9359,
"step": 2001
},
{
"epoch": 2.3693400414323764,
"grad_norm": 5.75,
"learning_rate": 1.260815953079075e-05,
"loss": 1.0906,
"step": 2002
},
{
"epoch": 2.3705238236164545,
"grad_norm": 6.125,
"learning_rate": 1.2600173232500935e-05,
"loss": 1.2762,
"step": 2003
},
{
"epoch": 2.3717076058005326,
"grad_norm": 7.0,
"learning_rate": 1.2592185155145023e-05,
"loss": 1.2793,
"step": 2004
},
{
"epoch": 2.3728913879846107,
"grad_norm": 5.28125,
"learning_rate": 1.2584195304188542e-05,
"loss": 0.8875,
"step": 2005
},
{
"epoch": 2.374075170168689,
"grad_norm": 6.90625,
"learning_rate": 1.2576203685098233e-05,
"loss": 1.3477,
"step": 2006
},
{
"epoch": 2.375258952352767,
"grad_norm": 7.375,
"learning_rate": 1.256821030334205e-05,
"loss": 1.0658,
"step": 2007
},
{
"epoch": 2.376442734536845,
"grad_norm": 6.25,
"learning_rate": 1.2560215164389149e-05,
"loss": 0.9216,
"step": 2008
},
{
"epoch": 2.377626516720923,
"grad_norm": 5.8125,
"learning_rate": 1.2552218273709889e-05,
"loss": 0.9907,
"step": 2009
},
{
"epoch": 2.3788102989050013,
"grad_norm": 6.1875,
"learning_rate": 1.254421963677582e-05,
"loss": 1.005,
"step": 2010
},
{
"epoch": 2.3799940810890794,
"grad_norm": 7.90625,
"learning_rate": 1.2536219259059703e-05,
"loss": 1.0935,
"step": 2011
},
{
"epoch": 2.381177863273158,
"grad_norm": 8.0625,
"learning_rate": 1.2528217146035478e-05,
"loss": 1.5337,
"step": 2012
},
{
"epoch": 2.382361645457236,
"grad_norm": 4.15625,
"learning_rate": 1.2520213303178274e-05,
"loss": 0.5092,
"step": 2013
},
{
"epoch": 2.383545427641314,
"grad_norm": 7.125,
"learning_rate": 1.251220773596441e-05,
"loss": 1.2299,
"step": 2014
},
{
"epoch": 2.3847292098253923,
"grad_norm": 6.625,
"learning_rate": 1.2504200449871378e-05,
"loss": 1.0632,
"step": 2015
},
{
"epoch": 2.3859129920094704,
"grad_norm": 5.5,
"learning_rate": 1.2496191450377844e-05,
"loss": 1.0878,
"step": 2016
},
{
"epoch": 2.3870967741935485,
"grad_norm": 6.25,
"learning_rate": 1.2488180742963654e-05,
"loss": 0.9378,
"step": 2017
},
{
"epoch": 2.3882805563776266,
"grad_norm": 5.8125,
"learning_rate": 1.2480168333109818e-05,
"loss": 0.959,
"step": 2018
},
{
"epoch": 2.3894643385617047,
"grad_norm": 5.9375,
"learning_rate": 1.2472154226298513e-05,
"loss": 0.8095,
"step": 2019
},
{
"epoch": 2.390648120745783,
"grad_norm": 7.03125,
"learning_rate": 1.2464138428013072e-05,
"loss": 1.0412,
"step": 2020
},
{
"epoch": 2.391831902929861,
"grad_norm": 6.25,
"learning_rate": 1.2456120943737996e-05,
"loss": 0.9365,
"step": 2021
},
{
"epoch": 2.393015685113939,
"grad_norm": 6.25,
"learning_rate": 1.244810177895893e-05,
"loss": 1.061,
"step": 2022
},
{
"epoch": 2.394199467298017,
"grad_norm": 5.46875,
"learning_rate": 1.2440080939162666e-05,
"loss": 0.6417,
"step": 2023
},
{
"epoch": 2.3953832494820952,
"grad_norm": 6.125,
"learning_rate": 1.2432058429837153e-05,
"loss": 0.9734,
"step": 2024
},
{
"epoch": 2.3965670316661734,
"grad_norm": 6.375,
"learning_rate": 1.2424034256471478e-05,
"loss": 1.3163,
"step": 2025
},
{
"epoch": 2.3977508138502515,
"grad_norm": 6.96875,
"learning_rate": 1.241600842455586e-05,
"loss": 1.3848,
"step": 2026
},
{
"epoch": 2.3989345960343296,
"grad_norm": 6.1875,
"learning_rate": 1.2407980939581665e-05,
"loss": 0.8747,
"step": 2027
},
{
"epoch": 2.4001183782184077,
"grad_norm": 6.15625,
"learning_rate": 1.239995180704138e-05,
"loss": 1.1132,
"step": 2028
},
{
"epoch": 2.401302160402486,
"grad_norm": 6.875,
"learning_rate": 1.2391921032428622e-05,
"loss": 1.2572,
"step": 2029
},
{
"epoch": 2.402485942586564,
"grad_norm": 6.875,
"learning_rate": 1.2383888621238132e-05,
"loss": 1.6051,
"step": 2030
},
{
"epoch": 2.4036697247706424,
"grad_norm": 6.59375,
"learning_rate": 1.2375854578965771e-05,
"loss": 1.1624,
"step": 2031
},
{
"epoch": 2.40485350695472,
"grad_norm": 5.75,
"learning_rate": 1.2367818911108518e-05,
"loss": 0.9578,
"step": 2032
},
{
"epoch": 2.4060372891387987,
"grad_norm": 6.46875,
"learning_rate": 1.2359781623164465e-05,
"loss": 1.3102,
"step": 2033
},
{
"epoch": 2.4072210713228768,
"grad_norm": 7.4375,
"learning_rate": 1.2351742720632798e-05,
"loss": 1.3265,
"step": 2034
},
{
"epoch": 2.408404853506955,
"grad_norm": 5.28125,
"learning_rate": 1.234370220901383e-05,
"loss": 0.9041,
"step": 2035
},
{
"epoch": 2.409588635691033,
"grad_norm": 6.4375,
"learning_rate": 1.2335660093808962e-05,
"loss": 1.1648,
"step": 2036
},
{
"epoch": 2.410772417875111,
"grad_norm": 6.40625,
"learning_rate": 1.232761638052069e-05,
"loss": 1.141,
"step": 2037
},
{
"epoch": 2.411956200059189,
"grad_norm": 7.6875,
"learning_rate": 1.2319571074652614e-05,
"loss": 1.5409,
"step": 2038
},
{
"epoch": 2.4131399822432673,
"grad_norm": 6.5625,
"learning_rate": 1.2311524181709417e-05,
"loss": 1.1061,
"step": 2039
},
{
"epoch": 2.4143237644273454,
"grad_norm": 7.6875,
"learning_rate": 1.2303475707196866e-05,
"loss": 1.199,
"step": 2040
},
{
"epoch": 2.4155075466114235,
"grad_norm": 5.84375,
"learning_rate": 1.2295425656621812e-05,
"loss": 1.082,
"step": 2041
},
{
"epoch": 2.4166913287955016,
"grad_norm": 6.0,
"learning_rate": 1.2287374035492184e-05,
"loss": 1.0405,
"step": 2042
},
{
"epoch": 2.4178751109795797,
"grad_norm": 6.125,
"learning_rate": 1.2279320849316991e-05,
"loss": 1.0831,
"step": 2043
},
{
"epoch": 2.419058893163658,
"grad_norm": 6.6875,
"learning_rate": 1.2271266103606305e-05,
"loss": 1.3222,
"step": 2044
},
{
"epoch": 2.420242675347736,
"grad_norm": 6.1875,
"learning_rate": 1.2263209803871271e-05,
"loss": 1.2136,
"step": 2045
},
{
"epoch": 2.421426457531814,
"grad_norm": 6.8125,
"learning_rate": 1.2255151955624092e-05,
"loss": 1.2208,
"step": 2046
},
{
"epoch": 2.422610239715892,
"grad_norm": 6.75,
"learning_rate": 1.2247092564378032e-05,
"loss": 1.2223,
"step": 2047
},
{
"epoch": 2.4237940218999703,
"grad_norm": 6.8125,
"learning_rate": 1.2239031635647417e-05,
"loss": 1.4074,
"step": 2048
},
{
"epoch": 2.4249778040840484,
"grad_norm": 7.53125,
"learning_rate": 1.2230969174947619e-05,
"loss": 1.1282,
"step": 2049
},
{
"epoch": 2.4261615862681265,
"grad_norm": 6.46875,
"learning_rate": 1.2222905187795053e-05,
"loss": 1.0305,
"step": 2050
},
{
"epoch": 2.4273453684522046,
"grad_norm": 7.625,
"learning_rate": 1.2214839679707193e-05,
"loss": 1.4201,
"step": 2051
},
{
"epoch": 2.428529150636283,
"grad_norm": 6.96875,
"learning_rate": 1.220677265620254e-05,
"loss": 1.3339,
"step": 2052
},
{
"epoch": 2.429712932820361,
"grad_norm": 6.28125,
"learning_rate": 1.2198704122800633e-05,
"loss": 1.0518,
"step": 2053
},
{
"epoch": 2.4308967150044394,
"grad_norm": 7.0,
"learning_rate": 1.2190634085022056e-05,
"loss": 1.372,
"step": 2054
},
{
"epoch": 2.4320804971885175,
"grad_norm": 5.5,
"learning_rate": 1.2182562548388414e-05,
"loss": 0.9632,
"step": 2055
},
{
"epoch": 2.4332642793725956,
"grad_norm": 6.09375,
"learning_rate": 1.2174489518422333e-05,
"loss": 1.0101,
"step": 2056
},
{
"epoch": 2.4344480615566737,
"grad_norm": 6.40625,
"learning_rate": 1.216641500064747e-05,
"loss": 1.2422,
"step": 2057
},
{
"epoch": 2.435631843740752,
"grad_norm": 7.3125,
"learning_rate": 1.2158339000588493e-05,
"loss": 1.2035,
"step": 2058
},
{
"epoch": 2.43681562592483,
"grad_norm": 5.4375,
"learning_rate": 1.2150261523771083e-05,
"loss": 1.0915,
"step": 2059
},
{
"epoch": 2.437999408108908,
"grad_norm": 5.40625,
"learning_rate": 1.2142182575721946e-05,
"loss": 1.0424,
"step": 2060
},
{
"epoch": 2.439183190292986,
"grad_norm": 6.15625,
"learning_rate": 1.2134102161968775e-05,
"loss": 1.0327,
"step": 2061
},
{
"epoch": 2.4403669724770642,
"grad_norm": 8.375,
"learning_rate": 1.212602028804028e-05,
"loss": 1.4762,
"step": 2062
},
{
"epoch": 2.4415507546611424,
"grad_norm": 6.625,
"learning_rate": 1.211793695946616e-05,
"loss": 1.3659,
"step": 2063
},
{
"epoch": 2.4427345368452205,
"grad_norm": 6.78125,
"learning_rate": 1.2109852181777117e-05,
"loss": 0.8306,
"step": 2064
},
{
"epoch": 2.4439183190292986,
"grad_norm": 5.9375,
"learning_rate": 1.210176596050484e-05,
"loss": 1.3483,
"step": 2065
},
{
"epoch": 2.4451021012133767,
"grad_norm": 5.84375,
"learning_rate": 1.2093678301182013e-05,
"loss": 0.9496,
"step": 2066
},
{
"epoch": 2.446285883397455,
"grad_norm": 6.0,
"learning_rate": 1.2085589209342287e-05,
"loss": 1.0806,
"step": 2067
},
{
"epoch": 2.447469665581533,
"grad_norm": 5.84375,
"learning_rate": 1.2077498690520316e-05,
"loss": 1.0137,
"step": 2068
},
{
"epoch": 2.448653447765611,
"grad_norm": 6.34375,
"learning_rate": 1.2069406750251713e-05,
"loss": 1.2828,
"step": 2069
},
{
"epoch": 2.449837229949689,
"grad_norm": 7.0625,
"learning_rate": 1.206131339407307e-05,
"loss": 0.7411,
"step": 2070
},
{
"epoch": 2.451021012133767,
"grad_norm": 7.125,
"learning_rate": 1.2053218627521946e-05,
"loss": 1.5642,
"step": 2071
},
{
"epoch": 2.4522047943178453,
"grad_norm": 7.25,
"learning_rate": 1.204512245613687e-05,
"loss": 1.8302,
"step": 2072
},
{
"epoch": 2.453388576501924,
"grad_norm": 6.46875,
"learning_rate": 1.2037024885457325e-05,
"loss": 1.0337,
"step": 2073
},
{
"epoch": 2.4545723586860015,
"grad_norm": 6.46875,
"learning_rate": 1.2028925921023753e-05,
"loss": 1.2193,
"step": 2074
},
{
"epoch": 2.45575614087008,
"grad_norm": 6.6875,
"learning_rate": 1.2020825568377556e-05,
"loss": 1.1718,
"step": 2075
},
{
"epoch": 2.456939923054158,
"grad_norm": 6.9375,
"learning_rate": 1.2012723833061077e-05,
"loss": 1.362,
"step": 2076
},
{
"epoch": 2.4581237052382363,
"grad_norm": 6.40625,
"learning_rate": 1.2004620720617613e-05,
"loss": 0.9497,
"step": 2077
},
{
"epoch": 2.4593074874223144,
"grad_norm": 5.59375,
"learning_rate": 1.1996516236591398e-05,
"loss": 0.8303,
"step": 2078
},
{
"epoch": 2.4604912696063925,
"grad_norm": 6.25,
"learning_rate": 1.1988410386527605e-05,
"loss": 1.3864,
"step": 2079
},
{
"epoch": 2.4616750517904706,
"grad_norm": 6.09375,
"learning_rate": 1.1980303175972343e-05,
"loss": 1.09,
"step": 2080
},
{
"epoch": 2.4628588339745487,
"grad_norm": 7.03125,
"learning_rate": 1.1972194610472655e-05,
"loss": 1.3681,
"step": 2081
},
{
"epoch": 2.464042616158627,
"grad_norm": 6.9375,
"learning_rate": 1.1964084695576496e-05,
"loss": 1.0458,
"step": 2082
},
{
"epoch": 2.465226398342705,
"grad_norm": 6.8125,
"learning_rate": 1.1955973436832768e-05,
"loss": 1.5279,
"step": 2083
},
{
"epoch": 2.466410180526783,
"grad_norm": 6.3125,
"learning_rate": 1.1947860839791276e-05,
"loss": 1.1974,
"step": 2084
},
{
"epoch": 2.467593962710861,
"grad_norm": 5.46875,
"learning_rate": 1.1939746910002745e-05,
"loss": 1.0416,
"step": 2085
},
{
"epoch": 2.4687777448949393,
"grad_norm": 6.25,
"learning_rate": 1.1931631653018812e-05,
"loss": 1.2608,
"step": 2086
},
{
"epoch": 2.4699615270790174,
"grad_norm": 6.625,
"learning_rate": 1.1923515074392022e-05,
"loss": 1.274,
"step": 2087
},
{
"epoch": 2.4711453092630955,
"grad_norm": 7.21875,
"learning_rate": 1.191539717967582e-05,
"loss": 1.4002,
"step": 2088
},
{
"epoch": 2.4723290914471736,
"grad_norm": 6.875,
"learning_rate": 1.1907277974424562e-05,
"loss": 1.5965,
"step": 2089
},
{
"epoch": 2.4735128736312517,
"grad_norm": 6.0625,
"learning_rate": 1.1899157464193493e-05,
"loss": 1.0529,
"step": 2090
},
{
"epoch": 2.47469665581533,
"grad_norm": 6.59375,
"learning_rate": 1.189103565453875e-05,
"loss": 1.1049,
"step": 2091
},
{
"epoch": 2.475880437999408,
"grad_norm": 7.25,
"learning_rate": 1.1882912551017362e-05,
"loss": 1.2907,
"step": 2092
},
{
"epoch": 2.477064220183486,
"grad_norm": 6.25,
"learning_rate": 1.1874788159187243e-05,
"loss": 0.9837,
"step": 2093
},
{
"epoch": 2.4782480023675646,
"grad_norm": 5.28125,
"learning_rate": 1.1866662484607186e-05,
"loss": 0.9067,
"step": 2094
},
{
"epoch": 2.4794317845516423,
"grad_norm": 5.40625,
"learning_rate": 1.1858535532836866e-05,
"loss": 0.9754,
"step": 2095
},
{
"epoch": 2.480615566735721,
"grad_norm": 7.46875,
"learning_rate": 1.1850407309436831e-05,
"loss": 1.4578,
"step": 2096
},
{
"epoch": 2.481799348919799,
"grad_norm": 6.78125,
"learning_rate": 1.1842277819968493e-05,
"loss": 1.4494,
"step": 2097
},
{
"epoch": 2.482983131103877,
"grad_norm": 6.03125,
"learning_rate": 1.183414706999414e-05,
"loss": 1.1405,
"step": 2098
},
{
"epoch": 2.484166913287955,
"grad_norm": 6.59375,
"learning_rate": 1.1826015065076912e-05,
"loss": 0.9289,
"step": 2099
},
{
"epoch": 2.4853506954720332,
"grad_norm": 7.21875,
"learning_rate": 1.1817881810780815e-05,
"loss": 1.5993,
"step": 2100
},
{
"epoch": 2.4865344776561114,
"grad_norm": 6.4375,
"learning_rate": 1.1809747312670711e-05,
"loss": 0.8541,
"step": 2101
},
{
"epoch": 2.4877182598401895,
"grad_norm": 7.0,
"learning_rate": 1.180161157631231e-05,
"loss": 1.4758,
"step": 2102
},
{
"epoch": 2.4889020420242676,
"grad_norm": 7.21875,
"learning_rate": 1.1793474607272164e-05,
"loss": 1.1511,
"step": 2103
},
{
"epoch": 2.4900858242083457,
"grad_norm": 6.1875,
"learning_rate": 1.1785336411117675e-05,
"loss": 1.3136,
"step": 2104
},
{
"epoch": 2.491269606392424,
"grad_norm": 6.96875,
"learning_rate": 1.1777196993417087e-05,
"loss": 1.3797,
"step": 2105
},
{
"epoch": 2.492453388576502,
"grad_norm": 6.0,
"learning_rate": 1.176905635973947e-05,
"loss": 1.2016,
"step": 2106
},
{
"epoch": 2.49363717076058,
"grad_norm": 5.46875,
"learning_rate": 1.1760914515654735e-05,
"loss": 0.9609,
"step": 2107
},
{
"epoch": 2.494820952944658,
"grad_norm": 6.625,
"learning_rate": 1.1752771466733622e-05,
"loss": 1.5196,
"step": 2108
},
{
"epoch": 2.496004735128736,
"grad_norm": 6.84375,
"learning_rate": 1.1744627218547683e-05,
"loss": 1.2287,
"step": 2109
},
{
"epoch": 2.4971885173128143,
"grad_norm": 6.34375,
"learning_rate": 1.1736481776669307e-05,
"loss": 1.0843,
"step": 2110
},
{
"epoch": 2.4983722994968924,
"grad_norm": 6.21875,
"learning_rate": 1.1728335146671683e-05,
"loss": 1.2175,
"step": 2111
},
{
"epoch": 2.4995560816809705,
"grad_norm": 7.0,
"learning_rate": 1.172018733412883e-05,
"loss": 0.9741,
"step": 2112
},
{
"epoch": 2.500739863865049,
"grad_norm": 7.125,
"learning_rate": 1.1712038344615564e-05,
"loss": 1.4846,
"step": 2113
},
{
"epoch": 2.5019236460491268,
"grad_norm": 4.75,
"learning_rate": 1.1703888183707513e-05,
"loss": 0.7569,
"step": 2114
},
{
"epoch": 2.5031074282332053,
"grad_norm": 6.4375,
"learning_rate": 1.1695736856981097e-05,
"loss": 1.1553,
"step": 2115
},
{
"epoch": 2.504291210417283,
"grad_norm": 5.84375,
"learning_rate": 1.1687584370013545e-05,
"loss": 1.2366,
"step": 2116
},
{
"epoch": 2.5054749926013615,
"grad_norm": 7.0,
"learning_rate": 1.1679430728382875e-05,
"loss": 1.0707,
"step": 2117
},
{
"epoch": 2.5066587747854396,
"grad_norm": 5.15625,
"learning_rate": 1.1671275937667895e-05,
"loss": 0.8996,
"step": 2118
},
{
"epoch": 2.5078425569695177,
"grad_norm": 7.9375,
"learning_rate": 1.1663120003448195e-05,
"loss": 1.5887,
"step": 2119
},
{
"epoch": 2.509026339153596,
"grad_norm": 6.65625,
"learning_rate": 1.1654962931304159e-05,
"loss": 1.0588,
"step": 2120
},
{
"epoch": 2.510210121337674,
"grad_norm": 5.8125,
"learning_rate": 1.1646804726816937e-05,
"loss": 0.9096,
"step": 2121
},
{
"epoch": 2.511393903521752,
"grad_norm": 5.53125,
"learning_rate": 1.1638645395568458e-05,
"loss": 0.8244,
"step": 2122
},
{
"epoch": 2.51257768570583,
"grad_norm": 5.875,
"learning_rate": 1.1630484943141428e-05,
"loss": 1.1549,
"step": 2123
},
{
"epoch": 2.5137614678899083,
"grad_norm": 7.40625,
"learning_rate": 1.1622323375119312e-05,
"loss": 1.3771,
"step": 2124
},
{
"epoch": 2.5149452500739864,
"grad_norm": 6.90625,
"learning_rate": 1.161416069708634e-05,
"loss": 1.1485,
"step": 2125
},
{
"epoch": 2.5161290322580645,
"grad_norm": 6.46875,
"learning_rate": 1.1605996914627509e-05,
"loss": 1.2642,
"step": 2126
},
{
"epoch": 2.5173128144421426,
"grad_norm": 7.09375,
"learning_rate": 1.1597832033328558e-05,
"loss": 1.3856,
"step": 2127
},
{
"epoch": 2.5184965966262207,
"grad_norm": 6.75,
"learning_rate": 1.1589666058775986e-05,
"loss": 1.4691,
"step": 2128
},
{
"epoch": 2.519680378810299,
"grad_norm": 7.28125,
"learning_rate": 1.1581498996557046e-05,
"loss": 1.5148,
"step": 2129
},
{
"epoch": 2.520864160994377,
"grad_norm": 6.25,
"learning_rate": 1.1573330852259724e-05,
"loss": 1.0071,
"step": 2130
},
{
"epoch": 2.522047943178455,
"grad_norm": 6.25,
"learning_rate": 1.156516163147275e-05,
"loss": 1.2114,
"step": 2131
},
{
"epoch": 2.523231725362533,
"grad_norm": 6.71875,
"learning_rate": 1.1556991339785595e-05,
"loss": 1.2458,
"step": 2132
},
{
"epoch": 2.5244155075466113,
"grad_norm": 6.46875,
"learning_rate": 1.1548819982788455e-05,
"loss": 1.2652,
"step": 2133
},
{
"epoch": 2.52559928973069,
"grad_norm": 4.71875,
"learning_rate": 1.1540647566072259e-05,
"loss": 0.5501,
"step": 2134
},
{
"epoch": 2.5267830719147675,
"grad_norm": 6.9375,
"learning_rate": 1.1532474095228667e-05,
"loss": 1.111,
"step": 2135
},
{
"epoch": 2.527966854098846,
"grad_norm": 6.09375,
"learning_rate": 1.1524299575850048e-05,
"loss": 0.7126,
"step": 2136
},
{
"epoch": 2.5291506362829237,
"grad_norm": 6.46875,
"learning_rate": 1.15161240135295e-05,
"loss": 0.9365,
"step": 2137
},
{
"epoch": 2.5303344184670022,
"grad_norm": 7.375,
"learning_rate": 1.1507947413860827e-05,
"loss": 0.9425,
"step": 2138
},
{
"epoch": 2.5315182006510804,
"grad_norm": 6.96875,
"learning_rate": 1.1499769782438547e-05,
"loss": 1.3125,
"step": 2139
},
{
"epoch": 2.5327019828351585,
"grad_norm": 6.125,
"learning_rate": 1.1491591124857873e-05,
"loss": 0.9106,
"step": 2140
},
{
"epoch": 2.5338857650192366,
"grad_norm": 5.75,
"learning_rate": 1.1483411446714744e-05,
"loss": 0.999,
"step": 2141
},
{
"epoch": 2.5350695472033147,
"grad_norm": 6.6875,
"learning_rate": 1.147523075360577e-05,
"loss": 1.1613,
"step": 2142
},
{
"epoch": 2.536253329387393,
"grad_norm": 6.34375,
"learning_rate": 1.1467049051128275e-05,
"loss": 1.1824,
"step": 2143
},
{
"epoch": 2.537437111571471,
"grad_norm": 6.75,
"learning_rate": 1.1458866344880266e-05,
"loss": 1.1473,
"step": 2144
},
{
"epoch": 2.538620893755549,
"grad_norm": 6.90625,
"learning_rate": 1.1450682640460434e-05,
"loss": 1.539,
"step": 2145
},
{
"epoch": 2.539804675939627,
"grad_norm": 5.84375,
"learning_rate": 1.1442497943468158e-05,
"loss": 0.6965,
"step": 2146
},
{
"epoch": 2.540988458123705,
"grad_norm": 4.84375,
"learning_rate": 1.1434312259503498e-05,
"loss": 0.7204,
"step": 2147
},
{
"epoch": 2.5421722403077833,
"grad_norm": 7.40625,
"learning_rate": 1.1426125594167179e-05,
"loss": 1.3211,
"step": 2148
},
{
"epoch": 2.5433560224918614,
"grad_norm": 5.90625,
"learning_rate": 1.1417937953060607e-05,
"loss": 1.132,
"step": 2149
},
{
"epoch": 2.5445398046759395,
"grad_norm": 6.15625,
"learning_rate": 1.1409749341785859e-05,
"loss": 1.2075,
"step": 2150
},
{
"epoch": 2.5457235868600177,
"grad_norm": 5.15625,
"learning_rate": 1.1401559765945657e-05,
"loss": 0.816,
"step": 2151
},
{
"epoch": 2.5469073690440958,
"grad_norm": 7.3125,
"learning_rate": 1.1393369231143406e-05,
"loss": 1.3949,
"step": 2152
},
{
"epoch": 2.548091151228174,
"grad_norm": 7.84375,
"learning_rate": 1.1385177742983155e-05,
"loss": 1.1718,
"step": 2153
},
{
"epoch": 2.549274933412252,
"grad_norm": 7.65625,
"learning_rate": 1.1376985307069606e-05,
"loss": 1.4287,
"step": 2154
},
{
"epoch": 2.5504587155963305,
"grad_norm": 5.90625,
"learning_rate": 1.1368791929008107e-05,
"loss": 1.2034,
"step": 2155
},
{
"epoch": 2.551642497780408,
"grad_norm": 7.375,
"learning_rate": 1.1360597614404663e-05,
"loss": 1.6369,
"step": 2156
},
{
"epoch": 2.5528262799644867,
"grad_norm": 6.9375,
"learning_rate": 1.1352402368865898e-05,
"loss": 1.2977,
"step": 2157
},
{
"epoch": 2.5540100621485644,
"grad_norm": 6.84375,
"learning_rate": 1.1344206197999095e-05,
"loss": 1.2282,
"step": 2158
},
{
"epoch": 2.555193844332643,
"grad_norm": 5.96875,
"learning_rate": 1.1336009107412162e-05,
"loss": 0.915,
"step": 2159
},
{
"epoch": 2.556377626516721,
"grad_norm": 6.59375,
"learning_rate": 1.1327811102713633e-05,
"loss": 0.971,
"step": 2160
},
{
"epoch": 2.557561408700799,
"grad_norm": 6.1875,
"learning_rate": 1.1319612189512665e-05,
"loss": 0.931,
"step": 2161
},
{
"epoch": 2.5587451908848773,
"grad_norm": 7.21875,
"learning_rate": 1.1311412373419051e-05,
"loss": 1.5765,
"step": 2162
},
{
"epoch": 2.5599289730689554,
"grad_norm": 5.71875,
"learning_rate": 1.1303211660043181e-05,
"loss": 0.993,
"step": 2163
},
{
"epoch": 2.5611127552530335,
"grad_norm": 6.53125,
"learning_rate": 1.1295010054996079e-05,
"loss": 1.1583,
"step": 2164
},
{
"epoch": 2.5622965374371116,
"grad_norm": 6.15625,
"learning_rate": 1.1286807563889366e-05,
"loss": 0.9841,
"step": 2165
},
{
"epoch": 2.5634803196211897,
"grad_norm": 6.8125,
"learning_rate": 1.1278604192335273e-05,
"loss": 1.3719,
"step": 2166
},
{
"epoch": 2.564664101805268,
"grad_norm": 7.28125,
"learning_rate": 1.1270399945946638e-05,
"loss": 1.4305,
"step": 2167
},
{
"epoch": 2.565847883989346,
"grad_norm": 7.25,
"learning_rate": 1.1262194830336888e-05,
"loss": 1.2178,
"step": 2168
},
{
"epoch": 2.567031666173424,
"grad_norm": 6.9375,
"learning_rate": 1.125398885112005e-05,
"loss": 1.4484,
"step": 2169
},
{
"epoch": 2.568215448357502,
"grad_norm": 6.9375,
"learning_rate": 1.1245782013910748e-05,
"loss": 1.4122,
"step": 2170
},
{
"epoch": 2.5693992305415803,
"grad_norm": 6.125,
"learning_rate": 1.1237574324324185e-05,
"loss": 1.066,
"step": 2171
},
{
"epoch": 2.5705830127256584,
"grad_norm": 6.5625,
"learning_rate": 1.1229365787976145e-05,
"loss": 0.851,
"step": 2172
},
{
"epoch": 2.5717667949097365,
"grad_norm": 6.78125,
"learning_rate": 1.1221156410482998e-05,
"loss": 1.3252,
"step": 2173
},
{
"epoch": 2.5729505770938146,
"grad_norm": 6.5625,
"learning_rate": 1.1212946197461688e-05,
"loss": 1.1035,
"step": 2174
},
{
"epoch": 2.5741343592778927,
"grad_norm": 7.25,
"learning_rate": 1.1204735154529728e-05,
"loss": 1.0301,
"step": 2175
},
{
"epoch": 2.5753181414619712,
"grad_norm": 6.9375,
"learning_rate": 1.1196523287305204e-05,
"loss": 1.2923,
"step": 2176
},
{
"epoch": 2.576501923646049,
"grad_norm": 6.1875,
"learning_rate": 1.118831060140676e-05,
"loss": 0.9894,
"step": 2177
},
{
"epoch": 2.5776857058301275,
"grad_norm": 5.71875,
"learning_rate": 1.1180097102453604e-05,
"loss": 0.9023,
"step": 2178
},
{
"epoch": 2.578869488014205,
"grad_norm": 4.84375,
"learning_rate": 1.1171882796065496e-05,
"loss": 0.5431,
"step": 2179
},
{
"epoch": 2.5800532701982837,
"grad_norm": 7.03125,
"learning_rate": 1.1163667687862755e-05,
"loss": 1.1523,
"step": 2180
},
{
"epoch": 2.581237052382362,
"grad_norm": 5.28125,
"learning_rate": 1.1155451783466244e-05,
"loss": 0.7373,
"step": 2181
},
{
"epoch": 2.58242083456644,
"grad_norm": 6.0,
"learning_rate": 1.1147235088497372e-05,
"loss": 1.1931,
"step": 2182
},
{
"epoch": 2.583604616750518,
"grad_norm": 6.90625,
"learning_rate": 1.1139017608578088e-05,
"loss": 1.4936,
"step": 2183
},
{
"epoch": 2.584788398934596,
"grad_norm": 6.90625,
"learning_rate": 1.1130799349330881e-05,
"loss": 1.1673,
"step": 2184
},
{
"epoch": 2.585972181118674,
"grad_norm": 5.84375,
"learning_rate": 1.1122580316378769e-05,
"loss": 0.9239,
"step": 2185
},
{
"epoch": 2.5871559633027523,
"grad_norm": 7.0625,
"learning_rate": 1.1114360515345301e-05,
"loss": 0.8895,
"step": 2186
},
{
"epoch": 2.5883397454868304,
"grad_norm": 6.28125,
"learning_rate": 1.1106139951854555e-05,
"loss": 0.9796,
"step": 2187
},
{
"epoch": 2.5895235276709085,
"grad_norm": 5.375,
"learning_rate": 1.1097918631531123e-05,
"loss": 0.8569,
"step": 2188
},
{
"epoch": 2.5907073098549867,
"grad_norm": 6.71875,
"learning_rate": 1.1089696560000128e-05,
"loss": 1.1965,
"step": 2189
},
{
"epoch": 2.5918910920390648,
"grad_norm": 5.90625,
"learning_rate": 1.108147374288719e-05,
"loss": 1.1392,
"step": 2190
},
{
"epoch": 2.593074874223143,
"grad_norm": 6.5,
"learning_rate": 1.1073250185818449e-05,
"loss": 1.2889,
"step": 2191
},
{
"epoch": 2.594258656407221,
"grad_norm": 6.28125,
"learning_rate": 1.1065025894420551e-05,
"loss": 1.2258,
"step": 2192
},
{
"epoch": 2.595442438591299,
"grad_norm": 5.34375,
"learning_rate": 1.1056800874320646e-05,
"loss": 0.76,
"step": 2193
},
{
"epoch": 2.596626220775377,
"grad_norm": 6.625,
"learning_rate": 1.1048575131146377e-05,
"loss": 1.3329,
"step": 2194
},
{
"epoch": 2.5978100029594553,
"grad_norm": 6.6875,
"learning_rate": 1.1040348670525889e-05,
"loss": 1.1724,
"step": 2195
},
{
"epoch": 2.5989937851435334,
"grad_norm": 5.875,
"learning_rate": 1.1032121498087807e-05,
"loss": 1.1201,
"step": 2196
},
{
"epoch": 2.600177567327612,
"grad_norm": 6.78125,
"learning_rate": 1.1023893619461253e-05,
"loss": 1.14,
"step": 2197
},
{
"epoch": 2.6013613495116896,
"grad_norm": 5.125,
"learning_rate": 1.1015665040275827e-05,
"loss": 0.7796,
"step": 2198
},
{
"epoch": 2.602545131695768,
"grad_norm": 5.96875,
"learning_rate": 1.1007435766161612e-05,
"loss": 1.0211,
"step": 2199
},
{
"epoch": 2.603728913879846,
"grad_norm": 5.78125,
"learning_rate": 1.0999205802749165e-05,
"loss": 1.3143,
"step": 2200
},
{
"epoch": 2.6049126960639244,
"grad_norm": 8.1875,
"learning_rate": 1.0990975155669516e-05,
"loss": 1.6023,
"step": 2201
},
{
"epoch": 2.6060964782480025,
"grad_norm": 6.96875,
"learning_rate": 1.0982743830554157e-05,
"loss": 1.6317,
"step": 2202
},
{
"epoch": 2.6072802604320806,
"grad_norm": 5.375,
"learning_rate": 1.0974511833035048e-05,
"loss": 0.7254,
"step": 2203
},
{
"epoch": 2.6084640426161587,
"grad_norm": 6.3125,
"learning_rate": 1.096627916874461e-05,
"loss": 1.0981,
"step": 2204
},
{
"epoch": 2.609647824800237,
"grad_norm": 7.15625,
"learning_rate": 1.0958045843315722e-05,
"loss": 1.4093,
"step": 2205
},
{
"epoch": 2.610831606984315,
"grad_norm": 7.03125,
"learning_rate": 1.0949811862381708e-05,
"loss": 1.2543,
"step": 2206
},
{
"epoch": 2.612015389168393,
"grad_norm": 6.5,
"learning_rate": 1.094157723157635e-05,
"loss": 1.0786,
"step": 2207
},
{
"epoch": 2.613199171352471,
"grad_norm": 7.625,
"learning_rate": 1.0933341956533865e-05,
"loss": 0.9444,
"step": 2208
},
{
"epoch": 2.6143829535365493,
"grad_norm": 6.375,
"learning_rate": 1.0925106042888915e-05,
"loss": 0.9605,
"step": 2209
},
{
"epoch": 2.6155667357206274,
"grad_norm": 6.09375,
"learning_rate": 1.0916869496276604e-05,
"loss": 0.7963,
"step": 2210
},
{
"epoch": 2.6167505179047055,
"grad_norm": 5.96875,
"learning_rate": 1.0908632322332464e-05,
"loss": 1.1557,
"step": 2211
},
{
"epoch": 2.6179343000887836,
"grad_norm": 5.96875,
"learning_rate": 1.0900394526692453e-05,
"loss": 0.8266,
"step": 2212
},
{
"epoch": 2.6191180822728617,
"grad_norm": 6.8125,
"learning_rate": 1.0892156114992963e-05,
"loss": 1.2309,
"step": 2213
},
{
"epoch": 2.62030186445694,
"grad_norm": 6.78125,
"learning_rate": 1.0883917092870797e-05,
"loss": 1.2328,
"step": 2214
},
{
"epoch": 2.621485646641018,
"grad_norm": 5.34375,
"learning_rate": 1.0875677465963186e-05,
"loss": 1.17,
"step": 2215
},
{
"epoch": 2.622669428825096,
"grad_norm": 6.09375,
"learning_rate": 1.0867437239907766e-05,
"loss": 1.1187,
"step": 2216
},
{
"epoch": 2.623853211009174,
"grad_norm": 5.28125,
"learning_rate": 1.085919642034259e-05,
"loss": 0.8455,
"step": 2217
},
{
"epoch": 2.6250369931932527,
"grad_norm": 7.0625,
"learning_rate": 1.0850955012906114e-05,
"loss": 1.162,
"step": 2218
},
{
"epoch": 2.6262207753773303,
"grad_norm": 6.53125,
"learning_rate": 1.08427130232372e-05,
"loss": 0.9175,
"step": 2219
},
{
"epoch": 2.627404557561409,
"grad_norm": 6.75,
"learning_rate": 1.0834470456975092e-05,
"loss": 1.1021,
"step": 2220
},
{
"epoch": 2.6285883397454866,
"grad_norm": 6.5625,
"learning_rate": 1.0826227319759458e-05,
"loss": 1.1384,
"step": 2221
},
{
"epoch": 2.629772121929565,
"grad_norm": 5.6875,
"learning_rate": 1.0817983617230326e-05,
"loss": 1.0669,
"step": 2222
},
{
"epoch": 2.630955904113643,
"grad_norm": 6.625,
"learning_rate": 1.080973935502813e-05,
"loss": 1.4744,
"step": 2223
},
{
"epoch": 2.6321396862977213,
"grad_norm": 5.40625,
"learning_rate": 1.0801494538793684e-05,
"loss": 0.7295,
"step": 2224
},
{
"epoch": 2.6333234684817994,
"grad_norm": 6.0625,
"learning_rate": 1.079324917416818e-05,
"loss": 1.05,
"step": 2225
},
{
"epoch": 2.6345072506658775,
"grad_norm": 7.125,
"learning_rate": 1.0785003266793173e-05,
"loss": 1.6694,
"step": 2226
},
{
"epoch": 2.6356910328499557,
"grad_norm": 7.375,
"learning_rate": 1.0776756822310615e-05,
"loss": 1.7085,
"step": 2227
},
{
"epoch": 2.6368748150340338,
"grad_norm": 6.0625,
"learning_rate": 1.0768509846362798e-05,
"loss": 0.9998,
"step": 2228
},
{
"epoch": 2.638058597218112,
"grad_norm": 5.90625,
"learning_rate": 1.0760262344592399e-05,
"loss": 0.9665,
"step": 2229
},
{
"epoch": 2.63924237940219,
"grad_norm": 6.125,
"learning_rate": 1.075201432264244e-05,
"loss": 0.7546,
"step": 2230
},
{
"epoch": 2.640426161586268,
"grad_norm": 6.1875,
"learning_rate": 1.0743765786156313e-05,
"loss": 1.2163,
"step": 2231
},
{
"epoch": 2.641609943770346,
"grad_norm": 6.34375,
"learning_rate": 1.0735516740777742e-05,
"loss": 1.0245,
"step": 2232
},
{
"epoch": 2.6427937259544243,
"grad_norm": 6.125,
"learning_rate": 1.0727267192150825e-05,
"loss": 1.0954,
"step": 2233
},
{
"epoch": 2.6439775081385024,
"grad_norm": 5.875,
"learning_rate": 1.0719017145919984e-05,
"loss": 0.8902,
"step": 2234
},
{
"epoch": 2.6451612903225805,
"grad_norm": 6.78125,
"learning_rate": 1.0710766607729983e-05,
"loss": 1.2902,
"step": 2235
},
{
"epoch": 2.6463450725066586,
"grad_norm": 6.71875,
"learning_rate": 1.0702515583225936e-05,
"loss": 1.4763,
"step": 2236
},
{
"epoch": 2.647528854690737,
"grad_norm": 5.5625,
"learning_rate": 1.069426407805328e-05,
"loss": 0.9611,
"step": 2237
},
{
"epoch": 2.648712636874815,
"grad_norm": 6.75,
"learning_rate": 1.068601209785778e-05,
"loss": 1.2726,
"step": 2238
},
{
"epoch": 2.6498964190588934,
"grad_norm": 5.84375,
"learning_rate": 1.0677759648285529e-05,
"loss": 0.968,
"step": 2239
},
{
"epoch": 2.651080201242971,
"grad_norm": 6.78125,
"learning_rate": 1.066950673498294e-05,
"loss": 1.1827,
"step": 2240
},
{
"epoch": 2.6522639834270496,
"grad_norm": 5.875,
"learning_rate": 1.0661253363596742e-05,
"loss": 0.9952,
"step": 2241
},
{
"epoch": 2.6534477656111277,
"grad_norm": 6.375,
"learning_rate": 1.0652999539773985e-05,
"loss": 1.0948,
"step": 2242
},
{
"epoch": 2.654631547795206,
"grad_norm": 5.65625,
"learning_rate": 1.064474526916202e-05,
"loss": 0.6706,
"step": 2243
},
{
"epoch": 2.655815329979284,
"grad_norm": 6.03125,
"learning_rate": 1.0636490557408501e-05,
"loss": 1.266,
"step": 2244
},
{
"epoch": 2.656999112163362,
"grad_norm": 6.0625,
"learning_rate": 1.0628235410161398e-05,
"loss": 0.7374,
"step": 2245
},
{
"epoch": 2.65818289434744,
"grad_norm": 6.1875,
"learning_rate": 1.0619979833068965e-05,
"loss": 1.1541,
"step": 2246
},
{
"epoch": 2.6593666765315183,
"grad_norm": 7.1875,
"learning_rate": 1.0611723831779756e-05,
"loss": 1.2227,
"step": 2247
},
{
"epoch": 2.6605504587155964,
"grad_norm": 8.1875,
"learning_rate": 1.0603467411942618e-05,
"loss": 1.9689,
"step": 2248
},
{
"epoch": 2.6617342408996745,
"grad_norm": 6.875,
"learning_rate": 1.0595210579206676e-05,
"loss": 1.0689,
"step": 2249
},
{
"epoch": 2.6629180230837526,
"grad_norm": 7.0,
"learning_rate": 1.0586953339221346e-05,
"loss": 1.2548,
"step": 2250
},
{
"epoch": 2.6641018052678307,
"grad_norm": 5.90625,
"learning_rate": 1.057869569763632e-05,
"loss": 0.7028,
"step": 2251
},
{
"epoch": 2.665285587451909,
"grad_norm": 7.28125,
"learning_rate": 1.057043766010156e-05,
"loss": 0.9881,
"step": 2252
},
{
"epoch": 2.666469369635987,
"grad_norm": 6.3125,
"learning_rate": 1.0562179232267304e-05,
"loss": 1.1419,
"step": 2253
},
{
"epoch": 2.667653151820065,
"grad_norm": 7.25,
"learning_rate": 1.0553920419784056e-05,
"loss": 1.0079,
"step": 2254
},
{
"epoch": 2.668836934004143,
"grad_norm": 5.90625,
"learning_rate": 1.0545661228302586e-05,
"loss": 0.9402,
"step": 2255
},
{
"epoch": 2.6700207161882212,
"grad_norm": 6.78125,
"learning_rate": 1.0537401663473916e-05,
"loss": 1.3335,
"step": 2256
},
{
"epoch": 2.6712044983722993,
"grad_norm": 6.75,
"learning_rate": 1.0529141730949334e-05,
"loss": 1.37,
"step": 2257
},
{
"epoch": 2.672388280556378,
"grad_norm": 5.53125,
"learning_rate": 1.0520881436380366e-05,
"loss": 1.0184,
"step": 2258
},
{
"epoch": 2.6735720627404556,
"grad_norm": 6.28125,
"learning_rate": 1.0512620785418796e-05,
"loss": 0.9751,
"step": 2259
},
{
"epoch": 2.674755844924534,
"grad_norm": 6.15625,
"learning_rate": 1.0504359783716652e-05,
"loss": 1.0907,
"step": 2260
},
{
"epoch": 2.6759396271086118,
"grad_norm": 6.53125,
"learning_rate": 1.0496098436926195e-05,
"loss": 1.1314,
"step": 2261
},
{
"epoch": 2.6771234092926903,
"grad_norm": 6.84375,
"learning_rate": 1.0487836750699924e-05,
"loss": 1.4835,
"step": 2262
},
{
"epoch": 2.6783071914767684,
"grad_norm": 7.53125,
"learning_rate": 1.0479574730690583e-05,
"loss": 1.3952,
"step": 2263
},
{
"epoch": 2.6794909736608465,
"grad_norm": 6.34375,
"learning_rate": 1.0471312382551121e-05,
"loss": 1.0181,
"step": 2264
},
{
"epoch": 2.6806747558449247,
"grad_norm": 7.28125,
"learning_rate": 1.0463049711934731e-05,
"loss": 1.2373,
"step": 2265
},
{
"epoch": 2.6818585380290028,
"grad_norm": 6.96875,
"learning_rate": 1.0454786724494819e-05,
"loss": 1.2083,
"step": 2266
},
{
"epoch": 2.683042320213081,
"grad_norm": 6.53125,
"learning_rate": 1.0446523425885008e-05,
"loss": 1.2205,
"step": 2267
},
{
"epoch": 2.684226102397159,
"grad_norm": 5.6875,
"learning_rate": 1.0438259821759133e-05,
"loss": 0.7115,
"step": 2268
},
{
"epoch": 2.685409884581237,
"grad_norm": 6.46875,
"learning_rate": 1.0429995917771247e-05,
"loss": 1.1987,
"step": 2269
},
{
"epoch": 2.686593666765315,
"grad_norm": 5.375,
"learning_rate": 1.042173171957559e-05,
"loss": 0.8469,
"step": 2270
},
{
"epoch": 2.6877774489493933,
"grad_norm": 5.78125,
"learning_rate": 1.041346723282662e-05,
"loss": 1.3682,
"step": 2271
},
{
"epoch": 2.6889612311334714,
"grad_norm": 10.3125,
"learning_rate": 1.0405202463178985e-05,
"loss": 1.4609,
"step": 2272
},
{
"epoch": 2.6901450133175495,
"grad_norm": 6.125,
"learning_rate": 1.0396937416287527e-05,
"loss": 1.0431,
"step": 2273
},
{
"epoch": 2.6913287955016276,
"grad_norm": 6.59375,
"learning_rate": 1.0388672097807282e-05,
"loss": 0.9281,
"step": 2274
},
{
"epoch": 2.6925125776857057,
"grad_norm": 5.6875,
"learning_rate": 1.0380406513393469e-05,
"loss": 1.0345,
"step": 2275
},
{
"epoch": 2.693696359869784,
"grad_norm": 7.3125,
"learning_rate": 1.0372140668701483e-05,
"loss": 1.3589,
"step": 2276
},
{
"epoch": 2.694880142053862,
"grad_norm": 6.125,
"learning_rate": 1.0363874569386907e-05,
"loss": 1.1661,
"step": 2277
},
{
"epoch": 2.69606392423794,
"grad_norm": 7.53125,
"learning_rate": 1.035560822110549e-05,
"loss": 1.3559,
"step": 2278
},
{
"epoch": 2.6972477064220186,
"grad_norm": 6.6875,
"learning_rate": 1.034734162951316e-05,
"loss": 1.2923,
"step": 2279
},
{
"epoch": 2.6984314886060963,
"grad_norm": 7.40625,
"learning_rate": 1.0339074800266005e-05,
"loss": 1.539,
"step": 2280
},
{
"epoch": 2.699615270790175,
"grad_norm": 6.09375,
"learning_rate": 1.0330807739020278e-05,
"loss": 1.21,
"step": 2281
},
{
"epoch": 2.7007990529742525,
"grad_norm": 6.15625,
"learning_rate": 1.0322540451432385e-05,
"loss": 0.957,
"step": 2282
},
{
"epoch": 2.701982835158331,
"grad_norm": 6.625,
"learning_rate": 1.0314272943158899e-05,
"loss": 0.9859,
"step": 2283
},
{
"epoch": 2.703166617342409,
"grad_norm": 5.53125,
"learning_rate": 1.030600521985653e-05,
"loss": 1.074,
"step": 2284
},
{
"epoch": 2.7043503995264873,
"grad_norm": 6.28125,
"learning_rate": 1.0297737287182144e-05,
"loss": 1.107,
"step": 2285
},
{
"epoch": 2.7055341817105654,
"grad_norm": 5.96875,
"learning_rate": 1.0289469150792753e-05,
"loss": 1.1007,
"step": 2286
},
{
"epoch": 2.7067179638946435,
"grad_norm": 6.78125,
"learning_rate": 1.0281200816345498e-05,
"loss": 1.441,
"step": 2287
},
{
"epoch": 2.7079017460787216,
"grad_norm": 6.3125,
"learning_rate": 1.0272932289497664e-05,
"loss": 1.0922,
"step": 2288
},
{
"epoch": 2.7090855282627997,
"grad_norm": 6.625,
"learning_rate": 1.0264663575906661e-05,
"loss": 1.3534,
"step": 2289
},
{
"epoch": 2.710269310446878,
"grad_norm": 6.34375,
"learning_rate": 1.0256394681230035e-05,
"loss": 1.0719,
"step": 2290
},
{
"epoch": 2.711453092630956,
"grad_norm": 8.0,
"learning_rate": 1.024812561112545e-05,
"loss": 1.2913,
"step": 2291
},
{
"epoch": 2.712636874815034,
"grad_norm": 7.375,
"learning_rate": 1.023985637125069e-05,
"loss": 1.3609,
"step": 2292
},
{
"epoch": 2.713820656999112,
"grad_norm": 6.21875,
"learning_rate": 1.023158696726366e-05,
"loss": 0.6447,
"step": 2293
},
{
"epoch": 2.7150044391831902,
"grad_norm": 6.84375,
"learning_rate": 1.022331740482237e-05,
"loss": 1.2455,
"step": 2294
},
{
"epoch": 2.7161882213672683,
"grad_norm": 5.75,
"learning_rate": 1.0215047689584943e-05,
"loss": 1.0227,
"step": 2295
},
{
"epoch": 2.7173720035513464,
"grad_norm": 9.0,
"learning_rate": 1.0206777827209609e-05,
"loss": 1.4218,
"step": 2296
},
{
"epoch": 2.7185557857354246,
"grad_norm": 7.09375,
"learning_rate": 1.0198507823354692e-05,
"loss": 1.3187,
"step": 2297
},
{
"epoch": 2.7197395679195027,
"grad_norm": 6.09375,
"learning_rate": 1.0190237683678614e-05,
"loss": 1.1648,
"step": 2298
},
{
"epoch": 2.7209233501035808,
"grad_norm": 5.25,
"learning_rate": 1.0181967413839901e-05,
"loss": 0.8839,
"step": 2299
},
{
"epoch": 2.7221071322876593,
"grad_norm": 6.0625,
"learning_rate": 1.0173697019497152e-05,
"loss": 1.1282,
"step": 2300
},
{
"epoch": 2.723290914471737,
"grad_norm": 6.90625,
"learning_rate": 1.0165426506309059e-05,
"loss": 1.3926,
"step": 2301
},
{
"epoch": 2.7244746966558155,
"grad_norm": 8.125,
"learning_rate": 1.0157155879934399e-05,
"loss": 1.5982,
"step": 2302
},
{
"epoch": 2.725658478839893,
"grad_norm": 5.1875,
"learning_rate": 1.014888514603202e-05,
"loss": 0.6999,
"step": 2303
},
{
"epoch": 2.7268422610239718,
"grad_norm": 7.03125,
"learning_rate": 1.0140614310260844e-05,
"loss": 1.5277,
"step": 2304
},
{
"epoch": 2.72802604320805,
"grad_norm": 6.15625,
"learning_rate": 1.0132343378279868e-05,
"loss": 0.9658,
"step": 2305
},
{
"epoch": 2.729209825392128,
"grad_norm": 5.9375,
"learning_rate": 1.012407235574815e-05,
"loss": 1.0067,
"step": 2306
},
{
"epoch": 2.730393607576206,
"grad_norm": 6.25,
"learning_rate": 1.0115801248324803e-05,
"loss": 1.0243,
"step": 2307
},
{
"epoch": 2.731577389760284,
"grad_norm": 6.0625,
"learning_rate": 1.0107530061669021e-05,
"loss": 0.8551,
"step": 2308
},
{
"epoch": 2.7327611719443623,
"grad_norm": 6.0625,
"learning_rate": 1.009925880144003e-05,
"loss": 1.2504,
"step": 2309
},
{
"epoch": 2.7339449541284404,
"grad_norm": 6.28125,
"learning_rate": 1.0090987473297113e-05,
"loss": 1.1004,
"step": 2310
},
{
"epoch": 2.7351287363125185,
"grad_norm": 5.9375,
"learning_rate": 1.0082716082899605e-05,
"loss": 0.9922,
"step": 2311
},
{
"epoch": 2.7363125184965966,
"grad_norm": 6.875,
"learning_rate": 1.0074444635906875e-05,
"loss": 1.3122,
"step": 2312
},
{
"epoch": 2.7374963006806747,
"grad_norm": 5.71875,
"learning_rate": 1.0066173137978333e-05,
"loss": 0.6825,
"step": 2313
},
{
"epoch": 2.738680082864753,
"grad_norm": 7.6875,
"learning_rate": 1.0057901594773432e-05,
"loss": 1.5614,
"step": 2314
},
{
"epoch": 2.739863865048831,
"grad_norm": 8.4375,
"learning_rate": 1.0049630011951643e-05,
"loss": 1.4846,
"step": 2315
},
{
"epoch": 2.741047647232909,
"grad_norm": 6.1875,
"learning_rate": 1.0041358395172473e-05,
"loss": 0.8773,
"step": 2316
},
{
"epoch": 2.742231429416987,
"grad_norm": 7.6875,
"learning_rate": 1.0033086750095455e-05,
"loss": 1.436,
"step": 2317
},
{
"epoch": 2.7434152116010653,
"grad_norm": 7.0,
"learning_rate": 1.0024815082380131e-05,
"loss": 1.1554,
"step": 2318
},
{
"epoch": 2.7445989937851434,
"grad_norm": 7.5625,
"learning_rate": 1.0016543397686059e-05,
"loss": 0.9821,
"step": 2319
},
{
"epoch": 2.7457827759692215,
"grad_norm": 5.46875,
"learning_rate": 1.0008271701672825e-05,
"loss": 0.753,
"step": 2320
},
{
"epoch": 2.7469665581533,
"grad_norm": 6.40625,
"learning_rate": 1e-05,
"loss": 1.0561,
"step": 2321
},
{
"epoch": 2.7481503403373777,
"grad_norm": 7.78125,
"learning_rate": 9.991728298327179e-06,
"loss": 1.2483,
"step": 2322
},
{
"epoch": 2.7493341225214563,
"grad_norm": 5.3125,
"learning_rate": 9.983456602313944e-06,
"loss": 0.7125,
"step": 2323
},
{
"epoch": 2.750517904705534,
"grad_norm": 7.375,
"learning_rate": 9.975184917619874e-06,
"loss": 1.241,
"step": 2324
},
{
"epoch": 2.7517016868896125,
"grad_norm": 7.0,
"learning_rate": 9.96691324990455e-06,
"loss": 1.3892,
"step": 2325
},
{
"epoch": 2.7528854690736906,
"grad_norm": 6.65625,
"learning_rate": 9.958641604827529e-06,
"loss": 1.4,
"step": 2326
},
{
"epoch": 2.7540692512577687,
"grad_norm": 7.875,
"learning_rate": 9.950369988048357e-06,
"loss": 1.3681,
"step": 2327
},
{
"epoch": 2.755253033441847,
"grad_norm": 6.0,
"learning_rate": 9.942098405226571e-06,
"loss": 1.1676,
"step": 2328
},
{
"epoch": 2.756436815625925,
"grad_norm": 6.59375,
"learning_rate": 9.933826862021669e-06,
"loss": 1.1843,
"step": 2329
},
{
"epoch": 2.757620597810003,
"grad_norm": 6.40625,
"learning_rate": 9.92555536409313e-06,
"loss": 0.7952,
"step": 2330
},
{
"epoch": 2.758804379994081,
"grad_norm": 7.4375,
"learning_rate": 9.917283917100396e-06,
"loss": 1.5425,
"step": 2331
},
{
"epoch": 2.7599881621781592,
"grad_norm": 6.6875,
"learning_rate": 9.909012526702889e-06,
"loss": 1.0534,
"step": 2332
},
{
"epoch": 2.7611719443622373,
"grad_norm": 6.1875,
"learning_rate": 9.900741198559971e-06,
"loss": 1.0728,
"step": 2333
},
{
"epoch": 2.7623557265463154,
"grad_norm": 7.46875,
"learning_rate": 9.89246993833098e-06,
"loss": 1.3768,
"step": 2334
},
{
"epoch": 2.7635395087303936,
"grad_norm": 7.125,
"learning_rate": 9.884198751675198e-06,
"loss": 1.3364,
"step": 2335
},
{
"epoch": 2.7647232909144717,
"grad_norm": 6.75,
"learning_rate": 9.875927644251856e-06,
"loss": 1.1443,
"step": 2336
},
{
"epoch": 2.7659070730985498,
"grad_norm": 6.5625,
"learning_rate": 9.867656621720137e-06,
"loss": 1.1828,
"step": 2337
},
{
"epoch": 2.767090855282628,
"grad_norm": 7.4375,
"learning_rate": 9.859385689739157e-06,
"loss": 1.384,
"step": 2338
},
{
"epoch": 2.768274637466706,
"grad_norm": 6.65625,
"learning_rate": 9.85111485396798e-06,
"loss": 1.4711,
"step": 2339
},
{
"epoch": 2.769458419650784,
"grad_norm": 6.15625,
"learning_rate": 9.842844120065603e-06,
"loss": 1.2126,
"step": 2340
},
{
"epoch": 2.770642201834862,
"grad_norm": 5.40625,
"learning_rate": 9.834573493690941e-06,
"loss": 0.9383,
"step": 2341
},
{
"epoch": 2.7718259840189408,
"grad_norm": 6.21875,
"learning_rate": 9.826302980502853e-06,
"loss": 1.0163,
"step": 2342
},
{
"epoch": 2.7730097662030184,
"grad_norm": 6.4375,
"learning_rate": 9.8180325861601e-06,
"loss": 1.1592,
"step": 2343
},
{
"epoch": 2.774193548387097,
"grad_norm": 6.3125,
"learning_rate": 9.809762316321388e-06,
"loss": 0.8533,
"step": 2344
},
{
"epoch": 2.7753773305711746,
"grad_norm": 5.65625,
"learning_rate": 9.801492176645313e-06,
"loss": 0.7429,
"step": 2345
},
{
"epoch": 2.776561112755253,
"grad_norm": 6.78125,
"learning_rate": 9.793222172790396e-06,
"loss": 1.4761,
"step": 2346
},
{
"epoch": 2.7777448949393313,
"grad_norm": 7.0625,
"learning_rate": 9.784952310415062e-06,
"loss": 1.3688,
"step": 2347
},
{
"epoch": 2.7789286771234094,
"grad_norm": 6.46875,
"learning_rate": 9.776682595177633e-06,
"loss": 1.3887,
"step": 2348
},
{
"epoch": 2.7801124593074875,
"grad_norm": 5.9375,
"learning_rate": 9.768413032736345e-06,
"loss": 1.0587,
"step": 2349
},
{
"epoch": 2.7812962414915656,
"grad_norm": 6.75,
"learning_rate": 9.760143628749312e-06,
"loss": 1.136,
"step": 2350
},
{
"epoch": 2.7824800236756437,
"grad_norm": 6.625,
"learning_rate": 9.75187438887455e-06,
"loss": 1.2838,
"step": 2351
},
{
"epoch": 2.783663805859722,
"grad_norm": 6.3125,
"learning_rate": 9.743605318769967e-06,
"loss": 1.1607,
"step": 2352
},
{
"epoch": 2.7848475880438,
"grad_norm": 6.5,
"learning_rate": 9.735336424093342e-06,
"loss": 0.9626,
"step": 2353
},
{
"epoch": 2.786031370227878,
"grad_norm": 6.1875,
"learning_rate": 9.727067710502341e-06,
"loss": 0.9949,
"step": 2354
},
{
"epoch": 2.787215152411956,
"grad_norm": 6.09375,
"learning_rate": 9.718799183654505e-06,
"loss": 1.222,
"step": 2355
},
{
"epoch": 2.7883989345960343,
"grad_norm": 6.28125,
"learning_rate": 9.710530849207249e-06,
"loss": 1.0208,
"step": 2356
},
{
"epoch": 2.7895827167801124,
"grad_norm": 5.9375,
"learning_rate": 9.702262712817857e-06,
"loss": 1.3664,
"step": 2357
},
{
"epoch": 2.7907664989641905,
"grad_norm": 6.15625,
"learning_rate": 9.693994780143475e-06,
"loss": 1.2969,
"step": 2358
},
{
"epoch": 2.7919502811482686,
"grad_norm": 6.28125,
"learning_rate": 9.685727056841108e-06,
"loss": 1.1192,
"step": 2359
},
{
"epoch": 2.7931340633323467,
"grad_norm": 5.71875,
"learning_rate": 9.677459548567617e-06,
"loss": 0.8762,
"step": 2360
},
{
"epoch": 2.794317845516425,
"grad_norm": 5.34375,
"learning_rate": 9.669192260979727e-06,
"loss": 0.923,
"step": 2361
},
{
"epoch": 2.795501627700503,
"grad_norm": 5.5625,
"learning_rate": 9.660925199733997e-06,
"loss": 1.0636,
"step": 2362
},
{
"epoch": 2.7966854098845815,
"grad_norm": 5.9375,
"learning_rate": 9.65265837048684e-06,
"loss": 0.9306,
"step": 2363
},
{
"epoch": 2.797869192068659,
"grad_norm": 6.78125,
"learning_rate": 9.64439177889451e-06,
"loss": 1.2471,
"step": 2364
},
{
"epoch": 2.7990529742527377,
"grad_norm": 6.4375,
"learning_rate": 9.636125430613096e-06,
"loss": 1.2942,
"step": 2365
},
{
"epoch": 2.8002367564368154,
"grad_norm": 6.25,
"learning_rate": 9.627859331298522e-06,
"loss": 1.0748,
"step": 2366
},
{
"epoch": 2.801420538620894,
"grad_norm": 7.3125,
"learning_rate": 9.619593486606533e-06,
"loss": 1.322,
"step": 2367
},
{
"epoch": 2.802604320804972,
"grad_norm": 5.78125,
"learning_rate": 9.611327902192718e-06,
"loss": 0.8874,
"step": 2368
},
{
"epoch": 2.80378810298905,
"grad_norm": 6.75,
"learning_rate": 9.603062583712475e-06,
"loss": 1.2388,
"step": 2369
},
{
"epoch": 2.8049718851731282,
"grad_norm": 5.625,
"learning_rate": 9.594797536821018e-06,
"loss": 1.2095,
"step": 2370
},
{
"epoch": 2.8061556673572063,
"grad_norm": 5.71875,
"learning_rate": 9.586532767173385e-06,
"loss": 1.1153,
"step": 2371
},
{
"epoch": 2.8073394495412844,
"grad_norm": 7.15625,
"learning_rate": 9.578268280424413e-06,
"loss": 1.289,
"step": 2372
},
{
"epoch": 2.8085232317253626,
"grad_norm": 6.09375,
"learning_rate": 9.570004082228757e-06,
"loss": 1.0524,
"step": 2373
},
{
"epoch": 2.8097070139094407,
"grad_norm": 6.1875,
"learning_rate": 9.561740178240868e-06,
"loss": 0.9151,
"step": 2374
},
{
"epoch": 2.8108907960935188,
"grad_norm": 7.65625,
"learning_rate": 9.553476574114993e-06,
"loss": 1.3566,
"step": 2375
},
{
"epoch": 2.812074578277597,
"grad_norm": 6.0625,
"learning_rate": 9.545213275505183e-06,
"loss": 0.9359,
"step": 2376
},
{
"epoch": 2.813258360461675,
"grad_norm": 5.8125,
"learning_rate": 9.53695028806527e-06,
"loss": 0.8209,
"step": 2377
},
{
"epoch": 2.814442142645753,
"grad_norm": 5.875,
"learning_rate": 9.528687617448884e-06,
"loss": 0.9029,
"step": 2378
},
{
"epoch": 2.815625924829831,
"grad_norm": 6.34375,
"learning_rate": 9.52042526930942e-06,
"loss": 0.8871,
"step": 2379
},
{
"epoch": 2.8168097070139093,
"grad_norm": 5.90625,
"learning_rate": 9.512163249300074e-06,
"loss": 0.8143,
"step": 2380
},
{
"epoch": 2.8179934891979874,
"grad_norm": 7.34375,
"learning_rate": 9.50390156307381e-06,
"loss": 1.1672,
"step": 2381
},
{
"epoch": 2.819177271382066,
"grad_norm": 6.0625,
"learning_rate": 9.495640216283352e-06,
"loss": 1.1058,
"step": 2382
},
{
"epoch": 2.8203610535661436,
"grad_norm": 8.375,
"learning_rate": 9.487379214581209e-06,
"loss": 1.765,
"step": 2383
},
{
"epoch": 2.821544835750222,
"grad_norm": 6.34375,
"learning_rate": 9.479118563619638e-06,
"loss": 1.0983,
"step": 2384
},
{
"epoch": 2.8227286179343,
"grad_norm": 6.40625,
"learning_rate": 9.470858269050669e-06,
"loss": 1.3971,
"step": 2385
},
{
"epoch": 2.8239124001183784,
"grad_norm": 5.96875,
"learning_rate": 9.462598336526086e-06,
"loss": 1.0095,
"step": 2386
},
{
"epoch": 2.825096182302456,
"grad_norm": 5.9375,
"learning_rate": 9.454338771697414e-06,
"loss": 0.8397,
"step": 2387
},
{
"epoch": 2.8262799644865346,
"grad_norm": 6.15625,
"learning_rate": 9.446079580215946e-06,
"loss": 1.1709,
"step": 2388
},
{
"epoch": 2.8274637466706127,
"grad_norm": 6.34375,
"learning_rate": 9.437820767732699e-06,
"loss": 1.1953,
"step": 2389
},
{
"epoch": 2.828647528854691,
"grad_norm": 5.4375,
"learning_rate": 9.429562339898446e-06,
"loss": 0.9987,
"step": 2390
},
{
"epoch": 2.829831311038769,
"grad_norm": 5.34375,
"learning_rate": 9.421304302363683e-06,
"loss": 0.8543,
"step": 2391
},
{
"epoch": 2.831015093222847,
"grad_norm": 5.875,
"learning_rate": 9.413046660778654e-06,
"loss": 1.1847,
"step": 2392
},
{
"epoch": 2.832198875406925,
"grad_norm": 7.0625,
"learning_rate": 9.404789420793327e-06,
"loss": 0.9492,
"step": 2393
},
{
"epoch": 2.8333826575910033,
"grad_norm": 9.5,
"learning_rate": 9.396532588057384e-06,
"loss": 1.3564,
"step": 2394
},
{
"epoch": 2.8345664397750814,
"grad_norm": 5.8125,
"learning_rate": 9.388276168220248e-06,
"loss": 1.1207,
"step": 2395
},
{
"epoch": 2.8357502219591595,
"grad_norm": 6.625,
"learning_rate": 9.380020166931037e-06,
"loss": 1.1249,
"step": 2396
},
{
"epoch": 2.8369340041432376,
"grad_norm": 7.25,
"learning_rate": 9.371764589838605e-06,
"loss": 1.3787,
"step": 2397
},
{
"epoch": 2.8381177863273157,
"grad_norm": 6.375,
"learning_rate": 9.363509442591502e-06,
"loss": 1.4053,
"step": 2398
},
{
"epoch": 2.839301568511394,
"grad_norm": 6.96875,
"learning_rate": 9.355254730837983e-06,
"loss": 1.2756,
"step": 2399
},
{
"epoch": 2.840485350695472,
"grad_norm": 6.34375,
"learning_rate": 9.34700046022602e-06,
"loss": 1.0073,
"step": 2400
},
{
"epoch": 2.84166913287955,
"grad_norm": 5.28125,
"learning_rate": 9.33874663640326e-06,
"loss": 0.8185,
"step": 2401
},
{
"epoch": 2.842852915063628,
"grad_norm": 5.5625,
"learning_rate": 9.330493265017062e-06,
"loss": 0.9028,
"step": 2402
},
{
"epoch": 2.8440366972477067,
"grad_norm": 6.71875,
"learning_rate": 9.322240351714475e-06,
"loss": 1.7147,
"step": 2403
},
{
"epoch": 2.8452204794317844,
"grad_norm": 6.25,
"learning_rate": 9.313987902142222e-06,
"loss": 1.1269,
"step": 2404
},
{
"epoch": 2.846404261615863,
"grad_norm": 6.4375,
"learning_rate": 9.305735921946724e-06,
"loss": 1.2338,
"step": 2405
},
{
"epoch": 2.8475880437999406,
"grad_norm": 6.53125,
"learning_rate": 9.297484416774066e-06,
"loss": 1.0978,
"step": 2406
},
{
"epoch": 2.848771825984019,
"grad_norm": 6.65625,
"learning_rate": 9.28923339227002e-06,
"loss": 1.3216,
"step": 2407
},
{
"epoch": 2.8499556081680972,
"grad_norm": 5.875,
"learning_rate": 9.280982854080021e-06,
"loss": 0.9065,
"step": 2408
},
{
"epoch": 2.8511393903521753,
"grad_norm": 5.6875,
"learning_rate": 9.272732807849177e-06,
"loss": 0.8997,
"step": 2409
},
{
"epoch": 2.8523231725362534,
"grad_norm": 6.625,
"learning_rate": 9.26448325922226e-06,
"loss": 1.2691,
"step": 2410
},
{
"epoch": 2.8535069547203316,
"grad_norm": 5.53125,
"learning_rate": 9.25623421384369e-06,
"loss": 1.0706,
"step": 2411
},
{
"epoch": 2.8546907369044097,
"grad_norm": 7.0625,
"learning_rate": 9.247985677357563e-06,
"loss": 0.9466,
"step": 2412
},
{
"epoch": 2.8558745190884878,
"grad_norm": 5.84375,
"learning_rate": 9.239737655407605e-06,
"loss": 1.0938,
"step": 2413
},
{
"epoch": 2.857058301272566,
"grad_norm": 7.84375,
"learning_rate": 9.231490153637203e-06,
"loss": 1.6677,
"step": 2414
},
{
"epoch": 2.858242083456644,
"grad_norm": 6.21875,
"learning_rate": 9.223243177689389e-06,
"loss": 1.167,
"step": 2415
},
{
"epoch": 2.859425865640722,
"grad_norm": 6.875,
"learning_rate": 9.214996733206827e-06,
"loss": 1.1343,
"step": 2416
},
{
"epoch": 2.8606096478248,
"grad_norm": 7.375,
"learning_rate": 9.206750825831826e-06,
"loss": 1.0821,
"step": 2417
},
{
"epoch": 2.8617934300088783,
"grad_norm": 6.46875,
"learning_rate": 9.198505461206318e-06,
"loss": 1.1438,
"step": 2418
},
{
"epoch": 2.8629772121929564,
"grad_norm": 7.0,
"learning_rate": 9.190260644971874e-06,
"loss": 1.284,
"step": 2419
},
{
"epoch": 2.8641609943770345,
"grad_norm": 6.71875,
"learning_rate": 9.182016382769678e-06,
"loss": 1.0553,
"step": 2420
},
{
"epoch": 2.8653447765611126,
"grad_norm": 6.78125,
"learning_rate": 9.173772680240547e-06,
"loss": 1.0545,
"step": 2421
},
{
"epoch": 2.8665285587451907,
"grad_norm": 7.1875,
"learning_rate": 9.165529543024911e-06,
"loss": 1.6063,
"step": 2422
},
{
"epoch": 2.867712340929269,
"grad_norm": 7.5625,
"learning_rate": 9.157286976762805e-06,
"loss": 1.1375,
"step": 2423
},
{
"epoch": 2.8688961231133474,
"grad_norm": 6.21875,
"learning_rate": 9.149044987093887e-06,
"loss": 1.082,
"step": 2424
},
{
"epoch": 2.870079905297425,
"grad_norm": 6.40625,
"learning_rate": 9.140803579657413e-06,
"loss": 1.0392,
"step": 2425
},
{
"epoch": 2.8712636874815036,
"grad_norm": 6.90625,
"learning_rate": 9.132562760092234e-06,
"loss": 1.4498,
"step": 2426
},
{
"epoch": 2.8724474696655813,
"grad_norm": 6.125,
"learning_rate": 9.124322534036816e-06,
"loss": 0.9699,
"step": 2427
},
{
"epoch": 2.87363125184966,
"grad_norm": 6.84375,
"learning_rate": 9.116082907129204e-06,
"loss": 1.3269,
"step": 2428
},
{
"epoch": 2.874815034033738,
"grad_norm": 6.40625,
"learning_rate": 9.107843885007042e-06,
"loss": 1.2253,
"step": 2429
},
{
"epoch": 2.875998816217816,
"grad_norm": 5.78125,
"learning_rate": 9.09960547330755e-06,
"loss": 1.0871,
"step": 2430
},
{
"epoch": 2.877182598401894,
"grad_norm": 6.8125,
"learning_rate": 9.091367677667538e-06,
"loss": 1.3727,
"step": 2431
},
{
"epoch": 2.8783663805859723,
"grad_norm": 6.9375,
"learning_rate": 9.083130503723397e-06,
"loss": 1.5471,
"step": 2432
},
{
"epoch": 2.8795501627700504,
"grad_norm": 5.34375,
"learning_rate": 9.074893957111087e-06,
"loss": 0.6225,
"step": 2433
},
{
"epoch": 2.8807339449541285,
"grad_norm": 6.28125,
"learning_rate": 9.06665804346614e-06,
"loss": 0.8826,
"step": 2434
},
{
"epoch": 2.8819177271382066,
"grad_norm": 6.125,
"learning_rate": 9.058422768423653e-06,
"loss": 1.2183,
"step": 2435
},
{
"epoch": 2.8831015093222847,
"grad_norm": 6.84375,
"learning_rate": 9.050188137618296e-06,
"loss": 1.286,
"step": 2436
},
{
"epoch": 2.884285291506363,
"grad_norm": 6.90625,
"learning_rate": 9.041954156684282e-06,
"loss": 1.3903,
"step": 2437
},
{
"epoch": 2.885469073690441,
"grad_norm": 7.0625,
"learning_rate": 9.033720831255391e-06,
"loss": 1.1627,
"step": 2438
},
{
"epoch": 2.886652855874519,
"grad_norm": 6.40625,
"learning_rate": 9.025488166964955e-06,
"loss": 1.1019,
"step": 2439
},
{
"epoch": 2.887836638058597,
"grad_norm": 5.59375,
"learning_rate": 9.017256169445847e-06,
"loss": 0.8802,
"step": 2440
},
{
"epoch": 2.8890204202426752,
"grad_norm": 5.53125,
"learning_rate": 9.009024844330489e-06,
"loss": 0.9479,
"step": 2441
},
{
"epoch": 2.8902042024267534,
"grad_norm": 6.5,
"learning_rate": 9.000794197250836e-06,
"loss": 1.2495,
"step": 2442
},
{
"epoch": 2.8913879846108315,
"grad_norm": 5.75,
"learning_rate": 8.992564233838388e-06,
"loss": 1.125,
"step": 2443
},
{
"epoch": 2.8925717667949096,
"grad_norm": 6.53125,
"learning_rate": 8.984334959724177e-06,
"loss": 1.2878,
"step": 2444
},
{
"epoch": 2.893755548978988,
"grad_norm": 4.71875,
"learning_rate": 8.976106380538752e-06,
"loss": 0.6566,
"step": 2445
},
{
"epoch": 2.894939331163066,
"grad_norm": 5.59375,
"learning_rate": 8.9678785019122e-06,
"loss": 1.0569,
"step": 2446
},
{
"epoch": 2.8961231133471443,
"grad_norm": 5.9375,
"learning_rate": 8.959651329474115e-06,
"loss": 1.0386,
"step": 2447
},
{
"epoch": 2.897306895531222,
"grad_norm": 7.65625,
"learning_rate": 8.951424868853623e-06,
"loss": 1.1062,
"step": 2448
},
{
"epoch": 2.8984906777153006,
"grad_norm": 6.53125,
"learning_rate": 8.943199125679357e-06,
"loss": 1.1579,
"step": 2449
},
{
"epoch": 2.8996744598993787,
"grad_norm": 5.5,
"learning_rate": 8.934974105579449e-06,
"loss": 0.978,
"step": 2450
},
{
"epoch": 2.9008582420834568,
"grad_norm": 6.4375,
"learning_rate": 8.926749814181555e-06,
"loss": 1.258,
"step": 2451
},
{
"epoch": 2.902042024267535,
"grad_norm": 6.8125,
"learning_rate": 8.918526257112815e-06,
"loss": 1.5111,
"step": 2452
},
{
"epoch": 2.903225806451613,
"grad_norm": 6.40625,
"learning_rate": 8.910303439999877e-06,
"loss": 1.2364,
"step": 2453
},
{
"epoch": 2.904409588635691,
"grad_norm": 6.25,
"learning_rate": 8.902081368468879e-06,
"loss": 1.1527,
"step": 2454
},
{
"epoch": 2.905593370819769,
"grad_norm": 7.09375,
"learning_rate": 8.893860048145446e-06,
"loss": 1.3789,
"step": 2455
},
{
"epoch": 2.9067771530038473,
"grad_norm": 8.0,
"learning_rate": 8.8856394846547e-06,
"loss": 1.1698,
"step": 2456
},
{
"epoch": 2.9079609351879254,
"grad_norm": 5.1875,
"learning_rate": 8.877419683621233e-06,
"loss": 0.8665,
"step": 2457
},
{
"epoch": 2.9091447173720035,
"grad_norm": 5.8125,
"learning_rate": 8.869200650669122e-06,
"loss": 0.8756,
"step": 2458
},
{
"epoch": 2.9103284995560816,
"grad_norm": 6.53125,
"learning_rate": 8.860982391421914e-06,
"loss": 1.2062,
"step": 2459
},
{
"epoch": 2.9115122817401597,
"grad_norm": 5.5625,
"learning_rate": 8.85276491150263e-06,
"loss": 0.8244,
"step": 2460
},
{
"epoch": 2.912696063924238,
"grad_norm": 7.375,
"learning_rate": 8.844548216533759e-06,
"loss": 1.2811,
"step": 2461
},
{
"epoch": 2.913879846108316,
"grad_norm": 7.0,
"learning_rate": 8.836332312137246e-06,
"loss": 1.3121,
"step": 2462
},
{
"epoch": 2.915063628292394,
"grad_norm": 6.625,
"learning_rate": 8.828117203934506e-06,
"loss": 1.1164,
"step": 2463
},
{
"epoch": 2.916247410476472,
"grad_norm": 5.90625,
"learning_rate": 8.8199028975464e-06,
"loss": 1.1892,
"step": 2464
},
{
"epoch": 2.9174311926605503,
"grad_norm": 6.28125,
"learning_rate": 8.811689398593245e-06,
"loss": 1.1518,
"step": 2465
},
{
"epoch": 2.918614974844629,
"grad_norm": 7.46875,
"learning_rate": 8.8034767126948e-06,
"loss": 1.304,
"step": 2466
},
{
"epoch": 2.9197987570287065,
"grad_norm": 7.0,
"learning_rate": 8.795264845470272e-06,
"loss": 1.5269,
"step": 2467
},
{
"epoch": 2.920982539212785,
"grad_norm": 7.21875,
"learning_rate": 8.787053802538316e-06,
"loss": 1.3274,
"step": 2468
},
{
"epoch": 2.9221663213968627,
"grad_norm": 7.0625,
"learning_rate": 8.778843589517005e-06,
"loss": 1.6598,
"step": 2469
},
{
"epoch": 2.9233501035809413,
"grad_norm": 7.09375,
"learning_rate": 8.770634212023861e-06,
"loss": 1.2165,
"step": 2470
},
{
"epoch": 2.9245338857650194,
"grad_norm": 5.875,
"learning_rate": 8.762425675675818e-06,
"loss": 1.2819,
"step": 2471
},
{
"epoch": 2.9257176679490975,
"grad_norm": 6.875,
"learning_rate": 8.754217986089253e-06,
"loss": 0.9995,
"step": 2472
},
{
"epoch": 2.9269014501331756,
"grad_norm": 8.3125,
"learning_rate": 8.746011148879951e-06,
"loss": 1.3236,
"step": 2473
},
{
"epoch": 2.9280852323172537,
"grad_norm": 6.15625,
"learning_rate": 8.737805169663113e-06,
"loss": 1.1501,
"step": 2474
},
{
"epoch": 2.929269014501332,
"grad_norm": 7.875,
"learning_rate": 8.729600054053367e-06,
"loss": 1.3133,
"step": 2475
},
{
"epoch": 2.93045279668541,
"grad_norm": 7.28125,
"learning_rate": 8.72139580766473e-06,
"loss": 1.4588,
"step": 2476
},
{
"epoch": 2.931636578869488,
"grad_norm": 6.65625,
"learning_rate": 8.713192436110635e-06,
"loss": 1.0934,
"step": 2477
},
{
"epoch": 2.932820361053566,
"grad_norm": 6.65625,
"learning_rate": 8.704989945003926e-06,
"loss": 1.3615,
"step": 2478
},
{
"epoch": 2.9340041432376442,
"grad_norm": 6.25,
"learning_rate": 8.696788339956819e-06,
"loss": 1.2157,
"step": 2479
},
{
"epoch": 2.9351879254217224,
"grad_norm": 5.78125,
"learning_rate": 8.688587626580954e-06,
"loss": 1.0072,
"step": 2480
},
{
"epoch": 2.9363717076058005,
"grad_norm": 5.0625,
"learning_rate": 8.680387810487337e-06,
"loss": 0.692,
"step": 2481
},
{
"epoch": 2.9375554897898786,
"grad_norm": 7.0625,
"learning_rate": 8.672188897286374e-06,
"loss": 1.3449,
"step": 2482
},
{
"epoch": 2.9387392719739567,
"grad_norm": 6.34375,
"learning_rate": 8.663990892587839e-06,
"loss": 1.0684,
"step": 2483
},
{
"epoch": 2.939923054158035,
"grad_norm": 6.3125,
"learning_rate": 8.655793802000905e-06,
"loss": 1.0667,
"step": 2484
},
{
"epoch": 2.941106836342113,
"grad_norm": 5.1875,
"learning_rate": 8.647597631134105e-06,
"loss": 0.8632,
"step": 2485
},
{
"epoch": 2.942290618526191,
"grad_norm": 6.9375,
"learning_rate": 8.639402385595342e-06,
"loss": 1.354,
"step": 2486
},
{
"epoch": 2.9434744007102696,
"grad_norm": 7.03125,
"learning_rate": 8.631208070991896e-06,
"loss": 1.3462,
"step": 2487
},
{
"epoch": 2.944658182894347,
"grad_norm": 6.5,
"learning_rate": 8.623014692930399e-06,
"loss": 1.3596,
"step": 2488
},
{
"epoch": 2.9458419650784258,
"grad_norm": 7.8125,
"learning_rate": 8.614822257016845e-06,
"loss": 1.4237,
"step": 2489
},
{
"epoch": 2.9470257472625034,
"grad_norm": 7.4375,
"learning_rate": 8.606630768856596e-06,
"loss": 1.5147,
"step": 2490
},
{
"epoch": 2.948209529446582,
"grad_norm": 5.53125,
"learning_rate": 8.598440234054345e-06,
"loss": 0.7361,
"step": 2491
},
{
"epoch": 2.94939331163066,
"grad_norm": 6.6875,
"learning_rate": 8.590250658214148e-06,
"loss": 1.1826,
"step": 2492
},
{
"epoch": 2.950577093814738,
"grad_norm": 7.96875,
"learning_rate": 8.582062046939396e-06,
"loss": 1.3799,
"step": 2493
},
{
"epoch": 2.9517608759988163,
"grad_norm": 7.09375,
"learning_rate": 8.573874405832828e-06,
"loss": 1.0446,
"step": 2494
},
{
"epoch": 2.9529446581828944,
"grad_norm": 6.21875,
"learning_rate": 8.565687740496507e-06,
"loss": 0.9808,
"step": 2495
},
{
"epoch": 2.9541284403669725,
"grad_norm": 6.5,
"learning_rate": 8.557502056531844e-06,
"loss": 1.1064,
"step": 2496
},
{
"epoch": 2.9553122225510506,
"grad_norm": 5.4375,
"learning_rate": 8.549317359539571e-06,
"loss": 0.9119,
"step": 2497
},
{
"epoch": 2.9564960047351287,
"grad_norm": 6.34375,
"learning_rate": 8.541133655119736e-06,
"loss": 1.0688,
"step": 2498
},
{
"epoch": 2.957679786919207,
"grad_norm": 6.71875,
"learning_rate": 8.532950948871728e-06,
"loss": 1.0381,
"step": 2499
},
{
"epoch": 2.958863569103285,
"grad_norm": 7.40625,
"learning_rate": 8.524769246394233e-06,
"loss": 1.1347,
"step": 2500
},
{
"epoch": 2.960047351287363,
"grad_norm": 6.96875,
"learning_rate": 8.516588553285258e-06,
"loss": 0.9746,
"step": 2501
},
{
"epoch": 2.961231133471441,
"grad_norm": 5.4375,
"learning_rate": 8.50840887514213e-06,
"loss": 1.0134,
"step": 2502
},
{
"epoch": 2.9624149156555193,
"grad_norm": 6.5625,
"learning_rate": 8.500230217561459e-06,
"loss": 0.9633,
"step": 2503
},
{
"epoch": 2.9635986978395974,
"grad_norm": 6.6875,
"learning_rate": 8.492052586139176e-06,
"loss": 0.8959,
"step": 2504
},
{
"epoch": 2.9647824800236755,
"grad_norm": 6.78125,
"learning_rate": 8.483875986470502e-06,
"loss": 1.2959,
"step": 2505
},
{
"epoch": 2.9659662622077536,
"grad_norm": 6.875,
"learning_rate": 8.47570042414995e-06,
"loss": 1.3671,
"step": 2506
},
{
"epoch": 2.9671500443918317,
"grad_norm": 8.6875,
"learning_rate": 8.467525904771336e-06,
"loss": 0.8344,
"step": 2507
},
{
"epoch": 2.9683338265759103,
"grad_norm": 6.71875,
"learning_rate": 8.459352433927743e-06,
"loss": 1.4816,
"step": 2508
},
{
"epoch": 2.969517608759988,
"grad_norm": 5.4375,
"learning_rate": 8.45118001721155e-06,
"loss": 0.7365,
"step": 2509
},
{
"epoch": 2.9707013909440665,
"grad_norm": 6.21875,
"learning_rate": 8.443008660214409e-06,
"loss": 1.4257,
"step": 2510
},
{
"epoch": 2.971885173128144,
"grad_norm": 5.53125,
"learning_rate": 8.434838368527254e-06,
"loss": 0.6069,
"step": 2511
},
{
"epoch": 2.9730689553122227,
"grad_norm": 7.21875,
"learning_rate": 8.42666914774028e-06,
"loss": 1.4487,
"step": 2512
},
{
"epoch": 2.974252737496301,
"grad_norm": 5.78125,
"learning_rate": 8.418501003442954e-06,
"loss": 1.0226,
"step": 2513
},
{
"epoch": 2.975436519680379,
"grad_norm": 6.96875,
"learning_rate": 8.410333941224017e-06,
"loss": 1.44,
"step": 2514
},
{
"epoch": 2.976620301864457,
"grad_norm": 6.125,
"learning_rate": 8.402167966671445e-06,
"loss": 1.2718,
"step": 2515
},
{
"epoch": 2.977804084048535,
"grad_norm": 6.4375,
"learning_rate": 8.394003085372496e-06,
"loss": 1.1982,
"step": 2516
},
{
"epoch": 2.9789878662326132,
"grad_norm": 7.3125,
"learning_rate": 8.385839302913661e-06,
"loss": 1.6151,
"step": 2517
},
{
"epoch": 2.9801716484166914,
"grad_norm": 6.71875,
"learning_rate": 8.377676624880688e-06,
"loss": 1.3398,
"step": 2518
},
{
"epoch": 2.9813554306007695,
"grad_norm": 6.78125,
"learning_rate": 8.369515056858575e-06,
"loss": 1.3009,
"step": 2519
},
{
"epoch": 2.9825392127848476,
"grad_norm": 7.1875,
"learning_rate": 8.361354604431544e-06,
"loss": 1.1456,
"step": 2520
},
{
"epoch": 2.9837229949689257,
"grad_norm": 7.28125,
"learning_rate": 8.353195273183068e-06,
"loss": 1.5974,
"step": 2521
},
{
"epoch": 2.984906777153004,
"grad_norm": 5.625,
"learning_rate": 8.345037068695844e-06,
"loss": 0.9518,
"step": 2522
},
{
"epoch": 2.986090559337082,
"grad_norm": 7.34375,
"learning_rate": 8.336879996551807e-06,
"loss": 1.3621,
"step": 2523
},
{
"epoch": 2.98727434152116,
"grad_norm": 5.625,
"learning_rate": 8.32872406233211e-06,
"loss": 0.988,
"step": 2524
},
{
"epoch": 2.988458123705238,
"grad_norm": 7.78125,
"learning_rate": 8.320569271617127e-06,
"loss": 1.7359,
"step": 2525
},
{
"epoch": 2.989641905889316,
"grad_norm": 6.34375,
"learning_rate": 8.31241562998646e-06,
"loss": 0.9107,
"step": 2526
},
{
"epoch": 2.9908256880733948,
"grad_norm": 6.71875,
"learning_rate": 8.304263143018906e-06,
"loss": 1.4946,
"step": 2527
},
{
"epoch": 2.9920094702574724,
"grad_norm": 7.375,
"learning_rate": 8.296111816292494e-06,
"loss": 1.3541,
"step": 2528
},
{
"epoch": 2.993193252441551,
"grad_norm": 6.53125,
"learning_rate": 8.287961655384439e-06,
"loss": 1.1389,
"step": 2529
},
{
"epoch": 2.9943770346256287,
"grad_norm": 5.71875,
"learning_rate": 8.279812665871171e-06,
"loss": 1.0001,
"step": 2530
},
{
"epoch": 2.995560816809707,
"grad_norm": 6.875,
"learning_rate": 8.271664853328319e-06,
"loss": 1.2689,
"step": 2531
},
{
"epoch": 2.996744598993785,
"grad_norm": 7.0,
"learning_rate": 8.263518223330698e-06,
"loss": 1.3522,
"step": 2532
},
{
"epoch": 2.9979283811778634,
"grad_norm": 6.78125,
"learning_rate": 8.255372781452322e-06,
"loss": 1.27,
"step": 2533
},
{
"epoch": 2.9991121633619415,
"grad_norm": 5.46875,
"learning_rate": 8.247228533266381e-06,
"loss": 0.8696,
"step": 2534
},
{
"epoch": 3.0,
"grad_norm": 6.4375,
"learning_rate": 8.239085484345265e-06,
"loss": 0.827,
"step": 2535
}
],
"logging_steps": 1.0,
"max_steps": 4220,
"num_input_tokens_seen": 0,
"num_train_epochs": 5,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 2.017988398822564e+17,
"train_batch_size": 12,
"trial_name": null,
"trial_params": null
}